Published in J. Environ. Qual. 32:2354-2363 (2003).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA
TECHNICAL REPORTS
Vadose Zone Processes and Chemical Transport
Aluminum Effect on Dissolution and Precipitation under Hyperalkaline Conditions
I. Liquid Phase Transformations
Nikolla P. Qafoku*,
Calvin C. Ainsworth,
James E. Szecsody and
Odeta S. Qafoku
O.S. Qafoku, Pacific Northwest National Lab., Interfacial Geochemistry Group, 902 Battelle Blvd., P.O. Box 999, MSIN: K3-61, Richland, WA 99352
* Corresponding author (nik.qafoku{at}pnl.gov).
Received for publication January 3, 2003.
 |
ABSTRACT
|
|---|
Substantial amounts of self-boiling, Al-rich, hyperalkaline, and saline high-level waste fluids (HLWF) were deposited to the vadose zone at the Hanford Site, in Washington State. The objective of this study was to investigate the effects of similar fluids on the extent of dissolution and precipitation in the sediments. Metal- and glass-free systems were used to conduct batch experiments at 323 K under CO2 and O2 free conditions. Base-induced dissolution of the soil minerals was rapid in the first 48 h as indicated by immediate releases of Si and Fe into the soil solution. Potassium release lagged behind and dissolution of K-bearing minerals (mica and K-feldspar) proceeded faster only after 2 to 3 d of the experiment. Silicon and Fe release exhibited high dependence on aqueous [Al] (rate orders <-1), because Al decreased free OH concentration in the contact solution and probably inhibited soil mineral dissolution. Initial K release exhibited low dependence on [Al] (fractional rate orders). Initial dissolution rates calculated based on Si release varied with aqueous [Al] from 29.47 to 4.35 x 10-12 mol m-2 s-1. Aluminum participated in the formation of the secondary phases (precipitation rates of 10-8 mol s-1) but the overall precipitation rate of alumino-silicate secondary phases was probably controlled by aqueous [Si] (rates of 10-9, and rate constants between 0.0054 and 0.0084 h-1). The changes in the soil solution chemistry (release of K, Si, Fe, and other elements) may play a significant role in the fate of radionuclides and contaminants like Cs, Sr, Cr, and U in the Hanford sediments.
Abbreviations: DOE, U.S. Department of Energy • EC, electrical conductivity • EGME, ethylene glycol monoethylene • ERDF, Environmental Restoration Disposal Facility • HLWF, high level waste fluids • ICP-AES, inductively coupled plasma–atomic emission spectroscopy
|
INTRODUCTION
|
|---|
THE PRODUCTION AND EXTRACTION of Pu and U was a major activity at the U.S. Department of Energy's (DOE) Hanford Site in Washington State, between the early 1940s and the late 1980s. Similar activities occurred at a number of other DOE facilities resulting in waste production, albeit to a lesser extent than Hanford. These include Idaho National Engineering Laboratory (Idaho), Los Alamos National Laboratory (New Mexico), Oak Ridge National Laboratory (Tennessee), Sandia National Laboratory (New Mexico), and the Savannah River Site (South Carolina). The HLWFs resulting from extraction of spent fuel rods at the Hanford Site were stored in 177 underground tanks. Sixty-seven tanks are known or suspected to have leaked, allowing millions of liters of HLWF to migrate into the underlying vadose zone. Of particular interest is HLWF, which leaked from tanks in the S-SX tank farm containing the self-boiling REDOX process waste (Jones et al., 2000). These fluids carried a large inventory of 137Cs, which generated a high heat load. Soils beneath these tanks are still at temperatures >323 K, to a depth of about 40 m. The HLWF in the S-SX tanks contained large concentrations of NaNO3, hydroxyl ions (pH > 13), dissolved Al, and substantial quantities of 137Cs, 60Co, 90Sr, 99Tc, 234, 238U, and Cr (Jones et al., 2000; Serne et al., 2001a).
The mineralogy of the Hanford vadose zone includes quartz and basaltic glass, plagioclase feldspars (members of the sequence from albite to anorthite), K-feldspars, mica (biotite and muscovite), smectite, chlorite, and kaolinite. All of these mineral phases may undergo dissolution on reaction with hyperalkaline and saline waste fluids. In the alkaline pH range, there is a basic dissolution due to hydroxide bound on the resulting deprotonated surface groups (Stumm, 1997). The adsorption of a catalyst, such as OH, to surface sites proximal to the metal may influence the bond strength and thus affects dissolution (Ganor and Lasaga, 1998). The presence of OH in higher concentrations in the sediments under hyperalkaline conditions permits the dissolution reaction to follow a more favorable path and enhances, as a consequence, the dissolution reaction rate. In addition, the silicate (quartz and feldspars) dissolution at high pH is also promoted in the presence of different cations, like Na+, which facilitates the release of Si from their structure (Blum and Stillings, 1995).
On the other hand, Al not only decreases the OH free concentration in the contact solution as a result of aluminate 
formation, but it also inhibits the base-promoted dissolution reactions (Ganor and Lasaga, 1998). Many of the tank fluids contained large concentrations of dissolved Al, which presumably stayed in solution as aluminate ions and aluminate polymers. It has been recently recognized that the dissolution reaction rate of alumino-silicates is controlled by the decomposition of a Si-rich, Al-deficient precursor complex (Oelkers et al., 1994). The hydrolysis requires the breaking of at least two different types of metal–oxygen bonds in sequential steps (Devidal et al., 1997). The first step is the reversible exchange of OH for Al releasing Al
-4 into the soil solution, which involves the breaking of Al–O bonds of the Al–O–Si groups producing siloxane surface groups. The second step is the reversible decondensation of siloxane groups forming the Si-rich surface precursors (Devidal et al., 1997). The detachment of this precursor (the breaking of the Si–O bonds) is the rate-limiting step that controls the rate of dissolution in many alumino-silicates (Oelkers et al., 1994; Devidal et al., 1997; Oelkers and Schott, 1999). The presence of Al-4 in the soil solution may affect the extent of the exchange reaction between OH and Al, as aluminate, in the first step. For these reasons, the effect of the aqueous Al concentration on dissolution rates must be taken into account explicitly (Oelkers et al., 1994).
Because of dissolution, Si, Fe, K, Al, Ca, Mg, and Mn may be released into the soil solution. The dissolution reaction products like Si and Fe are likely to over-saturate the soil solution with respect to thermodynamically stable secondary phases that may precipitate under such extreme conditions. Because Si, Fe, and Al take part in the structure of many ubiquitous secondary phases, the precipitation rate should depend on their concentrations in the soil solution. However, the nature of the secondary phases formed is not known and it would be necessary to determine experimentally which of these elements present in the soil liquid phase plays a nontrivial role in controlling the rate of precipitation in the Hanford sediments. In addition, the amount of elements mobilized and immobilized on dissolution and precipitation depends on the extent of these processes in the Hanford sediments under hyperalkaline conditions, which is not known.
Published accounts on the effects of aqueous Al, hyperalkalinity, and hypersalinity on the dissolution and precipitation of soil mineral assemblages are rare. In light of the potential environmental uncertainties surrounding leaks and discharges from high-level waste at Hanford, other DOE sites, and in Europe, we investigated the impact of Al-rich, high pH and saline solutions on the extent of dissolution and precipitation in the sediments. For this reason, changes in the aqueous phase chemical composition, i.e., Si, Fe, K, Al, and other elements' concentrations were closely followed. We also studied the effect of Al concentration in the contact solutions on the multiple phase dissolution reactions. The partial rate laws for this effect were experimentally obtained.
|
MATERIALS AND METHODS
|
|---|
Sediments
The sediments used in these experiments that were collected in February 2000 from the 200 Area at the Hanford Site in Washington are very similar to native uncontaminated sediments found underneath the waste tanks at the Hanford Site. The fine and course sand sediment was collected at the 7-m depth in the Environmental Restoration Disposal Facility (ERDF) pit, while the fine sandy silt sediment was collected at the 13-m depth in the Navy submarine reactor core pit. These samples were taken at these location and depth to yield material very similar with those above the Plio-Pleistocene layer under tanks. A mixture (50:50 on a wt. basis) of these two sediments was used in the Batch 1 experiments. Serne et al. (2001b) characterized the mineralogy of the Hanford sediments underneath the waste tanks at the Hanford Site, which is dominated by quartz (30–80%) and plagioclase feldspar (5–20%), with minor amounts (<10%) of K feldspar and amphibole. The clay fraction (<2 µm) is dominated by four clay minerals: illite (mica, 15–40 wt. %), smectite (30–40 wt. %), chlorite (15–20 wt. %), and kaolinite (<10 wt. %), with minor amounts of quartz, feldspar, and amphibole (5–10 wt. %).
Batch Experiments with Hanford Sediments
Sodium hydroxide and NaNO3 are two of the principal electrolyte components present in the waste tanks, which could affect the solubility of aluminum hydroxide compounds (Felmy et al., 1999). Since the NaOH–NaNO3–H2O system is the simplest chemical system that should be representative of the more complex mixed electrolytes present in the tanks, at least regarding the solubility of aluminum hydroxide compounds (Felmy et al., 1999), our contact solutions were all 1 M NaOH and 1 M NaNO3. Four solutions were used in these experiments, in which Al concentration was the only variable; these solutions were respectively 0.055, 0.11, 0.165, and 0.22 M Al, which was added as Al(NO3)3·9H2O in the 1 mol L-1 NaOH and 1 mol L-1 NaNO3 background solution. Aluminum decreased the free OH aqueous concentration in solution because four free OH are consumed to form an aluminate molecule under hyperalkaline conditions as it is presented in Table 1.
We conducted two similar experiments with the same solutions. The first experiment consisted of 20 treatments: 4 Al concentrations by 5 time periods (3, 7, 14, 21, and 42 d), and was conducted in duplicates. The second one was conducted to investigate the far from equilibrium (initial) dissolution in the first 3 d of contact time, where the contribution of reverse reaction (precipitation) is most likely negligible. This experiment consisted of 36 treatments: 4 Al concentrations by 9 time periods (20 min, 1, 2, 4, 8, 16, 24, 36, and 48 h). Both these experiments are hereafter referred to as the Batch 1 study.
Metal- and glass-free systems were used to conduct batch experiments at 323 K. The equivalent of 0.02000 kg oven dry mixtures of the two Hanford sediments were added into 250-mL plastic FEP batch reactor containers that were selected to be resistant to high base concentrations. The specific surface area of the sediments determined with the ethylene glycol monoethylene (EGME) method was 51285 m2 kg-1. The solutions were prepared at 323 K in a glove box under O2 and CO2 free conditions. All solutions were prepared with ultra-pure DI-water and reagent-grade chemicals. Two hundred mL of the respective solutions were mixed with the sediment in the 250-mL containers (soil/solution ratio of 1:10) inside a glove box under O2 and CO2 free conditions. The 250 mL containers were put inside 20-L HDPE plastic containers where He2 gas was flowing continuously. The 20 L containers were accommodated inside a thermostat chamber where the temperature was kept at 323 K for the duration of the experiment. In both batch experiments, the samples were shaken slowly by hand every day. In the short time period treatments (<1 d) the samples were shaken at the beginning and the end of the experiment. At the end of each time period, the respective 250-mL containers were taken out of the thermostat chamber and centrifuged at 48.33 rotations per second for 540 s to separate the phases. pH and electrical conductivity (EC) were measured immediately in the supernatants. A solid-state pH electrode was used to minimize interferences that may affect pH readings in hyperalkaline and saline solutions. To avoid possible supernatant precipitation, which results from exposure to the cooler room temperature, two subsamples from each sample of supernatant were immediately diluted to ratios 1:10 and 1:100 on a mass basis (for the 1:10 dilutions, 0.002 kg of supernatant were mixed with 0.018 kg of DI-water, and for the 1:100 dilutions, 0.0002 kg of supernatant were added to 0.0198 kg of DI-water) and stored in vacuum before analyses. All operations were performed under O2 and CO2–free conditions. Aluminum, Si, K, Fe, Ca, Mg, and Mn contents were determined in supernatants using inductively coupled plasma atomic emission spectroscopy (ICP-AES).
Dissolution Kinetics in the Sediments
The chemical reaction for the base-induced mineral dissolution in the sediments treated with Al-rich, hyperalkaline, and saline solutions may be written as follows:
 | [1] |
where the concentrations of different chemical elements released on dissolution are written in square brackets. Because the sediments at Hanford are comprised of different soil minerals, the dissolution rate calculated based on Si, Fe, and K release into the soil solution may be written as the sum of the rate expressions that represent the dissolution rate for each soil mineral (the release of Si is usually taken as an example):
To determine the reaction order using the initial rate method, the concentration of one of the reactants is varied, keeping the concentrations of all other reactants in excess and relatively constant for the duration of the experiment. The dual effect of Al concentration first, on decreasing the free OH concentration in the aqueous phase and second, on inhibiting the dissolution becomes apparent by examining the reaction rate dependence on the solution concentration of Al. The initial NaOH and NaNO3 concentrations in the contact solution were 1 mol L-1 in all the treatments, and the dissolution rate orders with respect to initial Al concentration in the contact solution were obtained experimentally from data collected in the Batch 1 study, using the partial rate law for Al effect of the form (Ganor and Lasaga, 1998):
 | [3] |
After taking log of both sides one can write:
 | [4] |
The dissolution rates were calculated using the finite differences method as follows:
 | [5] |
where [Si]t+
t is the Si concentration in the soil solution (mol L-1) at time (t + t), and [Si]t is the Si concentration in the soil solution (mol L-1) at time t. The rate was expressed in the units mol m-2 s-1 and log(rate) was plotted against log [Al]in (log initial Al concentration) to determine the rate order p.
Precipitation Kinetics
The first-order kinetic model was applied to data collected from the batch experiments to calculate the rate constants. The integrated form of that model is:
 | [6] |
where the [A] is the concentration of the reactant at time t, [A]0 is the initial concentration of the reactant, and k is the rate constant for the precipitation reaction. For graphical solution the log of this equation can be taken as (Brezonik, 1993):
 | [7] |
or
 | [8] |
A plot of log[A] vs. t should yield a straight line for reactions following first-order kinetics. The value of k is obtained from the slope, k = -2.3 x (slope). The reaction half-life was calculated as follows:
 | [9] |
This method was used to calculate the rate constant of Al precipitation reactions for all treatments in the Batch 1 study.
The Si and Fe precipitation reaction orders were calculated from data collected in the Batch 1 study, using the "partial rate law" method (Ganor and Lasaga, 1998) presented above:
 | [10] |
After taking log on both sides one may get:
 | [11] |
The Al, Si, and Fe precipitation rates were calculated using the finite difference method, which, in the case of Al, was calculated as follows:
 | [12] |
where [Al]t+t is the Al concentration in the soil solution (mol L-1) at time (t + t), and [Al]t is the Al concentration in the soil solution (mol L-1) at time t. The normalized rate to 1 L was expressed in the units of mol s-1 and, in the case of Si and Fe, the log(rate) was plotted against log [Al]in (log initial Al concentration) to determine the rate order p.
|
RESULTS
|
|---|
Silicon, Iron, Potassium, and Aluminum Fate in the Batch 1 Study
The plots in Fig. 1a and 2a
may be divided in three sections, each corresponding to the first (0–2 d), second (2–7 d), and the third (7–42 d) time periods, which describe best the dynamics of Si and Fe release into or removed from the soil solution. Base-induced dissolution of Si- and Fe-bearing soil minerals at 323 K was especially rapid in the first 4 h of contact time indicated by steeper slopes (Fig. 1b and 2b) and greater dissolution rates (Table 2). The dissolution rates decreased in the time period from 4 h to 2 d (Table 2). Peak concentrations of 0.0060, 0.0028, 0.0018, and 0.0013 mol Si L-1 and 7.86, 5.03, 2.90, and 1.31 x 10-5 mol Fe L-1 were observed in the 0.055, 0.11, 0.165, and 0.22 mol L-1 Al treatments, respectively, and the rate of dissolution was controlled by Al concentration in the contact solution (Table 2). Since increasing trends of concentration with time were observed in the first 2 d of the experiment (Fig. 1b and 2b), this suggests that dissolution was faster than precipitation during this time period.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 1. Changes in Si concentration with time and Al concentration: (a) from 0 to 42 d, and (b) from 0 to 3 d.
|
|
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 2. Changes in Fe concentration with time and Al concentration: (a) from 0 to 42 d, and (b) from 0 to 3 d.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Initial dissolution rates based on Si, Fe, and K released into the soil solution (standard deviations in parenthesis).
|
|
A sharp decrease in Fe and Si concentrations was observed after 2 and 3 d, respectively (Fig. 1a and 2a). It is likely that either the initial Si and Fe source ran out of mass, or the rate of secondary phase precipitation increased significantly in this time period. The Fe concentration in the 0.22 mol L-1 Al treatment remained almost invariable in the time period from 3 to 7 d.
Both dissolution and precipitation processes appear to control Si and Fe concentrations in the third time period from 7 to 42 d, and Si and Fe concentrations reached respective steady state plateaus of approximately 0.0027 and 0.00003 mol L-1 in the 0.055 mol L-1 Al treatment, and about 0.00070 and 0.000016 mol L-1 in the other treatments. While an increase in Al concentration beyond the 0.11 mol L-1 concentration affected Si and Fe release in the first 3 d, this was not observed when the steady state conditions were established.
A significant increase in Si concentration was observed after 21 d in the 0.055 mol L-1 Al treatment, which reached a new maximum of 0.0105 mol L-1 by the end of the experiment (42 d). Since Al concentration in the soil solution decreased from 55 to 9.636 mmol L-1 and 0.741 mmol L-1 by 21 and 42 d, respectively, both the inhibitory effect of Al on dissolution and the rate of precipitation of alumino-silicate secondary phases should have been reduced after 21 d. As a result, a good portion of Si mobilized on mineral dissolution was accumulated in the soil solution during the time period from 21 to 42 d.
Potassium release was sluggish in the first 2 d (Fig. 3b)
, suggesting K-bearing minerals (mica and K-feldspars) did not undergo immediate dissolution and rates were almost the same in different Al treatments (Table 2). Because Na was present in the contact solution in a very high concentration (2 mol L-1), exchange reactions between K and Na were probably taking place in the first 4 h, accounting for the slight increase in the K concentration during this time period (Fig. 3b). The initial Al concentration in the contact solution appears to control K release after 2 d as it is clearly shown in Fig. 3a and 3b. Potassium concentration in the soil solution increased steadily indicating that base-induced dissolution of K-bearing minerals continued for the duration of the experiment (42 d). Peak K concentrations of 0.00397, 0.002973, 0.002078, and 0.001247 mol L-1 were observed at the end of the experiment in the 0.055, 0.11, 0.165, and 0.22 mol L-1 Al(NO3)3 treatments, respectively, while the greatest rate of K release was observed in the time period from 2 to 3 d (Table 2). Since K is less likely to precipitate in secondary minerals and increasing trends were not observed for the Si and Fe concentration after 3 d of the experiment, this also suggests that Si and Fe released on dissolution were consumed in precipitation reactions.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 3. Changes in K concentration with time and Al concentration: (a) from 0 to 42 d, and (b) from 0 to 3 d.
|
|
Aluminum concentration in the soil solution remained almost invariant in the first 2 d of the experiment (Fig. 4)
. Decreasing Al concentrations with time were observed only after 3 d in all treatments, and aqueous Al concentration was drastically reduced by 42 d: for example, in the treatment where the initial Al concentration was 0.055 mol L-1, Al concentration in the soil solution decreased to approximately 0.000741 mol L-1 by 42 d.
Small amounts of Ca, Mg, and a very small amount of Mn were released into the soil solution, but their concentrations remained low and approximately constant throughout the experiment (data not shown).
Dependence of Dissolution Rate on Aqueous Aluminum Concentration
The Al effect on dissolution was immediately visible (by only 20 min of contact time) because different amounts of Si were mobilized in different Al treatments. As a result, Si release by 4 h exhibited high dependence on initial Al concentration in the contact solution (rate order of -1.338) indicating a strong negative effect of Al concentration in dissolution (Fig. 5a) . The same effect was observed by Day 3 of the experiment (Fig. 5b). The dissolution reaction rate order with respect to initial Al concentration in the soil solution was approximately -1, which clearly indicated that Al hindered the dissolution of Si-bearing minerals.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 5. Dissolution reaction order with respect to initial Al concentration in the soil solution, based on Si release by (a) 4 h and (b) 3 d. The straight lines are the linear regression fits and dotted lines are the 95% confidence intervals.
|
|
Similarly to Si, the rate of Fe release in the first 4 h exhibited high dependence on initial Al concentration (rate order of -1.498, Fig 6a)
. It is quite likely that Si and Fe were released from the same source that underwent immediate dissolution in the first hours of the experiment (most likely smectite). Poor correlation was found between initial Al concentration and Fe release rates by Day 2 (R2 = 0.857). No correlation was found between the log rates of K release in different treatments by 4 h, and rate dependence on initial Al concentration was low (fractional rate order) (Fig. 7a and 7b)
.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 6. Dissolution reaction order with respect to initial Al concentration in the soil solution, based on Fe release by 4 h. The straight lines are the linear regression fits and dotted lines are the 95% confidence intervals.
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 7. Dissolution reaction order with respect to initial Al concentration in the soil solution, based on K release by (a) 2 d and (b) 3 d. The straight lines are the linear regression fits and dotted lines are the 95% confidence intervals.
|
|
Aluminum Precipitation Rates
The pseudo-first order kinetic model was applied to data collected in the second experiment of the Batch 1 study to calculate the rate constant and reaction half-life (the time required to consume half of the reactant's initial concentration) in each treatment (Fig. 8)
. In addition, Al precipitation rates were calculated in different time periods and treatments (Table 3). From these results, one can clearly see that first, the reactions' half-lives were different for different initial Al concentrations in the contact solution (Fig. 8); second, the precipitation rates were similar, and varied within the same order of magnitude in all treatments (Table 3). Both these observations indicated that Al precipitation was not following first-order kinetics.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 8. Semilog plot of log[Al] vs. time in different treatments of the Batch 1 study: (a) 0.055, (b) 0.11, (c) 0.165, and (d) 0.22 mol L-1 Al. The straight lines are the linear regression fits and the dotted lines are the 95% confidence intervals.
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Aluminum precipitation rates (normalized per L) based on Al removal from the soil solution in the Batch 1 study.
|
|
Unlike Al, the respective Fe and Si precipitation reactions followed a first-order kinetics. The magnitude of the precipitation rates calculated with Fe and Si data depended on the respective Fe and Si aqueous concentrations in the soil solution (Table 4). Because the initial aqueous Al concentration controlled the rate of base-induced mineral dissolution (via decreasing free OH concentration in the contact solution and inhibition), it also controlled the rate of Fe and Si release into the soil solution. As a result, Fe and Si precipitation rates were strong inverse functions of the initial Al concentration in the soil solution (rate orders >-1) (Fig. 9)
. The pseudo-first order rate constants of Si precipitation reaction were independent of the aqueous Si concentration (Table 5).
View this table:
[in this window]
[in a new window]
|
Table 4. Precipitation rates (normalized per L) based on Si and Fe removal from the soil solution in the Batch 1 study.
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 9. Precipitation reaction order with respect to initial Al concentration in the soil solution based on the amount of Si and Fe removed from the soil solution in the time periods from 3 to 7 d (Si), and 2 to 3 d (Fe).
|
|
|
DISCUSSION
|
|---|
Results from this study show that Si and Fe concentrations in the soil solution increase rapidly in the first 4 h because of mineral dissolution, continue to increase in the next 44 h, decrease after 2 or 3 d, and reach equilibrium well below peak concentrations, after approximately 7 d. Potassium release is sluggish in the first 4 h, and K concentration remains approximately constant in the first 2 d. Most likely smectite is serving as a Si and Fe source in the first days. Potassium-bearing minerals (mica and K-feldspar) do not undergo dissolution in the first 2 d, but they undergo dissolution after the first 2 d.
It is generally believed that for most silicates and oxides the dissolution reaction is surface controlled (Sparks, 1999). This means that the surface reactions steps (adsorption of solutes, interlattice transfer of reaction species, surface chemical reactions, and removal of reactants from the surface) are slower compared with the other steps of the dissolution process (mass transfer of dissolved reactants from bulk solution to the mineral surfaces and mass transfer of products into the bulk solution) (Stumm and Wollast, 1990). In our experiment, the Si and Fe release in the first 3 d followed a parabolic rate law (Fig. 1b and 2b), which usually indicates transport-controlled dissolution reactions (Stumm, 1992; Sparks, 1999). Because these experiments are conducted under unusually high IS conditions, the possibility that diffusion may have limited the rate of dissolution, making it apparently transport-controlled, should not be excluded. On the other hand, the parabolic rate law may be an artifact caused by the immediate dissolution of fine particles present in the sediments (smectite), resulting in a higher initial dissolution rate or a pseudo-parabolic kinetics. In addition, the removal of Si and Fe from the soil solution on precipitation may contribute to making the trends parabolic. Precipitation is certainly promoted in batch experiments where reaction products supersaturate the soil solution. As will be discussed in the following paper (Qafoku et al., 2003), the SEM micrographs reveal the presence of secondary coatings in the sediments by Day 3.
The results presented above show that Al concentration in the soil solution does not change in the first 2 d of the experiment, but decreases substantially in the following days. Mineral dissolution predominates in the first 2 d and, as a result, some Al should have been released into the soil solution over this time period. However, it appears that little Al is released during the early phase dissolution. In another study conducted by our research group, in which an Al-free, 1 mol L-1 NaOH and 1 mol L-1 NaNO3 contact solution is used, the Al concentration in the soil solution by Day 7 is approximately 0.0051 mol L-1. This indicates that even in the case when no Al is present in the initial contact solution, the amount of Al release on dissolution is relatively small. Because the aqueous Al concentration remains invariable in the first 2 d, it is quite likely that both dissolution and precipitation, even though at slow rates, are controlling Al concentration in the soil solution in the first 2 d of the study. These results also imply that the precipitation rate of pure Al secondary phases is slow in this experiment. Because diagnostic peaks of gibbsite or other Al-pure phases are not present in the XRD patterns as will be discussed in the following paper (Qafoku et al., 2003), gibbsite is either not formed in abundant quantities, or is structurally amorphous and cannot be detected with XRD. Another reason gibbsite may not be identified is due to the formation of very small crystals that often eluded XRD detectors.
Silicon and Fe release exhibit a strong inverse dependence on initial aqueous Al concentration in the soil solution (rate orders >-1), indicating that Al affects inversely dissolution. The dependence of K release on Al is less pronounced (fractional rate orders). We closely monitored the changes in soil solution pH during the Batch 1 experiment and found that pH varied in the ranges: 13.80 to 14.00, 13.42 to 13.78, 13.34 to 13.68, and 13.10 to 13.35 in the respective four Al treatments. The free OH concentration in the soil solution is enough to keep the soil solution pH > 13.10 in all the treatments. However, the OH-/Al-4 concentration ratio changes significantly in these treatments, as it is shown in Table 1. By decreasing the free OH concentration in the contact solution, Al decreases the inhibitor/catalyst OH-/Al-4 aqueous molar ratio, making the OH catalytic effect on dissolution less effective. This is probably the main reason why its inhibitory effect was well manifest in these experiments and the rate orders calculated from partial rate laws of inhibition were >-1. This corroborates the results by Bauer (1998), who found a good correlation between the rate of kaolinite and smectite dissolution at 308 and 353 K and the OH-/Al-4 activity ratio in solution. These results are also in accordance with those presented by Oelkers et al. (1994) that found that the logarithm of the dissolution rate of alumino-silicate minerals was a linear function of the logarithm of Al concentration at constant pH, over a wide range of conditions. Similarly, the rate of dissolution of albite at pH = 9 exhibited a strong inverse dependence on dissolved Al in solution (Oelkers et al., 1994). In our experiment, the effect of Al on dissolution was so well pronounced because it encompasses the effects Al has on decreasing the OH-free concentration in the contact solution, and the inhibitory effect of Al-4 on dissolution.
Clearly, precipitation is controlling Al concentration in the soil solution after 2 d of the experiment (Batch 1 study). But the reaction half-life depends on initial Al concentration and the precipitation rate is not a function of initial Al concentration in the soil solution. Therefore, Al precipitation is not following first-order kinetics; rather, it appears to be following zero-order kinetics. The aqueous Al concentration is not controlling the precipitation rate in these experiments. Even though Al participates in the formation of the secondary phases, it appears that its concentration is in excess in all treatments of the Batch 1 study; hence, the apparent zero-order kinetics. It is quite likely then, that other elements present in the soil solution and that coprecipitate with Al are controlling the precipitation rate in these experiments.
The decrease in Al concentration with time that start by approximately Day 3 correspond to the decreasing trends observed for Fe and Si during the same time period. But unlike Al, Si precipitation reaction follows first-order kinetics. Because the decreasing trends of Fe and Si concentrations as a result of precipitation does not occur at the same time, but after 2 and 3 d, respectively, their precipitation reactions probably follow different paths, i.e., Fe most likely does not coprecipitate with Si. In addition, since the precipitation rate of Al-pure secondary phases is slow, it is also unlikely that Al and Fe coprecipitates in Fe-substituted Al secondary phases. Most likely, Fe precipitates of pure and/or Al substituted Fe(III) secondary phases are formed. The sediments' color at the end of the experiment changes to a more reddish hue (from 2.5 Y 6/0 to 2.5 Y 6/4 in the Munsell color book), which suggests the formation of Fe(III) precipitates. The reddish color is clearly visible in the Al treatments of both experiments, which indicates that Fe(III) secondary phases are formed when Al was present in the soil solution.
The changes in the soil solution chemistry [the release of K, Si, Fe(II), Fe(III), and other elements] caused by dissolution should play a significant role in the fate and transport of different radionuclides like Cs, Sr, and U and contaminants like Cr that were present in significant amounts in the HLWF at Hanford. For example, Fe(II) released on dissolution of Fe(II) bearing minerals (biotite and chlorite) may be involved in redox reactions with Cr(VI), K released on dissolution may compete for surface adsorption sites with Cs and Sr, and the presence of dissolved Si may result in the complexation of radionuclides in solution (Moll et al., 1998; Reich et al., 1998).
|
ACKNOWLEDGMENTS
|
|---|
This research was supported by the U.S. Department of Energy (DOE) through the Environmental Management Sciences Program (EMSP). The Pacific Northwest National Laboratory (PNNL) is operated for the DOE by Battelle Memorial Institute under Contract DE-AC06-76RLO 1830. We would like to recognize the contributions made to this study by other collaborators in this project, Dr. Samuel J. Traina of the University of California, Merced, and Dr. Gordon E. Brown, Jr. of Stanford University. Appreciation is extended to geologist Bruce Bjornstad for the collection of sediment samples representative of those beneath the S-SX tank farm at Hanford, and to the associate editor and an anonymous reviewer for their helpful comments. The research described in this paper was performed in part in the Environmental Molecular Sciences Laboratory, a national scientific user facility sponsored by the U.S. Department of Energy's Office of Biological and Environmental Research and located at Pacific Northwest National Laboratory in Richland, WA.
|
REFERENCES
|
|---|
- Bauer, A. 1998. Kaolinite and smectite dissolution rate in high molar KOH solutions at 35° and 80°C. Appl. Geochem. 13:905–916.
- Blum, A.E., and L.L. Stillings. 1995. Feldspar dissolution kinetics. p. 291–351. In A.F. White and S.L. Brantley (ed.) Reviews in mineralogy. Vol. 31. Chemical weathering rates of silicate minerals. Mineralogical Soc. of Am., Washington, DC.
- Brezonik, P.L. 1993. Chemical kinetics and process dynamics in aquatic systems. Lewis Publ., Boca Raton, FL.
- Devidal, J.L., J. Schott, and J.L. Dandurand. 1997. An experimental study of kaolinite dissolution and precipitation kinetics as a function of chemical affinity and solution composition at 150 degree C, 40 bars, and pH 2, 6.8, and 7.8. Geochim. Cosmochim. Acta 61:5165–5186.
- Felmy, A.R., J.R. Rustad, M.J. Marvin, and R. de la Bretonne. 1999. A chemical model for the major electrolyte components of the Hanford Waste Tanks: The binary electrolytes in the system: Na–NO3–NO2–SO4–CO3–F–PO4–OH–Al(OH)4–H2O. Publ. TWRS-PP-94-090. Pacific Northwest National Lab., Richland, WA.
- Ganor, J., and A.C. Lasaga. 1998. Simple mechanistic model for inhibition of a dissolution reaction. Geochim. Cosmochim. Acta 62:1295–1306.
- Jones, T.E., R.A. Watrous, and G.T. Maclean. 2000. Inventory estimates for single-shell tank leaks in S and SX tank farms. Publ. RPP-6285, Revision 0. Pacific Northwest National Lab., Richland, WA.
- Moll, H., G. Geipel, V. Brendler, G. Bernhard, and H. Nitsche. 1998. Interaction of uranium(VI) with silicic acid in aqueous solutions studied by time-resolved laser-induced fluorescence spectroscopy (TRLFS). J. Alloys Compd. 271:765–768.
- Oelkers, E.H., and J. Schott. 1999. Experimental study of kyanite dissolution rates as a function of chemical affinity and solution composition. Geochim. Cosmochim. Acta 63:785–797.
- Oelkers, E.H., J. Schott, and J.L. Devidal. 1994. The effect of aluminum, pH, and chemical affinity on the rates of aluminosilicate dissolution reactions. Geochim. Cosmochim. Acta 58:2011–2024.[Web of Science]
- Qafoku, N.P., C.C. Ainsworth, J.E. Szecsody, D.L. Bish, D.E. McCready, J.S. Young, and O.S. Qafoku. 2003. Aluminum effect on dissolution and precipitation under hyperalkaline conditions: II. Solid phase transformations. J. Environ. Qual. 32:2364–2372 (this issue).[Abstract/Free Full Text]
- Reich, T., H. Moll, T. Arnold, M.A. Denecke, C. Henning, G. Geipel, G. Bernhard, H. Nitsche, P.G. Allen, J.J. Bucher, N.M. Edelstein, and D.K. Shuh. 1998. An EXAFS study of uranium(VI) sorption onto silica gel and ferrihydrite. J. Electron Spectrosc. Relat. Phenom. 96(1–3):237–243.
- Serne, R.J., G.V. Last, G.W. Gee, H.T. Schaef, D.C. Lanigan, C.W. Lindenmeier, R.E. Clayton, V.L. LeGore, R.D. Orr, M.J. O'Hara, C.F. Brown, D.B. Burke, A.T. Owen, I.V. Kutnyakov, and T.C. Willson. 2001a. Geologic and geochemical data collected from vadose zone sediments from borehole SX 41-09-39 in the S/SX waste management area and preliminary interpretations. Publ. PNNL-2001-2. Pacific Northwest National Lab., Richland, WA.
- Serne, R.J., H.T. Schaef, B.M. Bjornstad, B.A. Williams, D.C. Lanigan, D.G. Horton, R.E. Clayton, V.L. LeGore, M.J. O'Hara, C.F. Brown, K.E. Parker, I.V. Kutnyakov, J.N. Serne, A.V. Mitroshkov, G.V. Last, S.C. Smith, C.W. Lindenmeier, J.M. Zachara, and D.B. Burke. 2001b. Characterization of uncontaminated sediments from the Hanford Reservation–RCRA borehole core and composite samples. Publ. PNNL-2001-1. Pacific Northwest National Lab., Richland, WA.
- Sparks, D.L. 1999. Kinetics and mechanisms of chemical reactions at the soil mineral/water interface. p. 135–191. In D.L. Sparks (ed.) Soil physical chemistry. CRC Press, Boca Raton, FL.
- Stumm, W. 1992. Chemistry of the solid–water interface. Processes at the mineral–water and particle–water interface in natural systems. John Wiley & Sons, New York.
- Stumm, W. 1997. Reactivity at the mineral–water interface: Dissolution and inhibition. Colloids Surf A 120:143–166.
- Stumm, W., and R. Wollast. 1990. Coordination chemistry of weathering. Kinetics of the surface-controlled dissolution of oxide minerals. Rev. Geophys. 28:53–69.
Related articles in JEQ:
- This Issue in Journal of Environmental Quality
JEQ 2003 32: 1931-1938.
[Full Text]
This article has been cited by other articles:
|
|
|
|
 
K. Mashal, J. B. Harsh, and M. Flury
Clay Mineralogical Transformations over Time in Hanford Sediments Reacted with Simulated Tank Waste
Soil Sci. Soc. Am. J.,
March 1, 2005;
69(2):
531 - 538.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
N. P. Qafoku, C. C. Ainsworth, J. E. Szecsody, D. L. Bish, J. S. Young, D. E. McCready, and O. S. Qafoku
Aluminum Effect on Dissolution and Precipitation under Hyperalkaline Conditions: II. Solid Phase Transformations
J. Environ. Qual.,
November 1, 2003;
32(6):
2364 - 2372.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
