Electrochemical reduction of CO2 to ethylene on Cu/Cu: XO-GO composites in aqueous solution

Article abstract of DOI:10.1039/d0ra02754e

Here, we present fabrication of Graphene oxide (GO) supported Cu/Cu_xO nano-electrodeposits which can efficiently and selectively electroreduce CO₂ into ethylene with a faradaic efficiency (F.E) of 34% and a conversion rate of 194 mmol g^-1 h^-1 at -0.985 V vs. RHE. The effect of catalyst morphology, working electrode fabricational techniques, the extent of metal-GO interaction and the oxide content in Cu/Cu_xO, was studied in detail so as to develop a protocol for the fabrication of an active, stable and selective catalyst for efficient electro-production of ethylene from CO₂. Moreover, a detailed comparative study about the effect of the GO support, and the nature of the cathodic collection substrate used for the electro-deposition is presented.

Full text of DOI:10.1039/d0ra02754e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Electrochemical reduction of CO2 to ethylene on
Cu/CuxO-GO composites in aqueous solution†
Cite this: RSC Adv., 2020, 10, 17572
a
b
Nusrat Rashid,
Mohsin Ahmad Bhat,
U. K. Goutamc
a
*
and Pravin Popinand Ingole
Here, we present fabrication of Graphene oxide (GO) supported Cu/CuxO nano-electrodeposits which can
efficiently and selectively electroreduce CO2 into ethylene with a faradaic efficiency (F.E) of 34% and
a conversion rate of 194 mmol gꢀ1 hꢀ1 at ꢀ0.985 V vs. RHE. The effect of catalyst morphology, working
electrode fabricational techniques, the extent of metal–GO interaction and the oxide content in Cu/
CuxO, was studied in detail so as to develop a protocol for the fabrication of an active, stable and
Received 25th March 2020
Accepted 24th April 2020
selective catalyst for efficient electro-production of ethylene from CO2. Moreover,
a detailed
DOI: 10.1039/d0ra02754e
comparative study about the effect of the GO support, and the nature of the cathodic collection
rsc.li/rsc-advances
substrate used for the electro-deposition is presented.
materials also plays a role in determining the selectivity as well
as total efficiency in reducing CO2.15
Introduction
Overall, the reported studies suggest that the optimization
and exploitation of the following aspects would be important
toward further improvements in ERC activity of Cu based elec-
trocatalysts; (i) altering the surface structure (morphology) of
Cu particles to increase the number of surface active sites, this
may also include the porous nature of the surface (ii) ne tuning
of the relative extent of metallic Cu to the oxide contents on Cu
surface, and (iii) selection of an appropriate support material
for the Cu deposits. In this regard, design and development of
ERC catalyst based on Cu-nanostructures with optimum oxide
content and an appropriate support material is an important
step towards effective and selective conversion of CO2 into
hydrocarbons. Among the various supports, graphene and its
related forms such as graphene oxide (GO) have been exten-
sively used for their advantageous physico-chemical properties
including16 high surface area, good electronic conduction
characteristics and easy to functionalize structure. We along
with other researchers have reported the successful immobili-
zation of noble metallic nanoparticles on graphene/GO sheets
that signicantly enhance the catalytic and sensing activities of
these nano-composites as the GO supports acts as both electron
acceptor as well as donor.17–20 The question is how well the GO
as support can be utilized to control the product of ERC
generated over Cu-NPs in contact with the graphene support.
Hence, in order to understand the effect of Cu-nanoparticles
supported on GO sheets for electrochemical conversion of
CO2 into hydrocarbons is important to probe. Motivated with
these properties of GO, we successfully attempted the synthesis
of Cu/CuxO-GO composite and its utilization for electro-
catalytic activity towards CO2 reduction. Furthermore, we
studied the effect of three different aspects of the catalysts viz.
The amount of atmospheric carbon dioxide is continuously
increasing, for example, CO2 concentration in the atmosphere
was recorded at its highest in May 2019, with the average at
414.7 parts per million (ppm) at NOAA's Mauna Loa Atmo-
spheric Baseline Observatory. In response, intense research
activity has been devoted globally to not only deal with carbon
emissions but also to utilize CO2 as a resource for benecial
processes through the development of economically viable
technologies for environmental remediation. In this regard, the
electrochemical reduction of CO2 (ERC) has attracted consid-
erable attention to address the problem of global climate
change.1–4 In particular, conversion of CO2 to hydrocarbons like
methane,5–7 ethane, ethylene, etc. which are energy dense fuels
and chemicals, would be an added advantage.
However, due to higher overpotentials, less selectivity and
low stability, the ERC needs catalysts capable of selectively
reducing CO2 at less overpotentials and with long run times.8,9
Various copper-based materials have been tested for electro-
chemical reduction of CO2 into hydrocarbons, and it has been
observed that the morphology and surface area of catalyst
directly affects the overall activity of the ERC.10–12 However, due
to the low stability of copper and copper oxide for ERC, the use
of support materials like carbon nanostructures have been
employed both to enhance the selectivity and to increase the
robustness of the catalyst.13,14 Furthermore, the support
aIndian Institute of Technology Delhi, 110016, India. E-mail: ppingole@chemistry.iitd.
ac.in
bUniversity of Kashmir Srinagar, 190006, India
cRaja Ramanna Centre for Advanced Technology Indore, 452013, India
† Electronic supplementary information (ESI) available. See DOI:
10.1039/d0ra02754e
17572 | RSC Adv., 2020, 10, 17572–17581
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
catalyst morphology, the extent of metal–GO interaction and the polished with a smooth polishing paper followed by electro-
oxide content in the Cu/CuxO on the electro-catalytic activity chemical polishing in 85% phosphoric acid, was used as a sacri-
towards ERC. Furthermore, an insight into the effect of fabri- cial anode. Pt plate (2 cm ꢁ 2 cm) cleaned in distilled water by
cation techniques for both catalyst and working electrode has ultra-sonication for 5 minutes was used as a collector (i.e. for
been attempted.
electrodeposition) electrode/cathode. A positive load of 15 V was
For this, a simple and scalable electrochemical deposition applied (Keithley 2231A-30-3 triple channel DC power supply) for
ꢂ
ꢂ
approach was used to fabricate the different electrodes based on 60 minutes at a xed temperature of 40 C (ꢃ1 C) in an elec-
Cu/CuxO and Cu/CuxO supported on GO sheets under varying trolytic bath containing 200 mM tri-sodium citrate aqueous
conditions. For example, pCu/CuxO,
acquired aer drying of electrodeposits (obtained with Pt-plate
a
powdered sample solution with electrolyte contents stirring continuously.
Aer 60 minutes the electrodeposits were tapped off from
cathode) was mixed with isopropanol and water mixture to platinum foil and collected in centrifuge tubes, washed multiple
make a catalyst ink which was then pressure sprayed on carbon times with copious amount of water followed by ethanol and
ber paper (CFP) to obtain pCu/CuxO modied CFP electrodes. nally by acetone before drying it under vacuum for 6 hours. The
Similarly, a catalyst ink of pCu/CuxO-GO, a powdered sample powder obtained as such was named as pCu/CuxO (where p
synthesized with Pt plate as a cathode but in presence of GO represents samples obtained in powder form at Pt plate cathode).
suspension was pressure sprayed on CFP to get pCu/CuxO-GO To synthesize the composite, 0.3 mg mLꢀ1 GO was added to the
modied electrode. For fabricating fCu/CuxO and fCu/CuxO- 200 mM aqueous tri-sodium citrate solution with all other
GO, the respective samples were directly electrodeposited on synthesis parameters like applied voltage, temperature of bath,
CFP from their electrodeposition baths, using CFP as a cathode concentration of tri-sodium citrate, and anode and cathode kept
(i.e. without removing the electrodeposit from CFP surface). For unaltered, the sample obtained was named as pCu/CuxO-GO
the comparison of mode of electrode fabrication, a few elec- (powder sample obtained with Pt as cathode and GO as
trodes were also fabricated in a layer-by-layer fashion. Thus, a support material). These materials were then pressure sprayed
overall, we have fabricated six different electrodes and on CFP for fabricating electrode for studying ERC.
comparatively studied their catalytic activity towards ERC.
Thin lms. For the fabrication of lms, a similar experimental
Among the studied samples, pCu/CuxO-GO nanocomposite setup was used except that the lms were now directly electro-
shows the best activity with an overall faradaic efficiency of ca. deposited on CFP in place of Pt for three minutes. Thus, obtained
72% at an applied potential of ꢀ0.985 V vs. RHE with ethylene modied CFP was rinsed with water and dried under the ow of
production at a rate of 194.4 mmol gꢀ1 hꢀ1 or 34% of F.E.
nitrogen gas, and named as fCu/CuxO (where f signies the
materials were obtained as lms directly on CFP). To synthesize
fCu/CuxO-GO (lm directly deposited at CFP with GO as support
material) composite, similar electro-deposition was carried out
but in presence of GO suspension (0.3 mg mLꢀ1) in otherwise
a replica of other synthesis parameters. Further to check the
effect of step-by-step electro-deposition, a CFP was rst modied
with GO (by pressure spraying) and then Cu/CuxO was electro-
deposited on it, and the electrode is named as fCu/CuxO@GO (a
thin lm of Cu/CuxO electrodeposited on a layer of GO which was
pressure sprayed on a CFP).
Experimental section
Materials
All the chemicals used for the synthesis and electrochemical
measurements were of analytical reagent grade and were used as
such without further purication unless otherwise mentioned.
Trisodium citrate (TSC) (99%) was procured from Merck India,
Limited. Graphene oxide dispersion in water (4 mg mLꢀ1) and
potassium hydrogen carbonate (KHCO3) (99.9%) were purchased
from Sigma Aldrich. Ultrapure CO2 and N2 gases were procured
from Sigma Gases, New Delhi and before using for electro-
chemical measurements were passed through a lter equipped
with an oxygen trap, gas purier and moisture trap to remove the
traces of impurities (Prama Instruments Pvt. Ltd.). Copper metal
sheet (99.9% pure, metal basis) was purchased from Alfa Aesar.
Carbon bre paper (CFP) sheets (TGPH-120) were procured from
Nickunj Exim Entp. Pvt. Ltd. India.
Material characterization
The solid-state properties of materials were analysed by X-ray
diffractometer (XRD) Bruker model D8 Advance with Cu Ka (l
˚
¼ 1.54 A) and a scan speed of 0.2 seconds/increment with 2q
range of 10ꢂ to 60ꢂ. A reference method was used to record XRD
patterns of the materials where an equal amount of sample was
spray dried on CFP for the powdered samples, and thin lm
samples were used as such for the measurements. The eld
emission scanning electron microscope (FESEM) attached with
Oxford-EDS system IE 250X max 80 with latest 80 mm2SDD
Synthesis of Cu/CuxO-GO composites in powder form and
thin lms
The Cu/CuxO-GO composites were prepared in two different detector model FEI Quanta 200 was used to image the
forms, powder and thin lms by using a DC electrophoretic morphology and elemental composition of the prepared mate-
anodization of metallic copper foil in presence of GO suspen- rials. Elemental mapping was done on Hitachi High technol
sion in aqueous tri-sodium citrate solution with slight modi- system. Renishaw plc micro Raman spectrometer under 514 nm
cation in our previously reported recipe.21
laser was used to get the characteristics of the composites. For
Powder form. Typically, for the synthesis of Cu/CuxO and its this, a certain amount of powder was scraped off from the CFP
composites, a metallic copper foil (2 cm ꢁ 2 cm) mechanically and pressed into a pellet. X-ray photoelectron spectroscopy
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 17572–17581 | 17573

View Article Online
RSC Advances
Paper
(XPS) measurements were carried using synchrotron radiation obtained in electro-reduction of CO2 were analysed by Thermo
facility hard X-ray photoelectron spectroscopy (HAXPES) Fischer Trace 1100 gas chromatograph with TCD and FID ana-
beamline BL-14, Indus-2 which has double-crystal mono- lysers for detection of hydrogen and other ERC products
chromator [Si (111)] with excitation energy of 4.311 keV and respectively aer passing through a Carboxene column at
ꢂ
equipped with Hemispherical analyser and detector system a ramping temperature of 50 C to 240 ꢂC.
(Phoibos 225, Specs make). The typical pressure in the experi-
Liquid products were analysed with Bruker Advance AV-III
mental station is 5 ꢁ 10ꢀ9 mbar. Survey spectra were recorded (400 MHz) NMR by diluting 0.5 mL catholyte with 0.1 mL D2O
over a range of 0 to 1000 eV for each composition. The energy (as NMR solvent) and 0.5 mL di-methylsulphoxide (DMSO)-as
calibration has been done with respect to the reference (C 1s internal standard. The solvent suppression method with
peak at a binding energy value of 284.8 eV adventitious carbon) a pre-saturation pulse was used to cater the overwhelming
for all edges.
solvent peak, which otherwise renders the ECR product peaks as
unobservable.
Electrochemical measurements
Calculation of faradaic efficiency
All the electrochemical measurements were carried out on
Metrohm Autolab PGSTAT302N with standard three electrode
system in a custom designed two compartment cell with
a Naon 117 (DuPont) membrane. The CFPs modied with Cu/
CuxO or Cu/CuxO-GO has been used as working electrode, Ag/
AgCl (saturated with KCl) as a reference electrode in cathode
chamber and Pt foil as counter electrode (separately) in an
anode chamber. All the potentials were referenced vis-a-vis
reversible hydrogen electrode using the formula ERHE, V ¼ E(Ag/
AgCl), V + 0.197 V + (0.059 ꢁ pH) V aer correcting for uncom-
pensated ohmic losses determined through impedance
measurements. 0.2 M KHCO3 saturated with either argon or
CO2 was used as an electrolyte medium for all the CO2 reduction
measurements, unless mentioned otherwise. 0.1 M HClO4 was
used to record the CVs (cyclic voltammograms) for electro-
chemical surface area (ECSA) measurements. In order to avoid
the effect of pH variation in CO2 purged solutions (generally, pH
decreases upon CO2 purging as solution becomes more acidic),
pH of the argon saturated electrolyte solutions was adjusted to
7.00 which was the pH of solutions saturated with CO2 (purging
for an hour with 20 mL minꢀ1 of ow rate).
The faradaic efficiency (F.E.) of the collected gaseous products
was calculated using eqn (1),
nxF Fxfm
F:E: ¼
(1)
I
where, nx is number of electrons involved in generation of
product under consideration (e.g. nx ¼ 2 for CO; 8 for CH4 and
12 for C2H4), F is Faraday's constant (96 485 C molꢀ1), Fx is the
volume fraction of the gaseous products (considering gases to
be ideal), fm is the molar ow rate (mol sꢀ1), and I is the cor-
responding current (in ampere, A). Sierra Instruments, USA
Make Smart-Track 100 Series Digital Mass Flow Controller
(C100L-DD-2-OV1-SV1-PV2-V1-S0) with Built-in 10 gas calibra-
tion system was used for controlling and measuring the ow
rate of CO2 purging. Faradaic efficiency of liquid products was
calculated by method mentioned elsewhere.22
For liquid products, the R factor was calculated as given
below
area under formic acid peak @ 8:4 ppm
R ¼
(2)
area under DMSO peak @ 2:5 ppm
R for different concentrations of formic acid was calculated, and
a calibration curve was plotted with R vs. concentration of formic
Preparation of electrodes
Working electrodes for ERC were fabricated in following ways; acid. The slope of the R vs. concentration of formic acid was
(i) electrodeposited thin lms (fCu/CuxO, fCu/CuxO-GO and further used for calculation of faradaic efficiency for formic acid
fCu/CuxO@GO) on CFP cathode were used as such without any production which is the primary reaction product in liquid form.
modications, (ii) for samples obtained as powder, 2.5 mg of
Formic acid conc. ¼ R/slope, Mꢀ1
each was dispersed in 3 : 1 iso-propanol and water mixture, with
5 mL Naon for 30 minutes and the dispersion was fed to an air
spray gun and coated as a lm on the CFP (2 cm ꢁ 1 cm) under
The total catholyte amounted to 50 mL in cathodic chamber;
therefore, the total concentration of formate produced was,
the ow of nitrogen as carrier gas. To understand whether direct
electro-deposition or mechanical fabrication of electrodes –
catalyst manually painted on the electrode is better, few more
electrodes were fabricated in the following manner. (iii) CFP
Nformic acid ¼ 0.05 L ꢁ formic acid concentration.
was modied with GO by pressure spray, and further air sprayed
with Cu/CuxO and was labelled as pCu/CuxO@GO. (v) Finally, to
Number of moles of electrons used to produce n moles of
understand the synergism between Cu/CuxO and GO, CFP
formic acid are
electrode was modied with only GO and was named as GO. All
eoutput ¼ Nformic acid ꢁ 2eꢀ
care was taken to mask the rest of the electrode surface so that
for all the fabricated electrodes identical area of CFP cathode is
exposed for equal geometric surface area, and also similar
sample loading was kept by calculating mass deposited in case
Total number of moles of electrons applied during the time
of lms by the amount of charge passed. The gaseous products of electrolysis was
17574 | RSC Adv., 2020, 10, 17572–17581
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
cathode, in the presence of GO, may dictate the interaction
within the GO framework and the deposited Cu/CuxO NPs,
though this needs further detailed investigation.
einput ¼ Q/F
where Q is the charge passed and F is the Faraday's constant.
The F.E. was calculated by,
From the FESEM images (Fig. 1), it is observable that the
presence of GO has limited the growth of Cu/CuxO composites
(pCu/CuxO-GO (Fig. 1(b))) and fCu/CuxO-GO (Fig. 1(d)) along the
(100) growth axis thus producing either cubes or small nger-
like spherical protrusions adhered to the surface of CFP. The
electrodeposited Cu/CuxO on the GO layer hand-painted on CFP
(fCu/CuxO@GO) shows the presence of Cu/CuxO spherical
particles on GO surface. The physical layer-over-layer (Cu/CuxO
over GO) i.e. pCu/CuxO@GO (Fig. 1(e)) shows structures like
pCu/CuxO on GO sheets. For the characterization of crystalline
features of the prepared catalysts, X-ray diffraction (XRD)
patterns (Fig. SI-2†) were recorded on the obtained materials
which suggest the formation of Cu and CuxO mixed phases in
all the samples. However, the collection substrate (cathode used
in electrodeposition) shows profound effect on the extent of
oxide content. Moreover, Raman spectroscopy data (Fig. SI-3†)
support the formation of a composite and deduce an idea about
the extent of reduction of GO into rGO (reduced graphene oxide
because of composite formation) (which also reects the extent
of defects in GO sheets manifested in ID/IG ratio) hence
a presentation of extent of metal–GO interactions.
Further, elemental mapping as well as EDX have been used
to demonstrate the distribution of the elements over the
observed structures. The colour maps and EDX shows (Fig. SI-
4†) the presence of copper, oxygen and carbon with a homoge-
nous distribution and free of any other impurities. The XPS
survey spectrum (Fig. 2(a)) of the materials shows peaks for
carbon, oxygen and copper only, indicating no other elements
of any nature are present. Table SI-2† lists the peak positions of
the copper, oxygen and carbon in different materials along with
the ratio of Cu2+ to Cu+. In all of the samples, Cu(II) state is
predominant, and its content is more in composites. From the
C 1s and O 1s detailed spectrum shown in Fig. SI-5 and SI-6,† it
can be concluded that oxygen functionalities of GO are used by
copper leaving more of sp2 hybridized carbon backbone in GO.
Another interesting observation in XPS is in the case of fCu/
CuxO@GO, which has an underlying GO layer. The charge
transfer between GO and copper has not taken place effectively
and hence has given rise to –C–O and –COOH peaks as observed
in copper oxides without GO. The peak position of Cu(I)
including Cu (0) at 932.6 ꢃ 0.1 eV in powdered Cu/CuxO-GO
composites can be seen, and peaks of Cu(II) at 933.8 ꢃ 0.2 eV
are also evident along with the associated satellite peaks. The
range of energies and the various binding energies are in good
agreement with literature reports.23–25
eoutput
F:E:formic
¼
ꢁ 100%
acid
einput
Results and discussion
The FESEM images of the samples (Fig. 1) shows self-assembly
of these catalysts depends upon the substrate used for electro-
deposition and whether or not GO was present in the electro-
lytic bath. While in absence of GO, the obtained Cu/CuxO
particles are self-assembled into either free cones or standing
dendrimers, the presence of GO restricts their growth to small
particles adhered to its surface. GO in solution because of its
electrical properties alters the solution conductivity which may
affect both mass transport as well as co-deposition of the copper
nanostructures. Effect of GO on synthesis of these catalyst is
elucidated in Fig. SI-1.† Fig. SI-1† presents the electrochemical
impedance spectroscopy (EIS) graphs obtained in reaction bath
with and without GO with Pt plate or CFP as working electrode
at open circuit potential (OCP). It is inferred from the graph that
there is a slight change in the conductivity of the solution with
Pt being more conductive as compared to CFP. The same
observation can be seen for presence of GO, the conductivity of
solution increases in presence of GO, owing to the electronic
properties of the GO. The decrease in solution resistance in GO
containing electrolyte is observed irrespective of the working
electrode used (Pt or CFP). This observation suggests that GO
not only plays a role in the growth of Cu/CuxO but also effects
the mass transport of the Cu cations towards the cathode, hence
guides the nano-morphology. Similarly, other inference from
the EIS data is that the cathode or collection substrate plays
a role in growth of the nanostructures. This observation indi-
cates that not only a carbonaceous support but also the elec-
trical properties of cathode play a crucial role in formation of
a catalyst. Lower resistance with Pt cathode as compared to CFP
To study the electro-catalytic activity of these samples
towards, ERC, cathodic linear sweep voltammetry (LSV) curves
were recorded for all materials both in presence as well as in the
absence of CO2. Fig. 3 depicts the overlay of the linear polari-
zation curves normalized with their respective electrochemical
surface area (ECSA) (the details of the ECSA measurements
which was done using capacitance measurements are given in
Fig. SI-7†). Solid lines in Fig. 3 represent the currents observed
in presence while dashed ones corresponds to LSVs recorded in
Fig. 1 FESEM images of different materials used as electrodes for ERC;
(a) pCu/CuxO, (b) pCu/CuxO-GO, (c) fCu/CuxO, (d) fCu/CuxO-GO, (e)
pCu/CuxO@GO (physical layer over layer composite), (f) fCu/Cux-
O@GO – film of Cu/CuxO electrodeposited on GO layer.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 17572–17581 | 17575

View Article Online
RSC Advances
Paper
Fig. 2 XPS spectra of the Cu/CuxO-GO composites (a) survey spectrum of the materials; and detailed spectra of Cu 2p, (b) pCu/CuxO, (c) pCu/
CuxO-GO, (d) fCu/CuxO, (e) fCu/CuxO-GO, and (f) fCu/CuxO@GO.
Fig. 3 LSV curves in 0.2 M KHCO3 electrolyte argon saturated (dashed lines) and CO2 saturated (solid lines); (a) pCu/CuxO (black traces) and pCu/
CuxO-GO (red traces), (b) thin films; fCu/CuxO (violet traces) and fCu/CuxO-GO (navy trace). The enlarged portions of the respective plots are
shown as an inset in the respective figures.
absence of CO2 (but in presence of argon). In presence of argon, respectively, it has been observed that pCu/CuxO-GO catalyst
the LSVs shows lower current than ECR along with higher onset leads to the hydrocarbons as ECR products whereas the current
potential before reasonable current is achieved. This current at on fCu/CuxO catalyst is due to mostly H2 evolution. Although,
higher potential obtained with argon-saturated 0.2 M KHCO3 the electrodes modied with fCu/CuxO-GO show the current
electrolyte can be attributed to the reduction of water which was densities of 25 mA cmꢀ2 in presence of CO2 the reaction prod-
also corroborated by gases collected from bulk electrolysis ucts obtained was majorly H2 gas. The electrodes fabricated by
where none of the CO2 reduction products other than hydrogen layering of catalyst samples are found to be less active for the
was found. When solution was saturated with CO2 and LSV CO2 reduction despite showing higher reduction currents per-
recorded there was a shi in the onset potential towards lower taining to more of HER (Fig. 4(d)). Therefore, based on these
negative potentials along with increased current density which observations, we found that pCu/CuxO-GO catalyst is the most
can be attributed to the electro-reduction of CO2.
active towards conversion of CO2 into hydrocarbons especially
From the measurements of currents at ꢀ0.985 V (RHE) we selectively into ethylene. Further detailed analysis of the elec-
can see that maximum reduction current is obtained with fCu/ trolysis products, viz. methane, CO, and H2 at different applied
CuxO (i.e. 45 mA cmꢀ2) followed by the pCu/CuxO-GO (i.e. 42 mA potentials on different samples is given in Fig. 4. It is pertinent
cmꢀ2) in the presence of CO2. However, from the electrolysis mention here that it is a well-known and reported fact that CO2
product analysis (Fig. 4) done using gas chromatography and saturation makes the aqueous electrolyte solutions more acidic.
NMR measurements on gaseous and liquid products, It is also well known that hydrogen evolution becomes more
17576 | RSC Adv., 2020, 10, 17572–17581
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Fig. 4 Production rates of ECR products on different catalysts in CO2-saturated 0.2 M KHCO3 under continuous CO2-purging for an hour. (a)
C2H4 at ꢀ0.985 V vs. RHE (error bars are calculated from standard deviation of three consecutive records), and (b) CH4, (c) CO, and (d) H2 at
different applied potentials.
prevalent in acidic solutions19,26,27 Hence, the production of H2 CuxO-GO (Fig. 3(a)) depicts a broad peak for the CO adsorption
gas is increased in presence of CO2 as compared to that in from 0.3 V to ꢀ0.6 V or in some cases up to ꢀ0.9 V; hence the
presence of Ar. However, the competitive production H2 can be presence of ethylene in the ERC product can be attributed to the
suppressed over the certain electro-catalyst surfaces that are *CO adsorption followed by dimerization. Importantly, similar
less active for hydrogen evolution but highly active for ECR. broader shoulders are observed for fCu/CuxO-GO but not for
Thus, the larger current density observed for the samples, fCu/ other materials like fCu/CuxO, fCu/CuxO@GO and pCu/Cux-
CuxO and fCu/CuxO-GO in presence of CO2 is attributed to the O@GO where the production of ethylene is either zero or very
lower local (more acidic) pH at the electro-catalyst surface. low. For pCu/CuxO-GO, the production of hydrogen is also
However, relatively lower hydrogen evolution and higher ERC decreased at higher negative potentials (Fig. 4(d)) to a mere F.E.
activity over pCu/CuxO.Go and pCu/CuxO is attributed to higher of 16% which can be attributed to the production of more
selective ERC catalytic activity of these electro-catalysts ethylene and increase in local pH. GO showed no appreciable
compared to hydrogen evolution.
It was observed that at very low overpotentials (i.e. ꢀ0.4 to production at ꢀ0.98 V (RHE) with lower reduction currents.
ꢀ0.7 V vs. RHE), CO and CH4 were detectable but around ꢀ0.8 V Fig. SI-8† shows the variation of total current density and
CO2 reduction at lower potentials and barely a F.E. of 20% CO
vs. RHE ethylene was also detected in composites but not at F.E. values with time for the ethylene production over pCu/
CuxO without GO. At ꢀ0.985 V vs. RHE ethylene was the main CuxO.GO electrode at an applied potential of ꢀ0.98 V vs. RHE.
product with a rate of 194.44 mmol gꢀ1 hꢀ1 with 34% F.E. on As can be noted from Fig. SI-8,† the total current density
pCu/CuxO-GO. Formic acid was the only liquid product detected decreases from initial values of ꢀ43 mA cmꢀ2 to ꢀ39 mA cmꢀ2
that too with a very low F.E. of ca. 4%. The production of aer 10 minutes but then remains constant for an hour long
ethylene is probably because of *CO coupling28 as can be duration. Similarly, the F.E. values for ethylene production
noticed from the lowered CO production in pCu/CuxO-GO remains almost constant over the studied duration of CO2
composites. It has been documented in the literature that electrolysis.
adsorbed CO shows a shoulder or mild peak before ECR and
The obtained trend in the activity of the prepared electro-
larger the CO adsorption broader the shoulder.22,29 Similar catalysts can be explained on the basis of the governing phys-
observations are noted in the present case as well, where pCu/ ico–chemical properties of these catalyst samples. For example,
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 17572–17581 | 17577

View Article Online
RSC Advances
Paper
the deposition substrate employed in an electrophoretic depo- similar enhancement in the ERC activity due to the presence of
sition is found to play an important role as it guides the self- GO supports in the catalysts. Moreover, we noticed that the
assembly of nanostructures and hence the morphology of the catalyst activities are affected by the extent of interaction
resultant electrodeposits. The comparison of the LSV polariza- between GO and copper oxide nanostructures which is man-
tion curves for pCu/CuxO, fCu/CuxO, pCu/CuxO-GO, and fCu/ ifested in variation of ID/IG ratio. pCu/CuxO-GO with highest ID/
CuxO-GO (Fig. 3) in CO2-saturated electrolyte solutions shows IG ratio (1.31) showed highest ethylene production followed by
that the catalysts obtained with Pt as a collection substrate fCu/CuxO-GO (1.01) and fCu/CuxO@GO (0.93). Again, the higher
depict better activity towards ERC than compared to CPF yield of hydrocarbon generation in pCu/CuxO-GO with higher
substrate. This could be due to the loose and porous nature of degree of defects can be explained on the basis of the higher
the materials obtained with Pt plate substrate (Fig. 1) which in stability to the intermediates or adsorbed CO giving enough
turn could be due to the higher hydrogen co-deposition on Pt- time for C–C coupling on its surface.
plate during electro-deposition of the samples compared to
Another substantial rationalization that explains the
CFP substrate. This has been well established in the litera- observed trend of activity is the presence of oxide lattices in the
ture30–32 that higher evolution of gases on the collection synthesized composites. Kanan et al.43 have reported that the
substrate leads to higher porosity of the obtained electrode- presence of Cu2O layers on Cu foil electrode with a higher
posits. It is also documented in the literature33–37 that the highly thickness of $3 mm results in lowering of overpotential by 0.5 V
porous catalysts provide the higher stability to the intermedi- than that of polycrystalline Cu. This lowering is mainly due to
ates or adsorbed CO giving enough time for C–C coupling on its enhancement in the roughness factor of the electrode surface.
surface, leading to an increased hydrocarbon production. Thus, While Lee et al.44 and Mistry et al.45 claimed residual Cu+ ions
the higher yield of hydrocarbons especially ethylene on pCu/ even aer ERC lead to higher ERC efficiencies by using in situ
CuxO-GO can be attributed to the loose and porous nature of XPS or X-ray Absorption near Edge Spectroscopy (XANES)
these electrodeposits that in turn is an aggregate function of the measurements. Other reports by using in situ XPS by Eilert
presence of GO and the nature of cathode material used to et al.46 suggested the presence of sub-surface oxide layers during
collect the electrodeposits. Besides, it can be noted from Fig. 1 ERC changes the electronic structure near Cu, thus increasing
that pCu/CuxO-GO have asymmetric and cone type shapes and the efficiency of ERC. The presence of oxides, along with
thus higher roughness factor (also observed from ECSA) which metallic structures, offer sites with multiple vacancies and
increases the total faradaic efficiency (72%) for CO2 conversion. increases the catalytic activity of the nanostructures towards
While doing the preliminary experiments, we noticed that ERC.47 However, whether such an enhancement is merely due to
the strategy employed for electrode fabrication has a profound the roughness factor or the oxygen content has to play any role
effect on the ERC. In general, several researchers, including us, is not clearly understood or well documented in the literature.
have preferred a layer by layer coating of electro-catalyst over the This seems to be very important, given the fact that the surfaces
support material.38,39 In contrast, few others have preferred of nanostructured materials play a critical role in their physico-
a mixed form of the electrocatalyst and a support material40 for chemical as well as electro-catalytic properties. Therefore, in
studying the electro-catalysis. In order to study the effect of order to have a better insight regarding the effect of oxide
electrode fabrication strategy, electrodes fabricated in three content in nano-dimensional copper based electro-catalysts, we
different ways viz. fCu/CuxO@GO, pCu/CuxO@GO and fCu/ tested the pCu/CuxO-GO nano-composites with varying O/Cu
CuxO-GO were investigated for their comparative activities. We ratios for their electro-catalytic activity towards ERC. Interest-
found that the electrode obtained with electrochemically co- ingly, we have noted that the inclusion of oxide form of Cu
deposited composites show better activity than the layered increases the efficiency of the catalyst in comparison to metallic
one. This comparison is also featured in their physical and Cu-rich catalyst for ERC. Table SI-2† shows Cu2+/Cu+ ratio ob-
electrical properties like XPS and XRD where co-deposited tained from XPS of the samples. It can be seen that samples
samples were found to yield more interconnected materials with GO show more of oxide content which could also assist in
which act synergistically to enhance the production of ethylene. increasing the efficiency of the composites.43
Furthermore, it was also noted that simple physical layering of
catalyst over GO support are less active than one where copper charge transfer resistance offered by pCu/CuxO-GO both at
has been electrochemically anchored to GO. onset as well as at OCP is minimum for ERC among the studied
Electrochemical impedance spectroscopy also showed that
GO-support has been reported to enhance the catalytic catalysts, other details for EIS are given in ESI (Fig. SI-9, SI-10a
activity and durability of metal and metal oxide NPs towards and b).† From the trends observed in current study, pCu/CuxO-
ERC41 as GO not only acts as a conducting scaffold for the GO was found to be the most active and selective catalyst for
catalyst but also increases the robustness of the material. It is CO2 conversion into ethylene with 34% F.E. at ꢀ0.985 V (RHE).
well established that during the formation of such hybrids, GO The observed catalytic efficiency appears to be a cumulative
undergoes reduction into reduced graphene oxide (rGO) effect of change of neighbouring electronic structure of copper
through charge transfer processes14,41 leading to the defect nanostructures on GO support,48 higher electrochemical active
formation in GO, manifested in the ratio of the intensity of D surface area, and higher oxide content which can lead to sub-
band (ID) to G band (IG), i.e. ID/IG ratio. These defects can be surface oxygen.45 The sub-surface oxygen can stabilize the
decorated with metal catalysts which causes a synergism in intermediates and thereby facilitating the *CO–CO* coupling. It
transport and conduction of the charges.42 We also found the was also noted that co-deposition of metal oxide and GO yields
17578 | RSC Adv., 2020, 10, 17572–17581
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Table 1 Comparative literature reports on different state of art copper based electrocatalysts for production of ethylene from CO2. Our finding
has also been included for comparison
S. no.
Electrocatalyst
Applied potential
F.E.C H (%)
Mass activity
Reference
2
4
1
2
3
4
5
6
Cu2O/NRGO
Cu2O/ILGS
Cl derived Cu
Cu on pyridinic-N rich graphene oxide
Cu/Cu2O
ꢀ1.40 V (RHE)
ꢀ1.30 V (RHE)
ꢀ1.06 V (RHE)
ꢀ0.90 V (RHE)
ꢀ1.10 V (RHE)
ꢀ0.98 V (RHE)
19.7
14.8
37.0
19.0
36.3
34.0
136.1 mmol gꢀ1 hꢀ1
—
—
—
49
50
51
52
—
53
Cu/CuxO-GO
194 mmol gꢀ1 hꢀ1
Present work
a better catalyst combination whether one uses CFP (fCu/Cu2O- oxide in the reduced electro-catalyst. Presence of oxides, along
GO) or Pt plate (pCu/Cu2O-GO) as a cathode during their elec- with metallic structures, offer sites with multiple vacancies and
trodeposition as compared to the electrodes obtained in a layer- increases the catalytic activity of the nanostructures 40. Similar
over-layer manner (both, physical layering of metal oxide results are noted in our study as well, where the oxide content in
nanostructures on GO (pCu/Cu2O@GO) and electrodeposited the copper oxide electro-catalysts as determined from the XRD
metal oxide nanostructures on physically adsorbed GO layer measurements carried out on these samples post-electrolysis
(fCu/Cu2O@GO)) on CFP. A layer-over-layer (pCu/Cu2O@GO) was found to decrease aer ERC (noted from the decrease in
electrode fabrication is least active for ERC than the one where relative peak intensities of copper oxide to metallic copper in
metal oxide nanostructures are electrodeposited on GO, among the XRD patterns).
the studied samples. This may be due to inefficient interaction
Overall, we can say that the enhanced activity of composites
of GO with the metal oxide nanostructures in case of layer-over- is also assisted by the sub-surface oxides that are remnant aer
layer electrodes which is explicitly conrmed from the Raman ERC as suggested by Lee et al.44 and Mistry et al.45 Also, GO
spectral characterization. Thus, the combination of the phys- provided a scaffold which increased the stability of the catalyst
ico–chemical properties of an electrodeposit viz. surface struc- as can be inferred from FESEM images of spent catalysts
ture (i.e. higher surface area), electronic interaction with the (Fig. SI-12†) where catalysts in composite with GO are not
support material (as conrmed from Raman spectroscopy), and morphologically changed. This could be because of potential of
larger oxide content (as conrmed from XPS, XRD, and EDX) GO to change electronic structure of the interacting metal oxide
was found to be the most appropriate combination that show nanostructures with it and acting as a scaffold and support for
the best activity among the studied materials.
the active metal catalysts.48 Table 1 summarizes the perfor-
Furthermore, as ERC was carried out under reducing mance comparison of the present catalyst with the results from
conditions, in order to study the effect of ERC on the crystal a few other recent copper/copper oxide based electro-catalysts
structure and the morphology of the catalyst, the electrode for ethylene production from electrochemical reduction of
catalyst materials were removed from the electrode surface and CO2. The comparison of the F.E. values and the mass activity
characterized using XRD and SEM imaging. The ratio of CuxO/ obtained at a respective applied potential suggests that the Cu/
Cu phase calculated from XRD overlays before and aer the ERC CuxO-based electro-catalyst presented in our study shows
(Fig. SI-11 is tabulated in Table SI-1†) which shows that the reasonably higher mass activity (194 mmol gꢀ1 hꢀ1) and
oxide content decrease aer post electrolysis which suggests comparable F.E. value of 34% at comparatively lower applied
that oxide phase is not highly stable under the high applied potential of ꢀ0.98 V vs. RHE.
negative potentials. We found that there is decrease in the ratio
In conclusion, we precisely, for the present study, succeeded in
of CuxO/Cu in all samples as summarized in Table SI-1.† It is craing a pCu/CuxO-GO composite through a simple, high
important to mention here that the literature related to the role yielding, and economic electro-deposition approach. The result-
of oxide layers/contents in Cu-based catalysts for ERC suggest ing catalysts exhibit a higher faradaic efficiency that too with
that the residual oxides are unstable in the oxide-derived Cu lower potentials for electrochemical reduction of CO2 to C2H4. In
catalysts during ERC and where most of the copper oxides are general, it has been found that the manifestation of several
reduced to Cu or Cu+ species however a small quantity of the parameters including oxide contents in an electrocatalyst, ID/IG
original oxide content remain in the sample even aer ERC. For ratio due to GO support and mode of fabrication of working
example, Lee et al.44 and Mistry et al.45 claimed that residual electrodes along with porous microstructures of the nano-
oxides are unstable in the oxide-derived Cu catalysts during ECR composite contributes crucially towards CO2 electro-reduction.
where the residual oxide content analyzed ex situ with secondary This study corroborates the synergism exhibited by copper oxide
ion-mass spectrometry (SIMS) was found to be only <1% of the and GO for enhanced ERC and promise as excellent catalysts for
original oxide content remained in the samples. Similarly, Eilert the same. These ndings can serve for preparation of promising
et al.46 on the basis of both ambient pressure X-ray photoelec- ERC catalysts by tuning specic physical and chemical properties.
tron spectrum (APXPS) and quasi in situ oxygen K-edge electron
energy-loss spectra (EELS) prove the absence of residual copper
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 17572–17581 | 17579

View Article Online
RSC Advances
Paper
19 S. A. Bhat, N. Rashid, M. A. Rather, S. A. Pandit, G. M. Rather,
P. P. Ingole and M. A. Bhat, ACS Appl. Mater. Interfaces, 2018,
10, 16376–16389.
Conflicts of interest
There are no conicts to declare.
20 A. N. Singh, H. Devnani, S. Jha and P. P. Ingole, Phys. Chem.
Chem. Phys., 2018, 20, 26719–26733.
21 K. Saini, H. Devnani, M. A. Bhat and P. P. Ingole, J.
Electrochem. Soc., 2017, 164, B674–B680.
22 C. S. Chen, A. D. Handoko, J. H. Wan, L. Ma, D. Ren and
B. S. Yeo, Catal. Sci. Technol., 2015, 5, 161–168.
23 G. Xie, M. Forslund and J. Pan, ACS Appl. Mater. Interfaces,
2014, 6, 7444–7455.
24 P. Singh, P. Nath, R. Kumar Arun, S. Mandal and N. Chanda,
RSC Adv., 2016, 6, 92729–92738.
Acknowledgements
NR acknowledges CSIR-UGC for the fellowship. PPI thanks DST-
SERB, Govt. of India for the nancial support vide SB/EMEQ-
339/2014 under the “Empowerment and Equity Opportunities
for Excellence in Science.” Authors are thankful to Central
Research Facility (CRF), and Nanoscale research facilities (NRF)
of IIT Delhi for assistance in material characterisation.
25 A. Gupta, R. Jamatia, R. A. Patil, Y.-R. Ma and A. Kumar Pal,
ACS Omega, 2018, 3, 7288–7299.
26 C. Chen, B. Zhang, J. Zhong and Z. Cheng, J. Mater. Chem. A,
2017, 5, 21955–21964.
References
1 W. Sheng, S. Kattel, S. Yao, B. Yan, Z. Liang, C. J. Hawxhurst,
Q. Wu and J. G. Chen, Energy Environ. Sci., 2017, 10, 1180– 27 K. K. Dey, S. Jha, A. Kumar, G. Gupta, A. K. Srivastava and
1185.
P. P. Ingole, Electrochim. Acta, 2019, 312, 89–99.
2 J. Wu, T. Shari, Y. Gao, T. Zhang and P. M. Ajayan, Adv. 28 A. H. Shah, Y. Wang, A. R. Woldu, L. Lin, M. Iqbal, D. Cahen
Mater., 2019, 31, 1804257. and T. He, J. Phys. Chem. C, 2018, 122, 18528–18536.
3 S. Chabi, K. M. Papadantonakis, N. S. Lewis and 29 Y. Hori, A. Murata, R. Takahashi and S. Suzuki, J. Chem. Soc.
M. S. Freund, Energy Environ. Sci., 2017, 10, 1320–1338. Chem. Commun., 1988, 17–19.
4 I. A. Rutkowska, E. Szaniawska, J. Taniewicz, A. Wadas, 30 K. I. Popov, S. S. Djokic, N. D. Nikolic, V. D. Jovic, K. I. Popov,
´
´
´
´
´
´
E. Seta, D. Kowalski and P. J. Kulesza, J. Electrochem. Soc.,
2019, 166, H3271–H3278.
5 B. Fang, Y. Xing, A. Bonakdarpour, S. Zhang and
S. S. Djokic, N. D. Nikolic and V. D. Jovic, in Morphology of
Electrochemically and Chemically Deposited Metals, Springer
International Publishing, 2016, pp. 171–203.
´
ˇ
D. P. Wilkinson, ACS Sustain. Chem. Eng., 2015, 3, 2381–2388. 31 N. D. NIKOLIC, ZASTITA Mater., 2010, 51, 197–2013.
´
´
´
6 Z. Xiong, H. Wang, N. Xu, H. Li, B. Fang, Y. Zhao, J. Zhang 32 N. D. Nikolic, K. I. Popov, L. J. Pavlovic and M. G. Pavlovic, J.
and C. Zheng, Int. J. Hydrogen Energy, 2015, 40, 10049–10062. Electroanal. Chem., 2006, 588, 88–98.
7 B. Fang, A. Bonakdarpour, K. Reilly, Y. Xing, F. Taghipour 33 J. J. Lv, M. Jouny, W. Luc, W. Zhu, J. J. Zhu and F. Jiao, Adv.
and D. P. Wilkinson, ACS Appl. Mater. Interfaces, 2014, 6,
15488–15498.
Mater., 2018, 30, 1803111.
34 W. Luo, W. Xie, M. Li, J. Zhang and A. Zuttel, J. Mater. Chem.
¨
8 Y. Hori, H. Konishi, T. Futamura, A. Murata, O. Koga,
A, 2019, 7, 4505–4515.
€
H. Sakurai and K. Oguma, Electrochim. Acta, 2005, 50, 35 W. Luo, J. Zhang, M. Li and A. Zu Ttel, ACS Catal., 2019, 9,
5354–5369.
3783–3791.
9 D. W. DeWulf, T. Jin and A. J> Bard, J. Electrochem. Soc., 1989, 36 K. Huang, J.-Y. Zhang, F. Liu and S. Dai, ACS Catal., 2018, 8,
136(6), 1686–1691. 9079–9102.
10 S. Liu, H. Tao, L. Zeng, Q. Liu, Z. Xu, Q. Liu and J. L. Luo, J. 37 Q. Zhu, X. Sun, D. Yang, J. Ma, X. Kang, L. Zheng, J. Zhang,
Am. Chem. Soc., 2017, 139, 2160–2163. Z. Wu and B. Han, Nat. Commun., 2019, 10, 3851.
11 N. Rashid, M. A. Bhat, P. P. Ingole and A. C. S. A. Materials, J. 38 H. R. Q. Jhong, F. R. Brushett and P. J. A. Kenis, Adv. Energy
CO2 Util., 2020, 38, 385–397.
Mater., 2013, 3, 589–599.
12 N. Rashid, M. A. Bhat, A. Das and P. P. ingole, J. CO2 Util., 39 N. Bhandary, A. P. Singh, P. P. Ingole and S. Basu,
2020, 39, DOI: 10.1016/j.jcou.2020.101178.
ChemPhotoChem, 2017, 1, 70–75.
13 E. Kecsenovity, B. Endrodi, P. S. Toth, Y. Zou, R. A. W. Dryfe, 40 S. Oh, J. R. Gallagher, J. T. Miller and Y. Surendranath, J. Am.
¨
´
´
K. Rajeshwar and C. Janaky, J. Am. Chem. Soc., 2017, 139,
Chem. Soc., 2016, 138, 1820–1823.
6682–6692.
41 R. Saito, M. Hofmann, G. Dresselhaus, A. Jorio and
M. S. Dresselhaus, Adv. Phys., 2011, 60, 413–550.
14 X. An, K. Li and J. Tang, ChemSusChem, 2014, 7, 1086–1093.
15 Y. Y. Birdja, R. E. Vos, T. A. Wezendonk, L. Jiang, F. Kapteijn, 42 S. Choi, C. Kim, J. M. Suh and H. W. Jang, Carbon Energy,
M. T. M. Koper, E. Vos, T. A. Wezendonk, L. Jiang, F. Kapteijn
and M. T. M Koper, ACS Catal., 2018, 8, 4420–4428.
16 M. Hu, Z. Yao and X. Wang, Graphene-Based Nanomaterials
for Catalysis, 2017.
17 S. A. Bhat, M. A. Rather, S. A. Pandit, P. P. Ingole and
M. A. Bhat, Analyst, 2015, 140, 5601–5608.
18 M. Liu, R. Zhang and W. Chen, Chem. Rev., 2014, 114, 5117–
5160.
2019, 1, 85–108.
43 C. W. Li and M. W. Kanan, J. Am. Chem. Soc., 2012, 134,
7231–7234.
44 S. Lee, D. Kim and J. Lee, Angew. Chem., Int. Ed., 2015, 127,
14914–14918.
45 H. Mistry, A. S. Varela, C. S. Bonifacio, I. Zegkinoglou,
I. Sinev, Y. W. Choi, K. Kisslinger, E. A. Stach, J. C. Yang,
P. Strasser and B. R. Cuenya, Nat. Commun., 2016, 7, 12945.
17580 | RSC Adv., 2020, 10, 17572–17581
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
46 A. Eilert, F. Cavalca, F. S. Roberts, J. Osterwalder, C. Liu, 49 H. Ning, Q. Mao, W. Wang, Z. Yang, X. Wang, Q. Zhao,
M. Favaro, E. J. Crumlin, H. Ogasawara, D. Friebel, Y. Song and M. Wu, J. Alloys Compd., 2019, 785, 7–12.
L. G. M. Pettersson and A. Nilsson, J. Phys. Chem. Lett., 50 H. Ning, W. Wang, Q. Mao, S. Zheng, Z. Yang, Q. Zhao and
2017, 8, 285–290. M. Wu, Acta Phys.-Chim. Sin., 2018, 34, 938–944.
47 L. Mandal, K. R. Yang, M. R. Motapothula, D. Ren, 51 T. Qin, Y. Qian, F. Zhang and B. L. Lin, Chinese Chem. Lett.,
P. Lobaccaro, A. Patra, M. Sherburne, V. S. Batista, 2019, 30, 314–318.
B. S. Yeo, J. W. Ager, J. Martin and T. Venkatesan, ACS 52 Q. Li, W. Zhu, J. Fu, H. Zhang, G. Wu and S. Sun, Nano
Appl. Mater. Interfaces, 2018, 10, 8574–8584. Energy, 2016, 26, 1–9.
48 M. N. Hossain, J. Wen and A. Chen, Sci. Rep., 2017, 7, 1–10. 53 N. Altaf, S. Liang, L. Huang and Q. Wang, J. Energy Chem.,
2020, 48, 169–180.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 17572–17581 | 17581