Liquid-phase hydrogenation of citral over Pt/SiO2 catalysts - II. Hydrogenation of reaction intermediate compounds

Article abstract of DOI:10.1006/jcat.1999.2804

Liquid-phase hydrogenation of the four principal reaction intermediates during citral hydrogenation (nerol, geraniol, citronellal, and citronellol) was studied at 298 and 373 K under 20 atm H₂ at concentrations of 0.5-1M in hexane. Decomposition of geraniol and nerol to form adsorbed CO on Pt as an inhibitor was proposed to be responsible for the activity minimum observed during citral hydrogenation. A decrease in the initial reaction rate as temperature increased from 298 to 373 K was exhibited during the hydrogenation of all four compounds. Furthermore, simultaneous hydrogenation of citronellal and geraniol at 298 K resulted in a continuous decrease in the rate of citronellal disappearance in contrast to the nearly constant rate of disappearance observed during hydrogenation of citronellal alone. Competitive hydrogenation of citral with either geraniol or citronellal showed that geraniol hydrogenation to citronellol was kinetically insignificant during citral hydrogenation at 373 K. The relative rates of hydrogenation of the compounds indicated that the hydrogenation rate of the C=C bond in this family of compounds was greater than that of the C=O bond, as evidenced by the ordering of the hydrogenation activity: geraniol > nerol > citronellol > E-citral > citronellal _ Z-citral. The product distributions obtained during hydrogenation of citronellal indicated that higher reaction temperatures favor activation of the C=O bond relative to the C=C bond.

Full text of DOI:10.1006/jcat.1999.2804

Journal of Catalysis 191, 181–191 (2000)
doi:10.1006/jcat.1999.2804, available online at http://www.idealibrary.com on
Liquid-Phase Hydrogenation of Citral over Pt/SiO2 Catalysts
II. Hydrogenation of Reaction Intermediate Compounds
Utpal K. Singh, Matthew N. Sysak, and M. Albert Vannice1
Department of Chemical Engineering, Pennsylvania State University, 107 Fens`ke Laboratory, University Park, Pennsylvania 16802-4400
Received September 3, 1999; revised December 16, 1999; accepted December 22, 1999
on the kinetics of citral hydrogenation. In the previous pa-
per, an unusual effect of temperature on product distribu-
tion and reaction rate was reported and a model was de-
veloped to explain the observed behavior. The minimum in
activity for citral hydrogenation that occurred between 298
and 423 K was attributed to Pt site blockage by adsorbed
CO formed by decomposition of the unsaturated alcohols
(geraniol and nerol) coupled with its subsequent desorp-
tion at higher temperatures (1). Literature regarding the
kinetics of liquid-phase hydrogenation of the reaction in-
termediates that can be formed during the hydrogenation
of , -unsaturated aldehydes is extremely scarce (2). As
with citral, the influence of reaction parameters on these
kinetics remains to be studied in a systematic fashion un-
der conditions shown to be free of transport limitations and
poisoning effects. Figure 1 presents the reaction pathways
that can occur to form intermediate compounds during cit-
ral hydrogenation. In the present paper, hydrogenation of
the reaction intermediates both as a single component in
the solvent and in a mixture of reactants was investigated
using citronellal (partially saturated aldehyde), nerol and
geraniol (unsaturated alcohol), and citronellol (partially
saturated alcohol). In addition to measuring turnover fre-
quencies (TOFs) to appropriately compare relative activi-
ties, emphasis was placed on examining temperature effects
for hydrogenation of reaction intermediate compounds.
Liquid-phase hydrogenation of the four principal reaction inter-
mediates formed during citral hydrogenation, i.e., nerol, geraniol,
citronellal, and citronellol, was studied at 298 and 373 K under
20 atm H2 at concentrations of 0.5 to 1.0 M in hexane. A decrease in
the initial reaction rate as temperature increased from 298 to 373 K
was exhibited during the hydrogenation of all four compounds, just
as reported earlier for citral; however, the decrease in rate at 373 K
was only one-half for citronellal whereas it was orders of magnitude
greater for nerol and geraniol. Furthermore, simultaneous hydro-
genation of citronellal and geraniol at 298 K resulted in a continuous
decrease in the rate of citronellal disappearance in contrast to the
nearly constant rate of disappearance observed during hydrogena-
tion of citronellal alone. Competitive hydrogenation of citral with
either geraniol or citronellal showed that geraniol hydrogenation to
citronellol is kinetically insignificant during citral hydrogenation
at 373 K. The initial activity for hydrogenation of the intermedi-
ates at 298 K follows the following trend: geraniol (1.9 s 1) > nerol
(1.5 s 1) > citronellol (0.21 s 1) > E-citral (0.13 s 1), citronellal
(0.13 s 1) > Z-citral (0.06 s 1). Based on the relative hydrogena-
tion rates of the intermediate alone versus its hydrogenation in
the presence of other reactants, the relative size of the adsorption
equilibrium constants for the various organic compounds appears
to be as follows: Kcitral > Kcitronellal > Kgeraniol, Knerol >Kcitronellol
>
==
K3,7-dimethyloctanol. This study indicates that activation of the C
O
bond should be performed at higher reaction temperatures to max-
c
imize selectivity to the unsaturated alcohols.
2000 Academic Press
INTRODUCTION
EXPERIMENTAL
SiO2 (Davison grade 57, 220 m2/g, 60–100 mesh) was
dried and calcined at 773 K for 4 h prior to the drop-
wise addition of an aqueous solution of chloroplatinic
acid (H2PtCl6, Aldrich 99.995% ) via an incipient wetness
(2cm3/g) method, then dried overnight in air at 393K. Three
SiO2-suupported catalysts with Pt loadings of 0.49, 1.44,
and 3.8 wt% were prepared and characterized by hydrogen
chemisorption as described previously (1), and respective
dispersions of 0.45, 0.41, and 0.67 were measured.
This is the second portion of a two-part series discussing
liquid-phase hydrogenation of citral and its intermediate
products over Pt/SiO2 catalysts. Hydrogenation of , -
unsaturated aldehydes represents a broad class of reactions
relevant to the specialtyand fine chemicalsindustrybecause
of the importance of selectivity control in systems contain-
==
==
ing conjugated C C and C O bonds. For example, much
work remains to be done to understand the influence of re-
action parameters and metal–support interactions (MSIs)
Hydrogenation experiments were conducted in a 100-ml
EZ-seal autoclave with automated data acquisition. The
1 To whom correspondence should be addressed.
181
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

182
SINGH, SYSAK, AND VANNICE
FIG. 1. Reaction network for citral hydrogenation.
pretreatment procedure is identical to that used during the pump (ISCO 500D). The hexane/catalyst slurry was stirred
study of citral hydrogenation and has been described before and allowed to equilibrate for 30 min prior to the introduc-
(1). Briefly, the catalyst was reduced in situ at 673 K for tion of the reactant at reaction temperature and pressure to
75 min prior to cooling to room temperature and storing obtain a totalreaction volume of60ml. The reactor pressure
overnight under a static H2 pressure of 20 atm.
was continuously maintained to within 5% of the set point
Geraniol (trans-3,7-dimethyl-2,6-octadienol), nerol (cis- (20 atm) using a Brooks 5860 pressure controller. A liquid
3,7-dimethyl-2,6-octadienol), citronellal (3,7-dimethyl-6- sample (ca. 0.5 ml) was periodically withdrawn through a
octenal), and citronellol (3,7-dimethyl-6-octenol) were dip tube inside the reactor, collected in a closed, N2-purged
used as received from Sigma with respective purities of 98, vessel, then analyzed in a Hewlett–Packard 5890 gas chro-
98, 85, and 95% . Both the reactants and the hexane solvent matograph equipped with a thermal conductivity detector
were degassed by sparging with high flow rates of nitrogen and a 10% Carbowax 20 M on Supelcoport column. Ini-
for 30 min prior to injection into the reactor. Competitive tial rates of citral disappearance were evaluated using the
hydrogenation reactions of geraniol and citronellal were slope of the linear portion of the temporal citral conversion
performed by mixing equal volumes of each reactant and profile (for citral conversions below 20% ). The selectiv-
sparging the mixture for 30 min with nitrogen prior to ad- ity to each product was calculated based on the following
dition into the reactor. Experiments were also performed equation:
in which an intermediate was introduced into the reaction
molar concentration of species i
system during the course of citral hydrogenation. As with
citral, extra precautions were taken to remove trace quan-
tities of oxygen dissolved in any reactant introduced into
the reactor. The standard reaction conditions in the present
study consisted of conducting the reaction using 10 ml of
reactant (5 ml of each reactant during competitive hydro-
P
Si =
.
products molar concentration of species i
RESULTS
Figure 2a displays the temporal concentration profiles of
genation) dissolved in 50 ml of hexane under a constant H2 reactants and products during geraniol hydrogenation at
pressure of 20 atm. n-Hexane (Fisher, 99.9% saturated C6 298 K. All results are under standard conditions unless oth-
hydrocarbons) was fed into the reactor from an enclosed erwise noted. Citronellol and 3,7-dimethyloctanol were the
environment to prevent any exposure to air at the reac- principal products. Trace quantities of trans-3,7-dimethyl-
==
tion temperature and pressure usinga high-pressure syringe 2-octenol from hydrogenation of the isolated C C bond

LIQUID-PHASE HYDROGENATION OF CITRAL OVER Pt/SiO2, II
183
hydrogenated to yield 3,7-dimethyloctanol. A reaction or-
der of 0.5 in geraniol concentration was obtained from a
differential treatment of the rate of geraniol disappearance
as shown in Fig. 2b.
Figure 3 displays the temporal concentration profile dur-
ing hydrogenation of geraniol at 373 K. These results in-
dicate that there is rapid hydrogenation of geraniol dur-
ing the first few minutes of reaction to yield citronellol
and give a conversion near 20% ; however, activity is sub-
sequently suppressed by orders of magnitude and a TOF
1
of only 0.005 s was obtained after this early period. An
initial TOF of 0.20 s 1 was observed with negligible deac-
tivation during hydrogenation of pure citronellol at 298 K;
however, reaction at 373 K resulted in rapid conversion of
15% after 2 min followed by deactivation and a TOF of
only 0.0004 s 1 during the latter period as shown in Fig. 4.
Figures 5a and 5b show the temporal concentration pro-
files during citronellal hydrogenation at 298 and 373 K. The
initial TOFs for citronellal disappearance were 0.13 s 1 at
1
298 K and 0.07 s at 373 K. In stark contrast to the ki-
netic behavior of geraniol, hydrogenation of citronellal at
298 and 373 K exhibited zero-order kinetics as shown in
Figs. 6a and 6b, respectively. The results in Figs. 5a and 6a
indicate that citronellal hydrogenation at 298 K exhibits
zero-order kinetics up to 50% conversion and a twofold
lower rate is observed for conversions greater than 50% ;
however, zero-order kinetics are maintained at this lower
rate. It should be noted that the y intercepts for citronellal
concentration profile in Figs. 5a and 5b are at 85% due to the
presence of isopulegol in the feed. It was necessary to cor-
rect for the feed impurity to properly evaluate selectivities
since isopulegol is a product of citronellal hydrogenation.
This was especially true at 298 K where the temporal isop-
ulegol concentration profile exhibits a continuous decrease
with time. Therefore, an approximate correction was made
by assuming that no isopulegol was formed under reaction
FIG. 2. (a) Temporal concentration profile during hydrogenation of
geraniol in hexane at 298 K over 1.44% Pt/SiO2 (20 atm H2 pressure, 1 M
geraniol, 98% pure). (b) Differential treatment of the rate of geraniol
disappearance.
in geraniol were detected, but this product could not be
quantitatively resolved from citronellol. During nerol hy-
drogenation, the GC peak for the cis isomer of the unsat-
urated alcohol (cis-3,7-dimethyl-2-octenol) could be sepa-
rated from that for citronellol, and the results revealed that
the solvent-free composition of this product was less than
5% of the total reaction mixture; also, it was rapidly hydro-
genated to the final saturated product after the initial stages
of the reaction to give nondetectable concentrations. The
initial TOF for geraniol hydrogenation was 1.8 s 1. Such
a large value for the TOF raises a concern about possible
mass transfer limitations; however, the Weisz–Prater cri-
terion for minimal mass transport limitations was satisfied
4
for geraniol hydrogenation as a value of 2.4 10 was
obtained for this dimensionless parameter at 298 K (3, 4).
Therefore, the kinetics displayed in Fig. 2a should be free
of any mass transport control. The results indicate that
geraniol hydrogenation occurs rapidly to yield primarily
FIG. 3. Hydrogenation of geraniol in hexane at 373 K over 1.44%
citronellol. At higher conversions, citronellol was further Pt/SiO2 (20 atm H2 pressure, 1 M geraniol, 98% pure).

184
SINGH, SYSAK, AND VANNICE
FIG. 4. Hydrogenation of citronellol in hexane at 373 K over 1.44% Pt/SiO2 (20 atm H2 pressure, 1 M citronellol, 95% pure).
conditions and the isopulegol present in the feed reacted 7b display the selectivity-versus-conversion profile for cit-
to form the menthol isomers. The selectivity at 373 K was ronellal hydrogenation at 298 and 373 K, respectively, and
evaluated by subtracting the isopulegol feed concentration dramatic differences can be seen between the two product
from that evaluated at each point in time. Figures 7a and distributions. Reaction at 298 K yielded dihydrocitronellal,
FIG. 6. Differential treatment of the rate of citronellal disappearance
FIG. 5. Hydrogenation of citronellal in hexane at (a) 298 K and
(b) 373 K over 1.44% Pt/SiO2 (20 atm H2 pressure, 1 M citronellal). Cit-
ronellal was obtained with a purity of 85% and used as is.
for reaction at (a) 298 K and (b) 373 K over over 1.44% Pt/SiO2 (20 atm H2
pressure, 1 M citronellal). Citronellal was obtained with a purity of 85%
and used as is.

LIQUID-PHASE HYDROGENATION OF CITRAL OVER Pt/SiO2, II
185
selectivity to 3,7-dimethyloctanol increased from 19 to 33%
over the same conversion range. No dihydrocitronellal or
3,7-dimethyloctanol was observed during reaction at 373 K,
and the selectivities to citronellol and isopulegol were 60
and 40% , respectively, and essentially constant with respect
to conversion.
Due to the drastic differences in the kinetic behavior of
citronellal and geraniol, the competitive hydrogenation of
citronellaland geraniolwasexamined, and the initialhydro-
genation rates for each intermediate alone and that during
competitive hydrogenation of an approximately equimolar
mixture of the two intermediates are presented in Table 1.
Asmentioned previously, the initialhydrogenation ratesfor
citral, citronellal, and geraniol decrease as the temperature
increases from 298 to 373 K. The temporal concentration
profiles for simultaneous hydrogenation of 0.5 M geraniol
and 0.4 M citronellal over the 1.44% Pt/SiO2 catalyst are
shown in Fig. 8a. The profiles during the first 3 h of reac-
tion are shown in the inset to Fig. 8a and these were used
to calculate the initial TOFs for disappearance of geraniol
and citronellal, which were 0.18 and 0.08 s 1, respectively.
There is a sustained decrease in the total rate of hydrogena-
tion and negligible activity exists after 40 h of reaction as
seen from the H2 uptake profiles in Fig. 8b which show that
the rate of H2 uptake during competitive hydrogenation of
1
geraniol and citronellal is near 0.5 s initially but even-
tually decreases to near zero after 40 h of reaction time.
Figure 9 displays the concentration profiles during hydro-
genation of a similar mixture at 373 K. A 20% conversion of
geraniol is observed during the first 5 min of reaction, and
after this high initial reaction rate, the concentration profile
levels out, yielding a much lower TOF of 0.005 s 1. The ini-
tial rate of disappearance of citronellal in the mixture was
0.03 s 1, which represented a twofold drop compared with
reaction with citronellal alone.
FIG. 7. Product distributions during hydrogenation of citral in hexane
at (a) 298 K and (b) 373 K over 1.44% Pt/SiO2 catalyst (20 atm H2 pressure,
1 M citral).
citronellol, and 3,7-dimethyloctanol while reaction at 373 K
yielded only isopulegol and citronellol. The selectivity to di-
hydrocitronellal at 298 K decreased from 58 to 33% as the
citronellal conversion increased from 17 to 70% while the
TABLE 1
Initial Turnover Frequencies for Hydrogenation of Reaction Intermediates at 298 and 373 K over
1.44% Pt/SiO2 Catalyst (20 atm H2, 1 M in Hexane)
Activity at 298 K
TOF at 298 K
Activity at 373 K
TOF at 373 K
1
1
(
mol/g cat/min)
(s
)
(
mol/g cat/min)
(s )
Citral
236
112
E-isomer: 0.13
Z-isomer: 0.06
18
9
E-isomer: 0.0098
Z-isomer: 0.0050
Geraniol
Nerol
3330 (852)a,b
2688
1.9 (1.2)a,b
1.5
9
4
0.005c
0.007c
Citronellal
Citronellol
Geraniol/nerol
(50 mol% )
Geraniol/citronellal
(50 mol% )
251 (196)a
140
0.13 (0.11)a
0.21
131
0.7
—
—
7
0.07
0.0004c
829
862
331
138
Geraniol: 0.46
Nerol: 0.48
Geraniol: 0.18
Citronellal: 0.08
—
—
Geraniol: 0.004.
Citronellal: 0.02
30
a Values in parentheses obtained with 0.5 M citral in hexane.
b Reaction conducted over 0.49% Pt/SiO2.
c Rapid initial deactivation.

186
SINGH, SYSAK, AND VANNICE
FIG. 8. (a) Temporal citral concentration profile and (b) temporal H2 uptake during competitive hydrogenation of citronellal and geraniol in
hexane over 1.44% Pt/SiO2 at 298 K (20 atm H2, 0.4 M citronellal, 0.5 M geraniol).
Figures 10a and 10b show the TOFs for H2 uptake dur-
ing reaction at 298 K before and after addition of geraniol
or citronellal. Citronellal addition after 40 min of reaction
had essentially no influence on the rate of H2 uptake and
the TOF for H2 uptake. Addition of geraniol, on the other
hand, resulted in a twofold increase in the total rate of H2
consumption which was due primarily to an increase in the
rate of conversion of geraniol to citronellol; however, the
TOF for H2 uptake eventually decreased to that observed
without the addition of geraniol.
Table 2 displays the effect of geraniol or citronellal ad-
dition on the net rate of disappearance of reactants and
formation of products for citral hydrogenation over 3.8%
FIG. 9. (a) Temporal citral concentration profile and (b) temporal H2
uptake during competitive hydrogenation of citronellal and geraniol in
hexane over 1.44% Pt/SiO2 at 373 K (20 atm H2, 0.4 M citronellal, 0.5 M
Pt/SiO2 at 373 K. During citral hydrogenation, the TOF for
citral disappearance was 0.006 s 1, while the TOFs for for-
mation of geraniol and nerol, citronellal, citronellol, and
geraniol).

LIQUID-PHASE HYDROGENATION OF CITRAL OVER Pt/SiO2, II
187
exhibited prior to the addition of citronellal. A fivefold en-
hancement in the net rate of isopulegol and citronellol for-
mation wasobserved after addition ofcitronellalasthe TOF
1
for isopulegol formation increased from 0.0008 to 0.004 s
and that for citronellol formation increased from 0.0008 to
0.0036 s 1. The net rate of formation of the unsaturated al-
cohol (geraniol and nerol) was unaffected, as seen by the
constancyin the net rate offormation ofthe productsbefore
and after addition of citronellal. It should be noted that the
TOF for citronellal disappearance, after the addition of cit-
ronellal to a citral hydrogenation reaction mixture, was 20-
fold lower than that during hydrogenation of 1 M citronellal
alone. Addition of geraniol to a citral hydrogenation reac-
tion mixture did not affect the rate of citral disappearance,
1
and a more modest rise occurred from 0.0008 to 0.0017 s
in the TOF for citronellol formation.
Table 1 also lists the specific activity (as TOF) at 298 K
for hydrogenation of either 1 M geraniol or 1 M nerol
separately as well as for simultaneous hydrogenation of
0.5 M geraniol and 0.5 M nerol. The initial TOFs for
hydrogenation of 1 M geraniol and 1 M nerol were 1.9 and
1.5 s 1, respectively. Simultaneous hydrogenation of 0.5 M
geraniol and 0.5 M nerol gave an initial TOF of 0.94 s 1 for
the rate of disappearance of geraniol + nerol with a reactiv-
1
ity ratio of unity. This is contrasted with the TOF of 1.2 s
obtained during hydrogenation of 0.5 M geraniol alone.
DISCUSSION
FIG. 10. Specific activity for H2 uptake before and after addition of
(a) 8.5 ml of geraniol and (b) 8.5 ml of citronellal during hydrogenation of
citral in hexane over 1.44% Pt/SiO2 at 298 K (20 atm H2, 1 M initial citral
concentration).
The results presented earlier for hydrogenation of citral
(10 mol% ) in hexane under reaction conditions similar to
those used in the present study have shown the absence of
transport limitations and poisoning effects (1). Because the
TOFs for all intermediates, except geraniol and nerol, were
similar to that for citral, the kinetic data from the present
study are also free of transport limitations. The unsaturated
alcohols (geraniol and nerol) exhibit very high TOFs; how-
ever, the Weisz–Prater criterion was satisfied and therefore
even the results for geraniol hydrogenation at 298 K are
free of transport limitations (3, 5). No evidence for the
leaching of metals from the reactor walls was found (3);
consequently, because similar compounds were used in the
present case with the same solvent, no leaching is antici-
pated under reaction conditions.
isopulegol were 0.0039, 0.0002, 0.0008, and 0.0008 s 1, re-
spectively. Addition of citronellal to a citral hydrogenation
reaction mixture at 373 K yielded a TOF for citral disap-
pearance of 0.0071 s 1, which was within experimental un-
certainty of the value for the rate of citral disappearance
TABLE 2
Effect of Addition of Geraniol or Citronellal on Hydrogenation
Rates at 373 K in Hexane over 3.8% Pt/SiO2 (20 atm H2, 1 M
Citral)
The reaction network in Fig. 1 shows the cis and trans
isomers of both 3,7-dimethyl-2-octenal (ENAL) and 3,7-
dimethyl-2-octenol as products. These products have not
been routinely reported in reaction networks for citral hy-
drogenation over Ru catalysts (6, 7). Although these prod-
ucts could not be unambiguously identified, indirect evi-
dence for their presence was obtained based on GC–MS
analysis and by the nerol hydrogenation kinetic data, which
allowed the resolution of a peak for cis-3,7-dimethyl-2-
octenol distinct from other product peaks (3).
1
TOF (1000
s
)
Prior to addition
of intermediate
Addition of
citronellal
Addition of
geraniol
Compound
Citral
Geraniol and nerol
Citronellal
(5.8)a
3.9
0.2
0.8
0.8
(7.1)a
3.9
(5.8)a
3.9
0.3
1.7
0
(3.9)a
3.6
Citronellol
Isopulegol
3.9
a Values in parentheses represent net rate of disappearance.

188
SINGH, SYSAK, AND VANNICE
The present study was conducted to understand the influ- tion of 2-methyl-2-propen-1-ol over Rh/AlPO4 catalysts at
ence ofthe reaction intermediate on the kineticsofcitralhy- 323 K where a rapid initial hydrogenation rate was observed
drogenation (1). Waghray and Blackmond have suggested yielding 22% conversion after which there was negligible
that a saturated aldehyde (3-methyl-butanal) can undergo activity (13). The behavior observed during geraniol hy-
decarbonylation to yield a carbonaceous species similar to drogenation was analogous to that observed during citral
that obtained with 3-methyl-2-butenal in the presence of hydrogenation at 373 K, i.e., a high initial rate that rapidly
H2, but which was not observed in the absence of H2 (8). decreased to a constant value during the first few minutes of
Therefore, they concluded that the carbonaceous species reaction. The high initial activity prior to the constant rate
and the adsorbed CO were due to decarbonylation of the attained during citral hydrogenation at 373 K manifested
products arising from 3-methyl-2-butenal hydrogenation itself by producing a nonzero intercept (1). Since this deac-
rather than reactant 3-methyl-2-butenal itself. Further evi- tivation occurs on the time scale of seconds, it is difficult to
dence that citral decomposition may not be responsible for properly characterize the reaction. The high initial reaction
this deactivation behavior via the production of an inhibitor rate during citral hydrogenation is associated primarily with
is based on the work of Davis and Barteau, who examined the formation of citronellal along with smaller quantities of
the effect of vinyl substituents on aldehydes and alcohols geraniol, and we propose that the subsequent decarbony-
by studying propanol, allyl alcohol, propionaldehyde, and lation of geraniol (or nerol) produces the rapid decrease in
acrolein (9–11) on Pd surfaces. They suggested that with activity to a constant value. Further evidence that geran-
vinyl substituents adjacent to alcohol and aldehyde func- iol and nerol, rather than citronellal, are involved in the
tionalities, the principal metal–organic interaction occurs deactivation process is provided by the competitive hydro-
via the oxygen moiety of the organic molecule. Further- genation of citronellal and geraniol mixtures. Citronellal
more, they noted that the new reaction pathways opened hydrogenation alone at 298 K exhibited zero-order kinet-
up by the vinyl substituents accounted for only 20% of the ics (Fig. 6a), while competitive hydrogenation of citronellal
total products. One new pathway opened for allyl alcohol, and geraniol at 298 K resulted in a continuous decrease in
but not propanol, was C–O bond scission to yield water the activity which eventually led to a nearly complete loss
and propylene, while a new pathway for acrolein, but not in activity, as seen in Fig. 8. This is further support for the
propanal, was hydrogenation of acrolein to propanal. In proposal that geraniol and nerol may be responsible for the
both cases the new reaction channels opened by the vinyl deactivation via inhibition during citral hydrogenation.
substituents yielded hydrocarbons that desorbed at tem-
The decomposition pathways for aliphatic alcohols on
peratures below those used in the present study (7). Re- Pt and Pd surfaces (10, 11, 15–17) and unsaturated alcohols
cently, Carlos de Jesus and Zaera noted that adsorption on Pd surfaces (17) have been studied and briefly addressed
of crotonaldehyde on Pt(111) yielded significant coordina- in the previous paper (1). The decomposition pathway for
==
tion via the C C bond via rehybridization; nevertheless, saturated alcohols occurs via an aldehyde intermediate to
==
the interaction via the C O bond was significant (12). In yield a surface acyl group, followed by decarbonylation to
consideration of these results, we suggest that any role in yield CO and carbonaceous species (10). In that context, it
the unusual temperature behavior due to citral decomposi- might be argued that decarbonylation of the corresponding
tion may be best understood by examining the behavior of unsaturated aldehyde (citral) should also occur. This pos-
its more aliphatic counterpart, citronellal. This is consistent sibility cannot be discounted, but in view of the previous
with the results at 373 K in the present study which show discussion, geraniol and nerol are considered to be more
that hydrogenation of geraniol and that of its more aliphatic likely candidates to produce the inhibiting surface species.
counterpart, citronellol, exhibited similar kinetic behavior,
It was previously assumed that each of the hydrogenation
i.e., a very high initial activity during the first few minutes steps in Fig. 1 was described by assuming quasi-equilibrium
of reaction followed by a sharp orders of magnitude drop in for adsorption of the reactants, competitive adsorption be-
the rate. As the following discussion illustrates, it is unlikely tween hydrogen and organic, and addition of a second H
that citronellal, and therefore citral, alone can account for atom as the rate-determining step. The general form of the
the observed activity minimum with temperature.
rate expression developed previously for each of the steps
Hydrogenation of either citronellal or geraniol individ- was (1)
ually exhibits an activity decrease as the reaction temper-
ature goes from 298 to 373 K, just as observed for citral
hydrogenation. The initial TOF for citronellal hydrogena-
kKH KOrg KOrg-H COrg PH
2
2
r =
,
[2]
2
(1 + KOrgCOrg
)
tion dropped by only one-half on going from 298 to 373 K where k is the rate constant, KOrg are the adsorption equi-
whereas the rate of geraniol disappearance at 373 K, af- librium constants for the organic reactants, and KOrg-H
ter the initial deactivation, was orders of magnitude lower is the equilibrium constant for formation of the half-
than the initial rate at 298 K. Similar behavior has been hydrogenated surface species. BOC estimates of the en-
observed by Campelo et al. during liquid-phase hydrogena- thalpy of reaction associated with addition of the first H

LIQUID-PHASE HYDROGENATION OF CITRAL OVER Pt/SiO2, II
189
==
==
atom to either a C C or a C O bond yield a similar value of the carbonyl functionality in , -unsaturated aldehydes
of 1 2 kcal/mol (18–20). Therefore, KOrg-H does not vary is difficult. Consequently, one can further conclude that at
much and the major differences in the rates among the dif- 298 K, Kcitronellal > Kgeraniol. In addition, from Fig. 10b, it
ferent compounds are expected to arise from the rate con- can be concluded that Kcitral > Kcitronellal because addition
stant k and the adsorption equilibrium constants KOrg. The of citronellal during citral hydrogenation at 298 K resulted
adsorption equilibrium constants for the reactants and in- in no change in reaction kinetics. However, the addition of
termediates of citral hydrogenation at 298 K can be ordered geraniol to a reaction mixture of citral at 298 K resulted in
as
a twofold increase in the rate of H2 uptake (Fig. 9a). The
increase in TOF for H2 uptake, as mentioned earlier, is at-
tributed to the larger rate constant for of hydrogenation of
geraniol to citronellol. Therefore, even though the surface
may be primarily covered by citral, the rate of H2 uptake
Kcitral > Kcitronellal > Kgeraniol
,
Knerol > Kcitronellol > K3,7-dimethyloctanol
based on the hydrogenation ratesofthe intermediatesalone goes up due to the more rapid hydrogenation of the small
along with competitive hydrogenation of a mixture of the amount of geraniol that does adsorb. Over time the rate
intermediates. The argument for this ordering is as fol- of H2 uptake decreases to that which would have been ob-
lows. Hydrogenation of 1 M geraniol in hexane at 298 K tained had no geraniol been introduced. This behavior is
yielded primarilycitronelloluntilthe geraniolwasdepleted, not completely understood but it may be due to a surface
at which point citronellol was further hydrogenated to 3,7- equilibration involving adsorption and decomposition of
dimethyloctanol. Such behavior suggests that at 298 K, geraniol being established on the time scale of the overall
K
geraniol >Kcitronellol >K3,7-dimethyloctanol. Thisisin accord with reaction.
what one would predict a priori. Since 3,7-dimethyloctanol The rate of citral disappearance at 373 K was not af-
has no unsaturated bonds, its adsorption equilibrium con- fected by addition of either citronellal or geraniol. This is
stant should be significantly lower than those of compounds consistent with the kinetics observed during hydrogenation
==
that contain unsaturated C C bondssuch ascitral, citronel- of citral alone at 373 K, which exhibited a zero-order de-
lal, and geraniol. Geraniol or nerol would be expected to pendence on citral concentration. Addition of citronellal
have greater absorption bond strength compared with cit- also did not affect the rate of disappearance of citral but
ronellol due to the presence of the vinyl substituent next a significant enhancement in the rate of citronellal hydro-
to the alcohol functionality, which provides additional co- genation to citronellol and isomerization to isopulegol did
ordination sites with the surface in addition to the alcohol occur (Table 2). Furthermore, the rate of disappearance of
funcitonality (9). The initial rates during competitive hy- citronellal, after its addition to a citral hydrogenation re-
drogenation of geraniol and citronellal at 298 K further in- action mixture, was almost 20-fold lower as compared with
dicate that citronellal is more strongly bound than geraniol. hydrogenation of citronellal alone at 373 K. This is strongly
Competitive hydrogenation of 0.5 M geraniol and 0.4 M cit- indicative of our earlier assertion that Kcitral > Kcitronellal
.
ronellal yielded an initial TOF for geraniol disappearance The results described in Table 2 also suggest that the rate
1
of 0.18 s 1, in contrast to a value of 1.2 s during hydro- of hydrogenation of geraniol to citronellol under citral hy-
genation of 0.5 M geraniol alone, thus representing a sixfold drogenation conditions is negligible. Addition of geraniol
decrease in the rate during competitive hydrogenation as to 1 M citral at 373 K did not affect the rate of formation/
opposed to hydrogenation of the alcohol alone. The ini- disappearance of geraniol or citronellol. However, addition
tial rate of disappearance of citronellal, on the other hand, of citronellal to an identical reaction mixture significantly
was 0.08 s 1 during competitive hydrogenation with geran- enhanced the rate of citronellal disappearance and rates
iol compared with 0.11 s 1 during hydrogenation of 0.5 M of isopulegol and citronellol formation. Therefore, hydro-
citronellal alone, which represents a decrease of only 30% genation of the unsaturated alcohol (geraniol + nerol) to
in the rate. The large decrease in the rate of geraniol disap- citronellol is kinetically insignificant during citral hydro-
pearance during competitive hydrogenation and the small genation at 373 K, and the primary route for citronellol
reduction in the rate of citronellal hydrogenation are strong formation is via citronellal hydrogenation.
evidence that the latter compound displaces geraniol from
The reaction temperature has a pronounced effect on
the catalyst surface. The higher rate for geraniol compared product distribution during hydrogenation of citronellal
with citronellal during competitive hydrogenation is due to alone. It was previously shown that the selectivity to geran-
a higher rate constant for the alcohol. This is in agreement iol during citral hydrogenation was greater at 373 K than
with results in the literature that indicate that the rate of at 298 K (1). Similarly, the results in the present study
==
==
hydrogenation ofa C C bond ismore than an order ofmag- indicate that activation of the C O bond can be accom-
==
nitude greater than that for hydrogenation of a C O bond plished much more selectively at higher reaction tempera-
==
(21). In fact, it is because activation of a C C bond is eas- tures, such as 373 K, than at 298 K where substantial activa-
== ==
ier than that of a C O bond that selective hydrogenation tion of C C bonds also occurs. This is clearly demonstrated

190
SINGH, SYSAK, AND VANNICE
by the different product distributions for citronellal hy- sumably due to steric constraints present for the cis isomer
drogenation at 298 and 373 K (Figs. 7a, 7b). At 298 K but not for the trans isomer. The rate of hydrogenation of
and a citronellal conversion of 20% , selectivities of 60% the trans isomer of the unsaturated alcohol (geraniol) is
1
1
==
for hydrogenation of the C C bond, 20% for hydrogena- 1.9 s compared with 1.5 s for the cis isomer (nerol).
== ==
tion of both C O and C C bonds, and only 20% for hy- Furthermore, the initial TOF for disappearance of E-citral
==
drogenation of the C O bond were obtained. At 373 K (trans isomer) is double that of Z-citral (cis isomer).
the only products detected were those involving selective
==
hydrogenation of the C O bond to give citronellol and
SUMMARY
isomerization of citronellal to the cyclic isopulegol. No
==
hydrogenation of the C C bond was detected. The more
The unusual activity minimum that occurs for citral hy-
drogenation as the temperature is increased from 298 to
373 K is duplicated during hydrogenation of citronellal,
geraniol, and citronellol. Decomposition of geraniol and
nerol to form adsorbed CO on Pt as an inhibitor is pro-
posed to be responsible for the activity minimum observed
during citral hydrogenation. Relative rates during hydro-
genation of the pure reaction intermediates as well as com-
petitive hydrogenation at 298 K allows the following or-
== ==
selective activation of C O bonds compared with C
C
bonds at higher reaction temperatures can be rationalized
in terms of bond dissociation energies as well as heats of
adsorption for coordination via each of the moieties. The
gas-phase bond dissociation energy of the C C bond is
approximately 141 kcal/mol compared with 174 kcal/mol
==
==
for the C O bond; thus, higher reaction temperatures
should enhance activation of the C O bond to a greater
==
extent (22). Furthermore, a calculated value of 14 kcal/mol
from BOC theory for the heat of adsorption of formal-
dehyde (20) is in close agreement with 17 kcal/mol reported
experimentally (23), as compared with 9 kcal/mol reported
for ethylene (24). Therefore, reaction at higher temper-
dering of the adsorption equilibrium constants: Kcitral
>
K
citronellal > Kgeraniol > Kcitronellol > K3,7-dimethyloctanol. The rel-
ative rates of hydrogenation of the compounds investigated
in the present study indicate that the hydrogenation rate of
==
the C C bond in this family of compounds is greater than
==
atures will favor hydrogenation of the C O bond since
==
that of the C O bond, as evidenced by the ordering of
a greater fraction of the adsorbed unsaturated aldehydes
the hydrogenation activity: geraniol > nerol > citronellol >
E-citral > citronellal > Z-citral. Furthermore, the product
distributions obtained during hydrogenation of citronellal
indicate that higher reaction temperatures favor activation
==
should be coordinated via the C O bond.
Hydrogenation of 0.5 M geraniol at 298 K yielded an ini-
tial TOF of 1.2 s 1. Simultaneous hydrogenation of 0.5 M
geraniol and 0.5 M nerol yielded a TOF of 0.5 s 1 for each of
the intermediates, representing a twofold drop as compared
with hydrogenation of the intermediate alone at an identi-
cal concentration of 0.5 M (Table 1). Such a decrease in rate
during competitive hydrogenation is expected due to com-
petitive adsorption between geraniol and nerol. However,
it is worthwhile to note that the total hydrogenation rate
of geraniol + nerol during the competitive hydrogenation
reaction is also half of that observed during hydrogena-
==
==
of the C O bond relative to the C C bond.
ACKNOWLEDGMENT
This study was supported by the DOE, Division of Basic Energy Sci-
ences, under Grant DE-FE02-84ER13276.
REFERENCES
1
tion of 1 M geraniol during which a TOF of 1.9 s was
1. Singh, U. K., and Vannice, M. A., J. Catal. 190 (2000).
2. Gallezot, P., and Richard, D., Catal. Rev. Sci. Eng. 40, 81
(1998).
3. Singh, U. K., Ph.D. thesis, Pennsylvania State University, in progress.
4. Weast, R. C., “Handbook of Chemistry and Physics,” 59th ed. CRC
Press, Boca Raton, FL, 1979.
5. Fogler, H. S., “Elements of Chemical Reaction Engineering,” 2nd ed.,
p. 62. Prentice–Hall, Englewood Cliffs, NJ, 1992.
6. Kulson, P., and Cerveny, L., Appl. Catal A 128, 13 (1995).
7. Mercadante, L., Neri, G., Milone, A., Donato, A., and Galvagno, S.,
J. Mol. Catal. 105, 93 (1996).
observed. This reduced activity during the competitive hy-
drogenation reaction is not well understood, but it may be
related to the inhibition involvingthe decarbonylation reac-
tion. Isomerization of geraniol and nerol to citronellal was
negligible under the conditions of the present study, in con-
trast to vapor-phase studies of crotonaldehyde in which sig-
nificant isomerization to butyraldehyde has been observed
(25, 26).
Based on the initial hydrogenation rate, the reactivity
of the compounds investigated in the present study can
be ordered as follows: geraniol > nerol > citronellol > E-
citral citronellal > Z-citral. As expected from literature,
8. Waghray, A., and Blackmond, D. G., J. Phys. Chem. 97, 6002
(1993).
9. Davis, J. L., and Barteau, M., J. Mol. Catal. 77, 109 (1992).
10. Davis, J. L., and Barteau, M., Surf. Sci. 235, 235 (1990).
11. Davis, J. L., and Barteau, M., Surf. Sci. 187, 387 (1987).
12. Carlos de Jesus, J., and Zaera, F., Surf. Sci., in press.
13. Campelo, J. M., Garcia, A., Luna, D., and Marinas, J. M., J. Catal. 113,
172 (1988).
==
the reactivity of the C C bond toward hydrogenation is
==
greater than that of the C O bond (21). Furthermore, it
is apparent that the trans isomer exhibits a slightly higher
hydrogenation rate as compared with the cis isomer, pre- 14. Shekhar, R., and Barteau, M., Catal. Lett. 31, 221 (1995).

LIQUID-PHASE HYDROGENATION OF CITRAL OVER Pt/SiO2, II
191
15. Rendulic, K. D., and Sexton, B. A., J. Catal. 78, 126 (1982).
16. Sexton, B. A., Rendulic, K. D., and Hughes, A. E., Surf. Sci. 121, 181
(1982).
17. Shekhar, R., and Barteau, M., Surf. Sci. 319, 298 (1994).
18. Dumesic, J. A., Rudd, D. F., Aparicio, L. M., Rekoske, J. E., and
Trevino, A. A., “The Microkinetics of Heterogeneous Catalysis,”
p. 129. Am. Chem. Soc., Washington, DC, 1993.
21. Beccat, P., Bertolini, J. C., Gauthier, Y., Massardier, J., and Ruiz, P.,
J. Catal. 126, 451 (1990).
22. Patil, A., Banares, M. A., Lei, X., Fehlner, T. P., and Wolf, E. E.,
J. Catal. 159, 458 (1996).
23. Abbas, N. M., and Madix, R. J., Appl. Surf. Sci. 7, 241 (1981).
24. Godbey, D., Zaera, F., Yeates, R., and Somorjai, G. A., Surf. Sci. 167,
150 (1986).
19. Sen, B., and Vannice, M. A., J. Catal. 78, 126 (1982).
20. Shustorovich, E., Adv. Catal. 37, 101 (1990).
25. Sen, B., and Vannice, M. A., J. Catal. 115, 65 (1989).
26. Dandekar, A. B., and Vannice, M. A., J. Catal. 183, 344 (1999).