The substrates in emitting structure were found to have an influence on the surface plasmon mediated light emission of ZnO films. Ag film mediated photoluminescence was quenched for ZnO on silicon substrate but enhanced for ZnO on quartz or sapphire substrate. Through a theoretical simulation, the quenching for ZnO on silicon substrate is ascribed to the power lost to the substrate mode nonradiatively at the expense of the power coupled to the SP mode. The substrate with a high refractive index may capture and dissipate the emitting power which limits the efficiency of SP mediated light extraction. Therefore, a proper arrangement of the refractive index of the substrate and emitting layers in the device structure is decisive for the SP coupled light emission enhancement. Base on the theoretical analysis, a four-layered structure was advanced to recover SP mediated emission enhancement from ZnO film on silicon substrate.
© 2008 Optical Society of America
It is well-known that an electric dipole placed in the proximity of metal surface may transfer energy to surface plasmon (SP), resulting in modification of its spontaneous emission decay rate . This effect has been used to enhance the photoluminescence (PL) of light emitters, further aiming to raise the emission efficiency of light emission diodes (LEDs) [2–6]. From the published reports [2–4], the morphology of metal films is found absolutely important for SP mediated emission enhancement, because after radiation energy is transferred to SP, only by scattering which bridges the momentum gap, the coupled SP energy can be transferred to free space radiation, otherwise it is only dissipated, inducing quenching of the emission. Thus, metal films with periodic grating or random corrugation was usually adopted to realize light emission enhancement.
In addition to the morphology of metal films, other factors such as the separation distance between metal films and light emitters and the energy matching between SP mode and emission band, are also found to be influential for SP coupled emission enhancement [5–7]. Whereas, there are few reports about the influence of emitting device structures on the SP coupled light emission. As we known, much of the power generated in light emitters is lost in LEDs before it is extracted to free space emission. Coupling with SP and then effective scattering radiatively is a promising way to recover part of the dissipated power, especially in top-emitting LED . However, other power dissipating channels existing in device structures may compete with the beneficial SP coupling, resulting in the failure of SP mediated light extraction through the metal films. So the investigation on the dependence of the SP coupling on emitting structures is indispensable for designing high efficiency SP enhanced LEDs. In this paper, the emphasis was put on the influence of substrates in emitting devices on the SP coupled emission. It was found that the SP mediated emission from ZnO film based on silicon substrate was quenched, but that from ZnO based on quartz or sapphire substrate were enhanced. It is believe that the competition between the power transferred to the substrate mode or to the SP mode resulted in the emission enhancement or quenching.
2. Sample fabrication and measurement
The fabrication started from the ZnO film (40 nm thick) deposition on three types of substrates (quartz, sapphire and Si) by reactive direct current sputtering. Then the films were annealed in O2 at 800 °C for 2 hours. Subsequently, Ag films were sputtered onto the ZnO films at 200 °C with a deposition time of 60s. The three samples are referred as quartz/ZnO/Ag, sapphire/ZnO/Ag and Si/ZnO/Ag, respectively. Before Ag sputtering, half of each sample was covered to work as the reference samples for comparing the PL intensity. The schematics of the Si/ZnO/Ag sample structure is illustrated in Fig. 1(a). The PL measurements were performed with a He-Cd laser excitation at 325nm with an incidence angle of 45°, and detected by a spectrometer (Acton SP2500i) on the Ag film side. Cross-section scanning electron microscopy (SEM) images of the samples were obtained from a HITACHI S-4800 microscope.
Figure 1(b) shows the cross-section SEM image of the quartz/ZnO/Ag structure. The Ag film thickness was about 40nm. It can be seen that Ag island microstructures were formed, which ensures that the energy coupled to the SP can be effectively scattered and transferred into radiative emission [9, 10]. Figure 1(c) illustrates the schematics of the structure of the four-layered sample (Si/SiO2/ZnO/Ag). In this sample, before the ZnO film and Ag film sputtering, the SiO2 layer with a thickness of 200nm was first formed on silicon substrate by thermal oxidation. A reference sample without Ag sputtering was also prepared for PL comparison.
3. Results and discussion
Figure 2 shows the PL spectra of the Ag sputtered ZnO films on three substrates and their respective reference samples. It can be seen that the PL peak occurs at about 380nm for the three samples, which is the band edge emission of ZnO films. The PL intensity is about 3 times higher for the sample quartz/ZnO/Ag and sapphire/ZnO/Ag than for the respective reference samples without Ag sputtering. However, for the sample Si/ZnO/Ag, the PL intensity was decreased after Ag sputtering.
As we known, the decay rates of a light emitter can be modified when it is situated near an interface or in a microcavity due to the modification of the photonic mode density (PMD) . In our experiment, the samples have a three-layered structure in which the ZnO film is situated between Ag layer and substrate. Therefore, the de-excited energy from the dipoles in ZnO may couple to the waveguide mode in the substrate and SPP mode at the ZnO/Ag interface. The coupled energy can be lost nonradiatively due to absorption in substrate waveguide and SPP absorption. In those cases, the radiative decay rate will be decreased and induces the quenching of PL. However, the SP coupled energy can also be scattered and reemitted into effective free space radiation. In this case, due to the high PMD associated with the SPP mode, the radiative decay rate of ZnO can be raised greatly, leading to the PL enhancement . So the PL intensity enhancement or quenching is mainly determined by three factors: power lost to the substrate waveguide mode, power coupled to the SP at ZnO/Ag interface, and the out-coupling efficiency from SP to free space emission. The third factor, i.e., out-coupling efficiency between SP and radiative emission, is mainly determined by the thickness and morphology of Ag layer . In our experiment, the Ag layers for the three samples were same, due to the same preparation condition. Furthermore, the surface roughness of ZnO on each substrate is small, so the influence from the interface between ZnO and Ag layer can be neglected. Thus, the PL enhancement or quenching for the three samples is mainly determined by the first two factors: power dissipated to the substrate mode and SP mode.
Using a well-established theory of the dipole radiation in multi-layered media [11, 12], we simulated how much power radiated from ZnO is coupled to substrate and SP modes in the three-layered structure. A convenient approach adopted in this theory was treating light emitters as source radiating dipoles. The reflections by the interfaces between the different materials that make up the multilayer structure act back on the source dipoles, and the effect of these reflections has on the power dissipated by the dipole was calculated. The obtained dissipated power as a function of the in-plane wave vectors is a measure of power lost to each mode in the layered structure [11, 12]. In our computation, the ZnO film is assumed as isotropic oriented dipoles situated in the median plane of the ZnO layer. The obtained power dissipation spectrum S (u) as a function of in-plane wavevector u is showed in Fig.3 (a). The refractive indexes of the layers at 380nm used in the calculation are 1.47, 1.76, 6.06+0.63i and 0.3+3.1i for quartz, sapphire, silicon and Ag island films respectively [13, 14].
It can be seen from Fig. 3(a) that a peak in u>1 region occurs in the power spectra for the three samples. This peak is related to the SP modes at the ZnO/Ag interface and indicates that much of relaxed power of the dipoles in ZnO is coupled to the SP modes. The oscillation observed in u<1 region is caused by the destructive or constructive interference of dipole field with the reflected ones from the interface [15, 16]. Comparing the spectra, two features are observed: (1) from the sample quartz/ZnO/Ag, sapphire/ZnO/Ag to Si/ZnO/Ag, the SPP mode shifts to the higher momentum. It is caused by the increase of the refractive index of the substrate which raises the effective dielectric constants experienced by the SPP mode ; (2) The amplitude of SP modes of quartz/ZnO/Ag and sapphire/ZnO/Ag is very close, and both larger than that of the Si/ZnO/Ag (note the log scale). Thus, it can be inferred that the power coupled to the SPP modes in the quartz/ZnO/Ag and sapphire/ZnO/Ag is stronger than that in the Si/ZnO/Ag.
Due to the lower refractive index for quartz and sapphire comparing to that of ZnO, no substrate waveguide mode was expected in the power spectra in Fig. 3(a). However, it was also not observed in the power spectrum of the Si/ZnO/Ag. We consider that it is probably due to the influence from the relative large imaginary part of the complex refractive index of silicon substrate which will dissipate most of the substrate mode power by absorption. Then the substrate waveguide mode may be broadened and screened in the spectrum .
To make clear the influence from the substrate mode and confirm our assumption, the power spectra were calculated by varying the real part of refractive index of the substrate with the imaginary part set to zero. The results are showed in Fig. 3 (b). As shown, the peaks related to the substrate waveguide modes indicated by the arrows are observed apart from the SP mode. It can be clearly seen that with the increase of the refractive index, the peaks shifts to the higher wavenumber. In addition, the power coupled to the substrate mode is at the expense of the power coupled to the SP mode. When the refractive index of the substrate is larger than that of the ZnO (n=2), the power dissipated to the waveguide mode is increased. A substantial portion of the power is coupled to the substrate mode, and it will be absorbed and has no contribution to the radiative emission. As a consequence, the power coupled to the SP mode is decreased, and then the out-coupled radiative emission by SP scattering is decreased accordingly. In a word, the emission intensity is determined by the competition between the power dissipated to the SP mode or to the substrate mode. When more power is coupled to the SP, the radiative scattered emission will be increased, inducing PL enhancement, just as that in the samples of the quartz/ZnO/Ag and sapphire/ZnO/Ag.
To overcome the quenching problem of the SP coupled emission in the ZnO on silicon substrate, a four-layered structure is proposed. As Fig. 1(c) shown, a layer of SiO2 was inserted between silicon substrate and ZnO film. This layer introduces two interfaces and has a low refractive index. The calculated power dissipation spectrum of this structure is shown in Fig. 4(a). The power spectrum of the three layered sample (Si/ZnO/Ag) is also included for comparison. It is found that the power amplitude of the SP mode for the four-layered structure was increased significantly comparing with that for the Si/ZnO/Ag. The power coupled to SP mode is greatly raised, and PL enhancement can be expected from this structure. To confirm the theoretical prediction, the PL of the structure was measured, and the results are illustrated in Fig. 4(b). The PL intensity of the Si/SiO2/ZnO/Ag is really enhanced more than two fold comparing to the control sample. Despite the fact that the enhancement ratio for this four-layer silicon based sample is less than that for the three-layer quartz or sapphire based sample, probably due to losses of absorption and scattering at the SiO2/Si interface, this structure conquers the problem of SP coupled emission quenching of ZnO on silicon substrate, which is promising in design SP enhanced silicon based emission devices.
In conclusion, the substrate in the emitting device structure was found to be influential on the SP mediated light emission through metal films. Theoretical simulation shows that competition between the power dissipated to the SPP and substrate mode was the cause of PL enhancement or quenching. Proper arrangement of the refractive index of the substrates and emitting layers in the devices is decisive for the SP coupled light emission enhancement.
The authors express their appreciations to the Natural Science Foundation of China (No. 60606001, National Basic Research Program of China (973 Program, No. 2007CB613403) and PCSIRT(IRT0651) project for the financial support.
References and links
1. R. R. Chance, A. Prock, and R. Silbey, “Lifetime of an emitting molecule near a partially reflecting surface,” J.Chem. Phys. 60, 2744–2748 (1974). [CrossRef]
2. K. Okamoto, I. Niki, A. Shvartser, Y. Narukawa, T. Mukai, and Axel Scherer,“Surface-plasmon-enhanced light emitters based on InGaN quantum wells,” Nat. Mater. 3, 601–605 (2004). [CrossRef] [PubMed]
3. C. J. Yates, I. D. W. Samuel, P. L. Burn, S. Wedge, and W. L. Barnes, “Surface plasmon-polariton mediated emission from phosphorescent dendrimer light-emitting diodes,” Appl. Phys. Lett. 88, 161105-1-3 (2006). [CrossRef]
4. P. Cheng, D. Li, Z. Yuan, P. Chen, and D. Yang, “Enhancement of ZnO light emission via coupling with localized surface plasmon of Ag island film,” Appl. Phys. Lett. 92, 041119-1-3 (2008). [CrossRef]
5. J. S. Biteen, D. Pacifici, N. S. Lewis, and H. A. Atwater, “Enhance radiative emission rate and quantum efficiency in coupled silicon nanocrystal-nanostructured gold emitters,” Nano Lett. 5, 1768–1773 (2005). [CrossRef] [PubMed]
6. D. Y. Lei, J. Li, and H. C. Ong, “Tunable surface plasmon mediated emission from semiconductors by using metal alloys,” Appl. Phys. Lett. 91, 021112-1-3 (2007). [CrossRef]
9. W. L. Barnes and P. T. Worthing, “Spontaneous emission and metal-clad microcavities,” Opt. Commun. 162, 16–20 (1999). [CrossRef]
10. C.-Y. Chen, D.-M. Yeh, Y.-C. Lu, and C. C. Yang, “Dependence of resonant coupling between surface plasmons and an InGaN quantum well on metallic structure,” Appl. Phys. Lett. 89, 203113-1-3 (2006). [CrossRef]
11. R. R. Chance, A. Prock, and R. Silbey, “Comments on the classical theory of energy transfer,” J. Chem. Phys. 62, 2245–2253 (1975). [CrossRef]
12. K. G. Sullivan and D. G. Hall, “Enhancement and inhibition of electromagnetic radiation in plane-layered media,” J. Opt. Soc. Am. B 14, 1149–1159 (1997). [CrossRef]
13. S. Kawabata, K. Ishihara, Y. Hoshi, and T. Fukazawa, “Observation of silver film growth using an in situ ultra-high vacuum spectroscopic ellipsometer,” Thin Solid Film 313–314, 516–521 (1998). [CrossRef]
14. Edward D. Palik, Handbook of optical constant of solid (Academic, 1985)
15. B. J. Soller and D. G. Hall, “Energy transfer at optical frequencies to silicon-based waveguiding structures,” J. Opt. Soc. Am. A 18, 2577–2584 (2001). [CrossRef]
16. T. Nakamura, M. Fujii, K. Imakita, and S. Hayashi, “Modification of energy transfer from Si nanocrystals to Er3+ near a Au thin film,” Phys. Rev. B 72, 235412-1-6 (2005) [CrossRef]
17. R. M. Amos and W. L. Barnes, “Modification of the spontaneous emission rate of Eu3+ ions close to a thin metal mirror,” Phys. Rev. B 55, 7249–7254 (1997). [CrossRef]
18. J. Kalkman, H. Gersen, L. Kuipers, and A. Polman, “Excitation of surface plasmons at SiO2/Ag interface by silicon quantum dots:experiment and theory,” Phys. Rev. B 73, 075317-1-8 (2006). [CrossRef]