Loading silica with metals (palladium or platinum) under mild conditions by using well-defined molecular precursors

Article abstract of DOI:10.1039/b903655e

The loading of pre-treated amorphous silica with platinum or palladium was carried out by using the molecular precursors Pt₂(μ-Cl) ₂Cl₂(CO)₂, or Pd₂(μ-Cl) ₂Cl₂(CO)₂, r

Full text of DOI:10.1039/b903655e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Loading silica with metals (palladium or platinum) under mild conditions
by using well-defined molecular precursors†
a
b
c
b
c
Lidia Armelao,* Daniela Belli Dell’Amico,* Roberto Braglia, Fausto Calderazzo, Fabio Garbassi,
c
b
Gianluigi Marra and Alessandra Merigo‡
Received 23rd February 2009, Accepted 18th May 2009
First published as an Advance Article on the web 8th June 2009
DOI: 10.1039/b903655e
The loading of pre-treated amorphous silica with platinum or palladium was carried out by using the
molecular precursors Pt
required amount of coordinated CO to carry out the formation of the metal particles upon contact with
moisture. The reactivity of the well-soluble mononuclear platinum complex cis-PtCl (CO) with
stoichiometric amounts of water was investigated either under N or CO. The metal nanoparticles
2
(m-Cl)
2
Cl
2
(CO)
2
, or Pd
2
(m-Cl)
2
Cl
2
(CO)
2
, respectively, which contain the
2
2
2
produced on the silica matrix have been characterized by X-ray photoelectron spectroscopy (XPS),
X-ray diffraction (XRD) and transmission electron microscopy (TEM). The catalytic performance of
the silica-supported metals thus produced was evaluated in the hydrogenation of cyclohexene.
Introduction
being converted to CO . The presence in the dinuclear precursors
2
of only one CO group per metal atom and, consequently, of two
reducing equivalents, was considered to be ideal for converting
the central metal atom to the zerovalent state, i.e. to M(s). A
further advantage in the use of the chloro-carbonyl derivatives of
palladium and platinum is that their reaction with moisture occurs
at room temperature or even lower. This feature was anticipated
to limit aggregation of the resulting metal particles, normally
favoured at higher temperatures.
The formation of metal nanoparticles is an expanding area of
research, in connection with: (a), the use of well-defined molecular
precursors; (b), the understanding of the aggregation phenomena;
1
(c) the application to several areas of surface science. When the
precursor contains a metal characterized by a positive standard
reducing potential, its prompt reduction (by heat or by exposure
to H
2
, or to CO/H O) is to be expected, with formation of metal
2
nanoparticles, whose dimensions depend, among other factors, on
Interest in this study also comes from comparing chemical
properties within metals characterized by a decreasing metal–
metal bond strength, as evaluated on the basis of the corre-
2
the temperature. In this regard, the observation was made that a
higher catalytic activity in the hydrogenation of several substrates
was found for catalysts obtained under milder reducing conditions.
The preparation of supported noble metal nanoparticles and the
study of their efficiency in hydrogenation catalysis is a topic of
-1
sponding enthalpies of formation of gaseous atoms (kJ mol ): Pt,
8
5
65.7; Pd, 376.6. As a test of the Pd/Pt comparison we decided
to investigate the catalytic performance of the platinum-loaded
silica with the corresponding palladium-containing system. A final
remark concerns the use of amorphous silica of a relatively high
surface area as a support.
3
current interest.
In earlier reports from these laboratories, palladium, platinum
and gold have been supported on silica by using the correspond-
4
ing N,N -dialkylcarbamato complexes as precursors. Also the
mononuclear chlorocarbonyls AuCl(CO) and cis-PtCl
2
(CO)
2
were
4c,5
used for the same purpose.
obtained with Pt (m-Cl) Cl (CO)
the precursors are similar from a molecular viewpoint (both are
This paper presents the results
6
,7
Results and discussion
2
2
2
2
and Pd
2
(m-Cl)
2
Cl
2
(CO)
2
.
As
Reaction of the molecular precursors with water
6b
chloride-bridging species), the Pd/Pt comparison was antici-
pated to be particularly interesting. These dinuclear derivatives
were expected to bind to the oxygen atoms on the silica surface by
splitting of the chloride bridges, with subsequent attack by water
thus reducing the platinum(II) and palladium(II) centres, CO
As explained in the Introduction, this study deals with loading
the silica surface with platinum and palladium by using the
dinuclear M (m-Cl) Cl (CO) (M = Pd, Pt), which are planar
2
2
2
2
6b
chloride-bridged species.
Taking into consideration that even dehydrated silica, see
Experimental, can be a source of water through condensation
of the surface silanol groups, it was of interest to know the
outcome of the reaction of the chlorocarbonyl derivatives of these
metals with a stoichiometric, or slightly higher, amount of water.
a
ISTM-CNR and INSTM–Dipartimento di Scienze Chimiche–
Universit a` di Padova–Via Marzolo, 1–35131, Padova, Italy. E-mail:
lidia.armelao@unipd.it
b
Dipartimento di Chimica e Chimica Industriale, Universit a` di Pisa, Via
Risorgimento 35, I-56126, Pisa, Italy. E-mail: belli@dcci.unipi.it
7b
Concerning the palladium derivative, we had previously noted
c
Istituto ENI Donegani, Via G. Fauser 4, I-28100, Novara, Italy
that Pd
formation of [PdCl(CO)]
product results when the PdCl
(H O/Pd molar ratios in the range 1–3). Recently, the assistance
2
(m-Cl)
2
Cl
2
(CO)
2
is reduced by CO in acetic anhydride with
†
This paper is dedicated to the memory of the late Prof. Carlo Carlini,
9
n
and later we have verified that the same
Universit a` di Pisa.
/CO system is contacted with water
‡
Present address: Lucchini L. R. S., Viale della Resistenza 2, I-57025
2
Piombino, Livorno, Italy
2
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5559–5566 | 5559

View Article Online
of water in the formation of [PdCl(CO)]
n
from PdCl
2
under CO
after being recovered by filtration, was subjected to a treatment
with water vapour at 40 C in order to carry out the exhaustive
reduction to zerovalent platinum, as for eqn (4).
10
◦
has been reported by another group.
The reduction of cis-PtCl (CO) in the presence of an excess of
2
2
11
water has been reported by Chatt and co-workers, the reaction
being carried out under carbon monoxide in ethanol as medium.
Pt
2
Cl
4
(CO)
2
+ 2H
2
O→2CO
2
+ 4HCl + 2Pt(s)
(4)
In the present study, the reaction of cis-PtCl
Pt (m-Cl) Cl (CO) in view of its higher solubility in hydrocarbons]
with 1, 2 and 3 equivalents of water was carried out both under
and under CO. Ethanol was excluded as solvent because of its
2
(CO)
2
[preferred to
The platinum-loaded silica samples, before (AM2-111) and after
(AM2-122) hydrolysis, have been characterized by XRD, XPS and
TEM analysis.
2
2
2
2
14
N
2
XRD studies have shown that metallic crystalline phases are
already present in the samples obtained by contacting the oxide
matrix with the toluene solution of the platinum precursor
(see Fig. 1). They are presumably formed by contacting water,
coming from the condensation of surface silanols or fortuitously
encountered in the course of the sample preparation. The broad
possible interference in the reduction process.
Under an atmosphere of pre-purified dinitrogen and at room
temperature, a toluene solution of the mononuclear carbonyl
derivative of platinum(II) undergoes reduction in the presence of
dissolved water, ligated carbon monoxide being partially converted
to carbon dioxide, see eqn (1). Disappearance of the typical
◦
peak centred at 2q ª 39.7 was indexed as the (111) reflection
carbonyl bands of cis-PtCl
2
(CO)
2
was observed at a H
(60% yield) being
obtained (nCO, 2060 cm ) which was recovered by filtration.
2
O/Pt molar
of the fcc structure of metallic platinum (JCPDS Card No.
◦
ratio of 3, a black solid analysing as [PtCl(CO)]
n
4-802,2000). Treatment with water vapour at 40 C promoted
-
1
further crystallization, as evidenced by the moderate sharpening
of the (111) peak and the slight intensity increase of the weak
2
cis-PtCl
2
(CO)
2
+ H
2
O→2/n[PtCl(CO)]
n
+ 2 HCl
◦
(200) platinum reflection (2q ª 46 ). The average crystallite size
+
CO + CO
2
(1)
was estimated, by using the Scherrer equation, to be ª 6 nm for
both samples, although it can be over-estimated, as only the larger
particles contribute to the diffraction.
The platinum-containing product of reaction (1), after drying in
vacuo, was identified through its reaction with [NBu ]Cl in CH Cl
yielding the already known compound [NBu
4
2
2
12
4
]
2
[Pt
2
Cl
4
(CO)
2
] in
good yields (≥70%), see eqn (2).
2
/n[PtCl(CO)]
n
+ 2[NBu
4
]Cl→[NBu
4
]
2
[Pt
2
Cl
4
(CO)
2
]
(2)
The formation of platinum(I) under the conditions of reaction
1) was therefore established. As [PtCl(CO)] is substantially
(
n
insoluble in the reaction medium, its further reduction to plat-
inum(0) presumably becomes unfavourable. Furthermore, the HCl
produced in the reaction interacts with H O, thus reducing its
2
activity in solution, and an excess of water is consequently required
to perform reaction (1).
Under CO at room temperature, the reaction between cis-
PtCl
ratio of 2.5 leading to the formation of the violet-black [Pt(CO)
≥97% yield). The CO-promoted reduction of platinum(II) can
therefore be represented by reaction (3).
2
(CO)
2
and water goes to completion with a H O/Pt molar
2
11
2 n
]
(
Fig. 1 XRD patterns of Pt-loaded silica samples before (AM2-111) and
after (AM2-122) exposure to water vapour.
cis-PtCl
2
(CO)
2
+ H
2
O + CO→1/n[Pt(CO)
2
]
n
+ 2HCl + CO
2
(3)
Further information on the chemical composition of both
samples was obtained by XPS analysis: the platinum content
was about 0.08 at.%, in agreement with the value of 0.76 wt%
obtained by ICP measurements (see Experimental section). In
Fig. 2, the XPS high resolution spectra of the Pt4f regions for
samples A (AM2-111) and B (AM2-122) clearly show the 4f 7/2
and 4f 5/2 spin–orbit splitting components. Both spectra show a
significant line broadening suggesting the presence of platinum
centres characterized by different oxidation states and/or chem-
ical environments. Peak deconvolution resulted in two doublets
centred at (BE Pt4f 7/2) ª71.5 eV and ª73.0 eV, assigned to Pt(0)
In separate experiments we observed that [PtCl(CO)]
n
reacts
in toluene with excess water: (a) under N with production of a
solid containing 95% of platinum; (b) under CO with formation
of [Pt(CO)
It is noteworthy that [Pt(CO)
loses CO (in vacuo or by thermal treatment) to finally yield
2
2 n
] .
2 n
] is not stable and irreversibly
13
platinum metal.
Loading of silica
When dehydrated silica was treated under a dinitrogen atmosphere
15
with a solution of the orange Pt (m-Cl) Cl (CO) at room temper-
ature, the platinum-containing fragments migrate from solution to
the SiO surface, as evidenced by the colour of the silica turning to
2
2
2
2
and Pt(II) species, respectively. The most remarkable difference
between the samples concerns the amount of metallic platinum,
that can be evaluated by comparing the intensity of the Pt(0) and
Pt(II) components.
2
orange. We propose adsorption of platinum(II) to be the primary
process, followed by grafting as a consequence of chloride-bridge
While the Pt(0)/Pt(II) ratio is ca. 1 in the as-prepared Pt-SiO
sample (A), it increases to ca. 2.2 after hydrolysis. Therefore,
contacting the toluene solution of Pt (m-Cl) Cl (CO) with the
2
splitting and formation of PtCl
2
(CO) fragments bonded to the
oxygen atoms of the silica surface. The platinum-grafted silica,
2
2
2
2
5
560 | Dalton Trans., 2009, 5559–5566
This journal is © The Royal Society of Chemistry 2009

View Article Online
Fig. 2 Pt4f XPS spectra of the Pt-loaded silica samples before
A, AM2-111) and after (B, AM2-122) exposure to water vapour.
(
silica matrix produces a reduction of about half of the Pt(II) ions,
whereas further Pt(II) → Pt(0) reduction is achieved through water
vapour treatment, up to at least 70% of the total platinum content.
In both samples, the presence of some Cl ions adsorbed on the
silica surface was also observed. As a matter of fact, the position
Fig. 4 XRD analysis of Pd-SiO
AM3-69) exposure to moisture.
2
samples before (AM3-66) and after
-
(
(
ª199 eV) of the Cl2p line is typical for chloride species.
TEM analysis of the particle size distribution in a significant
to moisture, the appearance of two diffraction peaks indicates
a remarkable variation of the sample microstructure. The peaks
◦
◦
region of the samples showed that the particles in AM2-111 had
a smaller diameter than in AM2-122. Fig. 3 shows images of
the platinum crystallites in the two samples (A: AM2-111; B:
AM2-122).
centred at 2q ª 40.1 and 2q ª 46.6 are indexed as the (111) and
(200) reflexes of the fcc structure of metallic palladium (JCPDS
Card No. 46-1043, 2000). The mean diameter of the Pd nano-
crystallites, as evaluated using the Scherrer equation, is ca. 6 nm:
as already anticipated, we expect that only the largest particles
contribute to the diffraction.
XPS investigation evidenced that metallic palladium is already
present in the as-prepared sample. Fig. 5 shows the XPS high
resolution spectra of the Pd3d regions, with the d5/2 and d3/2
spin–orbit splitting components, for the as-prepared Pd-SiO
A, AM3-66) and after hydrolysis (B, AM3-69). A moderately
2
(
broad band-shape profile can be observed especially for the as-
prepared sample, suggesting the presence of palladium centres
with different oxidation states and/or chemical surroundings. The
changes between sample A and sample B in the Pd3d spectra
became more apparent by fitting two doublet components to
the data. In both cases, the positions of the components at
Fig. 3 TEM images of platinum aggregates on silica obtained by
using Pt
2
2 2 2
(m-Cl) Cl (CO) as precursor. A: sample AM2-111; B: sample
(
BE Pd3d5/2) ª335.2 eV and ª337.0 eV, corresponding to Pd(0)
AM2-122.
15
and Pd(II) species, respectively. It is worth highlighting that,
by comparing the intensities of the Pd(0) and Pd(II) compo-
nents, metallic palladium represents ca. 65% of the total metal
In order to study the corresponding phenomenon with pal-
ladium, the same silica support was contacted with an orange
content in the as-prepared Pt–SiO
5% after hydrolysis. Therefore, contacting the toluene solution
of the Pd (m-Cl) Cl (CO) compound with the silica matrix is
2
sample (A) increasing up to
toluene solution of Pd
2
(m-Cl)
2
Cl
2
(CO) . The process was carried
2
8
out at room temperature under a CO atmosphere in order
to contrast the spontaneous decarbonylation of the palladium
2
2
2
2
effective in promoting the Pd(II)-to-Pt(0) conversion, which is
7b
complex before its contact with the support. The migration of
the palladium-containing fragments from the yellow solution to
the SiO surface was evidenced by the colour transfer from the
2
solution to the support. As in the case of the platinum loading,
◦
this step was followed by treatment at 40 C in the presence of
water vapour. Also the palladium-loaded silica samples, before
(
AM3-66) and after (AM3-69) hydrolysis, have been characterized
by XPS, XRD and TEM analysis.
Concerning the sample microstructure, from the XRD diffrac-
tion patterns reported in Fig. 4 it can be clearly observed that
crystallization depends on sample treatment. As a matter of fact,
an amorphous pattern without any detectable crystalline phase is
observed for the Pd-SiO
with the toluene solution of Pd
2
obtained by contacting the silica matrix
(m-Cl) Cl (CO) . After exposure
Fig. 5 Pd3d XPS spectra of the Pd-loaded silica samples before (AM3-66,
A) and after (AM3-69, B) exposure to water vapour.
2
2
2
2
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5559–5566 | 5561

View Article Online
almost complete after controlled exposure to moisture under mild
conditions. XPS analysis reveals a palladium content of 0.12 at.%
in the as-prepared sample (in agreement with the ICP elemental
analysis, see Experimental section), which apparently decreases to
of experiments was performed by adding fresh alkene when the
hydrogenation was complete, showing that the catalytic activity
was maintained and even enhanced. The catalytic performance
expressed as the H /M molar ratio per hour, M being the total
2
-
1
-1
~
0.06 at.% after treatment with moisture: a possible explanation
is the aggregation of Pd atoms to larger particles which are not
completely detected by XPS. The Cl2p region was observed in the
metal content, corresponds to 890 h and 340 h for the platinum-
and the palladium-loaded silica, respectively. Plots reporting the
[absorbed H ]/metal molar ratio vs. time are shown in Fig. 7 and
2
as-prepared Pd–SiO
2
(A, AM3-66) sample, centred at a position
8 of the Experimental section.
typical for chloride species (ª199 eV), thus suggesting that some
Cl ions remained adsorbed on the silica surface. After hydrolysis
In Table 1 these data are compared with some literature
information for the same reaction carried out with palladium-
or platinum-based catalysts under similar conditions.
-
(
B, AM3-69), the chloride was practically undetected.
TEM analysis of the particle size distribution in a significant
region of the sample AM3-66 showed that the particles had a
smaller diameter than in AM3-69 (see Fig. 6).
It appears that the catalytic efficiency is only moderately affected
by the nature of the molecular precursor. For instance, the presence
of CO, Cl or Br in the precursor does not compromise the activity
of the final catalyst. Nevertheless, it is interesting to note that in
the experiments reported in the present paper (items 1 and 5 of
Table 1) the pre-treatment of the catalyst is limited to its exposure
◦
to water vapour at 40 C (see Experimental), while in other cases a
◦
pre-treatment at higher temperature (200–250 C) is necessary to
observe rate independence of the olefin concentration (for instance
items 3 and 6).
Commercial catalysts perform similarly.
Experimental
General information and instrumentation
Fig. 6 TEM images of the palladium aggregates (A: AM3-66; B:
AM3-69).
All operations were performed under an atmosphere of pre-
purified dinitrogen, or carbon monoxide, as specified. Solvents
were dried by conventional methods prior to use. The starting ma-
Catalytic hydrogenations
terials cis-PtCl
2
(CO)
2
, Pt
2
(m-Cl)
2
Cl
2
(CO)
2
and Pd
2
(m-Cl)
2
Cl
2
(CO)
2
6
,7
The catalytic activity of these materials has been evaluated in the
hydrogenation of cyclohexene, carried out at room temperature
and at atmospheric pressure in cyclohexane as solvent. The
reactions, which were monitored by measuring the hydrogen
uptake, are independent of the olefin concentration. A sequence
were prepared according to the literature. Commercial silica
(Grace, EP 17G, surface area 325 m g , pore volume 1.82 cm g )
was treated at 160 C over P
2
-1
3
-1
◦
4
O10 for 12 h in vacuo up to a constant
weight and then stored in flame-sealed vials under an atmosphere
of dry N . On the basis of our past experience this type of
2
Fig. 7 Molar ratio (absorbed H
2
)/Pt vs. time for the platinum-catalyzed reduction of cyclohexene. For the reaction conditions, see text.
5
562 | Dalton Trans., 2009, 5559–5566
This journal is © The Royal Society of Chemistry 2009

View Article Online
a
Table 1 Efficiency of palladium- and platinum-based catalysts in cyclohexene hydrogenation
◦
/M)] h-1
Metal
Precursor
T/ C
P/MPa
[(H
2
Ref.
1
2
3
4
5
6
7
8
9
Pt/SiO
Pt/SiO
2
2
2
Pt
PtCl
Pt(O
2
(m-Cl)
(CO)
CNEt
2
Cl
2
2
(CO)
2
24.9
20
20
75
24.9
24
25
20
30
0.1
0.1
0.1
0.4
0.1
0.1
0.2
0.1
0.1
890
720
340
250
340
400
300
217
200
This work
5a
4a
16
2
Pt/SiO
2
2
)
2
L
2
b
PtO
2
Pd/SiO
Pd/SiO
Pd/pol
Pd colloid
Pd/C 1%
2
2
Pd
Pd(O
PdCp(C
[NR [PdCl
2
(m-Cl)
2
Cl
2
(CO)
(NHEt
2
This work
2
CNEt
2
)
2
2
)
2
4b
17
18
19
3
H
5
)
4
]
2
2
Br
2
]
c
a
The column before the last one reports the experimentally observed H
2
/M molar ratio per hour. In experiments 1 and 5, described in this paper, no
◦
pre-treatment of the catalyst was carried out. The Pt- and Pd-loaded silica samples of experiments 2, 3 and 6 had been treated at 210–230 C before use. The
Pd(0) coordination polymer (Pd/pol) used in experiment 7 was prepared by reacting PdCp(C
3
H
17
5
) with 4,4¢-diisocyanobiphenyl and was pre-hydrogenated
4
(
15 h) before use. The Pd colloid of experiment 8 was prepared by reacting PdCl
2
with N(C
8
H
)
Br in 1:2 molar ratio in refluxing THF and then treating
b
c
the obtained solution with H
2
(P = 1 bar, room temperature) for 14 d. Adams’ catalyst. Commercialized by Aldrich.
Fig. 8 Molar ratio (absorbed H )/Pd vs. time for the palladium-catalyzed reduction of cyclohexene. For the reaction conditions, see text.
2
pre-treatment eliminates most of the hydrogen-bonded water
fluoroethylene (PCTFE) mulls, as specified in each case, prepared
under exclusion of moisture. Diffuse Reflectance Infrared Fourier
Transform (DRIFT) spectra were measured with the same in-
strument by mixing the samples with dry KBr under an inert
atmosphere and by rapid transfer to the cell (Spectra Tech).
The composition of the samples was investigated by XPS. The
analyses were performed with a Perkin-Elmer U 5600-ci spectrom-
from the surface, leaving most of the hydroxyl groups ≡Si–OH,
which are chemically reactive towards the metal precursor. The
-
1
total silanol content (typically, between 2.8 and 3.1 mmol g
corresponding to 6 OH groups per square nanometer) was
◦
estimated from the mass loss upon further heating at 850 C.
Elemental analyses (C, H, N) were performed at Dipartimento
di Scienze Farmaceutiche, Universit a` di Pisa, with a C. Erba mod.
a
eter using non-monochromatized Mg-K radiation (1253.6 eV).
-
8
1
106 elemental analyzer. Pt, Pd and Cl elemental analyses on
The working pressure was ≤ 5 ¥ 10 Pa. The spectrometer was
calibrated by assuming the binding energy (BE) of the Au 4f 7/2 line
at 83.9 eV with respect to the Fermi level. The standard deviation
for the BE values was 0.15 eV. The reported BE’s were corrected
for the charging effects, assigning a value of 284.8 eV to the C1s
the molecular precursors were carried out by reduction of the
products with sodium formate: metal black was filtered off, dried
and weighed. The Volhard method was applied to the filtrate. Pd
and Pt content in the metal-loaded silica was obtained by ICP-AES
with a Perkin-Elmer Plasma II instrument. The gas-volumetric
measurements were carried out according to the method described
21
line of adventitious carbon. Survey scans were obtained in the
0–1200 eV range. Detailed scans were recorded for the C1s, N1s,
O1s, Cl2p, Si2p, Pd3d and Pt4f regions. The analysis involved
20
by Calderazzo and Cotton.
22
IR spectra were recorded with a mod. 1725X FT-IR Perkin-
Elmer spectrophotometer in solution or as Nujol or polychlorotri-
Shirley-type background subtraction, non-linear least-squares
curve fitting, adopting Gaussian–Lorentzian peak shapes and
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5559–5566 | 5563

View Article Online
peak area determination by integration. The atomic compositions
were evaluated from peak areas using sensitivity factors supplied
by Perkin-Elmer, taking into account the geometric configuration
of the apparatus. The uncertainty on atomic composition values is
ca. 5–10%. The powders were pressed into pellets and introduced
into the XPS analytical chamber through a fast entry lock system.
XRD measurements have been performed using a Philips PW
a H
2
O/Pt molar ratio of 2.5, when the platinum chloro-carbonyl
had disappeared (IR) from the solution. Each addition of water
was accompanied by an almost immediate precipitation of a
solid. The violet precipitate of [Pt(CO)
stream (221 mg, 97% yield), showed IR (Nujol) bands at 2071 and
1889 cm .
2
n
] , after drying in a CO
-
1
In another experiment carried out under CO, preformed
1
820 diffractometer (Cu-K
a
radiation, 40 kV, 50 mA) equipped
n
[PtCl(CO)] (0.228 mg, 0.88 mmol of platinum), prepared as
described above, was suspended in 50 mL of toluene and 2.7 mmol
of water was added. The initially black solid became violet in about
15 min: after some hours stirring the violet solid, [Pt(CO) ] , was
2 n
filtered, briefly dried in a CO stream and sealed in vials under CO
(90% yield).
◦
with a thin film attachment (glancing angle = 0.5 ). Selected
angular ranges were step-scanned several times up to a satisfactory
signal-to-noise ratio. For the microstructure characterization, an
improved profile fitting method using a pseudo-Voigt representa-
tion for the line profiles was used. The average crystallite size was
calculated using the Scherrer equation, taking into account the
instrumental broadening.
TEM measurements were carried out with a Jeol JEM 2010
instrument operating at 200 kV. A small amount of specimen was
ground in a mortar to a very fine powder, which was deposited on
a lacy carbon film supported on a standard copper grid. The whole
Loading silica with metals
Platinum. In a typical experiment, a solution of Pt (m-
2
-
1
Cl)
2
Cl
2
(CO)
2
(0.25 g, 0.85 mmol of platinum, nCO = 2130 cm )
in toluene (150 mL) was added to pre-treated silica (7.9 g
sequence of operations was carried out under a dry N atmosphere.
2
corresponding to 25.5 mmol of silica-bonded hydroxyl groups,
The time required for sample preparation was minimized in order
to avoid deterioration or contamination.
≡
SiOH). A prompt colour fading of the solution was observed,
and, after 15 h stirring, an IR spectrum of the liquid phase showed
-
1
the band of the unconverted material at 2130 cm to be definitely
of lower intensity than initially. The suspension was filtered and
the pale-yellow solid was dried in vacuo, finally yielding the dry
pale-yellow substance (AM2-111, 7.67 g). Anal. (%): C, 0.49; H,
0.18; Pt, 0.76. DRIFT: 2116 cm . The XRD and XPS spectra
are in Fig. 1 and 2, respectively. For TEM data, see Fig. 3A: the
particle diameters range from 0.3 to 2.6 nm.
Reaction of cis-PtCl
2
(CO)
2
with H O in toluene
2
Under N
.3 mmol) in toluene (50 mL) was treated with a solution of water
150 mL, 8.3 mmol) in toluene (400 mL). The toluene solution of
2
atmosphere. A solution of cis-PtCl
2
(CO) (2.68 g,
2
8
(
-
1
water was added dropwise to the solution containing the platinum
precursor. After 1 h stirring an IR spectrum of the solution showed
the presence of a consistent amount of unreacted cis-PtCl
Subsequent direct additions (75 mL each, 4.15 mmol) of H
performed with a micro-syringe up to a final H O/Pt molar ratio
2
(CO)
O were
2
.
A portion of the platinum-loaded silica (1.26 g) was exposed at
40 C for 8 h to water vapour originated from an aqueous solution
of NaOH (2.5%). The silica became gradually grey. After drying
◦
2
2
of 3, under IR monitoring. At every addition, the amount of
the black solid was observed to increase and simultaneously the
in vacuo in the presence of P O10, no DRIFT bands attributable
4
to platinum-bonded carbonyl groups were observed. The final
substance was classified as AM2-122. TEM data, see Fig. 3B,
showed particle diameters ranging from 0.5 nm to 6.8 nm.
Samples of silica thus obtained were used for the catalytic
hydrogenation experiments.
concentration of cis-PtCl
2
(CO) in the liquid phase decreased (IR),
2
up to complete disappearance of the typical carbonyl stretching
-
1
◦
bands at 2168 and 2127 cm . After further stirring at 50 C for
0 min, the suspension was filtered and the solid was dried in vacuo
1.23 g, 57% yield based on the platinum content). Anal. Calcd
%) for [PtCl(CO)] : Pt, 75.4; Cl, 13.7. Found: Pt, 77.1; Cl, 11.8.
3
(
(
n
Palladium. Under an atmosphere of carbon monoxide, an
-
1
IR (Nujol): nCO, 2060 cm . The black solid thus obtained was
orange solution of 0.181 g of Pd
2
Cl
4
(CO) in 50 mL of toluene
2
subjected to the following experiments.
-
1
(
0.88 mmol of palladium, nCO = 2161 cm ) was added to a
(
a) The product (280 mg, 1.1 mmol of Pt) was introduced
under a N atmosphere into a Schlenk tube containing CH Cl
25 mL) and [NBu ]Cl (313 mg, 1.1 mmol) was added to
the resulting suspension. After 12 h, IR bands attributable to
suspension of 9.1 g of silica in 100 mL of toluene. Within 24 h
stirring at room temperature, the liquid phase of the mixture
became colourless and the resulting yellow solid was recovered
by filtration, dried in vacuo and sealed under dinitrogen (AM3-66,
2
2
2
(
4
-
1
[
NBu
suspension was filtered, and the filtrate was evaporated to dryness
under reduced pressure. The solid residue of [NBu [Pt Cl (CO)
was dried in vacuo (424 mg, 72% yield). Anal. Calcd (%) for
NBu [Pt Cl (CO) Pt : C, 38.1; H, 6.8; N, 2.6.
Found: C, 37.9; H, 6.7; N, 2.7. IR (Nujol): nCO, 2043 and 2022 cm .
b) Under N atmosphere, the product [PtCl(CO)] was reacted
4
]
2
[Pt
2
Cl
4
(CO)
2
] were observed at 2048 and 2028 cm . The
-
1
7.35 g). Anal. (%): C, 0.48; H, 0.23; Pd, 0.76. DRIFT: 1978 cm .
The XRD and XPS spectra are in Fig. 4 and 5A, respectively. For
TEM data see Fig. 6A.
4
]
2
2
4
2
]
◦
A portion of the solid, maintained at 40 C, was contacted for
[
4
]
2
2
4
2
], C34
H
72Cl
4
N
2
2
8
h with water vapour generated from a 2.7% solution of NaOH.
-
1
After drying in vacuo at room temperature, the reflectance spectra
of the solid showed no bands attributable to palladium-bonded
carbonyl groups. The resulting grey-coloured silica (AM3-69) was
sealed under dinitrogen in glass vials. TEM data, see Fig. 6B,
showed particle diameters ranging from 2.4 to 6.1 nm.
Samples of silica thus obtained were used for the catalytic
hydrogenation experiments.
(
2
n
with an excess of water in toluene; HCl was evolved and the
formation of platinum metal (95% yield) was established.
Under CO atmosphere. To a solution of cis-PtCl
2
(CO)
2
(
290 mg, 0.90 mmol) in toluene (35 mL), additions (8 mL each,
0
.45 mmol) of H
2
O were performed with a micro-syringe up to
5
564 | Dalton Trans., 2009, 5559–5566
This journal is © The Royal Society of Chemistry 2009

View Article Online
Catalytic hydrogenation of cyclohexene
Platinum-catalyzed hydrogenation.
Acknowledgements
A
reactor containing
This work was financially supported by the research projects CNR-
INSTM PROMO, FIRB-MIUR RBNE033KMA “Molecular
compounds and hybrid nanostructured materials with resonant
and non-resonant optical properties for photonic devices”, FISR-
MIUR “Inorganic hybrid nanosystems for the development and
the innovation of fuel cells”, CARIPARO 2006 “Multi-layer
optical devices based on inorganic and hybrid materials by
innovative synthetic strategies” and by Ministero dell’Universita
e della Ricerca Universitaria (MIUR), 2004–6. Thanks are due to:
Chimet, S.pA., Badia al Pino, Arezzo, Italy, for loans of platinum
and palladium, and for the analyses of the metal content in most
2
5 mL of cyclohexane saturated with dihydrogen and a flame-
sealed thin-walled glass ampoule containing the platinum-loaded
-
3
silica (0.143 g, 5.6 ¥ 10 mmol of platinum) was connected to
a gas-volumetric apparatus under a dihydrogen atmosphere. By
◦
operating at 24.9 C, 0.7 mL of a 1.69 M solution of cyclohexene
in cyclohexane were introduced (1.18 mmol, corresponding to
a cyclohexene/Pt molar ratio of 211). When the ampoule was
`
broken, the absorption of H
2
started. It was recorded as a function
of time and found to be substantially linear, corresponding to 14.8
-
1
-1
(
(
mmol of H
absorbed H
2
) ¥ min ¥ mmolPt . The plot of the molar ratio
)/Pt vs. time of a typical experiment is in Fig. 7.
2
of the silica-supported systems; Prof. E. Tondello, Universita di
Padova, for his interest and encouragement; Prof. P. Colombo and
Prof. A. Martucci, Universit a` di Padova, for XRD measurements
and helpful discussions.
`
Palladium-catalyzed hydrogenation.
A reactor containing
2
0 mL of cyclohexane saturated with dihydrogen and a flame-
sealed thin-walled glass ampoule containing the palladium-loaded
-
3
silica (0.134 g, 9.6 ¥ 10 mmol of metal) was connected with
a gas-volumetric apparatus under a dihydrogen atmosphere. By
References
◦
operating at 24.9 C, 0.7 mL of a solution of cyclohexene in
1
(a) P. K. Jain, X. Huang, I. H. El-Sayed and E. M. El-Sayed, Acc. Chem.
Res., 2008, 41, 1578; (b) I. Favier, S. Massou, E. Teuma, K. Philippot,
B. Chaudret and M. G o´ mez, Chem. Commun., 2008, 3296; (c) G. A.
Samorjai, F. Tao and J. Y. Park, Top. Catal., 2008, 47, 1; (d) R. W.
Murray, Chem. Rev., 2008, 108, 2688; (e) D. Astruc, Transition-metal
Nanoparticles in Catalysis: From Historical Background to the State-
of-the Art in Nanoparticles and Catalysis,ed. D. Astruc, Wiley-VCH
Verlag, Weinheim, 2008, pp. 1–48; (f) H. B o¨ nnemann and K. S.
Nagabhushana, Metal Nanoclusters: synthesis and strategies for their
size control in Metal Nanoclusters in Catalysis and Materials Science,
The Issue of Size Control, ed. B. Corain, G. Schmid and N. Toshima,
Elsevier B. V., Amsterdam, 2008, pp. 21–48.
cyclohexane (1.69 M, 1.18 mmol) were introduced, corresponding
to a cyclohexene/Pd molar ratio of 123. After breakage of the
ampoule, the absorption of H
2
vs. time was measured and found
to be substantially linear and corresponding to 5.6 (mmol of H
2
)
1
-1
¥
min- ¥ mmolPd . The plot of the molar ratio (absorbed H
)/Pd
2
vs. time is in Fig. 8.
Conclusion
2
3
D Poondi and M. A. Vannice, J. Catal., 1996, 161, 742.
For palladium–catalyzed hydrogenations, see: (a) Y. Gao, C.-A. Chen,
H.-M. Gau, J. A. Bailey, E. Akhadov, D. Williams and H.-L. Wang,
Chem. Mater., 2008, 20, 2839; (b) A. M. Doyle, S. K. Shaikhutdinov and
H.-J. Freund, Angew. Chem., Int. Ed., 2005, 44, 629; (c) For platinum-
catalyzed hydrogenations, see: F. O. Ernst, R. B u¨ chel, R. Strobel and
S. E. Pratsinis, Chem. Mater., 2008, 20, 2117; (d) T. Onoe, S. Iwamoto
and M. Inoue, Catal. Commun., 2007, 8, 701.
Our approach to the preparation of the palladium- or platinum-
containing catalysts was based on an attempt to minimize the
intrinsic differences between the two metals in the process of
loading on the inorganic matrix. As the metal particles were
prepared by a substantially identical process at room temperature,
starting from molecular precursors of the same type, we are con-
fident that our goal was accomplished, although we cannot avoid
differences induced by the different nature of the metal atoms;
particularly important is the different mass and the different
momentum which may affect the migration phenomenon on the
4 (a) L. Abis, D. Belli Dell’Amico, C. Busetto, F. Calderazzo, R. Caminiti,
C. Ciofi, F. Garbassi and G. Masciarelli, J. Mater. Chem., 1998, 8, 751;
(
b) L. Abis, D. Belli Dell’Amico, C. Busetto, F. Calderazzo, R. Caminiti,
F. Garbassi and A. Tomei, J. Mater. Chem., 1998, 8, 2855; (c) L. Abis, L.
Armelao, D. Belli Dell’Amico, F. Calderazzo, F. Garbassi, A. Merigo
and E. A. Quadrelli, J. Chem. Soc., Dalton Trans., 2001, 2704.
5 (a) D. Belli Dell’Amico, F. Calderazzo, C. Ciofi, F. Garbassi, L.
Grande and G. Masciarelli, J. Cluster Sci., 1998, 9, 473; (b) A. Merigo,
Ph.D. Thesis, University of Pisa, 2000.
23
surface of the support. Previous suggestions have been made that
palladium should possess a catalytic activity similar to, or even
higher than, that of platinum in hydrogenation reactions. However,
6 (a) B. P. Andreini, D. Belli Dell’Amico, F. Calderazzo and M. G.
Venturi, J. Organometal. Chem., 1988, 354, 357, and references therein;
24
earlier papers have reported conflicting evidence about this point.
It is important to point out that both platinum and palladium
crystallize in the fcc structure and that the corresponding atomic
(
b) F. Bagnoli, D. Belli Dell’Amico, F. Calderazzo, U. Englert, F.
Marchetti, A. Merigo and S. Ramello, J. Organomet. Chem., 2001,
22, 180, and references therein.
6
˚
radii (for CN = 12) have similar values (A): platinum, 1.39;
7 (a) D. Belli Dell’Amico and F. Calderazzo, Inorg. Chem., 1981, 20,
1310; (b) D. Belli Dell’ Amico, F. Calderazzo and N. Zandon a` , Inorg.
Chem., 1984, 23, 137.
CRC Handbook of Chemistry and Physics, 82nd edn, ed. D. R. Lide,
Boca Raton, 2001–2, pp. 9–63.
25
palladium, 1.37. Thus, metal particles of the same dimensions
should contain the same number of atoms and the same percentage
of exposed atoms. In our case the average diameters of the metal
particles in the platinum- and palladium-loaded silica used in the
catalytic tests are 1.6 and 4.5 nm, respectively. It follows that
the percentage of exposed metal atoms should be about 60 and
8
9 S. Ramello, Chemistry Degree Thesis, University of Pisa,
996.
0 I. V. Fedoseev and A. S. Gordeev, Zh. Neorg. Khim., 2007, 52,
2; I. V. Fedoseev and A. S. Gordeev, Chem. Abstr., 2007, 148,
134337.
1
1
1
3
0% in the platinum and palladium samples, respectively. Thus,
the apparent lower activity of palladium can be attributed to a
morphological aspect rather than to an intrinsic property. We can
therefore confirm the conclusions by Boudart that the catalytic
activity of the two metals is similar.
11 (a) G. Booth, J. Chatt and P. Chini, Chem. Commun., 1965, 639; (b) G.
Booth and J. Chatt, J. Chem. Soc. A, 1969, 2131.
1
2 (a) P. L. Goggin and R. J. Goodfellow, J. Chem. Soc., Dalton Trans.,
973, 2355; (b) A. Modinos and P. Woodward, J. Chem. Soc., Dalton
Trans., 1975, 1516.
23
1
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 5559–5566 | 5565

View Article Online
1
1
1
1
3 D. Belli Dell’Amico, F. Calderazzo, M. Dittmann, L. Labella, F.
19 M. V. Seregina, L. M. Bronstein, O. A. Platonova, D. M. Chernyshov,
P. M. Valetsky, J. Hartmann, E. Wenz and M. Antonietti, Chem. Mater.,
1997, 9, 923.
Marchetti, E. Schweda and J. Str a¨ hle, J. Organomet. Chem., 1999, 583,
1
62, and references therein.
4 Applications of Physical Methods to Inorganic and Bioinorganic Chem-
istry, ed. R. A. Scott and C. M. Lukehart, J. Wiley, Chichester, West
Sussex, 2007.
5 J. M. Moulder, W. F. Stickle, P. E. Sobol and K. D. Bomben, Handbook
of X-ray Photoelectron Spectroscopy, Perkin-Elmer Corporation, Eden
Prairie, MN, 1992 [http:www.//srdata.nist.gov/xps/].
6 C. W. Scheeren, G. Machado, J. Dupont, P. F. P. Fichtner and S. R.
Texeira, Inorg. Chem., 2003, 42, 4738.
20 F Calderazzo and F. A. Cotton, Inorg. Chem., 1962, 1,
30.
21 D. Briggs and M. P. Seah, Practical Surface Analysis, J. Wiley,
Chichester, Vol. 1, 1990.
22 D. A. Shirley, Phys. Rev. B, 1972, 5, 4709.
23 (a) M. Boudart and D. J. Sajkowski, Faraday Discuss., 1991, 92, 57;
(b) M. Boudart, Chem. Rev., 1995, 95, 661.
24 (a) E. Segal, R. J. Madon and M. Boudart, J. Catal., 1978, 52, 45;
(b) E. E. Gonzo and M. Boudart, J. Catal., 1978, 52, 462.
25 A. F. Wells, Structural Inorganic Chemistry, 5th edn, Clarendon Press,
Oxford, 1984, p. 1288.
1
1
7 I. Feinstein-Jaffe and A. Efraty, J. Mol. Catal., 1986, 35, 285.
8 H. B o¨ nnemann, R. Brinkmann, R. K o¨ ppler, P. Neiteler and J. Richter,
Adv. Mater., 1992, 4, 804.
5
566 | Dalton Trans., 2009, 5559–5566
This journal is © The Royal Society of Chemistry 2009