PDFlib PLOP: PDF Linearization, Optimization, Protection Page inserted by evaluation version www.pdflib.com –
[email protected]
The Enzymatic Reaction-Fractionation Process in Supercritical Carbon Dioxide M A I N MARTY, SOPHIE MANON, DONG PYO JU, DIDIER COMBES, AND JEAN-ST~PHANE CONDOREF Dkpartement de Ginie Biochimique et Alimentaire URA CNRS 544 Institut National des Sciences Appliquies Complm Scientifique de Rangueil 31077 Toulouse Cedex, France Above a critical pressure (Pc) and a critical temperature (Tc), pure components are no longer gaseous or liquid. This other state is called a supercritical fluid. Because of their high density, leading to good solvent abilities, Hammond et al.,* in 1985, were the first to succeed in the use of supercritical fluids (SCF) used as solvents for enzymatic reactions. Besides, the low viscosity and high diffusion coefficient of these fluids lead to good transfer properties, which is of great interest when immobilized catalysts are used. Supercritical carbon dioxide (SCC02) is the most commonly used fluid because it is nontoxic, nonflammable, and cheap. Last but not least, it exhibits mild critical coordinates (Pc = 7,4 MPa, Tc = 31"C), compatible with biological materials. A brief overview of the literature in this field2essentially demonstrates that there is no conclusive evidence of the superiority of the supercritical carbon dioxide solvent upon organic solvents from the point of view of enzymatic stability and kinetics. For instance, in our laboratory, we have studied the esterification reaction between oleic acid and ethan01,~and the transesterification reaction between geraniol and propylacetate: both catalyzed with an immobilized lipase, the Lypozyme@@ in SCCO2 (130 bars, 40°C), and in n-hexane (40°C). The stability is high and similar in both solvents: 90% of residual activity after 6 days. We have observed an equal sensitivity with respect to the water content adsorbed on the enzymatic support (optimum activity at 10% g/g of support), even if the water solubility is different in each medium. Finally, the same enzymatic mechanism, Ping-Pong-Bi-Bi with inhibition by one substrate, the alcohol, and a similar catalyst activity were demonstrated in both solvents. An important exception of the similarity of SCF and organic solvents concerns reactions of oxidations that take advantage of the total miscibility of oxygen in supercritical fluids. In addition, there is not the slightest doubt that the natural and the nontoxic nature of the SCC02 is of great interest, particularly appreciated in the food and health industries. According to us, an actual development of the supercritical technology applied in catalysis would lie in the integration of downstream processing. Indeed, a great advantage of SCF upon a liquid solvent is the variability of its solvent ability versus temperature and pressure. Operating a cascade of depressurizations at the outflow "Towhom all correspondence should be addressed. 408
409
MARTY et nl.: SUPERCRITICAL CARBON DIOXIDE
Comparison of Ester Recovery in SCCO2 and in n-Hexane (after hypothetic n-hexane elimination)
TABLE 1.
SCCOZ n-Hexane
Ester Recovery Yield (%) 84.6 100
Ester Concentration (g/L) 772 380
Ester Molar Purity (%) 71.5 (ester/acid: 6.1) 14 (esterlacid: 1.3)
of a continuous reaction vessel, combined with temperature variations, allows a decrease, step by step, of the fluid density. Consequently, solvent capacity of the fluid decreases, leading to the fractionation of products and residual substrates. Marty er aL6 have demonstrated the feasibility of this coupling reaction-fractionation during the continuous esterification of oleic acid by ethanol. TABLE1 emphasizes the interest of SCC02compared with n-hexane for postreactional fractionation. With a conversion rate of oleic acid of about 60%, 85% of the produced ethyl oleate is recovered with a molar purity of 71.5%. The same process operated in n-hexane would lead, after complete elimination of the solvent, to a molar purity of only 15%. In addition to these quantitative advantages, we are reminded that residual traces of hexane could become a drawback, in view of obtention of a “clean” product. Besides, at the outflow of the last separation vessel in SCCO2, after repressurization, the pure solvent is recycled, which allows the use of low quantities of solvent. In the case of an enzymatic reaction, where the conversion is strongly limited by the thermodynamic equilibrium, one may want to take advantage of the ease of postreactional fractionation offered by the supercritical solvent, by extracting selectively one component in order to shift the equilibrium. In this case when considering the equilibrated reaction A + B *-, C + D, one can favor the synthesis of C and D by extracting selectively C and/or D (the ideal case being to extract the aimed product) from the reaction medium. The simplest system using pressure variation as the 1. separating agent is schematized on FIGURE
+
Reaction
vessel
-
FIGURE 1. Experimental device for the coupling reaction-fractionation. The equilibrated C + D. Tr and Ts are the temperatures of the reaction and separation reaction is A + B vessel, respectively. Pr and Ps are the pressures of the reaction and separation vessel, respectively.
ANNALS NEW YORK ACADEMY OF SCIENCES
410
This “extractive reaction” process with the supercritical fluid was first proposed by Doddema el al.’ who performed continuous extractive transesterification. In their case, ethanol, an inhibitor of the reaction, was extracted by operating a complex process involving two counter-current columns and two enzymatic reaction vessels. Despite its complexity, this process resulted in a continuous loss of one substrate that had to be continuously fed in the second reactor. Another attempt was made by Adshiri et aL8 for interesterification of tricaprylin, where the degree of incorporation of oleic acid was improved by the selective removal of one product in the course of a batch reaction. They operated a counter current column with temperature gradient. 100.0
I
geranylacetate
0.1 70
75
80
85
90
95
100 105 110 115 120
Pressure (bars) FIGURE 2. Selectivity of propylacetate and geranylacetate versus geraniol as a function of pressure and temperature of SCCO2.
In our study, we attempted to operate the simple system shown in FIGURE1 for batch transesterification: geraniol propylacetate c* geranylacetate propanol. This reaction had been previously studied in our laboratory4 and had been shown to be strongly limited by a thermodynamic equilibrium when SCCO2 was used as a solvent (30% maximum conversion). The logical approach is to investigate first the performance of separation obtained with a single depressurization step, as described on FIGURE 1, the goal being to selectively condense the geranylacetate all along its production to recycle produced propanol and unreacted substrates. This was done by using synthetic mixtures of geraniol, geranylacetate, and propylacetate (equimolar,
+
+
MARTY et d.: SUPERCRITICAL CARBON DIOXIDE
411
15 mM). The fluid phase that enters the separator and the condensed phase were assayed by HPLC, and data were processed by using the concept of selectivity (Sx) with respect to the desired component, that is, geraniol. The definition of this parameter is
where X represents geranylacatate or propylacetate. Go1 represents geraniol. [ISCCO, and [ILlq,are the concentrations in the supercritical and liquid phases, respectively. Results are given on FIGURE2. The concept of selectivity enables us to state that the further from 1 the value of this selectivity of one component is, the easier the separation of this component from geraniol will be. Values of selectivities for propylacetate, from FIGURE 2, are always very close to 1 and indicate a very difficult separation that condemns a possible success of the proposed process. Similar results were obtained by Doddema et al.,’ who found a selectivity for nonylacetate, with respect to nonanol, very close to 1. This explains the complexity of the process, briefly described in the previous paragraph, that they had to operate. When data can be obtained, upon selectivity, they are a prerequisite for assessing the validity of the concept of the extractive reaction in SCCO2. In our case, these preliminary results question the necessity of developing complex procedures for overcoming the lack of selectivity of a simple depressurization step. Work is in progress in our laboratory to investigate the system fulfilling this last criterium. REFERENCES
5.
6. 7.
8.
HAMMOND, D. A., M. KAREL,A. M. KLIBANOV & V. J. KRUKONIS.1985. Enzymatic reactions in supercritical gases. Appl. Biochem. Biotechnol. 11: 393-400. 0.& M. RANTAKYL~. 1991. Biocatalysis in supercritical COz. Chemtech, April: AALTONEN, 24&248. MARTY, A,, W.CHULALAKSANANUKUL, R. M. WILLEMOT & J. S. CONDORET. 1992. Kinetics of lipase-catalyzed esterification in supercritical COz. Biotechnol. Bioeng. 3 9 273-280. CHULALAKSANANUKUL, W.,J. S. CONDORET & D. COMBES.1993. Kinetics of the lipasecatalyzed transesterification of geraniol in supercritical COz. Enzyme Microb. Techno]. 15: 691697. RANDOLPH, T. W.,H. W.BLANCH & J. M. PRAUSNITZ. 1988. Enzyme-catalyzed oxidation of cholesterol in supercritical carbon dioxide. Am. Inst. Chem. Eng. J. 3 6 1354-1360. MARTY,A., W. CHULALAKSANANUKUL, D. COMBES& J. S. CONDORET. 1994. Continuous reaction-separation process for enzymatic esterification in supercritical carbon dioxide. Biotechnol. Bioeng. 43: 497-504. H. J., R. J. J. JANSSENS, J. P. J. DE JONG,VAN DER LUGT,J. P. & H. H. M. DODDEMA, OOSTROM.1990. Enzymatic reactions in supercritical carbon dioxide and integrated product-recovery. 5th European Congress on Biotechnology, Copenhagen. Christiansen et al., Ed.: 239-242. A D S H I R I , T., H. AKIYA, L. C. CHIN, K. ARAl & K. FUJtMOTO. 1992. Lipase-catalyzed interesterification of triglycerides with supercritical carbon dioxide. J. Chem. Eng. Jpn 25: 104-105.
