DOC TOR A L T H E S I S
ISSN: 1402-1544 ISBN 978-91-7439-019-3 Luleå University of Technology 2009
Annamaria Vilinska Microbial Adhesion and Surface Modifications of Sulphide Minerals Relevant to Flotation and Flocculation
Department of Chemical Engineering and Geosciences Mineral Processing
Microbial Adhesion and Surface Modifications of Sulphide Minerals Relevant to Flotation and Flocculation
Annamaria Vilinska
Microbial Adhesion and Surface Modification of Sulphide Minerals Relevant to Flotation and Flocculation
Annamaria Vilinska Luleå University of Technology Department of Chemical Engineering and Geosciences Division of Extractive Metallurgy
Printed by Universitetstryckeriet, Luleå 2009 ISSN: 1402-1544 ISBN 978-91-7439-019-3 Luleå www.ltu.se
ABSTRACT Biological processes have been attracting attention in mineral processing industry due to their lower operating costs, environmental acceptability and flexibility in adaptation. Biohydrometallurgical methods to treat various sulphide and oxide ores have been developed during the years and these processes have been adopted by certain mineral industries. Biobeneficiation processes involving the separation of value minerals from ores and materials using conventional flotation and flocculation methods have been shown to be promising in recent years. There is a tremendous potential to use microorganisms as flocculants, flotation collectors and/or depressants. A great progress in flotation of sulphides has been realised by using mineral environment – native bacterial strains, such as acidophilic sulphur oxidizing bacteria of Thiobacillus genus. In this work the alterations of surface properties of pyrite and chalcopyrite after biological conditioning with At. ferrooxidans and L. ferrooxidans cells and with the cells adapted to higher concentrations of Cu and Zn ions, were investigated. Both strains are acidophilic, iron oxidizing microorganisms, with natural occurrence in ore deposits and mine water and high affinity towards sulphide minerals. The changes in surface properties of wild and metal ions adapted bacteria and minerals after bacterial treatment were evaluated by zeta-potential and adsorption, diffuse reflectance FT-IR and XPS studies. The flocculation of particles from settling behaviour in aqueous suspension and Hallimond microflotation tests were performed to quantify the effect of cell treatment on separation processes. Yeast cells of Saccharomyces cerevisiae and Yarrowia lipolytica were tested as a coagulating agent for very fine sulphide mineral separation from silicates. Measurement of contact angles on cell lawn and mineral surfaces using different test liquids were done to determine the hydrophobic/hydrophilic character of bacterial and mineral surfaces, and to estimate the different components of surface free energy; Lifshitz-van der Waals and polar, and polar component divided into acid and base components contributing to the total surface free energy. In addition, the Hamaker constant that is essential to construct the DLVO potential energy diagrams of bacterial cell interaction to mineral, has been estimated from the free energy of bacterial adhesion to minerals, which is determined from contact angle data. Although both strains are iron oxidizing, their genetic and metabolic pathways differ and consequently their surface properties. While ferrous grown At. ferrooxidans tends to be slightly negatively charged in the entire pH range studied, L. ferrooxidans cells exhibited higher magnitude of zeta potential and a clear iso-electric point (IEP) at pH 3.3. After zinc and copper ion adaptation the cells surface became less negatively charged and the IEP shifted to pH 2.2 after copper adaptation and to pH 3 after zinc ions adaptation. This shift in IEP is not due to adsorption of cations on the cells surface but because of altered composition of surface compounds as revealed by XPS analysis.
The DRIFT spectra of minerals treated with cells showed the absorbance bands corresponding to cells surface chemical composition identifying cells adhesion to minerals. At. ferrooxidans and L. ferrooxidans DRIFT spectra are comparable and the metal ions adaptation resulted only in minor changes in absorption peak shapes and intensities. The contact angle data showed that both cells have similar hydrophobic/hydrophilic properties. Copper and zinc adaptation increases the total surface energy and the polar character of bacterial surface. The total surface energy of S. cerevisae is higher but less polar compared to other bacteria. However, all cells possess a dominant electron donating character.
i
At. ferrooxidans cells altered the surface properties of pyrite and chalcopyrite in different ways, where the IEP shifted to acidic and basic regions in case of pyrite and chalcopyrite respectively. The adhesion of L. ferrooxidans cells on minerals lowered the IEP of minerals, approaching close to that of the cells IEP. The changes in surface charge properties of minerals are corroborated with the results of settling tests. L. ferrooxidans cells selectively depressed the flotation of chalcopyrite but the flotation ability of pyrite remained intact. Yeast cells were successful in selective coagulation of sphalerite and galena fine particles but not silicates. Flotation found to be suitable method for the separation of selectively formed sulphide biocoagulates. The microbial adhesion is essential and critical for the success of bioflotation and bioflocculation processes. The adsorption of bacterial strains onto mineral surfaces is observed to be a fast process and the adsorption densities differed on pyrite, chalcopyrite and sphalerite minerals. Iron (II) grown At. ferrooxidans cells adhere more on pyrite than chalcopyrite but the cells adapted to copper and zinc ions adhered similarly on these minerals. Adhesion to sphalerite is the lowest in all the cases of bacteria. The adsorption of L. ferrooxidans cells on chalcopyrite is higher and also depresses its flotation more compared to pyrite flotation. Differences in adhesive characteristics are explained by electrostatic forces arising due to surface charges and/or considering the character of surfaces involved in the thermodynamic approach. But the extended DLVO theory incorporating dispersive, electrostatic and acid base interaction energies is found to be the more suited for microbial adhesion predictions onto minerals and was in a good agreement with the experimental flotation and flocculation results.
Keywords: Bioflotation; Bioflocculation; Sulphide minerals; Zeta-potential; Adhesion; Contact angle; Surface energy; DLVO theory; Acidithiobacillus ferrooxidans; Leptospirillum ferrooxidans, Adaptation.
ii
ACKNOWLEDGEMENTS
Firts of all, I would like to express my sincere gratitude to my supervisor, Professor K. Hanumantha Rao for the oportunity to explore this subject, his guidance, supervision, advices and support throughout my work. I would like to thank Professor Eric Forssberg for making it possible for me to come here and pursue my doctoral studies. I would like to thank to Dr. Bertil Pålsson for his caring guidance, special thanks to Ulf Nordström for his help with the surface area measurements and Maine Ranheimer for introduction to the world of infra red spectroscopy. I would also like to express my thanks to my friends and coleagues at the Department of Chemical Engineering and Geosciences for making the time spent valuable. The financial support from the EU BioMinE project (contract no. IP NMP2-CT-2005500329) and from Kempestiftelsen foundation in the form of scholarship is gratefully acknowledged. I wish to thank to my friend Ranjan Dwari for the endless discussions in our labs and outside them. I would like to thank to Amjad Alhalaweh for his kindness and cooperation. My thanks extend to the staff of Institute of Montaneous Sciences and Environmental Protection of Technical University in Košice for the continued help in my studies. My thanks also extend to the Aquatic Biotechnology group of the Biofilm centre of University of Duisburg-Essen in Duisburg for the six months spent with them and for the kind help and guidance with microbiology and atomic force microscopy. I wish to express my sicere gratitude to my small family and to those also who are not with us. My very special thanks is addressed to my Mamcsu, who gave me a lot of love and support from the first moments of my life till now thousands of kilometers away. Once more I would like to express my gratitude to my mom and Nagyika and Nagyapa by dedicating this thesis to them.
Annamaria Vilinska October 2009 Luleå
iii
LIST OF APPENDED PAPERS
Paper I Vilinska, A., Hanumantha Rao, K., Forssberg, K.S.E., 2008. Miccoorganisms in Flotation and Flocculation of Minerals – An overview. In: proceedings of XXIV International Mineral Processing Congress, D.Z. Wang et al. (Eds.), Science press, Beijing, 2008, 1, pp. 22-39. Paper II Vilinska A. and Hanumantha Rao K., 2009. Surface characterisation of Acidithiobacillus ferrooxidans and copper and zinc ions adapted cells. Geomicrobiology, submitted. Paper III Vilinska A. and Hanumantha Rao K., 2009. Adhesion to sulphide minerals by wild-type and of Cu- and Zn-adapted cells of Acidithiobacillus ferrooxidans. Applied Microbiology and Biotechnology, submitted. Paper IV Vilinska, A., Hanumantha Rao, K., 2009. Surface Thermodynamics and Extended DLVO Theory of Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite. The Open Colloid Science Journal. 2, pp. 1-14. Paper V Vilinska A. and Hanumantha Rao K., 2009. Surface thermodynamics and extended DLVO theory of Leptospirillum ferrooxidans cells adhesion on pyrite and chalcopyrite. Journal of Colloid and Interface Science, to be submitted. Paper VI Vilinska, A., Hanumantha Rao, K., Forssberg, K.S.E., 2007. Selective coagulation in chalcopyrite/pyrite mineral system using Acidithiobacillus group bacteria. Advanced Materials Research. 20-21, pp. 366-370. Paper VII Vilinska, A., Hanumantha Rao, K., 2008. Leptosririllum ferrooxidans-sulfide mineral interactions with reference to bioflotation and bioflocculation. Transactions of Nonferrous Metals Society of China. 18, pp. 1403-1409.
v
Paper VIII Halit Z. Kuyumcu, Tina Bielig, Annamaria Vilinska and K. Hanumantha Rao, 2009. Biocoagulation and its Application Potentials for Mineral Bioprocessing. The Open Mineral Processing Journal. 2, pp. 1-11. Paper IX Hanumantha Rao, K., Vilinska, A., and Chernyshova, I.V., 2009. Microorganisms in bioflotation and bioflocculation: potential application and research needs. Advanced Materials Research. 71-73, pp. 319-328.
vi
Contents Abstract………………………………………………………………………………….. Acknowledgements………………………..…………………………………………….. List of Appended Papers……………………………………………………………….. Contents…………………………………………………………………………………. 1 INTRODUCTION…………………………………………………………………….. 1.1 Microorganisms in mineral beneficiation……………………………………….. 1.1.1 Bacteria…………………………………………………………………….…... 1.1.1.6 Genus Thiobacillus ………………………………………………………... 1.1.1.7 Acidithiobacillus…………………………………………………………… 1.1.1.8 Acidithiobacillus ferrooxidans…………………………………….………. 1.1.1.9 Genus Leptospirillum……………………………………………………… 1.1.1.10 Leptospirillum ferrooxidans……………………………………………… 1.1.2 Yeast…………………………………………………………………………… 1.2 Bioflotation and bioflocculation studies……………………………..………….. 1.2.1 Microorganisms as collectors…………………………………………………. 1.2.2 Microorganisms as depressants………………………………………….……. 1.2.3 Chemolithotrophic bacteria…………………………………………………… 1.2.4.1 Electrokinetic studies………………………………………………………. 1.2.4.2 Spectroscopy studies……………………………………………….………. 1.2.4.3 Hydrophobicity and surface energy studies……………………….……….. 1.3 Microbial Adhesion………………………………………………………………. 1.4 Theoretical background of bacterial adhesion………………………………….. 1.4.1 Surface energy and contact angles…………………………………….……… 1.4.1.1 Equation of state…………………………………………………………… 1.4.1.2 Geometric mean approach………………………………………..………... 1.4.1.3 Acid-base method according to van Oss and Good………………………... 1.4.2 Thermodynamic approach…………………………………………………….. 1.4.3 DLVO approach……………………………………………………………….. 1.4.4 New approaches……………………………………………………………….. 2 MATERIALS AND METHODS………………………………………………..…… 2.1 Mineral samples preparation…………………………………………………...... 2.2 Bacterial cultivation…………………………………………………………….… 2.3 Zeta potential measurements………………………………..…………………… 2.4 Contact angle measurements…………………………………………………….. 2.4.1 Bacteria………………………………………………………………………… 2.4.2 Minerals………………………………………………………………………... 2.5 FTIR measurements……………………………………………………………… 2.6 XPS measurements……………………………………………………………….. 2.7 Adsorption measurements…………………………………………………..……. 2.8 Settling/Flocculation experiments…………………………………..…………… 2.9 Flotation experiments…………………………………………………………..… 3 RESULTS AND DISCUSSION……………………………………………………… 3.1 Characterisation of minerals and microorganisms……………………………... 3.1.1 Bacterial growth characterisation…………………………………………….. 3.1.2 Zeta-potential…………………………………………………………………... 3.1.3 Surface energies of cells and minerals……………………………………..…. 3.1.4 FTIR studies…………………………………………………………………… 3.1.5 XPS studies………………………………………………………….…………. vii
i ii v vii 1 1 1 1 2 2 2 3 3 4 4 4 5 5 6 6 7 8 8 9 9 10 10 10 12 13 13 13 14 14 14 14 15 15 15 16 16 17 17 17 17 19 21 23
3.2 Microbial Adhesion to minerals………………………………………………….. 3.2.1 Adsorption studies……………………………………………………………... 3.2.2 Theoretical approaches of bacterial cells adhesion on minerals…………….. 3.2.2.1 Thermodynamic approach…………………………………………………. 3.2.2.2 Extended DLVO theory approach…………………………………………. 3.3 Bioflotation and Bioflocculation…………………………………………………. 3.3.1 Flocculation studies…………………………………………………………… 3.3.2 Flotation studies……………………………………………………………….. SUMMARY……………………………………………………………………………... FUTURE WORK……………………………………………………………………….. REFERENCES…………………………………………………………………………..
viii
25 25 28 28 29 38 38 40 43 45 47
1 INTRODUCTION The use of various microorganisms in mineral beneficiation has shown to be promising in recent years. Biobeneficiation processes mainly compose of bio-flotation and bio-flocculation where selective removal of minerals is achieved by application of microorganisms and their products. Several studies showed that the microorganisms could substitute traditional chemical reagents. The adhesion of microorganisms to mineral surfaces results in alteration of the surface chemistry of minerals relevant to beneficiation processes due to a consequence of biofilm formation on the surface. Several types of heterotrophic and chemolithotrophic bacteria have been successfully investigated in laboratory biobeneficiation studies. The present research has been focussed on the use of microorganisms associated with sulphide minerals that is acidophilic sulphur and iron oxidizing Acidithiobacillus ferrooxidans and iron oxidizing Leptospirillum ferrooxidans. Both species are widely researched and employed in biohydrometallurgical operations.
1.1 Microorganisms in mineral beneficiation 1.1.1 Bacteria Thiobacilli are a collective group of microorganisms in acidic mines and are active in the leaching of sulphide minerals. Their ability to oxidise Fe2+ to Fe3+ ions and elemental sulphur in acidic solution is familiar. The most important bacterium is the Acidithiobacillus ferrooxidans and its close associates are At. thiooxidans and Leptospirilium ferroxidans. These bacteria obtain energy through the oxidation of sulphide minerals and thus were used in the solubilisation of metal sulphides. In the present studies relevant to biobeneficiation of sulphide minerals, these mineral habitat species of genus Thiobacillus and Leptospirillum have been chosen. 1.1.1.1 Genus Thiobacillus (Beijerinck 1904) The genus Thiobacillus cells according to Bergey’s manual of systematic bacteriology (1989) are Gram-negative and rod shaped small cells (0.5 x 1-4 ȝm) with some motile cells containing polar flagella. These cells derive energy from the oxidation of sulphur compounds (sulphides, sulphur, thiosulfate). All species fix CO2 by the Benson-Calvin cycle and are capable of autotrophic growth and some species are obligately chemolitotrophic. Optimal growth conditions vary between pH 2 and 8, and temperature within the range of 20–43°C. Natural occurrence of the bacteria is in different environments containing oxidizable sulphur compounds. The factor unifying the species of Thiobacillus into one genus was their ability to grow by using the energy through sulphur oxidation, which indicates a very heterogeneous group. The genus was established on the basis of the discovery of T. Thioparus and T. Denitrificans and later T. Thiooxidans. Also Acidithiobacillus ferrooxidans (formerly Thiobacillus ferrooxidans) was classified on the basis of its autotrophic sulphur metabolising capacity, rather than on its iron oxidation capacity. The later one is a classic chemolithotropic Thiobacillus, according to its ability to oxidise sulphur, DNA base composition and morphology. However, a number of morphologically and genomically diverse strains were obtained during the isolation of At. ferrooxidans. The iron oxidizing vibrios Leptospirillum 1
ferrooxidans were isolated in Bulgaria and Armenia (Markosyan 1972; Balashova et al 1974). After 16S rRNA gene sequence analysis some species of Thiobacillus were reassigned. The species of Ȗ-subclass of Proteobacteria were divided into three groups: Acidithiobacillus, Thermithiobacillus and Halothiobacillus (Kelly and Wood 2000). 1.1.1.2 Acidithiobacillus (Kelly and Wood 2000) Obligately acidophilic (with optimum pH below 4) aerobic, Gram negative and rod shaped bacteria, and motile by one or more flagella belongs to the Ȗ-subclass of Proteobacteria. Reduced sulphur compounds are used to support autotrophic growth, where some species are able to oxidise ferrous iron or use metal sulfides to obtain energy and some other species oxidize hydrogen. Optimum temperature for growth is 30–35°C for mesophilic species and 45°C for moderately thermophilic species. Acidithiobacillus thiooxidans represents this type of strain, At. ferrooxidans and At. caldus are the other representatives of the genus. 1.1.1.3 Acidithiobacillus ferrooxidans (Temple and Colmer 1951) Usually these bacterial species are single rods of 0.5x1 ȝm size, motile (single polar flagellum), obligate chemolithotroph and autotroph. Growing on ferrous iron, pyrite, sulphide minerals, sulphur, thiosulphate or tetrathionate is strictly aerobic. Optimum temperature for the growth is 30-35°C, and it’s also capable of growth within 10–37°C. Optimum pH is about 2.5, but tolerates a pH range 1.3–4.5. They are usually isolated from environments with occurrence of oxidizable iron, sulphide minerals and sulphur, they are prevalent in acid mine drainage, mineral leach dumps and heaps. ATCC 23270 represents this type of strain and several culture collections are genomically closely related to ATCC 23270, while some are not taxonomically related to others.
Fig. 1. AFM visualisation of At. ferrooxidans deposited on a Millipore filter paper, 50x50 ȝm (A) and 5x5 ȝm (B).
1.1.1.4 Genus Leptospirillum As first described by Markosyan in 1972, Leptospirillum was an acidophilic and ferrous iron oxidising vibrio-shaped bacterium isolated in Armenia from mine water. According to the phylogenetic studies, this strain was different from At. ferrooxidans and has its special position separate from any other division of the Bacteria such as Proteobacteria, Gram2
positive bacteria and spirochaetes. The description of Leptospirillum was vibroid or spirilla shaped gram-negative cells of 0.2–0.6x0.9–3.5 ȝm size, motile by a single polar flagellum, aerobic and acidophilic. They grow in a pH range 1.3–4.0 and mesophilic or moderately mesophilic conditions (maximum temperature 55°C). Growth is chemoautotrophic by oxidation of ferrous iron and the cells are unable to utilize sulphur or thiosulphate. Growth on pyrite is possible for some species, when the genus consist of two species L. ferrooxidans and L. thermoferrooxidans (Hippe 2000). 1.1.1.5 Leptospirillum ferrooxidans (Markosyan 1972; Balashova 1974) Leptospirillum ferrooxidans are small curved-rod shaped Gram-negative cells of 0.3–0.6x1.0– 3.3 ȝm size and motile by a single polar flagellum. They grow in acidic environment on mineral medium containing ferrous iron at an optimum pH 2.5–3.0. Only Fe2+ ions are used as an energy source and unable to utilize sulphur or organic compounds for growth. The cells are aerobic and mesophilic. DSM 2705 represents this type of strain and it was isolated from the mines of Alaverda copper deposit in Armenia (Hippe 2000).
Fig. 2. AFM visualisation of L. ferrooxidans deposited on a Millipore filter paper, 20x20 ȝm (A) and 5x5 ȝm (B).
1.1.2 Yeast Yeast cells are unicellular Fungi and thus eukaryotic cells. Eukaryotic cells are bigger in size and have more complex structure with real nucleus and organelles. Cell wall of yeast cells lacks peptidoglycan and is mostly composed of polysaccharides such as chitin, glucans and mannans. Most yeast species are capable of forming spores. Saccharomyces cerevisiae (Meyen ex Hansen 1883) is a key participant in fermentation processes and the main component of commercial yeasts. S. cerevisiae cells are ellipsoidal shape of few micrometer sizes (5–10x1–7 ȝm) and grow aerobically and anaerobically using organic carbon as energy source. Reproduction time is rather short (1.5–2.0 hours) and thus eukaryote became a model cells in various studies. Their genome was completely sequenced. Yarrowia lipolytica (van der Walt and von Arx 1980) as indicated from their name is capable to degrade lipids to more simple organic products. Y. lipolytica is isolated from different food media and forming different shapes and sizes (viz., ovoid, ellipsoidal, spheroidal, and cylindrical) during its life cycles. 3
Fig. 3. AFM of Saccharomyces cerevisiae (black arrow) and Sphalerite (white arrow)
1.2 Bioflotation and bioflocculation studies Bacterial strains used in mineral beneficiation are mostly heterotrophic and this type of bacteria requires organic carbon as energy source for growing the culture. These bacteria produce huge amounts of secretion products compared to chemolithotrops which grow through the oxidation of inorganic compounds. The need of large-scale bacterial mass with associated secretion products to fully cover mineral surfaces and alter the surface properties could be the reason for using heterotrophic microorganisms. The rational procedure is to choose surface specific cells that would preferentially adhere to a desired mineral surface causing their selective flocculation or changing the wettability character either to promote or depress the mineral in flotation process. 1.2.1 Microorganisms as collectors Mycobacterium phlei is a fast growing rod shaped prokaryotic cell with mycolic acid covered surface and highly hydrophobic. Raichur et al (1996) used this bacterium to separate mineral matter from coal by flocculation and flotation. The adhesion of the bacteria increased the contact angles of coal and therefore hydrophobicity. In adition to coal fines M. Phlei is observed to be a good flocculant for phosphates and hematite (Smith and Mishra 1991). Zheng et al (2001) used the same species in the flotation separation between dolomite and apatite. The presence of M. phlei also increased the contact angles and hydrophobicity of hematite (Dubel et al 1992). Rhodococcus opacus surface also dominates hydrophobic character and was used in hematite–quartz flotation system by Mesquita et al (2003) and magnesite–calcite flotation by Botero et al (2007). 1.2.2 Microorganisms as depressants Water is a crucial substance for life and thus the majority of organisms have adapted to its presence. To possess a hydrophilic surface is beneficial for microorganisms living in water based environment and thus if adhered to a mineral surface will decrease its hydrophobicity and therefore flotation. Bacillus subtilis cells behaved as a depressant for dolomite and apatite according to Zheng et al (2001). Activated magnesite tailing were found to be depressed after Aspergillus niger 4
treatment (Gawel et al. 1997). P. polymyxa cells were found to increase the ash rejection of coals (Vijayalakshmi et al 2002). Cells of P. polymyxa and different metabolite products (protein fraction and polysaccharides) were tested on different oxide minerals for their flocculation and flotation behaviour (Natarajan and Deo 2001). Phalguni et al (1996) reported calcium and iron removal from alumina. Desiliconisation of calcite, alumina and iron oxide (Deo and Natarajan 1997) and selective separation of silica and alumina from iron ores (Deo and Natarajan 1998) were reported. Cells and metabolites of P. polymyxa were applied in flotation and flocculation of sulphide minerals (Patra and Natarajan 2006; Sharma and Hanumantha Rao 1999; Sharma et al 2000; Sharma et al 2001). 1.2.3 Chemolithotrophic bacteria Acidithiobacillus ferrooxidans and Acidithiobacillus thiooxidans bacteria are representatives of the Thiobacillus genus of proteobacteria and typical for acidic sulphide mineral environments. Specific origin and growing conditions predetermines such species use in hydrometallurgy and even in mineral processing. Selective flotation and flocculation in galena–sphalerite system was achieved after At. thiooxidans conditioning but after a long interaction time compared to At. ferrooxidans (2 hours). Different solubility of the metal sulphate products on the minerals contributed to different flotation results (Santhiya et al 2001). Desulfurisation of coal by depression of pyrite in the flotation of coal from associated minerals was achieved by Misra et al (1996) and Attia et al (1993) using At. ferrooxidans. Yelloji Rao et al (1992) reported the separation of sulfuric acid treated galena from sphalerite using At. ferrooxidans. Selective depression of pyrite flotation in the presence of other sulphide minerals was reported in several laboratory scale studies (Chandraprabha et al 2005; Sharma et al 1999; Mishra et al 1996) and bench scale floation tests (Hosseini et al 2005). The efficiency of the latter process is influenced with a change in the growth substrate of bacteria (Sharma et al 2000; Natarajan and Das 2003). 1.2.4 Chemolitotrophic bacteria and minerals characterisation Besides flocculation and flotation studies using bacteria, the chemical, spectroscopic and electrokinetic characterisation of microorganisms was carried out by several researchers. 1.2.4.1 Electrokinetic studies Characterisation of surface charge by zeta-potential or electrophoretic mobility measurements provide the surface composition of cells and define the conditions at which the minerals can coagulate or support the adsorption of bacterium and/or collector onto mineral surface. Acidithiobacillus ferrooxidans grown in the presence of Fe2+ ions showed an iso-electric point (IEP) at pH 2 (Natarajan and Das 2003; Sharma et al 2003) or at pH 2.3 (Chandraprabha et al 2004a; Chandraprabha et al 2004b) or a small magnitude of negative potential in the entire acidic pH region (Sharma et al 1999). Solid substrate grown cells had higher values of IEP and also higher magnitude of negative potential. Elemental sulfur grown At. ferrooxidans cells exhibited IEP at a higher pH 4 (Natarajan and Das 2003), pH 3.8 (Devasia et al 1993) and pH 3.2 (Das et al 1999; Sharma et al 1999). Mineral grown At. ferrooxidans had IEP at pH 3.0–3.5 (Sharma et al 2000; Sharma et al 2003; Mishra et al 1996). Acidithiobacillus thiooxidans cells grown on S0 exhibited an IEP at pH 2.9 (Chandraprabha and Natarajan 2006) and pH 3.2 (Santhiya et al 2001). The location of IEP could determine the surface 5
composition of cells (Rijnaarts et al 1995) and the surfaces mainly consisting of glucuronic acids and other polysaccharide compounds have their IEP around pH 2.0–2.8. The IEP of bacteria is located above pH 3.2, if the surface is rich in proteins rather than polysaccharides. The presence of phosphate groups on cells surface leads to an IEP less than pH 2. In general, the IEP of mineral particles after bacterial conditioning shifts to the direction of the IEP of bacterial cells. Thus the minerals with low IEP increased in value after bacterial treatment (Chandraprabha and Natarajan 2006; Santhiya et al 2001), while minerals with higher IEP decreased in value (Sharma et al 1999; Sharma et al 2003). The effect is significant only on sulphide minerals, which represents an energy source for the cells. Other minerals such as quartz were unaffected by the bacterial treatment (Natarajan and Das 2003). The effect of bacterial conditioning is time dependent and in all cases the electronegative character of mineral was decreased. The shift in mineral’s IEP after conditioning with At. ferrooxidans follows the amount of adsorbed cells on the surface (Devasia et al 1993; Chandraprabha et al 2005). The extent of change also depends on the growth conditions of the cells and the IEP change effect on S0 or pyrite grown cells was higher than Fe2+ ions grown At. ferrooxidans (Das et al 1999; Sharma et al 1999; Sharma et al 2000). 1.2.4.2 Spectroscopy studies Fourier transform infrared (FTIR) spectra confirmed the presence of CH, CH2, CH3, NH, NH2, NH3, COOH and CONH groups on the surface of all cells in all studies. Absorption peaks were more intense on solid substrate (S0 and minerals) grown cells spectra, which indicate higher protein content and therefore higher hydrophobicity of the cell surface (Das et al 1999; Sharma et al 2000; Sharma et al 2003). IR spectra of pyrite and chalcopyrite interaction with cells showed increased amide bands, but the change was less significant on chalcopyrite spectra. IR spectra confirmed decreased adsorption of collector if the minerals are preconditioned with bacteria (Sharma et al 1999). X-ray photoelectron spectroscopy (XPS) studies of bacterial cells confirmed the presence of only elemental S on At. thiooxidans cells and the presence of FeO and FeSO4 in only Fe2+ grown At. ferrooxidans cells. The N/C ratio indicated higher amount of amide groups for S grown cells (Sharma and Rao 2005). 1.2.4.3 Hydrophobicity and surface energy studies Conditions of cells growth determine the amount and composition of surface polymers and therefore the surface properties of At. ferrooxidans cells grown in Fe2+ ions and mineral substrates are altered. Direct measurements of cell surface hydrophobicity and hydrophilicity were carried out by liquid-liquid partition in aqueous and organic phase by using nhexadecane. Fe2+ grown At. ferrooxidans cells were relatively more hydrophilic than S0 grown cells, and Acidithiobacillus thiooxidans grown in S0 exhibited the highest hydrophobicity (Natarajan and Das 2003). Pyrite and chalcopyrite grown cells were more hydrophobic compared to Fe2+ grown cells of At. ferrooxidans, since higher amounts of mineral substrate grown cells transferred to organic liquids (Devasia et al 1993; Das et al 1999). Surface free energies calculated from contact angle data of different liquids on bacterial lawns were slightly higher for At. thiooxidans relative to At. ferrooxidans (Sharma and Rao 2005). The surface energies of P. polymyxa increased after mineral adaptation. The increase of acid 6
part of surface energy is remarkable and caused the increase of acid-base part of surface energy, making the surfaces more polar and hydrophilic after sulphide mineral adaptation. P. polymyxa cells grown in pyrite and chalcopyrite substrates had the highest increase in electron accepting character (Sharma et al 2001).
1.3 Microbial Adhesion Microbial adhesion on solid substrate is important in various fields of science especially in biobeneficiation where the effect in flotation/flocculation is a function of adhered cell numbers. Adhesion of microbial cells depends upon its electrostatic, van der Waals and acid/base interactions with the mineral surface. All these interactions are a function of microbial and mineral surface properties such as surface charge, surface hydrophobicity, van der Waals component of microbial cell surface energy, etc. Adhesion of microbial cells on the mineral surface causes the formation of biofilm and imparts its own surface properties to the mineral. This leads to alteration in surface properties of the mineral. Adsorption of Acidithiobacillus ferrooxidans to minerals was found to be efficient and highest towards pyrite followed by other sulphide minerals (Chandraprabha et al 2005). Adsorption kinetics studies also correlated well with different adsorption densities on sulphide minerals with a maximum adsorption on pyrite. In all studies, adsorption towards pyrite was found very fast with an equilibrium state reached within 10-15 minutes (Natarajan and Das 2003; Das et al 1999; Sharma et al 1999). At. ferrooxidans cells adhesion on pyrite depends on the growth conditions and the substrate in which they were grown. Under acidic conditions, S0 grown cells adhered more or attained adsorption saturation faster and reached higher saturation level compared with Fe2+ grown cells (Natarajan and Das 2003). At a neutral pH, Fe2+ grown cells had higher affinity towards pyrite than S0 grown cells (Das et al 1999; Sharma et al 1999). The adsorption on non-sulfide minerals such as quartz was negligible (Natarajan and Das 2003). The amount of adhered cells on sulfide mineral surface depends on the actual strain of At. ferrooxidans, some strains are more specific and others varied the adhesion on pyrite (Harneit et al 2006). Variations in the attachment abilities of different strains of L. ferrooxidans onto pyrite are also reported (Ghauri et al 2007). Adhesion on the mineral surface is not a random process and the cells of At. ferrooxidans and L. ferrooxidans preferentially adsorb onto the sites with surface imperfections (Rohwerder et al 2003) and both cells possess chemotaxic abilities to Fe2+/ Fe3+ gradients (Acuña et al 1992; Meyer et al 2002). Besides surface imperfections, cells tend to create regular pentagonal or hexagonal structures on the mineral surface (Sand et al 2009; Sanhueza et al 1999) depending on the history of mineral formation and crystal structure. Full coverage of mineral surface is usually not achieved and an “attachment anomaly” was described by Sand et al (2009). The surface is covered to a maximum of 5-10% when there are still available unattached (planktonic) cells in the solution. The biofilm is then produced from the division of already attached cells rather than newly attached cells. This suggests that the whole mineral surface is not suitable for attachment and only some sites posses the characteristics favouring the adhesion (polarity difference, hydrophobic interactions).
7
As mentioned earlier the cells adhere to some surfaces depending on the growth medium. The reason for different adhesive capacities of the same strain grown under different conditions is probably due to the production of different composition of extracellular polymeric substances (EPS) and their important role in adhesion. The amount of EPS is minimal on iron grown cells, increased several times for the unattached planktonic cells and several folds higher for attached cells on the pyrite surface. The composition of EPS also slightly varied when different substrates were used and it was found to be helpful in the attachment of cells to minerals (Harneit et al 2006). At. ferrooxidans cells grown in the presence of pyrite produced EPS rich in sugars, fatty acids, glucuronic acid and Fe3+ ions. Elemental sulphur grown cells produced EPS of different composition of less sugars and uronic acids but more fatty acids (Gehrke et al 1998).
1.4 Theoretical background of bacterial adhesion 1.4.1 Surface energy and contact angles The process of adhesion of microorganism to a solid surface is a function of properties of the two interacting objects. Surface chemical composition determines the polar or hydrophobic (dispersive) character of surface and possible electrostatic potential. Surface energy and its components give valuable information about the surface character and are used in different theoretical methods to predict the adhesion. The adhesion itself is a result of different physico-chemical forces depending on the physical properties of the interacting bodies and biological forces, which act between the mineral particles and the cells. Physico-chemical forces contain van der Waals interactions, electrostatic interactions due to surface charges, acid-base interactions due to different surface energy and electron acceptability properties of the interacting surfaces. To estimate the effect of physico-chemical interactions two main approaches are known: thermodynamic approach using the energetic balance and the extended DLVO approach incorporating different forces into the final force. To determine the contact angle on a solid material, the basic method is to place a drop of liquid onto a flat surface of the material. The relationship between the interfacial tensions at the three phase contact point is (Young 1805):
JS
J SL J L cos T
(1)
where Ȗs and Ȗl represents the surface tension/energies of the solid and liquid phases, Ȗsl represents the interfacial tension between the phases and ș is the contact angle containing the liquid with the surface. Usually the contact angles are determined for flat surfaces, where a polished piece of mineral is used, however in real systems the mineral is in the form of powder. To measure the contact angle of powder, the Washburn equation is used. When a column of powder is in contact with liquid, the pores between the particles act like small capillaries and the rise of liquid is measurable. The rise is expressed by the Washburn equation:
I2 t
J 1.r. cosT 2K
(2) 8
where I represents the mass of the liquid flow, Ȗl is the surface tension of liquid, r is the capillary radius, Ș is the liquid viscosity and ș is the advancing angle. Capillary radius r could be replaced by capillary constant, determined by using a virtually completely wetting liquid with contact angle 0 and therefore cos ș equal to 1. Regardless of the method for determining the contact angle, there are still two unknowns in Young’s equation, namely the interfacial surface tension and the surface energy of solid. To determine the surface energies therefore several approaches are available. 1.4.1.1 Equation of state Considering that the surface energy of solid is a function of interfacial tension and surface tension of a liquid and using a huge amount of experimental data, an empirical equation combined with Young’s equation was obtained (Neumann et al. 1974): cos T
1 2 J S / J L e E ( J L J S )
2
(3)
where ȕ was calculated to be 0.0001247 mJ/m2. Fowkes (1962) divided the surface energies into two components, dispersive and polar part, while only the dispersive part was considered to be effective:
J SL
J S J L 2 J SD J LD
(4)
1.4.1.2 Geometric mean approach Taking interfacial tension to be a function of liquid and solid surface energies, the number of unknowns in Young’s equation was reduced. According to Good and Girifalco (1957), the interfacial tension is expressed as
J SL
J S J L 2) J S J L
(5)
where ɮ is an interaction parameter depending on the system. Owens, Wendt, Kaelbe combined it with the geometric mean approach to compute the dispersive and polar interactions of interfaces and obtained (Owens and Wendt 1969)
J SL
J S J L 2 J SD J LD 2 J SP J LP
(6)
1.4.1.3 Acid-base method according to van Oss and Good Van Oss and Good expressed the polar part of surface energy by Lewis acid-base model, the polar component of surface energy is equalled to electron-donating and electron-accepting properties of the interacting phases (van Oss et al 1987):
J AB
2 J J
J SL
J S J L 2( J SD J LD J S J L J S J L )
)
(7) (8)
The following equation is obtained while combining Young’s equation with above: 9
(1 cos T )J L
2( J SD J LD J S J L J S J L )
(9)
By measuring the contact angle of at least 3 different liquids with defined acid and base components, the surface energy components of solids could be determined. 1.4.2 Thermodynamic approach According to thermodynamic laws, a system will undergo change and proceeds only towards an energetically favoured state; the state of which after the change will have a lower total energy. In the case of bacterial adhesion to a solid surface, the energetic state of bacteria-solid system has to be lower than bacteria and solid, which can be expressed in terms of free energy of adhesion, ǻGadh (Absolom et al 1983; Busscher et al 1984): ǻGadh = Ȗbs – Ȗbl – Ȗsl
(10)
where Ȗ represents interfacial free energy for different interfaces: bacteria-solid (Ȗbs), bacterialiquid (Ȗbl) and solid-liquid (Ȗsl). Adhesion is energetically favoured only if ǻGadh is negative. For calculation of free energy of adhesion, the interfacial free energies are necessary. By measuring the contact angles with different liquids with known surface energy components, Ȗbl and Ȗsl could be calculated. There are different approaches in literature to determine solids surface energy from contact angles data, but the van Oss acid-base approach (Van Oss et al 1986) was followed since this approach is found to give consistent results (Sharma and Rao 2002; Sharma and Rao 2003) besides providing bacterial cells surface electron-donating and electron-accepting characteristics. For calculation of ǻGadh, the free energy of adhesion was divided into two parts: Lifshitz - van der Waals (LW) component and acid-base (AB) component: ǻGadh = ǻGadhAB + ǻGadhLW
(11)
The components of free energy of adhesion can be calculated from the interfacial tensions as follows: 'Gadh
LW
AB
'Gadh
2§¨ J bv ©
·¸§¨ J LW J LW ·¸ (12) lv ¹© sv ¹ 2§¨ Jbv Jsv ·¸§¨ Jbv Jsv ·¸ 2§¨ Jbv Jlv ·¸§¨ Jbv Jlv ·¸ 2§¨ Jsv Jlv ·¸§¨ Jsv Jlv ·¸ (13) ¹ © ¹© ¹ © ¹© ¹ © ¹© LW
J lv
LW
1.4.3 DLVO approach Classical DLVO approach (Deryagin and Landau 1941; Verwey and Overbeek 1955) includes Lifshitz-van der Waals interactions and electrostatic interactions. LW forces are short range and always attractive, and these are weak forces between neutral stable molecules. The Coulombic electrostatic interactions are long range and could be attractive or repulsive, depending on the surface charge of interacted particles. Acid-base interactions were added later by van Oss (Van Oss et al. 1986) and to involve electron donating-accepting abilities of different materials. The microbial adhesion to solid surfaces is then described by a sum of van der Waals, electrostatic and acid-base forces operating between the surfaces approaching close to one another: 10
Gtotal = GLW + GEL + GAB
(14)
Calculation of these forces depends on the geometry of interacting phases and for a spheresphere system, the following equations were used: Lifshitz van der Waals energy: § · ¨ ¸ § x 2 xy x ·º¨ Aª y y 1 LW ¸ ¨ ¸ (15) « 2 2 ln¨ 2 G ¸» 12 ¬ x xy x x 2 xy x y © x xy x y ¹¼¨¨ 1 1.77§¨ 2SH ·¸ ¸¸ © O ¹¹ © Electrostatic interaction energy:
G EL
º SHa1a 2 ] 1 ] 2 ª 2] 1] 2 1 e NH ln ln 1 e 2NH » « NH a1 a 2 ¬] 1 ] 2 1 e ¼
(16)
Acid-base interaction energy: GAB = ʌaȜǻGadhABe[(do-H)/ Ȝ]
(17)
where H is the separation distance, a1 and a2 are the radii of the interacting particles, ȗ1 and ȗ2 are the zeta-potentials of the interacting particles, ț is the double layer thickness-1, A is the Hamaker constant, d0 is the minimum separation distance between the two surfaces (0.157 nm), Ȝ is the correlation length of molecules in liquid (0.6 nm) and x = H/(a1 + a2), y = a1/a2 For interaction energy calculations, the parameters such as zeta potential, particle size radius and double layer thickness are known or measurable; the only unknown parameter is the Hamaker constant. There are two different methods to evaluate the Hamaker constant that influence the value of Lifshitz van der Waals interaction energy (Van Oss et al. 1986 Van Oss et al. 1987). First method is microscopic approach, where the total interaction is assumed to be the sum of all interactions between the atom pairs, the second method is a macroscopic approach, where the particles and the medium are considered as continuous phases. In the presence of a liquid medium, the van der Waals energy between the particles differs and therefore, the Hamaker constant has to be replaced by an effective Hamaker constant. In case of two different particles (1 and 2) in a medium (3), the effective Hamaker constant is calculated by: A123 = A12 + A33 – A13 – A23
(18)
The value of A12 is considered as a geometric mean of A11 and A22: A12 = (A11.A22)1/2
(19)
A123 = (A111/2 – A331/2)(A221/2 – A331/2)
(20)
Fowkes (1964) proposed the following equation for the calculation of individual Hamaker constants for solid surfaces: 11
A11 = 6ʌr2ȖsLW
(21)
where r represents the intermolecular distance and ȖsLW represents the dispersive component of the surface energy of the solid. According to Fowkes, the value of 6ʌr2 equals to 1.44x10-18 m2 for water and systems with the volume element as metal atoms, CH2 and CH groups which have nearly the same size. From the individual Hamaker constants calculated from the dispersive component of surface energies, the effective Hamaker constant can thus be calculated. Another possible way of obtaining the Hamaker constant is from Lifshitz-van der Waals component of free energy of adhesion, ǻGadhLW, calculated from the contact angle measurements. Then, the effective Hamaker constant is given by: A =- 12 ʌd02 ǻGadhLW
(22)
1.4.4 New approaches The electrostatic interaction energy determination in the DLVO theory was derived for mineral particles using the assumption of rigid solid-liquid interface. While it is the case for minerals, microorganisms often posses a highly heterogenic polymeric surface and therefore could not be considered as rigid particles. A soft particle theory was adopted instead (Ohshima 1994; Gaboriaud et al 2008). According to the soft particle assessment, the surface is composed of a thick polymeric layer, which is ion permeable and may influence the measured surface charge. Generally because the ions located inside the polymers are influencing the electrophoretic mobilities, the actual measurements give higher charge values than really present on the surface (Poortinga et al 2002; Chandraprabha et al 2009). This “overestimation” is more prevalent at higher magnitudes of measured charge (Chandraprabha et al 2009). On the other hand the electrostatic interaction calculations are also modified and the results are increased compared to classical approach (Poortinga et al 2002). Bacterial surface is usually not smooth but consists of several cell surface structures as fimbriae, pilli and S-layers. The first two are short filamentous structures and may be considered as particles of smaller size entering to interaction with mineral surface than the cell. The electrostatic interaction is decreased for smaller particle sizes. Cell surface is highly heterogenic regarding its chemical composition and topography resulting in heterogeneity of forces across the cell body. Atomic force microscopy studies showed differences in force measurements (Fang et al 2000) and cells created aggregates by joining at a special place of their surfaces is easily observable under optical microscope.
12
2 MATERIALS AND METHODS 2.1 Mineral samples preparation Pure mineral samples of chalcopyrite, pyrite, galena and sphalerite were obtained from Gregory, Bottley & Lloyd, UK. The samples were crushed and ground in agate mill. After grinding, the minerals were sorted by wet sieving to obtain size fractions suitable for flotation tests and contact angle measurements (–106+38 ȝm) and a fine fraction minus 38 ȝm. A portion of minus 38 ȝm fraction was further ground and a size fraction of –5 ȝm was obtained by a micron filter cloth sieving in ultrasonic bath, which was used in zeta potential, adsorption, FT-IR and coagulation measurements. The minerals were acid washed with HCl to clean the surface from the oxidation products and with ethanol to clean the surface from any organic compounds. The samples were dried and stored at -10ºC until use. The surface area was measured for both size fractions with BET method (Flowsorb II 2300) and were determined to be 0.06 and 1.02 m2/g for coarse and fine fractions of pyrite, and 0.17 and 1.9 m2/g for coarse and fine size fractions of chalcopyrite, and 1.75 m2/g for fine sphalerite respectively. 2.2 Bacterial cultivation Bacterial cells of Acidithiobacillus ferrooxidans type strain ATCC 23270 was cultured in 9K medium (Silverman and Lungren, 1959) at the following composition: 44.5 g/l of FeSO4.7H2O, 3 g/l (NH4)2SO4, 0.5 g/l MgSO4.7H2O, 0.5 g/l K2HPO4 and 0.1 g/l KCl. The bacterial strain Leptospirillum ferrooxidans DSM 2391 was grown in modified Leptospirillum (HH) medium (40 g/l FeSO4.7H2O, 0,132 g/l (NH4)2SO4 53 mg MgCl2.6H2O, 27 mg KH2PO4, 0,147 g CaCl2.2H2O, 62 ȝg MnCl2.2H2O, 68 ȝg ZnCl2, 64 ȝg CoCl2.6H2O, 31 ȝg H3BO3, 10 ȝg Na2MoO4, 67 ȝg CuCl2.2H2O). Metal ions adapted cells were grown in the presence of 0.3 M Cu2+ (CuSO4.5H2O) or 0.3 M Zn2+ (ZnSO4.7H2O) ions enriched medium. The bacteria were adapted to grow in the presence of the metals by serial subculturing in media containing successively increased concentrations of the metal ions. They were considered to have been adapted when their growth rate and maximum cell yield was comparable to that in simple 9K medium. The pH of the medium was maintained at pH 2 with H2SO4 for At. ferrooxidans and pH 1.8 for L. ferrooxidans. All solutions were made with deionised (distilled) water and sterilized in autoclave at 125°C for 15 minutes. Iron sulphate solution was prepared separately and filtered through Millipore filter, to remove all possible particles and cells. Sterilized Erlenmeyer flasks were filled with 200 ml of medium solution and inoculated with 20 ml of bacterial solution. The flasks were continuously shaken in an orbital shaker at 150 rpm and at 30°C. At exact time intervals, the following parameters were measured: pH, Fe+2 content and number of cells per 1 ml of solution to estimate and define the growth phase of solution. The Fe+2 ions were estimated by volumetric titration method and the cell number was counted using a microscope in Neubauer counting chamber with a defined volume of each square. The bacterial cells used were at the end of exponential growth phase. A 20 ml of the solution is used as an inoculum for the next culture and the rest is filtered through Whatman filter paper to remove all the precipitates and finally through Millipore filter. The bacterial cell mass was washed with pH 2 water to remove all the trapped metal ions and metabolites. The cell mass so collected was used in the investigations. 13
Saccharomyces cerevisiae cell were grown in YPD medium at the following composition; 10 g/l yeast extract, 20 g/l peptone, 20 g/l glucose at 30°C. Collected cells were rinsed with 0.9% solution of NaCl in deionized water. 2.3 Zeta potential measurements Zeta potential measurements were made using ZetaCompact equipped with video and Zeta4 software. The software allows the direct reading of zeta-potential calculated from the electrophoretic mobilites using Smoluchowski (1921) equation. The result is a particle distribution diagram, from which the mean mobilties are recalculated to zeta-potential values. Pyrite, chalcopyrite and sphalerite of -5 ȝm particle size at a concentration of 0.025 g/100 ml were used. Ionic strength of 10-2 M was maintained with KNO3. The solution pH was adjusted using HNO3 and KOH. Solutions with particular pH and constant ionic strength were prepared and then the mineral was added. After 30 minutes conditioning, the pH of the suspension was recorded again and regarded as the pH of the measurement. Zeta potential measurements of bacterial cells were performed similarly; At. ferrooxidans culture grown in different media and L. ferrooxidans grown in HH medium was used. Collected cell mass was washed with pH 2 deionised water solution of H2SO4. Different bacterial concentrations were attempted, but 2.5x107 cells/ml was found suitable for the instrument and accordingly used. The measurements of minerals interacted with bacterial cells were realized, the mineral was conditioned with bacteria for 30 minutes at cell concentration of 0.7x107 cells/ml, where the cells per mineral ratio was 2x1010cells/g of mineral. 2.4 Contact angle measurements 2.4.1 Bacteria Precipitate-free cell solution was filtered through millipore filter paper using vacuum filtration to obtain a uniformly distributed cell layer on the whole area of filter paper. When filtered, the filter paper was placed on a Whatman filter paper for a short time, to remove the excess moisture. After moisture removal, the filter paper is dry enough to mount on a microscope slide glass with a help of double sided adhesive tape. The samples were air dried to remove the rest of water and contact angles were measured. Contact angles were measured by placing a drop of liquid on the bacterial lawn surface; the dynamic contact angle was recorded with KrĦss Easy drop and evaluated with Drop Shape analysis software. A drop of liquid is placed on a flat surface with the help of a syringe and the advancing angle is recorded on a CCD camera. Experimentally a bacterial lawn of several hundred layers was estimated as necessary to obtain a stabilised contact angle. Each measurement was repeated 3 times, the arithmetic mean was calculated as a final contact angle for the particular liquid. 2.4.2 Minerals Contact angle on solid powders was determined by sorption measurements using KrĦss Tensiometer K100 and KrĦss LabDesk 3.1 software. Sorption measurements are estimated using rise of liquids in powder pores as in capillary tubes applying the Washburn equation. Modifying the equation by replacing the capillary radius by a capillary constant, the contact angle can be calculated. The capillary constant was determined using n-Hexane with virtual 14
0° angle, giving a cos ș = 1. Using the obtained capillary constant, the contact angles for other liquids are calculated. Pyrite, chalcopyrite, sphalerite and galena powder samples of 106+38 ȝm size fraction was used, 1 g of mineral was placed into a glass sample tube and was carefully and equally packed each time. Each measurement was repeated at least 3 times and a mean value was reported in the results. For the estimation of surface energy with van Oss acid-base method, contact angles of at least three different liquids are necessary with defined acid and base components. Four standard liquids used were: water, 1-Bromonaphtalene, Diiodomethane and Formamide. Surface tensions and the dispersive and polar part contributions, and polar part divided into acid-base components as reported in the literature were used (Bellon-Fontaine et al 1990; Ström et al 1987). 2.5 FTIR measurements The finest fraction (-5 ȝm) of pyrite and chalcopyrite was used for FT-IR studies. The spectra of cells alone and minerals interacted with cells were recorded. Cells of At. ferrooxidans grown in iron (II) media, copper and zinc enriched iron (II) media and L. ferrooxidans were studied. Cells were collected by filtration, washed to remove precipitates, concentrated by centrifugation (3000 rpm for 30 s) and dried. To preserve the biological material air-drying was done overnight. Minerals were conditioned with cells for 30 minutes using intensive stirring, at pH 2 and 23ºC. After conditioning, the suspensions were filtered and the amount of cells in the supernatant was determined to estimate the amount of adsorbed cells. The mineral samples were washed with deionised water to remove the loosely holding cells and were air dried. Drying was done at room temperature and the samples were stored in a desiccator. For recording the spectra, the diffuse reflectance IR method was used, with KBr matrix. For the cells 2.5 wt% concentration was used, for pyrite 2.5 wt% and for chalcopyrite 1.5 wt% was found to be sufficient. Measurement was done with 100 scans using a Perkin-Elmer FTIR spectrometer. 2.6 XPS measurements Bacterial samples were analyzed in a natural wet environment avoiding freeze drying and potential changes on the surface. The bacterial paste was placed on a sample holder and introduced to a liquid nitrogen pre-cooled chamber while the analysis was preformed at 150°C. The spectra were collected using Kratos axis Ultra electron spectrometer. An Al cathode operated at 150 W was used as the X-ray source. Possible charge created on the surface of the sample was controlled by the spectrometer charge neutralization system. The binding energy (BE) scale was referenced to the C 1s peak of aliphatic carbon at 285.0 eV. 2.7 Adsorption measurements Adhesion tests were carried out using the smallest-sized mineral fraction prepared (-5 ȝm) because of its’ high surface area. The surface area of different minerals was estimated by BET. Mineral – liquid ratio was kept constant during all adsorption tests at 1 g/100 ml while the concentration of bacteria was changed. The adhesion characteristics were studied at 5x106, 107, 5x107, 108, 5x108, 109 and 5x109 cells/ml. L. ferrooxidans and At. ferrooxidans cells and At. ferrooxidans cells grown in the presence of 0.3 M zinc and 0.3 M copper ions were studied simultaneously. In order not to compromise the viability of the bacteria the 15
experiments were run at an acidic pH by keeping the pH at 2 with H2SO4. After 30 min of agitation, the cells remaining suspended in the liquid phase were estimated by a direct microscopic count. 2.8 Settling/Flocculation experiments Settling tests were performed using Turbiscan LAb and Tlab EXPERT 1.13 software. The minerals used were pyrite and chalcopyrite, both ground to -5 ȝm particle size. Mineral concentration was 5 g/l for long time experiments and 2.5 g/l for short time experiments. Bacterial concentration was 2x108 cells/ml and this represents 4x1010 cells/g of mineral. Measurements were made under different pH conditions and at constant temperature of 30°C maintained by the instrument. The pH was adjusted with HNO3 or KOH addition. Predetermined amount of mineral was added to deionised water at a particular pH and after 30 minutes of agitation the final pH was measured again, and considered as the pH of the measurement. In case of measurements with bacteria, before this step, a pre-agitation of mineral with bacteria was done for 30 minutes and the pH was adjusted. The temperature was maintained before and during measurement. The agitation time was kept constant according to previous adsorption studies. The cylindrical glass measurement cell is filled with the suspension. The light source is a pulsed near infrared. Two synchronous optical sensors receive the light transmitted through the sample and light backscattered by the sample respectively at the same time. The optical reading head scans the length of the sample (up to 55 mm), acquiring transmission and backscattering data every 40 μm. From the transmission and backscattering data curves are created according to the scan length. These curves provide the transmitted and backscattered light flux in percentage relative to standards as a function of the sample height in mm. These profiles provide a macroscopic fingerprint of each sample at a specific time. In a schematic way, a transmission is used to analyze clear to turbid dispersions and backscattering is used to analyze opaque dispersions. 2.9 Flotation experiments Pyrite and chalcopyrite of -106+38 ȝm size fraction were used to study the ability of At. ferrooxidans and L. ferrooxidans cells to influence the flotation. Xanthate PIX collector of 5x10-5 M concentration was previously found to be effective with recoveries over 95%. Different cell concentrations and pH values were used, cell agitation time was fixed to 30 minutes prior to collector addition and for collector adhesion another 5 minutes were used. Flotation tests were performed in a glass Hallimond tube using 1 g of mineral dispersed in 100 ml deionised water at a particular cell concentration and pH. The air flow rate was kept constant at 200 ml/min for one minute of flotation time.
16
3 RESULTS AND DISCUSSION 3.1 Characterisation of minerals and microorganisms 3.1.1 Bacterial growth characterisation The growth curve of At. ferrooxidans grown in 9K medium is presented in Fig. 4. The cell number was increasing until the Fe2+ ions were completely oxidized, hence the medium with higher Fe2+ content produce liquor with higher amount of cells. The pH of the solution was relatively stable and only a slight increase was observed at the beginning due to the consumption of H+ ions by the cells, but later it was constant at around pH 2.
Fig. 4. Growth characteristics of At. ferrooxidans in 9K medium.
The decrease of Fe2+ ions with time is due to bacterial oxidation to Fe3+ as follows: 2Fe2+ + 1/2O2 + 2H+ ĺ 2Fe3+ + 2e- + H2O Fe3+ is then abiotically hydrolyzed to Fe(OH)3, causing a recovery of H+ ions: Fe3+ + 3H2O ĺ Fe(OH)3 + 3H+ Formation of precipitates was observed and Fe3+ ions in the presence of sulphate and potassium ions causes jarosite KFe3(SO4)2(OH)6 precipitation (Herbert 1997). Copper and zinc ions adapted At. ferrooxidans cells exhibited the same growth times as unadapted iron(II) grown cells. 3.1.2 Zeta-potential The zeta-potential of iron grown and copper and zinc ions adapted At. ferrooxidans cells and L. ferrooxidans cells is shown in Fig. 5. Unadapted At. ferrooxidans cells were negatively charged in the entire pH range studied without displaying an iso-electric point. An extrapolation of the curve shows that the iso-electric point is probably located at or below pH 1. Cells after adaptation to copper and zinc metal ions were positively charged below pH 2.2 and pH 3 respectively. The iso-electric point below pH 3 indicates an anionic polysaccharide surface. Only the zinc grown cells had a higher iso-electric point which is typical for peptidoglycan surface but still too low to be considered proteinaceous surface (Rijnaarts et al 1995). The magnitude of electronegativity was less for zinc grown cells. A shift in iso-electric point and decrease in surface charge of bacteria could be due to the adsorption of specific 17
ions, in this case the metal cations, and secondly a change in surface chemical composition due to secretion of different compounds caused by the varied growth environment. On the other hand L. ferrooxidans exhibited much wider magnitude of potential change as pH was varied, below pH 3 the surface was positively charged and at higher pH, the surface had a negative potential. The IEP was located at pH 3.3. Chemical groups on cell surface protonate or deprotonate depending on the pH and are responsible for a change in surface charge. The IEP is generally higher and typical for solid substrate grown At. ferrooxidans cells with higher amount of EPS and proteins on the surface. Higher magnitude of surface charge, either positive or negative, influences the adsorption process due to electrostatic interaction forces, which depend on the charge polarity and charge magnitude of interacting particles.
At.f. Fe grown
20
At.f. Zn grown
15
At.f. Cu grown
Zeta-potential (mV)
10
L.f. Fe grown
5 0 -5 0
2
4
6
8
12
10
-10 -15 -20 -25 -30 pH
Fig. 5. Zeta-potential of Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans relative to pH. Pyrite with A. f.
30
Pyrite Calcopyrite with A. f. Chalcopyrite
Zeta potential (mV)
20
10
0 1
2
3
4
5
6
7
8
9
10
11
-10
-20
-30
pH
Fig. 6. Zeta potential of minerals, and minerals interacted with At. ferrooxidans cells.
The zeta potentials of pure pyrite and chalcopyrite were similar, both exhibited IEP around pH 7.5 and the charge magnitude below and above IEP is considerably high (Fig. 6). After At. ferrooxidans cells interaction with minerals, there was no significant change in zeta-potential 18
and the most obvious change was the shift in IEP. In the case of pyrite, the IEP moved to acidic pH value whereas it shifted to a more basic pH for chalcopyrite. While before bacterial interaction pyrite and chalcopyrite had the same IEP at pH 7.5, but after bacterial treatment, the IEPs were differed by 2.5 pH units difference. The IEP defines one of the conditions for coagulation since uncharged particles coagulate easily in the absence of strong repulsive electrostatic forces. Therefore, modifying the pH of the suspension, it is possible to coagulate either pyrite or chalcopyrite selectively from a mixed composition with bacterial treatment.
chalcopyrite & L. f.
30
chalcopyrite pyrite
Zeta potential (mV)
20
pyrite & L. f. 10 0 1
3
5
7
9
11
-10 -20 -30 pH
Fig. 7. Zeta potential of minerals, and minerals interacted with L. ferrooxidans.
Zeta potentials of L. ferrooxidans cells interacted with minerals were measured after the minerals are acid washed to remove the surface oxidation products, and the zeta-potential curves of the mineral shown in Fig. 7 were not identical with the previous ones. Pyrite exhibited IEP at pH 7.5 and chalcopyrite at pH 6.5, which are comparable with previous results (Fig. 6). After L. ferrooxidans addition, the IEPs dramatically decreased to pH 4.5 for pyrite and to pH 3 for chalcopyrite. Generally the zeta-potential after bacterial addition was decreased, following the behaviour of L. ferrooxidans curve for both minerals. The zeta potentials until pH 5 were only slightly negative and at higher alkaline pH, the magnitude of negative potential was higher. Chalcopyrite remained at low potential in wide pH region and only of above pH 9 the negative surface charge started to increase dramatically. At basic pH, the zeta potentials of minerals treated with cells were more or less the same with pure mineral potentials. The bacterial treatment of minerals affected the surface charge properties and the zeta-potential curves closely matching to L. ferrooxidans cells potentials suggest the cells adhesion on particles. 3.1.3 Surface energies of cells and minerals The contact angles of different sulphide minerals and microorganisms are presented in Table 1. All minerals showed relatively high contact angle with water compared to cells, demonstrating hydrophobic character of the mineral surfaces. From the data it is clear that the surface of cells is more hydrophilic compared with the minerals and after interaction, the surface hydrophobicity of minerals is reduced. The surface energies were calculated using van Oss acid-base approach to determine not only the dispersive and polar components of surface energies, but also to define the acid and base character of surfaces. 19
Surface energies of solids and microorganisms were similar; varied from 40 to 55 mJm-2. Surface energy of pyrite was the highest relative to other sulphide minerals. All minerals showed relatively low polar surface energy component relative to the total surface energy, demonstrating hydrophobic character of the mineral surfaces. The total surface energy decreased in the direction pyrite > galena > chalcopyrite > sphalerite. The polar contribution to the surface energy followed the order pyrite > chalcopyrite > galena > sphalerite, which correlates with the amount of hydrophobic sulphur species on the surface of these minerals. The results also show that the electron donating character of the minerals is dominant but in a reverse order to that of the polar contribution. The electron donating and accepting character were more balanced for pyrite and chalcopyrite. Surface energies of cells were comparable (48.34 and 55.36 mJm2) but the dispersive and polar parts differed. L. ferrooxidans cells had higher polar surface energy component than At. ferrooxidans but the At. ferrooxidans exhibited higher dispersive part component. Both bacterial species were more polar than the minerals and this is due to their predominant surface functional groups. Electron donating character was dominating with a much higher value of Ȗ- than Ȗ+ due to the occurrence of compounds containing –OH groups (sugars, carboxylic acids, lipids, etc.) characteristic for bacterial surface (Gehrke et al 1998). Table 1. Contact angle and surface energy data of L. ferrooxidans, yeast cells and At. ferrooxidans cells grown under different conditions and sulphide minerals. Contact angle ș DiiodoBromoMaterial Water methane naphtalene Formamide Pyrite 61.00 --10.40 8.15 Chalcopyrite 71.76 --39.66 43.46 Sphalerite 68.70 --30.88 53.40 Galena 67.02 --3.27 37.81 At.f. Fe grown 33.00 53.50 36.00 32.50 At.f. Zn grown 26.57 --40.00 29.73 At.f. Cu grown 7.50 --35.55 14.80 L.f. Fe grown 30.50 55.30 42.30 32.50 S. Cerevisiae 33.53 --18.97 26.34
Surface energy Ȗ (mJ.m-2) total 53.75 42.39 40.64 49.93 48.55 49.54 55.36 48.34 51.66
Ȗd/ȖLW 43.87 34.93 38.50 44.53 34.31 34.78 36.67 32.43 42.21
ȖP/ȖAB 9.88 7.47 2.13 5.40 14.24 14.76 18.68 15.91 9.50
Ȗ+ 4.01 2.20 0.08 0.85 1.11 1.06 1.55 1.36 0.53
Ȗ6.08 6.34 14.51 8.60 45.57 51.46 56.43 46.47 42.00
The water contact angle on At. ferrooxidans cells after metal ion adaptation has decreased for both Cu and Zn ions, indicating that the adapted cell surface became more hydrophilic. The total surface energy of cells increased in the order of unadapted cells zinc adapted copper adapted. The increase in the polar surface energy component of the cells followed the same order. Thus At. ferrooxidans cells surface became more polar with metal ion adaptation. For both cases of metal ion adaptation, the electron donating character was enhanced but the electron accepting character decreased for zinc adapted cells and increased for copper adapted cells. The presence of metal ions in the growth medium has probably changed the cell’s surface chemical composition due to enhanced secretion of metabolites rich in –OH and – COOH groups contributing to the enhanced polar surface structure and electron donating character. S. cerevisiae cells expected to be hydrophilic due to their higher water-wetting abilities, but a low contact angle with a non-polar liquid display the non-polar character of their surfaces. A significant portion of the surface energy attributes to the polar part where the electron 20
donating part is prominent. The strong electron donating character of the microorganisms can be explained by the predominance of polysaccharide compounds and extra-cellular polymeric substances (EPS) rich in –OH groups on the cells surface. 3.1.4 FTIR studies FTIR spectra recorded for the bacterial samples are presented in Figs. 8 and 9 and the assignment of absorption bands are summarized in Table 2. Copper and zinc adapted cells spectra were almost overlapping while iron grown cells produced low intensity signal. There was also some change in the shape of the absorption band peaks, especially the one at 3300 cm-1 and around the 1700-1500 cm-1 region. A broad band around 3300 cm-1 could represent different groups but considering the composition of cell surfaces, the –OH groups seems reasonable due to high sugar contents. The hydrocarbons are represented by numerous peaks between 3069 and 2855 cm-1 and one asymmetric deformation at 1456 cm-1. The 1700 cm-1 region is typical for carbonyl groups (aldehydes, ketones, esters), the actual frequencies differed few cm-1 for different samples, probably due to different environment around the functional group. Nitrogenous substances are distinguished by the existence of different bands from 1653 to 1500 cm-1. Copper and iron grown cells spectra showed more recognizable nitro group bands (1558 and 1519 cm-1) compared to zinc grown. At. ferrooxidans spectrum of iron grown cells displayed more bands at this region characteristic of different C=N or C-NO2 groups. At lower wave numbers some sulphur compounds and CH-NH2 bands were seen.
Fig. 8. FTIR absorbance diagram of At. ferrooxidans cells. Cells were grown in iron(II) containing medium (Fe grown) and in iron media enriched with copper ions (Cu grown) and zinc ions (Zn grown).
21
Fig. 9. FTIR absorbance diagram of L. ferrooxidans.
DRIFT spectra of At. ferrooxidans and L. ferrooxidans were almost identical with no changes in peak positions. The hydrocarbon groups bands (around 3000 cm-1) and N substances bands (1650, 1540 cm-1) were more intense in At. ferrooxidans spectrum, but the spectra were unable to depict the difference between At. ferrooxidans and L. ferrooxidans surface chemical composition. The bands identified on both cell samples represent typical organic compounds such as hydrocarbon groups (CH3, CH2, and CH), carbonyl groups, nitrogenous substances and sulphur compounds. Table 2. FT-IR band identification for At. ferrooxidans cells (Colthup et al 1990; Schrader 1995).
Wavenumber (cm-1) 3300 3069 2959 2925 2875 2855 1733 1717 1653 1558 1540 1519 1456 1234 1159 1075 1046
Groups -OH, -NH2, -NH, CŁC =CH -CH3 -CH2 -CH3 -CH2 C=O C=O -C=N C-NO2 NO C-NO2 CH3 CH2-S SO2 CH2-NH2, CH2-OH CH-NH2
22
Description
aliphatic CH3 aliphatic CH2 aliphatic CH2 carbonyl group (aldehydes) carbonyl group (ketones, esters) 1st amide band aliphatic nitro aliphatic nitro (2nd amide band) aromatic nitro aliphatic CH3 (asymetric deformation)
Fig. 10. FTIR spectra of pyrite conditioned with a) At. ferrooxidans and b) L. ferrooxidans.
Spectra of pyrite interacted with At. ferrooxidans and L. ferrooxidans display new absorption bands (Fig. 10); characteristic peaks from bacterial spectra appeared, ascertaining the adsorption of cells on the mineral surface. The characteristic peaks for pure pyrite were reduced after interaction with cells. At. ferrooxidans spectrum bands were seen at lower cell concentration, while the bands of L. ferrooxidans became significant only at higher cell concentration. On the other hand, L. ferrooxidans bands around 3300 cm-1 region were visible, but were not recognisable in At. ferrooxidans spectra. DRIFT spectra of chalcopyrite and chalcopyrite after interaction with cells were also recorded, but there were no new bands in the mineral spectra interacted with At. ferrooxidans and L. ferrooxidans cells suggesting that the cells are not adsorbed on the chalcopyrite surface. The adsorption results showed that a comparable adsorption of cells onto chalcopyrite also takes place and the cells are not washed out during the samples preparation for FTIR spectra. The absorbance bands of cells might have been masked by the mineral high intensity bands. 3.1.5 XPS studies Since the X-rays used for XPS analysis penetrate up to a depth of 10 nm of the bulk solids, the spectral data represent the bacterial surface layer. The presence of different metal ions during bacteria growth has not changed the composition of the surface markedly and only minor changes are noticed (Fig.11 and Table 3). Some functional groups were affected to a minor degree, but carbon groups associated with oxygen showed significant changes (Fig. 9). 23
The C-(O, N) groups remained similar after copper ions adaptation but were increased for zinc adapted cells. The percentage of C=O and C-N-O groups decreased for zinc ions adapted cells but their amount increased in the case of copper adapted cells. Organic oxygen has decreased in the order of unadapted, zinc and copper adapted cells. The Fe-ions were expected to be found on cells surface either as ions or precipitate but were found only on the iron grown cells. In the case of adapted cells, Fe content was below the detection limit. The iron found for the case of original cells was probably in the form of ions and not as precipitates, because no potassium was detected in XPS results. The iron was present in Fe2+ and Fe3+ oxidation states, consistent with the chemical shifts of the peaks and in the ratio of 1:2. No copper or zinc metal ions were detected on the cell surface.
Atomic concentration (%)
30
Fe grown Zn grown Cu grown
25 20 15 10 5
II& III ) Fe (
Su lp ha te Ph os ph at e
rg an ic ) (o
N
S
N
(p ro
to na te d)
O H (o r ga O ni (o c) rg an ic ,H 2O )
O
O
C
-O ,N =O ,N -C -O C
C
C
-( C
,H
)
0
Fig. 11. XPS analysis of different At. ferrooxidans cells. 0,5
Fe grown
Atomic ratios
0,4
Zn grown Cu grown
0,3 0,2 0,1 0,0 N/C
P/C
O/C
CH/C
C(O,N)/C C=O/C COOH/C
NH+/N
Fig. 12. XPS analysis of different At. ferrooxidans cells, relative elemental visualisation.
The nitrogen-carbon ratio was the highest for unadapted cells followed by copper grown and zinc grown cells of At. Ferrooxidans (Fig. 12). The proteins were mostly present embedded in the cell wall and the decrease of nitrogen compound compared with carbon suggests the secretion of some polysaccharides on the cells surface. The amount of total oxygen to total carbon decreased in the order of unadapted, zinc and copper adapted cells. The amount of aliphatic carbon from the total carbon decreased after cell adaptation; C-(O, N) groups increased for zinc adapted cells and remained almost equal after copper adaptation. The number of carboxylic groups to the total carbon remained the same after zinc adaptation but 24
decreased for copper adapted cells. Unadapted At. ferrooxidans cells had a much higher protonated nitrogen groups compared to metal ions adapted cells. Table 3. XPS analysis of different At. ferrooxidans cells. Line
C 1s
O 1s N 1s S 2p 3/2 P 2p 3/2 Fe 2p 3/2
Fe grown BE, AC, eV at.% 285.0 28.86 286.7 26.38 288.2 9.82 289.3 1.04 531.6 4.86 533.1 23.08 400.2 4.09 401.8 0.76 163.6 0.08 168.7 0.36 133.4 0.09 709.6 0.56 711.8 Fe(III) : Fe(II) 2 :1
Zn grown BE, AC, eV at.% 285.0 28.76 286.7 29.1 288.2 8.74 289.3 1.13 531.6 3.83 533.1 23.57 400.2 3.96 401.9 0.48 163.6 0.06 168.7 0.3 133.5 0.12
Cu grown BE, AC, eV at.% 285.0 29.21 286.6 26.85 288.2 11.42 289.6 0.84 531.6 3.01 533.1 23.48 400.2 4.17 402.0 0.55 163.6 0.09 168.7 0.28 133.6 0.1
-
-
no Zn
no Cu
identification C-(C,H) (Beamson and Griggs 1992) C-(O,N) (Clark and Thomas 1976) C=O N-C-O (Russat 1988) COOH (Gelius et al 1970) Organic (Peeling et al 1978) organic, H2O (Boulanger et al 1988) N (Hendrickson et al 1969) proton. N (Lindberg and Hedman 1975) Organic S (Beyer et al 1981) Sulphate (Lindberg et al 1970) Phosphate (Fluck and Weber 1974) Fe(II) (Mills and Sullivan 1983) Fe(III) (McIntyre and Zetaruk 1977)
Extracellular polymeric substances (EPS) of At. ferrooxidans cells surface are mainly composed of sugars, lipids and loosely associated fatty acids (Kinzler et al 2003; Harneit et al 2006) and the composition is dependent on the growth media. Based on chemical structures, sugars contain higher number of C-O groups per total amount of carbon in molecule than fatty acids. The increase of this group on the surface of zinc adapted cells may indicate an increase of sugar compounds. At the same time, the amount of C-H prevalent in lipids was decreased. Copper adapted cells lowered the carbon, C-(H, C), content and it was compensated by the increased production of C=(O)/N-C-O rich substances. The increase of oxygen rich compounds is not consistent with the decrease of oxygen content of the samples, but the decrease of nitrogenous compounds abundant in the cell wall area suggests some increased secretion of EPS and thus the decrease of signal intensity from an area of higher oxygen abundance. Increased coverage of cell surface by EPS is probably the reason for decreased rate of protonated nitrogen by changing the environment around the protein. Changing the growth conditions changes the amount and composition of surface related compounds and different metal ions induced the production of different compounds. 3.2 Microbial Adhesion to minerals 3.2.1 Adsorption studies In Fig. 13 the adsorption density of iron grown At. ferrooxidans cells on different minerals as a function of equilibrium concentration is presented. Complete surface saturation was not reached within the applied concentration range of the cells, but above 108 cells/ml concentrations there was a visible levelling-off of the curves indicating the proximity of a saturation concentration. The curves show a clear difference in adsorption preferences of the cells to minerals. The adsorption density of cells on pyrite was the highest followed by chalcopyrite and sphalerite. Comparing the cells surface coverage on minerals and assuming 25
their adherence to pyrite was 100%, then chalcopyrite was covered by 80.6% and sphalerite by 47.2% at equilibrium.
Fig. 13. Unadapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite.
Zinc adapted At. ferrooxidans cells adhered more to pyrite and chalcopyrite than sphalerite before attaining the surface saturation level (Fig. 14). The isotherm for sphalerite indicates close to a complete surface saturation at and above 108 cell/ml concentration. The selective and higher adhesion of unadapted cells to pyrite completely disappeared with the cells after adaptation to zinc ions. The zinc adapted cells adhered to chalcopyrite to the same extent as to pyrite, but their adsorption onto sphalerite decreased to 41.3% compared to pyrite and chalcopyrite. At. ferrooxidans cells grown in the presence of copper ions adhered the least to all three minerals used in this study (Fig. 15). The cells almost reached surface saturation above 107 cell/ml equilibrium concentration. The cells adhesion to chalcopyrite and pyrite was comparable at higher cell concentrations. Considering this adsorption density as surface saturation, the adhesion of the cells on sphalerite corresponds to 47.5% surface coverage.
Fig. 14. Zinc adapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite.
26
Fig. 15. Copper adapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite.
The results showed that cells grown in the presence of copper ions adhered to the minerals to a lesser extent than cells grown in their absence. Cells adhesion to pyrite and chalcopyrite was higher with zinc adapted cells but led to a loss of selectivity towards pyrite. Unadapted At. ferrooxidans cells were effective towards sphalerite but the adsorption density was still the lowest among all the minerals. Adsorption isotherms of L. ferrooxidans on pyrite and chalcopyrite with respect to equilibrium cell concentration are presented in Fig. 3 of Paper 7. Adsorption was found to be fast onto both minerals and the adhesion kinetics is comparable with At. ferooxidans. The adsorption of At. ferrooxidans cells onto pyrite was higher compared to chalcopyrite, whereas L. ferrooxidans cells adhered more to chalcopyrite than pyrite. Full surface coverage was not reached in the concentration range studied and at the highest equilibrium concentration. While considering the geometrical cell dimensions of 1.65 ȝm length and 0.35 ȝm of breadth, the cells adsorption density for full surface coverage in horizontal orientation corresponds to 1x1012 cells mņ2. This approximate calculation reveals that the adsorption of cells at the highest adsorption data point in the isotherms of either mineral equals to just 10% of the surface coverage. Similar behaviour was described by Sand et al (2009) as adsorption anomaly. Generally At. ferrooxidans cells adhere the most to pyrite (Das et al 1999 Sharma et al 1999) but the adhesion is strain sensitive (Ghauri et al 2007). Thiobacillus group bacteria are known to specifically adsorb on surface defects and imperfections (Rohwerder 2003). Since chalcopyrite surface area is nearly twice to that of pyrite, it is presumed to contain higher surface imperfections than pyrite and therefore higher adsorption of cells on chalcopyrite. As the species are different, their surface properties are also different, which influence the thermodynamic balance of adhesion and electrostatic attraction-repulsion forces. While considering biological forces also for cells adhesion on surface defects (Rohwerder 2003), chalcopyrite surface consists more surface imperfections (dislocations, steps, kinks, vacancies, etc.) as its surface area is higher and therefore the adsorption is higher. 27
3.2.2 Theoretical approaches of bacterial cells adhesion on minerals 3.2.2.1 Thermodynamic approach The calculated free energy of adhesion values for At. ferroxidans – mineral systems in aqueous medium is presented in Table 4. The results show that the adhesion is favoured for all cells onto pyrite and chalcopyrite but not on sphalerite. On thermodynamic grounds, the adhesion of cells on sphalerite is energetically not favourable. For all minerals, the ǻGtot follows the order of unadapted cells zinc adapted copper adapted, making the adhesion most apparent for the unadapted cells of At. ferrooxidans. The values of dispersive component of free energy, ǻGLW, follow a reverse order because of the increased dispersive contribution by the bacteria to the system after their cultivation in the presence of metal ions. Metal ions adaptation also induced more polarity to cells surface with predominant electron donating character and thus a decrease in the likelihood of cells adhesion (ǻGAB value increased) is expected. The free energy of adhesion values illustrate that both unadapted and adapted cells favour adhesion onto pyrite (lowest ǻGtot) followed by chalcopyrite and sphalerite. This is due to an increase in dispersive free energy between mineral and bacteria when the cells surface becomes more polar character. Table 4. Free enegy of adhesion for different bacteria – mineral systems.
Mineral – Bacteria System Pyrite - At. ferrooxidans Pyrite - Cu grown At. ferrooxidans Pyrite - Zn grown At. ferrooxidans Pyrite - L. ferrooxidans Chalcopyrite - At. ferrooxidans Chalcopyrite - Cu grown At. ferrooxidans Chalcopyrite - Zn grown At. ferrooxidans Chalcopyrite - L. ferrooxidans Sphalerite - At. ferrooxidans Sphalerite - Cu grown At. ferrooxidans Sphalerite - Zn grown At. ferrooxidans
Free energy of adhesion ǻGadh (mJ.m-2) ǻGadhLW -4.65 -5.42
ǻGadhAB -10.29 -4.66
ǻGadhtotal -14.93 -10.08
-4.80 -4.01 -2.95 -3.44 -3.05 -2.55 -3.65 -4.26 -3.77
-7.83 -9.30 -8.10 -1.70 -5.21 -7.06 6.30 14.03 10.27
-12.63 -12.25 -12.11 -5.14 -8.26 -9.61 2.65 9.78 6.50
L. ferrooxidans cells surface was more hydrophilic relative to At. ferrooxidans cells, and tends to stay in aqueous phase more. The values of free energy of adhesion were smaller and the adhesion was less preferred compared with At. ferrooxidans. The higher electron acceptance ability of pyrite can explain the higher values of free energies of adhesion compared with chalcopyrite and sphalerite, because the cells are strong electron donors and are attracted to an electron acceptor surface more. Free energy of adhesion computed by LW-AB approach for different mineral-yeast cells are presented in Table 5. The surface energy data of quartz and Y. lipolytica were used from the literature (Aguedo et al 2005; Dwari et al 2009). The Lifshitz-van der Waals free energy component was negative for all the systems studied where the systems with Y. lipolytica exhibited less negative. The acid-base free energy component was negative only for the systems with galena and positive for sphalerite and quartz systems. The total free energy of 28
adhesion was negative and adhesion was favourable for the systems S. cervisiae-galena, S. cerevisiae-sphalerite and Y. lipolytica-galena. For Y. lipolytica-sphalerite system, the total free energy was slightly positive foreseeing a problematic adhesion, but the free energies of adhesion were highly positive for the cell-quartz systems leading to an unfavourable adhesion. The likelihood of adhesion followed in the order of galena > sphalerite > quartz. This order is evidently due to the decreasing non-polar character and increasing electron donor capabilities of the minerals in the same direction. These calculations are based on the fact that the particles and cells are dispersed in water and the mineral-cell adhesion/coagulation leads to a lower energy system. The highly hydrophilic Y. lipolytica cannot motivate to leave water environment to adhere on a hydrophobic surface and the free energies of adhesion are less negative compared to S. cerevisiae. Table 5. Free energy of adhesion values for different yeast – mineral systems.
Yeast-Mineral system S. Cerevisiae - Galena S. Cerevisiae - Sphalerite S. Cerevisiae - Quartz Y. Lipolytica - Galena Y. Lipolytica - Sphalerite Y. Lipolytica - Quartz
Free energy of adhesion ǻGadh (mJ.m-2) ǻGadhLW ǻGadhAB ǻGadhtotal -7.33 -6.49 -13.81 -5.61 2.92 -2.69 -2.68 9.32 6.64 -0.76 -3.38 -4.14 -0.58 1.44 0.86 -0.28 4.76 10.98
3.2.2.2 Extended DLVO theory approach For calculation of interactive energy components by extended DLVO approach, the parameters such as Hamaker constant, surface potential and free energy of adhesion are essential and these parameters have been determined experimentally. Hamaker constants were calculated by two methods and the computed values are presented in Table 6. Table 6. Calculated values of Hamaker constants for mineral – bacteria systems.
Mineral – Bacteria System Pyrite - At.f. Fe grown Pyrite - At.f. Cu grown Pyrite - At.f. Zn grown Pyrite – L. f. Fe grown Chalcopyrite - At.f. Fe grown Chalcopyrite - At.f. Cu grown Chalcopyrite - At.f. Zn grown Chalcopyrite – L.f. Fe grown Sphalerite - At.f. Fe grown Sphalerite - At.f. Cu grown Sphalerite - At.f. Zn grown
A (method 1)
A11
4.32 5.04 4.46 3.73 2.74 3.20 2.83 2.37 3.39 3.96 3.51
49.40 52.80 50.10 46.7 49.40 52.80 50.10 46.7 49.40 52.80 50.10
A22 A33 A (method 2) [10-21 J] 63.20 1.77 63.20 2.21 63.20 1.85 63.2 1.40 37.00 50.30 0.96 (Evans and 50.30 1.20 Wennerström 50.30 1.00 1994) 50.30 0.76 55.40 1.29 55.40 1.61 55.40 1.36
Method 1 represents the calculation using LW part of free energy of adhesion in aqueous system and the effective Hamaker constant was obtained using Equation 22. Method 2 is based on calculating the effective Hamaker constant from the individual Hamaker constants 29
for homogeneous phases. Hamaker constants for bacteria A11 and mineral A22 were estimated using the dispersive part of surface energies obtained from contact angle measurements and the water Hamaker constant A33 was from the literature. Hamaker constants calculated according to the methods 1 and 2 are of course different but the values are comparable and following the same trend. The Hamaker constants for pyrite - bacteria systems were higher than the chalcopyrite and sphalerite systems. The sphere-sphere geometry of particles for calculating the interaction forces is more suitable for mineral-bacteria system, although the shape of bacillus cells is not exactly spherical. However considering the size of interacting bodies, the spherical model is reasonable. The energy is expressed in kT units in the interaction energy versus distance diagrams presented in Figs. 16 to 27. For outlining the influence of various parameters, a model system was chosen with the following parameters: pH 2, 0.01 M KNO3 electrolyte concentration, Hamaker constant estimated by Method 1 and 1 ȝm size for particles and cells. Influence of Hamaker constant: Two different approaches were used for the Hamaker constant evaluation and as a result two different Hamaker constants were obtained for the same system. Lifshitz - van der Waals forces are always attractive, relatively weak and becoming significant at shorter separation distances. From Equation 15, it is clear that the LW interaction force is proportional to the value of Hamaker constant. The difference in the two Hamaker constants obtained by the methods 1 and 2 is rather small. It is also clear from Figure 16 that the LW interactions, regardless of the value of Hamaker constant, are the weakest force in the total interaction energy between At. ferrooxidans and minerals. The Hamaker constant with either of the two values has no significant influence on the character of total interaction force. Hamaker constants obtained for pyrite – At. ferrooxidans system were higher and therefore the LW forces were higher compared with chalcopyrite – At. ferrooxidans. This is due to a higher electron acceptance character of pyrite (Table 1). The LW interactions have influence on total interaction energy, but the difference between the values determined by different methods was only few kT and was negligible compared with the strong attractive acid–base interactions. Regardless of the method used for Hamaker constant estimation, the adhesion of bacteria onto minerals was favorable. Due to a small difference between the calculated Hamaker constants by the two methods, only the value of A1 is used in further calculations. B 500 LW interaction A2 LW interactions A1 Eelectrostatic Interaction Acid-base Interaction
400 300 Energy (kT)
200 100 0 -100 0
20
40
60
80
100
-200 -300 -400 -500 Distance (Å)
Fig. 16. Pyrite – At. ferrooxidans interaction energy diagram at pH 2, 0.01M ionic strength and 1 ȝm particle size. LW forces calculated using Hamaker constants derived by Method 1 and Method 2.
30
Influence of the ionic strength: Ionic strength is inversely proportional to electrical double layer thickness surrounding the particle and increasing the ionic strength, the double layer compresses and zeta potential decreases. Equation 16 illustrates that the electrostatic force is a function of double layer thickness and zeta potential, and thus by increasing the ionic strength, the two parameters decrease leading to a drop in electrostatic force. Zeta-potential was measured only at one electrolyte concentration; however potential energy diagrams were computed at varied ionic strength to identify its influence on bacterial adhesion behavior. Since the zeta-potential decreases with increasing ionic strength, the electrostatic force is greatly reduced. Strong long-range attractive force can be achieved at low electrolyte concentration and the modeled 0.001 M ionic strength led to an attractive force of 50 kT at a distance of 10 nm (100 Å) between pyrite and At. ferrooxidans. Depending on the magnitude of surface charge, electrostatic force can be strong or weak and therefore, a change in ionic strength displayed a great impact on the interaction forces (Figure 17). In the calculations, zeta-potentials measured at 0.01M KNO3 were only used and a change in zeta potential value due to a varied ionic strength was not incorporated and the calculations are rather simplified. In reality, zeta potential decreases with compressed double layer and the difference between different electrolyte concentrations are even more observable. Low electrolyte concentrations exaggerate the electrostatic force and they are preferable when the electrostatic force is weak, and higher concentrations are necessary to achieve fast adhesion or repulsion. C
0 0
20
40
60
80
100
Energy (kT)
-100
-200
-300 Total Interaction 0.001 Total Interaction 0.01 Total Interaction 0.1
-400
-500 Distance (Å)
Fig. 17. Pyrite – At. ferrooxidans interaction energy diagram at pH 2, LW force calculated using A1 and 1 ȝm particle size. Total interactions are presented at three different ionic strength 0.1M, 0.01M and 0.001M KNO3.
Influence of particle size: The size of bacterial cells is more or less the same (within one strain) and the mineral particles vary from sub-micron to 5 ȝm size. The particle size was therefore varied in the simulation of interaction energy. Particle size was varied in the Lifshitz - van der Waals, electrostatic and acid-base interaction forces, while the bacterial size was kept constant due to its constant cell size. In the LW force calculation (Eq. 15), the particle size is hidden in parameters x and y and a change from 1 to 4 ȝm particle size had only a micro effect on the LW force. The curves at different particle sizes were overlapping as presented in Fig. 18, and any difference in force was not observable. The effect of particle size within the tested and simulated conditions can be considered as negligible. On the other hand, the effect of particle size on electrostatic force was more distinguishable. From Equation 16, it is clear that the force is proportional to a1a2/(a1+a2) and by keeping one 31
parameter constant and increasing the another, the term is approaching 0 for very small particle sizes and 1 for high particle sizes. Using much higher particle size relative to the cells will increase the force only to some limit and using much smaller particles relative to cell size will minimize the electrostatic force. The simulated particle sizes were 1, 2 and 4 ȝm (Fig. 18) and with increasing particle size, the electrostatic force also increased and became significant at smaller separation distances. In the present model system, electrostatic force was attractive since the particle and bacteria were charged oppositely but the attractive force was not the strongest and the acid-base interactions were far the strongest. Therefore the alteration of particle size has not changed the electrostatic force and the total interaction energy markedly (Fig. 7c, Paper 4). However, the change is observable between bigger particles where the total attractive force is higher and this effect could be more predominant if the particles are charged more. A
0 20
40
Energy (kT)
0
60
80
100
LW Interaction 4 ȝm LW Interaction 2 ȝm LW Interaction 1 ȝm Electrostatic Interaction 4 ȝm Electrostatic Interaction 2 ȝm Electrostatic Interaction 1 ȝm -100 Distance (Å)
Fig. 18. Pyrite – At. ferrooxidans interaction energy diagram at pH 2, 0.01M ionic strength and LW force calculated using A1. Electrostatic and LW forces calculated using different particle sizes 1, 2 and 4 ȝm
Influence of pH: The pH of the medium in which the particles are interacting is not a variable in the interaction equations but greatly influence the surface potential and therefore the electrostatic force of interaction. At. ferrooxidans cells were charged negatively in the entire pH range studied whereas the minerals are positively charged below pH 7 and negatively charged above this pH suggesting an attractive force at lower acidic pH values and repulsive force at alkaline pH values. Pyrite zeta potentials were higher in magnitude and accordingly the electrostatic force obtained for pyrite was higher compared with chalcopyrite. At pH 2 and 7 the potentials were similar, this resulted in an overlap of electrostatic interaction curves. At pH 5, pyrites was strongly positively charged, which resulted in a strong electrostatic attraction towards cells, and at pH 10, the bacterial cells and minerals were repelling each other due to negative surface charges (Fig. 19). Based on lower surface charge characteristics of chalcopyrite compared to pyrite (Fig. 6), the electrostatic force for chalcopyrite was also lower (Fig. 20). The total interaction force for both the systems was attractive at pH 2, 5, and 7 for all the separation distances while at pH 10 only after approaching a distance of 2.7 nm (27 Å) overcoming the energetic barrier. The adsorption tests were carried out at pH 2 and the total force was attractive for both the mineral–bacteria systems. Because of higher acid–base and electrostatic interactions between the cells and pyrite, the adhesion of cells onto pyrite was more. 32
C
200 100
Energy (kT)
0 0
20
40
60
80
100
-100 -200 Total Interaction pH 10 Total Interaction pH 7 Total Interaction pH 2 Total Interaction pH 5
-300 -400 -500 Distance (Å)
Fig. 19. Pyrite – At. ferrooxidans interaction energy diagram at 0.01M ionic strength, LW force calculated using A1 and 1 ȝm particle size. Total interactions at pH 2, 5, 7 and 10. C
200 100
Energy (kT)
0 0
20
40
60
80
100
-100 -200 Total Interaction pH 10 Total Interaction pH 7 Total Interaction pH 2 Total Interaction pH 5
-300 -400 -500 Distance (Å)
Fig. 20. Chalcopyrite – At. ferrooxidans interaction energy diagram at 0.01M ionic strength, LW force calculated using A1 and 1 ȝm particle size. Total interactions at pH 2, 5, 7 and 10.
At. ferrooxidans after Cu- and Zn- ions adaptation: Adhesion of unadapted cells to mineral surfaces was most favoured in the presence of pyrite because of a strong attractive acid-base interaction and a strong electrostatic attraction between oppositely charged surfaces. The total interaction energy was attractive towards chalcopyrite but weaker compared with pyrite (Fig. 21). Lower polarity of the chalcopyrite surface compared to that of pyrite decreased the attractive acid-base forces and also minimised the electrostatic force due to reduced surface charge. Interaction energy between sphalerite and cells was positive below 4 nm (40 Å) separation distances which impaired adhesion or at least decreased it. The repulsive force stemmed from the repulsive acid–base interaction force and negatively charged particle and cell surface. The attractive LW forces were too weak to affect the repulsive forces.
33
300 Fe grown A. ferrooxidans
200
G (kT)
100
0 0
20
40
60
80
100
-100 pyrite chalcopyrite
-200
sphalerite
-300
Separation (Å)
Fig. 21. Total interaction forces between ferrous grown At. ferrooxidans and sulphide minerals.
Zinc grown cells were positively charged and also the minerals at pH 2 inducing repulsive electrostatic forces. The total interaction energy was, however, attractive with respect to pyrite and chalcopyrite because of LW and acid-base forces (Fig. 22). The cell adhesion on these minerals (Fig. 14) is in agreement with the interaction energy curves. Sphalerite is negatively charged at pH 2 and the electrostatic force favours cells adhesion, but below 1.5 nm (15 Å) distance the total interaction became repulsive due to a strong acid–base repulsive force. 300 Zn grown A. ferrooxidans
200
G (kT)
100
0 0
20
40
60
80
100
-100 pyrite chalcopyrite
-200
sphalerite
-300
Separation (Å)
Fig.22. Total interaction forces between zinc grown At. ferrooxidans and sulphide minerals
The positive surface charge of copper grown cells was very weak and the cells electrostatic repulsive force had no significant influence on the total interaction energy for pyrite and chalcopyrite (Fig. 23). Acid–base and LW interactions became decisive for cells adhesion on these minerals and both were attractive. The total interaction force was stronger for pyrite because the 'GadhAB value was greater. Cells adhesion on sphalerite in the present experiments was expected to be very weak because repulsive forces arose at greater distances from the mineral. The attractive forces resulting from the negative charge on sphalerite were 34
weaker than those of Zn–grown cells. Thus, the extent of adhesion of Zn–grown cells to sphalerite was expected to be slight. 300 Cu grown A. ferrooxidans
200
G (kT)
100
0 0
20
40
60
80
100
-100 pyrite chalcopyrite
-200
sphalerite
-300
Separation (Å)
Fig. 23 Total interaction forces between copper grown At. ferrooxidans and sulphide minerals.
The adhesion forces for the copper grown cells are lower according to the adhesion results (Figs. 14 and 15) and the zinc grown cells should maintain the same level as unadapted cells. From the potential energy diagrams it is clear that the strongest total interactions between the minerals and cells are for the unadapted and zinc grown cells and weaker interactions for copper grown bacteria. However the increased adhesion of copper grown cells on chalcopyrite was not reflected in the DLVO calculations. The electrostatic repulsive forces were too weak to have any effect compared to more charged pyrite. If the soft particle model is used instead of Smoluchowski, the result is a decrease of surface potential. However at low surface potential as in the case of copper grown cells this decrease is negligible (Chandraprabha 2009). Depending on the model (constant surface potentials or constant fixed charge densities) the electrostatic interaction energies may increase (Poortinga et al 2002). According to the adhesion experiments carried out with copper grown cells, the electrostatic forces had a greater influence on the total energetic balance than predicted by the DLVO theory. At. ferrooxidans cell surface is visualised as smooth (Ehrlich 1990 Mangold et al 2008) and therefore there is no decrease in electrostatic interaction as expected (Elimelech and O Melia, 1990). In reality, the bacterial surface is very heterogeneous in terms of chemical composition and topography which provoke different interactive forces at different areas of the surface (Fang et al 2000). The present results however showed that the extended DLVO theory was capable to reflect most of the changes in the adhesion abilities of unadapted and metal ions adapted cells onto minerals. L. ferrooxidans: At pH 2, where the adsorption tests were carried out, the surface potential of chalcopyrite was lower, causing a lower repulsive electrostatic force relative to pyrite particles, which contributed to the total energetic balance as a lower energetic barrier between the cells and mineral particles. Lower repulsive forces in combination with the possibility of higher occurrence of iron rich dislocation sites at the surface of chalcopyrite may result in higher adsorption densities onto chalcopyrite, compared with pyrite. However this result was not observed in the previous studies with similar species, but the extended DLVO theory 35
confirmed and explained the experimental results (Figs. 24 and 25). For achieving selective adhesion onto pyrite, a pH modification is necessary, where the surface potential of pyrite is higher relative to chalcopyrite and the cells are charged negatively. Above pH 4.5, pyrite exhibited higher zeta potential than chalcopyrite and the cells are beyond their IEP, resulting in stronger attractive electrostatic force toward pyrite and higher adsorption densities. Insufficient forces could be increased by manipulating the electrolyte concentration or increasing the particle size. By changing the growing conditions or substrate the surface properties of the bacteria are expected to change, thus these results are valid only for the Fe2+ grown L. ferrooxidans, but the alteration of the growth conditions represents a possible way to regulate the adhesion forces. 300 Pyrite - L. ferrooxidans
200
G (kT)
100
0 0
20
40
60
80
100
-100 Lifshitz - van der Waals Electrostatic
-200
Acid-Base Total
-300
Separation (Å)
Fig. 24. Pyrite – L. ferrooxidans interaction energy diagrams at 0.01 M KNO3, pH 2, 1 ȝm particle size and LW interactions calculated using A1. 300 Chalcopyrite - L. ferrooxidans
200
G (kT)
100
0 0
20
40
60
80
100
-100 Lifshitz - van der Waals Electrostatic
-200
Acid-Base Total
-300
Separation (Å)
Fig. 25. Chalcopyrite – L. ferrooxidans, interaction energy diagrams at 0.01 M KNO3, pH 2, 1 ȝm particle size and LW interactions calculated using A1.
36
Yeast cells: The interaction energy versus separation distance curves between S. cerevisiae and minerals are shown in Fig. 26. For galena–cell system, the different interacting forces were negative resulting in a total adhesive force. The electrostatic force was rather weak while the LW forces were relatively significant. The acid–base forces were the strongest and attractive controlling the total interaction force for this system. Below the separation distance of 25 Å the forces were attractive resulting in adhesion. In the case of sphalerite, the electrostatic and LW forces were negative while the acid–base forces were positive creating an energetic barrier. Electrostatic forces were very weak while the LW forces were weaker compared to galena and adhesion was possible after overcoming the energetic barrier. The LW forces were attractive but the electrostatic and acid–base forces were repulsive for quartz–cell system. The strong acid–base repulsive interactions were responsible for the inability of cells adhesion on quartz. 700
S. cerevisiae
500
G (kT)
300
100
-100 0
20
40
60
80
100
quartz galena
-300
sphalerite
-500
Separation (Å)
Fig. 26. S. cerevisiae - galena, sphalerite, quartz total interaction energy diagrams.
Fig. 27 shows the interaction energy curves between Y. lipolytica cells and minerals versus separation distance. The LW part of free energy of adhesion for Y. lipolytica–mineral systems was weaker compared to S. cerevisiae cells adhesion and the energy curves indeed display very weak LW forces for all Y. lipolytica systems. In the case of galena, the electrostatic forces were stronger and attractive due to a higher and opposing zeta-potential of Y. lipolytica cells and galena. The acid–base interactions were also attractive leading to attractive total interaction energy and the cells adhesion forces towards galena were significant. Although the electrostatic forces were attractive and stronger for Y. lipolytica–sphalerite system, the repulsive acid–base interactions created an energy barrier of 300 kT which needs to overcome for adhesion to occur. Very strong acid–base repulsive forces between Y. lipolytica cells and quartz render their coagulation impracticable (Fig. 27).
37
500 Y. Lipolytica
400 300 200 G (kT)
100 0
-100
0
20
40
60
80
100
-200 quartz
-300
galena sphalerite
-400 -500
Separation (Å)
Fig. 27. Y. lipolytica - galena, sphalerite and quartz interaction energy diagrams.
3.3 Bioflotation and Bioflocculation 3.3.1 Flocculation studies Transmission diagrams (Fig. 28) obtained from Turbiscan showed the percentage of light transmitted trough the sample, well dispersed suspension is unable to transmit the light beam and therefore the transmission is very low and approaching zero. When sedimentation occurs, the lower amount of dispersed particles allows the beam to pass and transmittance is increased. The suspensions were scanned for 60 minutes, but only the first 10 minutes and few millimetres from the suspension meniscus are presented for easier interpretation. Each minute one scan was performed and is presented as a line in the diagram; the final diagrams showed the progress of suspension “clearing” with time. The transmission data was after At. ferrooxidans addition at pH 5 for pyrite and at pH 8.1 for chalcopyrite. Sedimentation of pyrite as well as chalcopyrite was enhanced after At. ferrooxidans addition. The transmission data after L. ferrooxidans treatment are shown for pyrite at pH 6.5 and chalcopyrite at pH 8. Sedimentation of pyrite was enhanced; however it showed some natural sedimentation because of near IEP of pure pyrite. Poor sedimentation of chalcopyrite at pH 8 was multiplied after bacterial addition. Higher number of adsorbed cells on chalcopyrite surface is probably responsible for higher effect of flocculation.
38
Figure 28. Transmission Graph of pyrite and chalcopyrite suspensions after 10 minutes settling time.
The effect of cell addition is obvious as the suspensions are “clearer”, however the transmission values cannot represent flocculation behaviour accurately. Much better expression of settling and flocculation phenomena is to use the increase in size of particle aggregates. The increase of particle size as a function of time is calculated from the obtained transmission and backscattering data from Turbiscan and from the photon mean free path. At. ferrooxidans adhesion on mineral surface has not lowered the potential; however it caused the shift in iso-electric point. Settling tests done with At. ferrooxidans confirmed this phenomenon. After bacterial treatment the minerals were not settling better, but only at the pH values where the IEP shifted and where maximum settling is observed (Fig. 1 in Paper 6). While pyrite moved the pH of highest particle size (fastest sedimentation) to acidic region by 1 pH unit, chalcopyrite moved to basic region for about 2 pH units from their original pH, causing a difference of 3 - 4 pH units between them. While pyrite at pH 4.5 was settling perfectly after bacterial addition, chalcopyrite sedimentation was poor. Zeta-potential measurements after L. ferrooxidans addition showed decreased surface potential and a decrease of IEP for both tested minerals, so the sedimentation was expected to be increased. The settling behaviour expressed in transmission of mineral suspensions (Fig. 8 Paper 7), increased after bacterial treatment. The increase for pyrite was less significant and 39
limited by pH, because at higher pH the zeta-potential of cells interacted pyrite particles changed sharply to negative potential. Chalcopyrite increased the settling more at all measured pH values, because the potential of chalcopyrite turned from negative to almost uncharged, exhibiting negative charge only above pH 10. 3.3.2 Flotation studies Flotation tests conditions were chosen from the previous tests carried out in the absence of bacterial cells. The effect of At. ferrooxidans and L. ferrooxidans cells on pyrite and sphalerite flotation as a function of pH is presented in Fig. 29. Pyrite flotation was more depressed with At. ferrooxidans and less with L. ferrooxidans at the same concentration. Chalcopyrite was depressed after L. ferrooxidans treatment while the A. ferrooxidans cells were less effective. pyrite
100
Flotation recovery (%)
chalcopyrite
80 pyrite + L. f. 10^7cell/ml
60
pyrite + At. f. 10^7 cell/ml
40
chalcopyrite + At. f. 10^7 cell/ml
20
chalcopyrite + L. f. 10^7 cell/ml
0 0
2
4
6
8
10
12
pH
Fig. 29. Flotation of pyrite and chalcopyrite using At. ferrooxidans and L. ferrooxidans 107 cell/ml preconditioning relative to pH.
The effect of different concentrations of cells on pyrite and chalcopyrite flotation is presented in Figs. 30 and 31 respectively. With the increasing initial cell concentration the cells surface coverage on minerals increases imparting hydrophilicity and therefore flotation recoveries decreased. Cells adhered on the mineral surface inhibit the adhesion of collector. The depression effect of L. ferrooxidans on pyrite flotation was seen above pH 6 at 107 cell/ml concentration where the recoveries decreased with increasing pH reaching 20% recovery at pH 10. At. ferrooxidans at the same concentration caused a moderate pyrite depression below pH 5, and above which, the flotation behaviour was comparable with L. ferrooxidans treatment. At higher At. ferrooxidans concentration of 5x107 cell/ml, the depression takes place even at lower pH values. Complete depression was achieved at 108 cell/ml concentration where the flotation recoveries were below 20%. Chalcopyrite flotation was seen to depress at higher and lower concentrations of At. ferrooxidans and L. ferrooxidans respectively. Complete depression was noticed in both cases. At. ferrooxidans at 107 cell/ml concentration caused a gradual decrease of flotation recoveries with increasing pH.
40
Flotation recovery (%)
100 80 60 40 20 0 0
2
4
6
8
10
12
pH pyrite
pyrite + L. f. 10^7cell/ml
pyrite + At. f. 5x10^7 cell/ml
pyrite + At. f. 10^8 cell/ml
pyrite + At. f. 10^7 cell/ml
Fig. 30. Flotation of pyrite using different concentrations of At. ferrooxidans and L. ferrooxidans relative to pH.
Flotation recovery (%)
100 80 60 40 20 0 0
2
4
6
8
10
12
pH chalcopyrite
chalcopyrite + At. f. 10^7 cell/ml
chalcopyrite + At. f. 10^8 cell/ml
chalcopyrite + L. f. 10^7 cell/ml
chalcopyrite + At. f. 5x10^7 cell/m
Fig. 31. Flotation of chalcopyrite using different concentrations of At. ferrooxidans and L. ferrooxidans relative to pH.
At. ferrooxidans cells may not be suitable for selective flotation because both minerals were depressed at the same time. In two cases, the recoveries of pyrite and chalcopyrite were a little bit different: pH 4 and 107 cell/ml, and pH 2 and 5x107 cell/ml. L. ferrooxidans cells were capable to completely depress chalcopyrite flotation and no effect on pyrite flotation at lower cell concentrations. The highest selectivity was achieved at pH 4. The depression of chalcopyrite after L. ferrooxidans treatment is in agreement with higher adherence of Leptospirillum cells on chalcopyrite. At. ferrooxidans adhesion on pyrite is higher than chalcopyrite, thus the flotation of pyrite is expected to be depressed more. The recoveries of pyrite were decreased and chalcopyrite increased relative to Leptospirillum, but do not correspond to the amount of At. ferrooxidans cells adhered and other factors might have influenced the flotation results besides the depression effect of cells. Selective separation of galena and sphalerite from quartz is presented in Paper 8.
41
SUMMARY Acidithiobacillus ferrooxidans and Leptospirillum ferrooxidans are iron oxidizing acidophilic species and their metabolic and genetic disparities tend to be crucial for interaction with sulphide minerals. The cells are capable to respond to the changes in their growth environment by surface alterations and increased biopolymer production. Iron grown At. ferrooxidans cells were negatively charged with the iso-electric point (IEP) located below pH 1. After adaptation to copper and zinc ions, the cells surface negativity was reduced. L. ferrooxidans cells displayed positive zeta-potential below pH 3.3 and above which, the magnitude of negative potential is relatively high compared to At. ferrooxidans. Different zeta-potential characteristics between the two bacteria suggest different composition of the cell surface layers. This difference in surface charge among the two bacteria reacted to minerals differently. L. ferrooxidans treatment of minerals caused a decrease in zetapotentials of pyrite and chalcopyrite, ascertaining cells adhesion to minerals. At. ferrooxidans treatment caused the shift in iso-electric points without affecting the minerals zeta-potential. The shift in IEP was towards basic and acidic pHs for pyrite and chalcopyrite respectively, creating a 3 to 4 pH units difference between the IEPs of the minerals. Metal ions adapted At. ferrooxidans cells became more positive as a result of altered chemical composition of the surface layers and not due to the adsorption of metal cations as revealed by X-ray photoelectron spectroscopy. Contact angle measurements of cells and surface energy estimations showed that the bacterial species have the same surface energies, with a prevalently electron donating character surface. Both strains were hydrophilic and the hydrophilicity of L. ferrooxidans was higher than At. ferrooxidans. Metal ion adaptation increased the surface energy, hydrophilicity and polar character of the surface. Adsorption studies showed different adsorption preferences of the bacteria to minerals. Preferential adsorption to pyrite surface was found in the case of iron grown At. ferrooxidans. Substantial adhesion of cells to chalcopyrite was however noticed and minor adhesion to sphalerite. Copper and zinc ions adaptation enhanced the adherence of At. ferrooxidans to chalcopyrite causing a loss of selectivity towards pyrite. Iron grown L. ferrooxidans adsorbed to chalcopyrite in higher adsorption density than to pyrite. Adhesion of At. ferrooxidans and L. ferrooxidans cells on both pyrite and chalcopyrite is preferred according to the results of theoretical thermodynamic and extended DLVO calculations. The total free energy of adhesion is negative for both bacteria-mineral systems, predicting that the adhesion is energetically favourable. The ability of the minerals to accept electrons and the cells strong electron donors is responsible for attractive acid-base interactions and contributing to the total attractive force in the extended DLVO calculations. Only the same surface charge polarity of mineral and bacteria caused repulsive forces, which are responsible for energetic barrier over the separation distance of 30 Å in case of L. ferrooxidans. The energy barrier could be lowered by increasing ionic strength, or by altering the pH where the electrostatic forces are attractive. At pH 2, where the adsorption tests were performed, the energetic barrier between chalcopyrite and L. ferrooxidans was lower relative to pyrite and L. ferrooxidans, because of lower magnitude of surface charge. At the same pH the total attractive force between pyrite 43
and At. ferrooxidans was higher compared with chalcopyrite and At. ferrooxidans, due to a higher electrostatic attractive force. These forces contributed to different adsorption densities of cells onto minerals in adsorption studies. The extended DLVO theory calculations were in good agreement with the adhesion of yeast cells to sulphide minerals and not to quartz. Theoretical calculations also clearly depicted the lowest adhesion of differently grown At. ferrooxidans cells to sphalerite. At. ferrooxidans cells treatment of minerals shifted the IEP of pyrite and chalcopyrite. The shifts of IEP and coagulation peaks are in opposite direction, and therefore selective flocculation of either mineral from mixed composition is possible. L. ferrooxidans increased the settling characteristics of both minerals, but the effect on chalcopyrite was higher including the adsorption behaviour which suggest that the cells can function as flocculants depending on the number of cells adhered on the surface. Flotation of minerals was influenced in the presence of bacterial cells. Chalcopyrite flotation was depressed in the presence of L. ferrooxidans, while the flotation of pyrite remained relatively high. The depression is dependent on the cell concentration and the higher depression of chalcopyrite was a result of higher adsorption densities of L. ferrooxidans on chalcopyrite surface. At. ferrooxidans was found unsuitable for selective flotation in pyrite–chalcopyrite system where both minerals were depressed. Yeast cells selectively coagulated fine sulfide minerals and the bio-coagulates are floated from quartz.
44
Future work
x
Study the mechanism of bacterial adhesion and attachment to mineral surfaces. Identify the influencing factors and surface sites suitable for bacterial attachment. It is known, that cells cover the mineral surface only to some extent and preferentially at surface structural defects as screw and step dislocations, and some species need precolonisation to be able to adhere on minerals. A systematic study on the effect of surface morphology on bacterial adhesion.
x
Interaction force measurements between mineral particles and bacteria in aqueous system by AFM. Forces acting between a mineral particle fixed to a cantilever tip and bacterial cells immobilized on a Millipore filter paper will give a clearer insight into the physico-chemical forces for bacterial adhesion. The method allows changes in the environment such as pH and ionic strength.
x
Study the surface EPS layers of mineral adapted At. ferrooxidans and L. ferrooxidans cells and their adhesion response to minerals. The amount and composition of cell’s EPS change when they are grown in the presence of minerals and thereby affecting the adhesion and surface characteristics.
x
Characterisation and isolation of the EPS of unadapted and mineral adapted bacteria. Use of mineral adapted cells and surface polymers isolated from mineral adapted cells as a flotation modifiers in sulphide flotation and flocculation agent. Identify the compound responsible for mineral depression or activation in flotation.
x
Study the mineral–bacteria–collector interfaces and bacteria–collector interactions and mechanisms causing selective depression or activation in flotation. Use of different sulphide minerals (galena, arsenopyrite) and mineral mixtures.
x
Selective flotation and flocculation studies using sulphide ores and bulk flotation concentrates to test the effectiveness of bacterial conditioning in real systems.
45
REFERENCES
Absolom, D.R., Lamberti, F.V., Policova, Z., Zingg, W., van Oss, C.J., Neumann, A.W., 1983. Surface thermodynamics of bacterial adhesion. Appl. Environ. Microbiol. 46, p. 90-97. Acuña, J., Rojas, J., Amaro, A.M., Toledo, H., Jerez, C.A., 1992. Chemotaxis of Leptospirillum ferrooxidans and other acidophilic chemolithotrophs: comparison with the Escherichia coli chemosensory system. FEMS Microbiol. Lett. 96, p.37–42. Aguedo, M., Waché, Y., Belin, J.M., Teixeira, J.A., 2005. Surface properties of Yarrowia lipolytica and their relevance to J.decalactone formation from methyl ricinoleate. Biotechnol. Let. 27, p. 417-422. Attia, Y.A.,Elzeky, M., Ismail, M., 1993. Enhanced separation of pyrite from oxidized coal by froth flotation using biosurface modification. Int. J. Miner. Process. 37, p. 61-71. Balashova, V. V., Vedenina, I. Ya., Markosyan, G. E. & Zavarzin, G. A., 1974. The autotrophic growth of Leptospirillum ferrooxidans. Microbiology (English translation of Mikrobiologiya) 43, p. 491-494. Beamson, G., Briggs, D., 1992. High Resolution XPS of Organic Polymers: the Scienta ESCA300 Database. New York: Wiley. Beyer, L., Kirmse, R., Stach, J., Szargan, R., Hoyer, E.Z., 1981. Metallkomplexe des Benzoyldithioessigsäuremethylesters und des N-Benzoylamino-dithiokohlensäureethylesters: Darstellung und Charakterisierung, ESCA- und EPR-Untersuchungen. Anorg Allg Chem. 476, p.7-15. Beijerinck, M. W., 1904. Ueber die Bakterien, welche sich im Dunkeln mit Kohlensäure als Kohlenstoffquelle ernähren konnen. Centralbl Bakteriol Parasitenkd Infektionskr Hyg Abt II 11, p. 593-599. Bergey, D.H., Holt, J.G., Staley, J.T., 1989. Bergey’s Manual of systematic bacteriology. Vol. 3. Baltimore : Williams & Wilkins. p. 1842 – 1857. Bos, R., van der Mei, H.C., Busscher, H.J.,1999. Physico-chemistry of initial microbial adhesive interactions - its mechanisms and methods for study. FEMS Microbiol. Reviews. 23 (2), p. 179–230. Botero, A.E.C., Torem, M.L., Mesquita, L.M.S., 2007. Fundamental studies of Rhodococcus opacus as a biocollector of calcite and magnesite. Miner. Eng. 20, p.1026-1032. Boulanger, P., Riga, J., Delhalle, J., Verbist, J.J., 1988 .An XPS study of hexagonal polyoxymethylene with various bulk morphologies: surface modification under X-ray exposure. Polymer. 29, p.797-801. Busscher, H.J., Weerkamp, A.H., van der Mei, H.C., van Pelt, A W., de Jong, H.P., Arends, J., 1984. Measurement of the surface free energy of bacterial cell surfaces and its relevance for adhesion. Appl. Environ. Microbiol. 48, p. 980-983. Camesano, T.A., Logan, B.E., 2000. Probing Bacterial Electrosteric Interactions Using atomic Force Microscopy. Environ. Sci. Technol. 34, p. 3354-3362.
47
Chandraprabha, M.N., Modak, J.M., Natarajan, K.A., 2009. Soft-particle model analysis of effect of LPS on electrophoretic softness of Acidithiobacillus ferrooxidans grown in the presence of different metal ions. Colloids Surf. B Biointerfaces. 69, p. 1-7. Chandraprabha, M.N., Natarajan, K.A., Modak, J.M, 2004a. Selective separation of pyrite and chalcopyrite by biomodulation. Colloids Surf. B Biointerfaces. 37, p. 93-100. Chandraprabha, M.N., Natarajan, K.A., Somasundaran, P., 2004b. Selective separation of ersenopyrite from pyrite by biomodulation in the presence of Acidithiobacillus ferrooxidans. J. Coll. Inter. Sci. 276, p. 323-332. Chandraprabha, M.N., Natarajan, K.A., Somasundaran, P., 2005. Selective separation of pyrite from chalcopyrite and arsenopyrite by biomodulation using Acidithiobacillus ferrooxidans. Int. J. Miner. Process. 75, p. 113-122. Chandraprabha, M.N., Natarajan, K.A., 2006. Surface chemical and flotation behavior of chalcopyrite and pyrite in the presence of Acidithiobacillus ferrooxidans. Hydrometalurgy. 83, p. 146-152. Clark, D.T., Thomas, H.R., 1976. Applications of ESCA to polymer chemistry. XI. Core and valence energy levels of a series of polymethacrylates. J. Polym. Sci. Polym. Chem. Ed. 14, p. 1701-1713. Colthup, N.B., daly, L.H., Wiberley, S.E., 1990. Introduction to infrared and Raman spectroscopy. Third edition. Academic Press. Das, A., Rao, K.H., Sharma, P., Natarajan, K.A, Forssberg, K.S.E., 1999. Surface chemical and adsorption studies using Thiobacillus ferrooxidans with reference to bacterial adhesion to sulfide minerals. Biohydrometalurgy nad the environment toward the mining of the 21st century, part A, p. 697-707. Deo, N., Natarajan, K.A., 1997. Interaction of Bacillus polymyxa with some oxide minerals with reference to mineral beneficiation and environmental control. Miner. Eng. 10, p. 13391354. Deo, N., Natarajan, K.A., 1998. Studies on interaction of Paenibacillus polymyxa with iron ore minerals in relation to beneficiation. Int. J. Miner. Process. 55, p. 41-60. Deryagin, B.V., Landau, L., Acta Physicochem. URSS 1941, 333, 55. Devasia, P., Natarajan, K.A., Sathyanarayana, D.N., Ramananda Rao, G., 1993. Surface chemistry of Thiobacillus ferrooxidans relevant to adhesion on mineral surfaces. Appl. Environ. Microbiol. 59, p. 4051-4055. Dubel, J., Smith, R.W., Mishra, M., Chen, S., 1992. Microorganisms as chemical reagents: The hematite system. Miner. Eng. 5, p. 547-556. Dwari, R.K., Hanumantha Rao, K., Somasundaran, P., 2009. Characterisation of particle tribo-charging and electron transfer with reference to electrostatic dry coal cleaning. Int. J. Miner. Process. 91, p. 100-110. Ehrlich, H.L., 1990. Geomicrobiology, 2nd edn. Marcel Dekker, New York
48
Elimelech, M., O’Melia, C.R., 1999. Effect of Particle Size on Collision Efficiency in the Deposition of Brownian Particles with Electrostatic Energy Barriers. Langmuir. 6, p. 11531163. Evans, D.F., Wennerström, H., 1994. The colloidal domain where physics, chemistry, biology and technology meet. Wiley-VCH. Fang, H.H.P., Chan, K-Y., Xu, L-C.J., 2000. Quantification of bacterial adhesion using atomic force microscopy. Microbiol Meth. 40, p. 89-97. Fluck, E., Weber, D. Z., 1974. Naturforsch. 29, p. 603. Fowkes, F.M., 1962. Determination of interfacial tensions, Contact angles and dispersion forces in surfaces by assuming additivity of intermolecular interactions in surfaces. J Phys. Chem. 66, p. 382. Fowkes, F.M., 1964. Attractive forces at interfaces. Ind. Eng. Chem. 56, p. 40-52. Gaboriaud, F., Gee, M.L., Strugnell, R., Duval, J.F.L., 2008. Coupled electrostatic, hydrodynamic and mechanical properties of bacterial interfaces in aqueous media. Langmuir. 24, p. 10988-10995. Gawel, J., Maliszewska, I., Sadowski, Z., 1997. The effect of biopretreatment on the flotation recovery of magnesite tailings. Miner. Eng. 10, p. 813-824. Gehrke, T., Hallmann, R., Kinzler, K., Sand, W., 1998. Importance of extracellular polymeric substances from Thiobacillus ferooxidans for bioleaching. Appl. Environ. Microbiol. 64, p. 2743-2747. Gelius, U., Heden, P.F., Hedman, J., Lindberg, B.J., Manne, R., Nordberg, R., Nordling, C., Siegbahn, K., 1970. Molecular Spectroscopy by Means of ESCA III. Carbon compounds. Phys. Scripta. 2, p. 70-80. Ghauri, M. A., Okibe N., Johnson, D. B., 2007. Attachment of acidophilic bacteria to solid surfaces: The significance of species and strain variations. Hydrometalurgy. 85, p. 72-80. Girifalco, L.A., Good, R.J., 1957. A theory for the estimation of surface and interfacial energies. I. Derivation and application to interfacial tension. J. Phys. Chem. 61, p. 904-909. Harneit, K., Goksel, A., Kock, D., Klock, J-H., Gehrke, T., Sand, W., 2006. Adhesion to metal surfaces by cells of Acidithiobacillus ferrooxidans, Acidithiobacillus thiooxidans and Leptospirillum ferrooxidans. Hydrometallurgy. 83, p. 245-254. Hendrickson, D.N., Hollander, J.M., Jolly, W.L., 1969. Nitrogen ls electron binding energies. Correlations with molecular orbital calculated nitrogen charges. Inorg. Chem. 8, p. 26422647. Herbert, R.B., 1997. Properties of goethite and jarosite precipitated from acidic groundwater, Dalarna, Sweden. Clay. Clay Miner. 45, p. 261-273. Hippe, H., 2000. Leptospirillum gen. nov. (ex Markosyan 1972), nom. rev., including Leptospirillum ferrooxidans sp. nov. (ex Markosyan 1972), nom. rev. and Leptospirillum thermoferrooxidans sp. nov. (Golovacheva et al. 1992). Int. J. Syst. Evol. Microbiol. 50, p. 501-503.
49
Hosseini, T.R., Kolahdoozan, M., Tabatabaei, Y.S.M., Oliazadeh, M., Noaparast, M., Eslami, A., Manafi, Z., Alfantazi, A., 2005. Bioflotation of Sarcheshmeh copper ore using Thiobacillus ferrooxidans bacteria. Miner. Eng. 18, p. 371-374. Kelly, D.P. and Wood, A.P., 2000. Reclassification of some species of Thiobacillus to the newly designated genera Acidithiobacillus gen. nov., Halothiobacillus gen. nov. and Thermithiobacillus gen. Nov. Int. J. Syst. Evol. Microbiol. 50, p. 511-516. Kinzler, K., Gehrke, T., Telegdi, J., Sand, W., 2003. Bioleaching - a result of interfacial processes caused by extracellular polymeric substances (EPS). Hydrometallurgy. 71, p. 83-88. Lindberg, B.J., Hamrin, K., Johansson, G., Gelius, U., Fahlmann, A., Nordling, C., Siegbahn, K., 1970. Molecular Spectroscopy by Means of ESCA II. Sulfur compounds. Correlation of electron binding energy with structure. Phys. Scripta. 1, p. 286-298. Lindberg, B.J., Hedman, J., 1975. Molecular spectrocopy by means of ESCA, IV. Group shifts for N, P and as compounds. Chem. Scr. 7, p. 155-166. Mangold, S., Laxander, M., Harneit, K., Rohwerder, T., Claus, G., Sand, W., 2008. Visualization of Acidithiobacillus ferrooxidans biofilms on pyrite by atomic force and epifluorescence microscopy under various experimental conditions. Hydrometallurgy. 94, p. 127-132. Markosyan, G. E., 1972. A new iron-oxidizing bacterium - Leptospirillum ferrooxidans nov. gen. nov. sp. Biol. J. Armenia. 25, p. 26-29. McIntyre, N.S., Zetaruk, D.G., 1977. X-ray photoelectron spectroscopic studies of iron oxides. Anal. Chem. 49, p. 1521-1529. Mesquita, L.M.S., Lins, F.F., Torem, M.L., 2003. Interaction of a hydrophobic bacterium strain in a hematite–quartz flotation system. Int. J. Miner. Process. 71, p. 31-44. Meyer, G., Schneider-Merck, T., Böhme, S., Sand, W., 2002. A simple method for investigations on the chemotaxis of A. ferrooxidans and D. vulgaris. Acta Biotechnol. 22, p. 391–399. Mills, P., Sullivan, J.L., 1983. A study of the core level electrons in iron and its three oxides by means of X-ray photoelectron spectroscopy. Phys. D: Appl. Phys. 16, p. 723-732. Mishra, M., Bukka, K., Chen, S., 1996. The effect of growth medium of Thiobacillus ferrooxidans on pyrite flotation. Min. Eng. 9, p. 157-168. Natarajan, K.A., Das, A., 2003. Surface chemical studies on „Acidithiobacillus“ group of bacteria with reference to mineral flocculation. Int. J. Miner. Process. 72, p. 189-198 Natarajan, K.A., Deo, N., 2001. Role of bacterial interaction and bioreagents in iron ore flotation. Int. J. Miner. Process. 62, p. 143–157. Neumann, A.W., Good, R.J., Hope, C.J., Sejpal, M., 1974. An equation-of-state approach to determine surface tensions of low-energy solids from contact angles. J. Coll. Inter. Sci., 49, p. 291-304. Ohshima, H., 1994. Electrophoretic mobility of soft particles. J. Coll. Inter. Sci. 163, p. 474483.
50
Owens, D.K., Wendt, R.C., 1969. Estimation of the surface free energy of polymers. J. Appl. Polym. Sci. 13, p. 1741-1747. Patra, P., Natarajan, K.A., 2006. Surface chemical studies on selective separation of pyrite and galena in the presence of bacterial cells and metabolic products of Paenibacillus polymyxa. J. Coll. Inter. Sci. 298, p. 720-729. Peeling, J., Hruska, F.E., McIntyre, N.S., 1978. ESCA spectra and molecular charge distributions for some pyrimidine and purine bases. Can. J. Chem. 56, p.1555-1561. Phalguni, A., Modak, J.M., Natarajan, K.A., 1996. Biobeneficiation of bauxite using Bacillus polymyxa, calcium and iron removal. Int J Miner Process. 48, p. 51-60. Poortinga, A.T., Bos, R., Norde, W., Busscher, H.J., 2002. Electric double layer interactions in bacterial adhesion to surfaces. Surf. Sci. Reports. 47, p. 1-32. Raichur, A.M., Misra, M., Bukka, K., Smith, R.W., 1996. Flocculation and flotation of coal by adhesion of hydrophobic Mycobacterium phlei. Coll. Surf. B 8, p. 13-24. Rijnaarts, H.M., Norde, W., Lyklema, J., Zehnder, A., 1995. The isolectric point of bacteria as an indicator for the presence of cell surface polymers that inhibit adhesion. Coll. Surf., B. 4, p. 191-197. Rohwerder, T. Gehrke, T., Kinzler, K., Sand, W., 2003. Bioleaching review part A: Progress in bioleaching: fundamentals and mechanisms of bacterial metal sulfide oxidation. Appl. Microbiol. Biotechnol. 63, p. 239-248. Russat, J., 1988. Characterization of polyamic acid/polyimide films in the nanometric thickness range from spin-deposited polyamic acid. Surf. Interface Anal. 11, p. 414-420. Sand, W., Florian, B., Noel, N. 2009. Mechanisms of bioleaching and the visualisation of these by combined AFM & EFM. Adv. Mat. Res. 71-13, p. 297-302. Sanhueza, A., Ferrer, I.J., Vargas, T., Amils, R., Sánchez, C., 1999. Attachment of Thiobacillus ferrooxidans on synthetic pyrite of varying structural and electronic properties. Hydrometallurgy. 51, p. 115-129. Santhiya, D., Subramanian, S., Natarajan, K.A., 2000. Surface chemical studies on galena and sphalerite in the presence of Thiobacillus thiooxidans with reference to mineral beneficiation. Miner. Eng. 13, p. 747-763. Santhiya, D., Subramanian, S., Natarajan, K.A., Rao, K.H, Forssberg, K.S.E., 2001. Biomodulation of galena and sphalerite surfaces using Thiobacillus thiooxidans. Int. J. Miner. Process. 62, p. 121-141. Schrader, B., 1995. Infrared and Raman spectroscopy : methods and applications. Weinheim: VCH. Sharma, P., Das, A., Rao, K.H., Forssberg, K.S.E., 1999. Thiobacillus ferrooxidans interaction with sulphide minerals and selective chalcopyrite flotation from pyrite. Advances in Flotation Technology. 1999, p. 147-165. Sharma, P., Das, A., Rao, K.H., Forssberg, K.S.E., 2003. Surface characterisation of Thiobacillus ferrooxidans cells grown under different conditions. Hydrometalurgy. 71, p. 285-292. 51
Sharma, P.K., Hanumantha Rao, K., 1999. Role of a heterotrophic Paenibacillus polymyxa bacteria in the bioflotation of some sulphide minerals. Miner. Metal. Process. 16, p. 35-41. Sharma, P.K, Hanumantha Rao, K., 2002. Analysis of different approaches for evaluation of surface energy of microbial cells by contact angle goniometry. Adv. Colloid Interface Sci., 98, p. 341-463. Sharma, P.K, Rao, K.H., 2005. Surface characterization of bacterial cells relevant to the mineral industry. Miner. Metal. Process. 22, p. 31-37. Sharma, P.K., Hanumantha Rao, K., Forssberg, K.S.E., Natarajan, K.A., 2001. Surface chemical characterisation of Paenibacillus polymyxa before and after adaptation to sulphide minerals. Int. J. Miner. Process. 62, p. 3-25. Sharma, P.K., Rao, K.H., Natarajan, K.A., Forssberg, K.S.E., 2000. Bioflotation of sulphide minerals in the presence of heterotrophic and chemolitotrophic bacteria. Proceedings of the XXI International Mineral Processing Congress. B8a, p. 94-103. Silverman, M.P. and Lungren, D.G., 1959. Studies on the chemoautotrophic iron bacterium Ferrobacillus ferrooxidans – I. An improved medium and harvesting procedure for securing high cell yields. J. Bacteriol. 77, p. 642–647. Smith, R.W., Mishra, M., 1991. Mineral bioprocessing – an overview. In: Mineral Bioprocessing, R.W. Smith and M. Mishra (Eds.), The Minerals, Metals and Materials Society, Pennsylvania, p. 3-26. Temple, K. L. and Colmer, A. R., 1951. The autotrophic oxidation of iron by a new bacterium, Thiobacillus ferrooxidans. J. Bacteriol. 62, p. 605-611. Van der Walt, J.P., von Arx, J.A., 1980. The yeast genus Yarrowia gen, nov. Antonie van Leeuwenhoek. 46, p. 517-521. Van Oss, C.J., Good, R.J., Chaudhury, M.K., 1986. The role of van der Waals forces and hydrogen bonds in “hydrophobic interactions” between biopolymers and low energy surfaces. J. Coll. Inter. Sci. 111, p. 378-390. Van Oss, C.J., Chaudhury, M.K., Good, R.J., 1987. Monopolar Surfaces. Adv. Coll. Inter. Sci. 28, p. 35-64. Verwey, E.J., Overbeek, J.T.G., 1955. Theory of the stability of lyophobic colloids. J. Coll. Sci. 10, p. 224-225. Vijayalakshmi, S.P., Raichur, A.M., 2002. Bioflocculation of high-ash Indian coals using Paenibacillus polymyxa. Int. J. Miner. Process. 67, p. 199-210. Yelloji Rao, M.K., Natarajan, K.A., Somasundaran, P., 1992. Effect of biotreatment with Thiobacillus ferrooxidans on the floatability of sphalerite and galena. Miner. Metal. Process. 9, p. 95-100. Young, T. 1805. An essay on the cohesion of fluids. Phil. Trans. 95, p. 65-87. Zheng, X., Arps, P.J., Smith, R.W., 2001. Adhesion of two bacteria onto dolomite and apatite: their effect on dolomite depression in anionic flotation. Int. J. Miner. Process. 62, p. 159-172.
52
Appended papers
Paper I Miccoorganisms in Flotation and Flocculation of Minerals – An overview. Vilinska, A., Hanumantha Rao, K., Forssberg, K.S.E., 2008. In: proceedings of XXIV International Mineral Processing Congress, D.Z. Wang et al. (Eds.), Science press, Beijing, 2008, 1, pp. 22-39.
Paper II Surface characterisation of Acidithiobacillus ferrooxidans and copper and zinc ions adapted cells. Vilinska A. and Hanumantha Rao K., 2009. Geomicrobiology, submitted.
Surface characterisation of Acidithiobacillus ferrooxidans and copper and zinc ions adapted cells A. Vilinskaa and K. Hanumantha Raob Division of Mineral Processing, Luleå University of Technology, 971 87 LULEÅ, Sweden e-mail:
[email protected] ,
[email protected]
Abstract Changes in surface chemical properties of Acidithiobacillus ferrooxidans after adaptation to high copper and zinc ion concentration were studied by surface sensitive techniques such as zeta-potential, XPS and FT-IR measurements. The adapted bacteria were also characterized by their surface energies and adhesion capacities on different sulphide minerals. Their surface negative charge was decreased due to changes in the structure of bacterial surface layers. The metal ions adapted cells secreted more extracellular polymeric substances with a modified composition compared to unadapted ferrous ions grown cells. Bacterial cells hydrophilic property increased after adaptation and their adhesion behaviour to sulphide mineral was altered. Keywords: Bacteria, Copper Ions, Zinc Ions, Adaptation, Surface Characterisation INTRODUCTION Microorganisms are used in different biotechnological industries and also in metallurgical industrial processes involving minerals. Biohydrometallurgical processes are extensively studied and bioleaching is a well established method to recover valuable metals from ores. Attempts to use microorganisms in mineral beneficiation are also well known and in this case, the physico-chemical characteristics of microbes play a significant role to induce surface alternations upon adhesion to minerals. There is no clear understanding on the manner, in which bacteria affect the surface reactivity and the mechanism of bacteria adsorption. What happens at mineral-bacteria interfaces in aqueous solution control many biotechnologies and it is essential to understand how bacteria select and adhere to favourable minerals. Flotation is the most studied process in biobeneficiation where bacteria or fungi cells are used as a collector or depressant depending on their surface character. Heterotrophic microorganisms are often chosen because of their fast growth rate and the presence of high amounts of secretion products on their surface. Paenibacillus polymyxa mostly contains hydrophilic surface and has been successfully used in numerous flotation and flocculation studies as a depressant (Phalguni et al. 1996 Deo and Natarajan 1998 Vijayakshmi et al. 2002), while Mycobacterium phlei and Rhodococcus opacus exhibit hydrophobic character due to mycolic acid on their surface and they are employed as a collector in flotation (Zheng et al. 2001 Botero et al. 2007). Representatives of the genus Acidithiobacillus are often used as bioreagents in sulphides processing. The iron and sulphur oxidizing bacillus shaped At. ferrooxidans cells are hydrophilic and they were capable to depress pyrite (Mishra et al. 1996 Sharma et al. 2001) as well as to flocculate minerals selectively (Mishra et al. 1996 Natarjan and Das 2003). The success of these processes depends mainly on the amount of adhered cells on the minerals and their surface chemical characteristics. To achieve selectivity in mineral 1
systems, either in flotation or flocculation, microorganisms have to adhere selectively onto minerals modifying their surface properties significantly. Bioflotation and bioflocculation are rather fast processes and there is no sufficient time for microorganisms to chemically react with the mineral surface and only the bacterial adhesion is considered as a driving mechanism to modify the surfaces. The nature of cells adhesion onto minerals and their surface hydrophilicity or hydrophobicity character can be regulated by changing the conditions of their growth environment. A change in the chemical composition or presence of mineral particulates in the growth medium can modify the adhesion properties of cells on minerals and their role to function as a flotation reagent (Sharma et al. 1999 Sharma and Rao 1999). Bacteria are very sensitive to any change in their growth environment and adapt to growth conditions with production of some specific compounds. Studies on cells surface characterisation during leaching and their impact on leaching efficiency showed that adaptation of At. ferrooxidans to chalcopyrite reduced the negative charge of bacteria and increased their growth and adhesion to the chalcopyrite surfaces as well as their leaching efficiency (Chen et al. 2008). In addition, the contact angle of the cell with water is decreased and became more hydrophilic and polar compared to wild counterparts (Xia et al. 2008). Some special structures were detected on the cells attached to arsenopyrite surface, they are most likely surface wessels or precipitates (Jones et al. 2002). During leaching some cells are attached on the surface (sessile cells) and some remained in the solution (planktonic cells), and the differences between those and only iron grown cells were studied by Harneit et al. (2006). The extracellular polymeric substances (EPS) present on the surface of At. ferrooxidans were estimated and the amount of EPS was found to be minimal on iron grown cells, increased several times for the unattached planktonic cells and several folds higher for attached cells on the pyrite surface. The composition of EPS is also slightly varied with different substrates used and this is found to be helpful in the attachment of cells to minerals (Harneit et al. 2006). Cells adaptation to different mineral substrates and to higher metal ions concentration for their tolerance during leaching was also studied. Sulphur grown cells exhibit some degree of hydrophobic character and their adhesion to chalcopyrite was observed to be higher than pyrite (Solari et al. 1992). At. ferrooxidans strain grown in the presence of copper has a delayed growth with prolonged lag phase but these cells increased the surface potential and the amounts of proteinious surface structures (Das et al. 1998). Chandraprabha et al. (2009) have grown cells in the presence of arsenic and copper metal ions and found changes in electrophoretic mobilities and charge densities due to a different surface polymer production. At. ferrooxidans cells grown in ferrous ions medium and in the presence of Cu and Zn ions have been characterized for their surface chemical composition and surface properties by FTIR, XPS, zeta-potential, contact angle and adsorption measurements in this study. The changes in surface properties occurred during metal ions adaptation have been presented and discussed with reference to their adhesion behaviour onto minerals. MATERIALS AND METHODS A. ferrooxidans cells (type strain ATCC 23270) were grown in a simple Mackintosh medium (40 g/l FeSO4.7H2O, 0,132 g/l (NH4)2SO4, 53 mg MgCl2.6H2O, 27 mg KH2PO4, 0,147 g CaCl2.2H2O, 62 ȝg MnCl2.2H2O, 68 ȝg ZnCl2, 64 ȝg CoCl2.6H2O, 31 ȝg H3BO3, 10 ȝg Na2MoO4, 67 ȝg CuCl2.2H2O) and metal ion enriched mackintosh media. Cells were grown in the presence of 0.3 M Cu2+ (CuSO4.5H2O) and Zn2+ (ZnSO4.7H2O) ions. The bacterial 2
adaptation was performed by a serial sub-culturing of stepwise increase of metal ions concentration and the cells were considered “adapted“ when the growth time and cell numbers were comparable to the growth in the medium. After reaching the stationary phase the cells mass was collected by Millipore filtration, washed with a solution of H2SO4 at pH 2 and centrifuged if necessary to obtain a concentrated cell suspension. Pure mineral samples of chalcopyrite, pyrite and sphalerite were crushed and ground in agate mill and sorted by wet sieving to obtain different size fractions. A portion of finest fraction was further ground and a size fraction of –5 μm was obtained by a micron filter cloth sieving in ultrasonic bath. The surface area was estimated by BET to be 1.02 m2/g for pyrite, 1.9 m2/g for chalcopyrite and 1.75 m2/g for sphalerite. Zeta-potentials of washed and precipitate free cells of unadapted and adapted At. ferrooxidans were measured with ZetaCompact in 10-2M KNO3. The instrument is supplied with Zeta4 software which allows direct reading of zeta-potentials calculated from the mobility using Smoluchovski equation. A cell concentration of 2.5x107 cells/ml was used as previously found to be suitable for the instrument. The pH was adjusted with HNO3 and KOH to different values. The measurement itself was realized after 30 minutes of conditioning at a particular pH and the pH was once more measured directly before the measurement and considered as the pH of the measurement. Samples were analyzed in a natural wet environment avoiding freeze drying and potential changes on the surface. The bacterial paste was placed on a sample holder and introduced to a liquid nitrogen pre-cooled chamber while the analysis was preformed at -150°C. The spectra were collected using Kratos axis Ultra electron spectrometer. An Al cathode operated at 150 W was used as the X-ray source. Possible charge created on the surface of the sample was controlled by the spectrometer charge neutralization system. The binding energy (BE) scale was referenced to the C 1s peak of aliphatic carbon at 285.0 eV. Cells of At. ferrooxidans grown in iron (II) media, copper and zinc enriched iron (II) media were studied by FTIR. Cells were collected by filtration, washed to remove precipitates, concentrated by centrifugation (3000 rpm for 30 s) and dried. To preserve the biological material air-drying was done overnight. The samples for diffuse reflectance infrared spectra measurements were prepared using KBr matrix. Spectra were recorded using Perkin-Elmer FTIR spectrometer at 2.5 wt% of dry bacterial mass at a 100 scan mode. Bacterial mass collected and washed was filtered through Millipore filter paper with a pore size of 0.22 ȝm which is smaller than the cells to obtain a uniform cell layer on the surface of filter paper. The filter paper with bacterial lawns is then mounted on a glass slide with the help of double sided tape and left for drying at room temperature. To obtain a stable and reproducible contact angle of bacteria, a cell lawn composed of more than 200 layers is used. The contact angle is estimated by the sessile drop method; a drop of test liquid is placed on the surface and the contact angle between the surface and drop is measured. For this purpose KrĦss Easy Drop equipment and Drop shape analysis software was used. The instrument is equipped with a CCD camera and a dynamic contact angle is measured and the stabilized contact angle was taken as a result. Each experiment was repeated 3 times and the arithmetic mean value was considered as the final contact angle for a particular test liquid. For the calculation of surface energy and its components the van Oss acid base method was used. For this calculation, contact angle data from 3 different liquids are required. In this 3
study 4 well defined liquids with known components were used: water, dioodomethane, 1bromonaphtalene and formamide. Surface tensions and the dispersive and polar part contributions, and polar part divided into acid-base components as reported in the literature were used (Bellon-Fontaine et al. 1990; Ström et al. 1987). Adhesion tests were carried out using the finest mineral sample prepared (-5 ȝm) because of its’ high surface area. The surface area of each minerals was estimated by BET. Mineral – liquid ratio was kept constant during all adsorption tests at 1 g/100 ml while the concentration of bacteria was changed. The adhesion characteristics were studied at 5x106, 107, 5x107, 108, 5x108, 109 and 5x109 cells/ml concentrations. At. ferrooxidans cells and At. ferrooxidans cells grown in the presence of 0.3 M zinc and copper ions were studied simultaneously. In order not to compromise the viability of the bacteria an acidic pH was decided to be used during the experiments and was kept at pH 2-3 with H2SO4. After 30 minutes of agitation the amount of remaining cells in the supernatant was estimated by counting under optical microscope. RESULTS AND DISCUSSION
Zeta-potential (mV)
The zeta-potential of iron grown, copper and zinc adapted At. ferrooxidans cells is shown in Fig. 1. Unadapted At. ferrooxidans cells were negatively charged in the entire pH range studied without displaying iso-electric point. An extrapolation of the curve shows that it is located at or below pH 1. Cells after adaptation to copper and zinc metal ions were positively charged below pH 2.2 and pH 3 respectively. The iso-electric point below pH 3 indicates an anionic polysaccharide surface and only the zinc grown cells is higher which is typical for peptidoglycan surface but still too low to have proteinaceous surface (Rijnaarts et al. 1995). The magnitude of electronegativity is decreased for zinc grown cells. A shift in iso-electric point and decrease in surface charge of bacteria could be due to the adsorption of specific ions, possibly metal cations in this case, and secondly a change in surface chemical composition due to secretion of different compounds caused by the varied growth environment. 20
Fe grown
15
Zn grown Cu grown
10 5 0 -5 0
2
4
6
8
10
12
-10 -15 -20 pH
Fig. 1. Zeta-potential of different At. ferrooxidans cells at 0.01M ionic strength as a function of pH. Since the X-rays used for XPS analysis penetrate up to a depth of 10 nm of the bulk solids, the spectral data represent the bacterial surface layer. Presence of different metal ions during bacteria growth has not changed the composition of the surface markedly and only minor 4
changes are noticed (Fig.2, Table 1.). Some functional groups were affected to a minor degree, but carbon groups associated with oxygen showed significant changes (Fig.3). The C(O, N) groups remained similar after copper adaptation but were increased for zinc adapted cells. The percentage of C=O and C-N-O groups were decreased for zinc adapted cells but their amount increased in case of copper adapted cells. Organic oxygen decreased in the order of unadapted, zinc and copper grown cells. The Fe-ions are expected to be found on cells surface either as ions or precipitate but were found only on the iron grown cells. In the case of adapted cells, they were below the detection limit. The iron found for the case of original cells was probably in form of ions and not as precipitates, because no potassium was detected in XPS results. The iron is present in Fe2+ and Fe3+ oxidation states, consistent with the chemical shifts of the peaks and in the ratio of 1:2. No copper or zinc metal ions were detected on the cell surface. Atomic concentration (%)
30
Fe grown Zn grown Cu grown
25 20 15 10 5
Su lp ha te Ph os ph at e Fe (II &I II)
rg an ic )
(o
S
N
O
N
(o rg
(p ro
to na te d)
) 2O
an ic ,
(o rg
H
an ic )
O H O C
O
-O ,N =O ,N -C -O C
C
C
-( C
,H
)
0
Fig. 2. XPS analysis of different At. ferrooxidans cells 0,5
Fe grown
Atomic ratios
0,4
Zn grown Cu grown
0,3 0,2 0,1 0,0 N/C
P/C
O/C
CH/C
C(O,N)/C C=O/C COOH/C
NH+/N
Fig. 3. XPS analysis of different At. ferrooxidans cells, relative elemental visualisation The nitrogen-carbon ratio was the highest for unadapted cells followed with copper grown and the lowest for zinc grown cells of At. Ferrooxidans (Fig.3). The proteins are mostly present embedded in the cell wall and the decrease of nitrogen compound compared with carbon suggests the secretion of some carbohydrates on the surface of the cells. The amount of total oxygen to total carbon was decreasing in the direction from unadapted to zinc and copper adapted cells. The amount of aliphatic carbon from the total carbon decreased after 5
cell adaptation, C-(O, N) groups increased for zinc adapted cells and remained almost equal after copper adaptation. The number of carboxylic groups to the total carbon has not changed after zinc adaptation but decreased for copper adapted cells. Unadapted At. ferrooxidans cells had a much higher protonated nitrogen groups compared to metal ions adapted cells. Table 1. XPS analysis of different At. ferrooxidans cells Line
C 1s
O 1s N 1s S 2p 3/2 P 2p 3/2 Fe 2p 3/2
Fe grown BE, AC, eV at.% 285.0 28.86 286.7 26.38 288.2 9.82 289.3 1.04 531.6 4.86 533.1 23.08 400.2 4.09 401.8 0.76 163.6 0.08 168.7 0.36 133.4 0.09 709.6 0.56 711.8 Fe(III) : Fe(II) 2 :1
Zn grown BE, AC, eV at.% 285.0 28.76 286.7 29.1 288.2 8.74 289.3 1.13 531.6 3.83 533.1 23.57 400.2 3.96 401.9 0.48 163.6 0.06 168.7 0.3 133.5 0.12
Cu grown BE, AC, eV at.% 285.0 29.21 286.6 26.85 288.2 11.42 289.6 0.84 531.6 3.01 533.1 23.48 400.2 4.17 402.0 0.55 163.6 0.09 168.7 0.28 133.6 0.1
bdl
bdl
no Zn
no Cu
identification C-(C,H) (Beamson and Griggs, 1992) C-(O,N) (Clark and Thomas, 1976) C=O N-C-O (Russat, 1988) COOH (Gelius et al. 1970) Organic (Peeling et al. 1978) organic, H2O (Boulanger et al. 1988) N (Hendrickson et al. 1969) proton. N (Lindberg and Hedman,1975) Organic S (Beyer et al. 1981) Sulphate (Lindberg et al. 1970) Phosphate (Fluck and Weber, 1974) Fe(II) (Mills and Sullivan, 1983) Fe(III) (McIntyre and Zetaruk, 1977)
The extracellular polymeric substances (EPS) of At. ferrooxidans cells were mainly composed of sugars, lipids and loosely associated fatty acids (Kinzler et al. 2003, Harneit et al. 2006) and the composition is dependent on the growth media. Based on chemical structures, sugars contain higher number of C-O groups per total amount of carbon in molecule than fatty acids. The increase of this group on the surface of zinc adapted cells may indicate an increase of sugar compounds. At the same time, the amount of C-H prevalent in lipids was decreased. Copper adapted cells lower C-(H, C) carbon amount was compensated by increased production of C=(O)/N-C-O rich substances. The increase of oxygen rich compounds is not consistent with the decrease of oxygen content of the samples, but the decrease of nitrogenous compounds abundant in the cell wall area suggests some increased secretion of EPS and thus the decrease of signal intensity from an area of higher oxygen abundance. Increased coverage of cell surface by EPS is probably the reason for decreased rate of protonated nitrogen by changing the environment around the protein. Changing the growth conditions changed the amount and composition of surface related compounds and different metal ions induced the production of different compounds. FTIR spectra recorded for the samples are presented in Fig. 4 and the identified peaks are summarized in Table 2. Copper and zinc adapted cells spectra were almost overlapping while iron grown cells produced weaker signal. There was also some change in the shape of the absorption band peaks, especially the one at 3300 cm-1 and around the 1700-1500 cm-1 region. A broad band around 3300 cm-1 can represent different groups but considering the composition of cell surfaces the –OH groups seems reasonable due to high sugar contents. The carbohydrates are represented by numerous peaks between 3069 and 2855 cm-1 and one asymmetric deformation at 1456 cm-1. The 1700 cm-1 region is typical for carbonyl groups (aldehydes, ketones, esters), the actual frequencies are differing few cm-1 for different samples, probably due to different environment around the functional group. Nitrogenous 6
substances are proved by the existence of different peaks from 1653 to 1500 cm-1. Copper and iron grown cells spectra showed more recognizable nitro group peaks (1558 and 1519 cm-1) compared to zinc grown cells. The spectrum of iron grown cells of At. ferrooxidans was not continuous at this region and more peaks were recognizable, but all represented different C=N or C-NO2 groups. At lower wave numbers some sulphur components and CH-NH2 bands were visible.
Fig. 4. FTIR absorbance diagram of At. ferrooxidans cells grown under different conditions Fe grown, copper adapted and zinc adapted. Table 2. FT-IR band identification for At. ferrooxidans cells (Colthup et al. 1990, Schrader 1995) Wavenumber (cm-1) Groups Description 3300 -OH, -NH2, -NH, CŁC 3069 =CH 2959 -CH3 aliphatic CH3 2925 -CH2 aliphatic CH2 2875 -CH3 2855 -CH2 aliphatic CH2 1733 C=O carbonyl group (aldehydes) 1717 C=O carbonyl group (ketones, esters) 1653 -C=N 1st amide band 1558 C-NO2 aliphatic nitro 1540 NO aliphatic nitro (2nd amide band) 1519 C-NO2 aromatic nitro 1456 CH3 aliphatic CH3 (asymetric deformation) 1234 CH2-S 1159 SO2 1075 CH2-NH2, CH2-OH 1046 CH-NH2 7
In Table 3 the measured contact angle data and calculated surface energies are presented. The contact angle on cells surface with water after metal ions adaptation has decreased for both Cu and Zn ions, developing a more hydrophilic surface for the adapted cells. The total surface energies were increasing in the order of unadapted, zinc adapted and copper adapted cells. The polar components contribution to surface energy followed the same order while the polar part in total surface energy was enhanced. The surface of At. ferrooxidans cells is thus becoming more polar in character after metal ion adaptation. For both cases of adaptation, the electron donating property was enhanced while the electron accepting character decreased for zinc adapted cells and increased for copper adapted cells. Thus the presence of metal ions in the growth media caused changes in surface chemical composition due to enhanced secretion of metabolites with polar structure and negative (electron donating) character. Table 3. At. ferrooxidans contact angle and surface energy results. Contact angle ș DiiodoBromomethane naphtalene formamide
Material
water
At.f. Fe grown At.f. Zn grown
33 26.57
53.5 -
36 40
At.f. Cu grown
7.5
-
35.55
Surface energy Ȗ (mJ.m-2) total
Ȗd/ȖLW
ȖP/ȖAB
Ȗ+
Ȗ-
32.5 29.73
48.55 49.54
34.31 34.78
14.24 14.76
1.11 1.06
45.57 51.46
14.8
55.36
36.67
18.68
1.55
56.43
In Fig. 5 the adsorption density of iron grown cells on different minerals as a function of equilibrium concentration is presented. Complete surface saturation was not reached within the cells concentration range studied, but above 108 cell/ml concentrations there was a visible levelling off the curves indicating the proximity of a saturation concentration. However, the curves show a clear difference in adsorption preferences of the cells to minerals. The cells adsorption density onto pyrite was the highest followed by chalcopyrite and sphalerite. Comparing the coverage of the surface of the different minerals by adhered cells, assuming their adherence to pyrite was 100%, chalcopyrite was covered by 80.6% and sphalerite by 47.2% at equilibrium.
Fig. 5. Unadapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite
8
Zinc adapted cells of At. ferrooxidans adhered mostly onto pyrite and chalcopyrite before reaching the saturation level but not onto sphalerite at the same level (Fig. 6). The isotherm for sphalerite indicates close to a complete surface saturation at and above 108 cell/ml concentration. The selective and higher adhesion of unadapted cells to pyrite has completely disappeared with the cells after adaptation to zinc ions. The zinc adapted cells adhere to chalcopyrite to the same extent as to pyrite but their adsorption onto sphalerite decreased to 41.3% compared to pyrite and chalcopyrite.
Fig. 6. Zinc adapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite At. ferrooxidans grown in the presence of copper ions showed the lowest adhesion capacities to all the three minerals used in this study (Fig. 7). The cells almost reached surface saturation above 107 cell/ml equilibrium concentration. The cell adhesion to chalcopyrite and pyrite is comparable at higher cell concentrations. Considering this adsorption density as surface saturation level, then the cells adhesion on sphalerite corresponds to 47.5% surface coverage.
Fig. 7. Copper adapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite
9
Adsorption results showed that copper grown cells lessened their adhesion on all the three minerals. The highest adhesion to pyrite and chalcopyrite was achieved with zinc adapted cells but led to a loss of selectivity towards pyrite. Unadapted cells of At. ferrooxidans are the most effective towards sphalerite but the adsorption density is still the lowest among all the tested minerals. At. ferrooxidans cells adapted to higher metal ion concentrations resulted in some distinct changes in surface properties. The negative surface charge was decreased due to increased production of extracellular polymeric substances. The adsorption of Cu and Zn cations on the surface cannot be accounted for a decrease in negative charge since the XPS analysis showed that the surfaces are devoid of these ions. XPS confirmed the presence of ferric and ferrous ions on the surface of iron grown cells but not on Cu and Zn ions adapted cells. FTIR spectra showed multiple peaks at C=N and C-NO2 absorbance band regions for the unadapted cells as well as a more intense peak of carbon triple bond for metal ions adapted cells suggesting the presence of an additional surface compound layer. The decrease of nitrogen content more prevalent around the area of cell wall supports the increased EPS production. Metal ions adaptation changes the composition of EPS by decreasing the amounts of lipids and increasing the production of sugars, thereby making the surfaces more hydrophilic. The presence of zinc and copper ions during bacterial growth induced the production of different surface chemical compounds. Contact angle measurements and surface energy calculations confirmed increased hydrophilicity after metal ions adaptation as well as increased polarity of adapted cells. Ferrous iron grown cells adhere well to minerals, mostly to pyrite, followed by chalcopyrite and sphalerite. Metal ions adapted cells have increased the adhesion on chalcopyrite, but the preference to pyrite and sphalerite was stable. Copper grown cells possess the lowest adherence ability onto minerals. Increased hydrophilicity of minerals due to bacterial adhesion may lead to depression in flotation and this effect is proportional to the amount of cells adhered on the minerals. Metal ions adaptation has not increased the gap between the cells adherence towards pyrite and chalcopyrite thus the final process cannot be selective. On the other hand zinc grown cells adhered worst to sphalerite while maintaining the original adhesion towards pyrite and same to chalcopyrite suggesting selective depression of sphalerite during flotation. Acknowledgements Financial support from Kempestiftelsen foundation in the form of scholarship to the author, Ms. A. Vilinska, is gratefully acknowledged. REFERENCES Beamson G, Briggs D. 1992. High Resolution XPS of Organic Polymers: the Scienta ESCA300 Database. New York: Wiley. 295 p. Bellon-Fontaine MN, Mozes N, Van Der Mei HC, Sjollema O. 1990. A comparison of Thermodynamic Approaches to predict the adhesion of dairy microorganisms to solid substrata. Cell Biophysics 17:93-106. Beyer L, Kirmse R, Stach J, Szargan R, Hoyer E.Z. 1981. Metallkomplexe des Benzoyldithioessigsäuremethylesters und des N-Benzoylamino-dithiokohlensäureethylesters:
10
Darstellung und Charakterisierung, ESCA- und EPR-Untersuchungen. Anorg Allg Chem. 476:7-15. Botero AEC, Torem ML, Mesquita LMS. 2007. Fundamental studies of Rhodococcus opacus as a biocollector of calcite and magnesite. Miner Eng. 20:1026-1032. Boulanger P, Riga J, Delhalle J, Verbist JJ. 1988. .An XPS study of hexagonal polyoxymethylene with various bulk morphologies: surface modification under X-ray exposure. Polymer. 29:797-801. Chandraprabha MN, Modak JM, Natarajan KA. 2009. Soft-particle model analysis of effect of LPS on electrophoretic softness of Acidithiobacillus ferrooxidans grown in the presence of different metal ions. Colloids Surf B Biointerfaces. 69:1-7. Chen M-L, Zhang L, Gu G-H, Hu Y-H, Su L-J. 2008. Effects of microorganisms on surface properties of chalcopyrite and bioleaching. Trans Nonferrous Met Soc China. 18:1421-1426. Clark DT, Thomas HR. 1976. Applications of ESCA to polymer chemistry. XI. Core and valence energy levels of a series of polymethacrylates. J Polym Sci Polym Chem Ed. 14:1701-1713. Colthup NB, Daly LH, Wiberley SE. 1990. Introduction to infrared and Raman spectroscopy. 3rd ed. Boston: Academic Press. 547 p. Das A, Modak JM, Natarajan KA. 1998. Surface chemical studies of Thiobacillus ferrooxidans with reference to copper tolerance. Antonie van Leeuwenhoek. 73:215-222. Deo N, Natarajan KA. 1998. Studies on interaction of Paenibacillus polymyxa with iron ore minerals in relation to beneficiation. Int J Miner Process. 55:41-60. Fluck E, Weber D Z. 1974. Naturforsch. 29:603. Gelius U, Heden PF, Hedman J, Lindberg BJ, Manne R, Nordberg R, Nordling C, Siegbahn K. 1970. Molecular Spectroscopy by Means of ESCA III. Carbon compounds. Phys Scripta. 2:70-80. Harneit K, Goksel A, Kock D, Klock J-H, Gehrke T, Sand W. 2006. Adhesion to metal surfaces by cells of Acidithiobacillus ferrooxidans, Acidithiobacillus thiooxidans and Leptospirillum ferrooxidans. Hydrometallurgy. 83:245-254. Hendrickson DN, Hollander JM, Jolly WL. 1969. Nitrogen ls electron binding energies. Correlations with molecular orbital calculated nitrogen charges. Inorg Chem. 8:2642-2647. Jones RA, Koval SF, Nesbitt HW. 2003. Surface alteration of arsenopyrite by Thiobacillus ferrooxidans. Geochimi Cosmochim Acta. 67:955-965. Kinzler K, Gehrke T, Telegdi J, Sand W. 2003. Bioleaching - a result of interfacial processes caused by extracellular polymeric substances (EPS). Hydrometallurgy. 71:83-88. Lindberg BJ, Hamrin K, Johansson G, Gelius U, Fahlmann A, Nordling C, Siegbahn K. 1970. Molecular Spectroscopy by Means of ESCA II. Sulfur compounds. Correlation of electron binding energy with structure. Phys Scripta. 1:286-298. Lindberg BJ, Hedman J. 1975. Molecular spectrocopy by means of ESCA, IV. Group shifts for N, P and as compounds. Chem Scr. 7:155-166. McIntyre NS, Zetaruk DG. 1977. X-ray photoelectron spectroscopic studies of iron oxides. Anal Chem. 49:1521-1529.
11
Mills P, Sullivan JL. 1983. A study of the core level electrons in iron and its three oxides by means of X-ray photoelectron spectroscopy. Phys D: Appl Phys. 16:723-732. Mishra M, Bukka K, Chen S. 1996. The effect of growth medium of Thiobacillus ferrooxidans on pyrite flotation. Miner Eng. 9:157-168. Natarajan KA, Das A. 2003. Surface chemical studies on „Acidithiobacillus“ group of bacteria with reference to mineral flocculation. Int J Miner Process. 72:189-198. Peeling J, Hruska FE, McIntyre NS. 1978. ESCA spectra and molecular charge distributions for some pyrimidine and purine bases. Can J Chem. 56:1555-1561. Phalguni A, Modak JM, Natarajan KA. 1996. Biobeneficiation of bauxite using Bacillus polymyxa, calcium and iron removal. Int J Miner Process. 48:51-60. Rijnaarts HM, Norde W, Lyklema J, Zehnder A. 1995. The isolectric point of bacteria as an indicator for the presence of cell surface polymers that inhibit adhesion. Colloids Surf B Biointerfaces. 4:191-197. Russat J. 1988. Characterization of polyamic acid/polyimide films in the nanometric thickness range from spin-deposited polyamic acid. Surf Interface Anal. 11:414-420. Schrader B. 1995. Infrared and Raman spectroscopy : methods and applications. Weinheim: VCH. 787 p. Sharma PK, Hanumantha Rao K, Forssberg KSE, Natarajan KA. 2001. Surface chemical characterisation of Paenibacillus polymyxa before and after adaptation to sulphide minerals. Int J Miner Process. 62:3-25. Sharma P, Das A, Rao KH, Forssberg KSE. 1999. .Thiobacillus ferrooxidans interaction with sulphide minerals and selective chalcopyrite flotation from pyrite. In: Parekh BK, Miller JD, editors. Advances in Flotation Technology. SME. p. 147-165. Solari JA, Huerta G, Escobar B, Vargas T, Badilla-Ohlbaum R, Rubio J. 1992. Interfacial phenomena affecting the adhesion of Thiobacillus ferrooxidans to sulphide mineral surfaces. Colloids Surf. 69:159-166. Ström G, Frederiksson M, Stenius P. 1987. Contact angles, work of adhesion and interfacial tensions at a dissolving hydrocarbon surface. J Coll Interf Science. 119:352-361. Vijayalakshmi SP, Raichur AM. 2002. Bioflocculation of high-ash Indian coals using Paenibacillus polymyxa. Int J Miner Process. 67:199-210. Xia L, Liu X, Zeng J, Yin C, Gao J, Liu J, Qiu G. 2008. Mechanism of enhanced bioleaching efficiency of Acidithiobacillus ferrooxidans after adaptation to chalcopyrite. Hydrometallurgy. 92:95-101. Zheng X, Arps PJ, Smith, RW. 2001. Adhesion of two bacteria onto dolomite and apatite: their effect on dolomite depression in anionic flotation. Int J Miner Process. 62:159-172.
12
Paper III Adhesion to sulphide minerals by wild-type and of Cu- and Zn-adapted cells of Acidithiobacillus ferrooxidans. Vilinska A. and Hanumantha Rao K., 2009. Applied Microbiology and Biotechnology, submitted.
Adhesion to sulphide minerals by wild-type and of Cu- and Znadapted cells of Acidithiobacillus ferrooxidans A. Vilinska and K. Hanumantha Rao Division of Mineral Processing, Luleå University of Technology, SE-971 87 LULEÅ, Sweden e-mail:
[email protected];
[email protected]
Abstract Adhesion to sulphide minerals by Acidithiobacillus ferrooxidans was determined experimentally and estimated by different theoretical treatments. The presence of copper and zinc ions in the growth medium caused some surface chemical changes on the bacteria that influenced adhesion to sulphide minerals. Differential adhesion to pyrite by cells was lost after adaptation of the cells to metal ions. Copper-adapted cells exhibited decreased adhesion to all studied minerals. Theoretical consideration of electrostatic forces alone was inadequate to explain these differences in adhesion whereas a thermodynamic consideration yielded a more consistent explanation but was not sensitive enough to explain observed changes in adhesion caused by alteration in chemical composition of the cell surface. Application of extended DLVO theory resulted in greatest correlation between theoretical calculations and experimental laboratory observations. Keywords: Acidithiobacillus ferrooxidans, Metal Ion Adaptation, Adhesion, Extended DLVO theory
Introduction Microorganisms facilitate many industrial biotechnologies including those used in the mineral industry for metal extraction. Biohydrometallurgical leaching operations have been widely studied and bioleaching became a well established method for metal extraction from sulphide ores, laterite ores and low-grade minerals. Attempts of using microorganisms for surface modification in mineral processing operations such as flotation and flocculation have been made in recent years. Several types of autotrophic and heterotrophic bacteria, fungi, yeasts and algae are being utilised in minerals biobeneficiation, mostly as flotation collectors or depressants depending on the nature of their surface chemical composition. Flotation is a rather fast process and there is not sufficient time for the microorganisms to chemically react with the mineral surface. The adhering bacteria covering the minerals surface impart their own surface properties to the mineral and become the driving force in mineral flocculation and flotation. Several studies showed that both heterotrophic (Phalguni et al. 1996 Deo and Natarajan 1998 Vijayakshmi et al. 2002) and chemoautotrophic (Mishra et al. 1996 Sharma et al. 2001) cells functioned as depressants because of their surface hydrophilic character. Only few strains have a hydrophobic surface and act as collectors (Zheng et al. 2001 Botero et al. 2007). The success of all these operations mainly depends on the selective adhesion to minerals and the character of bacterial surface. Bacterial adhesion and charge transfer are crucial for biooxidation processes that occur in bioleaching. It is important to understand the mechanisms by which this adhesion and charge transfer occurs. During bioleaching some cells are attached to the mineral surface (sessile cells) and some remain in the solution (planktonic cells) (Harneit et al. 2006). Periodic attachment/detachment cycles also occur on some minerals (Yelloji Rao et al. 1997). At.
1
ferrooxidans attached to mineral surfaces have a greater amount of extracellular polymeric substances (EPS) than planktonic cells. The composition of EPS varies depending on the environment in which the cells grow. A change in composition of the growth medium may promote cell attachment to minerals (Harneit et al. 2006). Cells of At. ferrooxidans while growing in the presence of chalcopyrite exhibited increasing adhesion to the ore and enhancement of leaching (Chen et al. 2008 Xia et al. 2008). The surface of sulphur-grown cells of At. ferrooxidans exhibited some hydrophobicity as reflected by their greater adhesion to chalcopyrite than pyrite (Solari et al. 1992). Thus, the extent of cell adhesion to a mineral and the hydrophobicity and/or hydrophilicity of the cells can be controlled by the conditions under which the cells are grown. Bacteria are very sensitive to any changes in their environment and react by secretion of specific compounds as a result of adaptation. The character of a mineral determines whether bacteria can adhere to it. At. ferrooxidans cells preferentially adhere to surface defects and imperfection on sulphide minerals (Rohwerder et al. 2003). Synthetically prepared pyrite with a different degree of crystallisation resulted in different adherence of At. ferrooxidans and the cells adhere preferentially to more amorphous sites of film with higher sulphur availability (Sanhueza et al. 1999). Previous studies have shown that under most physiological conditions the bacterial cell surface carries a net negative charge, which, along with electrostatic forces, hydrophobic, entropic, acid-base and van der Waals interactions and H-bonding are important in bacterial adhesion (Poortinga et al. 2002 Sharma and Rao 2003). The adhesion of bacteria to minerals can be assessed by theoretical thermodynamic and extended DLVO approaches incorporating all these interaction forces. This paper presents the results of At. ferrooxidans cells, both unadapted and adapted to higher copper and zinc ions concentrations, adherence to pyrite, chalcopyrite and sphalerite minerals predicted on a theoretical basis and determined experimentally. Analysis of adhesion based on theory Several approaches are available to estimate the adhesive forces theoretically. Most widely used method is assuming the electrostatic forces acting between two charged particles as the most important. The charge of the particles is estimated by titration or electrophoretic mobilities. This method suffers from incorrect estimations of surface charge and does not consider adhesion when the interacting surfaces have the same polarity. More precise approaches to estimating the effect of physico-chemical interactions are the thermodynamic approach using the energetic balance and extended DLVO approach incorporating different forces into the final force. Thermodynamic approach According to thermodynamic laws, a system will undergo change and proceeds towards an energetically favoured state; the state of which after the change will have a lower total energy. In the case of bacterial adhesion to a solid surface, the energetic state of bacteria-solid system has to be lower than that of the bacteria and the solid phase, which can be expressed in terms of free energy of adhesion, ǻGadh (Absolom et al. 1983 Busscher et al. 1984). For calculation of ǻGadh, the free energy of adhesion has been divided into two parts: Lifshitz - van der Waals (LW) component and acid-base (AB) component: ǻGadh = ǻGadhAB + ǻGadhLW
(1)
2
The components of free energy of adhesion can be calculated from the interfacial tensions as follows: 'Gadh
LW
AB
'Gadh
2§¨ J bv ©
·¸§¨ J LW J LW ·¸ lv ¹© sv ¹ 2§¨ Jbv Jsv ·¸§¨ Jbv Jsv ·¸ 2§¨ Jbv Jlv ·¸§¨ Jbv Jlv ·¸ 2§¨ Jsv Jlv ·¸§¨ Jsv Jlv ·¸ © ¹© ¹ © ¹© ¹ © ¹© ¹ LW
J lv
LW
(2) (3)
where JLW represents the dispersive components of surface energy and J+ and J- the acid and base components of interacting bodies (bacteria, solid phase, liquid phase). DLVO approach The classical DLVO approach (Deryagin and Landau 1941 Verwey and Overbeek 1955) includes Lifshitz-van der Waals interactions and electrostatic interactions. LW forces are short range and always cause attraction. These are weak forces between neutral stable molecules. The Coulombic electrostatic interactions are long range and could cause attraction or repulsion, depending on the surface charge of interacted particles. Acid-base interactions were introduced later into the DLVO approach by van Oss (Van Oss et al. 1986) to involve electron donating-accepting abilities of different materials. The microbial adhesion to solid surfaces is then described by a sum of van der Waals, electrostatic and acid-base forces operating between the surfaces approaching close to one another: Gtotal = GLW + GEL + GAB
(4)
Calculation of these forces depends on the geometry of interacting phases and for a spheresphere system, the following equations were used: Lifshitz van der Waals energy: G LW
§ x 2 xy x Aª y y « 2 2 2 ln¨¨ 2 12 ¬ x xy x x xy x y © x xy x
§ ¨ ·º¨ 1 ¸¸» ¨ y ¹¼ § 2SH ¨ 1 1.77¨ © O ©
· ¸ ¸ ·¸ ¸¸ ¹¹
(5)
Electrostatic interaction energy:
G EL
º SHa1a 2 ] 1 ] 2 ª 2] 1] 2 1 e NH ln ln 1 e 2NH » « NH ] ] a1 a 2 1 e 2 ¬ 1 ¼
(6)
Acid-base interaction energy: GAB = ʌaȜǻGadhABe[(do-H)/ Ȝ]
(7)
where H is the separation distance, a1 and a2 are the radii of the interacting particles, ȗ1 and ȗ2 are the zeta-potentials of the interacting particles, ț is the double layer thickness-1, A is the Hamaker constant, d0 is the minimum separation distance between the two surfaces (0.157 nm), Ȝ is the correlation length of molecules in liquid (0.6 nm) and x = H/(a1 + a2), y = a1/a2 For interaction energy calculations, the only unknown parameter is the Hamaker constant. There are two different methods to evaluate the Hamaker constant that influence the value of
3
Lifshitz-van der Waals interaction energy (Van Oss et al. 1986 Van Oss et al. 1987). First method involves a microscopic approach, where the total interaction is assumed to be the sum of all interactions between the atom pairs. The second method involves a macroscopic approach, where the particles and the medium are considered as a continuous phases. In the presence of a liquid medium, the van der Waals energy between the particles differs and therefore, the Hamaker constant has to be replaced by an effective Hamaker constant. In case of two different particles (1 and 2) in a medium (3), the effective Hamaker constant A123 is calculated by: A123 = (A111/2 – A331/2)(A221/2 – A331/2)
(8)
For the calculation of individual Hamaker constants for solid surfaces, Fowkes (1964) proposed the following equation: A11 = 6ʌr2ȖsLW
(9)
where r represents the intermolecular distance and Ȗs of the surface energy of the solid.
LW
represents the dispersive component
Another possible way of obtaining the Hamaker constant is from the Lifshitz-van der Waals component of free energy of adhesion, ǻGadhLW, calculated from the contact angle measurements. In this instance, the effective Hamaker constant is given by: A =- 12 ʌd02 ǻGadhLW
(10)
Materials and methods Bacterial mass preparation At. ferrooxidans cells (type strain ATCC 23270) were grown in a simple Mackintosh medium (40 g/l FeSO4.7H2O, 0,132 g/l (NH4)2SO4, 53 mg MgCl2.6H2O, 27 mg KH2PO4, 0,147 g CaCl2.2H2O, 62 ȝg MnCl2.2H2O, 68 ȝg ZnCl2, 64 ȝg CoCl2.6H2O, 31 ȝg H3BO3, 10 ȝg Na2MoO4, 67 ȝg CuCl2.2H2O) and metal ion enriched Mackintosh media. The metal-ion enriched Mackintosh media contained 0.3 M Cu2+ (CuSO4.5H2O) and Zn2+ (ZnSO4.7H2O) ions. The bacteria were adapted to grow in the presence of the metals by serial subculturing in media containing successively increased concentrations of the metal ions. They were considered to have been adapted when their growth rate and maximum cell yield was comparable to that in simple Mackintosh medium. After reaching the stationary phase the cells mass was collected by Millipore filtration, washed with a solution of H2SO4 of pH 2 and centrifuged if necessary to obtain a concentrated cell suspension. Pure mineral samples of chalcopyrite and pyrite and sphalerite were crushed and ground in agate mortar and pestle. The particles of the ground minerals were sorted by wet sieving to obtain different size fractions. A portion of the smallest-sized fractions was further ground and a size fraction of –5 μm was obtained by sieving through a filter cloth with micron-sized pores in an ultrasonic bath. The surface area was estimated by BET to be 1.02 m2/g for pyrite, 1.9 m2/g for chalcopyrite and 1.75 m2/g for sphalerite.
4
Adhesion tests Adhesion tests were carried out using the smallest-sized mineral fraction prepared (-5 ȝm) because of its’ high surface area. The surface area of different minerals was estimated by BET. Mineral – liquid ratio was kept constant during all adsorption tests at 1 g/100 ml while the concentration of bacteria was changed. The adhesion characteristics were studied at 5x106, 107, 5x107, 108, 5x108, 109 and 5x109 cells/ml. At. ferrooxidans cells and At. ferrooxidans cells grown in the presence of 0.3 M zinc and 0.3M copper ions were studied simultaneously. In order not to compromise the viability of the bacteria we decided to run the experiments at an acidic pH by keeping the pH at pH 2 with H2SO4. After 30 min of agitation, we estimated the cells remaining suspended in the liquid phase by a direct microscopic count. Zeta potential measurements Zeta-potentials of washed and precipitate free cells of unadapted and adapted At. ferrooxidans and minerals were measured with ZetaCompact in 10-2M KNO3. The instrument measuring zeta-potentials was supplied with Zeta4 software which allows direct reading of zetapotentials calculated from the mobility using the Smoluchovski equation. A cell concentration of 2.5x107 cells/ml was used, which was previously found to be suitable for the instrument. Pyrite, chalcopyrite and sphalerite of -5 ȝm particle size at a concentration of 0.025 g/100 ml was used. The pH was adjusted with HNO3 and KOH to different values. The measurement itself was realized after 30 minutes of conditioning at a particular pH and the pH was once more measured directly before the measurement and considered as the pH of the measurement. Contact angle measurements Washed bacterial cells were collected by filtration through a Millipore membrane having a pore size of 0.22 μm through which the cells could not pass, in order to obtain a uniform layer of cells on the surface of the membrane. The filter paper with bacterial lawns was then mounted on a glass slide with the help of double sided tape and left to dry at room temperature. To obtain a stable and reproducible contact angle of bacteria, a cell lawn composed of more than 200 layers was used. The contact angle is estimated by the sessile drop method; a drop of test liquid is placed on the tested surface and the contact angle between the tested surface and drop is measured. For this purpose KrĦss Easy Drop equipment and Drop shape analysis software was used. The instrument is equipped with a CCD camera. We determined a stabilized contact angle with it. Each experiment was repeated 3 times and the arithmetic mean value was considered as the final contact angle for a particular test liquid. The contact angle on solid powders was determined by sorption measurements using KrĦss Tensiometer K100 and KrĦss LabDesk 3.1 software. The Washburn equation was used to measure the contact angle on powder samples. When powder contained in a column is in contact with liquid, the pores between the particles act like small capillaries and the rise of liquid is measurable. The capillary constant of the Washburn equation was determined using n-Hexane (low energy wetting liquid) that wets the solids completely. Using the obtained capillary constant, the contact angles for other liquids were determined. Mineral samples of 106+38 μm size fraction were used. A 1 g sample of mineral was placed into a glass sample tube and was carefully and uniformly packed each time. Each measurement was repeated at
5
least 3 times. The results were reproducible within ±3 degrees deviation and reported as a mean value. For the calculation of surface energy and its components the van Oss acid-base method was used. For this calculation, contact angle data using 3 different liquids are required. In this study 4 well defined liquids with known components were used: water, di-iodomethane, 1bromonaphtalene and formamide. Surface tensions and the dispersive and polar part contributions, and polar part divided into acid-base components as reported in the literature were used (Bellon-Fontaine et al. 1990; Ström et al. 1987). Results and Discussions Adhesion studies In Fig. 1 the adsorption density of iron grown cells on different minerals as a function of equilibrium concentration is presented. Complete surface saturation was not reached within the tested concentration range of the cells, but above 108 cells/ml concentrations there was a visible levelling off the curves indicating the proximity of a saturation concentration. The curves show a clear difference in adsorption preferences of the cells to minerals. The adsorption density of cells on pyrite was the highest followed by chalcopyrite and sphalerite. Comparing the coverage of the surface of the different minerals by adhered cells, assuming their adherence to pyrite was 100%, chalcopyrite was covered by 80.6% and sphalerite by 47.2% at equilibrium.
Fig. 1. Unadapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite. More Zn-adapted cells adhered to pyrite and chalcopyrite than sphalerite before saturating the surfaces of these minerals (Fig. 2). The isotherm for sphalerite indicates close to a complete surface saturation at and above 108 cell/ml concentration. The selective and higher adhesion of unadapted cells to pyrite completely disappeared with the cells after adaptation to zinc ions. The zinc adapted cells adhered to chalcopyrite to the same extent as to pyrite, but their adsorption onto sphalerite decreased to 41.3% compared to pyrite and chalcopyrite. At. ferrooxidans grown in the presence of copper ions adhered the least to all three minerals used in this study (Fig. 3). The cells almost reached surface saturation above 107 cell/ml
6
equilibrium concentration. The cell adhesion to chalcopyrite and pyrite was comparable at higher cell concentrations. Considering this adsorption density as surface saturation, the adhesion of the cells on sphalerite corresponded to 47.5% surface coverage.
Fig. 2. Zinc adapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite. The results showed that cells grown in the presence of copper adhered to the minerals to a lesser extent than cells grown in its absence. Most extensive adhesion to pyrite and chalcopyrite was achieved with zinc adapted cells but led to a loss of selectivity towards pyrite. Unadapted cells of At. ferrooxidans were the most effective towards sphalerite but the adsorption density was still the lowest among all the tested minerals.
Fig. 3. Copper adapted At. ferrooxidans cells adhesion isotherm for pyrite, chalcopyrite and sphalerite.
Zeta-potential studies The zeta-potential of iron grown, copper and zinc adapted At. ferrooxidans cells is shown in Fig. 4. Unadapted At. ferrooxidans cells were negatively charged in the entire pH range studied without displaying an iso-electric point. An extrapolation of the curve shows that the 7
Zeta-potential (mV)
isoelectric point is probably located at or below pH 1. Cells after adaptation to copper and zinc metal ions are positively charged below pH 2.2 and pH 3 respectively. The iso-electric point below pH 3 indicates an anionic polysaccharide surface. Only the zinc grown cells had a higher iso-electric point which is typical for peptidoglycan surface but still too low to have proteinaceous surface (Rijnaarts et al. 1995). The magnitude of electronegativity was less for zinc grown cells. A shift in iso-electric point and decrease in surface charge of bacteria could be due to the adsorption of specific ions, in this case the metal cations, and secondly a change in surface chemical composition due to secretion of different compounds caused by the varied growth environment. 20
Fe grown
15
Zn grown Cu grown
10 5 0 -5 0
2
4
6
8
10
12
-10 -15 -20 pH
Fig. 4. Zeta-potential of different At. ferrooxidans cells at 0.01M ionic strength as a function of pH. The zeta-potentials of the minerals and cells at pH 2 corresponding to experimental conditions for cell growth and adhesion to minerals are presented in Table 1. These values were used to assess bacterial adhesion behaviour by DLVO theoretical analysis. Iron grown At. ferrooxidans and sphalerite were negatively charged while zinc and copper grown cells as well as pyrite and chalcopyrite were positively charged. Taking into account only the electrostatic forces arising from different surface charges, unadapted cells should adhere to pyrite and chalcopyrite but not to sphalerite, because sphalerite like the unadapted cells is negatively charged. Metal ions adapted cells should prefer sphalerite over pyrite and chalcopyrite because the latter are positively charged similar to cells at pH 2. This approach is greatly simplified since several interacting forces are involved in bacterial adhesion. These have been incorporated in extended DLVO theory. Table 1.The zeta-potentials of minerals and At. ferrooxidans cells at pH 2. Material
Zeta-potential (mV)
At. ferrooxidans At. ferrooxidans Zn grown At. ferrooxidans Cu grown Pyrite Chalcopyrite Sphalerite
-4.98 15.22 1.81 13.8 9.13 -35
8
Contact angle and surface energy studies In Table 2 the measured contact angles with polar and non-polar test liquids, and the surface energies of minerals and At. ferrooxidans cells calculated from contact angle data are presented. All minerals showed a relatively high contact angle with water and lower polar surface energy component relative to the total surface energy, demonstrating a hydrophobic character of the mineral surfaces. The polar part of surface energy and total surface energy decreased in the direction pyrite > chalcopyrite > sphalerite, correlating with the amount of hydrophobic sulphur species on the surface of these minerals. The results also show that the electron donating character of the minerals is more predominant. Table 2. Contact angle and surface energy of At. ferrooxidans grown under different conditions, and of sulphide minerals.
Material water Pyrite 61.00 Chalcopyrite 71.76 Sphalerite 68.70 At.f. Fe grown 33.00 At.f. Zn grown 26.57 At.f. Cu grown 7.50
Contact angle ș DiiodoBromomethane naphtalene --10.40 --39.66 --30.88 53.50 36.00 --40.00 --35.55
Surface energy Ȗ (mJ.m-2) formamide 8.15 43.46 53.40 32.50 29.73 14.80
total 53.75 42.39 40.64 48.55 49.54 55.36
Ȗd/ȖLW 43.87 34.93 38.50 34.31 34.78 36.67
ȖP/ȖAB 9.88 7.47 2.13 14.24 14.76 18.68
Ȗ+ 4.01 2.20 0.08 1.11 1.06 1.55
Ȗ6.08 6.34 14.51 45.57 51.46 56.43
The water contact angle on At. ferrooxidans cells after metal ion adaptation decreased for both Cu and Zn ions, indicating that the adapted cell surface became more hydrophilic. The total surface energy of cells increased in the order of unadapted cells zinc adapted copper adapted. The increase in the polar surface energy component of the cells followed in the same order. Thus the surface of At. ferrooxidans became more polar with metal ion adaptation. For both cases of metal ion adaptation, the electron donating character was enhanced but the electron accepting character decreased for zinc adapted cells and increased for copper adapted cells. The presence of metal ions in the growth medium has probably changed the cell’s surface chemical composition due to enhanced secretion of metabolites rich in –OH and – COOH groups contributing to the enhanced polar surface structure and negative electron donating character. Thermodynamic analysis of cells adhesion In Table 3 the calculated free energy of adhesion for At. ferrooxidans – mineral systems in aqueous media is presented. The results show that the adhesion is favoured for all cells onto pyrite and chalcopyrite but not on sphalerite. On thermodynamic grounds, the adhesion of cells on sphalerite is energetically not favourable. For all minerals, the ǻGtot follows the order of unadapted cells zinc adapted copper adapted, making the adhesion most apparent for the unadapted cells of At. ferrooxidans. The values dispersive component ǻGLW follow a reverse order because of the increased dispersive contribution by the bacteria to the system after their cultivation in the presence of metal ions. Metal ions adaptation also induced more polarity to cells surface with predominant electron donating character and thus a decrease in the likelihood of cells adhesion (ǻGAB value increased) is expected. The free energy of adhesion values illustrate that both unadapted and adapted cells favour adhesion onto pyrite (lowest ǻGtot) followed by chalcopyrite and sphalerite. This is due to an increase in dispersive
9
free energy between mineral and bacteria when the cells surface becomes more polar in character. Table 3. Free enegy of adhesion for different At. ferrooxidans – mineral systems. Free energy of adhesion ǻGadh (mJ.m-2)
Mineral – Bacteria System Pyrite - At. ferrooxidans Pyrite - Cu grown At. ferrooxidans Pyrite - Zn grown At. ferrooxidans Chalcopyrite - At. ferrooxidans Chalcopyrite - Cu grown At. ferrooxidans Chalcopyrite - Zn grown At. ferrooxidans Sphalerite - At. ferrooxidans Sphalerite - Cu grown At. ferrooxidans Sphalerite - Zn grown At. ferrooxidans
ǻGadhLW -4.65 -5.42
ǻGadhAB -10.29 -4.66
ǻGadhtotal -14.93 -10.08
-4.80 -2.95 -3.44 -3.05 -3.65 -4.26 -3.77
-7.83 -8.10 -1.70 -5.21 6.30 14.03 10.27
-12.63 -12.11 -5.14 -8.26 2.65 9.78 6.50
Extended DLVO theory of cells adhesion Potential interaction energy between minerals and cells was calculated following the extended DLVO theory. For this calculation, the Hamaker constant for computing van der Waals force was needed. We calculated it using microscopic and macroscopic approaches. The values obtained from both methods were comparable but microscopic approach generally gave lower value within the same order of magnitude (Table 4). Although the results were consistent in both methods, the values obtained by macroscopic approach were used in the construction of DLVO interaction energy curves as a function of separation distance. Table 4. Hamaker constants calculated by microscopic (Method 1) and macroscopic (Method 2) method for different At. ferrooxidans – mineral systems.
Mineral – Bacteria System Pyrite - At.f. Fe grown Pyrite - At.f. Cu grown Pyrite - At.f. Zn grown Chalcopyrite - At.f. Fe grown Chalcopyrite - At.f. Cu grown Chalcopyrite - At.f. Zn grown Sphalerite - At.f. Fe grown Sphalerite - At.f. Cu grown Sphalerite - At.f. Zn grown
A11
A22
49.40 52.80 50.10 49.40 52.80 50.10 49.40 52.80 50.10
63.20 63.20 63.20 50.30 50.30 50.30 55.40 55.40 55.40
A33 [10-21 J]
37.00 (Evans and Wennerström 1994)
A (method 1) A (method 2)
1.77 2.21 1.85 0.96 1.20 1.00 1.29 1.61 1.36
4.32 5.04 4.46 2.74 3.20 2.83 3.39 3.96 3.51
In addition to LW dispersive forces, the electrostatic forces are decisive for adhesion and are dependent on the surface charge developed under experimental conditions such as pH and ionic strength. The adhesion behaviour was investigated at pH 2, which is the optimum for cell growth. Accordingly the zeta-potential at this pH value was used in the calculations. Adhesion of unadapted cells to mineral surfaces was most favored in the presence of pyrite because of a strong attractive acid-base interaction and a strong attraction between oppositely charged surfaces. The total interaction energy was attractive towards chalcopyrite but weaker 10
compared with pyrite (Fig. 5). Lower polarity of the chalcopyrite surface compared to that of pyrite decreased the attractive acid-base forces and also minimised the electrostatic force due to reduced surface charge. Interaction energy between sphalerite and cells was positive below 4 nm (40 Å) separation distances which impaired adhesion or at least decreased it. The repulsive force stemmed from the repulsive acid-base interaction force and negatively charged particle and cell surface. The attractive LW forces were too weak to affect the repulsive forces. Zinc grown cells are positively charged and also the minerals at pH 2 inducing repulsive electrostatic forces. The total interaction energy was, however, attractive with respect to pyrite and chalcopyrite because of LW and acid-base forces (Fig. 6). The cell adhesion on these minerals (Fig. 2) is in agreement with the interaction energy curves. Sphalerite is negatively charged at pH 2 and the electrostatic force favours cells adhesion, but below 1.5 nm (15 Å) distance the total interaction became repulsive due to a strong acid-base repulsive force. The positive surface charge of copper grown cells was very weak and the cells electrostatic repulsive force had no significant influence on the total interaction energy for pyrite and chalcopyrite (Fig. 7). Acid-base and LW interactions became decisive for cells adhesion on these minerals and both were attractive. The total interaction force was stronger for pyrite because the 'GadhAB value was greater. Cell adhesion on sphalerite in our experiments was expected to be very weak because repulsive forces arose at greater distances from the mineral. The attractive forces resulting from the negative charge on sphalerite were weaker than those of Zn-grown cells. Thus, the extent of adhesion of Zn-grown cells to sphalerite was expected to be slight. The adhesion forces for the copper grown cells are lower according to the adhesion results (Fig. 3) and the zinc grown cells should maintain the same magnitude as unadapted cells. From the potential energy diagrams it is clear that the strongest total interactions between the minerals and cells are for the unadapted and zinc grown cells and weaker interactions for copper grown bacteria. However the increased adhesion of copper grown cells on chalcopyrite is not reflected in the DLVO calculations. The electrostatic repulsive forces are too weak to have any effect compared to more charged pyrite. The electric double layer interactions in bacterial adhesion need reassessment. In calculations the assumption of bacterial cells as rigid particles is adopted, however the cells could be hardly considered as rigid. Bacteria surface composed of different polymers or surface structures such as fimbriae or fibrils. These layers are relatively thick (50 nm) and can be considered as ion penetrable and a soft particle model (Ohshima 1994 Gaboriaud et al. 2008) is more applicable. If the soft particle model is used instead of Smoluchowski, the result will be a decrease of surface potential. However at lower surface potential as in case of copper grown cells this decrease is negligible (Chandraprabha 2009). Depending on the model (constant surface potentials or constant fixed charge densities) the electrostatic interaction energies may increase (Poortinga et al 2002). According the adhesion experiments carried out with copper grown cells, the electrostatic forces had a greater influence on the total energetic balance than predicted by the DLVO theory.
11
500 Pyrite
400 300 200 G (kT)
100 0
-100 0
20
40
60
80
100
80
100
80
100
-200 -300 -400 -500
500 Chalcopyrite
400 300 200 G (kT)
100 0
-100 0
20
40
60
-200 -300 -400 -500
500 Sphalerite
400 300 200 G (kT)
100 0
-100 0
20
40
60
-200
Lifshitz - van der Waals
-300
Electrostatic
-400 -500
Acid-Base Total
Separation distance (Å)
Fig. 5. Ferrous grown At. ferrooxidans DLVO energy diagrams with sulphide minerals.
12
500 Pyrite
400 300 200 G (kT)
100 0
-100 0
20
40
60
80
100
80
100
80
100
-200 -300 -400 -500
500 Chalcopyrite
400 300 200 G (kT)
100
0 -100 0
20
40
60
-200 -300 -400 -500
500 Sphalerite
400 300 200 G (kT)
100
0 -100 0
20
40
60
-200
Lifshitz - van der Waals
-300
Electrostatic
-400 -500
Acid-Base Total
Separation distance (Å)
Fig. 6. Zinc grown At. ferrooxidans DLVO energy diagrams with sulphide minerals.
13
500 Pyrite
400 300 200 G (kT)
100 0
-100 0
20
40
60
80
100
80
100
80
100
-200 -300 -400 -500
500 Chalcopyrite
400 300 200 G (kT)
100 0
-100 0
20
40
60
-200 -300 -400 -500
500 Sphalerite
400 300 200 G (kT)
100 0
-100 0
20
40
60
-200
Lifshitz - van der Waals
-300
Electrostatic
-400 -500
Acid-Base Total
Separation distance (Å)
Fig. 7. Copper grown At. ferrooxidans DLVO energy diagrams with sulphide minerals.
14
In reality, the bacterial surface is very heterogeneous in terms of chemical composition and topography which provoke different interactive forces at different areas of the surface (Fang et al. 2000). The presence of surface structures (fimbriae, fibrils) may also reduce the electrostatic forces by decreasing the particle size (Elimelech and O Melia, 1990). Other forces influencing the adhesion are steric interactions from polymers present on the bacterial surface (Camesano and Logan 2000). At. ferrooxidans cells are characterized as rod shaped with a polar flagellum and mostly visualised with a smooth surface lacking surface structures such as fimbriae (Ehrlich 1990 Mangold et al. 2008) free from all suppositions. Using different approaches may increase the electrostatic repulsive forces and their effect for the copper grown cells and decreases the interaction energy difference between pyrite and chalcopyrite systems. Conclusions Predicting the adhesion of bacteria onto minerals is a difficult task without considering all the physical and chemical interacting forces. Surface charge is often considered as the main driving force towards adhesion and the magnitude of charge is found to play a critical role. Based on charge consideration the adhesion of At. ferrooxidans cells on sphalerite should have been greater but was the least among the tested materials. The adhesion also occurred even when both the interacting surfaces possessed the same charge polarity. The thermodynamic approach clearly highlighted the decreased adherence of cells towards sphalerite but failed to explain small changes in adhesion behaviour of the pyrite and chalcopyrite systems. Extended DLVO theory incorporating LW and electrostatic forces, and acid-base forces were found to be well-suited to elucidating the cell adhesion to minerals. This theory explained the experimental adhesion results from this study to a large extent and predicted the increased cell adhesion on chalcopyrite after zinc adaptation of the cells. Acknowledgements Financial support from Kempestiftelsen foundation in the form of scholarship is gratefully acknowledged. References Absolom DR, Lamberti FV, Policova Z, Zingg W, van Oss CJ, Neumann AW (1983) Surface thermodynamics of bacterial adhesion. Appl Environ Microbiol 46:90-97 Botero AEC, Torem ML, Mesquita LMS (2007) Fundamental studies of Rhodococcus opacus as a biocollector of calcite and magnesite. Miner Eng 20:1026-1032 Busscher HJ, Weerkamp AH, van der Mei HC, van Pelt A W, de Jong HP, Arends J (1984) Measurement of the surface free energy of bacterial cell surfaces and its relevance for adhesion. Appl Environ Microbiol 48:980-983 Camesano TA, Logan BE (2000) Probing Bacterial Electrosteric Interactions Using atomic Force Microscopy. Environ Sci Technol 34:3354-3362 Chandraprabha MN, Modak JM, Natarajan KA (2009) Soft-particle model analysis of effect of LPS on electrophoretic softness of Acidithiobacillus ferrooxidans grown in the presence of different metal ions. Colloids Surf B Biointerfaces 69:1-7
15
Chen M-L, Zhang L, Gu G-H, Hu Y-H, Su L-J (2008) Effects of microorganisms on surface properties of chalcopyrite and bioleaching. Trans Nonferrous Met Soc China 18:1421-1426 Deo N, Natarajan KA (1998) Studies on interaction of Paenibacillus polymyxa with iron ore minerals in relation to beneficiation. Int J Miner Process 55:41-60 Deryagin BV, Landau L (1941) Acta Physicochem 55:333 Elimelech M, and O’Melia CR (1999) Effect of Particle Size on Collision Efficiency in the Deposition of Brownian Particles with Electrostatic Energy Barriers. Langmuir 6:1153-1163 Evans DF, Wennerström H (1994) The Colloidal Domain Where Physics, Chemistry, Biology and Technology Meet. Wiley-VCH, New York Fang HHP, Chan K-Y, Xu L-C, J (2000) Quantification of bacterial adhesion using atomic force microscopy. Microbiol Meth 40:89-97 Fowkes FM (1964) Attractive forces at interfaces. Ind Eng Chem 56:40–52 Gaboriaud F, Gee ML, Strugnell R, Duval JFL (2008) Coupled electrostatic, hydrodynamic and mechanical properties of bacterial interfaces in aqueous media. Langmuir 24:1098810995 Ehrlich HL (1990) Geomicrobiology, 2nd edn. Marcel Dekker, New York Harneit K, Goksel A, Kock D, Klock J-H, Gehrke T, Sand W (2006) Adhesion to metal surfaces by cells of Acidithiobacillus ferrooxidans, Acidithiobacillus thiooxidans and Leptospirillum ferrooxidans. Hydrometallurgy 83:245-254 Mangold S, Laxander M, Harneit K, Rohwerder T, Claus G, Sand W (2008) Visualization of Acidithiobacillus ferrooxidans biofilms on pyrite by atomic force and epifluorescence microscopy under various experimental conditions. Hydrometallurgy 94:127-132 Mishra M, Bukka K, Chen S (1996) The effect of growth medium of Thiobacillus ferrooxidans on pyrite flotation. Miner Eng 9:157-168 Phalguni A, Modak JM, Natarajan KA (1996) Biobeneficiation of bauxite using Bacillus polymyxa, calcium and iron removal. Int J Miner Process 48:51-60 Poortinga AT, Bos R, Norde W, Busscher HJ (2002) Electric double layer interactions in bacterial adhesion to surfaces. Surf sci reports 47:1-32 Rohwerder T, Gehrke T, Kinzler K, Sand W (2003) Bioleaching review part A: Progress in bioleaching: fundamentals and mechanisms of bacterial metal sulfide oxidation. Appl Microbiol Biotechnol 63:239-248 Sanhueza A, Ferrer IJ, Vargas T, Amils R, Sánchez C (1999) Attachment of Thiobacillus ferrooxidans on synthetic pyrite of varying structural and electronic properties. Hydrometallurgy 51:115-129 Sharma PK, Hanumantha Rao K, Forssberg KSE, Natarajan KA (2001) Surface chemical characterisation of Paenibacillus polymyxa before and after adaptation to sulphide minerals. Int J Miner Process 62:3-25 Sharma P, Hanumantha Rao K (2003) Adhesion of Paenibacillus polymyxa on chalcopyrite and pyrite: surface thermodynamics and extended DLVO theory. Colloids Surf B Biointerfaces 29:21-38
16
Solari JA, Huerta G, Escobar B, Vargas T, Badilla-Ohlbaum R, Rubio J (1992) Interfacial phenomena affecting the adhesion of Thiobacillus ferrooxidans to sulphide mineral surfaces. Colloids Surf 69:159-166 Van Oss CJ, Good RJ, Chaudhury MK (1986) The role of van der Waals forces and hydrogen bonds in “hydrophobic interactions” between biopolymers and low energy surfaces. J Colloid Interf Sci 111:378-390 Van Oss CJ, Chaudhury MK, Good RJ (1987) Monopolar surfaces. Adv Colloid Interf Sci 28:35-64 Verwey EJ, Overbeek JTG (1955) Theory of the stability of lyophobic colloids. J Colloid Sci 10:224-225 Vijayalakshmi SP, Raichur AM (2002) Bioflocculation of high-ash Indian coals using Paenibacillus polymyxa. Int J Miner Process 67:199-210 Xia L, Liu X, Zeng J, Yin C, Gao J, Liu J, Qiu G (2008) Mechanism of enhanced bioleaching efficiency of Acidithiobacillus ferrooxidans after adaptation to chalcopyrite. Hydrometallurgy 92:95-101 Yelloji Rao KM, Natarajan KA, Somasundaran P (1997) Growth and attachment of Thiobacillus ferrooxidans dirung sulphide mineral leaching. Int J Miner Process 50:203-210 Zheng X, Arps PJ, Smith RW (2001) Adhesion of two bacteria onto dolomite and apatite: their effect on dolomite depression in anionic flotation. Int J Miner Process 62:159-172
17
Paper IV Surface Thermodynamics and Extended DLVO Theory of Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite. Vilinska, A., Hanumantha Rao, K., 2009. The Open Colloid Science Journal. 2, pp. 1-14.
The Open Colloid Science Journal, 2009, 2, 1-14
1
Open Access
Surface Thermodynamics and Extended DLVO Theory of Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite A. Vilinska and K. Hanumantha Rao* Division of Mineral Processing, Luleå University of Technology, SE-971 87 LULEÅ, Sweden Abstract: The adhesion of Acidithiobacillus ferrooxidans bacterial cells have been assessed by following the thermodynamic and extended DLVO theoretical approaches. Surface potential, interfacial tension and contact angle parameters that are necessary for the calculation of free energy of adhesion have been determined experimentally. The Hamaker constant involved in the Lifshitz-van der Waals interaction energy has been estimated by microscopic and macroscopic methods. The free energy of adhesion found to be negative on pyrite and chalcopyrite minerals indicating the adsorption of bacterial cells on these minerals. The potential energy diagrams of total interaction energy versus separation distance curves also illustrate the feasibility of bacterial cells adhesion except in a pH region where the bacterial species and minerals possess similar surface charge with high magnitude. The present theoretical analysis of bacterial adhesion on mineral surfaces found to be in good agreement with the experimental results and previous findings in the literature. Thus, the bacterial adhesion behavior on minerals can be judged and explained by considering the physico-chemical interaction forces.
Keywords: Acidithiobacillus ferrooxidans, sulfide mineral, adhesion, surface thermodynamics, extended DLVO theory. 1. INTRODUCTION Conventionally, physico-chemical methods are used in mineral processing, but now-a-days, biological processing routes are sought to solve the problems associated with lean grade ores and where the traditional methods fail to separate the minerals from complex ores. Acidophilic sulfur or iron oxidizing bacteria are used in bio-hydrometallurgical operations of leaching and in biobeneficiation processes such as bioflotation and bioflocculation [1]. Bio-leaching processes of the extraction of metals from sulfide and oxide minerals have been developed during the years and have been adopted by certain industries. Compared to bio-leaching processes, bio-beneficiation processes are relatively new and are under intense investigation in recent years and most of the studies are so far confined to laboratory [2]. Since the bacteria adhere to mineral surface within few minutes and alter the surface properties that are essential in mineral beneficiation techniques, the microorganisms have formidable applications in flotation and flocculation processes. The adhesion of bacterial cells on mineral surface is essentially one of the most important aspects determining the success of either bioleaching or biobeneficiation processes. Selective bacterial adhesion on minerals is a crucial factor for selectively modifying the surfaces relevant to flotation and flocculation processes. Microorganisms have a tremendous influence on their environment through the transfer of energy, charge, and materials across a complex mineral-solution interface. Natural phenomena driven by the bacteria-mineral interactions are incredibly diverse, including major environment and geochemical processes. Much of the impetus to study the *Address correspondence to this author at the Division of Mineral Processing, Dept of Chemical Engineering and Geosciences, Luleå University of Technology, Luleå, Sweden; Tel: +46-920 491705; Fax: +46- 920 97364; E-mail:
[email protected],
[email protected] 1876-5300/09
mineral-bacteria interactions arises from the expected impact on many technological areas, including protection against bacterial infection and biofouling [3], and minerals bioprocessing comprising bioremediation of organic and inorganic contaminants [4], bio-leaching and bio-beneficiation processes [5]. Common to many of these processes, the adhesion of bacteria on mineral surfaces occurs and reacts with mineral-water interface. Some types of bacteria utilize minerals as the terminal electron acceptor in the respiratory cycle and others recover energy from minerals by the enzymatic oxidation. However, the specific mechanisms of adhesion and charge transfer reactions remain a subject of debate [68]. The charge transfer mechanisms can be direct involving the adhesion of bacteria to the surface of sulfide minerals inducing oxidative/reductive dissolution [9], indirect mechanism which is mediated by ferric ions [10] and cooperative mechanism including direct and indirect mechanisms [8]. Previous studies have shown that under most physiological conditions the bacterial cell surface carries a net negative charge [1, 11], while, along with electrostatic force, hydrophobic, entropic, acid-base, and van der Waals interactions and H-bonding are important in bacterial adhesion on mineral surfaces [12, 13]. The cells adhesion is generally a function of biological properties (chemotaxis, specific surface site adhesion) [14] besides physico-chemical forces between the interacting phases. Taking into account only the physico-chemical forces, there are two main approaches to judge or predict the bacterial adhesion on mineral surfaces [15]:
Thermodynamic approach
Extended DLVO approach
1.1. Thermodynamic Approach According to thermodynamics, a system spontaneously undergoes a change from state A to state B if the change is 2009 Bentham Open
2 The Open Colloid Science Journal, 2009, Volume 2
Vilinska and Rao
advancing towards a lower energy state. In case of mineralbacteria adhesion, allow that state A represents mineral particles and cells dispersed in liquid, while state B represents a new phase of bacteria adsorbed on mineral. Then the change from a dispersed to adhered state will be spontaneous only if the energetic state of the adhered system B is lower. Mathematically, the free energy of adhesion Gadh can be expressed by Gadh = bs – bl – sl
(1)
where represents interfacial free energies for different interfaces: bacteria-solid (bs), bacteria-liquid (bl) and solidliquid (sl) [16, 17]. Adhesion is energetically favoured only if Gadh is negative. For calculation of Gadh, the free energy of adhesion was divided into two parts: Lifshitz van der Waals component and acid-base component. Gadh = GadhAB + GadhLW
(2)
Calculation according to LW-AB approach using the Lifshitz-van der Waals dispersive and acid-base components of surface energies is:
Gadh and
LW
= 2
( 2( 2(
(
bv
LW
lv
)( )( )(
Gadh AB = 2 bv + sv + + bv + sv
lv
+
lv
+
bv
bv
sv
)(
LW
sv lv lv
) )
sv
LW
lv
LW
)
(3)
)
Classical DLVO approach [18, 19] includes Lifshitz van der Waals (LW) interactions, and electrostatic interactions. LW forces are always attractive and strong at shorter distances between neutral stable molecules. Coulombic electrostatic interaction forces could be attractive or repulsive depending on the surface charge of interacting particles. Acidbase interactions were added later by van Oss [20] to involve the electron donating-accepting abilities of different materials. Thus microbial adhesion to solid surfaces can be described by a sum of van der Waals, electrostatic and acidbase forces as a function of separation distance. (5)
Calculation of these forces depends on the geometry of interacting phases and for sphere-sphere system, the following equations were used:
G LW
y y
x 2 + xy + x + x 2 + xy + x + y 1 A = (6) x 2 + xy + x
12 2 H 1 + 1.77 +2 ln 2 x + xy + x + y
Electrostatic interaction energy:
G EL =
a1a2 (1 + 2 ) 21 2 1 + e H 2 H )
+ ln 1 e H + ln (1 e a1 + a2 1 2
For interaction energy calculation, the parameters such as zeta potential, particle size radius, double layer thickness are known or measurable, while for calculation of Hamaker constant there are two different methods available. The Hamaker constant value that influences the Lifshitz-van der Waals interaction energy could be obtained by microscopic and macroscopic approaches [21, 22]. According to the first microscopic approach, the total interaction is assumed to be the sum of all interactions between the atom pairs. In the second macroscopic approach, the particles and the medium are considered as a continuous phases. In the presence of a liquid medium, the van der Waals energy between the particles is different, and therefore the Hamaker constant must be replaced by an effective Hamaker constant. In case of two different particles 1 and 2 dispersed in a medium 3, the effective Hamaker constant is calculated by:
(7)
(9)
A12 is regarded as a geometric mean of A11 and A22: A12 = (A11.A22)1/2
(10)
Then, A123 = (A111/2 – A331/2)(A221/2 – A331/2)
(11)
For calculation of individual Hamaker constants for solid surfaces, Fowkes [22] proposed the following equation: A11 = 6r2sD
(12)
where r represents the intermolecular distance and sD represents the dispersive component of the surface energy of the solid. According to Fowkes, for water and systems with the volume which have nearly the same size, the value of 6r2 is equal to 1.44x10-18 m2. From the calculation of individual Hamaker constants using the known values of dispersive component of surface energies, the effective Hamaker constant can be evaluated. Another possible way of obtaining Hamaker constant is using Lifshitz-van der Waals component of free energy of adhesion GadhLW, which is obtained from the contact angle measurements. The effective Hamaker constant is then given by [22]: A = 12 d02 GadhLW
Lifshitz van der Waals interaction energy:
(8)
where H – separation distance, a – radius of solid particle, – zeta-potential, – double layer thickness-1, A – Hamaker constant, d0 – minimum separation distance between 2 surfaces (0.157 nm), – correlation length of molecules in liquid (0.6 nm) and x = H/(a1 + a2), y = a1/a2
A123 = A12 + A33 – A13 – A23 (4)
1.2. Extended DLVO Approach
Gtotal = GLW + GEL + GAB
Acid-base interaction energy: GAB = aGadhABe[(do-H)/ ]
(13)
2. MATERIALS AND METHODOLOGY 2.1. Mineral Samples Preparation Pure mineral samples of chalcopyrite and pyrite were obtained from Gregory, Bottley & Lloyd, UK. The samples were crushed and ground in agate mill. After grinding, the minerals were sorted by wet sieving to obtain suitable coarse size fraction for flotation and contact angle measurements (– 106+38 m) and a fine fraction of minus 38 m. A portion of minus 38 m was further ground and a size fraction of –5 m was obtained by a micron filter cloth sieving in ultra-
Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite
sonic bath, which was used in zeta potential and adsorption measurements. The minerals were acid washed with HCl to clean the surface from the oxidation products and with ethanol to clean the surface from any organic compounds. The samples were dried and stored at -10ºC until use. The surface area was measured for both size fractions with BET method (Flowsorb II 2300) and determined to be 0.06 and 1.02 m2/g for coarse and fine fractions of pyrite and, 0.17 and 1.9 m2/g for coarse and fine size fractions of chalcopyrite, respectively. 2.2. Bacterial Cultivation Bacterial cells of Acidithiobacillus ferrooxidans were cultured in 9K medium with the following chemical composition: 44.5 g/l of FeSO4.7H2O, 3 g/l (NH4)2SO4, 0.5 g/l MgSO4.7H2O, 0.5 g/l K2HPO4 and 0.1 g/l KCl. The medium pH was maintained with H2SO4 at pH 2. All solutions were made with deionised (distilled) water and sterilized in autoclave at 125°C for 15 minutes and iron sulphate solution was prepared separately and filtered through Millipore filter, to remove all possible particles and cells. Sterilized Erlenmeyer flasks were filled with 200 ml of medium solution and inoculated with 20 ml of bacterial solution. The flasks were continuously shaken in an orbital shaker at 150 rpm and at 30°C. At different time intervals, the pH, Fe+2 content and number of cells per 1 ml of medium solution were measured for estimating and defining the bacterial growth in the solution. The Fe+2 ions were determined by volumetric titration method and the cell number was counted using a microscope in Neubauer counting chamber with a defined volume of each square. The bacterial cells used were at the end of exponential growth phase. A 20 ml of the solution is used as an inoculum for the next culture and the rest is filtered through Whatman filter paper to remove all the precipitates and finally through Millipore filter. The bacterial cell mass was washed with pH 2 water to remove all the trapped metal ions and metabolites. The cell mass so collected was used in the investigations. 2.3. Adsorption Measurements The -5 Dm size fractions of minerals were used for adsorption studies. Tests were performed at pH 2, at temperature 23ºC, at constant solid-liquid ratio (1:100) and at varied initial cell concentrations of 107, 5x107, 108, 5x108, 109 cells/ml. Cells and minerals were interacted together for 30 minutes by conditioning the suspension using a magnetic stirrer. After conditioning, the suspension was filtered to separate the solids and the bacterial cells remaining in the liquid phase were estimated by direct counting method. 2.4. Zeta Potential Measurements Zeta potential measurements were made using ZetaCompact equipped with video and Zeta4 software. The software allows the direct reading of zeta-potential calculated from the electrophoretic mobilites using Smoluchowski equation. The result is a particle distribution diagram, from which the mean mobilities are recalculated to zeta-potential values. Pyrite and chalcopyrite of -5 m particle size at a concentration of 0.025 g/100 ml was used. Ionic strength of 10-2 M was maintained with KNO3. The solution pH was adjusted using HNO3 and KOH. Solutions with a specified pH and
The Open Colloid Science Journal, 2009, Volume 2
3
constant ionic strength were prepared and then the mineral was added. After 30 minutes conditioning, the pH of the suspension was recorded again and regarded as the pH of the measurement. Zeta potential measurements of bacterial cells were performed similarly. Initially, A. ferrooxidans collected cell mass was washed with deionised water solution of pH 2 adjusted with H2SO4. Different bacterial concentrations were attempted and a cell population of 2.5x107cells/ml found to be suitable for the instrument and accordingly used. 2.5. Contact Angle Measurements Bacteria Precipitate-free cell solution was filtered through millipore filter paper using vacuum filtration to obtain a uniformly distributed cell layer on the whole area of filter paper. When filtered, the filter paper was placed on a Whatman filter paper for a short time to remove the excess moisture. After moisture removal, the filter paper is dry enough to mount on a microscope slide glass with a help of double sided adhesive tape. The samples were air-dried to remove the rest of water moisture and contact angles were measured while placing a drop of liquid on the bacterial lawn surface. The dynamic contact angle was recorded with Krss Easy drop equipment and evaluated with Drop Shape analysis software. A drop of liquid is placed on a bacterial lawn flat surface with the help of a syringe and advancing angle is recorded on a CCD camera. Experimentally a bacterial surface of 600-700 layers thick found to be sufficient to obtain a stabilised contact angle. Each measurement was repeated 3 times and the arithmetic mean was considered as the final contact angle for a particular liquid. Minerals Contact angle on solid powders was determined by sorption measurements using Krss Tensiometer K100 and Krss LabDesk 3.1 software. The Washburn equation is used to measure the contact angle on powder samples. When a column of powder bed is in contact with liquid, the pores between the particles act like small capillaries and the rise of liquid is measurable. The rise of liquid is expressed by the Washburn equation:
I 2 1 .r.cos = t 2
(14)
where I represents the mass of the liquid flow, l is the surface tension of liquid, r is the capillary radius, is the liquid viscosity and is the advancing angle. Capillary radius r could be replaced by capillary constant and can be determined by using a low energy wetting liquid (n-Hexane) that wets the solids completely. In this case, cos equal to 1 and the capillary constant is thus determined. Using the obtained capillary constant, the contact angles for other liquids were determined. Pyrite and chalcopyrite powder samples of -106+38 m size fraction were used. A 1 g of mineral was placed into a glass sample tube and was carefully and equally packed each time. Each measurement was repeated at least 3 times and the results found to be reproducible within ±3 degrees deviation and a mean value was reported in the results.
4 The Open Colloid Science Journal, 2009, Volume 2
Vilinska and Rao
Fig. (1). Growth characteristics of A. ferrooxidans in 9K medium.
beginning due to consumption of H+ by the cells, but after ferric hydroxide precipitation, it becomes constant around pH 2.
For the estimation of acid and base surface energy components with van Oss method, contact angles of at least three different liquids with well characterized acid and base components are necessary. Four standard liquids of water, 1bromonaphtalene, diiodomethane and formamide were used. The values of liquids surface tension and the dispersive and polar part contributions to the surface energy, and the polar part divided into acid and base components of surface energy as reported in the literature were used [23, 24].
The decrease of Fe2+ ions with time is due to their bacterial oxidation to Fe3+ as follows:
3. RESULTS AND DISCUSSION
Fe3+ + 3H2O Fe(OH)3 + 3H+
3.1. Bacterial Growth Characterisation
Formation of colloidal precipitates was observed and the Fe3+ ions in the presence of sulphate and potassium ions can also cause jarosite KFe3(SO4)2(OH)6 formation [25].
The growth curve of A. ferrooxidans in 9K ferrous iron media is presented in Fig. (1). The cell number is increasing until the Fe2+ ions are completely oxidized and the medium with higher Fe2+ content is expected to produce liquor with higher amount of cell population. The pH of solution is relatively stable and only a slight increase is observable at the
2Fe2+ + 1/2O2 + 2H+ 2Fe3+ + 2e- + H2O Fe3+ is then abiotically hydrolyzed to Fe(OH)3, causing a recovery of H + ions:
3.2. Adsorption Studies Adsorption isotherms of A. ferrooxidans on pyrite and chalcopyrite with respect to equilibrium cell concentration are presented in Fig. (2). Adsorption was found to be a fast
Fig. (2). Adsorption isotherm of A. ferrooxidans on pyrite and chalcopyrite.
Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite
process onto both minerals and the isotherms show a saturation adsorption of about 1010 cells/m2 at and above 106 cells/ml concentrations. This adsorption density observed to be well below a full surface coverage in a horizontal orientation while considering the geometrical cell dimensions (i.e., length x breadth: 1.65 x 0.35 ?m). In the concentration range studied and at the highest equilibrium concentration, full surface coverage was not reached. The adherence of A. ferrooxidans cells is higher to pyrite compared with chalcopyrite. As previously reported [26], there are differences between the adhesion ability of different strains of the same acidophilic species toward pyrite and other sulphide minerals. The results in literature generally showed higher cells adhesion onto pyrite [26, 27]. The adhesion is however dependent on thermodynamic balance of the mineral-bacteria system, surface charge of the particles and biological forces such as chemotaxis (directed move toward Fe2+) and preferential adsorption of cells to specific sites on mineral surface. For a better understanding, zeta potential and contact angle were determined so as to calculate the total interaction force according to extended DLVO approach pre-judging bacterial adhesion on mineral surfaces. 3.3. Zeta-Potential The magnitude of surface charge of either positive or negative influences the adsorption process because of electrostatic force, which depends on the charges of interacting particles. Zeta-potential of A. ferrooxidans cells (Fig. 3) is almost constant in the whole pH range measured and displayed negative charge without exhibiting any iso-electric point (IEP).
The Open Colloid Science Journal, 2009, Volume 2
The zeta potential curves of pure pyrite and chalcopyrite as a function of pH is similar, and both exhibited IEP around pH 7.5 (Fig. 3). Pure pyrite exhibited an IEP at pH 7.5 and above this pH the negative potential of the mineral increased with increasing pH. The reported IEP of pyrite as determined by electrophoresis fall in between pH 3.5 and 7.5 [13, 2830], and this variation could arise from several factors such as origin-mineralogy, sample preparation, surface oxidation and aging in water. The IEP of chalcopyrite displayed at almost identical pH as pyrite, but the magnitude of surface charge is lower at acidic and basic pH values. 3.4. Surface Energies of Cells and Minerals The liquid contact angles on mineral samples were determined by sorption measurements using the Washburn equation (Equation 14). Typical liquid sorption curves are presented in Fig. (4) as the square of liquid mass versus time. From the contact angle data presented in Table 1, it is clear that the surface of bacterial cells is more hydrophilic compared with minerals and after bacterial interaction, the surface hydrophobic character of minerals is reduced. The surface energies were calculated using van Oss acid-base approach to determine not only the dispersive and polar components of surface energies, but also to define the acid and base character of surfaces. Surface energies of solids and bacterial cells are comparable and fall in the range 42.39 to 53.75 mJm-2. Surface energy of pyrite is higher compared to chalcopyrite and therefore, the dispersive and polar components of surface energy are also higher but they are proportional to chalcopyrite. Chalcopyrite displayed marginally
30
Pyrite Chalcopyrite
20 Zeta potential (mV)
A. ferrooxidans 10 0 1
3
5
7
9
11
-10
-20 -30
pH Fig. (3). Zeta potential of pyrite, chalcopyrite and A. ferrooxidans with respect to pH. Table 1.
Contact Angles Data and Surface Energy Values Determined by Acid-Base Approach Surface Energy (mJ.m-2 )
Contact Angle
Material Water
Diiodomethane
Bromonaphtalene
5
Formamide
Total
d/LW
P/AB
+
-
Pyrite
61
---
10.4
8.15
53.75
43.87
9.88
4.01
6.08
Chalcopyrite
71.76
---
39.66
43.46
42.39
34.93
7.47
2.2
6.34
A. ferrooxidans
33
53.5
36
32.5
48.55
34.31
14.24
1.11
45.57
6 The Open Colloid Science Journal, 2009, Volume 2
Vilinska and Rao
Fig. (4). Sorption curves of pyrite and chalcopyrite using different test liquids for contact angle estimation.
a higher electron donating character, because of higher part and lower + part compared with pyrite. The electron donating and accepting character is more balanced for pyrite. Surface energy of bacterial cells is within the range of minerals (48.55 mJm2) but the dispersive and polar parts are different. Bacterial cells are more polar in nature than the minerals and this is due to their predominant surface functional groups. Electron donating character of cells is dominating with a much more higher value of - than + due to the occurrence of compounds containing –OH groups (sugars, carboxylic acids, lipids) characteristic of bacterial surface [31].
Table 3.
3.5. Theoretical Approaches of Bacterial Cells Adhesion on Minerals
For calculation of interactive energy components of extended DLVO approach, the parameters such as Hamaker constant, surface potential and free energy of adhesion are essential and these parameters have been determined experimentally. Hamaker constants were calculated by two methods and the calculated values are presented in Table 3. Method 1 represents the calculation using LW part of free energy of adhesion in aqueous system and the effective Hamaker constant was obtained using Equation 13. Method 2 is based on calculating the effective Hamaker constant from the individual Hamaker constants for homogeneous phases. Hamaker constants for bacteria A11 and mineral A22 were estimated using the dispersive part of surface energies obtained from contact angle measurements and the water Hamaker constant A33 was from the literature [32]. Hamaker constants calculated according to the methods 1 and 2 are of course different but the values are comparable and following the same trend. The Hamaker constant for pyrite - A. ferrooxidans system is higher than the chalcopyrite - A. ferrooxidans.
3.5.1. Thermodynamic Approach Free energy of adhesion computed by LW-AB approach for different mineral-bacteria cells are presented in Table 2. The Lifshitz-van der Waals component and acid-base component of free energy of adhesion is negative for all mineralbacteria systems and therefore the total free energy of adhesion is also negative. Thus the bacterial adhesion on minerals is energetically favourable and feasible. The higher electron acceptance ability of pyrite is responsible for the higher values of free energies of adhesion compared with chalcopyrite. The cells are strong electron donors and are attracted to an electron acceptor pyrite surface more. Combining all these factors, the attractive force between pyrite and A. ferrooxidans is stronger than between chalcopyrite and A. ferrooxidans. Table 2.
Free Energy of Adhesion Free Energy of Adhesion Gadh (mJ.m-2) GadhLW
GadhAB
Gadhtotal
Pyrite & A. f.
-4.65
-10.29
-14.93
Chalcopyrite & A. f.
-2.95
-8.10
-12.11
Calculated Hamaker Constants Method 1 A1
Method 2 A11
A22
A33
A2
[x10-21 J] Pyrite & A. f.
4.32
49.4
63.2
37
1.77
Chalcopyrite & A.f.
2.74
49.4
50.3
37
0.95
3.5.2. Extended DLVO Approach
The sphere-sphere geometry of particles for calculating the interaction forces is the most suitable for mineralbacteria system, although the shape of bacillus cells is not exactly spherical. However considering the size of interacting bodies, the spherical model is reasonable. The energy is expressed in kT units in the interaction energy versus dis-
Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite
tance diagrams presented in Figs. (5-9). For outlining the influence of various parameters, a model system was chosen with the following parameters: pH 2, 0.01M KNO3 electrolyte concentration, Hamaker constant estimated by Method 1 and 1 @m size for particles and cells. Influence of Hamaker Constant Two different approaches were used for the Hamaker constant evaluation and as a result two different Hamaker constants were obtained for the same system. Lifshitz - van der Waals forces are always attractive, relatively weak and becoming significant at shorter separation distances. From Equation 6, it is clear that the LW interaction force is proportional to the value of Hamaker constant. The difference in the two Hamaker constants obtained by the methods 1 and 2 is rather small. It is also clear from Fig. (5) that the LW interactions, regardless of the value of Hamaker constant, are the weakest force in the total interaction energy between A. ferrooxidans and minerals. The Hamaker constant with either of the two values has no significant influence on the character of total interaction force. Hamaker constants obtained for pyrite-A. ferrooxidans system are higher and therefore the LW forces are higher compared with chalcopyrite-A. ferrooxidans. This is due to a higher electron acceptance character of pyrite (Table 1). The LW interactions have influence on total interaction energy, but the difference between the values determined by different methods is only few kT and is negligible compared with the strong attractive acid-base interactions. Regardless of the method used for Hamaker constant estimation, the adhesion of bacteria onto minerals is favorable. Due to a small difference between the calculated Hamaker constants by the two methods, only the value of A1 is used in further calculations. Influence of the Ionic Strength Ionic strength is inversely proportional to electrical double layer thickness surrounding the particle and increasing the ionic strength, the double layer compresses and zeta potential decreases. Equation 7 illustrates that the electrostatic force is a function of double layer thickness and zeta potential, and thus by increasing the ionic strength, the two parameters decrease leading to a drop in electrostatic force. Zeta potential was measured only at one electrolyte concentration; however potential energy diagrams were computed at varied ionic strength to identify its influence on bacterial adhesion behavior. Since the zeta-potential decreases with increasing ionic strength, the electrostatic force is greatly reduced for both systems (Fig. 6A). Strong long range attractive force can be achieved at low electrolyte concentration and the modeled 0.001M ionic strength led to an attractive force of 50 kT at a distance of 10 nm (100 Å) between pyrite and A. ferrooxidans. Depending on the magnitude of surface charge, electrostatic force can be strong or weak and therefore, a change in ionic strength displayed a great impact on the interaction forces (Fig. 6B, C). In the calculations, zeta potentials measured at 0.01M KNO3 was only used and a change in zeta potential value due to a varied ionic strength was not incorporated and the calculations are rather simplified. In reality, zeta potential decreases with compressed double layer and the difference between different electrolyte concentrations are even more observable. Low electrolyte concentrations exaggerate the electrostatic force and they are
The Open Colloid Science Journal, 2009, Volume 2
7
preferable when the electrostatic force is weak, and higher concentrations are necessary to achieve fast adhesion or repulsion. Influence of Particle Size The size of bacterial cells is more or less the same (within one strain) and the mineral particles vary from submicron to 5 @m size. The particle size was therefore varied in the simulation of interaction energy. Particle size is varied in the Lifshitz - van der Waals, electrostatic and acid-base interaction forces, while the bacterial size was kept constant due to its constant cell size. In the LW force calculation, the particle size is hidden in parameters x and y and a change from 1 to 4 @m particle size had only a micro effect on the LW force. The curves at different particle sizes are overlapping as presented in Fig. (7A), and any difference in force is not observable. The effect of particle size within the tested and simulated conditions can be considered as negligible. On the other hand, the effect of particle size on electrostatic force is more distinguishable. From Equation 7, it is clear that the force is proportional to a1.a2/(a1+a2) and by keeping one parameter constant and increasing the another, the term is approaching 0 for very small values and 1 for high particle size. Using much higher particle size relative to the cells will increase the force only to some limit and using much smaller particles relative to cell size will minimize the electrostatic force. The simulated particle sizes were 1, 2 and 4 @m (Fig. 7) and with increasing particle size, the electrostatic force also increased and became significant at smaller separation distances. In the present model system, electrostatic force is attractive since the particle and bacteria are charged oppositely but the attractive force is not the strongest and the acid-base interactions are far the strongest. Therefore the alteration of particle size has not changed the electrostatic force and the total interaction energy markedly (Fig. 7C). However, the change is observable between bigger particles where the total attractive force is higher and this effect could be more predominant if the particles are charged more. Influence of pH The pH of medium in which the particles are interacting is not a variable in the interaction equations but greatly influence the surface potential and therefore the electrostatic force of interaction. A. ferrooxidans cells are charged negatively in the entire pH range studied whereas the minerals charged positively below pH 7 and negatively above this pH suggesting an attractive force at lower acidic pH values and repulsive force at alkaline pH values. Pyrite zeta potentials are higher in magnitude and accordingly the electrostatic force obtained for pyrite is higher compared with chalcopyrite. At pH 2 and 7 the potentials are similar that resulted an overlap of electrostatic interaction curves. At pH 5, pyrite is strongly positively charged, what resulted in strong electrostatic attraction towards cells and at pH 10, the bacterial cells and minerals are repelling each other due to negative surface charges (Fig. 8). Based on lower surface charge characteristics of chalcopyrite compared to pyrite (Fig. 3), the electrostatic force for chalcopyrite is also lower (Fig. 9). The total interaction force for both the systems is attractive at pH 2, 5, and 7 for all the separation distances while at pH 10 only after approaching a distance of 2.7 nm (27 Å) overcoming the energetic barrier. The adsorption tests were carried out at
8 The Open Colloid Science Journal, 2009, Volume 2
A
Vilinska and Rao
0
Energy (kT)
0
20
40
60
80
100
-50
LW Interaction using A2 LW Interaction using A1 -100 Distance (Å)
B 500 LW interaction A2 LW interactions A1 Eelectrostatic Interaction Acid-base Interaction
400 300
Energy (kT)
200 100 0 -100 0
20
40
60
80
100
80
100
-200 -300 -400 -500 Distance (Å)
C
0 0
20
40
60
Energy (kT)
-100
-200
-300 Total Interaction Energy A2 -400 Total Interaction Energy A1 -500 Distance (Å)
Fig. (5). Pyrite – A. ferrooxidans interaction energy diagrams at pH 2, 0.01M ionic strength and 1 &m particle size. LW forces calculated using Hamaker constants derived by Method 1 and Method 2 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite
A
The Open Colloid Science Journal, 2009, Volume 2
9
0 0
20
40
60
80
100
Energy (kT)
-50
-100 Electrostatic Interaction 0.001 -150
Electrostatic Interaction 0.01 Electrostatic Interaction 0.1
-200 Distance (Å)
B
0 0
20
40
60
80
100
Energy (kT)
-100
-200 LW Interaction Electrostatic Interaction 0.001 Electrostatic Interaction 0.01 Electrostatic Interaction 0.1 Acid-base Interaction
-300
-400
-500 Distance (Å)
C
0 0
20
40
60
80
100
Energy (kT)
-100
-200
-300 Total Interaction 0.001 Total Interaction 0.01
-400
Total Interaction 0.1 -500 Distance (Å)
Fig. (6). Pyrite – A. ferrooxidans interaction energy diagrams at pH 2, LW force calculated using A1 and 1 )m particle size. Electrostatic interactions are presented at three different ionic strength 0.1M, 0.01M and 0.001M KNO3 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
10 The Open Colloid Science Journal, 2009, Volume 2
A
Vilinska and Rao
0 20
40
Energy (kT)
0
60
80
100
/:,QWHUDFWLRQȝP /:,QWHUDFWLRQȝP /:,QWHUDFWLRQȝP (OHFWURVWDWLF,QWHUDFWLRQȝP (OHFWURVWDWLF,QWHUDFWLRQȝP (OHFWURVWDWLF,QWHUDFWLRQȝP -100 Distance (Å)
B
0 0
20
40
60
80
100
Energy (kT)
-100
-200 LW Interaction (OHFWURVWDWLF,QWHUDFWLRQȝP
-300
(OHFWURVWDWLF,QWHUDFWLRQȝP (OHFWURVWDWLF,QWHUDFWLRQȝP
-400
Acid-base Interaction -500 Distance (Å)
C
0
Energy (kT)
0
20
40
60
80
100
-100 7RWDO,QWHUDFWLRQȝP 7RWDO,QWHUDFWLRQȝP 7RWDO,QWHUDFWLRQȝP
-200 Distance (Å) Fig. (7). Pyrite – A. ferrooxidans interaction energy diagrams at pH 2, 0.01M ionic strength and LW force calculated using A1. Electrostatic and LW forces calculated using different particle sizes 1, 2 and 4 &m (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite
A
The Open Colloid Science Journal, 2009, Volume 2
200
11
Electrostatic Interaction pH 10 Electrostatic Interaction pH 7 Electrostatic Interaction pH 2 Electrostatic Interaction pH 5
Energy (kT)
100
0 0
20
40
60
80
100
-100
-200
B
Distance (Å)
200 100
Energy (kT)
0 0
20
40
60
80
100
-100 LW Interaction Electrostatic Interaction pH 10 Electrostatic Interaction pH 7 Electrostatic Interaction pH 2 Electrostatic Interaction pH 5 Acid-base Interaction
-200 -300 -400 -500
Distance (Å)
C
200 100
Energy (kT)
0 0
20
40
60
80
100
-100 -200 Total Interaction pH 10 -300
Total Interaction pH 7 Total Interaction pH 2 Total Interaction pH 5
-400 -500 Distance (Å)
Fig. (8). Pyrite – A. ferrooxidans interaction energy diagrams at 0.01M ionic strength, LW force calculated using A1 and 1 'm particle size. Electrostatic interactions at pH 2, 5, 7 and 10 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
12 The Open Colloid Science Journal, 2009, Volume 2
A
Vilinska and Rao
200
Electrostatic Interaction pH 10 Electrostatic Interaction pH 7 Electrostatic Interaction pH 2 Electrostatic Interaction pH 5
Energy (kT)
100
0 0
20
40
60
80
100
-100
-200
B
Distance (Å)
200 100
Energy (kT)
0 0
20
40
60
80
100
-100 LW Interaction Electrostatic Interaction pH 10 Electrostatic Interaction pH 7 Electrostatic Interaction pH 2 Electrostatic Interaction pH 5 Acid-base Interaction
-200 -300 -400 -500
Distance (Å)
C
200 100
Energy (kT)
0 0
20
40
60
80
100
-100 -200 Total Interaction pH 10 Total Interaction pH 7
-300
Total Interaction pH 2 Total Interaction pH 5
-400 -500 Distance (Å)
Fig. (9). Chalcopyrite – A. ferrooxidans interaction energy diagrams at 0.01M ionic strength, LW force calculated using A1 and 1 'm particle size. Electrostatic interactions at pH 2, 5, 7 and 10 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
Acidithiobacillus ferrooxidans Cells Adhesion on Pyrite and Chalcopyrite
pH 2 and the total force is attractive for both the mineralbacteria systems. Because of higher acid-base and electrostatic interactions between the cells and pyrite, the adhesion of cells onto pyrite is more. Acid-Base Interactions Acid-base interactions depend on the acid and base surface energy component characteristics of the surfaces and it is a parameter in the calculation of acid-base component of free energy of adhesion. The acid-base part of free energy of adhesion is significantly negative for both mineral-bacteria systems because of different surface characteristics of cells and minerals. Cell surfaces are distinctively polar and strong electron donors, while the minerals possessed some electron acceptance ability. Acid-base interactions are the strongest and therefore are the decisive forces and attractive for both the systems. Because of higher electron accepting character of pyrite, the acid-base interaction energies are higher in this case and the cells adhesion onto pyrite is more preferred (Fig. 2). CONCLUSIONS The adhesion behavior of ferrous grown A. ferrooxidans cells onto pyrite and chalcopyrite is assessed by thermodynamic and extended DLVO theory approaches and verified experimentally. The influence of parameters involved in the physico-chemical interaction forces is verified and a favorable condition for cells adhesion onto minerals by varying these parameters is illustrated and discussed. For a stronger acid-base and weaker LW interaction systems, the influence of parameters on interaction energy is rather limited and difficult to alter the circumstances. The other significant force is the electrostatic. Modification of this force is possible through the variation of electrolyte concentration and pH where a change in surface potential occurs. In agreement with the adsorption studies, both the theoretical approaches yielded favorable condition of bacterial cells adhesion onto mineral surfaces. Experimentally determined differences between the adhesion of cells on pyrite and chalcopyrite were rather small and it was also the case in the computed interaction energy values, and in both cases the adhesion of cells onto pyrite was little higher. The selective adhesion necessary for bio-beneficiation processes could thus be achieved by changing the conditions influencing the electrostatic force. At pH 2 of the natural growth condition of cells, the minerals have similar surface potentials but at a pH greater than 4.5, the difference between the minerals is large manifesting higher electrostatic force between the cells and pyrite. At high alkaline pH 10, the electrostatic force is partially repulsive and this pH is not conducive for physiological attributes of the cells. Manipulating the electrolyte concentration is influencing the electrostatic force proportionally for both the mineral-bacteria systems and additional pH change is necessary for selective cells adhesion onto pyrite. Another possible way to influence the bacterial adhesion is to change the surface properties of the bacteria itself by changing the conditions of growth. A. ferrooxidans cells grown in the presence of minerals secrete different amounts of extracellular polymeric substances thereby the surface properties and adhesion characteristics can be altered.
The Open Colloid Science Journal, 2009, Volume 2
13
ACKNOWLEDGEMENTS The financial support from the EU BioMinE project (contract no. IP NMP2-CT-2005-500329) is gratefully acknowledged. REFERENCES [1]
[2]
[3] [4] [5]
[6] [7] [8] [9]
[10] [11] [12]
[13]
[14]
[15]
[16] [17]
[18]
[19] [20]
[21] [22]
Vilinska, A.; Hanumantha Rao, K.; Forssberg, E. Selective coagulation in chalcopyrite/pyrite mineral system using Acidithiobacillus group bacteria. Adv. Mater. Res., 2007, 20-21, 366-370. Vilinska, A.; Hanumantha Rao, K.; Forssberg, E. In: Wang, D.Z.; Sun, C.Y.; Wang, F.L.; Zhang, L.C.; Han, L.; Eds.; Microorganisms in Flotation and Flocculation of minerals – an overview. Proceedings of XXIV Internation mineral Processing Congress. Beijing, China, Science Press: Beijing, China, 2008; pp. 22-39. Gristina, A. G. Biomaterial centered infection: Microbial adhesion versus tissue integration. Science, 1987, 237, 1588-1595. Lovley, D.R. Anaerobes to the rescue. Science, 2001, 293, 14441446. Sharma, P.K. Surface studies relevant to microbial adhesion and bioflotation of sulphide minerals. Doctoral thesis, Lulea University of Technology, Sweden, 2001. Ehrlich H.C.; Brierley, C.L. Microbial Mineral Recovery, McGraw-Hill: New York, 1990. Sand, W.; Gerhke, T.; Jozsa, P.G.; Shippers, A. (Bio)chemistry of bacterial leaching - direct vs indirect bioleaching. Hydrometallurgy, 2001, 59, 159-175. Tributsch, H. Direct versus indirect bioleaching. Hydrometallurgy, 2001, 59, 177-185. Schippers, A. Sand, W. Bacterial leaching of metal sulfides proceeds by two indirect mechanisms via thiosulfate or via polysulfides and sulfur. Appl. Environ. Microbiol., 1999, 65, 319-321. Schippers, A.; Jozsa, P.G.; Sand, W. Sulfur chemistry in bacterial leaching of pyrite. Appl. Environ. Microbiol., 1996, 62, 3424-3431. Sharma, P.K.; Das, A.; Hanumantha Rao, K.; Forssberg, K.S.E. Surface characterisation of Thiobacillus ferrooxidans cells grown under different conditions. Hydrometalurgy, 2003, 71, 285-292. Sharma, P.K.; Hanumantha Rao, K. Adhesion of Paenibacillus polymyxa on chalcopyrite and pyrite: surface thermodynamics and extended DLVO theory. Colloids Surf. B Biointerfaces, 2003, 29, 21-38 Sharma, P.K.; Hanumantha Rao, K. Analysis of different approaches for evaluation of surface energy of microbial cells by contact angle goniometry. Adv. Colloid Interface Sci., 2002, 98, 341463. Rohwerder, T.; Gehrke, T.; Kinzler, K.; Sand, W. Bioleaching review part A: Progress in bioleaching: fundamentals and mechanisms of bacterial metal sulfide oxidation. Appl. Environ. Microbiol., 2003, 63, 239-248. Bos, R.; van der Mei, H.C.; Busscher, H.J. Physico-chemistry of initial microbial adhesive interactions - its mechanisms and methods for study. FEMS Microbiol. Rev., 1999, 23 (2), 179-230. Absolom, D.R.; Lamberti, F.V.; Policova, Z.; Zingg, W.; van Oss, C.J.; Neumann, A.W. Surface thermodynamics of bacterial adhesion. Appl. Environ. Microbiol., 1983, 46, 90-97. Busscher, H.J.; Weerkamp, A.H.; van der Mei, H.C.; van Pelt, A W.; de Jong, H.P.; Arends, J. Measurement of the surface free energy of bacterial cell surfaces and its relevance for adhesion. Appl. Environ. Microbiol., 1984, 48, 980-983. Deryagin, B.V.; Landau, L. Bacteria sulfide mineral interactions with reference to flotation and flocculation. Acta Physicochem., URSS, 1941, 55, 333. Verwey, E.J.; Overbeek, J.T.G. Theory of the stability of lyophobic colloids. J. Colloid Sci., 1955, 10, 224-225. Van Oss, C.J.; Good, R.J.; Chaudhury, M.K. The role of van der Waals forces and hydrogen bonds in “hydrophobic interactions” between biopolymers and low energy surfaces. J. Colloid Interface Sci., 1986, 111, 378-390. Van Oss, C.J.; Chaudhury, M.K.; Godd, R.J. Monopolar surfaces. Adv. Colloid Interface Sci., 1987, 28, 35-64. Fowkes, F.M. Attractive forces at interfaces. Ind. Eng. Chem., 1964, 56, 40-52.
14 The Open Colloid Science Journal, 2009, Volume 2 [23]
[24]
[25] [26]
[27]
Vilinska and Rao
Bellon-Fontaine, M.N.; Mozes, N.; Van Der Mei, H.C.; Sjollema, O. A comparison of thermodynamic approaches to predict the adhesion of dairy microorganisms to solid substrata. Cell Biophys., 1990, 17, 93-106. Ström, G.; Frederiksson, M.; Stenius, P. Contact angles, work of adhesion and interfacial tensions at a dissolving hydrocarbon surface. J. Colloid Interface Sci., 1987, 119, 352-361. Herbert, R.B. Properties of goethite and jarosite precipitated from acidic groundwater, Dalarna, Sweden. Clay. Clay Miner., 1997, 45, 261-273. Harneit, K.; Göksel, A.; Kock, D.; Klock, J.H.; Gehrke, T.; Sand, W. Adhesion to metal sulfide surfaces by cells of Acidithiobacillus ferrooxidans, Acidithiobacillus thiooxidans and Leptospirillum ferrooxidans. Hydrometallurgy, 2006, 83, 245-254. Sharma, P.K.; Das, A.; Hanumantha Rao, K.; Forssberg, K.S.E. In: Parekh, B.K.; Miller, J.D.; Eds.; Thiobacillus ferrooxidans interaction with sulfide minerals and selective chalcopyrite flotation from pyrite. Advances in Flotation Technology, Denver, Colorado, USA, SME, NY, USA, 1999; pp. 147-165.
Received: November 28, 2008
[28]
[29]
[30]
[31] [32]
Das, A.; Hanumantha Rao, K.; Sharma, P.K.; Natarajan, K.A.; Forssberg, K.S.E. In: Amils, R.; Ballester, A., Eds.; Surface chemical and adsorption studies using Thiobacillus ferrooxidans with reference to bacterial adhesion to sulfide minerals, Biohydrometalurgy and the Environment toward the mining of the 21st Century, part A, IBS-1999; Madrid, Spain; Elsevier, 1999; pp. 697-707. Chandraprabha, M.N.; Natarajan, K.A.; Somasundaran, P: Selective separation of pyrite from chalcopyrite and arsenopyrite by biomodulation using Acidithiobacillus ferrooxidans. Int. J. Miner. Process., 2005, 75, 113-122. Natarajan, K.A.; Das, A: Surface chemical studies on Acidithiobacillus group of bacteria with reference to mineral flocculation. Int. J. Miner. Process., 2003, 72, 189-198. Gehrke, T.; Hallmann, R.; Kinzler, K.; Sand, W. Importance of extracellular polymeric substances from Thiobacillus ferooxidans for bioleaching. Appl. Environ. Microbiol., 1998, 64, 2743-2747. Evans, D.F.; Wennerström, H. The colloidal domain where physics, chemistry, biology and technology meet. Wiley-VCH, 1994.
Revised: January 6, 2009
Accepted: January 9, 2009
© Vilinska and Rao; Licensee Bentham Open. This is an open access article licensed under the terms of the Creative Commons Attribution Non-Commercial License (http: //creativecommons.org/licenses/bync/3.0/) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, provided the work is properly cited.
Paper V Surface thermodynamics and extended DLVO theory of Leptospirillum ferrooxidans cells adhesion on pyrite and chalcopyrite. Vilinska A. and Hanumantha Rao K., 2009. Journal of Colloid and Interface Science, to be submitted.
Surface thermodynamics and extended DLVO theory of Leptospirillum ferrooxidans cells adhesion on sulfide minerals A. Vilinskaa and K. Hanumantha Raob Division of Mineral Processing, Luleå University of Technology SE-971 87 LULEÅ, Sweden1 E-mail:
[email protected],
[email protected]
Abstract The adhesion behaviour of Leptospirillum ferrooxidans bacteria onto pyrite and chalcopyrite was assessed by the surface thermodynamics and extended DLVO theory approaches. Zetapotential and contact angle of powdered solids and bacterial cells that are essential parameters in the calculations were determined experimentally. Hamaker constant in the Lifshitz-van der Waals interactions was obtained by the microscopic and macroscopic methods. Adsorption tests were carried out at pH 2 of cells optimum growth conditions in order to estimate the amount of cells adsorbed onto mineral surfaces and to compare with the theoretical predictions. The free energy of cells adhesion on both minerals was found to be negative illustrating that the adhesion is energetically favoured based on thermodynamic considerations. The DLVO interaction energy curves also showed that the total interaction energy between bacteria and mineral particle was negative and attractive when they were oppositely charged and in the cases of same polarity, the total interaction energy was attractive at closer separation distances after overcoming an energetic barrier caused by the repulsive electrostatic forces. Adsorption results were found to be in good agreement with theoretical predictions emphasizing that the bacterial adhesion on minerals can be judged from the consideration of physico-chemical interaction forces. Keywords: Leptospirillum ferrooxidans, Thermodynamics, Extended DLVO theory
Bacteria,
Sulfide
mineral,
Adhesion,
Surface
1. Introduction Nowadays, biotechnological approaches based on integral green-chemistry methods are more favoured in mineral processing technologies than the conventional physico-chemical methods. As the biotechnological applications are becoming widely popular, it is important to understand and define the fundamental issues determining the success of a technological process. Microorganisms have a tremendous influence on their environment through the transfer of energy, charge, and materials across a complex biotic mineral-solution interface. The studies of mineral-bacteria interaction arises from the expected impact on many technological areas, including biosorption, bioleaching, bioremediation and bioflotation. Common to all these microbiological processes is the bacterial need to access, adhere to, and react with the mineral-water interface. Previous studies have shown that under most physiological conditions the bacterial cell surface carries a net negative charge, while, along with electrostatic forces, hydrophobic, entropic, acid-base, and van der Waals interactions and H-bonding are important in the bacterial adhesion. The microbe attachment to the mineral surface is followed/accompanied by expression of extracellular polymeric substances (EPS)
1
enabling the adhesion. In addition, the functional groups of EPS can form chemical bonds with the surface. Microorganisms are being used in flotation and flocculation of minerals. The biomodification of mineral surfaces relevant to these processes involves the complex action of microorganism on the mineral surface. There are three different mechanisms by means of which the biomodification can occur: i) attachment of microbial cells to the solid substrate, ii) oxidation reactions and iii) adsorption and/or chemical reaction with the EPS. Autotrophic and heterotrophic bacteria, yeasts, fungi and algae have been used in minerals biobeneficiation. Leptospirillum ferrooxidans is a chemolitotrophic acidophilic ferrous iron oxidizing bacteria of similar properties to Acidithiobacillus ferrooxidans. Both strains are comprehensively investigated in bioleaching of sulfides and A. ferrooxidans in addition was applied as a possible bioreagent in classical mineral processing technologies such as flotation. A. ferrooxidans was found to depress the flotation of pyrite and the process was selective in the presence of other sulfide minerals [1-6]. Also, the cells positively influenced the settling and flocculation of sulfides [4,7]. Similar promising results were achieved with Acidithiobacillus thiooxidans [8] and the use of other acidophilic species such as L. ferrooxidans is possible and needs further investigation. The bacterial adhesion and a change in surface properties relevant to flotation and flocculation were found to be dependent on several factors, in particular the cell concentration. There was a dependence of a change in surface properties (zeta potential change) on the amount of adsorbed cells on the minerals [4,5,9], and the amount of adsorbed cells is different for different minerals. The adhesion of cells on the mineral surface is therefore crucial for the biomodification and determining the success of the whole process. The adhesion behaviour of A. ferrooxidans onto sulfide minerals through surface thermodynamics and extended DLVO theory was investigated very recently by us [10] and in this paper, an analysis of L. ferrooxidans adhesion on sulfides has been presented by the same theoretical approaches and an attempt has been made to verify the inferences experimentally. 1.1 Bacterial Adhesion onto Minerals 1.1.1 Thermodynamic approach According to thermodynamics, a system will undergo change and proceeds only towards an energetically favoured state; the state of which after the change will have a lower total energy. In the case of bacterial adhesion to a solid surface, the energetic state of bacteria-solid system has to be lower than bacteria and solid, which can be expressed in terms of free energy of adhesion, ǻGadh [11,12]: ǻGadh = Ȗbs – Ȗbl – Ȗsl
[1]
where Ȗ represents interfacial free energy for different interfaces: bacteria-solid (Ȗbs), bacterialiquid (Ȗbl) and solid-liquid (Ȗsl). Adhesion is energetically favoured only if ǻGadh is negative. For calculation of free energy of adhesion, the interfacial free energies are necessary and they can be obtained from the contact angle data with different test liquids with known surface energy components. Different approaches were reported in literature to determine solids surface energy from contact angles data, but the van Oss acid-base approach [13-16] was followed since this approach was found to give consistent results [17,18] besides providing bacterial cells surface electron-donating and electron-accepting characteristics.
2
The free energy of adhesion, ǻGadh, was conveniently divided into two parts: Lifshitz - van der Waals (LW) component and acid-base (AB) component. ǻGadh = ǻGadhAB + ǻGadhLW
[2]
The components of free energy of adhesion can be calculated from the interfacial tensions as follows: 'Gadh
LW
AB
'Gadh
2§¨ J bv ©
·¸§¨ J LW J LW ·¸ lv ¹© sv ¹ 2§¨ Jbv Jsv ·¸§¨ Jbv Jsv ·¸ 2§¨ Jbv Jlv ·¸§¨ Jbv Jlv ·¸ 2§¨ Jsv Jlv ·¸§¨ Jsv Jlv ·¸ © ¹© ¹ © ¹© ¹ © ¹© ¹ LW
J lv
LW
[3] [4]
1.1.2. DLVO approach Classical DLVO approach [19,20] includes Lifshitz-van der Waals interactions and electrostatic interactions. LW forces are short range and always attractive, and these are weak forces between neutral stable molecules. The Coulombic electrostatic interactions are long range and could be attractive or repulsive, depending on the surface charge polarity of interacting particles. Acid-base interactions were added later by van Oss [21] to involve electron donating-accepting abilities of different materials. The microbial adhesion to solid surfaces is then described by a sum of van der Waals, electrostatic and acid-base forces operating between the surfaces approaching close to one another. Gtotal = GLW + GEL + GAB
[5]
Calculation of these forces depends on the geometry of interacting phases and for a spheresphere system, the following equations were used: Lifshitz van der Waals energy: G LW
§ x 2 xy x Aª y y « 2 2 2 ln¨¨ 2 12 ¬ x xy x x xy x y © x xy x
§ ¨ ·º¨ 1 ¸¸» ¨ y ¹¼ § 2SH ¨ 1 1.77¨ © O ©
· ¸ ¸ ·¸ ¸¸ ¹¹
[6]
Electrostatic interaction energy:
G EL
º SHa1a 2 ] 1 ] 2 ª 2] 1] 2 1 e NH ln ln 1 e 2NH » « NH ] ] a1 a 2 1 e 2 ¬ 1 ¼
[7]
Acid-base interaction energy: GAB = ʌaȜǻGadhABe[(do-H)/ Ȝ]
[8]
where H is the separation distance, a1 and a2 are the radii of the interacting particles, ȗ1 and ȗ2 are the zeta-potentials of the interacting particles, ț is the double layer thickness-1, A is the Hamaker constant, d0 is the minimum separation distance between the two surfaces (0.157 nm), Ȝ is the correlation length of molecules in liquid (0.6 nm) and x = H/(a1 + a2), y = a1/a2.
3
For interaction energy calculations, the parameters such as zeta potential, particle size radius and double layer thickness are known or measurable; the only unknown parameter is the Hamaker constant. There are two different methods to evaluate the Hamaker constant that influence the value of Lifshitz van der Waals interaction energy [13,21]. First method is microscopic approach, where the total interaction is assumed to be the sum of all interactions between the atom pairs and the second method is a macroscopic approach, where the particles and the medium are considered as continuous phases. In the presence of a liquid medium, the van der Waals energy between the particles differs and therefore, the Hamaker constant has to be replaced by an effective Hamaker constant. In case of two different particles (1 and 2) in a medium (3), the effective Hamaker constant is calculated by: A123 = A12 + A33 – A13 – A23
[9]
The value of A12 is considered as a geometric mean of A11 and A22: A12 = (A11.A22)1/2
[10]
A123 = (A111/2 – A331/2)(A221/2 – A331/2)
[11]
For the calculation of individual Hamaker constants for solid surfaces, Fowkes [22] proposed the following equation: A11 = 6ʌr2ȖsLW
[12]
where r represents the intermolecular distance and ȖsLW represents the dispersive component of the surface energy of the solid. According to Fowkes, the value of 6ʌr2 equals to 1.44x10-18 m2 for water and systems with the volume element as metal atoms, CH2 and CH groups which have nearly the same size. From the individual Hamaker constants calculated from the dispersive component of surface energies, the effective Hamaker constant can thus be calculated. Another possible way of obtaining the Hamaker constant is from Lifshitz-van der Waals component of free energy of adhesion, ǻGadhLW, calculated from the contact angle measurements. The effective Hamaker constant is then given by: A =- 12 ʌd02 ǻGadhLW
[13]
Hamaker constants were calculated by the above two methods, while the Hamaker constant used for water was from the literature. For evaluation of electrostatic interaction energies, the zeta-potential values obtained at a particular ionic strength as a function of pH were used in place of surface potentials. The ionic strength used in the experimentation determines the double layer thickness-1.
4
2. Materials and Experimental methods 2.1 Minerals Pure natural pyrite and chalcopyrite crystals supplied by Gregory, Bottley and Lloyd, UK, were used in the present studies. The crystals were broken into small pieces and dry ground in an agate mortar. The product was sieved to obtain suitable size fractions of ņ106+38 ȝm for contact angle measurements and ņ5 ȝm for zeta-potential and adsorption studies. Prepared samples were washed with HCl for cleaning the surface oxidised species and stored in a refrigerator. The BET specific surface area of these coarse and fine samples was 0.06 m2 gņ1 and 1.02 m2 gņ1 for pyrite and, 0.17 m2 gņ1 and 1.90 m2 gņ1 for chalcopyrite, respectively. 2.2 Bacteria The bacterial strain Leptospirillum ferrooxidans DSM2391 was grown in a modified Leptospirillum (HH) medium (40 g/l FeSO4.7H2O, 0.132 g/l (NH4)2SO4, 53 mg MgCl2.6H2O, 27 mg KH2PO4, 0.147 g CaCl2.2H2O, 62 ȝg MnCl2.2H2O, 68 ȝg ZnCl2, 64 ȝg CoCl2.6H2O, 31 ȝg H3BO3, 10 ȝg Na2MoO4, 67 ȝg CuCl2.2H2O) at pH 1.8 and at 30°C on a rotary shaker maintaining 150 rpm. Collected cell mass was filtered and washed with water at pH 2 to obtain cells devoid of possible precipitates and metabolites. 2.3. Contact angle determination 2.3.1 Bacteria Contact angle measurements on Leptospirillum ferrooxidans cells grown in ferrous iron solution were performed. The cells were collected at the exponential growth phase and filtered through Whatman filter paper to remove the precipitates. Precipitate-free cell solution was filtered through Millipore filter paper using vacuum filtration to obtain uniformly distributed cell layer composing several bacterial lawns on the whole area of filter paper. After filtration, the Millipore filter paper was placed on a Whatman filter paper for a short time to remove the excess moisture. When the moisture is removed, the filter paper is dry enough to be fixed on a microscope slide glass with a help of double sided adhesive tape. The samples were air-dried to remove the rest of water and contact angles were measured. Contact angles were measured by placing a drop of test liquid on the bacterial lawn surface and the dynamic contact angle was recorded with KrĦss Easy Drop equipment associated with Drop Shape Analysis software. Experimentally a bacterial lawn of 600-700 layers was found sufficient to obtain a stabilized contact angle. Each measurement was repeated 3 times and the arithmetic mean was reported in the results as the final contact angle of a particular liquid. 2.3.2 Sulfide minerals Contact angle on solid powders was determined by sorption measurements with the aid of KrĦss Tensiometer K100 and KrĦss LabDesk 3.1 software, which incorporate Washburn method. Sorption measurements define the rise of liquid flow in powder bed pores as in capillary tube. The rise is expressed by the Washburn equation:
I2 t
J 1 .r. cosT 2K
[14]
5
where I represent the mass of the liquid flow, Ȗl is the surface tension of liquid, r is the capillary radius, Ș is the liquid viscosity and ș is the advancing angle. Capillary radius r could be replaced by capillary constant, which is determined by using a virtually completely wetting liquid with contact angle 0 (cos ș = 1). The capillary constant was determined using n-hexane that wets the solids completely. With the experimentally determined capillary constant, the contact angles for other test liquids are calculated from the sorption curves. Pyrite and chalcopyrite powder samples of -106+38 ȝm size fraction were used. 1 g of mineral was placed in a glass cylindrical sample holder and was carefully and equally packed. Each measurement was repeated at least 3 times and a mean value was considered as the final contact angle. 2.4. Zeta potential measurements Zeta potential measurements were realized with ZetaCompact, using Smoluchowski equation. Pyrite and chalcopyrite ground to -5 μm particle size were used in the measurements. A solid concentration of 2.5 mg l-1 was used. Desired ionic strength (10-2 M and 10-3 M) was maintained with KNO3. Solution pH was adjusted with the addition of HNO3 and KOH. Solutions with a desired pH and constant ionic strength were prepared and then the particular mineral was added. The suspension was conditioned for 30 minutes and the pH of the suspension was recorded again and regarded as the pH of the measurement. Zeta potential measurements of bacterial cells were performed similarly. L. ferrooxidans cells grown in MM media were collected and the cell mass was washed with pH 2 de-ionized water. Different bacterial concentrations were attempted initially and 2.5x107 cells ml-1 was found suitable for the instrument. 2.5. Adsorption measurements The ņ5 ȝm size fraction of minerals was used for adsorption studies. The adsorption of bacterial cells on pyrite and chalcopyrite was carried out in 250 ml Erlenmeyer flasks. Tests were performed with 1 g of mineral in 100 ml water containing varied initial cell concentrations of 107, 5x107, 108, 5x108, 109 cells mlņ1. After 30 minutes of interaction time, the cells in liquid phase were estimated using a Neubauer counter under a microscope. The amount of cells adsorbed on mineral surfaces was determined by the difference in the concentration of cells before and after adsorption.
3. Results and Discussion 3.1 Bacterial growth characterisation Growth curve of L. ferrooxidans grown in ferrous iron medium is presented in Fig. 1. The cell number was increasing until the Fe2+ ions were completely oxidized; hence the medium with higher Fe2+ content is expected to produce liquor with higher amount of cells. The pH of solution was relatively stable, only a slight increase was observed at the beginning, due to consumption of H+ ions by the cells, but it became constant thereafter around pH 2 as the precipitation started.
6
Figure 1. Growth characteristics of L. ferrooxidans The decrease of Fe2+ ions with time is due to its bacterial oxidation to Fe3+ as follows: 2Fe2+ + 1/2O2 + 2H+ ĺ 2Fe3+ + 2e- + H2O Fe3+ is then abiotically hydrolyzed to Fe(OH)3, causing a recovery of H+ ions Fe3+ + 3H2O ĺ Fe(OH)3 + 3H+ Formation of precipitates was observed as the Fe3+ ions in the presence of sulphate and potassium ions causes jarosite KFe3(SO4)2(OH)6 formation [20]. 3.2 Adsorption studies Leptospirillum ferrooxidans cells adhesion on pyrite and chalcopyrite were carried out to determine the extent of cells adsorption on mineral surfaces. The adsorption isotherms of cells for pyrite and chalcopyrite at pH 2 are shown in Fig. 2. The adsorption density increased with increasing equilibrium concentration of cells. When the initial cell concentration was 109 cells mlņ1, nearly all the cells are adsorbed on pyrite and chalcopyrite corresponding to 80.6% and 99.9% of the initial cells respectively. The isotherms display a linear increase in adsorption density from an equilibrium cells density of 104 cells mlņ1 and there was no levelling off adsorption within the cell concentration range studied. While considering the geometrical cell dimensions (i.e., length x breadth: 1.65 x 0.35 ȝm), the maximum adsorption density attained was about 1011 cells mņ2 which corresponds to just 10% of surface coverage. The adsorption density of A. ferrooxidans cells on pyrite was found to be faster and higher compared to chalcopyrite [1,2,4,5,7]. It was also reported that the same species of bacterial cells arising from different strains had differences in adsorption behaviour [24]. Although the present adsorption isotherms are limited to lower surface coverage, L. ferrooxidans cells adsorption on chalcopyrite was higher than pyrite at any of the equilibrium cells concentration. Acidithiobacillus group bacteria are known to specifically adsorb on surface defects and imperfections [25]. Since chalcopyrite surface area is nearly twice that of pyrite, it is presumed to contain higher surface imperfections than pyrite and therefore higher adsorption of cells on chalcopyrite.
7
Fig. 2. Adsorption isotherms of L. ferrooxidans cells on pyrite and chalcopyrite.
3.2. Zeta potential measurements The zeta-potentials of pure Leptospirillum ferrooxidans cells and pyrite, and chalcopyrite as a function of pH are shown in Fig. 3. Pure cells exhibited an iso-electric point (IEP) at pH 3.3 and beyond this pH the magnitude of negative potential increases with a rise in pH. The magnitude of zeta-potentials is relatively high compared to A. ferrooxidans cells grown under similar conditions [1-4,7,9]. The presence of functional groups such as carboxyl, amino and hydroxyl, and their ionisation impart surface charge to the cells. The presence of ammonium containing polymers (proteins) and polysaccharides containing phosphate and/or carboxylic groups in the surface layers of bacteria and a charge equivalence of these anionic and cationic acid/base groups determines the IEP of bacterial cells. The positive surface potential at low pH values is obviously due to the protonation of ammonium groups and the increase in negative charge beyond pH 3.3 is caused by the dissociation of anionic functional groups. The cells are thus expected to adsorb on mineral surface through electrostatic interactions and as well through specific chemical interactions of functional groups on surface metal ions besides metabolic reasons.
Zeta potential (mV)
30
Pyrite
20
Chalcopyrite
10
L.ferrooxidans
0 -10
1
3
5
7
9
11
-20 -30 -40
pH
Fig. 3. Zeta potential of pyrite, chalcopyrite and L. ferrooxidans relative to pH
8
Pure pyrite exhibited an IEP at pH 7.5 and above this pH the negative potential of the mineral increased with increasing pH. The IEP of chalcopyrite was almost identical to that of pyrite, but the magnitude of chalcopyrite surface charge was relatively lower in acidic and basic pH values than pyrite. 3.3 Contact angles and surface energies of interacting phases The liquid contact angles on mineral samples were determined by sorption measurements using the Washburn equation. Typical liquid sorption curves are presented in Fig. 4 as the square of liquid mass versus time. The surface tensions of test liquids are presented in Table 1 and the surface energies of solids and bacterial cells calculated from the contact angle data are shown in Table 2. The total surface energy of either solids or bacterial cells was nearly the same, varied from 42.39 to 53.75 mJm2. Surface energy of pyrite was higher compared with chalcopyrite and therefore the dispersive and polar components of surface energy were also higher, but they were proportional to chalcopyrite. Chalcopyrite had a higher electron donating character, because of higher Ȗ- part and lower Ȗ+ part compared with pyrite. The electron donating and accepting character was relatively balanced for pyrite. The total surface energy for the bacterial cells (48.34 mJm2) lied within the range of minerals surface energy, but the dispersive and polar parts were different. L. ferrooxidans was slightly more polar character than the minerals. Electron donating character was dominating with a much higher value of Ȗ- than Ȗ+ due to the occurrence of compounds containing –OH groups of sugars, carboxylic acids and lipids, characteristic for bacterial surface [24]. Table 1. Surface energies of test liquids used for contact angle measurements [28,29]. Surface energy Ȗ [mJ.m-2] Total Ȗd(ȖLW) ȖP(ȖAB) Ȗ+ Ȗ-
Water
Diiodomethane Bromonaphtalene Formamide
72.8 21.8 51 25.5
50.8 50.8 0 0
44.6 44.6 0 0
58 39 19 2.3
25.5
0
0
39.6
Table 2. Contact angles data and surface energy values determined by acid-base approach.
Material Water Pyrite 61 Chalcopyrite 71.76 L.ferrooxidans 30.5
Contact angle ș DiiodoBromomethane naphtalene Formamide --10.4 8.15 --39.66 43.46 55,3 42.3 32.5
Surface energy Ȗ (mJ.m-2) total 53.75 42.39 48.34
Ȗd/ȖLW 43.87 34.93 32.43
ȖP/ȖAB 9.88 7.47 15.91
Ȗ+ 4.01 2.2 1.36
Ȗ6.08 6.34 46.47
9
Fig. 4. Sorption curves of pyrite and chalcopyrite using different test liquids for contact angle estimation. The free energy of adhesion of different mineral-bacteria cells are presented in Table 3. Calculations were done using the LW-AB approach. The Lifshitz-van der Waals component and acid-base component of free energy of adhesion was negative for both mineral-bacteria systems. Therefore the total free energy of adhesion was negative and the adhesion was energetically favourable. The higher electron acceptance ability of pyrite can explain the higher values of free energies of adhesion compared with chalcopyrite, because the cells are strong electron donors and are attracted more towards an electron acceptor. Combining these factors, the highest attractive force is expected between pyrite and L. ferrooxidans and smallest between chalcopyrite and L. ferrooxidans. Table 3. Free energy of adhesion.
Pyrite and L. f. Chalcopyrite and L. f.
Free energy of adhesion ǻGadh (mJ.m-2) ǻGadhLW ǻGadhAB ǻGadhtotal -4.01 -9.30 -12.25 -2.55 -7.06 -9.61
3.4 Extended DLVO approach of bacterial adhesion The necessary Hamaker constants for LW interaction energy calculation were evaluated by two methods and the values are presented in Table 4. Method 1 represents the calculation using LW part of free energy of adhesion in aqueous system and the effective Hamaker constant in this case was obtained using Eq. 13. Method 2 is based on calculating the effective Hamaker constant from the individual Hamaker constants for the same interacting phases. Hamaker constants for bacteria A11 and mineral A22 were calculated using the dispersive part of surface energies obtained from contact angle measurements and water Hamaker constant A33 from the literature was used [27]. Hamaker constants obtained with different methods were of course different, but the values were comparable and followed the same trend. The
10
Hamaker constant was higher for pyrite-L. ferrooxidans system, and lower for chalcopyrite-L. ferrooxidans. Table 4. Calculated Hamaker constants. Method 1 A1
Method 2 A11
A22 [x10
Pyrite and L. f. Chalcopyrite and L. f.
3.73 2.37
46.7 46.7
-21
63.2 50.3
A33
A2
37 37
1.40 0.76
J]
For mineral-bacteria interacting system, the sphere-sphere equation of DLVO forces is perceived to be the most suitable. Bacillus shape cells are not exactly spherical but considering the width/length ratio, the spherical model is more appropriate. The significance of different variable parameters is assessed using a standard model (pH 2, 0.01 M KNO3, 1 ȝm particle and bacteria size and Hamaker constant determined by method 1) with a change in one parameter while keeping the rest constant. 3.4.1 Influence of Hamaker constant For evaluation of the Hamaker constant, two different approaches were used and as a result two different Hamaker constants were obtained for the same system. Lifshitz van der Waals forces are always attractive, relatively weak and becoming significant only at closer separation distances. The LW interaction force is proportional to the value of Hamaker constant and the calculated values of Hamaker constants are different; but the differences are rather small. From Fig. 5, it is clear that the LW interactions were the weakest in the total interaction force between L. ferrooxidans and pyrite or chalcopyrite, regardless of the value of Hamaker constant used. The Hamaker constant obtained by the two methods did not alter the magnitude of total interaction force significantly. Hamaker constants obtained for pyrite - L. ferrooxidans system were higher and therefore the LW forces were higher compared with chalcopyrite - L. ferrooxidans system. This is caused by the higher electron acceptance character of pyrite surface. Higher LW interactions had a higher impact on total energetic balance, but the difference between the values determined by the two methods was few kT and negligible compared with the strong attractive acid-base interactions. Regardless of the method used for Hamaker constant estimation, bacterial adhesion on both minerals was favourable below the separation distance of 2.3 – 2.5 nm (23-25 Å), where the acid-base forces were becoming dominant. 3.4.2 Influence of the ionic strength Ionic strength (electrolyte concentration) is inversely proportional to double layer thickness and increasing the ionic strength causes the double layer to compress, which in turn decreases the zeta potential. The electrostatic force is a function of the double layer thickness and zeta potential and thus by increasing the ionic strength, these two parameters decrease at the same time leading to a drop in the electrostatic force. Zeta potential was measured only at 0.01 M ionic strength, but a simulation of the ionic strength change was used in the calculations. Because of the proportionality, the increase in ionic strength greatly decreased the electrostatic force for both systems (Fig. 6). Strong long range repulsive force can be achieved using low concentrations of electrolyte and the modeled 0.001M KNO3 lead to a repulsion of
11
about 100 kT at a distance of 10 nm (100 Å in diagram) between pyrite and L. ferrooxidans. Electrostatic force can be a strong force depending on the surface potential and therefore using different ionic strength has a great impact on the total balance of forces, as presented in Figs. 5b and 5c. Low electrolyte concentration increased the energetic barrier and at the highest electrolyte concentration, the electrostatic repulsion was decreased and made the total resultant force attractive. However, the calculations were simplified as the zeta potential measured at 0.01M KNO3 electrolyte concentration only was used and a change in the zeta potential value due to a change in ionic strength was not incorporated. In reality the zeta potential will be lower with compressed double layer and the differences at different electrolyte concentrations will be more significant. A higher electrolyte concentration lowers the repulsive force as desired in this case by increasing the bacterial adhesion on minerals. 3.4.3 Influence of particle size The size of bacterial cells could be the same for a single strain, but the fine mineral particles consist of submicron to 5 ȝm sizes and therefore only the variation of mineral particle size was incorporated in the simulations. Particle size as a variable parameter is included in the Lifshitz van der Waals force equation, electrostatic force calculation and acid-base interaction force, and a constant cell size of 1 ȝm was incorporated. In the LW force calculation, the particle size is hidden in the parameters and the simulated change from 1 to 4 ȝm particle size had only a minimal effect on the LW force. It can be seen from Fig. 7 that the interaction energy values were overlapping at different particle sizes and differ less than 1%. The effect of particle size on the LW interactions within the examined and simulated conditions can thus be considered negligible. On the other side, electrostatic force is more influenced with a change in particle size. From equation 7, it is clear that the force is proportional to a1a2/(a1+a2) and when one parameter is kept constant and increasing the other, the term is approaching 0 for very small particle sizes and 1 for high particle sizes. A higher particle size relative to the cells increases the force only to some extent and with a much smaller particle relative to cell size minimizes the electrostatic force. The simulated particle sizes were 1, 2 and 4 ȝm (Fig. 7) and with increasing particle size, the electrostatic force was seen to increase and becoming significant at smaller separation distances. In the case of present mineralbacteria system at pH 2, the electrostatic force was repulsive since the particles and cells were positively charged, and the repulsive electrostatic force was dominant at a separation distance of 3 nm (30 Å), below which, the acid-base interactions started to exert influence on the total force and became attractive after creating an energetic barrier of 30-70 kT depending on the particle size. The energetic barrier could be lowered and minimized by decreasing the particle size.
12
A
0
Energy (kT)
0
20
40
60
80
100
-50
LW Interaction using A2 LW Interaction using A1
-100 Distance (Å)
B
500 LW interaction A2 LW interaction A1 Eelectrostatic Interaction Acid-base Interaction
Energy (kT)
400 300 200 100 0 -100 0
20
40
60
80
100
80
100
-200 -300 -400 -500 Distance (Å) C
100 0
Energy (kT)
0
20
40
60
-100 -200 -300 Total Interaction Energy A2
-400
Total Interaction Energy A1
-500 Distance (Å) Fig. 5. Pyrite - L. ferrooxidans interaction energy diagrams at pH 2, 0.01 M KNO3 and 1 ȝm particle size. LW forces calculated with using Hamaker constants derived by Method 1 and Method 2 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
13
A
200 150
Energy (kT)
100 50 0 -50 0
20
40
60
80
100
Electrostatic Interaction 0.001 Electrostatic Interaction 0.01 Electrostatic Interaction 0.1
-100 -150 -200
Distance (Å) B
LW Interaction Electrostatic Interaction 0.001 Electrostatic Interaction 0.01 Electrostatic Interaction 0.1 Acid-base Interaction
500
Energy (kT)
400 300 200 100 0 -100 0 -200
20
40
60
80
100
80
100
-300 -400 -500 Distance (Å) C
200 100
Energy (kT)
0 -100
0
20
40
60
-200 Total Interaction 0.001 Total Interaction 0.01 Total Interaction 0.1
-300 -400 -500 Distance (Å)
Fig. 6. Pyrite – L. ferrooxidans, interaction energy diagrams at pH 2, LW force calculated using A1 and 1 ȝm particle size. Electrostatic interactions are presented at three different ionic strength 0.1M, 0.01M and 0.001M KNO3 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
14
A
300
Energy (kT)
200 100 0 0
20
40
60
80
100
LW Interaction 4 ȝm LW Interaction 2 ȝm LW Interaction 1 ȝm Electrostatic Interaction 4 ȝm Electrostatic Interaction 2 ȝm Electrostatic Interaction 1 ȝm
-100 -200 -300 Distance (Å)
B 500
LW Interaction Electrostatic Interaction 4 ȝm Electrostatic Interaction 2 ȝm Electrostatic Interaction 1 ȝm Acid-base Interaction
Energy (kT)
400 300 200 100 0 -100 0 -200
20
40
60
80
100
-300 -400 -500 Distance (Å)
Energy (kT)
C
100
0 0
20
40
60
80
100
Total Interaction 4 ȝm Total Interaction 2 ȝm Total Interaction 1 ȝm
-100
-200 Distance (Å) Fig. 7. Pyrite – L. ferrooxidans, interaction energy diagrams at pH 2, LW force calculated using A1 and 0.01 M KNO3. Electrostatic and LW interactions were calculated using three different particle sizes 1, 2 and 4 ȝm (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
15
A
500
Electrostatic Electrostatic Electrostatic Electrostatic
400
Energy (kT)
300 200
Interaction pH 10 Interaction pH 7 Interaction pH 2 Interaction pH 5
100 0 -100 0
20
40
60
80
100
-200 -300 -400 -500
Distance (Å)
B LW Interaction Electrostatic Interaction pH 10 Electrostatic Interaction pH 7 Electrostatic Interaction pH 2 Electrostatic Interaction pH 5 Acid-base Interaction
500
Energy (kT)
400 300 200 100 0 -100 0 -200
20
40
60
80
100
-300 -400 -500 Distance (Å) C
200
Energy (kT)
100 0 -100
0
20
40
60
80
100
-200 -300 -400
Total Interaction pH 10 Total Interaction pH 7 Total Interaction pH 2 Total Interaction pH 5
-500 Distance (Å) Fig. 8. Pyrite – L. ferrooxidans interaction energy diagrams at 0.01 M KNO3, LW interactions calculated using A1 and 1 ȝm particle size. Electrostatic interactions at pH 2, 5, 7 and 10 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
16
A
500
Electrostatic Electrostatic Electrostatic Electrostatic
400
Energy (kT)
300 200
Interaction pH 10 Interaction pH 7 Interaction pH 2 Interaction pH 5
100 0 -100 0
20
40
60
80
100
-200 -300 -400 -500
Distance (Å)
B LW Interaction Electrostatic Interaction pH 10 Electrostatic Interaction pH 7 Electrostatic Interaction pH 2 Electrostatic Interaction pH 5 Acid-base Interaction
500
Energy (kT)
400 300 200 100 0 -100 0 -200
20
40
60
80
100
-300 -400 -500 Distance (Å) C
200
Energy (kT)
100 0 -100
0
20
40
60
80
100
-200 -300 -400
Total Interaction pH 10 Total Interaction pH 7 Total Interaction pH 2 Total Interaction pH 5
-500 Distance (Å) Fig. 9. Chalcopyrite – L. ferrooxidans, interaction energy diagrams at 0.01 M KNO3, LW interactions calculated using A1 and 1 ȝm particle size. Electrostatic interactions at pH 2, 5, 7 and 10 (A), with reference to other interactions (B) and their influence on the total interaction energies (C).
17
3.4.4 Influence of pH The pH of solutions is not directly included as a variable in any equation, but it greatly influences the surface potential of the particles and therefore similar influence is expected on electrostatic force. L. ferrooxidans cells are positive below pH 3 and above which they were negative, whereas the minerals are charged positively until pH 7 and negative at higher pH values. This suggests an attractive electrostatic force between pH 3 and pH 7 and outside this pH range, the interaction force is repulsive. The magnitude of zeta potential for pyrite was higher and thus the electrostatic forces for pyrite system were higher compared to chalcopyrite. At pH 5 and 7 the potentials were similar and accordingly the electrostatic interaction curves were nearly the same. The cells and pyrite were repelling each other at pH 2 and pH 10 due to identical surface potentials (Fig. 8). Because of similar character of zeta potential curves of both minerals (Fig. 3), the electrostatic interactions were also similar for chalcopyrite, but the force was lower (Fig. 9). The total force for both systems was thus attractive at pH 5 and 7 at all separation distances and at pH 2 and pH 10 only at closer distances of 2-3 nm (20-30 Å) after overcoming the energetic barrier. The adsorption tests were carried out at pH 2, where an energetic barrier exhibits at a distance above 2.5 nm (25 Å) for pyrite and 3 nm (30 Å) for chalcopyrite. The interacting cells and mineral particles have to overcome this energy barrier to reach a distance where the total force is attractive for both minerals due to strong attractive AB interactions. Because of lower surface potential of chalcopyrite the energetic barrier was lower and the adhesion of cells to chalcopyrite was seen to commence at a lower cell concentration than pyrite (Fig. 2). 3.4.5 Acid-base interactions Acid-base interactions depend on the electron donating and accepting characteristics of the surfaces and they have been included in the calculation of free energy of adhesion. The acidbase part of free energy of adhesion was negative and high for both systems, due to different character of cells and minerals surface. Cells were strong electron donors while the minerals were capable to accept electrons resulting in strong interaction forces at closer distances. Acid-base interactions were the strongest among all the interactive forces and therefore the decisive forces and were attractive in all cases. As a result of higher electron accepting character of pyrite, the acid-base interaction energies between cells and pyrite were perceived to be higher. 4. Conclusions The adhesion of L. ferrooxidans onto pyrite and chalcopyrite has been examined by surface thermodynamics and extended DLVO theory methods and compared with experimental observations. At pH 2, where the adsorption tests were carried out, the cells and minerals were positively charged resulting to repulsive electrostatic forces. However, the surface potential of chalcopyrite was lower, thereby causing a lower repulsive electrostatic force relative to pyrite particles that contributed to the total energetic force balance to a lower energetic barrier between the cells and chalcopyrite particles. Lower repulsive forces in cooperation with the possibility of higher occurrence of iron rich dislocation sites at the surface of chalcopyrite might have resulted in higher adsorption densities onto chalcopyrite, compared with pyrite. However this result is contrary to the previous studies with Thiobaccillus ferrooxidans species where the cells adhesion on pyrite was found to be greater than chalcopyrite, but the extended DLVO theory confirmed and explained the present experimental results. For achieving selective adhesion onto pyrite, experimental conditions
18
need to be at a pH value where the surface potential of pyrite is higher relative to chalcopyrite and the cells are negatively charged. Above pH 4.5, pyrite exhibited higher zeta-potential than chalcopyrite and the cells are beyond their IEP, resulting in stronger attractive electrostatic force toward pyrite and higher adsorption densities. These results are valid only for the Fe2+ ions grown L. ferrooxidans. Insufficient interacting forces could be increased by manipulating the electrolyte concentration or using coarser particle sizes. Alternatively, with a change in the culture growth conditions or in the presence of mineral substrate, the surface properties of the bacteria are expected to change and represent a possible way to regulate the adhesion forces. Acknowledgements The financial support from the EU BioMinE project (contract no. IP NMP2-CT-2005-500329) is gratefully acknowledged. References [1] A. Das, K.H. Rao, P. Sharma, K.A. Natarajan, K.S.E. Forssberg, in: A. Amils, A. Ballester, (Eds.), Elsevier, Amsterdam, 1999, p. 697. [2] P. Sharma, A. Das, K.H. Rao, K.S.E. Forssberg, in: B.K. Parekh, J.D. Miller (Eds.), SME, Denver, 1999, p.147. [3] P.K. Sharma, K.H. Rao, K.A. Natarajan, K.S.E. Forssberg, in: P. Massacci, (ed.), Elsevier, Amsterdam, 2000, p. B8a-94. [4] M.N. Chandraprabha, K.A. Natarajan, J.M. Modak, Colloids Surf., B, 37 (2004) 93. [5] M.N. Chandraprabha, K.A. Natarajan, P. Somasundaran, Int. J. Miner. Process., 75 (2005) 113. [6] T. Nagaoka, N. Ohmura, H. Saiki, Appl. Environ. Microbiol., 65 (1999) 3588. [7] K.A. Natarajan, A. Das, Int. J. Miner. Process., 72 (2003) 189. [8] D. Santhiya, S. Subramanian, K. A. Natarajan, K. H. Rao, K. S. E. Forssberg, Int. J. Miner. Process., 62 (2001) 121. [9] P. Devasia, K.A. Natarajan, D.N. Sathyanarayana, G. Ramananda Rao, Appl. Environ. Microbiol., 59 (1993) 4051. [10] A. Vilinska, K. Hanumantha Rao, Open Colloid Sci. J., 2 (2009) 1. [11] D.R. Absolom, F.V. Lamberti, Z. Policova, W. Zingg, C.J. van Oss, A.W. Neumann, Appl. Environ. Microbiol., 46 (1983) 90. [12] H.J. Busscher, A.H. Weerkamp, H.C. van der Mei, A W. van Pelt, H.P. de Jong, J. Arends, Appl. Environ. Microbiol., 48 (1984) 980. [13] C.J. van Oss, M.K. Chaudhury, R.J. Good, Adv. Colloid Interface Sci., 28 (1987) 35. [14] C.J. van Oss, M.K. Chaudhury, R.J. Good, Chem. Rev., 88 (1988) 927. [15] C.J. van Oss, Colloids Surf., A, 78 (1993) 1. [16] C.J. van Oss, Colloids Surf., B, 5 (1995) 91. [17] P.K. Sharma, K.H. Rao, Adv. Colloid Interface Sci., 98 (2002) 341.
19
[18] P.K. Sharma, K.H. Rao, Colloids Surf., B, 29 (2003) 29. [19] B.V. Deryagin, L. Landau, Acta Physicochem., 55 (1941) 333. [20] E.J. Verwey, J.T.G. Overbeek, J. Colloid Sci., 10 (1955) 224. [21] C.J. van Oss, R.J. Good, M.K. Chaudhury, J. Colloid Interface Sci., 111 (1986) 378. [22] F.M. Fowkes, Ind. Eng. Chem., 56 (1964) 40. [23] R.B. Herbert, Clays Clay Miner., 45 (1997) 261. [24] M. A. Ghauri, N. Okibe, D. B. Johnson, Hydrometallurgy, 85 (2007) 72. [25] T. Rohwerder, T. Gehrke, K. Kinzler, W. Sand, Appl. Environ. Microbiol., 63 (2003) 239. [26] T. Gehrke, R. Hallmann, K. Kinzler, W. Sand, Appl. Environ. Microbiol., 64 (1998) 2743. [27] D.F. Evans, H. Wennerström, The Colloidal Domain Where Physics, Chemistry, Biology and Technology Meet, Wiley-VCH, New York, 1994. [28] M.N. Bellon-Fontaine, N. Mozes, H.C. van der Mei, J. Sjollema, O. Cerf, P.G. Rouxhet, H.J. Busscher, Cell Biophys., 17 (1990) 93. [29] G. Ström, M. Frederiksson, P. Stenius, J. Colloid Interface Sci., 119 (1987) 352.
20
Paper VI Selective coagulation in chalcopyrite/pyrite mineral system using Acidithiobacillus group bacteria. Vilinska, A., Hanumantha Rao, K., Forssberg, K.S.E., 2007. Advanced Materials Research. 20-21, pp. 366-370
Advanced Materials Research Vols. 20-21 (2007) pp. 366-370 online at http://www.scientific.net © (2007) Trans Tech Publications, Switzerland
Selective coagulation in chalcopyrite/pyrite mineral system using Acidithiobacillus group bacteria A. Vilinskaa, K. Hanumantha Raob and K.S.E. Forssbergc Division of Mineral Processing, Luleå University of Technology, SE-971 87 LULEÅ, Sweden a
[email protected],
[email protected],
[email protected]
Keywords: Sulphide Minerals, Acidithiobacillus ferrooxidans, Zeta-potential, FT-IR, Coagulation
Abstract. Acidithiobacillus ferrooxidans cells grown in ferrous ions were used to study the surface modification of pyrite and chalcopyrite, with focus on coagulation of very fine particles (-5 µm). The zeta-potential studies of the minerals, before and after bacterial treatment, showed that the cells have a distinct influence on the surface charge of pyrite and chalcopyrite. The maximum coagulation of particles determined by Turbiscan as a function of pH correlated well with the zetapotential results. Using diffuse reflectance FT-IR spectroscopic studies, the adhesion of cells showed a varied influence on these minerals. The results demonstrate that Acidithiobacillus ferrooxidans interact with pyrite and chalcopyrite differently, allowing selective coagulation of one mineral from the other under different pH conditions. Introduction Electrophoretic or zeta-potential measurements performed on A. ferrooxidans cells presented usually IEP (isoelectric point) around pH 2 or pH 2.3 for ferrous iron grown cells [1-9]. The magnitude of negative potential is usually around 10 mV [5,7]. Cells grown in the presence of minerals exhibited higher IEP, from pH 3 to 3.5, and higher magnitude of potential [3-5,7]. Changes of isoelectric points of pyrite and chalcopyrite depend on their original value. Pyrite and chalcopyrite with low original IEP (around pH 2-3) move their IEP after bacterial treatment to higher pH values (around pH 4). The change is time dependent, and the electronegative character of minerals is reduced [1,2,8-10]. Minerals with higher measured IEP (around pH 5-7) shift their IEP after cell conditioning to lower pH values (around pH 4) [3-5,7]. Successful selective bioflocculation of artificial mixtures of pyrite and nonsulphide minerals is possible with application of A. ferrooxidans [1], and also selective flocculation of pyrite from arsenopyrite mixture was reported [8]. Most attention has focused on reducing pyrite flotation by adding of A. ferrooxidans cells together with collector [3,5-11]. At the same time flotation of chalcopyrite is almost unaffected [35,7,9,10]. Selective depression of pyrite was achieved also in differential flotation studies with artificial mixtures of pyrite and chalcopyrite [9]. During bench scale experiments, flotation of pyrite was also depressed [12]. Experimental Methods Mineral Samples. Pure mineral samples of chalcopyrite and pyrite were ground to a -5 µm particle size, which was suitable for zeta-potential and coagulation measurements. Bacterial Cultivation. Bacterial strains of Acidithiobacillus ferrooxidans were cultured in 9K medium with following composition: 44,5 g/l of FeSO4.7H2O, 3 g/l (NH4)2SO4, 0,5 g/l MgSO4.7H2O, 0,5 g/l K2HPO4 and 0,1 g/l KCl. The medium was maintained at pH 2 with H2SO4. The collected cell mass was filtered and thoroughly washed with pH 2 water. Coagulation Measurements. Settling tests were performed using Turbiscan LAb and Tlab EXPERT 1.13 software. The mineral concentration was 5 g/l and the bacterial concentration was 2x108 cells/ml, representing 4x1010 cells/g of mineral. Measurements were performed under different pH conditions. The temperature remained constant at 30°C, maintained by the instrument. All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of the publisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 130.240.59.55-03/08/07,18:20:27)
Advanced Materials Research Vols. 20-21
367
Zeta-potential Measurements. Zeta-potential measurements were determined with ZetaCOMPACTTM. Pyrite and chalcopyrite particles at concentration of 0,025 g/100 ml were used. Ionic strengths of 10-2 M and 10-3 M were maintained with KNO3. A cell concentration of 2.5x107 cells/ml was found to be the most suitable for the instrument. The minerals were pretreated with cells for 30 minutes; this time was chosen with respect to previous works regarding cell adhesion [1,3,5,8]. A cell concentration of 0.7x107 cells/ml represented 2x1010 cells/g of mineral. FT-IR Measurements. The infrared spectra of Fe2+ grown cells and minerals conditioned with different amount of cells were recorded using a Perkin-Elmer FT-IR spectrometer. Results Coagulation Studies. Turbiscan instrument scans the glass tube filled with suspension, with two sensors - one receiving light transmitted through the sample and the second receiving light backscattered by the sample. This enables the detection of particle migration, and analysis of particle size change with time. Fig. 1 illustrates that untreated pyrite particles settle best around pH 6; below pH 4 and above pH 9 the suspension is stable with particles settling slowly or not at all. Pyrite treated with cells has a settling peak at pH 5; the suspension is stable below pH 3 and above pH 8. There is a visible shift of the peak and the curve to the more acidic area after bacterial treatment. Pyrite
60 min
100 80
Pyrite w ith cells 120
30 min
15 min
100
15 min
80
5 min
5 min
60 40 20 0
60 40 20 0
4
6
pH
8
10
Chalcopyrite 120 100
12
6
pH 8
10
Chalcopyrite w ith cells
12
120
60 min
30 min
100
30 min
80
15 min
60
5 min
5 min
60
4
60 min
15 min
80
2
40 20
particle size (µm)
2
particle size (µm)
60 min
30 min particle size (µm)
particle size (µm)
120
40 20 0
0 2
4
6
pH
8
10
12
2
4
6
8
10
12
pH
Fig. 1. Particle size increase with time for pure minerals (pyrite and chalcopyrite) and mineral-cell systems with respect to pH
Untreated chalcopyrite has the best settling behavior around pH of 6.5, similar to the untreated pyrite; below pH 5.5 and over pH 8.8 the suspension is stable. Treated chalcopyrite shows maximum settling around pH 8.5 with a shift of 2 pH units toward the alkaline direction. The treated suspension is stable below pH 4.5 and over pH 10. The bacterial treatment therefore increases the pH range under which the suspension is able to settle.
368
Biohydrometallurgy: From the Single Cell to the Environment
Zeta-potential Studies. Zeta-potential of A. ferrooxidans cells alone remains slightly negative within the measured pH range with no significant differences observed between 10-2 M and 10-3M ionic strength (Fig. 2). The magnitude and shape of the zeta-potential curve is similar to previously published results for Fe2+ grown A. ferrooxidans [5,7]. The zeta-potential of pure pyrite and A. ferrooxidans treated particles are nearly the same, with the two curves overlapping and following each other. However, the isoelectric point is clearly different. It moved approximately one pH unit in the acidic direction with the bacterial cells decreasing the isoelectric point of pyrite. A
-5 -10 -15 0,001M
-20
0,01M 3
5
pH
7
9
20 10 0 -10 -20
-25 1
B
30 Zeta potential (mV)
Zeta potential (mV)
0
11
-30
1
3
5
7
9
11
Pyrite w ith cells Pyrite Calcopyrite w ith cells Chalcopyrite pH
Fig. 2. Zeta-potential of A. ferrooxidans cells (A), pyrite and chalcopyrite particles and mineral-cell systems (B) with respect to pH
A. ferrooxidans cells decreased the zeta-potential of chalcopyrite particles below pH 6. A ferrooxidans treatment of chalcopyrite, shifted the isoelectric point from pH 7.4 to pH 9. FT-IR Studies. FT-IR spectrum of Fe2+ grown cells of A. ferrooxidans showed several peaks typical for organic substances such as -OH group of alcohols and water, hydrocarbons (=C-H, aliphatic –CH3 and CH2 , RO-CH3) and nitrogenous substances. Cell-treated pyrite showed some additional peaks compared with pure pyrite and some peaks become less intensive. Small peaks characteristic of CH3 and CH2 groups appeared. N compound peaks also changed. IR spectra of cell treated and untreated chalcopyrite were difficult to distinguish from each other, because the wave numbers and intensities of the peaks were the same. There were no new peaks after treatment, but some peaks were identical to those observed with pyrite. A change in characteristic groups on pyrite surface was noted, but the change on chalcopyrite was hardly observable. Previously published work also reported a smaller affect of cells on chalcopyrite compared with pyrite [5]. Discussion Particles of pyrite and chalcopyrite react differently to bacterial treatment, and their surface properties change in different ways. The IEP is shifted to the acidic region in case of pyrite and toward that alkaline direction for chalcopyrite. Pyrite FT-IR spectra changed significantly after bacterial treatment, while the spectra of chalcopyrite were almost unaltered after the treatment. Particle size measurements showed selective reaction of pyrite and chalcopyrite with bacterial treatment; and the results of zeta-potential measurements confirmed this. The isoeletric points of pyrite and chalcopyrite moved in opposite direction upon treatment with A. ferrooxidans, altering their original similar coagulation points from about pH 6-6.5, to two different points with greater than 3 pH units difference. After bacterial treatment chalcopyrite coagulated best around pH 8.5, whereas the pyrite suspension remained stable, indicating selective coagulation of chalcopyrite.
Advanced Materials Research Vols. 20-21
369
Fig. 3. Fe2+ grown A. ferrooxidans FT-IR absorbance spectrum
Fig. 4. FT-IR absorbance spectra of the minerals and minerals treated with A. ferrooxidans cells in different concentrations
370
Biohydrometallurgy: From the Single Cell to the Environment
Acknowledgement The financial support from the EU BioMinE project (contract no. IP NMP2-CT-2005-500329) is gratefully acknowledged. References [1] [2]
K.A. Natarjan, A. Das: Int. J. Mineral Processing Vol. 72 (2003), p. 189 P. Devasia, K.A. Natarjan, D.N. Sathyanarayana, G. Ramananda Rao: Applied and Environmental Microbiology Vol. 59 (1993), p. 4051 [3] A. Das, K.H. Rao, P. Sharma, K.A. Natarjan, K.S.E. Forssberg, in: Biohydrometalurgy and the environment toward the mining of the 21st century, edited by R. Amils and A. Ballester, Part A, Elsevier (1999), p. 697 [4] P. Sharma, A. Das, K.H. Rao, K.S.E. Forssberg: Hydrometalurgy Vol. 71 (2003), p. 285 [5] P. Sharma, A. Das, K.H. Rao, K.S.E. Forssberg in: Advances in Flotation Technology, edited by B.K. Parekh and J.D. Miller, SME (1999), p. 147 [6] M. Mishra, K. Bukka, S. Chen: Minerals Engineering Vol. 9 (1996), p. 157 [7] P.K. Sharma, K.H. Rao, K.A. Natarjan, K.S.E. Forssberg in: Proc. XXI International Mineral Processing Congress, edited by P. Massacci, Elsevier (2000), p. B8a-94 [8] M.N. Chandraprabha, K.A. Natarjan, P. Somasundaran: Journal of Colloid and Interface Science Vol. 276 (2004), p. 323 [9] M.N. Chandraprabha, K.A. Natarjan, J.M. Modak: Colloids and Surfaces B: Biointerfaces Vol. 37 (2004), p. 93 [10] M.N. Chandraprabha, K.A. Natarjan, P. Somasundaran: Int. J. Miner. Process Vol. 75 (2005), p. 113 [11] T. Nagaoka, N. Ohmura, H. Saiki: Applied and Environmental Microbiology Vol. 65 (1999), p. 3588 [12] T.R. Hosseini, M. Kolahdoozan, Y.S.M. Tabatabaei, M. Oliazadeh, M. Noaparast, A. Eslami, Z. Manafi, A. Alfantazi: Minerals Engineering Vol. 18 (2005), p. 371
Paper VII Leptosririllum ferrooxidans-sulfide mineral interactions with reference to bioflotation and bioflocculation. Vilinska, A., Hanumantha Rao, K., 2008. Transactions of Nonferrous Metals Society of China. 18, pp. 1403-1409.
Leptosririllum ferrooxidans-sulfide mineral interactions with reference to bioflotation nad bioflocculation A. VILINSKA, K. HANUMANTHA RAO Division of Mineral Processing, Luleå University of Technology, SE-971 87 LULEÅ, Sweden Received 20 September 2008; accepted 5 November 2008 Abstract: The adhesion of ferrous ions grown Leptospirillum ferrooxidans cells on pyrite and chalcopyrite minerals was investigated through adsorption, Zeta-potential and diffuse reflectance FT-IR measurements. The influence of bacterial species on minerals floatability was determined by Hallimond flotation tests while the flocculation behaviour was examined by Turbiscan measurements. The minerals iso-electric point (pH 6.5í7.5) after interaction with bacterial cells shifted towards cells iso-electric point (pH 3.3), indicating the chemical nature of cells adsorption on mineral surfaces. The FT-IR spectra of minerals treated with bacterial cells showed the presence of all the cell functional groups signifying cells adsorption. The bacterial cells adsorption on chalcopyrite was higher compared with pyrite, which agreed with cells greater depression effect on chalcopyrite flotation and pronounced flocculation behaviour in comparison with pyrite. Key words: bacteria; sulphide mineral; Zeta-potential; adhesion; FT-IR; flocculation; flotation
1 Introduction Recent developments in biotechnology have given promise of not only aiding hydrometallurgical operations but also providing means for bioremediation of environmental problems generated by the mineral industries. Many other uses of microorganisms are potentially possible. These include the use of microorganisms in flocculation and flotation of minerals, where the adhesion of bacterium and/or extracellular polymeric substances(EPS) to minerals induces a change in surface properties. The biomodification of mineral surfaces involves the complex action of microorganisms on the mineral surface. There are three different mechanisms by means of which the bio-modification can occur: 1) attachment of microbial cells to the solid substrate[1í3], 2) oxidation reactions[4] and 3) adsorption and/or chemical reaction with the metabolite products [5]. Several types of autotrophic and heterotrophic bacteria, fungi, yeast and algae have been tested in minerals biobeneficiation. However, the interfaces between biological and geological materials, as well as means to design and manipulate that interface is not unexplored. It is necessary to study the biotic interfaces focusing on the mineral side of the interface,
i.e., on response of chemical composition and structure, net charge/potential and wettability of the mineral surface to the bacterium presence. Chemolitotrophic bacterial strains with natural abundance in mineral environment and mine water can be selectively attached to sulphides, thereby essentially modifying the surfaces and the follow-up adsorption of flotation reagents. Leptospirillum ferrooxidans used in this study is a chemolitotrophic acidophilic bacterium which is isolated from acid mine drainage water[6í7]. This bacterium is capable of iron oxidation similar to Acidithiobacillus ferrooxidans but cannot able to oxidise sulphur. Both strains of microorganisms were successfully used for bioleaching of sulphide ores and concentrates[8í9] while A. ferrooxidans was applied in some studies regarding bioflotation and bioflocculation of sulphide minerals. Studies showed that the xanthate flotation of pyrite was greatly reduced by the application of A. ferrooxidans[10í17] whereas the effect on chalcopyrite was found to be marginal, thereby selective flotation of chalcopyrite from pyrite has been suggested. Flocculation and settling behaviour of sulphide minerals is also enhanced in the presence of A. ferrooxidans with negligible effect on non-sulphide minerals[18]. A. ferrooxidans cells with a natural affinity towards sulphide minerals, mostly pyrite, can positively influence
Corresponding author: K. HANUMANTHA RAO; Tel: +46-920-491705; E-mail:
[email protected]
1404
A. VILINSKA, et al/Trans. Nonferrous Met. Soc. China 18(2008)
their surface properties and enhance the separation process. L. ferrooxidans isolated from same sources as that of A. ferrooxidans and originally assumed to be A. ferrooxidans, lacks sulphur oxidation ability. Therefore, the mechanism of L. ferrooxidans cells adhesion to mineral surface and their influence on surfaces properties are perceived to be different. This work presents the results on surface chemical studies of the adhesion of L. ferrooxidans cells on pyrite and chalcopyrite, and the changes in surface properties caused by bacterial interaction are discussed with reference to flotation and flocculation behaviour of minerals.
2 Experimental 2.1 Minerals Pure natural pyrite and chalcopyrite crystals supplied by Gregory, Bottley and Lloyd, UK, were used in the present studies. The crystals were broken into small pieces and dry ground in an agate mortar. The product was sieved to obtain suitable size fractions of 38í106 ȝm for flotation experiments and ˘5 ȝm for zeta-potential, adsorption and FT-IR studies. Prepared samples were washed with HCl for cleaning the surface oxidised species and stored in a refrigerator until being used. The BET specific surface area of these coarse and fine samples was 0.06 m2/g and 1.02 m2/g for pyrite and, 0.17 m2/g and 1.90 m2/g for chalcopyrite, respectively. 2.2 Bacteria The bacterial strain Leptospirillum ferrooxidans DSM2391 was grown in a modified Leptospirillum HH medium (40 g/L FeSO4·7H2O, 0.132 g/L (NH4)2SO4, 53 mg MgCl2·6H2O, 27 mg KH2PO4, 0.147 g CaCl2·2H2O, 62 ȝg MnCl2·2H2O, 68 ȝg ZnCl2, 64 ȝg CoCl2·6H2O, 31 ȝg H3BO3, 10 ȝg Na2MoO4, 67 ȝg CuCl2·2H2O) at pH 1.8 and at 30 ć on a rotary shaker maintaining 150 r/min. Collected cell mass was filtered and washed with pH 2 water to obtain cells devoid of possible precipitates and metabolites. 2.3 Zeta potential measurements Zeta potential measurements were made with a Zeta Compact (Cad Instrumentation, France) equipped with video and image analysis system. The mineral concentration in the suspension was 0.25 g/L and a constant ionic strength of 0.01 mol/L was maintained with KNO3. Interaction time for cell-mineral system was 30 min and a cell concentration of 3h1010 mLí1 was used for both minerals. Measurements were performed as a function of pH adjusted with KOH and HNO3. 2.4 Adsorption measurements The ˘5 ȝm size fraction of minerals was used for
adsorption studies. The adsorption of bacterial cells on pyrite and chalcopyrite was carried out in 250 mL Erlenmeyer flasks. Tests were performed with 1 g of mineral in 100 mL water containing varied initial cell concentrations of 107, 5h107, 108, 5h108, 109 mLņ1. After 30 min of interaction, the cells in liquid phase were estimated using a Neubauer counter under a microscope. The amount of cells adsorbed on mineral surfaces was determined by the difference in the cells concentration in liquid phase before and after adsorption. 2.5 Diffuse reflectance FT-IR measurements The solid samples after bacterial cells adsorption measurements were filtered, air-dried and subjected to FTIR for recording the spectra. The spectra were obtained with a Perkin-Elmer 2000 spectrometer at 4 cmí1 resolution with a narrow band liquid N2-cooled MCT detector. The samples were prepared by dispersing the air-dried sample in KBr at a concentration of 2.5% for the cells and minerals. The absorbance units were defined by the decimal logarithm of the ratio of pure finely powdered KBr reflectance to the sample. 2.6 Settling measurements Settling tests were conducted by using Turbiscan ma2000 instrument, which scans the entire height of the sample cell and measures the stability and instability of suspensions. The detectors in Turbiscan receives the transmitted and backscattered light from a cylindrical sample cell every 40 ȝm on a maximum height of 55 mm. The profile obtained characterise the suspension homogeneity, particle concentration and mean diameter. A solid concentration of 2.5 g/L of ˘5 ȝm size particles of pyrite and chalcopyrite was used to study the sedimentation behaviour of particles. Tests were made at different pH values of the mineral suspensions after bacterial conditioning of 4h1010 cells/g. A blank experiment in the absence of bacteria was always carried out. 2.7 Flotation tests The single-mineral flotation tests were conducted in a Hallimond tube using 1 g of either pyrite or chalcopyrite of 38í106 ȝm size fraction. The mineral samples were first conditioned with a predetermined Leptospirillum ferrooxidans cells concentration in 100 mL of water at a specified pH for 30 min. Then potassium isopropyl xanthate collector was added and the sample was conditioned further for 5 min. The entire solution was transferred into a Hallimond tube and floated for 1 min. An airflow rate of 200 mL/min was applied during flotation. The influence of initial cell concentration on sulphides flotation was examined at 0.5h10í4 mol/L xanthate collector concentration.
A. VILINSKA, et al/Trans. Nonferrous Met. Soc. China 18(2008)
3 Results and discussion 3.1 Zeta potential studies The Zeta-potentials of pure Leptospirillum ferrooxidans cells and pyrite, and pyrite after interaction with cells as a function of pH are shown in Fig.1. Pure cells exhibited an iso-electric point(IEP) at pH 3.3 and beyond this pH the magnitude of negative potential increased with a rise in pH value. The magnitude of Zeta-potentials is relatively high compared to A. ferrooxidans cells grown under similar conditions [10í16, 18í19]. The presence of functional groups such as carboxyl, amino and hydroxyl, and their protonation and/or dissociation as a function of pH impart surface charge to the cells. The presence of ammonium containing polymers (proteins) and polysaccharides containing phosphate and/or carboxylic groups in the surface layers of bacteria and a charge balance of these anionic and cationic acid/base groups determine the IEP of bacterial cells. The infrared spectrum of cells showed the presence of these groups originating from cell wall components of lipo-polysaccharides, lipoprotein and bacterial surface proteins. The positive surface potential at low pH values is obviously due to the protonation of ammonium groups and the increase in negative charge beyond pH 3.3 is caused by the dissociation of anionic functional groups. The cells are thus expected to adsorb on mineral surface through electrostatic interactions and as well through specific chemical interaction of functional groups on surface metal ions besides metabolic reasons.
1405
factors such as origin/mineralogy, sample preparation, surface oxidation and aging in water. After interaction with bacterial cells for 30 min, the IEP shifted to pH 5 and the Zeta-potential below pH 3 and above pH 7 nearly corresponds to that of the cells potential. The pyrite Zeta-potential decreased in the entire pH region and the curve moved towards the pure cells Zeta-potential, indicating the adsorption of cells on pyrite surface. However, higher Zeta-potentials between pH 3 and 7 compared to cells illustrate that either the pyrite surface is not fully covered at the present cell concentration of 3 h1010 mLí1 or the presence of ferric ions on the surface due to bacterial oxidation of pyrite. The Zeta-potential of chalcopyrite before and after interaction with bacterial cells are presented in Fig.2. The IEP of chalcopyrite displayed at pH 6.5 which moved to pH 3 after interaction with cells, coinciding to the IEP of cells. The Zeta-potentials between pH 3 and 6 are close to zero, which is seen similar to pyrite potentials after interaction with cells. Beyond pH 6, the Zeta-potentials of chalcopyrite interacted with cells are similar to pure chalcopyrite potentials. There is hardly any influence of cells on chalcopyrite potentials in the basic pH region. In general, the Zeta-potentials of mineral-bacteria system exhibit on the characteristic curves between pure cells and minerals [11í12,14,19í20].
Fig.2 Zeta-potentials of L. ferrooxidans cells and chalcopyrite, and chalcopyrite treated with cells as function of pH
Fig.1 Zeta-potentials of L. ferrooxidans cells and pyrite, and pyrite treated with cells as function of pH
Pure pyrite exhibited an IEP at pH 7.5 and above this pH the negative potential of mineral increased with increase in pH. The reported IEP of pyrite as determined by electrophoresis fall between pH 3.5 and 7.5 [10í12, 14í18], and this variation could arise from several
3.2 Adsorption studies The Leptospirillum ferrooxidans cells adhesion on pyrite and chalcopyrite were carried out to determine the extent of cells adsorption on mineral surfaces. The adsorption isotherms of cells for pyrite and chalcopyrite at pH 4 are shown in Fig.3. The adsorption density increased with increasing equilibrium concentration of cells. At the beginning, when the initial cell concentration was 109 mLí1, nearly all the cells on pyrite and chalcopyrite are adsorbed corresponding to 80.6% and 99.9% of the initial cells, respectively. The isotherms
1406
A. VILINSKA, et al/Trans. Nonferrous Met. Soc. China 18(2008)
display a linear increase in adsorption density from an equilibrium cells density of 104 mLí1 and there was no levelling off adsorption within the cell concentration range studied. While considering the geometrical cell dimensions of 1.65 ȝm in length and 0.35 ȝm in breadth, the cells adsorption density for full surface coverage in horizontal orientation corresponds to 1h1012 mí2. This approximate calculation reveals that the adsorption of cells at the highest adsorption data point in the isotherms of either mineral equals just 10% of the surface coverage. Adsorption measurements beyond 1h108 mLí1 equilibrium cells concentration are thus needed to establish the saturation level of cells adsorption on pyrite and chalcopyrite.
chains, respectively. The bands at 3 290, 3 070, 1 650 and 1 539 cmí1 represent the stretching and bending vibrations of üNH groups of proteins (amideĉand Ċ). The carbonyl absorption of carboxylic groups is identified with the bands at 1 730 and 1 712 cmí1[23]. These absorbance bands are comparable to the bands in the spectrum of Acidithiobacillus ferrooxidans cells [10í11,14,24], illustrating similar surface chemical composition of these two Thiobacillus group bacterial cells.
Fig.4 Diffuse reflectance FT-IR spectrum of Leptospirillum ferrooxidans cells
Fig.3 Adsorption isotherms of L. ferrooxidans cells on pyrite and chalcopyrite
The adsorption density of A. ferrooxidans cells on pyrite was found to be faster and higher compared to chalcopyrite[10,12,15í18]. It was also reported that the same species of bacterial cells arising from different strains had differences in adsorption behaviour[21]. Although the present adsorption isotherms are limited to lower surface coverage, L. ferrooxidans cells adsorption on chalcopyrite was higher than pyrite at any of the equilibrium cells concentration. Thiobacillus group bacteria are known to specifically adsorb on surface defects and imperfections[22]. Since chalcopyrite surface area is nearly twice that of pyrite, it is presumed to contain higher surface imperfections than pyrite and therefore higher adsorption of cells on chalcopyrite. 3.3 FT-IR studies The diffuse reflectance FT-IR spectrum of Leptospirillum ferrooxidans cells depicts several absorbance bands (Fig.4) composing protein, lipid, extracellular polysacchrides and nucleic acids. The absorbance bands at 2 959, 2 925 and 2 856 cmí1 characterise asymmetric CH3 stretching, asymmetric CH2 stretching and symmetric CH2 stretching of hydrocarbon
The spectra of pyrite and chalcopyrite treated with bacterial cells showed that with increasing cells concentration the presence of cells absorbance bands increased in intensity. Typical spectra in the case of pyrite are shown in Fig.5. These spectra characterise the adsorption of cells on pyrite with the presence of majority of the bacterial cells absorbance bands where the intensity of peaks is dependent on cell concentration. The spectra of chalcopyrite treated with cells are also
Fig.5 Diffuse reflectance FT-IR spectra of pyrite treated with increasing bacterial cell concentration
A. VILINSKA, et al/Trans. Nonferrous Met. Soc. China 18(2008)
1407
similar, bearing most of the cells absorbance bands. However, the polysaccharide absorbance bands observed on pyrite are missing in the case of chalcopyrite, probably masked due to strong absorption character of chalcopyrite. 3.4 Flocculation and flotation studies Transmission diagrams of the percentage of light transmitted through the pyrite and chalcopyrite suspensions in the presence and absence of L. ferrooxidans cells are shown in Figs.6 and 7 respectively. The light beam cannot transmit through a well dispersed suspension and therefore zero transmission represents stable suspension. When sedimentation occurs, the lower amount of dispersed particles allows the beam to pass through the sample and the transmittance increases. The suspensions were scanned covering the entire height of the sample holder for 10 min with one scan per minute. The diagrams show the state of stability or instability of suspensions with time and the higher transmittance in the presence of L. ferrooxidans indicates the sedimentation of particles. The adhesion of cells on minerals caused flocculation of both minerals at all pH values measured, ranging from acidic to basic. The cell surface polymers
Fig.7 Transmission data of chalcopyrite suspensions for 10 min settling time
Fig.6 Transmission data of pyrite suspensions for 10 min settling time
are thought to be responsible to bridge the particles and to cause flocculation. The transmission data shown for pyrite were at pH 6.5 and there is some natural sedimentation. Since this pH is close to pyrite iso-electric point, the particles are coagulated to some extent. The presence of cells caused flocculation of pyrite particles and therefore sedimentation was enhanced. The data shown for chalcopyrite were at pH 8 and the suspension is very stable in the absence of bacterial cells while their presence multiplied the sedimentation of particles. The higher flocculation of chalcopyrite compared with pyrite corresponds to the adsorption data where the bacterial cells adsorption density on chalcopyrite was also higher. From the transmission data of Turbiscan, it is possible to evaluate the flocculation and settling phenomena of particles in the suspension. Accordingly, the recorded transmission data at a fixed time interval and fixed position for pyrite and chalcopyrite suspensions are shown in Fig.8. Bacterial cells addition increased the settling of both minerals, although the effect on pyrite is lower. The highest settling for cells treated pyrite was experienced at pH 5 and 6.5, and below pH 7 the suspension is 5í6 times more transparent in average, compared with untreated pyrite. Above pH 8
1408
A. VILINSKA, et al/Trans. Nonferrous Met. Soc. China 18(2008)
the pyrite suspensions alone and treated with cells are dispersed. Chalcopyrite exhibited higher effect of bacteria on settling behaviour and the particle size measured from transmission data (Fig.8) was found to increase significantly at all pH values studied. The increase of sedimentation is the highest at pH 8, where the settling can be considered to be 100%, due to the fact that the highest reachable transmission for this kind of experiment is about 50%í60%. When the suspension was stirred at high speed before the measurement, a further increase in the settling was observed. Pyrite flocculation and sedimentation are increased after A. ferrooxidans and A. thiooxidans addition has no effect on non-sulphide minerals[18] and moderate effect on other sulphide minerals[15,19í20].
Fig.9 Xanthate flotation of pyrite and chalcopyrite in presence of L. ferrooxidans cells
4 Conclusions
Fig.8 Transmission of pyrite and chalcopyrite suspensions as function of pH at fixed position after 10 min
Hallimond flotation tests were conducted according to previously established optimum reagents and pH conditions. The highest recovery of almost 100% for both minerals is obtained at pH 4 and 0.5h10í4 mol/L collector concentration, and these conditions were maintained for the cells interacted with pyrite and chalcopyrite. Flotation recoveries of both minerals are decreased (Fig.9) in the presence of cells but the depression of chalcopyrite is much higher than pyrite. A 95% chalcopyrite recovery is decreased to 25% at a cell concentration of 2.5h108 mLí1, while pyrite at the same conditions exhibited 67% recovery. The depression of minerals was found to depend on cell concentration. Several authors reported that flotation of pyrite interacted with A. ferrooxidans decreases while chalcopyrite still exhibits floatability at the same conditions; however, the cells adsorption on pyrite was found to be significantly higher [10,12,16í17] in these studies. CHANDRAPRABHA et al[16í17] observed that chalcopyrite keeps its flotation ability as a result of specific interaction between collector and mineral.
Biobeneficiation studies on pyrite-chalcopyrite flotation/flocculation system using L. ferrooxidans cells showed greater impact on chalcopyrite than pyrite. The higher amount of adsorbed cells on chalcopyrite surface explains this phenomenon. The higher cell adsorption density on chalcopyrite corroborates with higher depression of chalcopyrite flotation and also higher enhancement of its settling behaviour compared with pyrite. Higher affinity of L. ferrooxidans to chalcopyrite is assigned to its higher surface defects and higher accessibility of surface Fe as an exclusive energy source because of higher surface area compared with pyrite.
Acknowledgements The financial support from the EU BioMinE project (contract No. IP NMP2-CT-2005-500329) is gratefully acknowledged.
References [1]
[2]
[3]
[4]
[5]
ATTIA Y A, ELZEKY M. Enhanced separation of pyrite from oxidized coal by froth flotation using biosurface modification [J]. Int J Miner Process, 1993, 37: 61í71. BOTERO A E C, TOREM M L, MESQUITA L M S. Fundamental studies of rhodococcus opacus as a biocollector of calcite and magnesite [J]. Miner Eng, 2007, 20: 1026í1032. HOSSEINI T R, KOLAHDOOZAN M, TABATABAEI Y S M, OLIAZADEH M, NOAPARAST M, ESLAMI A, MANAFI Z, ALFANTAZI A. Bioflotation of Sarcheshmeh copper ore using Thiobacillus ferrooxidans bacteria [J]. Miner Eng, 2005, 18: 371í374. SANTHIYA D, SUBRAMANIAN S, NATARAJAN K A, HANUMANTHA RAO K, FORSSBERG K S E. Biomodulation of galena and sphalerite surfaces using Thiobacillus thiooxidans [J]. Int J Miner Process, 2001, 62: 121í141. NATARAJAN K A, DEO N. Role of bacterial interaction and bioreagents in iron ore flotation [J]. Int J Miner Process, 2001, 62: 143í157.
A. VILINSKA, et al/Trans. Nonferrous Met. Soc. China 18(2008) [6] [7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
MARKOSYAN G E. A new iron-oxidizing bacteriumüLeptospirillum ferrooxidans nov. gen. nov. sp [J]. Biol J Armenia, 1972, 25: 26í29. BALASHOVA V, VEDENINA I Y, MARKOSYAN G E, ZAVARZIN G A. The autotrophic growth of Leptospirillum ferrooxidans [J]. Mikrobiologiya, 1974, 43: 491í494. SAND W, ROHDE K, SOBOTKE B, ZENNECK C. Evaluation of Leptospirillum ferrooxidans for leaching [J]. Appl Environ Microbiol, 1992, 58(1): 85í92. GIAVENO A, LAVALLE L, CHIACCHIARINI P, DONATI E. Bioleaching of zinc from low-grade complex sulfide ores in an airlift by isolated Leptospirillum ferrooxidans [J]. Hydrometallurgy, 2007, 98(1/2): 117í126. DAS A, HANUMANTHA RAO K, SHARMA P K, NATARAJAN K A, FORSSBERG K S E. Surface chemical and adsorption studies using Thiobacillus ferrooxidans with reference to bacterial adhesion to sulfide minerals [C]// International Biohyrdometallurgy Symposium 1999, Biohydrometallurgy and the Environment toward the Mining of the 21st Century, part 9A. Madrid, 1999: 697í707. SHARMA P K, DAS A, HANUMANTHA RAO K, FORSSBERG K S E. Surface characterisation of Thiobacillus ferrooxidans cells grown under different conditions [J]. Hydrometallurgy, 2003, 71: 285í292. SHARMA P K, DAS A, HANUMANTHA RAO K, FORSSBERG K S E. Thiobacillus ferrooxidans interaction with sulphide minerals and selective chalcopyrite flotation from pyrite [C]// SME Annual Meeting, Advances in Flotation Technology. Denver, 1999: 147í165. MISHRA M, BUKKA K, CHEN S. The effect of growth medium of Thiobacillus ferrooxidans on pyrite flotation [J]. Minerals Engineering, 1996, 9: 157í168. SHARMA P K, HANUMANTHA RAO K, NATARAJAN K A, FORSSBERG K S E. Bioflotation of sulphide minerals in the presence of heterotrophic and chemolitotrophic bacteria [C]// Proceedings of the XXI International Mineral Processing Congress, B8a. Rome, 2000: 94í103. CHANDRAPRABHA M N, NATARAJAN K A, SOMASUNDARAN P. Selective separation of ersenopyrite from pyrite by biomodulation in the presence of Acidithiobacillus
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
1409
ferrooxidans [J]. Journal of Colloid and Interface Science, 2004, 276: 323í332. CHANDRAPRABHA M N, NATARAJAN K A, MODAK J M. Selective separation of pyrite and chalcopyrite by biomodulation [J]. Colloids nad Surfaces B: Biointerfaces, 2004, 37: 93í100. CHANDRAPRABHA M N, NATARAJAN K A, SOMASUNDARAN P. Selective separation of pyrite from chalcopyrite and arsenopyrite by biomodulation using Acidithiobacillus ferrooxidans [J]. Int J Miner Process, 2005, 75: 113í122. NATARAJAN K A, DAS A. Surface chemical studies on Acidithiobacillus group of bacteria with reference to mineral flocculation [J]. Int J Mineral Processing, 2003, 72: 189í198. DEVASIA P, NATARAJAN K A, SATHYANARAYANA D N, RAMANANDA RAO G. Surface chemistry of Thiobacillus ferrooxidans relevant to adhesion on mineral surfaces [J]. Applied and Environmental Microbiology, 1993, 59: 4051í4055. BLAKE R C, SHUTE E A, HOWARD G T. Solubilization of minerals by bacteria: Electrophoretic mobility of Thiobacillus ferrooxidans in the presence of iron, pyrite and sulfur [J]. Appl and Environmental Microbioogy, 1994, 60: 3349í3357. GHAURI M A, OKIBE N, JOHNSON D B. Attachment of acidophilic bacteria to solid surfaces: The significance of species and strain variations [J]. Hydrometallurgy, 2007, 85: 72í80. ROHWERDER T, GEHRKE T, KINZLER K, SAND W. Bioleaching review (part A): Progress in bioleaching: fundamentals and mechanisms of bacterial metal sulfide oxidation [J]. Appl Microbiol Biotechnol, 2003, 63: 239í248. COLTHUP N B, DALY L H, WIBERLEY S E E. Introduction to infrared and Raman spectroscopy [M]. Boston: Academic Press, 1990: 547. VILINSKA A, HANUMANTHA RAO K, FORSSBERG K S E. Selective coagulation in chalcopyrite/pyrite mineral system using Acidithiobacillus group bacteria [J]. Advanced Materials Research, 2007, 20/21: 366í370. (Edited by YANG Bing)
Paper VIII Biocoagulation and its Application Potentials for Mineral Bioprocessing. Halit Z. Kuyumcu, Tina Bielig, Annamaria Vilinska and K. Hanumantha Rao, 2009. The Open Mineral Processing Journal. 2, pp. 1-11.
The Open Mineral Processing Journal, 2009, 2, 1-11
1
Open Access
Biocoagulation and its Application Potentials for Mineral Bioprocessing Halit Z. Kuyumcu1,*, Tina Bielig1, Annamaria Vilinska2 and K. Hanumantha Rao2 1 Department of Mechanical Process Engineering & Solids Processing, Technical University of Berlin, Sekr. BH 11, Strasse des 17. Juni 135, D-10623 BERLIN, Germany, 2 Division of Mineral Processing, Department of Chemical Engineering & Geosciences, Luleå University of Technology, SE-971 87 LULEÅ, Sweden
Summary: The well-known sorting processes like density separation, separation in magnetic or electric fields and flotation, are not suitable to apply successfully within a particle-size range smaller than 10 m. Due to insufficient selectivity of above mentioned enrichment processes the concentrate recovery at this particle size range is extremely poor, which influences accordingly the techno-economic efficiency of mineral processing negative. Based on a process design idea, investigations confirm that the biocoagulation of microorganisms and solid particles can be used to generate coarser sized coagulates which are more suitable for sorting. Experimental investigations showed that microorganisms like Saccharomyces cerevisiae and Yarrowia lipolytica and sulphide particles like galena and sphalerite below 10 m coagulate effectively. Theoretical thermodynamic and extended DLVO theory calculations are in good agreement with microorganisms adhesion onto metal sulphides but not on silicates and selective biocoagulation of sulphides. Furthermore it has been demonstrated that flotation is suitable for the separation of the selectively formed biocoagulates.
1. INTRODUCTION The technical relevant sorting processes like density, magnetic or electrical separation and flotation, require a narrow particle-size range for a sufficient selectivity in order to eliminate overlapping effects. Because of the rapid decrease of the mass forces and increase of the surface energetic state with decreasing particle size the well-known sorting processes are not applicable to an effective separation of particle sizes smaller than 10 m. Moreover, the necessity of sorting processes for finely dispersed solid systems is increasing. Additionally, new sorting processes are useful in the cleaning of wastewater to remove a variety of suspended particles [1]. For obtaining a selective enrichment and recovery of sorting products in that particle size range, usually chemical dissolution or leaching processes are used. Bioleaching methods to treat copper, uranium, zinc and refractory gold ores, complex sulphides, manganese ores and industrial minerals have been developed and those processes have been adopted by the industry. The major hurdle for wider acceptance of these processes is their very slow rate of the biooxidation step of minerals. Among the several ideas on the mechanism of biooxidation, one possible explanation is that the adhesion of the bacteria on the minerals induces the oxidation process. Mechanisms and biochemical fundamentals of bacterial metal sulphide oxidation has been reported by Sand et al. [2].
*Address correspondence to this author at the Department of Mechanical Process Engineering & Solids Processing, Technical University of Berlin, Sekr. BH 11, Strasse des 17. Juni 135, D-10623 BERLIN, Germany; E-mail:
[email protected] 1874-8414/09
Generally, the possibilities of material separation by using microorganisms can be explained based on several microbially-mediated transformation mechanisms for dissolved metals and dispersed solid particulate materials. The mechanisms are exemplified in Table 1 and Fig. (1). Biosorption and bioaccumulation are based on interactions of microbial cells and soluble metals and metalloids. Microbial cell surface and organelles offer a large number of active functional groups and possible physicochemical mechanisms of interaction. The result is the immobilization of the metals and metalloids on the microbial biomass. Sometimes the binding is reversible. Biosorption is defined as selective or non selective sequestering of dissolved metals by microbial cells and refers mostly to passive physicochemical mechanisms of inactive (non-metabolizing) metal uptake by microbial biomass. Various mechanisms may occur, e.g. complexation, coordination, chelation, ion exchange, adsorption, microprecipitation and reductionoxidation [3]. Bioaccumulation is referred to as active absorption (as opposed to the passive physicochemical adsorption-retention assigned to biosorption). Passive adsorption is rapid and independent of the presence of specific nutrients, whereas active absorption is slow and nutrient dependent. Bioaccumulation of metals into a cell generally requires metabolic energy, enzymatic activity and specific transport systems to move material through a cell membrane. It also depends on the tolerance of the microorganisms to the concentration of possibly toxic elements in the intracellular cytoplasm and other subcellular components but may be dependent or independent of metabolism [4]. 2009 Bentham Open
2 The Open Mineral Processing Journal, 2009, Volume 2
Table 1.
Kuyumcu et al.
Mechanisms of Microbially-Mediated Transformations
Biosorption
Bioaccumulation
Biocoagulation
exocellular or pericellular binding of metals
intracellulare uptake of metals
adhesion on surface
Mechanisms:
Active Transport requires:
Thermodynamic approach
Ion exchange
specific membrane transport systems
Complexation
Chelation
metabolic energy
DLVO approach
Coordination
enzymatic activity
Lifshitz-van der Waals attractive forces
Precipitation
Adsorption-Retention
electrostatic forces
Reduction-Oxidation
acid/base interaction forces
interfacial free energy
x+ x+
dissolved metals x+ x+
x+ x+
Biosorption
Bioaccumulation
exocellular or pericellular binding of metals
intracellulare uptake of metals
microbial cell Biocoagulation adhesion on surface
mineral particles
Fig. (1). Microbially-mediated transformations.
1.1. Bacteria - Mineral Interactions Bacterial adhesion on mineral substrates can be assessed by surface thermodynamics and extended DLVO theory of Lifshitz-van der Waals, electrostatic and acid-base interactions [5-7]. According to thermodynamics, mineral particles and bacterial cells dispersed in a liquid medium change to an adhered system (= coagulation) if such a change leads to a lower energy state. The free energy of adhesion (Gadh) can be expressed by
Gadh = bs bl sl
(1)
where represents interfacial free energies for bacteriasolid (bs), bacteria-liquid (bl) and solid-liquid (sl) [5]. For the calculation of the free energy of adhesion (Gadh), the Lifshitz-van der Waals component (LW) and the acid-base component (AB) have to be considered: LW
AB
Gadh = Gadh + Gadh .
(2)
Calculation according to LW-AB approach using the Lifshitz-van der Waals dispersive and acid-base components of surface energies is: LW
Gadh = 2
(
LW bv
LW
lv
)(
LW sv
lv
LW
+
+
)
(3)
and
(
AB
+
+
)(
sv
lv
Gadh = 2 bv sv
(
+
+
2 sv lv
)(
bv
)
) (
sv 2 bv lv
)(
bv
lv
) (4)
Classical DLVO approach [8, 9] includes Lifshitz-van der Waals (LW) interactions, and electrostatic interactions. LW forces are always attractive and strong at shorter distances between neutral stable molecules. Coulombic electrostatic interaction forces could be attractive or repulsive depending on the surface charge of interacting particles.
Biocoagulation and its Application Potentials for Mineral Bioprocessing
The Open Mineral Processing Journal, 2009, Volume 2
Mineral particles
3
Bacteria cells dispersed in liquids
Adhesion = Biocoagulation selective/nonselective
Biooxidation
Bioleaching
Bioflocculation
Bioflotation
Sorting
Fig. (2). Mineral Bioprocessing of dispersed systems based on biocoagulation.
Acid-base interactions were added later by van Oss [10] to involve the electron donating-accepting abilities of different materials. Thus microbial adhesion to solid surfaces can be described by a sum of van der Waals, electrostatic and acid-base forces as a function of separation distance Gtotal = GLW + G EL + G AB .
(5)
Calculation of these forces depends on the geometry of interacting phases and for sphere-sphere system, the following equations were used: Lifshitz-van der Waals interaction energy: x 2 + xy + x y y + 2 + 2 ln 2
2 x + xy + x + y 12 x + xy + x x + xy + x + y
1 2 H
1 + 1.77 (6) G
LW
=
A
Electrostatic interaction energy: G
EL
=
(
)
H 1+ e 2 H ) (7)
+ ln 1 e H + ln (1 e 1 2
a1a2 1 + 2 2 1 2 a1 + a2
Acid-base interaction energy: GAB = aGadhABe[(do-H)/ ]
(8)
where H - separation distance, a - radius of solid particle, - zeta-potential, - inverse of double layer thickness, A - Hamaker constant, d0 - minimum separation distance between 2 surfaces (0.157 nm), - correlation length of molecules in liquid (0.6 nm) and x = H/(a1 + a2), y = a1/a2 For interaction energy calculation, the parameters such as zeta-potential, particle radius, double layer thickness are known or measurable, while for calculation of Hamaker constant there are two different methods available. The Hamaker constant value that influences the Lifshitz-van der Waals interaction energy could be obtained by microscopic and macroscopic approaches [11, 12], but according to our previous studies [13] the macroscopic approach was found to be more relevant. Macroscopic method [12] determines the Hamaker constant using the Lifshitz-van der Waals component of free energy of adhesion GadhLW:
A = -12 d02 GadhLW .
(9)
1.2. Bio-separation Processes Based on bacterial adhesion mechanism, several separation processes can be achieved and potentially used for bioprocessing, if the process can be performed selectively, refer to [14, 15]. The most interesting technological potentials of the mineral bioprocessing of dispersed systems are shown in Fig. (2). Several approaches to separate solid particles with microbes were investigated in the last decades. The possibility of using bacteria and fungi as flocculants for Florida phosphatic clays was investigated early in 1963 [16]. Flocculation of phosphate slimes, hematite and coal by using Mycobacterium phlei was tested in 1991 [17]. Mycobacterium phlei produce extracellular polymers and surfactants under certain conditions, which can cause flocculation of the microorganisms themselves or of other solids, as exemplarily reported by Long et al. [18]. Several studies have shown the potential of microorganisms to be used as flotation collectors or modifiers. The depression of pyrite in coal flotation has been reported by Capes et al. [19]. The selective separation of sphalerite from pyrite has been studied in the presence of Acidobacillus ferrooxidans [20]. The surface energies of 140 bacterial and seven yeast cell surfaces following Fowkes, Equation of state and Geometric mean and Lifshitz-van der Waals acid-base (LW-AB) approaches have been extensively studied and evaluated by Rao et al. [21]. The design idea to realize a biocoagulation based sorting process is composed of three main steps (Fig. 3): 1. the selective biocoagulation of microorganisms and finesized mineral particles, 2. the separation of the biocoagulates out of the slurry and 3. the detachment of microorganisms and particles, aiming to recover and recycle the biomass (optional). 2. MATERIALS AND METHODS 2.1. Materials Preliminary tests were carried out with different solid particles and microorganisms. The results presented here are based on the use of the sulphide minerals galena (PbS) and sphalerite (ZnS). The basic modules of both mineral structures are ions. Both lattices belong to the AB-lattice-type.
4 The Open Mineral Processing Journal, 2009, Volume 2
Kuyumcu et al.
Selective Biocoagulation
Sorting
Redispersing and Separation Microorganisms Sorting products
Fig. (3). Assumed process design.
Pure minerals were used as test materials. First of all the minerals were crushed and milled. Afterwards the fraction < 10 m was separated to be used in the tests. After some preliminary trials two different yeast cultures, Yarrowia lipolytica and Saccharomyces cerevisiae were selected as microorganisms. For them growth curves were determined. The optimal conditions for the coagulation tests were achieved in the middle of the exponential growth phase, when 108 cells per millilitre medium were counted. At that point of time, the yeasts were harvested. The following growing media were used: Sabouraud-Glucose-Broth(SGB)-medium for Saccharomyces cerevisiae. The medium consisted of peptones of meat and milk proteins (casein, nitrogen and carbon source) and glucose (energy source, 2 %). The pH-value of the used medium was 5.7 ± 0.2. Yeast-Extract-Peptone-Glucose(YEP)-medium for Yarrowia lipolytica. The medium consisted of a mixture of 1 % yeast-extract, 2 % peptone from casein and 2 % glucose. 2.2. Contact Angle Measurements The contact angles on yeast cells and mineral powders were obtained by sessile drop method and sorption measurements respectively. For the sessile drop method was used for the yeast cells, a uniform thick layer of cells were created by filtration through a Millipore filter paper. The filter paper was then fixed on a glass slide with the help of double sided tape. The samples were air-dried to remove the rest of water moisture and contact angles were measured while placing a drop of liquid on the bacterial lawn surface. The dynamic contact angle was recorded with Krss Easy drop equipment and evaluated with Drop Shape analysis software. A drop of liquid is placed on a bacterial lawn flat surface with the help of a syringe and advancing angle is recorded on a CCD camera. Each measurement was repeated 3 times and the arithmetic mean was considered as the final contact angle for a particular liquid. Contact angle on solid powders was determined by sorption measurements using Krss Tensiometer K100 and Krss LabDesk 3.1 software. The Washburn equation is used to measure the contact angle on powder samples. When a column of powder bed is in contact with liquid, the pores between the particles act like small capillaries and the rise of liquid is measurable. The capillary constant in Washburn
equation was determined using n-Hexane (low energy wetting liquid) that wets the solids completely. Using the obtained capillary constant, the contact angles for other liquids were determined. Mineral samples of -106+38 m size fraction were used. A 1 g of mineral was placed into a glass sample tube and was carefully and equally packed each time. Each measurement was repeated at least 3 times and the results found to be reproducible within ±3 degrees deviation and a mean value was reported in the results. For the estimation of acid and base surface energy components by van Oss method, contact angles of at least three different test liquids with well characterized acid and base components are necessary. Three standard liquids of water, 1-bromonaphtalene and formamide were used. The values of liquids surface tension and the dispersive and polar part contributions to the surface energy, and the polar part divided into acid and base components of surface energy as reported in the literature were used [22, 23]. 2.3. EPS Analysis The essential groups of eukaryotic cell walls are the hydroxyl-, carbonyl-, carboxyl-, amino- and phosphate groups. In dependency on the environmental conditions, these groups have an influence on the electrochemical charge of the cell wall [24, 25]. Additionally the microbes produce extracellular polymeric substances (EPS) that may have an influence on the environmental conditions. They are mostly composed of polysaccharides, proteins, nucleic acids and lipids. At “Biofilm Centre” of University of Duisburg-Essen the extracellular polymeric substances (EPS) of the yeasts Saccharomyces cerevisiae and Yarrowia lipolytica were analysed. The results are summarized in Table 2. It was found that the EPS composition of both yeasts is very similar. Both of the yeast cultures are enabling selective biocoagulation of the tested minerals. 2.4. Zeta-Potential Analysis Microscopic investigations and zeta-potential analysis were done immediately after adjustment of the defined condition (pH-value, electrical conductivity) for the yeasts and the prepared minerals. For those purposes a light-scattering microscope (Jenoptik) and a Malvern zeta-sizer were used. The Malvern zeta-sizer “Nano Z” uses a combination of electrophoresis and Laser Doppler Velocimetry to determine
Biocoagulation and its Application Potentials for Mineral Bioprocessing
The Open Mineral Processing Journal, 2009, Volume 2
EPS of the Yeasts Saccharomyces Cerevisiae and Yarrowia Lipolytica
Table 2.
Yarrowia Lipolytica
Saccharomyces Cerevisiae
EPS
1000 mg
1000 mg
Carbohydrates
898 ± 26 mg
803 ± 66 mg
Proteins
82 ± 11 mg
56 ± 9 mg
Uronic acid
4.2 ± 1.1 mg
4.6 ± 1.0 mg
the zeta-potential. The principle is the measurement of the particle velocity in a liquid of known ionic strength when applying an electrical field. The relation between velocity and electrical strength is called electrophoretic mobility and can be used to calculate the zeta-potential. The zeta-potential of the single minerals, yeasts as well as the zeta-potential of the biocoagulates at a given pH-value are obtained from the maximum value of the measured values distribution from several tests. Fig. (4) shows an example of a zeta-potential measurement for Yarrowia lipolytica at a pH-value of 2. Hence, zeta-potential measurements of the single yeasts and biocoagulates at different pH-values were made in addition to measurements of the single sulphide minerals. The isoelectric point of each material used could be determined. Fig. (5) shows the results with Yarrowia lipolytica and galena and sphalerite respectively [26]. 2.5. Thermodynamic Analysis of Cell Adhesion The contact angle of test liquids on different surfaces and their surface energies are presented in Table 3. For Yarrowia
lipolytica yeast cells, the surface energy values that reported by Aguedo et al. [27] are shown. Sulphide minerals comprise considerably higher dispersive component of surface energy than quartz. The polar component of sulphide surfaces is much lower compared to quartz. The electron donating property of minerals is in the order of quartz > sphalerite > galena, which is understandable since quartz surface mostly contain silanol groups. Quartz has a well balanced surface, with almost equal dispersive and polar components of surface energy. Saccharomyces Cerevisiae cells surface has higher dispersive component of surface energy than Yarrowia Lipolytica cells but reverse is the polar component. All the studied surfaces have a dominant electron donating part. Saccharomyces cerevisiae cells expected to be hydrophilic due to their higher water-wetting abilities, but a low contact angle with a non-polar liquid display the nonpolar character of their surfaces. A significant portion of the surface energy attributes to the polar part where the electron donating part is prominent. The non-polar and polar parts of Yarrowia lipolytica surfaces are more balanced with more
Fig. (4). Zeta-potential of Yarrowia lipolytica at pH 2 in 0.03 mol potassium chloride. 5
20 + 2 mV
0 2
2,5
3
3,5
4
4,5
5
5,5
6
-10 -15
-18 mV
10
6,5
-5
Biocoagulates
-20
G. -25
Zeta potential [mV]
Zeta potential [mV]
5
+ 2 mV
0
Biocoagulates -10
-20
-22 mV
Sph.
-30 -30
Y.l.
Y.l. -40
-35
Galenit Galena
Yarrowia Yarrowia lipolytica lipolytica
Galenit + Yarrowia lipolytica Biocoagulates
pH value
2 2,5 5 Figure
Sphalerite Sphalerit
Fig. (5). Zeta-potential of Yarrowia lipolytica, the minerals and biocoagulates.
3
3,5
Yarrowia lipolytica
4
4,5
5
Biocoagulates Yarrowia + Sphalerit
5,5
6
pH value
6,5
6 The Open Mineral Processing Journal, 2009, Volume 2
Table 3.
Kuyumcu et al.
Contact Angle Data and Surface Energy Values Determined by Acid-Base Approach Surface Energy (mJ.m-2 )
Contact Angle Bromo-Naphtalene
Forma-Mide
Total
d/LW
P/AB
+
67.02
3.27
37.81
49.93
44.53
5.40
0.85
8.60
68.70
30.88
53.40
40.64
38.50
2.13
0.08
14.51 28.16
Material
Water
Galena Sphalerite
-
Quartz
38.24
51.87
3.43
56.44
29.17
27.27
6.60
S. Cerevisiae
33.53
18.97
26.34
51.66
42.21
9.50
0.53
42.00
56.20
23.60
32.60
8.00
33.40
Y. Lipolytica
dominating polar part. The strong electron donating character of the microorganisms can be explained by the predominance of polysaccharide compounds and extra-cellular polymeric substances (EPS) rich in -OH groups on the cells surface.
2.6. Extended DLVO Analysis of Cell Adhesion The sphere-sphere geometry of particles for calculating the interaction forces is more suitable for mineralmicroorganism system, although the shape of particles and cells is not exactly spherical. Considering the size of interacting bodies, the spherical model is reasonable. The energy is expressed in kT units in the potential energy diagrams between the cells and minerals presented in Figs. (6 and 7). A model system was chosen for calculations with the following parameters: pH 2.5, 0.01M electrolyte concentration and 1 sphalerite > quartz. This order is evidently due to the decreasing nonpolar character and increasing electron donor capabilities of the minerals in the same direction. These calculations are based on the fact that the particles and cells are dispersed in water and the mineral-cell adhesion/ coagulation leads to a lower energy system. The highly hydrophilic Yarrowia lipolytica cannot drive to leave water environment to adhere on a hydrophobic surface and the free energies of adhesion are less negative compared to Saccharomyces cerevisiae. Table 4.
Table 5.
Zeta-Potential Values for Different Particles at pH 2.5 Material
Zeta-Potential (mV)
Galena
2
Sphalerite
2
Quartz
-5
S. Cerevisiae
-4,5
Y. Lipolytica
-18
Free Energy of Adhesion and Calculated Hamaker Constants Free Energy of Adhesion Gadh (mJ.m-2)
Bacteria-Mineral System
Hamaker Constant (10-21 J)
GadhLW
GadhAB
Gadhtotal
S. Cerevisiae - Galena
-7.33
-6.49
-13.81
S. Cerevisiae - Sphalerite
-5.61
2.92
-2.69
5.22
S. Cerevisiae - Quartz
-2.68
9.32
6.64
2.49
Y. Lipolytica - Galena
-0.76
-3.38
-4.14
0.70
6.81
Y. Lipolytica - Sphalerite
-0.58
1.44
0.86
0.54
Y. Lipolytica - Quartz
-0.28
4.76
10.98
0.26
Biocoagulation and its Application Potentials for Mineral Bioprocessing
The interaction energy versus separation distance curves between Saccharomyces cerevisiae and minerals are shown in Fig. (6). For galena-cell system, the different interacting forces are negative resulting in a total adhesive force. The electrostatic force is rather weak while the LW forces are relatively significant. The acid-base forces are the strongest and attractive controlling the total interaction force for this system. Below the separation distance of 25 A the forces are attractive resulting in an adhesion. In the case of sphalerite, the electrostatic and LW forces are negative while the acid-base forces are positive creating an energetic barrier. Electrostatic forces are very weak while the LW forces are weaker compared to galena and adhesion is possible after overcoming the energetic barrier. The LW forces are attractive but the electrostatic and acid-base forces are repulsive for quartz-cell system. The strong acid-base repulsive interactions are responsible for the inability of cells adhesion on quartz. 500
The Open Mineral Processing Journal, 2009, Volume 2
Fig. (7) shows the interaction energy curves between Yarrowia lipolytica cells and minerals versus separation distance. The LW part of free energy of adhesion for Yarrowia lipolytica-mineral systems is weaker compared to Saccharomyces cerevisiae cells adhesion and the energy curves indeed display very weak LW forces for all Yarrowia lipolytica systems. In the case of galena, the electrostatic forces are stronger and attractive due to a higher and opposing zeta-potential of Yarrowia lipolytica cells and galena. The acid-base interactions are also attractive leading to attractive total interaction energy and the cells adhesion forces towards galena are significant. Although the electrostatic forces are attractive and stronger for Yarrowia lipolytica-sphalerite system, the repulsive acid-base interactions created an energy barrier of 300 kT which needs to overcome for adhesion to occur. Very strong acid-base repulsive forces between Yarrowia lipolytica cells and quartz render their coagulation impracticable (Fig. 7C). 500
A
A Lifshitz - van der Waals
Lifshitz - van der Waals
300
Acid-Base
Total
200
300
100
0 -100 0
20
40
60
80
100
G (kT)
Electrostatic
Acid-Base
Electrostatic
200 G (kT)
400
400
7
Total
100
0 -100 0
-200
-200
-300
-300
20
40
60
80
100
-400
-400
Distance (Å)
-500
Distance (Å)
-500
B
B
500
900 700
Lifshitz - van der Waals
400
Lifshitz - van der Waals
Electrostatic
300
Electrostatic
Acid-Base
500
Acid-Base
200
Total
Total
G (kT)
G (kT)
100
300
0
-100
100
0
20
40
60
80
100
-200
-100 0
20
40
60
80
100
-300
-400
C
500 Lifshitz - van der Waals
C Lifshitz - van der Waals
Electrostatic
400
Electrostatic
300
Acid-Base
300
Acid-Base
Total
200
100 0
-100 0
20
40
60
80
100
G (kT)
400
200 G (kT)
Distance (Å)
-500
Distance (Å)
-500
500
-300
100
0 -100 0
-200
-200
-300
-300
20
40
60
80
100
-400
-400 -500
Total
Distance (Å)
Fig. (6). Saccharomyces cerevisiae - galena (A), sphalerite (B), quartz (C) interaction energy diagrams.
-500
Distance (Å)
Fig. (7). Yarrowia lipolytica - galena (A), sphalerite (B), quartz (C) interaction energy diagrams.
8 The Open Mineral Processing Journal, 2009, Volume 2
Kuyumcu et al.
3. BIOCOAGULATION TESTS The above in detail explained characteristics of the cell walls and the mineral surfaces allow the coagulation between the yeasts and the sulphide minerals, which was validated by microscopic pictures (Figs. 8 and 9). A large sized floc formation also was observed by using both of the yeast cultures. Fig. (9) shows pictures of single biocoagulates taken by means of an atomic force microscope (AFM) applied in the “Biofilm Centre” of University of Duisburg-Essen.
The microscopic analyses show differences between both yeasts. The examinations with sphalerite demonstrate that Saccharomyces cerevisiae adhere at first to larger particles, whereas smaller ones are attached later on. Afterwards large coagulates and flocks are formed. For Yarrowia lipolytica it was observed, that stress during the coagulation process leads to the formation of hyphae. The optimum results of coagulation were observed in an acidic environment at about pH 2. Tests with two binary mixtures sphalerite/quartz and galena/quartz demonstrated the selectivity of the coagulation process (Fig. 10) and corroborate the observations [14]. Galena was chosen for the further investigations.
10 μm Saccharomyces cerevisiae and Sphalerite (ZnS)
10 μm Yarrowia lipolytica and Sphalerite (ZnS) Saccharomyces cerevisiae and Galena (PbS)
Fig. (8). Microscopic pictures of the biocoagulates.
Fig. (9). AFM of Saccharomyces cerevisiae (black arrow) and Sphalerite (white arrow) 10 μm
PbS
10 μm
SiO2
SiO2 ZnS
Yarrowia lipolytica and Galena, Quartz
Fig. (10). Selective biocoagulation
Yarrowia lipolytica and Sphalerite, Quartz
Biocoagulation and its Application Potentials for Mineral Bioprocessing
Flotation was found to be a suitable process for the separation of microorganisms. Depending on the type of microorganisms an attachment between air bubbles and biocoagulates can be expected through natural hydrophoby of microorganisms and biocoagulates without adding any surfactants. For sorting the biocoagulates column flotation has been used. Column flotation is especially suited to sort the very small biocoagulates, considering low shearing forces between the sulphides and the microorganisms. For the flotation tests a lab-scale flotation column was built and used for batch tests as well as for continuous flotation tests. The experimental setup is shown in Fig. (11).
9
velocity (R/t) under variation of the pH-value and the ratio of mineral mass to cell concentration in the pulp. Recovery PbS 100 90 80 70
Recovery [%]
4. FLOTATION TESTS
The Open Mineral Processing Journal, 2009, Volume 2
60 50 40 30 20 10 0 0
10
20
30
40
50
60
70
t (time) [min]
The experimental work was carried out as follows:
S. cerevisiae
1. Breeding of the test yeast in an external fermenter, 2. Conditioning of the yeast-mineral-suspension to form biocoagulates, 3. Feeding the suspension into the flotation (batch-wise / continuous), 4. Flotation by generating air bubbles without adding surfactants, 5. Sampling of the concentrate in time intervals, 6. Analyses of the concentrates and tailings, 7. Interpretation of the results. The results of the flotation tests show that the designed column flotation, depending on the operating conditions, can lead to a recovery in the concentrates of more than 90 % without using any additional chemicals, e.g. collectors. Without additional chemicals a pH-value of 5 was reached. Fig. (12) exemplary shows under these conditions with a constant solids concentration of PbS and a process time of 60 minutes a PbS-recovery of 96 % with Saccharomyces cerevisiae and of 78 % with Yarrowia lipolytica. Thus further tests were carried out with Saccharomyces cerevisiae. In batch tests the influence of different process parameters on the flotation results, i.e. the recovery of mineral concentrate, was investigated. Fig. (13) exemplary shows the cumulative recovery of mineral concentrate and the flotation
Fig. (12). Cumulative recovery of PbS with both yeasts, pHvalue: 5.
Basically a pH-value of 2 is more suitable for the flotation of the biocoagulates. The flotation results at a pHvalue of 5 show no significant decline of the recovery of mineral concentrate (Fig. 13). Due to the advantages regarding the process conditions the continuous flotation tests were carried out at a higher pH-value of 5. Fig. (14) exemplary shows the cumulative recovery of mineral concentrate by variation of the mineral mass to cell concentration ratio in the pulp by continuous flotation. The PbS-recovery can reach up to 94 % depending on this ratio. The higher the ratio of mineral mass to cell concentration the higher is the flotation velocity (R/t) as seen in Fig. (14). The maximum value of the velocity was found to be after approximately 25 minutes [28]. 5. CONCLUSIONS Microbial adhesion on mineral substrates can be assessed by surface thermodynamics and extended DLVO theory of Lifshitz-van der Waals, electrostatic and acid-base interactions. According to thermodynamics, mineral particles and microbial cells dispersed in a liquid medium shift to an adhered particle system if this coagulation leads to a lower energy state. For interaction energy calculation, the parame-
Yeast, Growing media
Mineral sample
Cultivation of the yeast culture
Preparation Grinding, Sizing < 10 μm
Conditioning /
Biocoagulation
Air
Flotation
Tailings
Fig. (11). Experimental setup of the flotation.
Y. lipolytica
Concentrate (biomass + mineral)
10 The Open Mineral Processing Journal, 2009, Volume 2
Kuyumcu et al.
Recovery PbS
Recovery PbS / Time
100
14
90
12 80
10 'R/'t [%/min]
Recovery [%]
70 60 50 40
8
6
30
4
20
2 10 0
0 0
5
10
15
20
25
30
0
5
10
15
t (time) [min] 9
pH 5; 31 mg PbS/10 cells
pH 2;31 mg PbS/10 cells
20
25
30
t (time) [min] 9
9
pH 5; 146 mg PbS/10 cells
pH 2;31 mg PbS/109cells
pH 5; 31 mg PbS/109 cells
pH 5; 146 mg PbS/109 cells
Fig. (13). Cumulative recovery (%)of (PbS) (left) and flotation velocity (R/t, %/min) (right) in batch tests with different pH-values and varied ratios of mineral mass to cell concentration (mg/109 cells). Recovery PbS
Recovery PbS / Time
100
6
90
5
'R/'t [%/min]
80
Recovery [%]
70 60 50
4
3
40
2
30 20
1
10
0
0 0
10
20
30
40
50
60
0
70
10
20
30
12
21
40
50
60
70
t (time) [min]
t (time) [min] 9
12
31 mg PbS/10 cells
21
9
31 mg PbS/10 cells
Fig. (14). Cumulative recovery (%) of (PbS) (left) and flotation velocity (R/t, %/min) (right) in continuous tests with varied ratio of mineral mass to cell concentration (mg/109 cells).
ters such as zeta-potential, particle radius, double layer thickness are known or measurable, while for calculation of Hamaker constant there are two different methods available. According to previous studies the macroscopic approach is more relevant. In the presence of the mixture of different mineral particles the coagulation of the bacteria and mineral particles can be obtained selectively if the interaction energy level differs or can be influenced accordingly. The procedure of selective biocoagulation of microorganisms, e.g. yeasts like Saccharomyces cerevisiae and Yarrowia lipolytica, and micro-dispersed solids, e.g. minerals like galena, sphalerite and quartz, has been analysed with respect to selective biocoagulation. The characteristics of the cell surfaces of the microorganisms and minerals, e.g. the electrostatic charge, the composition of extracellular polymeric substances, the contact angle and the adhesion as well as their influence on the selective biocoagulation were studied. The experimental investigations showed that the microorganisms and the sulphide particles below 10 m coagulate effectively. The experimental findings corroborate the theoretical analysis of the adhesion mechanism between sulphide particles and yeast cells and their selective coagulation. Furthermore it could be shown that flotation is suitable for the separation of the selectively formed biocoagulates. With the designed column flotation satisfying recovery rates
are reached. Prospective research work will involve further steps for the optimisation of the process and scaling-up. ACKNOWLEDGEMENT The research work has been executed as a part of the Integrated EC Project “Biotechnology for Metal bearing materials in Europe (BioMinE)”. The authors thank the European Commission for the financial support of the work. REFERENCES [1]
[2] [3]
[4] [5]
[6]
F. Glombitza and H. Z. Kuyumcu, “Mikrobielles Sortieren“, in 1. Kolloquium Sortieren, Innovationen und Anwendungen, 1999, pp. 141-149. W. Sand and E. R. Donati, Eds., Microbial processing of Metal sulphides. Dordrecht: Springer, 2007. M. Tsezos, “Biological removal of ions: Principles and applications”, in Biohydrometallurgy: From the single cell to the environment, A. Schippers, W. Sand, F. Glombitza and S. Willscher, Eds., Stafa-Zurich: Trans Tech Publications Ltd., 2007, pp. 589-596. G. Rossi, Biohydrometallurgy. Hamburg, New York: McGrawHill, 1990. P. K. Sharma and K. H. Rao, “Adhesion of paenibacillus polymyxa on chalcopyrite and pyrite: surface thermodynamics and extended DLVO theory”, Colloids and Surfaces B: Biointerfaces, vol. 29, pp. 21-38, 2003. D. R. Absolom, F. V. Lamberti, Z. Policova, W. Zingg, C. J. van Oss and A. W. Neumann, “Surface thermodynamics of bacterial adhesion”, Applied and Environmental Microbiology, vol. 46, pp. 90-97, 1983.
Biocoagulation and its Application Potentials for Mineral Bioprocessing [7]
[8] [9] [10]
[11] [12] [13]
[14]
[15]
[16] [17] [18]
H. J. Busscher, A. H. Weerkamp, H. C. van der Mei, A. W. van Pelt, H. P. de Jong and J. Arends, “Measurement of the surface free energy of bacterial cell surfaces and its relevance for adhesion”, Applied and Environmental Microbiology, vol. 48, pp. 980-983, 1984. B. V. Deryagin and L. Landau, Acta Physicochem., URSS, vol. 55, pp. 333, 1941. E. J. Verwey and J. T. G. Overbeek, “Theory of the stability of lyophobic colloids”, Journal of Colloid Science, vol. 10, pp. 224225, 1955. C. J. van Oss, R. J. Good and M. K. Chaudhury, “The role of van der Waals forces and hydrogen bonds in “hydrophobic interactions” between biopolymers and low energy surfaces”, Journal of Colloids Surface B Biointerfaces Science, vol. 111, pp. 378-390, 1986. C. J. van Oss, M. K. Chaudhury and R. J. Good, ”Monopolar surfaces”, Advvanced Colloid Interface Science, vol. 28, pp. 35-64, 1987. F. M. Fowkes, “Attractive forces at interfaces”, Ind. Eng. Chem., vol. 56, pp. 40-52, 1964. A. Vilinska and K. Hanumantha Rao, “Surface thermodynamics and extended DLVO theory of Acidithiobacillus ferrooxidans cells adhesion on pyrite and chalcopyrite”, The Open Colloid Science Journal, vol. 2, pp. 1-14, 2009. H. Z. Kuyumcu, J. Pinka, T. Bielig, “Investigations on the sorting of very fine particles by Biocoagulation”, in Biohydrometallurgy: From single cell to the environment, A. Schippers, W. Sand, F. Glombitza and S. Willscher, Eds., Stafa-Zurich: Trans Tech Publications Ltd., 2007, pp. 337-340. A. Vilinska, K. Hanumantha Rao, and K. S. E. Forssberg, “Selective coagulation in chalcopyrite/pyrite mineral system using Acidothiobacillus group bacteria”, Advanced Maternal Research, vol. 20-21, pp. 366-370, 2007. J. H. Gary, I. L. Fled and E. G. Davis, “Chemical and physical beneficiation of Florida phosphate slimes”, USBM RI 6163, pp. 3334, 1963. R. W. Smith, M. Misra and J. Dubel, “Mineral bioprocessing and the future”, Minerals Engineerings, vol. 4, pp. 1127-1141, 1991. S. Long and F. Wagner, “Structure and properties of biosurfactants”, in Biosurfactants and Biotechnology, N. Kosaric, W. L.
Received: March 18, 2009
The Open Mineral Processing Journal, 2009, Volume 2
[19]
[20]
[21]
[22]
[23]
[24] [25]
[26]
[27]
[28]
11
Cairns and N. C. C. Gray, Eds., New York: Marcel Dekker, 1987, pp. 21-46. C. E. Capes, A. E. McIllhinney, A. F. Sirianni and I. E. Puddington, “Bacterial oxidation in upgrading pyritic coals”, Canadian Mineral and Metal Bulletin, vol. 66, pp. 88-91, 1973. R. J. Desphande, S. Subramanian and K. A. Natarajan, “Effect of bacterial interaction with Thiobacillus ferrooxidans on the floatability of phalerite and pyrite”, in International Seminar on Mineral Processing Technology, 2001, pp. 159-161. K. Hanumantha Rao and P. K. Sharma, “Analysis of different approaches for evaluation of surface energy of microbial cells by contact angle goniometry”, Advances. Colloid and Interface Science, vol. 98, pp. 341-463, 2002. M. N. Bellon-Fontaine, N. Mozes, H. C. van der Mei and O. Sjollema, “A comparison of Thermodynamic Approaches to predict the adhesion of dairy microorganisms to solid substrata”, Cell Biophysics, vol. 17, pp. 93-106, 1990. G. Ström, M. Frederiksson and P. Stenius, “Contact angles, work of adhesion and interfacial tensions at a dissolving hydrocarbon surface”, Journal of Colloid Interface Science, vol. 119, pp. 352361, 1987. P. B. Dengis and P. G. Rouxhet, “Surface properties of top- and bottom-fermenting yeast”, Yeast, vol. 13, pp. 931-943, 1997. D. Q. Lin, P. J. Brixius, J. J. Hubbuch, J. Thommes and M. R. Kula, “Biomass/adsorbent electrostatic interactions in expanded bed adsorption: a zeta potential study”, Biotechnology and Bioengineering, vol. 83, no. 2, pp. 149-157, 2003. J. Pinka, H. Z. Kuyumcu and F. Glombitza, “Untersuchungen zur Sortierung durch Biokoagulation“, in 4. Kolloquium Sortieren, Innovationen und Anwendungen, 2005, pp. 113-122. M. Aguedo, Y. Waché, J. M. Belin and J. A. Teixeira, “Surface properties of Yarrowia lipolytica and their relevance to .decalactone formation from methyl ricinoleate”, Biotechnology Letters, vol. 27, pp. 417-422, 2005. H. Z. Kuyumcu, J. Pinka and T. Bielig, “A novel sorting process for fine-sized sulphide minerals using biocoagulation”, in International Workshop on Bioprocessing of Minerals, Changsha, China, 2008.
Revised: April 24, 2009
Accepted: May 6, 2009
© Kuyumcu et al.; Licensee Bentham Open. This is an open access article licensed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/bync/3.0/) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, provided the work is properly cited.
Paper IX Microorganisms in bioflotation and bioflocculation: potential application and research needs. Hanumantha Rao, K., Vilinska, A., and Chernyshova, I.V., 2009. Advanced Materials Research. 71-73, pp. 319-328.
$GYDQFHG0DWHULDOV5HVHDUFK9ROV SS RQOLQHDWKWWS ZZZVFLHQWLILFQHW 7UDQV7HFK3XEOLFDWLRQV6ZLW]HUODQG
Microorganisms in bioflotation and bioflocculation: potential application and research needs K. Hanumantha Rao1a, A. Vilinska1b and I.V. Chernyshova2c 1
Division of Mineral Processing, Luleå University of Technology, Luleå, Sweden 2Henry Krumb School of Mines, Columbia University, New York, USA a
[email protected] ,
[email protected],
[email protected]
Keywords: microorganisms, oxide and sulfide minerals, flotation, flocculation
Abstract. Conventionally, physico-chemical methods are used in mineral processing for recovering value minerals from ores. The ageing of ore processing tailings and waste rocks, and mining tailings contamination by chemical reagents constitute a major threat to the environment. It is imperative to develop novel economically more efficient and environmentally benign methods of flotation and waste processing, exploiting the intriguing and exciting ability of bacteria to selectively modify the surface properties of solids. Microorganisms have not only facilitate hydrometallurgical leaching operations but have also show a great promise in mineral beneficiation processes such as flotation and flocculation. Several laboratory investigations revealed that microorganisms could function similar to traditional reagents. Microorganisms have a tremendous influence on their environment through the transfer of energy, charge, and materials across a complex biotic mineral-solution interface. The biomodification of mineral surfaces involves the complex action of microorganism on the mineral surface. The manner, in which bacteria affect the surface reactivity and the mechanism of bacteria adsorption, is still unknown and accumulation of the primary data in this area is only starting. The bio-flotation and bio-flocculation processes concern the mineral response to the bacterium presence, which is essentially interplay between microorganism and the physicochemical properties of the mineral surface, such as the atomic and electronic structure, the net charge/potential, acid-base properties, and wettability of the surface. There is an urgent need for developing basic knowledge that would underpin biotechnological innovations in the natural resource (re)processing technologies that deliver competitive solutions. Minerals bioprocessing Metal and energy extractive industries play a strategic role in the economic development of all countries. At the same time, these industries present the major threat to the environment. In fact, mining waste is one of the largest waste streams and is responsible for 18% of overall waste generation (OSCE data 1998). In particular, one of the most severe ecological problems is associated with the emission of sulphur dioxide to air from the roasting and smelting of sulphide concentrates and with the multidimensional environmental pollution produced in the course of ageing of ore processing tailings and waste rocks [1]. Oxidation of metal sulphides in mines, mine dumps, and tailings (acid mine drainage, AMD) is a notorious source of acidity, heavy metals (e.g. Fe2+, Mn2+, Al3+, Cu2+, Pb2+, Cd2+, Zn2+, etc.), and oxyanions (CrO42–, AsO3–, etc.) contamination for streams and groundwater, which poses a serious threat of short- and long-term environmental degradation. Mining tailings can be additionally contaminated by chemical reagents—collectors, modifies, frothers, flocculants, etc. The use in mineral processing, of substances potentially dangerous to the environment and/or human health was approx. 257.000 tons, which amounted to 1 % of the total tonnage of material produced. On the other hand, certain waste contains useful for the industry elements, which invokes an important problem of recycling. To protect public health and safety as well as the environment, the disposal, safety, recycling, and remediation of mining waste require very high costs on processing, containment monitoring, and diverting capital for
%LRK\GURPHWDOOXUJ\
development programmes. These costs negatively impact market competitiveness of metal and energy extractive industries. To overcome the problem, the mineral industries have made large efforts to reduce overall costs through rationalisations of the conventional (chemical and physical) schemes of ore and ore waste processing, e.g., towards limiting the use of dangerous substances, decreasing waste streams and improving waste disposal and recycling practice. The most promising new approach based on integral green-chemistry methods is the biotechnological approach. Effective autotrophic and heterotrophical biosolubilization/bioleaching was observed for sulfide ores, wastes and low-grade minerals, including soils and muds, filter dust/oxides, lateritic ores, copper converter slag, fly ash and electronic waste materials. Alternatively, bacteria can immobilize through a number of mechanisms various components of solutions/suspensions/emulsions, offering remediation, recovery or detoxification applications. Reduction of Se(VI), Cr(VI), U(VI), and Te(VI) by dissimilatory metal-reducing bacteria that use metals as terminal electron acceptors leads to the precipitation and long-term immobilization of these harmful to humans and wildlife ions. A combined reduction-deacidification approach can be applied to protecting the environment from AMD if one properly exploits the ability of metal- and sulfate-reducing bacteria to suppress oxidative solubilization of minerals and eliminate acidity from the system. Another known biotechnologically promising immobilization mechanisms are biosorption and binding of the solution components with peptides, proteins, polysaccharides and other biomolecules. In addition, several recent investigations revealed that adapted bacteria associated with ore deposits can selectively be attached to sulphides, thereby essentially modifying the surface properties relevant to bioflotation and bioflocculation processes. Almost all metal–microbe interactions have been examined hitherto as a means for removal, recovery or detoxification of inorganic and organic metal or radionuclide pollutants. However, except for some mechanisms, notably bioleaching, which has already been employed at a commercial scale, practical exploitation of the biotechnological potential of most of the above bioprocesses is still far from feasibility. There have been several attempts to commercialize biosorption using microbial biomass but success has been limited, primarily due to competition with commercially produced ion exchange media. One of the reasons of such a situation is that all the bioprocesses mentioned have been developed empirically, focused either on the particular biochemical aspects or on the process engineering. Remarkably, an understanding of interfacial phenomena at the molecular level remains elusive for all but the simplest systems. As a result, knowledge-based control, optimisation, and design of the biotic interfaces for solving a given applied problem from the first principles are still considered as a matter of the future. Microorganisms in biobeneficiation In biobeneficiation processes a wide range of microorganisms were used, from simple structure prokaryotes represented mostly by bacteria and archaea (Sulfolobus) to more complex eukaryotes represented by different fungi strains (A. niger, C. parapsilosus). Chemolitotrophic microorganisms capable to derive energy by the oxidation of inorganic compounds or minerals became the first choice for the surface modification of minerals in biobeneficiation for several researchers. The most used microorganisms are the representatives of the genus Acidithiobacillus, which could be considered as autotrophs because of the use of inorganic carbon source. The main advantage of these species is the ability to occupy the mineral sites and oxidize sulphur and ferrous compounds. The redox processes mediated by bacteria in iron and sulfur cycles are presented in Figs. 1 and 2 respectively. Sulfolobus species are also chemolithotrophic which derive energy from sulfur metabolism and could be both autotrophic and heterotrophic. Heterotrophic microorganisms requiring organic carbon as their carbon source can be considered as chemoorganotrophs which derive energy from oxidation of organic compounds. These microbes are more capable to produce secretion products. Metabolic products secreted by some organisms are shown in Table 1. The need of large scale bacterial mass with associated secretion products to fully cover mineral surfaces and
$GYDQFHG0DWHULDOV5HVHDUFK9ROV
alter the surface properties could be the reason for using heterotrophic microorganisms by many researchers.
Fe2+
Fe0
Fe3+
Fig. 1. Redox bacterial-chemical cycle of iron.
S0
SH group of proteins DMSO ↔ DMS
SO42-
H2 S DMSO ↔ DMS
SH group of proteins
S0 Fig. 2. Redox cycle of sulfur – oxidations are shown in yellow (grey) arrows and reductions in red (dark). Reaction in which no redox changes occurs, are in white. DMSO – dimethylsulfoxide, DMS – dimethylsulfide. Historically Gram staining method was an important step towards the identification and characterization of bacteria to Gram-positive and Gram-negative. The Gram-positive and Gramnegative bacterial cell surface structures are presented in Fig. 3. Gram-positive bacteria surface is composed mainly of peptidoglycan, linked together by a peptide interbridge. Many gram-positive bacteria are having several sheets of peptidoglycan creating a thick layer. Usually 90% of the outer membrane is peptidoglycan, while the rest is teichoic acid and some wall-associated proteins. Gram-negative bacteria, in addition to a peptidoglycan, have a second lipid bilayer. This layer contains lipids and polysaccharides linked together, creating a lipopolysaccharide layer. Some of these lipids may be toxic to humans, responsible for the pathogenic properties of some bacterial species.
%LRK\GURPHWDOOXUJ\
Table 1. Metabolic products secreted by some microorganisms. Microorganisms
Major metabolic products
Pseudomonas Sp. Bacillus polymyxa Bacillus circulans B. mucilaginosus Thiobacillus Sp. Aspergillus niger Penicillium Sp. Bacillus Sp. Yeasts and algae
Citric, oxalic and gluconic acids Acetic acid Succinic, formic, fumaric and maleic acids Exopolysaccharides Sulfuric acid, proteins Citric, oxalic and gluconic acids Citric and oxalic acids Amino acids Proteins and nucleic acids
Fig. 3. Bacterial cell surface structure [2,3]. Except Acidithiobacillus species, all bacteria used in mineral beneficiation were gram-positive strains. A summary of biobeneficiation results reported in literature is presented in Table 2. In majority of the studies, the adhesion of cells increased the hydrophilic property of mineral surfaces and caused depression in flotation. Some species (Mycobacterium, Rhodococcus) capable to produce mycolic acid are responsible for hydrophobic surface properties and these were used as collectors to float the minerals. Regardless of the cell’s surface composition, microorganisms often produce some additional surface layers called extracellular polymeric substances (EPS). The amount of EPS produced varies from little for chemolitotrophic species to higher amounts for heterotrophic species. The amount of EPS and surface properties could be influenced by the growth conditions [4]. The magnitude of flotation and/or depression is determined by the amount of adhered cells and flotation selectivity depends on selective adhesion of cells to minerals.
$GYDQFHG0DWHULDOV5HVHDUFK9ROV
Table 2. Summary of biobeneficiation results in literature. Microorganism Mycobacterium phlei
Rhodococcus opacus
Flotation
Flocculation
Sulphur and ash removal from coal [5]
Sulphur and ash removal from coal[5]
Dolomite selectively depressed from apatite [6]
Flocculant for phosphates and hematite [8]
Increased hematite hydrophobicity [7]
Increased hematite settling [7]
Selective flotation of hematite from quartz [9] Selective floatation of mangesite from calcite [10]
Bacillus subtilis
Dolomite and apatite depression from quartz [6]
Aspergillus niger
Depressor for activated magnesite tailings [11]
Candida parapsilosus
Selective calcite flocculation [12]
Corynebacterium xerosis
Flocculant for fluorite and calcite [13]
Bacillus circulans
Silica removal from bauxite [14]
Bacillus mucilaginosis Silica removal from bauxite [15,16] Paenibacillus polymyxa
Calcium and iron removal from alumina [17]
Ash removal from coal [30]
desiliconisation of calcite, alumina and iron oxide [18] Separatation of silica and alumina from iron ores [19] Polysacharide products selectively depress hematite, corrundum and calcite, Proteine products increase the floatability of kaolinite and quartz [20]
Polysacharide products flocculate hematite, corrundum and calcite, Proteine products increase the flocculation of kaolinite and quartz [20]
Depressant for pyrite and chalcopyrite [21,22]
Flocculant for pyrite and chalcopyrite [21-23,31]
Galena selectively depressed from sphalerite [23-26]
Flocculant for galena and sphalerite [23-26]
Pyrite and chalcopyrite moderate separation [27-29] Sulfolobus caldarious Pyrite removal from coal [32] Acidithiobacillus thiooxidans
Selective galena - sphalerite flotation [33,34]
Galena floculation from sphalerite [33]
Acidithiobacillus ferrooxidans
%LRK\GURPHWDOOXUJ\
Coal desulfurization [35,36] Pyrite chalcopyrite and arsenopyrite sellective flotation [37-39] Pyrite depression [29,37-41]
Sellective flocculant for pyrite [37] Sellective flocculation of pyrite from quartz [43]
Fe(II) grown assure better selectivity for pyrite/chalcopyrite [29,40] Selective removal of pyrite from other copper sulfide minerals [41] Galena sphalerite separation [42]
Microbial cells adhesion on mineral surfaces Microorganisms adhere to mineral surfaces for various reasons. S. oneidensis utilize minerals as the terminal electron acceptor in the respiratory cycle. Bacteria of Thiobacilli group recover energy from minerals by the enzymatic oxidation. Common to both microbiologic processes is the bacterial need to access, adhere to, and react with the mineral-water interface. Previous studies have shown that under most physiological conditions the bacterial cell surface carries a net negative charge, while, along with electrostatic forces, hydrophobic, entropic, acid-base, and Van der Waals interactions and H-bonding are important in the bacterial adhesion [44]. The microbe attachment to the mineral surface is followed/accompanied by expression of extracellular polymeric substances (EPS) enabling the adhesion, e.g., by trapping near surface or structural ions, changing thereby the charge of the bacterium envelope and/or that of the mineral surface. In addition, EPS can form chemical bonds with the surface and intermediate/promote the nutrition/respiration chemical reactions. However, the specific mechanisms of adhesion and charge transfer reactions remain a subject of debate [45]. In particular, the charge transfer mechanism can be direct (also called contact or enzymatic), indirect (mediated by quinine-containing shuttle compounds, or Fe3+ ions), and cooperative (includes both direct and indirect mechanisms). For example, it was found that pyrite oxidation by T. ferrooxidans and T. thiooxidans results in the formation of distinct oxidized surface species distributed non-uniformly over the pyrite surface and occurred at isolated regions of the surface that were not correlated with step edges or other topographical features [46]. In contrast, patterns of the hematite degradation by S. oneidensis differ from the microbe footprints and are associated with such structural defects as screw and step dislocations, which implies non-local electron transfer process or electron migration to defect sites from the point of biotic discharge. Either case, the mechanisms of biotic degradation, being based on indirect data, remains entirely speculative. The biomodification of mineral surfaces involves the complex action of microorganism on the mineral surface. There are three different mechanisms by means of which the biomodification can occur: i) attachement of microbial cells to the solid substrate, ii) oxidation reactions and iii) adsorption and/or chemical reaction with the metabolite products (EPS). Several types of autotropic and heterotrophic bacteria, fungi, yeasts and algae are implicated in minerals biobeneficiation. The information on the mechanisms of both bacterium adsorption and reagent (collector) adsorption in the presence of the adsorbed bacteria is necessary. However, this area is a dark spot at the moment. The tremendously complex problem of gaining insight into the mechanisms of adsorption of living organisms is met by a not-less-complex surface chemistry in a pulp as far as even in the absence of adsorbed bacteria the sulphide surface species are unstable. Hence exploring bioflotation processes is a real challenge for a researcher.
$GYDQFHG0DWHULDOV5HVHDUFK9ROV
Research needs in bioflotation and bioflocculation The interface between biological and geological materials, as well as means to design and manipulate that interface is virtually completely unexplored. Previous studies of interactions of bacteria with metal oxides and sulfides were focused either on the particular biochemical aspects or on the process engineering. To provide a new insight into the phenomena, the whole problem of the mineral-bacterium interaction can be conveniently divided into two problems: biochemical and geochemical. The former is related to the bacterium side of the interface (the bacterium envelope) in terms of particular mechanisms of sensing/recognition and response to extracellular minerals, molecule-specific pathways of charge transfer, and genomic mechanisms of regulation of these processes. The geochemical problem concerns the mineral response to the bacterium presence, which is essentially interplay between microorganism and the physicochemical properties of the mineral surface, such as the atomic and electronic structure, the net charge/potential, acid-base properties, and wettability of the surface. Understandably, a better knowledge on the ‘mineral’ side of the problem, that is, on the mineral response to the bacterium presence, is necessary to make headway in biobeneficiation processes. Whatever the mechanism of a particular biotic redox reaction is, reduction of Fe(III) oxides by S. oneidensis consists in the electron transfer from a donor D (an active membrane-bound cytochrome of S. oneidensis) present in the bacterium to the conduction band of the oxide, which requires: ΔE = E(D/D+) – EF > 0, where E(D/D+) is the redox potential of the heme center (protoporphyrin rings containing a central hexacoordinate iron atom) of the cytochrome and EF is the Fermi level of the oxide. Analogously, upon oxidation of pyrite by T. ferrooxidans, electron can transfer from structural Fe2+ ions of pyrite to the Fe(II) oxidase of the iron-oxidizing bacterium (electron acceptor A) provided: ΔG = E(A–/A) – EF, where E(A–/A) is the redox potential of the oxidase cofactor and EF is the Fermi level of the sulfide. The position of the Fermi level at the mineral surface depends on pH of the aqueous phase and/or the mineral potential and on the position and density of the surface states. It follows that biocatalyzed electron-transfer reactions on metal oxides and sulfides can be controlled through variation of these parameters, which require a systematic study in order to find answers to the following questions: i) Are the redox mechanisms established for abiotic interfaces valid in the case of bacterium-mediated reduction/oxidation of the mineral? ii) Is direct bacteria-mineral contact necessary for oxidative or reductive dissolution of sulfides, iron oxides, or manganese oxides? iii) Do bacteria exploit the energetic perturbations present in the surface defects to adhere and oxidize/reduce the mineral? If yes, defects of which type (point vs. clustered) are more preferable by bacteria? iv) Which type of surface defects bacteria generates on their own? What is the mechanism of generation of these defects? v) What are the roles of different kinds of the initial and bacterium-induced mineral heterogeneity in the overall biogeochemical interaction? vi) How size of the mineral particle influences the microbe adhesion and the rate of the biotic oxidation/reduction of the mineral? vii) How the mineral potential (regulated either by pH or potentiostatically) influences the attachment of the whole bacteria and the EPS components and the heterogeneous charge transfer reaction? viii) How the mineral potential affects forms of collector adsorption and surface equilibrium of model redox couples in the presence of the bacteria? Though the mechanisms of adsorption of the microbe’s extracellular polymeric substance (EPS) components onto solid surfaces are generally known, they are system specific [4]. The mechanisms of the interaction of the biopolymers produced by bacteria grown in the presence of i) culture media, ii) specific minerals and iii) both mineral and surfactant (flotation collector) need special attention focusing on the applicability to bioflotation. Another environmentally sound and cost-
%LRK\GURPHWDOOXUJ\
effective biochemical innovation in sulphide flotation could be replacement of Na2S and reduced sulphur compounds, which are widely used in flotation practice as regulator (reductants), by sulphate reducing bacteria. Clearly, to develop a biochemical control of sulphide flotation, it is necessary first of all to understand difference in floatability of the sulphide surfaces reduced chemically and biochemically, which require intensified effort. Summary The interface between biological and geological materials, as well as a means to design and manipulate that interface, is currently virtually completely unexplored [47]. The lack of knowledge could be recognized to the complexity and technical constraints of the problem. Progress in this area is hampered by a deficit of direct, not to mention in situ, spectroscopic molecular/atomic-level data about the mineral–bacterium interfaces. Biomolecular surface science specific to minerals biobeneficiation is just an emerging area of research. Given the recent developments, there is substantial reason to believe that this will be an area of tremendous future growth. References [1]
Information on http://eippcb.jrc.es
[2]
S.M. Hammond, P.A. Lambert and A.N. Rycroft: The bacterial Cell Surface (Kapitan Szabo. Washington, 1984)
[3]
I.C. Hancock, in: Microbial Cell Surface Analysis, edited by N. Mozes, P.S. Handley, H.J. Busscher and P.G. Rouxhet, VCH, New York (1991), p. 22.
[4]
T. Gehrke, R. Hallmann, K. Kinzler and W. Sand: Appl. Environ. Microbiol. Vol. 64 (1998), p. 2743.
[5]
A.M. Raichur, M. Misra, K. Bukka and R.W. Smith: Coll. Surf. B, Vol. 8 (1996), p. 13.
[6]
X. Zheng, P.J. Arps and R.W. Smith: Int. J. Miner. Process. Vol. 62 (2001), p. 159.
[7]
J. Dubel, R.W. Smith, M. Mishra and S. Chen: Miner. Eng. Vol. 5 (1992), p. 547.
[8]
R.W. Smith and M. Mishra, in: Mineral Bioprocessing, edited by R.W. Smith and M. Mishra, The Minerals, Metals and Materials Society, Pennsylvania, (1991) p. 3.
[9]
L.M.S. Mesquita, F.F. Lins and M.L. Torem: Int. J. Miner. Process. Vol. 71 (2003), p. 31.
[10] A.E.C. Botero, M.L. Torem and L.M.S. Mesquita: Miner. Eng. Vol. 20 (2007), p. 1026. [11] J. Gawel, I. Maliszewska, Z. Sadowski: Miner. Eng. Vol. 10 (1997), p. 813. [12] I.A.H. Schneider, M. Misra and R.W. Smith, in: Reagents for better metallurgy, edited by P.S. Mulukutla, TMS, Warrendale (1994), p. 197. [13] S.R. Haas, F.R. Nascimento, I.A.H. Schneider and C. Gaylarde: Revista de Microbiologia Vol. 30 (1999), p. 225. [14] V.I. Groudeva and S.N. Groudev: Travaux ICSOBA Vol. 13 (1983), p. 257. [15] G.I. Karavaiko, Z.A. Avakyan, L.V. Ogurtsava and O.F. Safanova, in Biohydrometallurgy, edited by J. Salley, R.G.L. McGready, L. Wichlacz, CANMET, Otawa (1989), p. 93. [16] L.V. Ogurtsava, G.I. Karavaiko, Z.A. Avakyan and A.A. Korenevskii: Microbiology Vol. 58 (1990), p. 774. [17] Phalguni, J.M. Modak and K.A. Natarajan: Int. J. Miner. Process. Vol. 48 (1996), p. 51.
$GYDQFHG0DWHULDOV5HVHDUFK9ROV
[18] N. Deo and K.A. Natarajan: Miner. Eng. Vol. 10 (1997), p. 1339. [19] N. Deo and K.A. Natarajan: Int. J. Miner. Process. Vol. 55 (1998), p. 41. [20] K.A. Natarajan and N. Deo: Int. J. Miner. Process. Vol. 62 (2001), p. 143. [21] P. Patra and K.A. Natarajan: Coll. Surf. B, Vol. 36 (2004a), p. 91. [22] P. Patra and K.A. Natarajan: Int. J. Miner. Process. Vol. 74 (2004b), p. 143. [23] P. Patra and K.A. Natarajan: J. Colloid Interface Sci. Vol. 298 (2006), p. 720. [24] S. Subramanian, D. Santhiya and K.A. Natarajan: Int. J. Miner. Process. Vol. 72 (2003), p. 175. [25] D. Santhiya, S. Subramanian and K.A. Natarajan: J. Colloid Interface Sci. Vol. 235 (2001b), p. 289. [26] D. Santhiya, S. Subramanian and K.A. Natarajan: J. Colloid Interface Sci. Vol. 256 (2002), p. 237. [27] P.K. Sharma and K. Hanumantha Rao: Miner. Metal. Process. Vol. 16 (1999), p. 35. [28] P.K. Sharma, K. Hanumantha Rao, K.A. Natarajan and K.S.E. Forssberg, in: Proc. XXI Int. Miner. Process. Congr., Developments in Mineral Processing, o. 13, edited by P. Massacci, Elsevier, (2000), B8a, p. 93. [29] P.K. Sharma, K. Hanumantha Rao, K.S.E. Forssberg and K.A. Natarajan: Int. J. Miner. Process. Vol. 62 (2001), p. 3. [30] S.P. Vijayalakshmi and A.M. Raichur: Int. J. Miner. Process. Vol. 67 (2002), p. 199. [31] P. Patra and K.A. Natarajan: Miner. Eng. Vol. 16 (2003), p. 965. [32] F. Kargi and J.M. Robinson: Biotechnol. Bioeng. Vol. 27 (1985), p. 41. [33] D. Santhiya, S. Subramanian and K.A. Natarajan: Miner. Eng. Vol. 13 (2000), p. 747. [34] D. Santhiya, S. Subramanian, K.A. Natarajan, K. Hanumantha Rao and K.S.E. Forssberg: Int. J. Miner. Process. Vol. 62 (2001), p. 121. [35] M. Mishra, K. Bukka and S. Chen: Miner. Eng. Vol. 9 (1996), p. 157. [36] Y.A. Attia, M. Elzeky and M. Ismail: Int. J. Miner. Process. Vol. 37(1993), p. 61. [37] M.N. Chandraprabha, K.A. Natarajan and P. Somasundaran: J. Colloid Interface Sci. Vol. 276 (2004), p. 323. [38] M.N. Chandraprabha, K.A. Natarajan and J.M. Modak: Coll. Surf. B. Vol. 37 (2004), p. 93. [39] M.N. Chandraprabha, K.A. Natarajan and P. Somasundaran: Int. J. Miner. Process. Vol. 75 (2005), p. 113. [40] P.K. Sharma, A. Das, K. Hanumantha Rao and K.S.E. Forssberg, in: Advances in Flotation Technology, edited by B.K. Parekh and J.D. Miller, SME/AIME, (1999) p. 147. [41] T.R. Hosseini, M. Kolahdoozan, Y.S.M. Tabatabaei, M. Oliazadeh, M. Noaparast, A. Eslami, Z. Manafi, A. Alfantazi: Miner. Eng. Vol. 18 (2005), p. 371. [42] M.K. Yelloji Rao, K.A. Natarajan and P. Somasundaran: Miner. Metal. Process. Vol. 9 (1992), p. 95.
%LRK\GURPHWDOOXUJ\
[43] K.A. Natarajan and A. Das: Int. J. Miner. Process. Vol. 72 (2003), p. 189. [44] P.K. Sharma and K. Hanumantha Rao: Adv. Colloid Interface Sci, Vol. 98 (2002), p. 341. [45] H.C. Ehrlich and C.L. Brierley: Microbial Mineral recovery (McGraw-Hill, New York 1990) [46] W. Sand, T. Gerhke, P.G. Jozsa and A. Schippers: Hydrometallurgy Vol. 59 (2001), p. 159. [47] G.E. Brown, Jr. V.E. Henrich, W.H. Casey, D.L. Clark, C. Eggleston, A. Felmy, D.W. Goodman, M. Grätzel, G. Maciel, M.I. McCarthy, K.H. Nealson, D.A. Sverjensky, M.F. Toney and J.M. Zachara: Chem. Rev. Vol. 99 (1999), p. 77.
