A catalytic route to lower alcohols from glycerol using Ni-supported catalysts

Article abstract of DOI:10.1039/c0gc00839g

The activity of Ni-supported catalysts on silica and alumina was studied for the transformation of glycerol to lower alcohols, primarily 1-propanol and ethanol. Pressure had a small effect on glycerol conversion whereas temperature had a significant effect on conversion and selectivity. Ni/SiO₂ gave quantitative conversion of glycerol at a lower temperature compared to Ni/Al₂O₃. Ni/SiO₂ also gave a higher selectivity to ethanol and propanol (63%) compared to Ni/Al₂O ₃ (52%) at a similar conversion. The higher activity of Ni/SiO ₂ can be ascribed to its smaller crystallite size ascertained from XRD analysis. The Ni/SiO₂ catalyst also showed improved reducibility compared to Ni/Al₂O₃. The used catalyst showed sintering of the Ni, which was confirmed by a loss of surface area and average crystallite size. The route to lower alcohols from glycerol is proposed to occur via 1,2-propanediol. Biomass is readily converted into biodiesel and glycerol and we have developed a route to convert glycerol into lower alcohols. We have therefore demonstrated that it is possible to convert biomass into biodiesel and lower alcohols (via glycerol) from the same crop, thus reducing the need to produce ethanol from a separate government-subsidised crop.

Full text of DOI:10.1039/c0gc00839g

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 1819
www.rsc.org/greenchem
PAPER
A catalytic route to lower alcohols from glycerol using Ni-supported
catalysts
a
b
b
c
a
Esti van Ryneveld, Abdul S. Mahomed, Pieter S. van Heerden, Mike J. Green and Holger B. Friedrich*
Received 22nd November 2010, Accepted 28th March 2011
DOI: 10.1039/c0gc00839g
The activity of Ni-supported catalysts on silica and alumina was studied for the transformation of
glycerol to lower alcohols, primarily 1-propanol and ethanol. Pressure had a small effect on
glycerol conversion whereas temperature had a significant effect on conversion and selectivity.
Ni/SiO
Ni/SiO
2
gave quantitative conversion of glycerol at a lower temperature compared to Ni/Al
also gave a higher selectivity to ethanol and propanol (63%) compared to Ni/Al
can be ascribed to its smaller
catalyst also showed improved
. The used catalyst showed sintering of the Ni, which was
2
3
O .
2
2
O
3
(
52%) at a similar conversion. The higher activity of Ni/SiO
crystallite size ascertained from XRD analysis. The Ni/SiO
reducibility compared to Ni/Al
2
2
2
O
3
confirmed by a loss of surface area and average crystallite size. The route to lower alcohols from
glycerol is proposed to occur via 1,2-propanediol. Biomass is readily converted into biodiesel and
glycerol and we have developed a route to convert glycerol into lower alcohols. We have therefore
demonstrated that it is possible to convert biomass into biodiesel and lower alcohols (via glycerol)
from the same crop, thus reducing the need to produce ethanol from a separate
government-subsidised crop.
ꢀ
R
6
1
. Introduction
PtRu, copper systems and Raney Ni. Surprisingly, the use
of supported Ni systems as catalysts towards the chemi-
cal transformation of glycerol, especially towards the forma-
tion of lower alcohols, has appeared less frequently in the
The interest in the catalytic conversion of renewable feedstocks
to a range of value-added chemical commodities has been
increasing. Studies on the hydrogenolysis of polyhydric alcohols
have been pursued since 1930. Glycerol is an important building
block in natural fats and oils, and recently the need has
arisen to find new applications for this by-product obtained
from the production of biodiesel. Glycerol can be converted
7–11
literature.
7
Bloom described a process for the hydrogenolysis of a 40 wt%
1
glycerol feedstock. Several Ni-supported catalysts were used,
and the major product was 1,2-PDO with a selectivity of up
to 72%. The supports that were used include carbon, silica and
alumina.
2
to hydrogen and synthesis gas by reforming, to 1,2-propanediol
(
1,2-PDO), 1,3-propanediol (1,3-PDO) and ethylene glycol (EG)
The hydrogenolysis of polyols (e.g. xylitol, sorbitol or glycerol)
to ethylene glycol in the presence of nickel on silica (or nickel on
3
via hydrogenolysis, and also to lower alcohols such as methanol,
ethanol, 1-propanol and 2-propanol.
Much work has been done towards the hydrogenolysis of
glycerol to 1,3- and 1,2-PDO, as well as oxidation of glycerol
to glyceric acid or dihydroxyacetone. However, routes to lower
alcohols, such as 1-propanol and ethanol, have been less
discussed. It is known that glycerol can be hydrogenolysed
using various heterogeneous systems including Rh, Ru, Pt,
4
8
alumina) and hydrogen was described by Bullock et al. High
conversions were observed, with 70% selectivity to 1,2-PDO,
1
5% to ethylene glycol (EG), and 11% of degradation products.
Huang et al. reported on the hydrogenolysis of solvent-
5
9
free glycerol to 1,2-PDO. Several catalysts were screened
and the most effective catalysts (41 wt% Ni/Al and 32
wt% Cu/ZnO/Al ) were further tested for vapour-phase
hydrogenolysis in a fixed bed. They found that Ni/Al is not
an effective catalyst for the production of 1,2-PDO because of
the high selectivity to CH and CO.
2
O
3
2
O
3
2
O
3
a
School of Chemistry, University of KwaZulu-Natal, Westville Campus,
Private Bag X54001, Durban, 4000, South Africa.
4
E-mail: friedric@ukzn.ac.za; Fax: +27 31-2603109; Tel: +27
The catalytic hydrogenolysis of glycerol to propanediols
using homogeneous Pd and Pt systems was also described
3
1-2603107
b
Sasol Technology, R&D, 1 Klasie Havenga Street, Sasolburg, 1948,
South Africa
10
by Drent and Jager. Typical selectivities obtained were 47%
c
to 1-propanol, 31% to 1,3-PDO and 22% to 1,2-PDO. The
hydrogenolysis was carried out under moderate reaction
School of Chemistry, Newcastle University, Bedson Building, Newcastle
upon Tyne, NE1 7RU, UK
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1819–1827 | 1819

View Article Online
◦
◦
◦
conditions with temperatures in the range of 140 C and
pressures around 40 bar H
at 350 C for 30 min and then cooled down to 80 C under the
2
.
same stream of argon. In the reduction experiment, 5% H
argon was used as a reducing agent at a flow rate of 50 ml min .
Under these reducing conditions, the temperature was ramped
up to 950 C at a rate of 10 C min .
In the TPD experiments, the catalysts (ca. 20 mg) were first
2
in
11
-1
Tomishige and co-workers reported a high selectivity of 42%
to 1-propanol and 13% to 2-propanol using supported metal
◦
◦
-1
catalysts (Rh, Ru, Pt, Pd on active carbon, SiO
2
, Al
2
O
3
) during
in a
the hydrogenolysis of aqueous glycerol solutions under H
batch reactor.
2
reduced under 5% H
2
in argon and thereafter treated with helium
Thus, much of the work that has been reported focused on the
production of 1,2-propanediol from glycerol using Cu supported
systems, and at times, Ni catalysts. The focus of this work was the
production of lower alcohols, primarily 1-propanol and ethanol,
from glycerol using Ni-supported catalysts in a continuous flow
fixed bed reactor.
for one hour to remove excess hydrogen. A 4% ammonia-in-
-
1
helium gas mixture was then passed (20 ml min ) over the
catalyst for 30 min. The excess ammonia was removed by
-
1
flushing the system with helium (30 ml min ) for 30 min and
then adsorbed ammonia was stripped off by the same stream of
-
1
◦
helium (30 ml min ) and a temperature ramp of up to 900 C
◦
-1
The global capacity for 1-propanol is about 275 kt, and it is
produced via the hydroformylation of ethylene to form propanal,
(at 10 C min ). The desorption profiles were recorded using a
TCD.
12
which is then hydrogenated to 1-propanol. Currently, 1-
propanol is a fairly expensive commodity chemical and is mainly
used as a solvent or in making n-propyl acetate.
Powder X-ray diffraction patterns were recorded on a Bruker
D8 Advance diffractometer equipped with a graphite monochro-
mator and operated at 40 kV and 40 mA. The source of radiation
◦
The world production of ethanol in 2007 was about 52 Mt,
and the average annual growth rate for ethanol for 2008–
was Co Ka. 2q covered the range between 10 and 90 at a
◦
-1
◦
speed of 1 min with a step size of 0.02 , and all data were
13
2
013 is estimated at 9%. The primary source of ethanol is
evaluated with DIFFRACPLUS (EVA, SEARCH, TOPAS)
◦
via the fermentation of sugar and grain crops (70% of global
software. Employing the peak at 50 , the average crystallite size
production), and the main market for ethanol is for fuel (80–
was calculated using the Scherrer equation.
13
9
0%).
The price of refined glycerol has shown a significant decline
Scanning electron microscopy (SEM) images were obtained
using a LEO 1450 Scanning Electron Microscope. Samples for
SEM images were coated with gold using a Polaron SC Sputter
Coater.
Transmission electron microscopy (TEM) images were taken
on a Jeol JEM 1010 transmission electron microscope operated
at a voltage of 100 kV. Samples were prepared by deposition
of a small amount of the catalyst between two formvar-coated
copper grids, and the images were captured with the MegaView
III Soft Imaging System.
over the last 5–7 years, and indeed a glycerol “lake” is being
produced as a by-product of biodiesel (15 Mt p.a. in 2010,
growing at around 10% p.a.; and 1 ton of glycerine is produced
for every 10 tons of biodiesel). Between 2004 and 2007, the
world capacity for refined glycerol grew at an average annual
14
rate of 8.7%, faster than world consumption. Biomass is
readily converted to biodiesel and glycerol, and we are therefore
proposing a route to convert the low-value glycerol to high-
value commodity alcohols, which also reduces the need to
produce ethanol from a separate government-subsidised crop,
thus freeing up land for food production.
Thermogravimetric analysis–differential scanning calorime-
try (TGA-DSC) was performed on the catalysts under a nitrogen
atmosphere using a SDT Q 600 TGA-DSC instrument. The
◦
temperature was increased from room temperature to 1000 C
◦
-1
at a rate of 20 C min .
2
. Experimental
2.1 Reagents
The Ni/Al
2
O
3
and Ni/SiO
2
catalyst systems were obtained as
2.3 Catalytic testing
commercial samples in a partially reduced state. The average
Ni content was between 45–55 wt% for Ni/Al and Ni/SiO
respectively. The glycerol was purchased from Sigma Aldrich
and used without further purification. The glycerol feed was
prepared as 60 wt% solution in water.
The catalytic reaction of glycerol was performed in a continuous
flow fixed-bed reactor in down-flow mode. The reactor tube was
stainless steel with an internal diameter of 20 mm and a length of
2
O
3
2
2
50 mm. The catalyst volume was 5 ml (ca. 8.5 g) and mixed with
an equal amount of carborundum. The catalyst had a particle
size distribution of 300–500 mm. The molar ratio of hydrogen to
2
.2 Catalyst characterisation
-
1
glycerol solution was 2 : 1 with a GHSV of 1060 h and a LHSV
-
1
The BET surface areas were measured by nitrogen physisorption
isotherms at 77 K using the standard multipoint method (eleven
points) on a Micromeritics Gemini instrument. Prior to the
analysis, samples were degassed in a stream of nitrogen at 473 K
for 24 h.
of 3.0 h . The catalytic reactions were carried out between
◦
230–320 C with pressures varying between 40–75 bar. Prior
◦
to the reaction, the catalyst was reduced at 180 C after which
the reactor was adjusted to operating conditions. The liquid
products and the unreacted glycerol were collected in catchpots
◦
◦
Temperature-programmed reduction (TPR) and NH
3
cooled to 3 C and -20 C respectively. The liquid products as
well as the gas products were collected at regular intervals, and
were analyzed on a GC (HP 6890) equipped with a FID using
a DB-1701 column. Another gas sample was injected on a GC
equipped with a TCD (Agilent 6850) using a Shincarbon-packed
temperature-programmed desorption (TPD) experiments were
carried out in a Micrometrics 2900 AutoChem II Chemisorption
Analyzer. Prior to the reduction in the TPR, the catalyst was
-
1
pretreated by being heated under a stream of argon (30 ml min )
1
820 | Green Chem., 2011, 13, 1819–1827
This journal is © The Royal Society of Chemistry 2011

View Article Online
Table 1 Textural properties of Ni-supported catalysts
BET surface Crystallite size
area (m g ) from XRD (nm)
Degree of
reducibility (%)
2
-1
Catalyst
Fresh Ni/Al
2
O
2
3
150.5
133.4
21.5
17.7
12.1
31.5
82.2
51.4
78.9
—
Fresh Ni/SiO
Used Ni/Al
Used Ni/SiO
2
O
3
2
12.9
—
column for CH
4
and CO evaluation. Mass balances were 100 ±
x
5
%.
3
. Results and discussion
3
.1 Catalyst characterisation
3
.1.1 BET surface area. Table 1 compiles the BET specific
surface area of the fresh and used Ni-supported catalysts.
Results showed that the surface area for the Ni/Al was
about 15% higher than Ni/SiO . As observed from the data,
2
O
3
2
there is a significant decrease in surface area between the fresh
and used catalysts. This decrease can probably be attributed
to the sintering of the reduced Ni particles which is caused by
high temperatures and the nature of the liquid phase of the
reaction which can result in carbon deposits on the catalyst
15
surface. This was also confirmed by XRD, for which Ni/SiO
2
showed crystallite size growth from 12 nm to 81 nm to form
agglomerates. The sintering of the Ni particles on the support
can contribute to the deactivation of the catalyst, due to the
reduction of the accessible catalytic surface and the blocking of
16
pores.
Fig. 1 (a): XRD pattern of the fresh Ni/Al
pattern of the used Ni/Al catalyst.
2
O
3
catalyst. (b): XRD
2
O
3
3
.1.2 Powder X-ray diffraction. Fig. 1 illustrates the
diffractograms obtained for the fresh Ni/Al and the used
Ni/Al catalysts. The XRD patterns for the fresh Ni/Al
2
O
3
2
O
3
2
O
3
catalyst shows NiO lines at 2q = 37.3 (d = 2.35), 44.3 (d = 2.09)
and 62.9 (d = 1.45) confirming the presence of “free” nickel
oxide.
3.1.3 Temperature-programmed reduction. TPR character-
isation is often used to analyse catalysts for surface Ni species,
20
24
reducibility and metal–support interaction. Fig. 3 shows the
The used Ni/Al
.03), 51.8 (d = 1.76), 76.3 (d = 1.25) and 89.1 (d = 1.09) which
were assigned to the metallic Ni phase, and peaks with 2q =
2
O
3
catalyst showed peaks with 2q = 44.5 (d =
TPR profile of fresh Ni/Al
reduction peaks (198, 280, 570, 700 and 800 C). Peaks observed
at temperatures lower than 400 C can be attributed to NiO not
2
O
3
, and this catalyst exhibited five
◦
◦
2
◦
1
9.9 (d = 4.60), 32.0 (d = 2.81), 36.2 (d = 2.41), 46.9 (d = 1.99),
associated or weakly associated with the support. The reduction
peaks at higher temperatures can be associated with the strong
◦
60.0 (d = 1.52), 61.1 (d = 1.41) and 83.5 (d = 1.15) which were
17
20
assigned to the alumina phase.
interaction of the NiO and the support. Rynkowski et al.
◦
The XRD patterns of the fresh Ni/SiO
2
and the used Ni/SiO
2
reported that the peak at 700 C is representative of the reduction
catalysts are shown in Fig. 2. The broad diffraction peaks can be
of nickel that had reacted with the support, forming nickel
18
◦
attributed to the presence of amorphous silica in the catalysts
aluminate, NiAl
2
O
4
, and the peak at 800 C is due to nickel
◦
21
22
◦
which is seen at 2q = 21.7 .
aggregates. Feng et al. also confirmed that the peak at 700 C
XRD analysis shows that the diffraction lines corresponding
to the NiO are slightly sharper on the Ni/Al than on the
Ni/SiO . This indicates the effect of the different support on
the size of the Ni crystallites, with Al favouring better
crystallinity and formation of larger Ni crystallites.
calculated Ni particle sizes (Scherrer equation), presented in
Table 1, indeed show that Ni particles supported on Al were
larger than Ni supported on SiO
is due to the reduction of a NiAl
2
O spinel or strongly interacting
4
2
O
3
NiO–g-Al
2
O
3
phase.
23
2
Yang et al. studied the effect of Ni loadings on the surface
Ni species. They observed, most importantly, that the number
of Ni species increased with increased Ni loading. The distinct
difference in the number of reduction peaks indicates that the
Ni loading has a prominent effect on the Ni species.
2
O
3
1
9,15
The
2
O
3
2
.
Fig. 4 shows the TPR profile for fresh Ni/SiO
2
. The peak at
◦
The sharpening of the Ni metal peaks in the XRD patterns of
the used catalysts suggests improved crystallinity likely due to
240 C was due to highly dispersed NiO on the support. Mile
24
◦
et al. assigned the peak at 336 C to the nickel oxide with
18
◦
the sintering of the Ni particles. This supports the BET data.
little or no interaction with the support, the peak at 520
C
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1819–1827 | 1821

View Article Online
2 3
Fig. 3 TPR profile of fresh Ni/Al O catalyst.
Fig. 2 (a): XRD pattern of the fresh Ni/SiO
of the used Ni/SiO catalyst.
2
catalyst. (b): XRD pattern
2
Fig. 4 TPR profile of fresh Ni/SiO catalyst.
2
◦
to the “free” nickel oxide, and the peak at 645 C to a strong
metal–support interaction.
image of the fresh Ni/SiO
2
catalyst showed bright spots on the
particles, which can be attributed to the Ni crystallites exposed
3
.1.4 Temperature-programmed desorption. The acidity of
26
on the surface. The used Ni/SiO catalyst showed large clusters,
2
the alumina or silica plays an important role in hydrogenolysis.
Table 2 gives a range of acid sites for the two different systems.
The Ni/Al
whereas, the Ni/SiO
62 C. Pattamakomsan et al. reported that their Al
supported catalyst showed desorption peaks around 170
and 340 C which corresponded to weak and medium acid
sites respectively. An additional peak was observed near 600 C
indicating strong acid sites of the Al
it is clear that the Ni/SiO catalyst had a slightly higher total
acidity, but showed a higher distribution of weak acid sites. The
Ni/Al catalyst showed a higher content of strong acid sites.
which can be attributed to sintering, and this was also reflected
in the BET measurements.
◦
2
O
3
catalysts showed peaks at 182, 366 and 603 C
Fig. 6 shows the TEM images of the fresh Ni/SiO
the used Ni/SiO catalyst. The fresh Ni/SiO catalyst showed
large spherical particles with smaller particles along the edges.
The used Ni/SiO catalyst has a thin sheet-like appearance with
2
catalyst and
2
catalyst showed peaks at 237, 368 and
◦
25
2
2
6
2
O
3
◦
C
◦
2
no distinct presence of spherical particles. Evidence of sintering
is clearly visible in the TEM image, Fig. 6(b). The electron
◦
2
3
O support. From Table 2
diffraction pattern (Fig. 7) of the fresh Ni/SiO catalyst shows
2
2
a ring pattern which is an indication of either an amorphous
or polycrystalline nature of the catalyst. A ring pattern which is
made up of spots usually implies small particle sizes, which were
2
O
3
3
.1.5 Electron microscopy. Fig. 5 presents typical SEM
confirmed by XRD. The diffraction pattern of the used Ni/SiO
catalyst suggests greater crystallinity.
2
images of the fresh and used Ni/SiO catalysts. The SEM
2
Table 2 Acidity of the Ni-supported catalysts
-
1
Acidity (mmol NH
Weak acid sites
3
g )
Catalyst
Medium acid sites
Strong acid sites
Total acidity
Fresh Ni/Al
Fresh Ni/SiO
2
O
2
3
0.085
0.942
0.669
0.852
1.170
0.381
1.924
2.175
1
822 | Green Chem., 2011, 13, 1819–1827
This journal is © The Royal Society of Chemistry 2011

View Article Online
Fig. 5 (a): SEM image of the fresh Ni/SiO
of the used Ni/SiO catalyst.
2
catalyst. (b): SEM image
2
Fig. 6 (a): TEM images of the fresh Ni/SiO
of the used Ni/SiO catalyst.
2
catalyst. (b): TEM images
2
3
.1.6 Thermogravimetric analysis. The thermogravimetric
plots for both the fresh Ni/Al
given in Fig. 8. The initial weight loss of about 7% up to 200 C,
for both unreduced catalysts, is attributed to the loss of water.
2
O
3
and Ni/SiO
2
catalysts are
◦
3
.2 Catalytic testing
3.2.1 Effect of pressure on glycerol hydrogenolysis. The
◦
A very slow but constant decrease in weight up to 700 C was
observed for both fresh catalysts, and can be attributed to the
extent of hydrogenolysis using a Ni/SiO catalyst, as a function
2
◦
27
loss of organic materials from preparation.
Fig. 8 also shows the TGA curves for the used Ni/Al
Ni/SiO
which was attributed to the loss of water. The Ni/Al
showed a significant weight loss of 21.9% up to 700 C. This can
be due to the decomposition of by-products which are formed
of pressure at constant temperature (230 C), is shown in Table 3.
The data showed that pressure does not have a significant
influence on the glycerol conversion as similarly observed by
2
O
3
and
◦
2
catalysts. Again a weight loss was seen up to 200 C,
6
2
◦
O
3
catalyst
Dasari et al.; however, a change in the selectivities of the polyols
was observed. The selectivity to the lower alcohols remained
essentially unchanged. The products included 1,2-propanediol,
acetol, ethylene glycol, ethanol and methane.
during the catalytic reaction and adsorb on the catalyst. The
◦
Ni/SiO
2
catalyst showed a weight loss of 9.9% up to 700 C,
The main product observed during glycerol hydrogenolysis
◦
and then a second weight loss of 8.3% up to 900 C, which is
due to the deposition of carbonaceous material.
over the Ni/SiO catalyst at the conditions expressed in Table 3
2
was 1,2-PDO. The selectivity to 1,2-PDO decreased with in-
creasing hydrogen pressure from 78% to 67%. At low pressures,
a selectivity of ~5% acetol was obtained, which decreased to
2% as the pressure increased. It is claimed that acetol is the
intermediate in the formation of 1,2-PDO using acid catalysts
The DSC plots for the fresh Ni/Al
2
O
3
and Ni/SiO
2
catalysts
are also shown in Fig. 8. For the Ni/Al
small endothermic peak below 200 C is attributed to the loss
of water, which is supported by the corresponding weight loss
2
O
3
catalyst, an initial
◦
4
in the TGA curve. A second small endothermic peak at about
via dehydration and subsequent hydrogenation. The selectivity
◦
3
00 C can be due to the loss of more water or due to the loss of
to 1,2-PDO showed a decrease with increasing pressure to the
advantage of EG. The higher EG selectivity is likely due to the
organic materials. An endothermic effect was observed for the
Ni/SiO catalyst corresponding to the dehydration of silica and
water evaporation in the temperature range 60–400 C.
29
2
degradation of 1,2-PDO to EG at high H pressures. It is also
2
◦
28
likely that the higher pressures favour direct hydrogenolysis of
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1819–1827 | 1823

View Article Online
Fig. 7 (a): Electron diffraction patterns of the fresh Ni/SiO
2
catalyst.
(
b): Electron diffraction patterns of the used Ni/SiO catalyst.
2
Table 3 Effect of pressure on glycerol hydrogenolysis over a Ni/SiO
2
◦
catalyst at 230 C
Selectivities to different
products (C mole basis)
4
0 bar H
2
60 bar H
2
75 bar H
2
Methane
Methanol
Ethanol
1.2
1.3
5.5
1.2
4.5
7.3
77.9
1.1
14.5
3.6
1.7
5.4
0.6
4.8
2.1
4.2
1.1
1
-Propanol
Acetol
Ethylene glycol
4.1
1.9
13.5
70.6
0.5
15.9
67.8
2.2
1
,2-PDO
a
Others
Conversion
16.2
13.0
a
Others = CO
2
, 1,3-PDO, condensation products, unknowns.
4
glycerol to EG. The increase in methane selectivity also supports
this pathway to EG.
Fig. 8 (a): TGA-DSC curve of the fresh Ni/Al
DSC curve of the used Ni/Al
2
O
3
catalyst. (b): TGA-
2
O
3
catalyst. (c): TGA-DSC curve of
the fresh Ni/SiO
catalyst.
2
catalyst. (d): TGA-DSC curve of the used Ni/SiO
2
3
.2.2 Effect of temperature on glycerol hydrogenolysis.
Table 4 shows the conversion of glycerol and selectivity to
products over the Ni/Al
at constant pressure (60 bar). From the results at 230 C, 1,2-
PDO, acetol and ethanol were observed as the major products.
2
O
3
system as a function of temperature
◦
The conversion was observed to increase as the temperature
◦
◦
◦
increased from 230 C to 320 C. At 230 C, the highest
1
824 | Green Chem., 2011, 13, 1819–1827
This journal is © The Royal Society of Chemistry 2011

View Article Online
◦
Table 4 Effect of temperature on glycerol hydrogenolysis over a
Ni/Al catalyst at 60 bar H
was increased to 320 C, which was accompanied by an increase
in the selectivity to ethanol and propanol. A total selectivity
of 69% to lower alcohols was obtained (methanol included).
Higher temperatures also produced higher selectivity to gaseous
products like methane, ethane, propane, carbon monoxide and
carbon dioxide, as well as heavy constituents. This was also
2
O
3
2
Selectivities to different products (C
mole basis)
◦
◦
◦
◦
◦
2
30 C
250 C
275 C
300 C
320 C
30
observed by Tomishige and co-workers, where at higher reac-
tion temperatures the selectivity to degradation products like
ethanol, methanol and methane over SiO
increased significantly.
Methane
CO
2.4
0.0
0.3
0.7
9.0
0.0
0.5
3.2
3.1
79.9
0.0
0.9
15.9
1.7
0.0
0.4
0.8
8.4
0.4
2.4
11.7
1.0
70.2
0.3
2.7
25.6
1.4
0.2
0.2
0.8
10.6
2.1
12.1
22.1
1.2
36.7
4.8
7.8
1.5
2.0
1.2
0.9
11.4
8.1
24.4
20.4
1.0
12.1
6.2
10.8
83.7
1.8
3.2
2.0
2.4
16.8
11.9
35.3
12.2
0.4
1.8
2.2
CO
2
2
-supported catalysts
Methanol
Ethanol
Acrolein
A conversion hysteresis (Fig. 9) was observed as temperature
◦
◦
1
-Propanol
was increased from 230 C to 320 C, for the different catalysts,
in favour of Ni/SiO
Acetol
Ethylene glycol
2
.
1
1
,2-Propanediol
,3-Propanediol
a
Others
10.0
96.1
Conversion
39.3
a
Others = ethane, propane, acetaldehyde, propenol, 2-propanol, con-
densation products, unknowns.
selectivity towards 1,2-PDO of 80% was achieved with a
selectivity of 9% to ethanol.
As the temperature increased, the 1,2-PDO selectivity de-
creased from 80% to 2%, while an increase in ethanol selectivity
was observed from 9% to 17%. At the same time, the 1-
propanol selectivity was observed to increase to 35%, as were
the selectivities to the dehydration products, acetol and acrolein.
The heavier by-products also increased from <1% to 9% as the
temperature increased.
2 2 3
Fig. 9 Conversion hysteresis for Ni/SiO and Ni/Al O systems.
The extent of hydrogenolysis using a Ni/SiO
a function of temperature at constant pressure (60 bar H
is shown in Table 5. The major products observed at 230 C
included 1,2-PDO, 1-propanol and ethanol.
2
catalyst, as
),
◦
At 300 C, a quantitative conversion was seen over the
2
◦
Ni/SiO
Ni/Al
The higher activity of the Ni/SiO
2
, while a conversion of 83.7% was observed over the
2
O
3
catalyst.
2
catalyst can be ascribed to
The data showed that temperature again had a significant
effect on the glycerol conversion. As the temperature of the
a higher density of Ni active sites for reaction. This is thought to
occur because of a weaker metal–support interaction, and thus
improved reducibility when compared to Ni/Al
by the TPR data.
Although both sintering and carbon deposits could affect the
activity of the catalyst over time, no significant deactivation of
the catalysts was observed over a period of 72 h.
◦
◦
reaction increased from 230 C to 320 C there was a uniform
increase in the glycerol conversion from 16% to 99%. The overall
selectivity to 1,2-PDO decreased significantly as the temperature
2
3
O , as suggested
Table 5 Effect of temperature on glycerol hydrogenolysis over a
Ni/SiO catalyst at 60 bar H
2 2
A possible explanation for the different product selectivities
observed for the two systems can be related to metal–support
interaction and the difference in acidity of the support.
The selectivity to 1,2-PDO (80%) was higher over the alumina-
supported catalyst than over the silica-supported catalyst (71%)
at similar conversion (~16%). This is likely due to the higher
Selectivities to different products (C
mole basis)
31
◦
◦
◦
◦
◦
2
30 C
250 C
275 C
300 C
320 C
Methane
CO
3.6
0.0
0.3
1.7
5.4
0.0
0.6
4.1
13.5
70.6
0.0
0.2
16.2
2.8
0.0
0.2
1.3
6.6
0.0
6.3
3.8
6.6
69.1
0.3
3.0
34.9
2.1
0.9
0.3
1.5
12.4
1.1
14.2
12.4
3.3
47.8
0.3
3.7
2.6
2.8
1.0
2.7
16.9
2.6
36.6
7.7
0.3
19.0
0.3
3.0
4.4
1.8
5.5
20.2
3.9
42.8
4.2
0.5
4.6
0.6
8.5
99.9
concentration of strong acid sites on the Ni/Al
2
O
3
compared to
CO
2
Methanol
Ethanol
Acrolein
Ni/SiO , which favoured the dehydration of glycerol to acetol
2
and subsequent hydrogenation of acetol to 1,2-PDO.
A higher EG selectivity (13%) was observed over the Ni/SiO
2
1
-Propanol
catalyst than the Ni/Al
2
3
O catalyst, likely due to the higher
Acetol
Ethylene glycol
propensity for C–C cleavage over the Ni/SiO catalyst. It has
2
1
1
,2-Propanediol
been claimed that nickel has a strong ability for breaking C–C
bonds, which can lead to shorter-chain products. More available
,3-Propanediol
a
Others
7.5
99.8
Conversion
69.3
Ni sites for the Ni/SiO catalyst may suggest a reason for the
2
higher EG selectivity for this system.
The selectivity to propanol and ethanol increased with
increasing temperature, with a subsequent decrease in EG and
a
Others = ethane, propane, acetaldehyde, propenol, 2-propanol, con-
densation products, unknowns.
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1819–1827 | 1825

View Article Online
1
,2-PDO selectivity. A similar trend was observed by Balaraju
Table 6 Difference in product selectivities as a function of conversion
29
over Ni/SiO
2
and Ni/Al
2
O
3
systems
et al. over a Ru/C catalyst. They observed that as the glycerol
conversion increased, the selectivity to 1,2-PDO decreased and
subsequently the selectivity to lower alcohols like ethanol and
methanol increased. However, it is generally proposed that the
Product selectivities (%)
a
b
Catalyst
Conversion (%)
15.9
Alcohols
Dehydration products
4
formation of EtOH occurs most likely via 1,3-PDO. The low
Ni/Al
2
O
3
10.2
11.6
23.5
36.7
54.5
8.6
14.3
28.1
56.3
68.5
3.2
selectivity to 1,3-PDO observed may suggest that this is not
necessarily the dominant pathway to EtOH. It is likely that due
to the high concentration of 1,2-PDO present in the product,
that this may be the likely intermediate to the propanols as well
as EtOH.
2
5.6
12.1
24.2
28.5
24.1
0.1
1.1
3.7
39.3
8
9
3.7
6.1
Ni/SiO
2
16.2
34.9
69.3
As the conversion over the two systems increased, a higher
propensity for dehydration was observed for the alumina-
9
9
9.8
9.9
7.2
8.5
32
supported catalyst, whilst a higher hydrogenation rate was
observed for the Ni/SiO catalyst. This is shown in Table 6,
which shows an increase in dehydration products, such as
acrolein and acetol, for the Ni/Al and alcohols, such as
ethanol and propanol, for Ni/SiO
The lighter compounds, CH , CO and CO
selectivity profile for both catalytic systems. The formation
of CO and CO is proposed to occur via decarbonylation
reactions of aldehydic intermediates coupled to the water gas
a
b
Alcohols = methanol, ethanol and propanol. Dehydration products =
2
acetol, acrolein and propenol.
2
O
3
2
.
The possible transformation of glycerol can occur via several
4
4
2
, showed a similar
pathways as shown in Scheme 1. The first, resulting from acid-
catalysed dehydration, can form 3-hydroxypropionaldehyde (A)
and acetol (B), which can be subsequently hydrogenated to
1,3-PDO and 1,2-PDO respectively. It has been reported that
glycerol can also be transformed via hydrogenolysis to form 1,3-
PDO (C), EG (D) and 1,2-PDO (E),
are of the opinion that these routes are not direct routes.
The subsequent diols can undergo further hydrogenolysis to
form primary alcohols such as 1-propanol (F), ethanol (G) and
methanol (H), as well as secondary alcohols like isopropanol
2
shift reaction. CH
or MeOH.
4
likely forms via direct hydrogenation of CO
3,33,34
although some authors
35,36
However, condensation to give heavier products, which in-
cluded ethers and aldol type compounds, was more pronounced
over the alumina-supported catalyst. This can be ascribed to an
increased rate of condensation due to the higher (strong) acidity
of the alumina support.
4
(J). C products such as methane (I), CO and CO can form via
1
2
Scheme 1 Reaction scheme showing glycerol transformation.
826 | Green Chem., 2011, 13, 1819–1827
1
This journal is © The Royal Society of Chemistry 2011

View Article Online
decomposition reactions or the water gas shift reaction. Heavier
products can form via condensation reactions of glycerol or the
aldehydic intermediates. In addition to the pathways described
in Scheme 1, we propose that ethanol is also likely to form via
4 T. Miyazawa, Y. Kusunoki, K. Kunimori and K. Tomishige, J. Catal.,
2
006, 240, 213–221.
5
R Garcia, M. Besson and P. Gallezot, Appl. Catal., A, 1995, 127,
1
65–176.
6 M. A. Dasari, P. P. Kiatsimkul, W. R. Sutterlin and G. J. Suppes,
1
,2-PDO.
Appl. Catal., A, 2005, 281, 225–231.
7
8
P. Bloom, Archer Daniels Midland Company, US 0103339 (2008).
R. Bullock, P. J. Fagan, E. M. Hauptman and M. Schalf (E. I. Du
Pont de Nemours and company), World Pat. WO 98241, 2001.
L. Huang, Y. L. Zhu, H. Y. Zheng, Y. W. Li and Z. Y. Zeng, J. Chem.
Technol. Biotechnol., 2008, 83, 1670–1675.
4
. Conclusion
9
The effect of increasing pressure was observed to have little
influence on glycerol conversion. However, the selectivity to 1,2-
1
1
0 E. Drent and W. W. Jager (Shell Oil Company), US Pat. 6080898,
2
000.
PDO decreased with increasing pressure while EG and CH
4
1 I. Furikado, T. Miyazawa, S. Koso, A. Shimao, K. Kunimori and K.
selectivity was observed to increase. The selectivity to lower
alcohols remained essentially unchanged.
Temperature was observed to have a significant effect on the
Tomishige, Green Chem., 2007, 9, 582–588.
12 P. D. Pavlechko, PEP Report No. 268: Higher alcohols from syngas,
SRI Consulting, 2009.
1
1
3 E. Linak, CEH Marketing Research Report: Ethanol, 2009.
4 S. Bizarri, M. Blagoev, and H. Mori, CEH Marketing Research
Report: Glycerin, 2008.
glycerol conversion with Ni/SiO
2
, giving a higher conversion
than Ni/Al except at the lowest temperature studied. With
2
O
3
1
1
5 E. S. Vasiliadou, E. Heracleous, I. A. Vasalos and A. A. Lemonidou,
Appl. Catal., B, 2009, 92, 90–99.
6 S. Rakass, H. Oudghiri-Hassani, P. Rowntree and N. Abatzoglou,
J. Power Sources, 2006, 158, 485–496.
respect to selectivity, 1,2-PDO was dominant at low tempera-
ture, whilst alcohols such as propanol, ethanol and methanol
dominated the product profile at higher temperature. At 320 C,
◦
a quantitative conversion of glycerol was observed, with a total
selectivity of 69% to lower alcohols which included 1-propanol,
ethanol and methanol. To the best of our knowledge, this is
the highest selectivity reported to lower alcohols from glycerol
using a continuous-flow fixed-bed reactor with an inexpensive
catalytic system.
The pathway to the lower alcohols is proposed to occur via
degradation of the intermediate polyols such as 1,2-PDO, 1,3-
PDO and possibly EG. A mechanistic study is in progress to
establish the route to the lower alcohols via glycerol.
Comparison of the product selectivity of the two catalysts
revealed that the intrinsic properties of the support, such as the
acidity of the support and the metal–support interaction, may
have had an influence.
17 S. Shang, G. Liu, X. Chai, X. Tao, X. Li, M. Bai, W. Chu, X. Dai, Y.
Zhao and Y. Yin, Catal. Today, 2009, 148, 268–274.
1
8 Z. Huang, F. Cui, H. Kang, J. Chen and C. Xia, Appl. Catal., A,
2
009, 366, 288–298.
19 A. Saadi, R. Merabti, Z. Rassoul and M. M. Bettahar, J. Mol. Catal.
A: Chem., 2006, 253, 79–85.
2
2
2
2
0 J. M. Rynkowski, T. Paryjczak and M. Lenik, Appl. Catal., A, 1993,
1
06, 73–82.
1 P. Kim, Y. Kim, H. Kim, I. K. Song and J. Yi, J. Mol. Catal. A:
Chem., 2005, 231, 247–254.
2 J.-T. Feng, Y.-J. Lin, D. G. Evans, X. Duan and D.-Q. Li, J. Catal.,
2
009, 266, 351–358.
3 R. Yang, X. Li, J. Wu, X. Zhang, Z. Zhang, Y. Cheng and J. Gou,
Appl. Catal., A, 2009, 368, 105–112.
24 B. Mile, D. Striling and M. A. Zammitt, J. Mol. Catal., 1990, 62,
1
79–198.
2
2
5 K. Pattamakomsan, K. Suriye, S. Dokjampa, N. Mongkolsiri, P.
Praserthdam and J. Panpranot, Catal. Commun., 2010, 11, 311–316.
6 A. E. Aksoylu, Z. Misirli and Z. I. Onsan, Appl. Catal., A, 1998, 168,
3
85–397.
Acknowledgements
27 D. N. Srivastava, N. Perkas, G. A. Seisenbaeva, Y. Koltypin, V. G.
Kessler and A. Gedanken, Ultrason. Sonochem., 2003, 10, 1–9.
28 G. A. Sharpataya, G. P. Panasyuk, G. P. Budova, Z. P. Ozerova, I. L.
Voroshilov and V. B. Lazarev, Thermochim. Acta, 1985, 93, 271–274.
2
3
3
3
3
All catalytic work was performed at Sasol Technology Research
and Development, and catalyst characterisation was carried out
at the University of KwaZulu-Natal. We would also like to thank
the Electron Microscope Unit as well as the Catalysis Research
Group at the University of KwaZulu-Natal for the help with the
characterisation data.
9 M. Balaraju, V. Rekha, P. S. Sai Prasad, B. L. A. Prabhavathi Devi,
R. B. N. Prasad and N. Lingaiah, Appl. Catal., A, 2009, 354, 82–87.
0 I. Furikado, T. Miyazawa, S. Koso, A. Shimao, K. Kunimori and K.
Tomishige, Green Chem., 2007, 9, 582–588.
1 L. Huang, J. Xie, R. Chen, D. Chu and A. T. Hsu, Mater. Res. Bull.,
2
010, 45, 92–96.
2 R. J. J. Nel and A. de Klerk, Ind. Eng. Chem. Res., 2009, 48, 5230–
References
5238.
3 I. Gandarias, P. L. Arias, J. Requies, M. B. Guemez and J. L. G.
Fierro, Appl. Catal., B, 2010, 97, 248–256.
1
2
3
Y. Zheng, X. Chen and Y. Shen, Chem. Rev., 2008, 108, 5253–
277.
T. Hirai, N. Ikenaga, T. Miyake and T. Suzuki, Energy Fuels, 2005,
5
34 J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel and
C. Rosier, Green Chem., 2004, 6, 359–361.
35 C. Montassier, J. C. Menezo, L. C. Hoang, C. Renaud and J. Barbier,
J. Mol. Catal., 1991, 70, 99–110.
36 E. P. Maris and R. J. Davis, J. Catal., 2007, 249, 328–337.
1
9, 1761–1762.
J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel and
C. Rosier, Green Chem., 2004, 6, 359–361.
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1819–1827 | 1827