SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sherman, L. A.
Right arrow Articles by Barak, P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Sherman, L. A.
Right arrow Articles by Barak, P.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Sherman, L. A.
Right arrow Articles by Barak, P.
Soil Science Society of America Journal 64:1959-1968 (2000)
© 2000 Soil Science Society of America

DIVISION S-2-SOIL CHEMISTRY

Solubility and Dissolution Kinetics of Dolomite in Ca–Mg–HCO3/CO3 Solutions at 25°C and 0.1 MPa Carbon Dioxide

Leslie A. Shermana and Phillip Barakb

a Dep. of Chemistry, Washington College, 300 Washington Ave., Chestertown, MD 21620-1197 USA
b Dep. of Soil Science, Univ. of Wisconsin-Madison, 1525 Observatory Dr., Madison, WI 53706-1299 USA

leslie.sherman{at}washcoll.edu


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 
Dolomite solubility in water has been measured by a number of different methods during the past several decades, yielding inconsistent and unreliable results that vary more than three orders of magnitude. The most commonly used best value for dolomite solubility in water is based on HCl solution calorimetry at 300.15 K, which is not confirmed by earlier determinations based on heat capacity of dolomite nor by more recent acid solution and metal oxide melt calorimetric measurements. In this study, the solubility of a high purity dolomite was measured directly by monitoring dissolution in Ca-Mg-HCO3/CO3 solutions designed to bracket the presumed solubility product of dolomite, pKs dolomite [= -log (Ca2+)(Mg2+)(CO2-3)2, where the values in parentheses are activities at equilibrium], between 16.0 and 19.0, at 0.101 MPa (1 atm) CO2 and 25°C. The use of gas-permeable, water-impermeable membranes over the dissolution vessels allowed for maintenance of an open system for CO2, with minimal water loss during the course of the 672-d experimental period. The dolomite dissolved congruently in Ca-Mg-HCO3/CO3 solutions with initial ion activity products [pIAPdolomite = -log (Ca2+)(Mg2+)(CO2-3)2, where the values in parentheses are measured activities] greater than 17.5. Both calcite and magnesian calcites can be ruled out as controlling solubility in these measurements. Based on statistical inference by comparison of alkalinity in unseeded (control) and dolomite-seeded solutions, the pKs dolomite is between 17.4 and 17.0, expressed as 17.2 ± 0.2. A previously proposed kinetic model of successive reactions for dolomite dissolution near equilibrium—a fairly rapid dissolution of the CaCO3 component in equilibrium and a rate-limiting protonation reaction dependent on the activity of the MgCO3 component—appears to fit the experimental data.

Abbreviations: BET, Brunauer–Emmett–Teller • EDX, energy dispersive x-ray • IAP, ion activity product • XRD, x-ray diffraction


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 
THE SOLUBILITY OF DOLOMITE, CaMg(CO3)2, is a critical component in the dynamics of many aqueous systems, as a significant portion of surface waters, soil water, and groundwater are in contact with carbonate minerals, primarily calcite and dolomite. The solubility of dolomite is considered in both static assessment of the saturation state of natural waters (Wicks and Herman, 1994; Bischoff et al., 1994), most often with the use of equilibrium models, and in process-oriented models such as dolomitization (Moore, 1989), magnesian calcite formation (Doner and Lynn, 1989), seawater–brine evaporation (Plummer and Parkhurst, 1990; Smith et al., 1990) and geomorphological processes (Birkeland, 1984). In addition, dolomite solubility is considered in assessing irrigation water quality (Thellier et al., 1990a, 1990b), in issues concerning application or disposal of waste products (Tedaldi and Loehr, 1992), acid precipitation in dolomitic environments (Eckstein and Hau, 1992), and evaluation of the origin of dolomite in soils (Bui et al., 1990; Dress and Wilding, 1987; Sobecki and Karathanasis, 1987).

Reported values of the solubility product of dolomite, pKs dolomite [= -log (Ca2+)(Mg2+)(CO32-)2, where the values in parentheses are activities at equilibrium], range from 16 to 19 (Fig. 1 , Table 1) . The free energy value listed in Robie et al. (1978) is often chosen as the current best value for dolomite (e.g., Sadiq and Lindsay, 1979; Nordstrom et al., 1990). The experimental methods and the supplementary thermodynamic data used in the calculations were reported later in Hemingway and Robie (1994). The enthalpy of formation of dolomite was determined by HCl solution calorimetry at 300.15 K; reference phases were prepared by heating reagent grade CaO and MgO. Hemingway and Robie (1994) acknowledged that several aspects of their experiment were problematic, among them the corrections and potential systematic errors based on the liberation of CO2 in the calorimeter. In addition, there is an assumption that the properties of the reference materials, in particular their crystallinity, are the same as the material used to determine the enthalpies of formation.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1 Literature values of the dolomite solubility product

 

View this table:
[in this window]
[in a new window]
  		Table 1 Literature values of the dolomite solubility product{ddagger}{ddagger}{ddagger}

 
Even though the value of Robie et al. (1978) is widely accepted, experimental confirmation by similar techniques is lacking. The enthalpy of formation determined differs by an order of magnitude from an earlier value based on measured heat capacity of dolomite, reported by one of the coauthors (Stout and Robie, 1963). Furthermore, the more recent acid solution and metal oxide melt calorimetric measurements of Navrotsky and Capobianco (1987) and Chai and Navrotsky (1993) do not confirm the value of Robie et al. (1978) and are themselves equivocal. Navrotsky and Capobianco (1987) noted that the saline acid solvent used in the acid solution measurements can be troublesome since the laws of dilute solutions cannot apply to the complexity of ionic interactions in the solution. As discussed by Hemingway (1991), the use of oxide melt calorimetry can be problematic for phases that liberate a volatile phase, as there will not be a stoichiometric match of the volatile component dissolved in the solvent between the reference phase and the phase of interest.

The numerous other studies of dolomite solubility neither yield consistent and reliable solubility products nor verify the value derived from the data of Robie et al. (1978). The most direct determination of the solubility of a mineral is the measurement of reaction equilibria in water (Nordstrom et al., 1990). Although several studies of dolomite dissolution in deionized water in contact with a known partial pressure of CO2 have been reported (Yanat'eva, 1952; Halla and Van Tassel, 1965; Garrels et al., 1960), the studies are unsatisfactory due to one or more experimental problems: (i) insufficient equilibration time; (ii) closed systems with respect to CO2; (iii) water loss; (iv) inconsistency between measured pH, partial pressure of CO2, and HCO-3 concentration; (v) inadequate measurements of important species; and, coupled to this, (vi) no proof of congruent dissolution or charge balance. Interpretations of in situ groundwater measurements (Hsu, 1963; Barnes and Back, 1964) are hampered by questions of attainment of equilibrium conditions with respect to dolomite. Free energies of dolomite have also been determined by linear programming (Berman, 1988) or least-squares analysis (Holland and Powell, 1990) on a chosen set of thermodynamic data; however, the results depend on the values chosen and the mathematical technique used. As stated by Nordstrom et al. (1990), "major carbonate mineral solubilities and their associated ion pairs are reliable except for dolomite..."

Quite unlike the many attempts at measuring dolomite solubility and despite the extensive study of the kinetics of dissolution of calcium carbonates, the kinetics of dolomite dissolution appears to have largely escaped study except for Busenberg and Plummer (1982), who measured weight loss of single dolomite crystals, and Chou et al. (1989), who analyzed dissolution products of fine dolomite particles. Both of these studies measured dolomite dissolution in response to pH (Chou et al., 1989) or pH and pCO2 (Busenberg and Plummer, 1982) far from presumed equilibrium with dolomite; in these studies, dolomite was allowed to dissolve into solutions that contained only that Ca2+, Mg2+, and alkalinity released into solution supplied only by dissolution of the dolomite itself.

Busenberg and Plummer (1989) suggested that the CaCO3 component of dolomite dissolves faster than the MgCO3 component and that dolomite dissolution is a two-step reaction largely rate dependent upon the dissolution of the MgCO3 component. The suggestion is supported by the four orders of magnitude difference in rate constants for magnesite and calcite found by Chou et al. (1989). Expanding on that hypothesis, Chou et al. (1989) developed a model for dissolution of dolomite based on successive reactions. In the first reaction, dissolution of the CaCO3 component is close to equilibrium because of slow dissolution of the MgCO3 component:

(1)
with the equilibrium constant

(2)
where XMgCO3 is activity of MgCO3 at the dolomite surface.

The second, rate-limiting reaction is the dissolution of the MgCO3 component:

(3)
with the forward dissolution rate, Rf

By substitution of Eq. [2], the forward dissolution rate is therefore given by

(4)

The objectives of the present study were (i) to determine a reliable value of the solubility product of dolomite by measuring dolomite dissolution in Ca-Mg-HCO3/CO3 solutions designed to bracket the presumed solubility product of dolomite, thereby bracketing the most undersaturated solution(s) in which dolomite did not dissolve and the least undersaturated solution in which dolomite did dissolve and (ii) to test the successive reaction model of Chou et al. (1989) for dolomite dissolution near equilibrium.


   Materials and methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results and discussion
 Conclusions
 REFERENCES
 
Sample Preparation and Characterization
Dolomite crystals from Butte, MT (Ward's Natural Science Establishment, Rochester, NY), were hand crushed with mortar and pestle and dry sieved to <0.25 mm. To remove fines, the dolomite powder was acid washed with 0.05 M HCl for 2 min (Busenberg and Plummer, 1982). The powder was rinsed thoroughly with deionized water, filtered (Whatman no. 44, Whatman International Limited, Springfield Mill, England), and freeze-dried.

Surface area of the dolomite was 0.25 m2 g-1, as determined by adsorption of N2 with a Quantasorb Sorption Analyzer (Quantacrome Corporation, Syosset, NY) and the Brunauer–Emmett–Teller (BET) adsorption equation (Adamson, 1976). Dolomite composition was determined by dissolution in excess concentrated HNO3 and analyzed for Ca, Mg, Fe, Al, and Mn by inductively coupled plasma atomic emission spectroscopy (ICP-AES). Reagent grade calcium carbonate (Fisher Scientific, Pittsburgh, PA) used as the calcite seed for the experiment had a BET surface area of 0.37 m2 g-1.

Dolomite and calcite samples were characterized by x-ray diffraction (XRD) and energy dispersive x-ray (EDAX) analyses (Minnesota Agricicultural Experiment Station). A Philips APD 3720 x-ray diffractometer (Eindhoven, the Netherlands) with a CuK{alpha} source was used for all XRD determinations. Random powder samples were run from 10 to 70° 2{theta} with a step size of 0.025° for 2 s step-1. Compositional spectra were performed using energy dispersive x-ray analysis with an EDAX model Power MX EDS unit (EDAX, Mahwah, NJ).

Experimental Procedures
Solutions of varying concentrations of Ca2+, Mg2+, and were prepared by mixing stock solutions of CaCl2, MgCl2, NaHCO3, and NaCl (certified ACS reagents; Fisher Scientific) stored under {approx}0.1 MPa ({approx}1 atm) CO2. Three sets of solutions with Mg/Ca ratios of 1:1, 2:1, and 4:1 were prepared with a pIAPdolomite [ = -log (Ca2+)(Mg2+)(CO2-3)2], where values in parentheses are measured activities] range of 16.0 to 19.0, at 0.094 MPa (0.94 atm) CO2 (i.e., 1 atm CO2 corrected for 262 m altitude at Madison, WI, and for saturation with water vapor) and 25.0°C (Table 2) . All solutions were set to have the molarity of alkalinity (HCO-3 + 2 CO2-3) equal to two times the molarity of Ca2+, with an ionic strength of ~0.1 M. Sufficient NaCl was added to the more dilute solutions to adjust to the target ionic strength.


View this table:
[in this window]
[in a new window]
  		Table 2 Initial solution compositions for Ca–Mg–HCO3–CO3 solutions for dolomite solubility study.{dagger}

 
Dolomite and calcite seed (0.5 g, sieved <0.25 mm) were each added to aliquots of 25-mL initial solutions in 40-mL polycarbonate reaction tubes. The calcite seed was used to test the reliability of the experimental design by comparison of the equilibrium pIAPcalcite reached by both dissolution and precipitation, with the well-established calcite solubility product of Plummer and Busenberg (1982). Controls were established in which no mineral seed was added. In addition, two tubes were set up with calcite and dolomite in deionized water. The head space of each reaction tube was flushed with CO2 and tubes were closed with friction-fit caps fabricated from the CO2-permeable, water-impermeable membranes (Biomedical Polymers, Leominster MA), similar to those used by Hooper and Kittrick (1986). The reaction tubes were placed in a hermetically sealed polycarbonate vessel through which water-saturated CO2 was continuously flushed at a rate of 20 L d-1, maintaining a positive pressure of 100 Pa (1-cm H2O head). The vessel was maintained at 0.09 MPa (0.94 atm) CO2 and 25.0 ± 0.2°C and shaken on an orbital shaker with a 2-h on/1-h off cycle for the duration of the experiment.

At 7, 14, 28, 56, 112, 224, 448, and 672 d, the membranes were removed under a CO2 atmosphere and 2-mL aliquots of the supernatant were withdrawn from each tube by syringe. The tubes were refilled with CO2 and recapped. The aliquots were filtered through 0.45-µm filters (Millipore Corp., Bedford, MA) by positive pressure, diluted, and immediately analyzed for carbonate alkalinity by Gran titration using 0.050 M HCl with an automatic burette (Dosimat 665, Metrohm Ltd., Herisau, Switzerland) and a pH meter (Model 720A, Orion Research, Boston, MA) by the procedure of Barak et al. (1996). Acidified samples after carbonate analysis were diluted for analysis of Ca2+ and Mg2+ by atomic absorption spectrophotometry (GBC Model 902, Perkin Elmer, Wellesley, MA) using matrix-matched standards. A subset of results was verified by ICP-AES analysis (Model 3400B, ARL Applied Research Laboratories, Franklin, MA).

Data Analysis
For the design of the initial solutions and interpretation of the results, ion activities were calculated using the chemical speciation program SPECIES (Barak, 1990), which uses the Davies equation for calculation of activity coefficients and constants (Table 3) at 25°C and zero ionic strength taken from Smith and Martell (1989) for calculating chemical speciation.


View this table:
[in this window]
[in a new window]
  		Table 3 Formation constants for chemical speciation from Smith and Martell (1989)

 
Rates of dolomite dissolution within each tube were calculated from alkalinity quantities at each time step, using the first derivative of a sliding, unequally spaced five-point cubic polynomial (Gorry, 1991) for the eight sampling periods. Alkalinity quantities were determined from measured alkalinity concentrations and solution volumes at each sampling time. Corrections for evaporation were applied to the sample volumes of dolomite-seeded tubes. Based on small changes in alkalinity concentrations of the unseeded blanks, daily evaporation through the plastic cap membranes was found to range from 1.3 to 2.8 µL d-1 and averaged 2.2 µL d-1.

To determine the solubility product, a statistical analysis was performed to compare the calculated dissolution rates of dolomite-seeded Ca-Mg-HCO3/CO3 solutions with the unseeded solutions. The calculated rate of change for the unseeded solutions, in the absence of carbonate dissolution, represents therefore the experimental error in the determination of the dissolution rates and lines representing 95% confidence intervals for the mean and prediction limits for individual points (Weisberg, 1985) were calculated. Calculated rates of the dolomite-seeded solutions were then compared with these limits and conclusions drawn regarding the null hypothesis that there was no difference in alkalinity dissolution between the control solutions and the dolomite-seeded solutions. Where indicated, a t-test was used in conjunction with the standard error of the intercept of regression lines to determine if the intercept was significantly different from zero at {alpha} = 0.05 (Weisberg, 1985).


   Results and discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
Conclusions
 REFERENCES
 
Sample Characterization
The identity of the dolomite was confirmed by the XRD analysis; within the spacings measured, 21 peaks were consistent with dolomite data from Joint Committee on Powder Diffraction Standards card no. 11-78 (Joint Committee on Powder Diffraction Standards, 1974) and no other peaks were observed. The strong acid digestion of the dolomite indicated the following mass percentages: 20.9% Ca, 12.7% Mg, 0.5% Fe, 0.2% Mn, and 63.4% CO3. The EDAX spectrum of dolomite particles (not shown) did not indicate inclusion of Fe or Mn with Ca and Mg, indicating that Fe and Mn were not carbonate impurities. The average composition of the dolomite is therefore calculated as Ca1.001Mg0.999 (CO3)2.

Precipitation
There was no evidence of precipitation of dolomite in the complete set of 21 solutions by 672 d. Precipitation did occur in two solutions that were initially oversaturated with respect to calcite, that is, with initial pIAPcalcite values greater than the calcite solubility product of Plummer and Busenberg (1982) (pKs calcite = 8.48 ± 0.02). A 1:1 reduction in Ca2+/alkalinity concentrations occurred by the first sampling date in these solutions, indicating precipitation; no further change in concentrations occurred (Fig. 2) . Initial pIAPcalcite values of 8.03 and 8.21 (x axis) increased to 8.39 and 8.38 (y axis), respectively (Fig. 3) , close to the log solubility product of calcite. (Decreases in pIAPcalcite, given on the y axis, from control conditions at any time, given on the x axis, reflect increases in Ca and alkalinity concentrations due to dolomite dissolution). When the Ca-Mg-HCO3/CO3 solutions were seeded with calcite, the equilibrium pIAPcalcite reached by both dissolution and precipitation of calcite (Fig. 4) also closely agreed with the log solubility product. The identity of the CaCO3 precipitate is therefore presumed to be calcite in both cases. Furthermore, measuring reasonable calcite solubility values by this CO2-permeable membrane technique adds weight to the soundness of applying this technique to dolomite studies.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 2 Dissolution and precipitation in dolomite-seeded Ca–Mg–HCO3/CO3 solutions and deionized water: change in Ca2+ concentration over values in control unseeded solutions ({Delta}Ca2+) against change in alkalinity over values in control unseeded solutions ({Delta}Alk)

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 3 Calculated pIAPcalcite [= -log (Ca2+)(CO2-3)] in dolomite-seeded against control unseeded Ca–Mg–HCO3/CO3 solutions with 1:1, 2:1, and 4:1 Mg/Ca at 7, 14, 28, 56, 112, 224, 448, and 672 d. Dashed line indicates no difference between seeded and unseeded control

 


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 4 Calculated pIAPcalcite in calcite-seeded against control unseeded Ca–Mg–HCO3/CO3 solutions with 1:1, 2:1, and 4:1 Mg/Ca at 7, 14, 28, 56, 112, 224, and 448 d. Dashed line indicates no difference between seeded and unseeded control

 
The lack of evidence for precipitation of dolomite from supersaturation is not unexpected, as dolomite has never been precipitated spontaneously at room temperature. Even so-called magic ingredients and seed crystals do not trigger crystallization according to the reactions Ca2+ + Mg2+ + 2CO2-3 = CaMg(CO3)2 (Usdowski, 1994). Only protodolomites (poorly ordered, high Ca dolomites) have reportedly precipitated from solutions at room temperature under laboratory conditions (Deer et al., 1992; Deelman, 1981; Busenberg and Plummer, 1989). Kinetic limitations relating to the high hydration energy of Mg2+ are probably the cause for the lack of dolomite precipitation under ambient conditions (Garrels et al., 1960; Hartman, 1982).

Dolomite Dissolution
The dolomite sample dissolved into deionized water and into 10 of the 21 Ca-Mg-HCO3/CO3 solutions, those with initial pIAPdolomite of 19.0, 18.5 and 18.0 in all series, as well as that solution with initial pIAPdolomite of 17.5 in the 4:1 series (Fig. 5) . Dissolution was indicated by a decrease in pIAPdolomite compared with the unseeded control solutions. Samples with initial pIAPdolomite <=17.0 showed no evidence of either dolomite dissolution or precipitation, and were therefore either at equilibrium or supersaturated. Increases in pIAPdolomite found in two samples reflect decreases in Ca and alkalinity concentrations due to calcite precipitation (see discussion above). Concentrations of Ca2+ and Mg2+ increased proportionately at a 1:1 ratio throughout the duration of the 672-d experimental period (Fig. 6a) , in accordance with the 1:1 Ca/Mg stoichiometry of the dolomite solid, indicating congruent dissolution. The sum of the increase in Ca2+ and Mg2+ concentrations equaled the increase in alkalinity concentrations (Fig. 6b), demonstrating that all ions required for charge balance were analyzed.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 5 Calculated pIAPdolomite [= -log (Ca2+)(Mg2+)(CO2-3)2] in dolomite-seeded against control unseeded Ca–Mg–HCO3/CO3 solutions with 1:1, 2:1, and 4:1 Mg/Ca at 7, 14, 28, 56, 112, 224, 448, and 672 d. Dashed line indicates no difference between seeded and unseeded control

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 6 Dissolution of dolomite in undersaturated Ca–Mg–HCO3–CO3 solutions with 1:1 Mg/Ca and in deionized water: (a) Change in Ca2+ over values in control unseeded solutions against change in Mg2+ concentrations over values in control unseeded solutions ({Delta}Ca2+ and {Delta}Mg2+, respectively). (b) Change in the sum of Ca2+ and Mg2+ concentrations over values in control unseeded solutions({Delta}Ca2+ + {Delta}Mg2+) against change in alkalinity over values in control unseeded solutions ({Delta}Alk)

 
For those dolomite samples dissolving into solutions with an initial pIAPdolomite of 16.5 to 19.0 with no evidence of calcite precipitation, the kinetic model of successive reactions of Chou et al. (1989) was tested by plotting dolomite dissolution rates based on changes in the quantity of alkalinity in the dissolution tubes against the term (H+)2/(Ca2+)(HCO-3) from Eq. [4]. The resulting plot (Fig. 7) indicates a relationship that is highly significant (r2 = 0.733, P < 0.001); moreover, the y intercept in this empirical relationship is not statistically different from zero, as predicted by the model, and the data are not segregated into 1:1, 2:1, and 4:1 Mg/Ca series. The results indicate that the rate equation of Chou et al. (1989)(Eq. [4] above), based on a two-step dissolution process for dolomite, with the dissolution of the MgCO3 component as the second, rate-limiting reaction, describes the data well.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 7 Evaluation of successive reaction kinetic model of Chou et al. (1989) based on dolomite dissolution near equilibrium. Rate of change in alkalinity in dolomite-seeded Ca–Mg–HCO3–CO3 solutions, pIAPdolomite ranging from 17.0 to 19.0, plotted against (H+)2/(Ca2+)(HCO-3)

 
Dolomite Equilibrium
The use of series of solutions designed to bracket the presumed solubility product follows a suggestion by Garrels and Wollast (1978) for magnesian calcite and is similar to the technique used by Lindsay and coworkers (Norvell and Lindsay, 1982; Ma and Lindsay, 1990; El-Falaky et al., 1991) for determining equilibrium metal activities in soil solutions. Mineral solubilities are ideally determined based on data collected from both supersaturated and undersaturated solutions, for example, Plummer and Busenberg (1982) for calcite. However, the repeated failure of dolomite to precipitate under laboratory conditions at room temperature in numerous studies (Usdowski, 1994), as well as here, precludes the use of these criteria for establishing the solubility in water of dolomite. On the other hand, careful use of data collected from undersaturation has been used for establishing the solubility of minerals such as vaterite and aragonite (Plummer and Busenberg, 1982), which did not successfully precipitate under laboratory conditions because of calcite precipitation.

Evidence for dolomite dissolution is sought by statistical inference using the null hypothesis that alkalinity is not increasing during the 672-d reaction period and bracketing dolomite solubility between the most saturated solution for which the null hypothesis can be refuted and the least saturated solution for which it cannot. Changes in alkalinity were used as the measure of dissolution since analytical errors in their determination by computer-assisted Gran titration, ±0.2 mM on the evaporation-corrected blanks or <2% of total alkalinity, were smaller than those of atomic absorption analysis of cations. Calculated rates of alkalinity change in the unseeded Ca–Mg–HCO3/CO2-3 solutions were used as a measure of experimental error in determining rates of alkalinity change. Control lines indicating 95% confidence and prediction limits were compared with the rate of alkalinity change as a function of pIAPdolomite for the Ca–Mg–HCO3/CO2-3 solutions seeded with dolomite (Fig. 8) . From this, it can be seen that solutions with a pIAPdolomite of 17.4 or greater fell outside the control lines and were dissolving. It is somewhat more difficult to choose a set of solutions representing the bottom bracketing value, but it can be conservatively stated that no evidence could be found that any of three samples with an initial pIAPdolomite of 17.0 or below were dissolving within the 672-d period of observation.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 8 Relationship between the rate of change in alkalinity in dolomite-seeded Ca–Mg–HCO3/CO3 solutions and pIAPdolomite. Control lines are 95% confidence and prediction lines from the unseeded Ca–Mg–HCO3/CO3 solutions, corresponding with the hypothesis of no alkalinity change due to dolomite dissolution. Lines connect individual dolomite-seeded solutions from 0 to 672 d

 
Although the solutions indicated above may be at, or nearly at, equilibrium, it cannot be ruled out a priori that mineral phases other than dolomite are dictating equilibrium. However, the three most oversaturated solutions in which dolomite was dissolving (initial pIAPdolomite 17.5) had p(Ca2+)(CO2-3) values between 8.76 and 9.03 (Table 4) , indicating considerable undersaturation with respect to calcite, for which the value of 8.48 was determined by Plummer and Busenberg (1982) and was found to be 8.50 in calcite-seeded solutions here. Compared with pKs Mg-calcite values (= -log[(Ca2+)1-x(Mg2+)x (CO2-3)]) appropriate for magnesian calcites Ca1-xMgx CO3 containing 5, 10, and 15% MgCO3 (Busenberg and Plummer, 1989), those solutions at or near dolomite equilibrium (initial pIAPdolomite 17.5 and 17.0) (Table 4) are undersaturated with respect to magnesian calcites, particularly the 4:1 Mg/Ca series, which was 2.03 to 4.07 times undersaturated with respect to magnesian calcites containing up to 15% MgCO3. Therefore, based on the undersaturation with respect to calcite and magnesian calcites of those solutions chosen herein to bracket the dolomite solubility product, it may be concluded that these solutions are not in equilibrium with other carbonate solids and are not determining the dolomite solubility results reported. It should be noted that dolomite is differentiated from the solid solution series of magnesian calcites [Ca1-xMgx(CO3)2] both by the degree of crystal ordering in dolomite (Lerman, 1965; Busenberg and Plummer, 1989) and by a large miscibility gap from 20 to 49% MgCO3 in the solid solution series (Busenberg and Plummer, 1989; Konigsberger and Gamsjager, 1992). The ideal structure is alternating sheets of Ca and Mg perpendicular to the threefold axis, separated by layers of CO3 (Deelman, 1981). Dolomite composition is normally close to ideal, but does permit up to 2 mole % excess CaCO3 (Reeder, 1983).


View this table:
[in this window]
[in a new window]
  		Table 4 Comparison of measured compositions of solutions near apparent dolomite equilibrium with calcite and magnesian calcite solubility products and calculated extent of undersaturation

 

   Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
REFERENCES
 
Dolomite solubility was measured in Ca-Mg-HCO3/CO3 solutions in a long-term study by maintaining a system open with respect to CO2, but closed with respect to water loss, by use of CO2-permeable membranes over the reaction vessels. The identity and purity of the dolomite was established by x-ray diffraction and chemical analysis. There was no evidence of dolomite precipitation by 672 d in the solutions, as expected. When dolomite was found to be dissolving, dissolution was congruent. Based on statistical inference by comparison with unseeded controls, the pKdolomite is between 17.4 and 17.0, expressed as 17.2 ± 0.2. Because these solutions were undersaturated with respect to calcite and magnesian calcites and, when dissolving into Ca-Mg-HCO3/CO3 solutions and in deionized water, dolomite dissolution was congruent, dolomite is the most likely mineral phase controlling solubility.

The results significantly narrow the range of the solubility product of dolomite, indicating that the majority of reported values in the literature are unreliable. The solubility products that are outside this range include both dissolution studies and hydrothermal studies, most notably the recent metal oxide and acid calorimetric work of Navrotsky and coworkers. The value from the dissolution study of Halla and Van Tassel (1965) calculated from Ca2+ and alkalinity data, pKdolomite = 17.0, does fall in the range; however, the value calculated from the same experiment using pH and CO2 data, pKdolomite = 16.6, does not. Our results support the value of Robie et al. (1978) and Hemingway and Robie (1994), which has not been confirmed by direct solubility measurements heretofore.Hemingway Robie 1973


   ACKNOWLEDGMENTS
 
We wish to thank Mr. G. Ahlstrand of the Electron Microscopy Facility of the Minnesota Agricultural Experiment Station for assistance with x-ray microanalysis and Dr. E. Nater (Dep. of Soil, Water, and Climate, Univ. of Minnesota) for his assistance in x-ray diffraction analysis and critical reading of the manuscript draft. Financial support from the Graduate School and the College of Agriculture and Life Sciences (as a Hatch Grant) of the University of Wisconsin-Madison is gratefully acknowledged.

Received for publication October 26, 1999.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 Geol Soc Am BullHome page
E. L. Williams, K. J. Szramek, L. Jin, T. C.W. Ku, and L. M. Walter
The carbonate system geochemistry of shallow groundwater-surface water systems in temperate glaciated watersheds (Michigan, USA): Significance of open-system dolomite weathering
GSA Bulletin, May 1, 2007; 119(5-6): 515 - 528.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
R. A. Kraimer, H. C. Monger, and R. L. Steiner
Mineralogical Distinctions of Carbonates in Desert Soils
Soil Sci. Soc. Am. J., September 29, 2005; 69(6): 1773 - 1781.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sherman, L. A.
Right arrow Articles by Barak, P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Sherman, L. A.
Right arrow Articles by Barak, P.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Sherman, L. A.
Right arrow Articles by Barak, P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome