Hydrogen Production from Aqueous Solutions of Urea with Ruthenium-based Catalysts

Article abstract of DOI:10.1002/cssc.201500112

An efficient catalytic system for hydrogen production from aqueous solution of urea was developed using ruthenium-based catalysts as an alternative for existing systems based on several H₂ carriers. Ru/SiO₂ showed the highest H₂ yield among various SiO₂-supported transition-metal catalysts. Optimization of the catalyst support revealed that Ru/Al₂O₃ exhibited the best catalytic performance, with a minimum amount of CO formed as a by-product (1.8 mol %). A mechanistic study suggested that the reaction proceeds mainly via two steps: urea hydrolysis (H₂NCONH₂+H₂O→2 NH₃+CO₂) followed by NH₃ decomposition (2 NH₃→3 H₂+N₂).

Full text of DOI:10.1002/cssc.201500112

DOI: 10.1002/cssc.201500112
Communications
Hydrogen Production from Aqueous Solutions of Urea
with Ruthenium-based Catalysts
[
a]
[b]
[b]
[c]
Shinya Furukawa, Ryohei Suzuki, Kazuyoshi Ochi, Tatsuaki Yashima, and
[
b]
Takayuki Komatsu*
An efficient catalytic system for hydrogen production from
aqueous solution of urea was developed using ruthenium-
based catalysts as an alternative for existing systems based on
several H carriers. Ru/SiO showed the highest H yield among
formation is induced not only from urea but also from water. Al-
though urea production includes a certain energy loss (2NH +
3
ꢀ
1
CO !H NCONH +H O, DH=ꢀ54 kcalmol ), this loss is com-
2
2
2
2
parable to that for the practically used methylcyclohexane/tolu-
2
2
2
ꢀ1 [5c]
various SiO -supported transition-metal catalysts. Optimization
ene system (DH=ꢀ49 kcalmol ). Despite these significant
practical benefits, very few studies on catalytic hydrogen pro-
duction from aqueous solutions of urea (HPAU) have been re-
ported. Recently, Rollinson et al. reported that Ni/Al O catalyst
2
of the catalyst support revealed that Ru/Al O exhibited the
2
3
best catalytic performance, with a minimum amount of CO
formed as a by-product (1.8 mol%). A mechanistic study sug-
gested that the reaction proceeds mainly via two steps: urea
hydrolysis (H NCONH +H O!2NH +CO ) followed by NH
2
3
gave an 88 mol% yield of H at 7008C from aqueous solutions
2
[7]
of urea. However, significant amounts of CO (30 C% at 7008C)
2
2
2
3
2
3
decomposition (2NH !3H +N ).
or CH (4.7 C% at 5008C) were produced as byproducts in this
3
2
2
4
system. To develop a more efficient and clean catalytic system
for HPAU, it is necessary to have an excellent catalyst that exhib-
Hydrogen is regarded as one of the cleanest alternative fuels
and energy carriers, compared to fossil fuels. The major chal-
lenges for the effective application of hydrogen are its efficient
its high H yields at low reaction temperatures with minimum
2
formation of CO and CH byproducts.
4
In this study, we investigate the catalytic performances of
various transition-metal catalysts and the effect of the catalyst
support in HPAU. A mechanistic study of the reaction mecha-
nism is also described. We report a highly efficient catalytic
system for hydrogen production based on urea.
[
1]
storage and safety of transportation. For this purpose, a varie-
ty of chemical compounds with high gravimetric hydrogen ca-
[
2]
[3]
pacities have been developed, such as ammonia, hydrazine,
[
4]
ammonia-borane, and several organic hydrides represented
[
5]
by methylcyclohexane. However, some drawbacks such as
high toxicity (ammonia and hydrazine) and chemical waste as
byproducts (boron species from ammonia-borane and benzene
from methylcyclohexane) hamper the widespread application.
In this context, the use of urea offers several potential advan-
tages. Aqueous solutions of urea are essentially non-toxic, easy
to transport, will release only H , N , and CO upon hydrolysis
At first, we investigated the catalytic performances of vari-
ous transition-metal catalysts in HPAU. A series of SiO -support-
2
ed transition-metal catalysts were prepared by a pore-filling
impregnation method (see the Supporting Information for a de-
tailed experimental section). X-ray diffraction (XRD) patterns of
these catalysts confirmed the formation of metallic particles
for each catalyst without formation of any oxide phases that
were observable by XRD (Supporting Information, Figure S1).
Crystallite sizes were mostly in the range 2–19 nm, except for
those of copper and ruthenium with much larger sizes (53 and
30 nm, respectively). Catalytic HPAU was carried out in a contin-
uous fixed-bed flow reactor by direct feeding of aqueous solu-
tion of urea with argon flow at 5008C (the set-up of the appa-
ratus is shown in the Supporting Information, Figure S2).
2
2
2
(
H NCONH +H O!3H +N +CO ), and have a hydrogen con-
2
2
2
2
2
2
tent of 7.95 wt%, sufficient to meet the target set by the U.S.
Department of Energy (DOE) for hydrogen storage in transport
[
6]
applications. The process has a further merit in that hydrogen
[a] Dr. S. Furukawa
Department of Chemistry
Graduate School of Science and Engineering
Tokyo Institute of Technology
Figure 1 shows H and CO yields over various silica-support-
2
ed transition-metal catalysts (M/SiO ; M=Ag, Co, Cu, Fe, Ni,
2
2
-12-1-E1-10, Ookayama, Meguro-ku, Tokyo, 152-8551 (Japan)
Pd, Pt, Rh, and Ru). Silver, copper, and iron were inactive for
this reaction. Platinum, palladium, and cobalt gave very low H2
yields with comparable by-production of CO. Although nickel
[
b] R. Suzuki, K. Ochi, Prof. T. Komatsu
Department of Chemistry and Materials Science
Graduate School of Science and Engineering
Tokyo Institute of Technology
showed a moderate H yield, a significant amount of CO was
2
2
-12-1-E1-10, Ookayama, Meguro-ku, Tokyo, 152-8551 (Japan)
formed. In contrast, ruthenium and rhodium exhibited good H2
E-mail: komatsu.t.ad@m.titech.ac.jp
yields with limited by-production of CO. The highest H yield,
2
[
c] Prof. T. Yashima
Advanced Research Institute for the Science and Humanities
Nihon University
8
6 mol%, was obtained when using ruthenium despite its
large crystallite size. Notably, by-production of CO on rutheni-
um was only 1.7 mol%, much lower than that for Ni
Surugadai 1-8-14, Chiyoda-ku, Tokyo, 102-8251 (Japan)
(
13 mol%). Thus, the catalyst survey revealed that ruthenium is
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201500112.
the best metal element for HPAU.
ChemSusChem 0000, 00, 0 – 0
1
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
These are not the final page numbers! ÞÞ

Communications
Figure 1. Catalytic performances of various silica-supported transition metal
catalysts in HPAU at 5008C.
Figure 2. Catalytic performances of various ruthenium-based catalysts in
HPAU at 5008C.
We subsequently studied the effect of catalyst support on
the catalytic performance in HPAU over ruthenium catalysts.
Several oxides (SiO , Al O , CeO , ZrO , MgO, SiO -Al O , SiO -
metallic ruthenium at 280.0 eV, which agreed with the corre-
[8]
sponding reported value (279.9 eV). Ru/SiO2 and Ru/SiO2-
Al O showed similar peaks at 280.5 and 280.7 eV, respectively.
2
2
3
2
2
2
2
3
2
MgO) and activated carbon (AC) were employed as catalyst
supports of ruthenium. XRD analysis (Supporting Information,
Figure S3) indicated the formation of metallic ruthenium parti-
cles with crystallite sizes of 30–40 nm for most oxides (SiO2,
Al O , CeO , ZrO , MgO). SiO -Al O and SiO -MgO afforded
2
3
The order of ruthenium electron-richness (Ru/SiO -Al O <Ru/
2
2
3
SiO !Ru/Al O ) is consistent with that of the basicity of the
2
2
3
supports, suggesting that a basic support makes ruthenium
more electron-rich. However, this order is not consistent with
the order of H2 yields (Ru/SiO -Al O !Ru/SiO ꢁRu/Al O ),
2
3
2
2
2
2
3
2
ruthenium particles of ca. 15 nm in size. A much smaller crys-
tallite size (ca. 2 nm) was obtained when AC was used. As
shown in Figure S4 (Supporting Information), the difference in
the crystallite sizes can be attributed to the specific surface
area of the supports. We also carried out transmission electron
microscopy (TEM) analysis for ruthenium catalysts having dif-
ferent crystallite sizes (Ru/Al O , Ru/SiO -Al O , and Ru/AC).
2
2
3
2
2
3
which indicates that the electronic state of ruthenium is not
the decisive factor for the catalytic performance.
It is important not only to develop an efficient catalytic
system, but also to understand the nature of the catalysis for
further improvement of the catalytic performance. Therefore,
we subsequently performed a mechanistic study on the reac-
tion pathway of urea hydrolysis over ruthenium-based cata-
lysts. As shown in Figure 2, urea conversions of higher than
90% were obtained even with oxide supports themselves
(CeO and SiO -Al O ). Moreover, a comparable urea conversion
2
3
2
2
3
Particle sizes of ruthenium were ca. 40, 10–20, and 2 nm, re-
spectively (Supporting Information, Figure S5), being consis-
tent with their crystallite sizes.
The catalytic performances of the prepared ruthenium-
based catalysts in HPAU are shown in Figure 2. As a control ex-
periment, the reaction was also performed with bulk rutheni-
um, CeO , and SiO -Al O and without any catalyst. Ru/MgO,
2
2
2
3
was shown in the absence of any catalyst. However, the prod-
ucts were NH and CO , hence no H was yielded in these
3
2
2
cases. This indicates that the reaction mainly consists of urea
2
2
2
3
Ru/SiO -MgO, and Ru/AC gave moderate H2 yields, whereas
hydrolysis (H NCONH +H O!2NH +CO ) and subsequent
2
2
2
2
3
2
Ru/SiO , Ru/ZrO , Ru/Al O , and Ru/CeO afforded higher than
NH decomposition (2NH !3H +N ) and that ruthenium is
2
2
2
3
2
3
3
2
2
8
6 mol% H yields. On the other hand, Ru/SiO -Al O and bulk
necessary for the second step. Note that some side reactions
yielding CO may be involved in the whole reaction, such as
direct urea decomposition (H NCONH !2H +N +CO) and
2
2
2
3
ruthenium showed very low H yields with yielding significant
2
amounts of CO. The highest H yield, 92 mol%, was obtained
2
2
2
2
2
by Ru/Al O with by-production of only 1.8 mol% CO. This cat-
the reverse water-gas shift reaction (RWGS, CO +H ÐCO+
2
3
2
2
alytic performance was maintained at least for 2 h of on-
H O). To understand the involvements of these reactions, we
2
stream (data not shown).No obvious dependence of H or CO
carried out several control reactions. Figure 3 shows CO yields
in RWGS and CO conversion in the water-gas shift (WGS) reac-
tion over several ruthenium catalysts. CO yields in HPAU are
also indicated for comparison. No correlation was observed be-
tween the CO yields in RWGS (RWGS activities) and those in
HPAU, suggesting that RWGS does not contribute to the pro-
duction of CO. On the other hand, the WGS activity negatively
correlated with CO yields in HPAU, indicating that there are
two pathways yielding and consuming CO; the latter corre-
sponds to WGS and the former would be direct urea decom-
2
yield was observed on acid-base property of the support and
crystallite size of ruthenium. It should be noted that the ruthe-
nium-based catalytic system provides higher H yield with min-
2
imum release of CO at much lower reaction temperature than
[
7]
the Ni-based system. Moreover, no hydrocarbon species such
as CH were detected for any of the ruthenium-based catalysts.
4
We subsequently performed an X-ray photoelectron spectros-
copy (XPS) analysis to characterize the electronic states of the
ruthenium-based catalysts. Figure S6 (Supporting Information)
shows Ru 3d XPS results of Ru/SiO , Ru/Al O , and Ru/SiO -
position. We also studied NH decomposition in the presence
2
2
3
2
3
Al O . Ru/Al O exhibited a 3d5/2 emission peak assigned to
of CO , assuming complete urea hydrolysis and no direct de-
2
3
2
3
2
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
2
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
In summary, we develop an efficient catalytic system for hy-
drogen production from aqueous solutions of urea (HPAU).
Among various transition metals supported on SiO , Ru/SiO
2
2
shows a much higher H yield and forms smaller amounts of
2
CO as by-product. Optimization of the catalyst support reveals
that Ru/Al O exhibits the highest H yield (92 mol%) with min-
2
3
2
imum by-production of CO (1.8 mol%, corresponding to
.36 vol%) at 5008C. For applications in polymer electrolyte
fuel cells (PEFCs), further improvements to reduce the CO con-
0
[10]
centration to ppm levels will be required. However, the ach-
ieved level of residual CO (0.36 vol%) is tolerable for fuel cells
Figure 3. CO conversion in WGS reaction and CO yields in RWGS reaction
and HPAU.
[11]
other than PEFC ones, such as phosphoric acid,
solid
[12]
[12]
oxide,
and molten carbonate fuel cells.
A mechanistic
study suggests that urea is predominantly hydrolyzed into NH3
and CO by the catalyst support, followed by NH decomposi-
2
3
tion over ruthenium to produce H . A part of the urea is direct-
2
ly decomposed into CO and N H and the subsequent WGS re-
x
y
action effectively reduces CO. The insights obtained in the
present study not only indicate the potential applicability of
urea as an H carrier but also open a new horizon of the effec-
2
tive H production system.
2
Keywords: ammonia · hydrogen · heterogeneous catalysis ·
ruthenium · urea
Scheme 1. Proposed reaction pathways involved in HPAU. Reactions (1)+(2)
indicate the main pathways of HPAU.
composition of urea. In this condition, only the equilibrium
amount of CO for WGS was generated for various ruthenium-
based catalysts compared to the case of HPAU (Supporting In-
formation, Figure S7). This result supports the notion that
direct urea decomposition is predominantly involved in the
formation of CO.
[1] J. M. Ogden, Annu. Rev. Energy Environ. 1999, 24, 227–279.
[
2] a) T. V. Choudhary, C. Sivadinarayana, D. W. Goodman, Catal. Lett. 2001,
2, 197–201; b) Y. Liu, H. Wang, J. Li, Y. Lu, H. Wu, Q. Xue, L. Chen, Appl.
7
Catal. A 2007, 328, 77–82; c) G. Li, M. Kanezashi, T. Tsuru, Catal.
Commun. 2011, 15, 60–63; d) F. Hayashi, Y. Toda, Y. Kanie, M. Kitano, Y.
Inoue, T. Yokoyama, M. Hara, H. Hosono, Chem. Sci. 2013, 4, 3124–
3
2
130; e) Z. Lendzion-Bielu n´ , W. Arabczyk, Catal. Today 2013, 212, 215–
19.
On the basis of these results, we propose the following reac-
tion Scheme for HPAU (Scheme 1). At first, urea is hydrolyzed
into NH and CO (1), which is catalyzed mainly by the support
[
3] a) S. K. Singh, X. B. Zhang, Q. Xu, J. Am. Chem. Soc. 2009, 131, 9894–
895; b) S. K. Singh, Q. A. Xu, Chem. Commun. 2010, 46, 6545–6547;
c) S. K. Singh, A. K. Singh, K. Aranishi, Q. Xu, J. Am. Chem. Soc. 2011, 133,
9
3
2
19638–19641; d) S. K. Singh, Z. H. Lu, Q. Xu, Eur. J. Inorg. Chem. 2011,
2232–2237; e) S. K. Singh, Y. Iizuka, Q. Xu, Int. J. Hydrogen Energy 2011,
36, 11794–11801.
itself. Subsequent NH decomposition generates H and N (2).
3
2
2
As a side reaction, CO is released by direct urea decomposition
(
(
3). This reaction should produce some hydronitrogen species
N H ), which will also be decomposed into N and H (4). The
[
[
4] a) M. Chandra, Q. Xu, J. Power Sources 2006, 156, 190–194; b) F. H. Ste-
phens, V. Pons, R. T. Baker, Dalton Trans. 2007, 2613–2626; c) F. Y.
Cheng, H. Ma, Y. M. Li, J. Chen, Inorg. Chem. 2007, 46, 788–794; d) M.
Zahmakıran, S. Ozkar, Appl. Catal. B 2009, 89, 104–110.
5] a) Y. G. Wang, N. Shah, G. P. Huffman, Energy Fuels 2004, 18, 1429–1433;
b) Y. Okada, E. Sasaki, E. Watanabe, S. Hyodo, H. Nishijima, Int. J. Hydro-
gen Energy 2006, 31, 1348–1356; c) S. Yolcular, O. Olgun, Catal. Today
x
y
2
2
formed CO is converted into CO through WGS reaction (5).
2
The presence of excess amounts of H O drives the WGS reac-
2
tion. Thus, by-production of CO is effectively reduced to the
level of WGS equilibrium, provided the WGS activity is suffi-
ciently high. However, when the catalyst possesses a low WGS
activity, such as Ru/SiO -Al O and bulk ruthenium, significant
2
008, 138, 198–202.
6] S. Satyapal, J. Petrovic, C. Read, G. Thomas, G. Ordaz, Catal. Today 2007,
20, 246–256.
[
[
[
[
2
2
3
1
amounts of CO remain as by-product. The observed high cata-
lytic performances of the ruthenium catalysts are probably re-
lated to the well-known high activity of ruthenium towards
7] A. N. Rollinson, G. L. Rickett, A. Lea-Langton, V. Dupont, M. V. Twigg,
Appl. Catal. B 2011, 106, 304–315.
8] V. Mazzieri, N. Figoli, F. C. Pascual, P. L’Argentiere, Catal. Lett. 2005, 102,
79–82.
[
8]
NH decomposition. However, it should be noted that several
3
9] S. F. Yin, B. Q. Xu, X. P. Zhou, C. T. Au, Appl. Catal. A 2004, 277, 1–9.
points differ from the case of NH decomposition or synthesis:
3
[10] A. Faur Ghenciu, Curr. Opin. Solid State Mater. Sci. 2002, 6, 389–399.
[11] R. H. Song, D. R. Shin, Int. J. Hydrogen Energy 2001, 26, 1259–1262.
[12] A. B. Stambouli, E. Traversa, Renewable Sustainable Energy Rev. 2002, 6,
(
1) Although NH decomposition activity generally depends on
3
[9]
the basicity of the support (electron-richness of ruthenium),
433–455.
no dependence of the H yield on acid-base property of the
2
support and electron-richness of ruthenium was observed in
our system. (2) WGS activity is involved in the inhibition of CO
by-production.
Received: January 21, 2015
Revised: March 5, 2015
Published online on && &&, 0000
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
3
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

COMMUNICATIONS
S. Furukawa, R. Suzuki, K. Ochi,
T. Yashima, T. Komatsu*
New to the urea: An efficient catalytic
system for hydrogen production from
aqueous solution of urea was devel-
oped using ruthenium-based catalysts
as an alternative for the existing system
&& – &&
Hydrogen Production from Aqueous
Solutions of Urea with Ruthenium-
based Catalysts
based on several H
gave the highest catalytic performance,
2 mol% of H yield at 5008C. A mecha-
carriers. Ru/Al O
2
2 3
9
2
nistic study revealed that the reaction
proceeds mainly via two steps; urea hy-
drolysis (H NCONH +H O!2NH +CO )
2
2
2
3
2
and the following NH decomposition
3
(
2NH !3H +N ).
3
2
2
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
4
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!