Annealing effect on the generation of dual mode acoustic waves in inclined ZnO films

Annealing effect on the generation of dual mode acoustic waves in inclined ZnO films

Ultrasonics 53 (2013) 1264–1269 Contents lists available at SciVerse ScienceDirect Ultrasonics journal homepage: www.elsevier.com/locate/ultras Ann...

2MB Sizes 0 Downloads 31 Views

Ultrasonics 53 (2013) 1264–1269

Contents lists available at SciVerse ScienceDirect

Ultrasonics journal homepage: www.elsevier.com/locate/ultras

Annealing effect on the generation of dual mode acoustic waves in inclined ZnO films H.F. Pang a,b, Y.Q. Fu b,⇑, R. Hou c, K.J. Kirk c, D. Hutson c, X.T. Zu a,⇑, F. Placido b a

School of Physical Electronics, University of Electronic Science and Technology of China, Chengdu 610054, PR China Thin Film Centre, Scottish Universities of Physics Alliance (SUPA), University of the West of Scotland, Paisley PA1 2BE, UK c Microscale Sensors Group, Scottish Universities Physics Alliance (SUPA), University of the West of Scotland, Paisley PA1 2BE, UK b

a r t i c l e

i n f o

Article history: Received 29 October 2012 Received in revised form 26 March 2013 Accepted 27 March 2013 Available online 6 April 2013 Keywords: ZnO film Annealing Ultrasonic wave Pulse-echo Inclined angle

a b s t r a c t ZnO films with different inclined angles on steel substrates were sputter-deposited by changing the substrate tilt angle during deposition and then used to fabricate ZnO film ultrasonic transducers. The ultrasonic performance of those devices was characterized using a standard pulse-echo method. A dual mode wave with both longitudinal and shear wave components was detected from the ZnO device at 0° inclined angle. At a columnar inclined angle of 31°, longitudinal wave excitation was suppressed with a nearly pure shear wave detected. Post annealing of the ZnO film improved the crystallinity and decreased the film stress. The dispersion of the received echoes was observed when the grain sizes of ZnO films were increased after annealing. The frequency components of the waveforms were analyzed and identified using a short time Fourier transform. Post-annealing of the ZnO films changed the primary frequency and enhanced the propagation of the relative high-frequency acoustic wave. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction Recently ZnO films have been widely studied for high performance acoustic wave based microsensors and microfluidics for lab-on-chip and environmental devices. Benefits include low cost, fast response, reduced reagent requirement and increased precision [1–3]. ZnO films normally grow in a hexagonal or wurtzite type crystalline structure and the (0 0 0 2) plane has the lowest surface free energy [4]. Therefore, in the absence of epitaxy between film and substrate, and without any external ion or plasma source, ZnO films grow in the (0 0 0 2) orientation on many different substrates. ZnO acoustic wave devices with a (0 0 0 2) film texture can be applied for sensing in air, gaseous and dry environments [1]. However, biosensors are often needed to detect chemical reactions in a liquid environment. If liquid is present on the sensing surface, excessive damping of the propagating wave occurs when the longitudinal mode (L-mode) wave couples into the liquid [5]. A common solution is to use an in-plane shear horizontal (SH) mode wave, thus dramatically reducing wave coupling into a liquid medium [6], but this normally requires deposition of the ZnO films with special orientations to generate the SH mode wave [7,8]. On the other hand, deposition of c-axis inclined films allows both longitudinal and shear wave modes (S-mode) to be generated on

a single device with different frequencies which could be individually controlled for microfluidic or sensing purposes [1,9–11]. Common methods used to deposit inclined ZnO films by magnetron sputtering include varying the substrate-tilting angle or the angle between the substrate and target [12,13]. Theoretical analysis has shown that pure thickness L-mode of a ZnO based acoustic devices with maximum coupling coefficient occurs at inclined angles of 0° and 65.4°, and maximum amplitude S-mode occurs at 41° [14]. A systematic experimental study on changes of wave modes as a function of film inclined angles (especially above 30°) for sputtered ZnO films has seldom been reported, and the annealing effect on the dual mode wave generation and propagation is not discussed. In this paper, we report that by changing substrate-tilt angle, ZnO films with different inclined crystal angles were deposited on ferritic carbon steel. Film ultrasonic transducers based on those films were fabricated and tested using an ultrasonic pulse-echo method. Annealing effect on the ultrasonic properties of the ZnO film transducers was investigated. Ultrasonic frequency analysis was performed using short time Fourier transformation (STFT) to identify the propagation modes and the multiple frequencies component were discussed. 2. Experimental

⇑ Corresponding authors. E-mail addresses: [email protected] (Y.Q. Fu), [email protected] (X.T. Zu). 0041-624X/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ultras.2013.03.010

ZnO films were deposited on ferritic carbon steel using DC magnetron sputtering from a rectangular Zn target (100 mm 

H.F. Pang et al. / Ultrasonics 53 (2013) 1264–1269

1265

Ef D2h 2mf 2 tan ho

ð1Þ

rf ¼

Fig. 1. Schematic of pulse echo experiment using the ZnO film deposited on ferritic steel plate, showing generation and propagation of longitudinal wave (L-mode) and shear wave (S-mode). Figure indicates generation of S-mode at edges of electrode by c-axis films with 0° inclined angle.

300 mm) at a power of 400 W, an Ar/O2 flow ratio of 90/45 sccm (standard cubic centimeter per minute) and a deposition temperature of 150 °C. Gas pressure during deposition was 4.8 mTorr. Substrates were tilted at different angles using a specially designed multi-angle plate holder, in order to obtain ZnO films with different inclined angles in one deposition whilst maintaining a similar distance from the target. The morphology and crystallinity of the films were characterized using scanning electron microscopy (SEM, Hitachi S4100) and X-ray diffraction (XRD, Siemens D5000 Cu Ka, 40 kV/ 30 mA), respectively. The average grain size was estimated from peak broadening using the Debye–Scherrer equation based on the XRD analysis with correction for instrumental broadening. Film stress was estimated using Eq. (1) based on XRD peak shift [15]:

where Ef and mf are the Young’s modulus and Poisson ratio of the ZnO films (124 GPa and 0.3, respectively), ho is the Bragg angle of stress free ZnO, h is the diffraction angle, and D2h = 2(h  ho). In order to evaluate the piezoelectric properties of the ZnO films, pulse-echo experiments were performed using an ultrasonic pulser/receiver (JSR Ultrasonics DPR300, 475V, USA) as illustrated in Fig. 1. The ZnO film was deposited on ferritic steel plates (area of 25  25 mm2 and thickness of 2.9 mm) and used to fabricate the film based transducer with a top electrode of silver paste (diameter of 2 mm). A negative spike pulse with amplitude of 100 V and duration of 0.2 ls was generated and applied to the ZnO film transducer. The film was excited by the pulse, and echo signals were produced and reflected by the back wall of the steel substrate when the ultrasonic wave propagated in the steel substrate. The receiver was set to a gain of 50 dB and a low frequency pass filter at 50 MHz. The high frequency pass filter was switched off. The echo signals were received from the ZnO film transducer. The electrical signal from the pulse/receiver was recorded using a digital oscilloscope (Agilent 54641A, Agilent Technologies UK Ltd., UK). All the equipments of the measurement system were connected using 50 X coaxial cables. Some of the films were annealed at 400 °C for 1 h in a furnace in air, and then were characterized using XRD. The transducers with post-annealing ZnO films were fabricated and measured using the above pulser/receiver system. In order to analyze the wave frequency, a STFT of the acoustic wave were performed for all the echo signals. 3. Results and discussions From SEM observation, the ZnO films deposited with substrate tilt angles of 0–60° show compact columnar structures as shown in Fig. 2a–e. The film surface shows equi-axial crystals for these films in Fig. 2f–h. This is because the ZnO crystals typically grow in a hexagonal wurzite structure and form long rods along the c-axis resulting in columnar grain structure. With tilting of the substrate during deposition, the ZnO columnar structure shows

Fig. 2. (a–e) Cross-sectional SEM images of ZnO films deposited at substrate tilt angle from 0° to 60°, and surface morphologies (f–h) of ZnO films with substrate tilt angles of 0°, 30° and 60°, respectively.

1266

H.F. Pang et al. / Ultrasonics 53 (2013) 1264–1269

Table 1 Results from analysis and ultrasonic testing of as-deposited and annealed ZnO films deposited with different substrate tilting angles. Substrate tilting angle (°)

Crystal inclined angle (°)

Film thickness (lm)

Average grain size (nm) as-deposited (annealed)

Film stress as (GPa) as-deposited (annealed)

0 15 30 45 60

0 8.5 ± 0.5 21 ± 1 31.1 ± 0.8 34.2 ± 1.0

1.6 3.7 3.4 2.7 1.9

20.1 16.7 17.2 14.7 20.8

1.4 1.1 1.0 0.8 0.5

(21.8) (19.3) (24) (24.3) (27.1)

Fig. 3. XRD analysis results for ZnO films of different inclined angles, as deposited and after annealing at 400° for 1 h.

different inclined angles as measured from SEM analysis and listed in Table 1. The inclined angles of the columnar structure are generally much less than the corresponding substrate-tilt angles. The maximum inclined angle of the columnar structure was 34° for a substrate tilt of 60°. Tilted c-axis growth can be explained by the diffusion limited crystal growth mode of the inclined ZnO films. The mobility of the adatoms is reduced by the oblique deposition angle and the film formation is controlled by the effects of shadowing and oblique ion impingement [1]. Surface morphology observation shows that the crystals deposited on a tilted substrate grow into an elliptical shape due to the inclined columns (see Fig. 2f– h). Post annealing of the films at 400 °C for 1 h did not change the film inclined angle or columnar structure as viewed in the SEM. XRD patterns of the ZnO films (both as-deposited and post-annealed ones) are shown in Fig. 3. For all the films deposited, (0 0 0 2)

(0.63) (0.83) (0.56) (0.32) (0.35)

Amplitude ratio L1/S1 5 2.85 1.25 0 0.5

orientation is dominant. With increasing inclined angle minor peaks are observed which correspond to (1  1 0 0) or (1  1 0 1) orientations. As-deposited films show a modest decrease in grain sizes with increase of inclined angle up to 31° (see Fig. 4a). The relatively larger crystal size at an inclined angle of 34° could be resulted from the elliptical shape of the inclined column. For the films with 0° inclined angle, there is an XRD peak shift of the (0 0 0 2) orientation to lower diffraction angle compared with bulk ZnO (JCPDS file 89-0510), indicating a large compressive stress (1.4 GPa as calculated from Eq. (1)) as calculated from XRD results shown in Fig. 4b. The sputter-deposited ZnO films of a few microns generally have large film stress (mostly intrinsic stress) and a lot of defects. This would have significant influences on the physical properties of the film, such as structure/mechanical stability, film adhesion, as well as the electrical resistance or ultrasonic performance of the device [16]. Post-annealing of the as-deposited ZnO film was commonly applied to increase crystal size, reduce the defects and enhance the coating performance [17,18]. Fig. 4a shows that the grain sizes of post-annealing ZnO films were increased compared with those of the as-deposited ZnO films, which has an important influence on the dispersion of the propagating acoustic pulse [19]. With an increase of the inclined angle, the film compressive stress decreases significantly (see Fig. 4b). This is attributed to many defects which are able to release the compressive stress in the film. As expected, film stress is lowered by post-annealing, with a larger reduction for films deposited at lower tilting angles due to the increased grains during the recrystallisation. Fig. 5a shows results of the pulse-echo test from the ZnO devices on steel substrates, in which L1 and L2 are the first and second longitudinal (L-mode) echoes from the back wall of the steel plate, and S1, S2 are the shear (S-mode) echoes. Time delays L and S mode echoes are 0.95 and 1.85 ls, respectively, because the L mode wave in steel has a velocity of 6105 m/s which is almost twice of that of the S mode wave in steel (3135 m/s). For the samples with 0° inclined angle, multiple echo signals can be detected from both the longitudinal and shear wave modes.

Fig. 4. Effect of inclined angle on ZnO film properties of as-deposited and post-annealed films, (a) grain size, and (b) film stress.

H.F. Pang et al. / Ultrasonics 53 (2013) 1264–1269

1267

Fig. 5. Pulse echo waveforms of (a) as-deposited ZnO films and (b) annealed ZnO films with different columnar inclination angle on steel substrate.

Fig. 6. (a) Dispersion of the waveforms of the transducer with the as-deposited and annealed ZnO films; and (b) the changes of the echo durations as a function of the corresponding variation of the grain size for ZnO films.

The amplitude ratio of the first longitudinal to first shear wave echo (L1/S1) is around 5. Generally for bulk acoustic waves generated by a c-axis aligned film device, the longitudinal wave is expected to be dominant. However in this study, both the L and S wave modes are observed. The generation of the S-mode is not the standard thickness shear mode resonance estimated in the range of 0.76–1.75 GHz for the ZnO films with a thickness from 1.6 to 3.7 lm. The shear peak is supposed to mainly result from lateral effects of the point ultrasonic source for the finite size of the top electrode which is much smaller than the bottom electrode (i.e. the bulk steel substrate). This phenomenon has also been observed in AlN film ultrasonic devices on steel substrates, confirmed using finite element modeling [20]. The conversion of the incident L-mode wave to the S-mode wave was also beneficial to the generation of the dual mode explained using the mass spring lattice model [21,22]. The crystallinity of the as-deposited ZnO films can be improved using heat treatment. The acoustic properties in the annealed ZnO films at 400 °C for 1 h show apparent changes in Fig. 5b. The longitudinal mode could not be identified from the waveform of the ZnO film with an inclined angle of 34° and only the shear wave was identified. With the increase of the film inclined angle the shear wave component increases, and the L1/S1 ratio decreases as shown in Table 1. For the as-deposited film with an inclined angle of 31°, the shear wave is dominant with an indiscernible longitudinal mode. With further increasing the film inclined angle to 34°, the

shear wave remained dominant but the L1/S1 ratio remains at 0.5. The previous analysis reported the shear mode with maximum amplitude was exited at 41° [12], however, even at this angle, a minor longitudinal wave component exists. Normally, the ZnO thin film devices are operating in a below-resonance mode in the ultrasonic tests. In this operation region, the ZnO thin film transducer will replicate the electrical signal from the pulser as a mechanical strain and the frequency content will then be modified by the electrical properties of the ZnO transducer and measurement system [23]. Although the columnar inclined angle was observed in the large angle of 31° and 34°, there is still a dominant c-axis orientation from XRD analysis, which prefers to scatter the excited ultrasonic wave in the lateral direction. Fig. 6 shows the annealing effect on the echo durations of the different samples. The dispersion of the echo signals were clearly observed if comparing the waveform of the as-deposited sample with that of the annealed one at an inclined angle of 0° (see Fig. 6a). The change of the durations of the echoes was calculated as a function of the corresponding variation of the crystal grain size with the results shown in Fig. 6b. The increased crystal grains could scatter the ultrasonic wave to modulate the time–frequency feature of the echoes in the film transducers. STFT is an advanced technique of the signal processing for the ultrasonic wave, which could provide detailed information of the acoustic emission in the frequency domain [24,25]. The Hanning window with 256 points corresponding to 1.28 ls was used to bal-

1268

H.F. Pang et al. / Ultrasonics 53 (2013) 1264–1269

Fig. 7. STFT spectrogram of the echo waveforms for the transducers with (a) as-deposited and (b) annealed ZnO film tilting the substrate at 0°, (c) as-deposited and (d) annealed ZnO film tilting the substrate at 15°, as-deposited (e) and (f) annealed ZnO film tilting the substrate at 30°, as-deposited (g) and (h) annealed ZnO film tilting the substrate at 45°, as-deposited (i) and (j) annealed ZnO film tilting the substrate at 60°; the arrow shows the primary frequency.

H.F. Pang et al. / Ultrasonics 53 (2013) 1264–1269

ance the appropriate resolutions of time and frequency. Fig. 7 shows the spectrograms of the waveforms for the transducers with as-deposited and annealed ZnO films. The primary frequencies of the transducers with as-deposited ZnO films are about 5 MHz, apart from the noise frequency near to 0 MHz, which correspond to the reduplication of the applied electric pulse in the frequency domain. The other frequency components of the reflected echoes are about 15, 20, 31 MHz and 37 MHz. When the ZnO films were annealed at 400 °C, the primary frequency of the spectrograms is changed to 31 MHz. Multiple components of the frequency were observed at 5, 18, 20, 25 MHz and 34 MHz in the range of 5– 45 MHz for the annealed film devices. The frequency of the shear peaks are about 20 MHz for the as-deposited and post-annealed ZnO film transducers with an inclined angle of 31° as shown in Fig. 7g and h. The increased amplitudes of those shear peaks indicate the enhanced effect for the conversion of the L-mode wave to the S-mode wave due to the relative variation of the grain size. Further work is being done to investigate this effect. 4. Conclusion ZnO films with inclined angles up to 34° were grown simultaneously on the ferritic steel substrates by changing the substrate tilt angles during film deposition. XRD analysis shows that the film stress was reduced and crystallinity was changed as a function of the inclined angle. Ultrasonic pulse-echo tests showed dual mode waves (longitudinal and shear) generated with decreasing ratio of longitudinal to shear amplitude when the inclined angle was increased. At an inclined angle of 31°, longitudinal wave excitation was suppressed with nearly pure shear wave observed. The dispersion of the received echoes was observed when the grain sizes of ZnO films were increased based on the annealing effect on the ZnO film transducers. STFT of the waveforms indicated the variations of the primary frequency and multiple frequency components with the changes in inclined angle and annealing temperature. The increased amplitudes of those shear peaks for the annealed sample at an inclined angle of 31° indicate the enhanced conversion of the L-mode wave to the S-mode wave due to the relative variation of the grain size and the improvement of the mechanical property after annealing. Acknowledgments The authors acknowledge financial support from Royal SocietyResearch Grant (RG090609), Carnegie Trust Funding, Royal Society

1269

of Edinburgh, Royal Academy of Engineering-Research Exchanges with China and India Awards, the Fundamental Research Funds for the Central Universities (ZYGX2009J046 and ZYGX2009X007), and the Sichuan Young Scientists Foundation (2010JQ0006). References [1] Y.Q. Fu, J.K. Luo, X.Y. Du, A.J. Flewitt, Y. Li, G.H. Markx, A.J. Walton, W.I. Milne, Sensors and Actuators B 143 (2010) 606–619. [2] P. Muralt, Journal of American Ceramic Society 91 (2008) 1385–1396. [3] X.Y. Du, Y.Q. Fu, S.C. Tan, J.K. Luo, A.J. Flewitt, W.I. Milne, D.S. Lee, N.M. Park, J. Park, Y.J. Choi, S.H. Kim, S. Maeng, Applied Physics Letter 93 (2008) 094105. [4] U. Özgür, Y.I. Alivov, C. Liu, A. Teke, M.A. Reshchikov, S. Dog˘an, V. Avrutin, S.-J. Cho, H. Morkoc, Journal of Applied Physics 98 (2005) 041301. [5] M. Link, J. Webber, M. Schreiter, W. Wersing, O. Elmazria, P. Alnot, Sensors and Actuators B 121 (2007) 372–378. [6] G. Wingqvist, Surface and Coatings Technology 205 (2010) 1279–1286. [7] G. McHale, Measurement Science and Technology 14 (2003) 1847–1853. [8] S. Bensmaine, L. Le Brizoual, O. Elmazria, J.J. Fundenberger, M. Belmahi, B. Benyoucef, Diamond and Related Materials 17 (2008) 1420–1423. [9] A. Fardeheb-Mammeri, M.B. Assouar, O. Elmazria, J.J. Fundenberger, B. Benyoucef, Semiconductor Science and Technology 23 (2008) 095013. [10] J. Weber, W.M. Albers, J. Tuppurainen, M. Link, R. Gabl, W. Wersing, M. Schreiter, Sensors and Actuators A 128 (2006) 84–88. [11] T. Yanagitani, M. Kiuchi, Journal of Applied Physics 102 (2007) 044115. [12] S. Bensmaine, L. Le Brizoual, O. Elmazria, J.J. Fundenberger, B. Benyoucef, Physica Status Solidi (a) 204 (2007) 3091–3095. [13] H.F. Pang, G.A. Zhang, Y.L. Tang, Y.Q. Fu, L.P. Wang, X.T. Zu, F. Placido, Applied Surface Science 259 (2012) 747–753. [14] C.-K. Jen, K. Sreenivas, M. Sayer, Journal of Acoustical Society of America 84 (1988) 26–29. [15] B.D. Cullity, Elements of X-ray Diffraction, Addison-Wesley, London, 1976. [16] M. Lematre, G. Feuillard, T. Delaunay, M. Lethiecq, IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control 53 (2006) 685–696. [17] S.-Y. Chu, W. Water, J.-T. Liaw, Journal of the European Ceramic Society 23 (2003) 1593–1598. [18] D.T. Phan, G.S. Chung, Applied Surface Science 257 (2011) 4339–4343. [19] A.G. Every, A.A. Maznev, Journal of the Acoustical Society of America 131 (2012) 4491–4499. [20] K.J. Kirk, J. Elgoyhen, J.P. Hood, D. Hutson, Journal of Electroceramics 27 (2011) 29–32. [21] H. Yim, Y. Sohn, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 47 (2000) 549–558. [22] Y.H. Kim, S.-J. Song, Journal of the Korean Physical Society 42 (2003) 111– 117. [23] K.J. Kirk, J. Elgoyhen, J.P. Hood, D. Hutson, R.S. Dwyer-Joyce, J. Zhang, B.W. Drinkwater, Insight 52 (2010) 184–187. [24] S.-C. Woo, N.S. Goo, Smart Materials and Structures 16 (2007) 1460–1470. [25] R. Piotrkowski, A. Gallego, E. Castroc, M.T. Garcıa-Hernandez, J.E. Ruzzante, NDT&E International 38 (2005) 260–267.