There are only a few reports in the literature on enzymatic hydrolysis of vegetable oil or triacylglycerol in supercritical carbon dioxide in a flow-through reactor. Hampson and Foglia [54] measured the extent of conversion of tripalmitin to palmitic acid in SC-CO₂ flowing through a bed of immobilized lipase from Candida antarctica, and showed the importance of enzyme moisture on the activity. Rezaei and Temelli [55,56] studied a Lipozyme-catalyzed hydrolysis of canola oil as a model reaction to develop an on-line extraction-reaction process, where fatty oil is extracted from oilseeds and converted to other valuable products using SC-CO₂. Sovová and Zarevúcka [57] studied the extent of conversion and possible Lipozyme specificity towards fatty acids in hydrolysis of blackcurrant oil with respect to its dependence on the pressure, temperature, carbon dioxide flow rate and moisture content, enzyme load and enzyme distribution in the reactor.

Zarevúcka et al. studied lipase hydrolysis of blackcurrant seed oil, which is rich in α- and γ-linolenic acid, which the human organism is unable to synthesize de novo [58]. Lipozyme, a lipase from Mucor miehei, immobilized on macroporous anionic resin [59,60] was used as a biocatalyst of the reaction. The reaction was performed in a continuous flow reactor at 10–28 MPa and 30–50 °C with carbon dioxide saturated with oil and water (55–100%) flowing up through the enzyme bed. The analysis of a product composition indicated unfavorable hydrodynamics with significant mixing in the reactor when solvent interstitial velocity was lower than 4 mL min⁻¹, while above this velocity value the flow pattern was near to the plug flow. Lipase stability was high, with no activity reduction observed during a long-term experiment. The reaction rate was a function of the ratio of enzyme loaded to a solvent volumetric flow rate. A complete hydrolysis of oil was achieved in the experiments carried out with the enzyme load of 0.8 g and CO₂ flow rate of 0.4–0.9 g min⁻¹. The effects of pressure (10–25 MPa) and temperature (30–40 °C) on the reaction rate were small, and the effects of CO₂ saturation with water and of enzyme distribution in the reactor were negligible. Lipozyme displayed specificity towards linolenic acids; the release of γ-linolenic acid was faster and that of γ-linolenic acid slower than the release of other constituent acids present in blackcurrant oil. The experimental system was also studied and analyzed by means of the HPLC-NMR hyphenated technique [61].

Because lipases were found to be stable in SC-CO₂, a lipase-catalyzed hydrolysis of sunflower oil was performed in two kinds of high-pressure reactors in this medium, a high-pressure batch stirredtank reactor (HP BSTR [62]) and a high-pressure continuous flat-shape membrane reactor (HP CFSMR [63]), in this medium. The reaction rates of enzyme-catalyzed reactions were higher than those of the conventional reactions. Higher conversion was achieved for hydrolysis in the HP BSTR after 48 h, but the maximum conversion in the HP CFSMR was achieved after only 1 h. In both cases the reaction parameters such as pressure, amount of biocatalyst and flow rate of the substrates (in HP CFSMR) had to be optimized.

Primožič et al. [64] studied the thermodynamic and kinetic properties of the immobilized lipase from Aspergillus niger (Lipolase 100T), and used this enzyme for catalysis of hydrolysis of sunflower oil in supercritical carbon dioxide. The optimal concentration of lipase was 0.0714 g/mL of CO₂–free reaction mixture, and the highest conversion of oleic acid (0.193 g/g of oil phase) and linolenic acid (0.586 g/g of oil phase) were obtained at 50 °C, 20 MPa, pH = 7, and an oil/buffer ratio of 1:1 (w/w).

The enzymic hydrolytic reaction in SC-CO₂ as a reaction medium to make glucose from starch was investigated [65]. The reaction rate was enhanced at higher temperature and pressure, especially near the critical point of the CO₂. α-Amylase and glucoamylase were found to be active in SC-CO₂.

The major advantage of immobilized enzymes consisted in an easier separation from the product solution. Reusability studies of immobilized enzyme preparation were performed at atmospheric pressure and in a system CO₂/H₂O at pressure of 10 MPa. The immobilized cellulase was successfully reused for more than 20 times without any significant loss of activity. Thermal stability studies of immobilized and crude cellulase in SC-CO₂ showed that the activity of immobilized cellulase preincubated in SC-CO₂ was much higher than the activity of crude cellulase exposed in SC-CO₂ at the same conditions. The residual activity of pre-incubated on aerogel immobilized cellulase at 110 °C was 172%.

The immobilization of cellulase in silica aerogel matrix has shown to be very efficient as it has improved the biocatalytic properties of cellulase such as activity, stability and its repeated application for hydrolysis of carboxymethyl cellulose [66]. An optimum temperature for the hydrolytic reaction performed at atmospheric pressure was found to be 40 °C with the optimum quantity of 40 g L⁻¹ of the immobilized cellulase. The activity of the immobilized cellulase rose for 110% compared to the activity of native cellulase in the reaction performed at atmospheric pressure. The incubation of the immobilized enzyme in SC-CO₂ (10.0 MPa and 35 °C) prior the hydrolytic reaction improved residual activity up to 461%. The immobilization of cellulase on aerogel matrix also improved the thermal stability of enzyme, as residual activity after incubation at 110 °C in SC-CO₂ was still more than 150%. The same incubation of the native cellulase at 50 °C deactivated the enzyme as the residual activity decreased to 30%. The reason for higher relative enzyme activities in the case of using a highpressure system CO₂/H₂O as a reaction medium was high dispersion of the cellulase in the silica aerogel matrix. The immobilized cellulase could be well reused without any significant loss of activity for at least 15 reaction cycles at atmospheric pressure and at least 20 reaction cycles when a highpressure system CO₂/H₂O was used as a reaction medium.

Proteinases are useful for regioselective hydrolytic transformations. Since papain is one of the few enzymes for organic synthetic transformations that originate from plant sources (the papaya in this case), proteinase from Carica papaya latex was tested for thermal stability at atmospheric pressure, as well as in SC-CO₂ and near-critical propane [42]. The activity of proteinase from Carica papaya at atmospheric pressure increased with the temperature increase from 20 °C to 60 °C. The optimum temperature was determined to be 50 °C. The activity of the enzyme treated with SC-CO₂ (30.0 MPa) also increased with increasing temperature, but the optimal temperature in this case was shifted to 40 °C.

Many examples of biocatalytic synthesis have been documented, though few were optimized in a bioreactor in spite of their importance for large-scale preparation [67–70]. For widespread industrial use to occur, the process has to be technically and economically feasible. Batch vessels are suitable for preliminary screening of enzymic reactions, though, in order to obtain suitable reaction performance data for up-scaling, flow reactors are indicated. Among continuous reactors, packed-bed bioreactors have been extensively investigated for use in industrial scale applications. The packed-bed reactor (PBR) is one of the most commonly employed for solid–fluid contacting in heterogeneous catalysis, because: (i) it facilitates the contact and subsequent separation (ii) and the continuous removal of inhibitory substances; (iii) it allows reuse of the enzyme without the need of a prior separation; (iv) it permits to handle substrates of low solubility by using large volumes containing low concentrations of substrate; (v) it leads to more consistent product quality and improved enzyme stability due to the ease of automation and control [71]; (vi) it is suitable for long-term and industrial-scale production, differently from a stirred-tank reactor where enzyme granules would be susceptible to breakage because of the mechanical shear stress; and (vii) it is more cost effective than the batch operation [72–74]. The ratio between substrate and enzyme is much lower in a PBR than in conventional batch reactors, and it results in higher reaction performance. A stirred tank reactor has to be emptied and refilled at the end of the batch run, leading to downtime and a loss of productivity and, in addition, is subject to suffer from batch-to batch variations. Thus, in a PBR, a purer and more reproducible product and a far greater productivity from a fixed amount of enzyme than the one achieved in the batch process may be expected.

Furthermore, the easily tunable solvating power of SC-CO₂ facilitates a relatively easy separation of reactants, products, and catalysts after reaction, whilst eventual separation of the reaction product from organic solvent is costly and time consuming and, thus, not suitable for industrial production. Furthermore, the high volatility of CO₂ allows it to be completely and easily removed from the product, resulting in an overall solvent-free reaction, required for industrial manufacture, cosmetic and pharmaceutical applications, as n-octyl oleate uses.

At present, the main disadvantage of the use of enzymes in industrial applications is their cost. To overcome this problem, the enzyme is employed in immobilized form because it would allow the reutilization of the enzyme in continuous process [75]. Immobilized enzyme preparations exhibit easy recovering and recycling for subsequent reuse and high biocatalytic density, and offer economic benefits in terms of being able to produce a great quantity of product per unit of enzyme consumed [76] with sufficient operational stability for industrial application.

Capsaicin (8-methyl-N-vanillyl-6-nonenamide) is the main pungent component in capsicum fruits. Its analog, palmitoyl vanillylamide, with similar properties, was synthesized through amidation by lipase in SC-CO₂. [55]. Among five lipases tested, immobilized Mucor miehei lipase, Lipozyme IM, was the most effective in catalyzing a synthesis of this analog. The reaction conditions for the analog synthesis were optimized at 50 °C, 17 MPa and pH 8, and the reaction proceeded for 23 h using vanillylamine hydrochloride and palmitic anhydride at a molar ratio of 5/15 as substrates. The reaction was catalyzed by Lipozyme IM at a concentration of 0.5% (w/w). The residual enzyme activity was about 40% and 15% after a 46- and 69-h repeated amidation reaction in a batch reaction under the optimized conditions, suggesting further modification of conditions was required.

Pyrrole was converted to pyrrole-2-carboxylate in SC-CO₂ using cells of Bacillus megaterium PYR 2910, and the yield of the carboxylation reaction in SC-CO₂ was 12 times higher than that under atmospheric pressure [78]. The cells of Bacillus megaterium PYR2910 [79–82] were employed for the CO₂ fixation reaction. The reaction was conducted by adding CO₂ to 10 MPa to the mixture of pyrrole, the cells, KHCO₃, and NH₄OAc in potassium phosphate buffer. For the reaction at the atmospheric pressure (0.1 MPa), the evolved CO₂ was released to keep the pressure atmospheric. However, the yield was much higher for the reaction proceeding in supercritical CO₂ than at atmospheric pressure. It was also confirmed by a control experiment without the cells that the non-biocatalytic carboxylation of pyrrole did not proceed. Therefore, the cell is surely catalyzing the CO₂ fixation reaction more effectively in SC-CO₂ than at atmospheric pressure.

The activity and stability of a cross-linked enzyme aggregate of Candida antarctica lipase B (CLEA-Calb) in SC-CO₂ have been studied by Dijkstra et al. [83]. The model reaction used is the esterification of isoamyl alcohol with acetic acid. The catalytic performance of CLEA-Calb is evaluated both in batch and in continuous experiments, with a focus on the effect of water production upon reaction. The results of the batch experiments show a decreasing initial activity of the CLEACalb with an increase in pressure. Moreover, CLEAs appear to be highly stable in SC-CO₂ and also remain active during continuous reactions.

Isoamyl acetate was successfully synthesized from isoamyl alcohol in SC-CO₂ by two different immobilized lipases (Novozym 435 from Candica antarctica and Lipozyme RM-IM from Rhizopus miehei) [84]. Among several tested reactants, including acetic acid and two different acetates, acetic anhydride gave best yields. An esterification extent of 100% was obtained in continuous operation using acetic anhydride as acyl donor and Novozym 435 as enzyme.

Long-chain fatty acid esters are useful functional molecules responding to the requirements of numerous fields of application in cosmetic, pharmaceutical and lubricant industry. Lipase-catalyzed production of n-octyl oleate by esterification of oleic acid with 1-octanol in dense CO₂, as reaction medium, was performed in bench-scale packed-bed bioreactor, in order to obtain suitable reaction performance data for up-scaling [85]. Lipase from Rhizomucor miehei (Lipozyme RM IM) was used as the biocatalyst. The experiments were planned to elucidate the effect of several process parameters, such as pressure, temperature, CO₂ and substrates flow rates. Pressure of 10 MPa, temperature of 50 °C, CO₂ flow rate of 210 L h⁻¹ and substrate flow rate of 18 mL h⁻¹ were predicted to be the optimum conditions: a maximum yield of about 93% was achieved. Performing the enzymic reaction in the continuously operating bioreactor, a long-term enzyme lifetime was observed and no decrease of the Lipozyme activity was registered over 50 days. A comparison with the experimental results obtained in a batch-wise mode was also proposed. Operating at the optimum reaction conditions, a higher ester yield than that obtained in batch-mode was detected. SC-CO₂ was shown to be a potential medium for the biosynthesis of n-octyl oleate for a large-scale continuous-mode production.

Cutinase from Fusarium solani pisi was encapsulated in sol-gel matrices prepared with a combination of alkyl-alkoxysilane precursors of different chain-lengths [86]. The specific activity of cutinase in a model transesterification reaction at a fixed water activity in n-hexane was the highest for the precursor combination tetramethoxysilane/n-butyltrimetoxysilane (TMOS/BTMS) in a 1:5 ratio, lower and higher chain lengths of the mono-alkylated precursor or decreasing proportions of the latter relative to TMOS leading to lower enzyme activity. The behavior of the gels in SC-CO₂ paralleled that exhibited in n-hexane, although cutinase activity was one order of magnitude lower (i.e., sol-gel encapsulation did not prevent the deleterious effect of CO₂) The impact that functionalization of some of the additives had on cutinase activity indicates that the enzyme/matrix interactions play an important role. Some of the best additives from the standpoint of enzyme activity were also the best from the standpoint of its operational stability (ca. 80% retention of enzyme activity at the tenth reutilization cycle). None of the additives that proved effective for cutinase could improve the catalytic activity of sol-gel encapsulated Pseudomonas cepacia lipase.

The comparative study on the activity of Fusarium solani pisi cutinase immobilized on zeolites NaA and NaY, in n-hexane, acetonitrile, supercritical ethane (SC-ethane) and SC-CO₂, at two different water activity (a_W) values set by salt hydrate pairs in situ and at acid-base conditions fixed with solidstate buffers of aqueous pK_a between 4.3 and 10.6 was performed [87]. The reaction studied was the transesterification of vinyl butyrate by (R,S)-2-phenyl-l-propanol. The transesterification activity of cutinase was highest and similar in SC-ethane and in n-hexane, about one order of magnitude lower in acetonitrile and even lower in SC-CO₂. Activity coefficients (γ) [88–90] generated for the two substrates indicated that they were better solvated in acetonitrile and thus less available for binding at the active site than in the other three solvents. The γ data also suggested higher reaction rates in SC-ethane than in n-hexane, as observed, and provided an evidence for a direct negative effect of SC-CO₂ on enzyme activity. Manipulation of the acid-base conditions of the media did not afford any improvement of the initial rates of transesterification relative to the blanks (no added acid-base buffer, only salt hydrate pair), except in the case of cutinase immobilized on zeolite NaA in SC-ethane at a_W = 0.7. A poor performance of the blank in this case and a great improvement observed in the presence of a basic buffer suggest a deleterious acidic effect in the medium which, an experiment without additives confirmed, was not due to the known acidic character of the salt hydrate pair used to set a_W = 0.7. In acetonitrile, increasing a_W was accompanied by a decrease in initial rates of transesterification, unlike in the other solvents. There was a considerable hydrolysis in acetonitrile, where initial rates of hydrolysis increased about 20-fold from a_W = 0.2 to 0.7. The hydrolysis was less pronounced in SC-ethane and in n-hexane, and only at a_W = 0.7, and in SC-CO₂ butyric acid was detected only at very long reaction times, in agreement with a generally low catalytic activity. Cutinase enantioselectivity towards the alcohol substrate was low and unaffected by any manipulation of medium conditions.

The catalytic activities of cutinase immobilized on zeolite NaY and Candida antarctica lipase B immobilized on an acrylic resin (Novozym 453) were measured in a model transesterification reaction in three imidazolium cation-based ionic liquids (RTILs), SC-ethane, SC-CO₂ and n-hexane, at a water activity a_W of 0.2 and 0.7 [91]. The transesterification activity of cutinase was the highest and similar in 1-n-butyl-3-methylimidazolium hexafluorophosphate ([C(4)mim][PF₆]), SC-ethane and n-hexane, more than one order of magnitude lower in SC-CO₂, and increased with an increase in a_W. Hydrolysis was not detected in SC-fluids and n-hexane, and was observed in RTILs at a_W 0.7 only. Both initial rates of transesterification and of hydrolysis of Novozym decreased with an increase in a_W. SC-CO₂ did not have a deleterious effect on Novozym activity, which was as high as in SC-ethane and n-hexane. The low reaction rates obtained in this case in RTILs suggested the existence of internal diffusion limitations absent in the cutinase preparation in which the enzyme is only adsorbed at the surface of the support. SC-CO₂ did not adversely affect the catalytic activity of cutinase suspended in [C(4)mim][PF₆], suggesting a protective effect of the RTIL. In the case of Novozym, remarkable increase in the rate of transesterification was obtained in the [C(4)mim][PF₆]/SC-CO₂ system, compared to the RTIL alone. This may reflect improved mass transfer of solutes to the pores of the immobilization matrix due to a high concentration of dissolved CO₂ and a reduction in viscosity of the RTIL. Immobilized Candida antarctica lipase B was used as catalyst to synthesize butyl butanoate from vinyl butanoate and 1-butanol in SC-CO₂ with excellent results [92]. The catalytic behavior of the enzyme immobilized on an acrylic support has been studied tank reactor, showing that a decrease in both the water content and the SC-CO₂ density enhanced the synthetic activity and selectivity.

Polyphenolic compounds, such as caffeic acid, offer many health benefits, but their industrial applications are limited because of their low solubility in water and instability under UV light or heat. In the study of Shin et al. [93] caffeic acid glucosides were produced under enzymic catalysis, in both aqueous buffer and aqueous-supercritical carbon dioxide media, using a recombinant sucrose phosphorylase (SPase) from Bifidobacterium longum. Under SC-CO₂, the amounts of the reaction products from the enzyme reaction were smaller than those in the aqueous reaction medium. However, this is the first report of the transglucosylation of caffeic acid by SPase, and also the first enzymic reaction with phenolic compounds conducted in a SC-CO₂ phase.

The esterification reaction of geraniol with acetic acid catalyzed by Novozym was studied in SC-ethane and in SC-CO₂ [94]. Water activity a_W had a very strong effect on enzyme activity, with reaction rates increasing up to a_W = 0.25 and then decreasing for higher a_W. Salt hydrate pairs could not prevent changes in a_W during the course of reaction but were able to control a_W to some extent and had a beneficial effect on both initial rates of esterification and conversion in SC-ethane. The enzyme was more active in SC-ethane than in SC-CO₂, confirming the deleterious effect of the latter already observed with some enzymes. The temperatures between 40 and 60 °C did not have a strong effect on initial rates of esterification, although reaction progress declined considerably in that temperature range. For the mixture of 50 mM acetic acid plus 200 mM geraniol, a 100% conversion was achieved at a reaction time of 10 h at 40 °C, 10.0 MPa, an a_W of incubation of 0.25, and a Novozym concentration of 0.55 mg mL⁻¹ in SC-ethane. Conversion was below 50% in SC-CO₂ at otherwise identical conditions. With an equimolar mixture of the two substrates (100 mM), 98% conversion was reached at 10 h of reaction in SC-ethane (73% conversion in SC-CO₂).

The monoterpene lavandulol, an significant industrial component of the essential oil of plants of genus Lavandula, has been successfully converted to lavandulyl acetate by enzymatic catalysis in SC-CO₂ using immobilized Candida antarctica lipase B (Novozym 435) [95]. The biotransformation was optimized with respect to substrate concentration, temperature and pressure/density. Conversion of up to 86% was observed at substrate concentration of 60 mM at 60 °C and 10 MPa. Increased temperature of the system resulted in lower enantioselectivity, whereas changes in pressure/density had little effect on this parameter. Immobilized Candida antarctica lipase B (Novozym 435) has been successfully used for number of esterification processes SC-CO₂. Michor et al. showed that Novozym 435 may be used for the acetylation of menthol at temperatures up to 50 °C and pressure 10 MPa [96]. Enzymic esterification of hexanol [97], ibuprofen [98] and geraniol [99] have also been reported using Novozym 435 in SC-CO₂, the last example employing acetic acid as acyl donor.

Among the various lipases, crude hog pancreas lipase is one of the cheapest commercial lipases available. Enzymic reactions using porcine pancreas lipase in SC-CO₂ have been reported for enantioselective esterification of glycidol [100].

The esterification of palmitic acid with ethanol was studied at various temperatures (35 °C–70 °C) in the presence of three lipases (Novozym 435, Lipolase 100T and hog pancreas lipase) in SC-CO₂ and under solvent-free conditions [101]. All enzymes showed an optimum temperature of 55 °C under both conditions. The conversion obtained in SC-CO₂ and under solvent-free conditions with Novozym at optimal conditions was 74 and 97%, respectively. Although a higher apparent yield was obtained under the solvent-free conditions due to higher substrate and enzyme concentrations, the reaction in SC-CO₂ is better because of lower enzyme loading, higher reaction rates and easier downstream processing.

Asymmetric reduction using alcohol dehydrogenases is usually conducted in aqueous media. There are often problems with solubility of substrates. Therefore, asymmetric reduction of various ketones using alcohol dehydrogenases was conducted in SC-CO₂. Whole cells of Geotrichum candidum were used for reduction so that the addition of co-enzymes was avoided [102,103]. The reduction of o-fluoroacetophenone in SC-CO₂ at 10 MPa was conducted using 2-propanol as reductant (hydrogen donor), which afforded (S)-1-(o-fluorophenyl)ethanol in 81% yield after 12 h. Additional substrates were used as substrates, and it was found that all of them were reduced by alcohol dehydrogenase in SC-CO₂ with 2-propanol with a very high enantioselectivity (more then 99%).

Compared to the studies using hydrolytic enzymes in SC-CO₂, dehydrogenase-catalyzed reactions in SC-CO₂ have not yet been developed. Only two reports have been found for biocatalytic reduction in SC-CO₂, and neither of these used flow reactors: asymmetric reduction using Geotrichum candidum [102], and reduction of butyroaldehyde by horse liver alcohol dehydrogenase with a fluorinated coenzyme [104]. The reduction of cyclohexanone was successful, and the biocatalyst was recycled up to four times with only a slight loss in activity. Recycling was not possible using the corresponding batch system because the biocatalysts cannot tolerate repeated depressurization at a very low temperature and separation of the product from the biocatalysts using organic solvents. This method was also applied for the asymmetric reduction of o-fluoroacetophenone, achieving excellent enantioselectivity (ee > 99%) and a higher space-time yield than the corresponding batch process.

The application of SC-CO₂ as a solvent in enzyme-catalyzed reactions has been a matter of considerable research because of its favorable transport properties, which accelerate mass-transferlimited enzymic reactions [105]. The inherent gas-like low viscosities and high diffusivities of supercritical fluids increase the rates of mass transfer of the substrates to the enzyme. Conversely, the liquid-like densities of supercritical fluids result in higher dissolution compared to those observed for gases. Unlike the behavior of gases and liquids, the physical properties of supercritical fluids can be adjusted over a wide range by a relatively small change in pressure or in temperature [106]. Moreover, it might be relevant to stress much lower expenses of solvent in the supercritical medium compared with those of conventional solvents [107]. Another advantage of the use of supercritical fluids (gases) as reaction media is the simple separation of products of the reactions and water in a continuously operated reactor on an industrial scale, due to their different solubility in these media. Since the solvent strength of a supercritical fluid can be varied by changing pressure and temperature, the solubility of substances can easily be regulated. It can be done continuously at the outlet of the reactor, allowing an integrated reaction and separation step, and thus simplifying the downstream processing and recycling of the solvent [107]. Examples of industrially important hydrolytic processes catalyzed by lipases are productions of soaps [108], fatty alcohols [108] and monoacylglycerols [109,110]. Table 1 summarizes the examples of enzymic synthesis in SC-CO₂ from the chapter.