We report the electrically pumped ultraviolet random lasing from ZnO nanorod arrays on Si. Metal-insulator-semiconductor structures in a form of Au/SiO2/ZnO-nanorod-array were fabricated on Si. Such devices exhibit random lasing when the Au electrode is applied with a sufficiently high positive voltage. In this context, in the region adjacent to SiO2/ZnO-nanorod-array interface, stimulated emission from ZnO occurs due to population inversion and, moreover, light is scattered by the nanorods and SiO2 films. Therefore, random lasing proceeds due to optical gain achieved by the stimulated emission and multiple scattering.
© 2009 Optical Society of America
ZnO features a large exciton binding energy of ~60 meV, substantially greater than the thermal energy (~26 meV) at room temperature (RT) . In this context, exciton recombination, as a more effective radiative process with respect to the electron-hole plasma process, can facilitate low-threshold stimulated emission [2,3]. The report on RT ultraviolet (UV) ZnO nanowire nanolasers  has greatly spurred the research enthusiasm in fulfilling lasing from diverse ZnO micro/nanostructures [5–12]. The laser action from the ZnO micro/nanostructures can be generally understood in terms of three feedback mechanisms. In a single nanowire/nanorod, Fabry-Pérot (F-P) resonator is naturally formed by the two smooth end facets acting as reflecting mirrors [3–5]. For the disordered structures, multiple scattering with optical gain contributes to random lasing [6–8]. For c-axis symmetrical micro/nanodisks and microrods, the lasing occurs via whispering-gallery mode (WGM) resonator in which the light wave circulates around due to multiple total internal reflection at the resonator boundaries [10–12]. For the ZnO nanorod/nanowire lasers based on the F-P/WGM resonators, other than the random lasers, the length and end-face morphology of the ZnO nonorods/nanowires should be well controlled. Up to now, to the best of our knowledge, all the reported laser actions from the ZnO nanorods/nanowires were created by optical pumping, which is not desirable for the practical application. Therefore, to realize the electrically pumped laser actions even the random lasing from the ZnO nanorods/nanowires is worth an effort.
Previously, we have reported the electrically pumped random lasing from c-axis oriented ZnO films taking advantage of metal-insulator-semiconductor (MIS) structures [13,14]. For the ZnO film, the grains are closely packed in the plane of film. In this case, optical scattering that is critical for random lasing proceeds via the grain boundaries. It is generally believed that the optical scattering is strong in a space with variation of refractive index (nr) . Evidently, it is quite difficult to substantially enhance the spatial variation of nr within ZnO film. While, for ZnO nanorod array, the spatial variation of nr can be flexibly modified because there are interspaces among the nanorods, which offer the possibility of filling foreign materials with nr different from that of ZnO. Therefore, ZnO nanorod array predominates over ZnO film for the enhancement of optical scattering. Nevertheless, the electrically pumped random lasing from ZnO nanorod arrays has not been reported as yet. Generally, current injection via a p-n junction is preferred for the electrical pumping of diode lasers. Unfortunately, it is quite difficult to fabricate ZnO-nanorod-based p-n junctions due to the difficulty in p-type doping of the ZnO nanorods. In this paper, we have electrically driven the ZnO nanorod arrays to generate random lasing by means of the Au/SiO2/ZnO-nanorod-array MIS structures. In our strategy, both the ZnO nanorod arrays and SiO2 film are prepared by simple chemical routes. Moreover, the difficulty in p-type doping of ZnO nanorods is avoided. To the best of our knowledge, the electrically pumped random lasing from ZnO nanorod arrays is firstly demonstrated. More importantly, the ZnO naonorod arrays allow accommodation of other optically functional materials. This virtue might be used to develop diverse optoelectronic devices based on the proof-of-concept random lasers presented herein.
2. Experimental details
Figure 1 shows the schematic diagram of a MIS device based on the ZnO nanorod array on Si substrate. In our experiment, the devices were fabricated through the procedures as described below. Firstly, ~50 nm thick ZnO thin films, which acted as the seed layers for the subsequent growth of ZnO nanorod arrays, were deposited on 1.5 × 1.5 cm2 (100) heavily arsenic-doped Si substrates (n-type, with an electron concentration of ~5 × 1019cm-3) by reactive DC sputtering. Secondly, the above-mentioned Si substrates coated with the ZnO films were vertically hung in a mixed solution heated to 90°C. The mixed solution was formed by adding 890 mg of Zn(NO3)2 and 420 mg of diethylenetriamine into 100 mL of deionized water. In this way, vertically aligned ZnO nanorod arrays were grown on the silicon substrates. After a growth time of 2 h, the silicon substrates covered with ZnO nanorod arrays were taken out of the solution. Then, they were ultrasonically cleaned in deionized water for 3 min to remove the unwanted ZnO particles adhering on the ZnO nanorod arrays. In order to improve the crystallinity, the as-deposited ZnO nanorod arrays on the silicon substrates were annealed at 700°C for 2 h. Thirdly, ~90 nm thick SiO2 films were deposited onto the ZnO nanorod arrays by a sol-gel process consisting of the following steps: i) spin-coating of a precursor sol consisting of TEOS: EtOH: H2O = 1: 10: 4 (molar ratio) in which a properly small amount of HNO3 was added as the catalyzer, ii) soft-bake at 60°C for 20 min to remove the solvents in the gel, iii) annealing at 650°C for 2 h under O2 ambient. Finally, ~20 nm thick Au film on the SiO2 film and ~100 nm thick Au film on the backside of silicon substrate were successively sputtered as the electrodes. Either electrode was patterned into a circle with a diameter of ~10 mm.
The EL spectra for the devices under different DC voltages were recorded at RT using an Acton spectraPro 2500i spectrometer with a lowest spectrum resolution of 0.5 Å and an accuracy of ± 2 Å. For the acquisition of spectrum, the scanning step size was 1 Å. Moreover, the output power was measured using a Newport 1931-C power meter with an 818-UV/DB detector (~1cm in diameter). For the measurement, the devices were brought face to face with the detector. The devices were ~2 cm apart from the detector. For such a measurement configuration, it is roughly calculated that only ~2% of the output power of a device is detected by the above-mentioned power meter.
3. Results and discussion
Figures 2a and 2b show the plan- and cross-sectional scanning electron microscopy (SEM) images of the ZnO nanorod arrays annealed at 700°C for 2 h. The rods are quite uniform in height (~2 μm). They are randomly located on the substrate. The rod diameters range from ~40 to 80 nm. The hexagon end planes of the nanorods indicate that the ZnO nanorods grow along the <0001> direction. Figure 2c shows the photoluminescence spectra of the ZnO naonorod arrays before and after 700°C/2 h annealing, measured with a He-Cd laser (325 nm) as an excitation source. After the annealing, the near-band-edge (NBE) UV emission is enhanced, indicating that the crystal qualities of the ZnO nanorods are improved by the annealing. Moreover, the visible emission is a little blue-shifted due to the annealing. This might be related to the variation of surface states that play important roles in the nanostructures .
Figure 3a shows the current-voltage (I-V) characteristic of a typical device. Because the sol-gel derived SiO2 film is much poorer in insulation performance with respect to the thermal-oxidation-grown SiO2 film, there is a remarkable current through the device at a sufficiently high forward bias voltage. This is critical for the EL from the device. Herein, forward/reverse bias means that the gate electrode of Au is connected to positive/negative voltage. As can be seen, the device exhibits a rectifying behavior to a great extent. It is found that the MIS devices based on ZnO nanorod arrays are electroluminescent only under forward bias.
Figure 3b shows the evolution of the EL spectra for a device with the increase of forward bias voltage. At 4 V, the EL spectrum is greatly spoiled by the noises, illustrating as a single broad spontaneous emission peak centered at around 383 nm, which is ascribed to the NBE UV emission from ZnO. As the voltage is a little bit increased to 4.5 V, a few of discrete narrow peaks emerge in the spectrum. The linewidth of these peaks is less than 2 Å. When the voltage is further increased to 5 V and above, more sharp peaks appear in the spectra. Moreover, the EL intensity is significantly increased. The detected output power as a function of injection current is shown in Fig. 3c. Above a threshold current, as shown by a solid line plotted to guide the eyes, the output power increases linearly with the injection current. Such a linear dependence is due to the gain saturation that forms an intrinsic aspect of an amplifying system above threshold . The narrow linewidth and the rapid increase of emission intensity, as shown in Figs. 3b and 3c, indicate the occurrence of lasing from the device. Note Fig. 3b that the spacing between the adjacent sharp peaks is not uniform. Therefore, we believe that the laser spikes in the EL spectra are ascribed to the random lasing from ZnO nanorod array. Evidently, the multiple sharp peaks in the spectra between 360 and 400 nm, as shown in Fig. 3b, represent different lasing modes.
It should be stated that the device exhibits no essential visible EL in the visible region. Figure 4 shows the EL spectrum in the wavelength range of 360–800 nm for the device applied with a forward bias of 8 V. As can be seen, in the visible wavelength region of 450–800 nm there is no discernable peak except the band centered at ~760 nm, which belongs to the second harmonics of the UV emission. This is also the case for the device applied with other forward bias voltages. The reason for no substantial visible EL from the device is elucidated below.
As the MIS structure based on the ZnO-nanorod-array is applied with a forward bias voltage, an electron accumulation region is formed near the ZnO/SiO2 interface. In other words, the electrons in the conduction band of ZnO accumulate near the ZnO/SiO2 interface, where the electron concentration is significantly higher than that inside the ZnO nanorod. On the other hand, as will be mentioned later, the holes generated in ZnO nanorod are also primarily populated near the ZnO/SiO2 interface. The details about the transport of carriers will be elucidated therein. Therefore, the electron-hole recombination and therefore the EL primarily proceed in the region near the ZnO/SiOx interface. This case has the following features: (1) the inter-band recombination rate increases dramatically due to the significant increase of electron concentration as a result of carrier accumulation, leading to the enhanced UV emission; (2) the defect states existing in the prime EL region are scarce since the accumulation region is extremely thin, leading to the remarkable suppression of defect-related visible emissions; (3) the self-absorption of UV light by the defect states is depressed since the UV light is primarily generated in the accumulation region.
Figure 5 shows a series of EL spectra taken at three successive measurements for the device applied with a forward bias of 8 V. Herein, it took ~60 s for each cycle of measurement. In the three spectra, the number and height of the sharp peaks are quite different. Moreover, the wavelengths corresponding to the sharp peaks change randomly. Such features of the emission spectra illustrate the intrinsic aspects of random lasing. It is reasonably believed that the electrically pumped random lasing from the ZnO nanorod array exhibits a different spectrum every moment. Definitely, in the random laser the light is multiply scattered and amplified by stimulated emission. For certain modes of the light, they achieve optical gain larger than the losses. As a consequence, the narrow emission spikes emerge in the emitted spectrum in the case of specific measurement configuration . In the device, the light emitted every moment is subsequently scattered with random walks. Therefore, the modes which can achieve optical gain larger than the losses are varying every moment in terms of wavelength and intensity. The scenario as described above can be believed to be vividly embodied in Fig. 5. Actually, in the case of optically pumped random laser action, it has been found that the narrow emission spikes in the emitted spectrum can change frequency randomly from one excitation pulse to another. This chaotic behavior is ascribed to the fact that the random lasing starts from spontaneous emission which is different at each shot .
So far, the electrically pumped random lasing from the ZnO nanorod arrays has been well demonstrated. In the following, the related mechanism is qualitatively elucidated, placing emphasis on how stimulated emission and multiple scattering occur in the Au/SiO2/ZnO-nanorod-array MIS device.
The device presented in this Express is actually an array of numerous MIS structures each of which is based on an individual ZnO nanorod. Figure 6a shows the schematic energy band diagram for a MIS structure based on an individual ZnO naonorod under forward bias. Herein, for the sake of simplicity, we tentatively neglect the effects of charges and defect sates in SiO2 film as well as the interface states on the band structure of the MIS structure. Under forward bias, the energy band of ZnO bends downward adjacent to the ZnO-nanorod/SiO2 interface, where electrons coming from the silicon substrate accumulate, with a concentration much higher than that in the bulk of ZnO nanorod. Regarding the injection of holes into the ZnO nanorod, despite being not essentially understood, we propose a possible pathway as follows. The sol-gel derived SiO2 film is considerably imperfect. Therefore, there are numerous carrier traps within the film. Under a substantial forward bias, an amount of electrons in the valence band of ZnO can be driven into the SiO2 film by the electric field. They are then captured by the traps in the SiO2 film. Thus, an equivalent amount of holes is generated in the valence band of ZnO. By the way, it should be stated that the electrons in the conduction band of ZnO can also be captured by certain traps in the SiO2 film. Such a trapping process contributes partly to the transport of electrons in the SiO2 film. The detailed mechanism for the transport of electrons in the SiO2 film can be referred to our previous report .
Understandably, both electron and hole concentrations increase with forward bias. Above a critical forward bias, the electron concentration in the band-downward region adjacent to the ZnO-nanorod/SiO2 interface is considerably high so that the quasi Fermi level of electron (EFn) enters into the conduction band. On the other hand, due to the considerably small hole mobility in ZnO, most of the injected holes in the ZnO nanorod are also populated in the band-downward region nearby the ZnO-nanorod/SiO2 interface, as shown in Fig. 6a. Under a sufficiently high forward bias, the hole concentration in the band-downward region is high enough to enable the quasi Fermi level of hole (EFp) to be close to the edge of valence band (case 1) and even to enter into the valence band (case 2). The above-mentioned scenario can be illustrated in terms of density of states diagram for the band-downward region, as schematically shown in Fig. 6b. Accordingly, with a high carrier injection level under forward bias larger than a critical voltage, the condition for population inversion in ZnO, that is, EFn - EFp > Eg (band gap), is satisfied. Consequently, the stimulated emission can occur in the device. Then, the spontaneous emission can be amplified.
It should be stated that the stimulated emission only occurs in the band-downward region because the carrier concentrations are too low elsewhere. Therefore, the light emission in the device is primarily localized in a quasi two-dimensional (2D) region close to the ZnO-nanorod-array/SiO2 interface. Moreover, the ZnO nanorod array acts as a 2D scattering system, as light is scattered by the nanorods in the plane perpendicular to the rods . In our device, the ZnO nanorods and the SiO2 films penetrating into the interspaces within the ZnO nonorod array, form a disorder network in the plane perpendicular to the nanorods. There is a refractive index mismatch between ZnO and SiO2, which facilitates the light scattering. Figure 6c shows a quasi-2D type of light transport in the region close to the nanorods/SiO2 interface, where multiple light scattering takes place. Although the picture of closed-loop random cavity can be well used to account for the random lasing , it is argued that an unrealistically high gain will be required to achieve the lasing threshold condition in such a loop since most of the energy is scattered out of the loop in each scattering act . Nevertheless, it has been theoretically derived that random resonators that correspond to certain arrangement of scatterers can form in a quasi 2D system with the random variations of nr . According to this viewpoint, we believe that the random resonators should form in our device where the random variations of nr are induced by the random distribution of the ZnO nanorods and SiO2 films. In this context, once the optical gain is larger than the losses when the device is applied with a sufficiently high forward bias, the device will lase thus leading to spectrally narrow emission.
Moreover, according to the mechanism presented in Ref. 22, we can alternatively interpret the electrically pumped random lasing from our devices. As mentioned above, the region adjacent to the ZnO-nanorod-array/SiO2 interface becomes a quasi 2D amplifying disordered system as the device is sufficiently forward-biased. Each individual sharp peak in Fig. 2b corresponds to a single spontaneous emission event that fortunately propagates with a very long light path due to the multiple scattering and picks up a gain larger than the losses through the stimulated emission. In other words, a number of spontaneously emitted photons experience extraordinarily multiple scattering to achieve gain that is larger than the losses, leading to laser spikes in the emitted spectrum.
In summary, we have demonstrated the electrically pumped random lasing from the ZnO nanorod arrays which acts as the component of semiconductor in the MIS devices in which Au and SiO2 are the metal gate and insulator layer, respectively. When the MIS devices are applied with a sufficiently high forward bias, in the region adjacent to the ZnO nanorods/SiO2 interface, the stimulated emission that provides optical gain occurs. In the mean time, the quasi 2D type of random light transport proceeds via multiple scattering facilitated by the random variations of nr, which are induced by the randomly distributed ZnO nanorods and SiO2 films. Therefore, the devices can be electrically pumped to achieve optical gain through stimulated emission together with multiple scattering. Above a threshold voltage/current, the devices exhibit random laser actions featuring a series of narrow spikes in the emitted spectra.
We thank the financial supports from Natural Science Foundation of China (No. 60776045), “973 Program” (No.2007CB613403), and Changjiang Scholars and Innovation Teams in Universities.
References and links
1. Ü. Özgür, Y. I. Alivov, C. Liu, A. Teke, M. A. Reshchikov, S. Doğan, V. Avrutin, S.-J. Cho, and H. Morkoç, “A comprehensive review of ZnO materials and devices,” J. Appl. Phys. 98(4), 041301 (2005). [CrossRef]
2. W. Wegscheider, L. N. Pfeiffer, M. M. Dignam, A. Pinczuk, K. W. West, S. L. McCall, and R. Hull, “Lasing from excitons in quantum wires,” Phys. Rev. Lett. 71(24), 4071–4074 (1993). [CrossRef] [PubMed]
3. Z. K. Tang, M. Kawasaki, A. Ohtomo, H. Koinuma, and Y. Segawa, “Self-assembled ZnO nano-crystals and exciton lasing at room temperature,” J. Cryst. Growth 287(1), 169–179 (2006). [CrossRef]
4. M. H. Huang, S. Mao, H. Feick, H. Q. Yan, Y. Y. Wu, H. Kind, E. Weber, R. Russo, and P. D. Yang, “Room-temperature ultraviolet nanowire nanolasers,” Science 292(5523), 1897–1899 (2001). [CrossRef] [PubMed]
5. J. C. Johnson, K. P. Knutsen, H. Q. Yan, M. Law, Y. F. Zhang, P. D. Yang, and R. J. Saykally, “Ultrafast carrier dynamics in single ZnO naonowire and nanoribbon lasers,” Nano Lett. 4(2), 197–204 (2004). [CrossRef]
6. S. F. Yu, S. P. Lau, W. I. Park, and G.-C. Yi, “Random laser action in ZnO nanorod arrays embedded in ZnO epilayers,” Appl. Phys. Lett. 84(17), 3241–3243 (2004). [CrossRef]
7. S. P. Lau, H. Y. Yang, S. F. Yu, H. D. Li, M. Tanemura, T. Okita, H. Hatano, and H. H. Hng, “Laser action in ZnO nanoneedles selectively grown on silicon and plastic substrates,” Appl. Phys. Lett. 87(1), 013104 (2005). [CrossRef]
8. S. P. Lau, H. Y. Yang, S. F. Yu, C. Yuen, E. S. P. Leong, H. D. Li, and H. H. Hng, “Flexible ultraviolet random lasers based on nanoparticles,” Small 1(10), 956–959 (2005).
9. B. S. Zou, R. B. Liu, F. F. Wang, A. L. Pan, L. Cao, and Z. L. Wang, “Lasing mechanism of ZnO nanowires/nanobelts at room temperature,” J. Phys. Chem. B 110(26), 12865–12873 (2006). [CrossRef] [PubMed]
10. J. Z. Liu, S. Lee, Y. H. Ahn, J. Y. Park, K. H. Koh, and K. H. Park, “Identification of dispersion-dependent hexagonal cavity modes of an individual ZnO nanonail,” Appl. Phys. Lett. 92(26), 263102 (2008). [CrossRef]
11. C. Czekalla, C. Sturm, R. Schmidt-Grund, B. Q. Cao, M. Lorenz, and M. Grundmann, “Whispering gallery mode lasing in Zinc oxide microwires,” Appl. Phys. Lett. 92(24), 241102 (2008). [CrossRef]
12. D. Wang, H. W. Seo, C.-C. Tin, M. J. Bozack, J. R. Willams, M. Park, and Y. Tzeng, “Lasing in whispering gallery mode in ZnO nanonails,” J. Appl. Phys. 99(9), 093112 (2006). [CrossRef]
13. X. Ma, P. Chen, D. Li, Y. Zhang, and D. Yang, “Electrically pumped ZnO film ultraviolet random lasers on silicon substrate,” Appl. Phys. Lett. 91(25), 251109 (2007). [CrossRef]
14. P. Chen, X. Ma, D. Li, Y. Zhang, and D. Yang, “Electrically pumped ultraviolet random lasing from ZnO-based metal-insulator-semiconductor devices: dependence on carrier transport,” Opt. Express 17(6), 4712–4717 (2009). [CrossRef] [PubMed]
15. S. Gottardo, S. Cavalieri, O. Yaroshchuk, and D. S. Wiersma, “Quasi-two-dimensional diffusive random laser action,” Phys. Rev. Lett. 93(26), 263901 (2004). [CrossRef]
16. A. B. Djurisic and Y. H. Leung, “Optical properties of ZnO nanostructures,” Small 2(8–9), 944–961 (2006).
17. D. S. Wiersma, “The physics and applications of random lasers,” Nat. Phys. 4(5), 359–367 (2008). [CrossRef]
18. S. Mujumdar, V. Türck, R. Torre, and D. S. Wiersma, “Chaotic behavior of a random laser with static disorder,” Phys. Rev. A 76(3), 033807 (2007). [CrossRef]
19. H. Cao, “Review on latest developments in random lasers with coherent feedback,” J. Phys. Math. Gen. 38(49), 10497–10535 (2005). [CrossRef]
20. H. Cao, Y. G. Zhao, S. T. Ho, E. W. Seelig, Q. H. Wang, and R. P. H. Chang, “Random laser action in semiconductor powder,” Phys. Rev. Lett. 82(11), 2278–2281 (1999). [CrossRef]