Water-soluble CdS quantum dots prepared from a refluxing single precursor in aqueous solution

Article abstract of DOI:10.1021/jp0470849

The first aqueous preparation of luminescent CdS QDs from a single precursor is reported. These water-soluble CdS quantum dots with tunable sizes between 25 and 40 Aì? were produced from a readily prepared [(2,2a?2-bipyridine)Cd(SC{O}Ph)₂] complex by simply refluxing in an aqueous solution. The as-prepared nanoparticles are fairly uniform in size without the need of size sorting, and exhibit a quantum confinement effect. The size of the nanoparticles is tunable by varying the capping agent-to-precursor molar ratio. The green- and yellow-emitting quantum dots obtained via this simple route are water soluble and have OH functionality that is suitable for further applications. The particle aging kinetics was investigated by monitoring the optical band edge absorption and was found to follow the Ostwald ripening mechanism. When the amount of capping agent used is 10 mmol, unprecedented dimerization of some QDs upon prolong reflux to form nanocrystals double the size is detected. Thus, a common red shift of the absorption band edge is not detected and the growth of larger QDs seems to arise from the coalescence of two QDs.

Full text of DOI:10.1021/jp0470849

J. Phys. Chem. B 2004, 108, 18569-18574
18569
Water-Soluble CdS Quantum Dots Prepared from a Refluxing Single Precursor in Aqueous
Solution
Z. H. Zhang, W. S. Chin,* and J. J. Vittal*
Department of Chemistry, National UniVersity of Singapore, 3 Science DriVe 3, Singapore 11753, Singapore
ReceiVed: July 2, 2004; In Final Form: September 18, 2004
The first aqueous preparation of luminescent CdS QDs from a single precursor is reported. These water-
soluble CdS quantum dots with tunable sizes between 25 and 40 Å were produced from a readily prepared
[(2,2′-bipyridine)Cd(SC{O}Ph)2] complex by simply refluxing in an aqueous solution. The as-prepared
nanoparticles are fairly uniform in size without the need of size sorting, and exhibit a quantum confinement
effect. The size of the nanoparticles is tunable by varying the capping agent-to-precursor molar ratio. The
green- and yellow-emitting quantum dots obtained via this simple route are water soluble and have OH
functionality that is suitable for further applications. The particle aging kinetics was investigated by monitoring
the optical band edge absorption and was found to follow the Ostwald ripening mechanism. When the amount
of capping agent used is 10 mmol, unprecedented dimerization of some QDs upon prolong reflux to form
nanocrystals double the size is detected. Thus, a common red shift of the absorption band edge is not detected
and the growth of larger QDs seems to arise from the coalescence of two QDs.
Introduction
growth of QDs in our proposed aqueous method, we have
investigated the growth kinetic using absorption spectroscopy.
During the past decade, nanostructured semiconductor materi-
als have been extensively studied due to their unique physico-
chemical and optical properties.1-5 Reliable methods which can
reproducibly give a high yield of uniformly sized nanocrystals
remain an active research goal in nanomaterials chemistry.6-12
In some cases, chemical manipulation toward both size- and
shape-tunable preparation is successfully achieved.6-8,11
Recently, injection of reagents into a hot coordinating solvent
such as tri-n-octylphosphine oxide (TOPO), at high temperatures
(200-400 °C), was used as an efficient route to prepare semi-
conductor quantum dots (QDs). For example, single precursors
dithio- and diselenocarbamate complexes were injected into hot
TOPO to synthesize various ME (M ) Cd, Zn, Pb; E ) S, Se,
Te) QDs.12 In all these reports, the crystallite size and size dis-
tribution of these TOPO-capped nanocrystals are very critically
dependent on the proper control of injection and reaction
temperature.13 The QDs obtained with this approach are only
soluble in nonpolar organic solvents, unless their surfaces are
further modified. This surface modification, however, often
increases the effective hydrodynamic radii of the QDs. For
biological14,15 and environmental applications,16 it is often
desirable for the prepared nanocrystals to be water soluble.
In this paper, we describe a simple method to prepare
monodispersed CdS QDs from a molecular precursor using
water as solvent. While there have been many reports on the
synthesis of CdS nanoparticles in water,17-25 as far as we know,
this is the first report of preparing CdS QDs in aqueous solution
directly from a single molecular precursor. Our approach
produces water-soluble QDs in relatively high yields, and the
particle size and its emitting color can be tuned easily by
changing the feed ratio between capping agent and the precursor.
The QDs prepared are fairly uniform in size, without the need
of size-sorting such as for those carried out in some earlier
reports on water-soluble preparations.17,19,23 To understand the
Experimental Section
A. Chemicals. 1-Thioglycerol (98%, Fluka), thiobenzoic acid
(95%, Fluka), ether, ethanol, chloroform (J. T.Baker), sodium
bicarbonate (Dumont), acetonitrile, cadmium acetate (98%,
Aldrich), 2-propanol (Merck), sodium hydroxide (99%, Dick-
son), and 2,2-bipyridine (99%, Aldrich) were used as received.
B. Preparation of [(2,2′-bpy)Cd(SC{O}Ph)2]. Sodium
bicarbonate (0.8 g) was dissolved in 20 mL of water and 30 mL
of acetonitrile, 0.7 mL of thiobenzoic acid was then added, and
the solution was stirred for 5 min. Subsequently, 0.468 g of
2,2′-bipyridine (bpy) and 0.8 g of cadmium acetate were added.
The mixture was stirred for 30 min and filtered. The crystals ob-
tained were washed with ethanol, dried, and recrystalized from
chloroform and ether. The compound was fully characterized
by NMR, microanalysis, IR spectroscopy, and thermogravim-
etry. Single crystals were grown by slow diffusion of methanol
into a DMSO solution of the compound. Crystal data: triclinic
space group, P1h; a ) 8.6754(4) Å, b ) 11.5986(6) Å, c )
11.7670(6) Å; R ) 76.238(1)°, â ) 75.934(1)°, γ ) 84.019(1)°;
Z ) 2; V ) 1114.1(1) Å3; F ) 1.618 g/cm3; R1 ) 0.0268; wR2
) 0.0643 for 280 parameters and 4923 data with Fo > 4σ(Fo).
C. Synthesis of CdS Nanoparticles in Aqueous Solution.
1-Thioglycerol (10 mmol) was dissolved in deionized water (25
mL) and the solution was adjusted to pH ca. 11 with 1 M sodium
hydroxide. The solution was placed in a three-necked flask fitted
with a septum and valve, then a certain amount of the precursor
[(2,2′-bpy)Cd(SC{O}Ph)2] was added into this solution. Sub-
sequently, the solution was deaerated with N2 bubbling and
refluxed at ca. 100 °C for 15-30 min. The solution turned from
turbidity to clear, and from colorless to pale yellow, indicating
the decomposition of precursor and the generation of CdS nano-
particles. Three batches of CdS nanoparticles were synthesized
with 0.25, 2.5, and 10 mmol of thioglycerol, respectively.
For kinetic studies, aliquots were taken from the refluxing
mixture at different time intervals and UV/vis spectra were
recorded immediately. Alternatively, the nanoparticles can be
* Address correspondence to this author. W.S.C.: e-mail chmcws@
nus.edu.sg; phone 65-6874-8031; fax 65-6779-1691. J.J.V.: e-mail
chmjjv@nus.edu.sg.
10.1021/jp0470849 CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/10/2004

18570 J. Phys. Chem. B, Vol. 108, No. 48, 2004
Zhang et al.
Figure 1. Structure of the precursor. Selected bond distances (Å) and
angles (deg): Cd(1)-S(1), 2.476(1); Cd(1)-S(2), 2.502(1); Cd(1)-
N(1), 2.331(2); Cd(1)-N(2), 2.339(2); Cd(1)‚‚O (1), 2.780(2); Cd(1)‚‚O
(2), 2.631(2); S(1)-Cd(1)-S(2), 129.93(2); N(1)-Cd(1)-N(2), 70.45(7).
isolated by first precipitation with the addition of 2-propanol,
followed by centrifugation and washing with ethanol and
acetone. The vacuum-dried yellow to pale yellow powder can
be used for XRD and TEM measurements or redissolved in
water for other studies.
Figure 2. FTIR transmission spectra of (a) thioglycerol, (b) the
precursor, and (c) a typical CdS QDs sample.
D. Characterization. Thermogravimetric analysis (TGA) was
recorded on a SDT 2960 Simultaneous DTA-TGA. Ap-
proximately 10 mg of the precursor was measured under a flow
of inert N2 gas (flow rate 90 mL/min) at a heating rate of 10
deg/min. IR spectra were recorded with an FTS 165 Bio-Rad
FTIR spectrophotometer in the range 4000-400 cm-1 on KBr
pellets or Nujol mulls. UV-vis absorption spectra were obtained
with a Shimadzu UV-2550 spectrophotometer. Photolumines-
cence (PL) spectra were collected with a Perkin-Elmer LS55
luminescence spectrometer, using a pulsed Xe lamp. Powder
X-ray diffraction (XRD) patterns of the nanoparticles were
recorded on a Siemens D5005 X-ray powder diffractometer with
Cu Ka radiation (40 kV, 40 mA). The powdered sample was
mounted on a sample holder and scanned with a step size of
2θ ) 0.01° in the range 20-60°. Transmission electron
micrographs (TEM) were obtained with a 300 kV Philips TEM
microscope. The samples were prepared by placing a drop of
the nanoparticles dispersed in acetone on a copper grid with
carbon film and were allowed to dry in desiccators.
When the precursor is decomposed into CdS and other
fragments in the basic refluxing solution, the surrounding
1-thioglycerol capping molecules readily chemisorbed onto the
growing particles and impeded further growth. The amount of
capping agent present in the reaction mixture was found to
directly control the size of QDs produced (vide infra). While
thiol groups are commonly used as capping agents, most of the
other colloidal or sol-gel preparations in aqueous solutions
produce nanoparticles with wide size distributions or low
product yields. The yield of our preparation is about 60-70%.
The prepared QDs are fairly monodispersed, and the results
shown in this paper were obtained directly without any size-
selective precipitation.
Figure 2 compares the IR spectra of 1-thioglycerol and the
precursor with an as-prepared CdS QDs sample. 1-Thioglycerol
(Figure 2a) has characteristic peaks at ∼2560 and 3400 cm-1
,
which are assigned to the SH and OH stretching vibrations, re-
spectively. The precursor (Figure 2b) has many vibration bands
in the range of 1500-1000 cm-1, mainly arising from the
stretching and bending modes of the bipyridyl group. The IR
spectrum of the as-prepared QDs in Figure 2c showed major
bands due to 1-thioglycerol, except for the SH stretching band
at ∼2360 cm-1. While the bipyridyl group was reported to also
act as a coordinating capping group,12a our IR analysis did not
indicate any sign of the bipyridyl group. The disappearance of
the -SH vibration of 1-thioglycerol clearly indicates that this
reagent is chemisorbed onto the QDs, and that the OH functional
groups remain intact on the particles. The QDs are thus hydro-
philic and water soluble, and the OH functionalities are available
on the surface for further chemical modification or attachment.
B. Morphology and Characterization of the QDs. High-
resolution TEM (HRTEM) images as shown in Figure 3 suggest
that spherical particles are obtained and the average sizes of
the QDs depend on the amount of capping agent used in the
preparation. Generally, a higher amount of capping agent
produces smaller QDs. From the HRTEM analysis, the size
distribution of our QDs is estimated as ca. (10%. This is
comparable to earlier reports on the preparation of CdS QDs in
aqueous solution after size-sorting precipitation.17,22-23 Clear
crystalline lattices of the QDs are observed in Figure 3d,
showing the 3.3 Å separated (111) planes extending over ∼30
Å. The side fringes are less prominent indicating incomplete
crystallization at the edges.
Result and Discussion
A. Formation of CdS Nanoparticles from the Precursor.
The precursor [(2,2′-bpy)Cd(SC{O}Ph)2] we used is an air-
stable crystalline complex, of which the molecular structure has
been determined by X-ray crystallography as illustrated in Figure
1. In this mononuclear compound, the Cd atom is in a highly
distorted tetrahedral geometry if weak interactions between Cd
and the carbonyl oxygen atoms are neglected.
TGA of the precursor in a nitrogen environment showed
weight loss in the region 200-240 °C corresponding to the for-
mation of CdS. Hence, in solvents such as TOP and TOPO, the
compound was found to decompose to CdS when the tempera-
ture is raised above 200 °C. In the aqueous preparation reported
here, nevertheless, the precursor readily decomposed in water
by refluxing at 100 °C in the presence of a base and 1-thiol-
glycerol. Bases were normally added in the preparation of thiol-
capped nanoparticles to facilitate the -S- capping.17-19,23 We
have found that the precursor did not decompose without the
base, even after refluxing for a long period in aqueous 1-thiol-
glycerol solution. Thus the basic condition has facilitated the
breaking down of the precursor, possibly via an initial attack
of the base onto the carbonyl carbon, which thus weakens the
C-S bond.

Water-Soluble CdS Quantum Dots
J. Phys. Chem. B, Vol. 108, No. 48, 2004 18571
Figure 3. HRTEM images of various CdS QDs prepared by refluxing the precursor for 15 min, using various amounts of 1-thiolglycerol: (a) 0.25
mmol, average particle size ) 37 Å; (b) 2.5 mmol, average particles size ) 30 Å; and (c) 10 mmol, average particle size ) 26 Å. (d) Higher
magnification image showing a single CdS nanocrystal with (111) lattices.
Figure 5. Optical absorption and PL spectra of CdS QDs prepared by
refluxing the precursor for 15 min, using various amounts of 1-thiolg-
lycerol: (a) 10, (b) 2.5, and (c) 0.25 mmol. Excitation wavelength for
the PL spectra is 380 nm.
Figure 4. XRD patterns for CdS QDs prepared by refluxing the
precursor for 15 min, using various amounts of 1-thiolglycerol: (a)
10, (b) 2.5, and (c) 0.25 mmol.
to ascertain whether it is a pure cubic or a mixture of cubic and
The wide-angle XRD patterns for the different QDs prepared
are shown in Figure 4. Diffraction peaks at 2θ ) 26.2°, 43.3°,
and 51.2° are clearly observed for samples b and c, and can be
readily assigned to the (111), (220), and (311) planes of bulk
cubic CdS. This is often the observed phase of CdS prepared
with thiols as the capping group.23,26,27 For the smallest QDs
shown in Figure 4a, diffraction peaks between 40° and 50° seem
to have merged into a broad peak. This is similarly observed in
other reports for very small CdS QDs, and it is often difficult
hexagonal phases.17,23 Applying the Debye-Scherrer formula28
using the width of (111) peaks, we estimated the average sizes
of the QDs as 17, 23, and 31 Å for samples a, b, and c,
respectively. These values follow the same trend, but are
consistently smaller than those obtained from TEM analysis.
This could be due to incomplete crystallization of our QDs at
the edges and surfaces.
In Figure 5, the optical absorption and photoluminescence
(PL) spectra of the various QDs prepared are presented. The

18572 J. Phys. Chem. B, Vol. 108, No. 48, 2004
Zhang et al.
al. studied the growth of ZnO nanoparticles in colloidal
suspensions and found that the growth kinetics followed the
Lifshitz-Slyozov-Wagner (LSW) theory for Ostwald ripening,
i.e., large particles grow at the expense of dissolution of smaller
ones.39 Peng et al. reported a “focusing and defocusing” process
for the growth of II-VI and III-V colloidal semiconductor
prepared by using injection method.13 They found that a critical
size (r*) existed at equilibrium and played an important role to
control the size distribution of the produced nanocrystals. When
the monomer concentration was depleted, r* became larger and
the size distribution broadened (“defocusing”). The injection
of additional monomer decreased r* to a smaller value, which
resulted in the process of “focusing”. Our method, while using
a single precursor, is slightly different from this study in the
sense that our precursor is continuously heated from room
temperature until the solution refluxes. Thus the supply of
monomers will be continuous throughout the heating period until
the precursor is depleted.
To follow the growth kinetics of QDs, it is necessary to
determine particle sizes as a function of time. In most kinetics
studies, particle size inferred from absorption onset was
conveniently used since good agreement with TEM analysis
could often be found.39,40 In Figure 7, two typical time evolutions
of the absorption band edge during our preparation are presented.
It should be pointed out that t ) 0 in the plots refers to the
onset of reflux, i.e., after continuous heating from room
temperature for ∼10-15 min. During this heating period, the
reaction mixture will slowly turn from a turbid to a clear
yellowish solution, indicating the onset of decomposition of the
insoluble precursor. The absorption band edge is already obvious
at t ) 0, suggesting that some nanoseeds are formed at the onset
of reflux. The absorption edge progressively shifts to higher
wavelength with time, indicating the gradual growth of these
QDs in size.
Figure 6. Schematic energy transfer diagram for the PL emission of
sample a.
sharp absorption edges and the well-defined excitonic features
exhibited confirm the narrow distribution of particle sizes. The
absorption onsets are dependent on the average sizes: 370 (sam-
ple a), 410 (sample b), and 425 nm (sample c), which correspond
to band gaps of 3.35, 3.02, and 2.92 eV, respectively. These
onsets clearly show a blue-shift relative to that of bulk CdS at
515 nm (2.42 eV) due to the quantum size confinement effect.
As shown in Figure 5, the PL emission maximum is also
red-shifting with increasing particle size; the λmax is respectively
510 (sample a), 560 (sample b), and 630 nm (sample c). It is
noted that all these samples emit at similar intensity, but the
emission peak width of sample c is relatively broader. A broad
emission in the 500-700-nm region is often detected for CdS
nanocrystals, and has been attributed to recombination of trapped
electrons/holes in some surface defect states.29-33 For example,
S-(s) trap states, which were located at about 1.5 eV above the
valence band, may be formed when photogenerated holes
migrate to the surface of CdS clusters.32 V+ defect states, on
S
the other hand, are formed by electrons trapped in anionic
vacancies and located about 0.7 eV below the conduction band
of CdS.33 In our samples, the emission λmax are red-shifted by
about 0.8-1.0 eV from their respective excitonic absorption
edges. One would thus expect the emission to arise from energy
transfer within the forbidden band gap as shown schematically
in Figure 6. It is commonly known that an improvement in
crystallinity is beneficial to luminescence.34 Since our TEM and
XRD results have indicated incomplete crystallization of the
QDs, lattice defects are expected and the emission peaks are
broadened. Indeed we believe that in sample c, which was
prepared by using the least amount of capping agent, substitu-
tional defects may also be present and thus give rise to an overall
broader band.35
At a fixed precursor concentration, the size-dependent growth
rate can be obtained by considering the Gibbs-Thompson
equation:13,39-41
2γVm
S ) S exp
(1)
(
)
r
b
rRT
Here, Sr and Sb are the solubility of the particles and the bulk
solid respectively, γ is the interfacial energy, r is the radius of
the particles, Vm is the molar volume of the materials, R is the
gas constant, and T is the temperature. When diffusion is the
rate-limiting step, the Gibbs-Thompson expression can be
inserted into Fick’s first law of diffusion and the growth law
based on the LSW model42,43 is obtained:
The effective mass approximation is commonly used to
estimate particle size from absorption onset of semiconductor
QDs.36-38 In the strong-confinement regime, the confinement
energy of the first excited electronic state can be approximated
by the Brus equation:
h2
1
1
1.8e2
4πꢀꢀ0r
bulk
E ) Eg
+
+
-
2
(
)
m0me* m0mh*
rav3 - rav,o3 ) Kt
(2)
8r
Here, rav is the average particle radius, rav,o is the average initial
particle radius, t is the reaction time, and K is a constant related
to the diffusion coefficient (D):39
Here, E is the absorption onset inferred from the UV spectra;
bulk
Eg
is the bulk band gap; h is Planck’s constant; r is the
particle radius; me* and mh* are the effective mass of electrons
and holes respectively; m0 and e are the mass and charge of a
free electron, respectively; ꢀ0 is the permittivity of free space;
and ꢀ is the relative permittivity. In this case, the particle size
of CdS QDs can be calculated by taking Egbulk ) 2.42 eV, ꢀ )
5.7, me* ) 0.19 and mh* ) 0.80.36 Thus, our prepared particle
diameters are calculated as 28, 33, and 35 Å, for sample a, b,
and c, respectively. These values are thus consistent with those
obtained from TEM analysis.
2
8γDVm Sb
K )
(3)
9RT
The equation thus predicts that the cubic term of rav is linearly
dependent on t. In Figure 8, we plot the cubic term of rav inferred
from the absorption band edges with time and the linear
relationship obtained thus confirms that the growth of CdS QDs
follows the LSW theory. The constant K value is calculated as
0.147((5) × 10-3 nm3/s for sample a and 0.128((5) × 10-3
nm3/s for sample b. The slight decrease of K in sample b agrees
C. Growth Kinetics of CdS QDs. While many reports
focused on the synthesis and characterization of QDs, relatively
few studies have reported on their growth kinetics. Searson et

Water-Soluble CdS Quantum Dots
J. Phys. Chem. B, Vol. 108, No. 48, 2004 18573
Figure 7. Time evolution for the absorption spectra of CdS QDs prepared with (a) 0.25 and (b) 2.5 mmol of 1-thioglycerol. Time ) 0 refers to
the onset of reflux of the reaction mixture.
Figure 8. Plot of (particle radius)3 versus time for CdS QDs prepared
with different amounts of 1-thioglycerol: 2.5 (stars) and 0.25 mmol (dots).
Figure 9. Time evolution for the absorption spectra of CdS QDs
prepared with 10 mmol of 1-thioglycerol. Prolong reflux shows the
growth of a new peak at longer wavelength.
with the expectation that the presence of more capping agents
will impede diffusion through the interfacial boundaries for
coalescence to occur.
growth model for the formation of CdS QDs in our reaction.
First, the precursor is decomposed and monomers (denoted as
Cd-S) are formed in the mixture. Some monomers come
together to form nanoseeds initially (denoted with a subscript x
or y) during reflux. The reaction is then followed by a
competition between coarsening and dissolution (Oswald ripen-
ing), depending on whether the crystal size is smaller or larger
than the equilibrium critical size (r*). The competition mech-
anism may be shown as follows:
When the amount of capping agent was increased to 10 mmol,
however, we observed an interesting phenomenon. Thus, as
shown in Figure 9, the onset peak now remains at the same
position instead of red-shifting for the first 1 h of reflux. From
there on an additional peak is observed at higher wavelength
and gradually grows in intensity. Thus after 5 h of reflux, the
absorption spectrum indicated two distinct peaks: one at 367
nm (3.38 eV) and another at 467 nm (2.66 eV). When these
band gap values are fitted into the Brus equation, two average
particle sizes, 23 and 48 Å, respectively, are obtained. We have
found that these two distinctly sized QDs can be easily separated
by size-selective procedures and the individual absorption spec-
trum can be measured when the separated QDs are re-dispersed
in water. In Figure 10, the individual spectrum is compared with
the original composite spectrum obtained after 5 h of reflux.
It is interesting to note that the larger QDs are about twice
the size of the smaller ones, and that the later practically does
not grow in size upon reflux. Such a phenomenon does not occur
for samples prepared with a lower amount of thioglycerol. Thus,
based on the above results, we can propose the following kinetic
m(Cd-S) + n(Cd-S) f
(CdS)x + (CdS)y + (m + n - x - y)(Cd-S)
(CdS)x f (CdS)x-1 + (Cd-S)
(CdS)y + (Cd-S) f (CdS)y+1
x < r* < y
Peng et al. has proposed that r* is the critical radius for which
the solubility of the nanocrystals is exactly equal to the concen-
tration of the monomers (i.e. the growth rate is zero).13 We have
found that effectively r* is inversely dependent on the amount

18574 J. Phys. Chem. B, Vol. 108, No. 48, 2004
Zhang et al.
(5) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Annu. ReV. Mater.
Sci. 2000, 30, 545.
(6) Manna, L.; Scher, E. C.; Alivisatos, A. P. J. Am. Chem. Soc. 2000,
122, 12700.
(7) (a) Peng, X. G.; Manna, L.; Yang, W. D.; Wickham, J.; Scher, E.;
Kadavanich, A.; Alivasatos, A. P. Nature 2000, 404, 59. (b) Yu, M. W.;
Peng, X. G. Angew. Chem., Int. Ed. 2002, 41, 2368. (c) Peng, Z. A.; Peng,
X. G. J. Am. Chem. Soc. 2001, 123, 183.
(8) (a) Jun, Y.-W.; Koo, J.-E.; Cheon, J. Chem. Commun. 2000, 1243.
(b) Jun, Y.-W.; Choi, C.-S.; Cheon, J. Chem. Commun. 2001, 101. (c) Lee,
S. M.; Jun, Y.-W.; Cho, S.-N.; Cheon, J. J. Am. Chem. Soc. 2002, 124,
11244. (d) Jun, Y.-W.; Jung, Y.-Y.; Cheon, J. J. Am. Chem. Soc. 2002,
124, 615. (e) Jun, Y.-W.; Lee, S. M.; Kang, N. J.; Cheon, J. W. J. Am.
Chem. Soc. 2001, 123, 5150.
(9) Joo, J.; Na, H. B.; Yu, T.; Yu, J. H.; Kim, Y. W.; Wu, F.; Zhang,
J. Z.; Hyeon, T. J. Am. Chem. Soc. 2003, 125, 11100.
(10) (a) Yan, P.; Xie, Y.; Qian, Y. T.; Liu, X. M. Chem. Commun. 1999,
1293. (b) Wang, W. Z.; Geng, Y.; Yan, P.; Liu, F. Y.; Xie, Y.; Qian, Y. T.
J. Am. Chem. Soc. 1999, 121, 4062.
(11) Some other examples on size and shape control: (a) Lee, S. M.;
Cho, S.-N.; Cheon, J. AdV. Mater. 2003, 15, 441. (b) Li, Y. C.; Li, X. H.;
Yang, C. H.; Li, Y. F. J. Mater. Chem. 2003, 13, 2641. (c) Peng, Z. A.;
Peng, X. G. J. Am. Chem. Soc. 2001, 124, 3343. (d) Barrelet, C. J.; Wu,
Y.; Bell, D. C.; Lieber, C. M. J. Am. Chem. Soc. 2003, 125, 11498.
(12) (a) Trindade, T.; O’Brein, P.; Zhang, X. Chem. Mater. 1997, 9,
523. (b) Ludolph, B.; Malik, M. A.; O’Brein, P.; Revaprasadu, N. Chem.
Commun. 1998, 1849. (c) Trindade, T.; O’Brien, P.; Pickett, N. L. Chem.
Mater. 2001, 13, 3843. (d) Nair, P. S.; Radhakrishnan, T.; Revaprasadu,
N.; Kolawole, G. A.; O’Brien, P. Chem. Commun. 2002, 564.
(13) Peng, X. G.; Wickham, J.; Alivisatos, A. P. J. Am. Chem. Soc.
1998, 120, 5343.
(14) Bruchez, M.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P.
Science 1998, 281, 2013.
(15) Chan, W. C. W.; Nie, S. M. Science 1998, 281, 2016.
(16) Mills, A.; Lehunte, S. J. Photochem. Photobiol. A 1997, 1, 108.
(17) Vossmeyer, T.; Katsikas, L.; Giersig, M.; Popovic, I. G.; Diesner,
K.; Chemseddine, A.; Eychmuller, A.; Weller, H. J. Phys. Chem. 1994,
98, 7665.
(18) Rockenberger, J.; Troger, L.; Kornowski, A.; Vossmeyer, T.;
Eychmuller, A.; Feldhaus, J.; Weller, H. J. Phys. Chem. B 1997, 101, 2691.
(19) Kho, R.; Torres-Martinez, C. L.; Mehra, R. K. J. Colloid Interface
Sci. 2000, 227, 561.
(20) Sondi, I.; Siiman, O.; Koester, S.; Matijevic, E. Langmuir 2000,
16, 3107.
(21) Sapra. S.; Nanda, J.; Sarma, D. D.; El-Al, F.; Hodes, G. Chem.
Commun. 2001, 2188.
(22) Qi, L. M.; Colfen, H.; Antonietti, M. Nano Lett. 2001, 1, 61.
(23) Rogach, A. L.; Kornowski, A.; Gao, M. Y.; Eychmuller, A.; Weller,
H. J. Phys. Chem. B 1999, 103, 3065.
(24) Barglik-Chory, C.; Buchold, D.; Schmitt, M.; Kiefer, W.; Heske,
C.; Kumpf, C.; Fuchs, O.; Weinhardt, L.; Stahl, A.; Umbach, E.; Lentze,
M.; Geurts, J.; Muller, G. Chem. Phys. Lett. 2003, 379, 443.
(25) Chen, Y. F.; Rosenzweig, Z. Anal. Chem. 2002, 74, 5132.
(26) Van Dijken, A.; Janssen, A. H.; Smitsmans, M. H.; Vanmaekel-
bergh, D.; Meijerink, A. Chem. Mater. 1998, 10, 3513.
(27) Nanda, J.; Sapra, S.; Sarma, D. D.; Chandrasekharan, N.; Hodes,
G. Chem. Mater. 2000, 12, 1018.
(28) Azaroff, L. V. X-ray diffraction; McGraw-Hill: New York, 1974.
(29) Chestnoy, N.; Harris, T. D.; Hull, R.; Brus, L. E. J. Phys. Chem.
1986, 90, 3393.
(30) Anderson, M. A.; Gorer, S.; Penner R. M. J. Phys. Chem. B 1997,
101, 5895.
Figure 10. Comparison between the absorption spectrum for the
separated QDs (solid) and the composite spectrum after 5 h of reflux
in Figure 9 (dotted).
of capping agent present, since the amount of capping agent will
affect the interfacial diffusion and hence the solubility of the
nanocrystals. There seems to exist a saturation amount of cap-
ping agent, S*, for the reaction. When the amount of capping
agent is >S*, the diffusion constant D will be lowered to near
zero, and then according to eqs 2 and 3, rav ) rav,o, i.e., the nano-
crystal practically does not grow once formed. During prolonged
reflux, some of these smaller QDs will probably collide with
each other to form dimers: (CdS)y + (CdS)y f (CdS)2y.
Since the dimers will have a lower surface-to-volume ratio
compared to the smaller QDs, we expect some capping
molecules are released into the reaction mixture in this process.
While the smaller QDs are more or less matured (i.e. not
growing) in the refluxing mixture, it will be interesting to
investigate modifications/reactions that could be further carried
out on these QDs. More work is now underway in this direction.
Conclusion
We have demonstrated a facile route to synthesize water-
soluble CdS QDs using a new single-source precursor. The pre-
cursor is air-stable and relatively easy to prepare and purify.
The produced QDs are fairly uniform and the particle size can be
adjusted by either controlling the capping agent-to-precursor
ratio or stopping the reaction at the appropriate time. The growth
of these QDs was found to follow Oswald ripening during the
reflux process. When the amount of capping agent used exceeds
some saturation value, an unprecedented dimerization of QDs
is detected. The first-grown QDs are found to have ceased grow-
ing except merging with one another to form QDs of double the
size. Two distinctly sized nanoparticles can thus be isolated after
prolonged reflux. This phenomenon is interesting as further mod-
ification of the QDs is now possible at elevated temperatures.
(31) Banin, U.; Bruchez, M.; Alivisatos, A. P.; Ha, T.; Weiss, S.;
Chemla, D. S. J. Chem. Phys. 1999, 110, 1195.
(32) Ginlet, D. S. Photoelectrochemistry and Electrosynthesis of Semi-
conductor Materials; Electrochemical Society: Pennington, NJ, 1988.
(33) Vuyesteke, A. A.; Sihonen, Y. T. Phys. ReV. 1959, 113, 40.
(34) Kitai, A. H. Solid State Luminescence: Theory, Materials and
DeVices; Hapman & Hall: New York, 1993.
(35) Misawa, K.; Yao, H.; Hayashi T.; Kobayashi, T. Chem. Phys. Lett.
1991, 183, 113.
(36) Brus, L. E. J. Chem. Phys. 1984, 80, 4403.
Acknowledgment. This research work was supported by a
National University of Singapore research grant (grant no.
R-143-000-167-112). We thank Ms. Chow Shue Yin in IMRE
for the HRTEM measurements.
(37) Brus, L. E. J. Phys. Chem. 1986, 90, 2555.
(38) Steigerwald, M. L.; Brus, L. E. Acc. Chem. Res. 1990, 23, 183.
(39) Wong, E. M.; Bonevich, J. E.; Searson, P. C. J. Phys. Chem. B
1998, 102, 7770.
(40) Curri, M. L.; Agostiano, A.; Manna, L.; Monica, M. D.; Catalano,
M.; Chiavarone, L.; Spagnolo, V.; Lugara, M. J. Phys. Chem. B 2000, 104,
8391.
References and Notes
(1) Bawendi, M. G.; Steigerwald, M. L.; Brus, L. E. Annu. ReV. Phys.
Chem. 1990, 41, 477.
(2) Tolbert, S. H.; Alivisatos, A. P. Annu. ReV. Phys. Chem. 1995,
46, 595.
(41) Ratke, L.; Voorhees, P. W. Growth and coarsening; Springer:
Berlin, Germany, 2002.
(42) Lifshitz, I. M.; Slyozov, V. V. J. Phys. Chem. Solids 1961, 19, 35.
(43) Wagner, C. Z. Elecktrochem. 1961, 65, 581.
(3) Beecroft, L. L.; Ober, C. K. Chem. Mater. 1997, 9, 1302.
(4) Linsebigler, A. L.; Lu, G. Q.; Yates, J. T. Chem. ReV. 1995, 95, 735.