Effect of particle size on selective hydrogenation of cinnamaldehyde by Pt encapsulated in mesoporous silica

Article abstract of DOI:10.1016/j.catcom.2012.08.017

Pt nanoparticles of various sizes, viz. 8, 4.9, 3.6 and 1.8 nm were encapsulated in 2D hexagonal mesoporous silica by in-situ synthesis as well as post synthetic modifications so that the final catalyst composition was ≤ 1 wt.% Pt/SiO₂. A kinetic analysis of the effect of particle size on selective hydrogenation of cinnamaldehyde was carried out on these catalysts. It was found that the materials, even at such low loading of Pt, were very active for the hydrogenation and selective for the desired product, cinnamyl alcohol. Among the different particle sizes, selectivity was found to be the highest on 8 nm particles. Kinetic analysis shows that the reaction follows a consecutive reaction pathway.

Full text of DOI:10.1016/j.catcom.2012.08.017

Catalysis Communications 28 (2012) 42–46
Contents lists available at SciVerse ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Effect of particle size on selective hydrogenation of cinnamaldehyde by Pt
encapsulated in mesoporous silica
Atul K. Prashar a, S. Mayadevi b, R. Nandini Devi a,
⁎
a
Catalysis and Inorganic Chemistry Division, National Chemical Laboratory, Dr. Homi Bhabha Road, Pune 411008, India
Chemical Engineering and Process Development Division, National Chemical Laboratory, Dr. Homi Bhabha Road, Pune 411008, India
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 July 2012
Received in revised form 11 August 2012
Accepted 13 August 2012
Available online 21 August 2012
Pt nanoparticles of various sizes, viz. 8, 4.9, 3.6 and 1.8 nm were encapsulated in 2D hexagonal mesoporous
silica by in-situ synthesis as well as post synthetic modifications so that the final catalyst composition
was ≤1 wt.% Pt/SiO2. A kinetic analysis of the effect of particle size on selective hydrogenation of
cinnamaldehyde was carried out on these catalysts. It was found that the materials, even at such low
loading of Pt, were very active for the hydrogenation and selective for the desired product, cinnamyl
alcohol. Among the different particle sizes, selectivity was found to be the highest on 8 nm particles.
Kinetic analysis shows that the reaction follows a consecutive reaction pathway.
Keywords:
Cinnamaldehyde hydrogenation
Selective hydrogenation
α,β-Unsaturated aldehydes
Mesoporous silica
© 2012 Elsevier B.V. All rights reserved.
Pt nanoparticles
Kinetic analysis
1. Introduction
changes with variations in particle size, we expect a reflection of this on
the selectivity of various products. Hence it is imperative that the active
Oxide supported noble metal nanoparticles have long played a
pivotal role in industrial catalysis [1,2]. The unique advantage of
such catalysts is that activity and selectivity can be fine tuned by con-
trolling the particle size and shape since active centres in metal
nanoparticles are present on the surface and the surface features of
metal nanoparticles change with shape and size [3,4]. This variation
in catalytic activity with changing particle size and shape is because
of the various molecular scale factors which include changes in sur-
face structure, electronic state, metal–support interaction and oxida-
tion states, especially in the range of 1–10 nm [5,6].
Selective catalytic hydrogenation of organic substrates containing a
number of unsaturated functional groups is such a process which shows
marked influence by particle size variations. Moreover, this is an impor-
tant step in industrial production of fine chemicals, especially the selec-
tive hydrogenation of α,β-unsaturated carbonyls like cinnamaldehyde,
which is crucial in the manufacture of pharmaceuticals, flavours and fra-
grances [7,8]. The desired reaction viz., chemoselective hydrogenation of
carbonyl bond in multi-unsaturated aldehydes and ketones is a difficult
task since thermodynamics favours C_C hydrogenation over C_O by
ca. 35 kJ/mol [9]. It is reported that adsorption energies of C_C and
C_O bonds on various crystal planes of metal particles are different
and this can be exploited for improving the selectivity of the catalyst
[10,11]. Since nature of crystal planes exposed on the nanoparticle surface
nanoparticles are as monodisperse as possible and with least tendency
for agglomeration and change in particle size during reactions. Conven-
tional methods of catalyst preparations have the drawback of non uni-
form particle sizes which severely affects the selectivity to desired
product. We have recently developed a simple in-situ method for encap-
sulating monodisperse nanoparticles within the channels of mesoporous
silica [12,13]. Such catalysts with uniform sized Pt nanoparticles isolated
exclusively within the channels of mesoporous compounds can be
considered to be good candidates for size sensitive reactions due to
monodispersity as well as stability. We have selected a series of catalysts
with Pt nanoparticles of varying sizes encapsulated in mesoporous silica
to study the effect of particle size on selective hydrogenation. In this
paper, we report the hydrogenation of cinnamaldehyde on Pt nano-
particles of sizes 8, 4.9, 3.6 and 1.8 nm encapsulated in 2D hexagonal
mesoporous silica and kinetic analysis to understand the reaction
pathways.
2. Experimental
2.1. Synthesis of Pt/mSiO2
Pt particles with average sizes of 1.8 and 3.6 nm were synthesised
by alcohol reduction method and incorporated in the channels of
mesoporous SBA-15 by sonication as reported elsewhere [14]. In
situ method as reported by us [12,13] was followed for the synthesis
of Pt with average sizes of 8 and 4.9 nm encapsulated in mesoporous
⁎
Corresponding author. Tel.: +91 20 25902271; fax: +91 20 25902063.
E-mail address: nr.devi@ncl.res.in (R. Nandini Devi).
1566-7367/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.catcom.2012.08.017

A.K. Prashar et al. / Catalysis Communications 28 (2012) 42–46
43
silica (Supplementary material). Pt wt.% of the catalysts with 1.8, 3.6,
4.9 and 8 nm Pt particles were found to be 0.96, 1, 0.72 and 0.7 wt.%
of Pt respectively, by elemental analysis.
was 0.001 (molar) and CAL/isopropanol ratio was 0.03 (by wt.) for
all the catalytic reactions. Catalysts were pre-activated at 100 °C for
2 h in 35 mL isopropanol under 34 bar hydrogen pressure (RT) be-
fore the reaction. 1.25 g of cinnamaldehyde in 40 mL of isopropanol
was charged in the reactor and heated to 60 °C under a H2 pressure
of 30 bar at RT. At intervals, aliquots of the reaction mixture were
taken from the reactor and GC analysis was performed on Agilent
Technologies 6890N Network GC system using an HP5 column.
2.2. Cinnamaldehyde hydrogenation
A 100 mL Parr autoclave equipped with a pressure gauge, stirrer
and thermocouple was used as the reaction vessel. Pt/substrate ratio
100
Cinnamaldehyde Conversion
90
80
70
60
50
40
30
20
10
Cinnamyl Alcohol
Hydrocinnamaldehyde
Phenyl Propanol
Pt/SBA-15 (1.8 nm)
Pt/SBA-15 (3.6 nm)
0
100
90
80
70
60
50
40
30
20
10
0
Pt/SBA-12 (4.9 nm)
Pt/SBA-15 (8 nm)
0
2
4
6
8
10
0
2
4
6
8
10
Time (h)
Time (h)
90
80
70
60
50
40
30
20
10
Cinnamyl Alcohol
Hydrocinnamaldehyde
Phenyl Propanol
Pt/SBA-15 (1.8 nm)
Pt/SBA-15 (3.6 nm)
0
90
Pt/SBA-12 (4.9 nm)
Pt/SBA-15 (8 nm)
80
70
60
50
40
30
20
10
0
40
50
60
70
80
90
100
40
50
60
70
80
90
100
Conversion
Conversion
Fig. 1. Cinnamaldehyde hydrogenation on Pt/mSiO2 catalysts with varying Pt particle sizes; top: conversion and product yield (%) vs. time, bottom: product selectivity vs. conver-
sion for 1.8, 3.6, 4.9 and 8 nm (■ CA; ● HCAL; ▲ PP; ▼ CAL).

44
A.K. Prashar et al. / Catalysis Communications 28 (2012) 42–46
3. Results and discussion
The above results and analysis of the graphs indicate that PP is
predominantly formed from HCAL and not at the expense of CA.
Hence the following scheme (Scheme 1) was considered for further
kinetic analysis. This can be considered as a set of parallel reactions
in which one of the steps (CAL to HCAL) is a series reaction which
continues further to form PP. As hydrogen used is in excess, the set
of reactions was considered as pseudo-first order reactions. Mass
transfer effects were not considered since stirring rate was adjusted
so that such effects were negligible. Moreover, adsorption was as-
sumed to be the rate limiting step and the following pathways were
adopted.
We have recently reported a novel in-situ method for encapsulation
of Pt nanoparticles exclusively within the channels of mesoporous silica
[12,13]. We have synthesised Pt nanoparticles of sizes 8 and 4.9 nm by
employing two different polymer templates, P123 and Brij-76 modified
with cationic surfactant CTAB for better metal precursor anchoring. The
silica channels act as moulds during the formation of the nanoparticles
and also prevent further agglomeration during reactions. We also syn-
thesised two more catalysts with Pt particles of sizes 1.8 and 3.6 nm
by encapsulating already formed nanoparticles in 2D hexagonal meso-
porous silica following the procedures reported elsewhere [14]. In
both of these types of catalysts, the particles are reasonably monodis-
perse and hence ideal for selective reactions dependent on particle sizes.
Selective cinnamaldehyde (CAL) hydrogenation was carried out on
the above four catalysts and a kinetic analysis was done based on the re-
sults. In the given reaction conditions, three products i.e. cinnamyl alco-
hol (CA), hydrocinnamaldehyde (HCAL), and phenyl propanol (PP) are
expected by hydrogenation of aldehyde, alkene and both aldehyde and
alkene groups respectively [10]. Hydrogenation was effected by molec-
ular hydrogen and transfer hydrogenation could be ruled out since no
such product peaks could be identified. In the case of all the four cata-
lysts, CA was obtained as the major product. Distribution of products
of the reaction with time and variation of product yield/selectivity
with conversion for all the catalysts are given in Fig. 1. After 1 h, 21%
yield of CA is obtained at 40% conversion for Pt/mSiO2(1.8 nm). Yield in-
creases with conversion up to a maximum of 52% after 10 h at 93% con-
version. HCAL yield increases up to 18% at 5 h at 74% conversion after
which it starts decreasing. Amount of PP increases as reaction proceeds
and at the end of the reaction i.e. at 10 h, it reaches 27%. Similar trends
were observed for the other three catalysts. PP amount is found to in-
crease steadily at first and as the reaction proceeds, it seems to increase
in tandem with a decrease in HCAL formation. This indicates that HCAL
gets hydrogenated with time as its concentration increases to give PP.
This points to a set of parallel and consecutive reactions. From these ob-
servations it is clear that the yield of CA increases with increase in par-
ticle size and follow the order 8 nm>4.9 nm>3.6 nm >1.8 nm. Since
the support, i.e. 2D mesoporous silica, is the same for all the four cata-
lysts this effect can be attributed solely to the particle size. It is notewor-
thy here that we could achieve very high selectivity of the order of 80%
in Pt/mSiO2(8 nm) catalyst with low loading of Pt whereas in conven-
tional catalysts higher loading is necessary for obtaining bigger nano-
particles. Selectivities of CA at 50% and 75% conversions on various
catalysts are given in Table 1.
3.1. Step-1 parallel reaction
At any time, the total concentration of HCAL formed consists of the
concentrations of HCAL and PP. Hence, it is possible to take the total
HCAL formed is equal to the moles of HCAL+moles of PP.
CHCALðTotalÞ ¼ CHCAL þ CPP
ð1Þ
Selectivity f or HCAL; s
¼
HCALðTotalÞ=CA
¼
ðCHCAL CPPÞ=CCA
þ
¼
k1=k3
Here CHCAL, CPP and CCA are the concentrations in moles of hydro-
cinnamaldehyde, phenyl propanol and cinnamyl alcohol, respectively.
The rate of disappearance of CA, according to first order parallel
reaction, can be written as,
−dðCCALÞ=dt
¼
k1CCAL
þ
k3CCAL
¼ ðk1 þ k3ÞCCAL
which, on integration gives
lnð1−XCALÞ ¼ ðk1 þ k3Þt:
ð2Þ
Here CCAL is the concentration of cinnamaldehyde, XCAL, its conver-
sion and t, the reaction time. (k1+k3) was obtained from the slope of
the linear plot of — ln(1−XCAL) versus t, passing through the origin.
k1/k3 was calculated from the average selectivity (Eq. (1)) obtained
from the experimental data. These values were used for determining
k1 and k3.
3.2. Step-2 series reaction
The rate constant k2 of the series reaction (formation of PP from
HCAL) was obtained from the standard equation for the concentra-
tion of the intermediate compound in a series reaction, which is
given by
As clear from the table, Pt/mSiO2 (1.8 nm) with the smallest par-
ticle size gives minimum selectivity while Pt/mSiO2 (8 nm) with larg-
est particle size gives maximum selectivity. Apparently, C_O bond is
hydrogenated faster than C_C bond, yielding more CA. Although
HCAL is also obtained as a result of hydrogenation of C_C bond ini-
tially, C_O bonds get further hydrogenated to give PP which is the
final product. Increase in selectivity of CA from 57% to 80% as particle
size increases from 1.8 nm to 8 nm can be attributed to change in the
extent of exposure of different crystallographic planes with increas-
ing particle size; [111] planes being more sterically hindering for
C_C approach to the surface [15–17] .
!
e−k1t
k2−k1 k2−k1
e−k2t
CHCAL ¼ CCAL0k1
−
or
CHCAL
CCAL0k1
e−k1t−e−k2t
k2−k1
¼
:
ð3Þ
CA
k3
Table 1
Selectivity of cinnamyl alcohol at 50% (S50) and 75% (S75) conversions on various
catalysts.
CAL
k1
Catalyst
Av. Pt particle size (nm)
S60
S75
Pt/mSiO2 (1.8 nm)
Pt/mSiO2 (3.6 nm)
Pt/mSiO2 (4.9 nm)
Pt/mSiO2 (8 nm)
1.8
3.6
4.9
8
50
68
73
87
57
67
68
79
k2
HCAL
PP
Scheme 1. Scheme for kinetic analysis of cinnamaldehyde hydrogenation.

A.K. Prashar et al. / Catalysis Communications 28 (2012) 42–46
45
higher concentration of CA in the product is obtained in the experi-
ments even though formation of PP is the most favoured reaction ki-
netically. This indicates that adsorption through C_O is the most
favoured geometry. However, HCAL formation being the most slug-
gish prevents higher yields of PP leading to selective hydrogenation
to CA. Parity plots between the experimental concentrations and the
predicted concentrations (CCAL/CCAL0 vs. CCAL′/CCAL0, CCHAL/CCAL0
vs. CCHAL′/CCAL0, CPP/CCAL0 vs. CPP′/CCAL0 and CCAL/CCAL0 vs.
CCAL′/CCAL0) are presented in Fig. 3. Only points in the linear in-
crease region of the yield vs conversion graphs were used in these
calculations. The figure shows that the points cluster around the Y =
X trend line indicating that the prediction matches reasonably well
with that of the experimental values, confirming the validity of the re-
action path way.
4. Conclusion
Fig. 2. Variation in rate constants for the series and parallel steps with the particle size.
Pt particles of varying sizes are isolated within the channels of
mesoporous silica compounds. Cinnamaldehyde hydrogenation showed
sensitivity to Pt particle sizes and the selectivity to cinnamyl alcohol in-
creased with increasing particle sizes. Cinnamyl alcohol was found to
be the major product with hydrocinnamaldehyde further forming fully
hydrogenated product phenyl propanol. A kinetic analysis based on a
set of parallel and consecutive reaction pathways revealed that the
rate of phenyl propanol formation is the highest and most sensitive
to Pt particle size. As particle size increases, complete hydrogenation
of hydrocinnamaldehyde is preferred; however, the sluggishness of
Eq. (3) was used for the estimation of k2 by trial and error method.
The values of k1, k2 and k3 obtained for different catalyst sizes are
given in the Supplementary material. Fig. 2. represents the variation
in rate constants with particle size.
It reveals that the rate constant for the formation of PP is higher
and more sensitive to the catalyst size than those of the other two
products. However, the formation of PP is restricted by the formation
of the intermediate HCAL. The rate constant for the formation of HCAL
is the lowest and decreases with increase in particle size. Hence a
Fig. 3. Plots of experimental versus theoretical concentrations (molar ratio) for the hydrogenation of cinnamaldehyde over Pt-mSiO2 catalysts with Pt nanoparticles of different
sizes at 60 °C and 30 bar H2 pressure.

46
A.K. Prashar et al. / Catalysis Communications 28 (2012) 42–46
[2] J.A. Anderson, M.F. Garcia, Supported Metals in Catalysis, Imperial College Press,
London, 2005.
[3] G.C. Bond, Chemical Society Reviews 20 (1991) 441–475.
[4] R.A. van Santen, Accounts of Chemical Research 42 (2009) 57–66.
[5] I. Lee, F. Delbecq, R. Morales, M.A. Albiter, F. Zaer, Nature Materials 8 (2009)
132–138.
hydrocinnamaldehyde formation leads to the enhanced yield of
cinnamyl alcohol.
Acknowledgement
[6] G.A. Somorjai, A.M. Contreras, M. Montano, R.M. Rioux, Proceedings of the National
Academy of Sciences of the United States of America 103 (2006) 10577–10583.
[7] K. Weissermel, H.J. Arpe, Industrial Organic Chemistry, Verlag Chemie, Weinheim,
1978.
[8] K. Bauer, D. Garbe, in: Ullman Encyclopedia, AII, VCH, New York, 1988, p. 141.
[9] C. Mohr, P. Claus, Science Progress 84 (2001) 311–334.
[10] F. Delbecq, P.J. Sautet, Journal of Catalysis 152 (1995) 217–236.
[11] P. Gallezot, D. Richard, Catalysis Reviews: Science and Engineering 40 (1998)
81–126.
Authors thank CSIR, India for funding.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://
dx.doi.org/10.1016/j.catcom.2012.08.017.
[12] A.K. Prashar, R.P. Hodgkins, R. Kumar, R.N. Devi, Journal of Materials Chemistry 18
(2008) 1765–1770.
[13] A.K. Prashar, S. Mayadevi, P.R. Rajamohanan, R.N. Devi, Applied Catalysis A: Gen-
eral 403 (2011) 91–97.
[14] R.M. Roux, H. Song, J.D. Hoefelmeyer, P. Yang, G.A. Somorjai, The Journal of Phys-
ical Chemistry. B 109 (2005) 2192–2202.
[15] M. Englisch, A. Jentys, J.A. Lercher, Journal of Catalysis 166 (1997) 25–35.
[16] R.I. Masel, Principles of Adsorption and Reaction on Solid Surfaces, Wiley, New
York, 1996.
References
[1] (a) G.C. Bond, Catalysis by Metals, Academic Press, London, 1962.;
(b) J. Sun, X. Bao, Chemistry - A European Journal 14 (2008) 7478–7488;
(c) R.J. White, R. Luque, V.L. Budarin, J.H. Clark, D.J. Macquarrie, Chemical Society
Reviews 38 (2009) 481–494;
(d) J.M. Campelo, D. Luna, F. Luque, J.M. Marinas, A.A. Romero, ChemSusChem 2
(2009) 18–45;
[17] C. Minot, P. Gallezot, Journal of Catalysis 123 (1990) 341–348.
(e) L. De Rogatis, M. Cargnello, V. Gombac, B. Lorenzut, T. Montini, P. Fornasiero,
ChemSusChem 3 (2010) 24–42.