Formation of carbon–nitrogen bonds in carbon monoxide electrolysis

Article abstract of DOI:10.1038/s41557-019-0312-z

The electroreduction of CO₂ is a promising technology for carbon utilization. Although electrolysis of CO₂ or CO₂-derived CO can generate important industrial multicarbon feedstocks such as ethylene, ethanol, n-propanol and acetate, most efforts have been devoted to promoting C–C bond formation. Here, we demonstrate that C–N bonds can be formed through co-electrolysis of CO and NH₃ with acetamide selectivity of nearly 40% at industrially relevant reaction rates. Full-solvent quantum mechanical calculations show that acetamide forms through nucleophilic addition of NH₃ to a surface-bound ketene intermediate, a step that is in competition with OH^– addition, which leads to acetate. The C–N formation mechanism was successfully extended to a series of amide products through amine nucleophilic attack on the ketene intermediate. This strategy enables us to form carbon–heteroatom bonds through the electroreduction of CO, expanding the scope of products available from CO₂ reduction.

Full text of DOI:10.1038/s41557-019-0312-z

Articles
https://doi.org/10.1038/s41557-019-0312-z
Formation of carbon–nitrogen bonds in carbon
monoxide electrolysis
1
,6
1,2,6
3,4,5,6
1
2
Matthew Jounyꢀ ꢀ , Jing-Jing Lv , Tao Chengꢀ ꢀ
, Byung Hee Koꢀ ꢀ , Jun-Jie Zhuꢀ ꢀ ,
William A. Goddard III ,4* and Feng Jiaoꢀ ꢀ1*
3
The electroreduction of CO is a promising technology for carbon utilization. Although electrolysis of CO or CO -derived CO
2
2
2
can generate important industrial multicarbon feedstocks such as ethylene, ethanol, n-propanol and acetate, most efforts have
been devoted to promoting C–C bond formation. Here, we demonstrate that C–N bonds can be formed through co-electrolysis
of CO and NH with acetamide selectivity of nearly 40% at industrially relevant reaction rates. Full-solvent quantum mechani-
3
cal calculations show that acetamide forms through nucleophilic addition of NH to a surface-bound ketene intermediate, a
3
–
step that is in competition with OH addition, which leads to acetate. The C–N formation mechanism was successfully extended
to a series of amide products through amine nucleophilic attack on the ketene intermediate. This strategy enables us to form
carbon–heteroatom bonds through the electroreduction of CO, expanding the scope of products available from CO reduction.
2
ecarbonization of the chemical industry presents a consider- carbon–nitrogen bond could be formed. The overall strategy for
able challenge in breaking our reliance on fossil resources and C–N bond formation is outlined in Fig. 1a.
1–4
D
reducing CO emissions . Driven by the increasingly rapid
Here, we report C–N bond formation that results from CO elec-
2
5
deployment of renewable power generation , the cost of renewable troreduction in the presence of NH . An electrochemical production
3
electricity has decreased considerably, making electrochemical CO
of acetamide with nearly 40% Faradaic efficiency was achieved at a
2
–2
reduction an attractive approach to produce sustainable fuels and current density of 300mAcm . Our previous full-solvent quantum
6
–8
23,24
chemicals . Multi-carbon (C ) chemicals are desirable products mechanical calculations
revealed that—under neutral or basic
2+
because they are more valuable than typical single-carbon products conditions—the reaction mechanism involves CO dimerization and
6
,9,10
such as CO (refs.
). Highly alkaline electrolytes are often used to sequential transfer of H from two surface waters to form the *(HO)
–
enhance C selectivity; however, the inevitable reaction of OH ions C=COH intermediate that subsequently proceeds through two sep-
2
+
with CO to form undesired carbonates at the electrode–electrolyte arate pathways to form C H and ethanol/n-propanol. We show now
2
2
4
1
1,12
interface disrupts the electrolysis process . This problem can be that *(HO)C=COH is also hydrolysed to *C=C=O, which in turn
addressed through a two-step process in which CO is first reduced reacts with NH to form intermediates that proceed to form acet-
2
3
to CO electrochemically at non-alkaline conditions, which is fol- amide while suppressing the formation of other C products. We
2
+
lowed by a CO reduction step to produce C chemicals in alkaline also successfully extended the range of C–N containing products
2+
environments. Regardless of CO or CO reduction, only four major to N-methylacetamide, N-ethylacetamide, N,N-dimethylacetamide,
2
C
products—ethylene, acetate, ethanol and n-propanol—have acetic monoethanolamide and aceturic acid. Our results provide
2
+
1
3–19
been reported in aqueous electrolytes . Although these products critical mechanistic insights into Cu-catalysed CO /CO electrore-
2
are of importance in current chemical industries, the ability to pro- duction and demonstrate the construction of carbon–heteroatom
duce chemicals beyond simple carbon species is critically important bonds in CO /CO electrolysis.
2
20,21
because valuable specialty chemicals often contain heteroatoms
.
In a recent study, we observed an enhanced acetate selectivity in Results and discussion
CO reduction in comparison with CO reduction under identical The concept of a nucleophilic NH addition during CO electroly-
2
3
22
conditions . Surface pH calculations and isotopic labelling studies sis was verified through a series of electrolysis experiments using a
suggested that acetate was favourably formed through nucleophilic three-compartment continuous flow cell with a well-defined triple-
–
attack of OH on a ketene-like intermediate under highly alkaline phase interface (Supplementary Fig. 1). Copper cathodes were pre-
conditions, which is in good agreement with a previous study by pared by coating Cu nanoparticles (NPs) onto a gas diffusion layer
1
3
Kanan et al. . As ketene is known to be highly reactive with nucleo- (GDL). The size distribution and monoclinic phase of the Cu NPs
philic agents, it is reasonable to assume that the ketene-like inter- were characterized using scanning electron microscopy (SEM),
mediate present on the Cu catalyst surface in CO reduction could X-ray diffraction (XRD) and X-ray photoelectron spectroscopy
–
readily react with other nucleophilic agents besides OH ions. We (XPS) (Supplementary Fig. 2). The Cu NPs are mainly highly crys-
postulated that if a nitrogen-containing nucleophilic agent such as talline metallic Cu with an average particle size of 50± 20nm, but
NH was introduced to the Cu-catalysed CO electrolysis system, a they also contain a small fraction of Cu oxides. The electroreduc-
3
1
2
Center for Catalytic Science and Technology, Department of Chemical and Biomolecular Engineering, University of Delaware, Newark, DE, USA. State
3
Key Laboratory of Analytical Chemistry for Life Science, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, China. Joint Center
for Artificial Photosynthesis, California Institute of Technology, Pasadena, CA, USA. Materials and Process Simulation Center, California Institute of
Technology, Pasadena, CA, USA. Institute of Functional Nano & Soft Materials, Jiangsu Key Laboratory for Carbon-based Functional Materials & Devices,
Joint International Research Laboratory of Carbon-based Functional Materials and Devices, Soochow University, Suzhou, Jiangsu, China. These authors
4
5
6
contributed equally: Matthew Jouny, Jing-Jing Lv, Tao Cheng. *e-mail: wag@caltech.edu; jiao@udel.edu
NATꢁRꢂ CHꢂMIꢃTRꢄ | VOL 11 | SEPTEMBER 2019 | 846–851 | www.nature.com/naturechemistry
8
46

NATure CHemisTry
Articles
a
b
CO2
Acetic acid
6
00
CO
CO + NH3
NH3
Acetamide
Ethanol
500
Ethylene
4
3
2
00
00
00
CO
C–C
C–N
100
0
Electrode
–0.45 –0.50 –0.55 –0.60 –0.65 –0.70 –0.75
Potential (V versus RHE)
c
d
e
1
1
8
6
C
C
O reduction
O reduction
1
00
CO
100
80
60
40
20
0
CO + NH
3
1
.0
.8
.6
Acetamide
8
6
4
2
0
0
0
0
0
0
0
O
NH2
0.4
0.2
0
–
0.45 –0.50 –0.55 –0.60 –0.65 –0.70 –0.75
–0.45 –0.50 –0.55 –0.60 –0.65 –0.70 –0.75
Potential (V versus RHE)
56 57 58 59 60 61 62 63 64 65
m/z (AMU
Potential (V versus RHE)
)
Acetamide
n-Propanol
Acetate
Ethanol
Ethylene
H2
Fig. 1 | Acetamide production from CO electrolysis with NH . a, A schematic that depicts NH inducing C–N bond formation from CO electrolysis. b,
3
3
Electrode polarization curves for electrolysis in 1ꢀM KOH under pure CO gas and a NH /CO ratio of 2:1. c,d, The Faradaic efficiencies versus the applied
3
1
6
18
potential for CO (c) and CO + NH (d). e, A mass spectrum comparing acetamide produced from C O and C O. The error bars represent the s.d. for at
3
least three independent measurements.
tion activity of CO was measured through steady-state galvanostatic liquid phase NH with appreciable Faradaic efficiency. To evaluate
3
electrolysis in a 1M KOH electrolyte. Under a pure CO gas feed, the stability of the Cu-catalysed CO electrolysis process in the pres-
a near-exponential polarization response was observed (Fig. 1b, ence of NH , an eight hour continuous experiment was performed
3
–
2
Supplementary Table 1), with up to ~80% C products for a at a total current density of 100 mA cm that led to stable pro-
2
+
–
2
total current density of 500mAcm . The major CO electrore- duction of acetamide (Supplementary Fig. 5 and Supplementary
duction products observed were ethylene, ethanol, acetate and Table 1). The spent Cu catalyst was characterized by SEM, XRD
n-propanol (Fig. 1c).
and XPS and no visible change was observed in the morphol-
After establishing the baseline of CO electrolysis activity, ogy and structure after electrolysis (Supplementary Fig. 6).
NH gas was fed together with CO in an NH :CO (mol/mol) X-ray photoelectron spectroscopy revealed some Cu oxides fol-
3
3
ratio of 2:1. In the presence of NH gas, the required poten- lowing electrolysis, which is probably due to exposure to air, as
3
2
7,28
tial to achieve the same current density increased by ~30 mV recent studies
have shown that Cu catalysts are fully metal-
(
Fig. 1b, Supplementary Table 1), possibly due to the reduced lic while under reducing potentials. This is further supported
2
5,26
CO partial pressure in the flow cell . Remarkably, the pres- by in situ X-ray absorption spectroscopy performed on a 5-nm-
2
9
ence of NH led to considerable production of acetamide, with thick Cu catalyst in 5 M ammonium hydroxide showing entirely
3
0
a Faradaic efficiency up to 38% and a partial current density Cu features (Supplementary Fig. 7). Furthermore, significant
–
2
of 114 mA cm at −0.68 V versus the reversible hydrogen elec- amounts of acetamide were also produced on other Cu-based
trode (RHE). Furthermore, the observed amounts of ethylene catalysts (Supplementary Fig. 8), suggesting that the formation
and alcohols were greatly decreased at moderate to high over- of acetamide is universal in Cu-catalysed CO electrolysis in the
potentials, whereas acetate selectivity was maintained (Fig. 1d). presence of NH . The use of CO as the feed gas resulted in spon-
3
2
Increasing the fraction of CO in the gas feed shifted selectivity taneous carbonate formation and the observation of no amides,
towards pure CO reduction products, but increasing the ratio indicating this chemistry may be difficult through direct CO
2
of NH beyond 2:1 did not significantly influence the acetamide reduction (Supplementary Fig. 9).
3
selectivity (Supplementary Fig. 3). Similar results were obtained
These experimental results strongly suggest that a surface ketene
using a mixture of ammonium hydroxide and KOH or KCl as intermediate is probably formed on the Cu catalyst surface during
–
the catholyte together with a pure CO gas feed (Supplementary CO electroreduction and nucleophilically attacked by either OH
Fig. 4). This suggests that acetamide can form in both gas and or NH to form acetate or acetamide, respectively, under highly
3
NATꢁRꢂ CHꢂMIꢃTRꢄ | VOL 11 | SEPTEMBER 2019 | 846–851 | www.nature.com/naturechemistry
8
47

Articles
NATure CHemisTry
†
∆G† = 0.55 eV
∆G
= 0.69 eV
O
O
Now we consider a new step starting with *(HO)C=COH. We
e–
O
O
O
OH
+
+ H2O
ϯ
C
C
C
C
C
C
find ΔG =0.57eV to form *C=C=O through a water-mediated
–
–
OH
pathway. The possibility of the *C=C=O intermediate was first
3
3
proposed by Koper , which was postulated as an intermediate in
the ethylene pathway. However, later full-solvent QM showed that
the formation of C H derives from *C=COH as in Fig. 2. Our new
†
∆G
= 0.24 eV
H2O
–
–
OH
+
2
4
O
C
C
QM calculations find that *C=C=O derives from dehydration of
∆G† = 0.57 eV
H2O
HO
OH
*
(HO)C=COH. Thus, in the competition with *C=COH from pro-
C
C
ton-coupled electron transfer, *C=C=O prefers high pH and less
negative potential. This is consistent with the experimental observa-
tion of exclusive acetate formation on Cu NP at pH 14 and –0.25V
–
∆G = – 0.26 eV
–
+
H O
– OH
2
–
∆G
†
= 0.61 eV
34
+
OH
+ NH3
∆G† = 0.53 eV
versus RHE .
†
∆G
= 0.72 eV
OH
We find that C–N bond formation arises from NH reacting
3
C
C
ϯ
with *C=C=O to form *C=C(OH)NH (ΔG =0.53eV) via a water-
2
mediated reaction pathway. We then discover that *C=C(OH)NH
H
2
–
∆G = 0.01 eV
O
O
HO
NH2
isomerizes into *CH–C(=O)NH through a keto–enol tautomerism
2
C
C
C
C
that is exergonic by –0.11eV. These two reactions are not electro-
−
chemical and *CH–C(=O)NH remains a 2e intermediate just as it
2
Ethylene
major)
Ethanol
(minor)
n-Propanol
(
(minor)
did for *C=C=O. The subsequent steps involve two proton-coupled
∆G = 0.28 eV
∆G = – 0.06 eV
electron transfers to the acetamide product as shown in Fig. 2. In
competition with NH addition to *C=C=O, we find that acetate
3
−
forms through direct reaction of OH with *C=C=O to form
−
ϯ
O
C
O
C
*C=C(OH)O (ΔG =0.72eV), which then undergoes a similar
–
keto–enol tautomerism and a subsequent proton-coupled electron
HC
O
HC
NH2
+
−
transfer. The presence of K will stabilize *C=C(OH)O on the
catalyst surface.
∆G = 0.29 eV
∆G = –0.11 eV
With these insights, we can extend the reaction networks of CO
2
4
24,30
Acetate
Acetamide
reduction to ethylene and ethanol
to include the branches of
–
acetamide and acetate from NH and OH addition, respectively
3
Fig. 2 | The mechanism for CO reduction on Cu that shows how it splits at
*(HO)C=COH] into two pathways. One pathway features *C=COH (and
produces ethylene, ethanol and n-propanol) whereas the other features
(Fig. 2). The observed suppression of ethylene and alcohols in the
[
presence of high concentrations of NH (Fig. 1c,d) is probably due
3
to decreased water availability hindering protonation of *(HO)
C=COH to form *C=COH, which results in an overall increase
in *C=C=O formation. Although acetamide formation from
*C=C=O is more energetically favourable than acetate formation,
*
C=C=O (and produces acetamide and acetate). The inset shows the
structure of *CH–C(=O)NH , the second reactive intermediate with
a C–N bond.
2
–
acetate production is maintained at high current densities as OH is
generated directly at the catalyst surface.
alkaline environments. This is further supported by a shift in selec-
As the key intermediate towards acetate and acetamide in
tivity from amide to acetate for CO electrolysis with NH in electro- Cu-catalysed CO electroreduction, ketene is also known to be
3
lytes with increasing KOH concentration (Supplementary Fig. 10). highly reactive with other amine-type nucleophilic agents. We
Furthermore, because the ketene intermediate contains one oxygen therefore investigated Cu-catalysed CO electrolysis in the pres-
originated from CO, the resulting acetate should contain two oxy- ence of additional amines in hope of producing the corresponding
gen atoms with one from CO and the other from water, which is in amides. We performed electroreduction of a pure CO gas feed using
22,30
good agreement with recent studies . In the case of a nucleophilic 5M solutions of methylamine, ethylamine and dimethylamine that
attack by NH , the oxygen in acetamide should originate from CO. contain 1M KCl as the supporting electrolyte. We used 1M KCl to
3
1
8
We verified the origin of oxygen in acetamide by conducting a C O enhance the ionic conductivity of the electrolytes. As shown in Fig.
18
isotopic labelling study, where O labelled acetamide was the domi- 3a–c, analogous results to the CO/NH system were obtained, where
3
nant product (Fig. 1e), consistent with the proposed ketene medi- substantial amounts of N-methylacetamide, N-ethylacetamide and
ated reaction mechanism.
N,N-dimethylacetamide were produced at high total current densi-
–2
To further elucidate the reaction mechanism, we used quantum ties of up to 300mAcm with peak Faradaic efficiencies of 42%,
mechanics (QM; Perdew–Burke–Ernzerhof D3 (PBE-D3) density 34% and 36%, respectively (Supplementary Table 2). The formation
1
functional theory) to investigate the electrocatalytic formation of acet- of these amides was confirmed using mass and H NMR spectrom-
amide in the presence of NH (see the Supplementary Information etry (Supplementary Figs. 11 and 12).
3
for simulation details). We used the full-solvent methods previously
The molar fraction of each C product (excluding hydrogen)
+
–2
2
23
applied to CO reduction and CO reduction on Cu(100), which is at 200mAcm in each amine system is shown in Fig. 3d. Data for
2
the dominant Cu surface under our conditions and the most active pure CO electrolysis were also shown for comparison. Similar to
24,30,31
for C–C coupling
. The explicit solvation calculations have been NH , the molar fractions for ethylene and ethanol decrease by about
3
demonstrated to be robust considering that simulations from the twofold and fourfold, respectively, for all amines tested. The trend
32
independent work of Bagger et al. reached very similar conclu- of the amide molar fraction across various amines is opposite to that
sions as our previous work. Our earlier QM full-solvent calculations of acetate and correlates well with the reactivity (or nucleophilicity)
2
4,31
showed that under neutral or basic conditions the reaction mecha- of the precursor amino group. The reactive N–H bond is weakest
nism involves CO dimerization and sequential transfer of H from for dimethylamine and strongest for NH , with methylamine and
3
35
–
two surface water to form *(HO)C=COH with an overall free energy ethylamine between . As the amine competes with OH to react
ϯ
barrier (ΔG ) at 298K of 0.69eV. This then leads to *C=COH with with the ketene intermediate, it is reasonable that dimethylamine
ϯ
ΔG =0.61eV that subsequently goes through two separate pathways produces the highest ratio of amide to acetate, whereas NH pro-
3
to form C H (65%) and ethanol/n-propanol (35%).
duces the lowest. Therefore, these observations further support the
2
4
NATꢁRꢂ CHꢂMIꢃTRꢄ | VOL 11 | SEPTEMBER 2019 | 846–851 | www.nature.com/naturechemistry
8
48

NATure CHemisTry
Articles
a
b
O
CO
e–
O
CO
H2N Me
Me
H2N
N
e–
N
H
H
3
2
2
1
1
5
0
00
50
00
50
00
0
300
250
200
150
100
50
1
00
100
80
60
40
20
0
8
6
4
2
0
0
0
0
0
0
–1.40
–1.45
–1.50
–1.55
–1.60
–1.40
–1.45
–1.50
–1.55
–1.60
Potential (V versus Hg/HgO)
Potential (V versus Hg/HgO)
Amide
n-Propanol
Acetate
c
O
Me
CO
e–
Me
Ethanol
Ethylene
Hydrogen
HN
N
Me
Me
d
0
0
0
0
0
0
0
.8
.7
.6
.5
.4
.3
.2
3
2
2
1
1
5
0
00
50
00
50
00
0
1
00
8
6
4
2
0
0
0
0
0
Amide
Ethylene
Alcohols
Acetate
0.1
0
–1.30
–1.35
–1.40
–1.45
–1.50
–1.55
None
NH3
CH NH Et–NH2 NH(CH3)2
3
2
Potential (V versus Hg/HgO)
Amine
Fig. 3 | ꢂlectrochemical production of longer amides from CO electrolysis in 5ꢅM amine solutions. a–c, The total current density and Faradaic efficiencies
for CO electrolysis in 1ꢀM KCl solution containing (a) 5ꢀM methylamine, (b) 5ꢀM ethylamine and (c) 5ꢀM dimethylamine. d, The molar production fraction
–2
for different C2+ products excluding hydrogen at 200ꢀmAꢀcm for CO reduction with various amines. Note that the NH fractions are calculated from Fig.
3
1
d. The error bars represent the s.d. for at least three independent measurements.
mechanism proposed earlier and provide important mechanistic recently3 . More importantly, the concept of nucleophilic attack of
8–40
insight into the Cu-catalysed CO electroreduction reaction.
a ketene intermediate in Cu-catalysed CO electroreduction enables
We were also able to further extend the range of products to acet- the formation of a much wider range of chemicals containing car-
amides that contain hydroxyl and carboxylate functional groups. bon–heteroatom bonds that cannot be built in conventional CO
Acetic monoethanolamide and aceturic acid were produced by per- electrolysis processes. The ability to produce heteroatom-contain-
forming CO electrolysis in solutions of ethanolamine and glycine, ing carbon species would greatly increase the potential of CO /CO
2
respectively (Supplementary Fig. 13). As these products contain electrolysis technologies for commercial applications.
reactive functional groups, they can be used as potential precursors
to build larger molecules with higher values. This opens up a wide Methods
Materials characterization. All chemicals were of analytical grade and used as
library of chemical transformations in which CO electrolysis can
received without further puriꢀcation unless otherwise noted. Commercial Cu NPs
play an important role. Although the goal of this work is to demon-
with 25nm diameter and Cu particles with 1µm diameter were used as catalysts
strate the concept of electrochemical C–N bond formation, future
in this work (both purchased from Sigma Aldrich). ꢁe microstructures of the
studies can identify and optimize the production of more species.
catalysts were characterized by ꢀeld emission SEM (Auriga, 1.5kV). Powder XRD
measurements were conducted on a D8 ADVANCE X-ray diꢂractometer (Bruker).
A ꢁermo Scientiꢀc K-Alpha XPS system was used and XPS ꢀtting was conducted
by CasaXPS soꢃware with the adventitious carbon peak being calibrated to
Conclusion
In summary, we demonstrate a new route to produce a variety of
acetamides through CO electrolysis at ambient conditions. These
products are commonly used in the polymer and pharmaceutical
2
84.8eV. All peaks were ꢀtted using a Gaussian/Lorentzian product line shape and
a Shirley background.
20,36
industries . In particular, N,N-dimethylacetamide has subtan- Flow cell electrolysis. Electrochemical measurements were conducted on an
Autolab potentiostat (PG128N) in a three-electrode system. Unless otherwise
tial usage as a polymerization solvent, and currently requires harsh
37
2
noted, an IrO (Alfa Aesar) coated GDL (Sigracet 29 BC, Fuel Cell Store) with
−2
synthesis conditions . Although the amines used in this work
a loading of 0.5mgcm was used as the counter electrode, whereas Ag/AgCl
(saturated KCl, Pine Research) or Hg/HgO (filled with 1M KOH, Pine Research)
are currently produced with methane-derived NH and metha-
3
nol, renewable NH synthesis is currently an area of great interest were used as the reference electrodes. The Cu NPs (or the Cu-coated GDL) were
3
and many promising production routes have been demonstrated applied as the working electrode, which was prepared by hand painting a catalyst
NATꢁRꢂ CHꢂMIꢃTRꢄ | VOL 11 | SEPTEMBER 2019 | 846–851 | www.nature.com/naturechemistry
8
49

Articles
NATure CHemisTry
ink. We weighed the GDL before and after deposition to record its actual catalyst
Cu(100) interface using 48 explicit water molecules (five layers, 1.21nm thick) on
−
2
2
loading and kept all of the electrodes with a loading of 0.6mgcm . To prepare the
catalyst ink, 25mg of catalyst was ultrasonically dispersed in the mixture of 3ml of
isopropanol, 500μl of multi-walled carbon nanotube solution (5mg multi-walled
carbon nanotube (>98%, Sigma-Aldrich) dispersed in 5ml THF) and 20μl of
Nafion (10wt% aqueous solution, Fuel Cell Store). The mixture was then sonicated
for 30min before dropcasting.
a 4×4 Cu(100) surface slab (three layers) with an area of 1.02nm . For the cases
that involve NH
3 3
, we also included the explicit NH in our calculation along with
47 H
2
O. The simulation protocol of free-energy calculations is the same as in our
24
previous work . To confirm the robustness of the calculation, the onset potential
for the hydrogen evolution reaction on Cu(100) was predicted to be –0.4V versus
43
RHE, which is close to the experimental value . More simulation details are
included in the Supplementary Information.
The electrolysis experiments were performed in a three-channel flow cell with
3
2
channels with dimensions of 2×0.5×0.15cm . The electrode area was 1cm and
the electrode to membrane distance was 1.5mm. An external Ag/AgCl or Hg/
HgO reference electrode located ~5cm from the cathode was used to measure the
cathodic half-cell potential. Electrolysis measurements were performed through
chronopotentiometry using an Autolab PGSTAT128N potentiostat/galvanostat. All of
the potential measurements were converted to the RHE using the following formula:
Data availability
The datasets generated during and/or analysed during the current study are
available from the corresponding author on reasonable request.
RHE =EAg/AgCl +0.222+0.059×pH (in volts) or ERHE =EHg/HgO +0.097+0.059×pH
Code availability
E
(
in volts), where the standard values for the reference electrodes were found by
The computational codes used in the current study are available from the
corresponding author on reasonable request.
calibration through cyclic voltammetry in a H -saturated 1M KOH electrolyte with
2
a Pt cathode and anode. The measured pH values of bulk electrolyte exiting the
flow cell—obtained using a pH meter (Apera Instruments)—were used for the RHE
conversions unless stated otherwise. The measured potential was iR corrected at 100%
by measuring the solution resistance between the reference electrode and cathode
Received: 21 January 2019; Accepted: 10 July 2019;
Published online: 23 August 2019
41
with the current-interrupt technique before each applied current . The device is
fabricated from acrylic and includes the gas channel for the following: feeding CO
References
1
.
Otto, A., Grube, T., Schiebahn, S. & Stolten, D. Closing the loop: captured
3
and NH ; anode and cathode channels for the flowing electrolyte; an anion exchange
CO
2
as a feedstock in the chemical industry. Energy Environ. Sci. 8,
membrane (FAA-3, Fumatech) for separating the anode and cathode; and solid acrylic
end pieces. Polytetrafluoroethylene gaskets were placed between each component
for sealing and the device was tightened using six bolts. The CO flow rate was set
3
283–3297 (2015).
2
.
.
Schiꢂer, Z. J. & Manthiram, K. Electriꢀcation and decarbonization of the
chemical industry. Joule 1, 10–14 (2017).
via a mass flow controller (Brooks GF40) and the NH flow rate was controlled by a
3
3
Katelhon, A., Meys, R., Deutz, S., Suh, S. & Bardow A. Climate change
mitigation potential of carbon capture and utilization in the chemical
industry. Proc. Natl Acad. Sci. USA 116, 11187–11194 (2019).
Montoya, J. H. et al. Materials for solar fuels and chemicals. Nat. Mater. 16,
rotameter (Cole Parmer, PMR1-010286). The electrolyte flow rates were controlled
via a peristaltic pump (Cole Parmer), with the catholyte and anolyte flow rates set
–
1
–1
to 0.5mlmin and 1mlmin , respectively. Amines were scrubbed from the effluent
4.
5.
6.
7.
gas from the flow cell using an acid trap (3M H
2
SO
4
solution) before entering the gas
7
0–81 (2016).
chromatograph.
Haegel, N. M. et al. Terawatt-scale photovoltaics: transform global energy.
Science 364, 836–838 (2019).
For CO electrolysis in the presence of NH
3
, the gas channel was co-fed with
3
CO and NH , with 1M KOH used as the catholyte and anolyte (Ag/AgCl reference
De Luna, P. et al. What would it take for renewably powered electrosynthesis
to displace petrochemical processes? Science 364, eaav3506 (2019).
electrode). For CO electrolysis in the presence of liquid phase amines, a pure CO
gas feed was used, with the catholyte comprising the reactants (NH O, CH NH
CH CH NH , CH NHCH , HOCH CH NH and NH CH COOH) and a supporting
3
H
2
3
2
,
2
Verma, S., Lu, S. & Kenis, P. J. A. Co-electrolysis of CO and glycerol as a
3
2
2
3
3
2
2
2
2
2
pathway to carbon chemicals with improved technoeconomics due to low
electrolyte (KOH or KCl), and a 1M KOH anolyte (Hg/HgO reference electrode).
A NiFe/Ni foam anode—prepared following a previously reported method —was
used as the anode electrode for the acetamide production stability test.
42
electricity consumption. Nature Energy 4, 466–474 (2019).
8
9
1
.
2
Nitopi, S. et al. Progress and perspectives of electrochemical CO reduction
on copper in aqueous electrolyte. Chem. Rev. 119, 7610–7672 (2019).
Jouny, M., Luc, W. & Jiao, F. General techno-economic analysis of CO
electrolysis systems. Ind. Eng. Chem. Res. 57, 2165–2177 (2018).
.
2
Product quantification. Gas products were quantified on a Multigas no. 5 GC
(
SRI Instruments) equipped with Hayesep D and Molsieve 5Å columns leading
0. Gao, D., Arán-Ais, R. M., Jeon, H. S. & Roldan Cuenya, B. Rational catalyst
to a thermal conductivity detector and a Hayesep D column leading to a flame
ionization detector. Hydrogen was quantified using the thermal conductivity
detector whereas ethylene and methane were detected using both a flame
ionization detector and a thermal conductivity detector. The Faradaic efficiency for
the products was calculated using the following equation:
and electrolyte design for CO
2
electroreduction towards multicarbon
products. Nat. Catal. 2, 198–210 (2019).
1
2
1. Dinh, C.-T. et al. CO electroreduction to ethylene via hydroxide-mediated
copper catalysis at an abrupt interface. Science 360, 783 (2018).
1
2
2. Verma, S. et al. Insights into the low overpotential electroreduction of CO to
nFxV
CO on a supported gold catalyst in an alkaline ꢄow electrolyzer. ACS Energy
Lett. 3, 193–198 (2018).
Faradaic efficiency ð%Þ ¼
*100
jtot
1
3. Li, C. W., Ciston, J. & Kanan, M. W. Electroreduction of carbon monoxide
to liquid fuel on oxide-derived nanocrystalline copper. Nature 508,
504–507 (2014).
where n is the number of electrons transferred, F is Faraday’s constant, x is the mole
fraction of product, V is thetotal molar flow rate of gas and jtot is thetotal current.
1
Liquid products were quantified using H NMR (Bruker AVIII 600MHz).
14. Verdaguer-Casadevall, A. et al. Probing the active surface sites for CO
reduction on oxide-derived copper electrocatalysts. J. Am. Chem. Soc. 137,
9808–9811 (2015).
1
The H NMR spectra were obtained using a pre-saturation method for water
suppression. Typically, 500µl of collected diluted catholyte was mixed with 100µl of
internal standard solution (25ppm (m/m) dimethyl sulfoxide (≥99.9%, Alfa Aesar)
15. Bertheussen, E. et al. Acetaldehyde as an intermediate in the electroreduction
of carbon monoxide to ethanol on oxide-derived copper. Angew. Chem. Int.
Ed. 55, 1450–1454 (2016).
2
or 250ppm (m/m) phenol (≥99%, Sigma-Aldrich) in D O). The electrochemical
reactions for amide formation are provided in Supplementary Table 3. Amide
production was further verified by GC–MS (gas chromatography–mass
spectrometry) (Agilent 59771A). The GC–MS spectral features were determined
by comparing the mass fragmentation patterns with those of the National Institute
of Standards and Technology library.
16. Garza, A. J., Bell, A. T. & Head-Gordon, M. Mechanism of CO reduction at
2
copper surfaces: pathways to C products. ACS Catal. 8, 1490–1499 (2018).
2
17. Pang, Y. et al. Eꢅcient electrocatalytic conversion of carbon monoxide to
propanol using fragmented copper. Nat. Catal. 2, 251–258 (2019).
1
8. Zhuang, T.-T. et al. Copper nanocavities conꢀne intermediates for eꢅcient
electrosynthesis of C3 alcohol fuels from carbon monoxide. Nat. Catal. 1,
946–951 (2018).
18
C O electrolysis. The labelled isotope experiment was performed by using
1
8
18
18
labelled C O gas (95 at% O, Sigma Aldrich) for electrolysis. The C O was
−1
extracted by a 30ml syringe and was injected into the flow cell at 5mlmin by
19. Zhang, H., Li, J., Cheng, M.-J. & Lu, Q. CO electroreduction: current
development and understanding of Cu-based catalysts. ACS Catal. 9,
49–65 (2018).
−
1
3
a syringe pump, along with NH at a flow rate of 10mlmin . Electrolysis was
−2
conducted at a constant current of 200mAcm for 5min and the catholyte was
collected for analysis by GC–MS.
20. Pattabiraman, V. R. & Bode, J. W. Rethinking amide bond synthesis. Nature
4
80, 471–479 (2011).
Model and methods. Electronic structure calculations were performed within
the density functional theory framework, as implemented in the Vienna ab initio
simulation program (a plane-wave pseudopotential package). The exchange
and correlation energies were calculated using the PBE functional within the
generalized gradient approximation. Spin polarization did not have an appreciable
effect on the overall energies. The PBE-D3 method was employed to correct for the
van der Waals interactions of water–water and water–Cu. We simulated the water–
21. Nagib, D. Nitrogen gets radical. Nat. Chem. 11, 396–398 (2019).
22. Jouny, M., Luc, W. & Jiao, F. High-rate electroreduction of carbon monoxide
to multi-carbon products. Nat. Catal. 1, 748–755 (2018).
23. Cheng, T., Xiao, H. & Goddard, W. A. Reaction mechanisms for the
electrochemical reduction of CO to CO and formate on the Cu(100) surface
2
at 298 K from quantum mechanics free energy calculations with explicit
water. J. Am. Chem. Soc. 138, 13802–13805 (2016).
NATꢁRꢂ CHꢂMIꢃTRꢄ | VOL 11 | SEPTEMBER 2019 | 846–851 | www.nature.com/naturechemistry
8
50

NATure CHemisTry
Articles
2
4. Cheng, T., Xiao, H. & Goddard, W. A. Full atomistic reaction mechanism
with kinetics for CO reduction on Cu(100) from ab initio molecular
dynamics free-energy calculations at 298 K. Proc. Natl Acad. Sci. USA 114,
41. Liu, K., Smith, W. A. & Burdyny, T. Introductory guide to assembling and
operating gas diꢂusion electrodes for electrochemical CO reduction. ACS
Energy Lett. 4, 639–643 (2019).
1795–1800 (2017).
42. Lu, X. & Zhao, C. Electrodeposition of hierarchically structured three-
dimensional nickel–iron electrodes for eꢅcient oxygen evolution at high
current densities. Nat. Commun. 6, 6616 (2015).
25. Kim, B., Ma, S., Molly Jhong, H.-R. & Kenis, P. J. A. Inꢄuence of dilute feed
and pH on electrochemical reduction of CO to CO on Ag in a continuous
2
ꢄow electrolyzer. Electrochim. Acta 166, 271–276 (2015).
6. Wang, L. et al. Electrochemical carbon monoxide reduction on
polycrystalline copper: eꢂects of potential, pressure, and pH on
selectivity toward multicarbon and oxygenated products. ACS Catal. 8,
43. Luc, W., Jiang, Z., Chen, J. G. G. & Jiao, F. Role of surface oxophilicity in
copper-catalyzed water dissociation. ACS Catal. 8, 9327–9333 (2018).
2
Acknowledgements
7445–7454 (2018).
F.J. would like to thank W. Luc for illustration assistance and E. Jeng for help with
preparation of the anode. M.J. and J.-J.L. also thank B. Murphy and Z. J. Wang for help
with GC–MS. The experimental work was financially supported by the US Department
of Energy under award no. DE-FE0029868. F.J. also thanks the National Science
Foundation Faculty Early Career Development program (award no. CBET-1350911).
J.-J.L. acknowledges financial support from Chinese Scholarship Council. T.C. and
W.A.G. were supported by the Joint Center for Artificial Photosynthesis, a DOE Energy
Innovation Hub, supported through the Office of Science of the US Department
of Energy under award no. DE-SC0004993. This work used the Extreme Science
and Engineering Discovery Environment, which is supported by National Science
Foundation grant no. ACI-1053575. This research used resources at the 8-ID Beamline of
the National Synchrotron Light Source II, a US Department of Energy Office of Science
User Facility operated by Brookhaven National Laboratory under contract no. DE-
SC0012704. The authors acknowledge E. Stavitski (8-ID Beamline, NSLS-II, Brookhaven
National Laboratory) for assistance in X-ray absorption spectroscopy measurements.
2
7. Lum, Y. W. & Ager, J. W. Stability of residual oxides in oxide-derived copper
1
8
2
catalysts for electrochemical CO reduction investigated with O labeling.
Angew. Chem. Int. Ed. 57, 551–554 (2018).
2
2
3
8. Scott, S. B. et al. Absence of oxidized phases in Cu under CO reduction
conditions. ACS Energy Lett. 4, 803–804 (2019).
9. Luc W. et al. Two-dimensional copper nanosheets for electrochemical
reduction of carbon monoxide to acetate. Nat Catal. 2, 423–430 (2019).
0. Lum, Y. W., Cheng, T., Goddard, W. A. & Ager, J. W. Electrochemical CO
reduction builds solvent water into oxygenate products. J. Am. Chem. Soc.
1
40, 9337–9340 (2018).
3
1. Cheng, T., Xiao, H. & Goddard, W. A. Free-energy barriers and reaction
mechanisms for the electrochemical reduction of CO on the Cu(100) surface,
including multiple layers of explicit solvent at pH 0. J. Phys. Chem. Lett. 6,
4767–4773 (2015).
3
3
3
2. Bagger, A., Arnarson, L., Hansen, M. H., Spohr, E. & Rossmeisl, J.
Electrochemical CO reduction: a property of the electrochemical interface.
J. Am. Chem. Soc. 141, 1506–1514 (2019).
Author contributions
F.J. conceived the idea and supervised the project. M.J. and J.J.-L. performed the
electrolysis experiments, analysed the data and wrote the first draft of the manuscript.
B.H.K. performed the SEM characterization. T.C. and W.A.G. performed the
computational modelling studies. All the authors contributed to the discussion of the
results and preparation of the manuscript. M.J., J.J.L. and T.C. have the right to list
themselves first in the bibliographic documents.
3. Calle-Vallejo, F. & Koper, M. T. M. ꢁeoretical considerations on the
2
electroreduction of CO to C species on Cu(100) electrodes. Angew. Chem.
Int. Ed. 52, 7282–7285 (2013).
4. Feng, X. F., Jiang, K. L., Fan, S. S. & Kanan, M. W. A direct grain-boundary-
activity correlation for CO electroreduction on Cu nanoparticles. ACS Cent.
Sci. 2, 169–174 (2016).
3
3
5. Dean J. A., Lange N. A. Lange’s Handbook of Chemistry (McGraw-Hill, 1999).
6. Lundberg, H., Tinnis, F., Selander, N. & Adolfsson, H. Catalytic amide
formation from non-activated carboxylic acids and amines. Chem. Soc. Rev.
Competing interests
M.J., J.-J.L. and F.J. have filed a patent application (international patent application
number: PCT/US 19/27012) that is based on the discovery presented in this work.
4
3, 2714–2742 (2014).
3
3
7. Watts, J. C. & Larson, P. A. “Dimethylacetamide” in Kirk–Othmer
Encyclopedia of Chemical Technology Online (John C. Wiley & Sons, 2002).
https://doi.org/10.1002/0471238961.0409130523012020.a01.pub2
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41557-019-0312-z.
8. McEnaney, J. M. et al. Ammonia synthesis from N
2
and H
2
O using a lithium
cycling electriꢀcation strategy at atmospheric pressure. Energy Environ. Sci.
Reprints and permissions information is available at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to W.A.G. or F.J.
1
0, 1621–1630 (2017).
3
4
9. Jiao, F. & Xu, B. J. Electrochemical ammonia synthesis and ammonia fuel
cells. Adv. Mater. 31, 1805173 (2019).
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
0. Chen, J. G. et al. Beyond fossil fuel-driven nitrogen transformations. Science
3
60, eaar6611 (2018).
© The Author(s), under exclusive licence to Springer Nature Limited 2019
NATꢁRꢂ CHꢂMIꢃTRꢄ | VOL 11 | SEPTEMBER 2019 | 846–851 | www.nature.com/naturechemistry
8
51