Communication

pubs.acs.org/JACS

Shape-Dependent Electrocatalytic Reduction of CO2 to CO on

Triangular Silver Nanoplates
Subiao Liu, Hongbiao Tao, Li Zeng, Qi Liu, Zhenghe Xu, Qingxia Liu,* and Jing-Li Luo*
Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta T6G 1H9, Canada
*
S Supporting Information

also been reported to enhance strongly the catalytic activity for

ABSTRACT: Electrochemical reduction of CO 2 CO2RR over metal-based catalysts, such as Au5−7 and Ag,8−13
(CO2RR) provides great potential for intermittent renew-
Downloaded via INDIAN INST OF TECH HYDERABAD on September 12, 2024 at 07:08:38 (UTC).

where FE and current density reach maximum values. Zhu et

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

able energy storage. This study demonstrates a predom- al.5 reported CO2RR on Au nanoparticles (NPs) and found that
inant shape-dependent electrocatalytic reduction of CO2 the 8 nm Au NPs show FE up to 90% at −0.67 V (vs reversible
to CO on triangular silver nanoplates (Tri-Ag-NPs) in 0.1 hydrogen electrode, RHE). Ag NP, an attractive alternative for
M KHCO3. Compared with similarly sized Ag nano- noble metal electrode, shows selectivity as high as that of Au at
particles (SS-Ag-NPs) and bulk Ag, Tri-Ag-NPs exhibited a lower cost. The size effects have also been observed by
an enhanced current density and significantly improved Cheonghee group8 and they noted that 5 nm Ag/C has the best
Faradaic efficiency (96.8%) and energy efficiency (61.7%), CO2RR performance and achieves a maximum FE of 79.2% at
together with a considerable durability (7 days). Addition- −0.75 V vs RHE. It has been demonstrated that the enhanced
ally, CO starts to be observed at an ultralow overpotential size-dependent FE and current density are related to the ratio
of 96 mV, further confirming the superiority of Tri-Ag- of edge-to-corner. Density functional theory (DFT) simulations
NPs as a catalyst for CO2RR toward CO formation. also suggest that edge sites are more preferred for CO evolution
Density functional theory calculations reveal that the than corner sites on sized-controlled noble metal NPs.5,8,9,14
significantly enhanced electrocatalytic activity and selec- Recently, Min Liu et al.15 also reported a field-induced reagent
tivity at lowered overpotential originate from the shape- concentration that enables highly efficient CO2RR resulting
controlled structure. This not only provides the optimum from local high electric fields. Besides the effects of size and
edge-to-corner ratio but also dominates at the facet of local electric filed, understanding the shape effects of metal NP
Ag(100) where it requires lower energy to initiate the rate- on CO2RR is also worth noting. There have been limited
determining step. This study demonstrates a promising studies on CO2RR over metal NPs regarding the influence of
approach to tune electrocatalytic activity and selectivity of particle shape, but further exploration is warranted because the
metal catalysts for CO2RR by creating optimal facet and presence of edge and corner sites varies as the shape changes.5,9
edge site through shape-control synthesis. Shape control has received extensive attention for Ag with
particular emphasis on triangular Ag nanoplate (Tri-Ag-NP)
because of their unique structure-related optical properties and
I ncreased utilization of fossil fuels has brought about record-
breaking levels of atmospheric carbon dioxide (CO2), which
causes climate change and environmental issues.1 To attenuate
potential applications.16
In this study, Tri-Ag-NPs were synthesized using a direct
chemical reduction method.17 The composite containing Tri-
our reliance on fossil fuels, sustainable and environmentally Ag-NPs and carbon black (CB) was fabricated on a glassy
friendly alternatives are desirable options. Electrochemical CO2 carbon electrode (GCE) for CO2RR. It was found that Tri-Ag-
reduction reaction (CO2RR) points to a promising direction NPs were particularly active for CO2RR toward CO formation
not only in decreasing CO2 accumulation but also in converting at an ultralow η. DFT calculations were employed to rationalize
intermittent renewable electricity into energy-dense fuels.2 the increased catalytic activity and selectivity of Tri-Ag-NPs
However, the more kinetically preferred H2 evolution reaction toward CO2RR.
(HER) always outcompetes the CO2RR and consequently, Tri-Ag-NPs were synthesized by chemically reducing an
decreases the selectivity of target product.3 The high over- aqueous solution of AgNO3 with NaBH4 in the presence of
potential (η) is also required for CO2RR to reach substantial H2O2 and trisodium citrate (Scheme 1).
reaction rates.4 Therefore, it is desirable to search for novel UV/vis spectra were recorded to investigate the trans-
catalysts capable of efficiently promoting CO2RR with high formation process of Tri-Ag-NPs by judging from the spectra
selectivity (Faraday efficiency, FE) and catalytic activity change, as shown in Figure 1a. Upon the rapid injection of
(current density, j) at low η. NaBH4, the color of the original transparent solution
Various metallic electrocatalysts have been experimentally immediately became pale yellow, suggesting the occurrence of
and computationally identified for CO2RR because the binding Ag reduction. Approximately 2 min after initiation, the pale
energy of intermediate (CO*) on the metal surface is relatively solution turned deep yellow, an indication of Ag NP formation,
weaker than that of H* derived from HER. This leads to the
selective evolution of CO rather than the competitive H2 on Received: November 23, 2016
these metal surfaces. The existence of optimal particle size has Published: February 2, 2017

© 2017 American Chemical Society 2160 DOI: 10.1021/jacs.6b12103

J. Am. Chem. Soc. 2017, 139, 2160−2163
Journal of the American Chemical Society Communication

Scheme 1. Synthesis Process and Digital Images of Tri-Ag- linear sweep voltammetry (LSV) results of Tri-Ag-NPs, SS-Ag-
NPs NPs and bulk Ag to distinguish the shape effects as well as CB
and GCE as references to differentiate their contributions for
CO2RR. The Tri-Ag-NPs exhibited approximately 2-fold higher
current densities (over 5.5 mA cm−2, normalized by the
geometrical surface area) relative to bulk Ag (3.0 mA cm−2),
and 1.0 mA cm−2 larger than SS-Ag-NPs. More importantly, a
much more positive onset potential was observed for the Tri-
Ag-NPs compared to the SS-Ag-NPs and to Bulk Ag. The
increment in current density, indicative of a promoted cathodic
kinetics for CO2RR, is not conclusive evidence since HER and
CO2RR are often interconnected. To verify the occurrence of
predominant CO2RR other than HER, and the enhanced
shape-dependent catalytic activity of Tri-Ag-NPs, potentiostatic
electrolysis was carried out at different applied potentials. The
achieved current densities are 30% less than the corresponding
values in LSV measurements under the same potential load.
This can be attributed to the combination of the presence of
extra current from the interface charging, and the formation of
Figure 1. (a) UV/vis spectra; (b) TEM image of Tri-Ag-NPs. a CO2 and/or CO partially depleted layer in the vicinity of the
catalyst due to the mass transport limitations.19,20 The outlet
as evidenced in the absorbance of the characteristic peak at gases were directly vented into a gas chromatography (GC)
∼400 nm,17 where it quickly increased to the maximum. instrument to analyze quantitatively the gas composition. To
Subsequently, the solution quickly changed to blue, where the confirm that the catalytic activity is not derived from bare GCE
corresponding intensity of the characteristic peak of Ag NPs at and CB, both samples were fabricated as the cathodes. GC
∼400 nm quickly decreased, indicating the gradual con- analyses indicated that H2 was the major product resulting from
sumption of Ag NPs during the transformation process. the competitive HER on GCE and CB (see Figure S7).
Another peak emerged at ∼600 nm and gradually red-shifted Tri-Ag-NPs, SS-Ag-NPs and bulk Ag were also performed
to the wavelength of 850 nm, indicative of the formation and under the same experimental conditions, and the potential-
growth of Tri-Ag-NPs, as confirmed by the TEM image in dependent FEs for CO formation are shown in Figure 2b. CO
Figure 1b. The Ag nanoplates obtained are enclosed by two (FE of 8.1%) started to generate at an onset potential of
(111) facets at both top and bottom surfaces, and three (100) −0.206 V for Tri-Ag-NPs, which was only 96 mV lower than
facets at the side faces which contain twin planes and stacking the theoretical equilibrium potential (−0.11 V) and remarkably
faults along the vertical direction, as demonstrated by previous lower than other reported Ag-based catalysts,8,21 whereas CO
studies.16,18 was not detectable under this ultralow η (96 mV) for SS-Ag-
To examine the CO2RR activity of Tri-Ag-NPs, the mixture NPs and bulk Ag. To achieve the equivalent FE for CO, SS-Ag-
containing catalysts and CB was quantitatively dropped onto a NPs (8.0%) and bulk Ag (8.6%) required a potential of −0.456
GCE (0.785 cm2). The polarization curves of all catalysts were V (η of 346 mV) and −0.656 V (η of 546 mV), respectively.
obtained by a sweeping potential between −0.056 and −1.156 The FEs of CO increased significantly for Tri-Ag-NPs, SS-Ag-
V vs RHE at a scan rate of 50 mV s−1. Figure 2a shows the NPs and bulk Ag when more negative potentials were applied.
Concurrently, Tri-Ag-NPs showed the best CO2RR perform-
ance and reached its maximum value of 96.8% with a much
lower η of 0.746 V as compared to SS-Ag-NPs (65.4%, 0.846 V)
and bulk Ag (57.2%, 0.946 V). Comparing the FEs of all
samples at a fixed potential of −0.856 V (Figure 2c), Tri-Ag-
NPs showed 3- and 1.5-fold higher values than SS-Ag-NPs
(64.9%) and bulk Ag (31.9%), respectively. This indicates that
the CO2RR is shape-dependent and the Tri-Ag-NPs are the
most active for CO formation as compared to other reported
Ag-based catalysts.8,9,21 Figure 2d shows the potential-depend-
ent CO partial current densities calculated based on the current
densities from potentiostatic measurements and the corre-
sponding CO FEs for all samples. It clearly reveals the exclusive
catalytic activity of Tri-Ag-NPs for the CO2RR toward CO
formation, where a 2.2-fold increase in the CO partial current
densities of Tri-Ag-NPs relative to SS-Ag-NPs further confirms
the shape-dependent effects.
For comparing the reaction kinetics for CO2RR and
subsequently confirming the increased catalytic activity of Tri-
Figure 2. (a) Cathodic LSV results; (b) FEs of CO at various applied Ag-NPs resulted from the shape control, Tafel plots for Tri-Ag-
potentials (inset shows the CO, CH4 and H2 overall FE for Tri-Ag- NPs, SS-Ag-NPs and bulk Ag are shown in Figure 3a.
NPs) and (c) CO FEs at fixed potential of −0.855 V; (d) CO partial The Tafel slope, an indication of kinetics for CO formation,
current density. was 153 mV dec−1 for Tri-Ag-NPs, closer to the value of 118
2161 DOI: 10.1021/jacs.6b12103
J. Am. Chem. Soc. 2017, 139, 2160−2163
Journal of the American Chemical Society Communication

degradation and the corresponding FEs of CO only fluctuated

slightly around 96% throughout the stability test. Moreover,
Tri-Ag-NPs show no morphological change after CO 2
reduction (Figure S9), further indicating the excellent stability
of Tri-Ag-NPs for CO2RR.
A better understanding on the origins credited for the
ultralow onset potential, and the high CO selectivity at lower η
for CO2RR over shape-controlled Tri-Ag-NPs relative to SS-
Ag-NPs, DFT calculations were further explored to investigate
the reactivity of different Ag features based on the computa-
tional hydrogen electrode model.5,9,14 Figure 4a shows the

Figure 3. (a) Tafel plot and (b) η as a function of various current

densities; (c) maximum energy efficiencies of Tri-Ag-NPs, SS-Ag-NPs
and bulk Ag; (d) long-term stability at a potential load of −0.856 V
and the corresponding FEs of CO and H2.

mV dec−1 expected for rate-determining step at the electrode22

as compared to a Tafel slope of 177 mV dec−1 for SS-Ag-NPs
and 197 mV dec−1 for bulk Ag. Both values were larger than the
one for Tri-Ag-NPs, indicating a poor kinetics for CO2RR.
Moreover, the formation of an adsorbed *COO− intermediate
(CO2 + e− → *COO−) on catalyst surface exclusively
determines the reaction rate for CO2RR because the anionic
radical is highly unstable. The negative reduction potential of
the CO2/*COO− redox couple, Eo = −1.49 vs RHE at pH 7, Figure 4. (a) Free energy diagrams for CO2RR to CO on different
not only creates a significant thermodynamic and kinetic facets and Ag55 cluster at −0.11 V; active adsorption site density on
bottleneck for the overall CO2RR but also promotes side (b) Tri-Ag-NPs and (c) SS-Ag-NPs as a function of particle size; (d)
reactions (e.g., HER) and consequently decreases the current proposed mechanism for CO2RR to CO on Tri-Ag-NPs.
efficiency for CO2RR. However, the onset potential shifted
anodically to −0.206 V on Tri-Ag-NPs (Figure 3b), a net Gibbs free energy (ΔG) diagram for CO2RR on different facets
decrease of 0.25 and 0.35 V in η compared to SS-Ag-NPs and and Ag55 cluster, where the total ΔG values required for the
bulk Ag, respectively. The exchange current density (io), a proposed four elementary reaction steps were simulated and
reflection of the free energy barrier required for CO2RR at the calculated (Figure 4d). The required η initially originates from
reversible potential as well as a measure of intrinsic rate of the formation of COOH* because an uphill energy barrier of
electron transfer between the electrolyte and the electrode, was the first proton-coupled electron-transfer step for all facets and
1.1 × 10−4 mA cm−2 for Tri-Ag-NPs, 1 order of magnitude the cluster was observed. Apparently, the required ΔG to form
higher than that for SS-Ag-NPs (3.8 × 10−5 mA cm−2) and bulk the COOH* on Ag(100) is significantly lower than that on
Ag (3.2 × 10−5 mA cm−2). It is also comparable to a recently Ag(111) and quite close to the one on Ag(110), suggesting a
reported value (1.0 × 10−5 mA cm−2) on 3 nm Ag/C for higher catalytic activity of Ag(100) for CO2RR. More
CO2RR to CO,8 further suggesting an improved shape- importantly, the facet of (100) is predominant in Tri-Ag-NPs
dependent catalytic activity. As a critical parameter in as compared to that in SS-Ag-NPs because Tri-Ag-NPs are only
benchmarking electrocatalyst for renewable energy storage, enclosed by the facets of (100) and (111). The free-energy step
the maximum energy conversion efficiencies of Tri-Ag-NPs, SS- subsequently becomes more thermodynamically facile for the
Ag-NPs and bulk Ag were evaluated (Figure 3c). The low η second proton-coupled electron transfer for the adsorbed CO*
together with the high FE of Tri-Ag-NPs contributed to an at all facets. Although the Ag55 cluster is slightly easier for the
energy efficiency of over 61.7%, much higher than SS-Ag-NPs formation of COOH* than all facets, it tends to overbind with
(42%) and bulk Ag (34%). This value is also comparable with CO* and consequently decreases the product evolution rate.
the most efficient existing platforms for the formation of Thus, both the dominance of (100) and the ease in the CO
CO,7,23 and consequently distinguishes Tri-Ag-NPs as a evolution deriving from the shape-controlled effect point to the
promising platform for CO2RR. decreased η and low onset potential for CO formation.
To address the stability of the developed catalyst, a major Additionally, studies have demonstrated that an increased edge-
concern for CO2RR, the long-term performance of Tri-Ag-NPs to-corner ratio yields enhanced catalytic activity and selectivity
was evaluated at a constant potential load of −0.856 V for 7 for CO2RR because the adsorption of CO2 and the subsequent
days. The outlet gases were analyzed each hour over a period of formation of intermediate COOH* are easier on edge sites
14 h by GC. The corresponding FEs of CO and H2 were also than corner sites.5,9,14 Thus, the edge-to-corner ratios of Tri-
determined (Figure 3d), and the inset exhibits the long-term Ag-NPs and SS-Ag-NPs as a function of the cluster diameter
stability test over 7 days. The current density maintained a were also analyzed (Figure 4b,c). It is found that Tri-Ag-NPs
steady value at around −1.25 mA cm−2 with negligible with a diameter range from 0 to 12 nm maintain a comparably
2162 DOI: 10.1021/jacs.6b12103
J. Am. Chem. Soc. 2017, 139, 2160−2163
Journal of the American Chemical Society Communication

higher edge-to-corner ratio. Although the ratio difference (12) Yoon, Y.; Hall, A. S.; Surendranath, Y. Angew. Chem. 2016, 128,
between Tri-Ag-NPs and SS-Ag-NPs narrows gradually at 15508−15512.
larger diameters, the density of catalytically active edge sites on (13) Lu, Q.; Rosen, J.; Zhou, Y.; Hutchings, G. S.; Kimmel, Y. C.;
Tri-Ag-NPs always exceeds that on SS-Ag-NPs. This ensures Chen, J. G.; Jiao, F. Nat. Commun. 2014, 5, No. 3242.
(14) Gao, D.; Zhou, H.; Wang, J.; Miao, S.; Yang, F.; Wang, G.;
sufficient active edge sites for CO2 reduction to CO and Wang, J.; Bao, X. J. Am. Chem. Soc. 2015, 137, 4288−4291.
consequently leads to superior selectivity of Tri-Ag-NPs. (15) Liu, M.; Pang, Y.; Zhang, B.; De Luna, P.; Voznyy, O.; Xu, J.;
In summary, we successfully synthesized Tri-Ag-NPs and Zheng, X.; Dinh, C. T.; Fan, F.; Cao, C.; de Arquer, F. P.; et al. Nature
investigated its shape-dependent electrocatalytic reduction of 2016, 537, 382−386.
CO2 to CO in 0.1 M KHCO3. Tri-Ag-NPs exhibit considerably (16) Xia, Y.; Xiong, Y.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int. Ed.
higher selectivity as compared to SS-Ag-NPs and bulk Ag. 2009, 48, 60−103.
Moreover, CO can be detected at an ultralow onset potential, (17) Zhang, Q.; Li, N.; Goebl, J.; Lu, Z.; Yin, Y. J. Am. Chem. Soc.
confirming the excellent catalytic activity of Tri-Ag-NPs toward 2011, 133, 18931−18939.
CO2RR. The durability test over 7 days further confirms the (18) Zeng, J.; Xia, X.; Rycenga, M.; Henneghan, P.; Li, Q.; Xia, Y.
excellent performance of Tri-Ag-NPs for efficient CO2RR. DFT Angew. Chem., Int. Ed. 2011, 50, 244−249.
(19) Kumar, B.; Asadi, M.; Pisasale, D.; Sinha-Ray, S.; Rosen, B. A.;
calculations indicate that the high selectivity of Tri-Ag-NPs at
Haasch, R.; Abiade, J.; Yarin, A. L.; Salehi-Khojin, A. Nat. Commun.
an ultralow overpotential is a consequence of both the 2013, 4, No. 2819.
optimum edge-to-corner ratio and the predominant Ag(100) (20) Asadi, M.; Kim, K.; Liu, C.; Addepalli, A. V.; Abbasi, P.; Yasaei,
facet in Tri-Ag-NPs. P.; Phillips, P.; Behranginia, A.; Cerrato, J. M.; Haasch, R.; et al. Science

■
*
ASSOCIATED CONTENT
S Supporting Information
2016, 353, 467−470.
(21) Hsieh, Y. C.; Senanayake, S. D.; Zhang, Y.; Xu, W.; Polyansky,
D. E. ACS Catal. 2015, 5, 5349−5356.
(22) Zhang, S.; Kang, P.; Ubnoske, S.; Brennaman, M. K.; Song, N.;
The Supporting Information is available free of charge on the House, R. L.; Glass, J. T.; Meyer, T. J. J. Am. Chem. Soc. 2014, 136,
ACS Publications website at DOI: 10.1021/jacs.6b12103. 7845−7848.
Experimental details (PDF) (23) Rosen, B. A.; Salehi-Khojin, A.; Thorson, M. R.; Zhu, W.;

■
Whipple, D. T.; Kenis, P. J.; Masel, R. I. Science 2011, 334, 643−644.

AUTHOR INFORMATION
Corresponding Authors
*[email protected]
*[email protected]
ORCID
Subiao Liu: 0000-0001-6075-3301
Zhenghe Xu: 0000-0001-8118-1920
Notes
The authors declare no competing financial interest.

■ ACKNOWLEDGMENTS
This work was supported by the Natural Sciences and
Engineering Research Council of Canada.

■ REFERENCES
(1) Buitenwerf, R.; Rose, L.; Higgins, S. I. Nat. Clim. Change 2015, 5,
364−368.
(2) Goeppert, A.; Czaun, M.; May, R. B.; Prakash, G. S.; Olah, G. A.;
Narayanan, S. J. Am. Chem. Soc. 2011, 133, 20164−20167.
(3) Qiao, J.; Liu, Y.; Hong, F.; Zhang, J. Chem. Soc. Rev. 2014, 43,
631−675.
(4) Costentin, C.; Robert, M.; Savéant, J.-M. Chem. Soc. Rev. 2013,
42, 2423−2436.
(5) Zhu, W.; Michalsky, R.; Metin, O.; Lv, H.; Guo, S.; Wright, C. J.;
Sun, X.; Peterson, A. A.; Sun, S. J. Am. Chem. Soc. 2013, 135, 16833−
16836.
(6) Koh, J. H.; Jeon, H. S.; Jee, M. S.; Nursanto, E. B.; Lee, H.;
Hwang, Y. J.; Min, B. K. J. Phys. Chem. C 2015, 119, 883−889.
(7) Chen, Y.; Li, C. W.; Kanan, M. W. J. Am. Chem. Soc. 2012, 134,
19969−19972.
(8) Kim, C.; Jeon, H. S.; Eom, T.; Jee, M. S.; Kim, H.; Friend, C. M.;
Min, B. K.; Hwang, Y. J. J. Am. Chem. Soc. 2015, 137, 13844−13850.
(9) Back, S.; Yeom, M. S.; Jung, Y. ACS Catal. 2015, 5, 5089−5096.
(10) Hatsukade, T.; Kuhl, K. P.; Cave, E. R.; Abram, D. N.; Jaramillo,
T. F. Phys. Chem. Chem. Phys. 2014, 16, 13814−13819.
(11) Salehi-Khojin, A.; Jhong, H.-R.; Rosen, M. B. A.; Zhu, W.; Ma,
S.; Kenis, P. J. A.; Masel, R. I. J. Phys. Chem. C 2013, 117, 1627−1632.

2163 DOI: 10.1021/jacs.6b12103

J. Am. Chem. Soc. 2017, 139, 2160−2163