Gas-phase hydrogenation of propionitrile on copper-lanthanide oxides

Article abstract of DOI:10.1016/j.molcata.2009.03.007

The hydrogenation of propionitrile on copper-lanthanide oxide catalysts (2Cu·CeO₂ and 4Cu·Ln₂O₃ (Ln = La, Pr, Nd)) was studied in the gas phase. The activity of the catalysts varies with the lanthanide in the order 2Cu·CeO

Full text of DOI:10.1016/j.molcata.2009.03.007

Journal of Molecular Catalysis A: Chemical 307 (2009) 37–42
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Gas-phase hydrogenation of propionitrile on copper-lanthanide oxides
Joaquim Badalo Branco , Danielle Ballivet-Tkatchenko , António Pires de Matosa
a,∗
b
a
Instituto Tecnológico e Nuclear, Unidade de Ciências Químicas e Radiofarmacêuticas, Estrada Nacional 10, 2686-953 Sacavém Codex, Portugal
Faculté des Sciences Mirande, Université de Bourgogne, 9 Avenue A. Savary, 21078 Dijon Cedex, France
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 July 2008
Received in revised form 22 January 2009
Accepted 6 March 2009
Available online 20 March 2009
The hydrogenation of propionitrile on copper-lanthanide oxide catalysts (2Cu·CeO and 4Cu·Ln O
2
2
3
(
Ln = La, Pr, Nd)) was studied in the gas phase. The activity of the catalysts varies with the lanthanide
in the order 2Cu·CeO2 > 4Cu·Pr2O3 > 4Cu·La2O3 ≥ 4Cu·Nd2O3, while the activation energies varies in the
opposite order, except for 2Cu·CeO2. The main product was the primary amine, n-propylamine. The for-
mation of the unstable imine CH3CH2N CHCH3 as a major product over 2Cu·CeO2 seems to be consistent
with the acidity of the catalyst. The catalysts were more selective than conventional copper impregnation
catalysts, Cu (10 wt.%) on SiO2, La2O3 or CeO2, that produce significant quantities of the secondary amine.
Keywords:
Copper
Supported catalysts
Propionitrile
©
2009 Elsevier B.V. All rights reserved.
Gas-phase hydrogenation
Primary amine
1
. Introduction
order to fully understand the mechanism of the gas-phase hydro-
genation of nitriles. For example, Volf and Pasek [12] noted
that on supported nickel catalysts the support affects the activ-
ity of nickel without changing the catalysts selectivity. Others
like Huang et al. [22] reported that the support acidity has no
effect on the catalysts selectivity, corroborating this observa-
tion.
The catalytic hydrogenation of nitriles is one of the most impor-
tant methods for preparing amines [1–3]. The reactions are usually
accomplished in the presence of heterogeneous transition metal
catalysts and the product distribution depends primarily on the
catalyst nature: cobalt and nickel catalysts are preferred for the
production of primary amines [4–13], copper and rhodium catalysts
gives predominantly secondary amines [7,14–19], and platinum and
palladium catalysts are quite selective for the production of tertiary
amines [7,9,17–19].
Nevertheless, the results obtained on the gas phase are clearly
different from those obtained on the liquid-phase. Rode et al. [31]
showed that supported platinum catalysts are highly selective for
the formation of diethylamine, whereas on liquid-phase reactions
the main product is triethylamine. Furthermore, the preparation
method has a strong influence on the catalyst selectivity [20,32,33].
As an example, Arai and co-workers [32] reported that Pd-based
catalysts supported on ZrO , CeO , MgO, SiO , Al O , ZnO, Ga O
The gas-phase hydrogenation of nitriles to produce amines
[
20–23] is an alternative to the classical liquid-phase reaction.
The results obtained are different from those of the liquid-phase,
namely what appear to be a better control of the selectivity by a
suitable choice of the active phase and support [21,24–26].
The nature of the support seems to have a strong influence
on the reaction selectivity, especially their acid–base properties.
This fact leads to the proposition of a bifunctional mechanism, in
which active sites for the hydrogenation are located on the metal,
and strong basic sites appear to increase the selectivity to primary
amines [26–29], while the acidic function of the support is responsi-
ble for the condensation reactions leading to secondary and tertiary
amines [27,30].
2
2
2
2
3
2
3
and In O3 were very selective for the gas-phase hydrogenation
2
of acetonitrile to ethylamine and such enhanced selectivity was
ascribed to the formation of Pd alloys PdZn, Ga5Pd, GaPd5, etc.
and to the synergism between the two metals. Coq et al. [26]
also reported an increase in the methylamine selectivity using
Co/Ni/Mg/Al layered double hydroxides as catalytic precursors and
attributed it to the formation of NiCo alloys. The lowest selectivity
to by-products is reached in the presence of bimetallic NiCo cata-
lysts where the synergism effect between Ni and Co appears to play
a major role. Therefore, there is a great deal of interest in the search
of new catalytic precursors.
However, this is a controversial subject in the literature and
it is evident that a great deal of work is still to be done in
In this context, we have recently reported that binary inter-
metallic compounds of lanthanide and copper LnCu2 (Ln = La, Ce,
Pr, Nd) are active and selective for the gas-phase hydrogena-
tion of propionitrile into n-propylamine [34]. This result is clearly
∗ Corresponding author. Tel.: +351 219946116.
E-mail address: jbranco@itn.pt (J.B. Branco).
1381-1169/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2009.03.007

3
8
J.B. Branco et al. / Journal of Molecular Catalysis A: Chemical 307 (2009) 37–42
different from those obtained in the liquid-phase with copper
based catalysts that gives the secondary amines as major prod-
ucts.
Impregnation supported copper catalysts, 10 wt.%, were
obtained by the incipient wetness method using an aqueous solu-
tion of copper (II) acetate monohydrate on SiO , La O or CeO . The
2
2
3
2
Binary intermetallic compounds of lanthanide or actinide met-
als combined with d metals (namely Ni, Co, Mn, or Fe), which
have drawn the attention of many researchers due to their
catalytic properties [35–42], are attractive candidates for the gas-
phase hydrogenation of nitriles. The intermetallic compounds were
described as a new type of supported catalysts precursors, which
can be more active and selective than those obtained by con-
ventional routes [36,37,40–46]. The support (lanthanide oxide
phase) sems to play an important role for their catalytic proper-
ties.
reagents were used as supplied: (CH COO) Cu·H O (Fluka, purity
3
2
2
≥99.0%), La O and CeO (Fluka, purity ≥99.0%), and SiO (Degussa
2
3
2
2
200, purity 99.9%). After impregnation, water was removed and the
◦
residue was dried in an oven at 150 C for 4 h and then treated with
◦
3
hydrogen at 750 C (H , 50 cm /min). The copper content on each
2
catalyst was determined at the Service Central d’Analyse du CNRS
(Solaize, France) by inductively coupled plasma spectrometry (ICP).
Found (wt.%): Cu, 9.2 ± 0.3; Si, 40.7 ± 1.2; calculated for 10% copper
loading on SiO : Cu, 10; Si, 42.1. Found (wt.%): Cu, 8.3 ± 0.3; La,
2
68.1 ± 2.1; calculated for 10% copper loading on La O : Cu, 10; La,
2
3
In our laboratories we have used the binary intermetallic
compounds LnCu2 (Ln = La, Ce, Pr, Nd) [36], ThCu2 and AnNi2
76.7. Found (wt.%): Cu, 9.5 ± 0.3; Ce, 69.3 ± 2.1; calculated for 10%
2
copper loading on CeO : Cu, 10; Ce, 73.3. BET specific areas (m /g):
2
(
An = Cu, Ni) [37] as catalytic precursors and the oxidized prod-
196.0, 157.4, 110.0, 83.7 and 43.3 for SiO , 10% Cu/SiO , CeO ,
2 2 2
ucts were described as heterobimetallic copper-lanthanide oxides,
10% Cu/CeO2 and 10% Cu/La O3 respectively. The XRD diffraction
2
e.g. 3CuO·Ln CuO or 2CuO·ThO . Such compounds exhibited
patterns after reduction show only the pure crystalline phase of
metallic copper.
2
4
2
selectivity for the 4-methylpentan-2-ol decomposition and the
lanthanide-containing phase, CeO2 or Ln CuO , seems to play
2
4
a role in the formation of the CuO active sites [36]. Unpub-
lished results show also that the copper-actinide oxides are active
for the oxidative coupling of methane using carbon dioxide as
reagent.
2.2. Catalysts characterization
2.2.1. Temperature-programmed desorption of carbon dioxide
(
CO -TPD)
2
The CO -TPD was used to determine the basicity of the sup-
After reduction, these heterobimetallic oxides were described
2
ports and catalysts. Prior to adsorption of CO , the catalysts (50 mg)
were cleaned by passing a He stream (50 mL min ) at 250 C for
as supported copper catalysts on lanthanide oxides, 2Cu·CeO and
2
2
−
1
◦
−
4
Cu·Ln O , which were active and selective for the mesityl oxide
2
3
1
3
5
0 min. The solids were then saturated with CO2 (17 mL min ) at
0 C for 1 h, followed by a pure He stream (50 mL min ) at the
hydrogenation to 4-methylpentan-2-one (methylisobutylketone,
MIKB) [47]. Their catalytic activity and selectivity were associ-
ated with the lanthanide-containing phase that seems to play an
important role in the formation of the copper or nickel active
sites.
◦
−1
saturation temperature for 3 h. Once a stable line was obtained, the
◦
◦
chemisorbed CO was desorbed by heating from 50 C up to 1000 C
2
◦
−1
at a rate of 10 C min . Desorption was monitored by a Shimadzu
A gas chromatograph equipped with a thermal conductivity detec-
9
On that account, we report here the results obtained for the gas-
phase hydrogenation of propionitrile over 2Cu·CeO and 4Cu·Ln O
◦
◦
tor (TCD, Tcolumn = 105 C, T
= 105 C, 150 mA). The amount of
detector
2
2
3
carbon dioxide desorbed was taken as a measure of the number
of basic centers while the temperature range at which the carbon
dioxide is desorbed is an indicator of the strength of the basic sites.
(
Ln = La, Pr, Nd). Samples were characterized by powder X-ray
diffraction (XRD), temperature-programmed desorption of carbon
dioxide (CO -TPD) and oxidative dehydrogenation–dehydration of
2
◦
The basic sites were classified as weak (Tm below 400 C), medium
2
-propanol.
◦
strength (Tm between 400 and 600 C) and strong basic sites (Tm
◦
over 600 C) according to the different carbon dioxide maximum
rate desorption temperatures, Tm [48,49].
2. Experimental
2
.2.2. Dehydrogenation/dehydration of 2-propanol
The oxidative dehydrogenation–dehydration of 2-propanol was
2.1. Catalysts preparation
used as a test reaction for effective acidity of catalysts. The catalytic
activity tests were carried out in a U-shaped quartz reactor (plug-
flow type) at 250 C, with a glass frit and an inside volume of 15 cm ,
operating continuously at atmospheric pressure under a gaseous
The copper-lanthanide oxide catalysts (2Cu·CeO , 4Cu·Ln O
2
2
3
(
Ln = La, Pr, Nd)) were prepared from LnCu2 (Ln = La, Ce, Pr, Nd), as
◦
3
described earlier [34,36,47]. Briefly, the intermetallic compounds
were first oxidized under air (Air Liquide, O :N = 20:80 (vol%),
2
2
3
−1
◦
−1
flow of 0.75 vol% of alcohol in air (N /O : 80/20 vol%, Air Liquide K,
purity 99.995%, 17 cm min ) at a heating rate of 10 C min up
to 1000 C, followed by reduction under pure hydrogen (Air Liq-
uide, purity 99.99%, 17 cm min ) at a heating rate of 10 C min
up to 550 C. The reduction products were characterized by XRD
and the diffraction patterns indicate pure crystalline phase of Cu
2 2
◦
purity 99.9995%). Mass flow controllers were used to control air
and He (Air Liquide N60, purity 99.9995%) flows. A thermocou-
ple was placed near the catalytic bed for continuous monitoring
of the sample temperature. The reaction was studied with an ade-
quate gas hourly space velocity (GHSV) of 2000 mLi-PrOH/g h at
3
−1
◦
−1
◦
along with either CeO2 or Ln O3 (Ln = La, Pr, Nd). The charac-
2
◦
T = 250 C, m = 50–100 mg. The outlet gas composition was analyzed
terization of the reduction end products by elemental analysis
confirms the stoichiometries (Cu ≈ 40 wt.%) and the SEM/EDX mea-
surements have shown that the catalysts surface are enriched with
copper, except for 4Cu·La O (molar ratio between the d and f
by on-line gas chromatography with a Carbopack C/0.1% SP-1000
packed column (L = 2.0 m, f = 1/8 in., 80–100 mesh) and a Shimadzu
9
A instrument equipped with a flame ionization detector (FID). The
2
3
reaction was performed under oxidative conditions to avoid cata-
lyst deactivation, as the surface metal oxides can lose oxygen under
reducing condition. None of the samples showed diffusion restric-
tions (Weisz–Prater criterion). The gaseous effluent was monitored
every 60 min using a 6-port gas sampling valve with a 0.250 cm3
elements at the catalysts surface is equal to 1.5 for 4Cu·La O
2
3
and higher in the case of 2Cu·CeO , 4Cu·Pr O and 4Cu·Nd O
2
2
3
2
3
(
3.0, 4.0 and 4.0 ± 0.1, respectively)) [47]. The BET specific sur-
face areas were determined on an Accusorb 2100E apparatus from
◦
Micromeritics using Kr or N2 adsorption at −196 C. The samples
−
3
◦
loop. The reaction rate (r ) has been calculated as the number of
were pretreated undervacuum (10 mbar) at 523 Cfor15 h. Found
i
2
2
moles of 2-propanol converted per second and per m of catalyst
(
2
m /g): 4Cu·La O , 2.9; 2Cu·CeO , 6.5; 4Cu·Pr O , 3.2; 4Cu·Nd O ,
2
3
2
2
3
2
3
−
1
−2
m ).
(mol s
.4.

J.B. Branco et al. / Journal of Molecular Catalysis A: Chemical 307 (2009) 37–42
39
2.3. Catalyst reactivity
2.3.1. Hydrogenation of propionitrile
The experimental procedure used for the gas-phase hydro-
genation of propionitrile was identical to that used for the
oxidative dehydrogenation–dehydration of 2-propanol with their
own specificities: (i) the fixed bed U-shaped glass reactor oper-
ates continuously at atmospheric pressure under a gaseous flow
of nitrile in hydrogen, (ii) low nitrile conversion to keep the reac-
tion in the kinetic regime, (iii) the starting intermetallic compounds
were treated in situ and (iv) unless otherwise stated, the reaction
was studied with an adequate gas hourly space velocity (GHSV) of
◦
1
000 mLPropionitrile/g h at T = 200 C, m = 10–150 mg, partial pressure
of propionitrile (PNitrile) = 0.01 bar and partial pressure of hydrogen
(
PH ) = 0.99 bar.
2
The outlet gas composition was analyzed on-line by gas chro-
matography using a Gas Chrom Q column (L = 2.0 m, f = 1/8 in.,
0–100 mesh; liquid phase AT223 + KOH (28 + 4, w/w%)) and an
Fig. 2. Effect of the catalysts basicity in the propionitrile reaction rate to n-
propylamine.
8
Intersmat 120 DFB instrument equipped with a flame ionization
detector (FID). GC–MS (Hewlett Packard 5971, 70 eV, electronic
impact, column 25 m, Ø = 0.2 mm, Gaz Chrom Q (80–100 mesh)
coated at 10% with Carbowax 20 M (film of 1.3 mm)) was used to
identify all the collected reaction products.
tion energies varies in the oposite order, except for 2Cu·CeO2
(Table 1). These kinetic data could indicate that different catalytic
species are present which exclude the predominant role of the Cu
phase; the lanthanide phases, Ln O and CeO appear to contribute
2
3
2
The reaction rate (r ) has been calculated as the number of moles
to the catalytic behavior of copper-lanthanide oxides [36].
i
2
−1
−2
of nitrile converted per second and per m of catalyst (mol s
m ).
Acid–base properties of metal oxides play a key role in catalysis
and can be characterized by different methods, e.g. test-reactions
or probe molecules adsorption. In this work, we have measured
the total number of basic sites, its density and strength distribu-
tion on the copper-lanthanide oxides by temperature-programmed
desorption using carbon dioxide as a molecular probe for basic sites
The extent of deactivation is defined as the fraction of the initial
activity loss at steady state. Apparent activation energies (Ea) were
−1
obtained by plotting ln r against 1/T (K ), using the equation of
i
Arrhenius.
(
CO -TPD) [49,50] and we have used the 2-propanol (IPA) oxidative
3
. Results and discussion
2
dehydrogenation/dehydration for the characterization of the cata-
lysts surface redox properties [51,52]. For comparison, the results
obtained with Ln O (Ln = La, Pr, Nd), CeO , ThO and MoO are also
3.1. Activity and stability of catalysts
2
3
2
2
3
reported. ThO2 and MoO3 are well known basic and acid oxides,
respectively.
Fig. 1 shows that the activity of the copper-lanthanide oxides
(
2Cu·CeO and 4Cu·Ln O (Ln = La, Pr, Nd)) for the gas-phase hydro-
2
2
3
According to the total basic site density measured by
genation of propionitrile decreases significantly with time on
stream. In all cases, the loss of activity was higher than 80% and the
primary amine (n-propylamine) the only product observed under
our experimental conditions.
CO -TPD (Table 2), the catalysts can be ranked as follows:
2
4
Cu·Nd O = 4Cu·La O > 4Cu·Pr O ꢀ 2Cu·CeO . The metal oxides
2
3
2
3
2
3
2
that presented an acidic character were CeO , 2Cu·CeO and MoO ,
2
2
3
whereas ThO2 has a basic surface in agreement with the literature
This deactivation can be related either to the formation of coke
or to modifications on the catalysts surface, namely due to particle
sintering [22]. Nevertheless, other contributions such as poison-
ing attributed to amines strongly adsorbed on the catalysts that
cover the active sites and/or formation of an overlayer must not be
discarded [20–22,33].
[
51,53,54]. Therefore, the acid–base properties of 2Cu·CeO are sig-
2
nificantly different from those of the other copper-lanthanide metal
oxide tested is this work that have basic properties.
The basicity (total number of basic sites) of copper-lanthanide
metal oxides clearly leads to the increase in the activity of the cata-
lysts (Fig. 2). The total number of weak and medium strength sites
The catalytic activity varies with the lanthanide in the order
◦
(
Tm below 600 C) seems to have also a good correlation with the
2
Cu·CeO > 4Cu·Pr O > 4Cu·La O ≥ 4Cu·Nd O , while the activa-
2
2
3
2
3
2
3
increase in the reaction rate (the number of strong basic sites is
always very low, except for 4Cu·La O ). On the other hand, the
2
3
decrease in total basic site density correlates with the increase in
the reaction rate but did not modify the catalysts selectivity towards
the primary amine.
Using IPA test reaction, the catalysts can be classified by
their propensity towards the dehydration or the dehydrogenation
activity to propene (PPE) or acetone (ACE), respectively. From a
mechanistic point of view, the acid site catalyzes the dehydra-
tion of 2-propanol, whereas both acid and basic sites catalyze the
dehydrogenation through a concerted mechanism [52]. For a num-
ber of catalysts, good correlations were found between the acidity
measured by the adsorption of ammonia and the 2-propanol dehy-
dration rate measured and the basicity measured by adsorption of
CO2 and the ratio of the 2-propanol dehydrogenation to the dehy-
dration rate [51].
In this work, the main products were ACE and PPE, and the
formation of other products such as diisopropyl ether (DIPE) was
Fig. 1. Conversion of propionitrile over the copper-lanthanide oxides as a function
of time on stream at 175 C.
◦

4
0
J.B. Branco et al. / Journal of Molecular Catalysis A: Chemical 307 (2009) 37–42
Table 1
◦
Activity and selectivity of the copper-lanthanide oxides for the gas-phase hydrogenation of propionitrile at 250 C.
Catalysts
2Cu·CeO2
4Cu·Pr2O3
4Cu·La2O3
4Cu·Nd2O3
roa, 10 (mol m
−8
−2 −1
s )
s
7051 (45971)
1410 (9190)
100
5547 (17917)
853 (2754)
51
1659 (4762)
392 (1124
77
1502 (3576)
39 (94)
107
a
−8
−2 −1
)
rss , 10 (mol m
b
−1
)
Ea (kJ mol
a
−1 −1
s .
Reaction rate at start of run (ro) and at steady state (rss); between parentheses the values in mol gCu
Determined between 150 and 250 C.
b
◦
Table 2
3.2. Reaction conditions
CO2-TPD distribution and strength of basic sites over the copper-lanthanide oxides.
Catalysts
CO2-TPDa (a.u./g)
Total
The reaction rate of the gas-phase hydrogenation of propionitrile
over the copper-lanthanide oxide catalysts increases with the tem-
WSS
MSS
SSS
perature, residence time (ꢀ) and hydrogen partial pressure (PH ).
2
2
CeO2
Cu·CeO2
0
0
371
0
0
0
0
0
0
0
378
91
2986
0
139
5939
0
However, the selectivity was not affected by these variables and
the propionitrile is selectively hydrogenated to n-propylamine. This
result is clearly different from that reported for the liquid-phase
hydrogenation of nitriles using copper based catalysts and agree
with the results previously reported for the gas-phase hydrogena-
tion of propionitrile on LnCu2 (Ln = La, Ce, Pr, Nd) [34].
We have found that the hydrogenation of propionitrile over the
copper-lanthanide oxides is a pseudo-first-order with respect to
hydrogen and a pseudo-zero-order with respect to propionitrile,
in agreement with the literature [34,55]. The independence of the
reaction rate from nitrile concentration is in line with previous
hydrogenation reactions, for which zero order with respect to nitrile
has been reported [25,55]. This fact suggests that the active sites
of the copper-lanthanide oxide catalysts were fully saturated with
nitrile, which could contribute to the primary amine selectivity.
Previous studies have also demonstrated that high hydrogen pres-
sures increases the reaction rates and can lead to a decrease in the
condensation reactions and therefore to a decrease in the formation
of secondary and tertiary amines [28].
0
371
2363
1438
5081
1734
6487
0
4
Cu·Pr2O3
Pr2O3
3328
2183
9470
2270
9397
5939
322
0
587
654
1403
536
2771
0
4
Cu·La2O3
La2O3
4
Cu·Nd2O3
Nd2O3
CuO
ThO2
MoO3
0
0
322
0
0
a
◦
SSS, Strong strength sites (600–1000 C); MSS, medium strength sites
◦ ◦
400–600 C); WSS, weak strength sites (<400 C).
(
never detected. This study confirms that the main property of
the copper-lanthanide oxides surface is their basicity, except for
2
Cu·CeO . 4Cu·Ln O (Ln = La, Pr, Nd) exhibited a high selectivity to
2
2
3
acetone (>96%), whereas 2Cu·CeO exhibited a significant selectiv-
2
ity to propene (25%).
The acidity of 2Cu·CeO is significantly higher than that of
2
4
Cu·Pr O and the acidity of 4Cu·La O and 4Cu·Nd O is lower
2
3
2
3
2
3
Nevertheless, in this work the selectivity in the gas-phase
hydrogenation of propionitrile over the copper-lanthanide oxide
than that of 4Cu·Pr O , which is reflected in the reaction rate of
2
3
the gas-phase hydrogenation of propionitrile (Fig. 3). We have also
studied the catalytic decomposition of 2-propanol over pure lan-
thanide oxides and the values obtained for the dehydrogenation
catalysts was not controlled by PH . In contrast, the increase in
2
PNitrile decreased the selectivity to n-propylamine; the selectivity
decreases with decreasing H /nitrile molar ratios. Fig. 4 presents a
(
>99% over Ln O ) and dehydration products (100% over CeO ) con-
2
2
3
2
plot of the selectivity vs. H /nitrile molar ratio (for clarity purposes
firms that in the case of the copper–cerium oxide catalyst the acidity
can be ascribed to the support.
2
we have omitted n-propylamine).
The results obtained over 4Cu·Ln O (Ln = La, Pr, Nd) show that
Therefore, the importance of the catalyst basicity as a precondi-
tion for a good selectivity seems to be a fact, whereas in the case of
the cerium compound the presence of an important acidic function
on the support and the existence of a synergism effect between cop-
per active phase and the cerium oxide support could play a major
role in the catalysts activity and selectivity (see below).
2
3
the decrease in the H /nitrile ratio increases the dipropylamine
2
selectivity. n-Propylamine is highly selectively (>90%) formed for
a hydrogen partial pressure (H /nitrile ratio ≥50), which corrob-
2
orate the hydrogen partial pressure effects. The decrease in the
H /nitrile ratio is always accompanied by the increase in the
2
Fig. 4. Effect of the H2/nitrile molar ratio in the reaction selectivity over the copper-
◦
lanthanide oxide catalysts at 200 C (PNitrile = 0.01–0.16 bar, on a gaseous flow of pure
Fig. 3. Effect of the catalysts acidity in the propionitrile reaction rate to n-
propylamine.
hydrogen at atmospheric pressure). Legend: DPA, dipropylamine (bold symbols); SI,
secondary imine (CH3CH2N CHCH3) (open symbols).

J.B. Branco et al. / Journal of Molecular Catalysis A: Chemical 307 (2009) 37–42
41
Table 3
Activity and selectivity of the impregnation supported copper catalysts for the gas-phase hydrogenation of propionitrile at 250 C.
◦
Catalysts
10% Cu/La2O3
10% Cu/SiO2
10% Cu/CeO2
roa, 10 (mol m
−8
−2 −1
s )
955 (41337)
105 (4547)
69
281 (44262)
14 (2213)
75
94 (402)
4 (294)
94
a
−8
−2 −1
rss , 10 (mol m s )
b
−1
)
Ea (kJ mol
Selectivity (%)
n-Propylamine
Dipropylamine
75
25
35
65
85
15
a
−1 −1
s ).
Reaction rate at start of run (ro) and at steady state (rss); between parentheses the values in (mol gCu
Determined between 150 and 250 C.
b
◦
dipropylamine selectivity and a minor production of the secondary
imine (CH CH N CHCH , <5%). Earlier studies have confirmed that
improved. To understand these results we must take in account: (i)
the nitrile hydrogenation mechanism [6,30], (ii) the nature of the
catalysts surface (active sites) and (iii) the bond strength between
reagent, intermediates and products and the active metal sites.
Copper-lanthanide oxides are better described as copper cata-
lysts supported on rare earth metal oxides that have a very strong
basic character [36]. Consequently, one possible reason for the
selective formation of the primary amine lies on the lack of acid
sites, which are known to catalyze condensation reactions that
could lead to secondary or tertiary amines [21,25]. A weaker interac-
tion between n-propylamine and the catalysts surface also explain
the higher selectivity to the primary amine [28]. Verhaak et al. have
demonstrated that basic supports favor the formation of primary
amines [30].
3
2
3
low space velocities and high H /acetonitrile ratios increase the
2
acetonitrile conversion and limit the yield of higher amines [21,28].
Over 2Cu·CeO the behavior is different and the decrease in the
2
primary amine selectivity was not followed by an increase in the
secondary amine selectivity but by a significant increase in the sec-
ondary imine selectivity. To the best of our knowledge, it is the first
time that a stable formation of an intermediate such as a secondary
imine is reported is the gas phase. On the other hand, the increase
in the hydrogen pressure is reported as a way to eliminate catalysts
deactivation [22], which was not the case on our materials.
3.3. Catalysts preparation method
The nature of the metal is probably the parameter that has
the major influence on the activity and selectivity of the cata-
lysts. Nevertheless, the support can affect the catalysts behavior
and the precise nature of the elementary processes at the surface
among adsorbed species is still a point of controversy in the liter-
ature and most of the intermediates of the nitrile hydrogenation
mechanism have not been confirmed with the exception of the
Schiff bases (secondary imines, enamines) [6,56] that were only
reported on supported metal catalysts where secondary reactions
to higher amines can continue on the support acid sites by conden-
sation between amine and protonated imine to yield a protonated
aminal. The secondary imine formed after NH3 release is in turn
hydrogenated on the metal sites to secondary amine (bifunctional
mechanism) [57–59]. However, this mechanistic model seems to
fail when applied to some catalytic systems where all the reaction
steps leading to higher amines seemed to take place only on the
metal sites [20,56].
Recently, Sachtler and co-workers [22,56,60,61] have studied the
absorption and hydrogenation of nitriles on supported transition
metals and proposed a new model for the nitrile hydrogenation
mechanism. Their key finding was that higher amines are formed
in the chemisorbed layer on the metal by a nucleophilic attack from
the carbon of a solution phase formed by the nitrile and an adsorbed
intermediate (M = NR, where M is the metal surface) and that the
hydrogen is donated by the nitrile molecule and not from adsorbed
hydrogen.
The performance of metallic catalysts depends on the prepara-
tion method and, as a result, on the metal distribution and degree
of reduction. In this work, the results of copper catalysts (10 wt.%)
loaded on La O and SiO2 were significantly different from those
2
3
of the copper-lanthanide oxides, namely a lower selectivity to n-
propylamine (Table 3).
The effects of other reaction conditions such as temperature,
gas hourly space velocity (GHSV) and the H /nitrile molar ratio on
2
the reaction rate were considerably different from those obtained
on the copper-lanthanide oxides. For example, at a low partial pres-
sure of nitrile (PNitrile) = 0.01 bar in a gaseous flow of pure hydrogen,
the decrease in the H /nitrile molar ratio decreases the rate of
2
formation of the primary amine that was selectively formed over
4
Cu·Ln O and 2Cu·CeO2 (Fig. 5).
2
3
Therefore, the products were substantially influenced by the
reaction conditions and by the catalysts preparation method and
in the case of the copper-lanthanide oxides the selectivity is clearly
In this work, both the copper and support sites are basic (except
for the cerium compound) and the hydrogen partial pressure study
shows that the availability of hydrogen (at a low nitrile partial pres-
sure) does not play a significant role in catalysts selectivity, whereas
the data from the nitrile partial pressure study confirm the forma-
tion of the secondary amine or the secondary imine intermediate
at high nitrile partial pressure (low H /nitrile ratio).
2
Therefore, the model that seems more consistent to explain
the catalytic behavior of the copper-lanthanide oxides is that of
Sachtler where high nitrile concentration could increase the forma-
tion of higher amines by the reaction on the overlayer involving the
amine meanwhile formed and an unsaturated surface intermediate.
Moreover, the stability of the partially hydrogenated reaction inter-
mediate (CH3CH2N CHCH3) over 2Cu·CeO2 can be also explained
Fig. 5. Effect of the H2/nitrile molar ratio and WHSV in the reaction selectivity to
n-propylamine (PA) over the impregnation supported copper catalysts at 200 C
◦
(PNitrile = 0.01–0.16 bar, on a gaseous flow of pure hydrogen at atmospheric pressure).
Legend: H2/nitrile (bold symbols); WHSV (open symbols).

4
2
J.B. Branco et al. / Journal of Molecular Catalysis A: Chemical 307 (2009) 37–42
by a higher adsorption strength of the C N bond on the catalyst
surface due to the acidity of the support that blocks the active sites
and slow down the hydrogenation to the secondary amine [62].
Nevertheless, further work is needed to elucidate the mechanism
operating in the gas-phase hydrogenation of propionitrile over this
type of catalysts, including the precise nature of the active copper
sites.
[17] P.N. Rylander, in: Catalytic Hydrogenation over Platinum Metals, Academic
Press, New York, 1967, p. 203.
[
[
18] P.N. Rylander, L. Hasbrouck, Engellard Ind. Tech. Bull. 11 (1970) 19.
19] P.N. Rylander, D.R. Steele, Engellard Ind. Tech. Bull. 5 (1965) 113.
[20] M. Arai, Y. Takada, Y. Nishiyama, J. Phys. Chem. B 102 (1998) 1968–1973.
[
21] P. Braos-Garcia, P. Maireles-Torres, E. Rodriguez-Castellon, A. Jimenez-Lopez, J.
Mol. Catal. A 193 (2003) 185–196.
[22] Y.Y. Huang, V. Adeeva, W.M.H. Sachtler, Appl. Catal. A 196 (2000) 73–85.
[23] I. Ortiz-Hernandez, C.T. Williams, Langmuir 23 (2007) 3172–3178.
[
24] M.C. Carrion, B.R. Manzano, F.A. Jalon, I. Fuentes-Perujo, P. Maireles-Torres, E.
Rodriguez-Castellon, A. Jimenez-Lopez, Appl. Catal. A 288 (2005) 34–42.
25] A. Chojecki, M. Veprek-Heijman, T.E. Muller, P. Scharringer, S. Veprek, J.A.
Lercher, J. Catal. 245 (2007) 237–248.
4
. Conclusions
[
The work covering copper-based catalysts shows that the forma-
[26] B. Coq, D. Tichit, S. Ribet, J. Catal. 189 (2000) 117–128.
27] A.C. Gluhoi, P. Marginean, U. Stanescu, Appl. Catal. A 294 (2005) 208–214.
[
tion of primary amines is normally accompanied by a significant
formation of secondary amines. This study shows that the sup-
ported copper-lanthanide oxides are active and very selective for
the gas phase propionitrile hydrogenation to n-propylamine. The
results obtained are consistent with the copper-lanthanide oxides
basicity that seems to play a key role in the behavior of this new type
of supported copper catalysts. One possible reason for the selective
formation of the primary amine lies on the lack of acid sites, which
are known to catalyze condensation reactions that could lead to sec-
ondary or tertiary amines. The model that seems more consistent
to explain the catalytic behavior of the copper-lanthanide oxides is
that of Sachtler.
[28] F. Medina, R. Dutartre, D. Tichit, B. Coq, N.T. Dung, P. Salagre, J.E. Sueiras, J. Mol.
Catal. A 119 (1997) 201–212.
[29] F.M. Cabello, D. Tichit, B. Coq, A. Vaccari, N.T. Dung, J. Catal. 167 (1997) 142–152.
30] M.J.F.M. Verhaak, A.J. Vandillen, J.W. Geus, Catal. Lett. 26 (1994) 37–53.
31] C.V. Rode, M. Arai, M. Shirai, Y. Nishiyama, Appl. Catal. A 148 (1997) 405–413.
32] N. Iwasa, M. Yoshikawa, M. Arai, Phys. Chem. Chem. Phys. 4 (2002) 5414–5420.
[33] M. Arai, Y. Takada, T. Ebina, M. Shirai, Appl. Catal. A 183 (1999) 365–376.
[
[
[
[
[
[
[
34] J.B. Branco, D. Ballivet-Tkatchenko, A.P. de Matos, J. Phys. Chem. C 111 (2007)
15084–15088.
35] J.S. Abell, in: K.A.G. LeRoy E. Jr. (Ed.), Handbook on the Physics and Chemistry
of Rare Earth, North-Holland Publishing Company, Amsterdam, 1989, p. 1.
36] D. Ballivet-Tkatchenko, J. Branco, A.P. Dematos, J. Phys. Chem. 99 (1995)
5
481–5484.
37] J. Branco, C. de Jesus Dias, A.P. Goncalves, T.A. Gasche, A.P. de Matos, Ther-
mochim. Acta 420 (2004) 169–173.
[
[
[
38] J.B. Branco, T.A. Gasche, A.P. Goncalves, A.P. de Matos, J. Alloys Compd. 323
(
2001) 610–613.
Acknowledgements
39] K.H.J. Buschow, in: K.A.G. LeRoy E Jr. (Ed.), Handbook on Physics and Chemistry
of Rare Earth, North-Holland Publishing Company, Amsterdam, 1984, p. 1.
40] A. Iandelli, A. Palenzona, in: K.A.G. LeRoy E. Jr. (Ed.), Handbook on the Physics
and Chemistry of Rare Earth, North-Holland Publishing Company, Amsterdam,
New York, Oxford, 1979, pp. 1–54.
We acknowledge financial support from the Centre National
de la Recherche Scientifique. One of us (J.B. Branco) thanks
BRITE/EURAM ECC programme for a research studentship (S/BREU-
00287). This work was also supported by the the Portuguese
Funda c¸ ão para a Ciência e a Tecnologia” under the contract
[41] G. Sandrock, S. Suda, L. Sclapbach, in: L. Sclapbach (Ed.), Topics in Applied
Physics, Springer–Verlag, Berlin, Heidelberg, New York, 1992, pp. 197–258.
9
“
[
42] W.E. Wallace, J. France, A. Shamsi, in: G.J. MaCarthy, J.J. Rhyne, H.S. Silber (Eds.),
Catalysis using Rare Earth and Actinide Intermetallics, Plenum Press, New York,
London, 1982, p. 561.
POCTI/QUI/35394/2000.
[
43] G.B. Atkinson, E.G. Baglin, L.J. Nicks, D.J. Bauer, in: R.G. Herman (Ed.), Catalytic
Conversions of Synthesis Gas and Alcohols to Chemicals, Plenum Press, New
York, London, 1984, p. 65.
References
[
44] E.G. Baglin, G.B. Atkinson, L.J. Nicks, Ind. Eng. Chem., Prod. Res. Dev. 20 (1981)
87–90.
[
[
[
1] A.M. Joshi, K.S. MacFarlane, B.R. James, P. Frediani, in: J.R. Kosak, T.A. Johnson
Eds.), Catalysis of Organic Reactions, vol. 53, M. Dekker (Chem. Ind.), New York,
994, p. 497.
2] P.N. Rylander, in: B. Elvers, S. Hawkins, M. Ravenscroft, G. Shultz (Eds.), Ull-
mann’s Encyclopedia of Industrial Chemistry, vol. A13, fifth ed., VCH, Weinheim,
(
1
[45] J.B. Branco, C.J. Dias, A.P. Gon c¸ alves, J. Alloys Compd. (2008),
doi:10.1016/j.jallcom.2008.11.113.
[46] V.T. Coon, T. Takeshita, W.E. Wallace, R.S. Craig, J. Phys. Chem. 80 (1976)
1878–1879.
[47] J.B. Branco, D. Ballivet-Tkatchenko, A.P.d. Matos, J. Alloys Compd. 464 (2008)
399–406.
[48] K.G. Allum, R.D. Hancock, I.V. Howell, T.E. Lester, S. Mckenzie, R.C. Pitkethly, P.J.
Robinson, J. Catal. 43 (1976) 331–338.
1
992, p. 493.
3] P. Scharringer, T.E. Muller, W. Kaltner, J.A. Lercher, Ind. Eng. Chem. Res. 44 (2005)
770–9775.
9
[4] M. Besson, J.M. Bonnier, D. Djaouadi, M. Joucla, Bull. Soc. Chim. Fr. (1990) 13–19.
[5] M. Besson, J.M. Bonnier, M. Joucla, Bull. Soc. Chim. Fr. (1990) 5–12.
[6] C. Debellefon, P. Fouilloux, Cat. Rev. Sci. Eng. 36 (1994) 459–506.
[7] H. Greenfield, Ind. Eng. Chem., Prod. Res. Dev. 6 (1967) 142–144.
[8] T.W. Russell, R.C. Hoy, J.E. Cornelius, J. Org. Chem. 37 (1972) 3552–3553.
[9] C. Barnett, Ind. Eng. Chem., Prod. Res. Dev. 8 (1969) 45.
[49] M.A. Aramendia, V. Borau, C. Jimenez, A. Marinas, J.M. Marinas, J.R. Ruiz, F.J.
Urbano, J. Mol. Catal. A 218 (2004) 81–90.
[50] J.A. Wang, X. Bokhimi, O. Novaro, T. López, F. Tzompantzi, R. Gómez, J. Navarrete,
M.E. Llanos, E. López-Salinas, J. Mol. Catal. A: Chem. 137 (1999) 239–252.
[51] A. Gervasini, A. Auroux, J. Catal. 131 (1991) 190–198.
[52] A. Gervasini, J. Fenyvesi, A. Auroux, Catal. Lett. 43 (1997) 219–228.
[53] P.F. Rossi, G. Busca, V. Lorenzelli, M. Lion, J.C. Lavalley, J. Catal. 109 (1988)
378–386.
[
10] M. Freifelder, in: Catalytic Hydrogenation in Organic synthesis-Procedures and
Commentary, Wiley-Interscience, New York, 1978, Chapter 6.
[
11] P.N. Rylander, in: Catalytic Hydrogenation in Organic Synthesis, Academic, New
York, 1979, Chapter 8.
12] J. Volf, J. Pasek, Stud. Surf. Sci. Catal. 27 (1986) 105.
13] W. Kiel, in: B. Elvers, S. Hawkins, M. Ravenscroft, G. Shultz (Eds.), Ullmann’s
Encyclopedia of Industrial Chemistry, fifth ed., VCH, Weinheim, 1985, p. 9.
[54] V. Solinas, I. Ferino, Catal. Today 41 (1998) 179–189.
[55] H. Li, Y. Wu, J. Zhang, W. Dai, M. Qiao, Appl. Catal. A 275 (2004) 199–206.
[56] Y.Y. Huang, W.M.H. Sachtler, Appl. Catal. A 182 (1999) 365–378.
[57] L.K. Freidlin, T.A. Sladkova, F.F. Englina, Kinet. Katal. 3 (1962) 417.
[58] C.F. Medina, D. Tichit, B. Coq, A. Vaccari, N.T. Dung, J. Catal. 167 (1997) 142.
[59] M.J.F.M. Verhaak, A.J. vanDillen, J.W. Geus, Catal. Lett. 26 (1994) 37.
[60] Y.Y. Huang, W.M.H. Sachtler, J. Catal. 188 (1999) 215–225.
[61] Y.Y. Huang, W.M.H. Sachtler, J. Catal. 190 (2000) 69–74.
[62] P. Kukula, K. Koprivova, J. Catal. 234 (2005) 161–171.
[
[
[
[
[
14] M. Freifelder, in: Pratical Catalytic Hydrogenation, Willey, New York, 1971, p.
46.
15] A. Galan, J. Demendoza, P. Prados, J. Rojo, A.M. Echavarren, J. Org. Chem. 56
1991) 452–454.
16] R.J. Peterson, Chem. Tech. Rev. (1977) 229.
3
(