Strong Electronic Coupling of Molecular Sites to Graphitic Electrodes via Pyrazine Conjugation

Article abstract of DOI:10.1021/jacs.7b10723

Glassy carbon electrodes were functionalized with redox-active moieties by condensation of o-phenylenediamine derivatives with o-quinone sites native to graphitic carbon surfaces. Electrochemical and spectroscopic investigations establish that these graphite-conjugated catalysts (GCCs) exhibit strong electronic coupling to the electrode, leading to electron transfer (ET) behavior that diverges fundamentally from that of solution-phase or surface-tethered analogues. We find that (1) ET is not observed between the electrode and a redox-active GCC moiety regardless of applied potential. (2) ET is observed at GCCs only if the interfacial reaction is ion-coupled. (3) Even when ET is observed, the oxidation state of a transition metal GCC site remains unchanged. From these observations, we construct a mechanistic model for GCC sites in which ET behavior is identical to that of catalytically active metal surfaces rather than to that of molecules in solution. These results suggest that GCCs provide a versatile platform for bridging molecular and heterogeneous electrocatalysis.

Full text of DOI:10.1021/jacs.7b10723

Subscriber access provided by READING UNIV
Article
Strong electronic coupling of molecular sites to
graphitic electrodes via pyrazine conjugation
Megan N. Jackson, Seokjoon Oh, Corey J Kaminsky, Sterling
B. Chu, Guanghui Zhang, Jeff Miller, and Yogesh Surendranath
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 07 Dec 2017
Downloaded from http://pubs.acs.org on December 7, 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.
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 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Strong electronic coupling of molecular sites to graphitic
electrodes via pyrazine conjugation
Megan N. Jackson†, Seokjoon Oh†, Corey J. Kaminsky†, Sterling B. Chu†, Guanghui Zhang‡, Jeffrey T.
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
Miller‡§, and Yogesh Surendranath†
† Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
‡ Davidson School of Chemical Engineering, Purdue University, West Lafayette, Indiana 47907, United States
§ Chemical Science and Engineering Division, Argonne National Laboratory, Argonne, Illinois 60439, United States
KEYWORDS. Electron transfer, outer-sphere, inner-sphere, surface immobilization, heterogeneous catalysis, tether, electro-
chemistry,
cyclic
voltammetry,
X-ray
absorption
spectroscopy
ABSTRACT: Glassy carbon electrodes were functionalized with redoxꢀactive moieties by condensation of oꢀphenylenediamine
derivatives with oꢀquinone sites native to graphitic carbon surfaces. Electrochemical and spectroscopic investigations establish that
these graphiteꢀconjugated catalysts (GCCs) exhibit strong electronic coupling to the electrode, leading to electron transfer (ET)
behavior that diverges fundamentally from that of solution phase or surfaceꢀtethered analogues. We find that: (1) ET is not obꢀ
served between the electrode and a redoxꢀactive GCC moiety regardless of applied potential. (2) ET is observed at GCCs only if the
interfacial reaction is ion-coupled. (3) Even when ET is observed, the oxidation state of a transition metal GCC site remains unꢀ
changed. From these observations, we construct a mechanistic model for GCC sites in which ET behavior is identical to that of
catalytically active metal surfaces rather than to that of molecules in solution. These results suggest that GCCs provide a versatile
platform
for
bridging
molecular
and
heterogeneous
electrocatalysis.
Introduction
chemistry,6,7 alkynyl linkages,8 and noncovalent linkages that
rely on πꢀπ interactions.9 However, these linkages provide
relatively poor electron coupling between the appended units
and the metallic electrode surface. For example, detailed studꢀ
ies on ferroceneꢀappended thiol SAMs display ET kinetics,
including solvent reorganization energies and electron tunꢀ
nelinfig preꢀfactors, that are in line with outerꢀsphere ET as
described by Marcus theory, suggesting that there is a tunnelꢀ
ing barrier to ET in these systems.3 Thus the electrochemical
and catalytic behavior of the tethered molecule closely resemꢀ
bles that of the dissolved species, not a metallic heterogeneous
surface site.
The efficient interconversion of electrical and chemical enꢀ
ergy requires catalysts capable of accelerating complex multiꢀ
electron reactions at electrified interfaces. These reactions can
be carried out at the metallic surface sites of heterogeneous
electrocatalysts or via redox mediation at molecular electroꢀ
catalysts. Molecular catalysts are straightforward to tune synꢀ
thetically and characterize spectroscopically, allowing for
unparalleled insight into their mechanisms of action. Similar
molecularꢀlevel insight into metallic heterogeneous catalysts
would be valuable not only on a fundamental level, but also on
a practical level since these catalysts are commonplace in
nearly all contemporary energy conversion devices, including
fuel cells and electrolyzers. However, it is difficult, if not imꢀ
possible, to obtain this level of understanding for most heteroꢀ
geneous catalysts because surface active sites are inherently
dynamic, difficult to modify at the molecular level, and hard to
identify, much less characterize. Thus, advances in electrocaꢀ
talysis require new strategies for controlling the surface reacꢀ
tivity of complex metallic heterogeneous interfaces at the moꢀ
lecular level.
We have developed a simple method for linking molecules
to graphitic carbon electrodes through conjugated aromatic
pyrazine linkages. We have shown that oꢀphenylenediamine
derivatives condense irreversibly with oꢀquinone moieties
found natively at edge planes and stepꢀedge defects of graphitꢀ
ic carbons to generate robust conjugated pyrazine linkages
between the appended molecule and the surface.10,11 The reꢀ
sulting graphite conjugated catalysts (GCCs) are active for
oxygen reduction in alkaline aqueous media.10 Additionally,
GCCs bearing a Re coordination compound are active cataꢀ
lysts for the selective reduction of CO2 to CO.11
Efforts toward molecular control of interfacial reactivity
have largely centered on immobilizing molecular redox sites
onto electrodes. A variety of methods have been developed for
attaching molecules to conductive surfaces, including thiolꢀ
based selfꢀassembled monolayers,1–3 diazonium grafts,4,5 click
Herein, we show that these conjugated linkages lead to
strong electronic coupling of the GCC sites to the graphitic
electrode that dramatically alters their electrochemical behavꢀ
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 9
ior relative to solutionꢀphase or surfaceꢀtethered molecular
N 1s XPS on GCCꢀRh is best fit to two peaks in a 1:1 ratio at
399.9 and 399.0 eV (Figure S4), again consistent with the
expected binding energies for metalꢀbound pyridinic N and
pyrazinic N. Together, these data suggest that GCCꢀRu and
1
2
3
4
5
6
7
8
analogues. Electrochemical and Xꢀray absorption spectroscopy
reveal that: (1) Classical outerꢀsphere ET is not observed beꢀ
tween the electrode and the GCC sites. (2) ET is observed at
GCCs if the interfacial reaction is ion-coupled. (3) Even when
ET is observed, the oxidation state of a transition metal GCC
site remains unchanged. From these observations, we infer that
GCC sites do not behave like their molecular analogues, but
rather as metallic active sites with molecular definition. Conꢀ
sequently, GCCs form a unique class of materials that comꢀ
bines the high structural fidelity and tunability of molecular
catalysts with the electronic properties of metallic extended
solids.
GCCꢀRh
surfaces contain
a
uniform array of
[RuII(dmbpy)2(phen)]2+ (phen
= 1,10ꢀphenanthroline) and
[RhIIICp*(phen)Cl]+ moieties conjugated to glassy carbon
electrodes through pyrazine linkages.
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
Results
Scheme 1. Synthesis of GCC-phenazine (top), GCC-Ru
(middle), and GCC-Rh (bottom).
Synthesis and characterization of GCCs
Chemically modified electrodes were prepared via treatment
of carbon surfaces with phenylenediamineꢀcontaining precurꢀ
sors (Scheme 1) using procedures described previously (full
synthetic details of electrode preparation are provided in the
Supporting Information).10 Following surface treatment, modiꢀ
fied electrodes were washed with ethanol and water, and
soaked in 0.1 M HClO4 to remove monoamineꢀlinked and
physisorbed species. Xꢀray photoelectron spectroscopy (XPS)
and N Kꢀedge Xꢀray absorption near edge structure (XANES)
spectroscopy establish that this procedure selectively generates
pyrazine linkages on graphitic carbon surfaces.10 Whereas
GCCꢀphenazine electrodes were prepared by treating glassy
carbon with oꢀphenylenediamine, GCCꢀRu and GCCꢀRh elecꢀ
trodes were prepared by treating glassy carbon with
Figure 1. (a) Cyclic voltammogram (10 mV s−1) of GCC phenaꢀ
zine recorded in 0.1 M NaOH. (b) E1/2 vs pH for dissolved phenaꢀ
zine (red) and GCCꢀphenazine (black). Values for dissolved
phenazine are reꢀplotted from literature values obtained from
linear sweep voltammograms recorded in 10% ethanolic aqueous
electrolyte.13 CVs of GCC phenazine were recorded at 10 mV s−1
in pHꢀadjusted 0.1 M borate/0.1 M phosphate/0.1 M formate
aqueous electrolyte.
Electrochemistry of GCC-phenazine
[RuII(dmbpy)2(phenda)]2+
bipyridine, phenda = 5,6ꢀdiaminoꢀ1,10ꢀphenanthroline) and
[RhIIICp*(phenda)Cl]+ (Cp* = pentamethylcyclopentadienyl),
respectively.
(dmbpy
= 4,4′ꢀdimethylꢀ2,2′ꢀ
XPS data establish the fidelity of the primary coordination
environments about the Ru and Rh centers in GCCꢀRu and
GCCꢀRh, respectively. Survey XPS scans of GCCꢀRu reveal a
Ru:N ratio of approximately 1:7 (Figure S1), and survey scans
of GCCꢀRh reveal a Rh:N:Cl ratio of 1:4:1 (Figure S2), conꢀ
sistent with the 1:8 Ru:N and 1:4:1 Rh:N:Cl ratios expected of
the conjugated species. Highꢀresolution N 1s XPS of GCCꢀRu
is best fit to two peaks at 399.6 and 398.2 eV in a 3.5:1 ratio
(Figure S3), in line with values expected for metalꢀbound
pyridinic N and pyrazinic N, respectively.10,12 Highꢀresolution
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
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
Figure 3. CVs of dissolved phenazine (5 mM, 100 mV s−1) (top)
and GCCꢀphenazine (10 mV s−1) (bottom) recorded in acetoniꢀ
trile electrolyte containing 0.1 M TBAPF6.
Figure 2. CVs of dissolved phenazine (5 mM, 100 mV s−1)
(top) and GCCꢀphenazine (10 mV s−1) (bottom) recorded in
acetonitrile electrolyte containing 0.1 M TBAPF6 and 0.2 M
tosylic acid.
While phenazine and GCCꢀphenazine display similar redox
behavior in protic electrolytes, they display radically different
electrochemical behavior in the absence of a proton donor. In
acetonitrile electrolyte containing 0.1 M TBAPF6 and no acid,
dissolved phenazine displays a reversible oneꢀelectron redox
wave at E1/2 = −1.61 V vs Fc+/0 (Figure 3, top); however,
GCCꢀphenazine electrodes display no discernible redox feaꢀ
tures beyond the background doubleꢀlayer charging current
(Figure 3, bottom). Importantly, redox features are not obꢀ
served even upon scanning to the solvent window (Figure S7),
increasing the electrolyte strength from 0.1 M to 0.5 M
TBAPF6 (Figure S8), exchanging Li+ for TBA+ (Figure S9),
exchanging Cl− for PF6− (Figure S10), or varying the scan rate
between 100 mV s−1 and 10 mV s−1 (Figure S11). In short,
unlike its molecular analogue, GCCꢀphenazine electrodes only
give rise to discernable redox waves in protic electrolytes.
In protic electrolytes, the electrochemical response arising
from GCCꢀphenazine is similar to that of molecular phenaꢀ
zine. In aqueous electrolyte, phenazine undergoes a twoꢀ
proton, twoꢀelectron reduction to dihydrophenazine at E1/2
=
−0.23 V vs RHE.13 Similarly, at pH 13, cyclic voltammograms
(CVs) of GCCꢀphenazine reveal a broad reversible wave at
E1/2 = 0.12 V vs RHE (−0.65 V vs NHE, Figure 1a). Despite
the relatively slow scan rate, this surface wave displays a large
peakꢀtoꢀpeak separation of 220 mV at 10 mV s−1, indicative of
a slow charge transfer process on the surface. The E1/2 of these
redox waves shift in a roughly Nernstian fashion by 58 and 56
mV per pH unit for phenazine and GCCꢀphenazine, respecꢀ
tively (Figures 1b and S5),13 suggesting that both processes
are protonꢀcoupled reactions involving an equal number of
protons and electrons. Similarly, in acetonitrile electrolyte
containing 0.1 M tetrabutylammonium hexafluorophosphate
(TBAPF6) and 0.2 M tosylic acid, CVs of phenazine display
two reversible redox waves at E1/2 = −0.29 V and 0.10 V vs
Fc+/0, attributed to two sequential oneꢀproton, oneꢀelectron
transfer processes (Figure 2, top),14 and CVs of GCCꢀ
phenazine display two analogous, albeit broad, redox features
at −0.19 V and 0.14 V (Figure 2, bottom). This broadness is
not due to significant degradation of the surface linkage. CVs
recorded in water following cycling in acetonitrile with tosylic
acid reveal wellꢀdefined waves of similar magnitude (Figure
S6). Instead, we attribute this broadness to the known slugꢀ
gishness of interfacial PCET in acetonitrile versus water.15,16
Together, these data indicate that in protic electrolytes, both
surfaceꢀbound GCCꢀphenazine units and dissolved molecular
phenazines undergo chemically reversible protonꢀcoupled ET
reactions.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
Figure 4. CVs of dissolved [RuII(dmbpy)2(phen)]2+ (5mM, 100
1
2
mV s−1) (top) and GCCꢀRu (10 mV s−1) (bottom)recorded in
acetonitrile electrolyte containing 0.1 M TBAPF6.
3
4
5
6
Electrochemistry of GCC-Ru
We now analyze the electrochemical behavior of metalꢀ
containing GCCs in comparison to soluble analogs. In soluꢀ
tion, [RuII(dmbpy)2(phen)]2+ undergoes a chemically reversiꢀ
ble, outerꢀsphere oneꢀelectron transfer at E1/2 = 0.79 V vs Fc+/0
,
7
which we assign to the RuIII/II couple (Figure 4, top). This
metalꢀbased outerꢀsphere ET is not observed for GCCꢀRu
electrodes. Instead, only background double layer charging
current is observed in the potential region over which the disꢀ
solved molecule is oxidized with the onset of the solvent winꢀ
dow at ~0.9 V (Figure 4, bottom). While this solvent window
rises earlier than that of the unfunctionalized electrode, the
magnitude of the current is strongly dependent on the trace
water concentration in the acetonitrile electrolyte and decreasꢀ
es with increasing scan rate. Together, these observations lead
us to postulate that this current arises from slow reaction with
adventitious water and is unrelated to outer sphere Ru oxidaꢀ
tion. As above, we do not observe reversible redox features for
GCCꢀRu upon increasing the electrolyte strength from 0.1 M
TBAPF6 to 0.5 M TBAPF6 (Figure S12), exchanging TBA+
for Li+ (Figure S13), or varying the scan rate between 100 mV
s−1 and 10 mV s−1 (Figure S14). CVs of GCCꢀRu recorded in
0.1 M HClO4 after polarization in acetonitrile establish that the
phenazine moiety remains intact (Figure S15). The amount of
Ru remaining on the surface after polarization was determined
by inductively coupled plasma mass spectrometry (ICPꢀMS)
following acid digestion of the surface (see Supporting Inforꢀ
mation for details). Noting the twoꢀelectron stoichiometry of
phenazine reduction, we find a Ru:phenazine surface concenꢀ
tration ratio of 0.9, indicating that the Ru surface sites remain
intact on the surface over the course of the measurement. Fig-
ure S15 compares the magnitude of the phenazine wave to the
CV trace in Figure 4 (bottom), indicating that, at these loadꢀ
ings, a surface RuIII/II redox wave, if it existed, would be easily
discernable over background double layer charging current.
Together, these data suggest that GCCꢀRu sites, unlike soluble
[RuII(dmbpy)2(phen)]2+, do not give rise to discrete surface
charge transfer waves.
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 5. CVs of dissolved [RhIIICp*(phen)Cl]+ (5 mM, 100 mV
s−1) (a, top) and GCCꢀRh (10 mV s−1) (a, bottom) recorded in
acetonitrile electrolyte containing 0.1 M TBACl. (b) in situ Rh Kꢀ
edge Xꢀray absorption near edge structure spectra of GCCꢀRh
recorded in acetonitrile electrolyte containing 0.1 M TBACl and
held at the open circuit potential (OCP, ~ −0.3 V), −1.43 V, and
−1.55 V vs Fc+/0
.
Electrochemistry and X-ray absorption spectroscopy of
GCC-Rh
In order to examine the behavior of metalꢀcontaining GCCs
that are able to undergo redoxꢀinduced ligand exchange, we
compared the redox chemistry of GCCꢀRh with its soluble
analog [RhIIICp*(phen)Cl]+. At −1.19
V
vs Fc+/0
,
[RhIIICp*(phen)Cl]+ undergoes twoꢀelectron reduction from
RhIII to RhI with simultaneous dissociation of the bound Cl−
ligand (Figure 5a, top).17,18 The analogous GCCꢀRh electrode
displays a reversible redox wave at −1.29 V in 0.1 M TBACl
in acetonitrile (Figure 5a, bottom). The electron stoichiometry
of this wave was determined by integrating the charge in CVs
of GCCꢀRh in 0.1 M TBACl and comparing it to the amount
of Rh on the surface as measured by acid digestion of the surꢀ
face and ICPꢀMS quantification. Using this procedure, we
measured an electron:Rh ratio of 1.01 (95% CI [0.74:1.28],
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
see Supporting Information for details). In TBABr and TBAI,
Mechanistic Model
1
the wave shifts by +60 mV and +40 mV (Figure S16), respecꢀ
tively, indicating the wave is sensitive to the identity of the
halide. We assign the cathodic wave to dissociation of an inꢀ
nerꢀsphere halide and the anodic wave to reassociation of the
majority halide in the electrolyte. This assignment is corroboꢀ
rated by XPS data that show a complete absence of Cl from
the surface upon polarizing a GCCꢀRh electrode in the presꢀ
ence of TBAPF6 (Figure S17). Furthermore, the reversible
wave observed in TBAX (X = Cl−, Br−, or I−) becomes chemiꢀ
cally irreversible upon introduction of CO to the solution
(Figure S18). Upon subsequent cycling of the electrode in
CO, the wave disappears altogether. XPS spectra taken after
polarizing in COꢀsaturated solution at −1.31 V vs Fc+/0 show a
N:Rh ratio of 4.6:1, suggesting that the Rh species remains inꢀ
tact in the presence of CO (Figure S19). We attribute the disꢀ
appearance of the CV feature in this medium to the irreversiꢀ
ble binding of CO to the Rh site upon halide dissociation,
which prevents the Rh from reassociating Cl−. This attribution
is consistent with the absence of Cl in the XPS, even after
scanning back to −0.86 V vs Fc+/0 (+0.34 V vs the GCCꢀRh
wave). Together, these data suggest that metalꢀbased GCC
sites can give rise to redox waves, provided that they are able
to undergo ion exchange with the solution.
The electrochemical and XAS studies described above esꢀ
tablish that the redox chemistry of GCCs is radically different
than that of dissolved molecular analogues. In particular, we
observe that (1) GCCs display discrete redox features only
when electron flow is coupled to ionꢀtransfer at the interface,
in analogy to metal surface sites, and (2) even during interfaꢀ
cial ionꢀcoupled electron transfer, the oxidation state of the
metal center undergoing ion exchange remains unchanged.
These surprising observations form the basis for a mechanistic
model in which molecules that are conjugated to graphite are
part of the electrode rather than merely appended to it.
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
We first note that the absence of a clear redox feature for
GCCꢀRu and GCCꢀphenazine in aprotic electrolyte could reꢀ
sult from the actual absence of charge transfer or from an exꢀ
treme level of broadening of the surface wave so as to make it
indistinguishable from the background doubleꢀlayer charging
current. Broadening could occur for several reasons, which we
exclude in turn. First, interactions between neighboring surꢀ
face sites have been shown to lead to wave broadening for a
variety of chemicallyꢀmodified electrodes;22 however, these
effects are typically observed when there are strong interacꢀ
tions between the appended redox moieties. Although we canꢀ
not entirely exclude the presence of lateral interactions beꢀ
tween GCC surface sites, the low coverage of phenazine
groups on GCCs (~0.25 nmol cm−2)10 is comparable to that
observed for dilute ferrocenyl mixed selfꢀassembled monolayꢀ
ers which display unbroadened surface redox features.22 Thus,
it is unlikely that the waves would broaden so significantly as
to make them unobservable. Second, CV waves can appear
broad when there is insufficient electrostatic screening. We
exclude this explanation because CVs collected at 100 mM
TBAPF6 and 500 mM TBAPF6 were identical, (Figures S8
and S12) and because no redox features were observed when
TBA+ was replaced with Li+ (Figures S9 and S13) or when
PF6− was replaced with Cl− (Figure S10). Finally, CV waves
can broaden due to kinetic sluggishness. We do not believe
that the absence of redox waves for GCCꢀphenazine and GCCꢀ
Ru is due to kinetic sluggishness because our observations are
independent of scan rate; despite varying the scan rate beꢀ
tween 100 mV s−1 and 10 mV s−1, we never observed redox
features (Figures S11 and S14). Additionally, both phenazine
and [RuII(dmbpy)2(phen)]2+ exhibit rapid charge transfer kinetꢀ
ics; the reported outerꢀsphere electrochemical ET rate conꢀ
stants for phenazine and [RuII(phen)3]2+ in acetonitrile are 2.0
× 10−2 cm s−1 and 1.7 cm s−1, respectively.14,23 Indeed, we exꢀ
pect the conjugated linkage between the GCC site and the
electrode to accelerate ET rates relative to electron tunneling
through solvent. Most convincingly, the cases where we do
observe GCC redox waves all involve ion transfer, and this
additional nuclear motion would, if anything, slow the rate of
the reaction relative to a simple outerꢀsphere ET. We conclude
that the absence of redox features in GCCꢀphenazine and
GCCꢀRu in aprotic acetonitrile electrolyte is most likely beꢀ
cause no ET is occurring.
Notably, the GCCꢀRh wave corresponds to a one-electron
process, in contrast to the known twoꢀelectron process for the
soluble Rh molecule. This discrepancy implies that the surface
redox wave observed for GCCꢀRh is distinct from the redox
activity of the molecular species. To directly probe the valency
of the GCCꢀRh sites during the observed redox process, we
monitored the oxidation state of Rh as a function of applied
potential via in situ Xꢀray absorption nearꢀedge structure
(XANES) spectroscopy. Based on literature precedent, the
reduction of RhIII to RhI is expected to shift the rising portion
of the Kꢀedge to lower energy by 2ꢀ3 eV.19,20 Indeed, we obꢀ
serve a significant difference in both the position and shape of
the Kꢀedge between [RhIIICp*(phen)Cl]+ and RhICp*(phen)
(Figure S20). The Rh Kꢀedge for [RhIIICp*(phen)Cl]+ has an
inflection point at 23,226.5 eV, while the Rh Kꢀedge for
RhICp*(phen) onsets earlier and has two inflection points at
23,225.6 and 23,231.7 eV (Figure S21). The different edge
shapes are attributed to a difference in coordination number
and geometry between the two compounds. We expect similar
changes for complexes tethered to the surface through a nonꢀ
conjugated insulating linker. While we were unable to collect
in situ XANES data on the analogous aliphatically tethered
[RhIIICp*(phen)Cl]+ complex due to the instability of the linkꢀ
age under reductive polarization, in situ XANES data collectꢀ
ed on electrodes modified with a noncovalently tethered
[RuII(dmbpy)2(bpy)]2+ (bpy = 2,2’ꢀbipyridine) complex21 reꢀ
veal a 0.7 eV shift in the Ru Kꢀedge upon polarization at 1.1 V
vs Fc+/0 (0.3 V beyond the RuIII/II wave) (Figure S22), exactly
in line with the 0.7 eV shift observed between RuII and RuIII
model complexes (Figure S23). Remarkably, in situ XANES
data collected on GCCꢀRh samples at the open circuit potenꢀ
tial (~ −0.3 V vs Fc+/0), 0.12 V negative of the GCCꢀRh redox
wave (−1.43 V vs Fc+/0), and 0.24 V negative of the redox
wave (−1.55 V vs Fc+/0) are all identical (Rh Kꢀedge of
23,229.0 eV) (Figure 5b), indicating that even though electriꢀ
cal polarization gives rise to current flow and halide dissociaꢀ
tion, it does not lead to a detectable change in the oxidation
state of Rh.
It is possible that ET is not observed in these cases because
the redox potential of the surface site has been shifted outside
the solvent window and is no longer accessible. Indeed, the
twoꢀelectron, twoꢀproton redox feature observed for GCCꢀ
phenazine in aqueous electrolyte occurs ~200 mV positive of
the reduction of phenazine in the same medium, and arguꢀ
ments based on electronꢀdelocalization would lead us to preꢀ
ACS Paragon Plus Environment

Journal of the American Chemical Society
dict a positive shift in nonaqueous electrolytes as well. Howꢀ
Page 6 of 9
ET driving force for this conjugated system. This lack of drivꢀ
ing force is consistent with doubleꢀlayer theory if and only if
the molecular fragment is electronically coupled to the elecꢀ
trode. In metallic electrodes, the metal atoms are strongly couꢀ
pled to the band states of the solid, and the effective conducꢀ
tivity of a metal ensures that the Fermi level of the surfaceꢀ
exposed metal atoms is identical to that of the bulk. Upon apꢀ
plication of a potential, the ensuing interfacial field gradient
raises the Fermi level of the solid and the energy levels of
surface atoms in synchrony; thus, the energy separation beꢀ
tween the donor/acceptor states of the surface atoms and the
Fermi level remain the same irrespective of the applied potenꢀ
tial.26 Our data indicate that the same phenomenon is occurring
in GCCs. In particular, the inability to observe outerꢀsphere
ET in GCCs indicates that the pyrazine linkage electronically
couples the molecular fragment to the electrode. This electronꢀ
ic coupling ensures that the separation between the Fermi level
of the carbon and the molecular donor/acceptor states of the
appended fragments remain largely invariant with applied
potential; i.e., varying the applied potential does not change
the difference in the potential between the electrode and the
GCC site, thereby eliminating the possibility of outerꢀsphere
ET between them (Figure 6, bottom).
1
2
3
4
5
6
7
8
ever, while phenazine in acetonitrile electrolyte displays a
redox feature at −1.61 V vs Fc+/0, we do not observe a redox
wave for GCCꢀphenazine between −2.5 and 1.0 V vs Fc+/0
(Figure S7). We believe that a shift in the E1/2 of phenazine of
> 2 V is unreasonable. Similarly, we would not expect a subꢀ
stantial
change
in
the
RuIII/II
potential
for
[RuII(dmbpy)2(phen)]2+ upon conjugation. Replacing a dipyriꢀ
do[3,2ꢀa:2′,3′ꢀc]phenazine ligand for 1,10ꢀphenanthroline in
pseudooctahedral RuII complexes has been shown to shift the
RuIII/II potential to positive values by less than 0.03 V.24,25 For
GCCꢀRu, we do not observe a redox wave despite scanning >
0.35 V positive of the expected value, too large of a shift to be
explained by changes in the π acidity of the surfaceꢀ
conjugated phenanthroline ligand. It follows that if no reversiꢀ
ble redox wave is observed even 0.35 V past the expected E°,
the variation in applied potential is not imposing a substantial
driving force for ET to the Ru center.
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. Putative interfacial free energy diagrams for unmodified
electrodes with dissolved RuII molecules (top) and electrodes
modified with conjugated RuII surface sites (bottom). The diagram
denotes the Fermi level of the electrode, EF, and the redox potenꢀ
tial of the molecule, E(RuIII/II), upon varying the applied potential
(left to right). The electrostatic potential across the electrochemiꢀ
cal double layer is indicated by the red dotted line. Varying EF
does not impact E(RuIII/II) for the dissolved molecule, leading to
classical outerꢀsphere ET (top) at the interface. For GCCs (botꢀ
tom), varying EF simultaneously shifts the energy levels of doꢀ
nor/acceptor states in the conjugated molecule by a similar magniꢀ
tude, and the driving force for ET remains unchanged.
Figure 7. Putative interfacial free energy diagrams for unmodified
electrodes with dissolved RhIII–Cl molecules (top) and electrodes
modified with conjugated RhIII–Cl surface sites (bottom). The
diagram denotes the Fermi level of the electrode, EF, upon varyꢀ
ing the applied potential (left to right) as well as the redox potenꢀ
tial of the molecule, E(RhIII/I), and the potential for halide dissociꢀ
ation from the surface E(RhIII–Cl /RhIII + Cl−). The electrostatic
potential across the electrochemical double layer is indicated by
the red dotted line. Varying EF drives twoꢀelectron transfer to Rh
via interfacial outerꢀsphere ET (top). For GCCs (bottom), varying
EF shifts the electrostatic potential of the surface, driving halide
dissociation, while simultaneously shifting the energy levels of
the Rh acceptor states, preventing Rhꢀcentered reduction.
We attribute this unique behavior of GCCs to the nature of
the pyrazine linkage. In the absence of a conjugated linkage,
the driving force for outer sphere ET is given by the difference
between the Fermi level of the electrode and the E0 of the moꢀ
lecular species. As the potential is varied, the electrode is
charged, causing a buildup of opposing charge near the elecꢀ
trode surface at the outer Helmholtz plane (OHP). This charge
buildup at the interface establishes an electric field gradient,
changing the Fermi level of the electrode relative to the enerꢀ
gies of the molecular donor/acceptor states residing at or beꢀ
yond the OHP (Figure 6, top). Thus, as the potential changes,
the driving force for ET also changes.
Even though our data suggest that there is negligible potenꢀ
tial drop between the electrode and the GCC site, there is a
potential drop between the GCC site and the solution. Thereꢀ
fore, varying the applied potential can alter the driving force
for ions to transfer between the GCC site and the solution,
provided that the surface site is able to bind the ion in quesꢀ
tion. Our data suggest that this is occurring for GCCꢀ
phenazine in protic media and GCCꢀRh in halide electrolytes.
The inability to do outerꢀsphere ET to GCC sites suggests
that varying the potential does not lead to the same change in
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
In protic electrolyte, the N sites on the phenazine unit of GCCꢀ
potential at a GCCꢀRh electrode simultaneously shifts the enꢀ
ergy of the Rh orbitals. In both cases, the metal site remains in
electrostatic equilibrium with the graphite electrode regardless
of the applied potential, and metalꢀcentered oxidation or reꢀ
duction does not occur. Consequently, ionꢀtransfer to GCC
sites occurs with coincident ET to carbonꢀbased orbitals.
1
2
3
4
5
6
7
8
phenazine can form and break bonds with protons. When the
applied potential is sufficiently negative, the electric field
drives protons to cross the double layer and bind to the N sites.
Since protons carry a positive charge, compensating electrons
must flow from the external circuit in order to maintain the
electrode potential.26 This compensatory current is observed as
a surface redox wave (Figures 1a and 2, bottom). Since all N
sites are expected to bind protons with similar affinity, ET
occurs over a relatively narrow potential range, leading to a
distinct wave above the background doubleꢀlayer charging
current.
Similarly, the Rh in GCCꢀRh binds a Cl− ion, and when the
potential is sufficiently negative, the electric field drives this
negatively charged ion across the double layer into solution.
Again, this ion dissociation leads to compensatory electron
flow from the external circuit, which we observe as a redox
wave (Figure 5a, bottom). A model for this behavior is deꢀ
picted in Figure 7, bottom, in which the E0 at which Cl− dissoꢀ
ciates is denoted as E(RhIIICl ꢀ RhIII + Cl) and does not vary
with the potential applied at the electrode. Importantly, in this
model, only the exchange of specifically adsorbed ions gives
rise to redox features, consistent with our observation of a 1 e−
stoichiometry in the wave. This model is also consistent with
the observed chemical irreversibility of the GCCꢀRh wave in
the presence of CO. We postulate that the strong binding affinꢀ
ity of CO with Rh leads to the irreversibility by preventing the
reꢀassociation of the halide.
The reactivity patterns of GCCs mimic those of metallic
surface sites. At metal surface sites, ET proceeds exclusively
via innerꢀsphere mechanisms in which electron flow is driven
by ions or molecules crossing the double layer. As a result, ET
is only observed when bonds are made or broken at metallic
surface sites, and no ET occurs in the absence of a species that
is able to adsorb to the surface. Our data suggest that ET at
graphiteꢀconjugated molecules also proceeds exclusively via
innerꢀsphere pathways, making GCCs mechanistically indisꢀ
tinguishable from authentic metallic surfaces.
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
Conclusion
The mechanistic studies described above suggest that GCCs
are a unique class of materials that combines the high structurꢀ
al fidelity and tunability of molecular catalysts with the elecꢀ
tronic properties of metallic extended solids. We attribute this
unique behavior of GCCs to the conductive aromatic linkage
that provides for sufficiently strong electronic coupling such
that there is negligible potential drop between the electrode
and the GCC site. Furthermore, we show that the oxidation
state of GCC sites can remain constant throughout a redox
event. Thus, the valency and reduction potential of the molecꢀ
ular analogue do not directly dictate the rate and driving force
for elementary and catalytic redox reactions at GCCs. Instead,
redox chemistry at GCCs is driven by the binding strength of
substrates and intermediates with the surface sites. These
unique characteristics of GCCs make their ET behavior indisꢀ
tinguishable from that of surface sites at metallic heterogeneꢀ
ous catalysts.
Perhaps even more surprisingly, the in situ XANES data
(Figure 5b) indicate that the oxidation state of the Rh remains
constant throughout the redox event for GCCꢀRh, suggesting
that a change in site valency is not necessary for charge transꢀ
fer from the external circuit. This observation is particularly
interesting because if the electron is not going to the metal, it
raises the question, “Where is the electron going?” The anꢀ
swer, of course, is the ligand, and for a GCC, the ligand is the
carbon electrode itself. Indeed, the metallic band structure of
graphitic carbon makes it redox nonꢀinnocent in the extreme.
We can therefore write the following equilibrium to describe
the charge transfer wave observed for GCCꢀRh (Figure 8):
Due to their metalꢀlike behavior, GCCs provide a versatile
platform with which to probe heterogeneous electrochemical
reactivity at the molecular level. Spectroscopic studies of acꢀ
tive sites on bulk or nanoscopic metal electrocatalysts are difꢀ
ficult if not impossible in many cases because the signal from
the active sites is dwarfed by a substantial bulk background.
Moreover, even surfaceꢀsensitive techniques provide signals
that reflect the ensemble average behavior of both active and
spectator surface sites. As exemplified by the in situ XANES
experiments reported here, GCCs overcome this limitation,
enabling molecularꢀlevel investigations into the thermochemꢀ
istry and kinetics of both elementary and catalytic electroꢀ
chemical reactions that form the basis for important energy
conversion technologies.
Figure 8. Proposed balanced reaction for the redox wave obꢀ
served at −1.29 V vs Fc+/0. The red “e−” represents an electron
residing in a graphiteꢀcentered band state of the solid and is drawn
in this particular ring symbolically.
in which the applied potential drives dissociation of the innerꢀ
sphere halide with electron flow to a graphite ligandꢀcentered
band state of the solid. Although we do not have a direct specꢀ
troscopic probe of the coordination sphere of the product, we
postulate that acetonitrile binds to the Rh upon halide dissociaꢀ
tion to complete its coordination sphere. Nonetheless, the reacꢀ
tion described in Figure 8 is consistent with the invariance of
the oxidation state at the Rh center upon polarization beyond
the redox wave.
ASSOCIATED CONTENT
Supporting Information. Full experimental details, synthetic
details, ICPꢀMS data, XPS data, and additional CVs and XAS
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
The behavior of the Rh sites in GCCꢀRh is consistent with
the model for GCCꢀRu (Figure 6), and is captured in Figure
7. Just as varying the potential at a GCCꢀRu electrode simulꢀ
taneously shifts the energies of the Ru orbitals, varying the
* yogi@mit.edu
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
(10)
Fukushima, T.; Drisdell, W.; Yano, J.;
Surendranath, Y. J. Am. Chem. Soc. 2015, 137,
10926.
ACKNOWLEDGMENTS
1
2
3
4
5
6
7
8
We gratefully acknowledge A. Jeremy Kropf and Joshua Wright
for assistance with the XAS experiments. We acknowledge Chrisꢀ
topher E. D. Chidsey and Richard L. McCreery for helpful discusꢀ
sions. This research was supported by the U.S. Department of
Energy, Office of Science, Office of Basic Energy Sciences, unꢀ
der award number DEꢀSC0014176. M.N.J. is supported by the
Department of Defense (DoD) through the National Defense Sciꢀ
ence & Engineering Graduate Fellowship (NDSEG) Program.
S.O. and C.J.K. were supported by the National Science Foundaꢀ
tion Graduate Research Fellowship under Grant No. 1122374.
Use of the Advanced Photon Source is supported by the U.S. Deꢀ
partment of Energy, Office of Science, and Office of Basic Enerꢀ
gy Sciences, under Contract DEꢀAC02ꢀ06CH11357. MRCAT
operations are supported by the Department of Energy and the
MRCAT member institutions. XPS investigations made use of
Shared Experimental Facilities supported in part by the MRSEC
Program of the National Science Foundation under award no.
DMRꢀ1419807. Support for ICPꢀMS investigations was provided
by a core center grant P30ꢀES002109 from the National Institute
of Environmental Health Sciences, National Institutes of Health.
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
Oh, S.; Gallagher, J. R.; Miller, J. T.; Surendranath,
Y. J. Am. Chem. Soc. 2016, 138, 1820.
Brisdon, B. J.; Griffin, G. F.; Pierce, J.; Walton, R.
A. J. Organomet. Chem. 1981, 219, 53.
Kaye, R. C.; Stonehill, H. I. J. Chem. Soc. 1952,
3240.
9
Sawyer, D. T.; Komai, R. Y. Anal. Chem. 1972, 44,
715.
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
Jackson, M. N.; Surendranath, Y. J. Am. Chem.
Soc. 2016, 138, 3228.
Bonin, J.; Costentin, C.; Louault, C.; Robert, M.;
Savéant, J.ꢀM. J. Am. Chem. Soc. 2011, 133, 6668.
Caix, C.; ChardonꢀNoblat, S.; Deronzier, A.;
Ziessel, R. J. Electroanal. Chem. 1993, 362, 301.
ChardonꢀNoblat, S.; Cosnier, S.; Deronzier, A.;
Vlachopoulos, N. J. Electroanal. Chem. 1993, 352,
213.
REFERENCES
(1)
Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R.
(19)
CipotꢀWechsler, J.; Covelli, D.; Praetorius, J. M.;
Hearns, N.; Zenkina, O. V.; Keske, E. C.; Wang,
G.; Whitesides, G. M. Chem. Rev. 2005, 105, 1103.
(2)
Beulen, M. W. J.; Kastenberg, M. I.; van Veggel, F.
C. J. M.; Reinhoudt, D. N. Langmuir 1998, 14,
7463.
R.;
Kennepohl,
P.;
Crudden,
C.
M.
Organometallics 2012, 31, 7306.
(20)
(21)
(22)
Desnoyer, A. N.; Behyan, S.; Patrick, B. O.; Dauth,
A.; Love, J. A.; Kennepohl, P. Inorg. Chem. 2016,
55, 13.
(3)
(4)
Chidsey, C. E. Science 1991, 251, 919.
Pinson, J.; Podvorica, F. Chem. Soc. Rev. 2005, 34,
429.
Gupta, A.; Blakemore, J. D.; Brunschwig, B. S.;
Gray, H. B. J. Phys. Condens. Matter 2016, 28,
94002.
(5)
(6)
MahoucheꢀChergui, S.; GamꢀDerouich, S.;
Mangeney, C.; Chehimi, M. M. Chem. Soc. Rev.
2011, 40, 4143.
Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T.
M.; Mujsce, A. M. J. Am. Chem. Soc. 1990, 112,
4301.
McCrory, C. C. L.; Devadoss, A.; Ottenwaelder,
X.; Lowe, R. D.; Stack, T. D. P.; Chidsey, C. E. D.
J. Am. Chem. Soc. 2011, 133, 3696.
(23)
(24)
Winkler, K.; McKnight, N.; Fawcett, W. R. J.
Phys. Chem. B 2000, 104, 3575.
(7)
(8)
(9)
Devadoss, A.; Chidsey, C. E. D. J. Am. Chem. Soc.
2007, 129, 5370.
Kobayashi, K.; Ohtsu, H.; Nozaki, K.; Kitagawa,
S.; Tanaka, K. Inorg. Chem. 2016, 55, 2076.
Sheridan, M. V.; Lam, K.; Geiger, W. E. J. Am.
Chem. Soc. 2013, 135, 2939.
(25)
(26)
Lever, A. B. P. Inorg. Chem. 1990, 29, 1271.
Blakemore, J. D.; Gupta, A.; Warren, J.;
Brunschwig, B. S.; Gray, H. B. J. Am. Chem. Soc.
2013, 135, 18288.
Schmickler,
W.;
Santos,
E.
Interfacial
electrochemistry; 2010.
Insert Table of Contents artwork here
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
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
9
ACS Paragon Plus Environment