Strategies for the design and operation of enzymatic reactors for the degradation of highly and poorly soluble recalcitrant compounds

1 Strategies for the design and operation of enzymatic reactors for the degradation of highly and poorly soluble recalcitrant compounds GEMMA EIBES, CARMEN LÓPEZ, M.TERESA MOREIRA, GUMERSINDO FEIJOO, & JUAN M. LEMA Dept. of Chemical Engineering, School of Engineering, University of Santiago de Compostela, Santiago de Compostela, Spain 2 Abstract The presence of recalcitrant compounds in both wastewaters and soils is an important environmental problem. The oxidative enzymes from white-rot fungi have been successfully utilised for the in vitro degradation of xenobiotics, such as the azo dye Orange II and the polycyclic aromatic hydrocarbon anthracene (compounds with high and low solubilities, respectively). Two different reactor configurations are proposed: i) an enzymatic membrane reactor for the treatment of soluble compounds consisting of a continuous stirred tank reactor coupled to an ultrafiltration membrane to facilitate the retention and recycling of enzyme; ii) a two-phase enzymatic reactor for the degradation of poorly soluble compounds consisting of an immiscible solvent, which contains the contaminant at high concentrations, and the aqueous phase containing the enzyme and cofactors involved in the catalytic cycle. In this paper, factors affecting the conception, design and operation of both systems are discussed and the experimental results concerning the efficiency and stability of the processes are presented. Keywords: Enzymatic membrane reactors, two-phase partitioning bioreactors, recalcitrant compounds, anthracene, Orange II, manganese peroxidase. Correspondence: Carmen López, Dept. of Chemical Engineering, School of Engineering, University of Santiago de Compostela, E-15782 Santiago de Compostela, Spain. Fax: 34-981-528050. E-mail: [email protected] 3 Features of enzymatic reactors Numerous advantages arise from the use of enzymes against microorganisms for environmental purposes. The advantages are as follows: i) enzymes can be active under a wider variety of conditions such as pH, ionic strength or temperature; ii) higher pollutant concentrations can be maintained in enzymatic reactors with no inhibitory problems; iii) operational times are reduced with no lag period due to microbial growth; iv) media composition is simpler and enzymatic requirements are low provided that the enzyme can be reused; v) control of the process is easier; and vi) no sludge is produced. On the contrary, the cost of enzyme production and its sensitivity to changes in the environmental conditions are the main weaknesses that have to be taken into account when favouring efficiency of the enzymatic process. In the present work, we studied the feasibility of using enzymatic reactors for the degradation of soluble and poorly soluble recalcitrant contaminants. Some typical examples of soluble compounds are nitroaromatic explosives, phenols, dyes, and among them, the azo dye Orange II, a recalcitrant compound commonly present in industrial wastewaters. Typical examples of low-soluble contaminants are polycyclic aromatic hydrocarbons (PAHs), pesticides and polychlorinated biphenyls (PCBs). Their high hydrophobicity makes these compounds environmentally persistent. Anthracene, a tricyclic PAH, was selected as a model compound due to its low solubility. Factors to consider for the operation of enzymatic reactors Selection of the enzyme The enzyme used as a catalyst for the degradation of compounds with recalcitrant character should exhibit high oxidation and ionisation potential and non-specific activity, thereby providing the capability to degrade a broad range of compounds such 4 as those present in contaminated effluents or soils. Because the enzyme-substrate interaction may be confined by the large size of the enzyme, the use of diffusible enzymes or related mediators is desirable. Moreover, extracellular enzymes are preferred since their production is easier and cheaper. All these characteristics are fulfilled by the ligninolytic enzyme referred to as manganese peroxidase (MnP). Catalytic cycle of the enzyme The first step required for successful application is a deep knowledge of enzyme characteristics, particularly the cofactors and cosubstrates that are involved in the catalytic cycle. MnP, an extracellular enzyme, was first discovered in Phanerochaete chrysosporium and is produced by a number of white-rot fungi (Tien and Kirk 1988). MnP catalytic cycle is similar to that of other peroxidases and involves a 2-electron oxidation (Figure 1). The initial oxidation of MnP by H2O2 produces an intermediate compound that promotes the oxidation of Mn+2 to Mn+3. Mn+3 is stabilised by organic acids and the Mn+3-organic acid complex acts as a strong diffusing oxidiser (oxidation potential: 1.54 V) (Kuan and Tien 1993). Excess H2O2 promotes enzyme inactivation. The value of this enzyme is supported by its ability to degrade a variety of complex compounds (Kuan et al. 1993; Martínez 2002). [Insert Figure 1 about here] The factors affecting the catalytic cycle are related to the presence of different cosubstrates and cofactors as well as environmental reaction conditions (Wariishi et al. 1988). These decisions affect not only the stoichiometry of the reaction, but also the kinetics and stability of the enzyme. Though the stoichiometry describes the minimum requirements of the cofactors, the proper selection of their concentrations and operational variables may accelerate reaction kinetics and reduce enzymatic 5 consumption. Two main goals have to be achieved: i) adequate conversion, which may be defined by environmental restrictions; and ii) high efficiency, defined as the amount of substrate degraded per unit of enzyme consumed during the reaction. H2O2 is directly involved both in the activation of the catalytic cycle at balanced concentrations and the deactivation of the enzyme at high concentrations (Timofeevski et al. 1998). A low concentration of H2O2 would imply kinetic or even stoichiometric limitations while the enzymatic activity would be protected towards inactivation (Mielgo et al. 2003a). On the other hand, organic acids present a similar crossed effect: a high concentration favours chelation of Mn3+ although it may affect the stability of the enzyme (Hofrichter et al. 1998), likely due to the endogenous formation of H2O2 during decarboxylation (Schlosser and Hofer 2002; Van Aken and Agathos 2002). Operational parameters With respect to the operational parameters affecting the continuous process, the hydraulic retention time (HRT), loading rate and environmental parameters (temperature, pH) were the most significant. With HRT, we should be aware of the desired conversion (degradation) of pollutant and kinetics, which is directly affected by the pollutant concentration. Thus, a higher loading rate could lead to a faster but a less efficient process. Increasing temperature could favour, in some cases, the reaction rate, but it may also increase enzyme deactivation. An appropriate temperature must be selected when trying to find a compromise between both aspects, as well as when considering economical factors. Operational pH must also be controlled at values near the optimal range. In the case of MnP, pHs lower than 4 and higher than 6 should be avoided. 6 Modelling and control of enzymatic reactors When changes in the influent flow or the pollutant concentration occur, one of the main aspects to consider for proper continuous operation of an enzymatic reactor is the system dynamics. The conception, design and operation of an efficient control system require a deep knowledge of the kinetic model, which depends on the enzyme. In the case of MnP, the substrate to be oxidised (pollutant) is not the primary substrate for the enzyme (H2O2), although the degradation rate is greatly influenced by the concentration of the substrate and, thus, this parameter must be considered as a variable (López et al. 2006). Furthermore, the study of reactor behaviour in unsteady-state conditions will be indicative of system stability as it depicts system resistance against alterations. Retention of the enzyme To minimise losses, the retention of enzyme in the reactor is a main factor when dealing with the design and operation of an enzymatic reactor. When the pollutant is highly soluble in water, the recovery of the enzyme from the effluent requires an additional system, whereas for low solubility compounds, the enzyme is trapped onto the aqueous phase. Below, we discuss different strategies depending on the solubility of the pollutants. Reactors treating highly soluble compounds. The configurations of enzymatic reactors treating soluble compounds can be classified according to the manner in which the enzyme is retained: i) immobilised onto a support, forming bigger structures that can be retained due to their size or ii) free in solution, being retained by a membrane. Immobilisation of the enzyme onto a support is usually complex and expensive, and increases processing costs. To improve the economical feasibility of immobilised enzyme reactors, a number of requirements should be met: the specific activity of the 7 derivative (units of enzyme per g of support) should be as high as possible; the support or membrane could be applied with a secondary function, such as the separation of substrates or products; and the support should have good mechanical resistance and minimum interaction with the substrates or products. Previous studies have determined a support based on agarose activated with glutaraldehyde groups as suitable for the immobilisation of MnP (Mielgo et al. 2003b). The immobilised enzyme was applied for the degradation of the dye Orange II in a continuous stirred tank reactor. However, there were some operational difficulties in this configuration with the free enzyme: high activity loss during the immobilisation process, lower reaction rates, more complex control, adsorption of the dye onto the support and, mainly, the ability to replace the deactivated immobilised enzyme with fresh enzyme. The second option corresponds to a continuous stirred tank reactor where the soluble enzyme is retained by means of an ultrafiltration membrane coupled to the reactor (Figure 2A) (López et al. 2004a). The main advantages of this configuration are: i) operation with free enzyme, avoiding limitations of mass transfer and, consequently, low kinetic rates; ii) retention of non-biodegradable molecules with high molecular weights; iii) ability of the products of degradation to cross the membrane, being discharged in the effluent; and iv) easy operation. Reactors treating poorly-soluble compounds. The in vitro degradation of poorly soluble compounds is limited by their enzyme availability. The addition of miscible cosolvents is a good approach to increase the solubility by several orders of magnitude. Eibes et al. (2005) established a system for the degradation of anthracene by MnP in a medium containing a mixture of acetone:water (36% v:v). In the mentioned work, 5 mg L-1 of anthracene were degraded after 6 h of operation under optimal conditions. However, 8 this system presented a number of limitations: i) the concentration of anthracene in the medium was limited by the amount of cosolvent; ii) higher amounts of solvent could lead to a higher inactivation of the enzyme; iii) the enzyme could not be recycled; and iv) the reuse of the solvent would require a separation process. To overcome these limitations, a completely different option, a two-phase partitioning bioreactor (TPPB), was considered (Figure 2B). In TPPBs, a solvent immiscible in the water phase with an appropriate partition coefficient is added. The substrate is almost completely dissolved in the organic phase, from which it diffuses to restore equilibrium, as the enzyme degrades the pollutant in the aqueous phase (Vrionis et al. 2002). TPPBs have been successfully used for the biological treatment of toxic and recalcitrant pollutants, such as polycyclic aromatic hydrocarbons. Janikowski et al. (2002) performed the degradation of anthracene and other PAHs in biphasic reactors in cultures of Sphingomonas aromaticivorans and dodecane as the organic phase. [Insert Figure 2 about here] The selection of the appropriate solvent is critical for the design of TPPBs because it greatly influences mass transfer and consequently degradation rates. The selected solvent should be inexpensive, readily available, and exhibit suitable physical and chemical properties (be immiscible, non-volatile, etc.) (Déziel et al. 1999; MacLeod and Daugulis 2003; Marcoux et al. 2000; Villemur et al. 2000). Furthermore, when dealing with enzymatic reactors, the possible interaction between the solvent and enzyme is critical. It is important that the solvent is not a substrate of the enzyme (MacLeod and Daugulis 2003) and its effect on enzymatic activity is as low as possible (Ross et al. 2000). The partition coefficient should enable the system to achieve the highest possible concentration of substrate in the aqueous phase. It has been stated that 9 solvents with high partition coefficients can sequester the target compound, thus limiting its biodegradation rate (Efroymson and Alexander 1995). The substrate transfer rate from the water-immiscible to the aqueous phase is another essential factor and has to be enhanced so as not to limit the overall degradation rate. Mass transfer is favoured by an increased surface area for partitioning, therefore, the rate of biodegradation in a TPPB is governed by the size of the interface between the two liquid phases (Ascón-Cabrera and Lebeault 1995; Köhler et al. 1994). The interfacial area is defined by equation 1: 6· ϕ = sm ad (1) where φ is the proportion of the organic phase in the reactor and dsm is the Sauter mean diameter of the solvent drops. Therefore, increasing the proportion of organic solvent or decreasing the diameter of the drops by increasing the agitation speed would augment the interfacial area. In vitro degradation of Orange II in an enzymatic membrane reactor The operation of a continuous enzymatic membrane reactor for the degradation of soluble compounds must consider the following steps: i) optimisation of the parameters involved in the catalytic cycle of the enzyme; ii) optimisation of the operational parameters in the enzymatic membrane reactor and iii) development of a control system. Optimisation of the parameters involved in the catalytic cycle The study of the effect of different variables involved in the MnP catalytic cycle on the degradation of the azo dye, Orange II, was performed in discontinuous experiments (Mielgo et al. 2003a). The authors studied the effect of the type and concentration of organic acid and concentrations of Mn2+, H2O2 and MnP, as well as pH and temperature 16 The equation, which describes the behaviour of anthracene in the organic phase (SS) considering the mass transfer and the degradation kinetics (as a pseudo-first order (Eibes et al. 2006b)), is shown in equation 5: 0 ln ln · · ·  = −  +  cat w L SS sw L cat S kV ka SS t k ka k V (5) The partition coefficient of anthracene in silicone oil (ksw) had been previously determined (ksw=5012, Table III) and the mass transfer coefficient (kLa) was correlated with the operational parameters, as described above. Finally the kinetic constant kcat was determined from batch experiments. The model was validated by comparing experimental data with model predictions in experiments at different agitation speeds and volumes of silicone oil (Figure 7). As presented previously, the highest degradation rate was obtained at 300 rpm and 30% silicone oil (v:v), oxidising 90% of anthracene present in the organic phase after 56 h. Conclusions and perspectives Enzymatic membrane reactors are a promising technology, as they are easy to operate and control. Degradation processes can be conducted continuously for more than 20 days without membrane cleaning or replacement. Furthermore, enzymatic reactors are versatile since there are a wide variety of commercially available membrane shapes, materials and modules. The selection process must primarily consider the characteristics of the effluent and the enzymatic molecular weight. The use of a two-phase partitioning bioreactor, which utilises a second immiscible phase for enzymatic degradation of poorly-soluble compounds, is advantageous because of simpler operation due to easy recovery of the solvent depleted in substrate and its reuse in subsequent experiments. Although a priori mass transfer 17 could be considered a limitation of this system, the selection of the appropriate solvent, as well as the study of the conditions resulting in the maximum efficiency enables us to obtain unprecedented degradation rates in enzymatic reactors. The challenges for those processes are now focused on the development and implementation of a control system and further scale-up for application on an industrial scale. Acknowledgments This work was funded by the Spanish Commission of Science and Technology (CICYT) (Project PPQ2001-3063) and by Xunta de Galicia (PGIDT02PXIC20905PN). Supports given to Carmen López from the Spanish Ministry of Education, Culture and Sport (AP2000-1712) and Gemma Eibes from the Spanish Ministry of Science and Technology (BES-2002-2809) are also greatly appreciated. 18 References Ascón-Cabrera MA, Lebeault JM. 1995. Interfacial area effects of a biphasic aqueous/organic system on growth kinetic of xenobiotic-degrading microorganisms. Appl Microbiol Biotechnol 43:1136-1141. Déziel E, Comeau Y, Villemur R. 1999. Two-liquid-phase bioreactors for enhanced degradation of hydrophobic/toxic compounds. Biodegradation 10:219-233. Efroymson RA, Alexander M. 1995. Reduced mineralization of low concentrations of phenanthrene because of sequestering in nonaqueous-phase liquids. Environ Sci Technol 29:515-521. Eibes G, Cajthaml T, Moreira MT, Feijoo G, Lema JM. 2006b. Enzymatic degradation of anthracene, dibenzothiophene and pyrene by manganese peroxidase in media containing acetone. Chemosphere 64:408-414.. Eibes G, Lu Chau T, Feijoo G, Moreira MT, Lema JM. 2005. Complete degradation of anthracene by Manganese Peroxidase in organic solvent mixtures. Enzyme Microb Technol 37(4):365-372. Eibes G, Moreira MT, Feijoo G, Daugulis AL, Lema JM. 2006a. Operation of a two phase partitioning bioreactor for the degradation of anthracene by the enzyme manganese peroxidase. Chemosphere (in press). Hofrichter M, Ziegenhagen D, Vares T, Friedrich M, Jäger MG, Fritsche W. 1998. Oxidative decomposition of malonic acid as basis for the action of manganese peroxidase in the absence of hydrogen peroxide. FEBS Lett 434:362-366. Janikowski TB, Velicogna D, Punt M, Daugulis AJ. 2002. Use of a two-phase partitioning bioreactor for degrading polycyclic aromatic hydrocarbons by a Sphingomonas sp. Appl Microbiol Biotechnol 59:368-376. Köhler A, Schüttoff M, Bryniok D, Knackmuβ HJ. 1994. Enhanced biodegradation of phenanthrene in a biphasic culture system. Biodegradation 5:93-103. 19 Kuan IC, Johnson KA, Tien M. 1993. Kinetic analysis of manganese peroxidase. The reaction with manganese complex. J Biol Chem 268:20064-20070. Kuan IC, Tien M. 1993. Stimulation of manganese peroxidase activity: a posible role for oxalate in lignin biodegradation. Proc Natl Acad Sci USA 90:1242-1246. López C, Moreira MT, Feijoo G, Lema JM. 2004a. Dye decolorization by manganese peroxidase in an enzymatic membrane bioreactor. Biotechnol Prog 20(1):74-81. López C, Moreira MT, Feijoo G, Lema JM. 2006. Dynamic model of an enzymatic membrane reactor treating xenobiotic compounds. Biotechnol Bioeng (submitted). López C, Valade AG, Combourieu B, Mielgo I, Bouchon B, Lema JM. 2004b. Mechanism of enzymatic degradation of the azo dye Orange II determined by ex situ 1H nuclear magnetic resonance and electrospray ionization-ion trap mass spectrometry. Anal Biochem 335:135-149. MacLeod CT, Daugulis AJ. 2003. Biodegradation of polycyclic aromatic hydrocarbons in a two-phase partitioning bioreactor in the presence of a bioavailable solvent. Appl Microbiol Biotechnol 62:291-296. Marcoux J, Déziel E, Villemur R, Lépine F, Bisaillon JG, Beaudet R. 2000. Optimization of high-molecular-weight polycyclic aromatic hydrocarbons' degradation in a two-liquid-phase bioreactor. J App Microbiol 88(4):655-662. Martínez AT. 2002. Molecular biology and structure-function of lignin-degrading heme peroxidases. Enzyme Microb Technol 30(4):425-444. Martínez MJ, Böckle B, Camarero S, Guillén F, Martínez AT. MnP isoenzymes produced by two Pleurotus species in liquid culture and during wheat-straw solid-state fermentation. In: Series AS, editor; 1996; New Orleans, LA. p 183196. 20 Mielgo I, López C, Moreira MT, Feijoo G, Lema JM. 2003a. Oxidative degradation of azo dyes by manganese peroxidase under optimized conditions. Biotechnol Prog 19(2):325-331. Mielgo I, Palma C, Guisán JM, Fernández-Lafuente R, Moreira MT, Feijoo G, Lema JM. 2003b. Covalent immobilisation of manganese peroxidases (MnP) from Phanerochaete chrysosporium and Bjerkandera sp. BOS55. Enzyme Microb Technol 32:769-775. Moreira MT, Palma C, Mielgo I, Feijoo G, Lema JM. 2001. In vitro degradation of a polymeric dye (Poly R-478) by manganese peroxidase. Biotechnol Bioeng 75(3):362-368. Ross AC, Bell G, Halling PJ. 2000. Organic solvent functional group effect on enzyme inactivation by the interfacial mechanism. J Mol Catal B: Enzym 8:183-192. Schlosser D, Hofer C. 2002. Laccase-catalyzed oxidation of Mn+2 in the presence of natural Mn+3 chelators as a novel source of extracellular H2O2 production and its impact on manganese peroxidase. Appl Environ Microbiol 68(7):3514-3521. Tien M, Kirk TK. 1988. Lignin peroxidase of Phanerochaete chrysosporium. Methods Enzymol 161:238-249. Timofeevski SL, Reading NS, Aust SD. 1998. Mechanisms for protection against inactivation of manganese peroxidase by hydrogen peroxide. Arch Biochem Biophys 356(2):287-295. Van Aken B, Agathos SN. 2002. Implication of manganese (III), oxalate, and oxygen in the degradation of nitroaromatic compounds by manganese peroxidase (MnP). Appl Microbiol Biotechnol 58(3):345-351. Villemur R, Deziel E, Benachenhou A, Marcoux J, Gauthier E, Lepine F, Beaudet R, Comeau Y. 2000. Two-liquid-phase slurry bioreactors to enhance the degradation of high-molecular-weight polycyclic aromatic hydrocarbons in soil. Biotechnol Prog 16(6):966-972. 21 Vrionis HA, Kropinski AM, Daugulis AJ. 2002. Enhancement of a two-phase partitioning bioreactor system by modification of the microbial catalyst: demonstration of concept. Biotechnol Bioeng 79(6):587-594. Wariishi H, Akaleswaran L, Gold MH. 1988. Manganese peroxidase from the basidiomycete Phanerochaete chrysosporium: spectral characterization of oxidized states and the catalytic cycle. Biochemistry 27:5365-5370. 22 Table I. Optimisation of reaction conditions for the discontinuous degradation of Orange II by MnP. V=25 mL; initial Orange II concentration=100 mg L-1. Parameter Range Optimal value Organic acid [acid] [Mn+2] H2O2 addition [H2O2] [MnP] pH T Oxal-Malon-Acet 1-50 mM 0-1000 µM FB-Continuous 50-200 µmol L-1 min-1 50-200 U L-1 4.5-8 20-30ºC Acetate 1 mM 33 µM Continuous 50 µmol L-1 min-1 200 U L-1 4.5 30ºC 23 Table II. Experimental conditions of assays E1-E4 in Figure 3. Experiment MnP (U L-1) H2O2 addition rate (µmol L-1 min-1) Organic acid HRT (min) E1 E2 E3 E4 - - 200 200 50 50 - 15 Acetic Acetic Acetic Oxalic 60 20 45 - 24 Table III. Log KSW of 15 different solvents. Solvent log Ksw Solvent log Ksw Silicone oil 3.7 Triacetin 4.8 Paraffin oil 4.3 Olive oil 4.9 Sunflower oil 4.3 Corn oil 4.9 Oleic alcohol 4.4 Ethyl acetate 5.0 Decanol 4.4 Biodiesel 5.0 n-Hexadecane 4.5 Marc olive oil 5.0 Dodecane 4.5 Undecanone 5.2 Engine oil 4.6 25 Table IV. Values of kLa obtained for experiments at different agitation rates and volumes of silicone oil. Agitation speed (rpm) kLa (min-1) 10% silicone oil 20% silicone oil 30% silicone oil 50 0.01 0.02 0.12 150 0.10 0.36 0.36 200 0.27 0.30 0.68 250 2.99 2.26 3.14 350 3.29 3.30 3.30 32 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 1 5 15 25 10 33 50 66 33 33 10 5 Degradation of anthracene (mg L-1h-1) 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 Efficiency (mg U -1) Figure 6 H 2 O 2 (μmol L-1min-1) Malonate (mM) Malonate (mM) control pH * E1 E2 E3 33 0 25 50 75 100 125 012 24 36 48 60 72 84 Time (h) Anthracene degradation (mg L -1 ) Figure 7