Platinum carbonyl derived catalysts on inorganic and organic supports: A comparative study

Article abstract of DOI:10.1016/j.jorganchem.2003.10.017

Fumed silica, silica gel, silica-alumina and cross-linked (5.5%) polystyrene have been functionalized with quaternary ammonium groups and the Chini cluster [Pt₁₂(CO)₂₄]^2- has been anchored onto these functionalized materials by ion pairing. A catalyst has also been prepared by the adsorption of Na₂[Pt₁₂(CO)₂₄] on unfunctionalized fumed silica. The catalytic activities of the resultant materials, and that of commercially purchased 5% platinum on alumina have been studied for the hydrogenation of a variety of unsaturated compounds. The substrates studied are: α-acetamidocinnamic acid, cyclohexanone, acetophenone, methyl pyruvate, ethyl acetoacetate, nitrobenzene and benzonitrile. Compared to the polystyrene supported catalyst, the inorganic oxide supported catalysts have higher surface areas and for most of the substrates have notably higher activities. The functionalized fumed silica-based catalyst gives higher conversions than functionalized silica gel and silica-alumina- based catalysts. In the hydrogenation of acetophenone and ethyl acetoacetate, the functionalized fumed silica-based catalyst show superior activity compared to the commercial platinum catalyst, and the catalyst made by conventional adsorption method. In benzonitrile hydrogenation with all the cluster-derived catalysts a hydrazine derivative is selectively formed, but when the commercial platinum catalyst is used benzyl amine is the main product.

Full text of DOI:10.1016/j.jorganchem.2003.10.017

Journal of Organometallic Chemistry 689 (2004) 309–316
www.elsevier.com/locate/jorganchem
Platinum carbonyl derived catalysts on inorganic
and organic supports: a comparative study
b
b
a,
b,
*
Himadri Paul , Susmit Basu , Sumit Bhaduri *, Goutam Kumar Lahiri
a
Reliance Industries Limited, Swastik Mills Compound, V.N. Purav Marg, Chembur, Mumbai 400071, India
b
Indian Institute of Technology-Bombay, Powai, Mumbai 400076, India
Received 10 September 2003; received in revised form 15 October 2003; accepted 16 October 2003
Abstract
Fumed silica, silica gel, silica–alumina and cross-linked (5.5%) polystyrene have been functionalized with quaternary ammonium
groups and the Chini cluster [Pt12(CO)24]2ꢀ has been anchored onto these functionalized materials by ion pairing. A catalyst has also
been prepared by the adsorption of Na2[Pt12(CO)24] on unfunctionalized fumed silica. The catalytic activities of the resultant
materials, and that of commercially purchased 5% platinum on alumina have been studied for the hydrogenation of a variety of
unsaturated compounds. The substrates studied are: a-acetamidocinnamic acid, cyclohexanone, acetophenone, methyl pyruvate,
ethyl acetoacetate, nitrobenzene and benzonitrile. Compared to the polystyrene supported catalyst, the inorganic oxide supported
catalysts have higher surface areas and for most of the substrates have notably higher activities. The functionalized fumed silica-
based catalyst gives higher conversions than functionalized silica gel and silica–alumina-based catalysts. In the hydrogenation of
acetophenone and ethyl acetoacetate, the functionalized fumed silica-based catalyst show superior activity compared to the com-
mercial platinum catalyst, and the catalyst made by conventional adsorption method. In benzonitrile hydrogenation with all the
cluster-derived catalysts a hydrazine derivative is selectively formed, but when the commercial platinum catalyst is used benzyl
amine is the main product.
Ó 2003 Elsevier B.V. All rights reserved.
Keywords: Platinum carbonyl cluster; Hydrogenation; Catalysis; Polystyrene; Inorganic oxides
1. Introduction
advantages of such an approach for making catalysts
remain to be firmly established.
Anchoring of homogeneous catalysts onto insoluble
polymeric supports has been much investigated for
achieving a simple and easy method for the separation
of the catalyst [1]. Supported carbonyl clusters in addi-
tion have the potential to generate small metal crystal-
lites of novel catalytic properties [2]. Because of these
two reasons catalysts made by supporting carbonyl
clusters on inorganic or organic polymeric materials
continues to be an area of active research [1,2]. While
considerable progress has been made elucidating struc-
tural details of cluster derived particles, the practical
The physico-chemical properties of the Chini clusters,
[Pt3(CO)6]2nꢀ (n ¼ 3; 4; 5), have been much studied and
we have reviewed and reported their use both as sup-
ported and as homogeneous catalysts [3a,3b,3c]. We had
also reported the use of commercial anion exchange
resin as a support material for a number of anionic
carbonyl clusters, including the Chini clusters, many
years ago [3d]. Later on this method of ion-paring Chini
clusters on suitable organic supports was found to be of
special interest because of the ability of the resultant
materials to catalyze hydrogen driven reductions of re-
dox active biomolecules, and asymmetric hydrogenation
of methyl pyruvate [4–6]. Very recently we communi-
cated the catalytic performance of [Pt12(CO)24]2ꢀ (1),
ion-paired on functionalized fumed silica, in hydroge-
nation reactions [7].
* Corresponding authors. Tel.: +9122-25767159; fax: +9122-
25723480.
E-mail addresses: sumit_bhaduri@ril.com (S. Bhaduri), lahiri@
ether.chem.iitb.ac.in (G.K. Lahiri).
0022-328X/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2003.10.017

310
H. Paul et al. / Journal of Organometallic Chemistry 689 (2004) 309–316
The work reported here has been carried out with
resultant materials may then be exchanged with (1).
This is effected by a treatment of the functionalized
support with a methanolic solution of the sodium salt
of (1). Repeated washing of the resultant material with
methanol removes all physically adsorbed sodium salt
of (1).
The designation of the supported materials with the
degree of functionalization determined from analytical
data is given in Scheme 1 and Table 1. While samples
(2), (3), (4) and (7) have been prepared by functional-
ization of the supports followed by anion exchange,
sample (5) has been prepared by conventional adsorp-
tion method with unfunctionalized fumed silica as the
support. To obtain meaningful comparative data on
catalytic performance (see later), the amount of plati-
num loading in (5) has been maintained nearly at the
same level as that of (2).
following objectives. First, to compare the performance
of the catalyst made by ion-pairing (1) on functionalized
silica on the one hand, and functionalized divinyl ben-
zene cross-linked polystyrene on the other. Although we
have recently reported the effect of cross-linking on the
catalytic performance, studies that compare the relative
performance of organic vs. inorganic supports have not
been reported [8]. Second within the family of func-
tionalized inorganic supports, we wanted to see the effect
of changing the support on catalytic performance. Im-
portantly, the results show that (1) anchored on func-
tionalized fumed silica gives a superior catalyst than
commercial platinum on alumina, and also a catalyst
made by simple adsorption of the cluster on unfunc-
tionalized fumed silica.
Evidence for the successful incorporation of func-
tional groups in the silica network comes from analytical
and spectroscopic (Si29 NMR (MAS), IR and UV–Vis)
data. The CPMAS spectra of silica before and after
treatment with (OMe)3Si(CH2)3Cl have been reported in
the literature [9]. Representative Si29 NMR (MAS)
spectra obtained from samples used in this work are
shown in Fig. 1. The chemical shifts of signals in
Fig. 1(a) and (b) agree well with the literature reported
values of (–O–)4Si, (–O–)3Si–OH, (–O–)3Si–CH2CH2
CH2Cl and (–O–)2Si–(OMe)(CH2CH2CH2Cl). Further
functionalization by treatment with NEt3 causes the
signals due to Si atoms with Si–C bonds to be broad-
ened. This broadening is presumably due to the coupling
of silicon nuclei with quadrupolar N14(I ¼ 1). On in-
corporation of (1) by anion exchange, all the signals are
merged and only one broad signal is observed. It is
2. Results and discussion
2.1. Preparation and characterization of the precatalysts
In the work presented here fumed silica, silica gel,
silica–alumina and commercially available chlorome-
thylated 5.5% divinyl benzene cross-linked polystyrene
have been used as the inorganic and organic support,
respectively. Commercially available platinum on alu-
mina has also been used for comparative performance
evaluation with the cluster-derived catalysts. As re-
ported earlier, the chloromethylated resin may be re-
acted with triethyl amine to give quaternary
ammonium functionality. The silica-based inorganic
supports require prior treatment with trimethoxychlo-
ropropyl silane (Scheme 1). The chloride anions of the
+
Cl
Si
Si
O
N
R Cl
3
OH OH OH
Si Si Si
O
O
O
O
O
(MeO) Si
Cl
3
NR
3
Si Si
Si
Si Si Si
- MeOH
Fumed Silica/
Silica Gel/
Silica - alumina
-
2
[Pt (CO)
12
]
24
(1)
Na [Pt (CO)
12 24
]
-
2
]
+
2
+
[Pt (CO)
R
3
N
N
R
3
Si
O
12
24
Si
O
O
O
O
O
Fumed Silica (5)
Si
Si Si Si
Si Si
(2) Fumed Silica
(3) Silica Gel
Scheme 1. Functionalization of the supports and synthesis of the supported cluster-based precatalysts. Catalyst (7) ¼ (1) supported on 5.5% divinyl
benzene cross-linked polystyrene through ion pairing.

H. Paul et al. / Journal of Organometallic Chemistry 689 (2004) 309–316
311
Table 1
Designations, functionalization and platinum content of supported clusters and commercial catalyst
Functionalizationa (%)
‘‘Chloro’’ functionality
Support
Designation
Total Platinumb (wt%)
NEt3
Pt12(CO)22ꢀ4
Fumed silica
Silica gel
Silica–alumina
(2)
(3)
(4)
(5)
54
14
7
0.12
0.04
0.05
None
2.8
1.5
1.4
2.7
5
21
5
None
Fumed silica without
functionalization
Alumina
None. Prepared by conventional
adsorption method
None. Commercial sample
87
(6)
(7)
None
65
None
0.3
5
3.8
Cross-linked (5.5%)
chloromethylated polystyrene
a Degree of functionalization calculated on the basis of analytical (C, H, N, Pt) data.
b By atomic absorption.
UV–Vis. bandsthatmatchwellwiththoseof[Pt12(CO)24]2ꢀ
(mCO ꢁ 2040 and 1860 cmꢀ1, strong charge transfer UV–
Vis band at ꢁ620 nm). However, the nature of the
support has a major effect on the stability of the sup-
ported cluster. Thus for (2), (3) and (4) even when they
are stored under CO, the characteristic IR and UV–Vis.
bands of (1) disappear quickly. In contrast under iden-
tical conditions, in (7) the spectral signatures of (1) re-
main intact for more than 24 h. In sample (5), the IR or
UV–Vis. bands of (1) are not seen even with a freshly
prepared sample. IR spectra of (2), (3) and (4) also show
mC–H and mO–H bands providing additional evidence for
incorporation of functional groups, and indicating re-
tention of some hydroxyl functionalities (see later).
Representative spectra are shown in Fig. 2.
Fig. 1. Si29 NMR (MAS) of (a) fumed silica and (b) fumed silica after
treatment with trimethoxychloropropyl silane.
The loss of the spectral signatures of (1) from (2), (3)
and (4) is especially rapid if the protocol of drying the
inorganic oxide (Section 4) is not strictly adhered to. As
mentioned earlier, from the IR spectra it is apparent
that fumed silica even after drying and functionalization
has some unreacted hydroxyl functionalities and is hy-
groscopic (Fig. 2). As (1) is known to react with Hþ,
reasonable to assume that this is due to further coupling
of the silicon nuclei with Pt195(I ¼ 1=2).
Evidence of initial retention of molecular identities by
(1) on all the supports comes from the characteristic IR
and UV–Vis. spectrophotometric signatures of the car-
bonyl anions [7]. Barring (5) and (6), all the other
samples when freshly prepared have IR signals and
Fig. 2. (a) IR (KBr disc) and (b) UV–Vis spectrum (reflectance) of freshly prepared (2).

312
H. Paul et al. / Journal of Organometallic Chemistry 689 (2004) 309–316
OHꢀ, and water, the greater stability of (1) in (7) may
partly be due to the hydrophobic environment of the
polystyrene matrix.
with (1). This may be due to the fact that to be able to
ion pair with the same anionic cluster molecule, the two
quaternary ammonium groups must be positioned close
to each other.
Sodium analyses in (2), (3), (4), (5) and (7) have been
carried out by both EDAX and atomic absorption.
Repeated analyses either by atomic absorption or by
EDAX do not show the presence of detectable amounts
of sodium in (2), (3), (4) and (7). In contrast sodium is
detected in (5). As shown in Scheme 1, in the proposed
formulations of (2), (3), (4) and (7), one cluster anion is
shown to pair up with two quaternary ammonium
groups. To maintain electroneutrality the cluster dian-
ion must ion pair with two singly charged cations. As
already mentioned, the anion exchange between Clꢀ and
(1) is effected by the treatment of the functionalized
support with the sodium salt of (1). The absence of so-
dium implies that for these species, alternative formu-
lations in which one cation is the quaternary ammonium
group and the other is a sodium ion, if present at all,
would be negligible in concentration. This however is
not the case for (5). Since no ion exchange is involved in
its synthesis, the presence of sodium is expected in (5),
and is found to be present in about 1–3% of the con-
centration of platinum.
Analytical data (C, H, N and Pt) of (2), (3), (4) and
(7) at different stages of preparation of the supported
species have been measured and used for determining
the degree of functionalization. The degree of func-
tionalization is given in Table 1 and may be interpreted
as follows. In (2) out of a thousand ‘‘SiO2’’ units ap-
proximately 540 are functionalized by trimethoxychlo-
ropropyl silane. Out of these 540 units about 70 are
quaternized by triethylamine, and only two out of these
70 take part in ion pairing with (1). The degree of
functionalization is notably less when the support is
changed from fumed silica to silica gel or silica–alumina.
Thus in (3) and (4) out of a thousand ‘‘SiO2’’ approxi-
mately 140 and 210 units, respectively are functionalized
by trimethoxychloropropyl silane.
Surface area, porosity and normalized platinum
concentration on the surface, as measured by EDAX, of
samples (2), (5) and (7) are given in Table 2. Compared
to (2) the degree of functionalization and platinum
loading in (3) and (4) are much less. Also, in catalytic
experiments (see later), (3) and (4) give lower conver-
sions than (2). For these reasons these measurements
were not carried out for (3) and (4). In terms of porosity
functionalized fumed silica-based catalyst (2), and un-
functionalized fumed silica-based catalyst (5) are very
similar. However, on functionalization the surface area
of fumed silica decreases by about 40%. The specific
surface area of functionalized polystyrene-based catalyst
(7) is more than an order of magnitude less than that of
(2) and (5). In terms of porosity also (7) is notably dif-
ferent from (2) and (5). While in the silica-based cata-
lysts about 30% of the pores are within the range of
ꢀ
80–100 A, (7) does not have pores in this range and has
considerably more small pores.
For a given catalyst, the surface platinum concen-
trations of several particles at a number of different lo-
cations have been measured by EDAX. From such
measurements the ranges of surface platinum concen-
trations of (2), (7) and (5) have been calculated. For
meaningful comparison the range is defined after nor-
malizing the average surface platinum concentration to
unity for each sample.
2.2. Catalytic hydrogenation-activity, selectivity and
recycle
Samples (2)–(7) have been tested as catalysts for the
hydrogenation of nitrobenzene, the ketonic functional-
ity of a- and b-ketoesters, unfunctionalized ketones,
benzonitrile and the olefinic double bond in a-acetami-
docinnamic acid. The conversion, activity and selectivity
data are given in Table 3. We had reported earlier that
the catalytic activity of anion exchanger supported (1)
increases substantially if it is subjected to mild thermal
activation [3d]. This is expected since supported (1) is
co-ordinatively saturated and mild thermal activation is
The analytical data of (7) indicate a somewhat dif-
ferent picture. On chloromethylation about 870 units are
functionalized, and on treatment with NEt3 out of this
870 units 650 units are quaternized. However, for all the
supports, inorganic or organic, very few of the quater-
nary ammonium groups, are accessible for ion pairing
Table 2
Porosity, EDAX and surface area of (2), (5) and (7)
Surface areaa (m2/gm)
Platinum concentration
a
ꢀ
Relative volume (%) [pore ranges (A)]
Catalyst
Large
(>500)
Medium
(80–100)
Small
Normalized range on
surface (EDAX datab)
(2)
(7)
(5)
17
10
15
30
53 (20–70)
90 (10–60)
51 (20–70)
0.82–1.31
0.68–1.24
0.57–1.67
180
12
Nil
34
320
a Measured by BET.
b The range is given after normalizing the average for a given sample to 1.

H. Paul et al. / Journal of Organometallic Chemistry 689 (2004) 309–316
313
Table 3
Conversion (%) data on hydrogenation of unsaturated substrates using (2) to (7) as catalystsa
Substrate (product)
Catalyst
(2)
(3)
(4)
(5)
(6)
(7)
Nitrobenzeneb (aniline)
100
100
100
100
100
100
77
100
100
76
100
100
26
100
100
9.5
19
2
Methyl pyruvatec (methyl lactate)
Ethyl acetoacetatec (ethyl-3-hydroxybutyrate)
Acetophenonec (1-phenylethanol)
Cyclohexanonec (cyclohexanol)
a-Acetamidocinnamic acidd (rac-N-acetylphenylalanine)
Benzonitilee (Benzylamine, N,N-dibenzylhydrazine)
100
100
85
57
87
60
62
56
40
24
69
96
56
58
70
86 (8,78)
73
99 (7,82)
79
40 (40,0)
5
45 (4,41)
100 (10,90)
80 (12,68)
a All hydrogenation reactions were carried out with 1 millimole substrate and 30 mg catalyst.
b 15 bar hydrogen, 2 ml methanol, 12 h, 300 K.
c 40 bar hydrogen, 1 ml methanol, 12 h, 300 K.
d 40 bar hydrogen, 2 ml methanol, 12 h, 300 K.
e 15 bar hydrogen, 1 ml methanol, 24 h, 373 K.
necessary to remove all the CO groups. Indeed IR
monitoring had clearly established that thermal activa-
tion causes complete disappearance of mCO. In contrast,
as already mentioned (2), (3) and (4) and (5) undergo
rapid CO loss, and do not require thermal treatment for
decarbonylation. However, to maintain a uniform pro-
tocol, all catalytic experiments with (2)–(7) have been
carried out after subjecting them to mild thermal acti-
vation (Section 4). The structures of the catalytic sites
after decarbonylation and in the thermal activated ma-
terials have not been investigated.
none and ethyl acetoacetate using (2), (5), (6) and (7) are
shown in Fig. 3.
It is clear that for both the substrates the rates of
conversion are in the order (2) > (5) > (6) > (7). With
(2) the rate of hydrogenation of acetophenone is about
three, five and 16 times that of (5), (6), and (7), respec-
tively. For the hydrogenation of ethylacetoacetate it is
about two and a half, seven, and 30 times that of (5), (6)
For nitrobenzene and methyl pyruvate hydrogena-
tion, all the catalysts give full conversions. This however
is not the case for the other substrates where the inor-
ganic oxide supported catalysts show higher activities
than the polystyrene-based one. Thus for a-acetamido-
cinamic acid, cyclohexanone, ethylacetoacetate and
acetophenone hydrogenation, the conversions obtained
with (2)–(5) are about 60–100% while with (7) they are
ꢁ2–20%. The overall higher activities of (2)–(5) than
that of (7) must partly be due to the higher surface areas
of the former class of catalysts. As mentioned earlier,
among (2), (3) and (4), (2) gives higher conversions than
the other two for all the substrates. For this reason as a
representative of functionalized inorganic oxide-based
catalyst, (2) has been used in all the comparative rate
studies (see later).
To find out if the specially prepared cluster derived
catalyst (2) offers any practical advantages over the
commercially available platinum catalyst (6), or the
catalyst prepared by simple adsorption i.e., (5), their
activities may be compared. The commercial platinum
catalyst (6) chosen for this purpose was 5% platinum on
alumina (Aldrich 2000–01, Catalogue No. 31,132–4). As
can be seen from Table 3, the conversions obtained with
(2) in most cases are more than those with (5) or (6). The
superior activity of (2) becomes evident when the time
monitored conversion values of all the catalysts are
compared. Time versus conversion plots for acetophe-
Fig. 3. Rate of hydrogenation of acetophenone by (2) (functionalized
fumed silica) (j), (5) (physical adsorption on fumed silica) (d), (6)
(commercial catalyst) (.) and (7) (functionalized polystyrene) (m)
under 40 bar hydrogen pressure at 300 K. Inset shows the rate of
hydrogenation of ethylacetoacetate. Least square fitted straight lines
are drawn to highlight the differences in apparent rates.

314
H. Paul et al. / Journal of Organometallic Chemistry 689 (2004) 309–316
and (7). It may be noted that while the variation in
surface platinum distribution of (2) and (7) are similar,
total platinum content of (7) in fact is about 30% more
than that of (2) (Tables 1 and 2). The significantly higher
activity of (2) compared to that of (7) is therefore must
be due to its much higher surface area.
In view of the fact that the surface area of (5) is
considerably more than that of (2), the higher activity of
the latter is surprising. As mentioned earlier the total
platinum contents of (2) and (5) are nearly the same,
differing only by about 3%. The observed difference in
activities is therefore not a result of difference in the
total platinum content.
conditions with all the cluster-based catalysts, a hydra-
zine derivative Ph–CH2–NH–NH–CH2–Ph, character-
ized by NMR and mass-spectrometry, is the main
product.
With these catalysts the selectivity towards Ph–CH2–
NH–NH–CH2–Ph lies within 80–90%. By using
butyronitrile as an additional nitrile substrate, the gen-
erality of this reaction has been established. With (2)
and (5) as the catalysts 100% conversions with more
than 90% selectivity to C4H9–CH2–NH–NH–CH2–
C4H9 are obtained. In contrast, with (6) as the catalyst
total conversion of butyronitrile exclusively to amyl-
amine is observed.
Chemical functionalization is known to bring about
a noticeable change in the morphology of the support.
Thus by TEM several nanotubes of about 200 nm
diameter have been observed in (2) but in (5) no such
regular shapes could be found [7]. This suggests that
chemical functionalization followed by ion pairing
takes place on the relatively more exposed and easily
accessible surface sites. This is expected to result in a
more uniform platinum distribution on the surface
than what may be achieved by the adsorption method.
In the adsorption method some of the clusters are
expected to diffuse deep inside the matrix and stay
buried.
This explanation is consistent with the EDAX data
(Table 2). Although the total platinum contents of (2)
and (5) are almost identical, the variation in the surface
platinum concentration of (5) is significantly more than
that of (2). As mentioned earlier EDAX analysis of (5)
also shows that sodium is present on the surface on an
average to about 1–3% of the concentration of platinum.
This suggests that the catalytic sites, derived from so-
dium associated (1), are probably less accessible to the
reactants and less active.
As the main reason for surface anchoring of (1)
through ion pairing is to generate a catalyst that may be
recycled, successive hydrogenation reactions using re-
cycled (2) have been performed. The hydrogenation of
methyl pyruvate using recycled (2) has been carried out
10 times and no drop in activity is observed. This sug-
gests that in the catalyst prepared by the ion pairing
method, the active sites are sufficiently strongly bound so
as to resist leaching under the experimental conditions.
Finally in nitrile hydrogenation reactions, the selec-
tivities of the cluster derived catalysts (2)–(5) on the one
hand, and the commercial catalyst (6) on the other, may
be compared. There are many reports in literature on
the hydrogenation of benzonitrile and the relevant
mechanistic aspects [10]. With conventional heteroge-
neous catalyst benzyl amine is the main product but
formation of secondary and tertiary amines are also
known. With the commercial catalyst (6), total conver-
sion of benzonitrile with selectivity towards benzyl
amine >95% is observed. Surprisingly under identical
There are many cluster complexes where structure
determination by single crystal X-ray shows that the
‘‘C–N’’ moiety interacts with more than one metal atom
[11]. Since high selectivity towards hydrazine derivatives
is observed with all the cluster derived catalysts, it is
possible that similar mechanism operates for all of them,
and one of the steps involved is a multimetal activation
of the ‘‘C–N’’ moiety as has been observed for discrete
cluster complexes. However to the best of our knowl-
edge in cluster organometallic chemistry there is no ex-
ample of the coupling of two ‘‘C–N’’ units through N–N
bond formation. The selective formation of hydrazine
derivatives as reported in this work must involve such a
coupling reaction.
3. Conclusion
Ion pairing of Chini clusters on functionalized silica is
found to be a viable method for making a novel hy-
drogenation catalyst. Such a catalyst is found to be more
active than the catalyst obtained by simple adsorption of
the cluster on the support. While the precise structures
of active sites in these catalysts are not known, the ac-
tivities of both the cluster-derived catalysts are found to
be more than that of a commercial platinum catalyst. In
benzonitrile hydrogenation unlike the commercial cat-
alyst that gives exclusively benzylamine, the cluster-
derived catalysts give hydrazine derivatives.
4. Experimental
Chloroplatinic acid was purchased from Johnson
Mathey, London. 5.5% Merifield polystyrene, methyl
pyruvate, methyl lactate, benzylamine and a-acetami-
docinnamic acid were purchased from Fluka, Switzer-
land. Fumed silica, silica–alumina (AldrichÕ2000–01;
31,132–4), 1-phenylethanol, ethyl-3-hydroxy butyrate
and 5% platinum on alumina were obtained from Al-
drich, USA. Silica gel (S.D. Fine Chemicals, India,
mesh 60–120), nitrobenzene, anilines, cyclohexanone,

H. Paul et al. / Journal of Organometallic Chemistry 689 (2004) 309–316
315
cyclohexanol, benzonitrile, ethylacetoacetate, acetoph-
enone, triethylamine and other organic solvents used in
this work were procured from S.D. Fine Chemicals,
India.
linked polystyrene by triethylamine was carried out in a
similar fashion.
4.3. Synthesis of (2)
Atomic absorption measurements were carried out on
GBC 902 double beam atomic absorption spectropho-
tometer. Microanalyses were carried out with a Carlo
Erba 1106 model (Italy) C, H, N analyzer. The anionic
carbonyl cluster [Pt12(CO)24]2ꢀ was synthesized accord-
ing to the literature reported procedure [12]. Hydroge-
nation experiments were carried out in an autoclave.
Hydrogenation of all substrates except a-acetamidocin-
namic acid and nitriles were monitored by gas chro-
matography (GC) using a SC-30 (mesh 1000–2000) GC
column with FID detector. Hydrogenation of a-acet-
amidocinnamic acid was monitored by 1H-NMR and
that of benzonitrile and butyronitrile was monitored by
GC–MS. All hydrogenated products except, N,N-dib-
enzylhydrazine were initially identified by using au-
thentic commercial samples of the expected products.
N,N-Dibenzylhydrazine was identified by mass spec-
trometry and 1H-NMR. Porosity and surface area
measurements were carried out on a BET surface area
analyzer. Solid state NMR (Si29 MAS) spectra were
recorded using 300 MHz FT-NMR spectrometer. IR
spectra were taken on a Nicolet Impact 400 spectro-
photometer. Reflectance spectra were measured using a
Shimadzu UV-260 spectrophotometer. SEM (EDAX)
measurements were carried out with a Camaca model
2003 instrument. All the solvents were dried and distilled
under nitrogen prior to use. All the reactions were car-
ried out under nitrogen atmosphere unless otherwise
stated.
To a deep blue solution of Na2[Pt12(CO)24] (200 mg)
in 15 ml methanol, the triethylamine modified func-
tionalized fumed silica (1 g) was added and the mixture
was stirred for 96 h under carbon monoxide atmosphere
at room temperature (300 K). The solid material was
filtered off and washed repeatedly with methanol and
dried under a flow of CO. Synthesis of (3), (4) and (7)
were carried out in a similar fashion except for (7) only
12 h were required for ion exchange.
4.4. Synthesis of (5)
To a deep blue solution of Na2[Pt12(CO)24] (200 mg)
in 15 ml methanol, fumed silica (1 g) was added and the
mixture was stirred for 96 h under carbon monoxide
atmosphere at room temperature (300 K). The solid
material was filtered off and washed repeatedly with
methanol and dried under a flow of CO.
4.5. Thermal activation of (7)
The supported material (7) (1 g) was taken in a three
necked round bottomed flask (25 ml) equipped with
nitrogen and vacuum adapter and was flushed with ni-
trogen to remove any residual oxygen. The system was
then evacuated under vacuum at 343 K for one hour and
the color of the beads changed from green to gray. The
system was then cooled to room temperature and the
activated material stored under nitrogen. The IR spec-
trum of the activated material showed complete absence
of inorganic carbonyl bands. It was used for hydroge-
nation reactions within one hour of its synthesis. (2)–(6)
were subjected to the same treatment before using them
as catalysts.
4.1. Functionalization of fumed silica
Fumed silica was dried at 200 °C under vacuum for
24 h. Dried fumed silica (1 g) was treated with neat 3-
chloropropyltrimethoxy silane (15 ml) at 160 °C in oil
bath for 96 h. It was then filtered and washed thor-
oughly using toluene followed by methanol. On func-
tionalization initial white color of the solid support
changed to pale yellow. The same procedure was fol-
lowed for the functionalization of silica–alumina and
silica gel.
4.6. Catalytic experiments with (2)
All hydrogenation reactions were carried out with 1
mmol of the organic substrate and 30 mg of (2) in
methanol. The hydrogen pressure, temperature, reaction
time, and amount of methanol used for different sub-
strates are given in Table 3. At the end of the catalytic
run the reaction mixture was subjected to GC or GC–
MS (for the hydrogenation of benzonitrile and butyro-
nitrile) and extent of conversion was calculated on the
basis of the area ratios of starting material and product.
Catalytic experiments with (3) to (7) were carried out in
a similar fashion. For the time-monitored conversion
data the amounts of (2), (5), (6) and (7) used were ad-
justed in such a way that the total platinum content in
each case was identical.
4.2. Modification of the functionalized fumed silica by
triethylamine
The chloropropylsilane functionalized fumed silica (1
g) was mixed with 10 ml triethylamine and 10 ml tolu-
ene. The mixture was heated at reflux for 96 h. It was
then filtered and washed thoroughly with toluene fol-
lowed by methanol. The same procedure was followed
for the functionalization of silica–alumina and silica gel.
The modification of the 5.5% chloromethylated cross-

316
H. Paul et al. / Journal of Organometallic Chemistry 689 (2004) 309–316
[3] (a) S. Bhaduri, G.K. Lahiri, D. Mukesh, H. Paul, K. Sarma,
Acknowledgements
Organometallics 20 (2001) 3329, and references therein;
(b) S. Bhaduri, Curr. Sci. 78 (2001) 1318, and references therein;
(c) S. Bhaduri, P. Mathur, P. Payra, K. Sharma, J. Am. Chem.
Soc. 120 (1998) 12127;
Financial assistance from Reliance Industries Limited
for this work is gratefully acknowledged. Special ac-
knowledgement is made to Sophisticated Analytical
Instrument Facility, Indian Institute of Technology,
Bombay, for providing the NMR, SEM and TEM
facilities.
(d) S. Bhaduri, K.R. Sharma, J. Chem. Soc., Dalton Trans. (1984)
2309.
[4] S. Bhaduri, K. Sarma, J. Chem. Soc., Chem. Commun. (1996) 207.
[5] S. Bhaduri, V.S. Darshane, K. Sarma, D. Mukesh, J. Chem. Soc.,
Chem. Commun. (1992) 1738.
[6] S. Bhaduri, G.K. Lahiri, P. Munshi, D. Mukesh, Cat. Lett. 65
(2000) 61.
References
[7] H. Paul, S. Bhaduri, G.K. Lahiri, Organometallics 22 (2003) 3019.
[8] H. Paul, S. Bhaduri, G.K. Lahiri, Indian J. Chem. 42A (2003) 2392.
[9] E.J.R Sudholter, R. Huis, G.R. Hays, N.C.M. Alma, J. Colloid
Interface Sci. 103 (1985) 554.
[1] (a) P. Panster, S. Wieland, in: B. Cornils, W.A. Herrmann (Eds.),
Applied Homogeneous Catalysis with Organometallic Com-
pounds, vol. 2, VCH, Weinheim, 2002, p. 646;
(b) U. Nagel, J. Albrecht, Top. Catal. 5 (1998) 3;
(c) B.C. Gates, Chem. Rev. 95 (1995) 511.
[10] S. Nishimura, in: Handbook of Heterogeneous Catalytic Hydro-
genation for Organic Synthesis, Wiley, New York, 2001, pp. 254–
284.
[11] (a) M.A. Andrews, H.D. Kaesz, J. Am. Chem. Soc. 99 (1977)
6763;
[2] (a) J.M. Thomas, B.F.G. Johnson, R. Raja, G. Sankar, P.A.
Midgley, Acc. Chem. Res. 36 (2003) 20;
(b) M.A. Andrews, C.B. Knobler, H.D. Kaesz, J. Am. Chem. Soc.
101 (1979) 7260;
(b) A.M. Argo, J.F. Odzak, F.S. Lai, B.C. Gates, Nature 415
(2002) 6872;
(c) M.A. Andrews, H.D. Kaesz, J. Am. Chem. Soc. 101 (1979)
7238;
(c) W. Zhou, J.M. Thomas, D.S. Shephard, B.F.G. Johnson, D.
Ozkaya, T. Maschmeyer, R.G. Bell, Q. Ge, Science 280 (1998)
705;
(d) R.D. Adams, D.A. Katahira, L.W. Yang, J. Organomet.
Chem. 219 (1981) 85;
(d) A.M. Argo, J.F. Goellner, B.L. Phillips, G.A. Panjabi, B.C.
Gates, J. Am. Chem. Soc. 123 (2001) 2275;
(e) R.D. Adams, Acc. Chem. Res. 16 (1983) 67;
(f) Z. Dawoodi, M.J. Mays, K. Henrick, J. Chem. Soc., Dalton
Trans. (1984) 433.
(e) A. Fukoka, N. Higashimoto, Y. Sakamoto, M. Sasaki, N.
Sugimoto, S. Inagaki, Y. Fukushima, M. Ichikwa, Catal. Today
66 (2001) 23.
[12] G. Longoni, P. Chini, J. Am. Chem. Soc. 98 (1976) 7225.