X. Zheng et al. / Materials Research Bulletin 44 (2009) 216–219
217
2. Experimental
The detailed efficient solvothermal procedure to InP hollow
spheres has been reported elsewhere [19]. Here, it is a simple
description for microwave-assisted process. The microwave
synthesis of InP microspheres was carried out in microwave
synthesis system (WX-4000, maximal temperature ꢀ240 8C and
maximal pressure ꢀ40 atm). Typical process is as follows:
InCl3Á4H2O (1.00 g), HAuCl4 ethanol solution (2 ml, 70.2 mM), P4
(0.32 g) and KBH4 (1.60 g) were placed in the microwave reaction
kettle, with ethylenediamine as solvent. Then the autoclave was
closed and put into the microwave synthesis system for ca. 30 min
under 180–220 8C and the microwave power was 600 W. When the
reactions were finished, the products were cooled to room
temperature naturally. Then the products were filtrated and first
washed with dilute hydrochloric acid (0.1 M) to remove residual In
and absolute ethanol three times. Finally the products were dried
in the vacuum oven at 50 8C. The InP products synthesized through
traditional solvothermal procedures are designated as T-InP, and
the samples synthesized by the microwave-assisted reaction are
designated as M-InP.
The final products were characterized by various techniques. X-
ray powder diffraction (XRD) was carried out on a Rigaku D/max rA
˚
a radiation (l = 1.54178 A). The
X-ray diffractometer with Cu K
scan rate of 0.058/s was applied to record the pattern in the 2
u
range of 20–808. The morphology and size of as-prepared products
were observed by scanning electronic microscopy (SEM) images,
which were performed on an X-650 scanning electronic micro-
analyzer.
3. Results and discussions
Fig. 1a is the X-ray powder diffraction (XRD) pattern of the
samples T-InP after HCl treatment, which is consistent with the
bulk InP reflection (JCPDS file No. 32-452) indicating the same
cubic zinc blende lattice structure. Fig. 1c is the corresponding XRD
pattern of M-InP, which is not obviously different from Fig. 1a.
Contrasts with the XRD pattern Fig. 1a and c, the products have
higher crystallinity under solvothermal condition than that under
microwave-assisted route. Here, Au catalyst component cannot be
detected in the XRD patterns, indicating its content is less than the
resolution limit of XRD. The presence of In before the HCl
treatment has been testified by XRD as shown in Fig. 1b for T-InP
samples and Fig. 1d for M-InP samples.
SEM images of the obtained samples after HCl treatment are
displayed in Fig. 2. Fig. 2a and b give the panorama of T-InP
micrometer hollow spheres under different magnifications.
Plentiful collapsed and hemispherical hollow spheres are also
found in close-up of Fig. 2b, indicating the hollow structure of the
spheres. Fig. 2c,d display the images of M-InP micrometer spheres.
Inserted in Fig. 2c is the typical collapsed hollow spheres for M-InP
samples. From the given SEM images, it can be seen that the
obtained hollow spheres from the two methods have nearly similar
Fig. 1. The XRD patterns (a) for T-InP samples and (c) for M-InP samples after HCl
treatment; the XRD patterns (b) for T-InP samples and (d) for M-InP samples
without HCl treatment.
elevated higher than the eutectic temperature. As reported, it is
impossible to produce 1D III–V structure at a lower reaction
temperature of 300 8C via the In–Au alloy catalysis route in
solution [21]. In the process of our experiment, the Au
nanoparticles are first reduced by the KBH4 from the HAuCl4
colloid. The indium is also reduced by the KBH4. The reaction
temperature (180–220 8C) is surprisingly lower than that of the
melting point of bulk gold and higher than that of the melting point
of In and P4 (In: 157 8C; P4: 44.1 8C). Such a low temperature is not
enough to activate the gold particles to catalyze the growth of one-
dimensional InP [4,5]. Due to the physical tendency to reach the
lowest energy level, the newly In liquid coat on Au nanoparticles to
form In/Au core/shell droplets which are used as the templates of
building the InP nanoparticles [19] rather than the formation of
Au–In alloy to direct the growth 1D structure [4,5,21]. Under the
thermal or microwave irradiation conditions, the reaction between
P4 molecules and In is on the surface of the In/Au droplets.
Thereafter, the formed InP nanoparticles undergo the solidification
to form compact InP layer on the surface of the Au/In core/shell
droplets, which block the further reaction of P4 with In molecules
in the beads. As a result, when removing the unreacted In by
diluted HCl solution, the inner Au cores separate from the
outermost InP shells, and finally produce InP hollow spheres.
Due to the loss of the support, some collapsed hollow spheres are
formed. In addition, the possible density gradient in the sealed
autoclave could lead to the ununiformity for the diameter of as-
grown InP hollow spheres. Further experiments are going on to
size, namely, ca. 3–6
mm in diameter and ca. 1 mm in wall
thickness. The openings and collapse are probably due to the
decomposition of templates by the washing-up of diluted HCl. In
addition, Fig. 2b and d gives further fine information that the
building blocks of InP hollow spheres are InP nanoparticles and the
porous surface nature. Carefully compared the samples, one can
find that there are obviously porous surface natures for T-InP
relative to the compact surface of M-InP samples.
It is reported that the noble metal Au with the in situ formed
alloy species functions as catalysts for the nucleation and further
directs the growth of 1D nanostructure via the VLS [4,5] or SLS
mechanism [21]. However, the reaction temperature must be