Encapsulation of supported metal nanoparticles with an ultra-thin porous shell for size-selective reactions

Article abstract of DOI:10.1039/c3cc44208j

A novel nanostructured catalyst with an ultra-thin porous shell obtained from the thermal decomposition of an aluminium alkoxide film deposited by molecular layer deposition for size-selective reactions was developed. The molecular sieving capability of the porous metal oxide films was verified by examining the liquid-phase hydrogenation of n-hexene versus cis-cyclooctene. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3cc44208j

Chemical Communications
www.rsc.org/chemcomm
Volume 49 | Number 86 | 7 November 2013 | Pages 10033–10194
ISSN 1359-7345
COMMUNICATION
Xinhua Liang et al.
Encapsulation of supported metal nanoparticles with an ultra-thin porous shell for size-selective reactions

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Encapsulation of supported metal nanoparticles with
an ultra-thin porous shell for size-selective reactions
Cite this: Chem. Commun., 2013,
49, 10067
Zeyu Shang,w Rajankumar L. Patel,w Brian W. Evanko and Xinhua Liang*
Received 4th June 2013,
Accepted 29th July 2013
DOI: 10.1039/c3cc44208j
www.rsc.org/chemcomm
A novel nanostructured catalyst with an ultra-thin porous shell
Normally, heterogeneous catalysts consist of small metal
obtained from the thermal decomposition of an aluminium alkoxide particles dispersed on a high surface area porous oxide support.
film deposited by molecular layer deposition for size-selective reac- The support maximizes the number of metal atoms on the
tions was developed. The molecular sieving capability of the porous surface, since metal atoms in the bulk are not involved in cata-
metal oxide films was verified by examining the liquid-phase hydro- lytic reactions. Traditional methods like hydrothermal synthesis4
genation of n-hexene versus cis-cyclooctene.
and sol–gel processes5 can produce inorganic coatings with few
defects, such as zeolite membranes and mesoporous films.
Heterogeneous catalysts are widely used (e.g., producing ferti- However, it is difficult for these methods to deposit ultra-thin
lizers, making gasoline from petroleum, and controlling auto- porous films inside the porous structure of the catalyst supports
motive exhaust pollution via the catalytic muffler). However, and control the thickness of the films with nanometer precision.
heterogeneous catalysts cannot selectively convert specific mole-
It is highly desirable to develop a new strategy to prepare an
cules in the reactant mixture to catalyze only desired reactions.1 ultra-thin porous film with controllable pore size and a limited
Size-selective catalysts with a metal core and porous oxide shell mass diffusion barrier. Lu et al.6 showed that ultra-thin porous
have a promising structure that can increase the reaction alumina films could be formed from dense atomic layer deposi-
selectivity through reactant molecular discrimination.1,2
ted (ALD) alumina films by thermal treatment at 700 1C. The
Current research activities in this area are mainly focused on pore size was about 2 nm. The formation of the porous structure
encapsulating unsupported catalysts or nanoparticles supported on was a result of the heat treatment process and there was no
dense catalyst supports.2,3 Nishiyama et al.2a applied an aqueous control over the pore size. Canlas et al.7 prepared shape-selective
solution of fumed silica, ethanol, and tetrapropylammonium sieving layers on an oxide catalyst surface by grafting the catalyst
hydroxide (TPAOH) to synthesize silicalite-1 coatings on spherical particles with bulky single-molecule sacrificial templates with
Pt/TiO2 particles with a diameter of 0.5 mm under hydrothermal submonolayer coverage, then partially overcoating the catalyst
conditions. The thickness of the silicalite-1 layer was about 40 mm. with alumina through ALD. The number of the nanocavities was
Yang et al.2b developed a size-selective catalyst with a core–shell controlled by the number of template molecules grafted on the
structure, wherein a Pd-containing silica core was surrounded by catalyst surface. The size of the nanocavities was controlled by
a silica shell in the presence of a cationic surfactant. This catalyst the size of the template molecule and the thickness of ALD films.
showed good activity and selectivity in the aerobic oxidation of Although they achieved good conversion and selectivity for
benzyl alcohol. Lu et al.3 encapsulated unsupported surfactant- photooxidation of certain alcohols, their technique requires
capped Pt nanoparticles in a zeolitic imidazolate framework rigid molecules with certain surface orientations as sacrificial
(ZIF-8) with an average pore size of about 1.2 nm and a film templates and many steps are needed to prepare the shape-
thickness of about 200 nm. The ZIF-8 porous coating greatly selective sieving layers.
increased the selectivity towards hydrogenation of n-hexene over
Recently, one novel method was developed by Liang et al. to
cis-cyclooctene. However, compared to the naked catalyst, there was prepare ultra-thin porous aluminium oxide films formed from
a 60% conversion loss for n-hexene hydrogenation mainly due to the calcination of aluminium alkoxide (alucone) films, which
the mass diffusion barrier from the relatively thick ZIF-8 coating.
were deposited by molecular layer deposition (MLD) using
trimethylaluminium (TMA) and ethylene glycol (EG) as precursors.8
The average pore size of porous alumina was about 0.6 nm
(more than 95%).8 MLD is a layer-by-layer gas phase thin film
coating technique, which has been utilized to deposit pure
polymer films or hybrid polymer films with nanometer-sized
Department of Chemical and Biochemical Engineering, Missouri University of
Science and Technology, Rolla, Missouri 65409, USA. E-mail: liangxin@mst.edu;
Tel: +1-573-341-7632
† Both authors contributed equally to this work.
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 10067--10069 10067

View Article Online
Communication
ChemComm
Fig. 1 Schematic representation of supported metal catalysts (a) before and
(b) after porous coating on all surfaces of the catalyst particles.
control of film thickness and well controlled film composition.9
The self-limiting nature of MLD makes it ideal for coating
porous substrates where line-of-sight gas phase deposition
methods fail. Herein, we prepared Pt/SiO2 catalysts encapsulated
by an ultra-thin porous alumina film by MLD, as schematically
shown in Fig. 1. Due to the conformal oxide coverage and sub-
nanometer pores, the reaction rates of larger reactants should be
slowed because they diffuse much more slowly than the smaller
reactants through the oxide layer. Since the porous film is only
Fig. 2 (a) Cross-sectional STEM image of the Pt nanoparticles on silica gel particles,
(b) Pt content, (c) surface area and pore volume, and (d) pore size distribution of the
Pt/SiO2 particles coated with different thicknesses of porous alumina films.
several nanometers thick, the mass diffusion barrier for the number of MLD cycles increased, as shown in Fig. 2b, because
smaller molecules is minimal. The combination of the catalytic the alumina film increased the catalyst weight.
properties of Pt nanoparticles and the molecular sieving cap-
ability of the porous oxide films was evaluated by examining the adsorption and desorption isotherms of catalyst particles at
liquid-phase hydrogenation of n-hexene versus cis-cyclooctene.
À196 1C. The specific surface areas of the samples were
A Quantachrome Autosorb-1 was used to obtain nitrogen
The Pt nanoparticles were deposited on mesoporous silica calculated using the BET method in the relative pressure range
particles by ALD using methylcyclopentadienyl-(trimethyl) of 0.05–0.25. The total pore volumes were calculated from the
platinum(IV) (MeCpPt-Me3) and oxygen as precursors in a adsorption quantity at a relative pressure of P/P0 = 0.99. The
fluidized bed reactor, as described in detail elsewhere.10 The pore size distribution curves were derived from the adsorption
silica particles are 30–75 mm in diameter with an average pore branches of the isotherms using the Barrett–Joyner–Halenda
size of 15 nm and a Brunauer–Emmett–Teller (BET) surface (BJH) method. As shown in Fig. 2c, the surface area of the
area of 270 m2 gÀ1. Three cycles of Pt ALD were carried out at Pt/SiO2 particles decreased from 266 to 219 m2 gÀ1 with 20 cycles
300 1C to obtain a Pt loading of 2.2 wt%, as measured using of MLD coating, and slightly increased with further deposition of
inductively coupled plasma-atomic emission spectroscopy MLD films. This increase is due to the contribution of the higher
(ICP-AES). A cross-sectional scanning transmission electron surface area porous alumina films, which can have a surface area
microscopy (STEM) image of Pt/SiO2 is displayed in Fig. 2a. as high as 1000 m2 gÀ1.8 The pore volume of the Pt/SiO2 particles
The white points in the picture are Pt nanoparticles. The Pt decreased from 0.99 to 0.81 cm3 gÀ1 with 20 cycles of MLD
nanoparticles are uniformly dispersed on the surface of silica coating, and slightly decreased again to 0.79 cm3 gÀ1 with
gel particles and inside of the porous structures. The average Pt further deposition of MLD films. As shown in Fig. 2d, the
particle size is about 2 nm. The Pt/SiO2 particles were coated average pore size of the silica gel particles was about 15 nm.
with three thicknesses of alucone MLD films, deposited with The number of mesopores decreased with the deposition of
20, 30, and 40 cycles of alternating reactions of TMA and EG in Pt nanoparticles, but there was no further decrease with the
a fluidized bed reactor at 160 1C.11 Three corresponding deposition of porous alumina films. Clearly, a large number of
thicknesses (B2, 3, and 4 nm) of porous alumina films were then micropores were formed after MLD coating, compared to silica
formed by oxidation at 400 1C in air, as described elsewhere.8 gel or Pt/SiO2 particles. The average pore size of the porous film
During the oxidation process, the samples were heated in air was estimated to be 0.6 nm.8
from room temperature to 400 1C at a rate of 1 1C minÀ1, kept at
The catalytic hydrogenation of olefins (n-hexene >99%, and
400 1C for one hour, and then cooled to room temperature at cis-cyclooctene >95%) was carried out in an ethyl acetate
the same rate. The organic component was removed completely solution under a static hydrogen atmosphere at 35 1C. The
and highly porous alumina films were formed. The carbon reactions were conducted in unstirred mini-batch reactors
chain length of the polymer component determines the pore assembled from 3/8 inch stainless steel Swageloks parts. Port
size, so MLD precursors with different carbon chain lengths connectors sealed with a cap on one end and one three-way
could lead to porous metal oxide films with different pore sizes. valve on the other end gave a reactor volume of about 2 mL. In a
The mass fraction of Pt in the catalyst decreased slightly as the typical run, n-hexene or cis-cyclooctene (0.08 g), ethyl acetate
c
10068 Chem. Commun., 2013, 49, 10067--10069
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
between Pt particles and the porous metal oxide films.12 About
42% of the Pt surface area was lost with the deposition of
40 cycles of MLD films.12 In this study, the reduction of the
conversion of n-hexene was 51% with the deposition of 40 cycles
of MLD films on Pt/SiO2, compared to the naked Pt/SiO2. It is
believed that the decline in the conversion of n-hexene was
mainly caused by the loss of Pt metal surface, rather than the
mass diffusion limitation resulting from the thin porous oxide
films. The porous structure allows smaller reactants to access the
encapsulated active sites, and inhibits or prevents the reactants
with larger molecular size from accessing the Pt sites. Since the
film is ultra-thin, the reactants and products of small molecules
can pass freely through the porous films. The molecular size of
H2 is so small that the size effects for H2 molecules can be
neglected. The size-selectivity effect results mainly from the
difference in the molecular size of olefins.
Fig. 3 Size-selective hydrogenation of n-hexene and cis-cyclooctene catalyzed
by Pt/SiO2 particles coated with different thicknesses of porous alumina films.
In summary, a novel strategy to prepare a supported size-
(0.78 g) and the Pt catalysts (B0.006 g) were added to the selective metal nanoparticle catalyst with an ultra-thin porous
reactor. All catalysts had an identical Pt loading even though shell was developed. The thickness of the porous oxide films
the total mass within the reactor increases as the number of could be well controlled at a subnanometer scale by applying
MLD cycles increases. The mass ratio of Pt to olefin was 0.15%. the MLD technique. The pore size of the film was about 0.6 nm.
The residual air in the reactor was expelled by flushing with The size selective effect of the porous alumina films was verified
hydrogen. The reactor was first pressurized to 20 psi with by the liquid-phase hydrogenation of n-hexene versus cis-
hydrogen and depressurized to atmosphere pressure. This process cyclooctene. This catalyst showed great selectivity in the hydro-
was repeated 50 times. After this flushing process, more than genation of olefins. Importantly, the mass diffusion limitation
99.999% of air was replaced by hydrogen gas. The control was not significant due to the ultra-thin films. The success of
experiments indicated that the mass loss of the reactants and making these materials by MLD opens up a new method for
the solvent during this flushing process was less than 0.5 wt%. preparing size-selective catalysts.
The reaction was carried out at 1 atm of hydrogen and 35 1C for
This work was partly supported by the University of Missouri
24 hours. The amount of hydrogen in the closed system was Research Board. The authors thank Dr Xinsheng Zhang at
more than enough for the hydrogenation reaction. After the the Environmental Research Centre at Missouri University of
reaction, the catalyst powder was filtered off and the filtrate was Science and Technology for the assistance with the GC analysis.
analysed using a gas chromatograph (Agilent, 6890N) equipped
Notes and references
with a 30 m DB-5 column and a FID detector to determine the
conversion and selectivity.
1 P. Collier, S. Golunski, C. Malde, J. Breen and R. Burch, J. Am.
The control experiments indicated that both silica gel particles
and alumina ALD films showed no catalytic activity for olefin
hydrogenation. The catalytic activity resulted solely from Pt.
The results are presented in Fig. 3. For the uncoated Pt/SiO2
catalyst, the conversion of n-hexene and cis-cyclooctene was
9.1% and 6.9%, respectively. The conversion of n-hexene
decreased with an B2 nm porous alumina film (20 cycles of
MLD), and decreased slightly more with further increases in
film thickness. The conversion of n-hexene fell to 4.5% after the
catalyst was coated with an B4 nm alumina film. In contrast,
the conversion of cis-cyclooctene decreased almost linearly as
the thickness of the porous alumina films increased, and no
obvious cis-cyclooctene conversion (o0.02%) was observed
with B4 nm of alumina films. Clearly, the naked Pt nano-
particles displayed indiscriminate catalysis of olefin hydroge-
nation. In contrast, the Pt nanoparticles encapsulated with a
porous alumina shell showed selectivity for catalytic hydro-
genation of n-hexene versus cis-cyclooctene due to the size
discrimination of the ultra-thin porous layer.
Chem. Soc., 2003, 125, 12414.
2 (a) N. Nishiyama, K. Ichioka, D. H. Park, Y. Egashira, K. Ueyama,
L. Gora, W. D. Zhu, F. Kapteijn and J. A. Moulijn, Ind. Eng. Chem.
Res., 2004, 43, 1211; (b) H. Q. Yang, Y. Z. Chong, X. K. Li, H. Ge,
W. B. Fan and J. G. Wang, J. Mater. Chem., 2012, 22, 9069.
3 G. Lu, S. Z. Li, Z. Guo, O. K. Farha, B. G. Hauser, X. Y. Qi, Y. Wang,
X. Wang, S. Y. Han, X. G. Liu, J. S. DuChene, H. Zhang, Q. C. Zhang,
X. D. Chen, J. Ma, S. C. J. Loo, W. D. Wei, Y. H. Yang, J. T. Hupp and
F. W. Huo, Nat. Chem., 2012, 4, 310.
4 Z. P. Lai, G. Bonilla, I. Diaz, J. G. Nery, K. Sujaoti, M. A. Amat, E. Kokkoli,
O. Terasaki, R. W. Thompson and M. Tsapatsis, Science, 2003, 300, 456.
5 D. Zhao, P. Yang, N. Melosh, J. Feng, B. F. Chmelka and G. D. Stucky,
Adv. Mater., 1998, 10, 1380.
6 J. L. Lu, B. S. Fu, M. C. Kung, G. M. Xiao, J. W. Elam, H. H. Kung and
P. C. Stair, Science, 2012, 335, 1205.
7 C. P. Canlas, J. L. Lu, N. A. Ray, N. A. Grosso-Giordano, S. Lee,
J. W. Elam, R. E. Winans, R. P. Van Duyne, P. C. Stair and
J. M. Notestein, Nat. Chem., 2012, 4, 1030.
8 X. H. Liang, M. Yu, J. H. Li, Y. B. Jiang and A. W. Weimer, Chem.
Commun., 2009, 7140.
9 (a) T. Yoshimura, S. Tatsuura and W. Sotoyama, Appl. Phys. Lett., 1991,
59, 482; (b) A. A. Dameron, D. Seghete, B. B. Burton, S. D. Davidson, A. S.
Cavanagh, J. A. Bertrand and S. M. George, Chem. Mater., 2008, 20, 3315.
10 X. H. Liang, Y. Zhou, J. H. Li and A. W. Weimer, J. Nanopart. Res.,
2011, 13, 3781.
11 X. H. Liang, D. M. King, P. Li, S. M. George and A. W. Weimer,
AIChE J., 2009, 55, 1030.
12 X. H. Liang, J. H. Li, M. Yu, C. N. McMurray, J. L. Falconer and
A. W. Weimer, ACS Catal., 2011, 1, 1162.
Previous studies of H2 chemisorption indicated that the
Pt dispersion decreased when the Pt nanoparticles were encapsu-
lated with porous alumina films due to the contact points
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 10067--10069 10069