Comparative studies of the reduction behavior of chromium(VI) by humic substances and their precursors

Material

Humic and fulvic acid (SHA, SFA) were extracted from peat soil (Shin-Shinotsu, Hokkaido, Japan) and purified according to the procedure recommended by the International Humic Substances Society (IHSS) [15]. An alkaline extract of peat was acidified with HCl up to pH <1. The precipitate and supernatant were separated by centrifugation (3,500 rpm, 15 min), and the precipitate was transferred into a dialysis tube (MW < 1,000 cut-off; Spectra/Pore, Houston, TX, USA). Dialysis was carried out against distilled water for 2 to 3 weeks with stirring. Finally, powdered humic acid (HA) was obtained by lyophilization. The supernatant was passed through an Amberite XAD-2 column (Organo, Tokyo, Japan; 50 mm ϕ × 320 mm). The fulvic acid (FA) was eluted by an aqueous solution of 0.1 mol/L NaOH. After neutralization of the fraction, the aqueous solution of FA was passed through a Sephadex G-10 column (50 mm ϕ × 330 mm; Pharmacia Biotech AB, Uppsala, Sweden), in which water was used as an eluent. The FA fraction was shaken with an H⁺-type cation exchanger (Amberite IR-500), and the aqueous phase was then evaporated. Finally, powdered FA was obtained by lyophilization.

Tannic acid and gallic acid were obtained from Wako Pure Chemical Industries (Osaka, Japan). These humic and fulvic acids were stored as a basic stock solution (0.01 mol/L NaOH) in a refrigerator.

Table 1 shows the results of elemental analysis and the values of reducing capacities of the DOCs. The reducing capacities were determined by a modified Walkley–Black method [16]. This method measures the amount of Cr(VI) reduced by DOCs under extremely acidic conditions. A mixture of 0.4 ml of 1.00 g/L DOCs and 0.2 ml of 0.10 mol/L Cr(VI) solution was added to 2.0 ml of concentrated H₂SO₄ in a glass vessel. The reaction vessel was kept at 90°C in a water bath, and the concentration of remaining Cr(VI) was monitored.

Determination of oxidation potential by cyclic voltammetry

The oxidation potential (E^ox_p) of each DOC was determined by cyclic voltammetry (CV) according to the method of Helburn and MacCarthy [17]. The pH of 0.1 mol/L KCl aqueous solution containing 1 g/L DOC was adjusted to 9 by 0.05 mol/L HEPES buffer. Five-milliliter aliquots of the solution were pipetted into an electrolysis glass cell and then deaerated by argon for 15 min. Subsequently, the CV voltammograms were recorded by scanning the potential from −1.0 to +0.9 V (vs Ag/AgCl) at 50 mV/s. A gold disk electrode for a working electrode, a platinum wire for a counter electrode, and an Ag/AgCl electrode for a reference electrode were used. The measurements were performed at room temperature (22°C).

Kinetic measurement

Since the reduction rate of Cr(VI) by humic substances is very slow [10], the temperature for the reaction was set at 50°C in this study in order to accelerate the reaction. The pH of the sample solutions, including 40 mg/L of each DOC, was adjusted by using sulfuric acid and sodium hydroxide. The pH of the sample was also checked at the end of the reaction in order to confirm a constant pH. The deviation of pH was about ±0.2 to 0.3. Fifty-milliliter aliquots of the sample solutions in the Erlenmeyer flask were stirred in a water bath. After reaching constant temperature, 5 mmol/L Cr(VI) was added by a micropipette to give 50 to 100 μmol/L Cr(VI) in the solution. Two-milliliter aliquots of the solution were pipetted into a 10-ml measuring flask to which 0.1 ml of concentrated sulfuric acid and 0.1 ml of 0.25% diphenylcarbazide (DPC) in acetone was added, and this solution was then diluted with water. After 20 to 30 minutes, the concentrations of Cr(VI) in the sample solutions were determined by measuring the absorbance at 542 nm. The reproducibilities of these experiments for GA and TA were about 10%. The statistical tool used in this work for curve fitting is ORIGIN version 4.1J (Microcal Software, Northampton, MA, USA).

Speciation analysis for chromium after reduction by cation exchange/DPC colorimetric method

The chemical species of Cr(VI) after reduction by DOCs were determined using a cation exchange/DPC colorimetric method reported previously [14]. The sample solution at pH 4.9, containing 40 mg/L DOC and 100 μmol/L Cr(VI), was shaken for 1 week in a water bath at 50°C. A 5-ml aliquot of the solution was passed through the column packed with 0.4 ml of the cation exchanger, Sulphopropyl Sephadex C-25 (C-25), and the effluent was then collected. In this procedure, the cationic species, inorganic Cr(III), were retained on the C-25, but the anionic species, inorganic Cr(VI) and Cr(III) complex with DOCs, passed though the column. The concentrations of Cr(VI) in the effluent were determined by the DPC colorimetric method, and the total concentration of Cr in the effluent was determined by the ICP-AES method. The concentration of Cr(III) complex species can be calculated by subtracting the concentration of inorganic Cr(VI) from total concentration of Cr in the effluent. The ability of each DOC to complex with Cr(III) was also evaluated by the above method using the C-25 resin [18].

UV-vis spectrometry. A UV-2000 type spectrophotometer (Shimadzu, Kyoto, Japan) was used to measure the UV-vis absorption spectra of the solutions. The concentrations of HA, FA, and TA were 20 mg/L and that of GA was 4 mg/L. The time dependency in the spectrum of GA from 400 to 200 nm was observed by reacting 25 μmol/L Cr(VI) and 4 mg/L GA (pH 4.9, 50°C).

Gel chromatography. Gel chromatography was carried out to confirm the polymerization of GA that accompanied the reduction of Cr(VI). The glass column (25 mm ϕ × 500 mm length) filled with Sephadex G-10 (exclusion limit 700, Pharmacia) was used for this. Blue dextran (MW > 2 × 10⁶) was used to determine the void volume of this column. Pure water was used as the mobile phase, the flow rate of which was 1 ml/min. The solution (pH 4.9) containing 200 mg/L GA and 1 mmol/L Cr(VI) was shaken about 6 d at 50°C, and then 2-ml aliquots of the solution were injected on the column and fractions of the effluent were collected at 5 min intervals. The absorbance of each fraction was measured at 260 nm, and the concentration of Cr in each fraction was determined by flame atomic absorption spectrometry.

Reduction rate of chromium(VI) by humic substances and their precursors

Figure 1 shows the kinetic curves for the reduction of Cr(VI) by various DOCs. As described in several reports [5-10] in the presence of all DOCs used in this study, the concentration of Cr(VI) in solution decreased with the reaction time and the reduction rate increased with decreasing pH. However, it is difficult to estimate and compare quantitatively the rate constants for each DOC since it is not possible to specify the molar concentrations of humic substances and TA because their molecular weights are not known. The reducing capacities of these DOCs indicate the total amounts of Cr(VI) reduced by the DOCs under strong acid conditions [16]. The reducing capacities of these DOCs are summarized in Table 1. The values for the DOCs obtained in the present work (89–110 mmol Cr/g of C) were slightly higher than that for fulvic acid from the Swanee River reported by Wittbrodt and Palmer (80 mmol Cr/g of C) [10]. The reducing capacity of each DOC can be correlated with the molar concentration of the reductive functional group F_red by following the equation

(1)

Therefore, F_red is useful to quantitatively compare the reduction rate of Cr(VI) by each DOC. The pseudo first-order kinetic constant (k_exp) for the reduction of Cr(VI) is defined by

(2)

The integrated form of Equation 2 can be written

(3)

where [Cr(VI)]_add indicates the concentration of Cr(VI) added initially. The k_exp values were calculated by curve fitting of the data points in Figure 1 to Equation 3. It is known that the k_exp values depend on not only the concentration of Cr(VI) but also the reactive functional group in the DOC [8-10]. Therefore, the rate equation can be represented as the empirical second-order reaction reported by Elovitz and Fish [8] as

(4)

where k_Fred indicates the empirical second-order kinetic constant considering the concentration of the reductive functional group in DOCs. Quantitative comparisons of the reduction rates between the DOCs are possible by using this k_Fred. The relationship between k_exp and k_Fred is derived by combining Equations 2 and 4 to get

(5)

Equation 5 shows a linear relationship between the k_exp values and the concentrations of the DOCs. Such a linear relationship was reported in the case of the reduction of Cr(VI) by various phenolic compounds [8]. As shown in Figure 2, we also confirmed this linearity for the case of SHA. The k_exp and k_Fred values, which were calculated by Equation 5, are summarized in Table 2. The k_Fred values increased with decreasing pH, which can be explained by the decrease in the redox potential of Cr(VI) [10]. The order of k_Fred at pH 5, which is near the pH of natural water, was GA > TA > SFA > SHA. The values for TA and GA were larger by two to three orders of magnitude than those for the humic substances. As shown in Figure 1, the reduction of Cr(VI) by the humic substances was not complete even after several hundred hours. However, the reduction of Cr(VI) by TA and GA was complete in several to 10 h. The Cr(VI) reduction rate for the low molecular weight DOC such as TA and GA was much faster than that for the humic substances.

In Table 2, although the concentration of the reductive functional groups for each DOC, [F_red], was similar, the k_Fred values were quite different. It has been reported that a positive correlation exists between the oxidation rate for a phenol derivative and the oxidation potential [19]. Thus, such differences in the reaction rates may be dependent on the oxidation potential of each DOC. We then determined the oxidation peak potential (E^ox_p) for these DOCs by CV (Table 2). We could not evaluate E^ox_p for SHA because a clear oxidation peak was not observed. The order of E^ox_p was GA < TA < SFA, and this order was consistent with that of the k^Fred values. This suggests that the DOC having the higher oxidation potential was hardly oxidized and, as a result, the reduction rate for Cr(VI) with the DOC becomes slower.

Speciation analysis of chromium after reduction of chromium(VI) with DOCs

Figure 3a shows the results of the analysis of Cr species after the reduction of Cr(VI) with each DOC at pH 5 for 1 week. In the case of SHA and SFA, about 50 to 60% of the Cr was found to be the unreduced Cr(VI) species and about 40% was complexed with the humic substances. In the case of TA, about 60% of the Cr was observed to be complexed with TA, and the ratio of the species complexed with TA was higher than those for the humic substances. In the case of GA, Cr(VI) was almost completely reduced and about 90% of the Cr was found to be complexed with GA. The ratio of the complexed species in the case of GA was the largest of these four DOCs.

Figure 3b shows a comparison of the percentages of the complexed Cr(III) species when Cr(VI) or Cr(III) were reacted with the DOCs. The percentages after the reduction of Cr(VI) were calculated as the percentages of the complexed species of the total Cr(III) species. In the cases of SHA and SFA, large differences were not observed between the percentages after the reduction of Cr(VI) and complexation with Cr(III). However, in GA and TA, the percentages of the complexed species after the reaction with Cr(VI) were larger than those after the complexation with Cr(III), especially in the case of GA; although the percentage of the complexed species after complexation with Cr(III) was 32%, the percentages of the complexed species after the reduction of Cr(VI) were about 90%, or 2.8-fold larger than those for the complexation with Cr(III). Therefore, TA and GA are more effective in decreasing the toxicity of Cr(VI) by the rapid reduction of Cr(VI), followed by complexation with the resulting Cr(III) produced by the reduction.

Oxidative polymerization of GA during reduction of chromium(VI)

In Figure 1c and d, the series of the kinetic data points for TA and GA deviated slightly from the calculated curve, which is indicated by arrows. These deviations can be attributed to different reducing sites for Cr(VI), which results in the reduction of Cr(VI) at a slower rate [10]. From the data of E^ox_p in Table 1 and k_Fred in Table 2, the higher oxidative potential of these compounds showed a larger molecule size and a slower reduction rate. Therefore, the deviation in Figure 1c and d suggests the production of larger molecular weight compounds from TA or GA. Phenolic compounds such as catechol are oxidized and then polymerized easily by dissolved oxygen [20, 21]. The TA and GA would also be expected to result in higher molecular weight materials by the oxidative polymerization during the reduction of Cr(VI). We attempted to confirm this for the polymerization of GA by UV-vis spectrophotometry and gel permeation chromatography.

UV-vis spectrum. Figure 4a shows the time dependency of a UV-vis spectrum of GA for a reaction of GA and Cr(VI) at 50°C. As described in the literature [22], two peaks were observed at 260 and 210 nm on the initial spectrum of GA. However, the absorbances at 260 and 210 nm decreased and the absorbance at 320 and 230 nm increased with increasing reaction time. After 50 min, the sharp peaks in the spectrum disappeared and the absorbance increased linearly with decrease in wavelength.

Figure 4b shows UV-vis spectra of the DOCs in the absence of Cr(VI). The TA showed two shoulders in the spectrum at 360 and 270 nm. However, humic substances, such as SHA and SFA, had no peaks in their spectrum. Comparing the curves in Figure 4a and b, it was found that the spectrum of GA changed to the TA type (Fig. 4a, 5 min) and finally the humic type (Fig. 4a, 51 min). These results suggest that GA is polymerized and converted to a high molecular weight compound as humic substances during the reduction of Cr(VI).

Gel chromatography. In order to clarify the polymerization of GA that occurs during the reduction of Cr(VI), we compared the molecular weight distribution of GA before and after the reduction of Cr(VI) by using gel permeation chromatography with a Sephadex G-10 column. The void volume for the column, which was measured by blue dextran (MW > 2 × 10⁶), was 76 ml. The elution volume (V_e) for the GA solution after the reduction of Cr(VI) was 76 ml both when the detection was carried out by measurement of the absorbance at 260 nm (Fig. 5a) and Cr by AAS (Fig. 5b). These V_e values coincided with the void volume (76 ml). On the other hand, the V_e values of GA itself (Fig. 5c) and Cr(VI) itself (Fig. 5d) were 350 ml and 145 ml, respectively. Therefore, it is thought that the molecular weight of GA after the reduction of Cr(VI) increases. These results suggest that GA was polymerized during the reduction of Cr(VI), producing high molecular weight compounds. Moreover, the comparison of the chromatograms in Figure 5b and d showed that the location of Cr in the gel chromatograph changed to that of the high molecular weight compounds after the reduction by GA. This result suggests that Cr, after the reduction, was taken up by the polymerized GA and appeared as the complexed species.