A novel metalorganic route for the direct and rapid synthesis of
monodispersed quantum dots of indium phosphide
Mark Green and Paul O’Brien*
Department of Chemistry, Imperial College of Science Technology and Medicine, South Kensington, London,
UK SW7 2AY. E-mail: p.obrien@ic.ac.uk: Fax No., 0171-594-5825
Received (in Cambridge, U.K.) 14th August 1998, Accepted 2nd October 1998
Nanometric particles of InP are readily prepared by the
nanoparticles resulted. At this point, addition of excess non-
solvent produces a nanodispersed powder of the semiconductor.
Electronic spectroscopy of this material showed a band edge of
1.92 eV, as determined by the direct band gap method13 with an
excitonic shoulder at ca. 2.72 eV (455 nm), a significant blue
shift from the bulk band gap of 1.27 eV.14 Luminescence
spectroscopy shows strong broad near band edge luminescence
with lmax. at 2.32 eV (534 nm) (Fig. 1) slightly red shifted from
the absorption. Size fractionation has little effect on band edge,
indicating the system is highly monodispersed. The sample
consists of a large number of relatively monodispersed dots,
with an average diameter of 7.24 ± 1.24 nm which is easily
observed in TEM experiments (Fig. 2).
t
decomposition of the complex In(PBu ) at 167 °C in
2
3
4
-ethylpyridine; the resulting materials show marked quan-
tum confinement effects, and was investigated using optical
absorption and photoluminescence spectroscopies, and
transmission electron microscopy.
There is considerable current interest in the synthesis of
compound semiconductors as isolated particles with dimen-
1
sions of the order of nanometers. Such materials are small
enough to show quantum confinement effect and the electronic
properties of the material depend on the size of the particles.
The majority of work to date has concerned on II–VI materials2
which are for many reasons easier to prepare than III–V or II–V
This work clearly shows that the use of defined precursors
has potential in the preparation of high quality quantum dots.
Indeed, the presently reported procedure is both simpler and
3
semiconductors. Initial attempts to prepare III–V materials
focused on the reaction of separate sources in a high boiling
point solvents4 or an electrospray method.5 However, the
properties of the materials were somewhat disappointing.
Recently better quality materials have been prepared by
thermolysis reactions in TOPO (tri-n-octylphosphine oxide), an
adaptation of the highly efficient route first described by
Murray et al. for the preparation of CdSe from dimethylcad-
2
mium and tri-n-octylphosphine selenide. These methods for
III–V TOPO capped quantum dots were initially developed by
6,7
Micic et al. and have been further exploited by Alivisatos and
co-workers8 and materials with near band edge luminescence
have been prepared. However, the methods used fairly ill-
,9
defined precursor systems such as an aged solution of InCl
3
–
TOPO in reaction with neat E(SiMe (E = As, P). Growth and
3 3
)
annealing to form the final crystalline material can take up to a
week (at 250 °C) and the resulting product is polydispersed and
contaminated with In
We have recently succeeded in using the phosphide com-
2 3
O waste material
t
10
pound [MeCdPBu
2
]
3
for the preparation of high quality
11
samples of nanocrystalline Cd
3
P
2
,
and have now developed
the use of related compounds for the synthesis of III–V
materials. We report here, the use of a single molecule precursor
in the ‘one pot’ preparation of nanometric InP quantum dots.
The problems encountered in the synthesis of III–V dots can
Fig. 1 Luminescence (a) and UV (b) spectra of Q-InP synthesised at
1
67 °C. Feature at ca. 2 eV is a second order excitation (lexc = 300 nm).
3
in part be attributed to covalent nature of the semiconductor
and the related effect of strong precursor–solvent interactions.
In these systems, nucleation and growth tend to be high
temperature processes and temporal separation of the two is
difficult, the resulting products are hence often polydispersed
and amorphous. The use of a reactive single-source precursor
overcomes this problem.
The single source precursor In(PBut
)
was prepared as
described by Jones and coworkers, by reacting InCl with 3
in hexane. Decomposition to InP was
effected by reflux in 4-ethylpyridine (20 ml) for 0.5 h (167 °C,
.4 g, 0.7 mmol).
2
3
3
t
12
equiv. of LiPBu
2
0
heat
In(PBut
)
––– –? Q-InP
2 3
Addition of a non-solvent (light petroleum) resulted in floccu-
lation of the InP quantum dots. The powder was then re-
dispersed in either pyridine or 4-ethylpyridine and centrifuged
to remove waste material, an optically clear solution of
Fig. 2 TEM of Q-InP dispersed on a copper grid: (A) bar = 200 nm,
(B) bar = 20 nm.
Chem. Commun., 1998, 2459–2460
2459
Green, Mark
