Physics Procedia 28 ( 2012 ) 28 – 32

1875-3892 © 2012 Published by Elsevier B.V. Selection and/or peer-review under responsibility of Universidade Federal de Juiz de Fora, Brazil.

doi: 10.1016/j.phpro.2012.03.665

15th Brazilian Workshop on Semiconductor Physics

Dark field optical imaging and photoluminescence spectra from ZnO microstructures obtained by spray pyrolysis

T. G. Silva1, E. Ribeiro, E. Silveira

Departamento de Física, Universidade Federal do Paraná, Caixa postal 19044, 81531-980 Curitiba – PR, Brazil 81

Abstract

In this work we present results on ZnO self-assembled microstructures obtained by modified ultrasonic spray pyrolysis technique. We show optical imaging and photoluminescence spectrum from a typical self-assembled ZnO microstructure grown on Si(100) substrate. The dendritic formation of ZnO clusters achieved with our technique allows us to study both structural defects and heat routes in ZnO films. After thermal treatment, enhanced green and red emissions were observed on PL spectrum of the ZnO microstructures. The energy band gap from a typical structure was 3.13 eV, lower than usual for thick ZnO film, and a clear dependence with substrate surface roughness was also observed.

© 2011 Published by Elsevier B.V.

Keywords: ZnO; microstructures; photoluminescence; spray pyrolysis; optical properties; self-assembled; dark field; clusters; red emission;green band.

1. Introduction

Zinc oxide is a well-known direct wide band gap semiconductor (3.37 eV at room temperature) with large exciton binding energy of 60 meV. It belongs to the group II-VI semiconductor materials and crystallizes in the hexagonal wurtzite structure. This type of structure presents a spontaneous electric polarization in the growth

1 Corresponding author. Tel.:+55-41-3361-3096; fax: +55-41-3361-3418 E-mail address: tgs04@fisica.ufpr.br.

Available online at www.sciencedirect.com

© 2012 Published by Elsevier B.V. Selection and/or peer-review under responsibility of Universidade Federal de Juiz de Fora, Brazil. Open access under CC BY-NC-ND license.

Open access under CC BY-NC-ND license.

T. G. Silva et al. / Physics Procedia 28 ( 2012 ) 28 – 32 29

direction, which allows both piezoelectric and pyroelectric properties to be present in ZnO. This multifunctionalbehavior makes ZnO a promising material for applications in optoelectronics and UV based devices [1], as well asgas sensors [2], surface acoustic waves (SAW) [3], transparent coatings and so forth.

Due its high optical transmittance in the visible range and low resistivity, ZnO has been widely studied to be usedas transparent conducting oxide (TCO) in solar cells [4]. The development of ZnO light-emitting devices has a disadvantage that the p-type ZnO is still difficult to obtain, although recent advances have been made by dopingZnO with nitrogen [5-10].

The origin and control of the green emission band and the improvement of the ZnO UV photoresponse areassumed to be the greatest challenge in the production of high quality ZnO films. Many researchers [11-15] havereported that the green band emission on photoluminescence spectra is related to oxygen vacancies or doping, whichare demonstrated to increase the carrier concentration. Dislocations, interstitial Zn, or Zn vacancies has also beenassumed to be usually intrinsic defects, which lowers the UV photoresponse signal [12].

In this work we demonstrate a different way to observe intrinsical defects in ZnO structures. For this, we obtained self-assembled ZnO clusters, which were grown on Si(100) substrates and annealed in a vacuumatmosphere. Optical images from the dendrictic ZnO formation and a photoluminescence spectrum from one typicalmicrostructure are showed, indicating that structural damages induced by thermal treatment enhances the green andred emission bands.

2. Experimental procedure

In order to obtain ZnO microstructures we used as precursor a solution of dihydrate zinc acetate(Zn(CH3COO)2·2H2O) diluted in ultrapure water at concentration of 0.05 M. The solution droplets were

generated by ultrasonic cavitations of a quartz crystal and fells on the silicon (100) substrate at room temperature.The deposition time was fixed to 1 hour. As result of this deposition, we obtained on the substrate several droplets ofzinc acetate solution, deposited randomly and with different sizes. After this, it was put inside a tubular oven pre-heated at 450ºC and the thermal treatment lasted for 1 hour in air atmosphere. Following that, we performed an annealing at 500ºC in vacuum. This procedure was made to create cracks and to increase the oxygen vacancies.

Before the deposition, the Si(100) substrates were merged on isopropyl alcohol and stirred in ultrapure water. Insome of our substrates we preserve the natural layer of SiO2, which let the substrate surface rough, whereas in the other we merged the silicon substrate for approximately 15 seconds in a hydrofluoric acid solution (HF 40%) to remove the SiO2 natural layer. This procedure let the silicon surface flat and hydrophobic. These two different waysof silicon cleaning were used to produce the microstructures.

Optical imaging was used to observe the roughness and formation of these structures. For this, we use anOLYMPUS BX51optical microscope coupled with an OLYMPUS UPMTVC camera. Dark-field images wereperformed to improve the contrast between the microstructures and the substrate.

Room temperature photoluminescence spectrum was obtained using as excitation source the 325 nm line from a He-Cd laser and the collected light was analyzed by a 0.5 m spectrometer with a GaAs photomultiplier.

3. Results and Discussion

In order to produce microstructures the use of spray pyrolysis is mandatory. Alternative tests using castingled to excessive large droplets and the final result was not satisfactory. The developing of our spray pyrolysistechnique allowed us to manipulate the precursor solution at a more controlled fashion, bringing to light interestingmorphologies of ZnO, as showed in Figure 1. There, optical image of the microstructures were obtained after thefirst thermal treatment at 450ºC and reveals interesting patterns of ZnO: from ramifications, as seen in Fig. 1(a), (b)and (c), to complex “backbone” structures, like in Fig. 1(d). Possibly, these types of patterns were caused by caloricroute during the thermal treatment procedure. Also, since ZnO is a pyroelectric material, we could imagine that this

30 T. G. Silva et al. / Physics Procedia 28 ( 2012 ) 28 – 32

kind of structure might be related to production of carriers during the thermal process; these considerations aresuggestive, although only speculative at this point.

Figure 1: Dark-field optical imaging from typical microstructures obtained on Si(100). (a) to (d) are representative of different dendritic morphologies found in our samples.

The next step was to investigate the influence of substrate surface conditions on microstructure formation.In order to avoid influence of crystalline nature of the Si substrate on ZnO growth, we did not remove native SiO2

layer before depositing the precursor solution. We modify our process in order to include a stirring bath of Sisubstrate in HF (40%) solution for removing the natural layer of SiO2 that is generally present in the silicon surface. This procedure leads to a substrate with a hydrophobic surface layer, smooth in the length scales relevant for this experiment. In Fig. 2 we present representative optical images of microstructures from a sample grown on (a)substrate without native oxide, and (b) substrate with SiO2 layer. Comparing the pattern shown in these two images,we observed that the ZnO deposited film is more homogeneous when substrate surface is not rough, what one would expect (the circular form is due to the effect of the evaporating water drop, during the thermal treatment). Fig. 2(b),on the other hand, showed again the backbone-like structure, with connecting ramifications from the two sides. Thisreinforces the idea of heat paths as main driving force for the dendrictic ZnO microstructures, and it will beinvestigated elsewhere using computational simulations of the growing process.

T. G. Silva et al. / Physics Procedia 28 ( 2012 ) 28 – 32 31

Figure 2: Optical micrographies showing the differences between the structures deposited on Si substrate (a) without the natural SiO2 layer, and

(b) with the oxide layer.

After we acquired these dark field optical images, we annealed a structure with a pattern similar to thoseshown in Fig. 1(b) at 500º C, in vacuum. After this thermal treatment, the structures presented several cracks, whichare believed to be the origin of the defect bands discussed below.

Figure 3: Room Temperature photoluminescence spectrum from a ZnO microstructure annealed at 500ºC.

The photoluminescence (PL) spectrum of this cracked structure is showed in Figure 3. Three separate PLba s ar

350 400 450 500 550 600 650 700

PL

Inte

nsity

(ar

b. u

nits

)

Wavelength (nm)

3.13 eV

nd e observed, and were fitted by gaussian line shapes. Green lines in Fig. 3 are the individual gaussians and red curve is the overall fitting to the data. The first contribution, around 390 nm, is identified as the band edge,which corresponds to an enegy gap of 3.13 eV, lower than the usual values for thin films and nanostructures, butvery close with the result found for ZnO powder by Maesiri et al. [15]. The second peak, at 510 nm, is due to theoxygen vacancies, possibly originated through the annealing in a vacuum chamber. The third structure, centered at630 nm, is called the red emission peak, is related to a sum of other intrinsic defects in ZnO films, already found in

32 T. G. Silva et al. / Physics Procedia 28 ( 2012 ) 28 – 32

recent works [16], although their physical origins are still under investigation. It is interesting to note that despite the annealing being performed in vacuum, the PL spectrum presents a well defined UV band gap emission.

4. Conclusion

Zinc oxide microstructures were grown on Si(100) substrates using the spray pyrolysis technique. Clusters of ZnO with micrometrical size showing ramifications of submicron size and backbone structures have been observed using dark-field optical imaging and the formation of these structures is shown to be dependent on the surface roughness. Photoluminescence spectrum of a typical microstructure shows the near band edge peak at 3.13 eV and two additional bands related to the oxygen vacancies and other unknown origins. As this kind of structure is self–assembled and composed of pure ZnO, is transparent in the visible range, and presents a larger superficial area compared to flat thin films, it can be useful in experiments where spread of light might be needed.

Acknowledgements

We would like to thank F. Iikawa for the assistance with the UV photoluminescence set up at Universidade de Campinas, C. M. Lepienski (UFPR) for optical microscope access, and I. A. Hümmelgen (UFPR) for the help with thermal treatment. We acknowledge CNPq and Fundação Araucária for the financial support.

References

[1] H. K. Yadav, K. Sreenivas, and Vinay Gupta, J. Appl. Phys. 107 (2010) 044507. [2] B. Baruwati, D. K. Kumar, and S. V. Manorama, Sensor Actuat. B 119 (2006) 676-682. [3] W.S. Wang, T.T. Wu, T.H. Chou, Y.Y. Chen, Nanotechnology 20 (2009) 135503. [4] J. Aranovich, A. Ortiz, R.H. Bube, J. Vac. Sci. Technol. 16 (1979) 4.

[5] A. Kaschner, U. Haboeck, Martin Strassburg, Matthias Strassburg, G. Kaczmarczyk, A. Hoffmann, C. Thomsen, Appl. Phys. Lett. 80 (2002) 11. [6] J. Liang-Zhao, Xiao-Min Li, Ji-Ming Bian, Wei-Dong Yu, Can-Yun Zhang, J. Cryst. Growth 280 (2005) 495-501.

[7] J. B. Wang H. M. Zhong, Z. F. Li, W. Lu, Appl. Phys. Lett. 88 (2006) 101913. [8] L. L. Kerr, X. Li, M. Canepa, A. J. Sommer, Thin Solid Films 515 (2007) 5282–5286. [9] J. P. Zhang, L. D. Zhang, L. Q. Zhu, Y. Zhang, M. Liu, X. J. Wang, J. Appl. Phys. 102 (2007) 114903. [10] J. Kennedy, B. Sundrakannan, R.S. Katiyar, A. Markwitz, Z. Li, W. Gao, Current Applied Physics 8 (2008)

291–294. [11] Y. Gong, T. Andelman, G. F. Neumark, S. O’Brien, I. L. Kuskovsky, Nanoscale Res. Lett. 2 (2007) 297–302. [12] M. D. McCluskey and S. J. Jokela, J. Appl. Phys. 106 (2009) 071101. [13] C. Jin, A. Tiwari, and R. J. Narayan, J. Appl. Phys. 98 (2005) 083707. [14] C. Bundesmann, N. Ashkenov, M. Schubert, D. Spemann,T. Butz, E. M. Kaidashev,M. Lorenz, M. Grundmann, Appl. Phys. Lett. 83 (2003) 10. [15] S. Maensiri, P. Laokul, and V. Promarak, J. Cryst. Growth 289 (2006) 102–106. [16] L. N. Dem’yanets, L. E. Li, and T. G. Uvarova, J. Cryst. Growth 287 (2006) 23-27.