Efficient solid-state light-emitting CuCdS nanocrystals synthesized in air

Article abstract of DOI:10.1002/anie.201409518

Semiconductor nanocrystals (NCs) possess high photoluminescence (PL) typically in the solution phase. In contrary, PL rapidly quenches in the solid state. Efficient solid state luminescence can be achieved by inducing a large Stokes shift. Here we report on a novel synthesis of compositionally controlled CuCdS NCs in air avoiding the usual complexity of using inert atmosphere. These NCs show long-range color tunability over the entire visible range with a remarkable Stokes shift up to about 1.25 eV. Overcoating the NCs leads to a high solid-state PL quantum yield (QY) of ca. 55% measured by using an integrating sphere. Unique charge carrier recombination mechanisms have been recognized from the NCs, which are correlated to the internal NC structure probed by using extended X-ray absorption fine structure (EXAFS) spectroscopy. EXAFS measurements show a Cu-rich surface and Cd-rich interior with 46% Cu^I being randomly distributed within 84% of the NC volume creating additional transition states for PL. Color-tunable solid-state luminescence remains stable in air enabling fabrication of light-emitting diodes (LEDs).

Full text of DOI:10.1002/anie.201409518

Angewandte
Chemie
DOI: 10.1002/anie.201409518
Light-Emitting Diodes
Efficient Solid-State Light-Emitting CuCdS Nanocrystals Synthesized
in Air**
Ali Hossain Khan, Amit Dalui, Soham Mukherjee, Carlo U. Segre, D. D. Sarma, and
Somobrata Acharya*
Abstract: Semiconductor nanocrystals (NCs) possess high
photoluminescence (PL) typically in the solution phase. In
contrary, PL rapidly quenches in the solid state. Efficient solid
state luminescence can be achieved by inducing a large Stokes
shift. Here we report on a novel synthesis of compositionally
controlled CuCdS NCs in air avoiding the usual complexity of
using inert atmosphere. These NCs show long-range color
tunability over the entire visible range with a remarkable
Stokes shift up to about 1.25 eV. Overcoating the NCs leads to
a high solid-state PL quantum yield (QY) of ca. 55%
measured by using an integrating sphere. Unique charge
carrier recombination mechanisms have been recognized
from the NCs, which are correlated to the internal NC structure
probed by using extended X-ray absorption fine structure
(EXAFS) spectroscopy. EXAFS measurements show a Cu-
rich surface and Cd-rich interior with 46% CuI being
randomly distributed within 84% of the NC volume creating
additional transition states for PL. Color-tunable solid-state
luminescence remains stable in air enabling fabrication of
light-emitting diodes (LEDs).
their size, shape, and composition-dependent photolumines-
cence (PL) properties, photostability, and high PL quantum
yields (QYs).[1–6] Recent progress in the synthetic strategies
have led to the prospect of designing low-cost fabrication
possibilities for high-efficiency LEDs using NCs as active
light-emitting layer.[6–14] The active layer of NC-based LEDs is
usually made from solution-processable routes which trans-
form to the solid state within the device. Although most of the
NCs are highly luminescent in solution form, the lumines-
cence rapidly quenches in the solid state.[14–17] Thus, achieving
intense solid-state luminescence from NCs is a key to enhance
the efficiency of NC-based LEDs. Major challenges encoun-
tered in achieving solid-state luminescence from NCs are
associated with self-absorption, nonradiative recombination
pathways, and crystal size distribution that leads to energy
transfer in solid proximity.[14–17] Consequently, it remains
important to develop new strategies for designing efficient
solid-state luminescent NCs by overcoming these detrimental
phenomena.
Recent advances in ternary and quaternary alloyed semi-
conductor NCs have shown advantageous features, such as
color-tunable PL, high PL QYs, and solution processability,
showing prospects for fabricating LEDs.[18–25] Nevertheless,
alloyed NCs show positional discontinuities of atoms at the
surface and interior of NCs, the influence of which on the
observed spectral transitions is largely unknown.[26] In addi-
tion, to the best of our knowledge, the design of solid-state-
luminescent alloyed NCs of I-II-VI category has not been
reported yet. Here we report on the synthesis of monodis-
perse ternary copper–cadmium–sulfide NCs simply by heat-
ing copper oleate and cadmium oleate with dodecanethiol in
air without the inert atmosphere conventionally required for
colloidal synthesis of NCs. These NCs readily show color-
tunable PL over the entire visible spectrum with remarkably
large Stokes shift. Overcoating the NCs with a CdS shell leads
to a high solid-state PL QY of about 55%. These NCs show
robust thermal and photostabilities sustaining nearly for
a year. A novel carrier recombination mechanism is observed
from the optical transitions which are correlated to the
internal structure of NCs probed by using EXAFS spectros-
copy. High solid-state PL QYenabled the fabrication of LEDs
with NCs as active layer, with a remarkable low turn-on
voltage. Electroluminescence (EL) obtained from the devices
is stable over a wide range of operating voltages offering the
possibility of meeting the technical requirements of NC-based
LEDs.
S
emiconductor nanocrystals (NCs) are promising candi-
dates for color-tunable light-emitting diodes (LEDs) owing to
[*] Dr. A. H. Khan, A. Dalui, Prof. D. D. Sarma, Dr. S. Acharya
Centre for Advanced Materials
Indian Association for the Cultivation of Science
Jadavpur, Kolkata 700032 (India)
E-mail: camsa2@iacs.res.in
S. Mukherjee, Prof. D. D. Sarma
Solid State and Structural Chemistry Unit
Indian Institute of Science
Bangalore 560012 (India)
Prof. C. U. Segre
Department of Physics & CSRRI, Illinois Institute of Technology
Chicago, IL 60616 (USA)
Prof. D. D. Sarma
Council of Scientific and Industrial Research—Network of Institutes
for Solar Energy (CSIR-NISE)
New Delhi 110001 (India)
[**] This work was supported by the DST India. We thank N. Beaver and
R. Obaid for EXAFS measurements. A.H.K and A.D. acknowledge
CSIR, India. C.U.S. acknowledges the National Science Founda-
tion’s Materials World Network program, grant number DMR-
086935. MRCAT operations are supported by the Department of
Energy and the MRCAT member institutions. Use of the Advanced
Photon Source at Argonne National Laboratory is supported by the
U.S. Department of Energy, Office of Science, Office of Basic Energy
Sciences, under contract number DE-AC02-06CH11357.
We synthesized the NCs in a single step by mixing copper
oleate, cadmium oleate, and 1-dodecanethiol in a molar ratio
of 1:2:16 in the noncoordinating solvent 1-octadecene. The
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201409518.
Angew. Chem. Int. Ed. 2015, 54, 2643 –2648
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2643

.
Angewandte
Communications
observed from binary NCs.[2,4,19] The PL excitation spectrum
(PLE) shows a peak at 3.18 eV, at the same position as in the
absorption spectrum (Figure 1c). In addition, PLE measured
by monitoring different energies of the broad PL peak show
the same peak position (Figure 1d), which suggests that the
broadness of the PL peak is not related to the NCs size
distribution. Interestingly, magnification of the spectral
region between 1.77 eV and 2.48 eV of the PLE spectra
shows another peak at about 2.0 eV, at the same position as
the PL maximum (Figure S3 and Figure 1d inset). The
overlapping of PLE and PL peaks suggests common states
of origin for both transitions at lower energies. The existence
of two PLE peaks suggests the presence of separate energy
states that contribute to the observed PL.
We have carried out X-ray photoelectron spectroscopy
(XPS) to investigate the oxidation state of Cu within the NCs
(Figure S4). A high-resolution XPS spectrum of Cu 2p state
shows doublet peaks with a separation of 19.7 eV and the
absence of any satellite features associated with 2p3/2 and 2p1/2
peaks are supportive of the presence of CuI state in the NCs.
The 3d electrons of Cu may create two possible states: CuI
(3d10) and CuII (3d9), among which the 3d9 is magnetically
active.[27] We have performed electron paramagnetic reso-
nance (EPR) measurements to identify the presence of
unpaired electrons of Cu in the NCs. No EPR signal is
observed for the NCs suggesting the absence of CuII states
(Figure S5). However, control experiments with only the
precursor mixtures (Cd, Cu-oleate) show EPR signals indi-
cating paramagnetic behavior (Figure S5). Hence the CuII
state, which is present in the precursor mixtures, is absent in
the NCs due to the reduction of CuII to CuI by the ligands.[28]
The presence of CuI level indicates the possibility of
isolated CuI state within the NCs when considered as dopant.
Another possibility is the formation of chemical bonds with
the CdS constituents forming an alloy. For an isolated CuI
state, the PL will solely depend on the energy gap between the
conduction band edge (CdSCB) to the CuI level and the
separation between CuI level and the valence band edge
(CdSVB) of the NCs will define the resultant Stokes shift
observed in optical processes (Figure S6 and SI).[27] Hence,
the energy gap between the CuI level and CdSVB should be
about 1.18 eV to match with the observed Stokes shift
(Figure S6b). However, our excitation-dependent PL meas-
urements rule out the existence of an isolated CuI state within
the NCs (Figure 2a). The excitation energies spanning from
4.4 eV to 2.4 eV show the highest PL intensity for the 3.5 eV
excitation which gradually decreases by lowering the excita-
tion (Figure 2a). Note that the PL maximum retains the same
position for all excitation energies. Notably, the absorption
spectrum showed a peak at 3.18 eV (Figure 1c), whereas the
PL can be obtained with excitation energies of up to 2.4 eV,
which is below 3.18 eV. These optical measurements suggest
the existence of additional energy levels within CdSVB and
CdSCB creating a reduced energy gap for PL. Hence, the
observed Stokes shift cannot be accommodated due to the
presence of these additional states. On the other hand,
alloying by chemical bond formation among the constituents
may occur by substituting CdII atoms from the cubic CdS
lattices or by the occupation at the interstitial positions of the
Figure 1. a) TEM images of CuCdS NCs. Top inset: HRTEM image.
Bottom inset: elemental analysis by EDS in TEM. b) XRD pattern with
reflections of zinc blende bulk CdS (dotted lines). c) UV/Vis absorp-
tion, PL (excitation 3.18 eV), and PLE (emission 2.0 eV) spectra of
CuCdS NCs in chloroform solution. d) PLE spectra monitored at
different emission energies (A, 2.431 eV; B, 2.296 eV; C, 2.175 eV; D,
2.006 eV; E, 1.068 eV; F, 1.878 eV; G, 1.796 eV). Inset: Comparison of
PL (excitation 3.18 eV) and PLE (emission 1.55 eV) spectra showing
overlap of the peaks at about 2 eV.
precursor mixture was annealed for 60 min at 2008C in air
(see the Supporting Information (SI) for details). The trans-
mission electron microscopy (TEM) image shows monodis-
perse NCs with a diameter of 2.5 Æ 0.2 nm (Figure 1a). The
high-resolution TEM (HRTEM) image shows that the NCs
are single-crystalline in nature with well-resolved lattice
spacing of 0.33 Æ 0.02 nm corresponding to (111) planes of
cubic bulk CdS (Figure 1a, inset). Energy-dispersive X-ray
spectroscopy (EDS) shows atomic ratios of 2:5:3 of Cu:Cd:S
in the NCs suggesting a possible Cu2Cd5S3 (hence onward
denoted as CuCdS) cation-rich composition (Figure 1a inset
and Figure S1). Quantitative elemental analyses using induc-
tively coupled plasma atomic emission spectroscopy (ICP-
AES) support the observed atomic ratio for Cu to Cd cations.
The TEM-EDS elemental analyses demonstrate sulfur defi-
ciency in comparison to the cations in the resultant NCs.
Powder X-ray diffraction (XRD) pattern of the NCs shows
(111), (220), and (311) reflections corresponding to the zinc
blende structure of bulk CdS (Figure 1b). However, the XRD
peaks are slightly shifted toward higher angles in comparison
to a zinc blende CdS NCs of the same size suggesting that the
reduced lattice spacing is possibly due to the incorporation of
Cu within CdS (Figure S2).
The absorption spectrum of the NCs shows a peak at
3.18 eV, whereas the PL exhibits a peak at 2.0 eV implying
a large Stokes shift of ca. 1.18 eV (Figure 1c). The observed
large Stokes shift can be accounted to the presence of Cu
within CdS creating additional lower-energy transition states.
One of the interesting features of the PL spectrum is the
broad nature of the peak compared to the narrow PL peaks
2644 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 2643 –2648

Angewandte
Chemie
tetrahedral environment (SI). We explored different possi-
bilities such as cationic-site substitution by Cu in a CdS lattice,
Cu occupation at interstitial positions, and surface precipita-
tion to understand the local environment of NCs (SI).
Additionally, we have accounted for the effect of metal–
metal correlations (Cu–Cu, Cu–Cd, and Cd–Cd) and surface
oxidation. The final model considers single scattering con-
tributions from the closest near neighbor Cu–S and Cd–S
correlations (Fit 1 and Fit 2, and Table S1 of the SI). Since the
nearest atomic connectivities are strongly dictated by the
local geometry, the shorter Cu–S distance can be well
attributed to the smaller ionic radius of Cu1+ (0.74 ꢀ)
compared to Cd2+ (0.92 ꢀ). The overall NC still maintains
a global cubic symmetry with a Cd-rich core and a Cu-rich
surface, thus lowering the Cu–S coordination number. On the
other hand, Cd is residing largely in the interior maintaining
tetrahedral connectivity in an ideal bulk cubic fcc lattice. Our
final results indicate Cu as substitution impurity at the Cd
sites, or occupying the S-coordinated high-symmetry inter-
stitial sites, or both, residing largely near to the surface of the
NCs.
Correlating the structural information from crystallogra-
phy and the EXAFS results with analytical calculations, we
calculated the fraction of Cu and Cd atoms in the core and
surface assuming 4-S coordination in the core and 2-S
coordination at the surface. We found a Cu-rich surface
(76% Cu and 24% Cd) and a Cd-rich interior (12% Cu and
88% Cd), which clearly explains the lower bond distance and
reduced coordination number for Cu–S correlations, whereas
near-ideal Cd–S correlations were found for a cubic CdS. Our
results suggest ca. 54% CuI residing at the surface, whereas
the remaining CuI ( ꢀ 46%) participates in alloying within
84% of the entire NC volume defining a Cu-rich surface.
Clearly, these results exclude the formation of a conventional
core–shell NC structure. Additionally, small disorder (pseudo
Debye–Waller factor values), low bond distance errors, and
small energy shifts rule out the possibility of gradient alloy
formation.[34] Hence we conclude that our NCs are inhomo-
geneous CuCdS alloys containing a dramatic difference of Cu
to Cd ratio at the surface (3.2:1) and interior (0.14:1).
Hence the band structure of NCs should contain signa-
tures of the alloyed states. However, the contribution from
the interior will be CdS-dominated, whereas the contribution
from surface will be Cu2S-like considering the Cd-rich interior
and Cu-rich surface. Based on this, we have modeled a band
diagram with bulk CdS and Cu2S while accounting the shift of
the VB and CB edges of NCs due to the confinement effect
accordingly to fit with the observed transition energies
(Figure 2d and SI).[27] The band diagram shows the existence
of alloyed states (Cualloy,VB and Cualloy,CB) on the top of CdSVB
and below the CdSCB (Figure 2d). Indeed the excitation-
dependent PL spectroscopy (Figure 2a) showed the existence
Figure 2. a) Excitation-dependent PL spectra of CuCdS NCs with
excitation energies spanning from 4.4 eV to 2.4 eV. b) Overlay plots of
R-space oscillations for both Cd-K and Cu-K edges for CuCdS NCs at
298 K. c) The real component of the R-space data and the correspond-
ing fits (lines) to the Cd-K edge (open circles) and the Cu-K edge
(open triangles) for CuCdS NCs. Hanning window over the fit range
(1.0–2.8 ꢀ) in R-space is shown by dotted lines. d) Schematic band
diagram of bulk CdS, Cu2S, and CuCdS NCs with the possible
absorption, PL, and PLE pathways. The VB edges of both CdS and
Cu2S are kept at energies as near to bulk values considering their large
hole effective masses,[32] whereas the CB edges are shifted to accom-
modate the confinement-induced band gap. Existence of additional
alloyCB and alloyVB states in the NCs are marked by horizontal lines.
The band gap of bulk CdS, Cu2S, and the Cu state positions are
collected from literature.[27,33]
CdS lattice by CuI. Both of these possibilities are feasible
considering the ionic radii of Cd2+ (0.92 ꢀ) and Cu1+ (0.74 ꢀ).
However, occupation of the interstitial positions would result
in an excess of cations in the NCs, which perhaps is reflected
by the elemental analyses showing distinct signature of anion
deficiency.
We have probed the internal structure of NCs by EXAFS
measurements at the Cu-K and Cd-K edges (SI). We
extracted bond distances, coordination numbers, pseudo
Debye–Waller factors, and energy shifts for relevant atom
pair correlations using cubic CdS and Cu1.95S (F-43m) for
modeling considering the extremities of alloy formation.[29,30]
The dissimilarities in major frequency components between
the k-space oscillations for the Cd-K and Cu-K edges of the
NCs suggest different local atomic connectivities around the
metal atoms (Figure S7). The magnitude of Fourier trans-
forms (R-space oscillations) shows an apparent shortening of
the bond distances (Figure 2b) corresponding to the well-
known phase shift.[31] The real component of the R-space and
the corresponding fits to the Cd-K and Cu-K edges for CuCdS
NCs are shown in Figure 2c. Strikingly enough, we find that
the local parameters of Cd closely match to cubic CdS,
whereas the Cu–S distances appear ca. 0.3 ꢀ shorter com-
pared to Cd–S distances (Figure 2b) and the corresponding
coordination number is lowered compared to an ideal
of energy states on the top of CdSVB and/or below the CdSCB
.
The absorption supposedly takes place from the states CdSVB
/
Cualloy,VB to CdSCB/Cualloy,CB, whereas the PL originates from
the Cualloy,CB to Cualloy,VB leading to two-level transition
pathways distinct from a type-I band structure.[12] The
absorption is governed by CdS states owing to the larger
absorption cross section. One might expect two PLE peaks
Angew. Chem. Int. Ed. 2015, 54, 2643 –2648
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2645

.
Angewandte
Communications
for such a band structure corresponding to the excitation of
electrons from CdSVB and Cualloy,VB. Our emission-dependent
PLE validates the band diagram (Figures 1d and S3).
The resultant transition gap of alloyed NCs can be varied
by changing the relative cationic ratio during the synthesis
(Figure S8) or the annealing time (Figure 3a–d). The absorp-
tion and PL spectra show a gradual redshift with increasing
annealing time (Figure 3a,b). The PL spectra of NCs at
remains fixed in the mid-gap region.[27] We have collected
aliquots to quantify Cd to Cu ratios at different stages of
annealing. ICP-AES results indicate a gradual variation of Cd
to Cu ratio within the NCs showing a maximum for 90 min
annealing (Figure 3d). An increase in diameter from ca. 2 nm
to ca. 3 nm is observed from TEM images for the longest
annealed NCs (Figure S9). The XRD patterns of all annealed
NCs show zinc blende structure with a shift toward smaller
angles (Figure S9). Interestingly, the solution-phase absolute
PL QY shows an optimal value of about 20% for 60 min of
annealing (Figure 3d). Further annealing beyond 60 min
results in a decrease of QYs. Hence, the continuous shift of
the PL peak and variation of QY is associated with the change
of composition and size of the NCs during annealing.[35] At the
beginning of the reaction, Cd preferentially dominates over
Cu and with annealing time Cu is eventually being ejected
from the interior to the surfaces by a “self-purification”
mechanism.[36] Longer annealing time promotes Cu residing
at the surfaces to insert within the interior owing to the
thermal energy decreasing the Cd to Cu ratio. The interplay of
Cd to Cu ratio at different stages of reaction obviously creates
intermediate states in the resultant alloyed NCs, which affects
the recombination dynamics. Excited-state lifetimes of charge
carriers by time correlated single photon counting (TCSPC)
spectroscopy shows an increment in lifetimes with annealing
reaching a maximum value of about 1 ms for orange-emitting
NCs following the same trend with QY (Figure S10).
We further increased the QYs of alloyed NCs by over-
coating a shell layer of CdS, which can simultaneously create
an alloy with the NCs by forming a thin shell through
diffusion at higher temperatures (Figure S11). To overcoat
the alloyed NCs, TOP-S complex solution (0.1 mmol) and
cadmium oleate (0.1 mmol) was added dropwise to the NC
solution at 1708C (SI). The absorption and PL spectra of the
core–shell alloyed NCs show a blueshift of ca. 0.1 eV
compared to the uncoated NCs, whereas retaining the same
crystallographic structure (Figure S12). We have achieved
a three-fold enhancement of solution-state PL QY by shelling
the alloyed NCs, which readily gives solid-state luminescence
with QY of about 55% (Figure 3e). Importantly, the solid-
state PL traces the solution phase suggesting a minimized
energy transfer within the NCs in the solid proximity (Fig-
ure S13a). The solid-state PL stability has been monitored by
checking the emission intensity with time, while keeping the
alloyed NCs on a quartz substrate at ambient conditions. The
stability plot shows that these NCs are substantially stable in
air for nearly a year retaining bright solid-state luminescence
(Figure 3 f). Thermogravimetric analysis indicates that the
NCs are stable up to 2508C proving their thermal robustness
(Figure S13b). In addition, the temperature-dependent PL
measurement shows that the PL intensity is retained upon
cooling down to room temperature suggesting that the
luminescence is stable (Figure S13c–e). The outstanding air
and thermal stability together with the large Stokes shift and
excellent PL tunability with high QY in the solid state further
imply applicability of the ternary alloyed NCs in solid-state
lighting.
Figure 3. a) Absorption and b) PL spectra (excitation at 3.18 eV) of
CuCdS NCs in chloroform at different stages of annealing. Inset of
Figure (a): Digital photographs at different annealing time showing the
true color with irradiation of 365 nm UV light. c) Spectral positions of
the absorption maxima and the PL maxima and corresponding fits
(black and red lines) with annealing time. (d) Dependence of PL QYs
and Cd to Cu ratio (measured from ICP-AES) of CuCdS NCs with
annealing time. e) Comparison of solution phase and solid-state PL
QY of CuCdS/CdS core–shell NCs showing solid-state PL QY of 55%.
(f) Solid-state PL stability curve with time. Insets, photographs of solid
state luminescence on quartz substrate and the arrows indicate
corresponding days.
different annealing time show peaks with variable full-width
at half-maximum ranging from 0.43 eV to 0.34 eV (Fig-
ure 3b). Blue emission emerges at the initial stage of
annealing, which evolves to red with longer annealing time
eventually spanning the entire visible spectrum. The plot of
absorption and PL maxima with annealing time shows an
increment in the Stokes shift up to 1.25 eV (Figure 3c). Such
large Stokes shift is beneficial for avoiding self-absorption.
Clearly the continuous change in Stokes shift differentiates
our NCs from doped systems in which the dopant level
To test the use of the alloyed NCs in fabricating LEDs, we
designed LED devices by using alloyed NCs as active material
2646 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 2643 –2648

Angewandte
Chemie
component. The simple device architecture contains orange
color emitting core–shell NCs as active layer sandwiched in
between TPD/PEDOT:PSS-coated indium tin oxide (ITO)
anode and ZnO/Al cathode (Figure 4a). The current–voltage
characteristics of the devices showed LED responses with an
effectively low threshold voltage below ca. 2 V (Figure 4b).
device yield. Second, the alloyed NCs possess advantageous
properties like a large Stokes shift and efficient solid-state
luminescence, which are useful for fabricating LEDs. Third,
NCs are stable in air for a year retaining high luminescent QY
and good thermal stability. Fourth, the emission wavelengths
are tunable, resulting in color control over the entire visible
spectrum, a feature favorable for fabricating color-tunable
LEDs. Although further scope remains for device engineer-
ing, our preliminary device fabrication shows a low turn-on
voltage and bright EL within a range of operating voltages.
We anticipate that the solid-state luminescence from the
alloyed NCs will explore newer possibilities of similar classes
of materials as a suitable candidate for LED fabrication.
Received: September 26, 2014
Published online: January 16, 2015
Keywords: alloys · colloid synthesis · EXAFS spectroscopy ·
.
fluorescence spectroscopy · light emitting diodes
[1] C. B. Murray, D. J. Norris, M. G. Bawendi, J. Am. Chem. Soc.
1993, 115, 8706 – 8715.
[2] L. H. Qu, X. G. Peng, J. Am. Chem. Soc. 2002, 124, 2049 – 2055.
[3] C. Burda, X. Chen, R. Narayanan, M. A. El-Sayed, Chem. Rev.
2005, 105, 1025 – 1102.
[4] I. Moreels, K. Lambert, D. Smeets, D. De Muynck, T. Nollet,
J. C. Martins, F. Vanhaecke, A. Vantomme, C. Delerue, G. Allan,
Z. Hens, ACS Nano 2009, 3, 3023 – 3030.
[5] C. She, A. Demortiere, E. V. Shevchenko, M. Pelton, J. Phys.
Chem. Lett. 2011, 2, 1469 – 1475.
[6] D. V. Talapin, J.-S. Lee, M. V. Kovalenko, E. V. Shevchenko,
Chem. Rev. 2010, 110, 389 – 458.
[7] V. L. Colvin, M. C. Schlamp, A. P. Alivisatos, Nature 1994, 370,
354 – 357.
Figure 4. a) Device structure and photograph of a working LED at
a driving voltage of 5 V showing uniform EL. The device is fabricated
using CuCdS/CdS core–shell NCs annealed for 60 min. b) Current
density versus voltage characteristics of different LED devices. c) Lumi-
nance characteristics of alloyed NCs LED at different voltages. Inset:
Cyclic voltammetry curve of CuCdS/CdS NCs in N2-purged dichloro-
methane solution at room temperature taken at a scan rate of
5 mVsÀ1. d) EL spectra of the core–shell alloyed NC LEDs with
increasing driving voltages. The PL spectrum (red line) matches with
the EL spectra.
[8] H. Mattoussi, L. H. Radzilowski, B. O. Dabbousi, E. L. Thomas,
M. G. Bawendi, M. F. Rubner, J. Appl. Phys. 1998, 83, 7965 –
7974.
[9] S. Coe, W. K. Woo, M. Bawendi, V. Bulovic, Nature 2002, 420,
800 – 803.
[10] J. S. Steckel, P. Snee, S. Coe-Sullivan, J. P. Zimmer, J. E. Halpert,
P. Anikeeva, L.-A. Kim, V. Bulovic, M. G. Bawendi, Angew.
Chem. Int. Ed. 2006, 45, 5796 – 5799; Angew. Chem. 2006, 118,
5928 – 5931.
[11] A. L. Rogach, N. Gaponik, J. M. Lupton, C. Bertoni, D. E.
Gallardo, S. Dunn, N. L. Pira, M. Paderi, P. Repetto, S. G.
Romanov, C. O’Dwyer, C. M. S. Torres, A. Eychmueller, Angew.
Chem. Int. Ed. 2008, 47, 6538 – 6549; Angew. Chem. 2008, 120,
6638 – 6650.
[12] B. N. Pal, Y. Ghosh, S. Brovelli, R. Laocharoensuk, V. I. Klimov,
J. A. Hollingsworth, H. Htoon, Nano Lett. 2012, 12, 331 – 336.
[13] L. Sun, J. J. Choi, D. Stachnik, A. C. Bartnik, B.-R. Hyun, G. G.
Malliaras, T. Hanrath, F. W. Wise, Nat. Nanotechnol. 2012, 7,
369 – 373.
Above the turn-on voltage, the luminance–voltage plot is
mostly linear in a double-logarithmic scale (Figure 4c). The
cyclic voltammetry performed on core–shell NCs solution
shows a low oxidation potential of 0.85 V (inset of Figure 4c),
which may have caused the low turn-on voltage of the device.
Luminance was detectable starting from a threshold voltage
of 2 V showing a value of ~ 280 CdmÀ2 at 9 V with an external
quantum efficiencies of 0.25% (Figures 4c and S14). The
devices show bright and stable EL over a wide range of
operating voltages (Figure 4d). The EL retraces the PL
suggesting that the EL is occurring from the alloyed NCs
(Figures 4d and S14c).
In summary, we have reported on the synthesis of new
compositionally controlled CuCdS NCs in air. We demon-
strate that the structure–property relationship dictates the
recombination processes in ternary alloyed NCs. Overcoating
these NCs with CdS results in high solid-state luminescence
with high QYs. Our alloyed NCs have several attractive
features for making them suitable candidates for solid-state
lighting devices. First, the synthesis is carried out in air by
simply annealing the precursor mixtures. This implies the
mass-scale production possibilities by avoiding usual syn-
thesis complexities and holds prospects for scaling up the
´
[14] Y. Shirasaki, G. J. Supran, M. G. Bawendi, V. Bulovic, Nat.
Photonics 2013, 7, 13 – 23.
[15] T.-W. F. Chang, A. Maria, P. W. Cyr, V. Sukhovatkin, L. Levina,
E. H. Sargent, Synth. Met. 2005, 148, 257 – 261.
[16] C. R. Kagan, C. B. Murray, M. Nirmal, M. G. Bawendi, Phys.
Rev. Lett. 1996, 76, 1517 – 1520.
[17] J. Kundu, Y. Ghosh, A. M. Dennis, H. Htoon, J. A. Hollings-
worth, Nano Lett. 2012, 12, 3031 – 3037.
[18] R. Xie, M. Rutherford, X. Peng, J. Am. Chem. Soc. 2009, 131,
5691 – 5697.
[19] M. D. Regulacio, M.-Y. Han, Acc. Chem. Res. 2010, 43, 621 – 630.
Angew. Chem. Int. Ed. 2015, 54, 2643 –2648
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2647

.
Angewandte
Communications
[20] S. Jun, E. Jang, Angew. Chem. Int. Ed. 2013, 52, 679 – 682;
Angew. Chem. 2013, 125, 707 – 710.
[27] R. Viswanatha, S. Brovelli, A. Pandey, S. A. Crooker, V. I.
Klimov, Nano Lett. 2011, 11, 4753 – 4758.
[28] C. Corrado, Y. Jiang, F. Oba, M. Kozina, F. Bridges, J. Z. Zhang,
J. Phys. Chem. A 2009, 113, 3830 – 3839.
[29] B. Ravel, M. Newville, J. Synchrotron Radiat. 2005, 12, 537 – 541.
[30] M. Newville, J. Synchrotron Radiat. 2001, 8, 322 – 324.
[31] B. K. Teo, EXAFS: Basic Principles and Data Analysis, Springer,
Berlin, 1986, pp. 26 – 27.
[32] L. E. Brus, J. Chem. Phys. 1984, 80, 4403 – 4409.
[33] A. Tang, F. Teng, Y. Wang, Y. Hou, W. Han, L. Yi, M. Gao,
Nanoscale Res. Lett. 2008, 3, 502 – 507.
[34] P. K. Santra, R. Viswanatha, S. M. Daniels, N. L. Pickett, J. M.
Smith, P. O’Brien, D. D. Sarma, J. Am. Chem. Soc. 2009, 131,
470 – 477.
[21] A. Singh, S. Singh, S. Levcenko, T. Unold, F. Laffir, K. M. Ryan,
Angew. Chem. Int. Ed. 2013, 52, 9120 – 9124; Angew. Chem.
2013, 125, 9290 – 9294.
[22] H. Zhong, Z. Bai, B. Zou, J. Phys. Chem. Lett. 2012, 3, 3167 –
3175.
[23] B. Chen, H. Zhong, W. Zhang, Z. Tan, Y. Li, C. Yu, T. Zhai, Y.
Bando, S. Yang, B. Zou, Adv. Funct. Mater. 2012, 22, 2081 – 2088.
[24] H. Shen, S. Wang, H. Wang, J. Niu, L. Qian, Y. Yang, A. Titov, J.
Hyvonen, Y. Zheng, L. S. Li, ACS Appl. Mater. Interfaces 2013, 5,
4260 – 4265.
[25] Y. Zhang, C. Xie, H. Su, J. Liu, S. Pickering, Y. Wang, W. W. Yu,
J. Wang, Y. Wang, J. Hahm, N. Dellas, S. E. Mohney, J. Xu, Nano
Lett. 2011, 11, 329 – 332.
[35] W. W. Yu, X. Peng, Angew. Chem. Int. Ed. 2002, 41, 2368 – 2371;
Angew. Chem. 2002, 114, 2474 – 2477.
[36] G. M. Dalpian, J. R. Chelikowsky, Phys. Rev. Lett. 2006, 96,
226802.
[26] S. Cadars, B. J. Smith, J. D. Epping, S. Acharya, N. Belman, Y.
Golan, B. F. Chmelka, Phys. Rev. Lett. 2009, 103, 136802.
2648 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 2643 –2648