Highly efficient Ru/carbon catalysts prepared by pyrolysis of supported Ru complex towards the hydrogen production from ammonia borane

Article abstract of DOI:10.1016/j.apcata.2016.08.018

Highly efficient carbon-supported Ru catalysts towards the H₂ production from the ammonia borane (AB: NH₃BH₃) hydrolysis reaction have been developed by using tri(2,2′-bipyridyl) ruthenium (II) chloride hexahydrate as metal precursor. Different temperatures were used to decompose the metal precursor, which afford the preparation of various nanoparticles sizes, ranging from 3.8 to 13.5?nm. The catalytic results revealed that both geometric and electronic features of the metal nanoparticles are crucial to attain optimum catalytic activity towards the H₂ production from the hydrolysis of AB. The optimum catalyst under the experimental condition used was that prepared at 800?°C, which efficiently boosts the hydrolysis of AB, achieving an average TOF number of 670?mol H₂/mol metal min, which is among the highest values ever reported for a monometallic Ru catalyst in this reaction up to date.

Full text of DOI:10.1016/j.apcata.2016.08.018

Applied Catalysis A: General 527 (2016) 45–52
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Highly efficient Ru/carbon catalysts prepared by pyrolysis of
supported Ru complex towards the hydrogen production from
ammonia borane
Miriam Navlani-Garcíaa, Kohsuke Moria,b,c,∗, Ai Nozakia, Yasutaka Kuwaharaa,b
,
Hiromi Yamashitaa,b,∗
a Division of Materials and Manufacturing Science, Graduate School of Engineering, Osaka University, 2-1 Yamada-oka, Suita, Japan
b Unit of Elements Strategy Initiative for Catalysts & Batteries, Kyoto University, Katsura, Kyoto 615-8520, Japan
c PRESTO, JST, Honcho, Kawaguchi, Saitama 332-0012, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 May 2016
Received in revised form 19 August 2016
Accepted 20 August 2016
Available online 22 August 2016
Highly efficient carbon-supported Ru catalysts towards the H2 production from the ammonia borane (AB:
NH3BH3) hydrolysis reaction have been developed by using tri(2,2ꢀ-bipyridyl) ruthenium (II) chloride
hexahydrate as metal precursor. Different temperatures were used to decompose the metal precursor,
which afford the preparation of various nanoparticles sizes, ranging from 3.8 to 13.5 nm. The catalytic
results revealed that both geometric and electronic features of the metal nanoparticles are crucial to
attain optimum catalytic activity towards the H2 production from the hydrolysis of AB. The optimum
catalyst under the experimental condition used was that prepared at 800 ◦C, which efficiently boosts the
hydrolysis of AB, achieving an average TOF number of 670 mol H2/mol metal min, which is among the
highest values ever reported for a monometallic Ru catalyst in this reaction up to date.
Keywords:
Hydrogen production
Ammonia borane
Ru nanoparticles
Metal complex
Nanoparticle size
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
consequence of these promising properties, important efforts have
been directed towards the fine tuning of the H2 release from AB,
Hydrogen (H2) is currently a worldwide accepted environmen-
tally clean energy carrier by combining with the proton exchange
membranes (PEM) technology, which offers a solution for the
energy problem as well as for the environmental pollution orig-
inated from the fossil fuels [1–3]. However, the hydrogen-based
economy still needs to face with important problems related to
controlled H2 storage and release [2,4]. In this sense, H2 can be
stored using both physical and chemical methods and the search
for an optimum material for the H2 storage is still the target of many
researches [5–7]. Among the huge variety of materials used for the
H2 storage, ammonia borane (NH3-BH3, AB) is considered as the
most promising chemical H2 storage material, which is due to its
remarkable theoretical gravimetric H2 content (19.6 wt.%, 6.5 and
13.1 wt.% for the first and second equivalent of H2, respectively),
low molecular weight (30.87 g/mol), stability in solid state, solu-
tion under ambient conditions and nontoxicity [8–14]. Then, as a
as well as the regeneration from the dehydrogenated “spent fuel”
materials [15,16]. The H2 release can take place by following differ-
ent pathways, both pyrolysis [17] or hydrolysis, being the last one
the most convenient route, as H2 can be released at room temper-
ature in presence of a suitable catalyst producing as much as 3 mol
of H2 per mol of AB by following Eq. (1) [8,18].
NH3−BH3 + 2H2O → NH4BO2 + 3H2
(1)
Various approaches, including both homogeneous and hetero-
geneous catalysts, have been already reported, but those catalytic
systems containing noble metals are considered as the most
favourable for this application [19,20]. Among them, Ru-based
catalysts have been claimed as one of the most effective cata-
lysts toward AB hydrolysis and great efforts have been devoted
to the optimisation of these catalytic systems by tuning either
the support used or the metal loading methodology [21–24]. Con-
cerning the studied supports, a wide variety of materials, such as
graphic carbon nitride [25], carbon black [26,27], carbon nanotubes
[28], Al2O3 [21,29], hydroxyapatite [30,31], Metal Organic Frame-
works [32], TiO2 [33] and so forth, have been already reported for
this application. In addition, various metal salts and experimental
∗
Corresponding authors at: Division of Materials and Manufacturing Science,
Graduate School of Engineering, Osaka University, 2-1 Yamada-oka, Suita, Japan.
E-mail addresses: mori@mat.eng.osaka-u.ac.jp (K. Mori),
yamashita@mat.eng.osaka-u.ac.jp (H. Yamashita).
http://dx.doi.org/10.1016/j.apcata.2016.08.018
0926-860X/© 2016 Elsevier B.V. All rights reserved.

46
M. Navlani-García et al. / Applied Catalysis A: General 527 (2016) 45–52
approaches have been considered in order to develop efficient cata-
lysts towards the AB hydrolysis. For instance, the in-situ generation
of Ru(0) nanoparticles reduced by AB was studied by Akbayrak et al.
[33], while a more complex system such as Ru@SiO2 nanospheres
was prepared by Yao et al. [34] via a one-pot synthetic route by
following a reversed-micelle technique, and the use of polymer-
capped Ru-based metal nanoparticles was reported by Rakap [35].
Furthermore, bimetallic nanoparticles, both in core-shell or alloy
configuration have been also investigated in the AB decomposition.
In this regard, several compositions such as RuCo [36–38], RuNi
[37], RuPt [35] and RuFe [37] have been proved to exhibit promis-
ing catalytic activity for H2 generation from the hydrolysis of AB.
However, despite the huge variety of approaches considered, the
search for an optimum catalyst to boost the AB hydrolysis is still
the goal of many researches, as is demonstrated by the increas-
ing number of publications dealing with this issue during the last
years.
2.2. Catalysts characterisation
Thermogravimetric analysis was performed by using a Rigaku
Thermoplus EVO2. Porous texture characterisation of the carbon
support and Ru/C catalysts was carried out by means of adsorp-
tion of N2 at −196 ◦C by using a BELSORP-max system (BEL Japan,
Inc.). Prior to the isothermal analysis, samples were degassed
under vacuum at 250 ◦C for 4 h in order to remove any adsorbed
impurities. Apparent surface area was calculated from N2 adsorp-
tion isotherms using the BET equation (SBET) at 10−3 < P/Po < 10−2
.
Total micropore volume (VDR(N2)) was calculated applying the
Dubinin–Radushkevich (DR) equation to the N2 adsorption data.
Ru loading was determined by using ICP-OES (inductively coupled
plasma-optical emission Spectroscopy).
Transmission Electron Microscopy (TEM) micrographs were
recorded using a Hitachi H-800 electron microscope equipped with
an energy-dispersive X-ray (EDX) detector, operated at 200 kV.
High Resolution Transmission Electron Microscopy (HR-TEM) was
used for the proper determination of the average nanoparticle size
of those samples which displayed smaller nanoparticles. To prepare
the samples for the TEM analysis, a small amount of the catalyst was
suspended in ethanol and sonicated in an ultrasonic bath for few
seconds. After that, a drop of this suspension was deposited onto a
copper grid and dried at room temperature. Ru K-edge XAFS spec-
tra were recorded using a fluorescence-yield collection technique
at the beam line 01B1 station with an attached Si (111) monochro-
mator at SPring-8, JASRI, Harima, Japan (prop. No. 2015A1149). The
EXAFS data were normalized by fitting the background absorption
coefficient, around the energy region higher than the edge of about
35–50 eV, with the smoothed absorption of an isolated atom. The
EXAFS data were examined using the Rigaku EXAFS analysis pro-
gram. X-ray photoelectron spectroscopy (XPS) data were recorded
using an ESCA 3400 Electron Spectrometer and the quantitative
analyses were done from the integrated intensities of the spectra.
Although Ru is normally analysed by following the signals from the
3d photoelectrons, herein the 3p spectra are used instead to avoid
interferences from the C 1 s spectrum of the carbon support.
Herein, we report a synthetic method of carbon-supported Ru
2+
nanoparticles (Ru/C) based on the impregnation of the Ru(bpy)3
precursor on carbon support and the subsequent precursor decom-
position by heat treatment at temperatures ranging from 600 to
1000 ◦C. This parameter is proven to have an important impact on
the final features of the catalysts, which ultimately reflected in the
catalytic performance of Ru/C towards the hydrogen production
from AB hydrolysis.
2. Experimental
2.1. Catalysts preparation
In order to prepare the Ru/C catalysts 0.037 g of tri(2,2ꢀ-
bipyridyl) ruthenium (II) chloride hexahydrate was dissolved in
150 mL of MeOH to obtain a target 0.5 wt.% of Ru content. Then
1 g of the support (commercial SHIRASAGI M (Osaka Gas Chemi-
cals Co. Ltd. Carbon support) was added to the solution and the
mixture was stirred for 2 h at room temperature. Afterwards, it
was vacuum dried at 40 ◦C to remove the solvent completely (a
filtration step was not used in order to avoid the possible metal
lost). The obtained powder was then placed in a glass tube and
heat treated at 600, 700, 800, 900 and 1000 ◦C, in each instance
for 2 h under N2 atmosphere and using a heating rate of 5 ◦C/min.
In order to favour the reduction of the metal, the catalysts were
reduced afterwards by using NaBH4 (Ru/NaBH4 molar ratio of 1/10)
and stirring for 3 h at room temperature. Then, the samples were
washed with water until neutral and dried at 100 ◦C overnight.
The as-prepared catalysts were denoted as Ru/C(600), Ru/C(700),
Ru/C(800), Ru/C(900) and Ru/C(1000), respectively. To elucidate
the importance of the metal precursor and carbon support in the
synthetic procedure used, two additional catalysts with the same
metal loading (0.5 wt.%) were prepared as reference samples. The
first one was a Ru/C-based catalyst prepared using Ru(NO)(NO3)3
as Ru precursor. The carbon support was impregnated for 2 h at
room temperature with the adequate amount of the metal precur-
sor aqueous solution and subsequently dried at 77 ◦C under vacuum
for 30 min to remove the solvent. Then, the sample was placed in
a glass tube and heat treated at 800 ◦C (5 ◦C/min) for 2 h under
N2 atmosphere. After that, the sample was reduced with NaBH4
by using the aforementioned procedure. This sample was denoted
as Ru/C(imp). The second additional catalyst was a Ru/SiO2-based
2.3. Catalytic tests
In order to assess the catalytic performance in the hydrolysis of
AB, 0.02 g of Ru-based catalysts and 10 mL of distilled water were
placed into a Schlenk-type reaction vessel (30 mL) connected with
a gas burette. The system was purged 3 times with nitrogen gas.
After this, the reaction was initiated by introducing 1 mL of aque-
ous solution of AB (containing 1.5 mmol of AB) and the mixture was
gently stirred while the evolution of the H2 production was moni-
tored by using the gas burette system. In the standard experiments,
the reaction temperature was maintained at 30 ◦C by using an oil
bath. The catalytic tests were performed three times to evaluate
the reproducibility of the reaction. Additional experiments at dif-
ferent reaction temperatures were also carried out with Ru/C(800)
to calculate the activation energy. The reusability of Ru/C(800) was
evaluated by conducting five consecutives runs. For this purpose,
the spent catalyst was kept in the reactor while a new aqueous
solution of AB (1 mL, containing 1.5 mmol of AB) was added and
the hydrogen output was monitored under the same experimental
condition as for the first reaction cycle.
3. Results & discussion
The thermogravimetric (TG) analysis profiles obtained with the
2+
2+
catalyst prepared from Ru(bpy)3 and commercial fumed silica
bare Ru(bpy)3 complex used as Ru precursor showed the sharp
decomposition peak at around 400 ◦C, as shown in Fig. S1(a). On the
by conducting the same experimental steps as for the Ru/C(800)
counterpart.
2+
other hand, the decomposition of the carbon-supported Ru(bpy)3

M. Navlani-García et al. / Applied Catalysis A: General 527 (2016) 45–52
47
Table 1
Ru loading (wt.%) and particle size of the Ru-based catalysts.
Table 2
Results of the XPS analysis in the Ru 3p core level spectra.
Sample
Ru loading (wt.%) ICP
Average Ru NP size (nm)
Sample
Metallic Ru (atomic%)
Oxidised (atomic%)
Ru/C(600)
Ru/C(700)
Ru/C(800)
Ru/C(900)
Ru/C(1000)
Ru/C(imp)
Ru/SiO2
0.56
0.56
0.56
0.63
0.57
0.55
0.47
3.8 1.0
6.2 2.3
7.0 3.4
11.1 6.1
13.5 11.2
3.3 0.6
2.0 0.4
Ru/C(600)
Ru/C(700)
Ru/C(800)
Ru/C(900)
Ru/C(1000)
Ru/C(imp)
Ru/SiO2
67.9
59.0
76.7
81.0
75.1
58.6
92.0
32.1
41.0
23.3
19.0
24.9
41.4
8.0
oxygen-containing functional groups and ␲-complex structures
(␲-sites), for the preparation of carbon-supported Pt catalysts from
Pt precursor [45]. According to this study, the oxidised Pt species
tend to be formed on the oxygen-containing functional groups,
while the metallic Pt nanoparticles are preferentially formed on
the ␲-sites. In our case, a larger proportion of metallic Ru nanopar-
ticles can be observed in Ru/C(800) prepared using Ru(bpy)32+ than
in Ru/C(imp) by using Ru(NO)(NO3)3, as evidenced by the rela-
tive intensity of the peaks corresponding to Ru O and Ru Ru in
the EXAFS spectra (See Fig. 2). This fact might indicate that the
nanoparticles tend to be preferentially formed in the ␲-sites when
complex before the calcination step gradually occurs and delayed
its decomposition temperature to higher degree, which is due to the
stronger interaction between carbon support and metal complex
(Fig. S1(b)). The results obtained from the N2 adsorption-desorption
isotherms indicated the Type I isotherms without blockage of the
porosity even after the Ru loading (See Fig. S2 and Table S1).
Fig. 1 shows the TEM images and the pertinent histograms with
the Ru nanoparticle size distribution for all Ru/C catalysts and the
data obtained from the counting of ∼200 nanoparticles is collected
in Table 1. The clear observation of Ru nanoparticles suggests that
the Ru complex decomposed to form Ru nanoparticles even when
the sample is prepared at the lowest temperatures. In all samples,
Ru nanoparticles have spherical shape in a homogeneous distribu-
tion on the carbon surface. However, important differences in the
average nanoparticle size and size distribution are observed; the
average Ru particle size increased with increasing the temperature
used to prepare the samples, which led to a notable broadening
of the nanoparticle size distribution. As expected, this effect is
especially important in sample Ru/C(1000), where nanoparticles
as large as 50 nm can be observed in the TEM image. This tendency
is in good agreement with results addressed elsewhere [39,40]. In
contrast, samples Ru/C(imp) and Ru/SiO2, treated at 800 ◦C (See Fig.
S3), showed smaller NP size and narrower size distribution than
2+
Ru(bpy)3 was used as metal precursor, while Ru(NO)(NO3)3 is
more prone to decompose near the oxygen-containing functional
groups.
To further analyse the electronic state of the Ru nanoparticles,
XPS characterisation was carried out by recording the spectra in
the 3p region. The results revealed the presence of two bands cor-
responding to the 3p3/2 and 3p1/2 transitions, having a binding
energy of about 462.5 and 485.5 eV, respectively. The deconvolu-
tion of both bands as a doublet, and the integration of the resulting
peaks, attributed to the presence of Ru in the metallic form and oxi-
dised state [46,47], gave information about the relative proportion
of reduced and oxidised Ru species in the catalysts (See Table 2).
These results pointed out the pivotal importance of both metal
complex decomposition temperature and the interaction between
the metal precursor and graphene structure of carbon support in
controlling the electronic features of the resulting Ru nanoparticles.
The comparison of samples Ru/C(800) and Ru/C(imp) suggested
the larger relative proportion of Ru(0) in Ru/C(800) compared to
Ru/C(imp), which empathised the importance of the interaction
2+
that prepared from Ru(bpy)3 (3.3, 2.0 and 7.0 nm for Ru/C(imp),
Ru/SiO2 and Ru/C(800), respectively). The results of the ICP charac-
terisation listed in Table 1 revealed that the final Ru loading (wt.%) is
very close to the targeted 0.5 wt.% for all the as-prepared catalysts.
X-Ray absorption measurements were performed in order to
elucidate the effect of the metal precursors on the Ru state in the
Ru/C(800) and Ru/C(imp) samples. Fig. 2 shows the Fourier Trans-
forms (FT) k3-weighted Ru K-edge extended X-ray absorption fine
structure (EXAFS) of these samples. As it can be seen, both cata-
lysts as well as RuO2 showed a first peak at around 1.5 Å due to
the Ru O bond [41]. In addition, a second peak was also observed
at around 2.4 Å, which can be assigned to the contiguous Ru Ru
bond in the metallic form [42–44]. It should be highlighted that the
relative intensity of these two peaks is different in both catalysts,
in which the contribution of Ru Ru bonds in the sample Ru/C(800)
2+
with the ␲-sites when Ru(bpy)3 is used as metal precursor.
3.1. Catalytic activity in the AB hydrolysis
The AB conversion profiles obtained with the Ru-catalysts are
plotted in Fig. 3. As it can be seen, the temperature used to prepare
the catalysts exhibits a remarkable effect on the catalytic activity of
the Ru/C samples. The conversion profiles indicated that no induc-
tion period was observed for all catalysts. Complete conversion of
AB was achieved by using Ru/C(800), Ru/C(900) and Ru/C(1000),
while AB hydrolysis did not take place completely in the cases of
Ru/C(600), Ru/C(700), Ru/C(imp) and Ru/SiO2. These results indi-
cated that relatively larger size of Ru nanoparticles is indispensable
for attaining complete reaction. As shown in Fig. 4, a volcano
type relationship was found between TOF values (calculated at
t = 5.5 min) based on Ru and the temperature used to prepare the
Ru/C catalysts. The best-performing sample achieved an average
TOF of 670 mol H2 mol Ru min−1, which, to the best of our knowl-
edge, is among the highest values ever reported for a monometallic
Ru catalyst up to date (See Table 3). The pivotal importance of the
metal precursor-support interaction is pointed out by the compari-
son of the catalytic activity shown by Ru/C(600)–(1000) and the two
additional reference samples, prepared with either different sup-
port or different metal precursor. As it can be seen in Fig. 3, in the
case of using the sample Ru/SiO2, the reaction was sluggish as com-
2+
prepared by using Ru(bpy)3 precursor is higher than that in the
Ru/C(imp) using Ru(NO)(NO3)3. This result suggests that the size of
Ru nanoparticles in the sample Ru/C(800) is larger than that in the
Ru/C(imp), which is in good agreement with the TEM results. It can
be said that kinetics of the reduction of Ru precursor under calcina-
tion conditions plays a crucial role in determining the particle size.
Generally, the rapid reduction generates more nuclei of Ru metals
in a shorter period and efficiently suppresses the growth of metal
nanoparticles, while slow reduction produces less nuclei, which
lead to the growth of metal nanoparticles. Thus, the difference in
the nanoparticles size can be ascribed to the different reduction
kinetics when different metal precursors are used, which in turn
might be related to different metal precursor-support interactions.
Concerning the electronic properties of the resulting nanoparti-
cles, the interaction between Ru precursor and the carbon support
might have an important influence in the oxidation state of the
Ru species. Dam and Bekkum proposed two types of ligand sites,

48
M. Navlani-García et al. / Applied Catalysis A: General 527 (2016) 45–52
Fig. 1. TEM micrographs and the corresponding histograms for the Ru/C catalysts: (a) Ru/C(600); (b) Ru/C(700); (c) Ru/C(800); (d) Ru/C(900); (e) Ru/C(1000). Notice that (a)
corresponds to a HR-TEM micrograph of sample Ru/C(600), while (b)–(e) are TEM images of sample Ru/C(700)–(1000), respectively.

M. Navlani-García et al. / Applied Catalysis A: General 527 (2016) 45–52
49
Table 3
Catalytic activity for Ru-based catalysts in the hydrolysis of AB.
Catalyst
TOF (min−1
)
Ea (kJ/mol)
Reference
(molH2/mol metal min)
Ru/C
Ru/C
Ru(0)@MWCNT
Ru/graphene
Ru/graphene
670
429.5
329
600
100
14.3
34.8
33
12.7
11.7
28
77
37.4
70
38.2
58
47.7
52
30.6
56.3
36.6
25.7
28.1
37.2
This study
[27]
[28]
[48]
[49]
[50]
[51]
[25]
[33]
[34]
[31]
[32]
[52]
[53]
[35]
[54]
[55]
[56]
[23]
Ru/Carbon thin film 70.5
Ru@X-NW
Ru/g-C3N4
Ru/TiO2
Ru@SiO2
Ru(0)@HAp
Ru/MIL-96
135
313.0
241
200
137
231
RuCu (1:1)/␥−Al2O3 16.4
RuCu/graphene
Pt–Ru@PVP
135
308
339.5
824.7
914
Ru@Ni/graphene
RuNi/Ti3C2X2
RuNi/TiO2
Ni0.74Ru0.26 alloy NPs 194.8
Fig. 2. Ru K-edge FT-EXAFS spectra for the Ru foil, RuO2, Ru/C(800) and Ru/C(imp)
catalysts.
3.0
2.5
2.0
1.5
1.0
0.5
3.0
2.5
Ru/C(800)_T=30 ºC
2.0
Ru/C(600)
Ru/C(800)_T=40 ºC
Ru/C(800)_T=45 ºC
Ru/C(800)_T=50 ºC
Ru/C(700)
1.5
Ru/C(800)
Ru/C(900)
1.0
0.0
0
Ru/C(1000)
2
4
6
8
Ru/C(imp)
0.5
Ru/SiO2
Time (min)
0.0
0
5
10 15 20 25 30 35 40 45 50 55 60
Fig. 5. Catalytic performance dependence of catalyst Ru/C(800) with the reaction
temperature.
Time (min)
pared to the Ru/C catalysts and the maximum conversion (∼97%)
is significantly delayed, while, even though the initial reaction rate
using the catalyst prepared by using Ru(NO)(NO3)3 (Ru/C(imp)) is
similar to those displayed by Ru/C(600)-(1000) catalysts, a total AB
decomposition was not achieved in that case.
Fig. 3. Catalytic activity of the Ru-based catalysts in the AB hydrolysis reaction at
30 ◦C.
The reaction temperature has a noticeable effect on the catalytic
performance of Ru/C(800) catalyst and the H2 generation rate was
enhanced with increasing the reaction temperature (See Fig. 5).
These experiments conducted at different temperatures were used
to calculate the activation energy (Ea) for the AB hydrolysis catal-
ysed by Ru/C(800) according to the Arrhenius Equations (Eq. (2)),
where A is the frequency factor, R is the ideal gas constant and
T is the reaction temperature (See Fig. S4). The activation energy
was calculated to be 14.3 kJ/mol, which is one of the lowest val-
ues among reported for Ru-based catalysts so far, highlighting the
superior catalytic performance of the present Ru/C catalytic system
(See Table 3).
900
800
700
600
500
400
300
200
100
Ea
0
ln k = ln A − ⁄RT
(2)
Ru/C(600)Ru/C(700)Ru/C(800)Ru/C(900)Ru/C(1000)
Furthermore, the reusability of Ru/C(800) was investigated by
performing five consecutive runs at 30 ◦C (Fig. 6). The results of
the TOF after 5.5 min and maximum conversion are summarised
in Table 4. The reaction proceeded smoothly and almost complete
conversion could be attained even in the fifth run, but the TOF value
decreased gradually, which was about 60% of the initial TOF in the
fifth run. The loss of activity could be tentatively attributed to the
Catalyst
Fig. 4. TOF number (mol H2/mol Ru min) calculated at t = 5.5 min for the Ru/C sam-
ples prepared by Ru complex decomposition. The data were average TOF numbers
calculated after performing three different catalytic tests.

50
M. Navlani-García et al. / Applied Catalysis A: General 527 (2016) 45–52
particles appear to have lower durability, which was attributed to
the more significant adsorption of B-containing species on the Pt
surface.
3.0
2.5
2.0
1.5
1.0
0.5
0.0
It is well-known that a plausible mechanism for the hydrol-
ysis of AB is initiated by the formation of an activated complex
species between NH3BH3 and the metal particle surface. This reac-
tion intermediate undergoes the attack by a H2O molecule, which
readily leads to concerted dissociation of the B N bonding and
hydrolysis of the resulting BH3 intermediate to produce the boric
acid product along with the H2 release [58,59]. Therefore, it seems
logical to think that the ability to form Ru-NH3BH3 activated com-
plex, which is strongly related to the Ru nanoparticles size, may
exert an important influence on the final catalytic performance. As
mentioned before, there is an optimum Ru nanoparticle size for
boosting the AB hydrolysis, which in our case appeared for sam-
ple Ru/C(800) with Ru particles of 7 nm. It is generally accepted
that the smaller metal particles lead to better catalytic performance
because they have a larger proportion of unsaturated surface atoms.
This would be in good agreement with the better performance
of sample Ru/C(800) compared to those attained by Ru/C(900)
and Ru/C(1000) whose average Ru particles sizes were 11.1 and
13.5 nm, respectively. However, it seems harder to explain the infe-
rior catalytic activity of Ru/C(600) and Ru/C(700) (3.8 and 6.2 nm,
respectively). In this point, it could be assumed that the ability to
form the Ru-NH3BH3 activated complex formed in the first stage
of the reaction might be higher in the smaller particle size, lead-
ing therefore to space exclusion and repulsive adsorbate–adsorbate
interaction. Such obstruction may inhibit the subsequent attack
by H2O molecules to reach the NH3BH3-based metallic complex.
Bearing these results in mind, it could be extracted that reaching
a balance between the higher reactivity and the space exclusion
originated from smaller particles is mandatory to find an optimum
catalyst towards the H2 production from the AB hydrolytic dehy-
drogenation.
Another explanation for the optimum catalytic performance
attained by the Ru/C(800) might be associated with the specific cat-
alytic site on the appropriate Ru particle size. It has been reported
that the atomic arrangement is strongly related to its particle size.
In this line, metals such as platinum, palladium, nickel or ruthenium
have been addressed to form so-called B5-type sites. The B5-type
sites are defined as consisting of five metal atoms in a certain three
dimensional arrangement with three Ru atoms in one layer (i.e., a
kind of terrace) and two further Ru in the layer directly above this at
a monoatomic step on a Ru (0001) terrace [60,61]. The majority of
such active sites were recognised to be favourably present at edges
related to low-coordinated surface atoms, e.g., edge atoms on (110)
and (113) planes and formed only on medium-sized nanoparticles
with a mean diameters of 1.5–7.0 nm [62]. The specific contribution
of such unique catalytic sites in the AB dehydrogenation cannot be
ruled out. Actually, it is accepted that the B5-type sites are respon-
sible for strong physical adsorption of nitrogen, thus efficiently
boosting structure-sensitive catalytic reaction exemplified by the
NH3 catalytic decomposition [60,63].
1stRun
2ndRun
3rdRun
4thRun
5thRun
0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15
Time (min)
Fig. 6. Reusability test of Ru/C(800) at 30 ◦C.
Table 4
Catalytic activity of sample Ru/C(800) in the reusability tests.
Run
TOF (mol H2/mol Ru min)
Conversion (%)
1st
718
575
517
460
417
100
100
100
100
97
2nd
3rd
4th
5th
following reasons: nanoparticle sintering under the experimental
condition used, changes in the electronic state of the Ru nanopar-
ticles and/or effect of the accumulated metaborate in the medium
[51]. In order to elucidate the main responsible for the gradual cat-
alytic activity decay, some additional experiments were conducted.
Firstly, the effect of the nanoparticle size increase after the catalytic
run on the catalytic activity decay was ruled out by recording TEM
micrographs after the first run (See Fig. S5). In this case the sample
showed an average Ru nanoparticle size of 7.1 3.1 nm, which is
very similar to that showed by the fresh counterpart (7.0 3.4 nm).
In addition post-reaction XPS analysis of sample Ru/C(800) after
five catalytic runs was also performed and the results indicated
that the relative proportion of metallic and oxidised species was
slightly different from that found in the fresh sample (76.7 and 67.1
atomic% of Ru(0) for the fresh and used catalyst, respectively). Inter-
estingly, XPS analysis revealed the presence of B-containing species
on the catalyst surface, which might be also responsible of the loss
of catalytic activity. Therefore, according to these experimental evi-
dences, the catalytic activity decay observed along the consecutive
runs can be attributed to these adverse factors but the negative
effect of the accumulated metaborate in the medium should not
be ruled out. The relatively poor stability displayed by the cata-
lyst indicates that further efforts should be directed towards the
improvement of the stability and recyclability of the present cat-
alytic system for practical application.
The catalytic performance displayed by the studied samples
might be related to geometric and electronic aspect of the Ru-
based catalysts. On the one hand, the volcano type relationship
between the catalytic activity and the temperature used to prepare
the catalysts might be attributed to the effect of this experimen-
tal parameter in controlling the final nanoparticle size. Sample
Ru/C(800) with a moderate average nanoparticle size among inves-
tigated, displayed better performance towards the hydrolysis of AB
than those catalysts with smaller or larger nanoparticles. Similar
phenomena has been claimed in the case of the Pt catalyst sup-
ported on CNTs for the hydrolysis of AB [57], where the catalyst with
nanoparticle size of 1.8 nm showed the optimum catalytic activity.
Furthermore, it was also found in that work that the smaller Pt
Furthermore, the electronic features of the resulting Ru
nanoparticles seem to play a crucial role in determining the final
catalytic activity. In this study, an optimal proportion of metal-
lic/oxidised Ru species, which was achieved in sample Ru/C(800),
is mandatory to attain high catalytic activity. These results indicate
that not only the presence of reduced Ru nanoparticles, but also
a moderate proportion of oxidised species is indispensable to effi-
ciently catalyse the AB hydrolysis. The importance of the oxidised
species in the present application has already been claimed else-
where [8,63,64]. According to previous studies, the metal-catalysed
hydrolysis of AB takes place by the following steps [65]: (1) NH3-
BH3 molecule and metal nanoparticles surface form an activated
complex where the interaction of the H2O molecules dissociates

M. Navlani-García et al. / Applied Catalysis A: General 527 (2016) 45–52
[2] A.F. Dalebrook, W. Gan, M. Grasemann, S. Moret, G. Laurenczy, Chem.
51
the B N bound; (2) NH3 ions release as ammonia hydroxide and
BH3− chemisorbed on the metal atoms transfers a H− to a free adja-
cent metal atom; (3) This H− leaves the metal site and reacts with
Commun. 49 (2013) 8735–8751.
[3] K. Mori, M. Dojo, H. Yamashita, ACS Catal. 3 (2013) 1114–1119.
[4] H.-L. Jiang, S.K. Singh, J.-M. Yan, X.-B. Zhang, Q. Xu, ChemSusChem 3 (2010)
541–549.
[5] R. Lan, J.T.S. Irvine, S. Tao, Int. J. Hydrogen Energy 37 (2012) 1482–1494.
[6] K. Koh, J.E. Seo, J.H. Lee, A. Goswami, C.W. Yoon, T. Asefa, J. Mater. Chem. A 2
(2014) 20444–20449.
[7] A.K. Hill, L. Torrente-Murciano, Appl. Catal. B 172–173 (2015) 129–135.
[8] K. Mori, T. Taga, H. Yamashita, ChemCatChem 7 (2015) 1285–1291.
[9] A. Staubitz, A.P.M. Robertson, I. Manners, Chem. Rev. 110 (2010) 4079–4124.
[10] Z. Li, G. Zhu, G. Lu, S. Qiu, X. Yao, J. Am. Chem. Soc. 132 (2010) 1490–1491.
[11] L. Pengyao, X. Zhengli, L. Zhaoyan, H. Jiale, L. Qingbiao, S. Daohua,
Nanotechnology 26 (2015) 025401.
H2O to produce H2 and OH−; (4) OH− reacts with boron in BH2
,
−
generating BH2(OH)−; 5) H− is transferred from BH2(OH)− to a free
adjacent metal atom and BH(OH)− is formed. Steps 4 and 5 are
−
repeated once more until B(OH)3 and H2 is released. According
to this mechanism, it has also been claimed that oxidised species
are crucial for this reaction, as they have been reported to act as
Lewis acid sites for OH− absorption due to the strong affinity of
these sites toward oxygen of the OH− group [66] and the resulting
polarization of the OH− groups would make them more prone to
react with the boron species attached to the reduced metal sites.
In our case, the importance of the oxidised species can be clearly
observed by comparing the poorer behaviour displayed by Ru/SiO2
sample, where Ru is mainly in the metallic state, with those showed
by Ru/C samples, where the contribution of Ru oxidised species is
much more important.
[12] M.A. Esteruelas, A.M. López, M. Mora, E. On˜ate, ACS Catal. 5 (2015) 187–191.
[13] L. Yang, J. Su, X. Meng, W. Luo, G. Cheng, J. Mater. Chem. A 1 (2013)
10016–10023.
[14] H. Cheng, T. Kamegawa, K. Mori, H. Yamashita, Angew. Chem. Int. Ed. 53
(2014) 2910–2914.
[15] O.T. Summerscales, J.C. Gordon, Dalton Trans. 42 (2013) 10075–10084.
[16] H. Li, Q. Yang, X. Chen, S.G. Shore, J. Organomet. Chem. 751 (2014) 60–66.
[17] A. Gutowska, L. Li, Y. Shin, C.M. Wang, X.S. Li, J.C. Linehan, R.S. Smith, B.D. Kay,
B. Schmid, W. Shaw, M. Gutowski, T. Autrey, Angew. Chem. Int. Ed. 44 (2005)
3578–3582.
[18] M. Yadav, Q. Xu, Energy Environ. Sci. 5 (2012) 9698–9725.
[19] H. Cheng, X. Qian, Y. Kuwahara, K. Mori, H. Yamashita, Adv. Mater. 27 (2015)
4616–4621.
4. Conclusions
[20] K. Fuku, R. Hayashi, S. Takakura, T. Kamegawa, K. Mori, H. Yamashita, Angew.
Chem. Int. Ed. 52 (2013) 7446–7450.
[21] H. Can, Ö. Metin, Appl. Catal. B 125 (2012) 304–310.
[22] H. Ma, C. Na, ACS Catal. 5 (2015) 1726–1735.
[23] G. Chen, S. Desinan, R. Rosei, F. Rosei, D. Ma, Chem. Eur. J. 18 (2012)
7925–7930.
[24] W.R.H. Wright, E.R. Berkeley, L.R. Alden, R.T. Baker, L.G. Sneddon, Chem.
Commun. 47 (2011) 3177–3179.
[25] Y. Fan, X. Li, X. He, C. Zeng, G. Fan, Q. Liu, D. Tang, Int. J. Hydrogen Energy 39
(2014) 19982–19989.
[26] S. Basu, A. Brockman, P. Gagare, Y. Zheng, P.V. Ramachandran, W.N. Delgass,
J.P. Gore, J. Power Sources 188 (2009) 238–243.
[27] H. Liang, G. Chen, S. Desinan, R. Rosei, F. Rosei, D. Ma, Int. J. Hydrogen Energy
37 (2012) 17921–17927.
[28] S. Akbayrak, S. Özkar, ACS Appl. Mater. Interfaces 4 (2012) 6302–6310.
[29] G.P. Rachiero, U.B. Demirci, P. Miele, Catal. Today 170 (2011) 85–92.
[30] H. Durak, M. Gulcan, M. Zahmakiran, S. Ozkar, M. Kaya, RSC Adv. 4 (2014)
28947–28955.
In conclusion, highly efficient Ru/C catalysts towards the
ammonia borane decomposition reaction have been successfully
prepared by using Ru(bpy)3 as metal precursor. The effect of
2+
the decomposition complex temperature in the final catalytic per-
formance was assessed by synthesising different samples with
temperatures ranging from 600 to 1000 ◦C. Additional catalysts
have also been evaluated to elucidate the role of the Ru-catalysts
features. The optimum catalyst under the experimental condition
used was shown to be the sample Ru/C(800) with average NP size of
7.0 nm, which efficiently boosts the hydrolysis of ammonia borane,
achieving 100% of conversion and with an average TOF number of
670 mol H2/mol metal min, which, as far as we are aware, is among
the highest values ever reported for a monometallic Ru catalyst in
the hydrolysis of ammonia borane up to date. The outstanding cat-
alytic performance displayed by the sample Ru/C(800) as compared
to those prepared by using different heat treatment temperatures,
as well as different metal precursor and catalyst support, points out
the importance of both electronic and geometric parameters in con-
trolling the final properties of the studied Ru/C catalysts towards
the efficient production of H2 from the hydrolytic dehydrogenation
of ammonia borane. We expect that this synthetic strategy will be
extendable in the deposition of other catalyst metals. The further
study is now under investigation in our laboratory.
[31] S. Akbayrak, P. Erdek, S. Özkar, Appl. Catal. B 142–143 (2013) 187–195.
[32] L. Wen, J. Su, X. Wu, P. Cai, W. Luo, G. Cheng, Int. J. Hydrogen Energy 39 (2014)
17129–17135.
˙
[33] S. Akbayrak, S. Tanyıldızı, I. Morkan, S. Özkar, Int. J. Hydrogen Energy 39
(2014) 9628–9637.
[34] Q. Yao, W. Shi, G. Feng, Z.-H. Lu, X. Zhang, D. Tao, D. Kong, J. Power Sources
257 (2014) 293–299.
[35] M. Rakap, Appl. Catal. A Gen. 478 (2014) 15–20.
[36] Li, C. Zeng, G. Fan, Int. J. Hydrogen Energy 40 (2015) 9217–9224.
[37] N. Cao, J. Su, X. Hong, W. Luo, G. Cheng, Chem. Asian J. 9 (2014) 562–571.
[38] N. Cao, J. Su, W. Luo, G. Cheng, Catal. Commun. 43 (2014) 47–51.
[39] T. Kawaguchi, W. Sugimoto, Y. Murakami, Y. Takasu, J. Catal. 229 (2005)
176–184.
[40] M. Tsypkin, J.L.G. de la Fuente, S. García Rodríguez, Y. Yu, P. Ochal, F. Seland, O.
Safonova, N. Muthuswamy, M. Rønning, D. Chen, S. Sunde, J. Electroanal.
Chem. 704 (2013) 57–66.
Acknowledgments
[41] D. Moonshiram, J.W. Jurss, J.J. Concepcion, T. Zakharova, I. Alperovich, T.J.
Meyer, Y. Pushkar, J. Am. Chem. Soc. 134 (2012) 4625–4636.
[42] A.M. Abdel-Mageed, S. Eckle, R.J. Behm, J. Am. Chem. Soc. 137 (2015)
8672–8675.
[43] J. Liu, S. He, C. Li, F. Wang, M. Wei, D.G. Evans, X. Duan, J. Mater. Chem. A 2
(2014) 7570–7577.
[44] S. Alayoglu, P. Zavalij, B. Eichhorn, Q. Wang, A.I. Frenkel, P. Chupas, ACS Nano
9 (2009) 3127–3137.
[45] H.E. Van Dam, H. Van Bekkum, J. Catal. 131 (1991) 335–349.
[46] D.J. Morgan, Interface Anal. 47 (2015) 1072–1079.
[47] E. Ertas, M. Gulcan, A. Bulut, M. Yurderi, M. Zahmakiran, Microporous
Mesoporous Mat. 226 (2016) 94–103.
The present work was partially supported by Grants-in-Aid
for Scientific Research (Nos. 26220911, 25289289, and 26630409,
26620194) from the Japan Society for the Promotion of Science
(JSPS) and MEXT. We acknowledge Dr. Eiji Taguchi and Prof. H.
Yasuda at the Research Center for Ultra-High Voltage Electron
Microscopy, Osaka University, for their assistance with the TEM
measurements. KM, YK and HY thank MEXT program “Elements
Strategy Initiative to Form Core Research Center”.
[48] C. Du, Q. Ao, N. Cao, L. Yang, W. Luo, G. Cheng, Int. J. Hydrogen Energy 40
(2015) 6180–6187.
Appendix A. Supplementary data
[49] N. Cao, W. Luo, G. Cheng, Int. J. Hydrogen Energy 38 (2013) 11964–11972.
[50] R. Fernandes, N. Patel, R. Edla, N. Bazzanella, D.C. Kothari, A. Miotello, Appl.
Catal. A 495 (2015) 23–29.
[51] S. Akbayrak, S. Ozkar, Dalton Trans. 43 (2014) 1797–1805.
[52] G.P. Rachiero, U.B. Demirci, P. Miele, Int. J. Hydrogen Energy 36 (2011)
7051–7065.
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.apcata.2016.08.
018.
[53] N. Cao, K. Hu, W. Luo, G. Cheng, J. Alloys Compd 590 (2014) 241–246.
[54] N. Cao, J. Su, W. Luo, G. Cheng, Int. J. Hydrogen Energy 39 (2014) 426–435.
[55] Li, C. Zeng, G. Fan, Int. J. Hydrogen Energy 40 (2015) 3883–3891.
[56] K. Mori, K. Miyawaki, H. Yamashita, ACS Catal. 6 (2016) 3128–3135.
References
[1] A. Bulut, M. Yurderi, Y. Karatas, M. Zahmakiran, H. Kivrak, M. Gulcan, M. Kaya,
Appl. Catal. B 164 (2015) 324–333.

52
M. Navlani-García et al. / Applied Catalysis A: General 527 (2016) 45–52
[57] W. Chen, J. Ji, X. Feng, X. Duan, G. Qian, P. Li, X. Zhou, D. Chen, W. Yuan, J. Am.
Chem. Soc. 136 (2014) 16736–16739.
[58] M. Chandra, Q. Xu, J. Power Sources 156 (2006) 190–194.
[59] M. Chandra, Q. Xu, J. Power Sources 159 (2006) 855–860.
[60] H. Bielawa, O. Hinrichsen, A. Birkner, M. Muhler, Angew. Chem. Int. Ed. 40
(2001) 1061–1063.
[62] S. Czekaj, J. Pin, Wambach, J. Phys. Chem. C 117 (2013) 26588–26597.
[63] W. Raróg-Pilecka, E. Mis´kiewicz, D. Szmigiel, Z. Kowalczyk, J. Catal. 231 (2005)
11–19.
[64] R. Fernandes, N. Patel, A. Miotello, R. Jaiswal, D.C. Kothari, Int. J. Hydrogen
Energy 37 (2012) 2397–2406.
[65] M. Chandra, Q. Xu, J. Power Sources 156 (2006) 190–194.
[66] N. Patel, R. Fernandes, A. Miotello, J. Catal. 271 (2010) 315–324.
[61] S. Dahl, E. Törnqvist, I. Chorkendorff, J. Catal. 192 (2000) 381–390.