2008
Biomacromolecules 2010, 11, 2008–2015
Lipase-Catalyzed Oligomerization and Hydrolysis of Alkyl Lactates: Direct Evidence in the Catalysis Mechanism That Enantioselection Is Governed by a Deacylation Step Hitomi Ohara,*,†,‡ Akihisa Onogi,† Masafumi Yamamoto,† and Shiro Kobayashi*,†,§ R&D Center for Bio-Based Materials, Department of Bio-Based Materials Science, and Center for Nanomaterials and Devices, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto 606-8585, Japan Received April 6, 2010; Revised Manuscript Received June 16, 2010
Lipase-catalyzed oligomerization of alkyl D- and L-lactate monomers (RDLa and RLLa, respectively) was studied for the first time. It has been found that the oligomerization occurs enantioselectively only for D-lactates to give oligomers up to heptamers of lactic acid (LA) in good to high yields by using primary C1 to C8 alkyl groups and sec-butyl group for D-lactate monomers. No reaction happened for all L-lactates in similar conditions. Lipasecatalyzed hydrolysis of alkyl D- and L-lactates was also examined, revealing that the hydrolysis took place for both D- and L-lactates, although L-lactates proceeded a couple of times slower. The hydrolysis results clearly demonstrate that the lipase catalysis mechanism involves an acyl-enzyme intermediate (EM) formation via the acylation step from both D- and L-lactates as a rate-determining step, and the subsequent deacylation step, a nucleophilic attack of water to the EM, takes place to produce free LA. On the other hand, in the oligomerization of D-lactates, the deacylation step, in which a sec-alcohol group of the monomer or of the propagating chain-end attacks to the EM, is only allowed for the sec-D-alcohol group to give a one-LA-unit-elongated oligomer. L-Lactates form the EM; however, the subsequent deacylation reaction with both the sec-L- and sec-D-alcohol groups does not take place, failing in the oligomerization to occur. These results provide with the first direct evidence in the lipase catalysis that the enantioselection is governed by the deacylation step. In the co-oligomerization between L- and D-lactates, the L-isomer retarded the reaction rate of the D-isomer, which was found due to the function of the former as a competitive inhibitor in the acylation step toward the latter.
Introduction Poly(lactic acid) (PLA) is currently one of the main polymers attracting much interest in polymer chemistry field, in particular, from viewpoint of “green plastics”1 as well as “green polymer chemistry”.2 Lactic acid (LA) can be produced via fermentation of corn, sugar canes, and so on, which are bio-based renewable resources. PLA is prepared either by ring-opening polymerization of lactide (a six-membered cyclic dimer of LA) normally using Sn(II) catalyst3 or by dehydration polycondensation of LA using an acid catalyst.4 Application studies of PLA are currently expanding for preparing new plastic materials with thermally more stable properties by designing stereocomplexes from PLLA and PDLA chains.5 It is not known so far, to our knowledge, to conduct polycondensation of an alkyl lactate (RLa) with liberating an alcohol for producing PLA, which is a transesterification polymerization.2h,j,k Therefore, pursuing such polycondenzation has been aimed in this study by employing lipase as catalyst, in relevant to a recent study on lipase-catalyzed enantioselective ring-opening polymerization of D,D-lactide.6 As to the lipasecatalyzed polyester synthesis, a number of studies have been published,2b,c,h,j,k,7 including the ring-opening polymerization and copolymerization of lactide,8 and of an O-carboxylic anhydride derived from LA.9 We report here the results on a new lipase (Novozym 435)-catalyzed enantioselective condensa* To whom correspondence should be addressed. E-mail:
[email protected] (H.O.);
[email protected] (S.K.). † R&D Center for Bio-Based Materials. ‡ Department of Bio-Based Materials Science. § Center for Nanomaterials and Devices.
tion oligomerization of alkyl lactates (RLa)s as well as their hydrolysis as the oligomerization-related reaction, and describe a first clear-cut reaction mechanism of the lipase catalysis. To our knowledge, alkyl lactates have not been used to date as a monomer in the polymerization chemistry.
Experimental Section Materials. Aqueous L-lactic acid (LLA, 90% solutions, HiPure 90) and D-lactic acid (DLA 90) were purchased from Purac Biochem b.v. (NLD). Immobilized lipase CA (Novozym 435) was obtained from Novo Nordisk A/S (DNK), whose catalyst activity was 10000 PLU/g. Novozym 435 sample was kept in a decompressed desiccator containing silica gel for 1 day, and then kept in a screw bottle with silica gel. The bottle was preserved in a refrigerator. All the alcohols of extra-purity grade and molecular sieves 3 Å were obtained from Kanto Kagaku Co. 1,4-Dioxane, tetrahydrofuran (THF), deuterochloroform (CDCl3), and other solvents were commercially available and used without further purification. Commercially available methyl L-lactate, ethyl L-lactate (Tokyo Chemical Ind. Co., Tokyo, Japan), and n-butyl L-lactate (Sigma-Aldrich Co., MO) were used. The other alkyl lactates (RLa) were prepared by esterification of LLA or DLA with a corresponding alcohol. Following alcohols were used: methyl (Me)-, ethyl (Et)-, n-propyl (Pr)-, n-butyl (Bu), sec-butyl (sBu)-, iso-butyl (iBu)-, n-pentyl (Pe)-, n-hexyl (Hx)-, n-heptyl (Hp)-, and n-octyl (Oc)-alcohols. A typical run was as followed. In a 500 mL flask, 50 g of D-lactic acid, 200 mL of ethanol, and 0.57 g of sulfuric acid were placed and kept in an oil bath of 120 °C for 8 h. The flask was equipped with a reflux cooler filled with molecular sieves 3 Å. After the reaction, the reaction mixture was distilled under reduced pressure three times for purification. Ethyl D-lactate (EtDLa) was obtained in 25% yields. For
10.1021/bm1003674 2010 American Chemical Society Published on Web 07/02/2010
Lipase-Catalyzed Oligomerization the reaction apparatus, a similar apparatus was used for hydrophilic alcohols (MeOH, EtOH, PrOH, and sBuOH), whereas a Dean-Stark apparatus for hydrophobic alcohols (BuOH, iBuOH, PeOH, HxOH, HpOH, and OcOH). In all preparations, alkyl lactates were prepared in 8-36% yields. Characterization of alkyl lactates was performed by measuring 1H NMR, ESI-TOF-MS, and GPC. Optical purity was determined by GC analysis using an optical resolution column, according to the relationships
D(%) ) [D*/(D* + L*)] × 100 L(%) ) [L*/(D* + L*)] × 100 where D* and L* denote the peak area value of D- and L-isomers in the GC chart. 1 H NMR data (CDCl3, 500 MHz, δ relative to TMS), the mass values, and the optical purity values are given as follows: MeLa (D and L): 1.40-1.44 (s, 3H, CH3), 2.70-2.95 (s, 1H, OH), 3.76-3.82 (d, 3H, CH3), 4.26-4.32 (q, 1H, CH); ms ) 104; D ) 99.5%, L ) 99.5%. EtLa (D and L): 1.28-1.32 (t, 3H, CH3), 1.40-1.43 (d, 3H, CH3), 2.82-2.90 (s, 1H, OH), 4.22-4.28 (m, 2H, CH3), 4.22-4.27 (m, 1H, CH); ms ) 118; D ) 99.5%, L ) 99.3%. PrLa (D and L): 0.94-0.97 (t, 3H, CH3), 1.41-1.44 (d, 3H, CH3), 1.66-1.73 (d, 2H, CH2), 2.89-2.93 (s, 1H, OH), 4.10-4.20 (m, 2H, CH2), 4.25-4.30 (q, 1H, CH); ms ) 132; D ) 99.7%, L ) 99.7%. BuLa (D and L): 0.92-0.97 (t, 2H, CH3), 1.36-1.44 (m, 3H, CH3), 1.41-1.44 (d, 3H, CH3), 1.62-1.68 (q, 2H, CH3), 2.81-2.83 (s, 1H, OH), 4.16-4.24 (d, 3H, CH3), 4.26-4.31 (q, 1H CH); ms ) 146; D ) 99.6%, L ) 99.2%. sBuLa (D and L): 0.89-0.92 (t, 3H, CH3), 1.22-1.26 (t, 3H, CH3), 1.38-1.42 (t, 3H, CH3), 1.55-1.65 (q, 2H, CH2), 2.80-2.99 (s, 1H, OH), 4.21-4.25 (q, 1H, CH), 4.91-4.95 (q, 1H, CH); ms ) 146; D ) 97.4%, L ) 99.9%. iBuLa (D and L): 0.92-1.0 0(d, 3H, CH3), 1.42-1.45 (d, 3H, CH3), 1.93-2.04 (m, 1H, CH), 2.85-2.91 (s, 1H, OH), 3.93-3.97, 3.98-4.35(q,d, 1H, CH), 4.26-4.32 (q, 1H, CH); ms ) 146; D ) 98.4%, L ) 98.5%. PeLa (D and L): 0.88-0.94 (t, 3H, CH3), 1.32-1.38 (m, 2H, CH2), 1.40-1.44 (d, 3H, CH3), 1.64-1.70 (d,t, 2H, CH2), 2.76-2.95(s, 1H, OH), 4.14-4.22(m, 2H, CH2), 4.24-4.30(q, 1H, CH); ms 160; D ) 98.7%, L ) 98.9%. HxLa (D and L): 0.87-0.92 (t, 3H, CH3), 1.26-1.38 (m, 2H, CH2), 1.40-1.44 (d, 3H, CH3), 1.63-1.69 (d,t, 2H, CH2), 2.70-2.94(s, 1H, OH), 4.14-4.22 (m, 2H, CH2), 4.24-4.30 (q, 1H, CH); ms ) 174, D ) 98.9%, L ) 99.4%. HpLa (D and L): 0.86-0.94 (t, 3H, CH3), 1.24-1.38 (m, 2H, CH2), 1.40-1.44 (d, 2H, CH2), 1.63-1.70 (d,t, 2H, CH2), 2.80-2.85 (s, 1H, OH), 4.14-4.22 (m, 2H, CH2), 4.24-4.30 (q, 1H, CH); ms ) 188; D ) 98.9%, L ) 98.8%. OcLa (D and L): 0.86-0.91 (t, 3H, CH3), 1.24-1.40 (m, 2H, CH2), 1.40-1.44 (d, 2H, CH2), 1.63-1.70 (d, t, 2H, CH2), 2.79-2.82 (d, 1H, OH), 4.12-4.22 (m, 2H, CH2), 4.24-4.30 (q, 1H, CH); ms )202; D ) 98.4%, L ) 99.6%. Oligomerization of Alkyl Lactates. A typical procedure is as follows. To EtDLa (5.0 mmol, 0.59 g) in a test tube containing a stirrer bar, 25 mg of Novozym 435 was added, and the mixture was allowed to react at 50 °C with stirring under a reduced pressure at 3.3 kPa. The consumption of the substrate was followed by sampling 20 µL from the reaction mixture occasionally and by GC analysis of the reaction mixture sample using n-propylbenzene as the standard substance. Characterization of the product oligoLAs was performed by 1H NMR and ESI-TOF-MS analyses. Co-Oligomerization between EtDLa and EtLLa. A mixture of the substrates (total amount of EtDLa + EtLLa ) 5.0 mmol, with varying the feed ratio), without or with 1,4-dioxane (0.50 mL) solvent, n-propylbenzene (200 µL, internal standard), and Novozym 435 (75 mg or 25 mg) was allowed to react at 3.3 or 101 kPa at 50 °C with stirring. The consumption of EtDLa and EtLLa was followed by the GC analysis of the sample taken out at 2 h or 10 min intervals from the reaction mixture.
Biomacromolecules, Vol. 11, No. 8, 2010
2009
Inhibition Type Test. A typical run was as followed. EtLLa (0.25 mmol) was added to the mixture containing EtDLa (2.0 mmol), 1,4dioxane (0.50 mL), n-propylbenzene (200 µL, internal standard), and Novozym 435 (25 mg). The reaction was performed at atmospheric pressure at 50 °C with stirring. The consumption of EtDLa was followed by the GC analysis of the sample taken out at 10 min intervals from the reaction mixture. Similar experiments were performed with varying the added amount of EtLLa from 0.0 to 0.25 to 0.50 mmol and with varying the substrate amount of EtDLa from 2.0, 3.0, 4.0, and 5.0 mmol, thus, for the total 12 runs. Hydrolysis of Alkyl Lactates. A typical run was as follows. To a mixture solution of EtDLa (5.0 mmol, 0.59 g) and 1,4-dioxane (0.50 mL) containing n-propylbenzene (200 µL, as internal standard), distilled water (5.0 mmol, 90 mg) and Novozym 435 (25 mg) were added. The mixture was homogeneous, except for the catalyst, and reacted at 50 °C with stirring. The consumption of EtDLa was followed by the gas chromatography (GC) analysis of the sample taken out at 10 min intervals from the reaction mixture. Analytical Methods. 1H NMR measurements were recorded on a spectrometer ARX-500 (500 MHz, Bruker BioSpin GmbH, GER). ESI-TOF-MS analysis was performed by using a micrOTOF instrument (Bruker Daltonik GmbH, GER). GC apparatus was equipped with an Rt-βDEXsm 0.25 mm × 30 m × 0.25 µm column (Restek Co., PA, U.S.A.) which was maintained at 80 °C with an FID detector. The injection and the detector temperatures were set to 200 °C. Molecular weights were determined by gel permeation chromatography (GPC) with a refractive index detector (HLC-8220, Tosoh Co.). The analytical conditions were as follows: columns, TSK gel Super HZ2000, HZ4000, and a guard column (Tosoh Co.); column temperature, 40 °C; solvent for elution, CHCl3; and flow rate, 0.35 mL/min. Polystyrene standards were used for calibration.
Results and Discussion Lipase-Catalyzed Enantioselective Oligomerization. For the purpose of screening the monomer reactivity and the catalyst activity, lipase (Novozym 435)-catalyzed polycondensation of 18 alkyl D- and L-lactates (RDLa and RLLa) was examined as shown Table 1. It is clear that the reaction of all the alkyl D-lactates was induced to give oligo(D-lactic acid)s (oligoDLAs) in good to high yields. There was a tendency that primary alkyl lactates of Et-, Pr-, and Bu- showed a higher reactivity compared with longer alkyl lactates like Pe-, Hx-, Hp-, and Oc-, and also a secondary alkyl lactate of sBuDLa showed a decreased reactivity. On the other hand, all alkyl L-lactates did not show any reactivity regardless of the primary or secondary structure of alkyl groups; nothing happened under the reaction conditions. The condensation oligomerization is definitely enantioselective (Scheme 1). Figure 1 shows 500 MHz 1H NMR spectra of the reaction mixture of iBuDLa, in which (A) clearly indicates signals due to the product oligomers and unreacted monomer; in an expanded version in (B), specific methine proton of the monomer appears at δ 4.28 (peak b), the terminal methine proton of oligomers at δ 4.36 (peak a), and internal methine protons of oligomers at δ 5.37 (peak a′). From the integrated ratio of peaks a, a′, and b, conversion of monomer (70%) as well as an average degree of oligomerization (n ) 2.72 in Scheme 1) could be calculated. In contrast, the reaction of iso-butyl L-lactate (iBuLLa) did not occur under the similar reaction conditions; the monomer recovered unchanged as seen in (C). Data of conversion values of all other lactates have been similarly obtained and included in Table 1. Figure 2 demonstrates the TOF-MS chart of the oligomers obtained from the reaction of iBuDLa indicating the formation
2010
Biomacromolecules, Vol. 11, No. 8, 2010
Ohara et al.
Table 1. Oligomerization of Alkyl Lactate Monomers by Lipase Catalysta code
alkyl group of alkyl lactate monomers
D or L of monomer
reaction pressureb (kPa)
conversion of monomerc (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
ethyl ethyl n-propyl n-propyl n-butyl n-butyl sec-butyl sec-butyl iso-butyl iso-butyl n-pentyl n-pentyl n-hexyl n-hexyl n-heptyl n-heptyl n-octyl n-octyl
D L D L D L D L D L D L D L D L D L
66 66 27 27 9.3 9.3 40 40 33 33 6.7 6.7 2.7 2.7 1.3 1.3 0.7 0.7
81 0 64 0 82 0 37 0 70 0 40 0 36 0 39 0 34 0
n values of oligomersd 2, 3, 4, 5 2, 3, 4, 5, 6 2, 3, 4, 5, 6, 7 2, 3, 4, 5 2, 3, 4, 5, 6 2, 3, 4, 5, 6 2, 3, 4, 5 2, 3, 4, 5 2, 3, 4
a Reaction of monomer (2.0 g) and Novozym 435 (0.20 g) at 80 °C with stirring for 24 h; without Novozym 435, no reaction took place. b Reaction was carried out under a reduced pressure, which was used so that a liberated alcohol might be evaporated from the reaction system. c Conversion (%) was determined from 1H NMR analysis of the reaction mixture after the reaction. d n values of Scheme 1 were determined from ESI-TOF-MS analysis of the products.
Scheme 1
from dimer up to hexamer of DLA (n values in Scheme 1). These data together with other reactions of lactates are cited in Table 1.
Figure 2. ESI-TOF-MS chart of the oligomers from the reaction of iBuDLa: the tallest peak, 313, corresponds to [M + Na]+ (n ) 3), the peak at 329 corresponds to [M + K]+ (n ) 3), and so forth.
Figure 3. Ethyl lactate (EtLa) conversion vs reaction time curves: EtDLa (5.0 mmol) ([) and EtLLa (5.0 mmol) (]), both with 25 mg of Novozym 435 at 50 °C with stirring at 3.3 kPa.
Figure 1. 1H NMR spectra (500 MHz, CDCl3, chemical shift in δ relative to TMS): (A) the reaction mixture of iBuDLa, (B) the expanded spectrum of the methine proton region, and (C) the reaction mixture of iBuLLa.
The lipase (Novozym 435)-catalyzed polycondensation of ethyl D- and L-lactate (EtLa) was investigated in more detail under a lower reaction temperature at 50 °C at a reduced pressure of 3.3 kPa. Under the reaction conditions, EtDLa was consumed in more than 90% after 12 h, whereas EtLLa was not consumed at all (Figure 3). The results clearly indicate that D-enantioselective condensations took place (Scheme 1). It was found that the reaction products were oligomers from DLA dimer up to hexamer, according to TOF-MS analysis as shown in Figure 4. Tri-, tetra-, and penta-DLA were main fractions in the products. These oligomers possessed an OH group at one end and an ethyl ester group at the other. Furthermore, the structure of the oligomers was supported by
Lipase-Catalyzed Oligomerization
Biomacromolecules, Vol. 11, No. 8, 2010
Figure 4. ESI-TOF-MS chart of the condensation products of EtDLa: the tallest peak, 357, corresponds to [M + Na]+ (n ) 4), the peak at 373 corresponds to [M + K]+ (n ) 4), and so forth.
2011
Figure 6. Mixture of an alkyl lactate (5.0 mmol), Novozym 435 (25 mg), 1,4-dioxane (0.50 mL), and n-propylbenzene (200 µL, the internal standard for the GC analysis), subjected to the reaction at 50 °C with stirring: MeDLa (9); EtDLa ([); PrDLa (2); BuDLa (b).
Figure 7. Plots of ln([S]0/[S]t) vs reaction time: MeDLa (9); EtDLa ([); PrDLa (2); BuDLa (b).
Figure 5. 1H NMR (500 MHz) spectrum of the oligomer products obtained from EtDLa (A) and that of the expanded methine proton region (B). 1
H NMR spectrum, showing a specific methine proton at the chain end appearing at δ 4.36, as shown in Figure 5. A similar oligomerization was performed with varying the substrates and the reaction conditions. Four alkyl D-lactate substrates, MeDLa, EtDLa, PrDLa, and BuDLa, were subjected to the reaction in 1,4-dioxane solvent. Results of reactions at 60 °C are given in Figure 6. It is to be noted that an alcohol component of the substrate did not much effect on the rate of the oligomerization. For the comparison of the rate value semiquantitatively, a reaction rate was calculated according to the following treatment: k′
kcat
E + S a ES f E + P -
d[S] ) k′[E][S] ) k[S] dt
(k′[E] ) k)
(1) (2)
where E, S, and P denote enzyme, substrate, and product, respectively, in the Michaelis-Menten eq 1. For simplicity, we followed only the very beginning of the reaction, and then [S]
Figure 8. Time-conversion curves of the reaction of MeDLa, EtDLa, PrDLa, and BuDLa at 50 °C at 3.3 kPa; subtrate amount (5.0 mmol each); Novozym (25 mg): MeDLa (9); EtDLa ([); PrDLa (2); BuDLa (b).
was much higher than [E], and hence, k of eq 2 can be regarded as a pseudo-first-order rate constant. Plots of the integrated form of eq 2 were made in Figure 7. Thus, values of k obtained were in the range of 3.4-4.4 × 10-4 s-1 for four substrates, MeDLa, EtDLa, PrDLa, and BuDLa. The reactivity of these four monomers was not much different under the reaction conditions. TOF-MS charts of the reaction mixture of these four lactates were similar to the observed in Figures 2 and 4, indicating the production of oligoDLAs from dimer to hexamer. Without using solvent, the above four alkyl D-lactates were subjected to the oligomerization at 50 °C at 3.3 kPa with Novozym 435 catalyst with stirring (Figure 8). Here, again EtDLa was the most reactive substrate; after 12 h around 90% of EtDLa were consumed to give oligoDLAs like shown in Figure 4. Co-Oligomerization of D-Lactate and L-Lactate. It was examined to co-oligomerize EtDLa and EtLLa (2.5 mmol each) at 50 °C at 3.3 kPa with using 75 mg of Novozym 435 (3-fold amount of the reaction of Figure 3). The time-conversion curves for the substrate consumption are given in Figure 9. It is very surprising that EtLLa was not consumed at all, while EtDLa was reacted, but only in a slow rate compared with the reaction of Figure 3; only 8% of the starting EtDLa was
2012
Biomacromolecules, Vol. 11, No. 8, 2010
Figure 9. Substrate consumption and time relationships in the bulk co-oligomerization between EtDLa and EtLLa (both 2.5 mmol) at 50 °C at a pressure of 3.3 kPa with Novozym 435 catalyst (75 mg): EtDLa ([); EtLLa (]).
Ohara et al.
Figure 11. Lipase-catalyzed hydrolysis of BuLa using water (8.0 mmol), Novozym 435 (125 mg), and n-propylbenzene (0.20 mL as internal standard) in THF (0.80 mL) at 50 °C under atomospheric pressure with stirring for the substrate: BuDLa (5.0 mmol) (b); BuLLa (5.0 mmol) (O). Scheme 2
Figure 10. Lineweaver-Burk plots to elucidate the inhibition type of EtLLa to the oligomerization of EtDLa. EtLLa [0.0 mmol (b); 0.25 mmol (2); or 0.50 mmol (9)] was added to the mixture of EtDLa ([S]: 2.0, 3.0, 4.0, or 5.0 mmol), 1,4-dioxane (0.50 mL), n-propylbenzene (0.20 mL as internal standard), and Novozyme 435 (25 mg), and the reaction was carried out at 50 °C at atmospheric pressure with stirring.
consumed even after 12 h. From TOF-MS analysis of the products, dimer and trimer were the main products, produced in an approximately 3:2 ratio. No higher oligomers were observed, being different from the results of Figure 3, where the oligomerization of EtDLa and EtLLa was independently conducted. For in more detail, experiments varying the feed molar ratio of EtDLa and EtLLa were carried out in 1,4-dioxane solution containing n-propylbenzene as internal standard at 50 °C at 101 kPa. The six ratios of D-isomer/L-isomer were 100:0, 90:10, 75: 25, 50:50, 25:75, and 10:90 as well as the six inverse ratios of L-isomer/D-isomer. In these solution experiments, EtDLa was noticeably consumed only in the reaction with feed ratios (Disomer/L-isomer) of 100:0 and 90:10 and consumed slightly in the 75:25 ratio reaction, after 60 min. In the reactions of the other three ratios, EtDLa was not consumed. In all other-way six runs, EtLLa was not consumed at all. These observations suggested inhibition effects of EtLLa on the reaction of EtDLa. Actually, the co-oligomerization did not take place; only oligomers from EtDLa were produced. To elucidate the inhibition function of EtLLa toward the oligomerization of EtDLa, EtLLa was added to the EtDLa reaction. The reaction rate, namely, the EtDLa consumption rate (ν0, mol L-1 s-1), was evaluated from the very beginning (
