CO-free hydrogen production via dehydrogenation of a Jet A hydrocarbon mixture

Article abstract of DOI:10.1016/j.jcat.2007.11.009

The dehydrogenation of a Jet A hydrocarbon mixture was investigated for the production of hydrogen for fuel cell applications. This process differs from steam/partial oxidation reforming of hydrocarbons in that the hydrogen produced is free of CO, and therefore can be used in low temperature fuel cells without further purification. Because of their relative importance with respect to the spectrum of hydrocarbons in the mixture, cyclohexane and decalin were investigated independently. A hybrid catalyst, Pt/γ-Al₂O₃-ZrO₂/SO^2-₄, was designed to enhance the hydrogen production yields.

Full text of DOI:10.1016/j.jcat.2007.11.009

Journal of Catalysis 253 (2008) 239–243
www.elsevier.com/locate/jcat
CO-free hydrogen production via dehydrogenation of a Jet A
hydrocarbon mixture
Bo Wang a, Gilbert F. Froment a, D. Wayne Goodman b,∗
a
Department of Chemical Engineering, Texas A&M University, College Station, TX 77843-3122, USA
b
Department of Chemistry, Texas A&M University, College Station, TX 77842, USA
Received 8 August 2007; revised 9 November 2007; accepted 13 November 2007
Abstract
The dehydrogenation of a Jet A hydrocarbon mixture was investigated for the production of hydrogen for fuel cell applications. This process
differs from steam/partial oxidation reforming of hydrocarbons in that the hydrogen produced is free of CO, and therefore can be used in low
temperature fuel cells without further purification. Because of their relative importance with respect to the spectrum of hydrocarbons in the
2−
mixture, cyclohexane and decalin were investigated independently. A hybrid catalyst, Pt/γ -Al O -ZrO /SO , was designed to enhance the
hydrogen production yields.
2
3
2
4
© 2007 Elsevier Inc. All rights reserved.
Keywords: Hydrogen; Dehydrogenation; Fuel cell; Cyclohexane; Decalin; Jet A
1. Introduction
[8–12]. The absence of carbon oxides eliminates the need for
further purification of the hydrogen product. Another advan-
tage of using liquid hydrocarbons as a H2 storage medium is
that the current distribution infrastructure can be used for the
transportation, storage, and dispersal of the liquid hydrocar-
bon. Since the dehydrogenation of cycloalkanes is a reversible
process, the dehydrogenated products (benzene, toluene, naph-
thalene, etc.) can be easily hydrogenated back to the parent
cycloalkane at local nodes in a hydrogen distribution system.
Since jet fuels are widely used in large quantities for mili-
tary applications, this fuel is an attractive source of hydrogen
for fuel cells in this environment. Steam/autothermal reform-
ing of these hydrocarbons to produce hydrogen involves the
co-production of large amounts of CO [13]. Typically used low
temperature fuel cells, e.g. proton exchange membrane (PEM)
fuel cells, are highly sensitive to poisoning by CO (tolerance is
at ppm levels). Several post-reforming procedures are required
for remediation of CO. These additional procedures are detri-
mental to the process economics. More importantly, the CO
removal treatments increase the complexity and bulkiness of
the PEM system and effectively minimize the main advantage
(compactness and weight) offered by PEMs. Accordingly fuel
reforming to produce CO-free hydrogen is currently the most
important problem associated with PEM fuel cells. It is there-
Onboard hydrogen storage imposes a serious challenge to
the implementation of fuel cells in transportation [1,2]. For
this purpose chemical hydrides have been explored for in situ
generation of hydrogen. Hydrolysis, reforming and dehydro-
genation are the most widely used reactions to produce hy-
drogen from hydrocarbon sources. Hydrolysis reactions utilizes
the oxidation of chemical hydrides with water to produce hy-
drogen [3,4]. Obviously water is needed onboard the vehicle
and thus increases the weight and decreases the useful vehi-
cle volume. Hydrolysis for hydrogen production has directed
considerable efforts toward sodium borohydride as the hydro-
gen storage medium. Steam reforming of natural gas is mainly
used for large scale hydrogen production in industry [5–7]. The
reaction is highly endothermic and hence requires a substan-
tial energy input. Another drawback is that steam reforming
generates CO which poisons the typical catalysts used in low
temperature fuel cells. Dehydrogenation of hydrocarbons is an
attractive method for hydrogen production since only hydrogen
and dehydrogenated hydrocarbons are formed in this reaction
*
Corresponding author.
E-mail address: goodman@mail.chem.tamu.edu (D.W. Goodman).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.11.009

240
B. Wang et al. / Journal of Catalysis 253 (2008) 239–243
fore desirable to explore other avenues for hydrogen production
with specific applications for the current PEM fuel cells. Hy-
drogen production via dehydrogenation of a Jet A hydrocarbon
mixture is an attractive route for the production of CO-free hy-
drogen and is investigated in this study for the first time. As
an added bonus, for fuel cells used in conjunction with inter-
nal combustion or turbine driven vehicles, the dehydrogenation
reaction can be carried out at or near the accessible exhaust tem-
peratures, i.e., the waste heat from the exhaust can be used to
drive the reaction. Preliminary data show that limited dehydro-
genation of Jet fuel does not significantly compromise its fuel
properties, thus allowing the fuel with its current composition
to serve as a hydrogen storage medium.
The dehydrogenation of hydrocarbons to produce hydrogen
has been reported on a number of supported metal catalysts in
different types of reactors. Ni and Pt are among the most used
metal catalysts. While Ni is much cheaper than Pt, the selectiv-
ity for the dehydrogenation reaction is much lower than that on
Pt, and, in particular, cracking reactions are pronounced on Ni
at high temperatures. Ichikawa et al. investigated the dehydro-
genation of cyclohexane, methylcyclohexane and decalin over
activated carbon supported Ni, Pt, and Ni–Pt catalysts using
a spray pulse mode reactor operating at 287–375 ◦C [8,9]. It
was assumed that pulse spray operation could increase the re-
action rate by suppressing the reverse reaction. The conversions
were 25–35% with the selectivity for dehydrogenation reactions
above 98.8%. Okada et al. reported results for alumina sup-
ported Pt catalysts for methylcyclohexane dehydrogenation in
a fixed bed reactor [10]. This catalyst could generate hydrogen
from methylcyclohexane with a conversion of 95% and toluene
selectivity above 99.9% at 320 ◦C. Recently, Huffman used
stacked-cone carbon nanotubes (SC-CNT) as a support medium
to prepare Pt and Pd catalysts for the dehydrogenation of cy-
clohexane and methylcyclohexane [11]. The catalysts exhibited
100% selectivity for dehydrogenation reactions. It was reported
that a 0.25 wt% Pt/SC-CNT catalyst had approximately the
same activity as a commercial 1 wt% Pt/Al2O3 catalyst. The
regeneration of the deactivated, carbon-supported catalyst is a
problem, and thus limits its potential.
2−
Fig. 1. Preparation of a Pt/γ -Al O -ZrO /SO hybrid catalyst.
2
3
2
4
impregnation method [14]. Three different kinds of high surface
area supports, γ -alumina, silica and activated carbon, were in-
vestigated for preparation of the catalysts.
The Pt/alumina-ZrO2/SO42−hybrid catalyst was prepared
from two components, 0.9 wt% Pt/alumina and sulfated zir-
conia. The sulfated zirconia was prepared according to a pub-
lished protocol (Fig. 1) [15]. Zirconium nitrate was precipitated
with 14 N ammonium nitrate at pH 10. The precipitate was fil-
tered, washed, and then dried overnight at 110 ◦C. Zirconium
hydroxide was then ion exchanged with 1 N sulfuric acid and
filtered. The solid was dried overnight at 110 ◦C and calcined
at 600 ◦C in air for 3 h. The hybrid catalyst was made by phys-
ically mixing 0.9 wt% Pt/alumina and sulfated zirconia with a
mass ratio of 6 to 4, then pelletized to particle sizes ranging
from 0.15–0.5 mm. The catalyst was loaded into the reactor,
and activated by calcination in air at 500 ◦C for 8 h with a sub-
sequent reduction in hydrogen at 300 ◦C.
2.2. Fixed-bed reactor setup
The dehydrogenation of cyclohexane, decalin and Jet A were
carried out over supported Pt catalysts at atmospheric pres-
sure in a fixed bed reactor described elsewhere [14]. A HP-
PLOT/alumina capillary column with a FI-detector was used to
analyze the hydrocarbon components for the dehydrogenation
of cyclohexane. A HP-5 capillary column was used to analyze
the hydrocarbons for the dehydrogenation of decalin and Jet A.
In the present study the dehydrogenation of Jet A to pro-
duce hydrogen was investigated over Pt/γ -Al2O3 and Pt-Sn/γ -
Al2O3 catalysts and the operating conditions optimized for this
range of reaction conditions. Because of their overall relevance
to the reaction chemistry, the dehydrogenation of cyclohexane
and decalin was also investigated. To generate additional hy-
drogen from paraffins, the major components of Jet A, catalytic
reforming of paraffins to aromatics on a bifunctional catalyst is
desirable. Accordingly, a Pt/γ -Al2O3-ZrO2/SO2− hybrid cata-
lyst was synthesized to increase the acidity of t4he catalyst and
thus increase the hydrogen production at low temperature.
3. Results and discussion
3.1. Definition of conversion, selectivity, and hydrogen
production yield
The definitions for the subsequently quoted conversions, se-
lectivities, and hydrogen production yields are as follows:
Conversions:
moles of i in − moles of i out
Xi =
× 100.
moles of reactant in
Selectivity:
2. Experimental
moles of product i formed
selectivity =
× 100.
moles of reactant converted
2.1. Catalyst preparation
Hydrogen production yield in Jet A dehydrogenation:
The catalysts used in the dehydrogenation of cyclohxane and
decalin were supported Pt catalysts, prepared by an evaporative
H2 production rate in mol × 2
yield (%) =
× 100.
Jet A molar feed rate × 21

B. Wang et al. / Journal of Catalysis 253 (2008) 239–243
241
Fig. 4. Effect of H /cyclohexane molar ratio on cyclohexane dehydrogenation
2
◦
over 0.8 wt% Pt/γ -alumina. Reaction conditions: 320 C and space time =
25 kg cat h/kmol.
Fig. 2. Cyclohexane dehydrogenation over various supported Pt catalysts. Tem-
◦
perature = 340 C, space time = 20 kg cat h/kmol, and H /cyclohexane = 1/2.
2
Fig. 5. Catalyst stability in cyclohexane dehydrogenation to produce H over
2
Fig. 3. Effect of temperature on cyclohexane dehydrogenation over 0.8 wt%
0.8 wt% Pt/γ -alumina. Reaction conditions: space time = 20 kg cat h/kmol and
Pt/γ -alumina. Reaction conditions: space time = 30 kg cat h/kmol and H /
2
H /cyclohexane = 1/2.
2
cyclohexane = 3.
3.2. The dehydrogenation of cyclohexane in a fixed-bed
reactor
3.2.1. Effect of various supports
Three different kinds of high surface area supports, e.g., γ -
alumina, silica, and activated carbon, were used for the prepa-
ration of the catalysts. Fig. 2 shows that the catalytic activity
decreases in the order γ -alumina > silica > activated carbon.
The catalyst prepared with a γ -alumina support also shows
good stability, with no decrease of cyclohexane conversion at
98.7% after 20 h. Therefore, γ -alumina was chosen as the sup-
port for the Pt catalysts used in the study of the dehydrogenation
of cylohexane and decalin.
Fig. 6. Effect of Pt loadings on cyclohexane dehydrogenation. Reaction condi-
◦
tions: 320 C, space time = 20 kg cat h/kmol, and H /cyclohexane = 3.
2
was tested in a reactant stream with a hydrogen to cyclohexane
molar ratio of 1:2.
3.2.2. Effect of temperature and H2/cyclohexane molar ratio
The dehydrogenation of cyclohexane was carried out at 260,
280, 300, 320, 340, 360, 380, and 400 ◦C, respectively. It is
shown in Fig. 3 that at space time of 30 kg cat h/kmol the
conversions increase from 51.7% at 260 ◦C to 100% for temper-
atures above 360 ◦C. Furthermore, trace amounts of cracking
products such as methane and ethane were detected at 380 and
400 ◦C.
Fig. 4 shows the conversion of cyclohexane as a function
of the H2/cyclohexane molar ratio at 320 ◦C. The dehydrogena-
tion of cyclohexane is not thermodynamically limited at 320 ◦C,
therefore the effect of hydrogen is not significant. The conver-
sion of cyclohexane decreases only slightly with an increase in
the H2/cyclohexane molar ratio. The addition of hydrogen is
known to minimize coking during dehydrogenation processes
[16,17]. Accordingly, the stability of a Pt/γ -alumina catalyst
3.2.3. Stability of the Pt/γ -alumina catalyst
The dehydrogenation was carried out over 0.8 wt% Pt/γ -
alumina catalyst at a cyclohexane space time of 20 kg cat h/
kmol, a hydrogen to cyclohexane molar ratio of 1 to 2, and reac-
tion temperatures in the range 320–410 ◦C. As shown in Fig. 5,
the conversions of cyclohexane at 320 (10 h), 350 (16 h), 380
(23 h), and 410 ◦C (22 h) were 94.4, 99.5, 99.8, and 99.9%, re-
spectively, with no apparent catalyst deactivation.
3.2.4. Effect of Pt loading and of the addition of Sn
Fig. 6 indicates that the conversions of cyclohexane on four
Pt/alumina catalysts with 0.3, 0.5, 0.8, and 1.0 wt% of Pt load-
ings are 41.0, 87.2, 94.6, and 97.2%, respectively. Fig. 7 shows
that the addition of Sn to a Pt/alumina catalyst decreases the

242
B. Wang et al. / Journal of Catalysis 253 (2008) 239–243
Fig. 9. Hydrogen production yield as a function of temperature in the dehydro-
Fig. 7. Effect of Sn/Pt molar ratio on cyclohexane dehydrogenation. Reaction
genation of Jet A on Pt/γ -alumina catalyst. Jet A = 10 ml/min, H = 10 ml/
2
◦
conditions: 320 C, space time = 20 kg cat h/kmol, and H /cyclohexane = 3.
min, and N = 40 ml/min.
2
2
Fig. 10. Effect of H /Jet A molar ratio on the dehydrogenation of Jet A at
2
◦
500 C. Feeds were as follows: Jet A = 10 ml/min, H = 10 ml/min, and N
Fig. 8. Hydrogen production yield as a function of reaction temperature in the
dehydrogenation of Jet A surrogate over a Pt/γ -alumina catalyst. Feed of Jet A
surrogate = 10 ml/min, H = 10 ml/min, and N = 40 ml/min.
2
2
= 40 ml/min.
2
2
tained at 480 ◦C is 11.6%, lower than the yield obtained in
the dehydrogenation of pure cyclohexane and decalin, 47.5 and
52.0%, respectively. Gas chromatographic analysis of the reac-
tant and product streams showed that the conversions of paraf-
fins were low. If the Jet A paraffin components, which account
for 60 molar% in the Jet fuel composition, could be reformed to
aromatics as in industrial naphtha reforming, the H2 yield could
obviously be substantially increased.
catalyst activity, presumably by coverage of the active Pt sites.
However, it has been reported in catalytic reforming work that
the addition of Sn prevents the sintering of Pt clusters and im-
proves catalyst stability toward deactivation by coking [18].
3.3. Dehydrogenation of decalin
The conversion of a decalin mixture, consisting of 24% cis-
and 76% trans-isomers, was also investigated on a Pt/alumina
catalyst. Isomerization between the two decalin isomers occurs
on the acid γ -Al2O3 sites, whereas both isomers are dehydro-
genated on the Pt sites into an intermediate, tetralin, which,
in turn, is dehydrogenated to naphthalene thereby generat-
ing more H2. With appropriate space times and temperatures
very high selectivities for naphthalene and low selectivities for
tetralin can be achieved. These results along with the kinetic
modeling are detailed in another paper [14].
3.4.2. Dehydrogenation of Jet A on various catalysts
3.4.2.1. Pt/γ -alumina catalyst The Jet A used in this experi-
ment consists of 18% aromatics, 20% naphthenes, 60% paraf-
fins, and 2% olefins. Fig. 9 shows the hydrogen production yield
on a 0.8 wt% Pt–0.3 wt% Sn/alumina catalyst as a function of
temperature. The maximum yield obtained at 500 ◦C is 12.8%.
The low yield indicates that paraffins in Jet A are not reformed
to aromatics, probably because the acidity of γ -alumina is not
sufficiently high to promote this reaction. Also, the catalyst is
prone to deactivation in Jet A dehydrogenation. The hydrogen
yield, defined as hydrogen produced per total hydrogen content
in Jet A, decreases from 12.0 to 7.2% after 180 min (Fig. 10).
Due to the deactivation of the catalyst, the H2/Jet A molar ratio
was increased to 10. The yield of hydrogen decreased from 11.2
to 9.2% after 180 min. Subsequently, a USY zeolite with acidity
substantially higher than alumina was used as a support for Pt.
3.4. Dehydrogenation of Jet A
3.4.1. Dehydrogenation of Jet A surrogate on Pt/γ -Al2O3
catalyst
The Jet A surrogate blend used in this experiment has a mo-
lar average formula of C10H19 and consists of six components
with the following weight percentages: 30% decane, 35% do-
decane, 14% methylcyclohexane, 6% decalin, 10% t-butyl ben-
zene, and 5% 1-methyl naphthalene. Fig. 8 shows the hydrogen
yield from a Jet A surrogate via dehydrogenation as a function
of temperature. The hydrogen yield, defined as the hydrogen
produced per total hydrogen content in the fuel, increases with
temperature up to 480 ◦C, then falls. The maximum yield ob-
3.4.2.2. Pt/USY catalyst Fig. 11 compares the dehydrogena-
tion of Jet A over various catalysts. In industry, bifunctional
catalysts with both metal and acid sites are employed in cat-
alytic reforming to convert paraffins into aromatic compounds,
producing H2 as a by product. In this process the metal sites
are responsible for the dehydrogenation of paraffins. The acid

B. Wang et al. / Journal of Catalysis 253 (2008) 239–243
243
4. Conclusions
The dehydrogenation of a Jet A hydrocarbon mixture was
investigated over Pt/γ -Al2O3 and Pt/γ -Al2O3-ZrO2/SO24− for
the production of hydrogen for low temperature fuel cell ap-
plications. The distinct advantage of this process is that the
hydrogen so produced is free of CO and can be used in typi-
cal low temperature fuel cells without further purification. For
dehydrogenation of Jet A on Pt/γ -Al2O3, the conversions of
paraffins, the major component in Jet A, are low, most likely
because the acidity of γ -alumina is not sufficiently high. Bi-
functional catalysts with both metal and acid sites are essential
to increase the hydrogen yield for this application. The acidity
can be increased by the addition of a strong acid component,
sulfated zirconia, or by using a strong acid support. The strong
acid sites promote the cracking ability of Pt-Sn/γ -Al2O3 as well
as the transformation of paraffins to aromatics, leading to an in-
crease in the hydrogen production. The strong acidity may also
promote coke formation and thus to rapid deactivation of the
catalyst. Nevertheless, the deactivated catalyst can be regener-
ated by oxidation at 500 ◦C.
◦
Fig. 11. Dehydrogenation of Jet A over various catalysts at 425 C. Feeds were
as follows: Jet A = 10 ml/min, H = 10 ml/min, and N = 40 ml/min.
2
2
2−
Fig. 12. Regeneration of Pt-Sn/γ -Al O -ZrO /SO hybrid catalysts used in
2
3
2
4
Acknowledgments
the dehydrogenation of Jet A by oxidation in air at various temperatures. Run
1—fresh catalyst; Run 2—oxidation at 500 C in air, then reduction at 300 C
in H ; and Run 3—oxidation at 500 C in air, then reduction at 300 C in H .
◦
◦
◦
◦
2
2
The authors gratefully acknowledge the support of the De-
partment of Energy, Office of Basic Energy Sciences, Division
of Chemical Sciences/Geosciences/Biosciences, Catalysis and
Chemical Transformations Program, and the Robert A. Welch
Foundation.
sites are essential to transform these dehydrogenated long chain
hydrocarbon intermediates into cycloalkanes. Acid sites may
also crack the paraffins to generate H2. The acidity of USY is
higher than that of γ -Al2O3 and consequently H2 production
over Pt/USY at 425 ◦C is 2% higher than that on Pt/γ -Al2O3
catalyst. However, due to the relatively small channel openings
of USY, deactivation is much more rapid.
References
[1] H.L. Hellman, R.V. Hoed, Int. J. Hydrogen Energy 32 (2007) 305.
[2] J.H. Wee, K.Y. Lee, J. Power Sources 157 (2006) 128.
[3] Y. Shang, R. Chen, Energy Fuels 20 (2006) 2142.
[4] J.V. Ortega, Y. Wu, S. Amendola, M.T. Kelly, U.S. patent 6 586 563,
2003.
3.4.2.3. Pt-Sn/γ -Al2O3-ZrO2/SO24− hybrid catalysts Sulfated
zirconia is a well known superacid catalyst and can be in-
corporated into a Pt/γ -alumina catalyst to increase its acidity.
Pt-Sn/γ -alumina-ZrO2/SO24− hybrid catalysts were prepared by
calcining a physical mixture of Pt-Sn/γ -Al2O3 to ZrO2/SO42−
consisting of a mass ratio of 6 to 4 at 550 ◦C for 12 h. For identi-
cal operating conditions the hydrogen production yield over this
hybrid catalyst is 5% higher than achieved with Pt/γ -alumina.
This is clearly not an optimized catalyst, therefore there is great
potential for further optimization.
[5] J.R. Ross, Catal. Today 100 (2005) 151.
[6] X. Wang, R.J. Gorte, Appl. Catal. A 224 (2002) 209.
[7] M.A. Rosen, Int. J. Hydrogen Energy 16 (1991) 207.
[8] R.B. Biniwale, N. Kariya, M. Ichikawa, Catal. Lett. 105 (2005) 83.
[9] N. Kariya, A. Fukuoka, T. Utagawa, M. Ichikawa, Appl. Catal. A 247
(2003) 247.
[10] Y. Okada, E. Sasaki, E. Watanabe, S. Hyodo, H. Nishijima, Int. J. Hydro-
gen Energy 31 (2006) 1348.
[11] Y.G. Wang, N. Shah, G.P. Huffman, Energy Fuels. 18 (2004) 1429.
[12] S. Hodoshima, H. Arai, Y. Saito, Int. J. Hydrogen. Energy 28 (2003)
197.
[13] A.F. Ibarreta, C.J. Sung, Int. J. Hydrogen Energy 31 (2006) 1066.
[14] B. Wang, D.W. Goodman, G.F. Froment, J. Catal., in press.
[15] H. Matsuhashi, H. Shibata, H. Nakamura, K. Arata, Appl. Catal. A 187
(1999) 99.
[16] E. Romero, J.C. Rodriguez, J.A. Pena, A. Monzon, Can. J. Chem. Eng. 74
(1996) 1034.
3.4.2.4. Catalyst regeneration The catalysts used in this
study are prone to deactivation. The hybrid catalyst was re-
generated by two ways as shown in Fig. 12: (1) regeneration by
oxidation in air at 500 ◦C, then reduction in H2 at 300 ◦C; and
(2) regeneration by oxidation in air at 300 ◦C, then reduction in
H2 at 300 ◦C. Fig. 12 shows that oxidation at 500 ◦C is essential
for complete remove of the deposited coke and full recovery of
the catalyst activity.
[17] S.J. Tauster, R.M. Koros, J. Catal. 27 (1972) 307.
[18] Y.K. Park, F.H. Ribeiro, G.A. Somorjai, J. Catal. 178 (1998) 66.