Carrier recombination spatial transfer by reduced potential barrier causes blue/red switchable luminescence in C8 carbon quantum dots/organic hybrid light emitting devices Carrier recombination spatia[.]
Trang 1Carrier recombination spatial transfer by reduced potential barrier causes blue/red switchable luminescence in C8 carbon quantum dots/organic hybrid light-emitting devices
Xifang Chen, Ruolin Yan, Wenxia Zhang, and Jiyang Fan
Citation: APL Mater 4, 046102 (2016); doi: 10.1063/1.4945722
View online: http://dx.doi.org/10.1063/1.4945722
View Table of Contents: http://aip.scitation.org/toc/apm/4/4
Published by the American Institute of Physics
Articles you may be interested in
Optimal nitrogen and phosphorus codoping carbon dots towards white light-emitting device
APL Mater 109, 083103083103 (2016); 10.1063/1.4961631
Optical spectroscopy reveals transition of CuInS2/ZnS to CuxZn1-xInS2/ZnS:Cu alloyed quantum dots with resultant double-defect luminescence
APL Mater 4, 126101126101 (2016); 10.1063/1.4971353
Trang 2Carrier recombination spatial transfer by reduced
potential barrier causes blue/red switchable
luminescence in C8 carbon quantum dots/organic
hybrid light-emitting devices
Xifang Chen, Ruolin Yan, Wenxia Zhang, and Jiyang Fana
Department of Physics and Jiangsu Key Laboratory for Advanced Metallic Materials,
Southeast University, Nanjing 211189, People’s Republic of China
(Received 11 March 2016; accepted 29 March 2016; published online 7 April 2016)
The underlying mechanism behind the blue/red color-switchable luminescence in the C8carbon quantum dots (CQDs)/organic hybrid light-emitting devices (LEDs)
is investigated The study shows that the increasing bias alters the energy-level spatial distribution and reduces the carrier potential barrier at the CQDs/organic layer interface, resulting in transition of the carrier transport mechanism from quantum tunneling to direct injection This causes spatial shift of carrier recom-bination from the organic layer to the CQDs layer with resultant transition of electroluminescence from blue to red By contrast, the pure CQDs-based LED exhibits green–red electroluminescence stemming from recombination of injected carriers in the CQDs C 2016 Author(s) All article content, except where other-wise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).[http://dx.doi.org/10.1063/1.4945722]
Colloidal quantum dots (QDs)-based light-emitting devices (LEDs) have attracted great interest
in recent years due to their wide applications in flat-panel display, solid-state lighting, and optical communication, etc.1Moreover, the continuous and smooth semiconductor QDs films used in the LEDs are readily to fabricate by using simple solution-processed deposition technique.2Since the first report on CdSe QD-LED in 1994,3various types of QDs such as CdSe/CdS, Cd1−xZnxSe1−ySy, ZnCdS/ZnS, PbS/CdS, ZnO, and Si QDs have been employed to fabricate QD-LEDs.4 10QDs have also been used as color converters in white LEDs.11,12However, there have been very few studies focusing on carrier transport and recombination characteristics of QD-LEDs On the other hand, compared with the wurtzite or zinc blende structured Cd-containing QDs, heavy-metal-free carbon QDs (CQDs) are more benign to human beings and the environment.13 , 14They have become a research focus owing to low cytotoxicity, exploited various synthesis methods, unexampled abundance of source materials on Earth, and robust near-infrared to near-UV luminescence.15 , 16CQDs usually exist
as nanodiamonds or graphite/graphene quantum dots.17We have recently synthesized luminescent
C8 (third carbon allotrope) CQDs with sizes ranging from about 3.5 nm down to below 1 nm.18
Here, we design and fabricate both pure and inorganic/organic hybrid LEDs based on C8CQDs and investigate their unusual carrier transport/recombination and electroluminescence (EL) properties The first-type designed device (named device A) is a pure CQDs-based LED It has a simple structure without carrier transport layers such that it can accommodate pure luminescence from the CQDs By contrast, the second-type device (named device B) contains carrier transport layers
to ensure higher quantum efficiency and more device stability.19 , 20 Previous study has indicated that the excimers and electromers are prone to form in organic hole-transport materials and their light emissions enlarge the EL spectral region.21 , 22 Although their light emissions are weak in usual semiconductor QD-LEDs due to energy transfer,23 , 24 but their existence makes it hard to discriminate the contribution of the QDs to electroluminescence of the conventional-structured
a Author to whom correspondence should be addressed Electronic mail: jyfan@seu.edu.cn
2166-532X/2016/4(4)/046102/9 4, 046102-1 © Author(s) 2016.
Trang 3046102-2 Chen et al. APL Mater 4, 046102 (2016)
LEDs The current device A has a three-layer structure with the following order: indium tin oxide (ITO), 100-nm-thick C8CQDs, and Al film (150 nm) Device B has a complex structure with the following order: ITO, CQDs embedded in poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS:CQDs, 70 nm), 1,3,5-tris(1-phenyl-1H-benzimidazol-2-yl) benzene (TPBI, 12 nm), and LiF/Al (3/150 nm) The excimer and electromer emissions are effectively prohibited in both devices The PEDOT:PSS acts as the transparent hole-injection layer.25 , 26 The TPBI film acts as the electron-transport layer The ITO film acts as the anode and the Al and LiF/Al films act as the cathode The ITO was deposited on a glass substrate by using the magnetron sputtering method It has a thickness of 135 nm, a sheet resistance of 10 Ω/sq, and a light transmittance of over 80% The CQDS and PEDOT:PSS:CQDs films were prepared by sol-gel spin-cast All the spin-coated films were subsequently annealed in air at 110◦C for 30 min to improve uniformity and stability The heating temperature for PEDOT:PSS is usually above 110◦C We adopted the lower temperature
in our experiment considering thermal stability of the CQDs The thickness of the spin-coated film
is controlled by altering the CQDs concentration and the spinning speed For fabrication of device
A, the spinning speed was chosen as 2000 rpm It is found that device A fabricated with 30 kg/m3
CQDs solution (100-nm-thick CQDs layer) exhibits the best electroluminescence performance For fabrication of device B, the spinning speed is chosen as 2500 rpm, and the concentration of the (PEDOT:PSS + CQDs) solution was selected as 12 kg/m3 The Al cathode in device A as well
as the TPBI film and LiF/Al cathode in device B was separately deposited with help of a shadow mask in high vacuum of 1 × 10−6 Torr by using a thermal-evaporation system The evaporation rate was 0.1 Å/s for LiF and TPBI films and 4–5 Å/s for Al film The active area of the device was 2 × 3 mm2as defined by the overlapped area of the ITO and Al electrodes The devices were not encapsulated and all the measurements were performed in air The atomic force microscopy (AFM) images and the thickness of the films were acquired by using a Dimension Icon atomic force microscope (Bruker Corporation) with resolution of 1 nm along X and Y axes and resolution
of 0.05–0.1 nm along Z axis The photoluminescence (PL) spectra were measured by using a Fluorolog 3-TCSPC spectrofluorometer (HORIBA JOBIN YVON) with a xenon lamp as the light source The electroluminescence spectra were measured by using an Acton SP-2358 Spectrometer (Princeton Instruments) The electrical characterization of the devices was performed by using a Keithley 2400 source meter
The AFM images (Fig 1) reveal that the CQDs film spin-coated on the ITO substrate (de-vice A) is smooth with a root mean square (RMS) roughness of 0.30 nm This value is close to that of the monolayer of close packed ZnCdSe alloyed QDs (0.4 nm) used in the QD-LED.27The PEDOT:PSS:CQDs layer on ITO in device B has a larger RMS roughness of 5.08 nm The AFM im-ages suggest that there is no phase separation for this layer, revealing that the CQDs are embedded
in the PEDOT:PSS matrix Figure2(a)shows the PL spectra of the CQDs dispersed in water As can
be seen, the PL peak shifts from 520 to 620 nm as the excitation wavelength increases from 420 to
580 nm The excitation dependence of the PL peak is not related to the quantum confinement effect because bulk C8carbon is an insulator with an indirect bandgap of 5.5 eV and a direct bandgap ranging from 6.5 to 11 eV,28 the probable luminescence stemming from interband transition of carriers in C8carbon should lie in the ultraviolet region29rather than in the current visible region The origin of the photoluminescence of the CQDs dispersed in water had been investigated in detail
by using various characterizations (PL, UV-vis absorption, infrared absorption, X-ray photoelec-tron spectroscopy) in our recent report.18The combined characterizations demonstrate that electron transition between the surface states associated with the C(==O)O functional groups generates the photoluminescence The addition of other chemical groups in the adjacent of the C(==O)O group usually makes the PL peak shift to red In particular, additional conjugation of the C==C double bond to the C(==O)O group causes considerable red shift of over 30 nm Hence, there are several separated emission peaks in the visible region resulting from different C(==O)O-related surface states The total PL spectrum resulting from their superimposition shows red shift with increasing excitation wavelength
The C8CQD-LEDs exhibit interesting electroluminescence properties Figure 2(b)shows the
EL spectra of device A under different biases The EL spectral region ranges from 400 to 900 nm The maximum of the EL spectrum lies at around 650 nm, and the intensity of the maximum
Trang 4FIG 1 Two-dimensional and three-dimensional AFM images of ((a) and (b)) close-packed CQDs film and ((c) and (d)) PEDOT:PSS:CQDs composite film spin-coated on ITO substrates.
increases first and decreases then with increasing bias The EL spectrum has a full width at half maximum (FWHM) of about 250 nm, which is much larger than that of the PL spectrum (about
100 nm) [Fig.2(a)] This is because that the injected carriers can occupy much more energy levels
in the case of EL compared with the photon-excited electrons in the case of PL The energy-level distribution of the injected carriers determines the line shape and maximum position of the EL spectrum The EL characteristics of device B are quite different from that of device A The features
of the EL spectra of device B for bias 4–8 V [Fig 2(c)] resemble that of the PL spectra of the TPBI film [Fig 2(a)], and they have the same peak position at around 390 nm and nearly equal linewidths These features prove that the violet EL stems from recombination of electrons and holes in the TPBI layer It becomes most intense at 5 V and then diminishes gradually with further increasing bias The EL spectral features totally change for bias >9 V [Fig.2(d)] On the one hand, the violet emission from TPBI vanishes gradually with increasing bias On the other hand, a broad
EL band lying in the green–red region arises, and it becomes more intense with increasing bias The measurement indicates that the pure PEDOT:PSS layer has no electroluminescence, hence, the green–red region EL must originate from the CQDs In practice, as far as the spectral region and linewidth are concerned, the green–red EL spectrum roughly resembles the EL spectrum of device A that arises from the CQDs The conversion of the electroluminescence band from blue region to green–red region with increasing bias in device B suggests that the recombination space
of the injected carriers transfers from the TPBI layer to the PEDOT:PSS:CQDs layer As can
be seen from Fig 2(d), for bias >12 V, there is only green–red EL band, suggesting under this circumstance, the injected electrons and holes recombine completely in the PEDOT:PSS:CQDs layer The characteristics of the three EL emission bands (red–green band from CQDs in device
A, blue band from TPBI, and red–green band from CQDs embedded in PEDOT:PSS in device B) are more clearly embodied by the curves of variation of the peak wavelength and intensity of these
Trang 5046102-4 Chen et al. APL Mater 4, 046102 (2016)
FIG 2 (a) Photoluminescence spectra of spin-coated TPBI film and CQDs in water (b) Electroluminescence (EL) spectra
of device A under di fferent driving biases (c) EL spectra of device B under (c) low bias [inset: EL image at 5 V] and (d) high bias [inset: EL image at 10 V].
emission bands with bias, as shown in Fig.3 Both the peak wavelength and intensity show different variation trends with bias for three emission bands For device A, the EL peak wavelength is nearly independent of bias, and the EL peak intensity versus bias curve is very narrow, both features are characteristic of the luminescence of an organic film In contrast, the red–green EL bands from both devices vary dramatically with bias, suggesting that there are considerable lower and higher energy levels involved in electroluminescence of the CQDs This is because, on the one hand, the CQDs have a loose size distribution, the conduction band minimum (relative to vacuum energy level) increases and the valence band maximum decreases with decreasing particle size due to the quantum confinement effect; on the other hand, there are several types of C(==O)O-related surface
FIG 3 (a) EL spectrum peak wavelength versus bias (b) EL spectrum peak intensity versus bias.
Trang 6states in the CQDs and they have different energy levels Although the red–green region EL from both devices stems from the CQDs, however, they show different variation trends with bias, such a
difference suggests that there are distinct energy-level spatial distributions and carrier transport and recombination mechanisms in these two devices of different spatial structures
There are some measures to improve the electroluminescence intensity of the devices for prac-tical applications For device A, the weak EL may be ascribed to the high potential barriers for carrier injection, especially the barrier for electrons at the CQDs-Al interface.30 Another reason lies in the fact that direct contact between the light emission layer and the electrodes leads to high rate of nonradiative Auger recombination at the interface.31For device B, the energy transfer between the CQDs and the adjacent polymer layers may result in significant diminishing of EL The quantum efficiency of the CQDs is another limitation factor Hence, suppressing the energy transfer efficiency and improving the quantum yield of the CQDs will lead to enhanced EL The multilayer structure may also affect the emission efficiency of the CQDs and optimized device structure with more favorable film quality is helpful The spotty EL image of the devices (insets of Fig.2) suggests nonuniformity of EL owing to nonuniform films resulting from irregular interfaces between domains of the same layer and irregular interfaces between different layers The EL image
in inset of Fig.2(d)contains a lighter emission band on the left-hand side because this edge area is more uniform as a result of shadow-mask covering during deposition of TPBI, LiF, and Al films Therefore, improvement of film uniformity will result in more uniform EL of the device
In order to reveal the underlying physics for the respective electroluminescence spectral fea-tures in pure and hybrid CQDs-based LEDs, especially for the strange and rare phenomenon of electroluminescence-color switching in device B, it is necessary to analyze the energy-level spatial distribution and the carrier transport and recombination mechanisms in these devices The left (top and bottom) panel in Fig.4shows schematic energy-level diagrams of two CQDs-based LEDs The work functions of the ITO anode and the Al cathode are 4.7 and 4.3 eV (below vacuum energy level), respectively.32 The PEDOT:PSS film improves the work function from 4.3 eV (ITO) to 5.0 eV.33,34The values of the lowest unoccupied molecular orbital (LUMO) and the highest occu-pied molecular orbital (HOMO) of TPBI are separately 2.8 and 6.3 eV The CQDs layer possesses at least two types of surface-state energy levels generated by different ester-related surface functional groups.18
Figure 5 displays the measured current density versus bias (J–V ) curves for both devices, and the semi-logarithmic coordinates are employed (the insets show the J–V curves with linear coordinates) The circled dots and solid lines correspond to the experimental data and fitted curves, respectively The different regions of J–V data are well fitted by using different equations and there is no any single equation that can fit the whole region of J–V data This suggests that the carrier transport and recombination mechanisms change with increasing bias in the same device Various models have been proposed to explain carrier transport properties in (especially organic) LEDs, including Fowler-Nordheim tunneling,35Poole-Frenkel emission,36trap-limited transport,37 and space charge-limited current38 models However, these models cannot account for the J–V characteristics of the current CQDs-based LEDs For device A, the J–V data for bias <4 V [zones I and II in Fig.5(a)] are fitted well by using the following equation:
where a and b are two empirical constants depending on the device parameters including mate-rial type and thickness of the layers This equation describes the direct quantum tunneling model derived from quantum mechanics, hence, the carrier transport and recombination under this circum-stance is governed by quantum tunneling.39It is well known that the transmission coefficient for an electron or hole (mass m, energy E) to penetrate the potential barrier (height Vb, width l) is
where ~ is the reduced Planck constant Since at zero bias, the potential barrier for the holes at the ITO/CQDs interface is much lower than the potential barrier for the electrons at the Al/CQDs interface, as shown in Fig.4 (top left), therefore, at low bias [zone I in Fig.5(a)], the holes have higher probability to penetrate the interface potential barrier compared with the electrons, thus, the
Trang 7046102-6 Chen et al. APL Mater 4, 046102 (2016)
FIG 4 Energy level diagrams for devices A and B showing process of electrons (filled circles) and holes (empty circles) transport as well as process of carrier recombination with resultant photon emission The blue shadow areas denote potential barriers for carriers at the interface of di fferent layers Each row of three diagrams corresponds to cases of increasing bias from left to right.
holes contribute a larger proportion to the quantum tunneling current As the bias is elevated to
be high enough [zone II in Fig 5(a)], the electron potential barrier at the Al/CQDs interface is reduced to be low enough such that the electron penetration probability is significantly improved; since the electron-transport ability of the metal layer is higher than the hole-transport ability of the ITO layer, hence, under this circumstance, the quantum tunneling of electrons overwhelms that of
FIG 5 Current density versus bias with logarithmic y-coordinates for (a) device A and (b) device B The insets show current density versus bias curves with linear y-coordinates In the fitting equations, a i and b j are constants depending on device parameters.
Trang 8holes, hence, the penetrating electrons begin to dominate the tunneling current (Fig.4: top middle) Such a transition of carrier type of tunneling current explains why the fit constants for zones I and II in Fig 5(a) are different Note that no remarkable electroluminescence was detected for bias <4 V because the current density under this circumstance does not exceed the threshold value
of the device, thus, the turn-on voltage of this device is 4 V, which is a little bigger than that of organic LEDs (about 3 V)40 , 41but is close to that of usual QD-LEDs.32 , 42The turn-on voltage can be further reduced by improving the quality of the CQDs film, optimizing the contact electrodes, and exploring better carrier-transport materials For bias >4 V [zone III in Fig.5(a)], the J–V data can
be well fitted by using the following equation:
where a an b are two empirical constants depending on the device parameters including material type and layer thickness This equation describes the direct injection model that was primarily proposed to explain the J–V characteristics of conventional PN junction diodes Therefore, the carrier transport for bias >4 V is governed by direct electron injection from the cathode and direct hole injection from the anode (Fig.4: top right) These injected electrons and holes encounter in the CQDs and recombine therein to emit photons of green–red spectral region
We now explain the more complex electroluminescence phenomenon in device B by analyzing the energy-level spatial distribution and associated carrier transport and recombination picture In device B, the potential barrier for holes to go into the CQDs layer from ITO anode is reduced
by 0.3 eV owing to introduction of a intermediate potential-barrier step ascribed to the added PEDOT:PSS layer (Fig.4: bottom left) This change is favorable for direct injection of holes into the CQDs layer and the following TPBI layer under driving voltage, and these advancing holes will recombine with the electrons at the TPBI/LiF-Al interface (near the Al layer) injected from the cathode This explains why the J–V data under bias <3 V [zone I in Fig.5(b)] are well fitted
by Eq (3) But the current density is small because of the limited hole-transport ability of the PEDOT:PSS:CQDs layer On the other hand, the 3-nm-thick LiF (insulator) layer between the Al and TPBI films constructs an electron potential barrier; under high enough bias (Fig 4: bottom middle), the electrons from the Al cathode are capable of penetrating this potential barrier by quantum tunneling to reach the conduction-band energy levels of TPBI This corresponds to the case of zone II in Fig 5(b), where the J–V data are well fitted by Eq (1) Because the electron potential barrier between the TPBI and PEDOT:PSS:CQDs layers is too high for bias <8 V, the electrons reaching the TPBI layer after quantum tunneling of the LiF potential barrier are unable
to penetrate the second potential barrier at the interface of the TPBI and PEDOT:PSS:CQDs layers, hence, they are trapped in the TPBI layer These trapped electrons recombine with the injected holes
in the TPBI layer coming from the anode, leading to emission of blue photons This is consistent with the observed blue electroluminescence at such bias [Fig 2(c)] For bias >8 V, the electron potential barrier at the interface of the TPBI and PEDOT:PSS:CQDs layers is considerably reduced;
as result, the transmission coefficient [Eq (2)] is highly improved and the number of electrons penetrating into the PEDOT:PSS:CQDs layer increases These penetrating electrons recombine with holes in the CQDs, leading to gradually enhanced electroluminescence in the green–red region This corresponds to the transition region between zone II and zone III in Fig 5(b) For bias >11 V, both the electron potential barrier at the interface of the TPBI and PEDOT:PSS:CQDs layers and the electron potential barrier at the TPBI/Al interface cease gradually (Fig.4: bottom right), this
is because that these two potential barrier spatial regions have the highest electric resistance and thus the applied voltage mainly falls in these two thin regions, thereby significantly reducing the barrier height Hence, the electron transport is dominated by direct injection across the whole device and the current density increases rapidly with bias This scenario corresponds to the case of zone III in Fig.5(b) Indeed, in this region, the J–V data are well fitted by Eq (3) Under this circum-stance, the electrons and holes recombine only in the CQDs, and this explains why there is only green–red electroluminescence in zone III Because the current density is very large in the mode of direct injection of electrons, so the electroluminescence intensity is highly improved in this region [Fig.2(d)] It should be noted that the electroluminescence performance of device B is sensitive to the thickness of the TPBI layer The turn-on voltage increases from 3 to 5 V as the thickness of
Trang 9046102-8 Chen et al. APL Mater 4, 046102 (2016)
the TPBI layer increases from 12 to 16 nm, meanwhile, the electroluminescence intensity decreases remarkably Such turn-on voltage is comparable to usual QD-LEDs and can be further reduced by using thinner TPBI layer; however, if the TPBI layer is too thin, the EL intensity will decrease and the normal-operation voltage range will become narrower Hence, there is an optimal thickness of TPBI film
The above discussions indicate that the operation mechanisms of the QDs-based LEDs can be revealed based on combined study of the electroluminescence spectra and the J–V curves of the LEDs in conjunction with analysis of the bias dependence of the energy-level spatial distribution These results also show that electroluminescence with tunable colors can be realized by using a single QDs-based LED device with intentionally designed multilayered structure in which there are multiple carrier potential barriers and thus the recombination spatial region of electrons and holes is transferable and controlled by varying the driving voltage
In summary, we have observed intriguing and peculiar phenomenon of blue/red switchable electroluminescence in inorganic/organic hybrid LED based on C8CQDs A complete model con-cerning carrier recombination spatial transfer is proposed to explain the experimental phenomenon based on analysis of the carrier transport and recombination mechanisms It was found that quan-tum tunneling and direct injection alternately dominate the carrier transport mechanism as the bias increases and this leads to transferable carrier recombination spatial region The pure CQDs-based LED exhibits totally green–red electroluminescence originating from the CQDs These findings largely improve our understanding of the electroluminescence mechanisms of the QDs-based LEDs This work was supported by the National Natural Science Foundation of China Nos 11274063 and 11574047 It is also supported by Jiangsu Key Laboratory for Advanced Metallic Materials (BM2007204)
1 Y Shirasaki, G J Supran, M G Bawendi, and V Bulovi´c, Nat Photonics 7, 13 (2013).
2 H Dondapati, D Ha, and A K Pradhan, Appl Phys Lett 103, 121114 (2013).
3 V L Colvin, M C Schlamp, and A P Alivisatos, Nature 370, 354 (1994).
4 X Dai, Z Zhang, Y Jin, Y Niu, H Cao, X Liang, L Chen, J Wang, and X Peng, Nature 515, 96 (2014).
5 Y Yang, Y Zheng, W Cao, A Titov, J Hyvonen, J R Manders, J Xue, P H Holloway, and L Qian, Nat Photonics 9, 259 (2015).
6 H Huang, C Dun, W Huang, Y Cui, J Xu, Q Jiang, C Xu, H Zhang, S Wen, and D L Carroll, Appl Phys Lett 106,
263506 (2015).
7 Y Z Xin, K Nishio, and K Saitow, Appl Phys Lett 106, 201102 (2015).
8 B N Pal, Y Ghosh, S Brovelli, R Laocharoensuk, V I Klimov, J A Hollingsworth, and H Htoon, Nano Lett 12, 331 (2012).
9 H Shen, W Cao, N T Shewmon, C Yang, L S Li, and J Xue, Nano Lett 15, 1211 (2015).
10 G J Supran, K W Song, G W Hwang, R E Correa, J Scherer, E A Dauler, Y Shirasaki, M G Bawendi, and V Bulovi´c,
Adv Mater 27, 1437 (2015).
11 L Yang, Y Liu, Y.-L Zhong, X.-X Jiang, B Song, X.-Y Ji, Y.-Y Su, L.-S Liao, and Y He, Appl Phys Lett 106, 173109 (2015).
12 X T Feng, F Zhang, Y L Wang, Y Zhang, Y Z Yang, and X G Liu, Appl Phys Lett 107, 213102 (2015).
13 C Sun, Y Zhang, Y Wang, W Liu, S Kalytchuk, S V Kershaw, T Zhang, X Zhang, J Zhao, W W Yu, and A L Rogach,
Appl Phys Lett 104, 261106 (2014).
14 X Y Xu, R Ray, Y l Gu, H J Ploehn, L Gearheart, K Raker, and W A Scrivens, J Am Chem Soc 126, 12736 (2004).
15 S N Baker and G A Baker, Angew Chem., Int Ed 49, 6726 (2010).
16 J Fan and P K Chu, Small 6, 2080 (2010).
17 L.-S Li and X Yan, J Phys Chem Lett 1, 2572 (2010).
18 X Chen, W Zhang, Q Wang, and J Fan, Carbon 79, 165 (2014).
19 B S Mashford, M Stevenson, Z Popovic, C Hamilton, Z Zhou, C Breen, J Steckel, V Bulovic, M Bawendi, S C Sullivan, and P T Kazlas, Nat Photonics 7, 407 (2013).
20 P O Anikeeva, C F Madigan, J E Halpert, M G Bawendi, and V Bulovi´c, Phys Rev B 78, 085434 (2008).
21 J Kalinowski, M Cocchi, D Virgili, V Fattori, and J A G Williams, Adv Mater 19, 4000 (2007).
22 Z Zhao, B Xu, Z Yang, H Wang, X Wang, P Lu, and W Tian, J Phys Chem C 112, 8511 (2008).
23 V T Forster, Ann Phys 6, 55 (1948).
24 Y Li, A Rizzo, M Mazzeo, L Carbone, L Manna, R Cingolani, and G Gigli, J Appl Phys 97, 113501 (2005).
25 K van de Ruit, R I Cohen, D Bollen, T van Mol, R Yerushalmi-Rozen, R A J Janssen, and M Kemerink, Adv Funct Mater 23, 5778 (2013).
26 Y H Huang, W L Tsai, W K Lee, M Jiao, C Y Lu, C Y Lin, C Y Chen, and C C Wu, Adv Mater 27, 929 (2015).
27 P O Anikeeva, J E Halpert, M G Bawendi, and V Bulovi´c, Nano Lett 9, 2532 (2009).
28 R L Johnston and R Hoffmann, J Am Chem Soc 111, 810 (1989).
29 P Liu, Y L Cao, C X Wang, X Y Chen, and G W Yang, Nano Lett 8, 2570 (2008).
Trang 1030 S Yang and M Jiang, Chem Phys Lett 484, 54 (2009).
31 W K Bae, Y S Park, J Lim, D Lee, L A Padilha, H McDaniel, I Robel, C Lee, J M Pietryga, and V I Klimov, Nat Commun 4, 2661 (2013).
32 W Ji, P Jing, W Xu, X Yuan, Y Wang, J Zhao, and A K Y Jen, Appl Phys Lett 103, 053106 (2013).
33 Q Sun, Y A Wang, L S Li, D Wang, T Zhu, J Xu, C Yang, and Y Li, Nat Photonics 1, 717 (2007).
34 X Yang, Y Divayana, D Zhao, K S Leck, F Lu, S T Tan, A P Abiyasa, Y Zhao, H V Demir, and X W Sun, Appl Phys Lett 101, 233110 (2012).
35 D Parker, J Appl Phys 75, 1656 (1994).
36 S Habermehl and C Carmignani, Appl Phys Lett 80, 261 (2002).
37 P E Burrows and S R Forrest, Appl Phys Lett 64, 2285 (1994).
38 F Torricelli, D Zappa, and L Colalongo, Appl Phys Lett 96, 113304 (2010).
39 J Warga, R Li, S N Basu, and L Dal Negro, Appl Phys Lett 93, 151116 (2008).
40 Q Huang, S Zhao, Z Xu, X Fan, C Shen, and Q Yang, Appl Phys Lett 104, 161112 (2014).
41 B S Kim and J Y Lee, Org Electron 21, 100 (2015).
42 J Song, J Li, X Li, L Xu, Y Dong, and H Zeng, Adv Mater 27, 7162 (2015).