Home
Search
Collections
Journals
About
Contact us
My IOPscience
Efficient white light emitting diodes based on Cu-doped ZnInS/ZnS core/shell quantum dots
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 Nanotechnology 25 435202 (http://iopscience.iop.org/0957-4484/25/43/435202) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 129.173.72.87 This content was downloaded on 08/10/2014 at 02:24
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 25 (2014) 435202 (8pp)
doi:10.1088/0957-4484/25/43/435202
Efficient white light emitting diodes based on Cu-doped ZnInS/ZnS core/shell quantum dots Xi Yuan1,2,3, Jie Hua1, Ruosheng Zeng4, Dehua Zhu5, Wenyu Ji2, Pengtao Jing2, Xiangdong Meng1, Jialong Zhao1 and Haibo Li1 1
Key Laboratory of Functional Materials Physics and Chemistry of the Ministry of Education, Jilin Normal University, Siping 136000, People’s Republic of China 2 State Key Laboratory of Luminescence and Applications, Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, People’s Republic of China 3 University of Chinese Academy of Sciences, Beijing 100039, People’s Republic of China 4 Key Lab for Functional Materials Chemistry of Guizhou Province, School of Chemistry and Materials Science, Guizhou Normal University, Guiyang 550001, People’s Republic of China 5 College of Mechanical and Electrical Engineering, Wenzhou University, Wenzhou 325035, People’s Republic of China E-mail: [email protected] and [email protected] Received 11 March 2014, revised 30 July 2014 Accepted for publication 27 August 2014 Published 7 October 2014 Abstract
We report the fabrication of efficient white light-emitting diodes (WLEDs) based on Cu : ZnInS/ ZnS core/shell quantum dots (QDs) with super large Stokes shifts. The composition-controllable Cu : ZnInS/ZnS QDs with a tunable emission from deep red to green were prepared by a one-pot noninjection synthetic approach. The high performance Cu : ZnInS QD-WLEDs with the colour rendering index up to 96, luminous efficacy of 70–78 lm W−1, and colour temperature of 3800–5760 K were successfully fabricated by integration of red and green Cu-doped QDs. Negligible energy transfer between Cu-doped QDs was clearly found by measuring the photoluminescence lifetimes of the QDs, consistent with the small spectral overlap between QD emission and absorption. The experimental results indicated low toxic Cu : ZnInS/ZnS QDs could be suitable for solid state lighting. S Online supplementary data available from stacks.iop.org/NANO/25/435202/mmedia Keywords: quantum dots, nanocrystals, Cu-doped quantum dots, white light-emitting diodes, energy transfer commercial phosphors (Y3Al5O12 : Ce3+) show a deficient emission in the red spectral region and a serious scattering effect, which limit the colour rendering index (CRI) and luminescent efficiency of the WLEDs [7–8]. Therefore, it is necessary to find novel materials for fabricating high colour rendering WLEDs with high efficiency. Semiconductor quantum dots (QDs) as potential converters for WLEDs exhibit size-tunable emissions, high photoluminescence quantum yields (PL QYs), low scattering and good colour saturation, compared to traditional phosphors [9–14]. The QD-WLEDs have been successfully demonstrated by employing the combination of red and green light-
1. Introduction Energy-efficient lighting has offered a promising option for energy saving because artificial lighting globally consumes about 20% of the total electrical energy in the world [1]. Environmentally friendly white light-emitting diodes (WLEDs) with higher efficiencies, longer lifetimes, and fast response times are considered as promising light sources to replace the traditional ones such as incandescent or fluorescent lamps [2–6]. Currently conventional WLEDs have been fabricated by integrating down-converting phosphors with blue InGaN/GaN LEDs. However, the most popular 0957-4484/14/435202+08$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
Nanotechnology 25 (2014) 435202
Xi Yuan et al
emitting CdSe QDs on blue InGaN/GaN LEDs [15–19]. The CdSe QD-WLEDs were produced for display backlighting and they showed luminous efficacy of 41 lm W−1 and colour reproducibility of 100% compared to the National Television Systems Committee colour space [15]. However, the toxicity of the cadmium element and the significant self-absorption and energy transfer in closely packed QDs with small Stokes shifts potentially hinders their ultimate research transformation and commercialization [14, 15, 20]. Recently, low toxic CuInS2-based QDs with large Stokes shifts and high PL QYs (60%–70%) were used for fabricating high colour rendering WLEDs [21]. The devices integrating green and red emitting CuInS2-based QDs with blue chips showed a CRI of 95 and luminous efficacy of ∼69 lm W−1 [21]. Despite all, the spectral overlap between the green CuInS2-based QD emission and red QD absorption in the WLEDs might result in significant energy transfer and re-absorption processes [20, 22], influencing the optoelectronic performances of the devices [23, 24]. Thus, it is desirable to develop the ideal QDs with low toxicity, high PL QY, and enough large Stokes shift for WLEDs. Transition metal ions (such as Mn and Cu) doped QDs exhibited good thermal and chemical stability [25], and super large Stokes shift [14, 26, 27]. Mn-doped QDs with a yelloworange emission and a PL QY of ∼68% were reported in our previous works [25, 28]. However, the Mn-doped QDs with the emission wavelength of (580–600 nm) are not suitable for fabrication of high colour rendering WLED. It is well known that the emission wavelength of Cu-doped QDs can be varied from blue to near-infrared by changing the size of QDs and composition of hosts [14, 26, 27, 29, 30]. The emission in Cudoped QDs is considered to come from the radiative recombination of the electron in the conduction band of host material and the hole in Cu T2 state [31, 32]. For example, Cu : ZnSe/S QDs show a tunable emission from 485 to 540 nm [33], Cu : ZnInSe from 540 to 660 nm [26], and Cu : InP from 630 to 1100 nm [34]. On the other hand, Wang et al fabricated high colour rendering WLED based on a blue LED combined with highly luminescent CdS:Cu/ZnS QDs and YAG : Ce phosphors. The as-fabricated WLEDs showed a high luminous efficacy of 37.4 lm W−1, good colour chromatics stability, and CRI value of 86 [14]. No re-absorption between the QDs and phosphors could occur, suggesting the advantage of doped QDs with super large Stokes shifts. More recently, the composition-tunable emission from 450 to 810 nm was reported from Cu doped ZnInS QDs with an absorption band edge of up to 540 nm [30], which is promising for high colour rendering WLED based on blue chips (450 nm). In this work, we report the integration of Cu : ZnInS/ZnS core/shell QDs for warm WLEDs with high colour rendering and high efficiency. The Cu : ZnInS/ZnS QDs with composition-tunable emission were synthesized by a one-pot noninjection approach. The performance parameters of the Cu : ZnInS QD-WLEDs with blue InGaN chips (450 nm) to excite QD resin films with various QD concentrations were measured. Further the PL lifetimes of Cu : ZnInS QDs was studied by time-resolved PL spectroscopy to understand the
PL quenching of the doped QDs and energy transfer processes between the QDs in WLEDs.
2. Experimental section 2.1. Materials
Zinc acetate (Zn(OAc)2, 99.99%), copper acetate (Cu(OAc)2, 99.99%), indium acetate (In(OAc)3, 99.99%), sulfur powder (S, 99.99%), dodecanethiol (DDT, 99.9%), oleylamine (OAm, 97%), 1-octadecene (ODE, 90%), were purchased from Aldrich. All chemicals were used without further purification. 2.2. Synthesis of Cu : ZnInS and Cu : ZnInS/ZnS QDs
The Cu : ZnInS and Cu : ZnInS/ZnS QDs used in this experiment were synthesized via a one-pot noninjection synthetic approach similar to our previous works [30]. In a typical procedure, Cu(OAc)2 (0.02 mmol), Zn(OAc)2 (0.2 mmol), In(OAc)3 (0.2 mmol), sulfur powder (1.6 mmol) together with 4.0 mL of DDT, and 6.0 mL of OAm were loaded in a 50 mL three-neck flask clamped in a heating mantle. The mixture was heated to 220 °C with a heating rate of 15 °C min−1 under argon flow and kept at this temperature for 10 min to allow the growth of Cu : ZnInS cores. Then the reaction mixture was allowed to cool to 100 °C, the Zn precursor (obtained by dissolving 0.4 mmol of Zn(OAc)2 in 0.2 mL of OAm and 0.8 mL of ODE) was then introduced into the reaction mixture. Following this, the system temperature was raised to 240 °C with a heating rate of 15 °C min−1 and kept at this temperature for 20 min to allow the overgrowth of ZnS shell around the preformed Cu : ZnInS cores. To obtain Cu : ZnInS cores with different Zn/In ratios, the Zn/In precursor ratio was varied while the gross amount of Zn and In precursors and all other variables were remained constant. The Zn/In precursor ratio was varied with the values of 1/1, 1/2, 1/3, 1/5 and 1/8 for green, yellow, orange, red, and deep red QDs, respectively. The obtained QDs were precipitated by adding methanol into the toluene solution and purified by repeated centrifugation and decantation. 2.3. Device fabrication
Cu : ZnInS/ZnS QDs (0.1–0.2 g) in chloroform (3 mL) was blended with 0.3 g of thermally curable silicone resin under vigorous stirring. Then this QD-resin mixture was placed on hot plate at 50 °C for 1 h to evaporate the chloroform. Subsequently, 0.15 g of a hardener was added to the above QDresin re-mixture. The composite was put into a vacuum oven for 30 min at 50 °C to eliminate any air bubbles. The 450 nm emitting InGaN-based blue LED (device efficacy of about 30 lm W−1 at 20 mA, Sanan Optoelectronics, China) without an silicone package was used to fabricate the QD-based LEDs. The QD paste was dispensed on a blue LED mold and then thermally cured by a two-step process of 100 °C for 30 min, followed by 120 °C for 1 h. For comparison, the fabrication of the WLEDs based on CdSe/ZnS QDs (Najing 2
Nanotechnology 25 (2014) 435202
Xi Yuan et al
spectrometer with an integrating sphere. All measurements were carried out at room temperature.
3. Results and discussion Figure 1 shows UV-visible absorption of Cu : ZnInS/ZnS core/shell QDs and the corresponding normalized PL spectra under excitation at 450 nm. The absorption spectra become steep around 350 nm due to the ZnS shell coating [21]. No distinct excitonic absorption peak is observed, which could be attributed to their irregular elemental distribution and/or intra band gap states [35]. Both the absorption onset and peak wavelength of the emission are red-shifted with increasing In/ Zn precursor ratio in the cores from 1/1 to 8/1. The red shift could be explained by the change in the band gap of the alloy structures with the different ratio of ZnS (bandgap of 3.7 eV) and In2S3 (bandgap of 2.1 eV), since the emission in Cudoped QDs derives from the radiative recombination of the electron in the conduction band of host material and the hole in Cu T2 state [29–32]. Therefore, the Cu : ZnInS QDs exhibit a super large Stokes-shift of about 140–180 nm, compared with that of undoped CdSe QDs about 20 nm [15]. The Cu : ZnInS/ZnS QDs exhibit an emission peak at 525, 560, 590, 630, and 670 nm, called as G-, Y-, O-, R-, and Dr-QDs, respectively, covering the green to red region of the visible spectrum, and they possess the PL QYs of 40–60%. The full width at half maximum of the Cu : ZnInS QD emissions are approximately 90–130 nm, which might be beneficial in solid state lighting due to the large spectral coverage [24]. Structural characterization for the typical green and red emissive Cu-doped QDs was carried out by means of TEM and XRD, which were shown in figure 2. The TEM images demonstrate that the G- and R-QDs are near-spherical in shape and monodispersed. The average diameters of the Gand R-QDs were estimated to be 3.3 and 3.9 nm, respectively. The QDs exhibit clear lattice fringes in the high-resolution TEM images as shown in the inset of figures 2(a), (b), which suggests the highly crystalline nature. The wide-angle XRD patterns for the Cu-doped QDs show the characteristic peaks of the zinc blende (cubic) structure, similar to those reported before [30]. The Cu : ZnInS QDs with small sizes, tunable broad emissions, high PL QYs, and super large Stokes-shifts are promising for fabrication of WLEDs [2, 36]. For the fabrication of QD-based LEDs, Cu : ZnInS/ZnS QDs with green, yellow, orange, red, and deep red emissions were utilized as colour converters and combined with a blue (450 nm) emitting LED chip. The normalized electroluminescence (EL) spectra of Cu : ZnInS/ZnS QD-based LED devices are shown in figure 3(a). To evaluate the quality of the light emitted from Cu : ZnInS QD-based LEDs, several parameters, such as Commission Internationale de l’Eclairage (CIE) chromaticity coordinates, CRI, correlated colour temperature (CCT), and luminous efficacy need to be determined [2, 37]. As seen in figure 3(b), the CIE colour coordinates of Cu : ZnInS QD-based LEDs with various concentrations of QDs in the silicone resin can cover the white light region. With increasing the concentration of QDs, the CIE
Figure 1. UV-visible absorption (a) and PL spectra (b) of
Cu : ZnInS/ZnS core/shell QDs in chloroform with emission peaks at 525 nm (black lines), 560 nm (red lines), 590 nm (green lines), 630 nm (blue lines), and 670 nm (cyan lines) under excitation at 450 nm.
Technology Corporation Limited, L51 and L55, PL QYs over 40%) was performed, which was similar to the method for Cu : ZnInS/ZnS QD-WLEDs. The weight ratio of green CdSe/ZnS QDs to the red ones was 7/1.
2.4. Characterization
The absorption spectra were recorded by a UV-3101PC UVvis-NIR scanning spectrophotometer (Shimadzu). The PL spectra were recorded by a Hitachi F-7000 spectrophotometer. The time-resolved PL spectra were measured by an FL920 fluorescence lifetime spectrometer (Edinburgh Instruments). The excitation source was a hydrogen flash lamp (nF900) with a pulse width of 1.5 ns. The size and shape of the QDs were measured by a Tecnai G2 transmission electron microscope (TEM) operated at 200 kV. X-ray diffraction (XRD) patterns of the QDs were collected by a Bruker XRD spectrometer with a Cu Kα line of 0.15418 nm at a scanning step of 0.02°. The PL QYs of QDs were estimated by comparing the integrated emission of the QD samples in solution with that of standard dye solutions with identical optical density. The optical properties of QDWLEDs were measured by an USB4000 Ocean Optics 3
Nanotechnology 25 (2014) 435202
Xi Yuan et al
Figure 2. TEM images of G-Cu : ZnInS/ZnS (3.3 nm) (a) and R-Cu : ZnInS/ZnS (3.9 nm) (b) core/shell QDs. The corresponding high
resolution TEM images are shown in the inset. (c) XRD patterns of G-Cu : ZnInS/ZnS and R-Cu : ZnInS/ZnS core/shell QDs.
remarkably enlarged Stokes shift for the doped QDs (140–180 nm) was observed, compared to the undoped ones (12 nm). To quantify the spectral overlap, we calculated the overlap integral J with the equation [40]:
coordinates of the LEDs using the same QDs show a nearly linear increase trend. The slope of these lines monotonously decreases from 2.23 to 0.61 as the QD emission wavelength changes from green to deep red. These lines could guide us for the WLED design and provide the advantages for colour displays. The device parameters for typical LEDs based on G-, Y-, O-, R-, and Dr-Cu : ZnInS/ZnS QDs operated at 20 mA are listed in table 1. The luminous efficiencies of the LEDs with green-, yellow-, orange-, red-, and deep red-emitting QDs were 87.2, 76.9, 53.5, 34.6, and 12.1 lm W−1, respectively. The superiority in luminous efficacy of a green doped QDbased LED relative to the red doped QD-based one mainly originates from the variety of colour sensitivity [37]. The CCT values were determined in a relatively wide range of 8110–1809 K, which shows a decrease with changing the QD emission wavelength from green to red. The CRI values for the WLEDs based on G- or Y- Cu : ZnInS QDs were in the range of 71–74. This indicates that the Cu : ZnInS QDs are good colour converters for the next generation of white solid state lighting. The blue chip-to-QD light conversion efficiency was calculated from the integrated spectral ratio of the QD emission intensity to the absorbed source light intensity, based on the reduction of the original blue LED spectrum [38]. The light conversion efficiency of Cu : ZnInS QD-based LEDs and PL QYs of the Cu-doped QDs in solution are shown in figure 4(a). The light conversion efficiency deceased from 57% to 39% for different colour Cu : ZnInS QD-based LEDs, which showed a similar trend as the light-emitting efficiencies of the QDs in solutions. As we know that both the strong energy transfer/self-absorption and the surface defects of QDs generated through the LED fabrication (such as high temperatures for silicone encapsulation) have a detrimental impact on the efficiencies of the LEDs [15, 20, 39]. The degree of energy transfer and self-absorption in the QD ensembles significantly depends on the spectral overlap between absorption and emission spectra [40]. The absorption and emission spectra for Cu-doped QDs and undoped QDs, such as CdSe, were shown in supplementary figure S1, a
J=
∫0
∞
FD (λ) εA (λ) λ4dλ ,
where FD(λ) is normalized donor emission spectrum, and εA(λ) is the acceptor molar extinction coefficient. The overlap integral J value for CdSe QDs is 3.434 × 1016 M−1 cm−1 nm4, while that for G-, Y-, O-, R-, and Dr-Cu : ZnInS QDs is 2.545 × 1014, 1.648 × 1014, 4.463 × 1014, 2.808 × 1014, and 2.362 × 1014 M−1 cm−1 nm4, respectively. The overlap integrals for the Cu-doped QDs are two orders of magnitude lower than that of the undoped QDs, owing to the super large Stokes shift in the doped QDs [41]. This significant difference suggests that energy transfer and self-absorption between the Cu : ZnInS QDs could be negligible. In addition, the luminescence decay curves of the G-Cu : ZnInS QDs in chloroform solution and in silicone resin (thermally cured as the LED fabrication) were measured by time-resolved PL spectroscopy, respectively, to investigate whether an additional nonradiative de-excitation channel of QDs occurred during the LED fabrication, as seen in figure 4(b). We analyzed the PL decays by a bi-exponential function with two time components (τi) and weights (Ai). The amplitude-weighted lifetimes were obtained by a relation τav = (A1τ1 + A2τ2)/(A1 + A2). The lifetimes of the G-Cu : ZnInS QDs in solution and silicone resin were estimated to be 432 and 419 ns, respectively, basically the same. This result confirmed that almost no additional nonradiative channels (e.g. energy transfer or thermally induced surface traps) occurred in the LEDs. This demonstrates that the energy transfer and thermally curing induced luminous efficacy reduction is not significant in Cu : ZnInS QD-based LEDs, which makes the Cu-doped QDs ideal emitters for WLEDs. In addition, the emission in Cudoped QDs is considered to come from the recombination of the electron in the conduction band with the hole in Cu T2 state. Due to the incorporation of an additional energy state by Cu-doping, the excited-state lifetime of the dopant emission is 4
Nanotechnology 25 (2014) 435202
Xi Yuan et al
Table 1. The luminous efficacy (LE), CIE colour coordinates, CCT, and CRI of the typical LEDs based on G-, Y-, O-, R-, and DrCu : ZnInS/ZnS QDs operated at 20 mA, respectively.
LE Samples G-Cu : ZnInS Y-Cu : ZnInS O-Cu : ZnInS R-Cu : ZnInS Dr-Cu : ZnInS
CIE −1
(lm W ) 87.2 76.9 53.5 34.6 12.1
(x, y) (0.2821, (0.3481, (0.3713, (0.3432, (0.3380,
0.3392) 0.3138) 0.2486) 0.1667) 0.1499)
CCT (K)
CRI
8110 4682 2645 1962 1809
71 74 59 — —
Figure 3. (a) Normalized luminescence spectra of Cu : ZnInS QD-
based LED devices operated at a forward current of 20 mA. Photographs of the 20 mA-driven QD-based LEDs fabricated with G-Cu : ZnInS/ZnS and R-Cu : ZnInS/ZnS QDs are shown in the inset, respectively. (b) CIE colour coordinates of Cu : ZnInS QDbased LEDs with various concentrations of QDs in the silicone resin. The black dashed line represents the blackbody locus curve.
Figure 4. (a) The PL QYs (black squares) of the Cu-doped QDs with different emission wavelength in chloroform solution, and light conversion efficiencies (blue circles) of Cu-doped QD-based LEDs. (b) Temporal change of PL intensities for G-Cu : ZnInS/ZnS QDs in chloroform solution (black squares), in resin (green circles), and in G- and R-Cu : ZnInS QD blend resin film (blue triangles), respectively. The solid red lines represent fitting curves.
longer than the excitonic emission of QDs, reaching up to several microseconds or hundred-nanoseconds [42]. Further, the EL spectra of the Cu-doped QDs exhibit a small red-shifts (around 10 nm) in comparison to the PL spectra of the corresponding QDs in solution, which may be caused by the different dielectric constant of the QD-surrounding medium (chloroform versus silicone resin) [13, 43]. It is noted that the red-shifts for Cu : ZnInS QDs are extremely smaller than that for alloyed CuInS2-ZnS QDs (about 75 nm) [21], possibly because strong energy transfer between CuInS2-ZnS QDs might result in a large red-shift of the emission [21]. Therefore, the experimental results indicate that the Cu-doped ZnInS QDs are ideal candidates for fabricating high efficiency WLED [2].
Although QD-WLEDs can be fabricated by using a single type of Cu : ZnInS/ZnS QDs, the colour rendering property (CRI of 59–74) does not still satisfy the requirements for specific lighting applications (e.g. indoor illumination, CRI > 80) [2, 19]. The combination of green and red emitting QDs can provide an effective way to improve the lighting quality of WLEDs. In general, strong energy transfer and reabsorption between green and red undoped QDs such as CdSe nanocrystals have a detrimental impact on the luminous efficacy of WLEDs with increasing the QD packing density or 5
Nanotechnology 25 (2014) 435202
Xi Yuan et al
Table 2. The CRI, luminous efficacy (LE), CCT, and CIE colour
coordinates of the as-fabricated WLEDs based on G- and RCu : ZnInS/ZnS QD blends with different weight ratios (from 1.6/1 to 0.9/1) operated at 20 mA. LE
CCT −1
G/R
CRI
(lm W )
(K)
1.6/1 1.4/1 1.2/1 1/1 0.9/1
94 95 96 96 96
78.3 74.7 72.5 71.2 70.5
5760 5198 4655 4147 3804
CIE (x, y) (0.3271, (0.3388, (0.3528, (0.3710, (0.3863,
0.3232) 0.3337) 0.3416) 0.3596) 0.3721)
figure 4(b), similar to that in G-QD neat resin film. The PL lifetime of G-Cu : ZnInS QDs in green and red QD blend resin films (403 ns) is nearly the same as that in green QD neat resin film (419 ns). The corresponding energy transfer efficiency is only 4%, obtained by an expression 1 − τdonoracceptor/τdonor. This indicates that energy transfer between Gand R-Cu : ZnInS/ZnS QDs can be negligible, making us to use the green and red Cu : ZnInS QDs together for the WLEDs without a significant reduction in luminous efficacy. In multiphase emitters-based WLEDs, the intensity ratio of the emissions with different colours has a great impact on the quality of devices. The performance parameters of the QD-WLEDs fabricated with the QD blend (G- and RCu : ZnInS with different weight ratio) and blue LED chips and operated at 20 mA are summarized in table 2. The typical EL spectra of the WLED with G- and R-Cu : ZnInS weight ratio of 1.4/1 at different forward current from 5 to 100 mA are shown in figure 5. Three distinct peaks at 450, 540, and 620 nm, corresponding to blue chip, green and red Cu-doped QDs, respectively, are clearly observed in the broad EL spectrum from 430 to 800 nm, as seen in figure 5(a). Owing to such a spectral extension, the G- and R-Cu : ZnInS QD blendbased WLEDs exhibit an improved CRI value of 94–96 as listed in table 2, and the corresponding luminous efficiencies of above 70 lm W−1, which are better than those for R- and GCdSe QD-based WLEDs in the literature (luminous efficacy of 41 lm W−1) [15]. Besides, for fair comparison, the CdSe/ ZnS QD-WLEDs were fabricated with the same type of blue LED as used in the Cu : ZnInS QD-WLEDs. The corresponding EL spectrum is shown in supplementary figure S2, the luminous efficacy is 58 lm W−1, the CRI value is 84, the CCT is 5043 K, and the colour coordinate is (0.3421, 0.3298). It is found that, at the similar coordinates (0.34, 0.33) and CCT (about 5100 K), the luminous efficacy of Cu : ZnInS QD-WLED (1.4/1) can reach 74.7 lm W−1, while the CdSe/ ZnS system can only reach 58 lm W−1. Obviously, the improved efficacy could be accounted by high PL QYs of the Cu : ZnInS/ZnS QDs and negligible energy transfer between the doped QDs. Furthermore, the optimized WLEDs exhibit a tunable CCT between 3800 and 5760 K by finely tuning the ratio and amount of the G-Cu : ZnInS QDs to the red ones, as seen in table 2, which can provide warm enough CCT (80) white light for indoor illumination applications [2, 19]. The CIE colour coordinates
Figure 5. (a) EL spectra of a QD-WLED fabricated with G- and R-
Cu : ZnInS QDs (weight ratio of 1.4/1) as a function of forward current from 5 to 100 mA. (b) CIE colour coordinates of the QDWLED under different forward bias currents. The black dashed line represents the blackbody locus curve.
loading fraction [15, 35]. Many undoped QD-based WLEDs suffered from the embarrassing situation [13, 44, 45]. Moreover, QDs with shorter wavelength need a high content ratio to the longer wavelength ones to obtain high CRI values due to the energy transfer. Therefore, the energy transfer between G- and R-Cu : ZnInS/ZnS QDs was specially investigated by time-resolved PL spectroscopy. The energy transfer process acting as an additional nonradiative de-excitation channel for the donor can be described by the following expression 1/ τdonor-acceptor = 1/τdonor + kET, where τdonor-acceptor and τdonor are the PL lifetimes of G-Cu : ZnInS QDs donor in green and red QD blend and green QD neat resin films, respectively, and kET is the energy transfer rate [40]. The luminescence decay curves of the G-Cu : ZnInS QDs in G- and R-Cu : ZnInS QD blend films (weight ratio is 1/1) were measured, as seen in 6
Nanotechnology 25 (2014) 435202
Xi Yuan et al
4. Conclusions In summary, we have fabricated high performance Cu : ZnInS/ZnS QD-WLEDs for the next generation solid state lighting. Highly luminescent Cu : ZnInS QDs with PL QYs of 40–60% and tunable emissions from green to deep red were obtained by effectively controlling the component of the QDs. The optimized warm WLEDs based on G- and RCu : ZnInS/ZnS core/shell QDs and blue LED chips exhibited high CRI up to 96, luminous efficacy of 70–78 lm W−1, CCT of 3800–5760 K, and CIE colour coordinates of (0.3388, 0.3337), which are the best values currently reported for QDWLEDs. No significant change in PL lifetimes of the GCu : ZnInS/ZnS QDs in G- and R-QD blend films was observed, compared with those in solution. This indicated that the energy transfer between G- and R-Cu : ZnInS QDs in WLEDs could be negligible. This resulted in a higher luminous efficacy in doped QD-WLEDs than the undoped QD devices. These experimental results suggest that low toxic Cu : ZnInS QDs are promising colour converting materials for fabricating high colour rendering WLEDs with high efficiency.
Figure 6. (a) CRI of the fluorescent lamp (Panasonic
YZ18RR6500K) and G- and R-Cu : ZnInS QD-WLED operated at 100 mA. Photographs of colourful candy are compared under the illumination with sun light (b), a fluorescent lamp (c), and a Cu : ZnInS QD-based WLED operated at 100 mA.
Acknowledgments
of the WLED with G/R weight ratio of 1.4/1 are obtained to be (0.3388, 0.3337) under 20 mA forward bias current, which are close to the equi-energy white point (0.3333, 0.3333). The values exhibit little change under different forward bias currents, as seen in figure 5(b), indicating good colour stability of the output light of the WLED. The detailed CRI values of R1-R9 and the general CRI (designated by the symbol Ra, the average value of R1-R8) of both the fluorescent lamp and the as-fabricated Cu : ZnInS QD-WLEDs operated at 100 mA are shown in figure 6(a). In comparison with the fluorescent lamp (Panasonic YZ18RR6500K) with Ra of 71, the WLED based on G- and R-Cu : ZnInS QDs exhibits a higher Ra of 95 and the improvement in the colour reproduction for the colour from R1 to R8. It is worth noting that the special CRI R9, corresponding to the colour rendering of the deep-red region, which is important for biomedical and painting appreciation applications [46, 47], is greatly improved to 97. This value is also much higher than that of the commercial YAG : Ce-based WLED (R9–15), because the emission wavelength of commercial red-emitting phosphors is commonly below 630 nm [14]. The R-Cu : ZnInS QDs with the broad emission band result in the enhancement of the colour rendering in both the red and deep-red spectral regions and show their potential applications in biomedical and painting appreciation. Further, the photographs of colourful candies are compared under the illumination with sun light, a fluorescent lamp, and the Cu : ZnInS QD-based WLED are shown in figures 6(b)–(d). This Cu : ZnInS QD-based WLED demonstrates chromatic light quality with a CRI of 95 and great colour saturation, which is similar to that of the sun light.
This work was supported by the National Natural Science Foundation of China (Nos. 11274304, 11204298, 61205025, 61275047, 21371071, 21461006 and 51472053), Program for the Development of Science and Technology of Jilin Province (No. 20120344), and the Research Project of Chinese Ministry of Education (No. 213009A).
References [1] Tsao J Y 2002 Light Emitting Diodes (LEDs) for General Illumination: An OIDA Technology Roadmap Update (Washington, DC, USA: Optoelectronics Industry Development Association (OIDA)) [2] Dai Q Q, Duty C E and Hu M Z 2010 Small 6 1577–88 [3] Ziegler J, Xu S, Kucur E, Meister F, Batentschuk M, Gindele F and Nann T 2008 Adv. Mater. 20 4068–73 [4] Schubert E F and Kim J K 2005 Science 308 1274–8 [5] Nakamura S, Mukai T and Senoh M 1994 Appl. Phys. Lett. 64 1687–9 [6] Li X F, Budai J D, Liu F, Howe J Y, Zhang J H, Wang X J, Gu Z J, Sun C J, Meltzer R S and Pan Z W 2013 Light: Sci. Appl. 2 e50 [7] Lin C C and Liu R S 2011 J. Phys. Chem. Lett. 2 1268–77 [8] Jang H S, Yang H, Kim S W, Han J Y, Lee S G and Jeon D Y 2008 Adv. Mater. 20 2696–702 [9] Coe S, Woo W K, Bawendi M G and Bulović V 2002 Nature 420 800–3 [10] Aboulaich A, Michalska M, Schneider R, Potdevin A, Deschamps J, Deloncle R, Chadeyron G and Mahiou R 2014 ACS Appl. Mater. Interfaces 6 252–8 [11] Dohnalová K, Poddubny A N, Prokofiev A A, Boer W D, Umesh C P, Paulusse J M, Zuilhof H and Gregorkiewicz T 2013 Light: Sci. Appl. 2 e47 7
Nanotechnology 25 (2014) 435202
Xi Yuan et al
[29] Zhang W J, Zhou X G and Zhong X H 2012 Inorg. Chem. 51 3579–87 [30] Zhang W J, Lou Q, Ji W Y, Zhao J L and Zhong X H 2014 Chem. Mater. 26 1204–12 [31] Mandal P, Talwar S S, Major S S and Srinivasa R S 2008 J. Chem. Phys. 128 114703 [32] Corrado C, Jiang Y, Oba F, Kozina M, Bridges F and Zhang J Z 2009 J. Phys. Chem. A 113 3830–9 [33] Jana S, Srivastava B B, Acharya S, Santra P K, Jana N R, Sarma D D and Pradhan N 2010 Chem. Commun. 46 2853–5 [34] Xie R G and Peng X G 2009 J. Am. Chem. Soc. 131 10645–51 [35] Chung W, Jung H, Lee C H and Kim S H 2012 Opt. Express 20 25071–6 [36] Bowers M J, McBride J R and Rosenthal S J 2005 J. Am. Chem. Soc. 127 15378–9 [37] Demir H V, Nizamoglu S, Erdem T, Mutlugun E, Gaponik N and Eychmüller A 2011 Nano Today 6 632–47 [38] Kim J K, Luo H, Schubert E F, Cho J, Sone C and Park Y 2005 Japan. J. Appl. Phys. 44 L649–51 [39] Phillips J M et al 2007 Laser Photonics Rev. 1 307–33 [40] Lakowicz J R 2006 Principles of Fluorescence Spectroscopy 3rd edn (Berlin: Springer) pp 443–527 [41] Norris D J, Efros A L and Erwin S C 2008 Science 319 1776–9 [42] Wu P and Yan X P 2013 Chem. Soc. Rev. 42 5489–521 [43] Song H and Lee S 2007 Nanotechnology 18 255202 [44] Li Y Q, Rizzo A, Cingolani R and Gigli G 2006 Adv. Mater. 18 2545–8 [45] Cheng G, Mazzeo M, Agostino S, Sala F, Carallo S and Gigli G 2010 Opt. Lett. 35 616–8 [46] Krames M R, Shchekin O B, Mueller-Mach R, Mueller G O, Zhou L, Harbers G and Craford M G 2007 J. Disp. Technol. 3 160–75 [47] Taguchi T and Kono M 2007 J. Light Vis. Environ. 31 149–51
[12] Sun M Y, Zhu D H, Ji W Y, Jing P T, Wang X Y, Xiang W D and Zhao J L 2013 ACS Appl. Mater. Interfaces 5 12681–8 [13] Wang X B, Li W W and Sun K 2011 J. Mater. Chem. 21 8558–65 [14] Wang X B, Yan X S, Li W W and Sun K 2012 Adv. Mater. 24 2742–7 [15] Jang E, Jun S, Jang H, Lim J, Kim B and Kim Y 2010 Adv. Mater. 22 3076–80 [16] Lita A, Washington A L II, Burgt L V D, Strouse G F and Stiegman A E 2010 Adv. Mater. 22 3987–91 [17] Chen H S, Hsu C K and Hong H Y 2006 IEEE Photonics Technol. Lett. 18 193–5 [18] Jun S, Lee J and Jang E 2013 ACS Nano 7 1472–7 [19] Nizamoglu S, Zengin G and Demira H V 2008 Appl. Phys. Lett. 92 031102 [20] Kundu J, Ghosh Y, Dennis A M, Htoon H and Hollingsworth J A 2012 Nano Lett. 12 3031–7 [21] Chen B K, Zhong H Z, Wang M X, Liu R B and Zou B S 2013 Nanoscale 5 3514–9 [22] Yuan X, Zhao J L, Jing P T, Zhang W J, Li H B, Zhang L G, Zhong X H and Masumoto Y 2012 J. Phys. Chem. C 116 11973–9 [23] Ku K H, Kim M P, Paek K, Shin J M, Chung S, Jang S G, Chae W-S, Yi G-R and Kim B J 2013 Small 9 2667–72 [24] Song W-S and Yang H 2012 Appl. Phys. Lett. 100 183104 [25] Yuan X, Zheng J J, Zeng R S, Jing P T, Ji W Y, Zhao J L, Yang W Y and Li H B 2014 Nanoscale 6 300–7 [26] Sarkar S, Karan N S and Pradhan N 2011 Angew. Chem. Int. Ed. 50 6065–9 [27] Maity A R, Palmal S, Basiruddin S K, Karan N S, Sarkar S, Pradhan N and Jana N R 2013 Nanoscale 5 5506–13 [28] Cao S, Zheng J J, Zhao J L, Wang L, Gao F M, Wei G D, Zeng R S, Tian L H and Yang W Y 2013 J. Mater. Chem. C 1 2540–7
8
Search
Collections
Journals
About
Contact us
My IOPscience
Efficient white light emitting diodes based on Cu-doped ZnInS/ZnS core/shell quantum dots
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 Nanotechnology 25 435202 (http://iopscience.iop.org/0957-4484/25/43/435202) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 129.173.72.87 This content was downloaded on 08/10/2014 at 02:24
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 25 (2014) 435202 (8pp)
doi:10.1088/0957-4484/25/43/435202
Efficient white light emitting diodes based on Cu-doped ZnInS/ZnS core/shell quantum dots Xi Yuan1,2,3, Jie Hua1, Ruosheng Zeng4, Dehua Zhu5, Wenyu Ji2, Pengtao Jing2, Xiangdong Meng1, Jialong Zhao1 and Haibo Li1 1
Key Laboratory of Functional Materials Physics and Chemistry of the Ministry of Education, Jilin Normal University, Siping 136000, People’s Republic of China 2 State Key Laboratory of Luminescence and Applications, Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, People’s Republic of China 3 University of Chinese Academy of Sciences, Beijing 100039, People’s Republic of China 4 Key Lab for Functional Materials Chemistry of Guizhou Province, School of Chemistry and Materials Science, Guizhou Normal University, Guiyang 550001, People’s Republic of China 5 College of Mechanical and Electrical Engineering, Wenzhou University, Wenzhou 325035, People’s Republic of China E-mail: [email protected] and [email protected] Received 11 March 2014, revised 30 July 2014 Accepted for publication 27 August 2014 Published 7 October 2014 Abstract
We report the fabrication of efficient white light-emitting diodes (WLEDs) based on Cu : ZnInS/ ZnS core/shell quantum dots (QDs) with super large Stokes shifts. The composition-controllable Cu : ZnInS/ZnS QDs with a tunable emission from deep red to green were prepared by a one-pot noninjection synthetic approach. The high performance Cu : ZnInS QD-WLEDs with the colour rendering index up to 96, luminous efficacy of 70–78 lm W−1, and colour temperature of 3800–5760 K were successfully fabricated by integration of red and green Cu-doped QDs. Negligible energy transfer between Cu-doped QDs was clearly found by measuring the photoluminescence lifetimes of the QDs, consistent with the small spectral overlap between QD emission and absorption. The experimental results indicated low toxic Cu : ZnInS/ZnS QDs could be suitable for solid state lighting. S Online supplementary data available from stacks.iop.org/NANO/25/435202/mmedia Keywords: quantum dots, nanocrystals, Cu-doped quantum dots, white light-emitting diodes, energy transfer commercial phosphors (Y3Al5O12 : Ce3+) show a deficient emission in the red spectral region and a serious scattering effect, which limit the colour rendering index (CRI) and luminescent efficiency of the WLEDs [7–8]. Therefore, it is necessary to find novel materials for fabricating high colour rendering WLEDs with high efficiency. Semiconductor quantum dots (QDs) as potential converters for WLEDs exhibit size-tunable emissions, high photoluminescence quantum yields (PL QYs), low scattering and good colour saturation, compared to traditional phosphors [9–14]. The QD-WLEDs have been successfully demonstrated by employing the combination of red and green light-
1. Introduction Energy-efficient lighting has offered a promising option for energy saving because artificial lighting globally consumes about 20% of the total electrical energy in the world [1]. Environmentally friendly white light-emitting diodes (WLEDs) with higher efficiencies, longer lifetimes, and fast response times are considered as promising light sources to replace the traditional ones such as incandescent or fluorescent lamps [2–6]. Currently conventional WLEDs have been fabricated by integrating down-converting phosphors with blue InGaN/GaN LEDs. However, the most popular 0957-4484/14/435202+08$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
Nanotechnology 25 (2014) 435202
Xi Yuan et al
emitting CdSe QDs on blue InGaN/GaN LEDs [15–19]. The CdSe QD-WLEDs were produced for display backlighting and they showed luminous efficacy of 41 lm W−1 and colour reproducibility of 100% compared to the National Television Systems Committee colour space [15]. However, the toxicity of the cadmium element and the significant self-absorption and energy transfer in closely packed QDs with small Stokes shifts potentially hinders their ultimate research transformation and commercialization [14, 15, 20]. Recently, low toxic CuInS2-based QDs with large Stokes shifts and high PL QYs (60%–70%) were used for fabricating high colour rendering WLEDs [21]. The devices integrating green and red emitting CuInS2-based QDs with blue chips showed a CRI of 95 and luminous efficacy of ∼69 lm W−1 [21]. Despite all, the spectral overlap between the green CuInS2-based QD emission and red QD absorption in the WLEDs might result in significant energy transfer and re-absorption processes [20, 22], influencing the optoelectronic performances of the devices [23, 24]. Thus, it is desirable to develop the ideal QDs with low toxicity, high PL QY, and enough large Stokes shift for WLEDs. Transition metal ions (such as Mn and Cu) doped QDs exhibited good thermal and chemical stability [25], and super large Stokes shift [14, 26, 27]. Mn-doped QDs with a yelloworange emission and a PL QY of ∼68% were reported in our previous works [25, 28]. However, the Mn-doped QDs with the emission wavelength of (580–600 nm) are not suitable for fabrication of high colour rendering WLED. It is well known that the emission wavelength of Cu-doped QDs can be varied from blue to near-infrared by changing the size of QDs and composition of hosts [14, 26, 27, 29, 30]. The emission in Cudoped QDs is considered to come from the radiative recombination of the electron in the conduction band of host material and the hole in Cu T2 state [31, 32]. For example, Cu : ZnSe/S QDs show a tunable emission from 485 to 540 nm [33], Cu : ZnInSe from 540 to 660 nm [26], and Cu : InP from 630 to 1100 nm [34]. On the other hand, Wang et al fabricated high colour rendering WLED based on a blue LED combined with highly luminescent CdS:Cu/ZnS QDs and YAG : Ce phosphors. The as-fabricated WLEDs showed a high luminous efficacy of 37.4 lm W−1, good colour chromatics stability, and CRI value of 86 [14]. No re-absorption between the QDs and phosphors could occur, suggesting the advantage of doped QDs with super large Stokes shifts. More recently, the composition-tunable emission from 450 to 810 nm was reported from Cu doped ZnInS QDs with an absorption band edge of up to 540 nm [30], which is promising for high colour rendering WLED based on blue chips (450 nm). In this work, we report the integration of Cu : ZnInS/ZnS core/shell QDs for warm WLEDs with high colour rendering and high efficiency. The Cu : ZnInS/ZnS QDs with composition-tunable emission were synthesized by a one-pot noninjection approach. The performance parameters of the Cu : ZnInS QD-WLEDs with blue InGaN chips (450 nm) to excite QD resin films with various QD concentrations were measured. Further the PL lifetimes of Cu : ZnInS QDs was studied by time-resolved PL spectroscopy to understand the
PL quenching of the doped QDs and energy transfer processes between the QDs in WLEDs.
2. Experimental section 2.1. Materials
Zinc acetate (Zn(OAc)2, 99.99%), copper acetate (Cu(OAc)2, 99.99%), indium acetate (In(OAc)3, 99.99%), sulfur powder (S, 99.99%), dodecanethiol (DDT, 99.9%), oleylamine (OAm, 97%), 1-octadecene (ODE, 90%), were purchased from Aldrich. All chemicals were used without further purification. 2.2. Synthesis of Cu : ZnInS and Cu : ZnInS/ZnS QDs
The Cu : ZnInS and Cu : ZnInS/ZnS QDs used in this experiment were synthesized via a one-pot noninjection synthetic approach similar to our previous works [30]. In a typical procedure, Cu(OAc)2 (0.02 mmol), Zn(OAc)2 (0.2 mmol), In(OAc)3 (0.2 mmol), sulfur powder (1.6 mmol) together with 4.0 mL of DDT, and 6.0 mL of OAm were loaded in a 50 mL three-neck flask clamped in a heating mantle. The mixture was heated to 220 °C with a heating rate of 15 °C min−1 under argon flow and kept at this temperature for 10 min to allow the growth of Cu : ZnInS cores. Then the reaction mixture was allowed to cool to 100 °C, the Zn precursor (obtained by dissolving 0.4 mmol of Zn(OAc)2 in 0.2 mL of OAm and 0.8 mL of ODE) was then introduced into the reaction mixture. Following this, the system temperature was raised to 240 °C with a heating rate of 15 °C min−1 and kept at this temperature for 20 min to allow the overgrowth of ZnS shell around the preformed Cu : ZnInS cores. To obtain Cu : ZnInS cores with different Zn/In ratios, the Zn/In precursor ratio was varied while the gross amount of Zn and In precursors and all other variables were remained constant. The Zn/In precursor ratio was varied with the values of 1/1, 1/2, 1/3, 1/5 and 1/8 for green, yellow, orange, red, and deep red QDs, respectively. The obtained QDs were precipitated by adding methanol into the toluene solution and purified by repeated centrifugation and decantation. 2.3. Device fabrication
Cu : ZnInS/ZnS QDs (0.1–0.2 g) in chloroform (3 mL) was blended with 0.3 g of thermally curable silicone resin under vigorous stirring. Then this QD-resin mixture was placed on hot plate at 50 °C for 1 h to evaporate the chloroform. Subsequently, 0.15 g of a hardener was added to the above QDresin re-mixture. The composite was put into a vacuum oven for 30 min at 50 °C to eliminate any air bubbles. The 450 nm emitting InGaN-based blue LED (device efficacy of about 30 lm W−1 at 20 mA, Sanan Optoelectronics, China) without an silicone package was used to fabricate the QD-based LEDs. The QD paste was dispensed on a blue LED mold and then thermally cured by a two-step process of 100 °C for 30 min, followed by 120 °C for 1 h. For comparison, the fabrication of the WLEDs based on CdSe/ZnS QDs (Najing 2
Nanotechnology 25 (2014) 435202
Xi Yuan et al
spectrometer with an integrating sphere. All measurements were carried out at room temperature.
3. Results and discussion Figure 1 shows UV-visible absorption of Cu : ZnInS/ZnS core/shell QDs and the corresponding normalized PL spectra under excitation at 450 nm. The absorption spectra become steep around 350 nm due to the ZnS shell coating [21]. No distinct excitonic absorption peak is observed, which could be attributed to their irregular elemental distribution and/or intra band gap states [35]. Both the absorption onset and peak wavelength of the emission are red-shifted with increasing In/ Zn precursor ratio in the cores from 1/1 to 8/1. The red shift could be explained by the change in the band gap of the alloy structures with the different ratio of ZnS (bandgap of 3.7 eV) and In2S3 (bandgap of 2.1 eV), since the emission in Cudoped QDs derives from the radiative recombination of the electron in the conduction band of host material and the hole in Cu T2 state [29–32]. Therefore, the Cu : ZnInS QDs exhibit a super large Stokes-shift of about 140–180 nm, compared with that of undoped CdSe QDs about 20 nm [15]. The Cu : ZnInS/ZnS QDs exhibit an emission peak at 525, 560, 590, 630, and 670 nm, called as G-, Y-, O-, R-, and Dr-QDs, respectively, covering the green to red region of the visible spectrum, and they possess the PL QYs of 40–60%. The full width at half maximum of the Cu : ZnInS QD emissions are approximately 90–130 nm, which might be beneficial in solid state lighting due to the large spectral coverage [24]. Structural characterization for the typical green and red emissive Cu-doped QDs was carried out by means of TEM and XRD, which were shown in figure 2. The TEM images demonstrate that the G- and R-QDs are near-spherical in shape and monodispersed. The average diameters of the Gand R-QDs were estimated to be 3.3 and 3.9 nm, respectively. The QDs exhibit clear lattice fringes in the high-resolution TEM images as shown in the inset of figures 2(a), (b), which suggests the highly crystalline nature. The wide-angle XRD patterns for the Cu-doped QDs show the characteristic peaks of the zinc blende (cubic) structure, similar to those reported before [30]. The Cu : ZnInS QDs with small sizes, tunable broad emissions, high PL QYs, and super large Stokes-shifts are promising for fabrication of WLEDs [2, 36]. For the fabrication of QD-based LEDs, Cu : ZnInS/ZnS QDs with green, yellow, orange, red, and deep red emissions were utilized as colour converters and combined with a blue (450 nm) emitting LED chip. The normalized electroluminescence (EL) spectra of Cu : ZnInS/ZnS QD-based LED devices are shown in figure 3(a). To evaluate the quality of the light emitted from Cu : ZnInS QD-based LEDs, several parameters, such as Commission Internationale de l’Eclairage (CIE) chromaticity coordinates, CRI, correlated colour temperature (CCT), and luminous efficacy need to be determined [2, 37]. As seen in figure 3(b), the CIE colour coordinates of Cu : ZnInS QD-based LEDs with various concentrations of QDs in the silicone resin can cover the white light region. With increasing the concentration of QDs, the CIE
Figure 1. UV-visible absorption (a) and PL spectra (b) of
Cu : ZnInS/ZnS core/shell QDs in chloroform with emission peaks at 525 nm (black lines), 560 nm (red lines), 590 nm (green lines), 630 nm (blue lines), and 670 nm (cyan lines) under excitation at 450 nm.
Technology Corporation Limited, L51 and L55, PL QYs over 40%) was performed, which was similar to the method for Cu : ZnInS/ZnS QD-WLEDs. The weight ratio of green CdSe/ZnS QDs to the red ones was 7/1.
2.4. Characterization
The absorption spectra were recorded by a UV-3101PC UVvis-NIR scanning spectrophotometer (Shimadzu). The PL spectra were recorded by a Hitachi F-7000 spectrophotometer. The time-resolved PL spectra were measured by an FL920 fluorescence lifetime spectrometer (Edinburgh Instruments). The excitation source was a hydrogen flash lamp (nF900) with a pulse width of 1.5 ns. The size and shape of the QDs were measured by a Tecnai G2 transmission electron microscope (TEM) operated at 200 kV. X-ray diffraction (XRD) patterns of the QDs were collected by a Bruker XRD spectrometer with a Cu Kα line of 0.15418 nm at a scanning step of 0.02°. The PL QYs of QDs were estimated by comparing the integrated emission of the QD samples in solution with that of standard dye solutions with identical optical density. The optical properties of QDWLEDs were measured by an USB4000 Ocean Optics 3
Nanotechnology 25 (2014) 435202
Xi Yuan et al
Figure 2. TEM images of G-Cu : ZnInS/ZnS (3.3 nm) (a) and R-Cu : ZnInS/ZnS (3.9 nm) (b) core/shell QDs. The corresponding high
resolution TEM images are shown in the inset. (c) XRD patterns of G-Cu : ZnInS/ZnS and R-Cu : ZnInS/ZnS core/shell QDs.
remarkably enlarged Stokes shift for the doped QDs (140–180 nm) was observed, compared to the undoped ones (12 nm). To quantify the spectral overlap, we calculated the overlap integral J with the equation [40]:
coordinates of the LEDs using the same QDs show a nearly linear increase trend. The slope of these lines monotonously decreases from 2.23 to 0.61 as the QD emission wavelength changes from green to deep red. These lines could guide us for the WLED design and provide the advantages for colour displays. The device parameters for typical LEDs based on G-, Y-, O-, R-, and Dr-Cu : ZnInS/ZnS QDs operated at 20 mA are listed in table 1. The luminous efficiencies of the LEDs with green-, yellow-, orange-, red-, and deep red-emitting QDs were 87.2, 76.9, 53.5, 34.6, and 12.1 lm W−1, respectively. The superiority in luminous efficacy of a green doped QDbased LED relative to the red doped QD-based one mainly originates from the variety of colour sensitivity [37]. The CCT values were determined in a relatively wide range of 8110–1809 K, which shows a decrease with changing the QD emission wavelength from green to red. The CRI values for the WLEDs based on G- or Y- Cu : ZnInS QDs were in the range of 71–74. This indicates that the Cu : ZnInS QDs are good colour converters for the next generation of white solid state lighting. The blue chip-to-QD light conversion efficiency was calculated from the integrated spectral ratio of the QD emission intensity to the absorbed source light intensity, based on the reduction of the original blue LED spectrum [38]. The light conversion efficiency of Cu : ZnInS QD-based LEDs and PL QYs of the Cu-doped QDs in solution are shown in figure 4(a). The light conversion efficiency deceased from 57% to 39% for different colour Cu : ZnInS QD-based LEDs, which showed a similar trend as the light-emitting efficiencies of the QDs in solutions. As we know that both the strong energy transfer/self-absorption and the surface defects of QDs generated through the LED fabrication (such as high temperatures for silicone encapsulation) have a detrimental impact on the efficiencies of the LEDs [15, 20, 39]. The degree of energy transfer and self-absorption in the QD ensembles significantly depends on the spectral overlap between absorption and emission spectra [40]. The absorption and emission spectra for Cu-doped QDs and undoped QDs, such as CdSe, were shown in supplementary figure S1, a
J=
∫0
∞
FD (λ) εA (λ) λ4dλ ,
where FD(λ) is normalized donor emission spectrum, and εA(λ) is the acceptor molar extinction coefficient. The overlap integral J value for CdSe QDs is 3.434 × 1016 M−1 cm−1 nm4, while that for G-, Y-, O-, R-, and Dr-Cu : ZnInS QDs is 2.545 × 1014, 1.648 × 1014, 4.463 × 1014, 2.808 × 1014, and 2.362 × 1014 M−1 cm−1 nm4, respectively. The overlap integrals for the Cu-doped QDs are two orders of magnitude lower than that of the undoped QDs, owing to the super large Stokes shift in the doped QDs [41]. This significant difference suggests that energy transfer and self-absorption between the Cu : ZnInS QDs could be negligible. In addition, the luminescence decay curves of the G-Cu : ZnInS QDs in chloroform solution and in silicone resin (thermally cured as the LED fabrication) were measured by time-resolved PL spectroscopy, respectively, to investigate whether an additional nonradiative de-excitation channel of QDs occurred during the LED fabrication, as seen in figure 4(b). We analyzed the PL decays by a bi-exponential function with two time components (τi) and weights (Ai). The amplitude-weighted lifetimes were obtained by a relation τav = (A1τ1 + A2τ2)/(A1 + A2). The lifetimes of the G-Cu : ZnInS QDs in solution and silicone resin were estimated to be 432 and 419 ns, respectively, basically the same. This result confirmed that almost no additional nonradiative channels (e.g. energy transfer or thermally induced surface traps) occurred in the LEDs. This demonstrates that the energy transfer and thermally curing induced luminous efficacy reduction is not significant in Cu : ZnInS QD-based LEDs, which makes the Cu-doped QDs ideal emitters for WLEDs. In addition, the emission in Cudoped QDs is considered to come from the recombination of the electron in the conduction band with the hole in Cu T2 state. Due to the incorporation of an additional energy state by Cu-doping, the excited-state lifetime of the dopant emission is 4
Nanotechnology 25 (2014) 435202
Xi Yuan et al
Table 1. The luminous efficacy (LE), CIE colour coordinates, CCT, and CRI of the typical LEDs based on G-, Y-, O-, R-, and DrCu : ZnInS/ZnS QDs operated at 20 mA, respectively.
LE Samples G-Cu : ZnInS Y-Cu : ZnInS O-Cu : ZnInS R-Cu : ZnInS Dr-Cu : ZnInS
CIE −1
(lm W ) 87.2 76.9 53.5 34.6 12.1
(x, y) (0.2821, (0.3481, (0.3713, (0.3432, (0.3380,
0.3392) 0.3138) 0.2486) 0.1667) 0.1499)
CCT (K)
CRI
8110 4682 2645 1962 1809
71 74 59 — —
Figure 3. (a) Normalized luminescence spectra of Cu : ZnInS QD-
based LED devices operated at a forward current of 20 mA. Photographs of the 20 mA-driven QD-based LEDs fabricated with G-Cu : ZnInS/ZnS and R-Cu : ZnInS/ZnS QDs are shown in the inset, respectively. (b) CIE colour coordinates of Cu : ZnInS QDbased LEDs with various concentrations of QDs in the silicone resin. The black dashed line represents the blackbody locus curve.
Figure 4. (a) The PL QYs (black squares) of the Cu-doped QDs with different emission wavelength in chloroform solution, and light conversion efficiencies (blue circles) of Cu-doped QD-based LEDs. (b) Temporal change of PL intensities for G-Cu : ZnInS/ZnS QDs in chloroform solution (black squares), in resin (green circles), and in G- and R-Cu : ZnInS QD blend resin film (blue triangles), respectively. The solid red lines represent fitting curves.
longer than the excitonic emission of QDs, reaching up to several microseconds or hundred-nanoseconds [42]. Further, the EL spectra of the Cu-doped QDs exhibit a small red-shifts (around 10 nm) in comparison to the PL spectra of the corresponding QDs in solution, which may be caused by the different dielectric constant of the QD-surrounding medium (chloroform versus silicone resin) [13, 43]. It is noted that the red-shifts for Cu : ZnInS QDs are extremely smaller than that for alloyed CuInS2-ZnS QDs (about 75 nm) [21], possibly because strong energy transfer between CuInS2-ZnS QDs might result in a large red-shift of the emission [21]. Therefore, the experimental results indicate that the Cu-doped ZnInS QDs are ideal candidates for fabricating high efficiency WLED [2].
Although QD-WLEDs can be fabricated by using a single type of Cu : ZnInS/ZnS QDs, the colour rendering property (CRI of 59–74) does not still satisfy the requirements for specific lighting applications (e.g. indoor illumination, CRI > 80) [2, 19]. The combination of green and red emitting QDs can provide an effective way to improve the lighting quality of WLEDs. In general, strong energy transfer and reabsorption between green and red undoped QDs such as CdSe nanocrystals have a detrimental impact on the luminous efficacy of WLEDs with increasing the QD packing density or 5
Nanotechnology 25 (2014) 435202
Xi Yuan et al
Table 2. The CRI, luminous efficacy (LE), CCT, and CIE colour
coordinates of the as-fabricated WLEDs based on G- and RCu : ZnInS/ZnS QD blends with different weight ratios (from 1.6/1 to 0.9/1) operated at 20 mA. LE
CCT −1
G/R
CRI
(lm W )
(K)
1.6/1 1.4/1 1.2/1 1/1 0.9/1
94 95 96 96 96
78.3 74.7 72.5 71.2 70.5
5760 5198 4655 4147 3804
CIE (x, y) (0.3271, (0.3388, (0.3528, (0.3710, (0.3863,
0.3232) 0.3337) 0.3416) 0.3596) 0.3721)
figure 4(b), similar to that in G-QD neat resin film. The PL lifetime of G-Cu : ZnInS QDs in green and red QD blend resin films (403 ns) is nearly the same as that in green QD neat resin film (419 ns). The corresponding energy transfer efficiency is only 4%, obtained by an expression 1 − τdonoracceptor/τdonor. This indicates that energy transfer between Gand R-Cu : ZnInS/ZnS QDs can be negligible, making us to use the green and red Cu : ZnInS QDs together for the WLEDs without a significant reduction in luminous efficacy. In multiphase emitters-based WLEDs, the intensity ratio of the emissions with different colours has a great impact on the quality of devices. The performance parameters of the QD-WLEDs fabricated with the QD blend (G- and RCu : ZnInS with different weight ratio) and blue LED chips and operated at 20 mA are summarized in table 2. The typical EL spectra of the WLED with G- and R-Cu : ZnInS weight ratio of 1.4/1 at different forward current from 5 to 100 mA are shown in figure 5. Three distinct peaks at 450, 540, and 620 nm, corresponding to blue chip, green and red Cu-doped QDs, respectively, are clearly observed in the broad EL spectrum from 430 to 800 nm, as seen in figure 5(a). Owing to such a spectral extension, the G- and R-Cu : ZnInS QD blendbased WLEDs exhibit an improved CRI value of 94–96 as listed in table 2, and the corresponding luminous efficiencies of above 70 lm W−1, which are better than those for R- and GCdSe QD-based WLEDs in the literature (luminous efficacy of 41 lm W−1) [15]. Besides, for fair comparison, the CdSe/ ZnS QD-WLEDs were fabricated with the same type of blue LED as used in the Cu : ZnInS QD-WLEDs. The corresponding EL spectrum is shown in supplementary figure S2, the luminous efficacy is 58 lm W−1, the CRI value is 84, the CCT is 5043 K, and the colour coordinate is (0.3421, 0.3298). It is found that, at the similar coordinates (0.34, 0.33) and CCT (about 5100 K), the luminous efficacy of Cu : ZnInS QD-WLED (1.4/1) can reach 74.7 lm W−1, while the CdSe/ ZnS system can only reach 58 lm W−1. Obviously, the improved efficacy could be accounted by high PL QYs of the Cu : ZnInS/ZnS QDs and negligible energy transfer between the doped QDs. Furthermore, the optimized WLEDs exhibit a tunable CCT between 3800 and 5760 K by finely tuning the ratio and amount of the G-Cu : ZnInS QDs to the red ones, as seen in table 2, which can provide warm enough CCT (80) white light for indoor illumination applications [2, 19]. The CIE colour coordinates
Figure 5. (a) EL spectra of a QD-WLED fabricated with G- and R-
Cu : ZnInS QDs (weight ratio of 1.4/1) as a function of forward current from 5 to 100 mA. (b) CIE colour coordinates of the QDWLED under different forward bias currents. The black dashed line represents the blackbody locus curve.
loading fraction [15, 35]. Many undoped QD-based WLEDs suffered from the embarrassing situation [13, 44, 45]. Moreover, QDs with shorter wavelength need a high content ratio to the longer wavelength ones to obtain high CRI values due to the energy transfer. Therefore, the energy transfer between G- and R-Cu : ZnInS/ZnS QDs was specially investigated by time-resolved PL spectroscopy. The energy transfer process acting as an additional nonradiative de-excitation channel for the donor can be described by the following expression 1/ τdonor-acceptor = 1/τdonor + kET, where τdonor-acceptor and τdonor are the PL lifetimes of G-Cu : ZnInS QDs donor in green and red QD blend and green QD neat resin films, respectively, and kET is the energy transfer rate [40]. The luminescence decay curves of the G-Cu : ZnInS QDs in G- and R-Cu : ZnInS QD blend films (weight ratio is 1/1) were measured, as seen in 6
Nanotechnology 25 (2014) 435202
Xi Yuan et al
4. Conclusions In summary, we have fabricated high performance Cu : ZnInS/ZnS QD-WLEDs for the next generation solid state lighting. Highly luminescent Cu : ZnInS QDs with PL QYs of 40–60% and tunable emissions from green to deep red were obtained by effectively controlling the component of the QDs. The optimized warm WLEDs based on G- and RCu : ZnInS/ZnS core/shell QDs and blue LED chips exhibited high CRI up to 96, luminous efficacy of 70–78 lm W−1, CCT of 3800–5760 K, and CIE colour coordinates of (0.3388, 0.3337), which are the best values currently reported for QDWLEDs. No significant change in PL lifetimes of the GCu : ZnInS/ZnS QDs in G- and R-QD blend films was observed, compared with those in solution. This indicated that the energy transfer between G- and R-Cu : ZnInS QDs in WLEDs could be negligible. This resulted in a higher luminous efficacy in doped QD-WLEDs than the undoped QD devices. These experimental results suggest that low toxic Cu : ZnInS QDs are promising colour converting materials for fabricating high colour rendering WLEDs with high efficiency.
Figure 6. (a) CRI of the fluorescent lamp (Panasonic
YZ18RR6500K) and G- and R-Cu : ZnInS QD-WLED operated at 100 mA. Photographs of colourful candy are compared under the illumination with sun light (b), a fluorescent lamp (c), and a Cu : ZnInS QD-based WLED operated at 100 mA.
Acknowledgments
of the WLED with G/R weight ratio of 1.4/1 are obtained to be (0.3388, 0.3337) under 20 mA forward bias current, which are close to the equi-energy white point (0.3333, 0.3333). The values exhibit little change under different forward bias currents, as seen in figure 5(b), indicating good colour stability of the output light of the WLED. The detailed CRI values of R1-R9 and the general CRI (designated by the symbol Ra, the average value of R1-R8) of both the fluorescent lamp and the as-fabricated Cu : ZnInS QD-WLEDs operated at 100 mA are shown in figure 6(a). In comparison with the fluorescent lamp (Panasonic YZ18RR6500K) with Ra of 71, the WLED based on G- and R-Cu : ZnInS QDs exhibits a higher Ra of 95 and the improvement in the colour reproduction for the colour from R1 to R8. It is worth noting that the special CRI R9, corresponding to the colour rendering of the deep-red region, which is important for biomedical and painting appreciation applications [46, 47], is greatly improved to 97. This value is also much higher than that of the commercial YAG : Ce-based WLED (R9–15), because the emission wavelength of commercial red-emitting phosphors is commonly below 630 nm [14]. The R-Cu : ZnInS QDs with the broad emission band result in the enhancement of the colour rendering in both the red and deep-red spectral regions and show their potential applications in biomedical and painting appreciation. Further, the photographs of colourful candies are compared under the illumination with sun light, a fluorescent lamp, and the Cu : ZnInS QD-based WLED are shown in figures 6(b)–(d). This Cu : ZnInS QD-based WLED demonstrates chromatic light quality with a CRI of 95 and great colour saturation, which is similar to that of the sun light.
This work was supported by the National Natural Science Foundation of China (Nos. 11274304, 11204298, 61205025, 61275047, 21371071, 21461006 and 51472053), Program for the Development of Science and Technology of Jilin Province (No. 20120344), and the Research Project of Chinese Ministry of Education (No. 213009A).
References [1] Tsao J Y 2002 Light Emitting Diodes (LEDs) for General Illumination: An OIDA Technology Roadmap Update (Washington, DC, USA: Optoelectronics Industry Development Association (OIDA)) [2] Dai Q Q, Duty C E and Hu M Z 2010 Small 6 1577–88 [3] Ziegler J, Xu S, Kucur E, Meister F, Batentschuk M, Gindele F and Nann T 2008 Adv. Mater. 20 4068–73 [4] Schubert E F and Kim J K 2005 Science 308 1274–8 [5] Nakamura S, Mukai T and Senoh M 1994 Appl. Phys. Lett. 64 1687–9 [6] Li X F, Budai J D, Liu F, Howe J Y, Zhang J H, Wang X J, Gu Z J, Sun C J, Meltzer R S and Pan Z W 2013 Light: Sci. Appl. 2 e50 [7] Lin C C and Liu R S 2011 J. Phys. Chem. Lett. 2 1268–77 [8] Jang H S, Yang H, Kim S W, Han J Y, Lee S G and Jeon D Y 2008 Adv. Mater. 20 2696–702 [9] Coe S, Woo W K, Bawendi M G and Bulović V 2002 Nature 420 800–3 [10] Aboulaich A, Michalska M, Schneider R, Potdevin A, Deschamps J, Deloncle R, Chadeyron G and Mahiou R 2014 ACS Appl. Mater. Interfaces 6 252–8 [11] Dohnalová K, Poddubny A N, Prokofiev A A, Boer W D, Umesh C P, Paulusse J M, Zuilhof H and Gregorkiewicz T 2013 Light: Sci. Appl. 2 e47 7
Nanotechnology 25 (2014) 435202
Xi Yuan et al
[29] Zhang W J, Zhou X G and Zhong X H 2012 Inorg. Chem. 51 3579–87 [30] Zhang W J, Lou Q, Ji W Y, Zhao J L and Zhong X H 2014 Chem. Mater. 26 1204–12 [31] Mandal P, Talwar S S, Major S S and Srinivasa R S 2008 J. Chem. Phys. 128 114703 [32] Corrado C, Jiang Y, Oba F, Kozina M, Bridges F and Zhang J Z 2009 J. Phys. Chem. A 113 3830–9 [33] Jana S, Srivastava B B, Acharya S, Santra P K, Jana N R, Sarma D D and Pradhan N 2010 Chem. Commun. 46 2853–5 [34] Xie R G and Peng X G 2009 J. Am. Chem. Soc. 131 10645–51 [35] Chung W, Jung H, Lee C H and Kim S H 2012 Opt. Express 20 25071–6 [36] Bowers M J, McBride J R and Rosenthal S J 2005 J. Am. Chem. Soc. 127 15378–9 [37] Demir H V, Nizamoglu S, Erdem T, Mutlugun E, Gaponik N and Eychmüller A 2011 Nano Today 6 632–47 [38] Kim J K, Luo H, Schubert E F, Cho J, Sone C and Park Y 2005 Japan. J. Appl. Phys. 44 L649–51 [39] Phillips J M et al 2007 Laser Photonics Rev. 1 307–33 [40] Lakowicz J R 2006 Principles of Fluorescence Spectroscopy 3rd edn (Berlin: Springer) pp 443–527 [41] Norris D J, Efros A L and Erwin S C 2008 Science 319 1776–9 [42] Wu P and Yan X P 2013 Chem. Soc. Rev. 42 5489–521 [43] Song H and Lee S 2007 Nanotechnology 18 255202 [44] Li Y Q, Rizzo A, Cingolani R and Gigli G 2006 Adv. Mater. 18 2545–8 [45] Cheng G, Mazzeo M, Agostino S, Sala F, Carallo S and Gigli G 2010 Opt. Lett. 35 616–8 [46] Krames M R, Shchekin O B, Mueller-Mach R, Mueller G O, Zhou L, Harbers G and Craford M G 2007 J. Disp. Technol. 3 160–75 [47] Taguchi T and Kono M 2007 J. Light Vis. Environ. 31 149–51
[12] Sun M Y, Zhu D H, Ji W Y, Jing P T, Wang X Y, Xiang W D and Zhao J L 2013 ACS Appl. Mater. Interfaces 5 12681–8 [13] Wang X B, Li W W and Sun K 2011 J. Mater. Chem. 21 8558–65 [14] Wang X B, Yan X S, Li W W and Sun K 2012 Adv. Mater. 24 2742–7 [15] Jang E, Jun S, Jang H, Lim J, Kim B and Kim Y 2010 Adv. Mater. 22 3076–80 [16] Lita A, Washington A L II, Burgt L V D, Strouse G F and Stiegman A E 2010 Adv. Mater. 22 3987–91 [17] Chen H S, Hsu C K and Hong H Y 2006 IEEE Photonics Technol. Lett. 18 193–5 [18] Jun S, Lee J and Jang E 2013 ACS Nano 7 1472–7 [19] Nizamoglu S, Zengin G and Demira H V 2008 Appl. Phys. Lett. 92 031102 [20] Kundu J, Ghosh Y, Dennis A M, Htoon H and Hollingsworth J A 2012 Nano Lett. 12 3031–7 [21] Chen B K, Zhong H Z, Wang M X, Liu R B and Zou B S 2013 Nanoscale 5 3514–9 [22] Yuan X, Zhao J L, Jing P T, Zhang W J, Li H B, Zhang L G, Zhong X H and Masumoto Y 2012 J. Phys. Chem. C 116 11973–9 [23] Ku K H, Kim M P, Paek K, Shin J M, Chung S, Jang S G, Chae W-S, Yi G-R and Kim B J 2013 Small 9 2667–72 [24] Song W-S and Yang H 2012 Appl. Phys. Lett. 100 183104 [25] Yuan X, Zheng J J, Zeng R S, Jing P T, Ji W Y, Zhao J L, Yang W Y and Li H B 2014 Nanoscale 6 300–7 [26] Sarkar S, Karan N S and Pradhan N 2011 Angew. Chem. Int. Ed. 50 6065–9 [27] Maity A R, Palmal S, Basiruddin S K, Karan N S, Sarkar S, Pradhan N and Jana N R 2013 Nanoscale 5 5506–13 [28] Cao S, Zheng J J, Zhao J L, Wang L, Gao F M, Wei G D, Zeng R S, Tian L H and Yang W Y 2013 J. Mater. Chem. C 1 2540–7
8
