Full-color emission from In2S3 and In 2S3:Eu3+ nanoparticles

Article abstract of DOI:10.1021/jp048107m

New observations on the luminescence of In₂S₃ and europium-doped In₂S₃ nanoparticles show a green (510 nm) emission from In₂S₃ and In_1.8Eu _0.2S₃ nanoparticles while a blue (425 nm) emission is observed from In_1.6Eu_0.4S₃ nanoparticles. Both the blue and green emissions have large Stokes shifts of 62 and 110 nm, respectively. Excitation with longer-wavelength photons causes the blue emission to shift to a longer wavelength while the green emission wavelength remains unchanged. The lifetimes of both the green and blue emissions are similar to reported values for excitonic recombination. When doped with Eu³⁺, in addition to the broad blue and green emissions, a red emission near 615 nm attributed to Eu³⁺ is observed. Temperature dependences on nanoparticle thin films indicate that with increasing temperature, the green emission wavelength remains constant, however, the blue emission shifts toward longer wavelengths. Based on these observations, the blue emission is attributed to exciton recombination and the green emission to Indium interstitial defects. These nanoparticles show full-color emission with high efficiency, fast lifetime decays, and good, stability; they are also relatively simple to prepare, thus making them a new type of phosphor with potential applications in lighting, flat-panel displays, and communications.

Full text of DOI:10.1021/jp048107m

J. Phys. Chem. B 2004, 108, 11927-11934
11927
Full-Color Emission from In2S3 and In2S3:Eu3+ Nanoparticles
Wei Chen,*,† Jan-Olov Bovin,‡ Alan G. Joly,§ Shaopeng Wang,† Fuhai Su,¶ and Guohua Li¶
Nomadics, Inc., 1024 South InnoVation Way, Stillwater, Oklahoma 74074, Department of Materials Chemistry,
Chemical Center, UniVersity of Lund, S-22100, Lund, Sweden, Pacific Northwest National Laboratory,
P.O. Box 999, Richland, Washington 99352, and National Laboratory for Superlattices and Microstructures,
Institute of Semiconductors, Chinese Academy of Sciences, P.O. Box 912, Beijing 100083, China
ReceiVed: May 2, 2004; In Final Form: June 2, 2004
New observations on the luminescence of In2S3 and europium-doped In2S3 nanoparticles show a green (510
nm) emission from In2S3 and In1.8Eu0.2S3 nanoparticles while a blue (425 nm) emission is observed from
In1.6Eu0.4S3 nanoparticles. Both the blue and green emissions have large Stokes shifts of 62 and 110 nm,
respectively. Excitation with longer-wavelength photons causes the blue emission to shift to a longer wavelength
while the green emission wavelength remains unchanged. The lifetimes of both the green and blue emissions
are similar to reported values for excitonic recombination. When doped with Eu3+, in addition to the broad
blue and green emissions, a red emission near 615 nm attributed to Eu3+ is observed. Temperature dependences
on nanoparticle thin films indicate that with increasing temperature, the green emission wavelength remains
constant, however, the blue emission shifts toward longer wavelengths. Based on these observations, the blue
emission is attributed to exciton recombination and the green emission to Indium interstitial defects. These
nanoparticles show full-color emission with high efficiency, fast lifetime decays, and good stability; they are
also relatively simple to prepare, thus making them a new type of phosphor with potential applications in
lighting, flat-panel displays, and communications.
1. Introduction
attention.6 However, much less work has been performed on
III-VI semiconductor nanoparticles such as In2S3.7-12 One of
Luminescent materials or phosphors can be found in a broad
range of everyday applications such as cathode ray tubes,
projection television screens, fluorescent tubes, X-ray detectors,
and biomedical probes.1 Efforts over the last three decades have
led to phosphor materials that operate close to their physical
limits.1 Thus, it cannot be expected that properties such as
quantum efficiency and spectral energy distribution will be
significantly improved or that distinctly better materials will
be found in the near future. However, the present phosphors
have some shortcomings that need to be addressed. For example,
Zn2SiO4:Mn2+ (willemite) is used as a green phosphor because
its green emission color is extremely saturated. However, its
luminescence decay lifetime is rather long, tens of milliseconds.1
In applications involving fast-moving images, this causes a green
afterglow that can blur the display. As a result of quantum size
confinement,2 nanoscale phosphors may offer significant im-
provements compared with conventional bulk phosphors. Lu-
minescent nanoparticles may have an increase in electron-hole
overlap, thus yielding greater oscillator strength and an enhance-
ment of luminescence quantum efficiency.3 Since the emission
decay lifetime is inversely proportional to the oscillator strength
of a transition,4 the lifetime is shortened with decreasing size.
Thus, high efficiencies with short decay times make nanopar-
ticles a new type of luminescent material with potential
applications.5
the advantages of In2S3 over other III-V semiconductors such
as InP, InAs, and GaAs is that In2S3 can be synthesized from
solution by chemical methods while sophisticated techniques
such as molecular beam epitaxy are needed for fabricating other
III-V semiconductors. The preparation of In2S3 from an
aqueous solution of InCl3 using H2S,7,12 Na2S,8,10 CS2,9 and CH3-
CSNH211 have been reported. Absorption spectra, photochemical
kinetics, and photosensitization effects have been reported in
In2S3 nanoparticle colloids.7-12 These results indicate that In2S3
nanostructured materials or thin films may find potential
applications as solar cell materials. Nagesha et al.8 reported the
luminescence of In2S3 nanoparticles with emission at 360-380
nm and a quantum yield of 1.5% and Xiong et al.9 have prepared
In2S3 nanophase dendrites with an emission peaking at 358 nm.
These UV emissions have been assigned to the excitonic or
band-to-band transitions of In2S3 nanoparticles.8,9 An orange
emission at 600 nm from In2S3 nanoparticles embedded in sol-
gel silica xerogel was reported and attributed to the luminescence
from In3+ ions.10 We have observed that small quantities of
metallic ions such as europium in the preparation solutions helps
to control the particle size and passivate the particle surfaces.
In this way, we have successfully synthesized In2S3 nanoparticles
with high quantum efficiencies in the blue, green, and red
spectral regions. In this paper, we discuss the structure,
luminescence, and potential applications of In2S3 nanoparticles
made using this novel recipe.
As a new type of luminescent material, II-VI semiconductor
nanoparticles such as CdTe and CdSe have received much
2. Experimental Details
* Corresponding author: wchen@nomadics.com.
† Nomadics, Inc.
The synthesis of indium-sulfide-doped nanoparticles is similar
to the scheme of arrested precipitation used in the preparation
of II-VI colloids with the exception that Mn2+ and Eu3+ ions
‡ University of Lund.
§ Pacific Northwest National Laboratory.
¶ National Laboratory for Superlattices and Microstructures.
10.1021/jp048107m CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/09/2004

11928 J. Phys. Chem. B, Vol. 108, No. 32, 2004
Chen et al.
are added to the solutions along with the chemical sources.
Typically, 4 mL of 0.1 M InCl3, 2 mL of 0.1 M Mn(NO3)2,
and 0.5 mL of thioglycolic (mercaptoacetic) acid were dissolved
in 50 mL of deionized water and degassed by N2 bubbling for
20 min. For preparation of In2S3:Eu3+-doped nanoparticles,
calculated amounts of Eu(NO3)3 were mixed with the InCl3.
The pH’s of the solutions were adjusted between 2.0 and 3.5.
Each mixture was vigorously stirred and 6 mL of an 0.1 M
aqueous solution of sodium sulfide was added. The solutions
were refluxed for 5 h, after which the doped nanoparticles (In1.8-
Eu0.2S3, In1.6Eu0.4S3) formed. The procedure for the preparation
of In2S3-undoped nanoparticles is the same except that no Eu3+
and Mn2+ ions were added to the precursor solution. We find
that the particle size can be modified by adding ions such as
Mn2+ and Eu3+ because metallic ions can inhibit the growth
rate of the nanoparticles.
The structure, size, and elemental composition of the nano-
particles were determined by high-resolution transmission
electron microscopy (HRTEM). The particles in solution were
brought onto holey carbon (with holes)-covered copper grids
for HRTEM observations. The HRTEM images of the particles
were obtained with a JEM-3000F electron microscope (300 kV)
with a structural information limit of 0.11 nm. The compositions
of the particles were analyzed using an energy-dispersive X-ray
spectrometer (EDS).
Figure 1. HRTEM images of deposited In2S3 nanoparticles showing
a particles size of about 4 to 5 nm.
The optical absorption spectra were obtained on a Hewlett-
Packard 8453 diode array spectrophotometer using 1-cm quartz
cuvettes. Luminescence emission, excitation, and temperature-
dependent luminescence were measured in a JY SPEX Fluorolog
3 Fluorometer. The quantum yield determination in this paper
is the same as reported by Xiong et al.9 The quantum yield was
determined by comparing the luminescence with Rhodamine B
in ethylene glycol (quantum yield near 100% at 580 nm). The
optical densities of an In2S3 dispersion in water and the standard
were adjusted to be within 0.05 absorbance units of each other
at the excitation wavelength of 390 nm. The quantum yield was
calculated according to the formula ηIn S ) ηS (IIn S /IS)*(ODS/
2
3
2 3
Figure 2. HRTEM images of deposited In1.6Eu0.4S3 nanoparticles
showing a particles size of about 2 to 3 nm.
ODIn S ), where ηIn S and ηS are the quantum yields of
2
3
2 3
nanoparticles and standard, IIn S and IS are the integrated area
of the luminescence peaks of the nanoparticles and the standard,
2
3
focused into a 1/8 m monochromator with a gated intensified
CCD camera to record the time-resolved spectra. Picosecond
lifetime data were collected using the output of a femtosecond
amplified Ti:Sapphire laser operating at 1 kHz. The 150-fs, 760-
nm output was frequency doubled in a 1 mm BBO crystal to
provide excitation at 380 nm and directed onto the samples.
The magic-angle polarized output emission was collected at right
angles and focused into the entrance of a streak camera
(Hamamatsu C5680). Appropriate filters were used to isolate
the wavelengths of interest. The time resolution was determined
to be 200 ps fwhm using a standard scattering material.
and ODIn S and ODS are the optical densities of the nanoparticles
2
3
and the standard at the excitation wavelength (390 nm).
However, we note that the quantum yield measured in this way
is not highly accurate as an absolute quantum yield, but gives
reliable values for the relative quantum yields of the different
particles. The purpose of measuring the quantum yield is to
compare the fluorescence efficiencies of the samples of different
sizes and to compare with published data in the literature. It is
also noted that the quantum efficiency of Rhodamine B is
excitation wavelength and solvent dependent. Its efficiency is
largest in ethylene glycol (close to 100%), while in other
solvents its yield is slightly lower. Recent measurements on
In2S3 nanoparticles using both Rhodamine B and p-quaterphenyl
as standards concluded that the quantum yield values obtained
with the different standards were similar even though the
emission wavelengths are quite different.9 Therefore it is
unlikely that the use of Rhodamine B as the quantum yield
reference leads to significant quantitative errors, particularly in
relative yield measurements.
3. Results
The In2S3 nanoparticles investigated with HRTEM (Figure
1) exhibit a cubic structure. The average size of the particles is
in the range of 4 to 5 nm. The (111) lattice spacing of the
particles is estimated to be around 0.3 nm from the HRTEM
images, which is consistent with the (111) spacing of bulk
In2S3,13 indicating a cubic cell of 0.54 nm. HRTEM images
(Figure 2) of the solution of In1.6Eu
S3 particles show that
0.4
The time-resolved spectral data were collected using a
nanosecond optical parametric oscillator/amplifier (OPO) op-
erating at a 10-Hz repetition rate and tunable between 220 and
1800 nm. The output was directed onto the particles and the
emission was collected at right angles to the excitation and
the average size of the particles is around 2 to 3 nm in diameter.
EDS analysis of aggregated particles such as the one in Figure
2 shows that the particles are composed of In, Eu, Mn, and S.
(Figure 3). Clearly the Mn and Eu ions added in the preparation
to inhibit particle growth are incorporated within or on the

Full-Color Emission from In2S3 and In2S3:Eu3+
J. Phys. Chem. B, Vol. 108, No. 32, 2004 11929
Figure 3. EDS analysis of an aggregate of In1.6Eu0.4 S3 nanoparticles, similar to the one in Figure 2. The analysis shows that the nanoparticles are
composed of Eu, Mn, In, and S.
Figure 4. Absorption spectra of In2S3, In1.8Eu0.2S3, and In1.6Eu0.4S3
nanoparticles.
Figure 5. Excitation spectra of In2S3, In1.8Eu0.2S3, and In1.6Eu0.4S3
nanoparticles. The monitored emission wavelength is at 515 nm for
In2S3 and In1.8Eu0.2S3, and is at 425 nm for In1.6Eu0.4S3 nanoparticles.
in the HRTEM images are most likely cubic close packed (ccp)
surface of the nanoparticles. The In1.6Eu0.4S3 particles observed
with a cubic cell of about 0.54 nm. HRTEM observations and
analysis confirm the formation of In2S3 nanoparticles in our
preparation.
Figure 4 displays the absorption spectra of In2S3, In1.8Eu0.2S3,
and In1.6Eu0.4S3 nanoparticles. For In2S3 and In1.8Eu0.2S3 par-
ticles, a strong absorption peak is observed at 390 nm with a
weak shoulder at 330 nm. For In1.6Eu0.4S3 particles, a strong
absorption peak is observed at 338 nm. Figure 5 displays the
excitation spectra and Figure 6 shows the emission spectra of
In2S3, In1.8Eu0.2S3, and In1.6Eu0.4S3 nanoparticles in aqueous
solution. For In2S3 and In1.8Eu0.2S3 nanoparticles, a strong green
emission at 510 nm is observed while for In1.6Eu0.4S3 particles
a strong blue emission is observed at 425 nm. The quantum
efficiencies for the green emission from In2S3, and In1.8Eu0.2S3
are 1.6% and 2.2%, respectively. The quantum efficiency for
the blue emission from In1.6Eu
S3 is 33%. The efficiency
0.4
from In2S3 nanoparticles is similar to that reported by Nagesha
et al.8 Interestingly, the efficiency from In1.6Eu0.4S3 nanoparticles
is one order of magnitude higher than that from pure In2S3
nanoparticles.
For In2S3 and In1.8Eu0.2S3 particles, two absorption bands at
310 and 400 nm are observed in the excitation spectra. For In1.6-
Eu0.4S3 particles, two absorption peaks are observed at 323 and
363 nm, respectively. The origins of the 310-nm peak in the
excitation spectra of In2S3 and In1.8Eu0.2S3 particles and the 323-
Figure 6. Emission spectra of In2S3 (a), In1.8Eu0.2S3 (b), and In1.6Eu0.4S3
(c) nanoparticles in aqueous solution following excitation at 397 nm.
nm shoulder in the excitation spectrum of In1.6Eu0.4S3 nano-
particles are not clear. The main excitation band shifts to shorter
wavelengths for In1.6Eu0.4S3 nanoparticles compared with that
of In2S3 and In1.8Eu0.2S3 particles, likely due to increased
quantum size confinement of the smaller particles. By comparing
the absorption spectra with the excitation spectra, it is observed

11930 J. Phys. Chem. B, Vol. 108, No. 32, 2004
Chen et al.
Figure 7. Emission spectra of In2S3 (a), In1.8Eu0.2S3 (b), and In1.6Eu0.4S3 (c) nanoparticle solutions at different excitation wavelengths as labeled in
the spectra.
2.5 ns while the green emission lifetimes depend on the sample
with the shortest decay from the pure In2S3 particles. The longer
nanosecond lifetime is slightly shorter than the excitonic
luminescence decay lifetime previously observed in In2S38 and
is similar to other semiconductor nanoparticles.14
4. Discussion
4.1. Luminescence of Indium Sulfide Nanoparticles in
Solution. Table 1 summarizes the absorption and emission
wavelengths of In2S3 and In2S3:Eu3+ nanoparticles in aqueous
solution. For the green emission, the Stokes shift is around 110
nm, while for the blue emission, the Stokes shift is about 62
nm. The energy gap of bulk In2S3 is between 2.0 and 2.2 eV
with the corresponding absorption edge from 620 to 550 nm.15,16
In comparison, the absorption edge of the In2S3 nanoparticles
is approximately 150-220 nm blue-shifted from bulk In2S3. This
blue-shift is likely due to quantum size confinement. As the
diameter of the semiconductor particle approaches the exciton
Bohr diameter, its electronic properties change. The Bohr radius
of the exciton in In2S3 may be calculated by2,4
Figure 8. Luminescence lifetime decay curves of In2S3 (monitored at
500 nm), In1.8Eu0.2S3 (monitored at 500 nm), and In1.6Eu0.4S3 (monitored
at 435 nm) nanoparticles following excitation at 380 nm.
that for each absorption band, the main peak in the excitation
spectrum is at the absorption edge of the corresponding
absorption peak in the absorption spectra. This is a common
phenomenon in semiconductors since the photon-excitation rate
is the greatest at the absorption edge where the absorption index
is fairly low and the dispersion is small.37
Figure 7 displays the emission spectra of In2S3, In1.8Eu0.2S3,
and In1.6Eu0.4S3 nanoparticle solutions at different excitation
wavelengths. Figure 6 shows clearly that in In2S3 nanoparticles,
the green emission position and shape do not vary significantly
following excitation at different wavelengths. For In1.8Eu0.2S3
nanoparticles, the green emission wavelength remains constant
at different excitation wavelengths as well. However, a new
emission peak at 425 nm appears following excitation at
wavelengths shorter than 370 nm (Figure 6b). The blue emission
at 425 nm is stronger in intensity than the green emission at
510 nm. Figure 6c shows the emission spectrum of In1.6Eu0.4S3
following excitation at different wavelengths. The emission peak
position moves to shorter wavelengths gradually as the excitation
wavelength decreases.
p2ꢀ 1
1
aB )
+
(1)
e2
[
]
me mh
where ꢀ is the dielectric constant (∼11), p is the Planck constant
and me and mh are the electron and hole effective mass,
respectively. Assuming that me ) mh ) µ ) 0.25 × 10-28 g,12
the Bohr radius of the exciton in In2S3 is calculated to be 33.8
nm. Since the particle sizes are significantly smaller than the
exciton Bohr radius, a large blue shift of the absorption edge is
observed.
Bulk In2S3 is normally not luminescent at room temperature.9
Recent observations on In2S3 micron-sized powders show a
strong emission peaking at about 595 nm at 80 K.17 In the
particles reported here, strong blue and green luminescence is
observed. In semiconductors, the most common types of
emission are from exciton-like transitions and free-to-bound and
donor-acceptor pair transitions, often mediated at defect sites.
Recently, nanoparticles of In2S3 have been reported to show
both UV8,9 and orange10 luminescence. The UV luminescence
has been assigned to excitonic emission8 while the red emission
Figure 8 shows the luminescence lifetime decays from In2S3,
In1.8Eu0.2S3, and from In1.6Eu0.4S3 nanoparticles, near the peak
wavelength of their respective emission following excitation at
380 nm. The lifetimes can be fit adequately with biexponential
decay functions and the results of this fit are listed in Table 2.
The decay lifetimes of the blue emission are about 700 ps and
has been attributed to the Indium interstitial defect, In3+ 10
. The
indium interstitial is also a common defect emitter in In-rich
CuInS2 thin films.18,19
Generally in nanoparticles, if the emission is from an exciton
or band-to-band transition, the emission wavelength maximum

Full-Color Emission from In2S3 and In2S3:Eu3+
J. Phys. Chem. B, Vol. 108, No. 32, 2004 11931
TABLE 1: Absorption and Emission Wavelengths of In2S3 and In2S3:Eu3+ Nanoparticles in Aqueous Solution
size
range (nm)
absorption
(nm)
excitation
(nm)
emission
(nm)
stokes-shift
(nm)*
samples
In2S3
In1.8Eu0.2S3
In1.6Eu0.4S3
efficiency (%)
4-5
4-5
2-3
330, 390
390
340
310, 400
310, 400
323, 363
510
512
425
110
112
62
1.6
2.2
33
*Measured from the excitation and emission spectra.
TABLE 2: Luminescence Lifetimes of In2S3 Nanoparticles
particle
wavelength (nm)
lifetime (ns)
amplitude ratio*
In2S3
In1.8Eu0.2S3
In1.6Eu0.4S3
500
500
435
0.4, 2.3
0.7, 3.7
0.7, 2.5
55
3
5
*Ratio of the amplitudes of the fast and slow lifetime components.
will vary with the excitation wavelength because the absorption
and emission wavelengths are size-dependent.20 Since the
nanoparticle sample contains a distribution of different sized
particles, the larger particles that have smaller band gaps absorb
and emit at longer wavelengths relative to smaller particles in
the distribution.20 However, if the emission is from a defect or
a luminescence center such as Mn2+ or Eu3+, the emission
energy is not dependent on the excitation energy because the
emission energies of these transitions are not size-dependent.
Figure 7 shows clearly that the green emission position and
shape of the In2S3 nanoparticles do not vary significantly
following excitation at different wavelengths. For In1.8Eu0.2S3
nanoparticles, the green emission wavelength also remains
constant following excitation at different wavelengths. However,
Figure 9. Time-resolved emission spectra of In1.8Eu0.2S3 nanoparticles
in aqueous solution taken at 0 ns, 100 ns, 200 ns, and 50 µs after
excitation at 395 nm.
shifted. Nonetheless, it is clear from eq 2 that the In1.6Eu0.4S3
particles have a significantly shorter radiative lifetime compared
with the other particles. This is consistent with our assignment
of the blue luminescence from the In1.6Eu0.4S3 particles as a
band-to-band transition and the green luminescence from the
other particles as defect luminescence that would be expected
to have a significantly longer lifetime. The nonradiative rate in
these samples may be entirely dominated by surface trapping
as is common in very small particles because of their necessarily
large surface-to-volume ratio. The fast, subnanosecond decays
observed may very well be associated with these trap states.
We note that in the In2S3 pure particles, the amplitude of this
component is significantly enhanced relative to the longer
nanosecond decay as compared to the europium-doped particles
(Table 2). Therefore, the incorporation of the Eu3+ ions may
passivate surface defect sites as has been reported previously.25
Based on the luminescence behaviors described above, the
blue emission from In2S3 nanoparticles is most likely due to
excitonic recombination and the green emission is from an
indium interstitial. The formation of indium interstitials in In2S3
is likely to occur because the particles are prepared in an acidic
solution with pH 2.0 and 3.5. Under these conditions, some
sulfur ions may escape as H2S gas.
Figure 6c shows that the peak of the blue emission of In1.6
-
Eu0.4S3 moves to shorter wavelengths gradually as the excitation
wavelength decreases. This is typical luminescence behavior
of an excitonic or band-to-band transition from a distribution
of semiconductor nanoparticles.20
The dependences of the emission intensity and emission
wavelength with the excitation power density can also provide
some valuable information about the origin of the emission. As
a general rule, the emission wavelength will shift toward shorter
wavelengths with increasing excitation intensity if the emission
is from donor-acceptor pair recombination.18,21-24 Power
density measurements on both In2S3 and In1.6Eu0.4S3 nanopar-
ticles (data not shown) show no shift in the emission wavelength
with power density. Therefore, donor and acceptor pair recom-
bination is not likely responsible for either the blue or green
emission.
The measured lifetimes in all three particles show fairly
similar decay times and all are at least biexponential in nature.
However, the relative quantum yields of the In1.6Eu0.4S3 particles
are about a factor of 10 higher than the other samples. These
particles also show the blue emission whereas the other particles
show the green emission. For purely single-exponential lumi-
nescence decays, the relation
From the EDS measurements, it is clear that both manganese
and europium are present in the nanoparticles as both are added
during the preparation procedure, however there is no obvious
spectroscopic signature for either species in the emission spectra.
The luminescence from Eu3+ is not easily observed over the
strong blue and green emission in Eu3+-doped In2S3 nanopar-
ticles in solution. However, red luminescence from Eu3+ is
observed in time-resolved spectra at longer times (microseconds)
once the green or blue emission from the In2S3 nanoparticles
has decayed away. Figure 9 displays the time-resolved emission
spectra taken at 0 ns, 100 ns, 200 ns, and 50 µs delay after
excitation. The emission bands peaking at 579.5, 593, 615, 648,
φ ) τobs/τrad
(2)
where φ is the quantum yield, and τobs and τrad are the observed
and radiative lifetimes, can be used to determine the radiative
lifetimes. While these relations may not be rigorously true for
multiexponential decays or decays fit to a distribution of
lifetimes, they can still be used to gain some insight into the
nature of the transitions. We note that the values for both the
quantum yield and observed lifetime of the In2S3 nanoparticle
samples presented here are quite similar to those reported by
Xiong,9 however the emission wavelength is significantly red-
5
5
5
5
and 698 nm are from the D0-7F0, D0-7F1, D0-7F2, D0-
7F3, and 5D0-7F4 transitions of Eu3+, respectively. (see Figure
11 for the details). The decay lifetime of Eu3+ emission in In2S3:

11932 J. Phys. Chem. B, Vol. 108, No. 32, 2004
Chen et al.
nanoparticles, the luminescence is predominantly from Mn2+
ions that are doped inside the particles. The Mn2+ species
adsorbed to the nanoparticle surfaces are not luminescent.28 In
other words, the dopant ions within the lattice are responsible
for emission while the ions adsorbed on the particle surfaces
provide surface passivation. Likewise, some Eu3+ most likely
resides on the surfaces of the nanoparticles, also offering surface
passivation. Evidence for this can be observed as the In1.8Eu0.2S3
sample begins to show a blue excitonic peak in the emission
(Figure 6) and the higher-doped In1.6Eu1.6S3 particles show
predominantly excitonic emission. This is likely because Eu3+
ions first passivate the surface so that the exciton does not trap
easily to the surface states. Higher Eu3+ doping may also replace
or reduce the In3+ interstitial defects, resulting in purely
excitonic emission with higher quantum efficiency. Strong
quantum size confinement and good surface passivation prob-
ably lead to the high luminescence efficiency from In1.6Eu0.4S3
nanoparticles.
4.2. Luminescence and Temperature Dependence of In-
dium Sulfide Nanoparticles Thin Films. To fabricate devices
that require full-color emission properties, it is desirable to have
red emission in addition to the blue and green emission of the
In2S3 nanoparticles. Eu3+ is a common dopant resulting in red
emission and used in fluorescent lighting and displays. In
addition, solids or thin film materials that can withstand elevated
temperatures may be required for some applications. Figure 10
shows the excitation and emission spectra of In1.8Eu0.2S3 and
In1.6Eu0.4S3 nanoparticle thin films that are made by evaporating
the nanoparticle solutions on glass slides. In addition to the
emission of In2S3 particles, strong luminescence from Eu3+ is
observed in the continuous-wave emission spectra from thin
films. In this case, the Eu3+ luminescence is not quenched by
nearby water molecules so that the luminescence is significantly
enhanced relative to the aqueous nanoparticles. In the excitation
spectra obtained monitoring the Eu3+ emission at 610 nm, the
f-f transitions of Eu3+ and the band-to-band transition of the
In2S3 host are observed.
Figure 10. Excitation and emission spectra of In1.8Eu0.2S3 and In1.6
-
Eu0.4S3 nanoparticle thin films. PLE-510 nm represents the photolu-
minescence excitation spectrum with monitored emission at 510 nm,
and PL-410 nm represents the photoluminescence emission spectrum
with excitation at 410 nm.
The changes in the emission wavelength as a function of
temperature can give important information about the lumines-
cence mechanism. Generally, if the energy levels of the emitter
are coupled to the crystal field strength or chemical bond or
pair distances, the emission wavelength is temperature depend-
ent. This behavior has been observed in the band-to-band
transitions and excitonic luminescence in semiconductors,31,32
the d-d transitions of transition metal ions,31,33,34 and the d-f
transitions of rare earth ions.31 If the transition is not strongly
influenced by the crystal field strength, then its emission
wavelength is temperature independent, as observed in the f-f
transitions of trivalent rare earth ions,14,31 and some interstitial
defects.35,36 Figure 11 displays the emission of both In2S3 and
In1.6Eu0.4S3 thin films at different temperatures. The results
shown in Figure 11 demonstrate that the 510-nm green emission
peak and the red emission from Eu3+ do not change in emission
wavelength while the 425-nm blue emission peak shifts to longer
wavelengths (lower energies) at higher temperatures. The red
shift of the blue emission with increasing temperature is a typical
temperature behavior of excitonic emission in semiconductors37
and supports our previous assignment. As the interstitial ions
are likely at the surfaces or interfaces among the nanocrystals,
the crystal field has little effect on its energy levels. This is
likely why the green emission wavelength does not change with
temperature.
Figure 11. Emission spectra of In2S3 and In1.6Eu0.4S3 nanoparticle thin
films at different temperatures below room temperature. The excitation
is at 325 nm.
Eu3+ nanoparticles is about 10 µs, much longer than either the
blue or green emissions from In2S3. In aqueous solutions, the
Eu3+ lifetime is significantly shortened due to quenching by
energy transfer to the O-H stretch26 of nearby water molecules.
No emission from Mn2+ is observed however, despite the
fact that its lifetime should be in the millisecond range. Surface-
adsorbed Mn2+ on ZnS27 or CdS28 nanoparticles is not lumi-
nescent, while Mn2+ doped into the lattice displays characteristic
broad orange luminescence. Due to the difference in charge, it
is unlikely that the Mn2+ can easily substitute for the indium in
the lattice, thus the most likely place for manganese to reside
is on the surfaces of the nanoparticles. Recent reports indicate
that ion doping can inhibit nanoparticle growth and passivate
the surface defect sites.25,29,30 In addition, in CdS:Mn2+-doped
The measurement of the luminescence intensity temperature
dependence can also provide useful information about the

Full-Color Emission from In2S3 and In2S3:Eu3+
J. Phys. Chem. B, Vol. 108, No. 32, 2004 11933
Luminescence intensities of In1.8Eu0.2S3 and In1.6Eu0.4S3
nanoparticle thin films at different temperatures above room
temperature show a gradual decrease with increasing temperature
likely due to thermal quenching (data not shown), although
strong luminescence is observed at temperatures up to 100 °C.
This indicates that the particles and luminescence properties are
quite stable in air, even at elevated temperatures. Eu3+-doped
phosphors are widely used in lighting and displays and the f-f
emission of Eu3+ is very stable due to shielding from outer-
shell electrons.31,35 The blue and green emissions are comparable
in stability to the Eu3+ emission, therefore these nanoparticles
may make good multicolor luminescent materials.
The blue, green, and red emissions from In2S3 and In2S3:
Eu3+ nanoparticles make them a potentially new type of full-
color phosphor. In fact, both the blue and green broad emission
bands cover a large portion of the full visible spectrum from
the blue to red wavelength ranges. In addition, the luminescence
decay lifetime is very fast and In2S3 nanoparticles are relatively
simple to prepare. Thus, In2S3-based nanophosphors may have
a bright future for full-color emission, flat panel displays, and
lighting.
5. Conclusions
In2S3 and In2S3:Eu3+ nanoparticles prepared by solution
techniques show blue, green, and red emission. Experimental
observations on the luminescence responses following changes
in the excitation wavelength and excitation power density, as
well as the luminescence lifetime decays and temperature
dependences, indicate that the blue emission of In2S3 nanopar-
ticles is due to exciton recombination and the green emission
to indium interstitial defects. The full-color emission, short
luminescence decays, and good stability make In2S3 nanopar-
ticles a new type of promising phosphor for flat-panel displays,
lighting, and optical communications.
Figure 12. The emission intensity changes of In2S3 (upper) and In1.6
-
Eu0.4S3 (lower, squares for the blue emission and triangles for the red
emission of Eu3+ at 612 nm) nanoparticle thin films at different
temperatures below room temperature. The excitation is at 325 nm.
stability of the particles as well as the photophysical origins of
the luminescence. Figure 12 shows the change in intensity at
different temperatures below room temperature for both In2S3
and In1.6Eu0.4S3 thin films. Overall, the green emission from
In2S3 increases in intensity with decreasing temperature. How-
ever, the intensity temperature dependence of the blue emission
from In1.6Eu0.4S3 is complex, decreasing from room temperature
to 20 K, while increasing at 10 K. The complex relationship
between the intensity of the blue emission with temperature may
be partly due to energy transfer to Eu3+ in In1.6Eu0.4S3
nanoparticles. Energy transfer from the blue band to Eu3+ ions
may occur because the blue emission band overlaps with the
absorption peaks of Eu3+. In this case, the intensity change of
the blue emission with temperature is related to both the phonon
induced nonradiative rate and the energy transfer rate. Generally,
the phonon-induced nonradiative rate decreases with decreasing
temperature, so the radiative emission should increase at lower
temperatures. However, the energy transfer rate is enhanced at
lower temperatures because the distance between the two centers
becomes smaller. As a result, the blue emission intensity
decreases at lower temperatures likely due to increase of energy
transfer to the Eu3+ ions. In addition, the carriers can transfer
between the energy levels responsible for the blue and red
emissions. The occupation of the carriers on the higher energy
level (responsible to the blue emission) will decrease with
decreasing temperature, which results in the decrease of the
intensity of the blue band. Therefore, the combination of
phonon-induced quenching and energy transfer leads to the
complex dependence of the blue emission intensity with
temperature. For the emission of Eu3+, both the changes in the
phonon-induced nonradiative rate and the energy transfer rate
enhance its intensity at lower temperatures. Therefore, the
luminescence of Eu3+ increases in intensity with decreasing
temperature.
Acknowledgment. W. C. and S. P. W. would like to thank
Nomadics, Inc., United States, the National Science Foundation
(NSF, Grant No. DMI-0132030), National Institutes of Health
(NIH, Grant No. 1R43CA94403-1), and the Department of
Energy (DOE, DE-FG02-04ER84023) for grants. Part of the
research described in this paper was performed at the W.R.
Wiley Environmental Molecular Sciences Laboratory, a national
scientific user facility sponsored by the Department of Energy’s
Office of Biological and Environmental Research and located
at the Pacific Northwest National Laboratory (PNNL). PNNL
is operated by Battelle for the U.S. Department of Energy under
contract DE-AC06-76RLO1830. J.-O. B. would like to thank
the Swedish Research Council for support. F. S. and G. L. would
like to thank the National Natural Science Foundation of China
for financial support (contract No. 60076012).
References and Notes
(1) Justel, T, H.; Nikol, H.; Ronda, C. Angew. Chem., Int. Ed. 1998,
37, 3084.
(2) Ekimov, A. I.; Efros, A. L.; Onuschenko A. A. Solid State Commun.
1985, 56, 921.
(3) Chen, W.; Sammynaiken, R.; Huang, Y. J. Appl. Phys. 2000, 88,
5188.
(4) (a)Knox, R. S. In Theory of Excitons, Solid State Physics Supple-
ments; Academic Press: New York, 1963. (b) Curthbert, J. D.; Thomas,
D. G. Phys. ReV. 1967, 154, 763.
(5) Chen, W.; Joly, A. G.; Wang, S. P. Luminescence of Semiconductor
Nanoparticles. In The Encyclopedia of Nanoscience and Nanotechnology;
Nalwa, H. S., Ed.; American Scientific Publishers: Los Angeles, 2004;
Vol. 4, pp 689-718.
(6) Alivisatos, A. P. J. Phys. Chem. 1996, 100, 13226.

11934 J. Phys. Chem. B, Vol. 108, No. 32, 2004
Chen et al.
(7) Kamat, P. V.; Dimitrijevic, N. M.; Fessenden, R. W. J. Phys. Chem.
1988, 92, 2324.
(23) Chang, Y.; Chu, J. H.; Tang, W. G.; Shen, W. Z.; Tang, D. Y. IR
Phys. Technol. 1996, 37, 747.
(8) Nagesha, D. K.; Liang, X. R.; Mamodov, A.; Gainer, G.; Eastman,
M. A.; Giersig, M.; Song, J.; Kotov, N. A. J. Phys. Chem. B 2001, 105,
7490.
(9) Xiong, Y. J.; Xie, Y.; Du, G. A.; Tian X. B.; Qian, Y. T. J. Solid
State Chem. 2002, 166, 336.
(24) Topper, K.; Bruns, J.; Scheer, R.; Weber, M.; Weidinger, A.;
Braunig, D. Appl. Phys. Lett. 1997, 71, 482.
(25) Pires, A. M.; Santos, M. F.; Davolos, M. M. R.; Stucchim, E. B. J.
Alloys Compd. 2002, 344, 276.
(26) Bartlett, J.; Cooney, R. P.; Kydd, R. A. J. Catal. 1988, 114, 58.
(27) Sooklal, K.; Cullum, B. S.; Angel, S. M.; Murphy, C. J. J. Phys.
Chem. 1996, 100, 4551.
(28) Counio, G.; Esnouf, S.; Gacoin, T.; Biolot, J.-P. J. Phys. Chem.
1996, 100, 20021.
(29) Han, J. P.; Mantas, P. Q.; Senos, A. M. R. J. Mater. Res. 2001, 16,
459.
(30) Logan, R. A.; Schumaker, N. E.; Henry, C. H.; Merritt, F. R. J.
Appl. Phys. 1979, 50, 5970.
(31) Wang, S. P.; Westcott, S. L.; Chen, W. J. Phys. Chem. B 2002,
106, 11203.
(32) Hunter, A. T.; McGill, T. C. J. Appl. Phys. 1981, 52, 5779.
(33) Tanaka, M.; Masumoto, Y. Chem. Phys. Lett. 2000, 324, 249.
(34) Chen, W.; Su, F.; Li, G. H.; Joly, A. G.; Malm, J.-O.; Bovin, J.-O.
J. Appl. Phys. 2002, 92, 1950.
(10) Ai, Z. P. Opt. Mater. 2003, 24, 589.
(11) Avivi, S.; Palchick, O.; Palchik, V.; Slifkin, M. A.; Weiss, A. M.;
Gedanken, A. Chem. Mater. 2001, 13, 2195.
(12) Yasaki, Y.; Sonoyama, N.; Sakata, T. J. Electroanal. Chem. 1999,
469, 116.
(13) Rooymans, C. J. M. J. Inorg. Nucl. Chem. 1959, 4, 119.
(14) Chen, W.; Joly, A. G.; Kowalchuk, M. C.; Malm, J.-O.; Huang,
Y.; Bovin, J.-O. J. Phys. Chem. B 2002, 106, 7034.
(15) Hara, K.; Sayama, K.; Arakawa, H. Sol. Energy Mater. Sol Cells
2000, 62, 441.
(16) Kim, W. T.; Kim, C. D. J. Appl. Phys. 1986, 60, 2631.
(17) Gorai, S.; Guha, P.; Ganguli, D.; Chaudhuri, S. Mater. Chem. Phys.
2003, 82, 974.
(18) Binsma, J. M.; Giling, L. J.; Bloem, J. J. Lumin. 1982, 27, 36.
(19) Wijesundara, R.; Siripala, W. Sol. Energy Mater. Sol Cells 2004,
81, 147.
(35) Grattan, K. T. V.; Zhang, Z. Y. Fiber Optic Fluorescence
Thermometry; Chapman and Hall: London, 1995.
(36) Halliday, D. P.; Potter, M. D. G.; Mullins, J. T.; Brinkman, A. W.
J. Cryst. Growth 2000, 220, 30.
(37) Moss, T. S. Optical Properties of Semiconductors; Butterworth
Scientific Publications: London, 1959.
(20) Hoheisel, W.; Colvin, V. L.; Johnson, C. S.; Alivisatos, A. P. J.
Chem. Phys. 1994, 101, 8455.
(21) Thomas, D. G.; Hopfield, J. J.; Augustiniak, W. M. Phys. ReV. B
1965, 140, 202.
(22) Lahlou, N.; Masse, G. J. Appl. Phys. 1981, 52, 978.