We report on the direct conversion of carbon dioxide (CO2) in a photoelectrochemical cell consisting of germanium doped gallium nitride nanowire anode and copper (Cu) cathode. Various products including methane (CH4), carbon monoxide (CO), and formic acid (HCOOH) were observed under light illumination. A Faradaic efficiency of ∼10% was measured for HCOOH. Furthermore, this photoelectrochemical system showed enhanced stability for 6 h CO2 reduction reaction on low cost, large area Si substrates.

Carbon dioxide (CO2), a stable molecule that is generated by the consumption of conventional fossil fuel, is a primary cause of global warming. During the past decades, atmospheric CO2 has been raised to a critical level with the growth of the global economy and population. The efficient conversion of CO2 into hydrocarbon fuel by using a renewable energy source has been intensively studied as a potential technology to decrease atmospheric CO2 concentrations.1–4 The best available option for this conversion is solar power, which is an abundant, clean, and sustainable energy source.5 In this regard, photocatalysis (PC) and photoelectrocatalysis (PEC) have been promising strategies because they can directly utilize solar energy.4,6–8 Comparing to PC, a PEC system can effectively separate the photo-generated holes and electrons to achieve higher photoelectric conversion efficiency, and the oxidation and reduction products can be collected at anodic and cathodic electrodes, respectively. The use of PEC systems for CO2 reduction has been reported.9–11 However, the conversion activity is still extremely low.

At the heart of a PEC cell is the photoelectrocatalyst, which converts incident photons into electron–hole pairs and then catalyzes oxidation or reduction reactions at the solid-liquid interface. Compared to conventional metal-oxide based photoelectrocatalysts, metal-nitride based catalysts (e.g., Ga(In)N) offer several important advantages, including a high level of stability in aqueous media and tunable energy bandgap across nearly the entire solar spectrum.12–22 Yotsuhashi et al. showed the promising potential of GaN for CO2 reduction,23–26 wherein GaN epilayers grown on sapphire or GaN substrate were used as photoanodes. For practical applications, however, it is highly desired that low cost, large area substrates such as Si can be utilized, instead of the expensive sapphire or GaN substrate. Moreover, the photoanode stability is severely limited by the presence of extensive defects and dislocations (∼108 cm−3, or higher) of conventional GaN-based epilayers. Compared to planar or powder catalysts, defect-free Ga(In)N nanowire photocatalysts can be formed directly on foreign substrates, including Si and SiOx.27 Such nanowire photocatalysts can exhibit significantly enhanced light absorption and carrier extraction efficiency, due to the large surface-to-volume ratios. Moreover, electrolyte can readily diffuse in the space among nanowires, thus leading to improved photoelectrocatalytic performance.18,28 Recent studies have revealed the extraordinary potential of Ga(In)N nanowires for solar-to-hydrogen conversion20,29–32 and for photo induced C–H bond activation.33,34 To date, however, there have been no reports on the use of GaN-nanowire based PEC system for CO2 reduction to the best of our knowledge. When used as a photoanode, n-type GaN nanowires are desired because of the upward surface band bending, which can promote photo-generated holes to the surface to take part in the oxidation reaction.30 The increased conductivity with n-type dopant incorporation also facilitates the transport of carriers and enhances the redox reactions.35 

Herein, we report on the direct CO2 conversion with the PEC reduction strategy, which involves the use of germanium doped gallium nitride nanowire arrays grown on Si substrate and copper (Cu) plate as anodic and cathodic electrodes, respectively. Various products including methane (CH4), carbon monoxide (CO), and formic acid (HCOOH) were observed. A Faradaic efficiency of ∼10% was measured for HCOOH. Moreover, the stability towards CO2 reduction is also excellent during the reaction (∼6 h), compared to previously reported GaN planar electrodes.

In this experiment, catalyst-free GaN nanowire arrays were grown directly on Si (111) substrate by plasma-assisted molecular beam epitaxy (PA-MBE) under nitrogen rich conditions. The growth conditions included a substrate temperature of 750 °C, nitrogen flow rate of 1.0 sccm, plasma forward power of 350 W, and Ga beam equivalent pressure of ∼6 × 10−8 Torr.13,36,37 The Ge effusion cell temperature was 1130 °C, which corresponds to Ge beam equivalent pressure (BEP) of ∼10−11 Torr, which is similar the previous reports.35,38 Due to the suitable atom size, Ge can be incorporated into GaN with less lattice distortion and can provide better nanowire morphology, compared to Si-doping.38 In this study, both Ge-doped and nominally undoped GaN nanowires were studied. Figure 1(a) shows the scanning electron microscopy (SEM) image of Ge doped GaN nanowires grown on a Si (111) substrate. The nanowires are vertically aligned to the substrate, with lengths and diameters in the ranges of 800 nm and 40–140 nm, respectively. Figures 1(b) and 1(c) show the transmission electron microscopy (TEM) images of Ge doped and undoped GaN nanowires. Comparing the two samples, it is noted that there is no obvious difference of morphology induced by Ge doping. The X-ray diffraction (XRD) results shown in Figure 1(d) confirm that both the Ge doped and undoped nanowires are grown along the c-axis by the presence of GaN (002) and (004) peaks in the XRD pattern, which also indicates the negligible difference of crystal structures between Ge doped and undoped samples. The optical property of the GaN is not affected by the Ge doping, which is supported by the photoluminescence (PL) results (not shown).

FIG. 1.

(a) SEM image of Ge doped GaN nanowires grown directly on Si(111) substrate. TEM images of (b) Ge doped and (c) undoped GaN nanowires. (d) XRD of the Ge doped and undoped GaN nanowires.

FIG. 1.

(a) SEM image of Ge doped GaN nanowires grown directly on Si(111) substrate. TEM images of (b) Ge doped and (c) undoped GaN nanowires. (d) XRD of the Ge doped and undoped GaN nanowires.

Close modal

A conventional three-electrode system was used in all the PEC measurements, schematically shown in Figure 2, which consists of GaN nanowires as the working electrode, a Cu plate (∼1 cm2) as the counter electrode, and an Ag/AgCl (filled with saturated KCl electrolyte) double junction electrode as the reference electrode. The Cu electrode can deliver a range of reaction products such as CO, HCOOH, and CH4, due to its medium hydrogen overpotential.9,39,40 An alloy of Ga–In eutectic (Sigma-Aldrich) was deposited on the backside of Si (111) substrate to form ohmic contact, which was subsequently connected with a Cu wire using silver paste. The entire sample except the nanowire surface was then covered by insulating epoxy to eliminate any current leakage. The nanowire sample surface area is ∼0.8 cm2. All the PEC measurements were carried out in an H type quartz reactor with the cathode and anode chambers separated by a Nafion film as the cation exchange membrane. The cathode and anode reactions were conducted in 1M KHCO3 (pH ∼ 8) and a mixture of 1M KBr and 25 mM sodium citrate (pH ∼ 7), respectively. The chamber for the cathode reaction was sealed, and the electrolytes were prepared by bubbling with pure CO2 for 30 min before the measurements. The current and potential were measured by an Interface 1000 potentiostat (Gamry Instruments). A 300 W Xenon lamp (Excelitas Technologies, light intensity ∼ 1.3 W cm−2 on the sample) was used as the light source.

FIG. 2.

Schematic diagram of the experimental setup used for the photoelectrochemical measurements. WE, RE, and CE refer to the working electrode, reference electrode, and counter electrode, respectively.

FIG. 2.

Schematic diagram of the experimental setup used for the photoelectrochemical measurements. WE, RE, and CE refer to the working electrode, reference electrode, and counter electrode, respectively.

Close modal

The PEC performance of CO2 reduction was first investigated by linear sweep voltammetry (LSV). Figure 3(a) shows the LSV of the Ge doped and undoped GaN nanowires under dark and illumination conditions at a scan rate of 20 mV s−1. There are negligible currents in the potential range from −1.2 V to 1.0 V in the dark condition (green and blue lines). In comparison, distinct photocurrent with an onset potential −1.17 V appears when Ge doped GaN nanowire photoanode was illuminated (black line). The current increases with the increase of potential and reaches ∼6.7 mA cm−2 at 1 V. The onset potential for undoped GaN nanowire photoanode was −1.15 V, which is close to the Ge doped GaN nanowire photoanode. However, the saturation photocurrent of undoped GaN nanowire photoanode was only ∼4.1 mA cm−2 at 1 V, indicating GaN nanowire with Ge doping is a promising photoanode material, due to enhanced conductivity.32,37 The electrochemical impedance spectroscopy (EIS) also confirms that the conductivity of GaN nanowires is enhanced by Ge doping, which is in agreement with the carrier density increase of Ge doped GaN nanowires calculated from Mott-Schottky plot (not shown). The lack of obvious current saturation is likely related to the enhanced nonradiative recombination induced by Ge doping, due to the relatively high doping concentration. With increasing bias voltage, the nonradiative recombination can be reduced, due to the faster carrier transfer rate to the nanowire/electrolyte interface, thereby leading to a small increase in the current density. We also observed similar LSV results in 1M HBr under one sun (∼0.1 W cm−2) irradiation (not shown), further confirming the difference in the I-V curves for the Ge-doped and undoped GaN nanowires is not due to variations in excitation conditions but directly related to their intrinsic properties. For the Si (111) substrate, even under light illumination, no obvious photocurrent was observed in the range from −1.2 V to 1 V (not shown), confirming that the measured photocurrent is generated by GaN nanowires rather than the Si (111) substrate. We have further investigated the stability of both Ge doped and undoped GaN nanowire photoanodes. Figure 3(b) shows the current-time curve of Ge doped and undoped GaN nanowire photoanodes at a constant voltage of 0 V vs. Ag/AgCl. It is seen that the current of both systems show small reduction with time. In 6 h, the photocurrent loss of both Ge doped and undoped GaN nanowire photoanodes is ∼30%. These results demonstrate that the GaN nanowires possess excellent stability as the photoanode, and the Ge doping does not affect the stability of GaN nanowires. The slow degradation is likely related to the etching of the GaN/Si interface, rather than GaN nanowires.41 

FIG. 3.

(a) Linear sweep voltammetry of Ge doped and undoped GaN nanowire working electrode with Cu counter electrode (vs. Ag/AgCl) at a scan rate of 20 mV s−1. The cathode and anode electrolytes are 1M KHCO3 and a mixture of 1M KBr and 25 mM sodium citrate, respectively. (b) Chronoamperometric curves of Ge doped and undoped GaN nanowire working electrodes with Cu counter electrode at 0 V (vs. Ag/AgCl).

FIG. 3.

(a) Linear sweep voltammetry of Ge doped and undoped GaN nanowire working electrode with Cu counter electrode (vs. Ag/AgCl) at a scan rate of 20 mV s−1. The cathode and anode electrolytes are 1M KHCO3 and a mixture of 1M KBr and 25 mM sodium citrate, respectively. (b) Chronoamperometric curves of Ge doped and undoped GaN nanowire working electrodes with Cu counter electrode at 0 V (vs. Ag/AgCl).

Close modal

To gain further insight into the photoelectrocatalytic activity of the Ge doped and undoped GaN nanowire photoanodes for CO2 reduction, we analyzed the reaction products in the cathode after 6 h chronoamperometric measurements. CH4 and CO were measured using the gas chromatograph equipped with flame ionization detector (FID) (Shimadzu GC-2014 with Porapak-Q-80/100 column), and H2 was determined using a gas chromatograph equipped with thermal conductivity detector (TCD) (Shimadzu GC-8A with molecular sieve column). The cathode electrolyte was analyzed by nuclear magnetic resonance (NMR) and gas chromatograph-mass spectrum. Figure 4(a) shows all the products generated in the PEC systems with Ge doped and undoped GaN nanowires photoanodes, respectively. Several different reduction products, including CH4, CO, HCOOH, and H2, were clearly measured. The amounts of CH4, CO, HCOOH, and H2 in the Ge doped GaN nanowires system are 0.16, 2.21, 7.24, and 60.6 μmol h−1 cm−2, respectively, normalized by the surface area of the photoanode. However, no CH4 signal was detected in the undoped GaN nanowires system, and the amounts of CO, HCOOH, and H2 decrease to 0.3, 2.37, and 51.1 μmol h−1 cm−2, respectively. According to the amounts of CO2 reduction products, the PEC system equipped with Ge doped GaN nanowires exhibits enhanced capability towards CO2 conversion compared to undoped GaN nanowires system. Moreover, we observed the existence of CH3OH in the electrolyte of cathode in both PEC systems, but it cannot be quantified due to the trace-level amount and the limitation of instruments.

FIG. 4.

(a) Illustration of the amount of reduction reaction products of Ge doped and undoped GaN nanowire working electrodes with Cu counter electrode, including CH4, CO, HCOOH, and H2 measured in 6 h. (b) The measured Faradaic efficiency for different reaction products.

FIG. 4.

(a) Illustration of the amount of reduction reaction products of Ge doped and undoped GaN nanowire working electrodes with Cu counter electrode, including CH4, CO, HCOOH, and H2 measured in 6 h. (b) The measured Faradaic efficiency for different reaction products.

Close modal

The selectivity of CO2 reduction in PEC systems should be concerned seriously because there is an intensive competition between the various CO2 conversion processes as well as the H2 evolution reaction in the cathode, which has become one of the biggest challenges for the development of this field.42 Therefore, it is necessary to evaluate the Faradaic efficiency of CO2 reduction products. Figure 4(b) shows the Faradaic efficiency of different products generated in the cathode. The Faradaic efficiencies of CH4, CO, HCOOH, and H2 in Ge doped GaN nanowires system are ∼0.9%, 3.1%, 10.2%, and 85.5%, respectively. In contrast, the Faradaic efficiencies of CO, HCOOH, and H2 in Ge doped GaN nanowires system are ∼0.54%, 4.35%, and 94.4%. The overall performance of Ge doped nanowire PEC CO2 reduction is also comparable to previous reports, wherein expensive sapphire and/or GaN substrates were utilized.23,24,26

In summary, we have demonstrated the use of Ge doped GaN nanowire electrode for the PEC reduction of CO2 into various valuable hydrocarbons, including CH4 and HCOOH. Compared to the previously reported planar semiconductor electrodes, such defect-free nanowires showed enhanced stability and, importantly, scalability, due to the direct integration with low cost, large area Si substrates. Moreover, with the incorporation of indium in the nanowire electrode, such a PEC system can operate efficiently under visible light irradiation, which is currently under investigation.

This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Climate Change and Emissions Management (CCEMC) Corporation. The authors wish to thank Mr. Y. Wang and Dr. S. Zhao at McGill University for their help with the measurements and the MBE growth, respectively.

1.
A.
Bachmeier
,
S.
Hall
,
S. W.
Ragsdale
, and
F. A.
Armstrong
,
J. Am. Chem. Soc.
136
,
13518
(
2014
).
2.
J.
Bonin
,
M.
Robert
, and
M.
Routier
,
J. Am. Chem. Soc.
136
,
16768
(
2014
).
3.
J.
Ettedgui
,
Y.
Diskin-Posner
,
L.
Weiner
, and
R.
Neumann
,
J. Am. Chem. Soc.
133
,
188
(
2010
).
4.
S. N.
Habisreutinger
,
L.
Schmidt-Mende
, and
J. K.
Stolarczyk
,
Angew. Chem., Int. Ed.
52
,
7372
(
2013
).
5.
J.
Ronge
,
T.
Bosserez
,
D.
Martel
,
C.
Nervi
,
L.
Boarino
,
F.
Taulelle
,
G.
Decher
,
S.
Bordiga
, and
J. A.
Martens
,
Chem. Soc. Rev.
43
,
7963
(
2014
).
6.
T.
Inoue
,
A.
Fujishima
,
S.
Konishi
, and
K.
Honda
,
Nature
277
,
637
(
1979
).
7.
K. R.
Thampi
,
J.
Kiwi
, and
M.
Gratzel
,
Nature
327
,
506
(
1987
).
8.
B.
Kumar
,
M.
Llorente
,
J.
Froehlich
,
T.
Dang
,
A.
Sathrum
, and
C. P.
Kubiak
,
Annu. Rev. Phys. Chem.
63
,
541
(
2012
).
9.
G.
Magesh
,
E. S.
Kim
,
H. J.
Kang
,
M.
Banu
,
J. Y.
Kim
,
J. H.
Kim
, and
J. S.
Lee
,
J. Mater. Chem. A
2
,
2044
(
2014
).
10.
J.
Cheng
,
M.
Zhang
,
G.
Wu
,
X.
Wang
,
J.
Zhou
, and
K.
Cen
,
Sol. Energy Mater. Sol. Cells
132
,
606
(
2015
).
11.
K.
Rajeshwar
,
N. R.
de Tacconi
,
G.
Ghadimkhani
,
W.
Chanmanee
, and
C.
Janáky
,
ChemPhysChem
14
,
2251
(
2013
).
12.
K.
Aryal
,
B. N.
Pantha
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
96
,
052110
(
2010
).
13.
J.-S.
Hwang
,
T.-Y.
Liu
,
S.
Chattopadhyay
,
G.-M.
Hsu
,
A. M.
Basilio
,
H.-W.
Chen
,
Y.-K.
Hsu
,
W.-H.
Tu
,
Y.-G.
Lin
,
K.-H.
Chen
,
C.-C.
Li
,
S.-B.
Wang
,
H.-Y.
Chen
, and
L.-C.
Chen
,
Nanotechnology
24
,
055401
(
2013
).
14.
W.-H.
Tu
,
Y.-K.
Hsu
,
C.-H.
Yen
,
C.-I.
Wu
,
J.-S.
Hwang
,
L.-C.
Chen
, and
K.-H.
Chen
,
Electrochem. Commun.
13
,
530
(
2011
).
15.
D.-H.
Tu
,
H.-C.
Wang
,
P.-S.
Wang
,
W.-C.
Cheng
,
K.-H.
Chen
,
C.-I.
Wu
,
S.
Chattopadhyay
, and
L.-C.
Chen
,
Int. J. Hydrogen Energy
38
,
14433
(
2013
).
16.
P. D.
Tran
,
L. H.
Wong
,
J.
Barber
, and
J. S. C.
Loo
,
Energy environ sci
5
,
5902
(
2012
).
17.
H. S.
Jung
,
Y. J.
Hong
,
Y.
Li
,
J.
Cho
,
Y.-J.
Kim
, and
G.-C.
Yi
,
ACS Nano
2
,
637
(
2008
).
18.
D.
Wang
,
A.
Pierre
,
M. G.
Kibria
,
K.
Cui
,
X.
Han
,
K. H.
Bevan
,
H.
Guo
,
S.
Paradis
,
A.-R.
Hakima
, and
Z.
Mi
,
Nano Lett.
11
,
2353
(
2011
).
19.
S.
Schäfer
,
S. A.
Wyrzgol
,
R.
Caterino
,
A.
Jentys
,
S. J.
Schoell
,
M.
Hävecker
,
A.
Knop-Gericke
,
J. A.
Lercher
,
I. D.
Sharp
, and
M.
Stutzmann
,
J. Am. Chem. Soc.
134
,
12528
(
2012
).
20.
L.
Caccamo
,
J.
Hartmann
,
C.
Fàbrega
,
S.
Estradé
,
G.
Lilienkamp
,
J. D.
Prades
,
M. W. G.
Hoffmann
,
J.
Ledig
,
A.
Wagner
,
X.
Wang
,
L.
Lopez-Conesa
,
F.
Peiró
,
J. M.
Rebled
,
H.-H.
Wehmann
,
W.
Daum
,
H.
Shen
, and
A.
Waag
,
ACS Appl. Mater. Interfaces
6
,
2235
(
2014
).
21.
M. G.
Kibria
,
F. A.
Chowdhury
,
S.
Zhao
,
B.
AlOtaibi
,
M. L.
Trudeau
,
H.
Guo
, and
Z.
Mi
,
Nat. Commun.
6
,
6797
(
2015
).
22.
K.
Maeda
and
K.
Domen
,
J. Phys. Chem. C
111
,
7851
(
2007
).
23.
S.
Yotsuhashi
,
M.
Deguchi
,
Y.
Zenitani
,
R.
Hinogami
,
H.
Hashiba
,
Y.
Yamada
, and
K.
Ohkawa
,
Appl. Phys. Express
4
,
117101
(
2011
).
24.
S.
Yotsuhashi
,
M.
Deguchi
,
Y.
Zenitani
,
R.
Hinogami
,
H.
Hashiba
,
Y.
Yamada
, and
K.
Ohkawa
,
Jpn. J. Appl. Phys.
51
,
02BP07
(
2012
).
25.
S.
Yotsuhashi
,
M.
Deguchi
,
H.
Hashiba
,
Y.
Zenitani
,
R.
Hinogami
,
Y.
Yamada
, and
K.
Ohkawa
,
Appl. Phys. Lett.
100
,
243904
(
2012
).
26.
S.
Yotsuhashi
,
M.
Deguchi
,
Y.
Yamada
, and
K.
Ohkawa
,
AIP Adv.
4
,
067135
(
2014
).
27.
S.
Zhao
,
M. G.
Kibria
,
Q.
Wang
,
H. P. T.
Nguyen
, and
Z.
Mi
,
Nanoscale
5
,
5283
(
2013
).
28.
E.
Garnett
and
P.
Yang
,
Nano Lett.
10
,
1082
(
2010
).
29.
M. G.
Kibria
,
H. P. T.
Nguyen
,
K.
Cui
,
S.
Zhao
,
D.
Liu
,
H.
Guo
,
M. L.
Trudeau
,
S.
Paradis
,
A.-R.
Hakima
, and
Z.
Mi
,
ACS Nano
7
,
7886
(
2013
).
30.
M. G.
Kibria
,
S.
Zhao
,
F. A.
Chowdhury
,
Q.
Wang
,
H. P.
Nguyen
,
M. L.
Trudeau
,
H.
Guo
, and
Z.
Mi
,
Nat. Commun.
5
,
3825
(
2014
).
31.
B.
AlOtaibi
,
H. P. T.
Nguyen
,
S.
Zhao
,
M. G.
Kibria
,
S.
Fan
, and
Z.
Mi
,
Nano Lett.
13
,
4356
(
2013
).
32.
B.
AlOtaibi
,
M.
Harati
,
S.
Fan
,
S.
Zhao
,
H. P. T.
Nguyen
,
M. G.
Kibria
, and
Z.
Mi
,
Nanotechnology
24
,
175401
(
2013
).
33.
L.
Li
,
X.
Mu
,
W.
Liu
,
X.
Kong
,
S.
Fan
,
Z.
Mi
, and
C.-J.
Li
,
Angew. Chem., Int. Ed.
53
,
14106
(
2014
).
34.
L.
Li
,
S.
Fan
,
X.
Mu
,
Z.
Mi
, and
C.-J.
Li
,
J. Am. Chem. Soc.
136
,
7793
(
2014
).
35.
M.
Schafer
,
M.
Gunther
,
C.
Langer
,
J.
MuBener
,
M.
Feneberg
,
P.
Uredat
,
M. T.
Elm
,
P.
Hille
,
J.
Schörmann
,
J.
Teubert
,
T.
Henning
,
P. J.
Klar
, and
M.
Eickhoff
,
Nanotechnology
26
,
135704
(
2015
).
36.
H. P. T.
Nguyen
,
S.
Zhang
,
K.
Cui
,
X.
Han
,
S.
Fathololoumi
,
M.
Couillard
,
G. A.
Botton
, and
Z.
Mi
,
Nano Lett.
11
,
1919
(
2011
).
37.
J.
Wallys
,
S.
Hoffmann
,
F.
Furtmayr
,
J.
Teubert
, and
M.
Eickhoff
,
Nanotechnology
23
,
165701
(
2012
).
38.
J.
Schörmann
,
P.
Hille
,
M.
Schäfer
,
J.
Müßener
,
P.
Becker
,
P. J.
Klar
,
M.
Kleine-Boymann
,
M.
Rohnke
,
M.
de la Mata
,
J.
Arbiol
,
D. M.
Hofmann
,
J.
Teubert
, and
M. J.
Eickhoff
,
J. Appl. Phys.
114
,
103505
(
2013
).
39.
Y.
Hori
,
K.
Kikuchi
, and
S.
Suzuki
,
Chem. Lett.
14
,
1695
(
1985
).
40.
M.
Jitaru
,
D. A.
Lowy
,
M.
Toma
,
B. C.
Toma
, and
L. J.
Oniciu
,
Appl. Electrochem.
27
,
875
(
1997
).
41.
S.
Fan
,
B.
AlOtaibi
,
S. Y.
Woo
,
Y.
Wang
,
G. A.
Botton
, and
Z.
Mi
,
Nano Lett.
15
,
2721
(
2015
).
42.
J.
Qiao
,
Y.
Liu
,
F.
Hong
, and
J.
Zhang
,
Chem. Soc. Rev.
43
,
631
(
2014
).