Novel method for determination of phenol ...

Novel method for determination of phenol degradation kinetics

Soonjae Lee, Dong-Ju Kim & Jae-Woo Choi

Bioprocess and Biosystems Engineering ISSN 1615-7591 Volume 36 Number 12 Bioprocess Biosyst Eng (2013) 36:1939-1945 DOI 10.1007/s00449-013-0970-y

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Bioprocess Biosyst Eng (2013) 36:1939–1945 DOI 10.1007/s00449-013-0970-y

ORIGINAL PAPER

Novel method for determination of phenol degradation kinetics Soonjae Lee • Dong-Ju Kim • Jae-Woo Choi

Received: 21 March 2013 / Accepted: 7 May 2013 / Published online: 22 May 2013 Ó Springer-Verlag Berlin Heidelberg 2013

Abstract In this study, we proposed a new method for estimating biokinetic parameters in phenol degradation kinetics. The new method relies on the new formulation of q–S relation where degradation rate q is calculated from the changes of substrate concentration S for each time segment during the course of entire degradation, while in the conventional method q is obtained from the slope of the straight line that is given as substrate concentration changes with time in a semi-logarithmic scale. Thus, this new method provided more data points than the conventional method. The q–S relations obtained from the new method and the conventional method were fitted with three inhibitory kinetic models of Haldane, Yano and Edwards. Simulation of degradation profile with each kinetic model and comparison with the observed profile revealed that the new method offered a better prediction with Edwards model as the best inhibitory model. Keywords Biokinetic Inhibitory model Phenol Simulation

Introduction Phenol is one of the harmful pollutants in wastewater discharged from wide industrial branches such as S. Lee D.-J. Kim (&) Department of Earth and Environmental Sciences, Korea University, Anam Dong 5-1, Seoul 136-701, Republic of Korea e-mail: [email protected] S. Lee J.-W. Choi Center for Water Resource Cycle Research, Korea Institute of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea

petrochemical, chemical, pharmaceutical and coal refining. The wastewater has to be treated before discharge in surface water. Among various methods, biological treatment of phenol in the wastewater has been upcoming alternative against conventional physical/chemical methods since it is cost effective and environmentally friendly. Biodegradation of phenol has been widely studied by many researchers. In general, phenol concentration in batch microcosm shows a non-linear decrease with time (Fig. 1a). And then the phenol degradation often follows the first-order kinetics dS ¼ qS dt

ð1Þ

where S is the substrate concentration (mg l-1), q the firstorder degradation rate (h-1), and t is time (h). Conventionally values of q were determined by the straight-line method (SLM) using the plot of ln (S/S0) vs. time (Fig. 1b). Several pairs of q and S0 can be obtained from the degradation data which are acquired for the degradation experiments performed with various initial phenol concentrations (S0). Dependency of q on S0 has been expressed by either noninhibitory or inhibitory models depending on the toxicity of substrate and nature and resistance of bacteria to the substrate. Banerjee and Ghoshal [1, 2] utilized the inhibitory models for the relationship between q and S0 and obtained biokinetic parameters such as maximum degradation rate, half saturation constant and inhibition constant from the relationship. The parameters have been applied in the dynamic model simulation of phenol degradation [3, 4]. However, the use of biokinetic parameters obtained from the q–S0 relation can be doubtful for its representativeness since the entire range of S (from S0 to 0) was used for the estimation of the degradation rate q for the condition of S0. The rate of phenol degradations can be also derived from

123

Author's personal copy 1940

Bioprocess Biosyst Eng (2013) 36:1939–1945

Fig. 1 a Typical degradation data obtained from batch microcosm test for initial substrate concentration S0 (substrate concentration S vs. time t), b conventional method for estimation of degradation rate, q, c new method for estimation of degradation rate from the time interval and d comparison of degradation kinetics with q as a function of S0 and S

the decrease in S during time intervals between the sampling times (Fig. 1c). For this case, the estimated q is corresponded against S of the time interval (q–S relation). Since S of the time interval is less than S0, the conventional relation between q and S0 may lead to incorrect estimation of q which can finally result in the overestimation of q (Fig. 1d). Therefore, in this study, we proposed a new method for determining q from the changes of phenol concentration for each time segment or step during the decline of S with time. The applicability of the proposed method was validated by comparison of simulated degradation profile using three different inhibitory models with the observed degradation profile.

Materials and methods Phenol degradation experiment For batch type experiment on phenol degradation by suspended cell, bacterial strain Pseudomonas putida F1 (Korea National Environmental Microorganisms Bank, Korea) was pre-cultured in 250 ml Erlenmeyer flasks containing 100 ml of LB medium at 30 °C in an orbital shaker at 140 rpm. Enrichment of P. putida F1 was done in a mineral salt medium (MSM) containing the following constituents per liter of distilled water: K2HPO4, 6 g; KH2PO4, 4 g; (NH4)2SO4, 2 g; MgCl2, 4.95 g; CaCl2, 1.25 g; H3BO3, 0.225 g; ZnSO47H2O, 0.6 g; NiSO4, 0.18 g; (NH4)6Mo7O244H2O, 0.134 g; CuSO45H2O,

123

0.0225 g; MnSO4, 0.375 g; CoCl2, 0.14 g; FeCl3 0.03 g [5]. One ml of culture grown on LB medium was transferred to 250 ml flasks containing 200 ml of MSM and incubated at 30 °C in an orbital shaker at 140 rpm. The bacterial cells in the late exponential-growth phase were harvested by centrifugation at 10,000 rpm at 4 °C for 15 min and re-suspended in fresh MSM and washed twice with MSM. After this, the cells were again centrifuged and re-suspended in MSM to a final concentration of 0.5 OD600, and used as a source of inoculum for the batch experiments. Batch experiment was conducted to investigate the phenol degradation using suspended cells at various initial concentrations of 50, 200, 500 and 1,000 mg l-1. Bacterial concentration corresponding to 0.5 OD600 of suspended cells was inoculated into 250 ml flasks containing MSM of 200 ml adjusted to the specified phenol concentrations. All flasks were sealed with parafilm and were shaken at 140 rpm at a temperature of 30 °C in the incubator until the experiment was completed. The suspension was sampled using a sterilized airtight glass syringe to determine phenol concentration at pre-determined times (0, 1, 2, 3, 4, 5, 6, 7, 8 and 9 days). Bacterial concentrations were determined by UV–visible spectrophotometer (Heyios ß, Thermo-Electron Corporation, Waltham, MA, USA) at 600 nm by measuring the absorbance of the cell solutions. Phenol was quantified using high-performance liquid chromatography (HPLC; Young Lin, Seoul, Korea) equipped with a fluorescence detector (M720), M925 pump, Rheodyne injector, and C18

Author's personal copy Bioprocess Biosyst Eng (2013) 36:1939–1945

1941

column (150 9 4.6 mm; Phenomenex, Torrance, CA, USA) at an excitation wavelength of 280 nm. The elution gradient was 40 % methanol/30 % distilled water with a flow rate of 1.0 ml min-1. The samples collected using 1 ml glass syringe were transferred to a 0.6-ml microtube, centrifuged at 3,000 rpm for 3 min, and 8 ll of aliquots of supernatant were injected to HPLC.

Table 1 Inhibitory kinetic models Kinetic models Haldane [6]

qmax S q ¼ KS þSþS 2 =K I

Yano et al. [7]

max S q ¼ KS þSþðSq2 =K I Þ ð1þS=KÞ

Edward [8]

q ¼ qmax ðexp ðS=KI Þ exp ðS=KS ÞÞ

qmax maximum degradation rate; KS, substrate affinity constant; KI, substrate inhibition constant, K, Yano constant

Estimation of degradation rate Straight-line method (SLM) Equation (1) dictating phenol degradation in the first-order rate can be expressed by q¼

1 dS S dt

ð2Þ

where S is concentration of substrate (mg l-1) at time t (h). The equation can be further reduced to q¼

dðln SÞ dt

ð3Þ

A plot of substrate concentration S vs. time in a semilogarithmic scale gives a straight line (Fig. 1b). The slope of this line is equal to q for each initial substrate concentration, S0. The phenol degradation kinetic can be obtained by corresponding the values of q against S0 [1, 2].

Fig. 2 Experimental data for degradation of phenol with initial concentrations from 50 to 1,000 mg l-1 by P. putida F1

Time segment method (TSM) During the microbial degradation, the rate of degradation changes continuously as the phenol concentration decreases because the degradation rates are dependent upon substrate concentrations. By applying forward finite difference approximation to Eq. (2), the degradation rates qi for the time interval from ti to ti?1 can be estimated using the following equation (Fig. 1c) qi ¼

1 Siþ1 Si Si Dt

ð4Þ

where Dt = ti?1 - ti. From the correspondence of qi against Si, phenol degradation kinetic data can be obtained. Degradation kinetics To represent the degradation kinetics of phenol, three inhibitory kinetic models, such as Haldane, Yano and Edwards models, were fitted to the q–S0 and q–S relations obtained by SLM and TSM to select the suitable models. The models are represented in Table 1 [6–8], where qmax is maximum degradation rate (h-1), KS is substrate affinity constant (mg l-1), KI is substrate inhibition constant (mg l-1), K is Yano constant (mg l-1).

Fig. 3 Degradation rates estimated from experimental data by straight-line method (SLM) and time segment method (TSM)

Verification of q–S0 and q–S relations The kinetic degradation parameters are frequently used to simulate degradation process under flow condition as well as batch condition. In these systems, the kinetic degradation can be simulated using the explicit form of Eq. (2) Siþ1 ¼ Si qi Si Dt

ð5Þ

123



In this study, the kinetic degradation of phenol was simulated using Eq. (5) and degradation rate calculated based on the three kinetic models and their parameters, where sufficiently small value of Dt (0.1 h) was used. The simulated results were compared with the observed results to verify the appropriateness of q–S0 and q–S relation, and to determine the most relevant model applicable to the phenol degradation kinetics observed in this study.

Results and discussion Observed phenol degradation The observed degradation of phenol is shown in Fig. 2. Phenol degradation exhibits different patterns depending on the initial concentration. For the case of S0 lower than 500 mg l-1, phenol was degraded completely in 150 h after lag period (ca. 24 h) while a negligible loss of phenol was observed at S0 = 1,000 mg l-1 due to the inhibitory effect of phenol on the bacterial cells. This inhibitory effect at phenol concentrations higher than 1,000 mg l-1 was reported by several studies [1, 2, 9–12]. Degradation rates Phenol degradation rates (q) estimated by SLM and TSM are shown in Fig. 3. It is noted that only four data points were obtained for the SLM since each q can be obtained from one degradation curve. Contrast to SLM, 19 data points were obtained for the TSM from the time intervals including lag period. In both methods, degradation rates showed a peak at around S0 = 50 mg l-1 and then declined sharply with the increased phenol concentration. This indicates that inhibitory model can be more appropriate for

Fig. 4 Fitting degradation rates estimated by a SLM and b TSM using Haldane, Yano and Edward models

the observed data set. Due to the employment of S0 in the q–S relation in SLM, SLM yielded mostly higher q values for a given S than TSM except for two data points (larger triangle in Fig. 3) which were obtained from the steepest

Table 2 Comparison of phenol degradation parameters obtained in this study and from previous researches Estimating q

Bacteria

a

Models

This study

Banerjee and Ghoshal [1]

TSM

SLM

SLM

SLM

P. putida

P. putida

AKG1

AKG2

H

Y

E

H

Y

E

H

Y

E

H

Y

E

qmax (h-1)

0.09

0.06

0.04

0.07

0.05

0.05

27.85

0.11

0.47

1.63

0.12

0.04

KS (mg l-1)

4.3

2.0

1.0

2.5

1.5

1.2

59,150

22.3

407

9.7

344.4

125.9

KI (mg l-1)

37.7

74.8

194.0

111.1

315.1

2.4

26,390

431.3

3,873

1,272

1,117

-1

K (mg l )

–

R2

0.76

599.3 0.77

–

–

0.76

0.95

4,008.5 16.3 0.99

–

–

33.7

–

–

276.2

–

0.99

0.64

0.59

0.84

0.82

0.86

0.79

The degradation rates were approximated using SLM or TSM. The parameters were obtained by fitting the Haldane, Yano and Edward models to the degradation rates H Haldane model, Y Yano model, E Edward model a

The strains AKG1 and AKG2 were identified as Bacillus cereus and named B. cereus MTCC9817 and B. cereus MTCC9818 [1]

123

Author's personal copy Bioprocess Biosyst Eng (2013) 36:1939–1945

1943

region of the degradation curve corresponding to S0 = 500 mg l-1.

Estimation of kinetic parameters The results of biokinetic model fit to the estimated q–S relation are shown in Fig. 4. For SLM, all the three inhibitory models (Haldane, Yano and Edwards) fit fairly well the q–S data with high goodness of fit (R2 0.95–0.99) (Fig. 4a). However, the kinetic curves showed difference in estimation of degradation rates in low and high concentrations. The difference between the model fits resulted from the lack of data point as well as characteristics of models. Model fits to the data obtained by TSM showed lower goodness of fit (R2 0.76–0.77) (Fig. 4b), since the data included the degradation rate in the lag period. The model fits estimate different degradation rates at high concentration due to different characteristics of the model. However, in the lower concentration, the degradation rates could be estimated similarly from the increase in the number of data point. It was noted that all three models gave lower q values for TSM than SLM for the S values ranging from 200 to 1,000 mg l-1, and Edwards model tends to yield lower estimation of q values at high S. Parameters estimated by model fit and the comparison with the previous studies are presented in Table 2. The values of qmax, KS obtained from the TSM are slightly higher than those from the SLM except for the values of KI. The parameters (qmax, KS and KI) estimated by TSM show more consistent values between the three models than those estimated by SLM. For instance, KI of Yano model was estimated as 4,008.5 by SLM. This value is not acceptable because the experimental data showed highly inhibited biodegradation of phenol at S0 = 1,000 mg l-1. Literature values of parameters estimated by SLM also showed inconsistency. For instance, the qmax, KS and KI values of Banerjee and Ghoshal [1] highly deviated among the models. Furthermore, the KI of Haldane and Yano models for AKG1 was estimated with unreasonably high values (59,200 and 26,000 mg l-1, respectively). The appropriateness of the estimated parameters is dependent on the preparation of data as well as characteristics of the models. The data prepared by TSM can derive more reasonable values of parameter by fitting increased number of data which can represent degradation processes more effectively.

Simulation of degradation profile The simulated degradation profiles using three biokinetic models obtained from the q–S relation based on the SLM

Fig. 5 Simulation of phenol degradation based on the a Haldane, b Yano and c Edward models using the parameters obtained by SLM

are compared with the observed data (Fig. 5). Surprisingly, none of the models could properly simulate the degradation profile at all concentrations used. The simulated profiles showed the highest deviation from that of the observed at the highest initial concentration, and the prediction was the worst for Haldane model. Comparison of the simulated vs. observed profile is shown in Fig. 6 for the case of TSM-based kinetic models. In contrast to SLM, the simulated profiles using the three inhibitory models of Haldane, Yano and Edwards can

123



Conclusion In this study, we proposed a new method for estimation of the first-order degradation rate coefficient q which is essential for determination of the first-order degradation kinetics q–S relation. The new method is based on the TSM where several q values are calculated from the q–S data for each measured time step. The q–S relation obtained from the TSM was compared with q–S0 relationship obtained from SLM where q is calculated from the slope of ln(S/S0) vs. time data. Three different inhibitory kinetic models (Haldane, Yano and Edwards) were applied to the q–S or q–S0 relation for parameter estimation. The parameters obtained by TSM showed more reasonable values with more consistency between the models than SLM due to the increase in the number of data point. Using the models and their parameters, dynamic degradation of phenol was simulated and compared with the observed degradation profile. Results showed that TSM offered a better prediction of phenol degradation than SLM. This is attributed the fact that phenol concentration decreases during dynamic degradation, and thus the degradation rate at time t is affected by the concentration at that time. Simulation based on TSM adopts degradation rates considering variation of phenol concentration during the reaction. On the other hand, SLM uses degradation rate q related to the initial concentration S0. Therefore, the estimation of degradation rate by TSM is more appropriate to describe dynamic phenol degradation than that by SLM. Simulating phenol degradation based on TSM, Edwards model gave the best result compared to other models. This may be attributed to the characteristic of Edwards model which shows tendency of yielding lower q values in the high concentration range.

References

Fig. 6 Simulation of phenol degradation based on the a Haldane, b Yano and c Edward models using the parameters obtained by TSM

describe the observed dynamic degradation of phenol more appropriately. Among the three inhibition models, Edwards model yielded the best match especially for high concentrations (S0 = 500 and 1,000 mg l-1). This is attributed to the fact that Edwards model and its parameters give lower q values for the high range of phenol concentration than the others as shown in Fig. 4.

123

1. Banerjee A, Ghoshal AK (2010) Phenol degradation by Bacillus cereus: pathway and kinetic modeling. Bioresource Technol 101:5501–5507 2. Banerjee A, Ghoshal AK (2011) Phenol degradation performance by isolated Bacillus cereus immobilized in alginate. Int Biodeter Biodegr 65:1052–1060 3. Banerjee I, Modak JM, Bandopadhyay K, Das D, Maiti BR (2001) Mathematical model for evaluation of mass transfer limitations in phenol biodegradation by immobilized Pseudomonas putida. J Biotechnol 87:211–223 4. Wolski EA, Durruty I, Haure PM, Gonza´lez JF (2012) Penicillium chrysogenum: phenol degradation abilities and kinetic model. Water Air Soil Pollut 223:2323–2332 5. Kim DJ, Choi JW, Choi NC, Mahendran B, Lee CE (2005) Modeling of growth kinetics for Pseudomonas spp. during benzene degradation. Appl Microbiol Biotechnol 69:456–462 6. Haldane JBS (1965) Enzymes. M.I.T. Press, Cambridge

Author's personal copy Bioprocess Biosyst Eng (2013) 36:1939–1945 7. Yano T, Nakahara T, Kamiyama S, Yamada K (1966) Kinetic studies on microbial activities in concentrated solutions. Part I Effect of excess sugars on oxygen uptake rate of a cell free respiratory system. Agric Biol Chem 30:42–48 8. Edwards VH (1970) The influence of high substrate concentrations on microbial kinetics. Biotechnol Bioeng 12:679–712 9. Chung TP, Tseng HY, Juang RS (2003) Mass transfer effect and intermediate detection for phenol degradation in immobilized Pseudomonas putida systems. Process Biochem 38:1497–1507

1945 10. Arutchelvan V, Kanakasabai V, Nagarajan S, Muralikrishnan V (2005) Isolation and identification of novel high strength phenol degrading bacterial strains from phenol-formaldehyde resin manufacturing industrial wastewater. J Hazard Mater 127:238–243 11. Juang RS, Tsai SY (2006) Enhanced biodegradation of mixed phenol and sodium salicylate by Pseudomonas putida in membrane contactors. Water Res 40:3517–3526 12. Bajaj M, Gallert C, Winter J (2009) Phenol degradation kinetics of an aerobic mixed culture. Biochem Eng J 46:205–209

123