Optimization of Simultaneous Chemical and Biological Mineralization ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1999, p. 2784–2788 0099-2240/99/$04.0010 Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Vol. 65, No. 6

Optimization of Simultaneous Chemical and Biological Mineralization of Perchloroethylene† ¨ YU ¨ KSO ¨ NMEZ,1 THOMAS F. HESS,1* RONALD L. CRAWFORD,1 ANDRZEJ PASZCZYNSKI,1 FATI˙H BU 2 AND RICHARD J. WATTS Center for Hazardous Waste Remediation Research, University of Idaho, Moscow, Idaho 83844,1 and Department of Civil and Environmental Engineering, Washington State University, Pullman, Washington 991642 Received 20 October 1998/Accepted 10 March 1999

Optimization of the simultaneous chemical and biological mineralization of perchloroethylene (PCE) by modified Fenton’s reagent and Xanthobacter flavus was investigated by using a central composite rotatable experimental design. Concentrations of PCE, hydrogen peroxide, and ferrous iron and the microbial cell number were set as variables. Percent mineralization of PCE to CO2 was investigated as a response. A second-order, quadratic response surface model was generated and fit the data adequately, with a correlation coefficient of 0.72. Analysis of the results showed that the PCE concentration had no significant effect within the tested boundaries of the model, while the other variables, hydrogen peroxide and iron concentrations and cell number, were significant at a 5 0.05 for the mineralization of PCE. The 14C radiotracer studies showed that the simultaneous chemical and biological reactions increased the extent of mineralization of PCE by more than 10% over stand-alone Fenton reactions. Halogenated organic chemicals have been introduced into the environment from a variety of sources, including the improper disposal of degreasers and solvents (14). Many of the halogenated compounds, such as perchloroethylene (PCE) and carbon tetrachloride, are persistent in the environment due to their resistance to microbial degradation and their toxicity to microorganisms. For the treatment of these recalcitrant compounds, a number of chemical processes have been investigated, including oxidation by Fenton’s reagent. Compound degradation by Fenton’s reagent may yield (i) partial mineralization (15), (ii) lowered toxicity (1), and (iii) increased susceptibility to biodegradation (4). Owing to their ability to lower the toxicity and increase the biodegradability of the parent compounds, chemical reactions have been coupled in sequence with biological reactions and have been the subject of much recent research. An extensive review is available elsewhere (19). However, the investigation of simultaneous chemical and biological transformation processes has received little attention. Simultaneous processes could have both economic and process advantages if applied to industrial pollution or in situ hazardous waste treatment in the environment. Fenton (8) discovered the strong oxidizing power of mixtures of hydrogen peroxide and ferrous iron solutions. Haber and Weiss (10) later identified the oxidizing element in Fenton’s reagent as a hydroxyl radical. The hydroxyl radical is a nonspecific, strong oxidant which reacts with most organic and biological molecules at near diffusion-controlled rates (.109 M21 s21) (7, 12). The classic Fenton’s reaction involves the addition of dilute hydrogen peroxide to a degassed, acidic ferrous iron solution, which generates hydroxyl radicals (equation 1). The degradation of organic chemicals by hydroxyl radicals then proceeds via hydroxylation, hydrogen atom abstraction, or dimerization (21).

H2O2 1 Fe21 3 Fe31 1 OHz 1 OH2

(1)

Some environmental applications of Fenton’s reagent involve reaction modifications, including the use of high concentrations of hydrogen peroxide, the substitution of different catalysts such as ferric iron and naturally occurring iron oxides, and the use of phosphate-buffered media and metal-chelating agents. These conditions, although not as stoichiometrically efficient as the standard Fenton’s reactions, are often necessary to treat industrial waste streams and contaminants in soils and groundwater (20). As a part of our overall simultaneous chemical and biological transformation study, PCE was chosen as a probe compound for experimental investigation. PCE was known to be degraded by Fenton’s reagent, yielding dichloroacetic acid (DCAA) (14, 18), a readily biodegradable compound. A major obstacle to combining the chemical and microbial reactions simultaneously was the toxicity of Fenton’s reagent to microorganisms. We previously showed that the toxic effects of Fenton’s reagent were sufficiently reduced by preacclimation of microorganisms to high concentrations of hydrogen peroxide, thereby allowing a significant number of microorganisms to survive throughout the course of treatment (3). In a later study, we investigated coexisting chemical and biological reactions used for mineralization of PCE (2). The results of that study, with 14C-labeled PCE, showed that the addition of microorganisms increased the extent of mineralization of PCE by more than 10% over that of a noninoculated control. This finding suggested that chemical and biological reactions could coexist and might be a viable alternative for the treatment of wastewaters containing PCE. In the present study, we investigated the effects of variables of coexistent chemical and biological reactions (concentrations of PCE, hydrogen peroxide, and ferrous iron and the initial cell number) on mineralization of PCE. Chemicals. PCE and DCAA-sodium salt were purchased from Aldrich Chemical Co. (Milwaukee, Wis.); 14C[1-2]-PCE was obtained from Sigma Chemical Co. (St. Louis, Mo.); hydrogen peroxide (30%) was purchased from Fisher Scientific (Fair Lawn, N.J.); and Ecolite scintillation cocktail was pur-

* Corresponding author. Mailing address: Center for Hazardous Waste Remediation and Research, University of Idaho, Moscow, ID 83844-0904. Phone: (208) 885-7461. Fax: (208) 885-7908. E-mail: tfhess @uidaho.edu. † Publication number 99301 of the Idaho Agricultural Experiment Station. 2784

VOL. 65, 1999

OPTIMIZATION OF PERCHLOROETHYLENE MINERALIZATION

2785

TABLE 1. Experimental design and results of 14C radiotracer experiments for the investigation of response surfaces describing the effects of the variables of PCE, hydrogen peroxide, and ferrous iron concentration and the cell number on the mineralization of PCE by simultaneous chemical and biological reactions PCE (mg/liter)

Fe(II) (MM)a

H2O2 (mg/liter)

Cells (log10/ml)

% Observed mineralization

% Recovery of 14C

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

3 7 3 7 3 7 3 7 3 7 3 7 3 7 3 7

0.15 0.15 0.45 0.45 0.15 0.15 0.45 0.45 0.15 0.15 0.45 0.45 0.15 0.15 0.45 0.45

25 25 25 25 75 75 75 75 25 25 25 25 75 75 75 75

5 5 5 5 5 5 5 5 7 7 7 7 7 7 7 7

49.1 40.4 44.3 46.1 54.4 74.1 63.8 57.8 46.4 39.6 38.4 47.1 62.2 52.6 54.2 58.4

100.8 87.4 98.9 99.2 94.3 110.7 111.8 99.2 99.3 93.5 94.6 105.5 100.4 90.9 97.1 98.0

17 18 19 20 21 22 23 24

1 9 5 5 5 5 5 5

0.3 0.3 0.0 0.6 0.3 0.3 0.3 0.3

50 50 50 50 0 100 50 50

6 6 6 6 6 6 4 8

51.2 59.7 17.4 48.6 20.0 55.7 53.4 40.4

88.3 92.9 97.2 87.5 97.3 85.2 96.4 90.1

25 26 27 28 29 30 31

5 5 5 5 5 5 5

0.3 0.3 0.3 0.3 0.3 0.3 0.3

50 50 50 50 50 50 50

6 6 6 6 6 6 6

46.7 55.5 60.8 52.8 50.0 60.9 54.4

82.1 89.2 96.1 89.9 86.8 98.7 90.6

C1 C2 C3 C4 C5

5 5 5 5 5

0.3 0.3 0.3 0.3 0.3

50 50 50 50 50

NCb NC NC NC NC

47.7 45.0 42.7 43.2

101.0 101.0 97.9 97.5

Expt no.

a b

Design location

Factorial design

Star points

Center

Control

Fe(II) solution contained 103 solution of NTA chelating agent. NC, no cells.

chased from ICN Pharmaceuticals (Costa Mesa, Calif.). Carbo-Sorb was obtained from Packard Instruments (Meriden, Conn.), and Ready Organic was purchased from Beckman (Fullerton, Calif.). Organism and culture conditions. A hydrogen peroxideresistant strain of Xanthobacter flavus FB71 (3) was used for the biotransformation of DCAA, the major product of PCE degradation by modified Fenton’s reagent (14, 18). Cells were grown in a continuous-flow fermentor (New Brunswick Scientific BioFlo III) in a 1.5-liter volume, at steady state, with growth medium replacement at a rate of 1 liter per day at 30°C. The fermentor was stirred at 200 rpm and aerated with 3 liters of sterile air flow per min. One liter of medium contained 300 mg of DCAA per liter, 200 mg of H2O2 per liter, 3.4 g of Na2HPO4, 1.5 g of KH2PO4, 0.25 g of NaCl, 0.5 g of NH4Cl, 0.12 g of MgSO4, 5.55 mg of CaCl2, 2 ml of Wolfe’s mineral solution (9), and 2 ml of vitamin supplement solution containing 50 mg each of biotin, thiamine HCl, and nicotinic acid per liter. Experimental design. In order to decrease the number of experiments while maintaining statistical significance, a central

composite, rotatable experimental design was developed as outlined by Cochran and Cox (5). The four-level design, shown in Table 1, included PCE, H2O2, and Fe(II)-chelate concentrations and the initial cell number as the experimental variables. The percent mineralization, determined by radiotracer analysis, was the response measured. In addition, microbial survival at the end of the each experimental point was surveyed qualitatively by using nutrient agar plates and the spread plate technique. The design contained three blocks of experiments based on a second-order design. Block 1 was a 23 factorial design that formed the first-order portion of the design. Block 2 provided the “star” points, which provided the second-order portion of the design, and block 3 was defined as the center of the experimental design, providing an estimate of the error of the measurements. Experimental setup and 14C radiotracer experiments. Experiments were performed according to the experimental design outlined in Table 1 and according to the procedures given below. In order to minimize the autooxidation of ferrous iron and the loss of PCE by volatilization, the experimental order was as follows: (i) 55-ml serum bottles were filled with M9

2786

¨ YU ¨ KSO ¨ NMEZ ET AL. BU

APPL. ENVIRON. MICROBIOL. TABLE 2. Parameter estimates for the second-order response surface equation (see text, equation 2) describing simultaneous chemical and biological mineralization of PCE Term

14

SE

Intercept (I)

29.311

8.097

Cells (C) Hydrogen peroxide (H) Ferrous iron (F)

5.774 0.888 144.181

1.724 0.186 31.031

20.681 20.006 2215.037

0.228 0.002 49.866

Cells2 (C2) Hydrogen peroxide2 (H2) Ferrous2 iron (F2) FIG. 1.

Estimate

C purge and trap apparatus.

buffer solution and distilled water so that the final solution would result in half-strength M9 buffer solution in 40 ml; (ii) hydrogen peroxide was added; (iii) microorganisms were added (except for controls); (iv) bottles were crimped; (v) PCE was added with a syringe to a resulting concentration of 5 mg/liter; and (vi) reactions were initiated by the addition of Fe(II)-nitrilotriacetic acid (NTA) solution. (Ferrous iron solution [10 mM] was prepared anaerobically by using 100 mM NTA solution degassed for 15 min with a 10-ml/min flow each of nitrogen and hydrogen.) The bottles were then incubated on a rotary platform shaker at 200 rpm at an ambient temperature for 72 h. Hydrogen peroxide and ferrous iron solutions were added every hour as indicated in Table 1 for the first 5 h of the experiments. At the end of 72 h of incubation, aliquots for microbial survival were withdrawn by a syringe and spread on nutrient agar plates. The remaining medium was acidified by the addition of 1.5 N HCl and purged with nitrogen at the flow rate of 60 ml/min for 30 min by using the purge and trap apparatus shown in Fig. 1. Volatile organic compounds were captured in three organic traps containing 15 ml of Ready Organic scintillation cocktail and analyzed directly for radioactivity. Carbon dioxide was captured by one 15-ml Carbo-Sorb trap, and 1 ml of this was transferred into Ecolite scintillation cocktail prior to counting. One milliliter of the remaining medium was mixed with 1 ml of cell solubilizer and incubated at 45°C for 30 min. The bottles were cooled for approximately 15 min prior to the addition of 15 ml of Ecolite scintillation cocktail for counting. All samples were counted with a Packard Tri-Carb 2100TR liquid scintillation counter by using 14C protocol. Counts per minute were converted to disintegrations per minute by using an efficiency plot for known 14C quench standards and a spectral index of the sample number. Results and discussion. The results of the experiments, the observed percent PCE mineralization, along with the experimental variable values, are shown in Table 1. A statistical analysis software package (JMP version 3.2, professional edition; SAS Institute, Inc., Cary, N.C.) was used to analyze the results, provide estimated mineralization values based on a second-order model fit to the data (Table 1), and generate response surface plots. A second-order model, typical for response surface equations (5), was reduced to the following form based on the results of this experimentation:

grams per liter), F is the ferrous iron concentration (millimolar), and a, b, c, d, e, and f are the coefficients of the experimental variables. Coefficients of the second-order model variables and the corresponding standard errors are given in Table 2. All model terms were analyzed by t test, and any term found to be insignificant at a 5 0.05 was excluded from the model. Although PCE was set as a variable in the preparation of the model, the statistical analysis showed that the concentration of PCE, within the tested boundaries of the model, had no significant effect on the model. In addition, all cross terms (e.g., F 3 C) were found to be statistically insignificant. It should be noted that the experimental points 1, 2, and 3 from Table 1 had to be rerun due to procedural problems that occurred during the original experimental run. Three control experiments on the center location (Table 1) were conducted in parallel with the repeated points to check for their validity. The mean and standard error values of additional controls (data not shown) were within the range of the original experiments, leading to the conclusion that the rerun points were valid. The regression analysis of the experimental results against estimated values using the model showed a corresponding coefficient of correlation (R2) of 0.72 (Fig. 2). Furthermore, based on the lack-of-fit analysis (Table 3), the second-order response model appeared to adequately fit the data; comparison of mean squares for the lack-of-fit and error terms showed values of similar order, indicating an adequate model fit (5).

% PCE mineralization 5 I 1 aC 1 bH 1 cF 1 dC2 1 eH2 1 fF2

(2)

where I is the intercept, C is the initial cell number (log cell number), H is the hydrogen peroxide concentration (in milli-

FIG. 2. Regression plot of observed data against predicted values from the second-order response surface model describing simultaneous chemical and biological mineralization of PCE.

VOL. 65, 1999

OPTIMIZATION OF PERCHLOROETHYLENE MINERALIZATION

TABLE 3. Lack-of-fit analysis for the second-order model given in Table 2 Source

Degrees of freedom

Sum of squares

Mean square

Lack of fit Pure error Total error

9 19 28

506.65 517.96 1,024.61

56.29 27.26 36.59

The optimal solution, concentrations of variables yielding the highest predicted mineralization of 62% within the boundaries of the tested model, were as follows: 104.2 cells/ml, 77 mg of hydrogen peroxide per liter, and 0.33 mM iron(II). In order to elucidate clearly the effects of the presence of microorganisms, sterile experiments with five replicate samples were included in the analysis, with the results given in Table 1. One of the datum points was determined to be an outlier by using Chauveret’s criterion, as described by Huggins (13), at a 5 0.05 and was excluded from the calculations and Table 2. The mean mineralization in the absence of microorganisms was found to be 44.6%, with the standard error of 1.13, which was substantially lower than the mineralization obtained under the same conditions in the presence of microorganisms (54.3 6 1.99%). These results statistically show that the extent of mineralization of PCE in simultaneous chemical and biological systems was increased by the presence of microorganisms. Moreover, the differences in mineralization results, with or without microorganisms, are in accordance with our previous study where we showed the coexistent chemical and biological mineralization of PCE with respect to time in a similar experimental setup (2). The typical response surface plot of the second-order model, generated by keeping one of the variables constant at the vicinity of the optimal point of the corresponding variable, is shown in Fig. 3. The extent of both PCE degradation and mineralization, as expected, decreased when the cell number

2787

increased beyond the optimal point, perhaps due to increasing cell populations which became a sink for hydrogen peroxide and the free radicals formed due to enzymatic quenching. The results showed that when the hydrogen peroxide concentration exceeded the optimum of 77 mg/liter, the extent of mineralization decreased slightly. This phenomenon may be attributed to (i) the lethal effect of high concentrations of hydrogen peroxide to microorganisms or (ii) the formation of other reactive oxygen species such as HO2z (hydroperoxyl radical) and O22 (superoxide) from the excess hydrogen peroxide (11, 16). Another observation on the effect of hydrogen peroxide was the partial degradation and mineralization of PCE even with no added hydrogen peroxide (Table 1, Fig. 3). Cohen (6) reported a series of reactions, in buffered, aerobic systems, yielding hydroxyl radicals and superoxide. In addition to the reaction sequences documented by Cohen (6), a further possible explanation for the observed results is the in vivo generation of reactive oxygen species (superoxide and hydroxyl radical) during the reduction of molecular oxygen to water (17). Excess ferrous iron concentrations can also act as a sink for hydroxyl radicals (6), generating Fe(III), and may be a final possible explanation for the decrease in PCE mineralization with peroxide and iron concentrations greater than the optimal point. The results of this study confirmed that chemical and biological reactions used for the mineralization of PCE were coexistent. The three statistically significant variables of the hydrogen peroxide and ferrous iron concentrations and the initial cell number showed optimal values, giving maximum PCE mineralization within the ranges studied. A fourth variable, the PCE concentration, was found to be statistically insignificant and did not affect mineralization extent. Although explanations for the experimental results were discussed in relation to values and mechanisms in the literature, further investigation is needed to clearly understand the combined chemical and biological interactions. For instance, the effects and interactions of medium formulations, such as vitamins and minerals, need to be investigated and optimized for PCE mineralization extent. Finally, for simultaneous abiotic and biotic reactions to be of practical use, the economics of the process need to be developed in relation to currently accepted active remediation schemes. This research was supported by National Science Foundation grant 9613258 from a joint program of the National Science Foundation and U.S. Environmental Protection Agency.

FIG. 3. Response surface plot for simultaneous chemical and biological mineralization of PCE at a ferrous iron concentration of ca. 0.3 mM.

REFERENCES 1. Barton, D. A., and E. P. Drake. 1994. Biodegradability of blow heat condensates with and without hydrogen peroxide. Water Sci. Technol. 29:229–238. 2. Bu ¨yu ¨kso ¨nmez, F., T. F. Hess, R. L. Crawford, A. Paszcynski, and R. J. Watts. 1998. Simultaneous abiotic-biotic mineralization of perchloroethylene (PCE), p. 277–282. In G. B. Wichramanayake and R. E. Hinchee (ed.), Designing and applying treatment technologies: remediation of chlorinated and recalcitrant compounds. Battelle Press, Columbus, Ohio. 3. Bu ¨yu ¨kso ¨nmez, F., T. F. Hess, R. L. Crawford, and R. J. Watts. 1998. Toxicity of modified Fenton’s reaction on Xanthobacter flavus FB71. Appl. Environ. Microbiol. 64:3759–3764. 4. Carberry, J. B., and T. M. Benzing. 1991. Peroxide pre-oxidation of recalcitrant toxic waste to enhance biodegradation. Water Sci. Technol. 23:367– 376. 5. Cochran, W. G., and G. M. Cox. 1992. Experimental designs. John Wiley and Sons, Inc., New York, N.Y. 6. Cohen, G. 1987. The Fenton reaction, p. 55–64. In R. A. Greenwald (ed.), CRC handbook of methods for oxygen radical research. CRC Press, Inc., Boca Raton, Fla. 7. Dorfman, L. M., and G. E. Adams. 1973. Document NSRDS-NBS-46. National Bureau of Standards, Washington, D.C. 8. Fenton, H. J. H. 1894. Oxidation of tartaric acid in presence of iron. J. Chem. Soc. 65:899–910. 9. Gherna, R., P. Pienta, and R. Cote. 1992. Catalogue of bacteria and phages,

2788

¨ YU ¨ KSO ¨ NMEZ ET AL. BU

18th ed. American Type Culture Collection, Rockville, Md. 10. Haber, F., and J. Weiss. 1934. The catalytic decomposition of hydrogen peroxide by iron salts. Proc. R. Soc. London 147A:332–351. 11. Halliwell, B., and J. M. C. Gutteridge. 1985. Free radicals in biology and medicine. Clarendon Press, Oxford, United Kingdom. 12. Hoigne, J., B. C. Faust, W. R. Haag, F. E. Scully, and R. G. Zepp. 1989. Aquatic humic substances as source and sink of photochemically produced transient reactants, p. 363–381. In I. Suffet and P. L. McCarty (ed.), Aquatic humic substances: influence on fate and treatment of pollutants. American Chemical Society, Washington, D.C. 13. Huggins, L. F. 1983. Analysis and interpretation, p. 1501–1513. In B. W. Mitchell (ed.), Instrumentation and measurement for environmental sciences. American Society of Agricultural Engineers, St. Joseph, Mo. 14. Leung, S. W., R. J. Watts, and G. C. Miller. 1992. Degradation of perchloroethylene by Fenton’s reagent: speciation and pathway. J. Environ. Qual. 21:377–381. 15. Li, Z. M., S. D. Comfort, and P. J. Shea. 1997. Destruction of 2,4,6-Trinitrotoluene by Fenton Oxidation. J. Environ. Qual. 26:480–487.

APPL. ENVIRON. MICROBIOL. 16. Pignatello, J. J., and Y. Sun. 1993. Photo-assisted mineralization of herbicide wastes by ferric ion catalyzed hydrogen heroxide, p. 77–84. In D. W. Tedder and F. G. Pohland (ed.), Emerging technologies in hazardous waste management III. American Chemical Society, Washington, D.C. 17. Ryckman, D. W., and M. D. Ryckman. 1980. Organizing to cope with hazardous material spills. J. Am. Water Works Assoc. 72:196–200. 18. Sato, C., S. W. Leung, H. Bell, W. A. Burkett, and R. J. Watts. 1993. Decomposition of perchloroethylene and polychlorinated biphenyls with Fenton’s reagent, p. 256–343. In D. W. Tedder and F. G. Pohland (ed.), Emerging technologies in hazardous waste management III. American Chemical Society, Washington, D.C. 19. Scott, J. P., and D. F. Ollis. 1995. Integration of chemical and biological oxidation processes for water treatment: review and recommendations. Environ. Prog. 14:88–103. 20. Tyre, B. W., R. J. Watts, and G. C. Miller. 1991. Treatment of four biorefractory contaminants in soils using catalyzed hydrogen perixide. J. Environ. Qual. 20:832–838. 21. Walling, C. 1975. Fenton’s reagent revised. Acta Chem. Res. 8:125–131.