G. Onyestyák et al. / Catalysis Communications 16 (2011) 184–188
185
short chain fatty alcohols can be obtained. Latter ones may be used
as transportation fuel replacing gasoline in contrast to short chain
n-paraffins obtainable by total hydrodeoxygenation. Long chain fatty
alcohols, which can be produced with lower hydrogen consumption
than synthetic hydrocarbons may be admixed to gasoil.
9 wt.% Ni. Composite catalysts were prepared by adding indium
(III) oxide to the Ni/alumina samples and grinding the mixture in
agate mortar. 163 m2/g was measured for this composite which spe-
cific surface area value means 206 m2/g related only to alumina indi-
cating that the surface increased after introducing the metal. Adkins
catalyst was a commercial product obtained from Süd-Chemie AG
(consisting of 72 wt.% CuCr2O4 and 28 wt.% CuO).
2. Experimental
X-ray patterns were recorded by a Philips PW 1810/1870 diffrac-
tometer with monochromatic Cu Kα radiation at elevated tempera-
tures using a high-temperature XRD cell (HT-XRD). The mean
crystallite size of the nickel particles was estimated by the Scherrer
equation [18]. The reducibility of the Ni2+ and In3+ cations in the
catalyst samples was investigated by temperature-programmed re-
duction (H2-TPR) in H2/Ar flow using a conventional TPR apparatus
equipped with a heat conductivity detector.
The catalytic hydrogenation of octanoic acid (OA)/as a model
compound with medium chain length/was studied in a high-
pressure flow through reactor in hydrogen atmosphere at 21 bar
total pressure (in general 20 bar hydrogen and 1 bar octanoic acid
partial pressures were applied) and 240–360 °C. The catalysts
have been pretreated in hydrogen flow in situ in the reactor at
450 °C and 21 bar for 1 h. The product mixtures, the effluents
were analyzed by GC (Shimadzu 2010) equipped with flame ioniza-
tion detector and a CP-FFAP CB capillary column. The activity and
selectivity of the catalysts studied can be characterized by the prod-
uct distributions represented by stacked area graphs as function of
different reaction parameters. Notice, that the distance between
two neighboring curves represents the concentration of a given
product in weight percents (Figs. 1, 3, 4, 7, 8).
3. Results and discussion
Octanoic acid hydroconversion on monometallic Ni/Al2O3 catalyst
shows selective hydrodecarbonylation resulting in heptane forma-
tion (Fig. 1A). CO formed beside water in hydrodecarbonylation is
completely hydrogenated to methane and some excess methane can
be also formed due to hydrogenolysis of heptane resulting in lower
hydrocarbons (e.g. hexane). On adding a large amount of In2O3
(10 wt.% In2O3 of the mass of monometallic Ni/Al2O3 catalyst) as a
co-catalyst the activity strikingly increases (Fig. 1B), and is main-
tained at a constant level simultaneously the selectivity changes con-
siderably (Fig. 1B) by elimination of hydrodecarbonylation and H2
consuming side-reactions. The transformation thus proceeds in a dif-
ferent consecutive deoxygenating pathway. Instead of hydrodecarbo-
nylation a step by step reduction with H2 proceeds and the series of
possible reactions are stopped at alcohol formation before mono-
molecular dehydration to alkenes (or bimolecular dehydration to
dioctyl-ether) and saturation of the formed olefins should take place:
γ-Al2O3 (Ketjen CK 300, Akzo-Chemie, BET: 199 m2/g) activated
at 550 °C was impregnated with NH4OH solution (Reanal Finechem-
ical Co.) of Ni (acetate)2 (Aldrich), dried, and calcined at 550 °C. After
impregnation the specific surface area changed to 180 m2/g however
this value corresponds to 207 m2/galumina. The sample designation,
e.g., 9Ni/Al2O3 formula represents a catalyst preparation containing
100
A
80
C8AC
OP
60
þH2
þH2
octanoic acid −H octanal → octanol −H
→
2O
octenes ðor dioctyl−etherÞ þ→H2 octane:
→
2O
CH4
H2O
C7-
40
20
100
80
60
40
20
0
While the carboxylic acid hydroconversion to alkanes increases
with the Ni-content of Ni/Al2O3 samples, for In2O3 doped samples at
constant indium content the conversion can be lower when the
Ni-content is above an optimum value (see Fig. 2). Decrease of nickel
loading below 9 wt.% is also disadvantageous resulting in the down-
fall of conversion and octanol yield followed by an increase of octanal
C8OL
C8AC
OP
100
H2O
(C8)2O
80
B
C
60
B
C8OL
A
40
C8=
20
0
0
2
4
6
8
10
Time on stream, h
240
260
280
300
320
Fig. 1. Octanoic acid hydroconversion over 9Ni/Al2O3 (A) and dopped with +10% In2O3
(B); catalysts are characterized by distributions of the main products as a function of
time-on-stream at 300 °C and 21 bar using stacked area graphs. WHSV of OA was
2.0 h−1. Notice, that the distance between two neighboring curves represents the con-
Temperature, °C
Fig. 2. Hydroconversion of octanoic acid characterized by the target product over
3Ni/Al2O3 +10% In2O3 (A), 9Ni/Al2O3 +10% In2O3 (B) and 18Ni/Al2O3 +10% In2O3
(C) catalysts; ocatanol yields are plotted between 240 and 320 °C at 21 bar total pres-
centration of a given product in weight percents. (symbols: C7−: heptane, C8=
:
octenes, C8OL: octanol, CH4: methane, (C8)2O: dioctyl ether, H2O: water, C8AC: octa-
noic acid, OP: other products).
sure. WHSV of OA was 2.0 h− 1
.