A hybrid structure of CdSe quantum dots (QDs) (λ = 640nm) spin-coated on the indium gallium nitride (InGaN) nanorod light-emitting diode (LED, λ = 525nm) is successfully fabricated. Experimental results indicate that the randomness and the minuteness of nanorods scatter the upcoming green light into the surrounding CdSe QDs efficiently, subsequently alleviating the likelihood of the emitted photons of red emission being recaptured by the CdSe QDs (self-absorption effect), and that increases the coupling probability of emission lights and the overall conversion efficiency. Moreover, the revealed structure with high color stability provides an alternative solution for general lighting applications of next generation.
© 2010 OSA
High luminescence efficiency and the size-tunable bandgap of colloidal CdSe quantum dots (QDs) have received considerable interest in the recent decade [1–8].The band structure of QDs can be altered according to the quantum confinement effect by varying the diameter, making it possible to determine the emission and absorption characteristics [9–11]. Consequently, colloidal QDs appear to be feasible as color conversion nanophosphors for solid state lighting, thus exacerbating the importance of the conversion efficiency of colloidal QDs. Indeed, the notion of using InGaN-based laser diodes and light-emitting diodes for the optical pumping of planar epitaxial structures with either CdSe QDs or CdZnSe QWs has been achieved for high efficiency green lasers and LEDs. By using CdSe QDs as an active element of the integral III-N/II-VI optical converter, an energy conversion efficiency as high as 14% is possible . Additionally, using CdZnSe QWs as a color-converting material with an InGaN-based blue LED can produce a green emission with an extremely high efficiency of 181lm/W . This study demonstrates the feasibility of a hybrid structure of CdSe QDs (λ = 640nm) spin-coated on the InGaN nanorod LEDs (λ = 525nm) to enhance its conversion efficiency. Despite a similar structure developed for ZnO nanorods to enhance its band-edge emission , this study discusses how to increase the overall conversion efficiency while adopting such a hybrid structure. Experimental results indicate that the conversion efficiency of this hybrid structure is more significantly enhanced (~1.7-fold) than that with the counterpart sample (CdSe QDs spin-coated on planar LED). This discrepancy is largely attributed to the random arrangement of a nanorod array that scatters an incoming green light onto the surrounding CdSe QDs efficiently, subsequently diminishing the self-absorption effect of CdSe QDs and increasing the conversion efficiency. Consequently, as is widely anticipated, such a hybrid structure of CdSe QDs spin-coated on the nanorod LEDs with high color stability provides an alternative solution for general lighting applications of next generation.
In this study, the InGaN nanorod LEDs were prepared by dry etching process. An LED wafer was grown on c-axis sapphire substrates by low-pressure metal-organic chemical vapor deposition. Following the growth of a GaN nucleation layer at 520°C, a 1.5μm-thick undoped GaN buffer layer and a 2.5μm-thick Si-doped n-type (5 × 1017cm−3) GaN cladding layer were grown at 1050°C. The active region consisted of five periods of 2.5nm In0.3Ga0.7N/10nm GaN multi-quantum wells (MQWs, λ = 525nm) with Si doping in GaN barriers to reduce the quantum-confined stark effect. Finally, a 300nm-thick Mg-doped p-type (3 × 1017cm−3) GaN cladding layer with a 50nm-thick AlGaN electron-blocking layer was grown on top at 1050°C. Consequently, a 5nm-thick nickel (Ni) metal film was deposited on the LED wafer by electron-beam evaporations. The sample underwent rapid thermal annealing at 850°C for 60sec in ambient nitrogen to yield randomly distributed Ni metal islands with nano-dimensions on the LED wafer surface, serving as a hard mask in the subsequent inductively-coupled-plasma (ICP) process. Then, 1.5 minutes of dry etching was conducted to form nanorod arrays on the top of p-type GaN. On the other hand, CdSe QDs were synthesized by a solgel process [15,16]. The typical synthesis procedure is as follows. Cd oleate complex, Cd(OA)2 were first prepared and obtained by reacting CdO with oleic acid at 150°C. A Se shot was then dissolved directly in trioctylphosphine (TOP) to form TOPSe. An excessive amount of TOPSe was loaded into a syringe, with its contents injected rapidly into a vigorously stirred reaction flask in a single injection through a rubber septum. The resulting reaction mixture of Cd(OA)2 and TOPSe was heated at 225°C to grow and obtain CdSe QDs. The CdSe QDs were then dissolved into a toluene solution with a concentration of 1%. Figure 1(a) plots the typical photoluminescence (PL) and absorption spectra of pure CdSe QDs. Accordingly, the main peak of CdSe QD emission is at λ = 640nm with a full width at half maximum (FWHM) of 45nm. Additionally, the absorption of CdSe ODs increases with a decreasing emitting wavelength of excitation source. In this study, we use InGaN green LEDs as the excitation source for CdSe QDs. Importantly, while the absorption of CdSe ODs of the green light is relatively lower than that of the blue or UV light, comparison is being made on the overall convention efficiency enhancement using the nanorod structure. A spin coating system was used to coat CdSe QDs onto the InGaN nanorod LED. For comparison, the sample of CdSe QDs spin-coated onto InGaN planar LED was also fabricated.
Figures 1(b) and (c) schematically depict the PL setup for assessing the hybrid structure of colloidal CdSe QDs spin-coated on planar (sample A) and nanorod (sample B) LEDs, respectively. Notably, the MQWs of sample B are still embedded in the p-type GaN, not exposed to the air, as shown in Fig. 1(c). In this study, both samples were excited by a semiconductor laser (λ = 405nm) from the backside of sapphire substrates. Therefore, green lights emitted from planar and nanorod LEDs were first excited by the incoming semiconductor laser. Red lights emitted from CdSe QDs spin-coated on top of both LEDs were then excited by green lights. Notably, the thicknesses of spin-coated CdSe QDs is maintained the same in both samples. Therefore, the variation in PL spectra observed in this study arose mainly from the nanorod structure. Consequently, the excited emissions, including green and red lights, were collected on top of both samples by an optical lens connected to an optical fiber and then analyzed by a spectrometer (Ocean Optics-USB4000).
3. Results and discussion
Figure 2(a) plots the PL spectra of planar and nanorod LEDs under an excited laser power of P = 20mW. The insert of this figure displays the SEM top view of the InGaN nanorod LED, where the scale bar represents 300 nm. According to this figure, the nanorod arrays are well aligned and have an average diameter ranging from 70 to 80nm and a length of around 250nm. Restated, thickness of the p-type GaN of our green LED is 300nm. Thus, for a nanorod 250nm in length, the MQWs are not exposed to the air, as previously discussed in reference to Fig. 1(c). Accordingly, a similar main peak appears in both samples at λ = 525nm with a FWHM of 33nm. Additionally, the peak intensity of PL spectrum of the nanorod LED substantially exceeds that of the planar LED. This finding is attributed mainly to the randomness and the minuteness of nanorods that eliminates the guiding modes that trap light, subsequently enhancing the extraction efficiency of a LED [17–19]. Figure 2(b) plots the integrated PL intensity as a function of excited laser power for both LEDs. The integrated PL intensity increases linearly with an increasing excited laser power for both samples of interest, indicating that among these samples, most energy of excitation laser is transferred to green emissions. Hence, the red emission of CdSe QDs pumped directly by the excitation laser is only slight and, therefore, can be omitted. Moreover, the integrated PL intensity of the nanorod LED is significantly higher than that of the planar LED, as shown in the insert of Fig. 2(b). On average, a two-folds enhancement of the integrated PL intensity is determined on the nanorod LED.
The scattering property of the nanorod LED was characterized by determining its total reflectance (Rtotal). Figure 3(a) schematically depicts the measurement setup. All reflectance measurements were performed for (1) incident lasers with the emitting wavelengths of λ = 532 and λ = 633nm, and (2) a broad range of incident angles (θ = 5°-70°), as well as for the (3) transverse electric (TE) and transverse magnetic (TM) polarizations of the incident light. The reflected light is collected by a commercially available integrating sphere. Next, the θ-dependence was investigated by mounting the sample at the sphere center, and the incident angle was varied by rotating the sample mount. The reflected light from the sample was collected by a calibrated silicon photon detector. Figure 3(b) shows Rtotal of the nanorod LED against θ for both TE (solid diamonds) and TM (solid dots), as evaluated by the incident laser of λ = 532nm. For comparison, this figure also shows Rtotal taken from a planar LED (open diamonds for TE and open dots for TM). Notably, Rtotal measured of the planar LED sample displays a fundamental behavior of external reflection ; for TM light, Rtotal gradually decreases with θ until reaching the Brewster angle, thus leading to lower reflectance than that of TE light. Meanwhile, for the planar LED sample, the intensity of Rtotal oscillated with θ. The oscillating fringes observed for both TE and TM light are attributed to the interference of incident light occurring at the GaN/air interface . Conversely, according to our results, Rtotal measured of the nanorod LED is relatively insensitive to θ and the polarization of incident light. Additionally, Rtotal of the nanorod LED for both TE and TM light are lower than that of the planar LED. The insert of Fig. 2(a) reveals that the dimension of individual nanorods is comparable to the optical wavelength of incident laser. Therefore, strong optical scattering is anticipated at the nanorod/air interface for both TE and TM incident light, and that is responsible for the insensitivity of Rtotal to polarization and angle, as well as its relatively lower value of Rtotal. Figure 3(c) shows Rtotal (θ) of both the nanorod and planar LEDs, as determined by the incident laser of λ = 633nm. According to this figure, Rtotal (θ) of both LEDs exhibit similar profiles determined by the incident laser of λ = 532nm, as discussed earlier the reference in Fig. 3(b). We can thus infer that the nanorod arrays can scatter the incident photons effectively with the emission wavelength ranging from λ = 532 to λ = 633nm.
Next, the optical properties of CdSe QDs spin-coated on both planar and nanorod green LEDs are examined. Figure 4 shows the PL spectra against an excited laser power for (a) sample A and (b) sample B, respectively. According to this figure, two distinct emission peaks appear for both samples. In Fig. 4(a), the PL intensity of green emission (planar LED) increases with an increasing excitation power from 5 to 50mW. However, the PL intensity of red emission (CdSe QDs) increases gradually with an increasing excitation power. Such saturated behavior of red emission diminishes the conversion efficiency of CdSe QDs and constrains its applications in general lighting. Conversely, according to Fig. 4(b), the PL peak intensity of green and red emissions in sample B increases simultaneously with an increasing excited laser power. This finding suggests that the saturated behavior of PL intensity of CdSe QDs in sample B is less significant than that of sample A. This same finding reveals that the composite of nanorods and CdSe QDs can affect the coupling of emission lights, ultimately increasing the conversion efficiency of CdSe QDs. The enhancement of conversion efficiency and saturation phenomena of PL intensity of CdSe QDs for both samples are discussed below.
Next, both samples are further compared in terms of conversion efficiency and saturation phenomenon. Figure 5(a) plots the integrated PL intensity of green and red emissions against excited laser power for both samples. In Fig. 5(a), these samples differ from each other in various ways. First, the red emission of sample B is considerably larger than that of sample A, although the much larger overall CdSe QDs volume in sample A than that in sample B at the same thickness of the QD layer. Indeed, as mentioned in our earlier discussion in reference to Fig. 3(b), the nanorod arrays randomly distributed on the top of p-GaN can effectively scatter the emitted photons of green emission into free space. Therefore, the higher extraction efficiency of green emission from the nanorod LED and the strongly expanded surface area of interaction of the excited green emission with the CdSe QDs make the hybrid structures based on the nanorod LED more efficient than that based on the planar LED. Second, the integrated PL intensity of green emission of sample A markedly exceeds that of the red emission, implying an inadequate conversion efficiency of CdSe QDs. Moreover, the integrated PL intensity of red emission of sample A is saturated gradually with an increasing excitation power. Above observations in sample A are owing to the self-absorption effect of CdSe QDs. In sample A, the red emission of CdSe QDs is locally pumped by the excited green emission, where the pumped area of CdSe QDs is roughly equivalent to the spot size of excited laser (8 × 103μm2). Therefore, with the increasing excited laser power, the emitted photons of red emission can be easily reabsorbed by CdSe QDs since most of them are confined within the region of laser spot size, thus strengthening the self-absorption effect of CdSe QDs and diminishing its overall conversion efficiency. Additionally, the strong self-absorption effect of CdSe QDs in sample A could also originate from its relatively larger overall CdSe QDs volume, as mentioned earlier. Finally, in sample B, the integrated PL intensity of red emission grows larger than that of the green emission for an excitation power of P<40mW. Moreover, the integrated PL intensity of red emission of sample B increases rapidly with an increasing excitation power. Importantly, the saturated phenomenon of red emission in sample B becomes less significant than sample A does. This finding suggests an enhanced conversion efficiency of CdSe QDs in this hybrid structure based on the nanorod LED. We believe that the enhanced conversion efficiency in sample B originates from the randomness and minuteness of the nanorod array that scatters emitted photons of both green and red emissions, as mentioned in our previous discussion in reference to Figs. 3(b) and 3(c). Consequently, the effective excited area of CdSe QDs expands due to the strong scattering of excited green emission at the interface of nanorod arrays, ultimately liberating the emitted photons of red emission outside the region of laser spot size. Furthermore, the likelihood of the emitted photons of red emission recaptured by the CdSe QDs is considerably decreased since most of them are effectively scattered by the nanorods as well. Consequently, in contrast with that of sample A, the self-absorption effect of CdSe QDs of sample B is obviously alleviated, leading to its less significant saturation phenomena of red emission and ultimately increasing the overall conversion efficiency.
Figure 5(b) plots the ratio of integrated PL intensity between red and green emission versus excited laser power for both samples. This figure reveals a slow decline for both samples. On average, the ratios are 0.68 and 1.18 for sample A and sample B with standard deviations of 0.12 and 0.23, respectively. Thus, with the same excited laser power, the conversion efficiency for sample B is increased by 1.7 folds over that for sample A.
Figure 6 plots (a) centroid wavelength and (b) FWHM for both samples as a function of the excited laser power. Here, a lower than 5nm offset in the green emission between planar and nanorod LEDs is related to compositional nonuniformities on the wafer. Additionally, the centroid wavelength for both green emissions undergo a blue-shift slightly with an increased excited laser power (525–523nm for the nanorod LED, 523–518nm for the planar LED). Conversely, with an increasing excited laser power, the peak emission of CdSe QDs is red-shifted for both samples (635–639nm for sample A and 636–638nm for sample B). In Fig. 6(b), the FWHM of CdSe QDs of sample A is always broader than that of sample B over all excited laser powers. However, no FWHM broadening of CdSe QDs is found for sample B. Restated, as compared to sample B, the larger long-wavelength shift of peak emission of CdSe QDs accompanying by the additional broadening of PL intensity in sample A is mainly arisen from its much larger self-absorption effect of CdSe QDs, and that is consistent with the previous discussion in reference to Fig. 5(a). Most important, above discussion also indicates that the hybrid structure based on the nanorod LED can function as a stable lighting source with acceptable color quality.
Figure 7 illustrates the variation of Commission Internationale de l’Eclairage (CIE) chromaticity coordinates for both samples with the excited laser power increasing from 5 to 50 mW. The variation of power-dependent PL spectra is reflected on the shift of CIE chromaticity coordinates from (0.427, 0.532) to (0.299, 0.628) and (0.467, 0.499) to (0.383, 0.571) for sample A and sample B, respectively. Again, as compared to sample B, the higher y-value observed on sample A is largely attributed to its early saturation of the red emission of CdSe QDs. Most important, the nanorod array randomly distributed on top of sample B surface evidently alleviates the saturation behavior of CdSe QDs, and that leads its much lower y-value than sample A with the same excited laser power.
In conclusion, this study presents a hybrid structure with a high conversion efficiency by combining green InGaN nanorod LEDs with luminescent colloidal CdSe QDs. The conversion efficiency for such a novel structure increases by 1.7 fold over that of the planar LED counterpart, which functions as the excitation source. We believe that the increased conversion efficiency originates from the randomness of the 2D nanorod array that scatters incoming green light efficiently into the surrounding CdSe QDs, thus alleviating the self-absorption effect of CdSe QDs and ultimately increasing the overall conversion efficiency. Results of this study concerning the optical properties of the nanorod LED significantly contribute to efforts to manufacture nanoscale photonic and optoelectronic devices.
The authors would like to thank the National Science Council of the Republic of China, Taiwan, for financially supporting this research under Contract No. NSC–98–2112– M–003–001–MY2.
References and links
1. P. Reiss, J. Bleuse, and A. Pron, “Highly luminescent CdSe/ZnSe core/shell nanocrystals of low size dispersion,” Nano Lett. 2(7), 781–784 (2002). [CrossRef]
3. C. Y. Huang, Y. K. Su, Y. C. Chen, P. C. Tsai, C. T. Wang, and W. L. Li, “Hybrid CdSe-ZnS quantum dot-InGaN-GaN quantum well red light-emitting diodes,” IEEE Electron Device Lett. 29(7), 711–713 (2008). [CrossRef]
4. H. S. Chen, C. K. Hsu, and H. Y. Hong, “InGaN-CdSe-ZnSe quantum dots white LEDs,” IEEE Photon. Technol. Lett. 18(1), 193–195 (2006). [CrossRef]
5. H. Song and S. Lee, “Red light emitting solid state hybrid quantum dot-near-UV GaN LED devices,” Nanotechnolgy 18(25), 255202 (2007). [CrossRef]
6. H. S. Chen, S. J. J. Wang, C. J. Lo, and J. Y. Chi, “White-light emission from organics-capped ZnSe quantum dots and application in white-light-emitting diodes,” Appl. Phys. Lett. 86(13), 131905 (2005). [CrossRef]
7. Y. Li, A. Rizzo, R. Cingolani, and G. Gigli, “White-light-emitting diodes using semiconductor nanocrystals,” Mikrochim. Acta 159(3-4), 207–215 (2007). [CrossRef]
8. S. Nizamoglu, G. Zengin, and H. V. Demir, “Color-converting combina tions of nanocrystal emitters for warm-white light generation with high color rendering index,” Appl. Phys. Lett. 92(3), 031102 (2008). [CrossRef]
9. A. P. Alivisatos, “Semiconductor clusters, nanocrystals, and quantum dots,” Science 271(5251), 933–937 (1996). [CrossRef]
11. H. F. Xiang, S. C. Yu, C. M. Che, and P. T. Lai, “Efficient white and red light emission from GaN/tris-(8-hydroxyquinolato) aluminum/platinum (II) meso-tetrakis(pentafluorophenyl) porphyrin hybrid light-emitting diodes,” Appl. Phys. Lett. 83(8), 1518–1520 (2003). [CrossRef]
12. S. V. Sorokin, I. V. Sedova, A. A. Toropov, G. P. Yablonskii, E. V. Lutsenko, A. G. Voinilovich, A. V. Danilchyk, Y. Dikme, H. Kalisch, B. Schineller, M. Heuken, and S. V. Ivanov, “High efficiency integral III-N/II-VI blue-green laser converter,” Electron. Lett. 43(3), 162–163 (2007). [CrossRef]
13. 3M claims record efficacy for green LED. LEDs magazine; 23 Nov 2009. http://www.ledsmagazine.com/news/6/11/24
14. L. J. Tzeng, C. L. Cheng, and Y. F. Chen, “Enhancement of band-edge emission induced by defect transition in the composite of ZnO nanorods and CdSe/ZnS quantum dots,” Opt. Lett. 33(6), 569–571 (2008). [CrossRef]
15. N. R. Jana, L. Gearheart, and C. J. Murphy, “Seeding Growth for Size Control of 5−40 nm Diameter Gold Nanoparticles,” Langmuir 17(22), 6782–6786 (2001). [CrossRef]
16. B. O. Dabbousi, J. Rodriguez-Viejo, F. V. Mikulec, J. R. Heine, H. Mattoussi, R. Ober, K. F. Jensen, and M. G. Bawendi, “(CdSe)ZnS Core−Shell Quantum Dots: Synthesis and Characterization of a Size Series of Highly Luminescent Nanocrystallines,” J. Phys. Chem. B 101(46), 9463–9475 (1997). [CrossRef]
17. I. Schnitzer, E. Yablonovitch, C. Caneau, T. J. Gmitter, and P. A. Schere, “30% external quantum efficiency from surface textured, thin-film light-emitting diodes,” Appl. Phys. Lett. 63(16), 2174–2176 (1993). [CrossRef]
18. Y. J. Lee, T. C. Lu, H. C. Kuo, and S. C. Wang, “High Light-Extraction GaN-based LEDs With Double Diffuse Surfaces,” IEEE J. Quantum Electron. 42(12), 1196–1201 (2006). [CrossRef]
19. Y. J. Lee, H. C. Kuo, S. C. Wang, T. C. Hsu, M. H. Hsieh, M. J. Jou, and B. J. Lee, “Increasing the extraction efficiency of AlGaInP LEDs via n-side surface roughening,” IEEE Photon. Technol. Lett. 17(11), 2289–2291 (2005). [CrossRef]
20. M. Born, and E. Wolf, Principles of optics 7th edition. Cambridge University Press, Cambridge, U.K., 46 (1999).
21. G. I. Surdutovich, J. Kolenda, J. F. Fragalli, L. Misoguti, R. Vitlina, and V. Baranauskas, “An interface method for the determination of thin film anisotropy,” Thin Solid Films 279(1–2), 119–123 (1999). [CrossRef]