Renewable Energy 50 (2013) 408e414

Contents lists available at SciVerse ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Production of renewable hydrogen by aqueous-phase reforming of glycerol over

NieCu catalysts derived from hydrotalcite precursors
Pablo V. Tuza, Robinson L. Manfro, Nielson F.P. Ribeiro, Mariana M.V.M. Souza*
Escola de Química, Universidade Federal do Rio de Janeiro (UFRJ), Centro de Tecnologia, Bloco E, Sala 206, CEP 21941-909, Rio de Janeiro/RJ, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: NieCu catalysts derived from hydrotalcite-like compounds were prepared and evaluated in aqueous-
Received 15 January 2012 phase reforming of glycerol. The catalysts were characterized by chemical composition, textural anal-
Accepted 9 July 2012 ysis, crystalline structure and reducibility. The reaction was carried out in a batch reactor with solution of
Available online 3 August 2012
10 wt.% glycerol, at 250 and 270
C. A maximum glycerol conversion of 60% was achieved at 270
C. In the
gas phase, the H2 selectivity was always higher than 80% and formation of CO was very low (<3%) at
Keywords:
250
C. The addition of Cu decreased the formation of methane. H2 is consumed during the reaction at
Hydrogen
270
C, mainly for Cu-containing catalysts, with simultaneous formation of propylene glycol in the liquid
Glycerol
Aqueous-phase reforming
phase. Acetol, lactic acid and acetaldehyde were also formed in the liquid phase, at both temperatures.
NieCu catalysts Ó 2012 Elsevier Ltd. All rights reserved.
Hydrotalcites

1. Introduction it is obtained as a by-product of biodiesel production by trans-

Renewable energy sources are becoming increasingly important
because of depleting of petroleum reserves and mounting envi-
ronmental concerns associated with fossil fuel utilization. In this
sense, the use of hydrogen as an alternative energy carrier has been
largely encouraged over the last years. Nowadays nearly 95% of the
world hydrogen is produced from fossil fuels [1], thus, it is neces-
sary to find renewable sources to produce hydrogen. Biomass is
a promising source for hydrogen production due to the carbon
neutral nature with environmental benefits.
Gasification and steam reforming have been widely investigated
for conversion of biomass to hydrogen, but these processes involve
high temperatures and many side reactions, which decrease the
hydrogen selectivity [2,3]. Aqueous-phase reforming (APR) for
production of hydrogen, first developed by the Dumesic group [4],
is an advantageous technology for feedstocks such as sugars,
alcohols and polyols, because it is carried out at low temperatures,
reducing the cost of the process and minimizing undesirable
decomposition reactions. Moreover, it produces H2 and CO2 in
a single reactor with low levels of CO, which is important for fuel
cell applications [5].
Several biomass-derived oxygenated compounds have been
used in APR process, such as methanol, sorbitol, glycerol, ethylene
glycol, and ethanol [4e9]. Glycerol is of particular interest because
* Corresponding author. Tel.: þ55 21 25627598; fax: þ55 21 25627596.
E-mail address: mmattos@eq.ufrj.br (M.M.V.M. Souza).
esterification of vegetable oils or animal fats. The transesterification
reaction produces approximately 110 kg of glycerol per ton of
biodiesel [10]. With increase in production of biodiesel, an excess of
glycerol is expected in the world market and therefore, it is
essential to find useful applications for glycerol. One of the prom-
ising ways to utilize this diluted glycerol from biodiesel plants is to
produce hydrogen. The hydrogen generation by APR of glycerol
takes place according to the Eq. (1):
C3H8O3 þ 3H2O / 3CO2 þ 7H2 (1)
Noble metal catalysts are commonly used to carry out the APR at
small scales because they are more active and less susceptible to
carbon deposition than non noble metal catalysts [6e8]. However,
large scale reforming processes use Ni-based catalysts that are far
cheaper and more available than noble metals. A good catalyst for
APR process has to be active in the cleavage of CeC bonds and
water-gas shift reaction, but has to inhibit the cleavage of CeO bond
and methanation reactions [4e6,11].
Ni shows a good activity for CeC scission, but it also exhibits
high rates of methanation [5]. Cu, on the other hand, is inactive for
CeC bond cleavage but has high activity for water-gas shift reac-
tion, thus favoring the selectivity to hydrogen. In the present work
Ni and Cu were selected to be included in catalyst formulations
applied to glycerol reforming in order to obtain high activity and
selectivity to hydrogen.
Hydrotalcite-like compounds (HTLCs) or layered double
hydroxides, also known as anionic clays, have a lamellar structure

0960-1481/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.renene.2012.07.006
 P.V. Tuza et al. / Renewable Energy 50 (2013) 408e414 409

with alternating positively charged mixed metal hydroxide sheets
and negatively charged interlayer anions along with water mole-
cules [12]. Thermal treatments of HTLCs give a stable, high surface
area, homogeneous mixture of oxides with very small crystal size,
which by reduction results in high metallic dispersion. NieMgeAl
mixed oxides prepared from hydrotalcite precursors have been
successfully used as catalysts for various reactions, including
reforming and oxidation of methane [13e15] and ethanol reform-
ing [9,16]. On the other hand, hydrotalcite-derived CueMgeAl
catalysts exhibit high activity for water-gas shift reaction [17]. To
our knowledge, NieCueMgeAl catalysts derived from hydrotalcites
have not been investigated in APR of glycerol yet.
In this context, the objective of this work is to study NieCu
catalysts prepared from hydrotalcite precursors in aqueous-phase
reforming of glycerol for hydrogen production. The effect of the
Cu loading (0e20 wt%) on the catalytic activity for reforming
reaction and hydrogen selectivity was investigated, correlating
with structural and morphological properties of the catalysts. By-
products formed in liquid phase were identified and the main
reaction routes were proposed.
2. Experimental
2.3. Catalytic tests
The reactions of APR of glycerol were carried out in an autoclave
batch reactor of 600 mL with 450 rpm of agitation, using 400 mg of
catalyst in 250 mL of 10 wt.% glycerol solution. Before the reaction,
the reactor was purged with He to remove the air inside. The
catalytic tests were performed at 250 and 270
C, resulting in
autogeneous pressures of 38 and 52 atm, respectively.
The catalysts were reduced ex situ under 75 mL min1 of 20%H2/
N2 up to 1000
C, using a heating rate of 10
C min1.
Gas products were analyzed online by gas chromatography (GC-
1000), equipped with a Hayesep D column and TCD. The products
detected in the gas phase were H2, CH4, CO2 and CO. The selectivity
to these products was calculated based on total moles produced in
the gas phase, without considering water.
In the liquid phase glycerol and by-products were quantified by
a Shimadzu HPLC with Aminex HPX 87H column, using 0.005 M
H2SO4 as eluent at 0.6 mL min1, and both UV and refractive index
detectors. Conversion was calculated based on the moles of glycerol
in the feed. Some liquid by-products were identified by gas
chromatography-mass spectrometry (GCeMS) using a Shimadzu
QP2010 instrument with DB5 column.

2.1. Catalyst preparation 3. Results and discussions

Hydrotalcite-like compounds were prepared by coprecipitation 3.1. Catalyst characterization

method at room temperature. An aqueous solution containing
appropriate amounts of nitrate precursors (Al/ The chemical composition of the synthesized samples is pre-
(Ni þ Cu þ Mg þ Al) ¼ 0.25 and [Ni þ Cu þ Mg þ Al] ¼ 1.5 M) was sented in Table 1 and Table 2. The composition of the as-
added dropwise to a vigorously stirred solution containing Na2CO3 synthesized samples is similar to those of the gel of synthesis,
and NaOH (CO2e 3 /Al
3þ
¼ 0.375 and OHe/Al3þ ¼ 6.3). The gel formed indicating an approximately complete incorporation of the cations
was aged for 18 h at 60
C. The resulting solid was filtered, washed in HTLC structure. The nickel loading of the sample without copper
with distilled water (70
C) until pH 7 and dried overnight at 100
C. is slightly lower than the nominal value, which can be due to
The Ni,Cu,Mg,Al-mixed oxides were produced by calcination of a small loss of nickel during washing.
HTLCs under flowing air (60 mL min1), using a heating rate of The XRD patterns of HTLC precursors (Fig. 1A) exhibit charac-
10
C min1, from room temperature to 500
C and keeping at this teristic planes of a well-crystallized HT in carbonate form (JCPDS
temperature for 2 h. 41e1428). The absence of other phases suggests that both Ni2þ
The NiO loading in calcined samples was always 20 wt% and CuO and Cu2þ have isomorphically replaced Mg2þ cations in the
loading was varied between 0 and 20 wt%. The samples will be brucite-like layers [9,18,19]. A clear improvement in the orderli-
referred to as Ni, Ni5Cu, Ni10Cu and Ni20Cu, according to CuO ness of the layer was noted with decreasing copper content, as
loading. indicated by both the increase in intensity and sharpness of
(110) and (113) reflections, observed around 60 and 62
, respec-
2.2. Catalyst characterization tively. This effect was expected due to Jahn-Teller distortion at
higher concentrations of copper, leading to poor long-range
The chemical composition of the synthesized samples was ordering [20].
determined by X-ray fluorescence (XRF) using a Rigaku (RIX 3100) The lattice parameters a (cationecation distance in the brucite-
spectrometer. like layer) and c ¼ 3c’(thickness of one brucite-like layer and one
X-ray powder diffraction (XRD) patterns were recorded in interlayer) were calculated based on the rhombohedral structure of
a Rigaku Miniflex II X-ray diffractometer equipped with a graphite hydrotalcites, as shown in Table 1. The lattice parameter a increases
monochromator using CuKa radiation (40 kV). The measurements with copper content since the ionic radius of Cu2þ (0.73 Å) is larger
were carried out with speed of 2
min1 and over the 2q range of 5
than that of Mg2þ (0.65 Å) [19]. There is also a decrease of the
and 90
. parameter c with increasing copper content, which can be
The textural characteristics, such as BET specific area and pore
volume (BJH method), were determined by N2 adsorption-
desorption at 196
C in a Micromeritcs ASAP 2000. Prior to the Table 1
analysis the samples were outgassed for 24 h at 200
C. Chemical composition, structural and textural characteristics of the as-synthesized
The reducibility of the catalysts was analyzed by temperature- samples.
programmed reduction (TPR), carried out in a microflow reactor Sample Al/(Ni þ Cu þ Mg þ Al)a a (Å) c (Å) SBET VPORE
operating at atmospheric pressure. The samples were firstly (molar) (m2/g) (cm3/g)
dehydrated at 150
C under flowing Ar before the reduction. A Ni 0.26 (0.25) 3.053 23.59 46.4 0.106
mixture of 1.53% H2/Ar flowed at 30 mL min1 through the sample, Ni5Cu 0.28 (0.25) 3.055 23.49 42.1 0.167
raising the temperature at a heating rate of 10
C min1 up to Ni10Cu 0.25 (0.25) 3.059 23.29 60.5 0.152
1000
C. The outflowing gases were detected by thermal conduc- Ni20Cu 0.24 (0.25) 3.064 23.20 62.0 0.145
a
tivity detector (TCD). The values in parentheses are nominal values.
410 P.V. Tuza et al. / Renewable Energy 50 (2013) 408e414

Table 2 above 800
C. The lines of MgO are broader in the sample of lower
Chemical composition, textural characteristics of the calcined catalysts and Ni Mg content, indicating a decrease in the size of the MgO crystal or
particle sizes of the reduced catalysts.
an increase of the Al3þ incorporation in MgO since the ionic radius
Sample NiOa (wt%) CuOa (wt%) MgO SBET VPORE Ni particle Dc of Al3þ (0.54 Å) is smaller than those of Mg2þ (0.65 Å) and Ni2þ
(wt%) (m2/g) (cm3/g) size (nm)b (%) (0.72 Å) [23]. Moreover, the reflections of MgO-solid solution are
Ni 16.7 (20) e 57.8 260.0 0.308 28.8 3.5 shifted towards wider 2q angles than in pure MgO, which also
Ni5Cu 19.1 (20) 5.3 (5) 48.0 211.3 0.480 9.4 10.7
indicates the presence of foreign ions in the oxide lattice [24].
Ni10Cu 19.2 (20) 11.3 (10) 44.8 203.4 0.416 9.8 10.3
Ni20Cu 19.4 (20) 22.5 (20) 33.7 118.4 0.325 14.3 7.0 XRD analysis of the catalysts was also carried out after reduction
a
at 1000
C (Fig. 1 C). The reduction caused the formation of Ni0, with
The values in parentheses are nominal values.
b
Calculated by the Scherrer equation using (200) plane of Ni.
reflections at 44.5
, 51.9
and 76.4
(JCPDS 4e850), showing that ex
c
Dispersion of Ni metal particles. situ activation was successful in reducing the active nickel phase.
Metallic nickel is formed by the reduction of NiO and MgeNieO
solid solution [9]. It can also be observed the formation of spinel
associated with a decrease in the interlayer distance, as suggested phases (MAl2O4 with M ¼ Mg, Ni or Cu) and a small contribution of
by Chmielarz et al. [21]. Cu0 at 43.3
for the sample with higher Cu loading (Ni20Cu).
After calcination at 500
C, the characteristic peaks of HT The Ni particle sizes were calculated by the Scherrer equation,
structure disappear and XRD patterns (Fig. 1 B) show the presence using the peak at 2q ¼ 51.9
, which corresponds to the (200) plane
of poorly crystallized MgO periclase phase at 42.9
and 62.3
(JCPDS of Ni fcc. The results are shown in Table 2, together with the cor-
45e946). The presence of NiO is only evidenced by the very small responding Ni dispersion, estimated according to Anderson [25].
peak at 37.2
, since the other peaks of this phase are overlapped The use of HTLC precursors results in high dispersion of metal
with MgO at 43.3
and 62.9
(JCPDS 47e1049). On the other hand, particles, as already observed by other authors [9,23,26]. The
whatever the Cu content, no crystalline Cu-containing phase addition of copper greatly improved the Ni dispersion, which can
evolved and no well-defined spinel phase was noted. These results be related to the increased reducibility of nickel in presence of
indicate that most of Ni, and Cu and Al oxides are well dispersed in copper, as will be shown by TPR.
MgO matrix, forming a solid solution, in accordance with the Textural characteristics of the as-synthesized samples (Table 1)
literature [18,20]. Kovanda et al. [22] reported that CuAl2O4 spinel showed that samples with higher amount of copper have larger BET
was formed from CueMgeAl hydrotalcites only after calcination surface areas, which is coherent with previous results [17].

Fig. 1. XRD patterns of the as-synthesized HTLCs (A), and catalysts after calcination (B) and reduction (C). Phases: D MgO, þ NiO, Ni, A spinel phases, O Cu0.
 P.V. Tuza et al. / Renewable Energy 50 (2013) 408e414
Adsorption-desorption isotherms of nitrogen for all HTLCs (not
shown) exhibited the type IV pattern with a broad hysteresis loop,
characteristic of mesoporous materials. There was an increase in
surface area and pore volume after calcination, as shown in Table 2,
which can be related to the destruction of the layered structure (as
observed by XRD) [9,17]. The surface areas of the Cu-containing
samples after calcination are smaller than the Ni sample, which
Kovanda et al. [22] related to the low sintering temperature of the
copper oxides.
TPR profiles of the catalysts are displayed in Fig. 2. The sample
without Cu showed only one reduction peak at 915
C, which can be
associated either to the reduction of Ni species on MgeNieO solid
solution [23] or to the reduction of NiAl2O4 spinel phase formed at
temperatures higher than 800
C during TPR analysis [18]. The Cu-
containing samples exhibited two reduction peaks, one at low
temperatures (340e370
C), associated with the reduction of Cu2þ to
Cu0, and another at high temperatures (725e775
C), related to
reduction of Ni species. The low-temperature peak shifts to lower
temperatures as the amount of Cu increases, suggesting that a large
amount of Mg stabilizes a large amount of Cu in the solid solution,
decreasing the reducibility of Cu2þ component [22]. On the other hand,
the increase in Cu content decreased the reduction temperature of Ni
species, which was attributed by Ashok et al. [27] to the spillover of
hydrogen produced by Cu, accelerating the nucleation of the Ni metal.
Table 3 shows the reduction degrees of CuO and NiO, calculated
by integration of the first and second TPR peaks, respectively,
considering the reduction of Cu2þ to Cu0 and Ni2þ to Ni0. The
addition of Cu increased the reducibility of Ni2þ species, in accor-
dance with Ashok et al. [27]. Another possibility is that the second
reduction peak also corresponds to the reduction of copper in
spinel phase (shown by XRD), as observed by Chmielarz et al. [28]
and, in this case, the reduction of copper species would be 100%
with an incomplete reduction of nickel species.
3.2. Catalytic tests
The glycerol conversion of the NieCu catalysts for aqueous-
phase reforming at 250 and 270
C is displayed in Fig. 3. At
250
C, the conversions increased with time up to about 6e8 h,
while at 270
C they increased continuously up to 10 h. The glyc-
erol conversion increased by increasing the reaction temperature
for all catalysts, as similarly observed in a previous study [29]. At
250
C, Ni catalyst provided the best performance, starting at 16% of
Fig. 2. TPR profiles of the catalysts.
411
Table 3
Reduction degrees of CuO and NiO calculated from temperature-programmed
reduction results.
Catalyst CuOa (%) NiOb (%)
Ni e 72.5
Ni5Cu 79.1 100
Ni10Cu 94.5 100
Ni20Cu 80.3 100
a
Calculated by integration of the first reduction peak.
b
Calculated by integration of the second reduction peak.
conversion and ending at approximately 34% after 12 h of reaction.
The conversions with the copper containing catalysts were not
much different from the Ni catalyst, with conversion of around 28%
after 12 h. At 270
C, all catalysts reached 59% of conversion, except
Ni5Cu that showed a maximum conversion of 45%.
The gas product distribution of the APR of glycerol at 250
C is
shown in Fig. 4. H2, CO2, CH4, and CO were the only reaction
products identified in the gas phase. The Ni5Cu and Ni10Cu cata-
lysts showed the best performance in terms of H2 production, since
they generated a H2 selectivity of about 90%, which remained
practically constant during 12 h of reaction. The H2 selectivity for
the other catalysts was about 80%. In relation to CO2, Ni5Cu
exhibited the lowest selectivity, 3% in the beginning of the reaction
and 8% at the end. The other catalysts presented CO2 selectivity of
approximately 15%. CO selectivities were always below 3% and the
maximum CH4 selectivity was 7% for Ni catalyst.
The higher selectivity of Ni5Cu and Ni10Cu catalysts for
production of hydrogen may be related to the lower Ni particle size.
The lower Ni particle size decreases the number of clusters with
multiple nickel atoms that are necessary for CO dissociation, which
leads to formation of methane, thus resulting in higher H2 selec-
tivity [29]. This effect is the same as the addition of Sn to nickel
catalysts reported by the Dumesic group [11,30]. Wawrzetz et al.
[31] showed that H2 and CO2 formation decreased in APR of glycerol
with increasing Pt particle size.
The aqueous-phase reforming of glycerol produces primarily H2
and CO2, according to reaction (1). The H2 selectivity obtained by
Shabaker et al. [11] in APR of glycerol using NieSn catalyst was 64%,
at 265
C, with 30% of CO2. These authors used a continuous reac-
tion system, and the liquid effluent contained less than 50 ppm of
by-products, such as aldehydes, alcohols and organic acids. In the
present study, the H2 production varied between 80 and 90%, while
CO2 was in the range of 5e30%, showing that H2/CO2 ratio was
much higher than the stoichiometry of the reforming reaction,
which indicates that the reaction is not purely APR, with
occurrence of different side reactions. Our tests were conducted
in a batch system, and high concentrations of liquid by-products
were detected. Some of these by-products are formed by dehy-
drogenation reactions, such as lactic acid (with concentrations as
high as 13 g L1) and acetaldehyde (concentrations up to 5 g L1), as
will be shown later. Thus, for a glycerol concentration in the feed of
100 g L1, these by-products are formed in significant amounts and,
consequently, the H2 formation by these side reactions is also
significant.
It can be seen from Fig. 4 that the formation of CO is very low,
indicating that the low temperature favors the water-gas shift
reaction (Eq. (2)), since Ni has a moderate activity and Cu is one of
the most active metals in this reaction [32]. Although it is well
known that Ni catalysts have high activity in methanation reactions
of CO and CO2 (Eqs. (3) and (4)) [5], the CH4 formation was not
significant in the present study. The addition of Cu decreased the
activity for methanation reaction, since Cu is not active for this
reaction [33].
412 P.V. Tuza et al. / Renewable Energy 50 (2013) 408e414

Fig. 3. Glycerol conversion as a function of time on stream for aqueous-phase reforming of 10 wt.% glycerol solution at 250 (A) and 270
C (B).

CO þ H2O / CO2 þ H2 (2) López et al. [34]. Acetol is produced by dehydration of glycerol
(Eq. (6)), as already observed by many authors
CO þ 3H2 / CH4 þ H2O (3) [31,35e37]. Acetaldehyde can be formed from dehydration of
ethylene glycol (Eq. (8)), which is produced by decarbonylation/
CO2 þ 4H2 / CH4 þ 2H2O (4) dehydrogenation of glycerol (Eq. (7)) [35]. Lactic acid was
produced at higher concentrations (11e13 g L1) than acetol
The main products detected in liquid phase for APR of glycerol (2e4 g L1) or acetaldehyde (4e5 g L1). The high
at 250
C were lactic acid, acetol (hydroxyacetone) and concentration of these by-products shows that formation of H2
acetaldehyde. The formation of lactic acid is related to dehy- by these side reactions is significant, justifying the high H2
drogenation of glycerol (Eq. (5)), as proposed by Ramirez- selectivities (Fig. 4).

Fig. 4. H2, CH4, CO2 and CO selectivities as a function of time on stream for aqueous-phase reforming of 10 wt.% glycerol solution at 250
C with (a) Ni, (b) Ni5Cu, (c) Ni10Cu and
(d) Ni20Cu catalysts.
 P.V. Tuza et al. / Renewable Energy 50 (2013) 408e414 413

O The distribution of products at 270
C followed the same trend for

OH OH OH + H2 (5) all catalysts: H2 production decreased with time, while CO2 increased,
OH OH but this tendency is more pronounced for Cu catalysts, especially
Ni5Cu. Hydrogen produced by APR can be consumed in the hydro-
genation of products from dehydration of glycerol, as acetol, 3-
hydroxypropanal [31], pyruvaldehyde and acetaldehyde [35].
OH OH OH + H2O (6) The main products detected in liquid phase for APR of glycerol at
OH O 270
C were lactic acid, acetol, acetaldehyde and also propylene
glycol. Propylene glycol comes from hydrogenation of acetol (Eq.
(9)) [31,36,38]. The production of lactic acid was higher for Ni
catalyst (achieving 19 g L1), while acetol formation increases with
OH OH OH + H2 + CO (7) Cu content: from 3 g L1 for Ni catalyst to 22 g L1 for Ni20Cu, at the
OH OH
end of the reaction. Small traces of other liquid by-products were
detected by GCeMS, mainly for Cu-containing catalysts at long
reaction times: 1,2 butanediol, cyclopentanone, 2,5 hexadione and
OH O + H2O (8) 4-hydroxy-3-hexanone.
OH
OH + H2 OH (9)
O OH

Fig. 5 presents the product distribution for APR of glycerol at

270
C. Ni catalyst generated an initial H2 selectivity close to 90%,
Propylene glycol was formed at higher concentrations on Cu-
which decreased to approximately 75% after 4 h of reaction, and
containing catalysts, which agrees with Dasari et al. [38] and Naka-
remained constant until the end of the reaction. CO2 formation
gawa and Tomishige [39], who showed that Cu is a better catalyst than
increased from 4% to 19% in 8 h of reaction and then decreased to
Ni for hydrogenolysis of glycerol to propylene glycol. Meher et al. [40]
15%. This catalyst produced little CO, around 2% after 12 h of reac-
showed specifically that Cu/Zn/Al mixed-oxide catalysts derived from
tion. However, methane production was greater under these
hydrotalcite precursors have high activity in glycerol conversion to
conditions, reaching a selectivity of 8%.
propylene glycol. Moreover, the production of propylene glycol
Reactions performed with Ni10Cu and Ni20Cu catalysts pre-
increases with reaction time, while the H2 selectivity decreases,
sented H2 selectivity initially around 92% and decreased to
confirming the occurrence of hydrogenation reactions.
approximately 53% after 12 h of reaction. Formation of CO2 was
below 2% in the beginning of the reaction, increasing to 33 and 42%
for Ni10Cu and Ni20Cu, respectively. Noteworthy is the product 4. Conclusions
distribution for Ni5Cu catalyst: the H2 selectivity decreased from 85
to 13% after 12 h of reaction, while CO2 increased from 3 to 78%. The NieCu catalysts derived from hydrotalcite-like compounds
CO selectivity was approximately 5% for Ni10Cu and Ni20Cu cata- showed great potential to be used in aqueous-phase reforming of
lysts, and reached 9% for Ni5Cu. The CH4 selectivity remained below glycerol, with good activity and high selectivity to hydrogen. Ni
1% for all Cu-containing catalysts. catalyst (without Cu) showed higher glycerol conversions, despite

Fig. 5. H2, CH4, CO2 and CO selectivities as a function of time on stream for aqueous-phase reforming of 10 wt.% glycerol solution at 270
C with (a) Ni, (b) Ni5Cu, (c) Ni10Cu and
(d) Ni20Cu catalysts.
414 P.V. Tuza et al. / Renewable Energy 50 (2013) 408e414
its lower Ni dispersion. Ni5Cu catalyst presented the highest H2
selectivity at 250
C, with the smallest amount of CO and CH4.
However, this catalyst provided the least desirable performance at
270
C because of low conversion and almost total consumption of
hydrogen. The consumption of H2 is related to hydrogenation of
acetol, formed by dehydration of glycerol, producing propylene
glycol. Lactic acid and acetaldehyde were also formed in liquid
phase, besides small traces of heavier by-products.
Acknowledgments
The authors thank Núcleo de Catálise/COPPE/UFRJ for providing
the FRX, BET analysis and also GCeMS; Angela Rocha for helping
with GCeMS; and CNPq for the financial support granted to carry
out this work.
References
[1] Dunn S. Hydrogen futures: toward a sustainable energy system. Int J
Hydrogen Energy 2002;27:235e64.
[2] Huber GW, Iborra S, Corma A. Synthesis of transportation fuels from biomass:
chemistry, catalysts, and engineering. Chem Rev 2006;106:4044e98.
[3] Adhikari S, Fernando SD, Haryanto A. Hydrogen production from glycerol: an
update. Energy Convers Manage 2009;50:2600e4.
[4] Cortright RD, Davda RR, Dumesic JA. Hydrogen from catalytic reforming of
biomass-derived hydrocarbons in liquid water. Nature 2002;418:964e7.
[5] Davda RR, Shabaker JW, Huber GW, Cortright RD, Dumesic JA. A review of
catalytic issues and process conditions for renewable hydrogen and alkanes
by aqueous-phase reforming of oxygenated hydrocarbons over supported
metal catalysts. Appl Catal B 2005;56:171e86.
[6] Shabaker JW, Davda RR, Huber GW, Cortright RD, Dumesic JA. Aqueous-phase
reforming of methanol and ethylene glycol over alumina-supported platinum
catalysts. J Catal 2003;215:344e52.
[7] Shabaker JW, Huber GW, Cortright RD, Dumesic JA. Aqueous-phase reforming
of ethylene glycol over supported platinum catalysts. Catal Lett 2003;88:1e8.
[8] Soares RR, Simonetti DA, Dumesic JA. Glycerol as a source for fuels and chemicals
by low-temperature catalytic processing. Angew Chem Int Ed 2006;45:3982e5.
[9] Cruz IO, Ribeiro NFP, Aranda DAG, Souza MMVM. Hydrogen production by
aqueous-phase reforming of ethanol over nickel catalysts prepared from
hydrotalcite precursors. Catal Commun 2008;9:2606e11.
[10] Behr A, Eilting J, Irawadi K, Leschinski J, Lindner F. Improved utilisation of
renewable resources: new important derivatives of glycerol. Green Chem
2008;10:13e30.
[11] Shabaker JW, Huber GW, Dumesic JA. Aqueous-phase reforming of oxygen-
ated hydrocarbons over Sn-modified Ni catalysts. J Catal 2004;222:180e91.
[12] Cavani F, Trifiró F, Vaccari A. Hydrotalcite-type anionic clays: preparation,
properties and applications. Catal Today 1991;11:173e301.
[13] Basile F, Basini L, D’Amore M, Fornasari G, Guarinoni A, Matteuzzi D, et al. Ni/
Mg/Al anionic clay derived catalysts for the catalytic partial oxidation of
methane: residence time dependence of the reactivity features. J Catal 1998;
173:247e56.
[14] Tsyganok AI, Tsunoda T, Hamakawa S, Suzuki K, Takehira K, Hayakawa T. Dry
reforming of methane over catalysts derived from nickel-containing MgeAl
layered double hydroxides. J Catal 2003;213:191e203.
[15] Fonseca A, Assaf EM. Production of the hydrogen by methane steam reforming
over nickel catalysts prepared from hydrotalcite precursors. J Power Sources
2005;142:154e9.
[16] Comas J, Dieuzeide ML, Baronetti G, Laborde M, Amadeo N. Methane steam
reforming and ethanol steam reforming using a Ni(II)eAl(III) catalyst
prepared from lamellar double hydroxides. Chem Eng J 2006;118:11e5.
[17] Ferreira KA, Ribeiro NFP, Souza MMVM, Schmal M. Structural transformation
of CueMgeAl mixed oxide catalysts derived from hydrotalcites during shift
reaction. Catal Lett 2009;132:58e63.
[18] Rodrigues ACC, Henriques CA, Monteiro JLF. Influence of Ni content in
physico-chemical characteristics of Ni, Mg, Al-hydrotalcite like compounds.
Mater Res 2003;6:563e8.
[19] Dussault L, Dupin JC, Dumitriu E, Auroux A, Guimon C. Microcalorimetry, TPR
and XPS studies of acidebase properties of NiCuMgAl mixed oxides using
LDHs as precursors. Thermochim Acta 2005;434:93e9.
[20] Kannan S, Dubey A, Knozinger H. Synthesis and characterization of CuMgAl
ternary hydrotalcites as catalysts for the hydroxylation of phenol. J Catal
2005;231:381e92.
[21] Chmielarz L, Kustrowski P, Rafalska-qasocha A, Majda D, Dziembaj R. Catalytic
activity of CoeMgeAl, CueMgeAl and CueCoeMgeAl mixed oxides derived
from hydrotalcites in SCR of NO with ammonia. Appl Catal B 2002;35:
195e210.
[22] Kovanda F, Jirátová K, Rymes J, Kolousek D. Characterization of activated Cu/
Mg/Al hydrotalcites and their catalytic activity in toluene combustion. Appl
Clay Sci 2001;18:71e80.
[23] Takehira K, Shishido T, Wang P, Kosaka T, Takaki K. Autothermal reforming of
CH4 over supported Ni catalysts prepared from MgeAl hydrotalcite-like
anionic clay. J Catal 2004;221:43e54.
[24] Fornasari G, Gazzano M, Matteuzzi D, Trifiró F, Vaccari A. Structure and
reactivity of high-surface-area Ni/Mg/Al mixed oxides. Appl Clay Sci 1995;10:
69e82.
[25] Anderson JR. Structure of metallic catalysts. London: Academic Press; 1975.
[26] Shishido T, Sukenobu M, Morioka H, Furukawa R, Shirahase H, Takehira K. CO2
reforming of CH4 over Ni/MgeAl oxide catalysts prepared by solid phase
crystallization method from MgeAl hydrotalcite-like precursors. Catal Lett
2001;73:21e6.
[27] Ashok J, Subrahmanyam M, Venugopal A. Hydrotalcite structure derived
NieCueAl catalysts for the production of H2 by CH4 decomposition. Int J
Hydrogen Energy 2008;33:2704e13.
[28] Chmielarz L, Kustrowski P, Rafalska-qasocha A, Dziembaj R. Influence of Cu,
Co and Ni cations incorporated in brucite-type layers on thermal behaviour of
hydrotalcites and reducibility of the derived mixed oxide systems. Thermo-
chim Acta 2003;395:225e36.
[29] Manfro RL, Costa AF, Ribeiro NFP, Souza MMVM. Hydrogen production by
aqueous-phase reforming of glycerol over nickel catalysts supported on CeO2.
Fuel Process Technol 2011;92:330e5.
[30] Huber GW, Shabaker JW, Dumesic JA. Raney NieSn catalyst for H2 production
from biomass-derived hydrocarbons. Science 2003;300:2075e7.
[31] Wawrzetz A, Peng B, Hrabar A, Jentys A, Lemonidou AA, Lercher JA. Towards
understanding the bifunctional hydrodeoxygenation and aqueous phase
reforming of glycerol. J Catal 2010;269:411e20.
[32] Grenoble DC, Estadt MM, Ollis DF. The chemistry and catalysis of the water
gas shift reaction: 1. The kinetics over supported metal catalysts. J Catal 1981;
67:90e102.
[33] Vannice MA. The catalytic synthesis of hydrocarbons from H2/CO mixtures
over the group VIII metals: V. The catalytic behavior of silica-supported
metals. J Catal 1977;50:228e36.
[34] Ramírez-López CA, Ochoa-Gómez JR, Fernández-Santos M, Gómez-Jiménez-
Aberasturi O, Alonso-Vicario A, Torrecilla-Soria J. Synthesis of lactic acid by
alkaline hydrothermal conversion of glycerol at high glycerol concentration.
Ind Eng Chem Res 2010;49:6270e8.
[35] King DL, Zhang L, Xia G, Karim AM, Heldebrant DJ, Wang X, et al. Aqueous
phase reforming of glycerol for hydrogen production over PteRe supported
on carbon. Appl Catal B 2010;99:206e13.
[36] Miyazawa T, Kusunoki Y, Kunimori K, Tomishige K. Glycerol conversion in the
aqueous solution under hydrogen over Ru/C þ an ion-exchange resin and its
reaction mechanism. J Catal 2006;240:213e21.
[37] Pompeo F, Santori G, Nichio NN. Hydrogen and/or syngas from steam
reforming of glycerol: study of platinum catalysts. Int J Hydrogen Energy
2010;35:8912e20.
[38] Dasari MA, Kiatsimkul PP, Sutterlin WR, Suppes GJ. Low-pressure
hydrogenolysis of glycerol to propylene glycol. Appl Catal A 2005;281:
225e31.
[39] Nakagawa Y, Tomishige K. Heterogeneous catalysis of the glycerol hydro-
genolysis. Catal Sci Technol 2011. http://dx.doi.org/10.1039/c0cy00054j.
[40] Meher LC, Gopinath R, Naik SN, Dalai AK. Catalytic hydrogenolysis of glycerol
to propylene glycol over mixed oxides derived from a hydrotalcite-type
precursor. Ind Eng Chem Res 2009;48:1840e6.