Controlling interparticle spacing among metal nanoparticles through metal-catalyzed decomposition of surrounding polymer matrix

Article abstract of DOI:10.1021/ja050735+

Systematic and reproducible control over average interparticle spacing of Pt, Ni, and Cu nanoparticles embedded in polyimide thin layers was achieved. The metal-catalyzed decomposition of polyimide matrixes surrounding metal nanoparticles causes a decrease in the composite layer thickness, while maintaining the size of nanoparticles. This ability provides an effective methodology for the preparation of metal/polymer nanocomposites with tailored microstructures and holds great promise toward the fundamental understanding of the physical interactions among metal nanoparticles. Copyright

Full text of DOI:10.1021/ja050735+

Published on Web 05/10/2005
Controlling Interparticle Spacing among Metal Nanoparticles through
Metal-Catalyzed Decomposition of Surrounding Polymer Matrix
Kensuke Akamatsu,*, Hiroyuki Shinkai, Shingo Ikeda, Satoshi Adachi, Hidemi Nawafune,* and
†,‡
‡
‡
†
,†,‡
§
Satoshi Tomita
Faculty of Science and Engineering and Graduate School of Science, Konan UniVersity, 8-9-1 Okamoto,
Higashinada, Kobe 658-8501, Japan, and PRESTO, Japan Science and Technology Agency (JST), Wako,
Saitama 351-0198, Japan
Received February 4, 2005; E-mail: akamatsu@center.konan-u.ac.jp
Nanocomposites containing inorganic nanoparticles have been
of great interest in recent years due to their potential application
1
for novel electronic, optical, and magnetic devices. The size of
the nanoparticles is one of the key parameters that determine the
properties of the nanocomposites. Additionally, physical interactions
among nanoparticles, such as electric and magnetic dipole interac-
tions, are recognized to be important in the next generation nano-
2
3
devices, including plasmon waveguides, biosensors, and ultrahigh
4
density data storage media. Since the physical interactions depend
on both the particle size and interparticle spacing, precise control
over the spacing under a fixed particle size is an important challenge
for understanding the physical interactions in the nanocomposite
systems. However, previous attempts at control over microstructures
of the nanocomposites did not enable one to independently control
the interparticle spacing and the nanoparticle size.5
Figure 1. (A) Schematic for controlling the interparticle spacing in the
nanocomposites. (B) Cross-sectional TEM images of nanocomposite films
containing Pt, Ni, and Cu nanoparticles prepared by annealing in a hydrogen
atmosphere at 270 °C (Pt), 290 °C (Ni), and 400 °C (Cu) for 30 min. Scale
bar: 50 nm. (C) Changes in nanoparticle size as a function of annealing
time. Lines serve as guides to the eyes.
Herein, we report a novel method to control the average
interparticle spacing among metal nanoparticles embedded in a
polyimide matrix based on an in situ approach using ion-doped
precursors.6 This is achieved by the metal-catalyzed decomposition
of the surrounding matrix, which induces a decrease in thickness
,7
(i.e., volume) of the composite layer, while the nanoparticle size
remains intact (Figure 1A). This holds great promise not only
toward the fundamental understanding of the physical interactions
among metal nanoparticles but also toward the development of
novel nanocomposites with tailored microstructures.
The approach works by the hydrolysis of polyimide resin using
an alkali solution, subsequent incorporation of metal ions through
ion exchange, followed by hydrogen-induced reduction of ions to
form metal nanoparticles. In a typical procedure, the PMDA-ODA-
type polyimide films were treated with aqueous KOH solution (5
Figure 2. Cross-sectional TEM images of nanocomposite thin films
containing Ni nanoparticles. The films were annealed at 290 °C for 30 (A),
6
0 (B), and 90 min (C). Scale bar: 500 nm. (D-F) High-magnification
images of the composite layers corresponding to images (A), (B), and (C),
respectively. Scale bar: 20 nm.
9
respectively. The nanoparticle size remains intact upon longer
M at 50 °C) for 5 min to form layers of potassium salts of poly-
(
annealing (Figure 1C). Longer annealing at the fixed temperature,
however, causes a decrease in composite layer thickness. Typical
examples are shown in Figure 2A-C. The thickness of the
composite layer containing Ni nanoparticles decreases from 1.4 to
0.7 µm as the annealing time increases from 30 to 90 min. It is
important to note that only the concentration of Ni nanoparticles
increases, as can be seen in the TEM images (Figure 2D-F). We
confirmed by tilting experiments during TEM observation that
anisotropic distribution, concentration gradient of the nanoparticles,
and interparticle contact were not observed in the present experi-
mental conditions.
The decrease in layer thickness with fixed nanoparticle size was
also observed for Pt and Cu samples, although the rate of decrease
in thickness was highly dependent on the species of metals and
the temperatures (Figure 3A). In addition, a decrease in film weight
upon annealing was observed. The time dependence of the decrease
in the layer thickness is quite similar to that of the decrease in film
weight (Figure 3B). We also found solid decomposition products
on the inner wall of the quartz tube after annealing. It should be
amic acid)6 with a thickness of 3 µm. The formed carboxylic
,8
acid groups can be used as ion exchange sites for subsequent doping
of metal ions using aqueous 10 mM PtCl (containing 1 M NH ),
0 mM NiCl , or 50 mM CuSO solutions, thus generating the ion-
2
3
5
2
4
-3
doped precursor layers (ion loading: 3.8 ( 0.2 mmol cm ). Forma-
tion of metal nanoparticles was achieved by annealing the precursors
at 270-400 °C in a hydrogen atmosphere in a quartz tube.
Metal nanoparticles (4-10 nm in diameter) are generated upon
hydrogen reduction, during which the modified layers are again
transformed to polyimide through a thermally induced dehydration
reaction (re-imidization).6b The reduction temperature depends on
the metal ions doped into the precursors. For example, 5.4 nm
diameter Pt nanoparticles are formed upon 270 °C annealing for
30 min (Figure 1B). Similarly, 7.3 nm Ni and 9.1 nm Cu
nanoparticles are formed upon annealing at 290 and 400 °C,
†
Faculty of Science and Engineering, Konan University.
‡
Graduate School of Science, Konan University.
§
PRESTO, JST.
7980
9
J. AM. CHEM. SOC. 2005, 127, 7980-7981
10.1021/ja050735+ CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
average interparticle spacings (surface-to-surface distance) of 6.5-
2.4 (Pt), 9.8-4.0 (Ni), and 13.4-6.0 (Cu) nm (longer to shorter
limits, obtained by lower to higher volume fractions). The estimated
distances become comparable and shorter than the corresponding
nanoparticle size, a regime where particle-to-particle interactions
are thought to play an important role in determining the physical
properties of the resulting nanocomposites. Efforts to explore optical
and magnetic properties are currently underway.
Figure 3. Changes in thickness (A) and weight (B) of the composite layer
as a function of annealing time. The weight was normalized to that of film
before annealing. The data obtained from the films acidified with HCl (1
vol %) after KOH treatment are also shown. The films were annealed at
This communication describes the systematic control of inter-
particle spacing among metal nanoparticles embedded in a high-
performance polyimide matrix via metal-catalyzed in situ decom-
position of the polyimide matrix. The strategy is quite general and
2
70 °C (Pt), 290 °C (Ni), 300 °C (HCl-acidified film), and 400 °C (Cu).
Lines serve as guides to the eyes.
1
4
extendable to other noble metals and transition metal alloys. The
ability to control interparticle spacing with a fixed amount of metal
and nanoparticle size enables the elucidation of the effect of particle-
to-particle interactions in studies of optical, electrical, and magnetic
properties of the nanocomposites. An understanding of these
physical properties and their relationship to the material micro-
structures is essential for initiating new technological applications
of metal/polymer nanocomposite materials, especially in the area
of optoelectronic and electromagnetic devices.
noted here that no decrease in layer thickness and weight was
observed for acidified films (without metallic ions, Figure 3). These
results allow one to conclude that the decrease in film thickness is
caused by metal-catalyzed decomposition of polyimide surrounding
the metal nanoparticles.
Similar trends have been observed for cellulose fibers containing
Pt nanoparticles10 and polyimide films containing Fe nanoparticles,11
both of which resulted in the formation of a carbon matrix through
carbonization (pyrolysis) of the matrix at high temperatures.
Oxidative degradation of thick polyimide films containing Ag
nanoparticles was also observed, but resulted in a broad size
Acknowledgment. This work was supported in part by JSPS.
S.I. is grateful for research fellowship support from JSPS.
Supporting Information Available: Film characterization and
FTIR spectra. This material is available free of charge via the Internet
at http://pubs.acs.org.
distribution and inhomogeneous distribution of the nanoparticles
in the film.7b In the current process, however, a different mechanism
for metal-catalyzed decomposition is operative. First, the molecular
structure after longer annealing was identified using infrared
spectroscopy as the fully cured polyimide (Figure S1), indicating
that carbonization of the matrix did not occur. Second, annealing
in inert atmosphere did not induce decomposition of the matrix
References
(1) (a) MacLachlan, M.; Ginzburg, M.; Coombs, N.; Coyle, T. W.; Raju, N.
P.; Greedan, J. E.; Ozin, G. A.; Manners, I. Science 2000, 287, 1460-
1463. (b) Huyuh, W. U.; Dittmer, J. J.; Alivisatos, A. P. Science 2002,
295, 2425-2427. (c) Coe, S.; Woo, W. K.; Bawendi, M.; Bulovic, V.
Nature 2002, 420, 800-803. (d) Tessler, N.; Medvedev, V.; Kazes, M.; Kan,
S.; Banin, U. Science 2002, 295, 1506-1508. (e) Korchev, A. S.; Bozack,
M. J.; Slaten, B. L.; Mills, G. J. Am. Chem. Soc. 2004, 126, 10-11.
2) Maier, S. A.; Kik, P. G.; Atwater, H. A.; Meltzer, S.; Harel, E.; Koel, B.
E.; Requicha, A. A. G. Nat. Mater. 2003, 2, 229-232.
(results not shown). Third, an FTIR spectrum of the decomposition
products shows the weak vibrational bands characteristic of NH
2
(
and CHO splits in addition to strong vibrational bands for polyimide
and an amide bond (Figure S2). From these results, although further
experiments are necessary to fully elucidate the decomposition
mechanism, the current process involves decomposition of the
matrix by partial hydrogenation of the imide ring to form the
aldehyde and amide bond (imide ring cleavage) and finally the
dialdehyde and amine (amide bond cleavage), generating oligomeric
(3) (a) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin,
C. A. Science 1997, 277, 1078-108. (b) Jin, R.; Wu, G.; Li, Z.; Mirkin,
C. A.; Schatz, G. C. J. Am. Chem. Soc. 2003, 125, 1643-1654.
(
4) (a) Sun, S.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A. Science 2000,
287, 1989-1992. (b) Kleemann, W.; Petracic, O.; Binek C.; Kakazei, G.
N.; Pogorelov, Y. G.; Sousa, J. B.; Cardoso, S.; Freitas, P. P. Phys. ReV.
B 2001, 63, 134423-134427. (c) Ebels, U.; Duvail, J. L.; Wigen, P. E.;
Piraux, L.; Buda, L. D.; Ounadjela, K. Phys. ReV. B 2001, 64, 144421-
144426.
(
low molecular weight) molecules that are vaporized during
annealing. Since the imidized PMDA-ODA matrix is in a glassy
) 420 °C)12 and the
state at the present temperature range (T
(
5) (a) Laurent, C.; Mauri, D.; Kay, E.; Parkin, S. S. P. J. Appl. Phys. 1989,
65, 2017-2020. (b) Heilmann, A.; Werner, J.; Schwarzenberg, D.; Henkel,
S.; Grosse, P.; Theiss, W. Thin Solid Films 1995, 270, 103-108. (c) Salz,
D.; Lamber, R.; Wark, M.; Baalmann, A.; Jaeger, N. Phys. Chem. Chem.
Phys. 1999, 1, 4447-4451.
g
catalytic reaction would occur only at the particle/matrix interface,
the nanoparticles do not have mobility to come in contact with
others so that the nanoparticles are effectively isolated without
coalescence.
The gradual decrease in the composite layer thickness, under a
fixed nanoparticle size and total number of metal atoms in the film,
indicates a gradual increase in the volume faction (f), and thus a
gradual decrease in the average interparticle spacing. The f is given
by following equation:
(
6) (a) Akamatsu, K.; Ikeda, S.; Nawafune, H.; Deki, S. Chem. Mater. 2003,
1
5, 2488-2491. (b) Ikeda, S.; Akamatsu, K.; Nawafune, H.; Nishino, T.;
Deki, S. J. Phys. Chem. B 2004, 108, 15599-15607.
(
7) (a) Wang, T. C.; Rubner, M. F.; Cohen, R. E. Chem. Mater. 2003, 15,
2
99-304. (b) Southward, R. E.; Thompson, D. W. Chem. Mater. 2004,
1
6, 1277-1284. (c) Gaddy, G. A.; Locke, E. P.; Miller, M. E.; Broughton,
R.; Albrecht-Schmitt, T. E.; Mills, G. J. Phys. Chem. B 2004, 108, 17378-
1
7383.
(
8) Thomas, R. R.; Buchwalter, S. L.; Buchwalter, L. P.; Chao, T. H.
Macromolecules 1992, 25, 4559-4568.
(
9) We confirmed by ICP measurement that there remained no ions for
annealed samples. During TEM observation, we also analyzed the area
between visible nanoparticles by nanobeam EDX and observed that no
signals of metals were detected for annealed samples. These results indicate
that all of the initially doped metallic ions were incorporated into the
nanoparticles.
f ) CL M F L-1
-
1
(1)
0
w
(
10) He, J.; Kunitake, T.; Nakao, A. Chem. Commun. 2004, 4, 410-411.
11) Kaburagi, Y.; Toriyama, T.; Yoshida, A.; Wakabayashi, H.; Hishiyama,
Y. J. Mater. Res. 2001, 16, 352-365.
where C, L , M , F, and L are initial ion loading, initial composite
0 w
(
layer thickness, atomic weight of metals, density of the metal, and
composite layer thickness after annealing, respectively. The average
interparticle distance between metal nanoparticles was estimated
using f and particle size by assuming that the particles are arranged
in a simple cubic lattice. In the present system, the f could be
controlled in the range of 5.0-18.3, 4.2-28.1, and 3.5-11.1% for
Pt, Ni, and Cu nanoparticles, respectively.13 These values give
(
12) Faupel, F.; Willecke, R.; Thran, A. Mater. Sci. Eng. 1998, R22, 1-55.
13) The values obtained by assuming the hcp and bcc lattice were only slightly
larger (1-2 nm) than those obtained by the simple cubic model.
(
(14) We have extended our process to Ag, Co, and Fe nanoparticles and
x x
Ni Co1-x and Ni Fe1-x alloy nanoparticles. Although the temperature at
which the decomposition of polyimide occurs depends on the species of
metal and alloy, general behavior is similar to that of Pt, Ni, and Cu.
JA050735+
J. AM. CHEM. SOC.
9
VOL. 127, NO. 22, 2005 7981