Site-Selective Growth of AgPd Nanodendrite-Modified Au Nanoprisms: High Electrocatalytic Performance for CO2 Reduction

Article abstract of DOI:10.1021/acs.chemmater.7b01813

Environmental impacts of continued CO₂ production have led to an increased need for new methods of CO₂ removal and energy development. Nanomaterials are of special interest for these applications, because of their unique chemical and physical properties that allow for highly active surfaces. Here, we successfully synthesize AgPd nanodendrite-modified Au nanoprisms in various shapes (nanoprisms, hexagonal nanoplates, and octahedral nanoparticles) by selective metal deposition. This strategy involves coupling galvanic replacement between Ag layers in Au@Ag core-shell nanoprisms and H₂PdCl₄ with a coreduction process of silver and palladium ions. Synthesis of AgPd nanodendrite-tipped (4.14-11.47 wt % Pd) and -edged (25.25-31.01 wt % Pd) Au nanoparticles can be controlled simply by tuning the concentration of H₂PdCl₄. More importantly, these multicomponent AgPd nanodendrite-modified Au nanoparticles show exceptional electrocatalytic performance for CO₂ reduction. AgPd nanodendrite-edged Au nanoprisms show more favorable potentials (-0.18 V vs RHE) than previously reported nanocatalysts for the reduction of CO₂ to formate, and exhibit higher faradaic efficiencies (49%) than Au, Au@Ag, and AgPd nanodendrite-tipped Au nanoprisms in aqueous electrolytes. Moreover, AgPd nanodendrite-modified Au nanoprisms show much higher selectivity and faradaic efficiency for CO₂ reduction to CO (85-87%) than Au and Au@Ag nanoprisms (43-64%) in organic electrolytes. The high performance of these particles for CO₂ reduction is attributed to the unique structure of AgPd nanodendrite-modified Au nanoprisms and the synergistic effect of Ag having an affinity for CO₂, efficient binding of hydrogen at Pd, and Au as a stable, conductive support. In addition, AgPd nanodendrite-edged Au nanoprisms show highly stable catalytic activity during long-term electrolyses (up to 12 h) and repetitive use. These exciting results indicate that AgPd nanodendrite-modified Au nanoparticles are promising for application in CO₂ conversion into useful fuels.

Full text of DOI:10.1021/acs.chemmater.7b01813

Subscriber access provided by NEW YORK UNIV
Article
Site-Selective Growth of AgPd Nanodendrite-Modified Au
Nanoprisms: High Electrocatalytic Performance for CO2 Reduction
Changsheng Shan, Erin T. Martin, Dennis G Peters, and Jeffrey M. Zaleski
Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.7b01813 • Publication Date (Web): 26 Jun 2017
Downloaded from http://pubs.acs.org on June 29, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8
8x29mm (300 x 300 DPI)
ACS Paragon Plus Environment

Chemistry of Materials
Page 2 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Site-Selective Growth of AgPd Nanodendrite-Modified Au
Nanoprisms: High Electrocatalytic Performance for CO Reduction
2
‡
‡
Changsheng Shan , Erin T. Martin , Dennis G. Peters*, and Jeffrey M. Zaleski*
Indiana University, Department of Chemistry, Bloomington, IN 47405
‡These authors contributed equally to the work.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
*
Email: zaleski@indiana.edu; peters@indiana.edu.
Abstract
Environmental impacts of continued CO
for new methods of CO removal and energy development. Nanomaterials are of
2
production have led to an increased need
2
special interest for these applications, due to unique chemical and physical properties
that allow for highly active surfaces. Here, we successfully synthesize AgPd
nanodendrite-modified Au nanoprisms in various shapes (nanoprisms, hexagonal
nanoplates, and octahedral nanoparticles) by selective metal deposition. This strategy
involves coupling galvanic replacement between Ag layers in Au@Ag core-shell
nanoprisms and H
2
PdCl with a co-reduction process of silver and palladium ions.
4
Synthesis of AgPd nanodendrite-tipped (4.14–11.47 wt% Pd) and -edged (25.25–31.01
wt% Pd) Au nanoparticles can be controlled simply by a tuning of the concentration of
H
2
PdCl
nanoparticles show exceptional electrocatalytic performance for CO
nanodendrite-edged Au nanoprisms show more favorable potentials (–0.18 V vs. RHE)
4
. More importantly, these multi-component AgPd nanodendrite-modified Au
2
reduction. AgPd
than previously reported nanocatalysts for the reduction of CO
2
to formate, and exhibit
1
ACS Paragon Plus Environment

Page 3 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
higher faradaic efficiencies (49%) than Au, Au@Ag, and AgPd nanodendrite-tipped
Au nanoprisms in aqueous electrolytes. Moreover, AgPd nanodendrite-modified Au
nanoprisms show much higher selectivity and faradaic efficiency for CO
CO (85–87%) than Au and Au@Ag nanoprisms (43–64%) in organic electrolytes. The
high performance of these particles for CO reduction is attributed to the unique
structure of AgPd nanodendrite-modified Au nanoprisms and the synergistic effect of
Ag having an affinity for CO , efficient binding of hydrogen at Pd, and Au as a stable,
2
reduction to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
2
conductive support. In addition, AgPd nanodendrite-edged Au nanoprisms show highly
stable catalytic activity during long-term electrolyses (up to 12 h) and repetitive use.
These exciting results indicate that AgPd nanodendrite-modified Au nanoparticles are
promising for application in CO conversion into useful fuels.
2
Introduction
Continued consumption of fossil fuels worldwide has led to two vital concerns:
development of new energy resources and mitigation of the environmental impacts of
CO
capture and conversion of CO
harnessing of CO
and for its conversion to usable fuels, including chemical reactions,7
2
production.1-6 Advances in sustainable energy and pollution control include the
into reusable forms. A variety of methods exists for the
2
2
photocatalysis,8,9 and electrocatalysis.10-13 Of these, electrocatalysis exhibits several
distinct advantages; the ability to control reaction products through choice of solvent,
electrode material, and applied potential, minimization of waste and byproducts, and
2
ACS Paragon Plus Environment

Chemistry of Materials
Page 4 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
generally mild reaction conditions make electrochemical treatment of CO
option.10,14
2
an attractive
Despite the appeal of electrochemical methods, achieving effective reduction of CO
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
comes with its own challenges. Electrochemical CO reduction often suffers from high
2
overpotentials, requiring large amounts of energy, as well as poor efficiency in the
presence of competing reactions such as hydrogen evolution and solvent–electrolyte
breakdown. To control reaction pathways and to decrease the overpotential of CO
2
reduction, unique electrocatalytic materials have been designed, often on the nanoscale
in order to maximize active sites.15-19 Metal nanocatalysts have been extensively
studied, including Au nanoparticles,15,20 Ag nanoparticles,21 nanoporous Ag,22 Pd
2
3
24
nanoparticles, and Au/Cu nanoparticles, but concerns of overpotential, selectivity,
and catalyst stability still exist.
When new electrocatalysts for CO reduction are designed, choice of material
2
and structure is paramount. Recent advances in nanomaterial synthesis have led to the
development of heterogeneous metal nanoparticles with exciting catalytic properties,
whose large areas, sharp tips, and metal coupling allow for wide application in catalysis
and sensing.25-28 Many heterogeneous metal nanoparticles are reported as core–shell
structures, which block the functionality of the inner metal core.29-32 Other
heterogeneous metal nanoparticles have special spatial arrangements, such as
nanoframes, nanodendrites, or site-selective deposition of metals on the metal cores,
which show high catalytic properties.27,33-40 In addition, metal nanoparticles with
nanodendritic structures have shown high catalytic properties due to larger surface areas
3
ACS Paragon Plus Environment

Page 5 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
and more active sites.37,41 Therefore, heterogeneous nanomaterials with nanodendritic
structures are particularly promising for catalysis due to a high number of active sites
and access to multiple metals at the surface. To our knowledge, there are only a few
reports for the successful synthesis of symmetrical, multi-branched heterogeneous
structures due to varying reduction rates of metal ions, galvanic replacement, and the
thermodynamic miscibility of metal phases.27,28 Therefore, controlled growth of multi-
branched heterogeneous structures still presents a challenge.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Here, we develop a facile method for site-selective growth of AgPd
nanodendrite-modified Au nanoarchitectures of different geometries for
electrochemical CO
replacement in the presence of H
silver and palladium. We control site-selectivity of this co-deposition by varying Ag
and H PdCl content, resulting in Au nanoparticles edged or tipped with AgPd
nanodendrite structures. The activity of these nanomaterials toward CO reduction is
2
reduction. Au@Ag core-shell nanoparticles undergo galvanic
2
PdCl and subsequent co-reduction and deposition of
4
2
4
2
explored by cyclic voltammetry in both aqueous and organic solvents, and remarkably
favorable reduction potentials at nanodendrite-edged and -tipped nanoprisms are
observed. The efficiency of these nanoprisms (Au, Au@Ag, AgPd nanodendrite-
modified Au) for bulk CO reduction is analyzed through controlled-potential
2
electrolyses, and reduction products are identified by means of gas chromatography and
NMR measurements. Faradaic efficiencies for the production of formic acid, CO, and
oxalic acid are reported, and mechanistic schemes are proposed for the reduction of
CO
2
at these novel nanodendrite-modified catalysts.
4
ACS Paragon Plus Environment

Chemistry of Materials
Page 6 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Experimental
Reagents.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Hydrogen tetrachloroaurate trihydrate (HAuCl
99%), palladium chloride (PdCl , 99%), sodium citrate (99%), L-ascorbic acid (AA，
9%), sodium borohydride (NaBH , 98%), sodium iodide (99.5%), sodium hydroxide
NaOH, 98%), and hexadecyltrimethylammonium bromide (CTAB, ≥99%) were all
4
•3H
2
O, 99.9%), silver nitrate (AgNO ,
3
2
9
4
(
purchased from Sigma–Aldrich and used as received. Hexadecyltrimethylammonium
chloride (CTAC, >95%) was purchased from Tokyo Chemical Industry Co., Ltd. (TCI)
and used without further purification. Millipore ultra-filtered water (18 MΩ•cm) was
used for all aqueous preparations. For electrochemical measurements, each of the
following chemicals (with purity given in parentheses) was purchased from Sigma–
Aldrich, or from a different commercial source when indicated, and was used as
received: carbon dioxide (Airgas, bone-dry grade), carbon monoxide (Airgas, 2000
ppm in N
monohydrate (98%), glycolic acid (98%), n-hexadecane (99%), deuterium oxide (D
99.9 atom D), N,O-bis(trimethylsilyl)trifluoroacetamide with 1%
2
), formic acid (J. T. Baker, 90%), oxalic acid (98%), glyoxylic acid
2
O,
%
trimethylchlorosilane (BSTFA:TCMS, Supelco, 99.5%), and 3-(trimethylsilyl)-1-
propanesulfonic acid sodium salt (97%). Purified water (Omnisolv, HPLC grade) and
acetonitrile (ACN, Omnisolv, 99.9%) were used as solvents for electrochemical
experiments.
Tetramethylammonium tetrafluoroborate (TMABF , TCI, 98%),
4
employed as supporting electrolyte in organic solvents, was recrystallized from a
5
ACS Paragon Plus Environment

Page 7 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
mixture of ethanol and water, then subsequently stored in a vacuum oven at 80 C for
48 h to remove traces of water. Lithium perchlorate (GFS Chemicals, reagent grade)
was employed as supporting electrolyte for aqueous solvents and used as received.
Deaeration of all solutions was accomplished with the aid of UHP-grade argon (Airgas).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Characterization of materials.
Transmission electron microscopy (TEM) was performed with a JEOL 1010
microscope operating at 80 kV. High-resolution transmission electron microscopy
(HRTEM), scanning transmission electron microscopy (STEM), and energy-dispersive
X-ray spectroscopy (EDX) data were collected on a JEOL JEM 3200FS instrument at
300 kV accelerating voltage. Scanning electron microscopy (SEM) and EDX data were
obtained by means of the SEM mode of a focused ion-beam instrument (Zeiss, Auriga
6
0) with EDX spectroscopy accessories. UV-Vis-NIR measurements were collected
with a Varian Cary 5000 spectrometer.
Synthesis of Au, Au@Ag, and Au/AgPd nanoprisms
Synthesis of Au nanoprisms. Gold (Au) nanoprisms were synthesized by the
three-step seed-mediated method reported previously.42,43 First, Au nanoparticle seeds
were prepared by reduction of HAuCl
4
with NaBH
4
. An aqueous solution of NaBH (1
4
mL, 100 mM, ice-cold water) was quickly added into a mixing solution containing 37
mL of deionized water, 1 mL of 10 mM aqueous sodium citrate, and 0.33 mL of 30 mM
HAuCl4, while being stirred vigorously. After 1 min, the resulting seed solution was
6
ACS Paragon Plus Environment

Chemistry of Materials
Page 8 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
aged for 2–3 h to allow hydrolysis of unreacted NaBH
growth solution containing 108 mL of 0.05 M CTAB, 22 µL of 0.1 M NaI, 3 mL of 10
mM HAuCl , 0.6 mL of 0.1 M NaOH and 0.6 mL of 0.1 M AA was divided into three
4
. Following the aging period, a
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
aliquots for the seed-mediated growth step. The first two solutions (solution 1 and
solution 2) contained 9 mL of the above growth solution, whereas solution 3 contained
the remaining 94 mL of growth solution. Nanoprism formation was initiated by fast
addition of 1 mL of seed solution to growth solution 1, and the mixture was shaken for
5 s. One mL of resulting solution 1 was then immediately added to solution 2 and
shaken for 5 s. The total contents of solution 2 were immediately added to solution 3
and shaken for 5 s. After this addition, the final solution changed from colorless to
deep magenta and was kept undisturbed overnight. Nanoprisms precipitated on the
bottom of the flask were collected by removal of the supernatant. Collected nanoprisms
were redispersed in 20 mL of aqueous 0.05 M CTAB solution.
Synthesis of Au@Ag core-shell nanoprisms. Au@Ag core-shell nanoprisms
were synthesized by means of a modified method from the literature.30 CTAB (9 mL,
0
.05 M), AA (0.4 mL, 0.1 M), and NaOH (0.6 mL, 0.1 M) were added to 3 mL of a
previously prepared suspension of Au nanoprisms. Aqueous AgNO solution (3.5 mL,
mM) was added with stirring to the above mixture at a rate of 160 µL/min with a
mechanical syringe pump (KD Scientific 200). After addition of AgNO , the resulting
3
1
3
Au@Ag core-shell nanoprisms were centrifuged at 7000 rpm for 10 min and then re-
dispersed in 3 mL of 0.05 M CTAB. These Au@Ag nanoprisms were grown further
by a single repetition of the above method, to yield a moderate coating of Ag, or 4
7
ACS Paragon Plus Environment

Page 9 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
repetitions, to produce nanoprisms with a higher Ag coating. A third set of Au@Ag
nanoprisms was created with a low Ag coating by use of only 0.6 mL of 1 mM AgNO
in a single step.
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of AgPd nanodendrite-tipped and -edged Au nanoprisms. AgPd
nanodendrite-tipped and -edged Au nanoprisms were prepared by addition of AA
solution and H
nanoprisms.34 Four different solutions were prepared containing 5, 10, 25, or 100 µL
of 2 mM AA. The volume of 2 mM H PdCl solution used was twice that of the AA
2
PdCl
4
solution to 1 mL of the previously prepared Au@Ag
2
4
solution in each case. Resulting solutions were left undisturbed at room temperature
for 3 h, then centrifuged at 7000 rpm for 10 min, washed with water once, and
redispersed in 100 µL of water. The nanoprism coating morphology depends on the
ratio of substrates used; the above procedure yields both tipped and edged nanoprisms.
Synthesis of Au, Au@Ag, Au/AgPd hexagonal nanoplates
Synthesis of Au hexagonal nanoplates.
Synthesis of Au and Au@Ag
hexagonal nanoplates was accomplished according to the modified method.44 Au
nanoprisms, synthesized by the previously described method, provided seeds to grow
Au hexagonal nanoplates. AA (22.5 mL, 4.73 mM) was added with stirring to a mixture
containing 65 mL of 0.05 mM CTAB, 37.5 µL of 0.1 M NaI, 2 mL of 30 mM
HAuCl
4
·3H O, and 10 mL of the Au nanoprism suspension at a rate of 0.5 mL/min by
2
means of a mechanical syringe pump (KD Scientific 200). After 3 h, the synthesized
8
ACS Paragon Plus Environment

Chemistry of Materials
Page 10 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
hexagonal nanoplates were collected by centrifugation at 3000 rpm for 10 min and
redispersed in 100 mL of 0.05 M CTAB.
Synthesis of Au@Ag hexagonal nanoplates. A growth solution containing 6
mL of 0.05 mM CTAB, 400 µL of 0.1 M AA, 600 µL of 0.1 M NaOH, and 3.2 mL of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1 mM AgNO was added to 6 mL of the previously prepared hexagonal nanoplate
3
suspension and vigorously stirred for 30 min. Then the resulting Au@Ag core-shell
nanoplate suspension was centrifuged at 3000 rpm for 10 min, and the product
redispersed in 6 mL of 0.05 M CTAB. Au@Ag nanoplates were grown further by one-
time repetition of the above growth method. Final Au@Ag core-shell hexagonal
nanoplates were dispersed in 12 mL of water.
Synthesis of AgPd nanodendrite-tipped and -edged Au hexagonal
nanoplates. AgPd nanodendrite-tipped and -edged Au hexagonal nanoplates were
synthesized by the same procedure as AgPd-modified Au nanoprisms. AA (25 µL, 2
mM) and H
2
PdCl (50 µL, 2 mM) were added sequentially to 1 mL of previously
4
prepared Au@Ag core-shell hexagonal nanoplate suspension to obtain AgPd
nanodendrite-tipped Au hexagonal nanoplates. Resulting mixtures were left
undisturbed at room temperature for 3 h, then centrifuged at 3000 rpm for 10 min,
washed with water once, and redispersed in 100 µL of water. Au@AgPd-edged
hexagonal nanoplates were prepared by the same method, except that 100 µL of 2 mM
AA and 200 µL of 2 mM H
2
PdCl were used.
4
9
ACS Paragon Plus Environment

Page 11 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
Synthesis of Au, Au@Ag, AgPd nanodendrite-tipped and -edged octahedral
nanoparticles
Synthesis of Au octahedral nanoparticles.
nanoparticles were synthesized according to Huang’s method.45,46 A gold seed solution
was prepared by fast addition of 0.45 mL of ice-cold 0.02 M NaBH to a mixture of 10
mL of deionized water and 80 µL of 30 mM HAuCl . After being stirred for 1 min, the
seed solution was aged for 1 h at 30 C in a water bath to decompose unreacted NaBH
Two growth solutions (1 and 2) were then prepared. First, CTAC surfactant (0.32 g)
was dissolved in 9.6 mL water. Subsequently, 80 µL of 30 mM HAuCl , 5 µL of 10
Gold (Au) octahedral
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
4
̊
4
.
4
mM KI, and 220 µL of 40 mM AA were added. Next, Au seed solution (25 µL) was
added to solution 1 and shaken for 5 s. The resulting solution 1 (25 µL) was then added
quickly to solution 2 and shaken for 10 s. Solution 2 was left undisturbed for 15 min
to allow for particle growth, then centrifuged at 7000 rpm for 10 min. The final
octahedral nanoparticles were redispersed in 10 mL of aqueous 0.05 M CTAB solution.
Synthesis of Au@Ag octahedral nanoparticles. Synthesis of Au@Ag
octahedral nanoparticles was similar to that of Au@Ag nanoprisms, except that a
suspension of Au octahedral seeds was used instead of Au nanoprisms. With stirring,
AgNO (3.5 mL, 1 mM) was added to a mixture of 3 mL of Au octahedral seed
3
suspension, 9 mL of 0.05 M CTAB, 0.4 mL of 0.1 M AA, and 0.6 mL of 0.1 M NaOH
at a rate of 160 µL/min by means of a mechanical syringe pump. After addition of
AgNO
3
, the resulting Au@Ag core-shell octahedral nanoparticles were centrifuged at
7
000 rpm for 10 min and then redispersed in 6 mL of water.
10
ACS Paragon Plus Environment

Chemistry of Materials
Page 12 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Synthesis of AgPd nanodendrite-tipped and -edged Au octahedral
nanoparticles. We prepared AgPd nanodendrite-tipped and -edged Au octahedral
nanoparticles by the same procedure as the AgPd nanodendrite-tipped and -edged Au
hexagonal nanoplates, except for substitution of Au hexagonal nanoplate seeds by Au
octahedral nanoparticles. Resulting mixtures were left undisturbed at room temperature
for 3 h, then centrifuged at 7000 rpm for 10 min, washed with water one time, and
redispersed in 100 µL water.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Electrochemical cells, electrodes, instrumentation, and procedures
Cyclic voltammetry. For cyclic voltammetry experiments, a Princeton
Applied Research Corporation (PARC) model 2273 or 273A potentiostat was used with
®
PowerSuite software. Cells and procedures used have been described in previous
publications.47,48 To prepare nanomaterial electrode surfaces, 10 µL of a nanomaterial
slurry was drop-cast onto a glassy carbon disk (Grade GC-20, 3.0-mm-diameter, Tokai
Electrode Manufacturing Company, Tokyo, Japan), which was press-fitted into a
2
Teflon rod to yield a geometric area of 0.071 cm . These electrodes were immersed in
deaerated aqueous solvent–electrolyte with a platinum coil counter electrode and a
saturated calomel reference electrode, and cleaned by means of repeated cycling of the
–1
potential between 0 and –1 V at a scan rate of 20 mV s . In organic solvent, the
reference electrode consisted of a cadmium-saturated mercury amalgam in contact with
DMF saturated with both sodium chloride and cadmium chloride; this electrode has a
potential of –0.76 V versus the aqueous saturated calomel electrode (SCE) at 25 °C.49-
11
ACS Paragon Plus Environment

Page 13 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
5
1
Geometric current densities were determined on the basis of the known surface area
of the glassy carbon support.
Controlled-potential electrolyses. Bulk electrolyses were performed with the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
aid of a PARC model 173 potentiostat. A locally written LabView program was used
for data collection, and the acquired data were processed with OriginPro 2015 software.
5
2
A two-compartment (divided) cell, described elsewhere, was modified to hold a wax-
impregnated graphite electrode (WIGE) prepared in-house. This working electrode,
2
with a geometric area of 0.28 cm , was modified by means of drop-casting 40 µL of a
nanomaterial slurry onto the working surface. Either a saturated calomel or cadmium
amalgam reference electrode was also present in the cathode compartment of the cell,
which was separated from the anode compartment by a sintered-glass disk and agar or
methyl cellulose plug. A carbon rod served as an auxiliary anode. Due to the volatility
of products arising from the reduction of carbon dioxide, sparging of argon through the
cathode compartment to remove oxygen was terminated as soon as the background
current reached a baseline level at the chosen potential, whereupon the cell was
saturated with carbon dioxide and sealed. Electrolyses performed in water lasted 2–6
h, whereas electrolyses in organic solvent typically lasted 45 min. Additional
experiments were performed at longer times under a constant flow of CO to analyze
2
catalyst stability. In all cases, geometric current densities with respect to the area of
the graphite support were reported as average current passed over the time of the
electrolysis. Only current passed after introduction of CO
2
to the cell was included in
these calculations, eliminating the contribution of background processes that occur
12
ACS Paragon Plus Environment

Chemistry of Materials
Page 14 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
during the pre-electrolysis step. Additionally, SEM images of the tip of a modified
electrode, which was cut with a razor from the graphite rod support, were obtained
before and after electrolysis. TEM images of nanomaterial catalyst removed from the
graphite support by sonication (10 s) in water were also obtained for comparison.
Electrolysis product analysis. Identification and quantitation of formic acid
was achieved by means of nuclear magnetic resonance (NMR) experiments. After an
electrolysis, the catholyte was diluted with deuterium oxide and 3-(trimethylsilyl)-1-
propanesulfonic acid sodium salt was added as an internal standard. Samples were
analyzed with a Varian 600-MHz spectrometer equipped with a cryogenic probe, and
data were processed with MestReNova software. Integration of formic acid and internal
standard peaks allowed for quantitative results as outlined in the literature.53
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Other products were detected by means of gas chromatography. For
quantitation of carbon monoxide, an Agilent Technologies 6890N gas chromatograph
fitted with an HP-PLOT Molesieve/5A column and a thermal conductivity detector
(TCD) were used with helium as a carrier gas. A gas-tight syringe was used to sample
the headspace of the electrochemical cell, and 0.50 mL of gas was injected directly onto
the column. Quantitation of carbon monoxide was achieved by means of a calibration
curve. Oxalic, glyoxylic, and glycolic acid were derivatized before analysis. Catholyte
was concentrated and treated with an equal volume of BSTFA:TCMS, then heated at
60 °C for one hour to prepare the respective trimethylsilyl derivatives. Detection of
these derivatives was accomplished with the aid of an Agilent 6890N gas
chromatograph fitted with a 30 m × 0.32 mm capillary column (J & W Scientific) that
13
ACS Paragon Plus Environment

Page 15 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
had a DB-5 stationary phase consisting of 5% phenylpolysiloxane and 95%
methylpolysiloxane and a flame ionization detector (FID). Internal standard (n-
hexadecane) was added for quantitation purposes; to determine gas chromatographic
response factors, a mock cell containing anticipated concentrations of products and
internal standard was allowed to stir (without application of a potential) for the same
length of time as an actual bulk electrolysis. Response factors and chromatographic
peak areas were measured with the aid of Agilent Chemstation software, and used to
determine product distributions according to a previously developed procedure.54
In all cases, faradaic efficiencies were calculated by comparison of total charge
passed during an electrolysis and the actual charge needed to produce detected
quantities of products according to known electrochemical mechanisms.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
Synthesis and Characterization of Nanoprisms. AgPd nanodendrite-modified
Au nanoprisms were synthesized by the coupling of galvanic replacement between
H
2
PdCl
process that used ascorbic acid (AA) as the reductant (Scheme 1). First, Au@Ag
nanoprisms were synthesized by slow addition of 1 mM AgNO to a mixture containing
4
and the Ag layers of Au@Ag core-shell nanoprisms with a co-reduction
3
AA, NaOH, and Au nanoprisms with an average edge length and thickness of 154.0 ±
8.4 nm and 8.3 ± 1.1 nm, respectively (Figures 1A, S1A). The resulting Au@Ag
nanoprisms have an Au:Ag ratio of 1:3, an average edge length of 200 ± 23 nm, and an
average thickness of 23.4 ± 3.7 nm (Figures 1B, S1C); the larger size of these prisms is
14
ACS Paragon Plus Environment

Chemistry of Materials
Page 16 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a result of the added Ag layer. Prominent features from Au and Ag in the EDX spectra
of Au@Ag nanoprisms (Figure S2) clearly indicate the successful deposition of Ag
layers on the Au nanoprisms. Au@Ag nanoprisms were further modified by addition
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of H
HRTEM images show that AgPd deposits on prism edges (Figures 1C, S3) and tips
Figures 1D–F) are composed of numerous dendrite nanocrystals with average sizes of
.9 ± 1.0 nm on the edges and 6.5 ± 1.2 nm at the tips. In Figure 1F, observed lattice
2
PdCl
4
to produce AgPd nanodendrite-modified Au nanoprisms. TEM and
(
6
d-spacings corresponding to the (111) and (200) planes of AgPd nanoparticles are
located between of the (111) and (200) planes of metallic Ag and Pd nanoparticles,55-57
indicating the formation of AgPd alloy nanoparticles.
Nanoprisms were further characterized by means of UV-Visible-NIR
measurements. A solution of Au nanoprisms exhibits two peaks at 790 nm and 1280
nm (Figure S4), corresponding to dipole and quadrupole surface plasmon resonances
(SPRs), respectively, which are consistent with previous reports.42,43 After deposition
of Ag layers to form Au@Ag nanoprisms, a new longitudinal dipole plasmon resonance
is observed at 870 nm.58-61 Additionally, two plasmon resonance peaks are observed at
545 nm and 340 nm; these correspond to in-plane quadrupole resonance and out-of-
plane quadrupole resonance of Ag nanoprisms, respectively.61 These results indicate
the successful formation of Au@Ag nanoprisms. After galvanic reaction of the Ag
layer with H
2
PdCl , these features red shift to 1312 nm and 1330 nm, indicating the
4
disappearance of Ag layers and the formation of AgPd nanodendrite-modified Au
nanoprisms.
15
ACS Paragon Plus Environment

Page 17 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
Structures and compositions of AgPd nanodendrite-modified nanoprisms were
further confirmed by STEM/EDX, XRD, and SEM/EDX spectroscopy. STEM and
EDX-mapping images show the triangular Au core and well-defined AgPd
nanodendrite-tipped (Figures 2A–D) as well as AgPd nanodendrite-edged structures
(Figures 2F–I). The uniformly dispersed signals of Ag and Pd in overlay images
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(Figures 2E, 2J) indicate the AgPd alloyed nanostructure, which is further confirmed
by XRD. XRD patterns of Au and Au@Ag nanoprisms clearly show four diffraction
peaks at ca. 2θ = 38.3, 44.5, 64.7, and 77.7, which can be indexed to (111), (200), (220),
and (311) facets of face-centered-cubic (fcc) metal, respectively (Figure S5). Compared
to Au and Au@Ag nanoprisms, the AgPd nanodendrite-tipped Au nanoprisms show
relatively wider peaks with a small blue shift, which is attributed to the formation of
AgPd nanodendrites. With more AgPd nanodendrites, the AgPd nanodendrite-edged
Au nanoprisms show diffraction peaks from both Au nanoprisms and AgPd
nanodendrites. Notably, peaks observed for AgPd nanodendrites are located between
those of Pd (JCPDS-46-1043) and Ag (JCPDS-04-0783), suggesting an AgPd alloy.
Content of Pd, Ag, and Au for each type of nanoprism were determined from EDX
spectra and found to be 8.80, 63.63, and 27.57 wt% for AgPd nanodendrite-tipped
nanoprisms (Figure S6A, B) and 25.51, 46.08, and 28.41 wt% for AgPd nanodendrite-
edged nanoprisms (Figure S6C, D), respectively. All above results indicate the
successful formation of AgPd nanodendrite-modified Au nanoprisms.
Controlled synthesis of AgPd nanodendrite-modified Au nanoprisms. Site-
selective growth of AgPd nanodendrites on Au nanoprisms is dependent on both the
16
ACS Paragon Plus Environment

Chemistry of Materials
Page 18 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
amount of Ag coating and the concentration of H
2
PdCl used for the galvanic
4
replacement step. With a low ratio of Ag to Au of 1:3 in the core-shell nanoprisms
Figures 3A, S2A), only AgPd nanodendrite-edged Au nanoprisms are produced, even
in the presence of different amounts of H PdCl (Figures 3B–E). At higher coatings of
Ag with ratios of Ag:Au of 3:1 (Figures 3F, S2B), both AgPd nanodendrite-tipped and
edged prisms are formed depending on the concentration of H PdCl (Figures 3G–J);
increasing the concentration of H PdCl leads from nanodendrite-tipped structures to
nanodendrite-edged structures. Similar results are observed for an Ag:Au ratio of 10:1
(
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4
-
2
4
2
4
(Figures 3K–O, S2C).
The effects of Ag content and H
2
PdCl concentration on structure follow from the
4
basic principle of this synthetic method: galvanic replacement of the Ag layer in
+
Au@Ag nanoprisms with H
2
PdCl
4
and co-reduction of H
2
PdCl
4
and Ag ions,
generated from the galvanic replacement reaction, at specific sites on nanoparticle
seeds. Reactivity of these sites varies; for example, Mulvaney and coworkers have
shown that high curvature sites on Au nanorods are the most reactive, due to a high rate
of collisions between these sites and micelles containing metal ions.62 In the case of
nanoprisms, Pd and Ag are expected to co-deposit favorably on high curvature sites at
edges and tips. Selectivity of these sites can also be tuned if one controls the rate of the
galvanic replacement reaction, in which the Ag layer, first formed by underpotential
deposition (UPD) of Ag on Au,63,64 is oxidized in the presence of Pd to release Ag+
ions and to create an excess of electrons on the particle surface. Electrons generated by
this reaction accumulate at regions of high curvature,65-69 resulting in preferential
2+
17
ACS Paragon Plus Environment

Page 19 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
2
+
reduction of Pd ions and subsequent co-deposition of Pd and Ag metal on tips and
edges. A slow rate of galvanic replacement, and thus limited generation of electrons,
allows accumulation in only the regions of highest curvature, the tips of nanoparticle
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
seeds. For this reason, particle synthesis with low concentrations of H
2
PdCl results
4
only in nanodendrite-tipped materials. Faster reaction rates will produce an
overabundance of electrons, resulting in a wider dispersion of charge and accumulation
at both tips and edges, allowing formation of both nanodendrite-tipped and –edged
materials.64
Formation of AgPd nanodendrite-modified Au nanoprisms follows the above
trend. When the Ag content of the Au@Ag nanoprism is high enough to support
continuous reaction (Ag:Au ratios of 3:1 and 10:1), the type of nanoprism formed is
2
+
dependent on the concentration of H
2
PdCl
4
present. Low availability of Pd in solution
(less than 50 L of 2 mM H
2
PdCl
4
) results in a slow galvanic reaction rate and limited
+
generation of electrons and Ag ions, allowing for migration of charge to the nanoprism
tips and subsequent deposition of Pd and Ag (Figures 3G–I, K–N). At higher
concentrations (200 L of 2 mM H
2
PdCl ), electrons are quickly consumed at tip sites,
4
2
+
+
leaving excess Pd and Ag ions to be reduced along available, electron-saturated edge
sites (Figures 3J, O). However, when the ratio of Ag:Au is low (1:3), even low
2+
concentrations of Pd overwhelm the Ag layer of the nanoprism. Fast consumption of
electrons does not allow for migration and accumulation of charge (electrons and Ag+)
at nanoprism tips, resulting in the exclusive formation of nanodendrite-edged
nanoprisms regardless of initial H
2
PdCl concentration (Figures 3B–E).
4
18
ACS Paragon Plus Environment

Chemistry of Materials
Page 20 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
These site-selective strategies can be further applied to Au nanoparticle cores with
other shapes, such as hexagonal nanoplates and octahedral nanoparticles. Au, Au@Ag,
AgPd nanodendrite-tipped and -edged Au nanoplates (Figures S7A–D, S8A–B, S9A,
S9C) and octahedral nanoparticles (Figures S7E-H, S10, S11A, S11C) were
characterized with the aid of TEM and SEM. Specific metal contents of Ag, Pd and Au
in these Au@Ag and AgPd nanodendrite-modified Au nanoplates and octahedral
nanoparticles were determined by EDX spectra (Figures S8–S11). The contents of Ag
and Au are 56.41 and 43.59 wt% for Au@Ag nanoplates (Figure S8) and 49.34 and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0.66 wt% for Au@Ag octahedral nanoparticles (Figure S10), respectively. The
contents of Pd, Ag, and Au are 4.14, 22.1, and 73.76 wt% for AgPd nanodendrite-tipped
Au nanoplates and 30.41, 21.41, and 48.19 wt% for AgPd nanodendrite-edge Au
nanoplates, respectively (Figure S9). In addition, the contents of Pd, Ag, and Au are
1
1.47, 28.34, and 60.19 wt% for AgPd nanodendrite-tipped Au octahedral
nanoparticles and 25.82, 31.01, and 43.17 wt% for AgPd nanodendrite-edged Au
octahedral nanoparticles (Figure S11). These results indicate the successful site-
selective growth of AgPd nanodendrites on different shapes of Au nanoparticles. This
strategy thus serves as a general method for site-selective nanodendrite growth based
on curvature-directed electron distribution, regardless of the shape of the central
nanoparticles.
In addition to allowing site-selective growth of nanodendrites, this synthesis
provides a simple, room-temperature method for preparation of noble metal catalysts.
24
70
Other noble metal electrocatalysts, including AuCu, carbon-supported PdPt, and
19
ACS Paragon Plus Environment

Page 21 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
Au@Pd core shell nanocatalysts,71 have been prepared with the aid of thermal
reduction, whereas the particles reported here were synthesized by galvanic
replacement and co-reduction of Ag and Pd ions at room temperature. Moreover,
previously reported noble metal electrocatalysts for CO2 reduction are primarily two-
component systems, whereas the AgPd nanodendrite-modified Au nanoprisms reported
in this work are unique, tri-component noble metal electrocatalysts where each
component plays a role.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Electrochemical behavior of CO
solution. Nanomaterials described above were drop-cast onto a carbon rod for use as
cathodes in an investigation of their efficiency for the electrochemical reduction of CO
2
at nanomaterial electrodes in aqueous
2
.
Shown in Figure 4 are cyclic voltammograms recorded at carbon electrodes modified
by deposition of prism, octahedral, and hexagonal nanomaterials composed of Au,
Au@Ag core-shell, AgPd nanodendrite-tipped, and AgPd nanodendrite-edged Au in
water containing 0.1 M LiClO
maintenance of the electrolyte solution, no buffer was used. Voltammograms were
recorded at pH 4, which was reached upon saturation with CO . For all geometries,
AgPd nanodendrite-edged Au nanoparticles show the least negative potential for CO
reduction; peak potentials are –0.18 V vs. RHE for prism and octahedral materials and
0.24 V vs. RHE for hexagonal nanoplates (Table 1). AgPd nanodendrite-tipped
nanoparticles also display excellent catalytic activity, with reduction potentials of –
.27, –0.31, and –0.26 V vs. RHE for prism, octahedral, and hexagonal nanomaterials,
respectively. Reduction of CO at solid Au nanoparticles and Au@Ag core-shell
4
as supporting electrolyte. To simplify preparation and
2
2
–
0
2
20
ACS Paragon Plus Environment

Chemistry of Materials
Page 22 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
particles is less favorable, with reduction potentials ranging from –0.32 to –0.56 V vs.
RHE.
Controlled-potential electrolyses were used to determine product distributions
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
for the reduction of CO
2
at nanoprism-modified electrodes and to explore possible
, headspace from
reaction pathways. For electrolyses in water containing 0.1 M LiClO
4
each electrolysis cell was analyzed by means of gas chromatography (GC) to detect
CO, and catholyte was analyzed by means of GC and NMR to determine solution-phase
products such as oxalic and formic acids. Bulk electrolyses were performed at peak
reduction potentials, determined by cyclic voltammetry, at which CO
2
reduction
proceeds with maximum efficiency compared to background processes. Electrolyses
7
2
yielded formate, which can be used in fuel cells, at all nanoprism-modified electrodes.
No CO or other CO reduction products were detected, indicating high selectivity for
production of formate. Faradaic efficiencies for CO reduction to formate at Au,
2
2
Au@Ag, AgPd nanodendrite-tipped Au, and AgPd nanodendrite-edged Au nanoprisms
are 8%, 10%, 21%, and 49%, respectively, as calculated from total current passed and
known n values for CO reduction reactions (Table 2). Efficiency of formate
2
production increased considerably with Pd content, as AgPd nanodendrite-tipped Au
prisms (8.8 wt% Pd) and AgPd nanodendrite-edged Au prisms (25.5 wt% Pd) showed
a two-fold and five-fold increase, respectively, over pure Au and Au@Ag nanoprisms.
AgPd nanodendrite-edged prisms thus show both the most favorable reduction potential
and the highest faradaic efficiency for formate production in water. Hydrogen
21
ACS Paragon Plus Environment

Page 23 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
evolution is believed to be the primary competing reaction in all cases, as the formation
of bubbles at the electrode surface is observed during the course of an electrolysis.
The stability of AgPd nanodendrite-modified Au nanoprisms was investigated
by means repetitive bulk electrolyses, as well as TEM and SEM characterization. Up
to three consecutive electrolyses were performed at each electrode for a total
electrolysis time of 12 h, with no significant change in current density or product
formation between the first and third use. SEM and TEM images of nanoparticles were
compared before and after a 4-h electrolysis (Figure S12). The morphology of AgPd
nanodentrite-edged Au nanoprisms was largely unchanged, indicating that AgPd
nanodentrite-edged Au nanoprisms are highly stable during long-term electrolysis.
In Table 3, electrolysis results have been compared to a number of existing
reports of other electrode materials. Reduction potentials at AgPd nanodendrite-edged
nanoprisms, the most favorable potentials observed in this study, are considerably more
positive than most reported values and on par with branched Pd nanoparticles reported
by Klinkova and coworkers.24,70,71,73-76 This improvement in reduction potential makes
AgPd nanodendrite-modified nanoprisms promising for scaled-up applications in
which applied potential and energy use are a concern. Current densities, calculated in
this work as an average current passed per geometric electrode area over the course of
electrolysis, are relatively low for AgPd nanodendrite-edged nanoprisms at only 0.2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
mA/cm , relative to those of some other materials. Klinkova and coworkers have
2
reported current densities as high as 22 mA/cm at branched Pd nanoparticles in water
containing 0.5 M KHCO , and other studies performed in bicarbonate electrolytes
3
22
ACS Paragon Plus Environment

Chemistry of Materials
Page 24 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
report current densities in the range of 4–10 mA/cm2.70,73,74,77 However, the use of
bicarbonate electrolytes in studies performed at pH 7 inflates the total availability if
CO
currents.78 This effect can be seen in a comparison of cyclic voltammetric currents at
AgPd nanodendrite-tipped prisms in LiClO and NaHCO electrolytes (Figure S13),
where current densities observed for CO reduction in bicarbonate electrolyte are nearly
0 times higher than in the absence of bicarbonate at negative potentials. Bicarbonate
2
, leading to higher CO
2
concentrations at the electrode surface and thus increased
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
3
2
1
is also known to be directly reduced by some metals, including Pd; Kanan and
coworkers observed high current densities for formate production at Pd materials in
bicarbonate buffer even in the absence of added CO
performed by Humphrey and coworkers with AuPd core-shell nanoparticles in
unbuffered 0.1 M NaSO
achieved current densities similar to this work.71 Similar
2
.79,80 In contrast, a recent study
4
trends are apparent in faradaic efficiencies reported at pH 7, which range from 10–97%,
whereas efficiencies at pH 4 range from 27–49% (Table 3). Hydrogen evolution at pH
4
is believed to account for lowered faradaic efficiency compared to work performed
at pH 7; in CV measurements, a small increase in current at pH 4 is seen compared to
pH 7 at a similar reduction potential (Figure S14). Still, efficiency at AgPd
nanodendrite-edged nanoprisms is high at 49% in comparison with materials
investigated under similar conditions.
Electrochemical mechanisms for the reduction of CO
reported for a number of solvent–electrolyte systems.81 Initially, adsorbed CO
in solution is reduced at the electrode surface to its corresponding radical anion, which
2
have been previously
2
or CO
2
23
ACS Paragon Plus Environment

Page 25 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
can then undergo further reaction along several established mechanistic pathways. For
all electrode materials explored, electrolysis of CO
2
in unbuffered aqueous solvent
radical
leads primarily to the production of formate, which is generated when the CO
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
anion is protonated by water before accepting a second electron to afford formate
(Scheme 2). This mechanism has been reported previously for the generation of formic
acid or formate at Pd electrodes, and its relevance is believed to be greatest under acidic
conditions.81-84
We propose that the unique structure and trimetallic composition of AgPd
nanodendrite-modified Au nanoprisms could be responsible for the relatively positive
reduction potential and high efficiency for CO
reported that nanomaterials with more corner and edge sites show high catalytic activity
for CO
reduction.85,86 Au and Au@Ag core shell nanoprisms exhibit only a small
2
reduction (Scheme 3A). It has been
2
number of these active sites due to their planar structure. In addition, Au surfaces are
also capable of especially strong adsorption of carbon species, often resulting in
8
6
poisoning of the catalyst surface; prevalence of hydrogen evolution at the planar Au
surface, a lack of sufficient corner and edge sites, and tight binding of CO and its
2
reduction products could explain the low yield of formate and poor catalytic activity on
Au nanoprisms. Au@Ag nanoprisms suffer from similar limitations.
In contrast, the addition of AgPd nanodendrites to particle tips results in a
dramatic increase in formate production and a considerably more favorable reduction
potential. This is due in large part to an increase in active edge and corner sites at AgPd
reduction.85,86 In addition,
nanodendrites, which have high catalytic activity for CO
2
24
ACS Paragon Plus Environment

Chemistry of Materials
Page 26 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
metal content and synergistic effects of Pd and Ag also play an important role. It has
been reported that short distances between active sites for surface stabilization of H+
and CO
physically adsorbed on Au, Ag, and Pd metals, providing many possible sites for
adsorption of CO can be adsorbed at Pd,
on mixed-metal particles.87 Although CO
2
can aid in the formation of C–H bonds for formate production.24 CO
2
can be
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
Pd is better known to have an affinity for hydrogen, resulting in high activity in
hydrogenation reactions as well as reversible adsorption and even incorporation of
hydrogen into Pd.88,89 It is well established Pd will primarily adsorb hydrogen even in
the presence of CO
combination with Ag and Au, it is apparent from our cyclic voltammograms at pH 4 in
the presence and absence of CO that a definite interaction with some form of hydrogen
occurs (S14). This would allow for optimized arrangement of both hydrogen and CO
2
.90 Though similar studies have not been reported for Pd in
2
2
on the particle surface and, concomitantly, a higher probability of their interaction at
the interface of the metals. Additionally, short distances between active sites for surface
+
stabilization of H and CO
2
should result in easier formation of C–H bonds for formate
production (Scheme 3B).24,87 When Pd content is increased to allow nanodendrite
formation over more of the particle surface, as is the case with AgPd nanodendrite-
edged Au nanoprisms, formate production and ease of reduction continue to improve,
indicating the higher catalytic activity of nanoparticles with more nanodendrites and
higher Pd content.
Although primary active sites for CO
2
reduction are presumed to be at the AgPd
nanodendrites, the Au core is also an important component of the nanoparticle. In
25
ACS Paragon Plus Environment

Page 27 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
addition to its role in nanoparticle synthesis, immobilization of highly active
nanodendrites on an Au core allows for optimized dispersion of their extensive edge
and corner sites, high electrical conductivity, and high chemical stability over the
course of electrolysis.91 The resulting nanodendrite structure and synergistic effects of
AgPd nanodendrite-modified Au nanoprisms leads to increased production of formate
and more positive reduction potentials (Scheme 3A).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Electrocatalytic behavior of CO
solvent. Electrocatalytic activities of the synthesized nanomaterials were also explored
in acetonitrile containing 0.05 M TMABF , to take advantage of the greater solubility
of CO in organic solvents; CO concentrations in acetonitrile have been reported as
approximately 10 times higher than in aqueous solution.92 Cyclic voltammograms for
reduction of CO at each electrode are shown in Figure 5, and a summary of peak
2
at nanomaterial electrodes in an organic
4
2
2
2
potentials is included in Table 1. Electrodes prepared with AgPd nanodendrite-tipped
and -edged Au nanomaterials of different geometries all show similar behavior. In each
case, a single cathodic peak is present between –1.58 and –1.69 V vs. SHE. Cyclic
voltammograms at Au and Au@Ag core-shell nanomaterials show multiple peaks,
indicating different reaction pathways and products formed upon reduction of CO . Au
2
nanomaterials exhibit one peak between –1.32 and –1.39 V vs. SHE, with a second
peak between –1.58 and –1.98 V vs. SHE. The potential of the more negative peak is
greatly dependent on particle geometry. Au@Ag core-shell nanomaterials show similar
behavior, though the first reduction peak is more negative than that for Au (–1.47 to –
1
.51 V vs. SHE).
26
ACS Paragon Plus Environment

Chemistry of Materials
Page 28 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Controlled-potential (bulk) electrolyses were used to determine the products of
reduction. Gas chromatographic analysis of solution headspace and catholyte
CO
2
treated with BSTFA:TCMS was used to detect all products except formate. Solution-
phase products were quantitated by comparison to an internal standard, whereas
quantitation of CO was achieved with a calibration curve (Figure S15). The presence
of formate was probed with NMR, and not detected for electrolyses in organic solvents.
Table 4 shows faradaic efficiencies for the formation of a range of products resulting
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
from electrolysis in acetonitrile containing TMABF , including CO (used as fuel for
4
9
3
solid oxide fuel cells as well as syngas), oxalic acid, and further reduction products
of oxalic acid. AgPd nanodendrite-tipped and -edged Au prisms show remarkable
efficiency for the production of CO gas, at 87 and 85%, respectively, after electrolysis
at –1.5 V vs. SHE. No other products were detected from electrolyses at these
electrodes, indicating high product selectivity. Electrolyses at Au and Au@Ag core-
shell prisms were performed at both –1.5 and –1.9 V vs. SHE to explore the products
generated at both peaks observed for cyclic voltammograms. At –1.5 V vs. SHE (the
same potential used for electrolysis at AgPd nanodendrite electrodes), Au and Au@Ag
core-shell nanoprisms show a faradaic efficiency for CO production of only 64% and
5
4%, respectively. The remaining current leads to the formation of oxalic acid, which
is then immediately reduced to glyoxylic and glycolic acid. Efficiency of CO
production was lower at –1.9 V vs. SHE, at only 45% for Au prisms and 43% for
Au@Ag core-shell prisms. The remaining products are again oxalic, glyoxylic, and
glycolic acid. Additionally, trends in faradaic efficiencies indicate that the less negative
27
ACS Paragon Plus Environment

Page 29 of 56
Chemistry of Materials
1
2
3
4
5
6
7
8
9
CO
2
reduction peak seen at Au and Au@Ag core-shell prisms is likely due to formation
of CO; however, even electrolyses at this potential generate a mixture of products.
These results indicate that AgPd nanodendrite-modified Au prisms are more selective
for production of CO than prisms without Pd.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In an organic solvent, CO
pathways than observed in aqueous conditions (Scheme 4). CO
electrode surface to its corresponding radical anion, as is the case in an aqueous
mechanism. CO radical anions can then form an intermediate adduct with an
unreduced CO molecule through coupling of carbon and oxygen before accepting a
second electron, resulting in the cleavage of a C–O bond to form CO gas and a
carbonate ion.82 A CO
radical anion may also undergo radical coupling with a second
2
reduction appears to follow a different set of reaction
2
is reduced at the
2
2
2
radical anion to form oxalate, which can then be protonated by adventitious water to
form oxalic acid.82 Oxalic acid is capable of undergoing a two-electron reduction in
+
the presence of H to form glyoxylate, followed by subsequent two-electron reduction
to afford glycolate.94
To understand better the mechanism by which oxalic acid and its reduction
products are formed at Au and Au@Ag core-shell nanoprisms, additional cyclic
voltammograms were recorded at Au@Ag core-shell nanoprisms in the presence and
absence of known quantities of water (Figure S15). The more negative cyclic
voltammetric peaks observed at Au@Ag core-shell prisms increase in current with the
+
addition of water, indicating an expected dependence on H . The absence of a second
voltammetric peak at AgPd nanodendrite Au electrodes and the lack of oxalic acid and
28
ACS Paragon Plus Environment