Facile heterogenization of a cobalt catalyst via graphene adsorption: Robust and versatile dihydrogen production systems

Article abstract of DOI:10.1039/c4cc02920h

A heterogeneous dihydrogen (H₂) production system has been attained by simply soaking electrodes made from electro-deposited graphene on FTO plated glass in solutions of a cobalt bis(dithiolate) compound. The resulting electrodes are active in weakly acidic aqueous solutions (pH > 3), have relatively low overpotentials (0.37 V versus platinum), show high catalytic rates (TOF > 1000 s^-1), and are resistant to degradation by dioxygen. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cc02920h

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Facile heterogenization of a cobalt catalyst via
graphene adsorption: robust and versatile
dihydrogen production systems†
Cite this: Chem. Commun., 2014,
50, 8065
Received 20th April 2014,
Accepted 30th May 2014
Shawn C. Eady, Sabrina L. Peczonczyk, Stephen Maldonado and Nicolai Lehnert*
DOI: 10.1039/c4cc02920h
www.rsc.org/chemcomm
A heterogeneous dihydrogen (H2) production system has been surfaces, and method of combination reported here require
attained by simply soaking electrodes made from electro- minimal synthetic effort and time to prepare. Resulting
deposited graphene on FTO plated glass in solutions of a cobalt catalyst-adsorbed graphene surfaces indicate activity for H2
bis(dithiolate) compound. The resulting electrodes are active in formation at reasonable overpotentials in weakly acidic aqu-
weakly acidic aqueous solutions (pH 4 3), have relatively low eous media. Most enticingly, this system shows resilience to O2
overpotentials (0.37 V versus platinum), show high catalytic rates exposure and resistance to catalyst leeching.
(TOF 4 1000 sÀ1), and are resistant to degradation by dioxygen.
Graphene oxide was prepared from flake graphite via
Hummer’s method,4 and an aqueous suspension was prepared
In the search for an economically viable solution for carbon- in a weakly alkaline carbonate-buffered system. Graphene-
neutral dihydrogen (H2) generation, the pool of low-cost early coated substrates were accessible by cyclic voltammetry (CV)
transition metal dihydrogen production catalysts, particularly in the graphene oxide suspension, using the substrate
hydrogenase models, has been fished extensively. A colourful (here: metal oxide-coated glass) as the working electrode
array of iron, nickel and cobalt catalysts (among others) has (Fig. S3, ESI†). For ease of study, initial testing was performed
been crafted, many of which show outstanding turnover on glass coated fluorine-doped tin oxide (FTO) electrodes, which
frequencies and catalytic lifetimes.1 Yet many catalytic systems provide a very generous cathodic window (BÀ1.3 vs. SCE) at very
are still being designed with synthetically challenging and low cost.5 To afford the catalyst-adsorbed surfaces, the graphene-
hence, costly ligand manifolds, and homogeneous compo- coated FTO surfaces were soaked in a 5 mM solution of catalyst
nents, which are both unsuitable for implementation at scale. in acetonitrile for 12 hours. Keeping simplicity in design, the H2
An ideal system would require minimal synthetic effort, only production catalyst selected for use here is a simple cobalt
inexpensive materials, and have the capability to be directly bis(dithiolate) complex that is afforded in high yield from a
tied to a renewable energy source to produce clean H2 without one-pot reaction with inexpensive materials (Chart 1).6
further modification. To achieve such a heterogeneous catalyst,
This compound has been shown by McNamara et al. to
adsorption to graphene provides an incredibly versatile option display impressive activity (TOF = 1400 hÀ1) and exemplary
open to nearly all aromatic systems, as shown for graphitic stability (TON = 6000) at low overpotentials for both electro-
systems in many examples.2 Additionally, graphene itself serves and photocatalytic (using Ru-bpy) dihydrogen generation in
as a means to access a number of different electrode materials, homogeneous phase.6 It is notable that no synthetic manipula-
as its electro-deposition on many substrates has been demon- tion is required for graphene adsorption of this compound,
strated.3 For implementation into (photo)cathode systems and which is ideal for applications at scale, in comparison with
eventual coupling to water oxidation half-cells, such a pliable other catalyst designs which often require custom introduction
catalyst interface is invaluable.
In this paper we report a facile means of designing a
graphene-interfaced heterogeneous catalyst system using
widely available, inexpensive materials. The catalyst, graphene
Department of Chemistry, University of Michigan, 930 North University Avenue,
Ann Arbor, Michigan 48109-1055, USA. E-mail: lehnertn@umich.edu
Chart 1 Co bis(dithiolate) compound (1) used for graphene adsorption,
(left), schematic representation of p-stacking interactions (right).
† Electronic supplementary information (ESI) available: Experimental details and
additional electrochemical data. See DOI: 10.1039/c4cc02920h
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 8065--8068 | 8065

View Article Online
Communication
ChemComm
reported in homogeneous phase, suggesting the catalyst is
unaltered upon adsorption to graphene.6
Analysis of activity upon increasing TFA addition shows a
linear correlation (under CV conditions; Fig. S5, ESI†) with no
activity saturation observed at the acid concentrations used.
Replacing the acidic solution after electrochemical analysis
with a new electrolyte-only solution (pH = 7.5) shows the same
redox couple seen in the first solution. Addition of TFA to the
second solution shows the same activity as observed in the
preceding run, and repetition of this process shows no signifi-
cant decrease of catalyst activity (Fig. S7, ESI†). These results
indicate that the catalyst-adsorbed surfaces are O2 stable
and resistant to leeching under the experimental conditions
(all manipulations done in air). XPS analysis of FTO/graphene/1
surfaces after electrochemical testing confirms these findings
(Fig. S1, ESI†). To assure these responses are not specific to
just TFA, electrochemical analyses of the catalyst-adsorbed
graphene surfaces with dilute hydrochloric acid solutions
were also performed. These tests elicit identical responses in
electrocatalytic behaviour (Fig. S8, ESI†).
The proton reduction mechanism of 1 on the graphene
surface appears to correspond to that reported in solution by
McNamara and coworkers. Upon addition of acid, no significant
change is seen in the reduction event at approximately À0.5 V;
however, a new wave appears at a potential roughly 0.25 V more
cathodic than the original wave (Fig. S9, ESI†). This is indicative
of an initial reduction of the cobalt anion to the dianion
preceding rapid protonation of the dianion. This protonation
event allows for a subsequent reduction unobserved in the
absence of acid. While the final protonation event was not
observed directly, these data suggest either an ECEC or ECCE
type mechanism. These results show that the activity profile for 1
on graphene closely mirrors that reported for 1 in homogenous
solution. These findings are indicative of the direct adsorption of
1 on the graphene surface and indicate a mechanism analogous
to the homogeneous catalyst.
Fig. 1 CoIII/II couple observed for FTO/graphene/1 in 0.1 M KPF6 aqueous
solution. The working electrode is catalyst-adsorbed graphene on FTO, the
auxiliary electrode is a platinum wire, and the reference is Ag/AgCl (sat. KCl).
of aromatic moieties for surface attachment (via involved
synthetic pathways).2b
CV analysis of catalyst-soaked FTO/graphene surfaces (after
extensive rinsing and sonication regiments) in aqueous
solutions indicates that the compound is bound, showing
quasi-reversible redox couples at an average potential of
À0.49 V. It is noteworthy that the redox couple of 1 is roughly
similar to that observed in solution (Fig. 1).6 Importantly, upon
addition of trifluoroacetic acid (TFA, pKa = 0.23), CVs of the
FTO/graphene/1 system show a sharp increase in current
during cathodic scanning at À0.73 V, a behaviour indicative
of electrocatalytic reduction of protons from TFA (Fig. 2). This
activity is observed with an overpotential of only 0.37 V as
compared to the reduction of TFA at a platinum electrode
under the same conditions (Fig. S6, ESI†). Testing of the same
graphene electrode in identical conditions prior to catalyst
adsorption showed minimal background current, confirming
this activity as being directly a result of the catalyst’s presence.
The half-wave potential (Ecat = À0.85 V) and catalytic peak
potential (ipc = À1.0 V) are both nearly identical to those
The bare FTO, FTO/graphene, and FTO/graphene/1 surfaces
were analyzed by X-ray photoelectron spectroscopy (XPS). The
high-resolution C 1s XPS data (Fig. 3a) of a bare FTO surface
Fig. 3 (a) High-resolution C 1s XPS data of (bottom) bare FTO, (middle)
FTO/graphene, and (top) FTO/graphene surfaces after soaking in a 5 mM
solution of 1 in acetonitrile for 12 hours. Co 2p (b) and S 2p (c) XPS data of
(bottom) graphene-deposited FTO glass and (top) the graphene surfaces
after deposition of 1 (using similar conditions).
Fig. 2 FTO/graphene/1 working electrode with the addition of trifluoro-
acetic acid. Counter electrode is a platinum wire and the reference is
Ag/AgCl (sat. KCl).
8066 | Chem. Commun., 2014, 50, 8065--8068
This journal is ©The Royal Society of Chemistry 2014

View Article Online
ChemComm
Communication
reduction precedes protonation, possibly indicating an
ECEC- or ECCE-type mechanism (Fig. S13, ESI†). It is therefore
interesting to note that the overall mechanism is similar in the
FTO/graphene/1 and graphite/1 systems, but that in the former
case a B0.2 V cathodic shift of the catalytic relative to the
first redox wave is observed. This may indicate an inherent
difference in the compound adsorption/interaction with
graphene compared to graphite.
The turnover frequency (TOF) of the immobilized catalyst
systems for dihydrogen production can be estimated using
direct comparison of cathodic peaks in the presence and
absence of acid (eqn (S1), ESI†). This method yields conserva-
tive estimates of 1007 sÀ1 and 701 sÀ1 for the FTO/graphene/1
system in TFA and HCl, respectively. It is noteworthy that these
rates dramatically exceed those reported for the compound in
homogeneous solution (1400 hÀ1). One problem for the
FTO/graphene/1 system at high acid concentrations is back-
ground acid reduction as well as tin oxide reduction at the
exposed FTO electrode surfaces (see acid controls, Fig. 2 and
Fig. S8, ESI†). Hence, acid-saturation conditions could not be
reached. Using the same method to calculate TOF for graphite/
1 systems gives a rate of 6182 sÀ1. It is noteworthy that this level
of activity is comparable to that of the renowned nickel
bis(diphosphine) catalysts (although at a comparably higher
overpotential), and is among the highest activities reported for
cobalt-based hydrogen production systems.9 This rate was
again determined in conditions where the catalyst activity was
not saturated, this time due to disturbance of the voltammo-
grams by H2 production and background proton reduction by
graphite at high acid concentrations (Fig. S14, ESI†). Therefore,
catalytic rates for heterogeneous systems of 1 were also analysed
by the ‘foot of the catalytic wave’ method (eqn (S2) and (S3),
ESI†). This method provides estimates of 5.77 Â 107 MÀ1 sÀ1 and
3.35 Â 107 MÀ1 sÀ1 for the graphene/1 systems in HCl and TFA,
respectively, and 3.68 Â 107 MÀ1 sÀ1 for the graphite/1 system
with TFA. While these estimates are exceptionally high
(as expected since this method is less reliable in instances of
substrate diffusion-limited activity), in combination with the
Fig. 4 Compound 1 adsorbed on HOPG (working electrode) with the
addition of TFA. A platinum wire is used as counter electrode and Ag/AgCl
(sat. KCl) as reference.
exhibit only features consistent with adventitious carbon.7 After
graphene deposition, signatures indicative of C–O (hydroxyl,
epoxy) groups at 286.7 eV and CQO (carbonyl groups) at
288.4 eV are prominent. These signatures are consistent with
previous reports of reduced graphene oxide on surfaces.8 After
soaking the graphene surface in catalyst, a peak at 287.7 eV
emerges in the C 1s spectra, corresponding to a C–S bonding
energy which would be expected for 1 (originating from the
dithiolate ligand). This is corroborated by the high-resolution
Co 2p and S 2p spectra (Fig. 3b and c) where features of Co and S
are clearly present after exposure to catalyst. A feature corres-
ponding to the Co–S energy is also present in the Co 2p data;
however, it is difficult to distinguish above the background.
To confirm the facile heterogenization of compound 1 on a
more controlled surface, adsorption on a highly-ordered
pyrolitic graphite electrode (HOPG, Pine Instrument Co.) was
also studied under the same conditions. For the purpose of
these studies, the highly-ordered nature of the graphite surface
was intended to simulate a sheet of graphene in terms of
electrostatic interactions. Here, compound 1 is seen to exhibit
a quasi-reversible redox couple at approximately À0.76 V
(Fig. S10, ESI†). Addition of TFA to graphite/1 shows catalytic
current at an onset potential slightly more cathodic than the
observed redox couple, approximately À0.77 (Fig. 4). This
current is absent at the same HOPG electrode prior to soaking
in catalyst, and is indicative of dihydrogen production from
TFA, with a peak catalytic potential of À1 V and a current half-
maximum potential of À0.92 V.
Increasing acid concentrations lead to a linear increase in
catalytic current (under CV conditions), with no activity saturation
observed at the acid concentrations studied (Fig. S11, ESI†).
Excitingly, analysis of 1 on graphite at higher TFA concentrations
(420 mM) showed such high levels of dihydrogen production
that peak catalytic currents were perturbed by gas bubbles at the
graphite electrode, still without reaching activity-limited currents
(Fig. S14, ESI†).
Fig. 5 Electrolysis of compound 1 adsorbed on graphite in 40 mM TFA
solution at À0.95 vs. SCE. Arrows indicate addition of supplemental acid
(40 mM aliquots).
The similarity between the catalytic onset potential of 1 and
its redox couple in the absence of acid on graphite suggests that
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 8065--8068 | 8067

View Article Online
Communication
ChemComm
results from eqn (S1) (ESI†) this demonstrates that heteroge- engineering. Applications in semiconductor systems are of
neous systems of 1 immobilized on graphitic surfaces have particular interest, and corresponding studies are underway.
impressive dihydrogen production activities.
Financial support was provided by the Rackham graduate
Electrolysis studies of 1 directly adsorbed on graphite were school and 3M (NTFG 5286067 to NL). We acknowledge Ms
performed to assess our heterogeneous system under multiple Kayla Pyper and Prof. Bart Bartlett’s group (University of
turnover conditions, with dihydrogen production monitored by Michigan) for use of their graphene oxide solutions and helpful
gas chromatography (Fig. 5). After 12 hours of a À0.95 V applied discussions. We thank Dr. Tanya Breault and Prof. Levi Thompson’s
potential in a 40 mM TFA solution, the graphite/1 system had group (University of Michigan) for help with dihydrogen
produced over 250 mmol of hydrogen and activity was still not quantification. SM acknowledges the Department of Energy
seen to subside. The current observed correlated closely with (DE-SC006628) for support of this work.
the evolved hydrogen (Fig. S16, ESI†) and a Faradaic efficiency
close to 100% was determined, indicating the exclusive use of
Notes and references
1 (a) C. d. Tard and C. J. Pickett, Chem. Rev., 2009, 109, 2245;
injected electrons for proton reduction. An initial rate of 2.88 Â
1016 [molecules H2] sÀ1 for H2 production was calculated for a
(b) S. Losse, J. G. Vos and S. Rau, Coord. Chem. Rev., 2010, 254, 2492.
graphite/1 electrode of 0.2 cm2 surface area. Unfortunately, the 2 (a) J. D. Blakemore, A. Gupta, J. J. Warren, B. S. Brunschwig and
H. B. Gray, J. Am. Chem. Soc., 2013, 135, 18288; (b) S. Dey, A. Rana,
S. G. Dey and A. Dey, ACS Catal., 2013, 3, 429; (c) G. Liu, B. Wu,
J. Zhang, X. Wang, M. Shao and J. Wang, Inorg. Chem., 2009, 48, 2383.
attempt to calculate a molecular TOF was prevented due to
difficulty in accurately quantifying the amount of 1 on the
graphite surface. Future work to determine the molecular TOF 3 (a) M. Hilder, B. Winther-Jensen, D. Li, M. Forsyth and
D. R. MacFarlane, Phys. Chem. Chem. Phys., 2011, 13, 9187;
(b) C. Liu, K. Wang, S. Luo, Y. Tang and L. Chen, Small, 2011,
7, 1203; (c) X. Wang, L. Zhi and K. Mullen, Nano Lett., 2007, 8, 323.
and long-term electrolysis studies are underway.
In summary, our initial studies show that heterogeneous
cobalt bis(dithiolate) electrocatalysts are easily afforded with- 4 W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
5 F. Wang, N. K. Subbaiyan, Q. Wang, C. Rochford, G. Xu, R. Lu,
A. Elliot, F. D’Souza, R. Hui and J. Wu, ACS Appl. Mater. Interfaces,
2012, 4, 1565.
out the need of time consuming and costly functionalization
of the ligand framework with large aromatic groups. These
systems display total catalysis in practically relevant pH ranges 6 W. R. McNamara, Z. Han, C.-J. Yin, W. W. Brennessel, P. L. Holland
and R. Eisenberg, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 15594.
7 D. J. Miller, M. C. Biesinger and N. S. McIntyre, Surf. Interface Anal.,
2002, 33, 299.
(pH 4 3), and no significant degradation or leeching is evident
from either acidic conditions or O2 exposure in the experiments
conducted. With the wide range of substrates available for 8 G. Sobon, J. Sotor, J. Jagiello, R. Kozinski, M. Zdrojek, M. Holdynski,
P. Paletko, J. Boguslawski, L. Lipinska and K. M. Abramski,
Opt. Express, 2012, 20, 19463.
9 S. Wiese, U. J. Kilgore, D. L. DuBois and R. M. Bullock, ACS Catal.,
graphene deposition, this technique ensures that hetero-
geneous dihydrogen-generation catalysis is viable on a variety
of materials, giving nearly limitless possibilities for materials
2012, 2, 720.
8068 | Chem. Commun., 2014, 50, 8065--8068
This journal is ©The Royal Society of Chemistry 2014