Influence of Bacillus subtilis Cell Walls and EDTA on Calcite

Yoshiko Fujita , George D Redden , Jani C Ingram , Marnie M Cortez , F.Grant Ferris , Robert W Smith. Geochimica et Cosmochimica Acta 2004 68 (15), 32...
0 downloads 0 Views 175KB Size
Download PDF
Environ. Sci. Technol. 2003, 37, 2376-2382

Influence of Bacillus subtilis Cell Walls and EDTA on Calcite Dissolution Rates and Crystal Surface Features A . K . F R I I S , †,§ T . A . D A V I S , † M. M. FIGUEIRA,‡ J. PAQUETTE,† AND A . M U C C I * ,† Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal, Quebec, Canada H3A 2A7, and Biotechnology Research Institute, NRC, 6100 Royalmount Avenue, Montreal, Quebec, Canada H4P 2R2

This study investigates the influence of EDTA and the Gram-positive cell walls of Bacillus subtilis on the dissolution rates and development of morphological features on the calcite {101h4} surface. The calcite dissolution rates are compared at equivalent saturation indicies (SI) and relative to its dissolution behavior in distilled water (DW). Results indicate that the presence of metabolically inactive B. subtilis does not affect the dissolution rates significantly. Apparent increases in dissolution rates in the presence of the dead bacterial cells can be accounted for by a decrease of the saturation state of the solution with respect to calcite resulting from bonding of dissolved Ca2+ by functional groups on the cell walls. In contrast, the addition of EDTA to the experimental solutions results in a distinct increase in dissolution rates relative to those measured in DW and the bacterial cell suspensions. These results are partly explained by the 6.5-8 orders of magnitude greater stability of the Ca-EDTA complex relative to the Ca-B. subtilis complexes as well as its free diffusion to and direct attack of the calcite surface. Atomic force microscopy images of the {101h4} surface of calcite crystals exposed to our experimental solutions reveal the development of dissolution pits with different morphologies according to the nature and concentration of the ligand. Highly anisotropic dissolution pits develop in the early stages of the dissolution reaction at low B. subtilis concentrations (0.004 mM functional group sites) and in DW. In contrast, at high functional group concentrations (4.0 mM EDTA or equivalent B. subtilis functional group sites), dissolution pits are more isotropic. These results suggest that the mechanism of calcite dissolution is modified by the presence of high concentrations of organic ligands. Since all the pits that developed on the calcite surfaces display some degree of anisotropy and dissolution rates are strongly SI dependent, the rate-limiting step is most likely a surface reaction for all systems investigated in this study. Results of this study emphasize the * Corresponding author telephone: (514)398-4892; fax: (514)3984680; e-mail: [email protected]. † McGill University. ‡ NRC. § Present address: Environment & Resources DTU, Technical University of Denmark (DTU). 2376

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 11, 2003

importance of solution chemistry and speciation in determining calcite reaction rates and give a more accurate and thermodynamically sound representation of dead bacterial cell wall-mineral interactions. In studies of natural aquatic systems, the presence of organic ligands is most often ignored in speciation calculations. This study clearly demonstrates that this oversight may lead to an overestimation of the saturation state of the solutions with respect to calcite and thermodynamic inconsistencies.

Introduction The characterization of mineral-bacteria interactions contributes to our understanding of many low-temperature geochemical processes, including dissolution, precipitation, and adsorption reactions that govern the rates of chemical weathering and regulate geochemical cycling. Numerous laboratory and field studies have demonstrated that the presence of certain bacteria influences the rate of mineral dissolution (1-4), but the dissolution mechanisms in aqueous solutions remain poorly characterized, even in abiotic systems. Bacteria are a common component in weathering environments. Metabolic activity, cell wall functional groups, and exudate biopolymers may all influence the dissolution rates of minerals (5). The bacterial cell wall contains several different organic acid functional groups, the most important of which are the carboxylic, phosphonic, and hydroxyl groups (6). These cell walls carry no net charge at low pH (below approximately 2) but become more negatively charged as pH is increased (7). The sites can bind cations in solution, but at the pH values most often encountered in natural aquatic environments (i.e., pH 5-9), it is chiefly the carboxylic and phosphonic sites that participate in metal sequestration (6, 8-10). This study focuses on the influence of the cell wall functional groups of a common, Gram-positive soil bacterium (Bacillus subtilis) on calcite dissolution rates, solute speciation, and micro-topographic dissolution features that develop on the {101h 4} surface of calcite crystals in aqueous solutions. An independent set of parallel experiments was also carried out in EDTA solutions at equimolar functional group concentrations to assess if bacterial cell wall functional groups could be viewed merely as dissolved organic acids (i.e., as studied by Orme et al.; 11). The objective was to determine whether dissolution rates are affected by a shift in the saturation state of the solution or a specific interaction with the bacterial cell walls. The influence of EDTA and B. subtilis on the calcite dissolution mechanism was assessed on the basis of the morphology of surface features (i.e., dissolution pits) that developed on the calcite {101h 4} crystal face upon reaction with our experimental solutions. These features were imaged by atomic force microscopy (AFM).

Methods Bacterial Growth, Preparation and Viability. A culture of Bacillus subtilis (168) was kindly provided by T. J. Beveridge (University of Guelph, Canada). The bacteria were incubated for 24 h on trypticase soy agar and then transferred to 0.3 L of trypticase soy broth to grow for another 48 h while being stirred on an orbital shaker (180 rpm) at 32 °C. Both the broth and the agar contained 0.5 w/v % of yeast extract. The B. subtilis cells were separated from their growth media by centrifugation before they were rinsed with a 0.1 N HNO3 solution and then with distilled, deionized water (DDW). 10.1021/es026171g CCC: $25.00

 2003 American Chemical Society Published on Web 05/03/2003

They were then rinsed in a 0.001 M EDTA solution and finally 5 times in DDW. This protocol, described by Fein et al. (9), serves to strip the cell walls of calcium ions and other substances acquired from the growth medium. The total bacterial concentration is reported as wet weight per liter of bacterial suspension after centrifugation at 6000 rpm (RCF ) 2560g; 12) for 30 min. This measurement was used to estimate the number of functional groups on the cell walls in units of functional group sites per wet weight of bacterial suspension (g), as reported by Fein et al. (9). The number of viable cells was determined by counting the number of cfu (colony-forming units/volume in agar plates). Approximately 1% of the cells were microbiologically viable before rinsing (estimated from 10 individual determinations). After multiple rinses, less than 1% of the initially viable cells remained viable, thus, less than 0.01% of the bacterial cells were microbiologically viable at the start of the experiments. Calcite Dissolution. Calcite crystal fragments of similar size (approximately 10 mm2) were cleaved parallel to the {101h 4} face from a large, optical-grade Iceland spar crystal using a razor blade. They were mounted on a glass slide with acetone-soluble Crystalbond 509 (obtained from SPI Metallography Supplies) to ensure that only the {101h 4} face was exposed to the solution. Tests carried out in the presence and absence of the adhesive revealed that it did not affect the pH of the solution or the surface-normalized rate of calcite dissolution in DDW. The surface area of each cleavage rhombohedron was estimated by measurement of the crystal dimensions with an optical microscope and a ruler with an error of (0.1 mm. Calcite dissolution, free-drift experiments were carried out in 0.5 L Erlenmeyer flasks in which three calcite crystals were immersed in 0.100 L of the experimental solutions. The flasks were stirred at 120 rpm on an orbital shaker at room temperature (22-25 °C). They were closed to the atmosphere (covered by Parafilm) to avoid microbial contamination from the ambient air and evaporation of the solution. The dissolution experiments were carried out from 5 min to 14 days, with most lasting 5 days. Experiments were conducted in either distilled water, EDTA in distilled water, or rinsed B. subtilis suspensions in distilled water. EDTA (ethylenediaminetetraacetic acid, disodium salt dihydrate, >99%, purchased from Aldrich) was used over the concentration range of 0.004-2.39 mM. The range in wet weight of centrifuged bacteria used in the experiments was 0.10-58.2 g/L. This corresponds to approximately 4.1 × 1011-2.3 × 1014 cells/L and concentrations that are comparable to the abundance of microbial life in natural aquifer systems (i.e., 7 × 109-7 × 1011 cells/L; 13). The low and intermediate bacterial concentrations used in this study fall within the range of microbial abundances encountered in natural weathering environments (i.e., 3 × 109-3 × 1013 cells/L; 14). The initial pH of the distilled water and EDTA solutions was adjusted to match the initial pH (i.e., approximately 4.5) of the bacterial suspensions by adding a maximum of 40 µL of 0.993 N HCl to the experimental solutions. The pH was measured with a combination glass electrode (Orion 910600) connected to a pH/ISE meter (Orion 710A) and calibrated with three NIST (National Institute of Standards and Technology) traceable buffers (i.e., 4.01, 7.00, and 10.0) at 25 °C. The dissolution experiments were interrupted after 1, 2, 6, 12, 24, 48, 72, 96, and 120 h by removing the calcite crystals from the solutions. At the end of each experiment, the pH and temperature were measured. Approximately 20 mL of the solution were recovered from the Erlenmeyer flask, transferred to polyethylene plastic containers, acidified with 0.167 mL of 12 M HNO3, and sealed with snap-on lids. Solutions isolated from the crystals and containing the dead bacterial cells were acidified and left to react for a minimum

of 1 h in order to liberate calcium attached to the cell walls (15). Subsequently, the bacterial suspensions were centrifuged at 3000 rpm (RCF ) 1280g; 12) for 15 min, and the supernatant was decanted and stored for later analyses. The procedure was repeated a second time to ensure that all the calcium had been stripped from the cell walls and the supernatants were combined. The total calcium concentration in the resulting solutions was determined by flame atomic absorption spectrophotometry (AAS) with a PerkinElmer 3100 spectrophotometer (50 ppb detection limit, reproducible to >95%). Estimates of the Calcite Dissolution Rates. The dissolution rates were estimated from the change in the total calcium concentration in the solution ([Ca2+]) divided by the time (t) and normalized to the total surface area of the three crystals (A):

rate )

∆[Ca2+] tA

(1)

The uncertainty on the rate measurements is estimated at (5% based on the cumulative errors of the calcium analyses ((3%) and crystal size measurements ((3%). Buhmann and Dreybrodt (16) demonstrated that the presence of a variety of ionic compounds (e.g., Na+, Cl-, Mg2+, and SO42-) at a concentration up to 1 mM does not significantly affect the calcite dissolution rate constants; thus, the ionic strength should not significantly influence the kinetics of dissolution (i.e., mechanism) over the range (i.e., I < 0.007 m) of our experimental solutions. Saturation State of the Experimental Solutions. PHREEQC (17) was used in this study to estimate the saturation index (SI) or saturation state (Ω) of our experimental solutions with respect to calcite. PHREEQC uses ion-association and Debye-Hu ¨ ckel expressions to account for the non-ideality of aqueous solutions (i.e., estimate ion activity coefficients) as a function of ionic strength. This aqueous speciation model is entirely adequate at the low ionic strengths of our experimental solutions (i.e., I < 0.007 m), but it does break down at higher ionic strengths (i.e., I > 0.5 m). The SI is defined as:

(

SI ) log

)

{Ca2+}{CO32-} ) log Ω K °sp

(2)

where { } represents the activities of the species and Ksp° is the thermodynamic calcite solubility constant at 25 °C and 1 atm total pressure, 10-8.48 (18). According to the above definition, the SI is zero when the solution is saturated with respect to calcite, SI < 0 when undersaturated, and SI > 0 when supersaturated. The error on SI is estimated as the cumulative uncertainties on the calcium analysis ((3%), the activity coefficient estimates obtained by PHREEQC ((4%), and the carbonate ion concentration ((3%). The cumulative uncertainty on the SI is estimated at (0.025 log unit. In the presence of organic ligands, the uncertainty will be larger due to errors in the ligand concentrations and their calcium binding constants. The chief constraints imposed on the model include the assumption that the system is closed to the atmosphere as well as the assignment of acid dissociation and metal complexation constants for the functional groups present on the bacterial cell walls. The system was considered closed to the atmosphere because Parafilm, which was used to cover the reaction flasks, was deformed during the experiment as the headspace gas expanded and contracted with changes in temperature. This assumption was verified by testing the internal consistency of the carbonate system (i.e., measured VOL. 37, NO. 11, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2377

TABLE 1. Deprotonation and Metal Stability Constants for EDTA and Functional Groups on B. subtilis Cell Walls Used in the PHREEQC Model Log K

Reference

EDTAH3- S EDTA4- + H+ EDTAH22- S EDTA4- + 2H+ EDTAH3- S EDTA4- + 3H+ EDTAH4 S EDTA4- + 4H+ B-COOH0 S B-OO- + H+ B-POH0 S B-PO- + H+ B-OH S B-O- + H+

Deprotonation Constants -11.25 -18.08 -20.36 -22.56 -4.82 ( 0.14 -6.9 ( 0.5 -9.4 ( 0.6

Daniele et al. in ref 31 Daniele et al. in ref 31 NIST in ref 31 NIST in ref 31 8 8 8

EDTA4- + Na+ S EDTANa3EDTA4- + H+ + Na+ S EDTAHNa2EDTA4- + Ca2+ S EDTACa2EDTA4- + H+ + Ca2+ S EDTAHCaB-COO- + Ca2+ S B-COOCa+ B-POO- + Ca2+ S B-POOCa+

Metal Stability Constants 2.7 11.44 10.7 16.0 2.8 4.2

Daniele et al. in ref 31 Daniele et al. in ref 31 31 Morel et al. in ref 31 15 21

Reaction

and calculated pH from alkalinity and pCO2) using PHREEQC. The stability constants used in the speciation model are listed in Table 1. Complexation reactions and constants with EDTA and cell wall functional groups were added to the PHREEQC database. The deprotonation constants for the cell wall functional groups of B. subtilis are similar to those reported in other studies (19, 20). The complexation constant for the calcium-phosphonyl complex of B. subtilis was estimated by the correlation technique described by Langmuir (21). This was necessary because the complexation constant of calcium to B. subtilis or other related bacterial species has not been measured experimentally and estimates are not available in the literature. The stability constants for various metals complexed with phosphoric acid (22) were correlated to the stability constants of metal-phosphonyl groups of the same metals bound to the cell wall of B. subtilis (9) and a value for the B-POOCa+ complex (log K ) 4.2, Table 1) was interpolated. Atomic Force Microscope (AFM). Reacted crystals were taken out of the Erlenmeyer flasks and briefly rinsed with a minimal amount of DW in order to remove solution salts, bacterial cells, exudates, etc. This avoided adherence of, for example, cells to the cantilever tip of the AFM and/or precipitation of salts that would have created morphological artifacts on the surface of the reacted crystals. The crystals were dried by holding a tissue at one of their edges in order to absorb most of the residual water from the surface. The crystals were air-dried at room temperature and stored in closed Petri dishes. AFM imaging of the reacted surfaces was carried out within 5 days of sampling using a Digital Instrument Dimension 3000 scanning probe microscope. The probe is a combined assembly of a single-crystal silicon tip (model TESP) attached to the end of a single beam cantilever mounted on a piezoelectric scanner. The probe was operated in tapping mode (4), and both height and phase images were captured. The scan parameters (i.e., scan angle, scale, and speed) as well as the original position of the sample (i.e., samples were physically rotated 90°) were varied before image capture in order to test for image artifacts (23). The digital images were processed (3rd order of flattening) using the image treatment software supplied by Digital Instruments.

Results and Discussion Calcite Dissolution Rates. The calcite dissolution rates in distilled water, EDTA, and B. subtilis decrease exponentially with time irrespective of the functional group concentration (Figure 1). The measured dissolution rate in distilled water is not sensitive to the initial pH over the range studied (i.e., 2378

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 11, 2003

FIGURE 1. Calcite dissolution rates in distilled water and in the presence of organic ligands at equimolar functional group concentrations. 4.14-5.75) because the buffer capacity of this solution is negligible but increases drastically following the dissolution of calcite. The pH increased rapidly during the first 24 h of reaction until it reached a plateau value, between 4.6 and 6.5, which differed according to the nature and concentration of the ligand. Over the range investigated (i.e., 0-2.39 mM EDTA and 0-58.2 g B. subtilis/L), calcite dissolution rates increased linearly with the ligand concentration (not shown). Rates determined in the presence of EDTA or dead bacterial cells are compared at equimolar acid functional group concentrations. For comparison purposes, only the carboxylic and phosphonic sites of the bacteria cell walls are

FIGURE 3. Calcite dissolution rates as a function of SI at equivalent functional group concentrations and in distilled water. Note that the rate scale is logarithmic. FIGURE 2. Evolution of SI during calcite dissolution in distilled water and in the presence of organic ligands at equimolar functional group concentrations. taken into consideration, since hydroxyl sites are not abundant and do not contribute significantly to the sequestration of calcium ions. Rates are compared at low, intermediate, and high functional group concentrations (i.e., 1.6 × 10-5, ∼4.0 × 10-3, and 9.5 × 10-3 M, respectively), corresponding to the experimental EDTA or bacterial cell concentrations. Results of experiments performed at the low functional group concentration (not shown; 0.004 mM EDTA or 0.10 g B. subtilis/L) show no significant difference between the calcite dissolution rates measured in the presence or absence of organic ligands (EDTA or B. subtilis). At intermediate (1.00 mM EDTA ) 4.0 mM sites; two concentrations of bacteria: 22.5 and 26.3 g/L corresponding to 3.7 and 4.3 mM of sites, respectively) and high (2.39 mM EDTA or 58.2 g B. subtilis/L) but equivalent functional site concentrations, the dissolution rates are higher in the presence of EDTA than in the bacterial cell suspensions (Figure 1A,B). The faster calcite dissolution rate in the presence of EDTA can be explained by the much greater stability of the calcium-EDTA complex. Its complexation constant is 6.5-8 orders of magnitude greater than the calcium complexes that form with functional groups on the bacterial cell walls (Table 1). Chemical Evolution of the Experimental Solutions. The SI was chosen as the master variable to compare the dissolution rates in the different experimental solutions because it provides an unbiased characterization of the degree of disequilibrium of the system. Thereby, a comparison of rates normalized to SI allows one to distinguish if factors other than the calcium and carbonate ion activity product

or the saturation state of the solution influence the dissolution rates. Under the free-drift conditions of our experiments, values of SI evolve toward saturation (i.e., SI ) 0) as dissolution proceeds (Figure 2), but it is possible to compare instantaneous rates at equivalent SI values. The calculated SI values reveal that all solutions remain undersaturated with respect to calcite (Figure 2) throughout the duration of the experiments. In all cases, the logarithm of the dissolution rate is inversely proportional to SI (Figure 3), thus, the calcite dissolution rate accelerates at higher degrees of undersaturation and is most likely dominated by surface reactions (24). The calcite dissolution rate is barely influenced by the presence of either EDTA or dead B. subtilis cells at the low functional group concentration (∼1.6 × 10-5 M sites, SI-normalized, not shown). At the intermediate and high functional group concentrations (Figure 3A,B), there is a distinct difference between dissolution rates measured in the presence of EDTA and the bacterial cell suspensions at comparable SI values. In contrast, the rates measured in the presence of B. subtilis cells are, within the uncertainty of our rate and SI estimates, undistinguishable from those obtained in distilled water. Although all three systems display a strong dependency on SI, the dissolution in the presence of EDTA clearly proceeds at an accelerated rate, likely because the reaction mechanism is modified. Fredd and Fogler (25) interpreted the accelerated dissolution of calcite in the presence of various organic chelators, including EDTA, as a direct “attack” on the crystal surface. This attack may only proceed as long as the functional groups of the EDTA are not completely saturated with calcium ions. Our speciation calculations reveal that, within the first hour, enough calcite was dissolved to saturate the EDTA sites at the low ligand concentration (i.e., 0.004 mM EDTA). Accordingly, at the low functional group concentration (not VOL. 37, NO. 11, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2379

FIGURE 5. (A) Idealized {101h4} calcite rhombohedron with the c axis oriented vertically. Nonequivalent edges and corners of type P (polar) and E (equatorial) of the {101h4} face are shown. (B) Schematic growth hillock or dissolution pit on an idealized single {101h4} face, with step A′ moving toward edges of type P (After ref 32).

FIGURE 4. Calcite dissolution rates as a function of the “true” and “apparent” SI, illustrating the influence of the organic ligands on the solution speciation. The lines correspond to the linear leastsquares fit to the rates measured in the presence and absence of bacterial cells in distilled water presented in Figure 3. shown), there is no significant difference between the calcite dissolution rates measured in the three systems under investigation. In contrast, in the intermediate and high EDTA concentration solutions, the ligand remained unsaturated throughout the dissolution experiments, with its carboxylic sites free to attack the calcite surface and enhance the dissolution rates relative to those measured in distilled water (Figure 3A,B). The solution speciation model described previously, used to estimate the saturation state of the experimental solutions with respect to calcite, accounts for the complexation of dissolved calcium with the organic ligands. In characterizing the speciation of natural solutions, however, the presence of organic ligands is generally ignored because their nature and concentration are not readily determined. Thus, the saturation state of the solutions estimated from a purely inorganic speciation model may be inaccurate and lead to observational inconsistencies (e.g., dissolution of calcite in an apparently supersaturated solution). To illustrate this, results of our dissolution rate experiments in distilled water and in the presence of dead B. subtilis cells at intermediate and high functional group concentrations are reproduced in Figure 4 as a function of the “true” (i.e., including organic complexation) and “apparent” (excluding organic complexation) SI. They clearly show that the complexation of dissolved Ca2+ by B. subtilis cell wall functional groups decreases the saturation state of the solutions and that the calcite dissolution rates measured in the presence of these ligands are nearly identical to those measured in distilled water (Figure 3) at equivalent, “true” SI. Ignorance of the complexation by the cell wall functional groups would have led us to conclude that their presence accelerates the calcite dissolution rate. 2380

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 11, 2003

FIGURE 6. Nomenclature for section analysis. (A) Asymmetric dissolution pit on an idealized {101h4} face. (B) Cross-section of idealized asymmetric dissolution pit (x/y ) degree of anisotropy). These results will likely be modified by the metabolic activity and the presence of exudates in systems with live bacteria. Atomic Force Microscopy. In most of the earlier AFM studies, dissolution features that developed on the surface of calcite were imaged after short reaction times (i.e., seconds to minutes; e.g., refs 26-28). This section focuses on the morphology of the calcite {101h 4} surface and more specifically on the anisotropy of the dissolution pits generated during the reaction following dissolution for between 5 min and 1 h in distilled water, EDTA solutions, and aqueous suspensions of dead B. subtilis cells. As dissolution takes place, pits typically develop as discrete, straight-edge features on the mostly flat {101h 4} cleavage surface of calcite (29). Figure 5 depicts the plan view of a rhombohedral dissolution pit (Figure 5B) with edges parallel to those of the morphological (or cleavage) plane of calcite (Figure 5A) (30). As dissolution proceeds further, the pits (Figure 5B) grow wider and deeper but usually at different rates along edges that are parallel and opposite to each other, and the resulting morphology can be described by its degree of anisotropy. The pit morphology (i.e., aspect ratios) that develops early on is preserved as the pit grows (27). Consequently, the steepness of the pit walls is directly related to its degree of anisotropy. A 3D image of the dissolution pit is generated by the AFM, from which a cross-sectional reconstruction along the symmetry axis is obtained (Figure 6). From this cross-section, the length, (x + y), is determined by the image treatment software (Digital Instruments). The degree of anisotropy is calculated as x, the horizontal length of the shortest flank, divided by y, the horizontal length of the longest flank (Figure 6). A value of x/y ) 1 represents a perfectly isotropic pit, whereas values deviating from 1 represent an increasing

anisotropy is expected given the fact that parallel steps on opposite sides of the pit are not related by the face symmetry of the calcite rhombohedron. At the low B. subtilis cell concentration (0.004 mM functional group sites), the average degree of anisotropy of the pits was 0.44 ( 0.12 (n ) 5; image not shown), but the surfaces were not imaged at the low EDTA concentration. At higher functional group concentrations (i.e., 4.0 mM sites as EDTA, Figure 7B, or 4.3 mM sites on B. subtilis cell walls, Figure 7C), the average values of pit anisotropy were found to be 0.81 ( 0.19 (n ) 4) and 0.94 ( 0.05 (n ) 5), respectively. These results suggest that, above a critical organic ligand concentration (i.e., >0.004 mM functional group sites), the dissolution mechanism changes. Despite the enhanced dissolution rates in the presence of EDTA compared to those measured in suspensions of B. subtilis cells, the anisotropy of dissolution pits revealed by AFM images is very similar. Isotropic dissolution pits probably did not form in the low concentration B. subtilis solution because all the cell wall binding sites were rapidly saturated with calcium. The use of AFM to quantitatively characterize the morphology of calcite dissolution pits that develop in the presence of natural (i.e., dead bacteria/cell walls) and synthetic organic ligands (i.e., EDTA) provides insights into the mechanisms that govern these reactions. This study shows that increasing concentrations of both bacterial cells and EDTA result in the development of different micro-topographic features on calcite mineral surfaces, including an increased isotropy of the dissolution pits.

Acknowledgments Financial support for this research was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC) through individual grants to A.M. Additional funds were provided by GEOTOP-UQAM-McGill, Knud Højgårds Fond, Ingeniørforeningen Danmark, Rudolph Als Fondet, Frants Allings Legat, and UniDanmark Fondet. The authors wish to thank Terry Beveridge for providing a laboratory stock culture of Bacillus subtilis 168. A.K.F. would also like to thank Prof. H. Vali for stimulating discussions on the interactions of biomolecules with the calcite surface; Associate Professor Rasmus Jakobsen for guidance on the use of PHREEQC; Glenn Poirier for operation of the AFM and interpretation of images; Constance Guignard, Glenna Keating, and Sandra Lalli for their technical assistance in the laboratory and analytical/intrumental instructions. Finally, we would like to acknowledge the three anonymous reviewers who provided critical and constructive comments on a previous version of this manuscript.

Literature Cited

FIGURE 7. Representative pits developed on calcite crystals upon dissolution in some of our experimental solutions: (A) distilled water, (B) 1 mM EDTA (i.e., 4.0 × 10-3 M functional groups), and (C) 26.3 g B. subtilis/L (i.e., 4.3 × 10-3 M functional groups). degree of anisotropy. A pit that typically develops following dissolution in distilled water displays an anisotropic morphology (Figure 7A). The average degree of anisotropy (i.e., x/y) of pits that developed in distilled water was 0.44 ( 0.20 (n ) number of distinct measurements ) 5). As indicated earlier, this

(1) Lee, J.-U.; Fein, J. B. Chem. Geol. 1998, 166, 193-202. (2) Hiebert, F. K.; Bennet, P. C. Science 1992, 258, 278-281. (3) Bennet, P. C.; Hiebert, F. K.; Choi, W. J. Chem. Geol. 1995, 132, 45-53. (4) Grantham, M. C.; Dove, P. M. Geochim. Cosmochim. Acta 1996, 60, 2473-2480. (5) Banfield, J. F., Nealson, K. H., Eds. Geomicrobiology: Interactions between Microbes and Minerals; Reviews in Mineralogy 35; Mineralogical Society of America: Washington, DC, 1997. (6) Beveridge, T. J.; Murray, R. G. E. J. Bacteriol. 1980, 141, 876887. (7) Harden, V. P.; Harris, J. O. J. Bacteriol. 1953, 65, 198-202. (8) Beveridge, T. J.; Murray, R. G. E. J. Bacteriol. 1976, 127, 15021518. (9) Fein, J. B.; Daughney, C. J.; Yee, N.; Davis, T. A. Geochim. Cosmochim. Acta 1997, 61, 3319-3328. (10) Daughney, C. J.; Fein, J. B. J. Colloid Interface Sci. 1998, 198, 53-77. (11) Orme, C. A.; Noy, A.; Wierzbicki, A.; McBride, M. T.; Grantham, M.; Teng, H. H.; Dove, P. M.; DeYoreo, J. J. Nature 2001, 411, 775-779. VOL. 37, NO. 11, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2381

(12) Perry J. H. Chemical Engineers’ Handbook; McGraw-Hill Book Company: New York and Toronto, 1963. (13) Chapelle, F. H.; Lovley, D. R. Ground Water 1992, 30, 29-36. (14) Albrechtsen, H.-J.; Winding, A. Microb. Ecol. 1992, 23, 303-317. (15) Fowle, D. A.; Fein, J. B. Chem. Geol. 2000, 168, 27-36. (16) Buhmann, D.; Dreybrodt, W. Chem. Geol. 1987, 64, 89-102 (17) Parkhurst, D. L. User’s guide to PHREEQC; U.S. Geological Survey Water-Resources Investigations Report 95-4227; USGS: Denver, CO, 1995. (18) Plummer, L. N.; Busenberg, E. Geochim. Cosmochim. Acta 1982, 46, 1011-1040. (19) Gonc¸ alves, M. L. S.; Sigg, L.; Reutlinger, M.; Stumm, W. Sci. Total Environ. 1987, 60, 105-119. (20) Daughney, C. J.; Siciliano, S. D.; Rencz, A. N.; Lean, D.; Fortin, D. Environ. Sci. Technol. 2002, 36, 1546-1553. (21) Langmuir, D. In Chemical Modeling in Aqueous Systems; Jenne, E. A., Ed.; ACS Symposium Series 93; American Chemical Society: Washington, DC, 1979; pp 353-387. (22) Martell, A. E. Critical Stability Constants, Vol. 3; Plenum Press: New York and London, 1976. (23) Eggleston, C. M. In Scanning Probe Microscopy of Clay Minerals; Nagy, K. L., Blum, A. E., Eds.; CMS Workshop Lectures 7; Clay Minerals Society: Boulder, CO, 1994; pp 3-90.

2382

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 11, 2003

(24) Morse, J. W.; Arvidson, R. S. Earth Sci. Rev. 2002, 58, 51-84. (25) Fredd, C. N.; Fogler, H. S. J. Colloid Interface Sci. 1998, 204, 187-197. (26) Dove, P. M.; Platt, F. M. Chem. Geol. 1996, 127, 331-338. (27) Liang, Y.; Baer, D. R.; McCoy, J. M.; LaFemina, J. P. J. Vac. Sci. Technol. 1996, 14, 1368-1375. (28) Teng, H. H.; Dove, P. M.; Yoreo, J. J. Geochim. Cosmochim. Acta 2000, 64, 2255-2266. (29) Klein, C.; Hurlbut, C. S., Jr. Manual of Mineralogy; John Wiley & Sons: New York, 1993. (30) Reeder, R. J. In Carbonates, Mineralogy and Chemistry; Reeder, R. J., Ed.; Reviews in Mineralogy 11; Mineralogical Society of America: Washington, DC, 1983; pp 1-47. (31) Anderegg, G. IUPAC Chemical Data Series 14; 1977. (32) Paquette, J.; Vali, H.; Mucci, A. Geochim. Cosmochim. Acta 1996, 60, 4689-4699.

Received for review September 19, 2002. Revised manuscript received February 26, 2003. Accepted March 13, 2003. ES026171G