Stephen E. Cabaniss

Size fractionation upon adsorption of fulvic acid on goethite : equilibrium and kinetic studies

Abstract We examined adsorption equilibrium and kinetics of an aquatic fulvic acid (XAD-8 resin extract) onto goethite (α-FeOOH). Molecular weight distributions were determined using high-pressure size exclusion chromatography (HPSEC). Overall adsorption isotherms and those of the most abundant intermediate molecular weight (IMW) fraction (1250–3750 Da) fit the Langmuir adsorption equation, as is commonly observed for humic substances. However, this overall fit masked the non-Langmuir isotherm shape of high and low molecular weight (HMW, LMW, respectively) fractions. We observed preferential adsorption of HMW fractions at low pH and of IMW fractions at higher pH. We also observed preferential adsorption of components with higher absorbance normalized to moles C (e 280 ), probably reflecting greater aromaticity. Over the first 6 h of adsorption experiments, adsorbed organic carbon increased and weight average molecular weight ( M w ) of the organic matter remaining in solution decreased, consistent with slower adsorption of higher molecular weight components. Observations of fractionation upon adsorption agreed well with a field study showing lower M w and lower e 280 organic matter in deeper ground water relative to surface and shallow ground water.

Hydrogeochemical controls on the variations in chemical characteristics of natural organic matter at a small freshwater wetland

This study investigated the spatiotemporal variability in dissolved organic carbon concentration ([DOC]), natural organic matter (NOM) weight average molecular weight (Mw), and absorptivity at 280 nm (e280, an estimator of aromaticity) at McDonalds Branch, a first-order stream that is a fen wetland. When ground-water discharge to the stream was predominant, the [DOC], Mw, and e280 were all relatively low. When soil pore water represented an important contribution to streamflow, not only was the [DOC] higher, but also the Mw and e280. Hence, the contribution of soil pore water relative to ground water controlled not only the concentration but also the average physicochemical characteristics of the NOM. Combined field and laboratory experiments suggested that preferential adsorption of higher Mw, more aromatic NOM components to mineral surfaces, most likely within the lower soil horizons, resulted in a lower Mw, more aliphatic ground-water NOM pool. This process ultimately affected surface water as ground water discharged to the lower reaches of the stream. Within the stream, higher molecular weight, more aromatic components were susceptible to photo-aggregation and photo-degradation, in the presence of Fe. Results from this small watershed study provide insight into climatic effects on surface-water NOM characteristics in a small freshwater fen. Low flow periods resulted in lower Mw, more aliphatic NOM derived primarily from ground-water discharge to the stream, whereas higher flow periods resulted in a higher Mw (by 150–500 Da, i.e., increased by as much as 25%), more aromatic downstream surface-water NOM pool. Hence, during future summer drought periods, as suggested by climate-change models for much of North America, surface-water NOM likely will be lower molecular weight, more aliphatic, and more hydrophilic with lesser metal binding and hydrophobic organic carbon (HOC) uptake abilities [Environ. Sci. Technol. 34 (2000) 1103], along with decreased ability to attenuate UV radiation. Because NOM plays an important role in attenuating UV radiation, research on climate-induced effects on aquatic ecosystems needs to consider not only decreases in NOM concentration and photochemistry but also hydrology-induced changes to the NOM physicochemical characteristics.

Aqueous infrared carboxylate absorbances: aliphatic monocarboxylates

Aqueous attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectra of 24 aliphatic monocarboxylates have asymmetric COO− stretch frequencies (νas) of 1541 cm−1 to 1678 cm−1, while the symmetric stretch frequencies range from 1202 cm−1 to 1417 cm−1. The pKa values of these acids are linearly correlated with νas according to the equation (pKa = 66.70-0.04018νas, σ = 0.266 pKa units, r2 = 0.968). This correlation should be useful in determining the pKa values of acids in complex mixtures and the site pKa values of carboxyl groups in polyacids. Semi-empirical molecular orbital calculations using the AM1 parameter set give νas values which correlate well with experimental νas (σ = 11.3 cm−1), although model νas values are ≈ 500 cm−1 too high; the PM3 parameter set predicts νas values that are only ≈ 300 cm−1 too high, but which do not correlate as well with experiment (σ = 17.6 cm−1).

Carboxylic acid content of a fulvic acid determined by potentiometry and aqueous Fourier transform infrared spectrometry

Abstract The carboxylic acid content of a fulvic acid is determined quantitatively using aqueous Fourier transform infrared spectrometry and potentiometry. The combined method minimizes interferences from other acidic or carbonyl functional groups in the fulvic acid mixture. Potentiometry alone provides only operational estimates of the acid content, which vary by up to 15% as the inert salt (sodium chloride) concentration increases from 1 to 500 mM. Quantitative IR spectra are collected in aqueous solutions fo varying pH using an attenuated total reflectance cell designed for liquids. Changes in the area ratio of IR peaks due to protonated and deprotonated carboxylic acid groups correlate strongly (r2 = 0.995) with base consumed in potentiometric titrations. The carboxylic acid content of this fulvic acid is 9.09 ± 0.28 (± 1 standard error) meq carboxylic acid per gram of carbon.

Aluminum binding to humic substances determined by high performance cation exchange chromatography

Abstract Aluminum binding to humic substances is studied using high performance cation exchange chromatography (HP-CEC) in both laboratory and field samples. HP-CEC provides independent quantification of both free and complexed Al(III), eliminating many of the uncertainties associated with previous measurements of aluminum speciation. We formulate and calibrate a simple aluminum complexation model, using solutions with variable total Al(III) (0.2–11 μM) and pH (4.0–8.2) and constant organic composition (12.4 mg/l Suwannee River fulvic acid, FA). The model is verified with HP-CEC data from a group of acid lakes and from literature data. The results indicate that Al(III) complexation by natural organic matter in several environments is similar to complexation by FA, and that Al dimers are potentially important species even at low environmental concentrations of Al. Furthermore, organic complexation of Al is not limited to acidic waters but appears to be important even at the alkaline pH values typical of marine systems.

Colorimetric flow-injection analysis of dissolved iron in high DOC waters

An iron flow-injection analysis system has been optimized for the analysis of iron in waters high in dissolved organic carbon. The method detects either dissolved iron(II) or total dissolved iron with a detection limit of 10 nM, precision of 0.65% at 1 microM, and a dynamic range of four orders of magnitude. There are minimal interferences (< 1%) from other metals at environmental concentrations. The iron(II) method measures iron(II) in the presence of excess iron(III) with less than 1% interference. When used with pre-acidified samples, the total dissolved iron method agrees well with electrothermal atomic absorption spectrometry for a variety of natural waters with a range of dissolved organic carbon (3-36 mg C/L) and iron (1-28 microM) concentrations. When used with samples at their ambient pH, the total dissolved iron method detects dissolved iron, but not colloidal iron (size fraction 0.05-0.45 micron).

Determination of trace aluminum in natural waters by flow-injection analysis with fluorescent detection of the lumogallion complex

Abstract A flow-injection analysis method is described for the determination of trace levels of aluminum using fluorescent detection of the Al-lumogallion complex. The method is very simple, requiring neither iron masking agents nor micelle forming compounds to improve sensitivity. The detection limit is 3.7 nM with linearity to 37 μM and a relative standard deviation at 1.1 μM of 0.8%. Naturally occurring substances which might be expected to react with aluminum or lumogallion, including citric acid, fulvic acid and iron(III), show no interference. The detection of all forms of aluminum except polymers is demonstrated, making the method useful as a speciation procedure for the determination of free and complexed forms of aluminum in acid impacted as well as non-impacted aquatic systems. Results of analyses of water samples from lakes having a variety of pH values are presented with comparison to results obtained by GFAAS.

Cu(II) binding by a pH-fractionated fulvic acid

Abstract The relationship between acidity, Cu(II) binding and sorption to XAD resin was examined using Suwannee River fulvic acid (SRFA). The work was based on the hypothesis that fractions of SRFA eluted from an XAD column at various pH’s from 1.0 to 12.0 would show systematic variations in acidity and possibly aromaticity which in turn would lead to different Cu(II) binding properties. We measured equilibrium Cu(II) binding to these fractions using Cu 2+ ion-selective electrode (ISE) potentiometry at pH 6.0. Several model ligands were also examined, including cyclopentane-1,2,3,4-tetracarboxylic acid (CP-TCA) and tetrahydrofuran-2,3,4,5-tetracarboxylic acid (THF-TCA), the latter binding Cu(II) much more strongly as a consequence of the ether linkage. The SRFA Cu(II) binding properties agreed with previous work at high ionic strength, and binding was enhanced substantially at lower ionic strength, in agreement with Poisson–Boltzmann predictions for small spheres. Determining Cu binding constants ( K i ) by non-linear regression with total ligand concentrations ( L T i ) taken from previous work, the fractions eluted at varying pH had K i similar to the unfractionated SRFA, with a maximum enhancement of 0.50 log units. We conclude that variable-pH elution from XAD does not isolate significantly strong (or weak) Cu(II)-binding components from the SRFA mixture.

Quantitative aqueous attenuated total reflectance Fourier transform infrared spectroscopy: Part II. Integrated molar absorptivities of alkyl carboxylates

A quantitative attenuated total reflectance Fourier transform infrared (ATR-FT-IR) spectroscopic method is developed for the analysis of total carboxylate concentration, [COO−], in aqueous solution. The short (12–13 μm) and highly reproducible pathlength of the ATR cell permits quantitative subtraction of the water peak at 1640 cm−1. Carboxylate quantitation is based on the area of the asymmetric stretching peak, which is nearly independent of compound structure. The molar absorptivity of alkyl carboxylates in water is 438 ± 58 l mol−1 cm−1, and the integrated molar absorptivity is 2.95 ± 0.08 × 104 l mol−1 cm−2 (n = 15 compounds, 0.1 M ≤ [COO−] ≤ 1.5 M). The [COO−] in solutions of mixed carboxylates is measured with a root mean square error of 2.4% and a small (+1.5) positive bias. The accuracy of the method is limited by the assumption that integrated absorbance is constant for all COO− groups.

Uncertainty propagation in geochemical calculations: non-linearity in solubility equilibria

Uncertainty in the results of geochemical equilibrium calculations may be estimated from input uncertainties using either derivative approximations or Monte Carlo simulations. While derivative methods are fast and convenient for many equilibria, inherently non-linear aspects of solubility equilibria may lead to a failure of the linear approximation and thus to asymmetric and/or bimodal uncertainty in calculated concentrations. In these cases, the use of derivative methods and the assumption of Gaussian uncertainty in the calculated concentrations misrepresent the propagated uncertainty. Equilibria with gibbsite, calcite, and the chromate analog of jarosite are examined.