Environ. Sci. Technol. 2009, 43, 5890–5895
Elimination of Sludge Odor by Oxidizing Sulfur-Containing Compounds with Ferrate(VI) C H U N H E , †,‡ X I A N G - Z H O N G L I , * ,† VIRENDER K. SHARMA,§ AND SHI-YU LI‡ Department of Civil and Structural Engineering, The Hong Kong Polytechnic University, Hong Kong, China, School of Environmental Science and Engineering, Sun Yat-sen University, Guangzhou 510275, China, and Chemistry Department, Florida Institute of Technology, Melbourne, Florida 32901
Received February 8, 2009. Revised manuscript received June 17, 2009. Accepted June 19, 2009.
Sulfur-containing compounds are one kind of odorant found in sewage treatment works, composting plants, refuse storage and transfer, landfill sites, and associated with various industries. In the present research, the reaction kinetics of ferrate(VI) (FeVIO42-, Fe(VI)) with 2-mercaptobenzothiazole (MBT), thiosemicarbazide (NH2NHC(S)NH2, TSC), and thiourea dioxide (NH2C(SO2)NH2, TUDO) were studied under alkaline conditions. Stoichiometry of Fe(VI) oxidation with hydrogen sulfide (H2S), TSC, and methyl mercaptan (CH3SH) were determined at neutral and alkaline pH (7.0-11.0). Stoichiometric molar ratios ([Fe(VI):[S]) were determined to be 2.5, 2.0, and 4.6 for sulfide, TSC, and CH3SH, respectively, at pH 9.0. TUDO and methyl sulfonic acid (CH3SO3H) were identified to be the main intermediates of TSC and CH3SH reactions with Fe(VI), respectively, at pH 9.0, while sulfate was one of the final products. A reaction scheme is given to explain the intermediates and products formed in the CH3SH degradation by Fe(VI). Experiments were also conducted to evaluate the odor emission of digested sludge from sewage treatment works in terms of chemical concentration and also odor concentration affected by the Fe(VI) dose. The potential of using Fe(VI) to achieve odor control in sludge treatment is briefly discussed.
Introduction Odor problems arising from sewage treatment works have continued to grow in significance over recent years (1, 2). Common odorous compounds include reduced sulfur, nitrogen compounds, organic acids, aldehydes, and ketones. Sulfur-containing compounds have been identified as predominant odorants in a wide range of odor emissions in sewage treatment works, composting plants, and rendering plants (3-5). Hydrogen sulfide (H2S) is produced as a byproduct from anaerobic bacterial reduction of sulfate and decomposition of sulfur-containing organic constituents of manure (3, 4). Methyl mercaptan (CH3SH) is a representative odorous compound with a very low odor threshold of around 0.4 ppb/v that arises mainly from various odor emission * Corresponding author phone: +852-27666016; fax: +85223346389; e-mail:
[email protected]. † The Hong Kong Polytechnic University. ‡ Sun Yat-sen University. § Florida Institute of Technology. 5890
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009
sources such as sewage and municipal solid waste (5-7). In addition to odor nuisance, H2S and CH3SH are also toxic to living organisms including plants (3-5). Benzothiazoles and thiols are organic sulfur-containing compounds, which are also of concern in wastewater (5, 8, 9). For example, 2-mercaptobenzothiazoles (MBT), found in wastewater effluents of rubber additive manufacturers and tanneries, is a toxic and poorly biodegradable compound. Considering the odorous and toxic nature of sulfur-containing compounds, a variety of approaches aimed at removing such compounds have been investigated (10). The most economical methods for sulfur-containing compound removal are biological processes (4, 11). However, biological processes take time to remove odors and also produce organic wastes as byproduct that need to be disposed of eventually. Chemical dosing with nitrate salts (KNO3, Ca(NO3)2) or iron salts such as ferric, ferrous, or their combination (FeCl3, FeCl2, Fe(NO3)3, Fe2(SO4)3) is common practice for minimizing the sulfide-associated odor emission in sewer systems (12). On the other hand, many chemical oxidation technologies have been applied to control odors from wastewater and sludge by adding chemical oxidants such as hydrogen peroxide, chlorine, hypochlorite, potassium permanganate, and ozone (13-15). For example, hydrogen peroxide has been applied to oxidize H2S, mercaptans, thiosulfate, and sulfur dioxide in wastewater, waste activated sludge, and thickened sludge (14, 15). Although a considerable reduction in odor released from the sludge could be achieved, oxidation processes still take time to react, and there are also concerns of unwanted oxidation products such as toxic chlorinated or ozonated byproduct from chlorination or ozonation processes (16, 17). Therefore, alternate oxidants with nontoxic byproduct are still desirable. Ferrate(VI) (FeVIO42-, Fe(VI)) has shown many advantageous properties such as higher reactivity and selectivity over traditional oxidant alternatives, and significant capability as a disinfectant and coagulant (16-19). Importantly, multifunctional properties can be utilized in a single Fe(VI) dose for treating sludge. Only limited research has previously been carried out to determine the odor reduction in sludge by Fe(VI) (20). This research is aimed at (a) determining the reaction rates of sulfur-containing compounds oxidation by Fe(VI) including relative reactivities and half-lives of oxidation, (b) seeking stoichiometric doses of Fe(VI) which will completely remove sulfur-containing compounds and form relatively nontoxic byproduct, and (c) exploring the potential of applying Fe(VI) for sludge odor control in sewage treatment works.
Material and Methods Potassium ferrate(VI) (K2FeO4) of high purity (>95%) was prepared according to the method reported earlier (19). The aqueous Fe(VI) solution was prepared by addition of solid K2FeO4 to 0.005 M Na2HPO4/0.001 M borate solution and Fe(VI) concentration was determined by the molar absorption coefficient (ε510 nm ) 1150 M-1cm-1) at pH 9.0 (17). Aqueous solutions of 2-mercaptobenzothiazole (MBT), thiosemicarbazide (NH2NHC(S)NH2, TSC), and thiourea dioxide (NH2C(SO2)NH2, TUDO) were prepared by dissolving their solids in deionized water. Solution pH was adjusted by addition of either H3PO4 or NaOH. In a series of stoichiometric experiments, equal volumes (10 mL) of Fe(VI) and sulfide (or TSC) solutions at different pH were mixed in which concentrations of substrates were 10.1021/es900397y CCC: $40.75
2009 American Chemical Society
Published on Web 07/07/2009
then filtered out from the liquid samples using 1.3 µm filter. The liquid portion of the sludge sample was used in carrying out the experiments. To determine the odor emission from the liquid sludge solution, 100 mL of sludge solution was placed in a Pyrex bottle and odor-free air was blown at a fixed flow rate of 400 mL min-1 through the bottle (SI Figure S2). Concentrations of H2S, CH3SH, and odor in the outlet gas were determined simultaneously at different time intervals. A fluorescence H2S analyzer (Teledyne Instruments, model 101E) and the methyl mercaptan sensor were used to determine the gaseous concentrations of H2S and CH3SH, respectively. For odor concentration, the gaseous sample was collected in a Nalophan 50 L sampling bag and an olfactometrytechnique(EuropeanStandardMethod,EN13725) using a dynamic forced-choice olfactometer (Olfacton-n2) was carried out (22). FIGURE 1. Percent of sulfate formed to consumed hydrogen sulfide as a function of [Fe(VI)]/[H2S] at pH 7.0, 9.0, and 11.0 at 25 °C. Condition: [Sulfide] ) 2.8 × 10-4 M. kept constant ([sulfide] ) 2.8 × 10-4 M; [TSC] ) 1.0 × 10-4 M) while [Fe(VI)] was varied in order to achieve molar ratios in the reaction mixtures, i.e., Fe(VI) concentrations ranged from 0.08 to 1.39 mM, to sulfide, and from 0.01 to 0.27 mM, to TSC. After reaction completion, the concentrations of substrates, intermediates, and products were analyzed in the solution using colorimetric and HPLC techniques. In performing the experiments of kinetics for determining rates for the reaction of Fe(VI) with MBT, the concentration of Fe(VI) was ∼5 µM, whereas the concentration range of MBT was 29-113 µM. In studying the oxidation of TUDO by Fe(VI) at various pH, the concentration of TUDO was 1.2 mM, whereas Fe(VI) concentrations ranged from 100-125 µM. The concentrations of MBT and TUDO reactions were monitored by measuring the absorbance of Fe(VI) at 510 nm wavelength, and rate constants represent the average of nine runs for each substrate concentration. The temperature of the reactions was controlled at 25.0 ( 0.1 °C. The oxidation experiments for CH3SH were conducted in a wet scrubbing column (Supporting Information (SI) Figure S1). CH3SH gas with a traceable concentration of 2000 ppm/v in air was supplied by BOC Gases and was used as an odorous gas source. A zero air generator (Thermo Environmental Inc., model 111) was used to supply a clean air stream as an odorfree gas source to dilute the 2000 ppm/v CH3SH gas in order to obtain 50 ppm/v CH3SH gas in a gas mixing chamber. The mixed synthetic foul gas at this concentration was introduced into the scrubber through a diffuser for 67 min with a total amount of CH3SH ) 3 × 10-6 mol. The careful control of gas flow rate was maintained at 20 mL min-1 using a mass flow controller to ensure slow passage of CH3SH gas through the wet scrubber. A methyl mercaptan sensor (Detcon DM-100CH3SH) was installed inside the mixing chamber for in situ monitoring of the CH3SH concentration during the experiments. This methyl mercaptan analyzer has a monitoring range of 0-100 ppm/v with reproducibility of (2%. The stoichiometry was examined by applying Fe(VI) dosages, ranged from 0.35 to 1.5 × 10-3 M at pH 9.0, to 50 ppm/v CH3SH gas. The concentration of sulfate ion as a product of Fe(VI) oxidation reactions was determined by Dionex ICS90 ion chromatography (IC) with an anion column (IonPac AS14A 4 × 250 mm). The anaerobically digested sludge sample with a solid content of 0.8% from a local sewage treatment plant was used to investigate the elimination of odor emission from sludge by Fe(VI) addition. To obtain reliable experimental data, homogeneous tests rather than heterogeneous tests were conducted, in which the raw sludge sample was well stirred in 400 mL of aqueous solution for several hours and
Results and Discussion Reactivity of Fe(VI) with Sulfur-Containing Compounds. Initially, the kinetics of the reaction of Fe(VI) with MBT was determined under pseudofirst-order conditions with MBT in excess at pH 9.0 and 25 °C. The absorbance profile of Fe(VI) at 510 nm versus time fits nicely to a single-exponential decay curve, indicating the reaction is the first-order with respect to Fe(VI). The plots of the pseudofirst-order rate constants, k1 (s-1) versus [MBT] were linear (SI Figure S3). This suggests that the rate law for this reaction is also the first-order with respect to MBT with an observed secondorder rate constant, k ) 2.3 × 104 M-1s-1. The kinetics was also studied for the reaction of benzothiazole (BT) with Fe(VI) using a similar approach at pH 9.0 and 25 °C. The secondorder rate constant, k, was determined to be 3.4 × 104 M-1s-1. Comparison of the reactivity between MBT and BT suggests that S is the most probable site of attack in MBT, resulting in a higher reactivity of MBT than that of BT with Fe(VI) (17). Next, the oxidation of TSC by Fe(VI) was examined at pH 9.0 and 25 °C. The reaction was rapid and complete before reaching the reaction mixing cell of stopped-flow spectrophotometer (dead time ) 10 ms), hence no rate constant for this reaction could be obtained. Finally, the rate of oxidation of TUDO by Fe(VI) was determined as a function of pH (9.6-11.7) at 25 °C. TUDO was determined as an intermediate of the oxidation of TSC by Fe(VI) and the rates explain the stoichiometry of TSC oxidation by Fe(VI) (see below). Similar to MBT, the rate law was firstorder with respect to both Fe(VI) and TUDO and the second-order rate constant, k, decreased nonlinearly from 4.8 × 102 M-1s-1 at pH 9.6 to 2.8 × 101 at pH 11.7 (SI Figure S4). The pH dependence of k in the studied pH range can be attributed to the speciation effects of Fe(VI). Three protonated forms of Fe(VI) have been suggested (H3FeO4+ S H+ + H2FeO4, pKa1 ) 1.6 ( 0.2; H2FeO4 S H+ + HFeO4pKa2 ) 3.5; HFeO4- S H+ + FeO42-, pKa3 ) 7.3 ( 0.1) (17). In the pH range studied, two forms of Fe(VI), namely, HFeO4- and FeO42-, react with TUDO. HFeO4 + TUDO f Fe(OH)3 + product(s)
(1)
FeO2+ TUDO f Fe(OH)3 + product(s) 4
(2)
The rate of disappearance of Fe(VI) is given by 2-d[Fe(VI)]/dt ) k1[HFeO4 ][TUDO] + k2[FeO4 ][TUDO] (3)
k can be substituted into eq 4 considering the equilibrium of Fe(VI) species. k ) k1RHFeO4 + k2RFeO4 VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
(4)
9
5891
where RHFeO4 ) [H+]/([H+] + Ka,HFeO4);RFeO4 ) Ka,HFeO4/([H+] + Ka,HFeO4) The values of the individual rate constants, k1 ) 1.16 ( 0.05 × 105 M-1s-1 and k2 ) 2.71 ( 1.16 × 103 of eq 4 were obtained by nonlinear regression of the data. Similar to the results from previous studies, protonated Fe(VI), (HFeO4-), reacts with TUDO faster than the deprotonated FeO42- species (17, 21, 23-25). The fraction of HFeO4species decreases with increase in pH and thus the rate of oxidation of TUDO by Fe(VI) decreases as well. The reactivity of Fe(VI) with sulfur-containing compounds, which may be found in wastewater and sludge, at pH 9.0 and 25 °C, are given in Table 1 (17, 21, 23, 25). The rate constants vary from 2 × 10-1 to 7.4 × 105 M-1s-1 with sulfide, cysteine, and mercaptans displaying the highest reactivity and conversely polythionates the lowest rates with Fe(VI). It seems that the nucleophilicity on the sulfur atom determines the rate of oxidation. Furthermore, the presence of electron withdrawing groups on the sulfur atom reduces the rate. For example, the SO3 moiety attached to the S(-2) atom in thiosulfate decreased the apparent rate constant in comparison with hydrogen sulfide. The calculated half-lives (t1/2) at a dose of 10 mg L-1 K2FeO4 for oxidation of most sulfur-containing compounds were in seconds. Importantly, sulfide and mercaptans, major sludge odor compounds had values of t1/2 e 1 s for oxidation by Fe(VI) at pH 9.0. Furthermore, oxidation rates increase with decrease in pH (17), hence removal of hydrogen sulfide would most likely occur with half-lives in sub milliseconds at pH < 7. Stoichiometry. The stoichiometric experiments on the removal of sulfide using Fe(VI) were conducted at pH 7.0, 9.0, and 11.0 by mixing known amounts of Fe(VI) and sulfide. The percentage of the sulfate formed due to sulfide consumption versus the molar ratio of Fe(VI) to total sulfide is shown in Figure 1. As the concentration of Fe(VI) increases, the percentage of sulfate produced increases and a molar ratio of Fe(VI) to sulfide of ∼2.5 was required for complete transformation of (S(-2)) to sulfate (S(+6))
FIGURE 2. Fe(VI) oxidation of TSC at different pH. Condition: [TSC] ) 1.0 × 10-4 M. in the studied pH range. This molar ratio is reasonably close to the ratio of stoichiometric equation (eq 5). 2+ 8HFeO4 + 3H2S + 6H2O f 8 Fe(OH)3 + 3SO4
2OH- (5) Because of the high reactivity of Fe(VI) with sulfide and possible intermediates such as thiosulfate, and sulfite (23), Fe(VI) was able to efficiently oxidize sulfide to sulfate in the pH range from 7.0-11.0. At different molar ratios of Fe(VI) to sulfide, an increase in solution pH at different initial pH values was observed (SI Figure S5). This is consistent with eq 5. The pH value changed from the initial values (8.4, 9.0, and 9.6) to as high as 9.5, 9.7, and 10.8, after reaction with Fe(VI) consecutively. Figure 2 shows the results of stoichiometric experiments for the oxidation of TSC at pH 7.0-11.3. As the concentration of Fe(VI) increased, the amount of TSC decreased with complete removal at 1:1 stoichiometry ([Fe(VI)]:[TSC]). This stoichiometry was independent of pH (Figure 2). This is not surprising because of the instantaneous reactivity of Fe(VI) with TSC in the studied pH range. Because of increase in reactivity of Fe(VI) with decreasing pH in the range of the
TABLE 1. Fe(VI) Oxidation of Sulfur-Containing Contaminants at 25 °C compound hydrogen sulfide cysteine dithionite 2-mercaptoethanesulfonic acid 2-mercaptobenzoic acid 2-mercaptobenzothiazole 3-mercaptopropionic acid mercaptonicotinic acid thioacetamide thiourea dioxide thiourea thiosulfate benzenesulfinate methionine cystine diethylsulfide thiodiethanol thioxane pentathionate trithionate tetrathionate
5892
9
pH
kapp, (M-1s-1)
t1/2 (10 mg L-1 K2FeO4)
H2S HSCH2CH(NH2)COOH S2O42-
9.0 9.0 9.0
7.4 × 105 6.0 × 105 6.3 × 104
1.9 ms 2.3 ms 0.22 s
23 17 25
C2H6O3S2
9.0
3.0 × 104
0.46 s
25
10.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
2.5 × 10 2.3 × 104 1.3 × 104 1.2 × 104 5.5 × 103 4.6 × 103 3.4 × 103 7.2 × 102 1.4 × 102 1.3 × 102
0.55 s 0.60 s 1.1 s 1.2 s 2.5 s 3.0 s 4.1 s 19.2 s 98.7 s 106 s
17 this study 25 25 17 this study 21 17 17 17
9.0
1.2 × 102
115 s
17
8.0 8.0 9.0 9.0 9.0 9.0
1.0 × 102 1.0 × 102 5.8 × 101 3.5 × 100 1.1 × 100 2.0 × 10-1
138 s 138 s 238 s 66 min 3.5 h 19.2 h
17 17 17 25 25 25
molecular formula
HOOCC6H4SH C7H5NS2 HSCH 2CH 2COOH C6H5NO2S CH3C(S)NH2 NH2(SO2)NH2 NH2CSNH2 S2O32C6H5SO2 CH3SCH2CH2CH(NH2)COOHOOCCH(NH2)CH2SSCH2(NH2)CHCOOH CH3(S)CH3 OHCH2CH2(S)CH2CH2OH C4H8OS S5O62S3O62S4O62-
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009
4
reference
occurred when the Fe(VI) to CH3SH concentration ratio reached a value of 4.6. In addition, once the stoichiometry of Fe(VI): CH3SH was greater than 4.6, the CH3SH concentration in the outlet gas was measured to be
