The Homogeneous Reduction of CO2 by [Ni(cyclam)]+: Increased Catalytic Rates with the Addition of a CO Scavenger

Article abstract of DOI:10.1021/ja512575v

The homogeneous electrochemical reduction of CO₂ by the molecular catalyst [Ni(cyclam)]²⁺ is studied by electrochemistry and infrared spectroelectrochemistry. The electrochemical kinetics are probed by varying CO₂ substrate and proton concentrations. Products of CO₂ reduction are observed in infrared spectra obtained from spectroelectrochemical experiments. The two major species observed are a Ni(I) carbonyl, [Ni(cyclam)(CO)]⁺, and a Ni(II) coordinated bicarbonate, [Ni(cyclam)(CO₂OH)]⁺. The rate-limiting step during electrocatalysis is determined to be CO loss from the deactivated species, [Ni(cyclam)(CO)]⁺, to produce the active catalyst, [Ni(cyclam)]⁺. Another macrocyclic complex, [Ni(TMC)]⁺, is deployed as a CO scavenger in order to inhibit the deactivation of [Ni(cyclam)]⁺ by CO. Addition of the CO scavenger is shown to dramatically increase the catalytic current observed for CO₂ reduction. Evidence for the [Ni(TMC)]⁺ acting as a CO scavenger includes the observation of [Ni(TMC)(CO)]⁺ by IR. Density functional theory (DFT) calculations probing the optimized geometry of the [Ni(cyclam)(CO)]⁺ species are also presented.

Full text of DOI:10.1021/ja512575v

Article
The Homogeneous Reduction of CO2 by [Ni(cyclam)]+:
Increased Catalytic Rates with the Addition of a CO Scavenger
Jesse D Froehlich, and Clifford P. Kubiak
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 25 Feb 2015
Downloaded from http://pubs.acs.org on February 26, 2015
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.
Journal of the American Chemical Society 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 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
The Homogeneous Reduction of CO2 by [Ni(cyclam)]+:
Increased Catalytic Rates with the Addition of a CO Scaven-
ger
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
Jesse D. Froehlich and Clifford P. Kubiak*
Department of Chemistry and Biochemistry, University of California, San Diego, 9500 Gilman Drive, La Jolla,
California 92093ꢀ0358
KEYWORDS: homogeneous CO2 reduction electrocatalysis spectroelectrochemistry Ni(cyclam).
ABSTRACT: The homogeneous electrochemical reduction of CO2 by the molecular catalyst [Ni(cyclam)]2+ is studied by
electrochemistry and infrared spectroelectrochemistry. The electrochemical kinetics are probed by varying CO2 subꢀ
strate and proton concentrations. Products of CO2 reduction are observed in infrared spectra obtained from spectroelecꢀ
trochemical experiments. The two major species observed are a Ni(I) carbonyl, [Ni(cyclam)(CO)]+, and a Ni(II) coordiꢀ
nated bicarbonate, [Ni(cyclam)(CO2OH)]+. The rateꢀlimiting step during electrocatalysis is determined to be CO loss
from the deactivated species, [Ni(cyclam)(CO)]+, to produce the active catalyst, [Ni(cyclam)]+. Another macrocyclic
complex, [Ni(TMC)]+, is deployed as a CO scavenger in order to inhibit the deactivation of [Ni(cyclam)]+ by CO. Addiꢀ
tion of the CO scavenger is shown to dramatically increase the catalytic current observed for CO2 reduction. Evidence
for the [Ni(TMC)]+ acting as a CO scavenger includes the observation of [Ni(TMC)(CO)]+ by IR. DFT calculations probꢀ
ing
the
optimized
geometry
of
the
[Ni(cyclam)(CO)]+
species
are
also
presented.
INTRODUCTION
CO2 reduction (rather than proton reduction) in organic
solvents with added proton sources. Molecular catalysts
that contain earthꢀabundant metal centers and operate
in aqueous solutions with high efficiency, turnꢀover
rates, and selectivity are rare. The macrocyclic complex,
The electrochemical reduction of CO2 to higher energy
products has received considerable attention recently.1ꢀ3
The majority of the wellꢀstudied systems produce CO or
formate as the major reduced product. These are considꢀ
ered to be intermediates on the path to liquid fuels. Utiꢀ
lizing CO2 as a feedstock to create liquid fuel would both
help to alleviate our dependence on fossil fuels as well as
help mitigate rising atmospheric CO2 levels by creating a
carbon neutral source of combustible energy.
[Ni(cyclam)]2+
(cyclam
=
1,4,8,11ꢀ
tetraazacyclotetradecane) and its analogs, is one such
example that has been extensively studied.13ꢀ15 This moꢀ
lecular catalyst system has the unique ability to maintain
efficient production of CO in an aqueous solution, howꢀ
ever, the overwhelming majority of these studies use a
mercury (Hg) working electrode. Hg has a very negative
potential for proton reduction giving it an extensive negꢀ
ative potential solvent window in water making it a popꢀ
ular electrode choice for aqueous reduction electrochemꢀ
istry. Furthermore, [Ni(cyclam)]+ has been shown to
adsorb onto the Hg surface which enhances its catalytic
activity for CO2 reduction.14,16 Our group’s recent work
demonstrated that [Ni(cyclam)]+ still has considerable
catalytic activity for CO production at an inert glassy
carbon electrode.17 The homogeneous (Hg free) reducꢀ
tion of CO2 by [Ni(cyclam)]+ has not been thoroughly
explored and is the subject of the present study.
The heterogeneous reduction of CO2 by metallic elecꢀ
trodes has been explored,4 however the mechanisms of
these systems can be difficult to study and often suffer
from poisoning of the electrode by the intermediates or
products of catalysis.5 Transition metal molecular cataꢀ
lysts can be used in a homogeneous fashion by acting as
an electron shuttle between an electrode and CO2 in soꢀ
lution. The molecular catalyst serves to stabilize interꢀ
mediate species allowing for less negative potentials to
be used in the reduction of CO2. In addition, these inꢀ
termediates can be characterized with solution based
techniques not available to heterogeneous systems
thereby providing more mechanistic insight into the reꢀ
duction of CO2.
EXPERIMENTAL PROCEDURES
Many different molecular catalyst systems have been
studied and can be found in recent reviews.6ꢀ9 Many of
these utilize expensive, rare earth metals. There are sevꢀ
eral notable catalysts that contain abundant first row
transition metals such as Mn, Fe, Co and Ni.6,10ꢀ12 Several
of these systems were shown to have high selectivity for
Cyclic voltammetry was performed on a Gamry Referꢀ
ence 600 potentiostat with a glassy carbon (GC) working
electrode, platinum counter and Ag/Ag+ (BASi) reference
electrode separated by a piece of vycor glass. 0.1M tetꢀ
rabutylammonium hexafluorophosphate (TBAPF6) was
ACS Paragon Plus Environment

Journal of the American Chemical Society
used as the electrolyte in all cyclic voltammetry and specꢀ
troelectrochemical experiments. All cyclic voltammoꢀ
gram potentials were converted to normal hydrogen elecꢀ
trode (NHE) reference by adding 0.54 V.
Page 2 of 11
1
2
3
4
5
6
7
8
9
The acetonitrile (ACN) used was dried over basic alumiꢀ
na with a custom dry solvent system. The CO2 used was
flowed through a column packed with Drierite during
sparging. Under 1 atm, the concentration of CO2 in satuꢀ
rated aqueous18 and ACN19 solutions are reported as
0.036 and 0.28 M respectively. The concentration of CO2
in a 1:4 water:ACN mixed solvent system is estimated to
be ≈190 mM.20 The concentration of CO in saturated
aqueous18 and ACN19 solutions are reported as 0.001 and
0.0083 M respectively. The concentration of CO in a 1:4
water:ACN mixed solvent system is estimated to be ≈6
mM. The CO2 concentration study utilized a flow meter
system with the flow rates Ar and CO2 controlled and
then mixed. The gas mixture was then flowed through an
ACN solution and then into a sparging needle in the CV
cell so as to minimize solvent evaporation. The ratio of
the flow rate of the two gasses was used to calculate the
equilibrium concentration of CO2 in solution. (Warn-
ing: small amounts of highly toxic Ni(CO)4 can be genꢀ
erated during the reduction of Ni complexes in the presꢀ
ence of CO.)
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
Scheme 1. The Proposed CO2 reduction catalytic
cycle for [Ni(cyclam)]2+
.
Electrochemistry in the Absence of Added Protons
The CV of [Ni(cyclam)]2+ in acetonitrile (ACN) under N2
and CO2 can be seen in Figure 1. With the addition of
CO2 to the [Ni(cyclam)]2+ solution in ACN the
Ni(II)/Ni(I) redox couple becomes irreversible and the
reduction peak has a slight increase in current and is
shifted positively by 30mV. The positive shift is due to
CO2 binding to Ni(I). The irreversibility may arise from
two possibilities: 1) CO2 binding is an irreversible proꢀ
cess or 2) There is a fast, irreversible chemical step (eq 2)
following CO2 binding or an ECC mechanism. Possibility
1) may be ruled out by consideration of the fast kinetics
of CO2 binding. Kelly et. al. reported the kinetics for the
process shown in eq 1 in aqueous solution to be kf = 3.2 x
107 Mꢀ1sꢀ1 and kr = 2.0 x 106 sꢀ1 for the forward and reverse
reactions respectively.22 These rate constants imply that
CO2 binding is a reversible process on the timescale of
CV. For possibility 2), the fast, irreversible chemical step
following CO2 binding could be an isomerization or proꢀ
tonation. It has been reported that reduction of a divaꢀ
lent Ni complex was accompanied by an isomerization
process.23 If it is due to a protonation step, the protons
would have to come from the electrolyte, solvent, residuꢀ
al water or [Ni(cyclam)] amine protons.
A schematic of the infraredꢀspectroelectrochemistry (IRꢀ
SEC) setup used can be seen in a recent report.21 The
electrodes in the IRꢀSEC cell consist of a glassy carbon
disk for the working electrode, a platinum counter and a
silver quasiꢀreference. IRꢀSEC samples were prepared in
a dry/N2 glove box by bubbling 10 mL of CO2 through a 1
mL solution of the Ni(II) species and 0.1M TBAPF6 and
injecting into the IRꢀSEC cell. The IR used was a Thermo
Scientific Nicolet 6700 FTꢀIR. The potential was held for
10 seconds before the IR scan was initiated. The scans
(16 scans) were completed in roughly 90 seconds.
RESULTS AND DISCUSSION
Electrochemistry
The cyclic voltammetry (CV) of [Ni(cyclam)]2+ in an
aqueous and mixed aqueous/acetonitrile solvent has
been reported previously.17 The mixed solvent system is
convenient because the Ni(II)/Ni(I) couple is observable
at a glassy carbon electrode whereas in an all aqueous
system, without addition of CO2, the reduction couple is
obscured by the solvent window (proton reduction). Afꢀ
ter saturation with CO2 both the aqueous and mixed solꢀ
vent systems show high Faradaic efficiencies for CO proꢀ
duction suggesting the active catalyst in the Ni(I) reꢀ
duced state preferentially reacts with CO2 rather than
protons. Although, in highly acidic solutions H2 producꢀ
tion dominates over CO production.
[Ni(cyclam)]+ + CO2 ⇌ [Ni(cyclam)(CO2)]+
[Ni(cyclam)(CO2)]+ → intermediate
(1)
(2)
The overall mechanism for CO2 reduction by
[Ni(cyclam)]2+ is summarized in Scheme 1.14 The Ni cenꢀ
ter is first reduced to generate the active catalyst,
[Ni(cyclam)]+, which in turn can react with CO2. In the
presence of a proton source, significant catalytic current
is observed. In order to further investigate the mechaꢀ
nism of CO2 reduction, the electrochemistry was studied
in a protic free organic solvent system and CO2 and proꢀ
ton concentration studies were undertaken.
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Figure
1.
Cyclic
voltammograms
of
1
mM
Figure 2. Effect of added H2O on the cyclic voltammogram.
1 mM [Ni(cyclam)](PF6)2, 0.1M TBAPF6 in ACN saturated
with CO2. Additions of aqueous 1mM [Ni(cyclam)](PF6)2 in
order to maintain constant catalyst concentration. Scan rate
= 0.1 V/s.
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
[Ni(cyclam)](PF6)2, 0.1 M TBAPF6 in ACN. Current normalꢀ
ized by square root of scan rate.
Interestingly, a reversible couple can be obtained under
CO2 at faster scan rates. Figure 1 also shows the CV of
[Ni(cyclam)]2+ under CO2 at various scan rates. Reversiꢀ
bility of the Ni(II)/Ni(I) couple is regained with scan
rates above 3 V/s. With a reversible couple, the CO2
binding constant (ꢀꢁꢂꢃ) can be calculated from the shift
in the Ni(II)/Ni(I) couple (ꢁE) under N2 and CO2 and eq
3.24
Electrochemistry with Added Protons
It is clear from the lack of significant catalytic current
(icat) in ACN that high turnꢀover frequencies (TOF) are
not possible without the addition of a proton source.
Multiple turnꢀovers are not possible if relying only on
trace water content in the ACN and no clear additional
reduction peaks are observed without added water indiꢀ
cating that the unprotonated CO2 adduct is not reduced
at these potentials. Figure 2 shows CVs with increasing
amounts of water as the proton source. As water is addꢀ
ed, two catalytic peaks develop. The first catalytic peak,
a, simultaneously increases in current density and shifts
positively in potential suggesting a proton dependent
electron transfer. This agrees with the reported mechaꢀ
nism in which the second reduction occurs only after
protonation of the CO2 adduct.14 At a water concentraꢀ
tion of 4 M peak a approaches the same potential as the
[Ni(cyclam)]2+ reduction peak under CO2 (ꢀ1.19V vs.
NHE). The second catalytic peak, b, increases reaching a
maximum current density and a peak potential of ꢀ1.59 V
at ≈4 M water, then slightly decreasing at 11 M water.
The current in peak b is assumed to be, in part, the reꢀ
duction of a [Ni(cyclam)(CO)]+ species produced by reacꢀ
tion of [Ni(cyclam)]+ with CO. Additional contributions
to the current in b could also be from CO2 or proton reꢀ
duction by a Ni(0) species or at the electrode. Evidence
of [Ni(cyclam)(CO)]+ forming during CO2 reduction by
[Ni(cyclam)]+ has been reported previously on a Hg elecꢀ
trode.14
ꢀꢁꢂ ꢄCOꢅꢆ ꢇ ꢈꢉꢊꢋꢌꢍ/ꢎꢏꢐ – 1
(3)
ꢃ
Using this electrochemical determination method KCO2
was found to be 6 Mꢀ1 which is in very good agreement
with the previous reported value (4±2 Mꢀ1).25 In addition
to an equilibrium constant, information about the rate of
the chemical step following CO2 binding can be obtained.
Assuming that with faster scan rates the chemical step
(eq 2) cannot proceed because the Ni(I) metal center is
reꢀoxidized to Ni(II) before the system can reach equilibꢀ
rium, the halfꢀlife (t1/2) is estimated to be ≈0.2 s under
these conditions. It should be noted that if water is addꢀ
ed to the solution the reversibility is lost even at higher
scan rates (Figure S1). This suggests that the chemical
step responsible for irreversibility is most likely protonaꢀ
tion of the CO2 adduct. Even though care was taken to
exclude water from the electrolyte solution by solvent
purification and drying of the CO2 stream used, there is
probably water in concentrations in excess of the catalyst
in solution (>1 mM). The oxygens of the CO2 adduct
would be basic in nature and are capable of abstracting a
proton under these conditions – especially if a protonꢀ
coupled electron transfer (PCET) mechanism is invoked.
A recent computational study concluded that PCET is the
most likely mechanism for the protonation of the CO2
adduct at the moderate potentials where catalysis is obꢀ
served.26
Further information of the nature of the second catalytic
peak, b, can be obtained by varying the CO2 concentraꢀ
tion. Figure 3 shows the CV of 5 mM [Ni(cyclam)]2+ in a
1:4 water:ACN solution. The first catalytic peak, a, quickꢀ
ly reaches a plateau as the CO2 concentration exceeds
that of the catalyst. This suggests that CO2 binding is not
a rate limiting step at high CO2 concentrations. Peak b
begins to appear at [CO2] > 5 mM suggesting that it is
related to CO2 reduction by a species other than
[Ni(cyclam)]+. The catalytic current under excess CO2
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 11
and proton source remains peak shaped. This implies
upon the further reduction [Ni(cyclam)(CO)]+. This speꢀ
1
2
3
4
that substrate consumption is not the cause of the peak
shape but is most likely due to an inhibition process such
as catalyst degradation.
cies is most likely a Ni(0) carbonyl. This similarity in the
CV reduction and oxidation peaks under CO and CO2
provide strong evidence that one or more stable nickel
carbonyl species are formed in the homogeneous reducꢀ
tion of CO2 at a glassy carbon electrode.
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
Figure 3. Cyclic voltammograms of 5 mM [Ni(cyclam)](Cl)2,
0.1M TBAPF6 in 1:4 water: ACN. Adding CO2 with flow meꢀ
ter. The positive scan portion has been removed for clarity.
Scan rate = 0.1 V/s.
Figure 4. Cyclic voltammograms of 5 mM [Ni(cyclam)](Cl)2,
0.1M TBAPF6 in 1:4 water:ACN. The solution was saturated
with CO (black) and then incrementally sparged with aliꢀ
quots of CO2 until saturated (red). Scan rate = 0.1 V/s. Inꢀ
set: zoomed region of return oxidation peaks.
Electrochemistry Under CO
[Ni(cyclam)]+ + CO ⇌ [Ni(cyclam)(CO)]+
(4)
IRꢀSEC
[Ni(cyclam)(CO)]+ + e– → Ni(0) carbonyl species (5)
In an effort to identify the nickel carbonyl species, IR
spectroelectrochemistry (IRꢀSEC) experiments were carꢀ
ried out. An IRꢀSEC cell with a glassy carbon working
electrode was constructed in order to match the CV exꢀ
periments. Figure 5a shows several peaks growing in as
the potential is decreased below ꢀ1.2 V (vs. Ag quasiꢀ
reference) under CO2. The peak at 1955 cmꢀ1 is attributed
to [Ni(cyclam)(CO)]+ which was verified by doing the
same experiment under CO (Figure S2). This
[Ni(cyclam)(CO)]+ species has also been observed before
by IR at 1955 cmꢀ1 in ACN by chemical reduction of
[Ni(cyclam)]2+ with a NaꢀHg amalgam and reaction with
CO.27 The peaks 1666, 1650 (shoulder) and 1615 cmꢀ1 that
concurrently grow in are assigned as a Ni(II) coordinated
bicarbonate species, [Ni(cyclam)(CO2OH)]+. The IR
spectrum of a solution of [Ni(cyclam)]2+ and tetraeꢀ
thylammonium bicarbonate in equal concentrations
yielded an identical spectrum (Figure S3). No free bicarꢀ
bonate was seen in the IRꢀSEC with [Ni(cyclam)]2+. The
origin of the bicarbonate in this electrocatalytic system is
most likely due to CO2 acting as a Lewis acid towards the
expelled OHꢀ in the CO forming step. This process most
likely helps promote the CꢀOH bond cleavage of the CO2
adduct. At more negative potentials there is a slight deꢀ
crease in the 1955, 1666 and 1615 cmꢀ1 peaks and a peak
at 2042 cmꢀ1 grows in. This peak at 2042 cmꢀ1 is assigned
as Ni(CO)4 by comparison with a genuine sample. The
same IRꢀSEC experiment was done using isotopically
labeled 13CO2 to verify these peaks were arising from reꢀ
action with CO2. The comparison between CO2 and 13CO2
can be seen in Figure 5b. With labeled 13CO2 the
[Ni(cyclam)(13CO)]+ species shows νCO at 1911 cmꢀ1. The
Since CO is the major product of the reduction of CO2 by
[Ni(cyclam)]2+, the electrochemistry was also studied
under CO. The CV of [Ni(cyclam)]2+ in a CO saturated 1:4
water:ACN solution (≈6mM CO) can be seen in Figure 4
(black curve). There is a large positive shift in the
Ni(II)/Ni(I) couple under CO relative to N2 due to CO
binding (eq 4). The CO binding constant (Kco) calculated
from this shift has been reported before in ACN27 and
water28,29 and is reported in Table 2. The Ni(II)/Ni(I)
couple remains reversible under CO if the scan is reꢀ
versed before ꢀ1.3 V. If scanned more negative, an irreꢀ
versible reduction peak is observed at ꢀ1.6 V which is a
similar potential as the second catalytic peak, b, seen
under CO2. The irreversibility of the second reduction
peak in Figure 4 suggests the formation of an unstable
Ni(0)(cyclam)(CO) species (eq 5). This formally 20 e–
species is predictably not a stable electronic configuraꢀ
tion and could degrade via ligand loss to form more staꢀ
ble Ni(0) carbonyl species. Some of these Ni(0) species
may also have CO2 reduction capabilities. As seen in Figꢀ
ure 4, as CO2 is incrementally sparged into the solution,
this second peak increases in intensity but the peak poꢀ
tential remains relatively constant. This implies that this
peak is due to the reduction of the same species whether
under CO or CO2. The same reꢀoxidation peaks at ≈ ꢀ1
and ꢀ0.3 V are present under both CO and CO2. Similar
anodic features were also reported with a Hg working
electrode.28,30 We ascribe the anodic peak at ≈ ꢀ1 V to be
the oxidation of [Ni(cyclam)(CO)]+ and the anodic peak
at ≈ ꢀ0.3 V to be the oxidation of a species that forms
ACS Paragon Plus Environment

Page 5 of 11
Journal of the American Chemical Society
peaks assigned to [Ni(cyclam)(13CO2OH)]+ appear at
Table 1. IR frequencies for peaks observed from
IR-SEC.
1624 and 1578 cmꢀ1. Ni(13CO)4 is observed at 1996 cmꢀ1. A
summary of the IR peaks can be found in Table 1. The
observation of considerable amounts of nickel carbonyl
products, [Ni(cyclam)(CO)]+ and Ni(CO)4, in the IRꢀSEC
experiment is clear evidence of catalyst deactivation by
the CO produced by the reduction of CO2. The formation
of [Ni(cyclam)(CO)]+ is predicted by the difference in
binding and rate constants for CO2 and CO to
[Ni(cyclam)]+ (Table 2). The magnitudes of the CO bindꢀ
ing and forward rate constants for eq 4 are ≈104 times
≈100 times the corresponding value for CO2 respectively.
These favorable thermodynamic and kinetic parameters
for CO binding predict that [Ni(cyclam)]+ will preferenꢀ
tially react with any CO in solution and become deacꢀ
tivated toward reaction with CO2. At more negative poꢀ
tentials the [Ni(cyclam)(CO)]+ formed will be further
reduced to generate Ni(0) carbonyl products including
Ni(CO)4. Catalytic deactivation by Ni(0) carbonyl prodꢀ
ucts has also been proposed with a Hg electrode.30
1
2
3
4
5
6
7
8
Assignment
νCO
ν13
∆ν
CO
(cmꢀ1)
(cmꢀ1)
1911
(cmꢀ1)
[Ni(cyclam)(CO)]+
1955
44
[Ni(cyclam)(CO2OH)]+ 1666
1624
1578
1996
42
1615
37
Ni(CO)4
2042
46
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
Table 2. Reported binding constant (KL) and
forward (kf) and reverse (kr) rate constants for
the binding of ligand L to [Ni(cyclam)]+.
L
Solv. KL (Mꢀ1)
CO2 ACN 4±2
aq. 16
kf (Mꢀ1sꢀ1)
kr (sꢀ1)
Ref.
25
3.2 x 107
2.0 x 106
22
27
CO ACN (2.8±0.6) x 105
28,29
29
aq.
aq.
7.5 x 105
(2±0.2) x 109
3 x 107
22
H+ aq.
60 (pKa=1.8)
5 x 105
Addition of a CO Scavenger
With such a high affinity for CO binding the concentraꢀ
tion of the active catalyst [Ni(cyclam)]+ is greatly deꢀ
creased with each catalyst turnover due to production of
CO. In an effort to determine how this catalyst deactivaꢀ
tion affects catalytic rates another Ni complex,
[Ni(TMC)]2+ (TMC
=
1,4,8,11ꢀtetramethylꢀ1,4,8,11ꢀ
tetraazacyclotetradecane), was used as a CO scavenger. It
was shown previously that [Ni(TMC)]+ was not an effecꢀ
tive CO2 reduction catalyst at a glassy carbon electrode.17
It also has a Ni(II)/Ni(I) reduction potential significantly
more positive than [Ni(cyclam)]2+ so it does not interfere
with the observed catalytic current from the reduction of
CO2 by [Ni(cyclam)]+. Finally, [Ni(TMC)]+ also has a
large binding constant for CO, KCO = (1.2±0.4) x 105,31 to
allow it to act as a CO scavenger. Figure 6 shows the efꢀ
fect of the concentration of [Ni(TMC)]2+ on the CV of
[Ni(cyclam)]2+ under CO2. As [Ni(TMC)]2+ is added there
are 4 major effects on the voltammogram:
1) The catalytic current of the two major reduction peaks
(a and b) are increased.
2) The second catalytic peak b is replaced by b’. b and b’
correspond to the reduction of [Ni(cyclam)(CO)]+ and
[Ni(TMC)(CO)]+ respectively.
3) The return oxidation peak of [Ni(TMC)] is shifted
from c’ to c. c’ and c correspond to the oxidation of
[Ni(TMC)(CO)]+ and [Ni(TMC)]+ respectively.
4) Catalytic peak a is shifted negative in potential.
Figure 5. a) IRꢀSEC (GC working electrode) of 20 mM of
[Ni(cyclam)](PF6)2 in ACNꢀd3 under CO2. b) IRꢀSEC comꢀ
paring CO2 and cꢀ13 labeled 13CO2.
All of these observations are consistent with the
[Ni(TMC)]+ acting as
a
CO scavenger freeing
[Ni(cyclam)]+ from deactivation by CO and allowing for
higher catalytic currents. The negative shift in peak a
which accompanies the higher currents can be explained.
The addition of the CO scavenger allows for a greater
number of catalytic turnꢀovers by [Ni(cyclam)]+. A faster
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 11
consumption of the local proton source would shift the
to [Ni(TMC)]+ must come from catalytic turnꢀover of
[Ni(cyclam)]+ because no significant peak is observed at
1967 cmꢀ1 with only 20 mM [Ni(TMC)]2+ under CO2 at
catalytic potentials (red curve in Figure 7). Two peaks at
1675 and 1646 cmꢀ1 corresponding to uncoordinated biꢀ
carbonate are also observed. The broad peaks observed
around 1850 cmꢀ1 are assumed to be Ni(0) carbonyl speꢀ
cies as they are also observed under CO and only grow in
at more negative potentials than the Ni(I) carbonyl speꢀ
cies are first observed. This spectral data further supꢀ
ports CO binding to [Ni(TMC)]+ under electrocatalytic
conditions thereby minimizing the free CO near the elecꢀ
trode surface that would otherwise deactivate the active
[Ni(cyclam)]+ species (Scheme 2). It would appear that
[Ni(TMC)(CO)]+ can also form Ni(0) carbonyl species at
these potentials.
1
2
3
4
5
6
7
8
catalytic peak more negative if a proton dependent elecꢀ
tron transfer was becoming the rate limiting step. This is
consistent with the data in Figure 2 in which additions of
H2O led to a positive shift in the catalytic peak a. The
relative rate of catalysis can be determined by measureꢀ
ment of the peak catalytic current (icat) divided by the
peak Faradaic current (ip) obtained from a CV under CO2
and N2 respectively. This ratio (icat/ip), though more soꢀ
phisticated electrochemical methods have been develꢀ
oped recently,32 is one of the simplest ways to measure
the relative catalytic rate. The icat/ip of peak a in Figure 6
for the various concentrations of [Ni(TMC)]2+ can be
seen in Table 3. A tenꢀfold increase is observed at the
highest concentration of [Ni(TMC)]2+ studied (20 mM).
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
Figure 6. Cyclic voltammograms of 1 mM [Ni(cyclam)](Cl)2,
0.1 M TBAPF6 in 1:4 water:ACN saturated with CO2 with
additions of [Ni(TMC)](Cl)2. Inset: Zoom of oxidation reꢀ
gion of [Ni(TMC)]. Scan rate = 0.1 V/s.
Figure 7. IRꢀSEC (GC working electrode) of 1 mM of
[Ni(cyclam)](PF6)2 and or 20 mM [Ni(TMC)](PF6)2 in ACN
under CO2 or CO. Potential = ꢀ1.3 V (vs. Ag quasiꢀreference).
Table 3. icat /ip of peak a in Figure 6.
[Ni(TMC)]2+ (mM) icata (ꢂA)
icat/ip
1
0
17 (N2)
55 (CO2)
108
0
3.2
1
6.4
2
145
8.5
5
235
13.8
21.4
28.4
36.7
10
15
20
364
482
Scheme 2. Proposed degradation pathway of
[Ni(cyclam)]2+ and inhibition of degradation
with [Ni(TMC)]+ as a CO scavenger.
624
a
The diffusion current from Ni(TMC)2+ was subtracted
from the peak to get icat
.
IRꢀSEC experiments were also conducted with added
[Ni(TMC)]2+. In Figure 7 a small peak at 1955 cmꢀ1 is
observed with only 1 mM [Ni(cyclam)]2+ under CO2 as
expected. However, when 20 mM [Ni(TMC)]2+ is added a
much larger peak at 1967 cmꢀ1, which corresponds to
DFT Calculations of the [Ni(cyclam)(CO)]+ Structure
There are no examples of crystal structures of nickel
tetraazamacrocyclic complexes containing CO as a ligꢀ
and. Therefore, to understand the favored geometry of
such a complex, gas phase density functional theory
(DFT) calculations were done using the functional BP86
[Ni(TMC)(CO)]+,
is
observed.
The
species
[Ni(TMC)(CO)]+ has been observed by IR at 1967 cmꢀ1
previously, generated from chemical reduction of
[Ni(TMC)]2+ and reaction with CO.31 The CO that binds
ACS Paragon Plus Environment

Page 7 of 11
Journal of the American Chemical Society
(more details found in supplemental information secꢀ
tion).
(2.0±0.2) x 109 Mꢀ1sꢀ1) followed by a much slower firstꢀ
order reaction with a rate constant of 1.8±0.2 sꢀ1. The
authors ascribed this to an isomerization process from
transꢀIII to transꢀI isomer of [Ni(cyclam)(CO)]+.29 This
would suggest that the transꢀI form is the more stable
isomer once a carbonyl ligand is bound.
1
2
3
4
Chart 1. Structures of the trans-I and trans-III
Isomers of [Ni(cyclam)]2+
5
6
7
8
9
It is clear that [Ni(cyclam)]+ forms a stable CO adduct by
this DFT calculation and by experiment (high KCO) and
that this species is formed in the electrocatalytic reducꢀ
tion of CO2. By examination of the SOMO of the DFT
structures in Figure 8 it can also be seen that reduction
of this species places an electron in an orbital which has
antiꢀbonding character with respect to the NiꢀN bonds
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
Figure
8
shows the optimized geometry of
thereby
weakening
these
bonds.
The
20e–
[Ni(cyclam)(CO)]+ for two different isomers, transꢀI and
transꢀIII, which are known to exist in a roughly 15% and
85% equilibrium, respectively, in aqueous solution.33 It
can be seen by the structures in Figure 8 that there are
significant distortions to the planar geometry with a
bound CO. For comparison, the DFT optimized strucꢀ
tures of [Ni(cyclam)]+ only show a slight deviation from
planarity with NꢀNiꢀN angles of 176 and 174° for transꢀI
and 179 and 179° for transꢀIII. The different isomers of
[Ni(cyclam)(CO)]+ distort differently as apparent by the
NꢀNiꢀN angles. The transꢀIII structure has a Ni core that
is displaced out of the plane of the cyclam nitrogens with
nearly equal NꢀNiꢀN angles of 146 and 147°. In the transꢀ
I structure, the NꢀNiꢀN angles are 130 and 160° resemꢀ
bling a more trigonal bipyramidal geometry with the CO
occupying an equatorial position. This geometry can help
stabilize Ni carbonyls by backbonding interactions.34 The
bending of the NꢀNiꢀN angle raises the energy of the dxz
and dyz metal orbitals due to overlap with cyclam ligand.
This increase in energy allows for better mixing with the
π* of the CO ligand and stronger back donation and is
observed in the SOMOꢀ2 orbital shown in Figure 8. Moꢀ
lecular mechanics calculations have shown that the
transꢀI isomer is more flexible and better suited for out
of plane distortions.35 EXAFS measurements on an unꢀ
saturated macrocyclic nickel complexes also showed out
of plane distortions.27 The transꢀIII structure has roughꢀ
ly equal dxz and dyz orbital energies while the transꢀI
structure has a dxz orbital significantly raised in energy
relative to the dyz. The critical angles, bond lengths and
relative energies of these DFT structures can be seen in
Table 4. The relative differences in the CO binding enerꢀ
gy (∆ECO) was modeled by finding the difference in the
total bonding energy (TBE) for geometry optimized
structures of [Ni(cyclam)]+, [Ni(cyclam)(CO)]+ and free
CO and applying eq 6. The absolute energies calculated
most likely carry a large error because solvent, counter
anions and entropy have been ignored. It is the relative
energies between the two isomers that can be of value
from these calculations.
Ni(0)(cyclam)(CO) species would most likely stabilize by
N ligand loss. This hypothesis is supported by DFT optiꢀ
mized structures of [Ni(cyclam)(CO)]0 which show sigꢀ
nificant lengthening of the NiꢀN bonds in the transꢀI
structure and cleavage of 2 of the NiꢀN bonds in the
trans III structure (Figure S4). The Ni(0) carbonyl could
then bind more CO (which stabilizes the Ni(0) oxidation
state) and eventually total cyclam ligand displacement
could be imagined leading to the observed Ni(CO)4 seen
in the IRꢀSEC experiments. All of these results are conꢀ
sistent with the deactivation of [Ni(cyclam)]+ by CO at
negative potentials.
Table 4. Critical angles and bond lengths for DFT
structures shown in Figure 8.
a
NꢀNiꢀN
angles
Avg. NiꢀN
(Å)
NiꢀCO
(Å)
ꢁECO
Isomer
trans I
(kJ/mol)
130° 160°
2.213
2.175
1.808
1.821
ꢀ97
ꢀ67
trans III 146° 147°
a∆ECO = CO binding energy. More negative values imply a
more favorable bonding interaction. Only the relative differꢀ
ence between the two isomers should be considered.
∆ECO = TBE[Ni(cyclam)(CO)]+ – (TBE[Ni(cyclam)]+ + TBECO
)
(6)
The transꢀI isomer has a much more favorable CO bindꢀ
ing energy relative to the transꢀIII which can also be
seen in the shorter NiꢀCO bond length for the transꢀI
isomer. Experimental evidence for this conclusion can
also be found. A study on the kinetics of addition of CO
to [Ni(cyclam)]+ showed a fast addition reaction (kCO
=
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
Ni(II/I) couple in solution.16 Second, the initial catalytic
1
rates are much faster on Hg than glassy carbon (without
added CO scavenger). The current results here would
suggest that the major limitation to higher catalytic curꢀ
rents in the homogenous system is deactivation of the
catalyst by CO. It would stand to reason that perhaps
when adsorbed on Hg, the deactivation of catalysis initiꢀ
ated by CO binding would be attenuated. The reason for
this decline in CO binding could be due to the constraint
of the macrocyclic ligand to bend out of a square planar
geometry. As shown by our DFT results, If the cyclam
ring is allowed to bend along one or both of the NꢀNiꢀN
angles to <180° then the CO adduct will be stabilized. To
further illustrate this stabilization effect, an additional
DFT calculation was done for [Ni(cyclam)(CO)]+ with
constraints on the NꢀNiꢀN angles to remain unchanged
from the [Ni(cyclam)]+ optimized structure (176, 174°
and 179, 179° for transꢀI and transꢀIII respectively). The
transꢀI angleꢀconstrained structure had a ∆ECO that was
47 kJ/mol higher (destabilized) relative to the non angleꢀ
constrained structure in Figure 8. A geometry converged
structure of the transꢀIII [Ni(cyclam)(CO)]+ was not
reached with these angleꢀconstraints instead favoring
breaking of the NiꢀCO bond. This DFT result emphasizes
the necessity for the bending out of square planar geomꢀ
etry to achieve a highly stable CO adduct. If the Hg surꢀ
face prevents such a distortion of the square planar
cyclam ligand then this stabilization of the
[Ni(cyclam)(CO)]+ might be prevented and could lead to
increased catalytic rates relative to what is seen without
Hg. Addition of the CO scavenger in the heterogeneous
system, in effect, helps to minimize the formation of the
[Ni(cyclam)(CO)]+ in the first place.
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
Figure 8. DFT optimized geometries and selected orbitals of
[Ni(cyclam)(CO)]+ of the transꢀI and transꢀIII isomers.
The true identity of the adsorbed species on Hg still reꢀ
mains unclear. Hg could lend electronic effects that enꢀ
courage CO2 binding.36,37 An increased CO2 binding conꢀ
stant for the adsorbed species would lead to higher TOF
because CO2 could better compete with CO for binding to
the Ni(I) metal center. Some reports hypothesize that the
adsorbed species is in the transꢀI geometry and that this
isomer has increased CO2 reduction activity.15 Past comꢀ
putational reports support that transꢀI has a higher afꢀ
finity for CO2 binding in gas phase17 and solvated.38
However, no difference was observed in the rate of reꢀ
ductive adsorption on Hg of a freshly prepared solution
of recrystallized [Ni(cyclam)]2+ (which is all transꢀIII)
and an aged solution (15% transꢀI).16 The authors conꢀ
cluded that the transꢀI isomer is not the species responꢀ
sible for the unusual reductive adsorption of
[Ni(cyclam)]+. Whether or not the isomeric form of
[Ni(cyclam)]+ has an effect on the catalytic reduction of
CO2 has not been thoroughly established experimentally.
However, its importance to the catalytic activity must be
considered. Isomerization can be base catalyzed, operatꢀ
ing with a deprotonation of the amine proton on the
cyclam ring followed by inversion.33 These basic condiꢀ
tions are met by examination of the overall electrocataꢀ
lytic reaction of [Ni(cyclam)]2+ with CO2 in the presence
of water (or other weak acid) (eq 7). Clearly there will be
a local increase in pH near the electrode surface due to
proton consumption and generation of OH– (or other
conjugate base). Alternatively, it has been reported that
reduction of some Ni(II) complexes is accompanied, or
followed by, isomerization processes.23 Therefore, under
Comparison with Catalysis on Mercury
The wealth of data of the CO2 reduction by
[Ni(cyclam)]2+ at a Hg electrode can be used to compare
with the present results at an inert glassy carbon elecꢀ
trode. It has been established that [Ni(cyclam)]+ is abꢀ
sorbed onto the Hg surface and this species differs in
electrocatalytic properties compared to the nonꢀ
adsorbed species. Very large icat/ip values are obtained at
low catalyst concentrations with an Hg electrode.14 Howꢀ
ever, this current quickly drops to near zero as the potenꢀ
tial is scanned more negative due to formation of insoluꢀ
ble Ni(0) carbonyl species which block the electrode
from adsorption of the active [Ni(cyclam)]+ catalytic speꢀ
cies.30 The formation of a [Ni(cyclam)(CO)]+ species durꢀ
ing CO2 reduction has been observed with a Hg elecꢀ
trode.14 However, there are two major differences with
the catalytic behavior of [Ni(cyclam)]+ when adsorbed on
Hg as compared to the homogeneous system described
here. First, the positive shift in potential of the onset of
the catalytic current under CO2 vs. the onset of the Faraꢀ
daic current under N2 is much greater for Hg (≈300
mV)14 than glassy carbon (≈40 mV). This is because the
Hg surface stabilizes the reductively adsorbed complex
allowing the catalytically active Ni(I) redox state to be
achieved at a much more positive potential than the
ACS Paragon Plus Environment

Page 9 of 11
Journal of the American Chemical Society
Supporting Information. Figures S1 – S4 and Synthetic
and DFT details. This material is available free of charge via
the Internet at http://pubs.acs.org.
electrocatalytic conditions, the barrier to isomerization
1
2
3
4
5
6
7
8
might be rather low. More experimental work will be
needed to determine if an isomerization process is imꢀ
portant to the catalytic activity of CO2 reduction – and if
so – is it beneficial or detrimental to the prolonged turnꢀ
over of the catalyst? The DFT results here suggest that an
isomerization is a possible pathway to catalyst deactivaꢀ
tion for formation of a more stable [Ni(cyclam)(CO)]+
species.
AUTHOR INFORMATION
Corresponding Author
* To whom correspondence should be addressed. Eꢀmail:
ckubiak@ucsd.edu.
9
Funding Sources
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
[Ni(cyclam)]2+ + 2e– + CO2 + H2O →
This material is based upon work supported by
the Air Force Office of Scientific Research through
the MURI program under AFOSR Award No. FA9550ꢀ10ꢀ1ꢀ
0572.
[Ni(cyclam)]2+ + CO + 2OH–
(7)
CONCLUSION
[Ni(cyclam)]2+ has proven to be a fast, efficient and seꢀ
lective catalyst for the reduction of CO2 to CO. The speꢀ
cies [Ni(cyclam)(CO)]+ was observed in appreciable
quantities during CO2 reduction by CV and IRꢀSEC. At
more negative potentials, Ni(CO)4 was also observed.
These results lead to the assumption that catalyst deactiꢀ
vation by CO is a major limitation to higher catalytic curꢀ
rents. This assumption was supported by the addition of
another macrocyclic Ni complex, [Ni(TMC)]2+, to act as a
CO scavenger. Addition of [Ni(TMC)]2+ leads to a subꢀ
stantial increase (up to 10 times) in the catalytic current
observed in a CV. Observation of [Ni(TMC)(CO)]+ by IRꢀ
SEC supports the role of the reduced [Ni(TMC)]+ species
as a CO scavenger to allow for [Ni(cyclam)]+ to remain in
its active state. DFT calculations were done on the transꢀ
I and transꢀIII isomers of the [Ni(cyclam)(CO)]+ species.
There is significant outꢀofꢀplane distortion observed in
both isomers but the transꢀI structure favored a geomeꢀ
try with an extreme angle on only one of the NꢀNiꢀN
bonds approaching a more trigonal bipyramidal geomeꢀ
try with the CO ligand in a equatorial position. The
[Ni(cyclam)(CO)]+ species can be further reduced to
Ni(0) carbonyl species which may degrade by cyclam
ligand loss, especially in the presence of excess CO.
ACKNOWLEDGMENT
Thanks to Matt Sampson and all of the Pubsketballers.
Thanks to Prof. Figueroa for assistance with DFT calculaꢀ
tions.
ABBREVIATIONS
ACN, acetonitrile; CV, cyclic voltammogram; DFT, Density
Functional Theory; IRꢀSEC, Infrared Spectroelectrochemisꢀ
try; TBAPF6, tetrabutylammonium hexafluorophosphate;
REFERENCES
(1) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.;
DuBois, D. L.; Dupuis, M.; Ferry, J. G.; Fujita, E.; Hille, R.;
Kenis, P. J. A.; Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.;
Portis, A. R.; Ragsdale, S. W.; Rauchfuss, T. B.; Reek, J. N. H.;
Seefeldt, L. C.; Thauer, R. K.; Waldrop, G. L. Chem. Rev. 2013,
113, 6621.
(2) Qiao, J.; Liu, Y.; Hong, F.; Zhang, J. Chem. Soc. Rev.
2014, 43, 631.
(3) Berardi, S.; Drouet, S.; Francas, L.; GimbertꢀSurinach, C.;
Guttentag, M.; Richmond, C.; Stoll, T.; Llobet, A. Chem. Soc.
Rev. 2014, 43, 7501.
(4) Hori, Y.; Wakebe, H.; Tsukamoto, T.; Koga, O.
Electrochim. Acta 1994, 39, 1833.
(5) Kedzierzawski, P.; Augustynski, J. J. Electrochem. Soc.
1994, 141, L58.
(6) Savéant, J.ꢀM. Chem. Rev. 2008, 108, 2348.
(7) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M.
Chem. Soc. Rev. 2009, 38, 89.
(8) Windle, C. D.; Perutz, R. N. Coord. Chem. Rev. 2012,
256, 2562.
(9) Costentin, C.; Robert, M.; Saveant, J.ꢀM. Chem. Soc. Rev.
2013, 42, 2423.
(10) Bourrez, M.; Molton, F.; ChardonꢀNoblat, S.; Deronzier,
A. Angew. Chem. Int. Ed. 2011, 50, 9903.
(11) Smieja, J. M.; Sampson, M. D.; Grice, K. A.; Benson, E.
E.; Froehlich, J. D.; Kubiak, C. P. Inorg. Chem. 2013, 52, 2484.
(12) Costentin, C.; Drouet, S.; Robert, M.; Savéant, J.ꢀM.
Science 2012, 338, 90.
We choose to interpret the results from this report to
suggest that the higher peak current densities on Hg are
due to suppression of the degradation pathway towards
Ni(0) carbonyl species. Even though Ni(0) carbonyl speꢀ
cies were observed on a Hg electrode and thought to inꢀ
hibit catalysis, the CO2 reduction at a Hg electrode ocꢀ
curs at a much more positive potential than Ni(0) species
can be formed. Therefore very high catalytic currents can
be observed at these more positive potentials. The adꢀ
sorbed species is likely to have increased energy barriers
for loss of planarity of the cyclam ring. The folding and
eventual loss of one or multiple NꢀNi bonds of the
cyclam ligand is a probable degradation path at negative
potentials leading to Ni(0) species with strong affinity
for CO. This selfꢀpoisoning behavior is presumed to be
minimized by interaction of the complex with the Hg
surface. However, the hypothesis that the Hg surface acts
to stabilize a more planar geometry of cyclam complexes
that more readily releases CO remains to be verified exꢀ
perimentally or computationally.
(13) Fisher, B. J.; Eisenberg, R. J. Am. Chem. Soc. 1980, 102,
7361.
(14) Beley, M.; Collin, J. P.; Ruppert, R.; Sauvage, J. P. J. Am.
Chem. Soc. 1986, 108, 7461.
(15) Schneider, J.; Jia, H.; Kobiro, K.; Cabelli, D. E.;
Muckerman, J. T.; Fujita, E. Energy Environ. Sci. 2012, 5,
9502.
(16) Balazs, G. B.; Anson, F. C. J. Electroanal. Chem. 1992,
322, 325.
(17) Froehlich, J. D.; Kubiak, C. P. Inorg. Chem. 2012, 51,
3932.
ASSOCIATED CONTENT
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
(18) Wilhelm, E.; Battino, R.; Wilcock, R. J. Chem. Rev.
1977, 77, 219.
(28) Fujihira, M.; Hirata, Y.; Suga, K. J. Electroanal. Chem.
Interfacial. Electrochem. 1990, 292, 199.
1
2
3
4
5
6
7
8
(19) Fujita, E.; Creutz, C.; Sutin, N.; Szalda, D. J. J. Am.
Chem. Soc. 1991, 113, 343.
(20) Tomita, Y.; Teruya, S.; Koga, O.; Hori, Y. J.
Electrochem. Soc. 2000, 147, 4164.
(21) Machan, C. W.; Sampson, M. D.; Chabolla, S. A.; Dang,
T.; Kubiak, C. P. Organometallics 2014, 33, 4550.
(22) Kelly, C. A.; Mulazzani, Q. G.; Venturi, M.; Blinn, E. L.;
Rodgers, M. A. J. J. Am. Chem. Soc. 1995, 117, 4911.
(23) Zilbermann, I.; Winnik, M.; Sagiv, D.; Rotman, A.;
Cohen, H.; Meyerstein, D. Inorg. Chim. Acta 1995, 240, 503.
(24) Gagne, R. R.; Allison, J. L.; Ingle, D. M. Inorg. Chem.
1979, 18, 2767.
(29) Kelly, C. A.; Mulazzani, Q. G.; Blinn, E. L.; Rodgers, M.
A. J. Inorg. Chem. 1996, 35, 5122.
(30) Balazs, G. B.; Anson, F. C. J. Electroanal. Chem. 1993,
361, 149.
(31) Szalda, D. J.; Fujita, E.; Sanzenbacher, R.; Paulus, H.;
Elias, H. Inorg. Chem. 1994, 33, 5855.
(32) Costentin, C.; Drouet, S.; Robert, M.; Savéant, J.ꢀM. J.
Am. Chem. Soc. 2012, 134, 11235.
(33) Connolly, P. J.; Billo, E. J. Inorg. Chem. 1987, 26, 3224.
(34) Macgregor, S. A.; Lu, Z.; Eisenstein, O.; Crabtree, R. H.
Inorg. Chem. 1994, 33, 3616.
(35) Thom, V. J.; Fox, C. C.; Boeyens, J. C. A.; Hancock, R. D.
J. Am. Chem. Soc. 1984, 106, 5947.
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
(25) Fujita, E.; Haff, J.; Sanzenbacher, R.; Elias, H. Inorg.
Chem. 1994, 33, 4627.
(26) Song, J.; Klein, E. L.; Neese, F.; Ye, S. Inorg. Chem.
2014, 53, 7500.
(27) Furenlid, L. R.; Renner, M. W.; Szalda, D. J.; Fujita, E. J.
Am. Chem. Soc. 1991, 113, 883.
(36) Sakaki, S. J. Am. Chem. Soc. 1990, 112, 7813.
(37) Sakaki, S. J. Am. Chem. Soc. 1992, 114, 2055.
(38) Schneider, J.; Jia, H.; Muckerman, J. T.; Fujita, E.
Chem. Soc. Rev. 2012, 41, 2036.
ACS Paragon Plus Environment

Page 11 of 11
Journal of the American Chemical Society
For Table of Contents Only
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
11
ACS Paragon Plus Environment