Palladium nanoparticles in situ generated in metal-organic films for catalytic applications

Article abstract of DOI:10.1039/c4ta01592d

Palladium nanoparticles were first in situ generated in metal-organic films for catalytic applications. Layer-by-layer assembly of metal-organic films consisting of rigid-rod chromophores connected by terminal pyridine moieties to palladium centers on solid substrates was presented. Bipyridyl and polypyridyl ligands were used as building blocks to explore the influence of different ligand structures on catalytic properties. Metal-organic films were characterized by UV-Vis spectra, atomic force microscopy (AFM), transmission electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS). The results show that the deposition mechanism of metal-organic films is perfect layer-by-layer self-assembling with complete surface coverage and regular growth. Moreover, the catalytic activity toward the hydrogenation of olefin was investigated. Based on XPS and TEM, the catalytic activity toward the hydrogenation of olefin was ascribed to the in situ formation of Pd nanoparticles from Pd ions in metal-organic films. This film material is an active catalyst for the hydrogenation of olefin under mild conditions. Furthermore, catalytic results indicated that monodentate bipyridyl ligands exhibited superior catalytic activity than tridentate polypyridyl ligands. Catalytic activity is related to the loading amount of catalysts and permeability. More importantly, this study points toward the potential application of metal-organic films as heterogeneous catalysts with easy separation and good recyclability. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4ta01592d

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Palladium nanoparticles in situ generated in metal–
organic films for catalytic applications†
Cite this: J. Mater. Chem. A, 2014, 2,
2185
1
Shuiying Gao, Minna Cao, Weijin Li and Rong Cao*
Palladium nanoparticles were first in situ generated in metal–organic films for catalytic applications. Layer-
by-layer assembly of metal–organic films consisting of rigid-rod chromophores connected by terminal
pyridine moieties to palladium centers on solid substrates was presented. Bipyridyl and polypyridyl
ligands were used as building blocks to explore the influence of different ligand structures on catalytic
properties. Metal–organic films were characterized by UV-Vis spectra, atomic force microscopy (AFM),
transmission electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS). The results show
that the deposition mechanism of metal–organic films is perfect layer-by-layer self-assembling with
complete surface coverage and regular growth. Moreover, the catalytic activity toward the
hydrogenation of olefin was investigated. Based on XPS and TEM, the catalytic activity toward the
hydrogenation of olefin was ascribed to the in situ formation of Pd nanoparticles from Pd ions in metal–
organic films. This film material is an active catalyst for the hydrogenation of olefin under mild
conditions. Furthermore, catalytic results indicated that monodentate bipyridyl ligands exhibited superior
catalytic activity than tridentate polypyridyl ligands. Catalytic activity is related to the loading amount of
catalysts and permeability. More importantly, this study points toward the potential application of metal–
organic films as heterogeneous catalysts with easy separation and good recyclability.
Received 2nd April 2014
Accepted 2nd June 2014
DOI: 10.1039/c4ta01592d
www.rsc.org/MaterialsA
12–14
onto solid slides through metal ion–ligand interactions.
Based on metal ion–ligand interactions between ruthenium
1
Introduction
Noble metal Pd nanoparticles (NPs) are widely used in hydro- ions and terpyridine ligands, metal–organic lms consisting of
genation reactions in the petroleum rening industry. Due to ruthenium ions were prepared and exhibited optoelectronic
1
5,16
the low storage capacity of noble metals, decreasing the loading properties.
of noble metals is particularly important. Therefore, Pd NPs are and cobalt ions, were also assembled into lms through coor-
usually loaded on solid substrates, such as carbon and Al
, as dinative interactions between tpy-groups and metal ions
Moreover, other metal ions, such as zinc ions
2 3
O
1–7
17
heterogeneous catalysts for recycling. During the past few leading to electrochromic properties. However, metal–organic
years, the application of catalytic lms has attracted signicant lms toward catalytic application have rarely been reported.
8
interest in heterogeneous catalytic technology. Metal–organic
There are several reports on metallic Pd lms using poly-

lms represent a new variety of catalysts due to their electronic, electrolytes as building blocks, which have been explored as
9,10
photonic and catalytic properties of metal ions.
However, matrices for the synthesis of zero-charge Pd NPs by NaBH
4
their applications in electro or photo and catalytic devices were reduction to generate large surface area heterogeneous cata-
18–22
limited by their crystalline form. Layer-by-layer (LbL) assembly lyst.
For example, Bruening and coauthors reported the
of metal–organic lms can be simply achieved by alternating selective hydrogenation of unsaturated alcohols using LbL PAA/
the adsorption of metal precursor and ligand molecules PEI–Pd(0) lms on alumina particles (PAA ¼ poly(acrylic acid),
1
1
18–20
through covalent or noncovalent interactions. Noncovalent PEI ¼ polyethylenimine).
Polymer matrices have also been
interactions, such as electrostatic interactions, and direct metal utilized as solid supports because of the nature of macromole-
ion–ligand interactions were most commonly involved in self- cule, which can control the size of nanoparticles and prevent
2
assembly LbL multilayers. For example, functional metal– the agglomeration of particles. Polyethylene glycol (PEG)
organic complexes were loaded within LbL multilayers by stabilized Pd NPs synthesized by Harraz and coauthors exhibi-
alternate adsorption of palladium ion and pyridyl ligand layers ted a remarkable catalytic activity toward hydrogenation reac-
1
tions. However, the in situ synthesis of Pd NPs within the LbL
metal–organic lms were rarely reported. In situ self-assembly
State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the
Structure of Matter, The Chinese Academy of Sciences, Fuzhou 350002, China. LbL multilayer nanoreactors possess several distinct advantages
E-mail: rcao@irsm.ac.cn
such as tunable structure, variable composition and good
chemical stability.
1
†
Electronic supplementary information (ESI) available: H NMR and GC-MS
spectra of the product. See DOI: 10.1039/c4ta01592d
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 12185–12193 | 12185

View Article Online
Journal of Materials Chemistry A
Paper
multilayers were prepared by alternately immersing quartz slide
precoated with PEI layer in K PdCl aqueous solution (5 mM,
2
4
10 mL) and ligand (5 mM, 10 mL) chloroform solution. The
substrates were washed with water and dried with nitrogen
stream aer each immersion.
2
.3 Analysis of Pd in (Pd–ligand) metal–organic lms
n
Quartz slide coated with (Pd–ligand)n was dissolved by aqua
regia (10 mL). Pd content was determined by inductively
coupled plasma OES spectrometer (ICP-OES).
Scheme 1 Chemical structures of the pyridyl ligands.
2
.4 Typical procedure for hydrogenation of olen
Based on metal ion–ligand interaction, small organic func-
tional ligands (Scheme 1) can be used as building blocks to
construct self-assembly LbL lms. Each metal ion is separated
and connected by organic ligands. Here, we report a self-
Hydrogenation reactions were carried out in a hydrogen atmo-
sphere (1.2 atm) at 34 C. Typically, olen (0.5 mmol), absolute
ꢁ
ethanol (10 mL) and quartz slide coated with (Pd–ligand)10 lm
were added to the tube. The tube was then sealed with a teon
cap and evacuated and backlled with H
mixture was stirred at a constant stirring speed (500 rpm) in the
tube under an H atmosphere. The conversion and yield were
determined by GC analysis (430 GC). The ethanol solvent was
removed under reduced pressure. The product was conrmed
by H spectroscopy.
assembly multilayer of metal–organic (Pd–ligand) lms con-
n
2
three times. The
sisting of pyridyl ligand building blocks with characteristic
catalytic activity toward the hydrogenation of olen. The pyridyl
ligands of different coordination mode such as monodentate
and tridentate ligands were chosen to investigate the inuence
of ligand structures on catalytic properties. X-ray photoelectron
spectroscopy (XPS) and transmission electron microscope
2
1
2
+
(
TEM) investigation revealed that noble metal ions (Pd ) in
metal–organic lms are in situ reduced to Pd NPs during the
hydrogenation of olen process. Catalytic lms were not des- UV-Vis absorption spectra were recorded on a quartz slide with a
2
.5 Instrument
orbed from slides and showed good recyclability.
Lambda 35 spectrophotometer (Perkin Elmer, USA). All the AFM
images were recorded on a single-crystal silicon slide using a
Veeco Multimode NS3A-02 Nanoscope III atomic force micro-
scope with silicon tips. Height images of the lms were recor-
ded using tapping-mode AFM. High-resolution X-ray
2
Experimental materials and
methods
2.1 Materials
photoelectron spectra (XPS) were collected at a takeoff angle of
ꢁ
45
using PHI Quantum 2000 scanning ESCA microprobe
Poly(ethylenimine) (PEI, MW ¼ 60 000, 50 wt% aqueous solu-
(Physical Electronics, USA) with an Ala X-ray line (1486.6 eV).
0
0000
0
0
00
tion), 4 ,4 -(1,4-phenylene)bis(2,2 :6 ,2 -terpyridine) (bitpy),
trans-4-phenyl-3-buten-2-one, nitrobenzene and ethyl cinna-
mate were purchased from Aldrich Chemical Co. Ligand 1,2-
di(4-pyridyl)ethane (bpe) and 2,4,6-tri(4-pyridyl)-1,3,5-triazine
Transmission electron microscopy (TEM) was generated on a
JEM-2010 operated at 120 kV. Small pieces of a multilayer lm
were scraped off the quartz substrate in ethanol and transferred
to a copper grid. Gas chromatography-mass spectrometry (GC-
(
tpt) were purchased from TCI Chemical Co. Ligand 1,3-di(4-
1
MS) was performed on 430 GC (Varian, USA). H NMR
0
pyridyl)propane was from Alfa Company. 4,4 -Bipyridyl (bpy),
(400 MHz) spectra were obtained on a Bruker AVANCE 400
1,2-di(4-pyridyl)ethylene (dpe), styrene, linalool, 4-nitrostyrene
spectrometer. Analysis of noble metal content was measured by
inductively coupled plasma OES spectrometer (ICP-OES)
0
and 4-bromostyrene were from J&K Chemical Company. 4 -(4-
0
0
00
Pyridyl)2,2 :6 ,2 -terpyridine (tpy) was synthesized according to
the ref. 23. Millipore-ltered water was used in all the experi-
ments (Milli-Q, 18.2 MU cm).
(Ultima 2, France).
3
Results and discussion
2.2 Preparation of (Pd–ligand)
n
metal–organic lms
3.1 Preparation and characterization
Quartz slides (size 25 mm ꢀ 12 mm ꢀ 1 mm) (or single-crystal Multilayer lms of (metal ion–ligand)
n
were prepared by alter-
ꢁ
silicon slides) were cleaned with a “piranha solution” at 80 C nately immersing quartz slide pre-coated with PEI layer in metal
for 40 min and thoroughly rinsed with water. Further purica- ion aqueous solution and pyridyl ligand solution. Metal ion
tion was carried out by immersion in an H
2
O–H
2
O
2
–NH
3
OH precursor is K
2
PdCl
4
. Self-assembly of (metal ion–ligand)
n
lms
ꢁ
(5 : 1 : 1) (v/v/v) bath for 30 min at 70 C. The clean slides were is based on the coordination bond between metal ions and the

rst immersed in PEI solution for 20 min. (Pd/bpy)
n
, (Pd/bpe)
n
,
terminal pyridyl groups of the ligand. Pd ions were connected
(
Pd/dpe)n and (Pd/bpp)n multilayers were prepared by alter- with ligand molecules. Metal–organic coordination interactions
nately immersing quartz slide pre-coated with PEI layer in are responsible for LbL deposition between palladium ions and
PdCl aqueous solution (5 mM, 10 mL) and ligand (5 mM, ligand molecules. The lms (Pd–ligand) were characterized by
0 mL) ethanol solution. (Pd/tpt)
K
1
2
4
n
n
, (Pd/tpy)
n
and (Pd/bitpy)
n
UV-Vis spectra. As shown in Fig. 1, different structures of
12186 | J. Mater. Chem. A, 2014, 2, 12185–12193
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
n
Fig. 1 UV-Vis spectra of (Pd–ligand) .
ligands result in various peaks of UV-Vis spectra. The absor- the optical absorption and the number of ligand layers, sug-
bance of the lms is ascribed to the transition of p to p or n to p gesting that the ligand is incorporated into the lms with
2
of ligand molecules. For nonconjugate ligands of bpp and bpe, regular growth (R is a statistical calculation used with linear
the characteristic peaks of UV-Vis spectra are not obvious. regressions). From UV-Vis spectra, the deposition mechanism
However, for conjugate groups of ligands (i.e., bpy, dpe, tpt, tpy of (Pd–ligand)
n
multilayers is perfect layer-by-layer, self-assem-
and bitpy), the characteristic peaks were clearly observed and bling with complete surface coverage by plotting absorbance vs.
24
varied from the difference of structures. The peaks were labelled number of Pd–ligand bilayers, which was further demon-
as in Fig. S1 (see ESI†). Furthermore, the UV absorbance of (Pd– strated by the AFMs given below. In other words, all the ligand
ligand) lms increases linearly with the number of bilayers of molecules coordinate with Pd when depositing the ligand layer.
n
metal ion–ligand, respectively (see Fig. S1 in ESI†). The lm To conrm the uniform LbL process, ICP-OES of Pd in 2, 4, 6, 8
2
growth shows a good linear correlation with R ¼ 0.99 between and 10 bilayers were measured (see ESI, Table S1†). Pd content
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 12185–12193 | 12187

View Article Online
Journal of Materials Chemistry A
Paper
2
Fig. 2 Typical height images of the films on single-crystal silicon substrate (scan area is 1.0 ꢀ 1.0 mm , Z range: 50 nm). (a) (Pd/bpy)
nm), (b) (Pd/bpe) (Rrms ¼ 6.36 nm), (d) (Pd/bpp) (Rrms ¼ 7.84 nm), (e) (Pd/tpt) (Rrms ¼ 6.33 nm), (f) (Pd/tpy)
(Rrms ¼ 6.20 nm), (c) (Pd/dpe)
7.61 nm), (g) (Pd/bitpy)
(Rrms ¼ 8.62 nm).
5
(Rrms ¼ 8.04
5
5
5
5
5
(Rrms
¼
5
increases linearly with increasing number of bilayers, and Pd mechanism of these lms is perfect layer-by-layer, self-assem-
content of each bilayer increases equally. This further demon- bling with complete surface coverage. Only complete surface
strates that the lms are uniform LbL process.
coverage results in a uniform surface without surface defects,
Surface morphology information was characterized by which is consistent with the results of UV-Vis spectra.
atomic force microscopy (AFM) images (Fig. 2). AFM images
provide valuable information regarding surface topographic
3
.2 Catalytic performance of lms for hydrogenation of
images at the molecular level. As seen in AFMs, the silicon
surface with (Pd–ligand) clearly appears to be very smooth, and
olen
5
large aggregate particles are not observed. Distribution of the
particles is uniform and compact. Moreover, the root-mean-
square (RMS) roughness changed slightly for different structure
Hydrogenation of alkenes is important due to its industrial
25,26
applications such as petroleum processing.
Using our as-
prepared lm catalysts, olen hydrogenation was explored.
Styrene hydrogenation was applied as a test reaction to study
the catalytic activity of the lms. Pd content was detected by
ICP-OES. Pd content and TON of different ligands are shown in
Table 1. Pd content of both (Pd/bpy)10 and (Pd/bpe)10 multi-
of ligands. For example, the RMS roughness of (Pd/bpe)
5
and
(
Pd/dpe) lms is 6.20 nm and 6.36 nm, respectively. The (Pd/
5
bitpy) lms with rigid conjugated ligand bitpy show a larger
5
RMS of 8.62 nm. In general, the RMS roughness of lms
changed very little, from 6.20 nm to 8.62 nm. The above
mentioned results further demonstrated that the deposition
ꢂ7
layers is 4.2 ꢀ 10 mol. For monodentate pyridyl ligands dpe
and bpp, it was found that Pd content is similar. The maximum
a
Table 1 Pd content and TON of different films
(Pd/bpy)10
(Pd/bpe)10
(Pd/dpe)10
(Pd/bpp)10
(Pd/tpt)10
(Pd/tpy)10
(Pd/bitpy)10
ꢂ7
ꢂ7
ꢂ7
4.1 ꢀ 10ꢂ7
3.8 ꢀ 10ꢂ7
5.3 ꢀ 10ꢂ7
5.8 ꢀ 10ꢂ7
Pd (mol)
TON
4.2 ꢀ 10
4.2 ꢀ 10
4.4 ꢀ 10
1190
1190
1136
1219
1316
943
862
a
Turnover number (TON): moles of substrate converted per mole of catalyst.
12188 | J. Mater. Chem. A, 2014, 2, 12185–12193
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
ꢂ7
content of Pd is 5.8 ꢀ 10 mol for tridentate ligand of bitpy.
For (Pd/tpt) , Pd content is minimal due to a larger steric
1
0
hindrance of tpt. Turnover number (TON) (moles of substrate
converted per mole of catalyst) can be calculated by
Number of molecules reacted
TON ¼
Number of catalyst
TON represents the maximum yield of products attainable
27
from a catalytic center. Therefore, TON of (Pd/bpy)10 lms is
calculated at 1190. The maximum TON value is from (Pd/bpp)10

lms due to small steric hindrance of monodentate pyridyl and
the increase in alkyl moieties. In contrast, the minimum TON
value is from (Pd/bitpy)10 lms due to tridentate ligand of bitpy.
Under the same reaction conditions, TON of the 5% Pd/C
ꢂ7
reference catalyst (9.8 ꢀ 10
mol) is 510. Compared with
traditional heterogeneous catalysts in the reported refer-
28,29
ences,
such as Pd NPs/PEG with 660 TON and Pd NPs/MOF-5
with 682 TON, our metal–organic lms not only have higher
TON values, but also are easily removed from the reaction
mixture without separation ltration. The reusability of
(
Pd/bpy)10 lms was also investigated. Aer the completion of a
catalytic reaction, solid glass slides modied with (Pd/bpy)10
lms were simply removed from the product, washed with

ethanol, dried, and then reused in another catalytic cycle using
fresh styrene–ethanol mixture. This process was repeated for at
least 10 cycles for the hydrogenation of styrene (see ESI,
Fig. S2†). The results show that our lm catalysts have good
recyclability and recycling is more convenient compared to
traditional heterogeneous catalysts.
To demonstrate if the reaction proceeded in a heterogeneous
or homogeneous manner, Pd leaching was measured. (Pd/
bpy)10 catalysts were removed during the reaction (15 min for
styrene hydrogenation reaction, 2 h for nitrobenzene hydroge-
nation reaction), and the reaction mixtures were allowed to
react under identical reaction conditions (30 min for styrene
hydrogenation and 4 h for nitrobenzene hydrogenation). The
substrate conversion remained unchanged aer the removal of
catalysts from the reaction systems. This demonstrates that no
soluble species leaching from lm act as true catalysts, indi-
cating that the lm catalysts proceeded in a heterogeneous
fashion. In addition, inductively coupled plasma-mass spec-
trometry (ICP-MS) of supernatant for the reaction mixtures
(10 mL) was performed. Aer 5.5 h, nitrobenzene was
completely hydrogenated into aniline by (Pd/bpy)10 lm cata-
lysts, and the aniline–ethanol supernatant was used to analyze
ICP-MS. The results show that the concentration of Pd in reac-
tion mixtures is very low (0.081 ppm). This demonstrates that
the amount of Pd leaching into the reaction system is negligible
and the lms have good stability.
Several substrate molecules were used as test substrates to
investigate the relationship between the substrate conversion
and the number of (Pd/bpy) bilayers (Fig. 3). For styrene, the
n
Fig.
3
Relationship between different substrate conversion and
results show that the hydrogenation of styrene proceeds
smoothly and completely to give ethylbenzene without any
detectable by-products during a duration of 30 min. The cata-
lytic activity of the lms increases with the number of bilayers.
number of (Pd/bpy)
n
bilayers for a certain time. (a) 30 min, (b) 110 min,
(c) 120 min, (d) 120 min.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 12185–12193 | 12189

View Article Online
Journal of Materials Chemistry A
Paper
However, when the number of (Pd/bpy)
n
bilayers is greater than the catalytic reactions. Results are summarized in Table 1.
4, the conversion of styrene hydrogenation increases slowly Selectivity is not listed because the selectivity of the product
(Fig. 3a), whereas the conversion of 4-bromostyrene hydroge- reaches higher than 99%. For example, 3,7-dimethyl-1,6-octa-
nation remains unchanged when the number of (Pd/bpy)n dien-3-ol is selectively hydrogenated to 3,7-dimethyl-6-octen-3-
bilayers is greater than 6 (Fig. 3b). In contrast, conversion ol (Table 2, entry 3), in which the allyl alcohol is hydrogenated
increases with number of bilayers for nitrobenzene and ethyl with difficultly due to steric hindrance of the a position to the
cinnamate (Fig. 3c and d). Therefore, catalytic activity of the double bond. Benzalacetone and ethyl cinnamate is selectively
lms correlates with number of bilayers. In other words, cata- hydrogenated (Table 2, entries 4–5). Nitrobenzene is reduced
lytic activity is related to loading amount of catalysts. Films with into aniline (Table 2, entry 1). Especially for cyclohexene
more bilayers have higher efficiency. In addition, there is a more hydrogenation, our pyridyl lm (Pd/bpy)10 exhibits catalytic
important factor inuencing catalytic activity for styrene and efficiency (Table 2, entry 7).
4-bromostyrene. Hydrogenation conversion remains unchanged
The inuence of different ligand structures on catalytic
aer the multilayers reach a certain bilayer. This indicates that properties was also explored. Results are listed in Table 3. For
the catalytic activity is also related to the permeability of the monodentate ligands, such as bpy, bpe and bpp, catalytic
substrate molecule into the lm. Some substrate molecules are activity is similar. With an increase of alkyl moieties (i.e.,
permeable through a certain number of bilayers but are difficult from bpy to bpp), catalytic activity increases. In the case of
to penetrate into the inner multilayers due to increasing pene- tridentate ligands, such as tpy and bitpy, catalytic activity is
tration resistance. When the multilayers reach more bilayers, the lower than monodentate ligands. This is because all three
substrate molecule can not enter the inner multilayers. There- pyridyl groups of tpy and bitpy coordinate to Pd ions. For
fore, both the loading amount of catalysts and the permeability nitrobenzene and cyclohexene hydrogenation, the lms (Pd/
simultaneously inuence catalytic activity.
bpp)10 have the highest TOF, indicating that the lms con-
structed by bpp molecules exhibited better catalytic activity.
The lms of tpy molecules have the lowest TOF. This is
related to the coordination mode of the ligand. One pyridyl
group of the monodentate ligand is coordinated to the Pd
3.3 Effect of substrates
In addition to styrene, hydrogenation of other olen substrates
was investigated using our lm catalysts to explore the scope of
a
Table 2 Hydrogenation of olefin and nitrobenzene over film catalysts (Pd/bpy)10
b
Entry
1
Substrate
Product
Time
5.3 h
Yield (%)
>99
85
2
3
7.5 h
3 h
95
91
4
8.5 h
5
6
5 h
>99
99
2.5 h
8.0 h
7
92
a
ꢁ
b
Reaction conditions: substrate (0.5 mmol), absolute ethanol (10 mL), stirring, H (1.2 atm), 34 C. Determined by GC methods.
2
12190 | J. Mater. Chem. A, 2014, 2, 12185–12193
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
a
Table 3 Hydrogenation of olefin and nitrobenzene over different film catalysts
b
ꢂ1
Entry
Substrate
Product
Catalyst
Time
Yield (%)
TOF (h )
(
(
(
Pd/bpy)10
Pd/bpe)10
Pd/dpe)10
5.3 h
5.0 h
5.0 h
5.0 h
5.75 h
9.0 h
7.5 h
8.0 h
7.0 h
6.5 h
6.0 h
9.0 h
13.0 h
7.6 h
>99
>99
>99
>99
>99
>99
>99
92
93
91
92
86
224
238
227
244
228
104
114
137
158
159
187
125
45
1
(Pd/bpp)10
(Pd/tpt)10
(Pd/tpy)10
(Pd/bitpy)10
(Pd/bpy)10
(Pd/bpe)10
(Pd/dpe)10
2
(Pd/bpp)10
(
(
(
Pd/tpt)10
Pd/tpy)10
Pd/bitpy)10
62
86
97
a
ꢁ
b
Reaction conditions: substrate (0.5 mmol), absolute ethanol (10 mL), stirring, H
substrate converted per mole of catalyst per h.
2
(1.2 atm), 34 C. Determined by GC methods. TOF: moles of
ion. In other words, the Pd ion is coordinated to two different
pyridyl groups. In general, Pd ions are four-connected. Thus,
there were two non-coordinate free vacancies. When catalytic
substrate molecules attack Pd ions, two non-coordinate free
vacancy of Pd ions are easily occupied by the substrate
molecules. With the increase of the alkyl group of mono-
dentate ligands, the steric hindrance becomes smaller. Thus,
it exhibits better catalytic activity. In contrast, for tridentate
ligands, the Pd ion is all coordinated to pyridyl groups and
have no coordinate free vacancy. When the catalytic substrate
molecule attacks Pd ions, it competes against coordinated
pyridyl groups. This largely increases the resistance and
results in lower catalytic activity.
Our metal complex lms were explored as highly active
and recyclable catalyst for the hydrogenation of olen under
mild conditions. The reasons as highly efficient catalyst were
investigated by TEM and XPS in detail. During the process of
olen hydrogenation, hydrogen gas was used as a reductant
for metal ions as well as a hydrogen source for hydrogenation
reactions. The specic reduction process was clearly
n
explained by (Pd/bpy) lms. The reduction of Pd(II) to Pd(0)
in the LbL multilayers was further characterized by XPS and
TEM. A signicant difference was recorded for the Pd/bpy
multilayers before and aer reduction from XPS. Peak posi-
tion and intensity for elements of Pd largely change before
and aer hydrogen reduction. Before the styrene hydroge-
nation reaction, peaks with binding energy 337.0/342.4 eV
Fig. 4 XPS spectra of Pd3d region for (Pd/bpy)10 film deposited on
(
Fig. 4a) ascribed to Pd(II) were observed. Aer the 10th cycle solid glass substrate before styrene hydrogenation (a) and after 10th
cycle of styrene hydrogenation (b).
of styrene hydrogenation,
a couple of new peaks at
3
35.3/340.5 eV with high intensity are ascribed to Pd(0)
3
0
(
Fig. 4b), indicating that Pd(II) ions are reduced to Pd(0).
A
couple of peaks at 337.0/342.4 eV with very low intensity are
Based on the above mentioned analysis, metal ions are
ascribed to Pd(II). With an increase in cycle number, it is gradually reduced into NPs. As shown in TEM images, for (Pd–
predicted that the peaks at 337.0/342.4 eV will completely ligand)10 lms NPs tend to aggregate aer hydrogenation
disappear and those at 335.3/340.5 eV will largely increase, (Fig. 5). Analysis of TEM results further demonstrates that Pd(II)
indicating that all the Pd(II) ions are reduced into Pd(0) ions are gradually converted to Pd(0) states during the process
(
Fig. 4b).
of hydrogenation. During the hydrogenation process, metal
ions are rst in situ reduced into atoms encapsulated in
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 12185–12193 | 12191

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 5 TEM images of films after nitrobenzene hydrogenation. (a) (Pd/bpy)10, (b) (Pd/bpe)10, (c) (Pd/bpp)10, (d) (Pd/tpt)10, (e) (Pd/tpy)10, (f) (Pd/
bitpy)10
.
31
multilayers in hydrogen gas aqueous solution. The single
atoms separated from bpy molecules have high energy and high
4
Conclusions
32
catalytic activity. The atoms tend to reduce its energy. In other In the present work, palladium NPs in situ generated in metal–
words, metal atoms with high energy tend to form small NPs organic lms were described for catalytic applications. Metal–
with slightly low energy and lower activity under external organic lms were characterized by UV-Vis spectra and AFM.
conditions. On the other hand, NPs are more easily dispersed by Metal–organic lms containing Pd ions show uniform and
the pyridyl ligand to avoid aggregation. Thus, it results in high regular growth. TEM results also further demonstrated that Pd
activity of lms.
NPs were in situ formed in metal–organic lms during catalytic
As shown in TEM images, for monodentate ligands (i.e., bpy, hydrogenation process. The resulting lms displayed catalytic
bpe and bpp) (Fig. 5a–c), NPs tend to aggregate aer hydroge- properties. The results show that catalytic lms exhibit superior
nation and become larger, whereas for tridentate ligands NPs catalytic activity toward the hydrogenation of various types of
do not aggregate easily, and the size of NPs is smaller (Fig. 5e olen under mild conditions. In view of this activity, mono-
and f). This is related to the coordination mode of the lms. For dentate bipyridyl ligands lms exhibit better performance than
monodentate ligands, Pd ions are coordinated to two pyridyl that of tridentate polypyridyl lms. Catalytic activity of lms is
groups, whereas for tridentate ligands Pd ions are coordinated related to number of bilayers and permeability. Moreover, our
to four or six pyridyl groups. Thus, Pd ions of tridentate ligands catalytic lms are superior compared with traditional hetero-
are more rmly bonded than those of monodentate ligands. geneous catalysts due to ease of catalyst separation and good
The dispersion of particles in (Pd/bitpy)10 lms is more recyclability.
uniform. As seen in AFMs (Fig. 6), obvious aggregates occur
aer hydrogenation, consistent with TEM images.
Acknowledgements
This work was nancially supported by 973 Program
(2011CB932504 and 2012CB821705), NSFC (21173222,
21221001 and 21203199), Fujian Key laboratory of Nano-
materials (2006L2005), the Key Project from CAS.
Notes and references
1
2
3
X. Wang, G. Wu, N. Guan and L. Li, Appl. Catal., B, 2012, 115,
–15.
F. A. Harraz, S. E. El-Hout, H. M. Killa and I. A. Ibrahim,
J. Catal., 2012, 286, 184–192.
M. Lim, K. A. De Castro, S. Oh, K. Lee, Y.-W. Chang, H. Kim
and H. Rhee, Appl. Organomet. Chem., 2011, 25, 1–8.
7
Fig. 6 Typical height images of films on solid glass substrate after
hydrogenation. (a) (Pd/bpe)10, (b) (Pd/bitpy)10
.
12192 | J. Mater. Chem. A, 2014, 2, 12185–12193
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
4
5
6
7
8
F. Menegazzo, M. Signoretto, G. Frison, F. Pinna, G. Strukul, 18 S. Kidambi, J. Dai, J. Li and M. L. Bruening, J. Am. Chem. Soc.,
M. Manzoli and F. Boccuzzi, J. Catal., 2012, 290, 143–150. 2004, 126, 2658–2659.
T. Shamim, M. Gupta and S. Paul, J. Mol. Catal. A: Chem., 19 S. Kidambi and M. L. Bruening, Chem. Mater., 2005, 17, 301–
009, 302, 15–19. 307.
C. A. McNamara, M. J. Dixon and M. Bradley, Chem. Rev., 20 S. Bhattacharjee and M. L. Bruening, Langmuir, 2008, 24,
002, 102, 3275–3300. 2916–2920.
M. Benaglia, A. Puglisi and F. Cozzi, Chem. Rev., 2003, 103, 21 Y. Wang and J.-K. Lee, J. Mol. Catal. A: Chem., 2007, 263, 163–
401–3430. 168.
A. Mentbayeva, A. Ospanova, Z. Tashmuhambetova, 22 Z. Ren, H.-L. Wang, Y.-Q. Cai, M. Chen and D.-J. Qian, Mater.
2
2
3
V. Sokolova and S. Sukhishvili, Langmuir, 2012, 28, 11948–
1955.
Chem. Phys., 2011, 127, 310–315.
23 J. H. Wang and G. S. Hanan, Synlett, 2005, 8, 1251–1254.
1
9
H. Krass, E. A. Plummer, J. M. Haider, P. R. Barker, 24 J.-F. Koenig and D. Martel, Thin Solid Films, 2008, 516, 3865–
N. W. Alcock, Z. Pikamenou, M. J. Hannon and
3877.
D. G. Kurth, Angew. Chem., Int. Ed., 2001, 40, 3862–3865.
25 D. Li, Chin. J. Catal., 2013, 34, 48–60.
10 H. Krass, G. Papastavrou and D. G. Kurth, Chem. Mater., 26 F. Pinto, S. Martins, M. Gonçalves, P. Costa, I. Gulyurtlu,
2
003, 15, 196–203.
A. Alves and B. Mendes, Appl. Energy, 2013, 102, 272–282.
27 S. Kozuch and J. M. L. Martin, ACS Catal., 2012, 2, 2787–
2794.
1
1
1 G. Decher, Science, 1997, 277, 1232–1237.
2 M. Altman, A. D. Shukla, T. Zubkov, G. Evmenenko, P. Dutta
and M. E. van der Boom, J. Am. Chem. Soc., 2006, 128, 7374– 28 F. A. Harraz, S. E. El-Hout, H. M. Killa and I. A. Ibrahim,
382. J. Catal., 2012, 286, 184–192.
3 J. Choudhury, R. Kaminker, L. Motiei, G. de Ruiter, 29 Y. Pan, D. Ma, H. Liu, H. Wu, D. He and Y. Li, J. Mater. Chem.,
7
1
M. Morozov, F. Lupo, A. Guilino and M. E. van der Boom,
J. Am. Chem. Soc., 2010, 132, 9295–9297.
4 S. Gao, Z. Zheng, J. L u¨ and R. Cao, Chem. Commun., 2010, 46,
2012, 22, 10834–10839.
30 S. Kidambi and M. L. Bruening, Chem. Mater., 2005, 17, 301–
307.
1
1
1
1
7
584–7586.
31 C. W. Scheeren, G. Machado, S. R. Teixeira, J. Morais,
J. B. Domingos and J. Dupont, J. Phys. Chem. B, 2006, 110,
13011–13020.
5 J. Pan, B. Tong, J. Shi, W. Zhao, J. Shen, J. Zhi and Y. Dong,
J. Phys. Chem. C, 2010, 114, 8040–8047.
6 W. Zhao, B. Tong, Y. Pan, J. Shen, J. Zhi and Y. Dong, 32 D.-J. Qian, T. Wakayama, C. Nakamura and J. Miyake, J. Phys.
Langmuir, 2009, 25, 11796–11801.
Chem. B, 2003, 107, 3333–3335.
7 A. Maier and T. Bernd, J. Phys. Chem. B, 2012, 116, 925–934.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 12185–12193 | 12193