Negative Thermal Quenching Behavior and Long Luminescence

Aug 21, 2008 - Negative Thermal Quenching Behavior and Long Luminescence Lifetime of Surface-State Related ... We suggest that a multiple trapping−d...
0 downloads 0 Views 951KB Size
Download PDF
14262

J. Phys. Chem. C 2008, 112, 14262–14265

Negative Thermal Quenching Behavior and Long Luminescence Lifetime of Surface-State Related Green Emission in ZnO Nanorods Haiping He, Zhizhen Ye, Shisheng Lin, Binghui Zhao, and Jingyun Huang* State Key Laboratory of Silicon Materials, Zhejiang UniVersity, Hangzhou 310027, People’s Republic of China

Haiping Tang State Key Laboratory of Silicon Materials, Zhejiang UniVersity, Hangzhou 310027, People’s Republic of China, and Department of Mechanical and Electrical Engineering, Baoji UniVersity of Arts and Sciences, Baoji, Shannxi 721007, People’s Republic of China ReceiVed: March 17, 2008; ReVised Manuscript ReceiVed: June 15, 2008

An experimental study of the luminescence intensity and time decay of the green band (GB) in ZnO nanorods as a function of temperature is reported. The GB intensity shows a negative thermal quenching behavior in the temperature range of 100-200 K. The increase of intensity is accompanied by a remarkable increase of lifetime with a maximum of ∼1.7 ms at 150 K, which is much longer than those documented values. We suggest that a multiple trapping-detrapping mechanism is responsible for the long lifetime. The results indicate that the surface state greatly influences the dynamics of the GB in ZnO nanostructures. Introduction

Experimental Details

One-dimensional (1D) ZnO nanostructures, especially nanowires or nanorods, have been a focus of current research due to their unique properties and prospect as building blocks for nano-LEDs1-3 and nanolasers.4 To explore the potentials of such nanostructures, it is essential to understand their optical properties which differ in relevant features from bulk crystals and the physics behind as well. Generally, 1D-ZnO nanostructures have a diameter from several tens to several hundreds of nanometers. Such sizes do not yet reach the quantum confinement region, but are small enough to exhibit the enhanced surface effects. It is expected that surface effects play a critical role in optical properties of semiconductor nanostructures and the performance of nanoscale optoelectronic devices. Indeed, a large number of publications have reported that the deep level emission around 2.4 eV in ZnO (usually called the green band, GB) is greatly enhanced in ZnO nanostructures, which is frequently attributed to surface states.5-11 Generally the strong GB is harmful to ZnObased short wavelength light-emitting devices. However, it may also find application as a possible UV-excited phosphor.7 Therefore the understanding of recombination dynamics becomes very important before we can either suppress or even make use of such emission in the future nano ZnO-based devices. So far little work on the recombination dynamics of the GB in ZnO nanostructures has been reported,6,11 and the mechanism of the decay behaviors is still unclear. In this work, temperature-dependent photoluminescence (PL) of ZnO nanorods was investigated. Negative thermal quenching behavior as well as a very long lifetime of ∼1.7 ms for the GB was observed. Thr trapping-detrapping mechanism based on surface states was proposed to interpret the experimental results. The present study may shed new light upon the understanding of the surface effect and the potential application of 1D-ZnO nanostructures.

ZnO nanorods were synthesized by a thermal evaporation method. The experimental setup was described in a previous work.12 Pure metallic Zn powders were used as the precursor. The source material was placed in a quartz boat, with a Si(111) substrate located 1 cm downstream to collect the product. The growth began by quickly inserting the quartz boat into the horizontal tube furnace at 680 °C in a nitrogen flow of 50 sccm (standard cubic centimeters per minute) for 5 min. Subsequently, oxygen was also introduced with a flux of 60 sccm for 30 min. This procedure produced a white layer on the Si substrates. Photoluminescence (PL) measurements were carried out on a FLS920 fluorescence spectrometer (Edinburgh Instruments). To avoid saturation of the GB, the sample was excited under low excitation density condition by using the 300 nm light from a 450 W Xe lamp as the excitation source. Time-resolved PL were excited by a µF920 microsecond flash lamp with a pulse width of ∼2 µs and a pulse period of 10 ms. The measuring temperature is 12-280 K. The room temperature decay of the GB was also measured by using the fourth (266 nm) harmonic of a Nd:YAG laser (spectra-physics, GCR 130) with a pulse width of 10 ns as the excitation source. The signal was detected with a Tektronix digital oscilloscope (model TDS 3052).

* Author to whom correspondence should be addressed. E-mail: [email protected].

Results and Discussion ZnO nanorods were synthesized by a thermal evaporation method. Scanning electron microscopy (SEM) revealed that the nanorods are randomly oriented and have diameters of ∼100 nm and length of ∼2 µm (Figure 1). The low-temperature (12 K) PL spectrum from the ZnO nanorods was plotted in Figure 2a. In the near band edge (NBE) region, a peak at 3.364 eV and a shoulder around 3.367 eV were observed. The 3.364 eV emission is well-known as excitons bound to donors (D0X).13 The shoulder at 3.367 eV has been reported by several groups14-17 and attributed to surface exciton (SX). With the temperature increasing, the intensity of both the D0X line and the SX line decreases quickly. The SX line quenches much faster and almost disappears at 50 K, as indicated in Figure 2b.

10.1021/jp8023099 CCC: $40.75  2008 American Chemical Society Published on Web 08/21/2008

Green Emission in ZnO Nanorods

J. Phys. Chem. C, Vol. 112, No. 37, 2008 14263

Figure 1. SEM image of ZnO nanorods. The scale bar is 2 µm.

Figure 3. (a) Temperature-dependent spectra of the GB and (b) integrated intensity of the GB vs reciprocal temperature. A normal thermal quenching behavior of the GB from ZnO nanorods is also plotted for comparison (red circles). The activation energy is estimated to be 114 meV. The inset shows the peak intensity ratio of the GB to the NBE as a function of temperature.

Figure 2. (a) The 12 K PL spectrum from ZnO nanorods. In the near band edge spectral range the resolution is set as 0.05 nm (∼0.45 meV), while in the visible range it is 0.5 nm. (b) PL spectra in the spectral range of excitonic recombination measured at various temperatures. D0X and SX lines can be identified.

The nanorods exhibit intense GB in the visible region. At 12 K, the peak intensity ratio of GB to NBE, PGB/PNBE, is about 0.13. With the temperature increasing, its value is greatly enhanced by 2-3 orders of magnitude, reaching ∼77 at 200 K and ∼14 at room temperature, as shown in the inset of Figure 3b. Figure 3 also plots the temperature-dependent spectra (Figure 3a) and integrated intensity of the GB. It can be divided into four temperature ranges. In the first range, 12-50 K, the GB intensity shows a slight increase with increasing temperature, while in the second range, 50-100 K, it decreases slightly. Interestingly, the GB intensity shows a clear increase with increasing temperature in the temperature range of 100-200 K, i.e., negative thermal quenching effect. Such a feature is clearly different from the normal thermal quenching behavior frequently observed in ZnO materials,18,19 as indicated in Figure 3b. Finally, it decreases abruptly by about 1 order of magnitude

as the temperature increases from 200 to 280 K. Some authors also reported similar temperature behavior for GB in ZnO (refs 19–21) and deep level emission in GaN.22 For example, Djurisic et al.20 found the GB intensity increases monotonously with the temperature from 12 to 200 K in ZnO nanoshells. However, the authors did not discuss the origin of such behavior. From Figure 3b, an activation energy of ∼114 meV for the thermal quenching could be estimated. This value may reflect the thermal activation of nonradiative recombination centers or ionization of donor/acceptor levels involved in the radiative recombination. To obtain more information about the origin of the negative thermal quenching of the GB, transients on the maximum of the GB were measured at various temperatures, as plotted in Figure 4. With the use of a deconvolution technique, the data can be fitted with multiexponential decay

I(t) ) A1 exp(-t/τ1) + A2 exp(-t/τ2) + A3 exp(-t/τ3) (1) where I is the luminescence intensity, τi the lifetime, and Ai the constant. For comparison, the average lifetime, 〈τ〉 ) ∑i Aiτi2/ ∑i Aiτi, was also calculated.23 The fitting parameters and results are listed in Table 1. Very interestingly, the lifetime first increases and then decreases with increasing temperature, both remarkably. At 150 K, the lifetime reaches a maximum value

14264 J. Phys. Chem. C, Vol. 112, No. 37, 2008

He et al.

Figure 5. Decay curves of the GB measured at 300 K by using a pulsed laser as the excitation source and an oscilloscope as the recorder.

Figure 4. Decay curves of the GB measured at various temperatures. The brown line represents the instruments response to the excitation source. The decay curves were fitted to eq 1 by using a deconvolution technique. The fitting parameters were listed in Table 1.

TABLE 1: Time-Resolved PL Decay Time Constants and Amplitude Ratios for the GB at Various Temperaturesa temp [K]

τ1 [µs]

WF1 [%]b

τ2 [µs]

WF2 [%]

τ3 [µs]

WF3 [%]

〈τ〉 [µs]

12 50 100 150 200 240 280

2.7 2.2 7.1 26.3 34.7 16.9 1.1

34.7 17.9 29.4 28.5 38.3 74.9 84.8

17.7 17.2 36.6 223.4 222.4 114.3 6.1

45.4 54.7 34.8 34.2 38.9 25.1 15.2

145.3 133.1 284.2 1886.4 1217.2

19.9 27.4 35.8 37.3 22.8

114.7 108.4 252.2 1707.3 947.5 84.5 3.6

a The data are derived from fitting the decay curves to eq 1. The definition of 〈τ〉 is according to ref 23. b WFi (Weight Factor) )Ai/ (A1+A2+A3), i ) 1, 2, 3.

of ∼1.7 ms, which is surprisingly long in comparison with the commonly reported value (in the µs range, ref 6 and therein). At room temperature, however, the average lifetime decreases very quickly to 3.6 µs, in agreement with most results in the literature. The microsecond-scale lifetime of the GB is further confirmed by a more conventional setup for PL transient measurements by using pulsed laser with short pulse width as the excitation source and the oscilloscope as the recorder. Figure 5 shows the result measured at 300 K. By fitting the data to eq 1, an average lifetime of 2.4 µs can be obtained, which is very close to the value of 3.6 µs. The discrepancy between the two values may result from the different measuring temperature. The lifetime of the GB in various ZnO materials, including bulk crystals, films, and nanostructures, has been extensively investigated.24-26 It was found that the lifetimes of GB change from several hundreds of picoseconds to microseconds or even to milliseconds, which depends strongly on the ZnO forms and growth methods. Guo et al.24 observed a lifetime of 200 ps for the GB in ZnO films and attributed the GB to tunnel-assisted DAP (donor-acceptor pair) recombination. Foreman et al.7 and Djurisic et al.20 reported the GB lifetime of ∼50 ns and 30-100

ns in sulfur-doped ZnO nanaowires and ZnO nanoshells, respectively, without discussing the recombination dynamics. Most work in the literature reported microsecond-lifetime for the GB. Among those works, Studenikin25 suggested that the slow component (a few microseconds) was explained as an electron-hole recombination in a donor-acceptor complex. van Dijken et al.6 investigated the dynamics of the radiative and nonradiative processes in ZnO nanoparticles. They suggested that the GB was due to a transition of a free electron to a deeply trapped hole. The dynamics of the GB was explained by a model in terms of hole transfer from the valence band to a hole trap, in which a surface trapping process played an important role. The energy-state model12 and the electron-phonon coupling model26 were also invoked by some researchers to explain the decay behavior of the GB in ZnO nanoparticles. However, only a few works focus on the effects of surface states on the recombination dynamics of the GB. There are several competing processes which can affect the temperature dependence of the GB intensity. First of all, the thermal activation of nonradiative recombination centers (NRC) at elevated temperatures results in a decrease of the GB intensity, as is the case in other luminescence bands. In the viewpoint of Garces et al.,27 the unstructured GB in ZnO is a DAP-like emission. Therefore, the delocalization of carriers from the donor or acceptor impurities at elevated temperatures also results in quench of the GB. On the contrary, the recapture of carriers by the impurities and subsequent recombination leads to an increase of the GB intensity. In the temperature range of 12-50 K for our sample, the slight increase of the GB intensity with temperature was accompanied by a sharp decrease of the D0X and SX intensity, suggesting that the recapture of carriers dissociated from excitons by the donor and acceptor impurities is responsible for the GB increase. In this temperature range, the lifetime slightly decreases due to the enhanced contribution of the nonradiative recombination, which is a much faster process. As the temperature further increases, however, the lifetime becomes much longer. This may indicate a carrier diffusion process that could arise from a multiple trapping-detrapping mechanism through surface states. Due to the high surface-to-volume ratio of the ZnO nanorods, the surface is expected to play an important role in the carrier transport and recombination. Surface states can serve as traps for photoexcited electrons and holes. It is generally

Green Emission in ZnO Nanorods

J. Phys. Chem. C, Vol. 112, No. 37, 2008 14265 Conclusions In summary, we observed negative thermal quenching behavior of the GB in ZnO nanorods in the temperature range of 100-200 K. With the temperature increasing, the lifetime of the GB increases abruptly with a maximum of ∼1.7 ms at 150 K. We suggest that the very long lifetime arises from the trapping-detrapping process of the photoexcited carriers by surface states. The present study demonstrated that surface states play a very important role in the optical properties of ZnO nanostructures and further tailoring them by suppressing or even using surface effects would be possible. Acknowledgment. The authors thank Prof. Jiahua Zhang for PL decay measurements. This work was supported by the National Natural Science Foundation of China under Grant No. 50572095 and the “973” Program under Grant No. 2006CB604906. References and Notes

Figure 6. (a) Low-temperature PL spectra and (b) integrated intensity of the structured GB from an annealed ZnO nanorod sample.

accepted that the GB in ZnO is due to the defects located in the surface region. For example, by using a metal plasma immersion ion plantation, Yang et al.11 found that the GB origin of ZnO nanostructures is about 0.5-7 nm in depth from the surface. To support the assumption that the results are indeed caused by surface states, temperature-dependent PL of annealed ZnO nanorods with approximately uniform size (∼200 nm in diameter) were also measured. The GB in the annealed nanorods sample is a structured one at low temperature (see Figure 6), clearly different from the one in the present case. The features of the structured GB are almost identical to those observed by Garces et al.27 in annealed ZnO crystal, which is believed to be of different origin from the unstructured GB. From Figure 3b and Figure 6b, one can find the thermal quenching behavior of the unstructured and structured GB is very similar, suggesting that the negative thermal quenching is independent of the nature of the GB and results from the surface states. Accordingly, the negative thermal quenching and decay behavior can be interpreted in terms of a multiple trapping-detrapping process. At low temperatures, no detrapping from the surface states is allowed. As the temperature increases, however, the release rate of the carriers increases remarkably, which results in an increase of the GB intensity. At temperatures above 200 K, the long lifetime component disappears, which may suggest overwhelmingly domination of the fast nonradiative recombination process over the trapping-detrapping process. This is also evidenced by the sudden decrease of the GB intensity in this temperature range. In a word, after being generated by photoexcitation, electrons and holes may be captured by three kinds of centers, that is, surface trap, NRC, and localized impurity levels. The competition between the three processes in different temperature ranges is responsible for the unusual thermal quenching and decay behaviors.

(1) Konenkamp, R.; Word, R. C.; Godinez, M. Nano Lett. 2005, 5, 2005. (2) Bao, J. M.; Zimmler, M. A.; Capasso, F.; Wang, X. W.; Ren, Z. F. Nano Lett. 2006, 6, 1719. (3) Jeong, M. C.; Oh, B. Y.; Ham, M. H.; Myoung, J. M. Appl. Phys. Lett. 2006, 88, 202105. (4) Zhou, H. J.; Wissinger, M.; Fallert, J.; Hauschild, R.; Stelzl, F.; Klingshirn, C.; Kalt, H. Appl. Phys. Lett. 2007, 91, 181112. (5) Shalish, I.; Temkin, H.; Narayanamurti, V. Phys. ReV. B 2004, 69, 245401. (6) van Dijken, A.; Meulenkamp, E. A.; Vanmaekelbergh, D.; Meijerink, A. J. Phys. Chem. B 2000, 104, 1715. (7) Foreman, J. V.; Li, J. Y.; Peng, H. Y.; Choi, S.; Everitt, H. O.; Liu, J. Nano Lett. 2006, 6, 1126. (8) Hsu, N. E.; Hung, W. K.; Chen, Y. F. J. Appl. Phys. 2004, 96, 4671. (9) Li, D.; Leung, Y. H.; Djurisic, A. B.; Liu, Z. T.; Xie, M. H.; Shi, S. L.; Xu, S. J.; Chan, W. K. Appl. Phys. Lett. 2004, 85, 1601. (10) Djurisic, A. B.; Leung, Y. H.; Choy, W. C. H.; Cheah, K. W.; Chan, W. K. Appl. Phys. Lett. 2004, 84, 2635. (11) Yang, Y.; Sun, X. W.; Tay, B. K.; Cao, P. H. T.; Wang, J. X.; Zhang, X. H. J. Appl. Phys. 2008, 103, 064307. (12) Tang, G. Q.; Xiong, Y.; Zhang, L. Z.; Zhang, G. L. Chem. Phys. Lett. 2004, 395, 97. (13) Meyer, B. K.; Alves, H.; Hofmann, D. M.; Kriegseis, W.; Forster, D.; Bertram, F.; Chresten, J.; Hoffmann, A.; Strabburg, M.; Dworzak, M.; Haboeck, U.; Rodina, A. V. Phys. Status Solidi B 2004, 241, 231. (14) Wischmeier, L.; Voss, T.; Ru¨ckmann, I.; Gutowski, J.; Mofor, A. C.; Bakin, A.; Waag, A. Phys. ReV. B 2006, 74, 195333. (15) Grabowska, J.; Meaney, A.; Nanda, K. K.; Mosnier, J. P.; Henry, M. O.; Ducle`re, J. R.; McGlynn, E. Phys. ReV. B 2005, 71, 115439. (16) Chang, P. C.; Chien, C. J.; Stichtenoth, D.; Ronning, C.; Lu, J. G. Appl. Phys. Lett. 2007, 90, 113101. (17) Yang, Y.; Tay, B. K.; Sun, X. W.; Sze, J. Y.; Han, Z. J.; Wang, J. X.; Zhang, X. H.; Li, Y. B.; Zhang, S. Appl. Phys. Lett. 2007, 91, 071921. (18) Su, F. H.; Liu, Y. F.; Chen, W.; Wang, W. J.; Ding, K.; Li, G. H.; Joly, A. G.; McCready, D. E. J. Appl. Phys. 2006, 100, 013107. (19) Klason, P.; Børseth, T. M.; Zhao, Q. X.; Svensson, B. G.; Kuznetsov, A. Y.; Bergman, P. J.; Willander, M. Solid State Commun. 2008, 145, 321. (20) Djurisic, A. B.; Leung, Y. H.; Tam, K. H.; Hsu, Y. F.; Ding, L.; Ge, W. K.; Zhong, Y. C.; Wong, K. S.; Chan, W. K.; Tam, H. L.; Cheah, K. W.; Kwok, W. M.; Philips, D. L. Nanotechnology 2007, 18, 095702. (21) Hauser, M.; Hepting, A.; Hauschild, R.; Zhou, H. J.; Fallert, J.; Kalt, H.; Klingshirn, C. Appl. Phys. Lett. 2008, 92, 211105. (22) Reshchikov, M. A.; Shahedipour, F.; Korotkov, R. Y.; Wessels, B. W.; Ulmer, M. P. J. Appl. Phys. 2000, 87, 3351. (23) Kamat, P. V.; Patrick, B. J. Phys. Chem. 1992, 96, 6829. (24) Guo, B.; Qiu, Z. R.; Wong, K. S. Appl. Phys. Lett. 2003, 82, 2290. (25) Studenikin, S. A.; Cocivera, M. J. Appl. Phys. 2002, 91, 5060. (26) Shi, J. Y.; Chen, J.; Feng, Z. C.; Chen, T.; Wang, X. L.; Ying, P. L.; Li, C. J. Phys. Chem. B 2006, 110, 25612. (27) Garces, N. Y.; Wang, L.; Bai, L.; Giles, N. C.; Halliburton, L. E.; Cantwell, G. Appl. Phys. Lett. 2002, 81, 622.

JP8023099