Selectivity patterns in heterogeneously catalyzed hydrogenation of conjugated ene-yne and diene compounds

Article abstract of DOI:10.1016/j.jcat.2011.10.003

Selectivity control in heterogeneously catalyzed hydrogenation of conjugated hydrocarbons (ene-yne and diene compounds) is a challenging task. Available studies on the topic mainly encircle 1,3-butadiene as the substrate and palladium as the catalyst, while more elaborated playground molecules and other metals remain largely unexplored. This study investigates the gas-phase hydrogenation of valylene (2-methyl-1-butene-3-yne) and isoprene (2-methyl-1,3-butadiene) over Pd, Pb-poisoned Pd, CO-modified Pd, Cu, Ni, and bimetallic CuNi catalysts. Chemoselectivity, regioselectivity, full hydrogenation, and CC bond formation/scission footprints of the catalytic systems at different inlet hydrogen-to-hydrocarbon ratios and conversion degrees have been rationalized. Complementary studies of 3-methylbutyne and 1-penten-4-yne hydrogenation were carried out in order to analyze (i) the impact of isomerization on the observed mono-olefin distribution in valylene/isoprene hydrogenation and (ii) the conjugation issue in partial ene-yne hydrogenation. Our results lead to an improved understanding of hydrogenation of polyunsaturated hydrocarbons and open doors to design more selective heterogeneous catalysts and related processes for this practically important class of reactions.

Full text of DOI:10.1016/j.jcat.2011.10.003

Journal of Catalysis 284 (2011) 165–175
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Selectivity patterns in heterogeneously catalyzed hydrogenation of conjugated
ene-yne and diene compounds
⇑
Blaise Bridier, Javier Pérez-Ramírez
Institute for Chemical and Bioengineering, Department of Chemistry and Applied Biosciences, ETH Zurich, Wolfgang-Pauli-Strasse 10, CH-8093 Zurich, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 8 November 2011
Selectivity control in heterogeneously catalyzed hydrogenation of conjugated hydrocarbons (ene-yne and
diene compounds) is a challenging task. Available studies on the topic mainly encircle 1,3-butadiene as
the substrate and palladium as the catalyst, while more elaborated playground molecules and other
metals remain largely unexplored. This study investigates the gas-phase hydrogenation of valylene
Keywords:
Heterogeneous catalysis
Selective hydrogenation
Ene-yne compounds
Valylene
Isoprene
Isomerization
(
2-methyl-1-butene-3-yne) and isoprene (2-methyl-1,3-butadiene) over Pd, Pb-poisoned Pd, CO-modi-
fied Pd, Cu, Ni, and bimetallic CuANi catalysts. Chemoselectivity, regioselectivity, full hydrogenation,
and CAC bond formation/scission footprints of the catalytic systems at different inlet hydrogen-to-hydro-
carbon ratios and conversion degrees have been rationalized. Complementary studies of 3-methylbutyne
and 1-penten-4-yne hydrogenation were carried out in order to analyze (i) the impact of isomerization on
the observed mono-olefin distribution in valylene/isoprene hydrogenation and (ii) the conjugation issue
in partial ene-yne hydrogenation. Our results lead to an improved understanding of hydrogenation of
polyunsaturated hydrocarbons and open doors to design more selective heterogeneous catalysts and
related processes for this practically important class of reactions.
Ó 2011 Elsevier Inc. All rights reserved.
1
. Introduction
The catalytic hydrogenation of targeted functional groups in
[18–23]. Nevertheless, available works in the literature have
mostly tackled mono-alkynes (e.g., ethyne, propyne, pentyne, and
hexyne). The hydrogenation of polyunsaturated hydrocarbons,
such as ene-yne and diene compounds, over a variety of heteroge-
neous catalysts is comparatively much less studied. An exception is
hydrocarbons is frequently used in multiple industrial sectors,
yielding a wide variety of intermediate and end-user products
[
1,2]. Therefore, chemo-, regio-, and/or stereoselectivity control
4
1,3-butadiene due to its practical relevance in C hydrorefining
upon hydrogenation of multi-functionalized substrates attracts
much interest for fundamental and applied research in catalysis
[5,6,24–29]. Hydrogenation of this diene experiences regioselectiv-
ity (the first hydrogenation can follow 1,2- or 1,4-addition mecha-
nisms to produce 1-butene or 2-butenes, respectively) and
stereoselectivity (cis- versus trans-2-butene) issues. Another
appealing substrate for hydrogenation studies is isoprene
(2-methyl-1,3-butadiene). This compound is a common structure
motif to a variety of other natural compounds, collectively named
isoprenoids. The presence of the methyl group in isoprene sup-
presses the stereoselectivity issue intrinsic to 1,3-butadiene hydro-
genation and produces an additional mono-olefin from the
1,2-addition. As it occurs with 1,3-butadiene and mono-alkynes,
isoprene hydrogenation has been mainly studied on palladium-
based catalysts [6,30–32].
Valylene (2-methyl-1-butene-3-yne) is the smallest branched
hydrocarbon possessing conjugated double and triple bonds and
could be a very interesting playground molecule to improve
knowledge on catalyst design for the hydrogenation of conjugated
systems (Fig. 1). In particular, the conversion of valylene to
isoprene is of practical interest [33,34], since isoprene is used to
produce cis-1,4-polyisoprene, a synthetic version of natural rubber.
[
3,4]. Specifically, the partial hydrogenation of highly unsaturated
hydrocarbons, that is, mono-alkynes and dienes, is a crucial cata-
lytic step for the purification of alkene streams and for fine chem-
ical production [5,6]. Although this class of reactions runs
industrially on Pd-based catalysts for several decades, the topic
has recently gained substantial interest due to the wide margin
to improve commercial systems. For instance, many modified Pd/
Al O catalysts used for alkyne hydrogenation in C AC cuts of
2 3 2 4
steam crackers display moderate alkene selectivities and insuffi-
cient lifetime due to fouling.
The application of modern experimental methods, controlled
routes for catalyst preparation, and computer simulations led to
an improved understanding of the factors governing selectivity in
alkyne hydrogenation on bare and modified Pd catalysts [7–17],
and also enabled the identification of alternative non-Pd catalysts
⇑
Corresponding author. Fax: +41 44 633 1405.
E-mail address: jpr@chem.ethz.ch (J. Pérez-Ramírez).
0
021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.10.003

1
66
B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
Fig. 1. Reaction scheme of valylene hydrogenation. Compound codes are VL = valylene, IP = isoprene, 3MBy = 3-methylbutyne, 2MB = 2-methylbutene, 2M2B = 2-methyl-2-
butene, 3MB = 3-methylbutene, and iC = isopentane. The sum of methylbutenes is denoted as MBs. Products coming from CAC bond formation (oligomers, OL) and scission
cracked products, CK) are not displayed for the sake of clarity.
5
(
Early work by Freidlin et al. [35,36] reported valylene hydrogena-
tion over Zn, Pd black, and Raney-type cobalt catalysts. Aduriz
et al. [32] studied the properties of supported palladium in the
gas-phase valylene hydrogenation on alloying with a second metal.
Isoprene was obtained with a maximal selectivity of 78% at 80%
fully-automated set-up equipped with mass-flow controllers
(Bronkhorst, EL-FLOW) to feed gases (H , CO, and He), a syringe
pump (Chemyx, Fusion 100) to feed the liquid hydrocarbons, an
electrically-heated oven with a 12 mm i.d. quartz micro-reactor,
and a gas chromatograph. The catalysts (sieve fraction 200–
2
valylene conversion on PdAPb/
a-Al
2
O
3
.
400 lm) were loaded in the tubular reactor and heated in He at
Herein, we have systematically investigated the gas-phase
hydrogenation of valylene, isoprene, 3-methylbutyne, and 1-pen-
ten-4-yne over various catalytic systems, with the aim of rational-
izing selectivity patterns associated with the type of metal and the
presence of modifiers. Bare Pd, CO-modified Pd, Pb-poisoned Pd,
Cu, Ni, and CuANi catalysts have been studied in a continuous-flow
reactor at ambient pressure covering a broad range of feed hydro-
gen-to-hydrocarbon ratios and conversion levels.
573 K for 30 min. Pd and PdAPb were evaluated as such, while
the other catalysts were subjected to an additional pre-treatment:
PdACO (0.1 vol.% CO in He, 348 K, 10 min), Cu, Ni, and CuANi
2
(5 vol.% H in He, 773 K, 30 min). The liquid hydrocarbon was evap-
orated at 353 K and mixed with the gas mixture prior to the reactor
inlet. The hydrogenation reactions were carried out at 348 K (Pd,
PdACO, PdAPb), and 523 K (Cu, Ni, CuANi). Reaction temperatures
and pre-treatment conditions were selected on the basis of previ-
ous tests over these specific catalysts in propyne hydrogenation
[
37]. The hydrogen-to-hydrocarbon ratio dependence and 3-meth-
ylbutyne and 1-penten-4-yne hydrogenation were carried out over
.3 g of catalyst. The inlet hydrocarbon concentration was
2.5 vol.%, the inlet H concentration varied in the range of 2.5–
2
. Experimental
0
Hydrogenation studies were conducted over six catalysts:
2
3
À1
commercial 1 wt.% Pd/
c
-Al
2
O
3
(Aldrich, ref: 20,570-2,
30 vol.%, and the total flow was kept at 84 cm STP min by bal-
ancing the feed mixtures with He. The influence of the degree of
hydrocarbon conversion on the product distribution was studied
using catalyst amounts in the range of 0.01–0.2 g. In the experi-
ments with a weight of catalyst 60.1 g, the bed was diluted with
SiC (sieve fraction 200–400
hydrocarbon concentration was 2.5 vol.%, the inlet H
tion was 2.5 vol.% (Pd, PdAPb) and 5 vol.% (PdACO, Cu, Ni, CuANi).
2
À1
SBET = 210 m g ), commercial Lindlar-type Pb-poisoned 5 wt.%
2
À1
3
Pd/CaCO (Alfa Aesar, ref: 043172, SBET = 13 m g ), and self-made
2
À1
3 3
Cu Al (65 wt.% Cu and 8 wt.% Al, SBET = 71 m g ), Ni Al (58 wt.%
2
À1
Ni and 8 wt.% Al, SBET = 206 m g ), and Cu2.75Ni0.25Fe (50 wt.%
2
À1
Cu, 5 wt.% Ni, and 18 wt.% Fe, SBET = 35 m g ). The catalysts are
denoted along the manuscript by their respective active metal
and modifiers, that is, Pd, PdACO (carbon monoxide was continu-
ously added to the feed mixture), PdAPb, Cu, Ni, and CuANi. The
alumina-supported Pd catalyst was characterized and evaluated
in the gas-phase hydrogenation of ethyne and propyne [13]. De-
tails on preparation and characterization of the Cu, Ni, and CuANi
samples, as well as on their catalytic response in the gas-phase
hydrogenation of propyne, can be found in recent publications
lm) to minimize by-passing. The inlet
2
concentra-
3
À1
The total flow was varied in the range of 20–270 cm STP min
yielding space velocities (SV) of 6000–1620,000 cm g
,
3
À1 À1.
h High
space velocities were necessary in order to attain a low degree of
conversion on the palladium-based catalysts. Hydrogenation tests
with Pd/c-Al O were also carried out in the presence of 0.1 vol.%
2 3
CO. Fresh catalyst was loaded for each hydrocarbon substrate,
while the hydrogen-to-hydrocarbon and conversion dependence
were studied without changing the catalytic bed. The catalytic per-
formance was measured in steady state, after equilibration for at
least 1 h at each experimental condition. The catalysts were stable
during time-on-stream which went up to 10 h and this was corrob-
orated by running the same experimental condition in different
stages. Additionally, tests with different particle sizes and different
[
23,37–39]. The reader is addressed to these references for details
on synthesis protocols and physico-chemical properties of the
samples.
The gas-phase hydrogenation of valylene (VL) (ABCR-Chemicals,
purity 97%), isoprene (IP) (ABCR-Chemicals, purity 99%), 3-meth-
ylbutyne (3MBy) (Acros Organics, purity 96%), and 1-penten-
4
-yne (ABCR-Chemicals, purity 95%) was studied at 1 bar in a

B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
167
total volumetric flows (at constant space velocity) were conducted
to discard internal and external mass-transport constraints. It was
further emphasized by the fulfillment of the Weisz–Prater crite-
rion. Products in the outlet stream were analyzed using an on-line
gas chromatograph (Agilent GC7890A) equipped with a GS-GasPro
column (and an additional HP-Plot Q for the separation of the
2
H :VL = 1 (stoichiometric ratio for partial hydrogenation to iso-
prene) and 348 K, valylene conversion over Pd was 80% but
increased to 100% at higher ratios (Fig. 2a). The high activity of pal-
ladium is expected attending to its unique ability to dissociate H ,
2
providing atomic hydrogen species that are active for hydrogena-
tion. In terms of product distribution, the isoprene selectivity of
1
-penten-4-yne hydrogenation products) and a flame ionization
2
Pd decreased from 53% to 4% upon increasing the H :VL ratio from
detector. The tubes from the syringe connection to the gas chro-
matograph were heated at 393 K to prevent reactants and products
condensation. The conversion, X, was determined as the amount of
converted hydrocarbon divided by the amount of hydrocarbon at
the reactor inlet. The selectivity, S, was calculated as the amount
of product formed divided by the amount of converted hydrocar-
bon. The selectivity to oligomers was obtained using the carbon
1 to 2. The sharp drop in the selectivity to the partially hydroge-
nated product occurred at the expense of an increased selectivity
to mono-olefins, from 23% to a maximum of 68% at H
Methylbutenes (MBs = 2MB + 2M2B + 3MB) result from the second
hydrogenation of valylene (Fig. 1). Further increase in the H :VL ra-
2
:VL = 2.
2
tio led to a progressive decrease in the selectivity to mono-olefins
(down to zero) with the concomitant production of the fully hydro-
genated alkane. Accordingly, the selectivity to isopentane exceeds
balance as S(oligomers) = 1 À
RS(all other products).
9
5% at H
2
:VL P 6. The observed selectivity pattern as a function of
pressure is characteristic of a consecutive
the inlet partial H
2
3
. Results and discussion
hydrogenation scheme. CAC coupling reactions also occurred dur-
ing valylene hydrogenation on palladium. Oligomers were formed
3
3
.1. Valylene and isoprene hydrogenation
with a selectivity of ca. 25% at H
-richer conditions. It can be noticed that the production of meth-
ylbutenes and oligomers ceased at the same H :VL ratio of 6. 3-
2
:VL = 1 and steadily decreased at
H
2
.1.1. Comparison of Pd and CuANi
2
Fig. 2 shows the conversion and product selectivity of Pd and
Methylbutyne (3MBy) was not observed on palladium, suggesting
that the first hydrogenation step is chemoselective, that is, it exclu-
sively forms isoprene by partial hydrogenation of the triple bond in
valylene. A full degree of IP conversion was measured in isoprene
CuANi catalysts in valylene and isoprene hydrogenation at differ-
ent hydrogen-to-hydrocarbon ratios. These two representative
samples were selected in order to describe in more detail the range
of products formed during hydrogenation of the conjugated hydro-
2
hydrogenation at all H :IP ratios studied (Fig. 2b). The highest
2
carbons and their dependence with the inlet H :HC ratio. At
selectivity to the partially hydrogenated methylbutenes (73%)
Fig. 2. Conversion and selectivity to products versus the feed hydrogen-to-hydrocarbon ratio in valylene (left graphs) and isoprene (right graphs) hydrogenation over (a, b) Pd
:HC = 1–12, balance He, SV = 16,800 cm3
g
À1 À1
h , and P = 1 bar.
at 348 K and (c, d) CuANi at 523 K. Other conditions: HC = 2.5 vol.% and H
2

1
68
B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
2
was obtained at H :IP = 1 and drastically decreased to zero at high-
2
was analyzed as a function of the feed H :HC ratio. The lead-
er ratios at the expense of isopentane production. As expected, the
selectivity to oligomers gradually decreased too.
The CuANi catalyst exhibited remarkable differences compared
with Pd (Fig. 2c and d). The hydrogenation activity of CuANi was
poisoned Pd catalyst displayed full hydrocarbon conversion under
all conditions studied. The copper-containing catalysts showed a
drastic increase in valylene conversion from 30% (Cu) and 49%
(CuANi) to ꢀ100% by increasing the H
presence of nickel promotes a higher conversion at a lower partial
pressure, as a consequence of the ability of Ni to dissociate H in
comparison with Cu [37,39]. PdACO and Ni are much less active
catalysts, showing 100% VL conversion at inlet H :VL ratios of 8
2
:VL ratio from 1 to 3. The
2
much lower; full conversion was reached at H :HC > 4 and 523 K.
CuANi displayed a higher selectivity to partially hydrogenated
products in both valylene and isoprene hydrogenation. The iso-
prene selectivity in valylene hydrogenation reached a maximum
H
2
2
2
of 53% at H
expense of mono-olefins, whose selectivity increased from 13% to
7% (Fig. 2c). CuANi possesses a broader H :HC window for selec-
tive operation than Pd, that is, the drop in selectivity to partially
hydrogenated products was much less pronounced on the former
catalyst. Isopentane production was negligible at any ratio.
Another distinctive feature of CuANi compared to Pd in valylene
hydrogenation was the detection of 3-methylbutyne (S(3MBy) =
2
:VL = 2 and decreased to 25% at H
2
:VL = 12 at the
and 10, respectively. The low activity of the CO-modified Pd cata-
lyst is assigned to the ability of CO to form very dense layers even
at a low partial pressure [13], which implies a reduction in the
number of adsorption sites for hydrogen and valylene. The rela-
tively low activity of Ni can be attributed to fouling (formation of
carbonaceous deposits), attending to the high selectivity to oligo-
mers (vide infra).
In previous studies of gas-phase hydrogenation of propyne over
the same catalysts [13,23,38,39], the alkyne conversion was com-
plete at lower hydrogen-to-hydrocarbon ratio compared to valyl-
ene hydrogenation. The higher hydrogen-to-hydrocarbon ratio
required to achieve full conversion of the polyunsaturated com-
pound agrees with Molnár et al. [3]. These authors stipulated that
conjugated hydrocarbons are strongly adsorbed on the surface due
6
2
3
% at H
higher than on Pd in the whole range of inlet partial H
During isoprene hydrogenation (Fig. 2d), the product distribution
over CuANi was hardly affected by the H :IP ratio. The selectivity
2
:VL = 1). The selectivity to oligomers on CuANi was slightly
2
pressures.
2
to MBs was practically constant at ca. 85%. The selectivity to olig-
omers decreased progressively with the hydrogen content in the
feed at the expense of isopentane production. Still, the S(iC
5
) was
to the contribution of the entire p-system and thus the adsorbed
as low as 12% at H :IP = 12 in comparison with the value of 90%
2
valylene strongly competes with hydrogen for the active sites, low-
ering its coverage. The conversion patterns in isoprene hydrogena-
tion (open symbols in Fig. 3, left) are comparable to those in
valylene hydrogenation, although higher activities are generally
observed upon decreasing the degree of substrate unsaturation.
on Pd. A final observation from Fig. 2 is that oligomer production
over the two metals was somewhat higher on hydrogenation of
the more unsaturated ene-yne compound than of the diene.
3
.1.2. Influence of H
2
:HC ratio
2
For example, the H :HC ratio for 100% valylene conversion over
Fig. 3 presents the overall comparison of catalytic systems in
PdACO (8) or Ni (10) is decreased in both cases to 2 for 100% iso-
prene conversion. This may be due to the fact that the adsorption
of isoprene was not so strong (in contrast to that of valylene) to
hinder hydrogen dissociation.
valylene (solid symbols) and isoprene (open symbols) hydrogena-
tion. The influence of Pd modification by continuous CO feeding or
Pb addition as well as the influence of the metal (Pd, Cu, Ni, CuANi)
Fig. 3. Conversion and selectivity to products versus the feed hydrogen-to-hydrocarbon ratio in valylene (solid symbols) and isoprene (open symbols) hydrogenation. The
:HC = 1–12, balance He, SV = 16,800 cm3
g
À1 À1
h , and P = 1 bar.
side panels indicate catalyst and reaction temperature. Other conditions: HC = 2.5 vol.% and H
2

B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
169
The most exciting differences between the catalysts is related to
their diverse selectivity patterns. Lead, a common selective poison
of palladium-based hydrogenation catalysts [3,6,12,24,40,41], is a
major and essential component of the widely applied Lindlar cata-
lyst [42]. Recent Density Functional Theory simulations [15] have
shown that by means of both electronic and geometric modifica-
tions, Pb (i) decreases the hydrogen coverage on Pd, (ii) hinders
the formation of sub-surface hydride and carbide species (that lead
to selective or unselective hydrogenation in mono-alkynes [8,13]),
and (iii) enhances the thermodynamic control of the reaction by
lowering the adsorption energies of surface moieties compared to
pure Pd. Nevertheless, at this particular condition, PdAPb was an
extremely poor catalyst for valylene hydrogenation to isoprene,
with practically zero selectivity. Additionally, the CAC bond forma-
tion channel leading to oligomers is enhanced (S(OL) = 75%) com-
pared to bare Pd that favors the production of isoprene under the
same condition. No methylbutyne was produced either. On the
other hand, the presence of lead in the palladium catalyst eased
(S(OL) ꢀ 75%). At H
2
:VL > 6, the selectivity to iC
pressure with the concomitant decrease in olig-
5
increased with
the inlet partial H
2
omerization. The product distribution of VL hydrogenation is
highly sensitive to the hydrogen coverage. A similar conclusion de-
rived from isoprene hydrogenation. At H
mono-olefins was 76% but dropped to 26% at H
pense of iC . Contrarily to valylene, isoprene hydrogenation on
2
:IP = 1, the selectivity to
2
:IP = 4, at the ex-
5
nickel led to little oligomerization, S(OL) ꢀ 10%.
The ability of nickel to crack hydrocarbons during the reaction
was quantified. Previous experiments and DFT calculations have
shown that step-edges of Ni(111) dissociate CAC bonds leading
to carbon-containing fragments on the surface [43]. Fig. 4 shows
the selectivity to oligomers, methylbutenes, and cracked hydrocar-
bons (C
hydrogenation as a function of the H
low hydrogen content, CAC formation was favored, while CAC scis-
sion was enhanced at high H pressure leaving a very narrow win-
1
4
AC ) analyzed by gas chromatography during valylene
2
:VL ratio. As expected, at
2
dow for selective hydrogenation. This feature largely restricts the
use of Ni in partial hydrogenation of unsaturated hydrocarbons.
As often done for palladium, nickel would require additional mod-
ifiers, e.g., pre-sulfidation, to moderate its ability to hydrogenate
unsaturated substrates in a wider range of operating conditions.
Cu and CuANi (circles and hexagons in Fig. 3, respectively)
showed comparable selectivity patterns in VL hydrogenation. The
promotion by nickel in CuANi originally identified in propyne
hydrogenation [23] also applies to valylene hydrogenation. At
the desorption of mono-olefins (S(MBs) = 78% at H
2
:VL = 10) and
minimized the formation of isopentane (S(iC
5
) < 20%) compared
to bare Pd.
Continuous addition of carbon monoxide is industrially applied
to enhance the selectivity of Pd catalysts in partial hydrogenation
of alkynes in C AC cuts of steam crackers [3,5,6]. Similarly to
2 4
Pb, CO improves the thermodynamic factor, decreases the hydro-
gen coverage, and prevents the formation of sub-surface hydrides
and carbides [13]. Our results indicated that, for the particular cat-
alysts evaluated here, CO has a stronger impact than Pb on decreas-
ing the hydrogen coverage. This is already clear from the much
2
H :VL = 1, oligomers comprised the major product over Cu
(S(OL) = 75%) and the selectivity to isoprene was limited to 10%.
Under the same experimental condition, and due to a higher
hydrogen coverage induced by the presence of small amounts of
nickel [23], isoprene became the main reaction product
(S(IP) = 50%) and the selectivity to oligomers decreased to 35%.
The selectivity to 3-methylbutyne decreased from 10% on Cu to
3% on CuANi. At a higher H :VL ratio, the selectivity to 3MBy over
2
both Cu and CuANi dropped to zero and the selectivity to oligo-
mers decreased to 22% (Cu) and 8% (CuANi). The IP selectivity over
higher H
10) compared with PdAPb (1). Besides, PdACO showed a signifi-
cant selectivity to isoprene (41% at H :VL = 3) in contrast to the
2
:VL ratio required for full valylene conversion on PdACO
(
2
zero IP production over the lead-poisoned Pd sample. Minor
amounts of 3-methylbutene (S(3MBy) ꢀ 1.5%) were observed (not
shown in Fig. 3). The production of IP and 3MBy highlights the im-
pact of CO on the so-called thermodynamic factor. Carbon monox-
ide enhances the differential energy of adsorption between
valylene and the products of the first hydrogenation. Upon increas-
Cu displayed a maximum of 46% at H
tivity steadily increased from 6% to 62% with the H
selectivity patterns of IP and MBs over CuANi and Cu very much
resemble at H :VL P 3.
2
:VL = 3, while the MBs selec-
2
:VL ratio. The
ing the inlet partial H
on PdACO increased up to ca. 65% at H
at higher ratios. Oligomers were produced to a larger extent over
PdACO than over PdAPb in the whole range of H :VL ratios
S(OL) ꢀ 90% at H :VL = 1 and still 22% at H :VL = 12). Oppositely,
the CO-modified Pd catalyst experienced minor over-hydrogena-
tion; there is practically no iC production at H :VL < 10. In iso-
2
pressure, the selectivity to methylbutenes
2
:VL = 6, remaining constant
2
Contrarily to palladium, in which the sub-surface chemistry of
carbide/hydride species plays an important role [8,13], the hydro-
2
(
2
2
5
2
prene hydrogenation, the modification of bare Pd catalyst by Pb
addition or by continuous CO feeding remarkably increased the
production of mono-olefins, being the main product at any H
ratios. PdACO was more selective for partial hydrogenation
S(MBs) = 75% versus 57% on PdAPb at H :IP = 10). Concomitantly,
PdAPb was more prone to over-hydrogenation (S(iC ) = 38% versus
10% on PdACO at H :IP = 10), while PdACO favored oligomeriza-
tion (S(OL) = 22% versus <5% on PdAPb at H :IP = 10). The modifi-
2
:IP
(
2
5
<
2
2
cation of Pd using 0.1 vol.% CO was chosen based on our previous
study in ethyne and propyne hydrogenation [13]. It cannot be ruled
out that an optimized CO addition upon increasing the substrate
complexity leads to a reduced CAC formation at the expense of
partially hydrogenated products. Nevertheless, the effect of CO
concentration in valylene and isoprene hydrogenation is beyond
the scope of this study.
The product distribution of valylene hydrogenation over Ni
markedly differed from the Pd catalysts (Fig. 3, squares). The selec-
tivity to isoprene was very low (ca. 10%) and methylbutyne was
produced with a selectivity not higher than 1%. The selectivity to
MBs presented a volcano shape with a maximum of 49% at
Fig. 4. Selectivity to oligomers, methylbutenes, and cracked products versus the
feed hydrogen-to-valylene ratio over Ni at 523 K. Other conditions: VL = 2.5 vol.%
À1 À1
2
H :VL = 8. Below this ratio, oligomers were the main product
and H
2
:VL = 1–12, balance He, SV = 16,800 cm3
g
h , and P = 1 bar.

1
70
B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
genation over Cu-based catalysts is governed exclusively by sur-
face species [37]. Therefore, the thermodynamic factor largely con-
trols the product distribution. At low inlet partial H pressure, the
2
difference in binding energy between VL and IP (or 3MBy) was suf-
ficient to induce high selectivity to the product of the first hydro-
but oligomers were produced at all conversions and reached a min-
imum of 19% at X(VL) = 90%. The amount of oligomers progres-
sively increased on decreasing the conversion, probably due to
slow deactivation of Pd by fouling. On the other hand, the higher
oligomerization at full degree of conversion is attributed to the
low space velocity used. If full conversion is achieved at a certain
depth within the catalytic bed, the low hydrogen coverage present
in the remaining part enhances the formation of partially hydroge-
nated intermediates inducing CAC bond formation. This phenome-
genation. However, at high
H
2
content, the reaction was
kinetically favored toward MBs. Nickel in the bimetallic CuANi cat-
alyst increased the hydrogen coverage to enhance the IP selectivity
at a lower H :VL ratio, although this did not lead to isopentane pro-
2
duction. The higher 3MBy selectivity in the absence of Ni suggests
the low hydrogen coverage as the key factor to enhance the pro-
duction of 3-methylbutyne on copper. During IP hydrogenation,
2
non was not observed upon varying the H :HC ratio (Fig. 3), since
above the stoichiometric ratio (needed to reach full conversion),
there was enough hydrogen in the feed to favor over-
hydrogenation.
PdAPb showed similar trends with important distinctive fea-
tures that further highlight the effect of lead poisoning. As men-
tioned above, the presence of lead favors the mono-olefin
Cu and CuANi exhibited their intrinsically high and H
dent selectivity to mono-olefins (93% and 82%, respectively) at
:IP P 2. Oligomers comprised the only side reaction over Cu,
while the presence of Ni favored further hydrogenation of MBs to
iC (selectivity of 12% at H :IP = 12). Furthermore, it is important
2
-indepen-
H
2
5
2
desorption (S(MBs) = 75%) and suppresses the formation of iC
5
to note that the small addition of Ni in the Cu-based catalyst was
not sufficient to enhance the cracking ability of the catalyst. This
is in line with our previous study [44], where we reported by
means of surface-sensitive in situ X-ray photoelectron and X-ray
absorption spectroscopies that Ni would most probably not form
large surface ensembles required to split CAC bonds.
compared to bare Pd at full degree of valylene conversion. Never-
theless, the most striking difference is the increase in IP selectivity
that reached ca. 70% at X(VL) = 82% and remained constant upon
increasing the space velocity. This in agreement with Aduriz
et al. [32] that reported an enhanced selectivity of the Pb-poisoned
Pd catalyst at different degrees of conversion. The poor IP selectiv-
ity and the high CAC bond formation compared to Pd observed in
Fig. 3 (H
which enhanced oligomer production. Nevertheless, the high
mono-olefin selectivity at H :VL = 12 emphasized Pb as an efficient
selective poison that enables the production of the MBs under ex-
treme conditions (large catalyst amount and high H :HC ratio).
Cu and CuANi exhibit remarkable differences compared to Pd as
the decrease in conversion at a constant H :VL ratio (Fig. 5) did not
strongly affect the product distribution. Selectivity to products at
low degree of conversion over the Cu-based catalysts were stable
up to X(VL) = 80% with an expected higher selectivity over CuANi,
S(IP) = 28% ant 60% for Cu and CuANi, respectively. Nevertheless,
upon further increase in conversion, the product distribution was
barely changed over Cu, while the IP selectivity decreased from
2
:VL = 1 and 2) is the proof of a sub-optimal space velocity
3
.1.3. Product distribution at variable conversion
The conversion and selectivity dependence on the feed hydro-
2
gen-to-hydrocarbon ratio in VL and IP hydrogenation enables a
clear distinction between the catalytic systems. Nevertheless, the
full degree of conversion at low hydrogen-to-hydrocarbon ratio
over Pd, PdAPb, Cu and CuANi implies that all reactants are con-
sumed at a certain point within the catalytic bed and further reac-
tions (hydrogenation, isomerization, and oligomerization) of the
products might occur on the remaining catalyst. Fig. 5 shows the
selectivity to products as a function of the degree of conversion
2
2
2
during valylene hydrogenation over the catalysts. The feed H :VL
ratio was established based on Fig. 3 and the degree of conversion
was varied by changing the space velocity. As expected
6
0% to 44% over CuANi. 3MBy was only observable over Cu
[
32,40,45,46], an increase in space velocity over all catalysts de-
(S(3MBy) = 8%).
creased valylene conversion and enhanced selectivity toward par-
tially hydrogenated products.
During hydrogenation over Pd, IP selectivity slightly decreased
from 52% to 44% at the expense of MBs from X(VL) = 28–90%,
respectively. A further increase to full conversion led to a drastic
decrease in isoprene selectivity, an increased in mono-olefin
3.2. Hydrogenation of 3-methylbutyne
The hydrogenation of 3-methylbutyne was carried out in order
to analyze the interplay between isomerization and intrinsic regi-
oselectivity in the obtained mono-olefin distribution during valyl-
5
selectivity, and to iC production (14%). 3MBy was not formed,
Fig. 5. Selectivity to products versus the degree of valylene conversion over Pd, PdAPb, Cu, and CuANi. Conditions: VL = 2.5 vol.% and H
2
:VL = 1 (Pd, PdAPb) and 2 (Cu,
CuANi), T = 348 K (Pd, PdAPb) and 523 K (Cu, CuANi). Other conditions are detailed in the Experimental section.

B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
171
Fig. 6. Conversion (striped bar) and selectivity to MBs (solid bar), iC
5
(open bar), and OL (gray bar) in hydrogenation of 3-methylbutyne (left) and isoprene (right) over the
3
À1 À1
catalysts. Reaction temperature and feed H :HC ratio for each catalyst are indicated. Other conditions: SV = 16,800 cm
2
g
h
and P = 1 bar.
ene and isoprene hydrogenation (see Fig. 1). Fig. 6 compares con-
version and selectivity in hydrogenation of 3MBy (left) and IP
(
right) over the six catalytic systems included in this study. These
tests were done at a fixed H :HC ratio, which was selected to max-
imize the production of methylbutenes. Marked differences in
MBy conversion and selectivity to MBs, iC , and OL were ob-
2
3
5
served. The conversion over Pd, PdAPb, Cu, Ni, and CuANi was
ca. 100%, and 71% over PdACO. At the conditions applied, Cu
proved to be the most selective catalyst to mono-olefins (65–
7
0%), closely followed by nickel- and palladium-based catalysts
(
50–60%). Pd, PdAPb and Ni produced iC . Similarly to the valylene
5
and isoprene tests, CO feeding suppressed over-hydrogenation at
the expense of a lower activity.
Through the comparison of conversion and product distribution
in 3MBy and IP hydrogenation at the same temperature and feed
2
H :HC ratio (Fig. 6), the intrinsic selectivity of the diene and
mono-alkyne to mono-olefins over the catalysts can be figured
out. The conversion of 3MBy was slightly higher as well as the
selectivity to oligomers. This is tentatively attributed to the stron-
ger adsorption of the alkyne with respect to the diene [47]. IP
hydrogenation led to an enhanced selectivity to MBs, while the
Fig. 7. Ratio of selectivities to 3-methylbutyne and isoprene in valylene hydroge-
nation over the catalysts. Reaction temperature and feed H :VL ratio for each
2
catalyst are indicated. Other conditions: SV = 16,800 cm3
g
À1 À1
h
and P = 1 bar.
5
iC production from both substrates was comparable. Therefore,
were generally highly chemoselective to isoprene, that is, the triple
bond is predominantly hydrogenated. Fig. 7 shows the selectivity
ratio of 3-methylbutyne and isoprene during valylene hydrogena-
it can be concluded that 3MBy is slightly more reactive than IP
and the alkyne functionality increases the oligomerization ability
of 3MBy, while mono-olefin production is favored during IP hydro-
genation. This is in line with Fig. 3, where more oligomers were
produced in VL hydrogenation than in IP hydrogenation.
2
tion over the catalysts. In this comparison, the feed H :VL ratio for
each catalyst was selected to maximize the production of 3MBy.
Only Pd and PdAPb were purely chemo specific, that is, 3-meth-
ylbutyne was not detected at the reactor outlet. In the other ex-
treme, Cu produced equivalent amounts of IP and 3MBy at
3.3. Selectivity patterns
2
H :VL = 1. PdACO and CuANi produced very low though measur-
In the case of valylene, chemoselectivity is regarded as the com-
petitive hydrogenation of C@C versus C„C, leading to 3-methylbu-
tyne or isoprene, respectively. The catalysts evaluated in this study
able amounts of 3-methylbutyne (S(3MBy) ꢀ 1–2%). Due to its
extensive oligomerization, the chemoselectivity pattern of Ni could
not be clearly established. The higher 3MBy production over

1
72
B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
PdACO with respect to Pd and the lower 3MBy production of
CuANi with respect to Cu highlight the H coverage as an important
factor governing chemoselectivity.
Isoprene hydrogenation raises an interesting feature during the
first and second hydrogen addition as it reveals the regioselectivity
patterns of each catalytic system. A regioselective reaction is one in
which a specific direction of bond making or breaking occurs pref-
erentially over all other possible directions. In the case of isoprene,
it is a favored reduction of one C@C bond over the other by 1,2-
addition of hydrogen (forming 2MB or 3MB) or by 1,4-addition
suppressed (PdACO) the ability of palladium for mono-olefin isom-
erization. Anderson et al. [12] reported that Pd modification by Pb
is more efficient in suppressing cis–trans isomerization than the
double-bond shift in the liquid-phase 2-hexyne and 1-hexyne
hydrogenation, respectively. Nevertheless, in gas-phase valylene
hydrogenation, Pb-poisoned Pd significantly reduced isomerization
compared to bare Pd. On the other hand, Silvestre-Albero et al. [45]
concluded that CO is inefficient in suppressing isomerization dur-
ing 1,3-butadiene hydrogenation using a feed H
ene = 2. Nevertheless, under our condition (H
2
:1,3-butadi-
:3MBy = 1),
2
(
forming 2M2B) (Fig. 1). Nevertheless, it is very complicated to un-
negligible isomerization of 3MB occurred, which further highlights
the influence of the H coverage on isomerization during the overall
hydrogenation process. Contrarily to Pd, Cu did not isomerize and
produced exclusively 3MB during 3MBy hydrogenation. This result
is in agreement with Koeppel et al. [47] that reported only 1-bu-
ravel the true regioselectivity patterns of each catalyst based on
the mono-olefin distribution at the reactor outlet during valyl-
ene/isoprene hydrogenation, since isomerization that lead to C@C
shift within the products may play an important role. Nevertheless,
the isomerization ability of the catalyst can be derived from the
mono-olefin distribution during partial hydrogenation of 3MBy.
In addition of being a probe molecule that belongs to the valylene
hydrogenation pathway, 3-methylbutyne possesses a terminal al-
2
tene production during 1-butyne hydrogenation over Cu/SiO .
The bimetallic CuANi catalyst was inactive for isomerization, too.
This result could have been anticipated as nickel itself showed
minor isomerization, with only 2% and 1% of 2M2B and 2MB,
respectively.
2
kyne that should lead exclusively to 3MB upon H addition.
Fig. 8 compiles the MBs distribution during VL, IP, and 3MBy
hydrogenation under the same condition that maximized the
methylbutene production. Pd was very prone to isomerize and pro-
duced 66% and 12% of 2M2B and 2MB, respectively during 3MBy
hydrogenation. The use of modifiers decreased (PdAPb) or even
No definitive conclusion can be derived on the intrinsic regiose-
lectivity of the catalysts that showed isomerization for system in
which it was not possible to quantify the C = C shift. As mentioned
above, inappropriate space velocity promotes consecutive (over-
hydrogenation) and/or side reaction (isomerization and oligomeri-
zation) that may lead to inconclusive interpretations of the mono-
olefin distribution. Fig. 9 shows the mono-olefin distribution
according to the degree of conversion during isoprene hydrogena-
2
tion over the catalysts. The H :IP feed ratio was established accord-
ing to Fig. 3 and the degree of conversion was varied by changing
the space velocity. As expected from the 3MBy hydrogenation, Cu
was the only catalyst that did not show any variation in its
mono-olefin distribution upon decreasing conversion. The most
drastic changes were observed over Pd. At 100% conversion,
2
M2B was the predominant MBs (88% of the mono-olefins). Upon
increasing the space velocity up to X(IP) = 65%, the relative 2M2B
production drastically decreased to 53% at the expense of both ter-
minal alkenes. Nevertheless from X(IP) = 65% to 28%, little variation
in the mono-olefin distribution was observed. Similar selectivity
patterns were derived during IP hydrogenation over PdAPb and
PdACO. 2M2B production concomitantly decreased with the con-
version degree at the expense of 2MB and 3MB. The only difference
was that below X(IP) = 85% and 80% for PdAPb and PdACO, respec-
tively, the mono-olefin distribution during IP hydrogenation be-
came conversion independent. Goetz et al. [40] reported that two
mechanisms are involved in isomerization. The Horiuti–Polanyi
type mechanism that is favored at high H coverage, and the in-
tra-molecular hydrogen shift mechanism that does not require
the participation of adsorbed hydrogen. Both mechanisms are
clearly identified over Pd. At low conversion (X(IP) < 65%), the H
coverage is sufficiently high to favor the former mechanism. Upon
increasing conversion, the lack of H on the surface enhances the in-
tra-molecular mechanism and therefore, the drastic variation of
mono-olefin at high degree of conversion is attributed to the co-
existing isomerization mechanisms. Contrarily, over PdAPb and
PdACO, the variation on the mono-olefin distribution only occurs
at high degree of conversion. This is attributed to the intra-molec-
ular isomerization mechanism (favored at low H coverage). In
agreement with 3MBy hydrogenation, both Pb and CO modifiers
reduce the degree of isomerization over Pd. The suppression of
the Horiuti–Polanyi isomerization mechanism at low conversion
enables the identification of regioselectivity. During IP hydrogena-
tion, 40% of H additions over PdAPb occurred through 1,4-addition
and 60% through 1,2-addition leading to equal amounts of 2MB
and 3MB. PdACO showed a different regioselectivity exhibiting
62% of 1,4-addition products.
Fig. 8. Mono-olefin distribution in hydrogenation of valylene, isoprene, and 3-
2
methylbutyne over the catalysts. Reaction temperature and feed H :HC ratio for
3
À1 À1
each catalyst are indicated. Other conditions: SV = 16,800 cm
g
h
and P = 1 bar.

B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
173
Fig. 9. Mono-olefin distribution versus the degree of conversion in isoprene hydrogenation over the catalysts. Conditions: IP = 2.5 vol.% and H
2
:IP = 1 (Pd, PdAPb) and 2
(
PdACO, Cu, Ni, CuANi), T = 348 K (Pd, PdACO, PdAPb) and 523 K (Cu, Ni, CuANi). Other conditions are detailed in the Experimental section.
Likewise, the mono-olefin distribution over Ni was conversion
independent from X(IP) = 38% to 75%, with 50%, 35%, and 15% of
M2B, 2MB, and 3MB, respectively. At high conversion, the in-
crease in 1,4-addition products reveals the presence of intra-
molecular isomerization. This is in close agreement with the re-
sults of 3MBy hydrogenation in which little isomerization was ob-
served at 100% conversion.
The mono-olefin distribution during IP hydrogenation over
CuANi is independent of the degree of conversion up to 80% and
shows an intrinsic regioselectivity resembling that of Cu (Fig. 9).
The 1,2-addition is favored toward 2MB production (50%). Never-
theless, a clear increase in 2M2B at the expense of 2MB at higher
conversions evidenced intra-molecular isomerization. This result
is not in disagreement with Fig. 8 as most of the 2M2B was pro-
duced from 2MB isomerization that was not produced during
during VL hydrogenation over PdACO should be taken with care
due to the very low conversion under this condition. Cu was the
only catalyst that allowed a clear interpretation of chemo- and reg-
ioselectivity according to the relative mono-olefin distribution.
Taking into account that 2MB and 2M2B are produced exclusively
from isoprene hydrogenation, and the composition of MBs is
3MB:2MB:2M2B = 1.5:3:1 and 8.5:3:1 during IP and VL hydrogena-
tion, respectively, we can conclude that 56% of VL was hydroge-
nated through a 3MBy intermediate on Cu. This result is in
agreement with the equivalent IP and 3MBy selectivity during val-
2
2
ylene hydrogenation at H :VL = 1.
3.4. Ene-yne versus simpler alkynes
Fig. 10 shows the selectivity toward the products of the first
hydrogenation of valylene, isoprene, and propyne, that is, iso-
prene + 3-methylbutyne, methylbutenes, and propene, respec-
tively. These selectivities are compared under the same
conditions that lead to the highest IP + 3MBy selectivity during
VL hydrogenation (see Fig. 3). At first glance, the selectivity to
the first hydrogenated product is lower in valylene hydrogenation
than in isoprene or propyne hydrogenation. These results can be
3
MBy hydrogenation. We assume that isomerization by intermo-
lecular hydrogen over CuANi is favored over the branched carbon
from 2MB to 2M2B than from 3MB to 2M2B.
Finally, the comparison of the MBs distribution in VL and IP
hydrogenation (Fig. 8) gives further insights into the chemoselec-
tivity of the catalytic systems. The higher 3MB production during
VL hydrogenation is the proof of the hydrogenation pathway with
3
MBy as intermediate. Therefore, Pd and PdAPb, which were de-
easily understood by the nature of the product from the first H
2
scribed as the only chemo specific catalysts, should be revised
and restricted to the only catalysts that did not produce 3MBy. This
is in agreement with the higher reactivity of the alkyne that once
produced on Pd and PdAPb surface does not desorb and gets fur-
ther hydrogenated. It has to be mentioned that the excessively
high relative production of 3MB compared to 2MB and 2M2B
addition on VL (IP + 3MBy) that are still highly unsaturated prod-
ucts and their strong adsorptions compete with VL, which drasti-
cally enhance further hydrogenation. On the other hand, the
products from the first hydrogenation of isoprene and propyne
possess only one unsaturation and hardly compete with the
reactant. In addition to chemoselectivity, regioselectivity, and

1
74
B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
all catalysts. However, marked differences in product distribution
were observed. The selectivity trends derived from valylene hydro-
genation were ascertained. Under the conditions applied, no over-
hydrogenation leading to pentane was evidenced in any case. Pd
was the most selective catalyst toward the non-conjugated diene
(S(1,4-pentadiene) = 51%). No 1,4-pentadiene was formed over
PdAPb which produced only pentenes and oligomers. The most
important point arises with the production of the conjugated 1,3-
pentadienes over Pd (2%), Cu (14%), and CuANi (16%). The presence
of the conjugated dienes is attributed to the isomerization of di-
olefin. This restores the conjugation, enhancing the stability of
the molecule. Although operating at full conversion may bias the
interpretation, the comparison from valylene and 1-penten-4-yne
highlights that a single carbon atom between the ene and the
yne functionalities is insufficient to overcome the ‘‘selectivity chal-
lenge’’ of partial hydrogenation of conjugated ene-yne compounds.
More separation between the double and triple bonds in the
hydrocarbon is required to minimize conjugation effects in par-
tially hydrogenated intermediates, which makes selectivity control
extremely difficult.
Fig. 10. Selectivity to the product of the first H
isoprene (striped bar), and propyne (gray bar). Reaction temperature and feed
:HC ratio for each catalyst are indicated. Other conditions: SV =
2
addition on valylene (solid bar),
H
2
3
À1 À1
1
6,800 cm
g
h
and P = 1 bar.
4
. Conclusions
The gas-phase hydrogenation of valylene and isoprene has been
systematically studied in a broad range of hydrogen-to-hydrocar-
bon ratios and degrees of conversion over heterogeneous catalysts,
including Pd, Pb-poisoned Pd, CO-modified Pd, Ni, Cu, and CuANi.
Favorable reaction paths, leading to chemoselectivity, regioselec-
tivity, isomerization, full hydrogenation, oligomerization, and
CAC bond scission footprints, are determined by the nature of
the metal and the modifier. Bare Pd and lead-poisoned Pd catalysts
were highly chemo specific and continuous CO feeding is a partic-
ular effective strategy to decrease Pd activity and widen the win-
dow for selective hydrogenation. Cu produced equivalent
amounts of isoprene and 3-methylbutene at a hydrogen-to-valyl-
ene ratio of 1. Ni-promoted Cu catalyst enhances isoprene selectiv-
ity and decreases oligomerization, while Ni-based catalyst was the
less suitable catalyst for ene-yne and diene hydrogenation due to
extensive oligomerization and cracking. The distribution of
mono-olefins from isoprene and 3-methylbutyne hydrogenation
was compared in order to understand the interplay between regi-
oselectivity and isomerization on the catalysts. Cu, CuANi, Ni, and
PdACO showed negligible isomerization during 3-methylbutyne
hydrogenation compared to Pd and PbAPd. Nevertheless, the
mono-olefin distribution dependence on the degree of isoprene
conversion highlighted that most catalysts promoted intra-molec-
ular isomerization and favored the 1,4-addition product at full con-
version. Cu was the only catalyst for which isomerization was not
observed, resulting in an enhanced selectivity to 1,2-addition
olefins. Low conversion levels enabled the study of catalytic perfor-
mance in the absence of isomerization. Consequently, the intrinsic
regioselectivity of PdAPb, PdACO, Ni, and CuANi was derived.
Comparison with simpler alkynes such as propyne or even iso-
prene emphasized the challenges in selectivity control during
hydrogenation of conjugated (valylene) or non-conjugated (1-pen-
ten-4-yne) ene-yne compounds. Our results derived from the
selectivity dependence on hydrogen-to-hydrocarbon ratio and
the degree of conversion highlight the importance of catalyst and
reactor design in ene-yne hydrogenation.
Fig. 11. Selectivity to products in 1-penten-4-yne hydrogenation over Pd, PdAPb,
Cu, and CuANi. Reaction temperature and feed H :1-penten-4-yne ratio for each
catalyst are indicated. Other conditions: SV = 16,800 cm
2
3
À1 À1
g
h , and P = 1 bar.
isomerization issues intrinsic to valylene, the selectivity control in
terms of consecutive hydrogenation remains a challenging task
compared to propyne or even isoprene hydrogenation. Upon
comparison between S(MBs) and S(propene) during isoprene and
propyne hydrogenation, respectively, an interesting feature ap-
peared. Due to the stronger adsorption of the alkyne with respect
to the diene, one could expect a higher relative selectivity to the
first hydrogenated product during propyne hydrogenation com-
pared to IP hydrogenation. Nevertheless, the minor steric hin-
drance of propyne coupled with the ability of the alkyne to form
oligomers decrease propene selectivity in propyne hydrogenation
compared to MBs selectivity in IP hydrogenation.
Complementarily to valylene hydrogenation, 1-penten-4-yne,
the smallest branched hydrocarbon possessing non-conjugated
ene-yne, was used to investigate the impact of conjugation in het-
erogeneously catalyzed hydrogenation. Fig. 11 shows the selectiv-
ity to products during 1-penten-4-yne hydrogenation over Pd,
PdAPb, Cu, and CuANi. These catalysts were chosen since they
are the most appropriate for ene-yne compounds hydrogenation
according to Fig. 3. The reactions were undertaken at a fixed
Acknowledgment
2
H :HC ratio, which was selected to maximize the partially hydro-
genated production. A full degree of conversion was achieved over
ETH Zürich is acknowledged for financial support.

B. Bridier, J. Pérez-Ramírez / Journal of Catalysis 284 (2011) 165–175
175
[
21] F. Studt, F. Abild-Pedersen, T. Bligaard, R.Z. Sorensen, C.H. Christensen, J.K.
Nørskov, Science 320 (2008) 1320.
References
[
[
[
[
[
[
[
[
22] Y. Azizi, C. Petit, V. Pitchon, J. Catal. 269 (2010) 26.
[
[
1] R.J. Farrauto, C.H. Bartholomew, Fundamentals of Industrial Catalytic
Processes, Blackie & Son Ltd., 1997. p. 411.
2] H.F. Rase, Handbook of Commercial Catalysts, CRC Press, Boca Raton, 2000. p.
23] B. Bridier, J. Pérez-Ramírez, J. Am. Chem. Soc. 132 (2010) 4321.
24] J. Goetz, M.A. Volpe, R. Touroude, J. Catal. 164 (1996) 369.
25] A. Sárkány, Appl. Catal., A 149 (1997) 207.
26] A. Hugon, L. Delannoy, C. Louis, Gold Bull. 41 (2008) 127.
27] A. Hugon, L. Delannoy, J.M. Krafft, C. Louis, J. Phys. Chem. C 114 (2010) 10823.
28] J.A. Alves, S.P. Bressa, O.M. Martínez, G.F. Barreto, Chem. Eng. J. 99 (2004) 45.
29] K. Pattamakomsan, F. Aires, K. Suriye, J. Panpranot, Reac. Kinet. Mech. Catal.
1
37.
[
[
[
3] Á. Molnár, A. Sárkány, M. Varga, J. Mol. Catal. A: Chem. 173 (2001) 185.
4] C.J. Kliewer, M. Bieri, G.A. Somorjai, J. Am. Chem. Soc. 131 (2009) 9958.
5] M.L. Derrien, in: L. Cˇ erven y´ (Ed.), Catalytic Hydrogenation, Stud. Surf. Sci.
Catal., vol. 27, Elsevier, Amsterdam, 1986, p. 613.
6] G.C. Bond, Metal-Catalysed Reactions of Hydrocarbons, Springer, New York,
103 (2011) 405.
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
30] T.-B. Lin, T.-C. Chou, Appl. Catal., A 108 (1994) 7.
31] J.-C. Chang, T.-C. Chou, Appl. Catal., A 156 (1997) 193.
32] H.R. Aduriz, P. Bodnariuk, B. Coq, F. Figueras, J. Catal. 129 (1991) 47.
33] J.Y. Ryu, J.A. Adams, W.A. Groten, US 7319176, 2008.
34] J.Y. Ryu, US 7838710, 2010.
2
005.
7] D. Teschner, E. Vass, M. Havecker, S. Zafeiratos, P. Schnoerch, H. Sauer, A. Knop-
Gericke, R. Schlögl, M. Chamam, A. Wootsch, A.S. Canning, J.J. Gamman, S.D.
Jackson, J. McGregor, L.F. Gladden, J. Catal. 242 (2006) 26.
8] D. Teschner, J. Borsodi, A. Wootsch, Z. Révay, M. Hävecker, A. Knop-Gericke,
S.D. Jackson, R. Schlögl, Science 320 (2008) 86.
[
[
35] L.Kh. Freidlin, V.I. Gorshkov, Dokl. Akad. Nauk SSSR 131 (1960) 1109.
36] L.Kh. Freidlin, Y.Y. Kaup, Izv. Akad. Nauk. SSSR, Ser. Khim. 8 (1964) 1501.
37] B. Bridier, N. López, J. Pérez-Ramírez, Dalton Trans. 39 (2010) 8412.
38] S. Abelló, D. Verboekend, B. Bridier, J. Pérez-Ramírez, J. Catal. 259 (2008) 85.
39] B. Bridier, N. López, J. Pérez-Ramírez, J. Catal. 269 (2010) 80.
40] J. Goetz, M.A. Volpe, C.E. Gigola, R. Touroude, J. Catal. 199 (2001) 338.
41] A. Borodzi n´ ski, G.C. Bond, Catal. Rev. Sci. Eng. 48 (2006) 91.
42] H. Lindlar, Helv. Chim. Acta 35 (1952) 446.
9] M. Ruta, N. Semagina, L. Kiwi-Minsker, J. Phys. Chem. C 112 (2008) 13635.
[
[
[
[
[
10] N. López, B. Bridier, J. Pérez-Ramírez, J. Phys. Chem. C 112 (2008) 9346.
11] D. Mei, M. Neurock, C.M. Smith, J. Catal. 268 (2009) 181.
12] J.A. Anderson, J. Mellor, R.P.K. Wells, J. Catal. 261 (2009) 208.
13] M. García-Mota, B. Bridier, J. Pérez-Ramírez, N. López, J. Catal. 273 (2010) 92.
14] M. Armbrüster, K. Kovnir, M. Behrens, D. Teschner, Y. Grin, R. Schlögl, J. Am.
Chem. Soc. 132 (2010) 14745.
43] R.T. Vang, K. Honkala, S. Dahl, E.K. Vestergaard, J. Schnadt, E. Lægsgaard, B.S.
Clausen, J.K. Nørskov, F. Besenbacher, Nat. Mater. 4 (2005) 160.
44] B. Bridier, J. Pérez-Ramírez, A. Knop-Gericke, R. Schlögl, D. Teschner, Chem. Sci.
[
15] M. García-Mota, J. Gómez-Díaz, G. Novell-Leruth, C. Vargas-Fuentes, L.
Bellarosa, B. Bridier, J. Pérez-Ramírez, N. López, Theor. Chem. Acc. 128
[
(
2011) 663.
2
(2011) 1379.
[
[
[
16] F.-M. McKenna, R.P.K. Wells, J.A. Anderson, Chem. Commun. 47 (2011) 2351.
17] M.W. Tew, H. Emerich, J.A. van Bokhoven, J. Phys. Chem. C 115 (2011) 8457.
18] J.C. Rodriguez, A.J. Marchi, A. Borgna, E. Romeo, A. Monzón, Stud. Surf. Sci.
Catal. 139 (2001) 37.
19] J.A. López-Sánchez, D. Lennon, Appl. Catal., A 291 (2005) 230.
20] Y. Segura, N. López, J. Pérez-Ramírez, J. Catal. 247 (2007) 383.
[
[
45] J. Silvestre-Albero, G. Rupprechter, H.J. Freund, J. Catal. 235 (2005) 52.
46] K. Pattamakomsan, K. Suriye, S. Dokjampa, N. Mongkolsiri, P. Praserthdam, J.
Panpranot, Catal. Commun. 11 (2010) 311.
[
47] R.A. Koeppel, J.T. Wehrli, M.S. Wainwright, D.L. Trimm, N.W. Cant, Appl. Catal.,
A 120 (1994) 163.
[
[