«THERMODYNAMIC MODELING OF METAL ADSORPTION AND MINERAL SOLUBILITY IN GEOCHEMICAL SYSTEMS A Dissertation Submitted to the Graduate School of the ...»
Organic C extractions of both fumigated and un-fumigated samples were performed using identical procedures. That is, each sample was transferred into a 500 mL glass media bottle, and 200 mL of 0.5 M K2SO4 was added. The bottles were then placed on an oscillating shaker for 30 minutes to thoroughly mix the sample with the K2SO4 extractant solution. The supernatant was filtered using Whatman No. 42 filters, and collected for C analysis. We used an organic C analyzer (Shimadzu TOC-5000) to analyze for organic C in the extracts. Sets of six calibration standards for the TOC analyzer were made from a 1000 ppm organic C solution as potassium biphthalate (Ricca Chemical Company, Arlington, TX) in the same 0.5 M K2SO4 matrix as the samples. The organic C concentrations in each calibration set were designed to be appropriate for the expected C concentrations in the various samples. Samples and standards were acidified with 1.0 uL of 1.0 M trace grade HC1 per mL of solution to remove dissolved inorganic C. The TOC instrument analyzed the C concentration in each sample 3 times, with a resulting error of less than 5% for all analyzed solutions.
Control blanks containing only 0.5 M K2SO4 and the HC1 acidification were included with every TOC set, and used to correct each dataset for any background C present in the reagents or from the instrument.
A set of two-component experiments was conducted using a constant 50 g of silica sand and between 0 and 200 mg of bacteria. The fumigation and extraction procedures for these experiments were identical to those conducted for the singlecomponent experiments.
3.3 Results and Discussion The results of the bacteria experiment are depicted in Figure 3.1, which relates the biomass initially placed in the Petri dish to the concentration of organic C extracted in the K2SO4 solution. Data for the fumigated samples and for the un-fumigated controls.
are both plotted, and the calculated difference (the concentration of TOC in the extract from the fumigated samples minus the concentration of TGC in the extract from the unfumigated controls) is also shown. The un-fumigated controls exhibit increasing extracted TOC with increasing initial biomass in the Petri dish, indicating that the extraction wash dissolves some of the organic C from un-treated bacteria, and that the concentration of organic C in the wash is controlled by the amount of biomass exposed to the wash solution. The fumigated bacterial samples exhibit enhanced TOC in the extraction solution relative to that measured in the extraction solutions for the unfumigated controls for all initial biomasses studied, and the enhancement increases with increasing initial biomass. The enhanced TOC in the fumigated samples indicates that the chloroform treatment exposes the extraction wash solution to higher levels of organic C than in the un-fumigated case, likely due to cell lysis from fumigation. For an ideal application of the fumigation-extraction approach, the un-fumigated control in a bacteria-bearing system would exhibit little or no extracted TOC in the K2SO4 wash solution. However, the difference in TOC concentrations between the fumigated samples and the un-fumigated controls also increases with increasing biomass, and this relationship strongly suggests that the fumigation procedure can be calibrated and used successfully in a bacteria-only system to relate extracted TOC to initial biomass in the sample.
Figure 3.1: Experimental results from the bacteria-only system, plotted in terms of the concentration of extracted TOC as a function of the initial mass of bacteria present in the sample.
For the non-bacterial soil components that we tested, the differences between the organic C concentrations that were extracted from the fumigated samples and the unfumigated controls should be zero for the fumigation-extraction procedure to be valid.
That is, because those single-component systems do not contain any bacterial cells, the fumigation should not lyse any cells and therefore should not introduce additional organic C to the K2SO4 wash solution. The humic acid (Figure 3.2) and the silica sand (Figure 3.3) systems show this relationship. In the case of humic acid, we expect the TOC concentration in fumigated and un-fumigated extraction solutions to increase with the amount of humic acid because humic acid is largely composed of organic C. The amount of TOC in the fumigated and un-fumigated extraction solutions from experiments with the same initial mass of humic acid is essentially the same, and is close to being equal to the entire mass of humic acid in each experiment, suggesting that almost all of the humic acid dissolves into the K2SO4 wash solution in each extraction.
The difference between the two concentrations is close to zero for all humic concentrations tested. This result suggests that the presence of humic acid in a soil does not affect the amount of C attributed to biomass using the fumigation-extraction method.
The results from the silica sand-only systems (Figure 3.3) demonstrate that virtually no TOC is extracted from each sample, and that this result is independent of initial sand mass and is independent of whether the sample was exposed to chloroform.
As is the case for the humic acid system, the difference between extractable TOC concentrations in the solutions from fumigated and un-fumigated samples is essentially zero and does not increase with increasing sand mass, indicating that the chloroform fumigation procedure does not add to the extractable C pool. The two-component
Figure 3.2: Fumigation results for humic acid-only samples, plotted in terms of the concentration of extracted TOC as a function of the initial mass of humic acid present in the sample.
Figure 3.3: Fumigation results for silica sand-only samples, plotted in terms of the concentration of extracted TOC as a function of the initial mass of sand present in the sample.
experiment that involved a constant 50 g of the silica sand and varying amounts of bacteria demonstrates that the method may be difficult to apply to a sand-rich soil or aquifer material (Figure 3.4). The resultsfromthe sand and bacteria system indicate that the presence of the silica sand reduces the efficiency of the fumigation-extraction procedure in extracting biomass C from the bacteria that are present in the sample. The silica sand does not sorb chloroform and its presence does not totally block cell lysis due to fumigation. However, for a given mass of bacteria, the concentrations of extracted C from both the fumigated and unfumigated samples are lower when the silica sand is present than it is for the corresponding bacteria-only experiments. The dependence of the extraction efficiency on silica sand content of the sample suggests that the fumigation extraction technique would need to be calibrated for a particular biomass:sand ratio, making it impractical for the determination of biomass in a sample where the ratio of biomass to mass of sand is unknown. However, Figure 3.3 also shows that the values of the difference between fumigated and unfumigated samples in the bacteria-only and in the bacteria and sand experiments are similar for a given biomass,. This result may indicate that the method could be used successfully for determining biomass C in a sandy soil, but more tests would be required to determine if the difference values are independent of silica sand content.
The clays that we studied exhibit markedly different behavior than the other soil components. If chloroform does not sorb onto mineral surfaces, then the model clays, like the silica sand, would show little or no total organic C upon extraction, and no difference between fumigated samples and un-fumigated controls. Unlike the results for silica sand and humic acid, our experiments with wet and dry montmorillonite (Figures
3.5 and 3.6) and kaolinite (Figure 3.7) strongly suggest that chloroform vapor sorbs
Figure 3.4: Results for fumigation experiments with a constant 50 g of silica sand and varying amounts of bacteria, plotted in terms of the concentration of extracted TOC as a function of the initial mass of bacteria present in the sample.
Figure 3.5: Fumigation results for montmorillonite-only samples (SWy-2), plotted in terms of the concentration of extracted TOC as a function of the initial mass of montmorillonite present in the sample.
Figure 3.6: Fumigation results for wet montmorillonite samples (SWy-2), plotted in terms of the concentration of extracted TOC as a function of the initial dry mass of montmorillonite present in the sample.
Figure 3.7: Fumigation results for kaolinite-only samples (KGa-lb), plotted in terms of the concentration of extracted TOC as a function of the initial mass of kaolinite present in the sample.
substantially to the clays during the 24 hours of fumigation. That is, the difference between the fumigated clay samples and the un-fumigated controls is not zero, but increases with the mass of clay that is fumigated. Different surface chemistries and a lower surface area for the silica sand relative to the clay samples likely explain why chloroform sorption onto silica sand was negligible in our experiments while we observed extensive chloroform sorption onto both types of clay. The wet montmorillonite experiments demonstrate that natural soil samples that contain hydrated clays also sorb chloroform during the fumigation-extraction method, adding to the organic C pool in the extraction solution. The wetted montmorillonite sorbs significantly less chloroform than the dry montmorillonite samples, but there is still a significant extent of chloroform sorption. The dry and wetted montmorillonite samples contain 5 and 83% water by mass, respectively. Natural soil samples typically contain clays in this hydration range, so our results indicate that these clays sorb significant concentrations of chloroform and that the fumigation-extraction approach can not be accurately used to determine biomass for these types of samples.
3.4 Conclusions The amount of chloroform sorbed in our experiments adds an average of 1540 ug • C per g montmorillonite and 350 \ig C per g kaolinite into the extraction solution. The fumigation-extraction method, applied to ten different naturally occurring soils by Vance et al. (1987), yielded between 60 and 1220 \ig biomass C per g soil. The concentration of sorbed chloroform in our clay experiments is comparable to this range, and our results.
indicate that sorbed chloroform could contribute substantially to the C pools of these soil samples. In soil samples that contain significant quantities of clay mineral surfaces, exposure of the sample to chloroform leads to chloroform sorption onto the clays. The subsequent K2SO4 wash desorbs the chloroform into the wash solution, and because chloroform contains organic C, its presence on the clays and in the wash solution contributes to the organic C concentration that is determined for the sample. Clearly, for soil samples with significant concentrations of clay minerals, the fumigation-extraction procedure yields results for biomass C that are artificially high due to the sorption artifacts.
Our results show that chloroform fumigation-extraction is not valid for soils containing substantial amounts of clay. It may be possible to use the fumigationextraction method to find estimates of biomass C concentrations for materials that are poor in clay, such as a silica sand groundwater aquifer that also contains humic acid and/or bacteria. However, because most soils contain substantial clay mineral fractions, our results indicate that the fumigation-extraction method is not accurate for determination of the biomass C in soils.
4.1 Introduction The fate of heavy metals in soils and aquifers can be controlled by their adsorption to solid components (e.g., Meng and Letterman, 1996; Ledin et al., 1997, 1999; Covelo, 2007; Lund et al., 2008), aqueous complexation with dissolved organic ligands (e.g., Liu and Gonzales, 1999; Buerge-Weirich et al., 2003), and the formation of ternary surface complexes (e.g., Zachara et al., 1994, AH and Dzombak, 1996; Fein, 2002). Although metal adsorption in multi-sorbent systems has been studied (e.g., Krantz-Rulcker et al., 1996; Ledin et al., 1997,1999; Fingler et al., 2004, Covelo et al., 2007), the application of a quantitative surface complexation model (SCM) approach to predict metal distribution among mixtures of geosorbents is complex and difficult (Davis et al., 1998). However, the SCM approach has distinct advantages to empirical models in that the models can be extrapolated to systems of different ionic strength, pH, and component composition (Bethke and Brady, 2000; Koretsky, 2000).
Two approaches can be used when applying SCMs to describe metal distribution in multi-sorbent systems: the component additivity (CA) approach and the general composite (GC) approach. The CA approach predicts the extent of adsorption in mixed systems based on the adsorption affinities of each solute-sorbent combination measured in isolated binary experiments, and the relative concentrations of sorption sites of each sorbent. Success of the CA approach requires that sorbents do not interact with each other, that all solutes in the system have access to all surfaces, and that the only surface complexes that form are those that form in single sorbent, single component systems as well. Previous applications of the CA approach have met with varying degrees of success. For example, Davis et al. (1998) attempted to use a SCM to predict Zn(II) adsorption onto a natural, well-characterized sedimentary mineral assemblage. The CA approach under-predicted Zn(II) uptake, likely due to difficulties in determining absolute site concentrations for each site type within the complex sediment studied. Davis et al.