Synthesis of CdSe and CdTe nanocrystals without precursor injection

Article abstract of DOI:10.1002/anie.200502279

With high quality and in high yield: The synthesis presented here affords CdSe and CdTe semiconductor nanocrystals without the need of a precursor injection. It allows the detailed control of the size and shape of the nanocrystals, as can be seen from the tetrahedral CdSe nanocrystals that have been prepared (TEM image). The method is suitable for industrial-scale preparations. (Figure Presented)

Full text of DOI:10.1002/anie.200502279

Communications
Semiconductors
of nanocrystals in the synthesis, which results in products with
poor monodispersity. However, if the precursors are too
stable, a very small number of nuclei may form, which leads to
uncontrollable particle growth. (Note that suitable nucleation
initiators could lead to a synthesis of monodisperse nano-
crystals in this case.) Therefore, when suitable nucleation
initiators are not available, only precursors with suitable
reactivity can result in monodisperse nanocrystals. In general,
the ideal precursors should exhibit no (or very low) reactivity
below the desired temperature for growth of the chosen
nanocrystals, but they should also exhibit high reactivity when
the temperature reaches the desired point.
DOI: 10.1002/anie.200502279
Synthesis of CdSe and CdTe Nanocrystals without
Precursor Injection**
Yongan Andrew Yang, Huimeng Wu,
Kathryn R. Williams, and Y. Charles Cao*
With the growing interest in applications based on nanoscale
[
1]
materials comes a need for an industrial-scale synthesis of
colloidal semiconductor nanocrystals with uniform size and
shape and well-confined surface passivation. To date, the most
successful and widely used nanocrystal synthesis relies on
To test this hypothesis, we first designed a noninjection
synthesis for making CdSe nanocrystals. We chose 2408C as
the growth temperature, which was in accordance with the
lesson learned from the injection-based synthesis of CdSe
[2,3]
[2,3]
rapid precursor injection.
However, it is difficult to scale
nanocrystals.
At this growth temperature, pure cadmium
up such a synthesis for making nanocrystals in large quantities
myristate is a suitable cadmium precursor because it has a
decomposition point at about 2268C. Selenium powder can
[
4]
[6]
(
e.g., tens to hundreds of kilograms). Therefore, there is a
need to develop nanocrystal synthesis without precursor
injection.
be an effective selenium precursor for two reasons: 1) it has a
[
6]
melting point at 2218C; 2) it is insoluble in octadecene
(ODE) at room temperature and becomes slightly soluble at
temperatures over 1908C. As selenium has low solubility in
ODE, it is unreactive with cadmium myristate up to 2008C. In
a typical synthesis (see Experimental Section for detailed
conditions), small particles (i.e., nuclei) suddenly appeared at
2108C (Figures 1a and b). As the nanocrystals grew, the
number of particles quickly increased to a maximum, and
then during further growth it continued to decrease. Signifi-
cantly, about 72% of the particles formed in the initial stage
disappeared after two hours of growth, while the size
distribution of the nanocrystals continued to become more
narrow throughout the synthesis (Figure 1c–e). Such size
Although many publications have described the synthesis
[
5]
of semiconductor nanocrystals without precursor injection,
only a few of them have led to nanocrystals with very good
[
5a–c]
monodispersity.
However, these nanocrystals exhibit
optical properties inferior to those of nanocrystals produced
by the injection method. They normally do not show as many
multiple exciton absorption peaks, which are critical for
nanocrystal applications in advanced optical and electronic
devices. Recently, we reported a non-injection-based syn-
[
4]
thesis of CdS nanocrystals. The sample quality (i.e., size
distribution and optical properties) is as good as that of the
best CdS nanocrystals synthesized by using precursor injec-
tion. Nucleation initiators are important to the CdS synthesis,
but unfortunately, such nucleation initiators are not readily
available for making nanocrystals with other compositions.
Herein, we report how to design a noninjection synthesis for
making high-quality CdSe and CdTe nanocrystals without
nucleation initiators.
The formation of high-quality nanocrystals is often
favored at high temperatures (e.g., > 2008C). This factor
creates a major challenge for making monodispersed nano-
crystals through a noninjection synthesis because such a
synthesis involves a period of increasing temperature over a
broad range (e.g., from room temperature to over 2008C). If
the reactivity of a precursor is too high, this broad change of
temperature often leads to concurrent nucleation and growth
[*] Dr. Y. A. Yang,H. Wu,Dr. K. R. Williams,Prof. Y. C. Cao
Department of Chemistry
University of Florida
Gainesville,FL 32611 (USA)
Fax: (+1)352-392-0588
E-mail: cao@chem.ufl.edu
[
**] Y.C.C. acknowledges the University of Florida and the ACS
Petroleum Research Fund for financial support. We thank Charles R.
Martin and Willard W. Harrison for helpful discussions,and Kerry
Siebein for technical assistance.
Figure 1. Temporal evolution of a) the fluorescence (f) spectrum,
b) the absorption spectrum,c) the diameter ( D) of the nanocrystals,
d) (and inset) the concentration (C) of the nanocrystals,and e) full
width at half-maximum (FWHM) of the fluorescence spectrum during
the CdSe synthesis. The size and concentration of the nanocrystals
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
[
2b,3b]
were obtained according to a reported procedure.
6
712
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6712 –6715

Angewandte
Chemie
focusing is not likely to arise from diffusion-limited growth
because this growth does not lead to a decreasing number of
[
2b]
nanocrystals. Additional investigation is required to under-
stand the detailed mechanism involved in this phenomenon.
The addition of oleic acid during the synthesis is important
to control the quality of the nanocrystals. Without oleic acid,
the nanocrystals show absorption spectra with a tail of higher
intensity at the low-energy end after their diameter reaches
around 3.0 nm and with further growth, they show broad
absorption spectra (Figure 2, dotted lines). Such spectral
Figure 2. Absorption spectra of CdSe nanocrystals with (solid lines)
and without (dotted lines) oleic acid a) after 16 min of growth and
b) after 60 min of growth.
Figure 3. a) XRD pattern of 4.0-nm-diameter ZB CdSe nanocrystals.
The standard diffraction peak positions and relative intensities of bulk
ZB CdSe are indicated. The inset shows an HRTEM image
(5 nm”5 nm) of the nanocrystal sample. b) Absorption (black) and
emission (gray) spectra of the ZB CdSe sample. c) Absorption (black)
and emission (gray) spectra of a wurtzite (W) CdSe nanocrystal
sample made by chalcogen precursor injection according to a reported
procedure. d,e) TEM images of ZB (d) and W (e) CdSe nanocrystals.
Scale bar: 10 nm.
changes are normally attributed to broadening of the particle
size distribution due to Ostwald ripening. However, in this
synthesis we found that the phenomenon was caused by
particle aggregation as it could be reversed by adding a small
amount of oleic acid into the growth solution (Figure 2, solid
lines). After this addition, nanocrystal growth turned out to
be extremely stable. The size distribution of the nanocrystals
further narrowed with particle growth (Figure 1), and the
narrow size distribution was maintained for at least 16 h at the
reaction temperature.
[3a]
those of small wurtzite nanocrystals made by Bawendi and co-
[
2a]
workers. In contrast, the noninjection method presented in
this paper can also result in larger CdSe nanocrystals, which
allow us to unambiguously identify ZB nanocrystals.
Without size sorting, typical CdSe nanocrystals from this
synthesis show sizes with a standard deviation (s) of less than
Indeed, zinc-blende and wurtzite CdSe nanocrystals can
be easily distinguished by their optical properties (Figure 3b–
e). For both cases, nanocrystals with an identical position of
the absorption peak for the first exciton transition band are
nearly identical in size. The Stokes shifts of the emission
peaks are nearly identical and PL quantum yields of band-gap
emissions are similar. However, the fine structures in their
absorption spectra are different. The gap between the first
(1S 1S ) and second (2S 1S ) exciton transition peaks for
5
% (see Figure 3d) and a quantum yield of photolumines-
cence (PL) of about 30 to 40%. Interestingly, these nano-
crystals have a zinc-blende (ZB) structure but not a wurtzite
(
W) structure. Although these two crystal phases produce
similar X-ray powder diffraction (XRD) results with nano-
crystals, the XRD pattern of 4.0-nm particles shows unam-
biguous evidence that the particles have a ZB structure
(
Figure 3a): a) the valley between the the (220) and (311)
3
/2
e
3/2
e
peaks is deep; b) the (400) peak is at 61.08; c) the (620) peak is
at 106.28. In the case of wurtzite crystals, there are no peaks in
ZB particles is clearly wider than that for W particles. The
peak positions for the exciton transitions with higher energies
also vary for these two types of CdSe nanocrystals. These
differences in the absorption spectra originate from the
different electronic band structures at the center of the
[
2a,4,6]
these positions.
with the high-resolution transmission electron microscopy
HRTEM) observation (JEOL-JEM 2010 operated at
00 kV): a square cross-fringe pattern represents an ordered
This structural assignment is consistent
(
2
[
8]
Brillouin zone for these two types of crystal phases.
distance of 0.22 nm, which corresponds to the lattice spacing
of the (220) faces in ZB CdSe (Figure 3a, inset). Such a cross-
fringe pattern does not exist in a wurtzite structure. Note that
ZB CdSe nanocrystals have been made recently by an
injection-based synthesis. However, this synthesis only led
to high-quality nanocrystals in very small sizes, and these
small nanocrystals show nearly identical XRD patterns to
To further demonstrate the power of noninjection syn-
thesis, we synthesized nonspherical CdSe nanocrystals. In the
case of a ZB structure, nonspherical particles are normally
cubes, tetrahedral pyramids, or triangular prisms, and these
particles have a higher surface energy than spherical par-
[
7]
[
9]
ticles. Therefore, making nonspherical particles requires
growth conditions with a high chemical potential, which we
Angew. Chem. Int. Ed. 2005, 44, 6712 –6715
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6713

Communications
thought could be achieved by increasing the per-particle
[
6]
concentration of the precursors. To create such conditions,
we chose a method to decrease the number of initial nuclei
while leaving the precursor concentration unchanged.
Because more reactive precursors can lead to fewer
[
4]
nuclei, we used tributylphosphine selenide (TBPSe) in
place of selenium powder for making CdSe nanocrystals.
Indeed, the number of nuclei at the initial stage in this
synthesis decreased by 75% as compared with the synthesis
with selenium powder; after 30 min, tetrahedral particles with
edge length of 5.0 nm and s of 7% were formed. The high
uniformity of these nanocrystals allows the formation of
ordered, triangular close-packed nanocrystal assemblies on
TEM grids (Figure 4).
Figure 5. a) Temporal evolution of the absorption spectrum in CdTe
nanocrystal synthesis. b) TEM image of CdTe nanocrystals with a first
absorption peak at 625 nm.
Experimental Section
Cadmium myristate was prepared by an ex situ method: a solution of
cadmium nitrate in methanol (0.05m, 40 mL) was added to a solution
of sodium myristate in methanol (0.025m, 240 mL). The resulting
white precipitate was washed twice with methanol and dried under
vacuum overnight to remove all solvents. Note that the purity of
cadmium myristate is important to the following CdSe nanocrystal
syntheses.
Synthesis of spherical CdSe nanocrystals: selenium powder
(
(
0.05 mmol, 100 mesh from Aldrich) and cadmium myristate
0.1 mmol) were added into a three-neck flask with ODE (5.0 g).
À2
The mixture was degassed for 10 min under vacuum ( ꢀ 3 ” 10 Torr)
at room temperature. Under a flow of Ar, the solution was stirred and
À1
heated to 2408C at a rate of 25Kmin . Serial quantitative aliquots
(
[
6]
50 mL) were taken for kinetic studies, and the time was counted as
zero when the temperature reached 2408C. After the particle
diameter had reached about 3.0 nm (i.e., the first absorption peak
at 550 nm), a solution of oleic acid in ODE (0.05m, 1.0 mL) was added
dropwise into the reaction solution to stabilize the growth of the
nanocrystals.
Figure 4. TEM image of tetrahedral CdSe nanocrystals.
The synthesis of tetrahedral CdSe nanocrystals followed a similar
protocol as above, but TBPSe was used instead of selenium powder
(
cadmium myristate: 0.1 mmol, and TBPSe in ODE: 0.05 mmol of
Moreover, the principle for designing a noninjection
synthesis can be applied to the synthesis of CdTe nanocrystals.
The key point is to find suitable precursors. The reactivity of
cadmium myristate with tellurium powder in ODE is too low,
whereas with tributylphosphine telluride (TBPTe) it is too
high. Neither case resulted in the formation of uniform
nanocrystals. We then decreased the reactivity of the cad-
mium precursor by replacing cadmium myristate with cad-
mium octadecylphosphonate. The reaction between cadmium
octadecylphosphonate and TBPTe just perfectly fits the need
for growth of monodisperse CdTe nanocrystals. A typical
synthesis led to CdTe nanocrystals with a size distribution of
less than 5%, which exhibit up to five exciton absorption
bands (Figure 5). Such particle quality is at least comparable
with that of the best particles made by the injection
TBPSe in a mixture of tributylphosphine and ODE, 0.012g and 0.5 g,
respectively ).
In the synthesis of CdTe nanocrystals, cadmium octadecylphosph-
onate was made in situ: a mixture of CdO (0.1 mmol), octadecyl-
phosphonic acid (0.22 mmol), and ODE (5.0 g) was heated to 3008C
to give a colorless solution. The solution was cooled to room
temperature, and TBPTe (0.20 mmol in 0.19 g of tributylphosphi-
ne, ) was added. The mixture was degassed, then stirred and heated
to 2408C (25Kmin ) under a flow of Ar. Note that adding oleic acid
is not needed during the CdTe synthesis.
[
2]
[
2a]
À1
Received: June 29, 2005
Published online: September 27, 2005
Keywords: crystal growth · fluorescence spectroscopy ·
.
nanotechnology · self-assembly · semiconductors
[
3b]
synthesis.
In summary, we reported the synthesis of high-quality
CdSe and CdTe nanocrystals without precursor injection. This
noninjection synthesis allows detailed control of nanocrystal
size and shape. In addition, because it exhibits easily control-
lable nanocrystal-growth kinetics, such a synthesis is very
important for a large-scale industrial preparation of CdSe and
CdTe nanocrystals in which the rates of heating and cooling of
reaction solutions are limited.
[
1] a) A. P. Alivisatos, Nat. Biotechnol. 2004, 22, 47; b) H. Weller,
Philos. Trans. R. Soc. London Ser. A 2003, 361, 229; c) M. Han, X.
Gao, J. Z. Su, S. Nie, Nat. Biotechnol. 2001, 19, 631; d) M. Kazes,
D. Y. Lewis, Y. Ebenstein, T. Mokari, U. Banin, Adv. Mater. 2002,
14, 317; e) L. Zhu, M. Zhu, J. K. Hurst, A. D. Q. Li, J. Am. Chem.
Soc. 2005, 127, 8968.
6
714
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6712 –6715

Angewandte
Chemie
[
2] a) C. B. Murray, D. J. Norris, M. G. Bawendi, J. Am. Chem. Soc.
993, 115, 8706; b) X. Peng, J. Wickham, A. P. Alivisatos, J. Am.
1
Chem. Soc. 1998, 120, 5343.
[
3] a) L. Qu, X. Peng, J. Am. Chem. Soc. 2002, 124, 2 049; b) W. W.
Yu, L. Qi, W. Guo, X. Peng, Chem. Mater. 2003, 15, 2854; c) R.
Xie, U. Kolb, J. Li, T. Baschꢀ, A. Mews, J. Am. Chem. Soc. 2005,
127, 7480.
[
[
4] Y. C. Cao, J. Wang, J. Am. Chem. Soc. 2004, 126, 14336.
5] a) S. L. Cumberland, K. M. Hanif, A. Javier, G. A. Khitrov, G. F.
Strouse, S. M. Woessner, C. Yun, Chem. Mater. 2002, 14, 1576;
b) N. Pradhan, S. Efrima, J. Am. Chem. Soc. 2003, 125, 2050;
c) D. V. Talapin, S. Haubold, A. L. Rogach, A. Kornowski, M.
Haase, H. Weller, J. Phys. Chem. B 2001, 105, 2260; d) Y. Wang,
N. J. Herron, J. Phys. Chem. 1991, 95, 525.
[
[
6] See the Supporting Information.
7] M. B. Mohamed, D. Tonti, A. Al-Salman, A. Chemseddine, M.
Chergui, J. Phys. Chem. B 2005, 109, 10533.
[
[
8] S. Ninomiya, S. Adachi, J. Appl. Phys. 1995, 78, 4681.
9] a) R. Jin, Y. C. Cao, E. Hao, G. S. Mꢀtraux, G. C. Schatz, C. A.
Mirkin, Nature 2003, 425, 487; b) F. Kim, S. Connor, H. Song, T.
Kuykendall, P. Yang, Angew. Chem. 2004, 116, 3759; Angew.
Chem. Int. Ed. 2004, 43, 3673.
Angew. Chem. Int. Ed. 2005, 44, 6712 –6715
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6715