356
D.E. Mainwaring et al. / Solid State Communications 140 (2006) 355–358
by passage through two columns containing anhydrous CaSO4
desiccant (Rowe Scientific Pty Ltd, Australia). Precautions
were taken to avoid the escape of excess H2S into the
atmosphere by attaching an aqueous CuSO4 bubbler to the exit
of the reaction flask.
TiCl4 was dissolved in a degassed ternary tridodecyl-
methylammonium iodide (TDAI)/hexanol/octane (8/8/84% by
weight) inverse micelle solution, at concentrations between
8 × 10−4 to 1 × 10−3 M. Following H2S addition, the mixture
was allowed to stir for 1 h, then the inverse micelle solution
containing the TiS2 nanoclusters was extracted with an equal
volume of acetonitrile. This dissolved the inverse micelles, with
the hexanol and TDAI entering the acetonitrile phase, while the
nanoclusters partitioned between the nonpolar octane phase and
the polar acetonitrile/TDAI/hexanol phase.
Fig. 1. The structure of a TiS cluster [2].
2
manifolds of individual, delocalized electronic states. The
energy shift of the band edge, as well as the separations between
conduction and valence band states, can be larger than 1 eV. In
addition to these delocalized electronic states, semiconductor
nanoclusters can also have localized electron and hole traps
that often accumulate at the nanocluster surface, and the large
surface-to-volume ratio can result in a high density of trap
states. The presence of a high density of localized states along
with finite relaxation rates between localized and delocalized
states can affect the observed nanocluster optical properties
The preparation of such nanoparticles can be achieved using
micelle structures, which are able to create a reaction template
in the appropriate size range. It is well known that size control
during colloidal growth requires the distinct separation of
nucleation from growth. If nucleation occurs continuously, in
either space or time, a wide range of cluster sizes will inevitably
occur. However, if confinement of the reactant precursors can
be achieved to limit growth to a restricted region of space, a
narrow size distribution of clusters will occur.
2.2. Characterization
High-resolution transmission electron microscopy (HRTEM)
using Philips CM 200 FEG microscope has been used to study
the microstructure of particles. Fluorescence measurements
were carried out on a Perkin-Elmer LS-50B.
3. Results and discussion
According to electronic energy calculations, bulk TiS2 is a
semiconductor with a band gap of about 2 eV and a smaller
indirect gap of 1.4 eV [12]. In agreement with the predictions,
the experimentally measured electronic spectrum of bulk TiS2
exhibits two main absorption bands between 0 and 3.5 eV [13,
14]. The low energy band is centred around 2.3 eV (550 nm)
with a width of 1.5 eV and can be ascribed to electronic
transitions from the upper p-bands into the lower triplet of non-
bonding d-bands followed by a window due to the separation
between the lower and upper non-bonding d bands.
Comparison of the spectrum of the clusters to that of the
bulk crystal reveals a large blueshift for the clusters. The
absorption peak at 360 nm in the cluster spectrum is associated
with the excitonic peak at 550 nm of the bulk, corresponding
to a blueshift of 0.85 eV arising from the quantum-sized
confinement effect of the TiS2 clusters.
Surfactant molecules can form segregated droplet-like
aggregates called inverse micelles whose dimensions are
typically 1–10 nm. The modifier ‘inverse’ refers to the negative
curvature between the surfactant–oil interface (the oil is the
continuous medium). Nucleation only occurs in the micelle
interior because of the total lack of solubility of charged species
in the low dielectric constant inert oils used as the continuous
2. Experimental
2.1. Sample preparation
Owing to the presence of I− (from the TDAI) during
the synthesis, it was possible that the 360 nm peak was
due to molecular species rather than nanoclusters, specifically
I−3 /I−5 . Two results indicate that the peak is due to TiS2
nanoclusters rather than I−3 /I−5 . Firstly, the blank spectrum
(synthesis, except without the TiCl4) shows no absorption
at wavelengths >290 nm in either the acetonitrile or octane
phases. Secondly I−3 /I5− is insoluble in the octane phase [15].
These results indicate that while it is possible to incur spectral
contamination in the acetonitrile phase, this does not occur in
the octane phase and the absorption is due exclusively to TiS2
nanoclusters.
The following reagents were used as received: titanium
tetrachloride (TiCl4, Aldrich), iron(II) sulphide 99% (Sigma,
Australia), concentrated hydrochloric acid (32% w/w) (Asia
Pacific Speciality Chemicals Limited, Australia), tridodecyl-
methyl ammonium iodide (TDAI). Titanium tetrachloride and
TDAI were stored in a glovebox.
Typical sample volumes were 10–20 mL. High-purity
solvents and reagents are necessary and all solvents and
reagents were rigorously purified. Octane (Aldrich) was
distilled from sodium, acetonitrile (Aldrich) was distilled from
phosphorous pentoxide and hexanol was distilled from iodine-
activated magnesium.
Fig. 2 shows the optical-absorption spectra of TiS2
nanoclusters dispersed in octane, and the fluorescence emission
spectrum in both the octane and acetonitrile phases. The
fate of photogenerated electron–hole pairs in a semiconductor
Dry gaseous H2S was used as the sulphidizing agent: this
was obtained by reaction of solid FeS with HCl (32% w/w
conc.) and, to prevent water contamination, the gas was dried