Enhanced near-band-edge emission from a-plane ZnO thin films on SrTiO

R A P I D C O M M U N I C A T I O N S

Enhanced near-band-edge emission from a-plane ZnO thin films on SrTiO

substrates

Guangheng Wu^1•Xiang Li^1• Meifeng Liu^1•Zhibo. Yan^1• Jun-Ming Liu^1,2

Received: 31 March 2015 / Accepted: 7 July 2015 / Published online: 7 August 2015 ÓSpringer-Verlag Berlin Heidelberg 2015

Abstract Epitaxial ZnO with controlled orientations is highly concerned, and we report epitaxial growth of (0001)- and (11–20)-oriented ZnO films, respectively, on (111)- and (100)-SrTiO₃ (STO) substrates using pulsed laser deposition. The epitaxial growth of (11–20)-ZnO films on (100)-STO, with in-plane orientation relationship of ZnO [0001]//STO [110] and ZnO [10–10]//STO [1–10], is realized. It is revealed that the (11–20)-ZnO films have much stronger photo-luminescence (PL) intensity than those of the (0001)-oriented ones, while the electric con- ductance remains comparable with the latter. The electrical and PL properties of the (11–20)-ZnO films are correlated with the in-plane polarity.

Zinc oxide (ZnO) is one of the most extensively investi- gated semiconductors [1–3]. In the ambient conditions, it crystallizes in wurtzite structure, which has a hexagonal unit cell with theP6₃mcspace group and lattice parameters a=0.32475 nm and c=0.52042 nm. Widespread inter- est in ZnO is fueled by its prospects in optoelectronics applications [4, 5]. It is known that ZnO and GaN both have wide band-gap (E_g) of *3.3 eV at 300 K. However, ZnO shows some advantages over GaN, such as the availability of high-quality ZnO single crystals and thin

films, resulting in a potentially low cost for processing relevant devices. The most attractive advantage of ZnO over GaN is the high exciton binding energy which is 25 meV for GaN but 60 meV for ZnO, while the room- temperature (RT) thermal energy is *25 meV [6]. This allows the specific preference of ZnO for intense near- band-edge exciton emission at higher temperatures.

Another well-recognized application associated with ZnO is transparent thin-film transistor [7], where the pro- tective covering preventing light exposure is eliminated since ZnO-based transistors are insensitive to visible light.

More importantly, a carrier density up to 2910²¹ cm^-3 can be reached in ZnO by heavy substitution [8], allowing the electrical properties to be modulated in a broad regime, while its optical transparency remains robust, making it useful for transparent electrodes [9,10]. These advantages even promise ZnO as a material for spintronics since good magnetic properties can be induced by magnetic substitu- tions [11].

Along this line, high-quality ZnO epitaxial films with different orientations are highly appealed for developing ZnO-based devices. Successful preparations of (0001)- oriented ZnO (c-ZnO) films on various substrates including sapphire have been well reported [12, 13]. However, defects are easily generated in epitaxial ZnO films on the sapphire due to the large lattice mismatch of 18 % [14], including misfit dislocations [15,16] and twin domains [14, 17,18]. On the other hand, increasing attentions are paid to obtain in-plane polar epitaxial ZnO films, such as (10–10)- oriented (m-plane) and (11–20)-oriented (a-plane) ones [19]. It is known that ZnO has its intrinsic electrical polarity along thec-axis in the wurtzite structure, and thus, the c-ZnO thin films inevitably suffer the quantum-con- fined Stark effect (QCSE), which reduces the exciton binding energy and limits the optical performance.

& Guangheng Wu

[email protected]

& Jun-Ming Liu

[email protected]

1 Laboratory of Solid State Microstructures, Nanjing University, Nanjing 210093, China

2 Institute for Advanced Materials, South China Normal University, Guangzhou 510006, China

DOI 10.1007/s00339-015-9342-6

Epitaxial growth of them-plane anda-plane ZnO (a-ZnO) with the in-plane polarity provides a solution to avoid this problem. Unfortunately, the growth of in-plane polar ZnO thin films often encounters issues such as high density of domain defects. Them-ZnO thin films often accommodate (0002)-, (10–13)-, and (11–22)-oriented domains embed- ded in the matrix [20–22], which is detrimental to the electrical and optical performance. The basal-plane stack- ing faults (BSF) of high density, typically[1910⁶cm^-1, are another type of major defects acting as nonirradiative recombination centers responsible for poor internal quan- tum efficiency (IQE) [25].

In this work, we intend to grow the epitaxiala-ZnO thin films on SrTiO₃(STO) substrates. For comparison, we also prepare thec-ZnO thin films on STO substrates. Different from other commonly used substrates, STO has perovskite structure and is usually chosen as substrate for a wide class of functional oxide films [24,25]. This fact allows a pos- sibility to integrate wurtzite ZnO with other perovskite materials. In the first step, ZnO/STO heterostructures may be also useful for alternative multifunctional devices due to the carrier limitation and coupling effect. Previous effort was made to deposit epitaxial ZnO thin films on STO substrates with various orientations [26]; however, open lectures on the optical and electrical properties for in-plane polar ZnO films are still lacking. These motivate us to search for a-ZnO on STO substrates with high performance.

In order to match the in-plane lattice symmetry between ZnO thin films and STO substrates, (100)- and (111)-STO were chosen for depositing the a- and c-ZnO thin films, respectively. For the a-ZnO, the in-plane lattices can be treated as a rectangle block with parameters of 0.52042 nm (cZnO) and 0.5624 nm (H3aZnO), matching to the extended square in-plane lattice of (100)-STO with parameter of 0.5523 nm (H2a_STO). On the other hand, for thec-ZnO, the hexagon in-plane lattice can be treated as a triangle lattice with parameter of 0.5624 nm (H3aZnO), matching to the triangle lattice of (111)-STO with a parameter of 0.5523 nm (H2a_STO).

Experimentally, ZnO thin films were deposited by pulsed laser deposition (PLD), using homemade ZnO ceramic targets. The deposition was carried out under oxygen of 15 Pa. The target-to-substrate distance was 60 mm. A KrF excimer laser (248 nm in wavelength) with fluency energy of 300 mJ and repetition rate of 2 Hz was used to ablate the ZnO target. A ZnO film of*200 nm in thickness can be obtained using 3600 pulses. After the deposition, the films were in situ annealed in the oxygen ambient of 10³Pa for 30 min to improve the crystallization of the thin films.

The cyrstallinity and epitaxy of the as-prepared ZnO thin films were characterized by X-ray diffraction (XRD)

h -2handUscans using the Cu-K_aradiation. For theU scan, the scanning angles of ZnO films were respectively corrected by the angles of STO. For the PL measurements, the ZnO films were excited by 250-nm photons from a xenon lamp with a grating selector and the PL spectra were obtained using the Shimadzu RF-5301PC spectrofluo- rophotometer. The carrier concentration and mobility were measured with the standard Hall method, and metal In was used as electrodes for the ohmic contact.

Figure1a shows the XRD h-2h spectra of two ZnO films on (111)-STO substrates obtained at two different deposition temperatures, T_s=400 and 600°C, labeled withc-ZnO-H andc-ZnO-L, indicating the preferredc-axis orientation and no substantial change of this spectrum upon differentT_s. However, the diffraction peaks fromc-ZnO-H has smaller full width at half maximum (FWHM) than that from c-ZnO-L, suggesting that c-ZnO-H is better crystal- lized. Therefore, further characterizations are performed on sample c-ZnO-H (c-ZnO).

Here, the {11–22} reflection family of ZnO is chosen for theUscan, while the {110} reflections of STO are chosen for the reference. As expected, the {110} reflections of (111)-STO show the threefold symmetry. Thec-ZnO thin film exhibits the sixfold symmetry. In details, the epitaxy relationship is described by ZnO [11–20]//STO [1–10] and ZnO [11–20]//STO [2-1-1]. The lattice mismatch between ZnO and STO in this epitaxy mode is only*1.9 %, much smaller than that between c-ZnO and sapphire, and the lattice match geometry is schematically shown in Fig.1c.

We are more interested in the epitaxial growth ofa-ZnO films. The crystallinity and orientation at differentTswere investigated, and remarkable dependence on T_s was observed. Figure2a shows two typical XRDh-2h pat- terns of the ZnO thin films at Ts=400 and 600 °C, respectively. Surprisingly, the thin film deposited at T_s =600 °C does show perfect (11–20)-orientation, namely thea-plane orientation, although the film deposited at T_s=400 °C remains, however, the (0001)-oriented.

Thisc-plane thin film is not epitaxial with the (100)-STO.

Indeed, one cannot find applicable lattice match between any low-index ZnO plane and (100)-STO. The c-plane orientation is due to the very low potential for the (0001) plane growth [27]. Therefore, we pay attentions to the a- plane thin films. First, we figure out the epitaxy of the thin films by performing the XRD Uscans. The {122} reflec- tions of STO, the {11–22} and {10–10} reflection families of thec- anda-ZnO thin films were respectively chosen for theUscans. The data for thea-ZnO are plotted in Fig.2b, where clearly the fourfold symmetry is identified, indicat- ing the epitaxial growth of the a-ZnO thin film on (100)- STO. The epitaxy relationship is described by ZnO [0001]//

STO [110] and ZnO [10–10]//STO [1–10] with a lattice mismatch as low as -5.8 % (minus represents tensile

stress) and 1.7 %, respectively. The lattice match on the interface is schematically drawn in Fig.2c.

We address the PL behaviors of thea-ZnO thin films.

For comparison, the PL spectra for thec-ZnO thin films are also presented. Figure3a shows the PL spectra from the two samples, given the identical pump intensity, and the inset shows the PL image from thea-ZnO. A typical PL spectrum for ZnO consists two parts: an intensity peak located around 3.2–3.6 eV and a broad brand around

*2.5 eV. The former is the near-band-edge (NBE) emis- sion attributed to the recombination of excitons through an exciton–exciton collision process, and the latter is attrib- uted to deep level defects, possibly be summed up to be zinc interstitials, oxygen vacancies, zinc vacancies, oxygen interstitials, or donor–acceptor transitions between defects complex. The intensity of NBE emission from thea-ZnO film is at least five times as that from the c-ZnO film, different from previous studies on in-plane polar ZnO films

on sapphire; for instance, Shi et al. [28] fabricated the c- ZnO, a-ZnO, and m-ZnO on c-Al₂O₃, r-Al₂O₃, and m- Al₂O₃, respectively, and found weaker NBE emissions from thea-ZnO andm-ZnO than that from thec-ZnO. On the other hand, for well-prepared GaN, the nonpolar ori- ented films and quantum wells usually benefit to enhanced PL intensities [29].

The differences of PL intensity between a-ZnO and c- ZnO thin films are affected by several factors, such as the polarity of the thin film, as well as the microstructures.

However, the effect of microstructure is ruled out. From Fig.3a, the broad band around 2.5 eV from both a-ZnO andc-ZnO thin films is nearly the same. This PL band is caused by defect in the thin films, suggesting that the defect densities ina-ZnO andc-ZnO thin films are similar to each other. This is confirmed with TEM and XPS experiments (not shown). Therefore, we can conclude that the a-ZnO andc-ZnO thin films are of same structural quality, and the Fig. 1 (color online)aXRD

h-2hpatterns for thec-ZnO- H andc-ZnO-L films on (111)- STO deposited atT_s=400 and 600°C,bcorresponding XRD uscan patterns of thec-ZnO-H and substrate andcin-plane atomic arrangement for the ZnO film on STO.Red open circles represent the ZnO lattice sites, andblue solid circlesrepresent the STO ones

Fig. 2 (color online)aXRD h-2hpatterns for the ZnO films on (100)-STO atT_s=400 and 600°C. Only the ZnO deposited atT_s=600°C is grown epitaxially, and the corresponding,bXRDuscan pattern andcin-plane atomic arrangement.Red open circles represent ZnO lattice sites, and blue solid circlesrepresent the STO ones

differences of PL intensity betweena-ZnO andc-ZnO thin films are mainly resulted from the polarity of the thin films.

The polarity dependence of the PL behavior is associ- ated with the band alignment across the surface and/or interface. Usually, the energy band bending occurs at the surfaces and interfaces, causing an electric fieldEperpen- dicular to the surface and interfaces as shown in Fig.3b.

When the film is illuminated to a light beam with photon energy larger than the optical band-gap, electrons in the valence band transmit to the conduction band, generating holes in the valence band. The photon-induced electrons and holes move oppositely from each other, preventing their recombination. As a result, the light emitting is sup- pressed. It becomes worse when a polar plane lines in the surface and/or interface, because the interruption of the polarizations may strengthen the band bending across the surfaces and/or interfaces. This is why thec-ZnO film has lower PL intensities than that of thea-ZnO film, observed in the present study.

Although the electrical transport in nonpolar wurtzite semiconductors was reported elsewhere [30], most of these reports focus on the junctions perpendicular to the surfaces and interfaces. Here, we perform the Hall measurement at room temperature to estimate the carrier concentration and

mobility in the a- and c-ZnO films, and the data are pre- sented in Fig.4. The carrier concentration is*5.6910¹⁹ for the a-ZnO film and 5.8910¹⁹ for the c-ZnO film, consistent with the reported value for undoped ZnO. The carrier mobility for thea-ZnO film is*22 cm³/Vs, lower than that from thec-ZnO (42 cm³/Vs).

We can calculate the conductance r with r =nle, where n is the carrier concentration and l is the carrier mobility. The calculated conductance for thec-ZnO is two times that for thea-ZnO. As mentioned above, thea-ZnO andc-ZnO thin films are of similar quality and the differ- ence of electron conduction is mainly caused by the polarity of the thin films. The in-plane conductance for an in-plane polar film is usually smaller than that from out-of- plane polar one. This is attributed to the scattering from the polar grain boundaries [31].

The c-ZnO and a-ZnO epitaxial films on (111)- and (100)-STO have been prepared. For the c-ZnO films, the epitaxy relationship is described by ZnO [10–10]//STO [1–10] and ZnO [11–20]//STO [-211], while the epitaxy relationship for the a-ZnO films is described with ZnO [0001]//STO [110] and ZnO [10–10]//STO [1–10]. Inter- estingly, the a-ZnO films show strong PL intensity and remain high electrical conductance. The electrical and PL properties of the as-prepared ZnO films are correlated with the polarity of ZnO, suggesting the potentials of the (11–20)-ZnO films in optoelectronic devices.

Acknowledgments This work was supported by the National 973 Projects of China (Grant No. 2011CB922101), the Natural Science Foundation of China (Grants No. 11234005 and No. 11374147, 50332006), and the Priority Academic Program Development of Jiangsu Higher Education Institutions, China.

References

1. H. Park, S. Chang, J. Jean, J.J. Cheng, P.T. Araujo, M.S. Wang, M.G. Bawendi, M.S. Dresselhaus, M.V. Bulovic, J. Kong, S.

Gradecak, Nano Lett.13, 233 (2013)

2. L.Y. Lim, S. Lany, Y.J. Chang, E. Rotenberg, A. Zunger, M.F.

Toney, Phys. Rev. B86, 235113 (2012)

3. K. Noori, F. Giustino, Adv. Funct. Mater.22, 5089 (2012) Fig. 3 (color online)aPL emission spectra for thea-ZnO andc-ZnO

films, and the insert is a real color image for thea-ZnO film excited with 250-nm light.bThe negative effect of energy barrier at surface/

interface on PL. When the film is illuminated to a light beam with photon energyhtlarger than theE_g, electrons (red open circle) in the valence band (VB) transmit to the conduction band (CB), generating holes (blue solid circle) in the VB (process i). The photon-induced electrons and holes move oppositely from each other (process ii and iii), preventing their recombination (process iv)

Fig. 4 (color online) Carrier concentrationnand mobilitylfor the a-ZnO andc-ZnO films

4. H. Long, S.Z. Li, X.M. Mo, H.N. Wang, H.H. Huang, Z. Chen, Y.P. Liu, G.J. Fang, Appl. Phys. Lett.103, 123504 (2013) 5. K.H. Baik, H. Kim, J. Kim, S. Jung, S. Jang, Appl. Phys. Lett.

103, 091107 (2013)

6. U. Ozgur, D. Hofstetter, H. Morkoc, Proc. IEEE98, 1255 (2010) 7. S.H.K. Park, C.S. Hwang, M. Ryu, S. Yang, C. Byun, J. Shin, J.I.

Lee, K. Lee, M.S. Oh, S. Im, Adv. Mater.21, 678 (2009) 8. U¨ . O¨zgu¨r, Y.I. Alivov, C. Liu, A. Teke, M.A. Reshchikov, S.

Dog˘an, V. Avrutin, S.J. Cho, H. Morkoc¸, J. Appl. Phys. 98, 041301 (2005)

9. H. Frenzel, A. Lajn, H. von Wenckstern, M. Lorenz, F. Schein, Z.P. Zhang, M. Grundmann, Adv. Mater.22, 5332 (2010) 10. Q.F. Zhang, C.S. Dandeneau, X.Y. Zhou, G.Z. Cao, Adv. Mater.

21, 4087 (2009)

11. K.M. Whitaker, M. Raskin, G. Kiliani, K. Beha, S.T. Ochsenbein, N. Janssen, M. Fonin, U. Rudiger, A. Leitenstorfer, A.D.R.

Gamelin, R. Bratschitsch, Nano Lett.11, 3355 (2011)

12. F. Vigue´, P. Venne´gue`s, C. Deparis, S. Ve´zian, M. Lau¨gt, J.P.

Faurie, J. Appl. Phys.90, 5115 (2001)

13. C.C. Kuo, W.R. Liu, B.H. Lin, W.F. Hsieh, C.H. Hsu, W.C. Lee, M. Hong, J. Kwo, Opt. Express21, 1857 (2013)

14. I. Ohkubo, A. Ohtomo, T. Ohnishi, Y. Matsumoto, H. Koinuma, M. Kawasaki, Surf. Sci.443, L1043 (1999)

15. W.Y. Shiao, C.Y. Chi, S.C. Chin, C.F. Huang, T.Y. Tang, Y.C.

Lu, Y.L. Lin, L. Hong, F.Y. Jen, C.C. Yang, B.P. Zhang, Y.

Segawa, J. Appl. Phys.99, 054301 (2006)

16. J. Narayan, K. Dovidenko, A.K. Sharma, S. Oktyabrsky, J. Appl.

Phys.84, 2597 (1998)

17. I.W. Kim, Y.B. Kwon, J.M. Yi, J.H. Je, G. Nouet, T. Wojtowicz, P. Ruterana, J. Kioseoglou, J. Vac. Sci. Technol. A 22, 2159 (2004)

18. C.S. Steplecarua, M.S. Martı´n-Gonza´leza, J.F. Ferna´ndezb, J.L.

Costa-Kra¨mer, Thin Solid Films518, 4630 (2010)

19. B.H. Lin, W.R. Liu, C.Y. Lin, S.T. Hsu, S. Yang, C.C. Kuo, C.H.

Hsu, W.F. Hsieh, F.S.S. Chien, C.S. Chang, ACS Appl. Mater.

Interfaces4, 5333 (2012)

20. K. Iwata, H. Asahi, K. Asami, R. Kuroiwa, S. Gonda, Jpn.

J. Appl. Phys.36, L661 (1997)

21. J.W. Lee, J.H. Kim, S.K. Han, S.K. Hong, J.Y. Lee, S.I. Hong, T.T. Yao, J. Cryst. Growth312, 238 (2010)

22. T.J. Baker, B.A. Haskell, F. Wu, J.S. Speck, S. Nakamura, Jpn.

J. Appl. Phys.45, L154 (2006)

23. Z. Vashaei, C. Bayram, P. Lavenus, M. Razeghi, Appl. Phys.

Lett.97, 121918 (2010)

24. T. Onuma, H. Amaike, M. Kubota, K. Okamoto, H. Ohta, J.

Ichihara, H. Takasu, S.F. Chichibu, Appl. Phys. Lett.91, 181903 (2007)

25. J.H. Noh, J.S. An, J.Y. Kim, C.M. Cho, K.S. Hong, H.S. Jung, J.

Eelctroceram.23, 497 (2009)

26. X.H. Wei, Y.R. Li, J. Zhu, W. Huang, Y. Zhang, W.B. Luo, H. Ji, Appl. Phys. Lett.90, 151918 (2007)

27. L. Lahourcade, E. Bellet-Amalric, E. Monroy, M. Abouzaid, P.

Ruterana, Appl. Phys. Lett.90, 131909 (2007)

28. K. Shi, A.L. Yang, J. Wang, H.P. Song, X.Q. Xu, L. Sang, H.Y.

Wei, S.Y. Yang, X.L. Liu, Q.S. Zhu, Z.G. Wang, J. Cryst. Growth 314, 39 (2011)

29. H. Fujisawa, H. Nonomura, M. Shimizu, H. Niu, J. Crystal.

Growth459, 237 (2002)

30. M.K. Rajpalke, T.N. Bhat, B. Roul, M. Kumar, S.B. Krupanidhi, J. Appl. Phys.112, 023706 (2012)

31. J.J. Zhu, T. Aaltonen, V. Venkatachalapathy, A. Galeckas, A.Y.

Kuznetsov, J. Cryst. Growth310, 5020 (2008)