Effect of electron irradiation on nanogroove-networked single-crystalline and dendritic polycrystalline platinum nanosheets prepared from lyotropic surfactant liquid-crystal templates

Article abstract of DOI:10.1016/j.materresbull.2007.05.019

Using transmission electron microscopy (TEM) electron irradiation effects were studied for nanogroove-network structured single-crystalline and dendritic polycrystalline Pt nanosheets 50-60 nm in size. These two nanosheets with nearly the same average groove-width or dendritic spacing of 1.3-1.4 nm were prepared from the mixed and single surfactant liquid crystalline templates, respectively. On exposure to electron beam for 20 min at the acceleration voltage of 200 kV, the nanogrooved nanosheets were morphologically little affected, but the dendritic ones were transformed into less branched polycrystalline structures with spacings distributed around ～1.7 nm. The shape transformation of the latter occurred by the combined mechanism of segmental migration and atomic diffusion. These observations indicate that the nanogrooved Pt nanosheets are highly stabilized by the grooved but crystallographically continuous Pt framework, leading to their extremely high thermo-resistance, in marked contrast to the polycrystalline dendritic structures constructed of crystallographically discontinuous linkages of nanoblocks.

Full text of DOI:10.1016/j.materresbull.2007.05.019

Materials Research Bulletin 43 (2008) 1282–1290
www.elsevier.com/locate/matresbu
Effect of electron irradiation on nanogroove-networked
single-crystalline and dendritic polycrystalline platinum
nanosheets prepared from lyotropic surfactant
liquid-crystal templates
Takumi Yoshimura a, Masafumi Uota b, Takeshi Kuwahara a, Daisuke Fujikawa a,
Hideya Kawasaki c, Go Sakai a, Tsuyoshi Kijima a,
*
a Department of Applied Chemistry, Faculty of Engineering, Miyazaki University, Miyazaki 889-2192, Japan
b CREST, Japan Science and Technology Cooperation, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan
c Department of Applied Chemistry, Faculty of Engineering, Kansai University, Suita, Osaka 564-8680, Japan
Received 6 September 2006; received in revised form 14 April 2007; accepted 20 May 2007
Available online 24 May 2007
Abstract
Using transmission electron microscopy (TEM) electron irradiation effects were studied for nanogroove-network structured
single-crystalline and dendritic polycrystalline Pt nanosheets 50–60 nm in size. These two nanosheets with nearly the same average
groove-width or dendritic spacing of 1.3–1.4 nm were prepared from the mixed and single surfactant liquid crystalline templates,
respectively. On exposure to electron beam for 20 min at the acceleration voltage of 200 kV, the nanogrooved nanosheets were
morphologically little affected, but the dendritic ones were transformed into less branched polycrystalline structures with spacings
distributed around ꢀ1.7 nm. The shape transformation of the latter occurred by the combined mechanism of segmental migration
and atomic diffusion. These observations indicate that the nanogrooved Pt nanosheets are highly stabilized by the grooved but
crystallographically continuous Pt framework, leading to their extremely high thermo-resistance, in marked contrast to the
polycrystalline dendritic structures constructed of crystallographically discontinuous linkages of nanoblocks.
# 2007 Elsevier Ltd. All rights reserved.
Keywords: A. Metals; C. Electron microscopy; D. Radiation damage
1. Introduction
Nanoscale platinum particles have attracted particular attention because of their potential use as catalysts in various
fields such as petroleum chemical processes, depoisoning of exhaust gas from automobiles, and polymer electrolyte
fuel cells [1]. The catalytic activity of Pt used for these applications depends on the size and the shape of the
nanoparticles including the arrangement of surface atoms [2]. Various methods have been therefore developed for
controlling the size and the shape of Pt nanoparticles [3], and fabricating unconventional Pt nanostructures including
nanorods [4], nanotubes [5], and 2D and 3D mesoporous solids [6,7]. Except for a controlled polyol process [8], most
* Corresponding author. Tel.: +81 985587311; fax: +81 985587320.
E-mail address: t0g102u@cc.miyazaki-u.ac.jp (T. Kijima).
0025-5408/$ – see front matter # 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.materresbull.2007.05.019

T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
1283
of these nanostructured Pt materials were prepared by templating approaches using 2D mesoporous silica or alumina
[4], silver wires [5], 3D mesoporous silica [6], and surfactant molecular assemblies [7]. Many catalytic reactions occur
at elevated temperatures and even in polymer electrolyte fuel cells operated at relatively low temperatures the catalysts
undergo grain growth due to migration of small particles [9]. Not only high catalytic activity but also high thermal
stability free from interparticle aggregation would be therefore desired for Pt nanopartcles as catalysts. El-Sayed et al.
observed the shape transformation of cubic and tetrahedral Pt nanocrystals with about 8 nm in diameter on a carbon
grid under transmission microscope [10]. The particle shapes showed no change at temperatures up to ꢀ350 8C, but a
small truncation developed at temperatures between 350–450 8C until the Pt nanocrsytals underwent explicitly the
shape transformation at about 600 8C much lower than the melting point of bulk Pt (1769 8C), leading to the shape
change and the coalescence of the surfaces of neighboring nanocrystals. Silver nanoparticles of triangle in shape with
the curvature of about 5 nm on mica exhibited the shape transformation at room temperature far lower than that of the
bulk [11]. Diffusion as dislocation of Pt clusters on Pt (1 1 1) was also observed for a series of clusters up to pentamers
at a temperature as low as 350 K [12]. Many other studies have been reported concerning the diffusion of two-
dimensional islands or large clusters on surfaces [13–24]. Electron microscopy was effectively used to observe the
shape transformation or qusimelting of metal nanoparticles under electron irradiation [18–24]. On the other hand, our
recent study demonstrated the synthesis of Pt nanotubes by the reduction of H2PtCl6 confined to lyotropic liquid
crystals (LCs) of polyoxyethlene-type mixed surfactants of polyoxyethylene (20) sorbitan monosteare (Tween 60) and
nonaethylene-glycol (C12EO9) [25]. Furthermore, the reduction of Pt salts confined to the hemicylindricl micelles of
the same polyoxyethlene-type single or mixed surfactants on graphite yielded Pt nanosheets with a uniform thickness
of as thin as 3.5 nm [26] and in situ atomic force microscopy (AFM) also revealed that the Pt nanosheets undergo the
shape transformation and the coalescence in preferred directions at room temperature [27]. More recently,
nanogroove-network structured single-crystalline nanosheets ꢀ3.5 nm thick were synthesized by the reduction of
Na2PtCl6 with NaBH4 using the same mixed surfactant LC templates, although a similar reaction using the Tween 60
based single surfactant LC templates yielded morphologically similar but 2D dendritic polycrystalline Pt nanosheets
[28]. We also found that the nanogrooved Pt nanostructures loaded on carbon exhibit fairly high electrocatalytic
activity for oxygen reduction reaction [28]. It would be thus of importance to characterize the structural stability of
these functionally unique Pt nanostructures obtained in the surfactant templating systems. Here we report the
extremely high resistance of nanogroove-network structured single-crystalline nanosheets to electron irradiation, in
marked contrast to 2D-dendritic polycrystalline nanosheets showing the shape transformation into 2D less branched
polycrystalline ones by the combined mechanism of segmental migration and atomic diffusion within the individual
nanosheets.
2. Experimental
2.1. Materials
Hydrogen hexachloroplatinate(IV) hexahydrate (H2PtCl6Á6H2O) was purchased from Wako Chemical Co. and it
was neutralized with NaOH or LiOH into sodium or lithium hexachloroplatinate (Na2PtCl6 or Li2PtCl6) prior to use.
Nonaethylene-glycol monododecyl ether (C12EO9) and Polyoxyethylene (20) sorbitan monostearate (Tween 60) were
purchased from Wako Chemical Co. and used without further purification. Other chemicals were of reagent grade.
Sodium borohydride (NaBH4; SBH) was also used as received from Wako Chemical Co. The chemical structure of
Tween 60 is shown in Fig. 1.
2.2. Synthesis
In the typical reaction, Na2PtCl6 (7.54 Â 10À4 mol), C12EO9 (7.54 Â 10À4 mol), Tween 60 (7.54 Â 10À4 mol), and
H2O (4.52 Â 10À2 mol) at a 1:1:1:60 molar ratio were mixed in a glass tube of 15 mm in inner diameter at 60 8C for
15 min, and then cooled to 20 8C and allowed to stand at that temperature for 30 min. SBH (7.54 Â 10À4 mol) was
dissolved in water (15.1 Â 10À3 mol) and then added to the precursory pasty mixture to keep at that temperature for
24 h. After the reaction mixture being mixed with water, the resulting solid was separated from the solution containing
unreacted Na2PtCl6 species, washed with water and then with ethanol prior to drying in air. Similar reactions using
Li2PtCl6 instead of Na2PtCl6 and Tween 60 alone instead of C12EO9 and Tween 60 were carried out as above.

1284
T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
Fig. 1. Chemical structure of Tween 60.
2.3. Characterization
All TEM and HRTEM images were taken with a JEOL-2010 microscope operated at 200keV. Small-angle XRD
patterns for pasty samples were measured by the transmission method using Cu Ka radiation on a Shimadzu XRD-
7000 diffractometer equipped with a glass plate of 1 mm thick with a slit of 1 mm wide filled with specimen.
2.4. Electron irradiation
The experiments were carried out as follows: The sample for investigation was placed on a micro-grid (Nisshin EM
Co., 200 mesh) after ultrasonication in ethanol. First, a high resolution TEM image of the area selected for the
subsequent irradiation was taken to characterize the original morphology of sample. Then the sample was irradiated
for a prescribed time up to 20 min with an electron beam generated using the acceleration voltage of 200 kV while the
TEM image of the same area was taken at 5-min intervals.
3. Results and discussion
The resulting products in the Tween 60 based mixed surfactant systems were identified as single-crystalline
nanosheets of ꢀ50 nm in size with a nanogroove-network as described elsewhere [28], according to their transmission
electron microscope (TEM) images (Fig. 2a and b): The continuous atomic fringes with a spacing of 0.23 nm observed
over 20 nm or above confirmed that the elliptic nanoleaves ꢀ2.5 nm wide and 2–4 nm long are connected with their
crystallographic alignment to form a single-crystalline nanosheet with a nanogroove-network (Fig. 2b). Based on the
TEM image, the single-crystalline nanosheets were characterized by the presence of nanogrooves with a narrow
distribution of groove-widths centered at ꢀ1.4 nm (Fig. 2a). Similar nanogroove-network structured single-crystalline
nanostructures were obtained by the reaction using Li2PtCl6 instead of Na2PtCl6 (Fig. 2f). This product showed the
development of nangrooves of as wide as ꢀ1.2 nm with a narrow distribution of groove-widths. On the other hand, the
resulting products in the Tween 60 based single surfactant systems were identified as 2D dendritic polycrystalline
nanosheets of ꢀ 60 nm in size (Fig. 3a–c). The dendritic nanosheets gave an average spacing of ꢀ1.3 nm but with a
broad distribution ranging from 0.8 to 2.4 nm. The crystallographic difference between the single-crystalline and the
polycrystalline nanosheets was also confirmed by their Fourier transform patterns (inset) indicative of a pair of spots
and four or more pairs of spots, respectively.
Solidification from a melt commonly occurs as directional growth of dendritic arrays and a typical freely grown
dendrite forms straightly developed primary stem and sidebranches [29]. In contrast, as seen from the TEM images,
the present nanogrooved and the dendritic nanosheets are formed of meandered branches because the nanosheets were
grown in a LC medium in which the path for platinate and reductant species to diffuse is spatially restricted and
complicated. It is also remarked that the nanogrooved nanosheets exhibit a narrow distribution of groove-widths,
whereas the dendritic nanosheets show not only a dendritic branching pattern (Fig. 3c) but also a wide distribution of
spacings characteristic of dendritic growth [29,30]. This fact suggests that the grooved nanosheets would grow by a
mechanism essentially different from that expected for the other ones. The striking difference between the structural
properties and growth patterns of the nanosheets produced in the Tween 60 based mixed and single surfactant systems
would be arisen from that between their precursory liquid crystalline phases identified as hexagonal and lamellar LCs,

T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
1285
Fig. 2. TEM or HRTEM images of the Pt nanoparticles (a and b) as grown and (c–e) electron irradiated in the lyotropic mixed surfactant Na2PtCl6/
12EO9/Tween 60 system and (f) as grown Pt nanoparticles in the lyotropic mixed surfactant Li2PtCl6/C12EO9/Tween 60 system, together with
C
distributions of groove-widths: irradiation time, (a, b, f) 0 min, (c) 10 min and (e) 15 min. The HRTEM image of the box area in (a) is given in (b).
The Fourier transform patterns taken at the initial and final stages of electron irradiation are given in inlets, showing two spots indicating that the
sheet is single crystalline.
respectively [27,28]. In the 2D-hexagonal array of cylindrical micelles for the former the chemical species or reduced
Pt clusters can move much more easily along the micellar axis than in a direction perpendicular to it, as observed in a
2D array of heimicylindrical micelles of Tween 60 [26]. This might be more favorable for the partial coalescence of Pt
nanoleaves with their crystallographic alignment. In the lamellar LCs of Tween 60 bimolecular layered micelles for
the other nanosheets the doped materials can move freely in the 2D-isotropic space, but reductant species are supplied
only through the edge of the lamellar LCs, resulting in the growth of crystallographically discontinuous dendrites. The
nanogroove-network structured single-crystalline nanosheets produced in the Na–platinate system showed a little
smaller groove-widths than those in the Li–platinate system. This might be also resulted from the structural difference
between their precursory LCs. Fig. 4 shows the XRD patterns of the M2PtCl6/C12EO9/Tween 60/H2O mixtures at a
x:1:1:60 molar ratio for the M = Li and M = Na systems. The major peak and a very weak band near 2u = 2.58 for both
mixtures can be attributed to the 100 and 110 reflections for a hexagonal structure with a = 7–7.4 nm. The peak
intensities of 100 reflection for the Li– and Na–platinate mixtures showed a more marked tendency to decrease with an
increase of x in that order. In accord with these tendencies, the parameter a of 7.4 nm for the precursory Li–platinate
LC phase at x = 1 was larger than 7.3 nm for the Na–platinate phase, probably because the cylindrical micelles in the
former phase containing hydrated Li ions is more expanded than those in the latter containing hydrated Na ions with a
smaller radius. The more expanded space formed by the 2D-hexagonally array of wider cylindrical micelles for the
Li–platinate system might lead to an increase in groove-width.
On electron irradiation for 5 min or more up to 20 min using the acceleration voltage of 200 kV the nanogroove-
network structured single-crystalline Pt nansheets prepared in the Na–platinate system showed no changes in
morphology, as confirmed by the groove-width distributions independent of irradiation time (Fig. 2a and c–f, Fig. 5).

1286
T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
Fig. 3. TEM or HRTEM images of the Pt nanoparticles (a and b) as grown and (d–g) electron irradiated in the lyotropic Tween 60 based single
surfactant Na2PtCl6/Tween 60 system, together with their distributions of spacings: irradiation time, (a and b) 0 min, (d) 5 min, (e) 10 min, (f)
15 min, and (g) 20 min. The HRTEM image of the box area in (a) is given in (b). The Fourier transform patterns taken at the initial and final stages of
electron irradiation are given in inlets. The branched framework of a dendritic nanosheet in (a) is illustrated in (c).
No structural change by electron irradiation was also observed for the nanogroove-network structured nanosheets
obtained in the Li–platinate system. In marked contrast, on exposure to electron beam for 5–20 min the 2D dendritic
polycrystalline nanosheets prepared in the Tween 60 based single surfactant system were gradually coalesced into less
branched dendritic polycrystalline nanosheets: the spacings of the dendritic nanosheets and their distribution were
appreciably increased from ꢀ1.3 nm and 0.8–2.4 nm at irradiation time t = 0 min into ꢀ1.7 and 1–2.6 nm at
t = 20 min, respectively, with an accompanying enrichment of closed spaces (Figs 3 and 5). Fig. 6 shows the
morphological change before and after irradiation as well as the detailed coalescence process of Pt nanoleaves or
islands 3–4 nm wide and 4–5 nm long with crystallographically different orientations in a selected area of the same
dendritic nanosheet sample. In this process, the left part of nanoleaf B is separated from the right part to move to
nanoleaf A, leading to their coalescence at irradiation time t = 10 min (Fig. 6e): at t = 5 min the nanoleaf B is just
about to be separated into two parts (Fig. 6d). Moreover, the coalescence is accompanied by the atomic rearrangement
so that the Pt atoms at the outer part of A are diffused to occupy the empty space located around the center of the
selected area. The round space remained unoccupied was fully occupied with Pt atoms at t = 35 min, resulting in a
single-crystalline leaf with a size of 6–8 nm (Fig. 6h). It is known that the grain growth of Pt clusters or nanoparticles
with different sizes proceeds by Ostwald ripening based on the diffusion of atoms from small to large particles [31] or
by the migration of small particles to coalesce with larger ones [32,33]. The shape transformation of the present 2D
dendritic Pt nanosheets by electron irradiation could be therefore explained by the combined mechanism of the small
leaflet migration and the atomic diffusion based on Ostwald ripening, as schematically shown in Fig. 7B. The model

T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
1287
Fig. 4. XRD patterns of the (A) Li2PtCl6/C12EO9/Tween 60/H2O and (B) Na2PtCl6/ C12EO9/Tween 60/H2O mixtures at a molar ratio of x:1:1:60 for
x = 0, 0.2, 0.4, 0.6, 0.8. and 1.0 (Cu Ka). (C) Plots of lattice parameter a as a function of platinate amount added for the Li–platinate (squares) and
Na–platinate (circles) LC phases.
(C) in Fig. 7 illustrates a decrease of the number of branches and broadening of dendritic spacings resulted from the
structural change of the nanosheets upon electron irradiation. Since the dendritic polycrystalline Pt nanosheets are an
aggregate of Pt nanoparticles with high edge surface and interfacial energies [27], the energy supplied by electron
irradiation would induce the movement of Pt nanoleaves (or islands) or atoms so as to minimize the total surface
energy of the Pt aggregate. The nanometer- or atomic-scale movement may be caused by the atomistic processes such
as evaporation–recondensation of single atoms between the islands and a surrounding two dimensional gas [17], edge
running of atoms around the island perimeter [34] or some other mechanism, although the detailed shape
transformation mechanism of the Pt nanosheets are not obvious at the present stage. On the other hand, the
nanogrooved single-crystalline Pt nanosheets are free from interfacial energy in comparison with the dendritic
nanosheets even if both structures are similar in shape (Fig. 7A), which would raise the thermodynamic stabilities of
the former nanosheets leading to their high resistance to electron irradiation for as long as 20 min.
Fig. 5. Plots of average groove-width and dendritic spacing as a function of irradiation time for the nanogroove-networked single-crystalline (&)
and dendritic polycristalline (*) nanosheets.

1288
T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
Fig. 6. Sequence of TEM images showing a selected area (circle) of a dendritic Pt nanosheet changing from an initial phase to the shape transformed
phase during electron irradiation: Irradiation time, (a and c) 0 min, (d) 5 min, (e) 10 min, (f) 20 min, (g) 30 min, and (b and h) 35 min.
Fig. 7. A schematic illustration of the shape transformation of (A) nanogrooved single-crystalline and (B and C) dendritic polycrystalline nanosheets
during electron irradiation.

T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
1289
Under irradiation of electron beam at an energy of 200 keV heating occurs predominantly through ionization
processes and radiation losses may be ignored [35]. It is therefore believed that the Pt nanoparticles exposed to electron
beam underwent the explicit shape transformation by the heating effect of the beam. Thus, the above observations
clearly indicate that the nanogrooved but single-crystalline Pt nanosheets are thermally much more resistant than the
polycrystalline Pt ones because the former nanostructures are highly stabilized by the grooved but crystallographically
continuous Pt framework, in marked contrast to the polycrystalline branched structures constructed of cry-
stallographically discontinuous linkages of nanoblocks.
4. Conclusions
We demonstrated that nanogroove-network structured single-crystalline Pt nanosheets are morphologically little
affected even by 20-min electron irradiation, whereas dendritic polycrystalline nanosheets exposed to electron beam
are transformed in shape into less branched polycrystalline ones. The shape transformation of the latter nanosheets
proceeds by the combined mechanism of segmental migration and atomic diffusion within the individual nanosheets to
minimize their surface and interfacial energies. The striking difference between the resistances of the nanogroove-
networked single-crystalline and dendritic polycrystalline Pt nanosheets to electron beam are attributed to whether
these two nanostructures are energetically stabilized by their crystallographically continuous Pt framework or
destabilized by their crystallographically discontinuous linkages of nanoblocks. Considering the heating effect caused
predominantly by electron irradiation, the present results reveal that the nanogroove-network structured single-
crystalline Pt nanosheets have extremely high thermo-resistance and hence promising as catalysts for fuel cells and
other applications.
Acknowledgement
This work was supported by Grant-in-Aid for the CREST of Japan Science and Technology Agency (JST).
References
[1] See for example
A. Rouxoux, J. Schulz, H. Patin, Chem. Rev. 102 (2002) 3757.
[2] See for example
(a) H. Arai, T. Seiyama, M. Harakawa, H. Tominaga, Catalysis Deactivation, Elsevier, 1980,, p. 167;
(b) G.A. Somrijai, Catal. Rev. 7 (1972) 87;
(c) R.R. Adzic, A.A.V. Tripkovic, W.E. O’Grady, Nature 196 (1982) 137;
(d) K. Yahikozawa, Y. Fujii, Y. Matsuda, K. Nishimura, Y. Takasu, Electrochim. Acta 36 (1991) 973.
[3] See for example
(a) T.S. Ahmadi, Z.W. Wang, T.C. Green, A. Henglein, M.A. EI-sayed, Science 272 (1996) 1924;
(b) J.M. Petroski, Z.W. Wang, T.C. Green, M.A. EI-sayed, J. Phys. Chem. B 102 (1998) 3316;
(c) T. Teranishi, M. Hosoe, T. Tanka, M.J. Miyake, Phys. Chem. B 103 (1999) 3818;
(d) N. Toshima, K. Nakata, H. Kitoh, Inorg. Chim. Acta 265 (1997) 149;
(e) H. Wakayama, N. Setoyama, Y. Fukushima, Adv. Mater. 15 (2003) 742.
[4] (a) M. Sasaki, M. Osada, N. Sugimoto, S. Inagaki, Y. Fukushima, A. Fukuoka, M. Ishikawa, Micropor. Mesopor. Mater. 21 (1998) 597;
(b) Z. Liu, Y. Sakamoto, T. Ohsuna, K. Hiraga, O. Terasaki, C. Ko, H.J. Shin, R. Ryoo, Angew. Chem. Int. Ed. 39 (2000) 3107;
(c) Y.J. Han, J.M. Kim, G.D. Stucky, Chem. Mater. 12 (2000) 2068;
(d) B.R. Martin, D.J. Dermody, B.D. Reiss, M. Fang, A. Lyon, M.J. Natan, T.E. Mallouk, Adv. Mater. 11 (1999) 1021.
[5] Y. Sun, B. Mayers, Y. Xia, Adv. Mater. 15 (2003) 641.
[6] H.J. Shin, R. Ryoo, Z. Liu, O. Terasaki, J. Am. Chem. Soc. 123 (2001) 1246.
[7] (a) G.S. Attard, C.G. Goltner, J.M. Corker, S. Henke, R.H. Templer, Angew. Chem. Int. Ed. Engl. 36 (1997) 1315;
(b) G.S. Attard, P.N. Bartlett, N.R.B. Coleman, J.M. Elliott, J.R. Owen, J.H. Wang, Science 278 (1997) 838;
(c) P.R. Birlin, J.M. Elioott, Y.E. Watson, Chem. Commun. 17 (2000) 1693.
[8] J. Chen, T. Herricks, M. Geissler, Y. Xia, J. Am. Chem. Soc. 126 (2004) 10854.
[9] M.S. Wilson, F.H. Garzon, K.E. Sickkafus, S. Gottesfeld, J. Electrochem. Soc. 140 (1993) 2872.
[10] Z.L. Wang, J.M. Petroski, T.C. Green, M.A. El-Sayed, J. Phys. Chem. B 102 (1998) 6145.
[11] J.C. Hamilton, Phys. Rev. B 55 (1997) R7402.
[12] T.R. Linderoth, S. Horch, L. Petersen, S. Helveg, E. Lgsgaard, I. Strnsgaard, F. Besenbacher, Phys. Rev. Lett. 82 (1999) 1494.
[13] J.C. Hamilton, M.S. Daw, S.M. Foiles, Phys. Rev. Lett. 74 (1995) 2760.

1290
T. Yoshimura et al. / Materials Research Bulletin 43 (2008) 1282–1290
[14] S.C. Wang, G. Ehrlich, Phys. Rev. Lett. 79 (1997) 4234.
[15] S.C. Wang, U. Ku¨rpick, G. Ehrlich, Phys. Rev. Lett. 81 (1998) 4923.
[16] W. Fan, X.G. Gong, W.M. Lau, Phys. Rev. B 60 (1999) 10727.
[17] J.M. Wen, S.L. Chang, J.W. Burnett, J.W. Evans, P.A. Thiel, Phys. Rev. Lett. 73 (1994) 2591.
[18] S. Iijima, T. Ichihashi, Phys. Rev. Lett. 56 (1986) 616.
[19] P. Williams, Appl. Phys. Lett. 50 (1987) 1760.
[20] P.M. Ajayan, L.D. Marks, Am. Phys. Soc. 60 (1988) 585.
[21] P.M. Ajayan, L.D. Marks, Phys. Rev. Lett. 63 (1989) 279.
[22] V.G. Gryaznov, A.M. Kaprelov, A.Y. Belov, Phi. Mag. Lett. 63 (1991) 275.
[23] N. Doraiswamy, L.D. Marks, Surf. Sci. 348 (1996) L67.
[24] S.A. Nepijko, M. Klimenkov, H. Kuhlenbeck, H.J. Freund, J. Vac. Sci. Technol. A 17 (2) (1999) 577.
[25] T. Kijima, T. Yoshimura, M. Uota, T. Ikeda, D. Fujikawa, S. Mouri, S. Uoyama, Angew. Chem. Int. Ed. Engl. 43 (2004) 228.
[26] H. Kawasaki, M. Uota, T. Yoshimura, D. Fujikawa, G. Sakai, M. Annaka, T. Kijima, Langmuir 21 (2005) 11468.
[27] H. Kawasaki, G. Sakai, T. Kijima, Appl. Surf. Sci. 253 (2006) 1512.
[28] G. Sakai, T. Yoshimura, S. Isohata, M. Uota, H. Kawasaki, T. Kuwahara, D. Fujikawa, T. Kijima, Adv. Mater. 19 (2007) 237.
[29] L. Makkonen, J. Crystal Growth 208 (2000) 772.
[30] M.E. Glicksman, M.B. Koss, E.A. Winsa, Phys. Rev. Lett. 73 (1993) 573.
[31] J.A. Bett, K. Kinoshita, P. Stonehart, J. Catal. 41 (1974) 124.
[32] C.G. Granqvist, R.A. Buhrman, J. Catal. 42 (1976) 477.
[33] K.F. Burton, H.R. Kunz, D.R. Rutt, Electrochim. Acta 23 (1978) 183.
[34] A.F. Voter, Phys. Rev. B 34 (1986) 6819.
[35] L. Reimer, Transmission Electron Microscopy: Physics of Image Formation and Microanalysis Springer Series in Optical Science, Springer,
Berlin, 1997, p. 563ff.