High-quality hydrogen from the catalyzed decomposition of formic acid by Pd-Au/C and Pd-Ag/C

Article abstract of DOI:10.1039/b803661f

Pd-Au/C and Pd-Ag/C were found to have a unique characteristic of evolving high-quality hydrogen dramatically and steadily from the catalyzed decomposition of liquid formic acid at convenient temperature, and further this was improved by the addition of C

Full text of DOI:10.1039/b803661f

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
High-quality hydrogen from the catalyzed decomposition of formic acid
by Pd–Au/C and Pd–Ag/Cw
Xiaochun Zhou, Yunjie Huang, Wei Xing,* Changpeng Liu, Jianhui Liao and Tianhong Lu
Received (in Cambridge, UK) 4th March 2008, Accepted 24th April 2008
First published as an Advance Article on the web 29th May 2008
DOI: 10.1039/b803661f
Pd–Au/C and Pd–Ag/C were found to have a unique character-
istic of evolving high-quality hydrogen dramatically and steadily
from the catalyzed decomposition of liquid formic acid at
convenient temperature, and further this was improved by the
addition of CeO2(H2O)x.
dramatically and steadily from the decomposition of formic
acid at low temperature. The alloy catalysts were also found to
be greatly promoted by CeO2(H2O)x.
Through a large number of filtrations, elements of the
copper group, especially Ag and Au, were found to be
dramatically effective in retaining the Pd activity for the
decomposition of formic acidz in reaction (1), which is shown
in Fig. 1 and Fig. S1 (ESIw). The experiment was performed to
measure the volume of reforming gas (H2 + CO2) during 2 h
with 30 mg Pd–Me/C (20 wt% Pd, nPd : nMe = 3 : 1, Me = Cu,
Ag and Au) at 365 K. For the Pd/C catalyst, reforming gas
was only evolved for the first few minutes and then the
catalytic decomposition reaction stopped. Although the vo-
lume of reforming gas of Pd/C is larger than Pd–Cu/C during
2 h, the reaction rate of Pd–Cu/C is not zero but 0.074 mL
minꢀ1. However, the volumes of reforming gas from Pd–Au/C
and Pd–Ag/C increase sharply at the beginning and then
continue to increase linearly with fast rate. Interestingly,
Pd–Au/C gives a maximum output of 260 mL reforming gas
among the four catalysts during 2 h. The original reaction rate
is extremely fast and about 80 mL reforming gas is evolved in
the first 5 min, which is 31% of the total reforming gas in 2 h.
The volume of reforming gas from Pd–Ag/C is only little
smaller than that of Pd–Au/C. The turnover numbers (TON)
of Pd–Au/C and Pd–Ag/C after the first 10 min are 27 and
17 hꢀ1, respectively. Table S1 (ESIw) shows detailed data of
the four catalysts after the first 10 min. The intercepts in Fig. 1
clearly reflect the original reaction activity of the catalysts, and
the slope of the line shows the steady reaction activity after
10 min. Consequently, both the original reaction activity and
Hydrogen is a promising source of energy having profound
applications in various fields due to its high energy density and
efficiency with low environmental load.1 Its application in
proton exchange membrane fuel cells (PEMFCs) is receiving
much attention recently.2 Generally, there are two ways to
supply hydrogen for transportable energy applications, one of
which is hydrogen storage technologies,3–5 while the other is
reforming hydrogen from small organic molecules such as
methanol and ethanol.6–8 Though an outstanding progress
has been achieved in these routes mentioned above, there are
some inherent shortcomings of them,3–8 such as low efficiency,
high operating temperature, huge volume and weight loading,
excessive CO content as well as inconvenience, preventing use
for applications. On the other hand, the energy densities of
small organic molecules is higher than hydrogen but their
electrochemical reactivities are much lower than that of hy-
drogen.9–11 The application of small organic molecules as fuels
is also limited at present. Obviously, it is of high significance to
evolve hydrogen from such small organic molecules with
convenient methods.6–8 Noteworthy, it was found that formic
acid can decompose to hydrogen and carbon dioxide at
convenient temperatures.12–15 In the literature, the decomposi-
tion of formic acid was catalyzed by two types of catalysts, Pd
catalysts,12–14 and those of other metal complexes.15 The
decomposition of formic acid follows two principal path-
ways,12
HCOOH - CO2m + H2m: DG = ꢀ48.4 kJ molꢀ1 (1)
HCOOH - COm + H2O: DG = ꢀ28.5 kJ molꢀ1 (2)
where reaction (1) producing hydrogen is the expected and
desired reaction. However, reaction (2) producing carbon
monoxide is a side-reaction. Unfortunately, those Pd/C cata-
lysts deactivate quickly due to the poisoning intermediates,
resulting in its failing in applications.12–14 In this paper, we
synthesised Pd–Au/C and Pd–Ag/C alloy catalysts, which
overcame the poisoning and evolved high-quality hydrogen
State Key Laboratory of Electroanalytical Chemistry, Changchun
Institute of Applied Chemistry, Chinese Academy of Sciences, 5625
Renmin StreetChangchun, 130022, PR China.
E-mail: xingwei@ciac.jl.cn
w Electronic supplementary information (ESI) available: Additional
characterization data. See DOI: 10.1039/b803661f
Fig. 1 The output volume of reforming gas (H2 + CO2) during 2 h
with 30 mg Pd–Me/C (20 wt% Pd, nPd : nMe = 3 : 1, Me = Cu, Ag and
Au), respectively, at 365 K and 5.00 ml 9.94 M formic acid–3.33 M
sodium formate solution.
ꢁc
This journal is The Royal Society of Chemistry 2008
3540 | Chem. Commun., 2008, 3540–3542

View Online
the steady reaction activity of Pd–Au/C are the highest ones
among the four catalysts.
Transmission electron micrographs (TEM) of Pd/C, Pd–Ag/
C and Pd–Au/C are shown in Fig. S2 (ESIw), and particle
distribution histograms of these catalyst samples were ob-
tained by measuring the metal particles shown in Fig. S3
(ESIw) for the nanoparticles deposited on the carbon support.
The average sizes of the three catalysts can be ranked as Pd/C
4 Pd–Ag/C 4 Pd–Au/C. The TEM results show that the
addition of Ag and Au reduces the size of nanoparticles.
Furthermore, the effect of Au is better than Ag, which may
be one of the reasons for the better performance of Pd–Au/C
compared to Pd–Ag/C.
Considering the moderate reaction rate and experimental
convenience, Pd–Ag/C was selected as a model catalyst to
study the effects of different factors in detail. First, the atom
proportions of Pd : Ag was found to have great influence on
the catalyst activity with a saddle shape as shown in Fig. S4
(ESIw). The highest activity appears at 3 : 3 (20 wt% Ag), and
Fig. 3 Lifetime test of 10 mg Pd–Ag/C (20 wt% Pd, nPd : nAg = 3 : 1)
during 240 h at 365 K with 60.0 mL 9.94 M formic acid–3.33 M
sodium formate solution.
tion of carbon monoxide is 80 ppm. Remarkably, the con-
centration is dramatically lower than that of conventional
reforming gas from methanol, ethanol, methane etc.,6–8 and
the quality of the hydrogen reaches PEMFC standard with the
CO concentration lower than 100 ppm.16
the reaction rate is 160 ml minꢀ1 gꢀ1 (Pd), and TON is 22 hꢀ1
.
Second, the dependence of the reaction rate on temperature
was also studied as shown in Fig. S5 (ESIw). The reaction rate
is doubled every 5.0 K and according to the Arrhenius
equation (3),
The lifetime of Pd–Ag/C shown in Fig. 3 had been tested
with 10 mg catalyst for 240 h at 365 K. The activity of Pd–Ag/
C and the reaction rate indicated high stability during the test.
Furthermore, the linear fit for the data shows that the activity
of Pd–Ag/C increases slightly with a rate of 1.59 ꢂ 10ꢀ3 ml
ln k = ꢀ(Ea/RT) + B
(3)
minꢀ1 gꢀ1 ꢀ1. The high stability is due to the fact that the
h
the activation energy for the decomposition reaction of formic
acid is calculated to be Ea = 115 kJ molꢀ1 as shown in Fig. S6
(ESIw). This indicates that the working temperature plays an
important role in the catalytic decomposition due to the
difference in the adsorption and desorption of poisoning
intermediates at different temperatures. Third, the effect of
proportion of formate/(formic acid and formate) on both
reforming rate and concentration of CO in reforming gas
was accurately measured using Fourier-transform infrared
absorption spectrometry (FTIR). A plot has been constructed
between CO content and the corresponding proportions as
shown in Fig. 2. The maximum appears at the proportion 0.3.
The reaction rate is a saddle shape14 with the maximum rate
300 ml minꢀ1 gꢀ1 at the proportion 0.7, where the concentra-
HCOONa solution is basic, while Pd and Ag are stable at
relatively higher pH value. The solution containing HCOOH
and HCOOꢀ as reducing agent is important to keep the Pd
and Ag in a reduced state, and consequently retains the
stability of the alloy catalyst Pd–Ag/C. In addition, we also
found that the activities of the catalysts are renewed comple-
tely after drying in air at 353 K. At this condition, poisonous
intermediates on Pd surface will be oxidized by air.
The characteristics of the elements of copper group, espe-
cially Ag and Au, to retain the catalytic activity of Pd for the
decomposition of formic acid are discussed primarily herein. It
has been reported that the reason for the disappearance of
activity of Pd/C is due to the hydrogen adsorption on Pd.13
Nevertheless, in our experiment, the catalyst activity only
decreased slightly after bubbling hydrogen into an aqueous
suspension of Pd/C, while it decreased completely after bub-
bling CO into an aqueous suspension of Pd/C. Obviously,
besides the effect of hydrogen adsorption on Pd,13 poisoning
intermediates (such as COad) produced by reaction (2) cover
the surface of the catalysts and occupy the active sites.17–19
Additionally, the long lifetime of catalysts Pd–Ag/C and
Pd–Au/C in the environment of CO implies that CO cannot
steadily adsorb on their surface even at 365 K. According to
some previous researches, the adsorption strength of CO on
metal surfaces ranks as Pd 4 Cu 4 Ag 4 Au,18 and Ag and
Au do not form stable complexes with CO.20 When Pd is
alloyed with Cu, Ag or Au as shown by XRD experiments in
Fig. S7 (ESIw), they effectively inhibit the adsorption of CO on
Pd.16,20–22 The occupied active sites are released due to CO
desorbing from the Pd surface,16,20–22 and consequently reac-
tion (1) is not inhibited. Thus, the concomitant CO poisoning
Fig. 2 Effect of proportion of formate to formate and formic acid on
the decomposition rate of formic acid and CO concentration in
reforming gas with 30 mg Pd–Ag/C (20 wt% Pd, nPd : nAg = 3 : 1)
at 365 K and 5.00 ml solution: (’) concentration of CO, (m) output
rate of reforming gas (H2 + CO2).
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 3540–3542 | 3541

View Online
hydrogen at low temperature, which is of high potential for
transportable, especially portable, hydrogen sources. The
breakthrough of the advanced catalysts in preventing CO
poisoning is a valuable contribution for advancement of
oxidation and between CO content and the corresponding
proportion of electro-oxidation of small organic molecules.16
The authors are grateful for the financial sponsors of State
Key Fundamental Research Program of China, (973 Program,
G2000026408), State Key High Technology Research
Program of China (863 Program, 2001AA323060,
2003AA517062, 2006AA05Z137), Nature Science Foundation
of China (20373068, 20433060). The authors are also grateful
for the help of Dongmei Wang, Liaohai Ge, Xinglin Li,
Guangfu, Zeng, Zhen Zhang, Xiangguang Yang, Fei Li and
Debashis Panda.
Fig. 4 The output volume of reforming gas at 365 K with 5.00 ml
9.94 M formic acid–3.33 M sodium formate solution. (Solid line) 60
mg PdAu/C–CeO2 10 wt% Pd, 50% CeO2, nPd : nAu = 1 : 1; (dashed
line) 60 mg PdAg/C–CeO2 10 wt% Pd, 50% CeO2, nPd : nAg = 1 : 1;
(dotted line) 30 mg PdAg/C 20 wt% Pd, nPd : nAg = 1 : 1.
Notes and references
z CAUTION: formic acid is harmful to skin or eyes and thus it is
important to prevent the leakage of formic acid during the experiments
and applications.
process during decomposition of small organic molecules has
been removed effectively by the Pd alloyed catalysts with
Au or Ag.
1. P. Seungdoo, M. V. John and J. G. Raymond, Nature, 2000, 404,
265.
2. P. Costamagna and S. Srinivasan, J. Power Sources, 2001, 102,
242.
In addition, the activities of Pd–Au/C and Pd–Ag/C here
can be enhanced enormously by co-deposition with CeO2-
(H2O)x (simply written as CeO2) as shown in Fig. 4. It can be
seen that the activities of 60 mg Pd–Me/C–CeO2 (10 wt% Pd,
50 wt% CeO2, Me = Au and Ag, nPd : nMe = 3 : 3) reach 1640
3. L. Schlapbach and A. Zuttel, Nature, 2001, 414, 353.
¨
4. J. Chen, N. Kuriyama, Q. Xu, H. T. Takeshita and T. Sakai,
J. Phys. Chem. B, 2001, 105, 11214.
5. D. Sun, S. S. Srinivasan, T. Kiyobayashi, N. Kuriyama and
C. M. Jensen, J. Phys. Chem. B, 2003, 107, 10176.
6. W. C. Shin and R. S. Besser, J. Power Sources, 2007, 164, 328.
7. G. A. Deluga, J. R. Salge, L. D. Schmidt and X. E. Verykios,
Science, 2004, 303, 993.
8. D. L. Hoang, S. H. Chan and O. L. Ding, J. Power Sources, 2006,
159, 1248.
9. E. Antolini, J. R. C. Salgado and E. R. Gonzalez, J. Electroanal.
Chem., 2005, 580, 145.
and 548 ml minꢀ1 ꢀ1, or 10.3 and 3.4 times that of 30 mg
g
Pd–Ag/C (20 wt% Pd, nPd : nAg=3 : 3), respectively. The
TONs of Pd–Au/C–CeO2 and Pd–Ag/C–CeO2 are 227 and 76
ꢀ1, respectively. With increasing the working temperature by
h
10 degrees, the reaction rate for Pd–Au/C–CeO2 became
extremely fast (46000 ml minꢀ1 ꢀ1, TON up to 832 hꢀ1),
g
which shows high potential in transportable application. Fig.
S8 (ESIw) shows the TEM of Pd–Ag/C–CeO2 and Pd–Au/
C–CeO2 catalysts. Looking at Fig. S2, the dispersion of
nanoparticles of Pd–Ag is improved by adding CeO2, while
it becomes poor for Pd–Au. This result implies that a higher
dispersion is not the key fact for the improvement of activity.
The promotion of CeO2 on the activity of Pd-based catalysts is
probably due to that the CeO2 produces cationic palladium
species, which show high activity in CO oxidation23 and
methanol decomposition.24 Another reason is due to that
CeO2(H2O)x on the Pd surface can induce the decomposition
of formic acid by a more efficient route, in which fewer
poisoning intermediates will be produced. Because of the
lowest adsorption strength of CO on Au surfaces, Au is most
effective to reduce the adsorption of CO on Pd surface.18 This
effect is further strengthened by the addition of CeO2. Pd–Au/
C–CeO2 thereby become much more active than Pd–Ag/
C–CeO2.
The yield of 1640 ml minꢀ1 gꢀ1 reforming gas indicates a
theoretical power density of 135 W gꢀ1 for PEMFC, which can
being further improved by increasing activity and decreasing
the working temperature by optimising the catalyst composi-
tion and working conditions. The main contribution of the
work is to inhibit the CO poisoning, and consequently provide
a new route to rapidly and conveniently evolve high quality
10. C. Rice, S. Ha, R. I. Masel and A. Wieckowski, J. Power Sources,
2003, 115, 229.
11. L. Colmenares, H. Wang, Z. Jusys, L. Jiang, S. Yan, G. Q. Sun
and R. J. Behm, Electrochim. Acta, 2006, 52, 221.
12. N. Akiya and P. E. Savage, AIChE J., 1998, 44, 405.
13. D. M. Ruthven and R. S. Upadhye, J. Catal., 1971, 21, 39.
14. S. P. Hill and J. M. Winterbottom, J. Chem. Technol. Biotechnol.,
1988, 41, 121.
15. (a) Y. Gao, J. Kuncheria, G. P. A. Yap and R. J. Puddephatt,
Chem. Commun., 1998, 2365; (b) R. S. Coffey, Chem. Commun.,
1967, 18, 923a; (c) J. R. Hyde and M. Poliakoff, Chem. Commun.,
2004, 1482; (d) K. Abdur-Rashid, S. E. Clapham, A. Hadzovic,
J. N. Harvey, A. J. Lough and R. H. Morris, J. Am. Chem. Soc.,
2002, 124, 15104.
16. T. J. Schmidt, Z. Jusys, H. A. Gasteiger, R. J. Behm,
U. Endruschat and H. Boennemann, J. Electroanal. Chem.,
2001, 501, 132.
17. F. S. Thomas and R. I. Masel, Surf. Sci., 2004, 573, 169.
18. F. Abild-Pedersen and M. P. Andersson, Surf. Sci., 2007, 601,
1747.
19. C. Xu and D. W. Goodman, J. Phys. Chem., 1996, 100, 245.
20. K. Judai, S. Abbet, A. S. Worz, U. Heiz, L. Giordano and
¨
G. Pacchioni, J. Phys. Chem. B., 2003, 107, 9377.
21. V. Joha
15.
22. S. Karski, I. Witon
Catal. A: Chem., 2005, 240, 155.
nek, N. Tsud, V. Matolın and I. Stara, Vacuum, 2001, 63,
´ ´ ´
´
ska, J. Rogowski and J. Go"uchowska, J. Mol.
23. G. Glaspell, L. Fuoco and M. S. El-Shall, J. Phys. Chem. B, 2005,
109, 17350.
24. W.-J. Shen and Y. Matsumura, J. Mol. Catal. A: Chem., 2000,
153, 165.
ꢁc
This journal is The Royal Society of Chemistry 2008
3542 | Chem. Commun., 2008, 3540–3542