Department of Materials Science, Faculty of Engineering, Tottori University, Tottori, Japan
The use of ionic liquids (ILs) to replace organic or aqueous solvents in biocatalysis processes has recently gained much attention and great progress has been accomplished in this area; lipase-catalyzed reactions in an IL solvent system have now been established and several examples of biotransformation in this novel reaction medium have also been reported. Recent developments in the application of ILs as solvents in enzymatic reactions are reviewed.
Ionic liquids (ILs) have very good properties as reaction medium in chemical reactions: they are non-volatile, non-flammable, have low toxicity and good solubility for many organic and inorganic materials.1 It has long been recognized that an enzymatic reaction proceeds in an aqueous buffer solution under appropriate pH conditions and an enzyme quickly loses its activity in a highly concentrated aqueous salt solution.2 Therefore it seems a foolish notion that enzymatic reaction occurs in a salt medium from the standpoint of biology. However, the use of ILs to replace traditional organic solvents in chemical reactions has recently gained much attention, and even as a novel reaction medium for biotransformation. Lipase-catalyzed reactions in an IL solvent system have now been established,113-6 and several types of non-lipase enzymatic reactions have also been reported recently. I wish to review recent progress in the area of "enzymatic reactions in an IL solvent system" in this chapter.
2. IONIC LIQUIDS AS A REACTION MEDIUM FOR BIOTRANSFORMATION
Cull and co-authors3a reported a microbe-mediated transformation of benza-mide from benzonitrile in a mixed solvent of IL, 1-butyl-3-methylimidazolium
CAL-B R NH2
Figure 1: The first enzymatic reaction conducted in a pure ionic liquid solvent system.
hexafluorophosphate ([bmim][PF6]), with water (1:4) in July 2000. Then Russell and co-authors3b reported that thermolysin-catalyzed amidation of CBz-asparagine with L-phenylalanine methyl ester proceeded in a mixed solvent of [bmim][PF6] with aqueous buffer solution. These examples showed that the IL had no inhibitory action against the enzymes because [bmim][PF6] was insoluble in water and enzymatic reactions took place in the water layer. The first example of enzymatic reaction in a pure IL solvent system was reported by the Sheldon group in December 2000.4 The authors successfully demonstrated two types of Candida antarctica lipase (CAL-B) catalyzed reaction in a pure IL: CAL-B catalyzed amidation of octanoic acid with ammonia and also the formation of octanoic peracid by the reaction of octanoic acid with hydrogen peroxide (Fig. 1).
However, the reactions were not enantioselective ones, though the most important aspect of the biocatalysis reaction should be in the enantioselective reaction. We5a and Kragl6 independently reported the first enantioselective lipase-catalyzed reaction in February-March 2001. Since lipase was anchored by the IL solvent and remained in it after the extraction work-up of the product, we succeeded in demonstrating that recyclable use of the lipase in the [bmim][PF6] solvent system was possible (Fig. 2).5a
Typically the reaction was carried out as follows: to a mixture of lipase in the IL were added this racemic alcohol and vinyl acetate as the acyl donor. The resulting mixture was stirred at 35°C and the reaction course was monitored by GC analysis. After the reaction, ether was added to the reaction mixture to form a biphasic layer, and product acetate and unreacted alcohol were extracted with ether quantitatively. The enzyme remained in the IL phase as expected (Fig. 2). Two months later, Kim and co-workers7a reported similar results and Lozano and Ibora7bd reported other examples of lipase-catalyzed reaction in June. Further Park and Kazlauskas7c reported full details of lipase-catalyzed reaction in an IL solvent
system in August 2001. Studies on the enzymatic reaction in an IL solvent system were thus launched in 2000-2001.
We initially tested Candida antarctica lipase using imidazolium salt as solvent because CAL was found to be the best enzyme to resolve our model substrate 5-phenyl-1-penten-3-ol (1a); the acylation rate was strongly dependent on the anionic part of the solvents. The best results were recorded when [bmim][BF4] was employed as the solvent, and the reaction rate was nearly equal to that of the reference reaction in diisopropyl ether. The second choice of solvent was [bmim][PF6]. On the contrary, a significant drop in the reaction rate was obtained when the reaction was carried out in TFA salt or OTf salt. From these results, we concluded that BF4 salt and PF6 salt were suitable solvents for the present lipase-catalyzed reaction.5a Acylation of 1a was accomplished by these four enzymes: Candida antarctica lipase, lipase QL from Alcaligenes, Lipase PS from Burkholderia cepacia and Candida rugosa lipase. In contrast, no reaction took place when PPL or PLE was used as catalyst in this solvent system. These results were established in March 2000 but we encountered a serious problem in that the results were significantly dependent on the lot of the ILs that we prepared ourselves. The problem was very serious because sometimes the reaction did not proceed at all. So we attempted to purify the ILs and established a very successful procedure (Fig. 3): the salt was first washed with a mixed solvent of hexane and ethyl acetate (2:1 or 4:1), treated with activated charcoal and passed into activated alumina neutral type I as an acetone solution. It was evaporated and dried under reduced
Me' LiTFSI LiCI
Stirred at RT for 6 h, ■*A then filtered through a celite pad and washed with acetone three times
The filtrate was evaporated and the residue was washed with a mixed solvent of hexane and ethyl acetate (2:1 or 4:1)
The filtrate was evaporated and the residue was dissolved in acetone and treated with charcoal and filtered through a Al203 (neutral type I, activated) short column
Evaporated, then dried under vacuo at 66.7 Pa for 24 h at 50°C
Figure 3: Purification protocol of imidazolium ionic liquid.
pressure at 50°C for 24 h to obtain very clean imidazolium salt. Using the ILs, we succeeded in obtaining more reproducible results; I recommend this as also being very useful to recycle the IL. In fact, we always recycle our ILs after the reaction and have not wasted any in the past. We are still using ILs that have a 7-year history. After establishing the reproducibility of our results of lipase-catalyzed reaction, we submitted our first paper in December 2000 and the paper was accepted on January 5, 2001. Although we lost the chance to be the first to publish in the field for this reason, we learned many things about ILs during that time and these are now important bases of our research group. Very pure ILs are commercially available now and we can use them, but I imagine that all research groups encountered the same problem in the early days of this field, because very clean ILs are required for a biocatalysis system compared to chemical reactions. This story highlights a very important point; we should pay attention to the quality of the IL when we evaluate the appropriate one for our desired biocatalyst reaction.
3. LIPASE-CATALYZED REACTION IN AN IONIC LIQUID SOLVENT SYSTEM
We succeeded in showing that recycling of the enzyme was indeed possible in our IL solvent system, though the reaction rate gradually dropped with repetition of the reaction process.5a Since vinyl acetate was used as acyl donor, acetaldehyde was produced by the lipase-catalyzed transesterification. It is well known that acetaldehyde acts as an inhibitor of enzymes because it forms a Schiff base with amino residue in the enzyme. However, due to the very volatile nature of acetaldehyde, it easily escapes from the reaction mixture and therefore has no inhibitory action on the lipase. However, this drop in reactivity was assumed to be caused by the inhibitory action of acetaldehyde oligomer which had accumulated in the [bmim][PF6] solvent system. In fact, it was confirmed that the reaction was inhibited by addition of acetaldehyde trimer.5c
One of the most important characteristics of IL is its wide temperature range for the liquid phase with no vapor pressure, so next we tested the lipase-catalyzed reaction under reduced pressure. It is known that usual methyl esters are not suitable for lipase-catalyzed transesterification as acyl donors because reverse reaction with produced methanol takes place. However, we can avoid such difficulty when the reaction is carried out under reduced pressure even if methyl esters are used as the acyl donor, because the produced methanol is removed immediately from the reaction mixture and thus the reaction equilibrium goes through to produce the desired product.8 To realize this idea, proper choice of the acyl donor ester was very important. The desired reaction was accomplished using methyl phenylth-ioacetate as acyl donor. Various methyl esters can also be used as acyl donor for these reactions; methyl nonanoate was also recommended and efficient optical resolution was accomplished. Using our system, we demonstrated the completely recyclable use of lipase. The transesterification took place smoothly under reduced pressure at 10 Torr at 40° C when 0.5 equivalent of methyl phenylthioacetate was used as acyl donor, and we were able to obtain this compound in optically pure form. Five repetitions of this process showed no drop in the reaction rate (Fig. 4).5b c Recently Kato reported nice additional examples of lipase-catalyzed reaction based on the same idea that CAL-B-catalyzed esterification or amidation of carboxylic acid was accomplished under reduced pressure conditions.9
However, it still remained a problem that the system could not be applied to volatile substrates. We hypothesized that oligomerization of acetaldehyde may be caused by the proton derived from the water molecule trapped by hydrogen bonding at 2-position of the imidazolium ring, because it was suggested that the acidity of the 2-position of imidazolium cation is very high.10 Hence we prepared 1-butyl-2,3-dimethylimidazolium (bdmim) salt and used it as solvent. As we expected, no accumulation of an acetaldehyde oligomer was observed in this solvent system by1 H NMR analysis. The reaction proceeded very smoothly and we were able to use the enzyme repeatedly 10 times while still maintaining perfect
0.5 eq. PhSCH,C02Me MeOH
1st run 3.5
3rd run 3.4
5th run 3.5
Figure 4: Lipase-catalyzed reaction under reduced pressure conditions.
Figure 5: Recyclable use of lipase in [bdmim][BF4] solvent system.
Figure 5: Recyclable use of lipase in [bdmim][BF4] solvent system.
enantioselectivity and high reactivity (Fig. 5).5d It has now been confirmed that we can use the enzyme more than 20 times for several months using this system.
Imidazolium PF6 or BF4 salts were frequently used as solvent for the present lipase-catalyzed reaction. However, these salts are very expensive, and we should develop cheaper ILs. Imidazolium alkyl sulfates might be good candidates because various types of alkyl sulfates can be easily prepared. The imidazolium alkyl sulfate was prepared starting from the corresponding ammonium alkyl sulfate as follows: ammonium alkyl sulfates ([NH4][RSO4]) are easily prepared by the reaction of an alcohol with sulfuric amide, and subsequent anion exchange reaction with [bmim][Cl] gave the corresponding [bmim][RSO4].5e The lipase-catalyzed trans-esterification proceeded in these solvent systems and optically pure acetate was obtained with excellent enantioselectivity. Ethoxyethyl sulfate and phenoxyethyl sulfate gave excellent results and the acetates were obtained with > 99% ee.5e However, interestingly, no reaction took place when [bmim][EtSO4] was used as solvent.
We recently prepared various types of differently fluorinated alkyl sulfate ILs and discovered that the hydrophobicity was dependent on the content ratio of the fluorine on the alkyl sulfate anion and 2,2,3,3,4,4,5,5-octafluoropentyl sulfate salts showed hydrophobic properties. Melting point and viscosity were also dependent on the fluorine contents of the anionic part, while conductivity was determined by the cationic part and not influenced by the fluorine contents. Efficient lipase-catalyzed transesterification was demonstrated using hydrophobic 1-butyl-3-methylimidazolium 2,2,3,3,4,4,5,5-octafluoropentyl sulfate ([bmim][C5F8]) as solvent (Fig. 6).11
Kim and co-workers12 recently reported an excellent example of dynamic kinetic resolution (DKR) using lipase-ruthenium combo catalyst in an IL solvent system (Fig. 7). Applied to this protocol, the authors succeeded in preparing (R)-ester or (S)-ester using lipase PS or subtilisin, respectively. An IL solvent system is truly appropriate for DKR because racemization takes place easily in a highly polar solvent.
Lipase-catalyzed reaction is useful for polyester synthesis and IL was employed successfully as solvent. Uyama and Kobayashi13 demonstrated an efficient polyester synthesis; lipase-catalyzed esterification of agipic acid with butan-1,4-diol proceeded smoothly in [bmim][BF4] solvent, particularly under reduced pressure conditions (Fig. 8). Further Russel14 and Nara15 independently reported efficient examples of the lipase-catalyzed polyester synthesis in an IL solvent system.
Figure 7: Efficient DKR reaction in an IL solvent system.
Figure 7: Efficient DKR reaction in an IL solvent system.
4. ACTIVATION OF LIPASE BY AN IONIC LIQUID
Lozano and co-workers7b reported an interesting stabilization effect of IL for lipase-catalyzed reaction; the authors discovered that the presence of an appropriate substrate was essential for stabilization of enzyme in an IL solvent. The half lifetime of native CAL was only 3.2 h in [emim][PF6] solvent, while it lengthened remarkably to 7500 h in the presence of the substrate. The authors succeeded in demonstrating an efficient lipase-recyclable use system based on scCO2 solvent (Fig. 9).16'17
In the reaction, it was essential to use an IL as a co-solvent. Lozano, Iborra and co-workers recently reported an interesting stabilizing effect of two types of water-immiscible ILs ([emim][TFSI] and [BuMe3N][TFSI]) for CAL-B-catalyzed transesterification of vinyl butyrate.18 The synthetic activity and the stability of the enzyme in these IL solvent systems were markedly enhanced as compared to those in hexane. CAL-B maintained its activity higher than 75% after 4 days of incubation in [emim][TFSI] solvent, while it showed an activity of only 25% when incubated in both water and hexane media at 50°C. Comparison of the ratio of a-helix and ß-sheet by CD spectra showed the activity was closely related with a-helix content which reduced to 31% immediately after lipase was added to hexane and had reached only 2% after 4 days in hexane. On the contrary, no significant reduction of a-helix content was
Substrate + Acyl donor
Figure 9: Lipase recyclable use reactor using a mixed solvent of scCO2 with IL.
obtained in [emim][TFSI] solvent. Based on these results, the authors concluded that a-helix content might play an important role in maintaining the enzymatic activity.
Polyethyleneglycol (PEG) treatment is known to cause stabilization of an enzyme. Goto and co-workers prepared PEG-coated lipase and demonstrated that transesterification of vinyl cinnamate with butanol proceeded smoothly using PEG-lipase PS in [omim][PF6] solvent system.19 The Russell group investigated the details of PEG treatment on the lipase activity in various types of ILs and reported that this activity was mainly dependent on the anionic part of the ILs (Fig. 10)14; high activity was obtained for imidazolium PF6 salt, while no activity was observed for [NO3], [OAc], [CH3SO3], [OTf] or [TFA] salt. The authors proposed that nitric anion or acetate anion might have a strong interaction with some parts of the enzyme protein due to the highly nucleophilic nature of these anions and caused deactivation of the enzyme activity.14
Sheldon's group also investigated the relationship of activity and IL species; CAL-B showed no catalytic activity in some ILs, such as [emim][EtSO4], [bmim][lactate], [EtNH3][NO3] or [bmim][NO3], though the enzyme was soluble in these solvents. On the other hand, high activity was obtained in [bmim][PF6] or [bmim][BF4] although the enzyme was insoluble in them. Interestingly deactivation was recorded for [emim][EtSO4], while the enzyme worked in [emim][MeSO4]. The authors speculated that this deactivation by the IL might be caused by conformational change due to the interaction of the anionic part of the IL with lipase protein based on the results of FT-IR analysis of the protein.20
Was this article helpful?