Immobilization of a molecular catalyst on carbon nanotubes for highly efficient electro-catalytic water oxidation

Article abstract of DOI:10.1039/c4cc06959e

Electrochemically driven water oxidation has been performed using a molecular water oxidation catalyst immobilized on hybrid carbon nanotubes and nano-material electrodes. A high turnover frequency (TOF) of 7.6 s^-1together with a high catalytic current density of 2.2 mA cm^-2was successfully obtained at an overpotential of 480 mV after 1 h of bulk electrolysis.

Full text of DOI:10.1039/c4cc06959e

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Immobilization of a molecular catalyst on carbon
nanotubes for highly efficient electro-catalytic
water oxidation†
Cite this: Chem. Commun., 2014,
50, 13948
Received 3rd September 2014,
Accepted 11th September 2014
Fusheng Li,a Lin Li,a Lianpeng Tong,a Quentin Daniel,a Mats Gothelidb and
¨
Licheng Sun*ac
DOI: 10.1039/c4cc06959e
www.rsc.org/chemcomm
Electrochemically driven water oxidation has been performed using onto multi-walled carbon nanotubes (MWCNTs) through electro-
a molecular water oxidation catalyst immobilized on hybrid carbon static p–p stacking interactions.12 However, difficulties in modifying
nanotubes and nano-material electrodes. A high turnover frequency the substrate catalyst and fabricating the device hindered further
(TOF) of 7.6 sꢀ1 together with a high catalytic current density of application of these methods. Therefore, novel and widely applic-
2.2 mA cmꢀ2 was successfully obtained at an overpotential of able approaches are urgently needed.
480 mV after 1 h of bulk electrolysis.
Carbon nanotubes have extraordinary electronic, thermal, optical,
and mechanical properties. There are two main routes to function-
The world today is eagerly demanding a renewable energy source alized CNTs, i.e., covalent and non-covalent functionalization. The
as a replacement for fossil fuels that cause environmental and former approach disrupts the p-networks of CNTs, which creates
climate concerns. Artificial photosynthesis, which captures and possible losses in their mechanical and electrical properties. The
stores solar energy, may provide a sustainable choice.1,2 Most latter approach is based on the adsorption of appropriate molecules
research efforts in this field have been focused on the develop- on the CNTs surface, which preserves their desired properties, and
ment of efficient water oxidation catalysts (WOCs) and proton this approach is widely used.13–15 Cao et al. modified CNTs with
reduction catalysts to reduce the energy barrier of water splitting surfactants such as sodium dodecyl sulfate (SDS),16 in their method,
reactions.3–5 Several other aspects, however, have to be consid- the hydrophobic aliphatic chain can interact with the surface of the
ered before the construction of a practical water splitting device. CNTs, which inspired us. Herein, we report a carbon nanotubes
One of the most challenging tasks is to employ these catalysts in based nano-material that was functionalized by molecular WOC
functional devices without affecting their performance evaluated Ru-C12 (Scheme 1) for highly efficient electrocatalytic water
under homogeneous system.
oxidation. The catalyst core is the well-defined mononuclear
For implementation of any WOC in electrocatalytic or photo- WOC Ru(pdc)(pic)3. The dodecyloxy group was used here in
electrocatalytic devices for water splitting, the catalysts have to be order to attach the catalyst to the MWCNT. A benefit of this
immobilized on the surface of a conductive electrode or semi- design is that the long hydrophobic aliphatic chain group can
conductor.6,7 One widely applied approach is to ligate molecular provide a linkage to the MWCNTs, resulting in an electro-
catalysts on the surface of the electrode through carboxylic or phos- catalytic nanomaterial. In addition, this dodecyloxy group will
phoric ester linkage.8–10 Other approaches, reported by our groups decrease the solubility of Ru-C12 in water, which is an impor-
recently, include the immobilization of catalysts [Ru(pdc)(pic)3 tant consideration for the application of the self-assembled
pdc = 2, 6-pyridine dicarboxylic acid] on a conductive carbon system between the Ru-C12 and MWCNTs.
surface through click chemistry11 and attachment of complex
To measure the electro-catalytic activity of Ru-C12 for water
[Ru(bda)(Py-pyrene)2, bda = 2,20-bipyridine-6,60-dicarboxylic acid] oxidation, cyclic voltammetry (CV) measurements were conducted
a Department of Chemistry, School of Chemical Science and Engineering,
KTH Royal Institute of Technology, 100 44 Stockholm, Sweden.
E-mail: lichengs@kth.se
b ICT Material Physics, Electrum 229, KTH Royal Institute of Technology,
16440 Kista, Sweden
c State Key Laboratory of Fine Chemicals, DUT-KTH Joint Education and
Research Center on Molecular Devices, Dalian University of Technology (DUT),
116024 Dalian, China
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4cc06959e
Scheme 1 Structures of complexes Ru-C12 and Ru(pdc)(pic)3.
13948 | Chem. Commun., 2014, 50, 13948--13951
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
Fig. 2 SEM image of the surface of [Ru-C12–MWCNTsCOOH@GDL]
electrode.
Fig. 1 Cyclic voltammograms (CV) of
[Ru(pdc)(pic)3] (blue line) in 0.1 M potassium phosphate buffer (pH 7.0, 10%
acetonitrile), with glassy carbon as the working electrode, Ag/AgCl as the
reference electrode, and Pt as the counter electrode. Scan rate is 100 mV sꢀ1
1 mM Ru-C12 (red line) and
is promising to prepare an electro-catalytic nanomaterial. In other
words, we can easily collect the precipitates and apply them on
some conductive substrates for further device making. In the
present work, the Ru-C12-CNTs precipitates were collected by
filtration through a gas diffusion layer (GDL). In such a simple
.
both in organic solvent and aqueous solution. Fig. S4 (ESI†) way, GDL-based electrodes were prepared whose size could be
shows CV curves of Ru-C12 and [Ru(pdc)(pic)3] in dichloro- tailored.
methane. The E1/2 Ru(II/III) of Ru-C12 is dramatically moved to a
SEM and EDX experiments show that the nano-structure is well
more negative potential compared to [Ru(pdc)(pic)3]. This is retained and presents a large specific surface area after Ru-C12–
caused by the electron donating ability of the dodecyloxy group, MWCNTsCOOH is adsorbed on GDL (Fig. 2). The Ru content is
which enhances the electron density of ruthenium center. Fig. 1 3.26 wt% on the [Ru-C12–MWCNTsCOOH@GDL] electrode
shows CV curves in aqueous pH 7.0 phosphate buffer solution (Table S1, ESI†), while there is no Ru on [MWCNTsCOOH@GDL]
of Ru-C12 and [Ru(pdc)(pic)3]. Ru-C12 exhibits one irreversible electrode (Table S2, ESI†). Fig. S9 and S11 (ESI†) show that the
Ru(II) to Ru(III) oxidation peak at Eon = 0.82 V vs. NHE, indicat- distribution of [Ru-C12–MWCNTsCOOH@GDL] is more uniform
ing the ligand exchange reaction of picoline by water molecule than that of bare [MWCNTsCOOH@GDL]. These observations
during the redox process as discovered earlier by our group confirmed that the electrodes were successfully prepared
with the core catalyst [Ru(pdc)(pic)3].
with the achievement of a Ru-C12–MWCNTsCOOH nano-
An onset of catalytic water oxidation current by Ru-C12 was material and a [Ru-C12–MWCNTsCOOH@GDL] composition
observed at around 1.2 V vs. NHE, which is very close to that by electro-active nanomaterial.
[Ru(pdc)(pic)3]. The catalytic current, however, showed more
Catalytic activities of water oxidation for the functionalized
enhancement in the case of Ru-C12 when potential beyond 1.2 V catalyst-CNTs@GDL electrodes were investigated in a 50 mM
was applied. Two possible reasons are ascribed to the stronger potassium phosphate buffer solution. Both the functionalized
catalytic current triggered by Ru-C12 than that by [Ru(pdc)(pic)3]: electrodes and the pristine electrode (control) were examined
(1) the electron donating effect of dodecyloxy substituent in Ru-C12 as working electrodes. Because the surfaces of CNTs are highly
is beneficial in overcoming the rate-limiting step in the catalytic hydrophobic, a rotating disk electrode was employed to get rid
cycle, assuming that Ru-C12 and [Ru(pdc)(pic)3] have the same of bubbles (generated oxygen) on the surface and to ensure an
catalytic pathway. (2) The dodecyloxy chain of Ru-C12 is attracted to effective contact between the surface of the working electrode
the surface of glassy carbon electrode, leading to higher concen- and water during bulk electrolysis.
tration of Ru-C12 than that of [Ru(pdc)(pic)3] in the interface layer
of the working electrode.
When [MWCNTsCOOH@GDL] was used as a blank, no significant
catalytic current could be observed below the applied potential until
Ru-C12 was successfully immobilized into two types of CNTs: 1.6 V (vs. NHE) in CVs (Fig. 3), which indicates that MWCNTsCOOH
MWCNTs and MWCNTs functionalized with carboxyl group is inactive for water oxidation. Furthermore, the performance of
(MWCNTsCOOH). After mixing MWCNTs with Ru-C12 in water, [MWCNTs@GDL] is similar to that of [MWCNTsCOOH@GDL], where
the resulting solution exhibited a light brown color, indicating almost no catalytic current can be observed. When [Ru(pdc)(pic)3]
that a certain amount of Ru-C12 still remained in the liquid was employed on CNTs electrodes, the formed [Ru(pdc)(pic)3–
phase. In contrast, upon mixing MWCNTsCOOH with the same MWCNTsCOOH@GDL] electrode shows only a slight enhance-
concentration of Ru-C12, a colourless solution with precipitation ment in the catalytic current in comparison to the blank electrode
was observed as shown in Fig. S5 (ESI†), which suggests a complete [MWCNTsCOOH@GDL]. In contrast, when Ru-C12 was employed,
graft of Ru-C12 onto the surface of MWCNTsCOOH. This is significant enhancement of catalytic current was observed for the
probably because, when the surface of MWCNTsCOOH immobi- formed [Ru-C12–MWCNTsCOOH@GDL] electrode.
lized by Ru-C12 becomes more hydrophobic, precipitate could be
The significant enhancement in catalytic current is ascribed
obtained from the solution. Notably, this precipitation behaviour to the large amount of catalyst loading on the electrode and to
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 13948--13951 | 13949

View Article Online
ChemComm
Communication
catalyst coverage is far less than the amount of Ru-C12 loading
(0.24 mg cmꢀ2, see ESI†), because only the outside superficial
layer of the hybrid material contacts with the aqueous electro-
lyte and responds to the CV sweep.
In comparison to our previous study, where [Ru(pdc)(pic)3]
was immobilized on conductive carbon surface via covalent
bonding with an effective loading of 1.2 ꢁ 0.4(10ꢀ10 mol cmꢀ2),11
the effective loading of catalyst in the present study was signifi-
cantly improved. More importantly, G0 in the present work was
estimated by CV performed under aqueous conditions, same as
where catalytic O2 evolution was carried out, whereas the loading
amount of catalyst in our previous work was estimated in organic
solvent. Therefore, we believe that the catalyst loading G0 repre-
sents the realistic amount of catalysts involved in electro-catalytic
water oxidation. Moreover, the value of G0 was expected to be
significantly enhanced if a conducting supporting material with
larger specific surface was used instead of GDL.
Fig. 3 CVs of [Ru-C12–MWCNTsCOOH@GDL] electrode (blue), [Ru-C12–
MWCNTs@GDL] electrode (green), [Ru(pdc)(pic)3–MWCNTsCOOH@GDL]
electrode (red), without catalyst only [MWCNTsCOOH@GDL] (light blue),
without catalyst only [MWCNTs@GDL] (pink), only GDL (black). All CV curves
were obtained by using rotating disk electrode (1000 rpm, 0.2 cm2) as
working electrode, Ag/AgCl as the reference electrode, and Pt as the counter
electrode in aqueous solution (phosphate buffer, pH 7.0, ionic strength = 0.1),
The electro-catalytic activity of [Ru-C12–MWCNTsCOOH@GDL]
was further evaluated by chronoamperometric experiments in a
pH 7.0 phosphate buffer. In a typical experiment, the current
density was monitored continuously under a sequence of applied
potential steps in the range from 1.12 to 1.32 V (Fig. 4). TOFs were
with scan rate of 100 mV sꢀ1
.
the long hydrophobic aliphatic chain assisted immobilization of calculated according to eqn (S2) (ESI†).
Ru-C12 on MWCNTsCOOH. [Ru-C12–MWCNTs@GDL] was used as
The catalytic current density measured on [Ru-C12–MWCNTs-
a reference in comparison to the [Ru-C12–MWCNTsCOOH@GDL] COOH@GDL] electrode as a function of Z followed a Tafel
electrode, whereas both of them exhibit an identical redox waves behaviour (the inset in Fig. 5). The logarithm of yielded TOFs
for the Ru(II) to Ru(III) and an onset potential of uprising varies linearly on the applied overpotentials from 300 to 500 mV.
current at around 1.15 V. These potentials are consistent with A TOF of 5.1 sꢀ1 was achieved at Z = 350 mV, while it reaches
our observation of the electrochemical analysis of WOC Ru-C12 13.7 sꢀ1 at Z = 500 mV. At 1.3 V, the calculated TOF is 12.4 sꢀ1
,
in homogeneous conditions. However, the catalytic current of which is the initial TOF for bulk electrolysis. Bulk electrolysis
[Ru-C12–MWCNTsCOOH@GDL] is much higher than that of experiments (which will be discussed later) show that the faradaic
[Ru-C12–MWCNTs@GDL] electrode. This might be due to the efficiency is nearly 100%; hence, the initial TOF obtained here
larger loading amount of catalyst on the surface of MWCNTsCOOH is credible.
than MWCNTs, as discussed above. Another reason could be
The overpotential Z refers to the additional voltage to the
attributed to the carboxyl group, –COOH, in MWCNTsCOOH, thermodynamically required oxidation potential of water at pH 7.0
which might form a hydrogen bonding network with water (E = 0.82 vs. NHE). Under an applied overpotential Z beyond 350 mV,
molecules, making the electrode surfaces have better contact
with water in the buffer solution.
To determine the strength of the interaction between
Ru-C12 and the surfaces of the CNTs and stability, the
[Ru-C12–MWCNTsCOOH@GDL] electrode was rotated (1000 rpm)
for 2 h without applying any potential. CV measurements (Fig. S6,
ESI†) were carried out before and after rotation, and they show
almost identical CVs, which suggests that both the catalysts
and the CNTs are strongly attached to the GDL, demonstrating
a steady and stable combination among the three components
in [Ru-C12–MWCNTsCOOH@GDL].
The effective coverage of functional WOCs (G0, mol cmꢀ2) on the
[Ru-C12–MWCNTsCOOH@GDL] electrode surface was estimated
according to an established method.17 According to eqn (S1) (ESI†),
the peaks currents of Ru(II) to Ru(III) redox couple and the scan rate
were supposed to follow a linear relationship. When the scan rates
were increased from 25 to 300 mV per second, the peak currents of
Fig. 4 Chronoamperometric current densities measured in phosphate
buffer (pH 7.0, IS = 0.1 M) at [Ru-C12–MWCNTsCOOH@GDL] (red) and
[MWCNTsCOOH@GDL] (black) under the application of sequential potential
Ru(II) to Ru(III) redox wave were proportional to the scan rate (Fig. S8,
ESI†). An average catalyst coverage of 8.5 ꢁ 0.4 (10ꢀ10 mol cmꢀ2
)
was obtained from three independent measurements. This average steps.
13950 | Chem. Commun., 2014, 50, 13948--13951
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
oxide nanoparticles,8,10 electrostatic interactions with CNTs
decorated by polyamidoamine ammonium dendrimers,19 and
our previous study where the complex [Ru(bda)(Py-pyrene)2]
was attached to MWCNTs through electrostatic or p–p stacking
interactions.12 [Ru-C12–MWCNTsCOOH@GDL] is clearly superior
to any other system in the table in terms of overpotential, current
density, TON and TOF.
In summary, a molecular WOC Ru-C12 with a long alkyl
chain has been successfully synthesized and fully characterized.
This complex has been immobilized on CNTs through the
hydrophobic interaction forming a hybrid nanomaterial, which
can be easily deposited onto GDL. This fabricated [Ru-C12–
MWCNTsCOOH@GDL] electrode not only preserves the water
oxidation activity of the molecular catalyst, but also exhibits an
essentially high initial TOF of 12.4 sꢀ1. Furthermore, an average
current density of 0.84 mA cmꢀ2, corresponding to an impress-
ive TOF of 3.5 sꢀ1, was achieved after 15 h bulk electrolysis at
overpotential of 480 mV in pH 7.0 buffer. This methodology
provides a simple and reliable approach for transformation
from a homogeneous catalyst to a functional electrochemical
electrode for a total water splitting cell (as shown by the video in
ESI†). Further research is being carried out to improve the
stability of the system towards practical implementation.
Fig. 5 TOF plot of [Ru-C12–MWCNTsCOOH@GDL] electrode as a func-
tion of overpotential Z. The inset is Tafel plots showing current density of
[Ru-C12–MWCNTsCOOH@GDL] electrode vs. Z.
an appreciable catalytic current was observed for [Ru-C12–
MWCNTsCOOH@GDL], while a negligible current was seen
for the control electrode [MWCNTsCOOH@GDL]. This difference
in current densities became quite significant at Z = 500 mV,
indicating a much more effective Faraday process that is attributed
to a more efficient catalytic water oxidation occurred on the
[Ru-C12–MWCNTsCOOH@GDL] surface.
Notes and references
Bulk electrolysis was carried out with a controlled potential
of 1.3 V vs. NHE on [Ru-C12–MWCNTsCOOH@GDL] in 50 mM
potassium phosphate buffer for a time period of 1 h and 15 h.
The corresponding current density was recorded and has been
shown in Fig. S14a (ESI†). Based on the current, the total charge
passed through the cell was calculated. The amount of charge
that passed the electrode was divided by 4F to get a theoretical O2
yield, which was 4.9 and 31.2 mmol for 1 h and 15 h, respectively.
The real oxygen evolution was monitored by a fluorescence-based
oxygen sensor fixed in the headspace of a gas-tight cell. At the end
of experiment, the amount of oxygen was calibrated by injecting
0.5 mL headspace gas sample into GC. Product of 4.7 mmol O2
after 1 h and 29.0 mmol O2 after 15 h were confirmed, corre-
sponding to a faradaic efficiency of 96% and 93%, respectively.
The average TOF based on the overall TON is 7.6 sꢀ1 during 1 h
and 3.5 sꢀ1 during 15 h electrolysis experiments (Fig. S14b, ESI†).
This decay of TOF during bulk electrolysis suggests a decrease in
the catalytic rate probably due to the decomposition of the catalyst.
Although not all catalysts in [Ru-C12–MWCNTsCOOH@GDL] are
able to remain intact, the stability of the immobilized catalysts is
critically improved in comparison to [Ru(pdc)(pic)3] evaluated
under homogeneous catalytic conditions.18
1 T. R. Cook, D. K. Dogutan, S. Y. Reece, Y. Surendranath, T. S. Teets
and D. G. Nocera, Chem. Rev., 2010, 110, 6474–6502.
2 M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi,
E. A. Santori and N. S. Lewis, Chem. Rev., 2010, 110, 6446–6473.
3 T. J. Meyer, Acc. Chem. Res., 1989, 22, 163–170.
4 A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278.
5 K. Maeda and K. Domen, J. Phys. Chem. Lett., 2010, 1, 2655–2661.
6 L. Duan, L. Tong, Y. Xu and L. Sun, Energy Environ. Sci., 2011, 4,
3296–3313.
7 P. D. Tran, V. Artero and M. Fontecave, Energy Environ. Sci., 2010, 3,
727–747.
8 Z. Chen, J. J. Concepcion, J. W. Jurss and T. J. Meyer, J. Am. Chem.
Soc., 2009, 131, 15580–15581.
9 Z. Chen, J. J. Concepcion, H. Luo, J. F. Hull, A. Paul and T. J. Meyer,
J. Am. Chem. Soc., 2010, 132, 17670–17673.
10 F. Liu, T. Cardolaccia, B. J. Hornstein, J. R. Schoonover and
T. J. Meyer, J. Am. Chem. Soc., 2007, 129, 2446–2447.
11 L. Tong, M. Gothelid and L. Sun, Chem. Commun., 2012, 48,
10025–10027.
12 F. Li, B. Zhang, X. Li, Y. Jiang, L. Chen, Y. Li and L. Sun, Angew.
Chem., 2011, 123, 12484–12487.
13 P. Kang, S. Zhang, T. J. Meyer and M. Brookhart, Angew. Chem.,
Int. Ed., 2014, 53, 8709–8713.
14 L. Vaisman, H. D. Wagner and G. Marom, Adv. Colloid Interface Sci.,
2006, 128–130, 37–46.
15 Y.-L. Zhao and J. F. Stoddart, Acc. Chem. Res., 2009, 42, 1161–1171.
16 L. Cao, F. Scheiba, C. Roth, F. Schweiger, C. Cremers, U. Stimming,
H. Fuess, L. Chen, W. Zhu and X. Qiu, Angew. Chem., Int. Ed., 2006,
45, 5315–5319.
17 A. J. Bard and L. R. Faulkner, Electrochemical methods: fundamentals
and applications, Wiley, New York, 1980.
18 L. Duan, Y. Xu, M. Gorlov, L. Tong, S. Andersson and L. Sun,
Chem. – Eur. J., 2010, 16, 4659–4668.
19 F. M. Toma, A. Sartorel, M. Iurlo, M. Carraro, P. Parisse, C. Maccato,
S. Rapino, B. R. Gonzalez, H. Amenitsch, T. Da Ros, L. Casalis,
A. Goldoni, M. Marcaccio, G. Scorrano, G. Scoles, F. Paolucci, M. Prato
and M. Bonchio, Nat. Chem., 2010, 2, 826–831.
After 15 h electrolysis experiments driven by a relatively low
potential 1.3 V (vs. NHE), the current density of 0.84 mA cmꢀ2
retained, a TON of 1.7 ꢂ 105, and an average TOF of 3.5 sꢀ1 were
obtained. Table S3 (ESI†) lists the performance of [Ru-C12–
MWCNTsCOOH@GDL] in the current work and that of pre-
viously reported anodes for electrocatalytic water oxidation,
including incorporated catalysts by covalent bond to metal
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 13948--13951 | 13951