Ether-functionalized ionic liquids for nonaqueous biocatalysis: Effect of different cation cores

Article abstract of DOI:10.1016/j.procbio.2019.03.018

Ether-functionalized ionic liquids (ILs) usually have low viscosities, and can be designed to be compatible with enzymes. However, there is a lack of understanding of the effect of different ether-functionalized structures on the enzyme activity. We systematically evaluated new ether-functionalized ILs carrying different cation cores (pairing with Tf₂N^? anions) in two Novozym 435-catalyzed reactions: (1) the transesterification of ethyl sorbate with 1-propanol at 50 °C; (2) the ring-opening polymerization (ROP) of ε-caprolactone at 70 °C. The lipase showed different activities: in the first reaction, [CH₃OCH₂CH₂-Et₃N][Tf₂N] and [CH₃OCH₂CH₂-Py][Tf₂N] gave the highest reaction rates; in the second reaction, [CH₃OCH₂CH₂-PBu₃][Tf₂N] produced the highest molecular mass (M_w up to 25,400 Da). The lipase's thermal stability in [CH₃OCH₂CH₂-Et₃N][Tf₂N] was found much higher than that in t-butanol. The fluorescence spectra of free lipase (excited at 280 nm) in these ILs reveal that the wavelength of the maximum emission peak occurred at 314 nm for both [CH₃OCH₂CH₂PBu₃][Tf₂N] and [CH₃OCH₂CH₂PEt₃][Tf₂N], which matched closely with that (313 nm) in aqueous phosphate buffer (pH 7.5, 20 mM), while other ether-functionalized ILs led to various degrees of red shifts. In summary, the lipase activity is not only dependent on the IL structure, but also on the substrate and other reaction conditions.

Full text of DOI:10.1016/j.procbio.2019.03.018

ProcessBiochemistryxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Process Biochemistry
journal homepage: www.elsevier.com/locate/procbio
Ether-functionalized ionic liquids for nonaqueous biocatalysis: Effect of
different cation cores
Hua Zhao⁎, Naphatsawan Kanpadee, Chanida Jindarat
Department of Chemistry and Biochemistry, University of Northern Colorado, Greeley, CO 80639, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Ether-functionalized ionic liquids (ILs) usually have low viscosities, and can be designed to be compatible with
enzymes. However, there is a lack of understanding of the effect of different ether-functionalized structures on
the enzyme activity. We systematically evaluated new ether-functionalized ILs carrying different cation cores
(pairing with Tf2N− anions) in two Novozym 435-catalyzed reactions: (1) the transesterification of ethyl sorbate
with 1-propanol at 50 °C; (2) the ring-opening polymerization (ROP) of ε-caprolactone at 70 °C. The lipase
showed different activities: in the first reaction, [CH3OCH2CH2-Et3N][Tf2N] and [CH3OCH2CH2-Py][Tf2N] gave
the highest reaction rates; in the second reaction, [CH3OCH2CH2-PBu3][Tf2N] produced the highest molecular
mass (Mw up to 25,400 Da). The lipase’s thermal stability in [CH3OCH2CH2-Et3N][Tf2N] was found much higher
than that in t-butanol. The fluorescence spectra of free lipase (excited at 280 nm) in these ILs reveal that the
wavelength of the maximum emission peak occurred at 314 nm for both [CH3OCH2CH2PBu3][Tf2N] and
[CH3OCH2CH2PEt3][Tf2N], which matched closely with that (313 nm) in aqueous phosphate buffer (pH 7.5,
20 mM), while other ether-functionalized ILs led to various degrees of red shifts. In summary, the lipase activity
is not only dependent on the IL structure, but also on the substrate and other reaction conditions.
Biocatalysis
Ionic liquid
Lipase
Transesterification
Ring-opening polymerization
Enzyme activity
Thermal stability
1. Introduction
conventional organic solvents) for biocatalysis due to their designable
structures and tunable properties [6–11]. To mimic the water-like en-
Enzymatic reactions are often reversible, and thus an enzyme can
synthesize or decompose the substrate molecules depending on the
reaction conditions. Reaction medium is one of the most critical factors
that can shift the reaction direction. As an example, hydrolases (such as
proteases, lipases and cellulases) catalyze the hydrolysis (i.e. decom-
position) of different substrates (e.g. peptides, lipids, and cellulose re-
spectively) in aqueous media; on the other hand, these reactions are
reversed (i.e. synthesis) in nonaqeuous solvents including organic sol-
vents, supercritical fluids, and ionic liquids (ILs) [1–3]. However, the
enzyme activities in nonaqueous organic solvents are usually lower by
several magnitudes than those in aqueous solutions. For instance, the
proteases α-chymotrypsin and subtilisin in anhydrous octane showed
104‒105 times lower activities than in water [4]. The underlying me-
chanism is not fully understood, but several important parameters have
been discussed including the substrate diffusion, accessibility of en-
zyme’s active site, structural changes of enzyme molecules, energetics
of substrate desolvation and transition state stabilization, conforma-
tional mobility, and pH optimization [5].
vironment for favorable enzyme-solvent interactions, different versions
of hydroxy- or ether-/glycol-functionalized ILs have been developed for
a variety of enzymatic reactions. A number of commercial glycol-
grafted tetraammonium-based ILs were evaluated by the Xu group
[12–17] as reaction media for the enzymatic glycerolysis; particularly,
Ammoeng 100 (known as [CPMA][MeSO4]1) and Ammoeng 102 were
able to dissolve triglycerides and afforded high lipase activities during
glycerolysis [13,14]; in addition, trioctylmethylammonium bis(tri-
fluoromethylsulfonyl)imide ([TOMA][Tf2N]) and its mixture with
Ammoeng 102 led to effective enzymatic glycerolysis [16–18]. Higher
transesterification activities were reported in [CPMA][MeSO4] (vs. non-
functionalized ILs) for both free and immobilized Candida antarctica
lipase B (CALB); however, lipases from Thermomyces lanuginosus (TLL)
and Rhizomuncor miehei (RML) became less active in [CPMA][MeSO4]
[19]. Higher activities of alcohol dehydrogenase were observed by the
Kroutil group [20] in hydroxy-functionalized ILs than in ordinary ILs
even when the IL concentrations were up to 50–90% (v/v). An aqueous
two-phase system (ATPS) was constructed from Ammoeng 110, and was
used to purify active enzymes (two alcohol dehydrogenases); the Kragl
Ionic liquids (ILs) are alternative nonaqueous media (vs.
⁎ Corresponding author.
E-mail address: hua.zhao@unco.edu (H. Zhao).
1 From the name of cocosalkyl pentaethoxy methylammonium methylsulfate.
https://doi.org/10.1016/j.procbio.2019.03.018
Received 9 February 2019; Received in revised form 9 March 2019; Accepted 19 March 2019
1359-5113/©2019ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:HuaZhao,NaphatsawanKanpadeeandChanidaJindarat,ProcessBiochemistry,
https://doi.org/10.1016/j.procbio.2019.03.018

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
group found the IL could stabilize the enzymes while dissolving more
hydrophobic substrates [21]. Tetrakis(2-hydroxyethyl)ammonium tri-
flouromethanesulfonate was prepared by Das et al [22], and was found
highly compatible with horseradish peroxidase than methanol (10-fold
more active) and conventional ILs (30–240-fold more active). An ether-
functionalized phosphonium, 2-methoxyethyl(tri-n-butyl)phosphonium
bis(trifluoromethylsulfonyl)imide ([MeOCH2CH2-Bu3P][Tf2N]) was
synthesized by the Itoh group [23], and this IL enabled a faster reaction
rate than diisopropyl ether for the transesterification of secondary al-
cohols catalyzed by lipase PS. Another group [24] reported the use of
two functionalized ILs (i.e. [C2OHmim][PF6] and [C5O2mim][PF6]) for
the esterification of glycerol with sinapic acid catalyzed by feruloyl
esterase, and obtained high yields (up to 72.5% and 76.7% respectively
in these two ILs). The Yuan group [25] synthesized several ether-
grafted imidazolium-type ILs, and reported that 1-(3-ethoxypropyl)-2,3-
dimethylimidazolium bis(trifluoromethylsulfonyl)imide enabled 99%
enantioselectivity and 50% conversion for the Novozym534-catalyzed
transesterification of rac-1-phenylethanol with vinyl acetate. Fan et al.
2. Materials and methods
2.1. Materials
Free Candida antarctica lipase B (CALB), which is a recombinant
from Aspergillus oryzae (catalog #62288, Lot# BCBP3380V), and
Novozym 435, known as CALB immobilized on acrylic resin (Catalog
#L4777, Lot # SLBW1544 and # SLBP0766V) were purchased from
Sigma-Aldrich (St. Louis, MO). 2-Bromoethyl methyl ether was the
product of BeanTown Chemical (Hudson, NH), and lithium bis(tri-
fluoromethylsulfonyl)imide (Li[Tf2N]) was the product of Matrix
Scientific (Columbia, SC), both of which were supplied by VWR
(Radnor,
PA).
1-Bromo-2-(2-methoxyethoxy)ethane,
1-(2-bro-
moethoxy)-2-(2-methoxyethoxy)ethane, ε-caprolactone, 2-chloroethyl
ethyl sulfide, and 2-chloroethyl phenyl sulfide were provided by TCI
America (Portland, OR). 1-Butyl-3-methylimidazolium bis(tri-
fluoromethylsulfonyl)imide ([BMIM][Tf2N], synthesis grade) and 1-
butyl-3-methylimidazolium hexafluorophosphate ([BMIM][PF6], high
purity) were produced by Merck KGaA (EMD Millipore Corporation,
Billerica, MA) and supplied by VWR (Radnor, PA).
[26] found that
a hydroxy-containing IL, N-methyl-N-(3-hydro-
xypropyl)pyrrolidinium bis(trifluoromethylsulfonyl)imide was more
effective than organic solvents (e.g. n-hexane and t-butanol) and non-
functionalized ILs for improving the biodiesel yield during the lipase-
catalyzed transesterification of soybean oil with methanol. The same
group [27] also demonstrated that hydroxy-containing imidazolium ILs
with short alky chains [such as 1-(3-hydroxypropyl)-3-methylimidazo-
lium bis(trifluoromethylsulfonyl)imide] could enhance the lipase ac-
tivity for the Michael addition synthesis of warfarin from 4-hydro-
xycoumarin and benzylideneacetone. The Itoh group [28] coated
Burkholderia cepacia lipase by pyridinium-based ILs carrying cetyl-
PEG10 sulfate (a glycol-functionalized anion), and reported that these
enzyme preparations led to high reaction rates and enantioselectivities
for the transesterification of a variety of secondary alcohols.
2.2. Preparation of several ether-functionalized ILs
Ether-functionalized ILs based on phosphonium, imidazolium, pyr-
idinium, alkylammonium, and piperidinium cations (see Table 1) were
recently prepared and characterized in our laboratory [34]: 2-meth-
oxyethyl-triethylphosphonium
bis(trifluoromethylsulfonyl)imide
([MeOCH2CH2-PEt3][Tf2N]), 2-methoxyethyl-tributylphosphonium bis
(trifluoromethylsulfonyl)imide ([MeOCH2CH2-PBu3][Tf2N]), 2-meth-
oxyethyl-tributylphosphonium
([MeOCH2CH2-PBu3][beti]),
bis(pentafluoroethanesulfonyl)imide
(2-methoxyethoxy)ethyl-tributylpho-
Previously, our group designed a series of hydroxy- or ether-func-
tionalized imidazolium- and alkylammonium-based ILs containing the
acetate anion [29,30]. We reported that several ether-functionalized ILs
could dissolve a variety of substrates (e.g. cellulose, sugars, ascorbic
acid, amino acids, betulinic acid, fatty acids, and triglycerides) that are
not typically soluble in common organic solvents; meanwhile, these ILs
enabled reasonably high enzymatic activities in several lipase-catalyzed
transesterification reactions [29,31,32]. The hydrophobic versions
(carrying Tf2N− anions) of three ether-functionalized ILs containing
10–15% (v/v) water led to high synthetic activities and selectivities in
the subtilisin-catalyzed transesterification of N-acetyl-L-phenylalanine
ethyl ester with 1-propanol [33]. Our recent study [34] synthesized
several new ether-functionalized ILs (based on phosphonium, imida-
zolium, pyridinium, alkylammonium, and piperidinium cations) car-
rying Tf2N− anions for the high-temperature enzymatic polymerization
reactions (70 °C for the polymerization of ε-caprolactone, and 130 °C for
the polymerization of L-lactide). However, there is no systematic com-
parison of how different ether-functionalized ILs affect the enzyme
stabilization, and a lack of mechanistic discussion of the IL structure-
enzyme activity relationship. This study intends to prepare a series of IL
cation cores functionalized by ether-chains, and then systematically
evaluate how these tailored structures stabilize the enzyme. To achieve
this objective, we screened these custom-made ILs by two CALB-cata-
lyzed synthetic reactions: the transesterification of ethyl sorbate with 1-
propanol at 50 °C, and the ring-opening polymerization (ROP) of ε-ca-
prolactone at 70 °C. We further probed the CALB-IL interactions by
fluorescence emission spectra for a molecular-level view of enzyme
stabilization in different types of solvents. Our results provide general
guidance for designing enzyme-compatible ether-functionalized ILs.
sphonium bis(trifluoromethylsulfonyl)imide ([Me(OCH2CH2)2-PBu3]
[Tf2N]), (2-(2-methoxyethoxy)ethoxy)ethyl-tributylphosphonium bis
(trifluoromethylsulfonyl)imide ([Me(OCH2CH2)3-PBu3][Tf2N]), 1-
ethyl-3-(2-methoxyethyl)imidazolium
bis(trifluoromethylsulfonyl)
imide ([MeOCH2CH2-Im-Et][Tf2N]), N-(2-methoxyethyl)pyridinium bis
(trifluoromethylsulfonyl)imide ([MeOCH2CH2-Py][Tf2N]), N-(2-meth-
oxyethyl)-N-methylpiperidinium
bis(trifluoromethylsulfonyl)imide
2-(Ethylmercapto)ethyl-tributylpho-
([MeOCH2CH2-Pip-Me][Tf2N]).
sphonium bis(trifluoromethylsulfonyl)imide ([CH3CH2SCH2CH2-PBu3]
[Tf2N]) and 2-(phenylmercapto)ethyl-tributylphosphonium bis(tri-
fluoromethylsulfonyl)imide ([PhSCH2CH2-PBu3][Tf2N]) were prepared
followed the same method [34] by refluxing 2-chloroethyl ethyl sulfide
or 2-chloroethyl phenyl sulfide with tributylphosphine first, followed
by the anion exchange from Cl− to Tf2N−
.
[CH3CH2SCH2CH2-PBu3][Tf2N]. 1H-NMR (400 MHz, CDCl3, [ppm])
δ = 0.97 (9H, m, 3CH3CH2CH2CH2-), 1.28 (3H, t, CH3CH2S-), 1.52
(12H, m, 3CH3CH2CH2CH2-), 2.21 (6H, m, 3CH3CH2CH2CH2), 2.50
(2H, m, -SCH2CH2-P), 2.63 (2H, m, CH3CH2S-), 2.81 (2H, m, -SCH2CH2-
P). 13C-NMR (101 MHz, CDCl3, [ppm]) δ = 13.20, 13.48, 14.23, 18.63,
19.09, 19.18, 19.64, 23.22, 23.26, 23.42, 23.47, 23.66, 23.81, 23.86,
24.01, 26.23, 118.21, 121.40.
[PhSCH2CH2-PBu3][Tf2N]. 1H-NMR (400 MHz, CDCl3, [ppm])
δ = 0.94 (9H, m, 3CH3CH2CH2CH2-), 1.43 (12H, m, 3CH3CH2CH2CH2-
), 2.16 (6H, m, 3CH3CH2CH2CH2), 2.47 (2H, m, -SCH2CH2-P), 3.18 (2H,
m, -SCH2CH2-P), 7.28–7.36 (5H, m, C6H5S-). 13C-NMR (101 MHz,
CDCl3, [ppm]) δ = 13.20, 13.62, 18.47, 18.94, 23.31, 23.36, 23.63,
23.67, 23.71, 23.79, 24.17, 24.32, 26.31, 26.36, 27.07, 27.72, 118.26,
121.46, 127.75, 129.24, 129.66, 130.52, 132.99.
2

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
Table 1
Enzymatic transesterification of ethyl sorbate with 1-propanola.
b
c
Trial
Solvent
Solvent water content (wt%)
Dynamic viscosity at 30 °C (mPa s)
Enzyme activity (μmol min−1 g‒1 CALB)
Selectivity
1
t-butanol
0.02
0.41
0.01
0.01
0.01
0.03
0.05
0.01
0.02
0.02
0.01
0.02
0.01
0.01
0.01
0.01
0.01
0.02
0.03
0.01
0.02
0.02
0.04
4.31 (25 °C) [39]
7.38
6.12
2.38
3.87
5.12
3.41
0.23
1.77
0.94
0.53
0.21
1.03
4.97
3.27
6.08
6.57
4.13
4.09
2.55
2.39
5.13
0.58
0.039
> 99%
> 99%
> 99%
86.6
2
t-butanol
‒
3
N,N-dimethylacetamide
[BMIM][PF6]
1.927 (25 °C) [39]
205.8
41.4
4
5
[BMIM][Tf2N]
> 99%
> 99%
> 99%
93.5
6
[BMIM][BF4]
85 [65]
26 [66]
122.5
‒
7
[BMIM][dca]
8
[CH3OCH2CH2-PBu3][Tf2N]
[CH3OCH2CH2-PBu3][Tf2N]
[CH3(OCH2CH2)2-PBu3][Tf2N]
[CH3(OCH2CH2)3-PBu3][Tf2N]
[CH3OCH2CH2-PBu3][beti]
[CH3OCH2CH2-PEt3][Tf2N]
[CH3OCH2CH2-Im-Et][Tf2N]
[CH3OCH2CH2-Py][Tf2N]
[CH3OCH2CH2-Et3N][Tf2N]
[CH3OCH2CH2-Pip-Me][Tf2N]
[CH3OCH2CH2-SMe2][Tf2N]
[CH3(OCH2CH2)2-SMe2][Tf2N]
[CH3(OCH2CH2)2-SMe2][Tf2N]
[CH3(OCH2CH2)3-SMe2][Tf2N]
[C6H5SCH2CH2-PBu3][Tf2N]
[CH3CH2SCH2CH2-PBu3][Tf2N]
9
> 99%
> 99%
98.3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
95.5
79.8
163.4
36.0
> 99%
> 99%
99.2
33.1
44.5
> 99%
> 99%
> 99%
90.4
61.4
84.2
d
47.1
‒
> 99%
> 99%
99.5
e
91.3
111.3
199.2
206.4
87.0
> 99%
Note: a The enzymatic reaction between ethyl sorbate (5 mM) and 1-propanol (0.67 M) was performed in 1.0 mL solvent catalyzed by 20 mg Novozym 435 at 50 °C. b
The water contents were determined by the coulometric Karl Fischer titration at 22 °C using Hydranal® Coulomat AG as the analyte. c The viscosity data at 30 °C were
d
e
acquired by an Anton Paar SVM 3000 viscometer. One literature reported 46.84 mPa s at 25 °C. One literature reported 33.10 mPa s at 25 °C.
2.3. Preparation of ether-functionalized sulfonium-based ILs
[ppm]) δ = 25.82, 44.23, 58.76, 65.11, 69.82, 70.09, 70.44, 71.54,
118.08, 121.27.
A literature method [35] was slightly modified. Bromoglycol (1.0
eq.) was added dropwise into an aqueous solution of sodium thio-
methoxide (21 wt%, 1.1 molar equiv.). The reaction mixture was cov-
ered by an aluminum foil and cooled in an ice bath for 24 h. Iodo-
methane (1.5 molar equiv.) was added dropwise to the reaction mixture
(containing 1.0 equiv. glycol thioether), and the biphasic mixture was
stirred at room temperature for 24 h. Into the reaction mixture, lithium
bis(trifluoromethylsulfonyl)imide (Li[Tf2N], 1.05 molar equiv.) dis-
solved in deionized water was added dropwise at room temperature. A
cloudy mixture formed instantly. The mixture was continuously stirred
for 20 min. After sitting for 2 h for the phase separation, the IL layer at
the bottom was separated and dissolved in dichloromethane, which was
thoroughly rinsed with deionized water three times. The silver nitrate
test of the aqueous layer indicated the absence of halides. The rinsed IL
was further washed by diethyl ether two times, and then dried in a
vacuum oven (25 mmHg, 80 °C) for seven days. 1H and 13C NMR con-
firmed the structure and purity of the ILs.
2.4. Enzymatic transesterification of ethyl sorbate with 1-propanol
In a typical setup, 50 μL of 1-propanol containing 100 mM ethyl
sorbate were mixed with 1.0 mL of IL in a capped glass vial. The final
concentrations of ethyl sorbate and 1-propanol were 5 mM and 0.67 M,
respectively. Following the addition of 20 mg Novozym 435, the reac-
tion mixture was gently agitated at 50 °C in an oil bath. Periodically
(every 15 min for the first hour), an aliquot (50 μL) of the mixture was
withdrawn and diluted with 1.0 mL methanol. After centrifuging for
2 min, the clear supernatant was injected into a HPLC. The concentra-
tion of propyl sorbate was calculated from its integrated area against
the standard curve for ethyl sorbate. As Novozym 435 can potentially
leach trace amounts of sorbic acid and sorbate ester [32], control ex-
periments were conducted without the addition of substrate (ethyl
sorbate) but with 50 μL of 1-propanol. Therefore, all activities reported
were the net activities after subtracting the control rates. LC-20AD
Shimadzu HPLC equipped with a SPD-20A UV–vis dual-wavelength
detector and an auto-sampler was used to analyze the reaction mix-
2-Methoxyethyl-dimethylsulfonium
bis(trifluoromethylsulfonyl)imide
([MeOCH2CH2-SMe2][Tf2N]). 1H-NMR (400 MHz, CDCl3, [ppm])
δ = 2.92 (6H, m, 2CH3S-), 3.90 (3H, m, CH3OCH2CH2S-), 3.51 (2H, m,
CH3OCH2CH2S-), 3.83 (2H, m, CH3OCH2CH2S-). 13C-NMR (101 MHz,
CDCl3, [ppm]) δ = 25.86, 44.19, 58.94, 66.43, 118.14, 121.32.
®
tures. The column employed was a Phenomenex Kinetex C18 column
(100 mm × 4.6 mm, particle size 2.6 mm). The flow rate was set at
1.0 mL min−1. The isocratic eluent consisted of 60 vol% methanol and
40 vol% water containing 1 vol% acetic acid; the UV detection wave-
length was set at 258 nm.
(2-Methoxyethoxy)ethyl-dimethylsulfonium
bis(trifluoromethylsulfonyl)
imide ([Me(OCH2CH2)2-SMe2][Tf2N]). 1H-NMR (400 MHz, CDCl3, [ppm])
δ = 2.95 (6H, m, 2CH3S-), 3.36 (3H, s, -SCH2CH2OCH2CH2OCH3), 3.54
(4H, m, -SCH2CH2OCH2CH2OCH3), 3.68 (2H, m, -SCH2CH2OCH2-
CH2OCH3), 3.97 (2H, m, -SCH2CH2OCH2CH2OCH3). 13C-NMR (101 MHz,
CDCl3, [ppm]) δ = 25.72, 26.10, 44.26, 58.61, 65.04, 70.38, 71.22,
118.09, 121.27.
2.5. Enzymatic ROP of ε-caprolactone
ε-Caprolactone (0.5 g; density of 1.03 g/mL) was mixed with
0.25 mL of solvent and 100 mg of Novozym 435 in a reaction vial. After
capping the viral, the reaction mixture was stirred (210 rpm) at 70 °C in
an oil bath. After 48 h, the mixture was removed from the oil bath and
allowed to cool to room temperature, followed by the addition of
2.0 mL of CDCl3 to dissolve the polyester under stirring. To analyze the
sample by 1H NMR, an aliquot (50 μL) was taken from the mixture and
diluted with 1.0 mL of CDCl3; the diluted sample mixture was cen-
trifuged. To analyze the sample by GPC, an aliquot (50 μL) was taken
(2-(2-Methoxyethoxy)ethoxy)ethyl-
dimethylsulfonium
bis(tri-
fluoromethylsulfonyl)imide ([Me(OCH2CH2)3-SMe2][Tf2N]). 1H-NMR
(400 MHz, CDCl3, [ppm]) δ = 2.95 (6H, s, 2CH3S-), 3.35 (3H, s, -SCH2-
CH2OCH2CH2OCH2CH2OCH3), 3.54 (4H, m, -SCH2CH2OCH2CH2-
OCH2CH2OCH3), 3.62 (4H, m, -SCH2CH2OCH2CH2OCH2CH2OCH3),
3.68 (2H, m, -SCH2CH2OCH2CH2OCH2CH2OCH3), 3.97 (2H, m,
-SCH2CH2OCH2CH2OCH2CH2OCH3). 13C-NMR (101 MHz, CDCl3,
3

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
from the reaction mixture and diluted with 1.0 mL THF, followed by a
centrifugation before the GPC injection. To collect the solid product,
chloroform was evaporated first and ice-cold methanol was added to
precipitate the polymer, followed by the separation of solid from liquid
using centrifugation or vacuum filtration. The polyester product was
dried in air for 24 h.
between ethyl sorbate (a flavoring ingredient) and 1-propanol (or other
sorbates with alcohols). This model reaction has several advantages
including: (1) due to the highly conjugated double bonds, the reactant
(ethyl sorbate), product (propyl sorbate), and byproduct (sorbic acid)
can be easily detected at micromolar concentrations (μM) by a UV
detector (e.g. at 258 nm) using the HPLC analysis; (2) the reaction is
moderately fast and thus exhibits linear reaction rates during the first
1–2 hours under normal biocatalysis temperatures (30–50 °C).
The mass of isolated/dried product was divided by the mass of ε-
caprolactone (0.5 g) to give the polyester yield. The ε-caprolactone
conversion was calculated from the 1H NMR spectrum (Bruker Avance
II 400 MHz NMR) by using the integrated peak areas of the methylene
groups next to the carbonyl group within ε-caprolactone (CL, 4.23 ppm)
and poly(ε-caprolactone) (PCL, 4.07 ppm) [36]. The mass-average
molecular mass (Mw) and polydispersity index (PDI = Mw/Mn) of the
polymer were measured by a GPC (LC-20AD Shimadzu HPLC) equipped
with SPD-20A UV–vis dual-wavelength detector operated at 210 nm,
and two Agilent PLgel MIXED-B (10 μm, 300 × 7.5 mm) columns eluted
with 1.0 mL/min THF at 30 °C [37]. The calibration curve was estab-
lished by polystyrene standards with 570 to 62,500 (Mw) [38].
Following this model reaction, we examined the initial reaction
rates in a number of organic solvents and ILs catalyzed by Novozym 435
(see Table 1). It is well known that Novozym 435 contains about 10 wt
% Candida antarctica lipase B (CALB) immobilized on the macroporous
acrylic beads [41,42]. Therefore, we calculated the initial reaction rate
(μmol min−1 g−1 CALB) based on the estimated CALB content (i.e.
∼2 mg free CALB in 20 mg Novozym 435) while many studies used the
total amount of CALB and acrylic beads to compute the reaction rate. In
our study, most solvents contained extremely low water contents (e.g.
0.01‒0.02 wt%). The reaction selectivity (synthesis vs hydrolysis) fa-
vored the transesterification reaction at low water contents (at least up
to 0.41 wt% water in t-butanol, see trial 2 in Table 1). Most reactions in
Table 1 achieved high selectivities of > 99%, and some variations in
the selectivity seemed not correlated with the low water content. The
initial rate in t-butanol with 0.02 wt% water (trial 1) was very high
(7.38 μmol min−1 g−1 CALB), which is used as a reference for com-
paring the enzyme activity (although unlike ILs, t-butanol may not
dissolve a variety of substrates and not be suitable for high-temperature
enzymatic applications; see section 3.3). The initial rate decreased by
17% when the water content in t-butanol increased to 0.41 wt% (trial
2). When the water content varied between 0.01 wt% and 0.03 wt%
(see trials 8 and 9, 19 and 20 in Table 1), there appeared no significant
change in reaction rates. Under the low water content for a specific
enzymatic reaction, the lipase activity mainly depends on the type of
solvents. N,N-Dimethylacetamide (DMA) was less compatible with the
enzyme (2.38 μmol min−1 g−1 CALB); three classic ILs [BMIM][PF6]
(3.87 μmol min−1 g−1 CALB), [BMIM][Tf2N] (5.12 μmol min−1 g−1
CALB), and [BMIM][BF4] (3.41 μmol min−1 g−1 CALB) were more li-
pase-activating than DMA. On the other hand, [BMIM][dca] showed a
low enzyme activity (trial 7 in Table 1); we also observed that [BMIM]
[dca] even dissolved the immobilized CALB including the acrylic beads.
It is well known that [BMIM][dca] is enzyme-denaturing due to the
strong hydrogen-bonding capability of dicyanamide anions (dca‒)
[29,31,43].
2.6. Fluorescence emission of free CALB in ILs
All fluorescence emission spectra were measured by a Hitachi F-
2500 fluorescence spectrophotometer. The lipase solution was excited
at 280 nm to observe the emission of tyrosine and tryptophan residues
at about 300 nm. A lipase (free CALB) solution (10 μL, 20 mg/mL in pH
7.5, 20 mM phosphate buffer) was added into 1.0 mL of solvent
(phosphate buffer, organic solvent, or IL) in a microcentrifuge tube. The
tube was turned upside down several times to dissolve or disperse the
aqueous enzyme in the solvent. The final lipase concentration was
0.2 mg/mL. The lipase solution was transferred to a quartz cuvette
(1.0 mL volume). Fluorescence measurements were made at room
temperature (22 °C) except that the measurement in t-butanol was de-
termined at 30 °C (due to its low melting point of 25.8 °C [39]). The
excitation monochromator slit was set to 5 nm, and the emission
monochromator slit width was also set at 5 nm. The PMT voltage was
controlled at 400 V and the response time was 0.08 s. The spectra were
scanned from 250 to 400 nm at a rate of 300 nm/min. The emission
spectrum of each solvent without the enzyme was also scanned at the
same condition, which was subtracted from the CALB emission spec-
trum in the same solvent.
3. Results and discussion
Ether-functionalized ILs displayed a wide range of biocompatibility
with the lipase CALB (trials 8–23 in Table 1). All tributyl phosphonium
ILs (trials 8–12 and 22–23) caused the lipase deactivation, resulting in
initial rates below 2.0 μmol min−1 g−1 CALB. This could be due to the
substrate ground-state stabilization by highly hydrophobic ILs carrying
tributyl groups, not necessarily due to the hydrophobic interaction
between tributyl groups and the protein leading to the enzyme in-
activation (see the fluorescence spectrum in section 3.4); less hydro-
phobic triethyl phosphonium (trial 13) is four-fold more lipase-acti-
vating than the tributyl phosphonium analogue (trial 8). Surprisingly,
longer glycol chains attached to tributyl phosphoniums (trials 8–11)
We evaluated the lipase CALB’s activities at two different tem-
peratures through two synthetic reactions: a transesterification at a
moderate temperature (50 °C), and a ring-opening polymerization at a
higher temperature (70 °C).
3.1. Transesterification activities in different solvents
In contrast to conventional lipase-catalyzed transesterification as-
says (such as ethyl butyrate with 1-butanol), our group [32,40] pre-
viously developed a more sensitive reaction assay (see Scheme 1)
Scheme 1. Lipase-catalyzed transesterification of ethyl sorbate with 1-propanol.
4

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
resulted in even lower enzyme activities: ([CH3OCH2CH2-PBu3]
[Tf2N] > [CH3(OCH2CH2)2-PBu3][Tf2N] > [CH3(OCH2CH2)3-PBu3]
[Tf2N]). However, this trend does not hold for the sulfonium ILs (trials
18–21), where the initial rates decreased in the order of
[CH3(OCH2CH2)3-SMe2][Tf2N] > [CH3OCH2CH2-SMe2]
longer glycol chains to sulfoniums (trials 13–15), or thioether groups to
phosphoniums (trials 16–17 vs 12) was unable to improve the mole-
cular mass of poly(ε-caprolactone), which is consistent with our earlier
findings where longer glycol chains were introduced to tributylpho-
sphoniums for the synthesis of polylactide and poly(ε-caprolactone)
[34]. The cationic headgroup of the IL imposed a significant impact on
the polyester's Mw: phosphonium (trial 3) > piperidinium (trial 9), al-
kylammonium (trial 8) > pyridinium (trial 7), sulfonium (trial 13, after
the enzyme batch conversion shown above) > imidazolium (trial 6).
Overall, the phosphonium-based [CH3OCH2CH2-PBu3][Tf2N] produced
a relatively high molecular mass of polylactone with a high yield. This
sequence of different cation cores is different from that concluded from
the transesterification reaction above, suggesting that different ILs have
profound effects on the lipase activation in different enzymatic reac-
tions; different substrate stabilization, IL stability and temperatures
could all be important factors. Polydispersity indexes (PDI) of poly(ε-
caprolactone) fall in the range of 1.2–2.0 (Table 2), which are typically
values for polyesters produced by enzymatic methods [46–50]. The
molecular mass distribution in enzymatic ROP reactions can be attrib-
uted to the relative rates of initiation to propagation, and the possible
chain transfer, termination, hydrolysis, and transesterification [50,51].
Due to such a complex reaction system, there is no apparent correlation
between PDI and the type of solvents (Table 2).
[Tf2N] > [CH3(OCH2CH2)2-SMe2][Tf2N]. Therefore, the role of glycol
chain is also cation-dependent. Different cation cores have drastic im-
pact on their lipase compatibility in a decreasing order of triethy-
lammonium (trial 16) > pyridinium (trial 15) > triethylphosphonium
(trial 13) > piperidinium (trial 17), sulfonium (trial 18) > imidazolium
(trial 14). In particular, high lipase activities at 50 °C were observed in
these ILs: [CH3OCH2CH2-Et3N][Tf2N] (6.57 μmol min−1 g−1 CALB),
[CH3OCH2CH2-Py][Tf2N]
(6.08 μmol
min−1 g−1
CALB),
[CH3(OCH2CH2)3-SMe2][Tf2N] (5.13 μmol min−1 g−1 CALB), [BMIM]
[Tf2N] (5.12 μmol min−1 g−1 CALB), and [CH3OCH2CH2-PEt3][Tf2N]
(4.97 μmol min−1 g−1 CALB). The inclusion of ether groups in IL
structures often leads to lower viscosities of the organic salts [31,44].
There is a poor correlation between the initial reaction rate and the
Tf2N‒-based-IL viscosity (see Table 1), implying that the viscosity might
play a role but not a determining factor.
3.2. High-temperature enzymatic polymerization of ε-caprolactone
The enzymatic ring-opening polymerization of ε-caprolactone was
conducted at a higher temperature (70 °C) (see Table 2). Dimethyla-
cetamide and ILs were used as co-solvents (0.25 mL with 0.5 g liquid ε-
caprolactone) because higher solvent contents (such as 0.5 and 1.0 mL)
usually led to lower molecular masses [34,45]. The polyester with high
molecular mass appeared not soluble in these co-solvents, and the re-
action mixture gradually solidified over time. Our earlier study sug-
gested that the enzymatic polymerization could be largely affected by
different batches of Novozym 435 and their water contents [45]. Higher
water contents in the enzyme and the solvent resulted in lower mole-
cular masses of polyesters [34,45]. Table 2 compared the data from two
different batches of Novozym 435 (SLBP0766V and SLBW1544). The
second enzyme batch (containing 0.68 wt% water as determined by
Karl-Fischer titration) produced higher molecular masses (Mw) than the
first batch (containing 1.09 wt% water) when comparing trial 10 and 2
(1.7 times higher) in [BMIM][PF6], and trial 12 and 3 (1.3 times higher)
in [CH3OCH2CH2-PBu3][Tf2N]. Therefore, a conversion factor (1.3–1.7)
should be considered when discussing data in trials 10–17 with trials
1–9 in Table 2. In general, the use of most co-solvents led to higher Mw
than the solventless condition (trial 1 in Table 2). The incorporation of
3.3. Thermal stability of Novozym 435
Fig.1 compares the thermal stability of Novozym 435 after its in-
cubation in t-butanol or [CH3OCH2CH2-NEt3][Tf2N] for 24 h (or 48 h)
at 50, 70 and 130 °C respectively. The remaining enzyme activity in t-
butanol after 24 h at 50 °C was 17% (of its initial activity shown as trial
1 in Table 1) and only 1% after 24 h at 70 °C. On the contrary, the
thermal stability of the lipase in [CH3OCH2CH2-NEt3][Tf2N] after 24 h
was much higher (86% at 50 °C, and 46% at 70 °C of its initial activity
shown as trial 16 in Table 1) although Novozym 435 lost most of its
activity after 24 h at 130 °C (1% remaining activity). A further in-
cubation of the lipase in [CH3OCH2CH2-NEt3][Tf2N] at 50 °C or 70 °C
for 48 h led no further significant change of its enzyme activity (81%
and 47% of the initial activity respectively). In conclusion, the thermal
stability of Novozym 435 in [CH3OCH2CH2-NEt3][Tf2N] is much higher
than that in t-butanol, which makes the use of ILs for biocatalysis more
advantageous than organic solvents like t-butanol. In general, ILs car-
rying anions of Tf2N−, PF6− and BF4− enable higher enzyme stabilities
than those polar (e.g. 1-butanol and 1-propanol) or even less polar
Table 2
Enzymatic ROP of ε-caprolactone under different reaction conditionsa.
Trial
Novozym 435 batch
Solvent (water content)
Conversion (%)
Yield (%)
Mw (Da)b
PDI
1
SLBP0766V
SLBP0766V
SLBP0766V
SLBP0766V
SLBP0766V
SLBP0766V
SLBP0766V
SLBP0766V
SLBP0766V
SLBW1544
SLBW1544
SLBW1544
SLBW1544
SLBW1544
SLBW1544
SLBW1544
SLBW1544
no solvent
96.8
87.8
96.9
52.1
70.1
95.7
94.3
96.4
96.1
98.6
99.0
99.8
99.0
60.0
71.2
94.0
95.9
37
55
48
15
30
42
32
11
11
38
54
56
65
33
37
56
55
13,800
18,500
18,900
9,100
1.71
1.94
1.59
1.19
1.35
1.60
1.34
1.39
1.69
1.55
1.75
1.53
1.89
1.35
1.46
1.73
1.65
2
[BMIM][PF6] (0.01 wt%)
3
[CH3OCH2CH2-PBu3][Tf2N] (0.02 wt%)
[CH3OCH2CH2-PBu3][beti] (0.02 wt%)
[CH3(OCH2CH2)2-PBu3][Tf2N] (0.02 wt%)
[MeOCH2CH2-Im-Et][Tf2N] (0.03 wt%)
[MeOCH2CH2-Py][Tf2N] (0.03 wt%)
[MeOCH2CH2-Et3N][Tf2N] (0.03 wt%)
[MeOCH2CH2-Pip-Me][Tf2N] (0.03 wt%)
[BMIM][PF6] (0.01 wt%)
4
5
14,300
12,300
16,200
17,300
18,100
31,400
29,100
25,400
21,700
14,700
19,000
21,500
14,800
6
7
8
9
10
11
12
13
14
15
16
17
Dimethylacetamide (0.01 wt%)
[CH3OCH2CH2-PBu3][Tf2N] (0.02 wt%)
[CH3OCH2CH2-SMe2][Tf2N] (0.01 wt%)
[CH3(OCH2CH2)2-SMe2][Tf2N] (0.02 wt%)
[CH3(OCH2CH2)3-SMe2][Tf2N] (0.02 wt%)
[C6H5SCH2CH2-PBu3][Tf2N] (0.02 wt%)
[CH3CH2SCH2CH2-PBu3][Tf2N] (0.03 wt%)
a
Note: general reaction conditions (unless otherwise noted): 0.5 g of ε-caprolactone, 0.25 mL of solvent, 100 mg of Novozym 435, gentle stirring (210 rpm) at
70 °C for 2 days. GPC-derived Mw values were based on results calibrated using polystyrene standards. Parts of the data (using Novozym 435 lot # SLBP0766V) were
b
published in our earlier paper [34]. Based on the GPC analysis.
5

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
300 nm (see Figs. S4‒S22 in the Supporting Information), particularly for
imidazoliums (Figs. S4‒S7, S17), piperidinium (Fig. S19), and multiple
alkoxy groups (Figs. S12, S13, and S20‒S22), which should all be
subtracted from the enzyme-in-IL spectra. Usually, the changes in the
maximal intensity of fluorescence and the red shift of the maximal
emission wavelength indicate the protein denaturation [62].
The fluorescence spectrum of CALB in phosphate buffer (pH 7.5,
20 mM) was used as a reference of the native protein structure. The
maximal intensity of fluorescence of CALB in t-butanol was higher than
that in buffer (Fig. 2a). Visually, we observed a slightly turbid mixture
formed; in turbid media, the light scattering and absorption often result
in weaker fluorescence intensity, not a stronger intensity [63]. One
possible explanation is that hydroxy groups of t-butanol interact with
the protein surface through hydrogen-boning to cause the hydrophobic
tyrosine and tryptophan residues to become more buried within the
core of the protein, resulting in higher fluorescence intensity. Coin-
cidentally, this corresponds to the highest CALB activity in Table 1.
Aqueous droplets of CALB in N,N-dimethylacetamide (DMA) formed a
clear solution, but the fluorescence maximal intensity was very weak
comparing with that in buffer (Fig. 2a and Fig. S3). The enzyme
quenching might explain the CALB activity in DMA was about 3-fold
lower than that in t-butanol (Table 1).
Fig. 1. Thermal stability of Novozym 435 in t-butanol and [MeOCH2CH2-Et3N]
[Tf2N]. Reaction conditions were as the following: 20 mg Novozym 435 in-
cubated in 1.0 mL t-butanol or [MeOCH2CH2-Et3N][Tf2N] at a fixed tempera-
ture for 24 or 48 h. After cooling to room temperature, ethyl sorbate (50 μL
100 mM in 1-propanol) was added and the reaction mixture was immersed in an
oil bath at 50 °C. The lipase activity was measured following the procedures in
section 2.4.
In hydrophobic ILs (containing anions of Tf2N−, PF6 or beti−),
−
aqueous droplets (10 μL, 20 mg/mL CALB) of CALB were dispersed into
fine droplets rather than fully solubilized in the media (1.0 mL);
therefore, the change in maximal intensity of fluorescence in hydro-
phobic ILs is not an absolute indication of protein denaturation. In the
case of even more hydrophobic [CH3OCH2CH2PBu3][beti] (Fig. S11),
the aqueous droplets of CALB were barely dispersed in the IL and no
CALB emission peak could be detected. On the other hand, the water
droplets of CALB become soluble in hydrophilic ILs, forming a ‘homo-
geneous’ solution. The addition of aqueous lipase droplets (10 μL) in-
stead of the enzyme power into the solvent was based on several con-
siderations: (1) after adding the droplets, the overall water content in
the solvent was 1% (v/v) which had minimum influence on the medium
environment; (2) water droplets can be better dispersed in hydrophobic
media (into tiny droplets) than the enzyme powder since the enzyme is
not soluble in hydrophobic solvents; (3) the final CALB concentration in
the solvent was 0.2 mg/mL; such a low concentration could be achieved
more uniformly and easily by adding the aqueous droplets than adding
the powder. As illustrated in Fig. 2a, the maximum emission occurred at
305 nm for CALB in [BMIM][Tf2N] (also in Fig. S4), and at 300 nm in
[BMIM][PF6] (Fig. S5), both of which showed blue shifts comparing
with that (313 nm) in buffer. However, no apparent emission peak was
seen for CALB in [BMIM][BF4] (Fig. S6) and [BMIM][dca] (Fig. S7).
This explained the relatively high lipase activities in [BMIM][Tf2N]and
[BMIM][PF6], but an extremely low activity in [BMIM][dca] (see
Table 1). It remains a mystery that fluorescence signal of CALB was
completely quenched in [BMIM][BF4] although the enzyme was still
active in this IL (Table 1). Fig. 2b compares the emission spectra of
CALB in [CH3OCH2CH2PBu3][Tf2N] and [CH3OCH2CH2PBu3]Br with
CALB in buffer. It is clear that CALB in [CH3OCH2CH2PBu3][Tf2N]
showed the characteristic peak at 314 nm, while the lipase in
[CH3OCH2CH2PBu3]Br showed a red-shifted peak at 325 nm, implying
the former IL is more protein-stabilizing than the bromide type. The
fluorescence intensity is lowest in [CH3OCH2CH2PBu3][Tf2N] due to
the poor miscibility of aqueous CALB in the IL (while aqueous CALB is
miscible in hydrophilic [CH3OCH2CH2PBu3]Br). Fig. 2c demonstrates
the influence of different cation cores on the emission signal of CALB.
Comparing with the emission peak of CALB in buffer at 313 nm, the red
shifts of emission wavelength in different ILs increase in the order of
organic solvents (e.g. n-hexane, dibutylether, and benzene), which can
be explained by the aprotic, non-coordinating, and neutral basicity nature
of the anions and their minimum perturbation of the protein (e.g. not
having strong hydrogen bonding with the enzyme) [52–56]. In addi-
tion, short-chain-ether/glycol-functionalized ILs could improve the
thermal stability of an enzyme even when the anion (e.g. acetate) was
basic and forming hydrogen bonds [31]. Therefore, [CH3OCH2CH2-
NEt3][Tf2N] has combined these two advantages, resulting in a high
thermal stability of Novozym 435. On the other hand, t-butanol is a
polar aprotic solvent and miscible with water; molecular dynamics si-
mulations suggest that t-butanol does not alert the substrate entrance
and the binding pocket size of CALB, and also maintains hydrogen
bonds between Ser105 and His224 residues (Ser105‒His224‒Asp187 as
the catalytic triad in CALB’s active site) [57], which explains the high
lipase activity in t-butanol (trial 1 in Table 1). However, at higher
temperatures (i.e. 50–70 °C), it appeared that CALB in Novozym 435
was detached from the polymer support over 24 h as we observed the
solution became cloudy; we suspect that this solvent could have a
strong interaction with the protein through hydrogen bonds, and/or
strip off water from the enzyme.
3.4. Fluorescence emission spectra of CALB in ILs
CALB consists of 317 amino acid residues with a molecular mass of
33,273 Da, and has a Ser105‒His224‒Asp187 catalytic triad in its ac-
tive site [58,59]. CALB has a high thermal stability, and may be active
under temperatures as high as 100 °C [60]. As fluorophores in the
protein, the tryptophan residues are selectively excited at 295 nm af-
fording the emission maximum in water near 350 nm; the excitation of
proteins at 280 nm leads to the energy absorption by both tryptophan
and tyrosine (the emission maximum of tyrosine in water is near
303 nm) [61]. Therefore, following the excitation of CALB in phosphate
buffer (pH 7.5, 20 mM) at 280 nm, the overall emission maximum of
tryptophan and tyrosine residues was observed at 313 nm, whose
fluorescence intensity was three times stronger than that being excited
at 295 nm (Fig. S1 in the Supporting Information). Therefore, this study
excited all other CALB samples at 280 nm, and all solvent emission
signals were subtracted (Figs. S2‒S22). The presence of water or t-bu-
tanol molecules in all samples gave a sharp peak of Raman scattering
band at ∼282 nm (interestingly, 0.2 mg/mL CALB in t-butanol led
to > 34 times increase in peak intensity at 281 nm, see Fig. S2(a)). It is
notable that many ILs themselves have strong emission signals at >
[CH3OCH2CH2PBu3][Tf2N]
(314 nm),
[CH3OCH2CH2PEt3][Tf2N]
(322 nm) < [CH3OCH2CH2-
(314 nm) < [CH3OCH2CH2SMe2][Tf2N]
Et3N][Tf2N]
(329 nm) < [CH3OCH2CH2Py][Tf2N]
(338 nm) < [
CH3OCH2CH2-Im-Et][Tf2N] (344 nm) < [CH3OCH2CH2-Pip-Me][Tf2N]
(no maximum). It is known that the solvent properties (such as polarity)
6

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
Fig. 2. Fluorescence emission spectra of CALB in different solvents (the solvent emission spectra were subtracted).
have a major impact on the tryptophan emission spectra and excited-
state ionization of tyrosine [61,64]. This order is not always consistent
with the CALB’s transesterification activities at 50 °C shown in Table 1:
[CH3OCH2CH2NEt3][Tf2N] > [CH3OCH2CH2Py][Tf2N] > [CH3OCH2-
protective role for the lipase.
4. Conclusions
CH2PEt3][Tf2N] > [CH3OCH2CH2-Pip-Me][Tf2N],
[CH3OCH2CH2-
Ether-functionalized ILs with different cation cores have been syn-
thesized and evaluated as solvents for two enzymatic reactions: the
transesterification of ethyl sorbate and 1-propanol, and the ring-
opening polymerization (ROP) of ε-caprolactone. Two ILs,
[CH3OCH2CH2-Et3N][Tf2N] and [CH3OCH2CH2-Py][Tf2N] afforded the
highest transesterification reaction rates (> 6.0 μmol min−1 g−1 CALB)
among all ILs studied. [CH3OCH2CH2-PBu3][Tf2N] produced the
highest molecular mass (Mw up to 25,400 Da) in the polymerization
reaction. We obtained different enzyme activity sequences for these two
reactions, implying that the lipase activity depends on both the solvent
and the specific reaction. The thermal stability of Novozym 435 in
[CH3OCH2CH2-Et3N][Tf2N] is much higher than that in t-butanol. The
fluorescence spectra of free CALB in different ILs allow a more in-
sightful view of the interactions between the lipase and the solvent. The
red shifts of emission wavelength in different ILs increase in the order of
SMe2][Tf2N] > [CH3OCH2CH2-Im-Et][Tf2N] > [CH3OCH2CH2PBu3]
[Tf2N]. The lipase activities at 70 °C for the ROP of ε-caprolactone (in
terms of Mw as discussed in Section 3.2) followed another sequence:
[CH3OCH2CH2PBu3][Tf2N] > [CH3OCH2CH2-Pip-Me][Tf2N] > [CH3O-
CH2CH2PEt3][Tf2N] > [CH3OCH2CH2Py][Tf2N], [CH3OCH2CH2SMe2]
[Tf2N] > [CH3OCH2CH2-Im-Et][Tf2N]. Therefore, the lipase activity is
also dependent on the specific substrates and different reaction condi-
tions.
Fig. 2d suggests that a longer ethoxy chain attached to phospho-
nium leads to diminishing characteristic emission peak, which corre-
lates well with lower CALB activities in [CH3(OCH2CH2)2PBu3][Tf2N]
and [CH3(OCH2CH2)3PBu3][Tf2N] (trials 10 and 11 respectively in
Table 1). The sulfonium ILs showed
a
different trend:
strong emission peak at
315 nm (Fig. S22), which correlates with a high CALB activity (trial 21
[CH3(OCH2CH2)3SMe2][Tf2N] showed
a
[CH3OCH2CH2PBu3][Tf2N]
(314 nm),
[CH3OCH2CH2PEt3][Tf2N]
(322 nm) < [CH3OCH2CH2-
in Table 1). Two thioether-substituted
phosphoniums
(314 nm) < [CH3OCH2CH2SMe2][Tf2N]
[PhSCH2CH2PBu3][Tf2N] and [CH3CH2SCH2CH2PBu3][Tf2N] showed
weak or no emission peak (Figs. 2d, S14 and S15), which explained the
very low Novozym 435 activities in Table 1 (trials 22 and 23 respec-
tively). However, the enzymatic ROP reaction in [PhSCH2CH2PBu3]
[Tf2N] still produced poly(ε-caprolactone) with Mw 21,500 Da and 56%
yield (trial 16 in Table 2), implying the substrate might provide some
Et3N][Tf2N] (329 nm) < [CH3OCH2CH2Py][Tf2N] (338 nm) < [CH3O-
CH2CH2-Im-Et][Tf2N] (344 nm) < [CH3OCH2CH2-Pip-Me][Tf2N] (no
maximum). The solvent properties (such as polarity) and/or the lipase-
solvent interactions could lead to the red shifts. Therefore, ether-func-
tionalized ILs can be custom-designed for individual biocatalytic reac-
tions.
7

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
Declarations of interest
None.
[24] C. Vafiadi, E. Topakas, V.R. Nahmias, C.B. Faulds, P. Christakopoulos, Feruloyl
esterase-catalysed synthesis of glycerol sinapate using ionic liquids mixtures, J.
Biotechnol. 139 (2009) 124–129, https://doi.org/10.1016/j.jbiotec.2008.08.008.
[25] H. Zhou, J. Chen, L. Ye, H. Lin, Y. Yuan, Enhanced performance of lipase-catalyzed
kinetic resolution of secondary alcohols in monoether-functionalized ionic liquids,
Bioresour. Technol. 102 (2011) 5562–5566, https://doi.org/10.1016/j.biortech.
2011.02.066.
Acknowledgement
[26] Y. Fan, X. Wang, L. Zhang, J. Li, L. Yang, P. Gao, Z. Zhou, Lipase-catalyzed synthesis
of biodiesel in a hydroxyl-functionalized ionic liquid, Chem. Eng. Res. Des. 132
(2018) 199–207, https://doi.org/10.1016/j.cherd.2018.01.020.
H. Z. acknowledges the startup fund support from the University of
Northern Colorado.
[27] Y. Fan, D. Cai, X. Wang, L. Yang, Ionic liquids: efficient media for the lipase-cata-
lyzed Michael addition, Molecules 23 (2018) 2154, https://doi.org/10.3390/
molecules23092154.
Appendix A. Supplementary data
[28] S. Kadotani, T. Nokami, T. Itoh, Enhanced activity and modified substrate-favor-
itism of Burkholderia cepacia lipase by the treatment with a pyridinium alkyl-PEG
sulfate ionic liquid, Tetrahedron 75 (2019) 441–447, https://doi.org/10.1016/j.tet.
2018.12.028.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.procbio.2019.03.018.
[29] H. Zhao, G.A. Baker, Z. Song, O. Olubajo, T. Crittle, D. Peters, Designing enzyme-
compatible ionic liquids that can dissolve carbohydrates, Green Chem. 10 (2008)
696–705, https://doi.org/10.1039/B801489B.
References
[30] S. Tang, G.A. Baker, S. Ravula, J.E. Jones, H. Zhao, PEG-functionalized ionic liquids
for cellulose dissolution and saccharification, Green Chem. 14 (2012) 2922–2932,
https://doi.org/10.1039/C2GC35631G.
[1] P.J. Halling, Biocatalysis in low-water media: understanding effects of reaction
conditions, Curr. Opin. Chem. Biol. 4 (2000) 74–80.
[2] S. Cantone, U. Hanefeld, A. Basso, Biocatalysis in non-conventional media—ionic
liquids, supercritical fluids and the gas phase, Green Chem. 9 (2007) 954–971,
https://doi.org/10.1039/B618893A.
[31] H. Zhao, C.L. Jones, J.V. Cowins, Lipase dissolution and stabilization in ether-
functionalized ionic liquids, Green Chem. 11 (2009) 1128–1138, https://doi.org/
10.1039/B905388C.
[3] J. Kadokawa, S. Kobayashi, Polymer synthesis by enzymatic catalysis, Curr. Opin.
Chem. Biol. 14 (2010) 145–153, https://doi.org/10.1016/j.cbpa.2009.11.020.
[4] A. Zaks, A.M. Klibanov, Enzymatic catalysis in nonaqueous solvents, J. Biol. Chem.
263 (1988) 3194–3201.
[32] H. Zhao, Z. Song, Migration of reactive trace compounds from Novozym® 435 into
organic solvents and ionic liquids, Biochem. Eng. J. 49 (2010) 113–118, https://
doi.org/10.1016/j.bej.2009.12.004.
[33] H. Zhao, Z. Song, O. Olubajo, High transesterification activities of immobilized
proteases in new ether-functionalized ionic liquids, Biotechnol. Lett. 32 (2010)
1109–1116, https://doi.org/10.1007/s10529-010-0262-4.
[5] A.M. Klibanov, Why are enzymes less active in organic solvents than in water?
Trends Biotechnol. 15 (1997) 97–101, https://doi.org/10.1016/S0167-7799(97)
01013-5.
[34] H. Zhao, L.O. Afriyie, N.E. Larm, G.A. Baker, Glycol-functionalized ionic liquids for
hight-emperature enzymatic ring-opening polymerization, RSC Adv. 8 (2018)
36025–36033, https://doi.org/10.1039/c8ra07733a.
[6] F. van Rantwijk, R.A. Sheldon, Biocatalysis in ionic liquids, Chem. Rev. 107 (2007)
2757–2785, https://doi.org/10.1021/cr050946x.
[7] Z. Yang, W. Pan, Ionic liquids: green solvents for nonaqueous biocatalysis, Enzyme
Microbial Technol. 37 (2005) 19–28, https://doi.org/10.1016/j.enzmictec.2005.
02.014.
[35] E. Coadou, P. Goodrich, A.R. Neale, L. Timperman, C. Hardacre, J. Jacquemin,
M. Anouti, Synthesis and thermophysical properties of ether-functionalized sulfo-
nium ionic liquids as potential electrolytes for electrochemical applications,
ChemPhysChem 17 (2016) 3992–4002, https://doi.org/10.1002/cphc.201600882.
[36] U. Piotrowska, M. Sobczak, E. Oledzka, C. Combes, Effect of ionic liquids on the
structural, thermal, and in vitro degradation properties of poly(ε-caprolactone)
synthesized in the presence of Candida antarctica lipase B, J. Appl. Polym. Sci. 133
(2016) 43728, https://doi.org/10.1002/app.43728.
[8] P. Domínguez de María, Z. Maugeri, Ionic liquids in biotransformations: from proof-
of-concept to emerging deep-eutectic-solvents, Curr. Opin. Chem. Biol. 15 (2011)
220–225, https://doi.org/10.1016/j.cbpa.2010.11.008.
[9] M. Moniruzzaman, K. Nakashima, N. Kamiya, M. Goto, Recent advances of enzy-
matic reactions in ionic liquids, Biochem. Eng. J. 48 (2010) 295–314, https://doi.
org/10.1016/j.bej.2009.10.002.
[37] M. Mena, A. López-Luna, K. Shirai, A. Tecante, M. Gimeno, E. Bárzana, Lipase-
catalyzed synthesis of hyperbranched poly-L-lactide in an ionic liquid, Bioprocess
Biosyst. Eng. 36 (2013) 383–387, https://doi.org/10.1007/s00449-012-0792-3.
[38] D. Omay, Y. Guvenilir, Synthesis and characterization of poly(D,L-lactic acid) via
enzymatic ring opening polymerization by using free and immobilized lipase,
Biocatal. Biotransform. 31 (2013) 132–140, https://doi.org/10.3109/10242422.
2013.795148.
[10] H. Zhao, Methods for stabilizing and activating enzymes in ionic liquids — a review,
J. Chem. Tech. Biotechnol. 85 (2010) 891–907, https://doi.org/10.1002/jctb.2375.
[11] H. Zhao, Protein stabilization and enzyme activation in ionic liquids: specific ion
effects, J. Chem. Technol. Biotechnol. 91 (2016) 25–50, https://doi.org/10.1002/
jctb.4837.
[12] Z. Guo, X. Xu, New opportunity for enzymatic modification of fats and oils with
industrial potentials, Org. Biomol. Chem. 3 (2005) 2615–2619, https://doi.org/10.
1039/B506763D.
[39] J.R. Rumble, CRC Handbook of Chemistry and Physics, 99th ed., CRC Press, Taylor
& Francis Group, New York, 2018.
[13] Z. Guo, B. Chen, R.L. Murillo, T. Tan, X. Xu, Functional dependency of structures of
ionic liquids: do substituents govern the selectivity of enzymatic glycerolysis? Org.
Biomol. Chem. 4 (2006) 2772–2776, https://doi.org/10.1039/b606900b.
[14] Z. Guo, X. Xu, Lipase-catalyzed glycerolysis of fats and oils in ionic liquids: a further
study on the reaction system, Green Chem. 8 (2006) 54–62, https://doi.org/10.
1039/B511117J.
[40] H. Zhao, G.A. Baker, S. Holmes, New eutectic ionic liquids for lipase activation and
enzymatic preparation of biodiesel, Org. Biomol. Chem. 9 (2011) 1908–1916,
https://doi.org/10.1039/c0ob01011a.
[41] F.C. Loeker, C.J. Duxbury, R. Kumar, W. Gao, R.A. Gross, S.M. Howdle, Enzyme-
catalyzed ring-opening polymerization of ε-caprolactone in supercritical carbon
dioxide, Macromolecules 37 (2004) 2450–2453, https://doi.org/10.1021/
ma0349884.
[15] B. Chen, Z. Guo, T. Tan, X. Xu, Structures of ionic liquids dictate the conversion and
selectivity of enzymatic glycerolysis: theoretical characterization by COSMO-RS,
Biotechnol. Bioeng. 99 (2008) 18–29, https://doi.org/10.1002/bit.21520.
[16] D. Kahveci, Z. Guo, B. Özçelik, X. Xu, Lipase-catalyzed glycerolysis in ionic liquids
directed towards diglyceride synthesis, Process Biochem. 44 (2009) 1358–1365,
https://doi.org/10.1016/j.procbio.2009.07.009.
[42] F. Deng, R.A. Gross, Ring-opening bulk polymerization of ε-caprolactone and tri-
methylene carbonate catalyzed by lipase Novozym 435, Int. J. Biol. Macromol. 25
(1999) 153–159, https://doi.org/10.1016/S0141-8130(99)00029-X.
[43] A.R. Toral, A.P. de los Ríos, F.J. Hernández, M.H.A. Janssen, R. Schoevaart, F. van
Rantwijk, R.A. Sheldon, Cross-linked Candida antarctica lipase B is active in dena-
turing ionic liquids, Enzyme Microb. Technol. 40 (2007) 1095–1099.
[44] S. Tang, G.A. Baker, H. Zhao, Ether- and alcohol-functionalized task-specific ionic
liquids: attractive properties and applications, Chem. Soc. Rev. 41 (2012)
4030–4066, https://doi.org/10.1039/C2CS15362A.
[17] Z. Guo, D. Kahveci, B. Özçelik, X. Xu, Improving enzymatic production of digly-
cerides by engineering binary ionic liquid medium system, New Biotechnol. 26
(2009) 37–43, https://doi.org/10.1016/j.nbt.2009.04.001.
[18] D. Kahveci, Z. Guo, B. Özçelik, X. Xu, Optimisation of enzymatic synthesis of dia-
cylglycerols in binary medium systems containing ionic liquids, Food Chem. 119
(2010) 880–885, https://doi.org/10.1016/j.foodchem.2009.07.040.
[19] T. De Diego, P. Lozano, M.A. Abad, K. Steffensky, M. Vaultier, J.L. Iborra, On the
nature of ionic liquids and their effects on lipases that catalyze ester synthesis, J.
Biotechnol. 140 (2009) 234–241, https://doi.org/10.1016/j.jbiotec.2009.01.012.
[20] G. de Gonzalo, I. Lavandera, K. Durchschein, D. Wurm, K. Faber, W. Kroutil,
Asymmetric biocatalytic reduction of ketones using hydroxy-functionalised water-
miscible ionic liquids as solvents, Tetrahedron Asymmetry 18 (2007) 2541–2546,
https://doi.org/10.1016/j.tetasy.2007.10.010.
[45] H. Zhao, G.A. Nathaniel, P.C. Merenini, Enzymatic ring-opening polymerization
(ROP) of lactides and lactone in ionic liquids and organic solvents: digging the
controlling factors, RSC Adv. 7 (2017) 48639–48648, https://doi.org/10.1039/
c7ra09038b.
[46] S. Kobayashi, K. Takeya, S. Sudu, H. Uyama, Lipase-catalyzed ring-opening poly-
merization of medium-size lactones to polyesters, Macromol. Chem. Phys. 199
(1998) 1729–1736.
[47] R. Marcilla, M. de Geus, D. Mecerreyes, C.J. Duxbury, C.E. Koning, A. Heise,
Enzymatic polyester synthesis in ionic liquids, Eur. Polym. J. 42 (2006) 1215–1221,
https://doi.org/10.1016/j.eurpolymj.2005.12.021.
[21] S. Dreyer, U. Kragl, Ionic liquids for aqueous two-phase extraction and stabilization
of enzymes, Biotechnol. Bioeng. 99 (2008) 1416–1424, https://doi.org/10.1002/
bit.21720.
[48] C. Wu, Z. Zhang, C. Chen, F. He, R. Zhuo, Synthesis of poly(ε-caprolactone) by an
immobilized lipase coated with ionic liquids in a solvent-free condition, Biotechnol.
Lett. 35 (2013) 1623–1630, https://doi.org/10.1007/s10529-013-1245-z.
[49] A. Duda, A. Kowalski, S. Penczek, H. Uyama, S. Kobayashi, Kinetics of the ring-
opening polymerization of 6-, 7-, 9-, 12-, 13-, 16-, and 17-membered lactones.
Comparison of chemical and enzymatic polymerizations, Macromolecules 35
(2002) 4266–4270, https://doi.org/10.1021/ma012207y.
[22] D. Das, A. Dasgupta, P.K. Das, Improved activity of horseradish peroxidase (HRP) in
‘specifically designed’ ionic liquid, Tetrahedron Lett. 48 (2007) 5635–5639,
https://doi.org/10.1016/j.tetlet.2007.06.022.
[23] Y. Abe, K. Kude, S. Hayase, M. Kawatsura, K. Tsunashima, T. Itoh, Design of
phosphonium ionic liquids for lipase-catalyzed transesterification, J. Mol. Catal., B
Enzym. 51 (2008) 81–85, https://doi.org/10.1016/j.molcatb.2007.11.010.
[50] A. Kumar, R.A. Gross, Candida antartica lipase B catalyzed polycaprolactone
8

H. Zhao, et al.
ProcessBiochemistryxxx(xxxx)xxx–xxx
synthesis: effects of organic media and temperature, Biomacromolecules 1 (2000)
T. Anthonsen, T.A. Jones, Crystallographic and molecular-modeling studies of li-
pase B from Candida antarctica reveal a stereospecificity pocket for secondary al-
cohols, Biochemistry (Mosc). 34 (1995) 16838–16851.
133–138, https://doi.org/10.1021/bm990510p.
[51] L.A. Henderson, Y.Y. Svirkin, R.A. Gross, D.L. Kaplan, G. Swift, Enzyme-catalyzed
polymerizations of ε-caprolactone: effects of initiator on product structure, propa-
gation kinetics, and mechanism, Macromolecules 29 (1996) 7759–7766.
[52] P. Lozano, T. De Diego, D. Carrié, M. Vaultier, J.L. Iborra, Over-stabilization of
Candida antarctica lipase B by ionic liquids in ester synthesis, Biotechnol. Lett. 23
(2001) 1529–1533.
[60] B. Réjasse, T. Besson, M.-D. Legoy, S. Lamare, Influence of microwave radiation on
free Candida antarctica lipase B activity and stability, Org. Biomol. Chem. 4 (2006)
3703–3707.
[61] J.R. Lakowicz, 16. Protein fluorescence, Principles of Fluorescence Spectroscopy,
3rd ed., Springer, New York, 2006.
[53] P. Lozano, T. de Diego, D. Carrie, M. Vaultier, J.L. Iborra, Enzymatic ester synthesis
in ionic liquids, J. Mol. Catalysis B: Enzym. 21 (2003) 9–13.
[62] P. Lozano, T. De Diego, J.L. Iborra, Dynamic structure/function relationships in the
α-chymotrypsin deactivation process by heat and pH, Eur. J. Biochem. 248 (1997)
80–85, https://doi.org/10.1111/j.1432-1033.1997.00080.x.
[54] P. Lozano, T. de Diego, J.-P. Guegan, M. Vaultier, J.L. Iborra, Stabilization of α-
chymotrypsin by ionic liquids in transesterification reactions, Biotechnol. Bioeng.
75 (2001) 563–569.
[63] Y. Chen, Z. Chen, J. Yang, J. Jin, J. Zhang, R. Yu, Quantitative fluorescence spec-
troscopy in turbid media: a practical solution to the problem of scattering and
absorption, Anal. Chem. 85 (2013) 2015–2020, https://doi.org/10.1021/
ac302815e.
[55] O. Ulbert, K. Bélafi-Bakó, K. Tonova, L. Gubicza, Thermal stability enhancement of
Candida rugosa lipase using ionic liquids, Biocatalysis Biotransform. 23 (2005)
177–183.
[64] A.S. Ladokhin, Fluorescence spectroscopy in peptide and protein analysis, in:
R.A. Meyers (Ed.), Encyclopedia of Analytical Chemistry, John Wiley & Sons Ltd,
Chichester, 2000, pp. 5762–5779.
[56] H. Noritomi, K. Minamisawa, R. Kamiya, S. Kato, Thermal stability of proteins in
the presence of aprotic ionic liquids, J. Biomed. Sci. Eng. 4 (2011) 94–99, https://
doi.org/10.4236/jbise.2011.42013.
[65] K.R. Harris, M. Kanakubo, L.A. Woolf, Temperature and pressure dependence of the
viscosity of the ionic liquid 1-butyl-3-methylimidazolium tetrafluoroborate: visc-
osity and density relationships in ionic liquids, J. Chem. Eng. Data 52 (2007)
2425–2430, https://doi.org/10.1021/je700370z.
[57] H.J. Park, K. Park, Y.J. Yoo, Understanding the effect of tert-butanol on Candida
antarctica lipase B using molecular dynamics simulations, Mol. Simul. 39 (2013)
653–659, https://doi.org/10.1080/08927022.2012.758850.
[58] J. Uppenberg, M.T. Hansen, S. Patkar, T.A. Jones, The sequence, crystal structure
determination and refinement of two crystal forms of lipase B from Candida ant-
arctica, Structure 2 (1994) 293–308.
[66] P.J. Carvalho, T. Regueira, L.M.N.B.F. Santos, J. Fernandez, J.A.P. Coutinho, Effect
of water on the viscosities and densities of 1-butyl-3-methylimidazolium dicyana-
mide and 1-butyl-3-methylimidazolium tricyanomethane at atmospheric pressure,
J. Chem. Eng. Data 55 (2010) 645–652, https://doi.org/10.1021/je900632q.
[59] J. Uppenberg, N. Oehrner, M. Norin, K. Hult, G.J. Kleywegt, S. Patkar, V. Waagen,
9