Surface Chemistry and Dissolution Kinetics of Divalent Metal

A surface complexation model (SCM) for divalent metal carbonates (Ca, Mg, Sr, Ba, Mn, Fe, Co, Ni, Zn, Cd, and Pb) is developed based on new electropho...
0 downloads 0 Views 96KB Size
Environ. Sci. Technol. 2002, 36, 426-432

Surface Chemistry and Dissolution Kinetics of Divalent Metal Carbonates O. S. POKROVSKY AND J. SCHOTT* Ge´ochimie: Tranferts et Me´canismes, CNRS (UMR 5563)-OMP-Universite´ Paul-Sabatier, 38 Rue des Trente-Six Ponts, 31400 Toulouse, France

A surface complexation model (SCM) for divalent metal carbonates (Ca, Mg, Sr, Ba, Mn, Fe, Co, Ni, Zn, Cd, and Pb) is developed based on new electrophoretic measurements and correlation between aqueous and surface reactions stability constants. This SCM postulates the formation of the following surface species: >CO3H0, >CO3-, >CO3Me+, >MeOH0, >MeO-, >MeOH2+, >MeHCO30, and MeCO3within the framework of a constant capacitance of the electric double layer. It can be used to describe the surfacecontrolled dissolution kinetics of divalent metal carbonates and allows determination of the order of dissolution reactions with respect to rate-controlling protonated carbonate surface groups in acid solutions (>CO3H0) and hydrated metal groups (>MeOH2+) in neutral to alkaline solutions. The reaction order with respect to protonated carbonate groups increases from 2 for MnCO3 and ZnCO3 to 4 for NiCO3, whereas for hydrated surface metals, it augments from 2 for ZnCO3 to ∼4 for MnCO3 and NiCO3. The dissolution rates at 5 e pH e 8 increase in the order Ni < Mg < Co < Fe < Mn < Zn < Cd < Sr e Ca ≈ Ba ≈ Pb and correlate nicely with water exchange rates from the aqueous solution into the hydration sphere of the corresponding dissolved cations. Such a correlation allows the generation for all carbonates of a model describing their dissolution/precipitation kinetics, including the effect of various ligands, provided that rate constants and their activation volumes for water exchange around Me(II)-ligand dissolved complexes are available.

Introduction The knowledge of the dissolution and precipitation kinetics of carbonate minerals at earth surface temperatures is of crucial importance for modeling the global cycles of C, Ca, Mg, Mn, Fe, and a number of trace elements in the biosphere. It has been shown for a number of minerals that the chemical nature of metal atoms and their hydration properties control minerals reactivity in aqueous solution. The Me-O bond energies exert a major control on metal oxides and some silicates dissolution rates (1-3). Accordingly, linear free energy relationships for solids and aqueous ions formation were successfully used to predict surface reaction-controlled dissolution rates of isostructural families of divalent metal oxides and orthosilicates (4, 5). A close approach assumes correlations between metal solvation number and dissolution rate for isostructural sulfates (6) or between the rate of metal * Corresponding author phone: +33 561 55 65 18; fax: +33 561 55 81 38; e-mail: [email protected]. 426

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 3, 2002

adsorption on hydrous oxide and dissolution rates of orthosilicates and oxides and the rate of water molecules exchange in the first hydration sphere of the corresponding cation (7-11). Such modeling approaches require a detailed knowledge of chemical speciation at the solid-solution interface. Surface complexation models (12-17) were shown to be essential tools for the quantitative description of the reactive surface species that control mineral dissolution. Recent developments of new techniques for high-resolution in-situ and ex-situ spectroscopy such as FT-IR X-ray reflectivity measurements (18, 19) additionally proved the validity of the surface speciation approach in describing chemical equilibria at the carbonate mineral-solution interface. The aim of this paper is to improve our knowledge of metal carbonates surface-controlled reactivity in aqueous solutions by generating a surface complexation model (SCM) for new carbonates (i.e., Ba, Sr, Zn, Mn, Ni, Cd, Pb) and to use this model to describe the dissolution kinetics of these minerals. Ultimately, the molecular mechanisms discussed in this study should allow prediction of the surface reactioncontrolled dissolution rates of all divalent metal-bearing minerals.

Experimental Section Mineral Samples. Nine metal carbonates have been investigated in this study. Their formulas, origin, crystal size, and specific surface areas are given in Table 1. Natural hydrothermal rhodochrosite (MnCO3), siderite (FeCO3), and smithsonite (ZnCO3) were obtained from mineral collections (Pyre´ne´es, France; Mexico; Ural, Russia). Large crystals (0.5-2 cm length) were ground in an agate mortar and sieved. The 50-100-µm size fraction of each mineral was ultrasonically cleaned to remove fine particles. Carbonates of cadmium, lead, cobalt, and nickel were hydrothermally synthesized during 2 months at 250 °C in titanium reactors from analytical reagent grade hydrous carbonates. Synthesis was performed in solutions of pH (25 °C) ∼4 and pCO2 ∼40 atm, which was achieved by addition of several grams of solid CO2 per 200 mL of distilled water in the reactor before the synthesis. Lowtemperature synthesis of BaCO3 and ZnCO3 was performed at 25 °C, pCO2 ) 1 atm and pH around 6 as provided by stoichiometric mixtures (i.e., 0.05 M) of NaHCO3 and metal chlorides in a closed plastic vessel. SrCO3 was synthesized in the same way but subsequently recrystallized during 2 months at 250 °C in 0.05 M NaHCO3. X-ray diffraction analyses confirmed that all the samples were well-crystallized anhydrous carbonates without any contamination of clays or oxides (natural samples) and hydrated metal carbonates/ hydroxide impurities (synthetic samples). Chemical analyses by atomic absorption spectroscopy and ICP-MS of investigated solids revealed g99.9% purity. Only for natural rhodochrosite and smithsonite were a maximum 0.5% of Ca and 0.2% of Fe detected. The specific surface area of initial powder was measured by Kr adsorption using the multi-point BET method. For MnCO3, ZnCO3, and NiCO3, the surface area was also measured after dissolution experiments, and no difference (within 15%) was detected. Analyses. NIST buffers (pH 4.008, 6.865, and 9.180 at 25 °C) were used for calibration of a combination pH-electrode. Precision of the pH measurement was (0.002 unit. Carbonate alkalinity was measured by potentiometric titration with HCl to pH 4.2 using a Gran method with a detection limit of 10-5 M and an uncertainty of 2%. Strontium, barium, manganese, iron, cobalt, zinc, cadmium, and lead concentrations were determined using flame and flameless atomic absorption (Perkin-Elmer 5100PC) with an uncertainty of 2%. Low 10.1021/es010925u CCC: $22.00

 2002 American Chemical Society Published on Web 01/04/2002

TABLE 1. Solids Used in This Study

solid

origin

SrCO3 BaCO3 MnCO3

synthesis at 250 °C synthesis at 25 °C natural cleavage crystals 2-5 cm (hydrothermal) natural, hydrothermal synthesis at 250 °C synthesis at 250 °C synthesis at 25 °C natural hydrothermal synthesis at 250 °C synthesis at 250 °C

FeCO3 CoCO3 NiCO3 ZnCO3 ZnCO3 CdCO3 PbCO3 a

crystal size (µm)

surface area (m2/g)

MeO-, >MeHCO30, >MeCO3-, >CO3Me+, >CO3H0, and >CO3-. The surface reactions for all metal carbonates together with crystallographic surface densities for each site on (104) plane are listed in Table 3. Six surface complexation reactions are necessary to characterize the MeCO3-H2O systems. Initial values for surface stability constants are derived from correlations VOL. 36, NO. 3, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

427

TABLE 3. Surface Complexation Model for MeCO3(s)-H2O-CO2 System log K°int (298.15 K, I ) 0) surface reaction >CO3H0 >CO3H0

-

Mg

Ca

H+

) >CO3 + -4.65 -5.1 + Me2+ ) -2.2 -1.7 >CO3Me+ + H+ (3) >MeOH0 - H+ ) >MeO-12.0 -12.0 (4) >MeOH0 + H+ ) >MeOH2+ 10.60 11.85 (5) >MeOH0 + CO32- + 2H+ ) 22.40 23.50 >MeHCO30 + H2O (6) >MeOH0 + CO32- + H+ ) 14.4 17.1 >MeCO3- + H2O (1) (2)

EDL capacitance in 0.01 M NaCl (F/m2) geometric surface sites density (µmol/m2)

31

referencea

16

a

9.84

17 8.22

16, 17

Sr

Ba

Mn

Fe

Co

Ni

Zn

Cd

Pb

-5.0 -2.5

-5.2 -2.5

-5.0 -2.6

-4.4 -1.6

-5.0 -2.0

-4.0 -2.0

-5.2 -1.8

-5.40 -1.80

-5.0 -2.4

-12.0 11.9 23.0

-12.0 12.1 24.5

-14.0 -10.4 -13.0 9.9 10.2 9.2 22.65 22.75 21.8

-11.7 9.5 19.5

16.0

16.15

14.15

14.65

14.0

14.0

13.65

13.15

14.15

17

17

17

31

31

31

17

17

17

7.73

7.01

pw

pw

9.19

15

9.49

15

9.72

9.99

9.72

8.64

7.58

pw

pw

pw

pw

pw

pw ) present work.

between surface and homogeneous solution equilibria and are further refined by fitting experimental electrokinetic data. The surface stability constants determined previously for calcium (17, 18), magnesium (16), iron and manganese (15) carbonates were used to derive the corresponding constants for the other metal carbonates. First, stability constants for hydrolysis reactions on metal sites (reactions 3 and 4 of Table 3) have been approximated from the very good positive correlation that exists between the logarithms of stability constants for surface (>MeOH2+) and solution (Me(OH)20(aq)) metal hydration (Figure 2A). For Mg, Sr, and Ba, the log K(Me(OH)20(aq)) values are unknown. To derive this constant, an additional correlation between the first and the second hydrolysis constants of divalent metals was used (Figure 2B). The relationship shown in Figure 2A allowed determination of stability constants for reaction 4. Reaction 3 has been shown to play a negligible role in carbonate minerals surface speciation at ambient conditions as >MeO- species becomes significant only at pH >12 (15-17). A high uncertainty, therefore, affects this constant. Van Cappellen et al. (15) suggested that reaction 3 was analogous to the reaction formation of Me(OH)3- from Me(OH)20 in homogeneous solution. Following this track, we assigned to the log K of reaction 3 the values of the third step hydrolysis reaction constants for Ni, Zn, Co, Cd, and Pb (22). Note that these values were found to be close to those for the same reaction on simple oxide-solution interface (for NiCO3 and ZnCO3; 24). For BaCO3 and SrCO3, log K (reaction 3) values were postulated the same as for MgCO3 and CaCO3. Stability constants for reactions describing bicarbonate and carbonate ions adsorption on metal sites with formation of >MeHCO30 and >MeCO3- species (reactions 5 and 6 of Table 3, respectively) were assumed to be similar to MeHCO3+(aq) and MeCO30(aq) formation reaction constants in homogeneous solution. Unlike for hydrolysis reactions, no clear correlation is observed between stability constants of aqueous metal carbonate complexes and carbonate adsorbed on metal surface sites. This can be caused by (i) high uncertainties for log K°int of reactions 5 and 6 (Table 3) of studied metal carbonate minerals (15-17) and (ii) important structural difference between carbonate-bearing surface and aqueous complexes. The surface acidity constants of >CO3H0 sites for all metal carbonates (reaction 1 of Table 3) were initially set equal to the value for calcite or rhodochrosite, which is also compatible with the acidity constant of carbonic acid in solution (15). The reaction constants for metal sorption on carbonate surface sites (reaction 2 of Table 3) were set equal to those for MeHCO3+(aq) formation from H2CO30 and Me2+ in solution. Further refinements of stability constants for 428

-11.0 -12.0 -11 8.8 10.2 8.3 21.15 22.15 19.65

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 3, 2002

FIGURE 2. (A) Plot of surface stability constants for metal hydrolysis vs second hydrolysis constant in homogeneous solutions for previously measured (refs 15-18, solid circles) and new (open circles) metal carbonates. (B) Correlation between the logarithms of first and second hydrolysis constants for divalent metals (solid circles). All thermodynamic data are from ref 22 except for Ca (15, 44). Second hydrolysis constants for Mg, Sr, and Ba determined from this correlation are shown as open circles. surface reactions 1, 2, 5, and 6 were achieved from the fitting for each mineral of zero charge conditions (pH, ∑CO2, and [Me2+]tot) as measured in the presence of various amount of dissolved bicarbonate and carbonate (reactions 5 and 6) and metal ions (reactions 1 and 2). It is important to note that the isoelectric point conditions for MnCO3 and FeCO3 measured in the present study (Table 2) were precisely reproduced using a set of surface stability constants for these

FIGURE 3. Examples of different metal carbonates dissolution rates at 25 °C as a function of pH measured in this study. minerals proposed earlier by Van Cappellen et al. (15) based on surface titration results (25). Homogeneous solution equilibria as well as carbonates surface speciation were calculated for each solution composition using K°int values listed in Table 3 and the MINTEQA2 code and thermodynamic database (26). The activity coefficients of free ions and charged complexes were approximated using the Davies equation. Dissolution Kinetics. Steady-state dissolution rates measured at 25 °C and ionic strength of 0.01 M in a variety of solutions with different inlet compositions are given in Table S2 of Supporting Information and depicted as a function of pH for several carbonates in Figure 3. Taking NiCO3 as an example, it can be seen that the data define four distinct regions for the pH dependence of dissolution rate. In strongly acidic solutions (pH e 3), the rate is independent of pH. In mildly acid conditions (3 e pH e 5), it increases with the activity of H+. At a pH of 5-8, the rate is pH-independent. In alkaline solutions (pH >8), the rate decreases with pH. Surface Speciation Control in Acid Solutions. The enhancement of dissolution rate by H+ at pH MeOH2+ species at pH CO3H0}m

(1)

where kCO3 and m are the kinetic constant and reaction order, respectively; and the units are mol cm-2 s-1 for RH+ and mol cm-2 for {>CO3H0}. For rhodochrosite and gaspeite, the data obtained at different pH, [Me2+]tot, and ∑CO2 are consistent with a second- and fourth-order dependency (m ) 2.0 for MnCO3 and 3.7 for NiCO3) of the proton-promoted dissolution rate on >CO3H0 concentration (Figure 4). This suggests that detachment of metals from the surface and thus mineral dissolution requires the protonation of two or four surface carbonates adjacent to hydrated Mn and Ni sites, respectively. It is important to note that at pH CO3- sites are protonated (Figure 3). Similar behavior was reported for magnesite (16). However, such a pH-independent dissolution in acid solutions can be hardly seen for other, more reactive, metal carbonates such as dolomite, calcite, or smithsonite because of significant contribution of transport control at these conditions.

FIGURE 4. Proton-promoted dissolution rate of rhodochrosite and gaspeite in 0.01 M NaNO3 at 25 °C and pH of 2-6 as a function of {>CO3H0}. The solid lines represent linear fits to the data using eq 1. The slopes of 2 and 4 suggest that the surface precursor complex is composed of metal centers surrounded by two and four protonated surface carbonate neighbors, respectively.

FIGURE 5. H2O promoted dissolution rate of ZnCO3, MnCO3, and NiCO 3 at 25 °C and pH of 5-11 as a function of {>MeOH2+}. The solid lines represent linear fit to the data using eq 2. The slopes close to 2 (ZnCO3) and 4 (MnCO3 and NiCO3) show that the carbonate mineral dissolution requires the hydration of two and four surface metals surrounding a surface carbonate site, respectively. Surface Speciation Control in Neutral to Alkaline Solutions. At near-neutral conditions, metal carbonate dissolution is controlled by the hydration of surface metal sites, which is similar to what is observed for magnesite and dolomite (16, 17). Metal carbonates surface speciation at pH >5 suggests that >MeOH2+ is the surface complex most likely to control dissolution at these conditions. This can be seen in Figure 5 where the dissolution rates of ZnCO3, MnCO3, and NiCO3 determined in this study in neutral to alkaline carbonaterich solutions at 25 °C (RH2O) are plotted as a function of >MeOH2+ concentration. The order dependence of RH2O on {>MeOH2+} is close to an integer consistent with

RH2O ) kMe{>MeOH2+}n

(2)

where kMe and n are the kinetic constant and the reaction order, respectively; and the units are mol cm-2 s-1 for RH2O and mol cm-2 for {>MeOH2+}. Equation 2 suggests that dissolution requires the hydration of two (smithsonite, n ) VOL. 36, NO. 3, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

429

TABLE 4. Parameters of Equation 5 for Divalent Metal Carbonatesa solid CaCO3 SrCO3 BaCO3 MnCO3 FeCO3 CoCO3 NiCO3 ZnCO3 CdCO3 PbCO3 MgCO3 (20) CaMg(CO3)2 (17, 29)

FIGURE 6. Rhodochrosite and smithsonite dissolution rates at 25 °C and 6 < pH < 10.3 as a function of carbonate ions activity. All solutions are strongly undersaturated with respect to MnCO3 and ZnCO3 (saturation index below 0.2).

m nd nd nd 2.05 nd nd 3.7 nd nd nd 3.97 2.00

log kCO3

n

log kMe

nd nd nd 7.81 nd nd 19.65 nd nd nd 23.08 9.21

1.0b

-1.0 ( 0.3 -1.35 ( 0.15 -1.05 29.7 ( 0.25 23.2 ( 0.15 10.95 ( 0.15 19.0 ( 0.05 6.34 ( 0.2 -2.3 ( 0.1 -1.6 ( 0.1 21.16 ( 0.1 12.65 ( 0.2

1.0c 1.0c 4.65 4.0d 4.0d 3.73 2.0 1.0 1.0 3.94 1.90

a Units of R and {>i} are mol cm-2 s-1 and mol cm-2, respectively. T nd ) not determined. b Pokrovsky and Schott, unpublished data. c Postulated like for calcite. d Postulated like for gaspeite and rhodochrosite.

2.0) or four (gaspeite, n ) 3.7) surface metal atoms surrounding a carbonate site. For rhodochrosite, n ≈ 4.5. Such high fractional order may be explained by the influence of chemical affinity (rhodochrosite saturation index is higher than 0.5) for dissolution rates MeOH2+} decreases as this species is replaced by >MeHCO30 and >MeCO3-:

>MeOH2+ + HCO3- ) >MeHCO30 + H2O

(3)

>MeOH2+ + CO32- ) >MeCO3- + H2O

(4)

As a result, carbonate and bicarbonate ions act as inhibitors of ZnCO3 and MnCO3 dissolution at far from equilibrium as illustrated in Figure 6. The inhibiting effect of HCO3- and CO32- ions on carbonates dissolution in strongly undersaturated solutions was reported earlier for dolomite and magnesite (17, 20) and recently for calcite (29). To construct a general rate equation for dissolution of divalent metal carbonates, the activated complex theory for far from equilibrium surface-controlled dissolution (1, 30) can be combined with transition state theory in order to account for the effect of chemical affinity (31, 32). Thus, taking into account the protonation reaction in acid region, metal centers hydrolysis reaction in neutral/alkaline solutions and the effect of chemical affinity at close to equilibrium conditions, the following overall rate equation can be proposed (see full derivation in refs 20 and 28):

RT ) [kCO3{>CO3H0}m + kMe{>MeOH2+}n] ×

[ ( )] 1-

Q K°sp

n

(5)

where m and n are the reaction orders with respect to surface protonated carbonate and hydrated metal centers, Q is the activity product, K°sp is the solubility product of the metal 430

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 3, 2002

FIGURE 7. H2O-promoted dissolution rate of carbonates at 25 °C and 5 e pH e 8, ∑CO2 < 10-4 M (RH2O ) RT) plotted against the first-order rate constant, k (s-1) for water exchange from the solvent into the hydration sphere of the corresponding dissolved cation. The solid line is a linear fit to the data. Lead is excluded from the regression as it falls of the line (see text). Dissolution rate data are from refs 27 (Ca, Ba), 17 and 20 (magnesite, dolomite), and this study (Sr, Zn, Cd, Pb, Mn, Fe, Co, Ni). carbonate, and the units of RT and {>i} are mol cm-2 s-1 and mol cm-2, respectively. The values of m, n, kCO3, and kMe for all divalent metal carbonates are listed in Table 4. Metal Hydration Control on the Minerals Reactivity. To check if carbonate dissolution rate correlates with the water exchange rate from the solvent into the hydration sphere of the corresponding dissolved cation, the dissolution rates of eight metal carbonates have been measured at 5 e pH e 8 and ∑CO2 e 10-4 M. At these conditions, rates are pH- and ∑CO2-independent and controlled solely by the metal hydration reaction (water-promoted). The logarithm of this dissolution rate (RH2O) is plotted in Figure 7 against the logarithm of the first-order reaction constant for water exchange rates from the solvent into the hydration sphere of the divalent cation. The values of water exchange rates constants for metals used in this figure are taken from refs 33 and 34. A very good linear correlation is observed with a slope close to 1 for all metals. Only cerussite falls off this correlation, which can be explained by (i) significant control of aqueous diffusion on cerussite dissolution rate when log R > -10 and (ii) formation on the reacting interface of

secondary phases such as hydrocerussite, which is more stable (and thus less reactive) than cerussite in aqueous solutions in contact with atmosphere (35). The correlation observed in the present study allows prediction of dissolution rates for metal carbonates in the presence of various ligands provided that kinetic data for at least one ligand are available. Indeed, a good correlation has been observed between the rate constant for water exchange around a metal-ligand complex and the rate of ligandpromoted dissolution of simple oxides (9, 10, 36). Adsorption of metal cations on carbonate minerals can also modify their dissolution rates such as was shown for calcite (37-39). This can be explained within the SCM framework as the adsorption or exchange with Ca of a metal cation can lead to the formation of more stable >MeOH2+ species whose detachment is rate-controlling. As a result, the relative effect of metal cations on far from equilibrium calcite and other carbonates dissolution should be linked to the water exchange rates in their hydration spheres. Finally, the model elaborated in this study allows prediction of precipitation rates of various metal carbonates at close to equilibrium conditions where eq 5 is valid. In accord with previous observations for calcite, magnesium calcite, and dolomite (40-42), it is postulated that the limiting step for carbonate crystallization is the dehydration of cation at the surface, i.e., the formation of surface metal sites, >MeOH2+, from hydrated adsorbed >CO3Me+‚nH2O species. The lower the rate of water molecules exchange in the first hydration sphere of a metal ion, the lower the dissolution, and, consequently, the crystallization rate constant (kMe in eq 5) of a metal carbonate. As a result, among all calciummagnesium carbonates, magnesite and dolomite have the lowest measurable precipitation rate (20, 28). At ambient conditions, the metals that exhibit the slowest rate of water molecules exchange in their hydration sphere such as Be (log kexchange ∼ 3.2), Ni, Mg, and Co either form hydrated carbonates or are incorporated as traces in calcite whereas all metals from Fe to Ba with the highest water exchange rates form individual anhydrous carbonates. It is interesting to note that the change from hydrous to anhydrous carbonate precipitation that occurs between FeCO3 and CoCO3 is likely to result from a different mechanism of water substitution in Fe and Co coordination spheres. Indeed, the significant increase of the value of the activation volume (∆V‡) for water molecules exchanges between Fe2+ and Co2+ (∆V‡ increases from +0.4 cm3 mol-1 for Fe to +6.1 and +7.2 cm3 mol-1 for Co and Ni, respectively; 33) reflects a change from associative to dissociative mechanism for water exchange. This later mechanism, which requires that the metal drops its coordination number by one to create a gap that is then filled by the incoming water molecule, yields more difficult and slower water/ligand exchanges (and thus dehydration) than the associative mechanism. The model developed in this study corresponds to an integral description of mineral powders represented by various crystallographic surface planes. Although for all metal carbonates the reaction rate can differ between different types of surface steps as shown for calcite (29), these differences become less important when considering the reactivity of the whole mineral. Shiraki et al. (43) found that microscopically measured rate of migration of the (1014) calcite plane is close to that macroscopically measured for the powder. In this regard, the surface speciation approach, which takes into account the hydration affinity of a metal atom, can be a very promising tool for predicting reactivities of various minerals such as oxides, simple silicates, carbonates, phosphates, and sulfides in aqueous solutions of environmental interest.

Acknowledgments The authors are grateful to J. Escalier and R. Freydier for careful technical assistance during the analytical part of this study. F. Thomas and B. Pre´lot (LEM, Nancy) are thanked for their help with electrophoretic measurements. Special thanks are due to M. Thibault for XRD analysis and providing an excellent sample of natural rhodochrosite. F. Fontan and I. Pekov are thanked for providing natural samples of smithsonite and siderite, respectively. We are extremely grateful to J.-C. Harrichoury and B. R. Tagirov for assistance during hydrothermal synthesis of carbonates. Finally, we would like to thank E. Oelkers for helpful discussions and two anonymous referees for their thorough and insightful comments on the manuscript.

Supporting Information Available Two tables giving the metal carbonates ζ-potentials measured by eletrophoresis and a summary of metal carbonate dissolution experiments. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Wieland, E.; Wehrli, B.; Stumm, W. Geochim. Cosmochim. Acta 1988, 52, 1969. (2) Velbel, M. Am. J. Sci. 1999, 299, 679. (3) Oelkers, E. H. Geochim. Cosmochim. Acta 2001, 65, 3703. (4) Sverjensky, D. A. Nature 1992, 358, 310. (5) Sverjensky, D. A.; Molling, P. A. Nature 1992, 355, 231. (6) Dove, P. M.; Czank, C. A. Geochim. Cosmochim. Acta 1995, 59, 1907. (7) Wehrli, B. E.; Ibric, S.; Stumm, W. Colloids Surf. 1990, 51, 77. (8) Westrich, H. R.; Cygan, R. T.; Casey, W. H.; Zemitis, C.; Arnold, G. W. Am. J. Sci. 1993, 293, 869. (9) Casey, W. H. J. Colloid Interface Sci. 1991, 146, 586. (10) Casey, W. H.; Westrich, H. R. Nature 1992, 355, 157. (11) Samson, S. D.; Stillings, L. L.; Eggleston, C. M. Geochim. Cosmochim. Acta 2000, 64, 3471. (12) Schindler, P. W.; Stumm, W. In Aquatic Surface Chemistry; Stumm, W., Ed.; John Wiley & Sons: New York, 1987; pp 83110. (13) Sposito, G. J. Colloid Interface Sci. 1983, 91, 329. (14) Stumm, W. Chemistry of the Solid-Water Interface; John Wiley & Sons: New York, 1992. (15) Van Cappellen, P.; Charlet, L.; Stumm, W.; Wersin, P. Geochim. Cosmochim. Acta 1993, 57, 3505. (16) Pokrovsky, O. S.; Schott, J.; Thomas, F. Geochim. Cosmochim. Acta 1999, 63, 863. (17) Pokrovsky, O. S.; Schott, J.; Thomas, F. Geochim. Cosmochim. Acta 1999, 63, 3133. (18) Pokrovsky, O. S.; Mielczarski, J. A.; Barres, O.; Schott, J. Langmuir 2000, 16, 2677. (19) Fenter, P.; Geissbuhler, P.; DiMasi, E.; Srajer, G.; Sorensen, L. B.; Sturchio, N. C. Geochim. Cosmochim. Acta 2000, 64, 1221. (20) Pokrovsky, O. S.; Schott, J. Geochim. Cosmochim. Acta 1999, 63, 881. (21) Pokrovsky, O. S.; Schott, J. Geochim. Cosmochim. Acta 2000, 64, 3313. (22) Martell, A. E.; Smith, R. M.; Motekaitis, R. J. Critically Selected Stability Constants of Metal Complexes Computer Database, Version 3.0; Texas A&M University: College Station, TX, 1997. (23) Stumm, W.; Morgan, J. J. Aquatic Chemistry, 3rd ed.; John Wiley & Sons: New York, 1996; 1022 pp. (24) Sverjensky, D. A.; Sahai, N. Geochim. Cosmochim. Acta 1996, 60, 3773. (25) Charlet, L.; Wersin, P.; Stumm, W. Geochim. Cosmochim. Acta 1990, 54, 2329. (26) Allison, J. D.; Brown, D. S.; Novo-Gradac, K. J. MINTEAQA2/ PRODEFA2, A Geochemical Assessment Model for Environmental Systems, Version 3.0; U.S. EPA: Athens, GA, 1991; 106 pp. (27) Chou, L.; Garrels, R. M.; Wollast, R. Chem. Geol. 1989, 78, 269. (28) Pokrovsky, O. S.; Schott, J. Am. J. Sci. 2001, 301, 597. (29) Lea, A. S.; Amonette, J. E.; Baer, D. R.; Liang, Y.; Colton, N. G. Geochim. Cosmochim. Acta 2001, 65, 369. (30) Morgan, J. J.; Stone, A. In Chemical Processes in Lakes; Stumm, W., Ed.; John Wiley & Sons: New York, 1985; pp 383-426. (31) Oelkers, E. H.; Schott, J.; Devidal, J.-L. Geochim. Cosmochim. Acta 1994, 58, 2011. VOL. 36, NO. 3, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

431

(32) Lasaga, A. C. Kinetic Theory in the Earth Sciences; Princeton University Press: Princeton, NJ, 1998. (33) Burgess, J. Ions in Solution: Basic Principles of Chemical Interactions; Ellis Horwood Ltd.: Chichester, U.K., 1988. (34) Margerum, D. W.; Cayley, G. R.; Weatherburn, D. C.; Pagenkopf, G. K. In Coordination Chemistry, Vol. 2; Martell, A., Ed.; ACS Monograph 174; American Chemical Society: Washington, DC, 1978; pp 1-220. (35) Bilinski, H.; Schindler, P. Geochim. Cosmochim. Acta 1982, 46, 921. (36) Ludwig, Chr.; Devidal, J.-L.; Casey, W. H. Geochim. Cosmochim. Acta 1996, 60, 213. (37) Terjesen, S. G.; Erga, O.; Thorsen, G.; Ve, A. Chem. Eng. Sci. 1961, 14, 277. (38) Nestaas, I.; Terjesen, S. G. Acta Chem. Scand. 1969, 23, 2519.

432

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 3, 2002

(39) Brown, C. A.; Compton, R. G.; Narramore, C. A. J. Colloid Interface Sci. 1993, 160, 372. (40) Lippmann, F. Sedimentary Carbonate Mineral; Springer-Verlag: Berlin, 1973. (41) De Boer, R. B. Am. J. Sci. 1977, 277, 38. (42) Mucci, A.; Morse, J. W. Geochim. Cosmochim. Acta 1983, 47, 217. (43) Shiraki, R.; Rock, P.; Casey, W. H. Aquat. Geochem. 2000, 6, 87. (44) Guimar, M. J.; Lito, H. M.; Filomena, M.; Camo˜es, G. F. C.; Covington, A. K. J. Solution Chem. 1998, 27, 925.

Received for review May 2, 2001. Revised manuscript received September 4, 2001. Accepted September 28, 2001. ES010925U