CO2 Reduction Promoted by Imidazole Supported on a Phosphonium-Type Ionic-Liquid-Modified Au Electrode at a Low Overpotential

Article abstract of DOI:10.1021/acscatal.7b03274

The catalytic conversion of CO₂ to useful compounds is of great importance from the viewpoint of global warming and development of alternatives to fossil fuels. Electrochemical reduction of CO₂ using aromatic N-heterocylic molecules is a promising research area. We describe a high performance electrochemical system for reducing CO₂ to formate, methanol, and CO using imidazole incorporated into a phosphonium-type ionic liquid-modified Au electrode, imidazole@IL/Au, at a low onset-potential of -0.32 V versus Ag/AgCl. This represents a significant improvement relative to the onset-potential obtained using a conventional Au electrode (-0.56 V). In the reduction carried out at -0.4 V, formate is mainly generated and methanol and CO are also generated with high efficiency at -0.6 ～ -0.8 V. The generation of methanol is confirmed by experiments using ¹³CO₂ to generate ¹³CH₃OH. To understand the reaction behavior of CO₂ reduction, we characterized the reactions by conducting potential- and time-dependent in situ attenuated total reflection surface-enhanced infrared absorption spectroscopy (SEIRAS) measurements in D₂O. During electrochemical CO₂ reduction at -0.8 V, the C-O stretching band for CDOD (or COD) increases and the C=O stretching band for COOD increases at -0.4 V. These findings indicate that CO₂ reduction intermediates, CDOD (or COD) and COOD, are formed, depending on the reduction potential, to convert CO₂ to methanol and formate, respectively.

Full text of DOI:10.1021/acscatal.7b03274

Subscriber access provided by READING UNIV
Article
CO2 Reduction Promoted by Imidazole Supported on a Phosphonium-
Type Ionic Liquid-Modified Au Electrode at a Low Overpotential
Go Iijima, Tatsuya Kitagawa, Akira Katayama, Tomohiko Inomata, Hitoshi Yamaguchi,
Kazunori Suzuki, Kazuki Hirata, Yoshimasa Hijikata, Miho Ito, and Hideki Masuda
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b03274 • Publication Date (Web): 24 Jan 2018
Downloaded from http://pubs.acs.org on January 24, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 24
ACS Catalysis
1
2
3
4
5
CO2 Reduction Promoted by Imidazole Supported on a
Phosphonium-Type Ionic Liquid-Modified Au Electrode at a Low
Overpotential
6
7
8
9
Go Iijima,† Tatsuya Kitagawa,¶ Akira Katayama,¶ Tomohiko Inomata,¶ Hitoshi Yamaguchi,† Kazunori
Suzuki,† Kazuki Hirata,† Yoshimasa Hijikata,† Miho Ito,† and Hideki Masuda¶*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
† Advanced Research and Innovation Center, DENSO CORPORATION, 500ꢀ1 Minamiyama, Komenokiꢀcho,
Nisshin 470ꢀ0111, Japan
¶ Graduate School of Engineering, Nagoya Institute of Technology, Gokisoꢀcho, Showa, Nagoya 466ꢀ8555,
Japan
*Corresponding email for H.M.: masuda.hideki@nitech.ac.jp
KEYWORDS: CO2 reduction; methanol generation; ionic liquidꢀmodified electrode; imidazole, SEIRAS
Abstract
The catalytic conversion of CO2 to useful compounds is of great importance from the viewpoint of global
warming and development of alternatives to fossil fuels. Electrochemical reduction of CO2 using aromatic
Nꢀheterocylic molecules is a promising research area. We describe a high performance electrochemical
system for reducing CO2 to formate, methanol and CO using imidazole incorporated into a
phosphoniumꢀtype ionic liquidꢀmodified Au electrode, imidazole@IL/Au, at a low onsetꢀpotential of –0.32
V vs. Ag/AgCl. This represents a significant improvement relative to the onsetꢀpotential obtained using a
conventional Au electrode (–0.56 V). In the reduction carried out at –0.4 V, formate is mainly generated,
and methanol and CO are also generated with high efficiency at –0.6 ~ –0.8 V. The generation of methanol
is confirmed by experiments using 13CO2 to generate 13CH3OH. To understand the reaction behavior of
CO2 reduction, we characterized the reactions by conducting potentialꢀ and timeꢀdependent inꢀsitu attenuated
total reflection surfaceꢀenhanced infrared absorption spectroscopy (SEIRAS) measurements in D2O.
During electrochemical CO2 reduction at –0.8 V, the C–O stretching band for CDOD (or COD) increases and
the C=O stretching band for COOD increases at –0.4 V. These findings indicate that CO2 reduction
intermediates, CDOD (or COD) and COOD, are formed, depending on the reduction potential, to convert
CO2 to methanol and formate, respectively.
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 24
1
2
Introduction
3
4
5
6
7
8
9
Conversion of CO2 to useful compounds and development of sustainable energy sources are subjects of
intense global research because CO2 is a major cause of global warming.1ꢀ3 Alternative sources of energy
will be required as fossil fuel sources that are exhausted in the future. It is envisioned that CO2 may be
useful for storing as alternative energy, such as solar, wind and other natural powers, by converting it to other
fuels. A number of methods for reducing CO2 electrochemically,4ꢀ8 photoꢀcatalytically9ꢀ14 and thermally15,16
have been proposed. Among such methods, efforts to promote electrochemical conversion of CO2 into
fuels such as CO, hydrocarbons and alcohol have attracted much attention and experienced the most progress.
Since Hori et al. reported the electrochemical conversion of CO2,4 many additional researchers have engaged
in electrochemical studies to further develop this field.5ꢀ14 However, almost all reduction systems
investigated thus far have a high overpotential and a low current efficiency and many of these systems
produce 2ꢀelectron reduced species such as formate and CO.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Recently, Bocarsly et al. have proposed a unique synthetic scheme for generating methanol from CO2
using aromatic Nꢀheterocycles (ANHs) such as pyridinium.17 In this case, the Faradic efficiency varies
within the range of 5 ꢀ 90% in a manner which depends on the type of ANH. Electrode materials containing
GaP as a semiconductor are poor in terms of reproducibility. Dyer et al. also studied CO2 reduction using
mercaptopteridine on a glassy carbon electrode to give formate, formaldehyde, and methanol.18 Although
the catalysts employed in these studies can reduce CO2, there are many problems such as the uncertainty of
the mechanism,19ꢀ28 the catalytic efficiencies29ꢀ31 and the conditions required for CO2 reduction.31 Most
notably, the reaction efficiency and catalytic mechanism have not been clarified in all cases. The current
efficiency is believed to be affected by the ability of the ANH to reach the electrode surface under
homogeneous conditions.23,24,32ꢀ34 Therefore, a method to improve access of ANH to the electrode to induce
CO2 reduction is required. JaiꢀXing Lu et al. have reported electrochemical CO2 reduction using a Pd
nanoparticle composite substituted with pyridine derivatives modified on the electrode.35,36 This system
reduces CO2 with high reactivity. However, some problems remain, including decomposition by the
cleavage of ether bonds in the pyridine derivatives under the electrochemical and aqueous conditions and an
inability to control entrapment of the pyridine derivatives. To develop an electrode surface appropriate for
close access by ANH, an appropriate electrode space is needed. Another problem is the uncertainty with
respect to reactivity of ANH with CO2 on the electrode surface. Bocarsly et al. proposed a proton shuttle
mechanism using pyridiniumꢀderived protons as the proton source.17 This scheme was later investigated
experimentally and theoretically by Bocarsly et al. and Batista et al.21,29 Carter et al. proposed a unique
process which includes two electron/hydrogen reductions of the γꢀposition carbon of pyridine on the basis of
theoretical calculations based on the relationship between the overpotential in CO2 reduction and the
reduction potential of ANH.19 Although several possible mechanisms were reported, evidence to confirm the
mechanism is lacking due to difficulties encountered in detecting intermediates using conventional analytical
2
ACS Paragon Plus Environment

Page 3 of 24
ACS Catalysis
1
2
methods.
3
4
5
6
7
8
9
Recently, we developed a unique Au electrode modified with a phosphoniumꢀtype ionic liquid (IL) as a
selfꢀassembled monolayer (IL/Au), which can entrap metal complexes within the vacant space generated
among the IL molecules, regardless of the net charge of the complexes and without forming any chemical
bonds.37 Using the IL/Au electrode, we previously succeeded in promoting fourꢀelectron reduction of
dioxygen by the diiron(II) complex entrapped in the vacant space between the IL molecules.38 This
arrangement could also be applied to retain a portion of the catalysts in close proximity to the electrode
surface at the nano level and to effectively promote electron transfer to the catalyst in the selfꢀassembled
monolayer. Ionic liquids are known to adsorb CO2 with high efficiency and stabilize the activated CO2
reduction intermediates.39,40 Use of the electrode system is expected to address problems in CO2 reduction.
In this study, we employed the IL/Au electrode incorporated with ANH (ANH@IL/Au) (Scheme 1), and
using ANH@IL/Au, we performed the electrochemical reduction of CO2. It is desirable to retain ANH on
the substrate without any direct binding to promote effective electron transfer from Au to ANH and/or CO2.
We evaluated the electrochemical reduction of CO2 and the reaction products using this ANH@IL/Au
electrode. Fortunately, we succeeded in reduction of CO2 to formate, methanol and CO in a high yield at a
low overpotential. We also studied the CO2ꢀreduction behavior on the ANH@IL/Au by using
potentialꢀdependent and timeꢀdependent inꢀsitu attenuated total reflection surfaceꢀenhanced infrared
absorption spectroscopy (ATRꢀSEIRAS). We propose a molecular mechanism which appears to be
responsible for inducing reduction of CO2 at a relatively low overpotential in our system.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Schematic illustration of the procedure for incorporation of ANH into the IL/Au electrode
Results and Discussion
Electrochemical CO2 reduction on bare Au and ANH@IL/Au electrodes
It was deemed appropriate to confirm that ANH@IL/Au electrodes are more effective than a bare Au
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 24
1
2
3
4
5
6
7
8
electrode in reduction of CO2 in terms of current efficiency and generation of products. Electrochemical
reduction of CO2 was studied using a series of ANH compounds through the following two typical electrodes
at pH 6.9; (a) electrochemical reduction of a homogenous solution containing 0.01 M ANH in 0.1 M NaClO4
aqueous solution using a bare Au electrode and (b) electrochemical reduction using the ANH@IL/Au
electrode in 0.1 M NaClO4 aqueous solution. The ANH compounds investigated in this study are imidazole,
1ꢀmethylimidazole, 2ꢀmethylimidazole, 1,2ꢀdimethylimidazole, pyridine, triazole, and pyrazole. In all
electrochemical reduction reactions, the main product was formate, which was detected by ion
chromatography, as shown in Figure 1. The current efficiency of formate generation using a bare Au
electrode was less than 5% in all other ANHs. On the other hand, in ANH@IL/Au systems,
imidazole@IL/Au showed the best performance for formate generation at all potentials studied (10 ꢀ 55%)
relative to other ANH@IL/Au electrodes. Fortunately, for imidazole@IL/Au, both formate and methanol
were detected and measured by gas chromatography as described below. Therefore, we decided to study
the electrochemical CO2 reduction using imidazole in detail.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Current efficiencies for formate produced by CO2 reduction in the presence of several ANH compounds. (a)
Bare Au electrode and (b) ANH@IL/Au electrode
Overpotential for CO2 reduction using bare Au and IL/Au with and without imidazole
We studied the electrochemical behavior of CO2 reduction using a bare Au electrode and an IL/Au
electrode with and without imidazole in a 0.1 M NaClO4 aqueous solution by cyclic voltammetry (CV) in the
range of 0.3 ~ –0.8 V vs. Ag/AgCl under CO2 (Figure 2) to confirm that imidazole plays a role in reducing
the overpotential for CO2 reduction. In the reaction systems without imidazole, both the bare Au electrode
and the IL/Au electrode systems have an onsetꢀpotential at –0.60 V, indicating that some extent of CO2
reduction occurs on the Au electrode surface at this potential. The onsetꢀpotential of CO2 reduction by the
bare Au electrode in the presence of imidazole was observed at –0.56 V vs. Ag/AgCl, which is comparable to
4
ACS Paragon Plus Environment

Page 5 of 24
ACS Catalysis
1
2
that of direct electrochemical CO2 reduction using a Pt electrode in the presence of imidazole (–0.51 V vs.
3
4
Ag/AgCl).17,29
Surprisingly, CV measurements using the IL/Au electrode with imidazole,
5
6
7
8
imidazole@IL/Au, under saturated CO2 indicate a catalytic current at –0.32 V vs. Ag/AgCl, as shown in
Figure 2. The onsetꢀpotential was not changed depending on the sweep rates (Figure S5). This
onsetꢀpotential is significantly lower than other onsetꢀpotentials reported previously.29.32 The significant
lowering of the onsetꢀpotential means significant lowering in the overpotential of CO2 reduction. These
facts clearly indicate that both imidazole and the modifiedꢀionic liquid contribute to lowering of the
overpotential of CO2 reduction. This clearly represents that such a low overpotential for CO2 reduction has
been observed.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to clarify the catalytic characteristics of imidazole for electrochemical CO2 reduction, we
compared the electrochemical CO2 reduction among imidazole and its derivatives (1ꢀmethylimidazole,
2ꢀmethylimidazole and 1,2ꢀdimethylimidazole). As described above (Figure 2), the onsetꢀpotential of CO2
reduction using the imidazole@IL/Au electrode has a significantly lower value, which is quite surprising in
comparison with other cases that indicate onsetꢀpotentials near –0.6 V vs. Ag/AgCl (Figure S6). The
cathodic current, when the bare Au electrode was used, showed a relatively large value for imidazole and all
its derivatives only under a CO2 atmosphere, as shown in Figure S6. This result indicates that imidazole
and its derivatives can promote CO2 reduction in a homogeneous solution.
In the case of
1-methylimidazole@IL/Au and 1,2-dimethylimidazole@IL/Au, any significant differences were not
observed between the CV measurement results under Ar and under CO2. On the other hand, the significant
cathodic current increasing were observed for the imidazole@IL/Au and the 2-methylimidazole@IL/Au
systems, indicating that NꢀH protons were important for CO2 reduction in IL/Au. This corresponds with the
suggestion suggested previously by Yan et al. that both the lone pair and the NꢀH proton of imidazole
contribute to selective CO2 reduction.41 In the comparison between the imidazole@IL/Au and the
2-methylimidazole@IL/Au systems, the bulky methyl group in 2ꢀmethylimidazole may be the reason why
the 2-methylimidazole@IL/Au electrode system did not give a low onsetꢀpotential. This consideration is
also supported by the finding that the imidazole@IL/Au and 2-methylimidazole@IL/Au systems produce
formate, methanol and CO with high current efficiency relative to other imidazole derivatives (Figure S7).
The results obtained for systems containing pyrazole and triazole also produce formate in good yield (Figure
1). Among all CV measurements, the most significant lowering in the overpotential of CO2 reduction was
measured for the imidazole@IL/Au electrode. This clearly demonstrates that imidazole has some
interactions, such as hydrogen bonding and van der Waals interactions, with CO2 on the electrode.
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 24
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. CV measurements under CO2 conditions; (a) bare Au, (b) bare Au in the presence of imidazole, (c) IL/Au,
(d) imidazole@IL/Au, as measured with a sweep rate of 10 mV/sec.
Controlled potential electrolysis measurements for CO2 reduction
To confirm the CO2 reduction, we conducted long running reactions using the imidazole@IL/Au
electrode. The controlled potential electrolysis measurements conducted using imidazole@IL/Au as a
working electrode were carried out at –0.4, –0.5, –0.6, –0.7, and –0.8 V vs. Ag/AgCl. The reaction
products were analyzed by gas chromatography, ion chromatography and gas chromatographyꢀmass
spectrometry (GCꢀMS) to identify the reaction products of CO2 reduction. As shown in the analytical
results obtained (Figures 3, S8 and S9), formation of methanol was clearly detected. Other product not to
be described in these figures was H2, which was main product under each condition. The current efficiency
of all reaction products was found to change in a manner which depends on the reaction potentials, as shown
in Figure S8, and the current efficiency for methanol was 26% at –0.7 V vs. Ag/AgCl. Although the
reaction using the bare Au electrode also afforded methanol, the current efficiency of CO2 reduction was less
than 1%. The comparison with the homogeneous reaction system demonstrates that this significant
enhancement in catalytic behavior has been achieved by utilizing the ionic liquid selfꢀassembled monolayer
supported with imidazole. It appears that electron transfer from the electrode to imidazole and/or CO2 has
been induced by their close contact on the electrode surface. To elucidate the source of methanol obtained,
we carried out the electrochemical reduction of CO2 using the imidazole@IL/Au electrode system in the
presence of 13CO2, as described in the previous reports.42,43 As expected, 13CH3OH was detected by GCꢀMS,
as shown in Figure S9.
6
ACS Paragon Plus Environment

Page 7 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
The durability of imidazole@IL/Au electrode was also evaluated by the plots of current density vs. time
and CV and XPS measurements before and after the electrolysis (Figure S10). Fortunately, the positive
shift in onsetꢀpotential and N1s peak originated from imidazole were detected even after electrolysis.
These findings clearly indicate the existence of imidazoles in the imidazole@IL/Au electrode system after
the reaction.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Current efficiencies for formate and methanol generated in the reaction at –0.8 V vs. Ag/AgCl using the
imidazole@IL/Au electrode and the bare Au electrode in the presence of imidazole (homogeneous system).
Spectroscopic observation of CO2 reduction on the imidazole@IL/Au surface
In order to investigate the electrochemical behavior of CO2 reduction on the ANH@IL/Au electrode, we
measured the inꢀsitu FTꢀIR spectrum of the ANH@IL/Au electrode system using potentialꢀdependent and
timeꢀdependent inꢀsitu attenuated total reflection surface enhanced infrared absorption spectroscopy
(ATRꢀSEIRAS).44 This technique characterizes the surface plasmon resonance behavior of the working
electrode. SEIRAS is a powerful method for observing the behavior of the molecules adsorbed on the
electrode surface. In SEIRAS measurements, the intensity in an IR band of a functional group for the
adsorbed molecule is enhanced 10ꢀ1000ꢀfold by the electric field at the electrode surface.44 Therefore,
characteristic vibration modes of the molecule adsorbed on the electrode surface can be observed. The
electrochemical behavior of the imidazole@IL/Au electrode is spectroscopically detectable using
potentialꢀdependent and timeꢀdependent inꢀsitu ATRꢀSEIRAS. As described above, the electrochemical
measurements using ANH@IL/Au electrodes fabricated with several ANH compounds have demonstrated
that the most efficient CO2 reduction reaction is induced by the imidazole@IL/Au electrode.
Potentialꢀdependent inꢀsitu ATRꢀSEIRAS measurements were performed in the range of 0.3 ~ −0.8 V vs.
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 24
1
2
3
4
5
6
7
8
Ag/AgCl against the working electrode in the 0.1 M NaClO4/D2O solution on the basis of a spectrum
measured at 0.3 V as background, and timeꢀdependent inꢀsitu ATRꢀSEIRAS at −0.4, −0.6, or −0.8 V vs.
Ag/AgCl during 600 seconds on the basis of the spectrum measured at 0 seconds as the background. D2O
was employed as the solvent to prevent significant peaks of CO2ꢀreduction intermediates from overlapping
with OHꢀdeformation vibration peaks of H2O. The reaction behavior of imidazole, CO2 and ionic liquid on
the electrode surfaces was monitored as described below.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(i) Reaction behavior of imidazole and CO2 on the bare Au electrode and on the IL/Au electrode as
evaluated by potential-dependent in situ ATR-SEIRAS
Figure 4(a) shows potential dependent in situ ATRꢀSEIRAS data obtained from the homogeneous system
using the bare Au electrode in the presence of 0.01 M imidazole in 0.1 M NaClO4/D2O solution under Ar and
CO2, and Figure 4(b) shows data obtained from the imidazole@IL/Au electrode in 0.1 M NaClO4/D2O
solution under Ar and CO2. In the homogeneous system, the IR band assignable to the NꢀD stretching
vibration, which was generated from HꢀD exchange in D2O solution, gives rise to a peak at 2920 cm–1,45
which was detected when it was swept in the negative potential region. This is more credible by detecting
the peak even when we measured in the same conditions using imidazoleꢀd4 in D2O solution (Figure S11(a)).
This peak has increased intensity under CO2. In contrast, the IR band assignable to D2O observed in the
range of 2400 ꢀ 2800 cm–1 is less intense under CO2 relative to its intensity under Ar. This may suggest that
the NꢀD(H) moiety of imidazole plays an important role in CO2 reduction. In the imidazole@IL/Au
system, some peaks observed in the range of 1900 – 2200 cm–1 are enhanced at negative potentials, while the
NꢀD stretching peak at 2920 cm–1 is less intense than it is in the homogeneous system. Another peak,
which was detected at 2100 cm–1 under both Ar and CO2, is thought to be attributed to the NꢀD stretching
mode of imidazole forming hydrogen bonds (NDꢀN), in place of the peak at 2920 cm–1 that was observed in
the homogeneous system.
The difference between imidazole@IL/Au and homogeneous systems is that for the former, the cationic
head groups of the modifiedꢀphosphonium type IL will be attracted and tilted towards the negativelyꢀcharged
electrode surface at negative potentials. As mentioned above, in SEIRAS, IR peak is enhanced by
resonance between vibration of compounds and Au nanoꢀparticles, in particular, vertical vibration against Au
surface is more enhanced. We measured SEIRAS using IL/Au electrode (Figures S11(b), (c)), and a peak
was detected at 2871 cmꢀ1, which was close to CH stretching vibration value of branched alkyl chain of IL
(2855 cmꢀ1, experimental section).46 From this finding, at an openꢀcircuit potential, the branched alkyl
chain of IL was thought to be distant from Au surface and had parallel vibration mode against Au. On the
other hand, by driving to negative potential, the peak at 2871cmꢀ1 was enhanced. It means that the IL tilted
towards the electrode surface enhanced the resonance between vertical vibration of alkyl chain and Au
nanoꢀparticles. As mentioned above, taking into consideration our proposal that the combination of the
8
ACS Paragon Plus Environment

Page 9 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
lone pair and the NꢀH proton of imidazole might be crucial points to reduce CO2, the peak at 2100 cm–1
(which is particularly enhanced under CO2 as well as the peak at 2920 cm–1 in the homogeneous system), is
attributable to the NꢀD stretching band of NDꢀN hydrogen bond between adjacent imidazoles prompted by
tilted IL at negative potentials (Scheme 2). From the compounds used for experiment, this assignment is
also probable. The following three compounds are attributable as a candidate of the peak; IL, CO2 (CO2
reduction intermediates) or imidazole. As described in the experimental section, Figures S1 and S3, IL
does not have any peaks around 2100 cm–1, so it could be ruled out of the candidates. Although CO2 or
CO2 reduction intermediate was also considered as origin, they could be out of candidates because of
detecting the peak at 2100 cmꢀ1 under not only CO2 but also Ar. As reported in other paper,45,47 NꢀD
stretching band of imidaozleꢀd1 forming hydrogen bonds (NDꢀN) was detected at 2132cmꢀ1, which was very
close to 2100 cmꢀ1. The hydrogen bonds formed between adjacent imidazoles could induce lowering of the
energy requirement for accepting an electron and releasing a proton, thereby resulting in lowering of the
onsetꢀpotential. The existence of imidazoles in the ionic liquidꢀmodified Au is supported by data obtained
by the XPS analysis (Figure S4); about 8 imidazole molecules are estimated to be contained in the vacant
space formed by the assembly of IL against one phosphonium atom. These imidazole molecules form
hydrogen bonds with each other in a previously suggested manner.48,49
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Interestingly, only under CO2, the peak at 1970 cm–1 is detected at negative potentials, which is assignable
to the CObridge vibration mode. Because, in general, vibration mode of CO on substrate has two patterns,
COatop and CObridge, where COatop means that C atom of CO is adsorbed on metal terminally and CObridge
means that C atom of CO is adsorbed on metal with 2ꢀhold bridge. They are observed in different
wavenumber regions as follows; the former is observed over 2000 cmꢀ1 and the latter is detected in the low
wavenumber region of 1800 ~ 2000 cmꢀ1.50ꢀ53 In this case, we detected the peak at 1970 cmꢀ1, so we
assigned the peak as CObridge
.
This mode is considered to be a stable intermediate among the CO2 reduction
intermediates.50ꢀ53 The schematic view of CO adsorption is shown in Scheme S1. In contrast, a previous
study reported that a CObridge stretching peak is not usually detected in the range of 1800 ꢀ 2200 cm–1 using
an Au electrode at negative potential.53 This peak is detectable when a Cu electrode is used.54 Therefore,
imidazole@IL/Au could reduce CO2 to stabilize CO as a CO2 reduction intermediate more efficiently than a
bare Au electrode.
In the reduction of CO2 using imidazole@IL/Au system, imidazole may contribute to the CO2 reduction
through supplying protons from water. Because SEIRAS using IL/Au without imidazole did not show any
D2O peak at any potentials, as shown in Figure S11(b). On the other hand, SEIRAS using
imidazole@IL/Au (Figure 4 (b)) gave large D2O peak, which was enhanced at negative potentials. So, this
observation clearly means that imidazole of imidazole@IL/Au attracted waters and conducted protons to Au
for CO2 reduction.
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 24
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Potentialꢀdependent in situ ATRꢀSEIRAS carried out in the range of 0.3 ~ –0.8 V vs. Ag/AgCl. (a) Spectra
obtained using the bare Au electrode in D2O solution containing 0.01 M imidazole (as homogeneous) under Ar or CO2,
(b) spectra obtained using the imidazole@IL/Au electrode under Ar or CO2, and (c) expanded version of spectra (b) in
the range of 1800 ~ 2250 cm–1 using the imidazole@IL/Au electrode under Ar or CO2.
10
ACS Paragon Plus Environment

Page 11 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Schematic illustration of forming hydrogen bonds among imidazoles accompanied by tilting of individual
molecules of the phosphonium type ionic liquid at negative potential of about –0.3 V vs. Ag/AgCl.
(ii) Reaction behavior of IL/Au as evaluated by time dependent in situ ATR-SEIRAS under Ar and CO2
conditions at –0.4, –0.6 and –0.8 V
To evaluate products treated at different potentials using imidazole@IL/Au, we carried out
timeꢀdependent in situ ATRꢀSEIRAS under Ar and CO2 conditions. The results are shown in Figure 5.
1
Under CO2, IR peaks in the range of 1200 ꢀ 1800 cm− increase in intensity at −0.4 V and peaks in the range
of 1300 ~ 1650 cmꢀ1 increase at −0.6 and −0.8 V. In these cases, significant spectral changes were not
observed in the upper range of 2500 cm–1 (Figure S12). Similarly, spectral changes were not observed for
the bare Au electrode in 0.1M NaClO4/D2O solution containing 0.01 M imidazole and for the IL/Au
electrode in 0.1M NaClO4/D2O solution (Figure S12). These results strongly suggest that the IR peaks
observed using imidazole@IL/Au under CO2 are attributed to intermediates or products obtained from CO2
reduction. Also, the IR peaks attributable to the CObridge mode and the NDꢀN hydrogen bonds of imidazoles
were not detected. This differs from the results obtained from the potentialꢀdependent in situ ATRꢀSEIRAS
measurements and indicates that the environments of the imidazole and CObridge molecules on the Au surface
were not changed during the reaction time of 600 seconds. We can infer from this result that imidazole and
CO are continually utilized to reduce CO2 at each potential. Our proposal for the reaction behavior of
imidazole, CO2 and modifiedꢀionic liquid on Au electrode at each potential is as follows.
First, the reaction at –0.4 V vs. Ag/AgCl gives rise to peaks assignable to vibration modes attributable to
1
COOD as a CO2 reduction intermediate; IR bands at 1740, 1371 and 1212 cm− are assignable to C=O
55,56
stretching, CꢀO stretching and ODꢀdeformation of COOD, respectively.
These peaks were also
assigned56 by Firet et al. in the electrochemical reaction of CO2 on an Ag electrode. However, a
onsetꢀpotential higher than −1.4 V vs. Ag/AgCl was required to detect them. This demonstrates that our
system is able to reduce CO2 with low activation energy.
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 24
1
2
3
4
5
6
7
8
Next, the reaction at –0.6 ~ –0.8 V vs. Ag/AgCl gives rise to IR peaks observed at 1331 and 1421 cm–1 as
vibration modes attributed to CO2 reduction intermediates as well. These peaks are respectively assignable
to CꢀO stretching of DCOD or COD57ꢀ62 and a symmetric stretching vibration mode of COO–.56,63ꢀ65 The
former is also confirmed from the observation that methanol is generated as detected by the controlled
potential electrolysis. Other peaks are assignable to products from CO2 reduction; IR bands at 1366, 1470
and 1590 cm–1 are respectively assignable to C=O stretching vibrations of carbonate, adsorbed carbonate and
an asymmetric stretching vibration of C=O of DCOO–.63,65,66 Generation of carbonate demonstrates
effective reduction of water and production of OH–, which continuously convert CO2 to carbonate.65 The
peak of DCOO– appearing at 1590 cm–1 is usually combined with CꢀO stretching at 1354 cm–1.
Unfortunately, the IR peak at 1354 cm–1 overlaps with the IR band of adsorbed carbonate in our system, but
it could be attributed to DCOO– from the observation that formate is detected by controlled potential
electrolysis at −0.6 ~ −0.8 V.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As mentioned above, the results of SEIRAS also showed durability of imidazole@IL/Au against aqueous
solution. Solubility of imidazole into aqueous solution and destruction of NDꢀN hydrogen bond are
scientifically concerned using imidazole@IL/Au. However, as shown in the wide range spectra using the
time dependent in situ SEIRAS (Figure S12(c)), there was neither change in the peaks attributable to
imidazole (2300 ~ 3000 cmꢀ1) nor that to D2O (2400 ~ 2800 cmꢀ1). Furthermore, Figures 4(a) and (b) show
strong interaction between imidazole and Au substrate at negative potentials, which is opposite behavior to
diffusion of imidazole into aqueous solution.
12
ACS Paragon Plus Environment

Page 13 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Time dependent in situ SEIRAS carried out at (a) –0.4 V, (b) –0.6 V and (c) –0.8 V vs. Ag/AgCl using the
imidazole@IL/Au electrode under Ar or CO2.
Mechanism of CO2-reduction at the respective potentials
So far, two pathways for producing formate from CO2 have been proposed in a previous report.67 One of
these pathways uses an indium substrate. It was proposed that formate is generated through
oxygenꢀterminated formation of a metalꢀcarbonate bond on the indium substrate. The second pathway uses
a bismuth substrate. This pathway includes carbonꢀterminated formation of a metalꢀCOOH bond on the
bismuth substrate. In our system, COOD was detected at lower potential of −0.4 V, which leads us to
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 24
1
2
3
4
5
6
7
8
propose the following mechanism: CO2 is reduced to formate via metalꢀCOOH(D) with a “protonꢀcoupled
electron transfer reaction (PCET)” as proposed previously.56,68 The PCET reaction is induced by the
activated imidazole and ionic liquid. Considering that the reaction proceeds at a relatively low potential of
−0.4 V, the free energy of COOH(D) bound to Au is believed to be lower in comparison with those at high
potential,61 which makes reduction unfavorable. Fang et al. also reported CO2 reduction to formate using
an Au electrode functionalized by 4ꢀpyridylethylmercaptan,69 which showed a high Faradaic efficiency to
produce formate relative to bare Au. However, the onsetꢀpotential of reducing CO2 has a higher potential at
approximately −0.98 V vs. Ag/AgCl, in spite of the functionalization of Au with 4ꢀpyridylethylmercaptan.
Therefore, in our system, the detection of formate at low potential is quite a surprising result in view of the
previous work.69 When the reaction is conducted at −0.6 ~ −0.8 V, CO2 is first reduced by one electron to
form COO–. Furthermore, with high negative potential, COO– can be reduced to other intermediates and
methanol. Interestingly, our system promotes additional reduction to methanol, although the Au substrate
has been known to only reduce CO2 to CO.4,70ꢀ73 The reduction behavior of CO2 bound to Au is explained
by the weak binding energy of CO against Au, which has weaker binding energy than Pt and Cu.72 In our
system, the binding energy of CO against Au is assumed to be comparatively weak as well, but it is possible
to reduce CO because of the efficient supply of protons provided by imidazole molecules and the effect of
the ionic liquid to stabilize the CO2 reduction intermediate.39,73,74 The stabilization of CO2 reduction
intermediate by ionic liquid was described by Chen et al., and they theoretically and experimentally
demonstrated that the cationic parts of ionic liquid have a substantial stabilization effect for the activated
CO2.73 Many CO2 reduction intermediates possess a dipole moment or a polarizability, which would be
stabilized by an electric field of the cationic parts. The detection of methanol using ILꢀmodified Au
substrate in our system is good consistent with the previously reported works.39,73,74
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
On the basis of the above results, we can propose a mechanism for reduction of CO2 to formate and
methanol. As described above, the reduction intermediates obtained at several potentials, such as CO (1970
cm–1 at −0.4 ~ −0.8 V), COOD (1740 cm–1 at −0.4 V), COO– (1421 cm–1 at −0.8 V) and DCOD (or COD)
(1331 cm–1 at –0.8 V), were observed depending on the specific operating potential. These species are
converted to formate or methanol by the controlled potential electrolysis. These observations were
accompanied by an increase in intensity of the IR peak at 2100 cm–1 which is assignable to the NꢀD
stretching vibration of imidazole molecules forming hydrogen bonds with each other. We propose the
reaction scheme illustrated in Scheme 3. In the first step, imidazole molecules are incorporated into the
selfꢀassembled ionic liquid monolayer of the imidazole@IL/Au electrode. In the next step, which is driven
by sweeping of the electrode potential to a negative region, the cationic head groups of the phosphonium
type ionic liquid are attracted towards the negativelyꢀcharged electrode. The cationic head groups retain the
imidazole molecules and induce formation of hydrogen bonds between adjacent imidazole molecules.
These interactions between the imidazole molecules contribute to the release of protons from the imidazole
14
ACS Paragon Plus Environment

Page 15 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
molecules which are used to reduce CO2 selectively. In the third step, the CO2 reduction intermediates are
stabilized by forming an ion pair with the tilted cationic ionic liquid, as was observed by in situ
ATRꢀSEIRAS. Interestingly, at low potential, the tilting of ionic liquid molecules to constrain imidazole
molecules and induce formation of hydrogen bonds between the induces the reduction of CO2 in the PCET
reaction.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Schematic illustration of the possible reactions among imidazole, CO2 and the phosphonium type ionic
liquid on the Au electrode. (i) Imidazole molecules are incorporated into the IL/Au electrode to form
imidazole@IL/Au, (ii) when induced to negative potentials, ILs are tilted toward the electrode, which promote the
formation of hydrogen bonds among imidazole molecules, and (iii) CO2 is reduced on Au surface with supply of
protons from imidazoles to begin to generate formate and methanol at ꢀ0.4 and ꢀ0.8 V, respectively.
Experimental
Synthesis of the phosphonium type ionic liquid
The phosphoniumꢀtype ionic liquid (IL) was synthesized by the previous method37 and purified by silica
gel chromatography. The preparation of IL was confirmed by NMR spectroscopy and FTꢀIR spectroscopy.
Preparation of the ionic liquid with the disulfide group
The procedures are as follows (see also the summary in Scheme S2).
Preparation of 12-bromododecyl(trihexyl)phosphonium bromide
1,12ꢀdibromododecane (18.3 mmol, 6004.7 mg) was completely dissolved in toluene (60 mL) under Ar.
A toluene solution (60 mL) of trihexylphosphine (18.3 mmol, 5242.5 mg) was added dropwise to this
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 24
1
2
3
4
5
6
7
8
solution over the course of 2 h. The resultant solution was stirred for 68 h under Ar at room temperature,
and then the reaction mixture was evaporated. The resulting highly viscous yellow oil was purified by
silica gel column chromatography (eluent: AcOEt/MeOH = 9/1). After evaporation to dryness, a viscous
yellow oil was obtained. Yield 46.3%. 1HꢀNMR (300 MHz, CDCl3, vs. TMS): 0.90 (m, 9H, ꢀCH3), 1.20 ꢀ
2.30 (m, 44H, ꢀCH2ꢀ), 2.45 (m, 8H, ꢀP+CH2ꢀ), 3.42 (t, 2H, BrCH2ꢀ). FTꢀIR (ATR(Ge), cm–1): 2954, 2926,
2855 (ν(CꢀH) of methylene and methyl groups), 1457 (δ(CꢀH) of methylene group). ESI TOFꢀMS (positive
mode): m/z = 535.5 [M – Br]+.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of 12-thiouroniumdodecyl(trihexyl)phosphonium dibromide
Thiourea (4.05 mmol, 308.3 mg) and 12ꢀbromododecyl(trihexyl)phosphonium bromide (3.30 mmol,
2031.5 mg) were dissolved in EtOH (50 mL) and refluxed for 2 days at 90 ºC. The resultant solution was
evaporated completely. The residue was dissolved in CHCl3 (200 mL) and washed with H2O (100 mL x 3)
and brine (100 mL). After drying over anhydrous MgSO4, a viscous pale yellow liquid was obtained by
evaporation. This liquid was purified by silica gel column chromatography (eluent: CHCl3/MeOH (9/1
v/v)). After evaporation to dryness, a viscous yellow liquid was obtained. Yield 33.0%. 1HꢀNMR (300
MHz, CDCl3, vs. TMS): 0.90 (m, 9H, ꢀCH3), 1.25ꢀ1.84 (m, 44H, ꢀCH2ꢀ), 2.31ꢀ2.58 (m, 8H, ꢀP+CH2ꢀ), 3.39 (t,
2H, BrCH2ꢀ), 9.00 (br, 4H, ꢀC+NH2). FTꢀIR (ATR (Ge), cm–1): 2956, 2923, 2853 (ν(CꢀH) of methylene and
methyl groups), 1652 (δ(NꢀH) of amino group), 1458 (δ(CꢀH) of methylene group), 743 (ν(CꢀS)). ESI
TOFꢀMS (positive mode): m/z = 529.7 [M – H – 2Br]+, 265.3 [M – 2Br]2+.
Preparation of bis[12-(trihexylphosphonium)dodecyl]disulphane dibromide
12ꢀThiouroniumdodecylꢀ(trihexyl)phosphonium dibromide (1.03 mmol, 712.5 mg) was dissolved in EtOH
(30 mL). An aqueous solution (10 mL) of NaOH (2.20 mmol, 88 mg) was added to this solution. The
resultant solution was stirred for 4 days at room temperature. After evaporation, the residue was dissolved
in CHCl3 (80 mL) and washed with H2O (50 mL x 3) and brine (50 mL). After drying over anhydrous
MgSO4, the solution was evaporated completely. The resulting yellow viscous liquid was purified by silica
gel chromatography (eluent: AcOEt/MeOH 9/1(v/v) (1st) and CHCl3/MeOH (9/1(v/v) (2nd)). After
evaporation to dryness, a yellow viscous liquid was obtained. Yield 76.5%. 1HꢀNMR (300 MHz, CDCl3,
vs. TMS): 0.90 (m, 18H, ꢀCH3), 1.25ꢀ1.71 (m, 88H, ꢀCH2ꢀ), 2.45 (m, 16H, ꢀP+CH2ꢀ), 2.68 (t, 4H, ꢀSCH2ꢀ).
FTꢀIR (ATR (Ge), cm–1): 2955, 2926, 2856 (ν(CꢀH) of methylene and methyl groups), 1465 (δ(CꢀH) of
methylene group), 748 (ν(CꢀS)). ESI TOFꢀMS (positive mode): m/z = 1054.0 [M – Br]+, 486.5 [M –
2Br]2+.
Preparation of bis[12-(trihexylphosphonium)dodecyl]disulfane bis(trifluoromethanesulfonate) (1)
Bis[12ꢀ(trihexylphosphonium)dodecyl]disulphane dibromide (0.394 mmol, 446.6 mg) was dissolved in
16
ACS Paragon Plus Environment

Page 17 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
CHCl3 (25 mL). To the solution, was added EtOH solution (15 mL) of NaTfO (0.860 mmol, 148.0 mg).
The resultant mixture was stirred at room temperature until the TLC spot of the starting material disappeared
(eluent: CHCl3/MeOH = 4/1, Rf = 0.7). After removal of the white precipitate (NaBr), the solution was
evaporated completely. The residue was dissolved in CHCl3 (100 mL) and washed with H2O (100 mL x 3).
After drying over anhydrous MgSO4, the solution was evaporated completely. A pale yellow viscous liquid
was obtained. Yield 82 %. 1HꢀNMR (300 MHz, CDCl3, vs. TMS): 0.90 (m, 18H, ꢀCH3), 1.25ꢀ1.69 (m,
88H, ꢀCH2ꢀ), 2.20 (m, 16H, ꢀP+CH2ꢀ), 2.69 (t, 4H, ꢀSCH2ꢀ). FTꢀIR (ATR (Ge), cm–1): 2957, 2927, 2855
(ν(CꢀH) of methylene and methyl groups), 1465 (δ(CꢀH) of methylene group), 1260 (ν(CꢀF) of TfO–), 1153
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
–
sulfonate (ν(SO3 ) of TfO–), 1030 (ν(S=O) of TfO–), 748 (ν(CꢀS)). ESI TOFꢀMS (positive mode): m/z =
1121.7 [M – TfO–]+, 486.3 [M – 2 TfO–]2+ (Figure S1).
Preparation of the IL-modified Au (IL/Au) electrode and incorporation of imidazole into the IL/Au
electrode
Preparation of IL/Au
Au film with a thickness of 2,000 Å was deposited onto a mica substrate with 14 × 14 mm at 1.0 Å s–1
with a JISꢀ300AK vacuum coater (Sinku of Technology Co., Ltd.). The Au film was annealed with a
hydrogen gas flame before it was dipped into each sample solution. The IL/Au electrode was prepared
according to the previously established procedure. The Au electrode modified by IL (IL/Au), was prepared
by direct casting of IL onto the Au electrode with addition of a few drops over the course of a few days.
After washing with EtOH and CHCl3, IL/Au was obtained. The quantity of IL molecules modified on
Au(111) was evaluated by electrochemical desorption,75 which was scanned from 0.2 V to −1.4 V vs.
Ag/AgCl in the 0.5M KOH solution. A reduction potential peak was observed at −1.0 V (Figure S2), and
the electric charge consumed by the reductive desorption of IL was calculated to be 34.6 ꢀ 62.2 ꢁC⋅cmꢀ2. The
calculated amount of IL modified on Au(111) is 8.3 x 10ꢀ10 mol/cm2.
Preparation of ANH@IL/Au
The ANH@IL/Au electrode, which was treated with both ionic liquid and ANH for immobilization of the
compounds onto the Au substrate, was obtained by immersing IL/Au into Milli Q water containing 0.01 M
ANH for 7 days. The incorporation of ANH (imidazole derivatives, pyridine, pyrazole, and triazole) into
IL/Au was carried out by immersing the IL/Au substrate in an aqueous solution (pH 6.9) containing 0.01 M
imidazole or other ANH compounds for a few days. Characterization of imidazole@IL/Au was performed
by RASꢀIR measurements (Figure S3), and the presence of imidazole held in the IL/Au cavity was
confirmed by detecting an NꢀC stretching vibration peak at 1450 cmꢀ1. Then, the amount of imidazole
incorporated into IL/Au was also determined by low temperature XPS measurements (Figure S4). At room
temperature, imidazole was evaporated under high vacuum in an XPS analysis chamber. There was no
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 24
1
2
3
4
5
6
7
8
signal indicating the imidazole nitrogen atom. Therefore, we repeated the analysis at low temperature (−
120 ℃), and detected signals from the imidazole nitrogen atom at 400 eV. From this measurement, it was
determined that there are 1~8 molecules per IL molecule, which is estimated to be 8.3 x 10ꢀ10 ~ 6.6 x 10ꢀ9
mol/cm2 on the surface, as estimated by comparing the peak area of the nitrogen atom with that of the
phosphorus atom of IL.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Apparatus
1H NMR spectra were measured on a Varian Gemini 300 MHz NMR spectrometer in CDCl3 or D2O with
TMS or DSS, respectively, as internal reference standards. The concentration of each sample was adjusted
to 10 mM in a sample tube (5 mmφ). Fourier transform infrared (FTꢀIR) spectra were recorded using a
JASCO FT/IRꢀ4200 spectrometer. Attenuated total reflection spectra were measured using a JASCO ATR
PRO410ꢀS spectrometer. Reflectionꢀadsorption spectra were measured using a JASCO RASꢀPRO410ꢀH
spectrometer. Gasꢀchromatography massꢀspectra (GCꢀMS) were obtained with a JEOL JMSꢀQ1050GC
instrument with an Inert Cap FFAP column. Gas chromatography with a flame ionization detector was
performed using a Shimadzu GCꢀ2014 instrument with an Agilent Inc. BꢀWAXETR column. Gas
chromatography with a thermal conductivity detector was performed using an Agilent Technologies 7890A
instrument with an Agilent Inc. TCꢀMolsieve 5A column. Ionic chromatography was performed using a
Thermo Scientific DIONEX IC20 Ion system with a Thermo Scientific Ion Pac AG22 column. XꢀRay
photoelectron spectra (XPS spectra), were measured using an ULVACꢀPHI XPS PHI5000 system with Al Kα
radiation (1486.6 eV). Samples were held in the analyzing holder and inserted into the front chamber.
Evacuation of the chamber was started after the holder’s temperature became less than –50°C. The analysis
was performed at –120°C after inserting the samples into the analyzing chamber.
Electrochemistry
Electrochemical measurements were performed using a BAS ALS model 660E analyzer. Cyclic
voltammetry was performed using either an Au(111) electrode or the ANH@IL/Au electrode as the working
electrode, a Pt wire as a counter electrode and Ag/AgCl (3 M NaCl) as reference electrode. An aqueous
solution of 0.1 M NaClO4 was used as the electrolyte.
Controlled potential electrolysis of CO2 using ANH
The homogeneous reaction was performed using MilliꢀQ water containing 0.1 M NaClO4 and 0.01 M
ANH. Pt plate and Au plate were used as counter and working electrodes, respectively. The potentials of
the working electrode were –0.4, –0.6, –0.8, and –1.2 V vs. Ag/AgCl, and the reaction time was 2.5 hrs.
The electrochemical reduction reactions using the ANH@IL/Au electrode were carried out at –0.4, –0.6, –
0.8, and –1.2 V vs. Ag/AgCl in Milli Q water containing 0.1 M NaClO4, and the reaction time was 2.5 hrs.
18
ACS Paragon Plus Environment

Page 19 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
The ANH molecules evaluated are imidazole, 1ꢀmethylimidazole, 2ꢀmethylimidazole, 1,2ꢀdimethylimidazole,
pyridine, triazole, and pyrazole. The analyses of formate, methanol and gas products from electrolysis were
performed by Ionic chromatography, Gas chromatography (Shimadzu GCꢀ2014) and Gas chromatography
(Agilent Technologies 7890A), respectively. Amounts of these products were estimated from each
calibration curve using the standard compounds for each instrument.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of the electrode for SEIRAS measurements
The electrode was prepared for SEIRAS measurements as illustrated in Scheme S3. The Si ATR prism
used for SEIRAS measurements was polished using 0.05 mmφ alumina for about 10 minutes and
subsequently cleaned in MilliꢀQ water using an ultrasonic sonicator for 30 minutes. The surface of the Si
prism was etched with a 40% NH4F solution, cleaned with MilliꢀQ water, and continuously treated with 5
mM Na(AuCl4)·2H2O, 50 mM Na2SO3, 15 mM Na2S2O3, 15 mM NH4Cl, and 1 wt% HF solution at 60°C to
prepare the Au thin layer on the Si prism. The resistance of the plated Au was about 10 ohms.
Immobilization of ionic liquids and ANH onto the Au thin layer was performed by dipping the electrode into
aqueous solutions containing these compounds.
Characterization of the SEIRAS film
The characterization of the SEIRAS film was employed using cyclic voltammogram with Au film prepared
on a Si ATR prism in 0.1 M H2SO4 aqueous solution. Scan rate was 20 mV s−1. Active surface area was
calculated by integrating the reduction peak of Au oxide for a theoretical value of 444 ꢁC/cm2 required for
reducing Au oxide monolayer (Figure S13).76 The roughness factor of Au film employed for SEIRAS
measurement was estimated as 4.18.
Measurements of in-situ ATR-SEIRAS
Potentialꢀdependent and timeꢀdependent inꢀsitu ATRꢀSEIRAS measurements were performed using the
ANH@IL/Au electrode, which was immobilized with both IL and ANH on the Au thin layer plated on a
hemispherical Si ATR prism. SEIRA spectra were recorded in the Kretschmann attenuated total reflection
(ATR) configuration41 using a JASCO FT/IR4200 Fourier transform IR spectrometer equipped with a
1
HgCdTe (MCT) detector, in which the resolution of spectra is 4 cm− . Electrochemical measurements were
carried out using a BAS ALS600 series electrochemical analyzer. Potentialꢀdependent and timeꢀdependent
measurements were performed using either the bare Au electrode plated on a Si ATRꢀprism or the
ANH@IL/Au electrode as the working electrode, using a Pt wire as a counter electrode and an Ag/AgCl (3
M NaCl) as a reference electrode. A 0.1 M NaClO4/D2O solution was used as the electrolyte, which was
saturated with either Ar or CO2.
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 24
1
2
Conclusion
3
4
5
6
7
8
9
In summary, we developed a system for reducing CO2 at a low onsetꢀpotential (–0.32 V vs. Ag/AgCl), and
succeeded in generation of methanol at –0.7 V with high current efficiency (26% yield) using an
imidazoleꢀincorporated ionic liquidꢀmodified Au electrode in an aqueous solution. This system represents a
significant improvement over the previous system which employs pyridine/pyridinium ion on an Au
electrode and is not capable of producing methanol for reduction of CO2.32 In the present system, we
determined that formation of hydrogen bonds among imidazole molecules in ionic liquid on Au substrate is a
key aspect of the electrochemical CO2 reduction. In order to elucidate the CO2 reduction behavior for the
imidazole@IL/Au electrode, we characterized the interaction/reaction behavior of molecules on the
electrode by conducting in situ ATRꢀSEIRAS measurements. Our proposed mechanism of CO2 reduction
to formate and methanol at different potentials (Scheme 3) is based on data obtained from a combination of
timeꢀdependent and potentialꢀdependent in situ ATRꢀSEIRAS methods. We believe that the mechanism for
altering electron transfer properties identified in this work will contribute to the development of high
performance catalysts for conversion of CO2 to useful compounds.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI:
@@@@@@@@@@@.
> Additional spectroscopic details and schemes (PDF)
Acknowledgments
This work was supported in part by the Japan Society for the Promotion of Science (JSPS) through the
Program for Advancing Strategic International Networks to Accelerate the Circulation of Talented
Researchers and the Japan Society for the Promotion of Science (JSPS) for a GrantꢀinꢀAid for Scientific
Research (B)(C) (No. 16H04117, 15K05606).
References
1. Albo J.; AlvarezꢀGuerra M.; Castañoa P.; Irabien A. Green Chem. 2015, 17, 2304−2324.
2. Kim J.; Johnson T. A.; Miller J. E.; Stecheld E. B.; Maravelias C. T.; Energy Environ. Sci. 2012, 5, 8417−
8429.
3. Lee S.; Kim D.; Lee J. Angew. Chem., Int. Ed. 2015, 54, 14701−14705.
4. Hori Y.; Kikuchi K.; Suzuki S. Chem. Lett. 1985, 14, 1695−1698.
5. Lewis N. S.; Nocera D. G. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15729−15735.
6. Nakata K.; Ozaki T.; Terashima C.; Fujishima A.; Einaga Y. Angew. Chem., Int. Ed. 2014, 53, 871−874.
7. Qiao J.: Liu Y.; Hong F.; Zhang J. Chem. Soc. Rev. 2014, 43, 631−675.
20
ACS Paragon Plus Environment

Page 21 of 24
ACS Catalysis
1
2
3
4
8. Costentin C.; Robert M.; Savéant J.ꢀM. Chem. Soc. Rev. 2013, 42, 2423−2436.
9. Licht S. Adv. Mater. 2011, 23, 5592−5612.
5
6
7
10. Roy S. C.; Varghese O. K.; Paulose M.; Grimes C. A. ACS Nano 2010, 4, 1259−1278.
11. Lewis N. S. Science 2007, 315, 798−801.
8
9
12. Kumar B.; Llorente M.; Froehlich J.; Dang T.; Sathrum A.; Kubiak C. P. Annu. Rev. Phys. Chem. 2012,
63, 541−569.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13. Tran P. D.; Wong L. H.; Barber J.; Loo J. S. C. Energy Environ. Sci. 2012, 5, 5902−5918.
14. Yui T.; Tamaki Y.; Sekizawa K.; Ishitani O. Top Curr. Chem. 2011, 303, 151−184.
15. Balaraman E.; Gunanathan C.; Zhang J.; Shimon L. J. W.; Milstein D. Nat. Chem. 2011, 3, 609−614.
16. Centi G.; Perathoner S. Catal. Today 2009, 148, 191−205.
17. Cole E. B.; Lakkaraju P. S.; Rampulla D. M.; Morris A. J.; Abelev E.; Bocarsly A. B. J. Am. Chem. Soc.
2010, 132, 11539−11551.
18. Xiang D.; Magana D.; Dyer R. B. J. Am. Chem. Soc. 2014, 136, 14007−14010.
19. Keitha J. A.; Carter E. A. Chem. Sci. 2013, 4, 1490−1496.
20. Marjolin A.; Keith J. A. ACS Catal. 2015, 5, 1123−1130.
21. Ertem M. Z.; Konezny S. J.; Araujo C. M.; Batista V. S. J. Phys. Chem. Lett. 2013, 4, 745−748.
22. Lim C.ꢀH.; Holder A. M.; Hynes J. T.; Musgrave C. B. J. Phys. Chem. Lett. 2015, 6, 5078−5092.
23. Peroff A. G.; Weitz E.; Van Duyne R. P. Phys. Chem. Chem. Phys. 2016, 18, 1578−1586.
24. Lim C.ꢀH.; Holder A. M.; Musgrave C. B. J. Am. Chem. Soc. 2013, 135, 142−154.
25. Costentin C.; Canales J. C.; Haddou B.; Savéant J.ꢀM J. Am. Chem. Soc. 2013, 135, 17671−17674.
26. Yan Y.; Zeitler E. L.; Gu J.; Hu Y.; Bocarsly A. B. J. Am. Chem. Soc. 2013, 135, 14020−14023.
27. Dridi H.; Comminges C.; Morais C.; Meledje J.ꢀC.; Kokoh K. B.; Costentin C.; Savéant J.ꢀM. J. Am.
Chem. Soc. 2017, 139, 13922–13928.
28. Dunwell M.; Yan Y.; Xu B. ACS Catal. 2017, 7, 5410–5419.
29. Liao K.; Askerka M.; Zeitler E. L.; Bocarsly A. B.; Batista V. S. Top. Catal. 2015, 58, 23−29.
30. Portenkirchner E.; Enengl C.; Enengl S.; Hinterberger G.; Schlager S.; Apaydin D.; Neugebauer H.; Knör
G.; Sariciftci N. S. ChemElectroChem 2014, 1, 1543−1548.
31. Rybchenko S. I.; Touhami D.; Wadhawan J. D.; Haywood S. K. ChemSusChem. 2016, 9, 1660−1669.
32. Lucio A. J.; Shaw S. K. J. Phys. Chem. C 2015, 119, 12523−12530.
33. Lebegue E.; Agullo J.; Morin M.; Belanger D. ChemElectroChem 2014, 1, 1013−1017.
34. Barton Cole E. E.; Baruch M. F.; L’Esperance R. P.; Kelly M. T.; Lakkaraju P. S.; Zeitler E. L.; Bocarsly
A. B. Top. Catal. 2015, 58, 15−22.
35. Yang H.ꢀP.; Qin S.; Wang H.; Lu J.ꢀX. Green Chem. 2015, 17, 5144−5148.
36. Yang H.ꢀP.; Yue Y.ꢀNa; Qin S.; Wang H.; Lu JꢀX. Green Chem. 2016, 18, 3216−3220.
37. Kitagawa T.; Inomata T.; Funahashi Y.; Ozawa T.; Masuda H. Chem. Commun. 2013, 49, 10184−10186.
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 24
1
2
3
4
38. Kitagawa T.; Nishino J.; Inomata T.; Ozawa T.; Funahashi Y.; Masuda H. Chem. Commun. 2016, 52,
4780−4783.
5
6
7
39. Rosen B. A.; SalehiꢀKhojin A.; Thorson M. R.; Zhu W.; Whipple D. T.; Kenis P. J. A.; Masel R. I.
Science 2011, 334, 643−644.
8
9
40. Aki S. N. V. K.; Mellein B. R.; Saurer E. M.; Brennecke J. F. J. Phys. Chem. B 2004, 108, 20355－
20365.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
41. Yan Y.; Gu J.; Bocarsly A. B. Aerosol Air Qual. Res. 2014, 14, 515–521.
42. Seshadri G.; Lin C.; Bocarsly A. B. J. Electro. Chem. 1994, 372, 145−150.
43. Boston D. J.; Xu C.; Armstrong D. W.; MacDonnell F. M. J. Am. Chem. Soc. 2013, 135, 16252−16255.
44. Osawa M. Bull. Chem. Soc. Jpn. 1997, 70, 2861−2880.
45. Perchard C.; Bellocq A. M.; Novak A. J. Chim. Phys. 1965, 62, 1344–1358.
46. Grossutti M.; Leitch J. J.; Seenath R.; Karaskiewicz M.; Lipkowski J. Langmuir 2015, 31, 4411−4418.
47. Sister M. C.; Walter J. L. Spectrochimica Acta 1968, 24, 237−252.
48. Rodriguez J.; Elola M. D.; Laria D. J. Phys. Chem. B 2015, 119, 9123−9128.
49. Costa P. S.; Miller D. P.; Teeter J. D.; Beniwal S.; Zurek E.; Sinitskii A.; Hooper J.; Enders A. J. Phys.
Chem. C 2016, 120, 5804−5809.
50. Furuya N.; Motto S.; Kunimatsu K. J. Electroanal. Chem. 1988, 239, 347−360.
51. Miki A.; Yeb S.; Osawa M. Chem. Comm. 2002, 21, 1500−1501.
52. Wuttig A.; Yaguchi M.; Motobayashi K.; Osawa M.; Surendranath Y. Proc. Natl. Acad. Sci. U. S. A. 2016,
113, E4585–E4593.
53. Dunwell M.; Lu Q.; Heyes J. M.; Rosen J.; Chen J. G.; Yan Y.; Jiao F.; Xu B. J. Am. Chem. Soc. 2017,
139, 3774−3783.
54. Wuttig A.; Liu C.; Peng Q.; Yaguchi M.; Hendon C. H.; Motobayashi K.; Ye S.; Osawa M.;
Surendranath Y. ACS Cent. Sci. 2016, 2, 522−528.
55. Xia X. H.; Liess H. D.; Iwasita T. J. Electroanal. Chem. 1997, 437, 233−240.
56. Firet N. J.; Smith W. A. ACS Catal. 2017, 7, 606–612.
57. Li M.; Liu P.; Adzic R. R. J. Phys. Chem. Lett. 2012, 3, 3480−3485.
58. Koziol L.; Mozhayskiy V. A.; Braams B. J.; Bowman J. M.; Krylov A. I. J. Phys. Chem. A 2009, 113,
7802−7809.
59. Iwasita T.; Nart F. C. Pro. Surf. Sci. 1997, 55, 271−340.
60. Bowman J. M.; Bittman J. S.; Harding L. B. J. Chem. Phys. 1986, 85, 911−921.
61. Liu X.; Xiao J.; Peng H.; Hong X.; Chan K.; Nørskov J. K. Nature Commun. 2017, 15438.
62. Xiao H.; Cheng T.; Goddard III W. A.; Sundararaman R. J. Am. Chem. Soc. 2016, 138, 483−486.
63. Arihara K.; Kitamura F.; Ohsaka T.; Tokuda K. J. Electroanal. Chem. 2001, 510, 128−135.
64. Sandoval A. P.; Orts J. M.; Rodes A.; Feliu J. M. J. Phys. Chem. C 2011, 115, 16439−16450.
22
ACS Paragon Plus Environment

Page 23 of 24
ACS Catalysis
1
2
3
4
5
6
7
8
65. Figueiredo M. C.; LedezmaꢀYanez I.; Koper M. T. M. ACS Catal. 2016, 6, 2382−2392.
66. Rodriguez P.; Kwon Y.; Koper M. T. M. Nature Chem. 2014, 4, 177–182.
67. Pander III J. E.; Baruch M. F.; Bocarsly A. B. ACS Catal. 2016, 6, 7824−7833.
68. Kortlever R.; Shen J.; Schouten K. J. P.; CalleꢀVallejo F.; Koper M. T. M. J. Phys. Chem. Lett. 2015, 6,
4073–4082.
9
69. Fang Y.; Flake J. C. J. Am. Chem. Soc. 2017, 139, 3399−3405.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
70. Hori Y.; Murata A.; Kikuchi K.; Suzuki S. J. Chem. Soc., Chem. Commun. 1987, 728−729.
71. Chen Y.; Li C. W.; Kanan M. W. J. Am. Chem. Soc. 2012, 134, 19969−19972.
72. Shi C.; Hansen H. A.; Lauscheb A. C.; Nørskov J. K. Phys. Chem. Chem. Phys. 2014, 16, 4720−4727.
73. Chen L. D.; Urushihara M.; Chan K.; Nørskov J. K. ACS Catal. 2016, 6, 7133−7139.
74. Cave E. R.; Montoya J. H.; Kuhl K. P.; Abram D. N.; Hatsukade T.; Shi C.; Hahn C.; Nørskovb J. K.;
Jaramillo T. F. Phys. Chem. Chem. Phys. 2017, 19, 15856−15863.
75. Yoshimoto, S.; Ono, Y.; Kuwahara, Y.; Nishiyama, K.; Taniguchi, I. J. Phys. Chem. C 2016, 120, 15803
−15813.
76. AngersteinꢀKozlowska, H.; Conway, B. E.; Hamelin, A.; Stoicoviciu, L. Electrochim. Acta 1986, 31,
1051−1061.
23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 24
1
2
Table of content
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
24
ACS Paragon Plus Environment