IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control ,

vol. 59, no. 2,

February

2012

295

Langasite Surface Acoustic Wave Sensors: Fabrication and Testing Peng Zheng, Member, IEEE, David W. Greve, Member, IEEE, Irving J. Oppenheim, Tao-Lun Chin, and Vanessa Malone Abstract—We report on the development of harsh-environment surface acoustic wave sensors for wired and wireless operation. Surface acoustic wave devices with an interdigitated transducer emitter and multiple reflectors were fabricated on langasite substrates. Both wired and wireless temperature sensing was demonstrated using radar-mode (pulse) detection. Temperature resolution of better than ±0.5°C was achieved between 200°C and 600°C. Oxygen sensing was achieved by depositing a layer of ZnO on the propagation path. Although the ZnO layer caused additional attenuation of the surface wave, oxygen sensing was accomplished at temperatures up to 700°C. The results indicate that langasite SAW devices are a potential solution for harsh-environment gas and temperature sensing.

I. Introduction

W

ireless harsh-environment sensing has attracted recent research interest because of applications in engines and combustion systems. Wireless sensing is advantageous when penetrations into the harsh environment are undesirable. Surface acoustic wave sensors fabricated on high-temperature piezoelectric substrates represent an attractive solution because they combine sensing with a wireless transponder function. Surface acoustic wave sensors have received considerable attention recently for harsh-environment applications. Substrate and metallization materials for harsh-environment SAW sensing were investigated by Horsteiner et al. [1]. Operation of langasite SAW delay lines up to 1000°C was reported by Mrosk et al. and a Ti-Si-N barrier layer was evaluated [2]. Prospects for harsh-environment SAW sensing with several substrate materials were also considered by Wolff et al. [3]. Fachberger et al. [4] compared the applicability of lithium niobate, gallium phosphate, and langasite as high-temperature SAW substrates. They observed increased acoustic loss at higher temperature in both gallium phosphate and langasite and preferred langasite at high temperatures. More recently,

Manuscript received August 30, 2011; accepted November 14, 2011. This work was performed in support of research on carbon storage at the National Energy Technology Laboratory under Research and Engineering Services contract DE-FE0004000. P. Zheng, D. W. Greve, I. J. Oppenheim, and T.-L. Chin are with the National Energy Technology Laboratory, Pittsburgh, PA (e-mail: dg07@ andrew.cmu.edu). P. Zheng is with the Department of Physics, Carnegie Mellon University, Pittsburgh, PA. D. W. Greve and T.-L. Chin are with the Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, PA. I. J. Oppenheim and V. Malone are with the Department of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, PA. Digital Object Identifier 10.1109/TUFFC.2012.2190 0885–3010/$25.00

significant effort has been directed at the development of high-temperature sensors. The surface wave velocity and attenuation were measured as a function of temperature by Shrena et al. [5], in which two SAW directions including (0°, 138.5°, 26.6°) were investigated. Increased acoustic losses at higher temperature were observed. High-temperature stability was studied for langasite SAW devices with Pt/Zr metallization [6] and with novel metallization and SiAlON passivation layers [7]. Langasite SAW devices were found to operate for extended times at temperatures of 750°C and above. Some recent work has been directed at wired and wireless sensing with langasite SAW devices. Wireless temperature sensing was reported by Wang et al. [8]. Tortissier et al. operated langasite SAW devices at temperatures up to 500°C with an integrated heater and deposited and monitored drying of a TiO2 sensing layer [9]. This paper did not report on gas sensing. Gas sensing for H2 and C2H4 was reported by Thiele and Pereira da Cunha [10], [11]. Canabal et al. [12] wirelessly observed reflection in langasite SAW devices up to 750°C in CDMA and frequencyswept continuous wave (CW) measurements up to 750°C. Prior work has shown that langasite SAW devices can be operated at elevated temperatures in both wired and wireless modes. We will report here on the development of oxygen-gas-sensing SAW devices for operation in the exhaust of combustion systems at temperatures in excess of 500°C. Section II describes the SAW sensor design and fabrication. Section III describes the technique used for high-resolution measurements of the surface wave velocity change. Section IV reports on wired and wireless temperature sensing. Section V describes measurements of the effect of surface layers on surface wave propagation, with particular emphasis on ZnO oxygen-sensing layers. In Section VI, we report observations of the oxygen sensing behavior of ZnO-coated SAW devices. II. SAW Sensor Design and Fabrication We consider here both temperature and oxygen sensors fabricated on langasite substrates. A degree of temperature sensitivity is intrinsic to the piezoelectric substrate; so a temperature sensor consists of interdigitated transducers on a langasite substrate (Fig. 1, top). In this work, gas sensing is achieved by adding a resistive sensing layer to the propagation path (Fig. 1, bottom). The sensing layer is ZnO, which is known to have a conductivity which depends on the oxygen partial pressure in the ambient atmo-

© 2012 IEEE

296

IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control ,

vol. 59, no. 2,

February

2012

Fig. 2. Typical SAW sensor design (mask ID A13 in Table I). The IDT on the far right was used as the exciting transducer. For sensor T5, the first reflector on the left was lost during wafer dicing. Fig. 1. SAW sensors: (top) temperature and (bottom) gas.

sphere. This sensing layer can be deposited either directly on the langasite substrate or on top of a spacer layer. A spacer layer may be desirable to control the magnitude of the change in velocity or to prevent reactions between the langasite substrate and the ambient atmosphere. The devices used in this work consist of an emitting interdigitated transducer (IDT) and several reflectors, spaced so that the first reflections do not coincide when they return to the emitting IDT. All IDTs had finger widths of 2 µm and 1:1 line-to-space ratios, yielding a surface acoustic wavelength of 8 µm. A typical design is shown in Fig. 2 and a summary of the design parameters for the various devices used in this work is given in Table I. Langasite substrates were purchased from Roditi International Corp. Ltd. (London, UK). Interdigitated transducers were patterned by lift-off using a 10-nm titanium adhesion layer followed by 100 nm of platinum. Initial characterization was performed by probing devices on the wafer and measuring S11 as a function of frequency using

a Rohde and Schwartz ZVB4 network analyzer (Munich, Germany). The resonant frequencies of the IDTs were 340 MHz for langasite (0, 138.5, 27), and 325 MHz for langasite (0, 138.5, 117). Fig. 3 shows the measured S11 transformed into the time domain for a SAW device. Three reflections are seen with arrival times consistent with a surface wave velocity of about 2700 m/s. III. SAW Device Measurements We will consider wired and wireless measurements of SAW sensors for temperature and gas concentration. In these sensors, it is necessary to measure small changes in the surface wave velocity. In previous studies of SAW sensors, various techniques have been used, including measurement of the shift in frequency of a SAW resonator [13], observation of the ringdown of a SAW resonator [14], measurement of S11 and transformation to the time domain using amplitude [8] or both amplitude and phase information [15], [16], measurement of S21 [17], applica-

TABLE I. Fabrication Parameters of SAW Devices.

Euler angle

Aperture

Pt electrode thickness

F3

(0, 138.5, 27)

50λ

100 nm

50

T1

A11

(0, 138.5, 117)

100λ

50 nm

50

T2

A11

(0, 138.5, 27)

100λ

50 nm

50

T3 T4

A3 A11

(0, 138.5, 27) (0, 138.5, 117)

100λ 100λ

50 nm 50 nm

50 50

T5

A13

(0, 138.5, 27)

50λ

50 nm

60

S1 G1 G2

A2 A2 F8

(0, 138.5, 27) (0, 138.5, 27) (0, 138.5, 27)

50λ 50λ 50λ

100 nm 100 nm 100 nm

50 50 50

G3

F5

(0, 138.5, 27)

50λ

100 nm

50

T6

F3

(0, 138.5, 27)

50λ

100 nm

50

Device ID

Mask ID

C1

Emitting transducer (finger pairs)

Reflectors (distance/finger pairs) 4.48 mm/30, 5.76 mm/50 0.72 mm/40, 2.16 mm/60 0.72 mm/40, 2.16 mm/60, 0.8 mm/50 0.72 mm/40, 2.16 mm/60, 1.44 mm/40, 3.60 mm/40 2 mm/50 2 mm/50 2.56 mm/30, 3.84 mm/30, 5.44 mm/50 2.56 mm/20, 3.84 mm/50 4.48 mm/30, 5.76 mm/50

5.12 mm/50, 1.44 mm/40, 1.44 mm/40, 2.88 mm/60 1.44 mm/40, 2.88 mm/60 2.88 mm/20,

3.2 mm/30, 4.48 mm/50, 3.2 mm/50, 5.12 mm/50,

zheng et al.: langasite surface acoustic wave sensors: fabrication and testing

Fig. 3. Example of reflections (S11 measured as a function of frequency and transformed into time domain) at room temperature (device C1 in Table I).

tion of orthogonal frequency coding [18], or analysis of the delay time of the reflected pulses (radar mode) [14], [19]. Many of these methods are adaptable to wireless sensing [20], [21]. We have adopted pulse-mode (radar) interrogation because it offers high phase resolution. In addition, this mode separates the exciting and reflected signals in time, allowing the use of a sensitive receiver and potentially extending the interrogation distance. In our experiments, the RF pulse generation and signal analysis are performed by a PXI-5670 vector signal generator and PXI-5661 vector signal analyzer, respectively (National Instruments Corp., Austin, TX). Details of the apparatus and signal analysis have been reported in a previous publication [22]. Fig. 4 shows a simplified block diagram of the generator and analyzer instruments. In the signal generator, a windowed sinusoid at the intermediate frequency is up-converted and amplified to create the exciting pulse. In the analyzer, the received signal is down-converted and detected to extract in-phase (I) and quadrature (Q) components. One complication is that the two local oscillator (LO) frequencies of the two instruments are different. As a result, the two local oscillators are not phase-coherent and, strictly speaking, we cannot perform coherent detection. However, in this application, we only need to determine the phase difference between two reflections. Apart from the phase noise of the oscillator, the LO phase remains constant during a single pulse train. As a result, we can accurately determine the difference between the phase of two reflections within a single pulse train. The phase difference between LO2 and the signal at time t can be determined from the I and Q components through

 Q(t)  θ(t) = arg  . (1)  I (t) 

Within the same pulse train, we can make two phase measurements θ(t1) and θ(t2). Each phase measurement will be in the range (π, −π), so the phase difference θ21 = θ2 − θ1 has the range (2π, −2π). We limit the range to

297

Fig. 4. Block diagram of the transient (radar-mode) measurement electronics. For wired-mode measurements, the T/R switch is connected directly to the SAW sensor.

(π, −π) and average many phase difference measurements to improve measurement resolution. Averaging θ21 gives an average near zero when the phase difference is near π or −π, so we perform exponential averaging on the real and imaginary parts of e jφ21, and then convert the averaged real and imaginary parts to the averaged phase difference φavg21. The transmit/receive switch in Fig. 4 is included because the pulse produced by the vector signal generator has a spurious tail. The tail is large enough to overload the signal analyzer and interfere with the signals of interest. It is therefore necessary to isolate the transmitter and receiver. This can be done using an isolator or a fast RF switch. The second option offers superior isolation (−20 dB versus −80 dB). Therefore, we used an Agilent P9402A solid state PIN diode switch (Agilent Technologies Inc., Santa Clara, CA) controlled by an Agilent 33120A function generator triggered by the vector signal generator to perform the switch function. The measurement system is controlled by a custom Labview (National Instruments Corp.) application. The phase difference is calculated within a measurement loop that executes every 2 ms. Phase measurements are displayed in a separate loop at a rate of 10 samples/s. To determine the resolution of the phase difference measurement, we averaged the sampled phase difference. The standard deviation of 3000 samples was less than 0.018 radians for pulses larger than −71 dBm in magnitude. This corresponds to an uncertainty of 90 ps in the time difference between two pulses. For a reflection with a 1 μs propagation time, it should be possible to detect velocity changes of 0.24 m/s. IV. Temperature Sensing We first consider temperature sensing using the bare langasite SAW device. We report here temperature sensing in both wired and wireless modes. For wired measurements, a sheathed mineral-insulated thermocouple wire (Super Omegaclad XL, Omega Engineering Inc., Stamford, CT) was used as the high-temperature RF cable to

298

IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control ,

vol. 59, no. 2,

February

2012

Fig. 5. Langasite SAW devices: (a) connected for wired measurements and (b) with a half-wave dipole antenna for wireless measurements (right).

connect the SAW sensor to the RF measurement equipment [Fig. 5(a)]. One thermocouple wire is grounded to the tube; the other wire is used for signal transmission. Thermocouple wire of this type is convenient because it can tolerate high temperatures and exhibits a characteristic impedance near 50 Ω. Platinum/silver paste (9595-A, ESL ElectroScience, King of Prussia, PA) was used to bond the langasite SAW devices to a 5 × 5 cm alumina substrate (CoorsTek Inc., Golden, CO). The conducting paste is also used to connect the high-temperature RF cable with the SAW-emitting IDT terminals. No impedance matching circuit was used. For wireless measurements, the SAW device was mounted on a ceramic substrate and attached to a λ/2 dipole antenna made from two pieces of stainless steel tube. The dipole elements are fastened onto the ceramic substrate with high-temperature cement. Fig. 5(b) shows the assembled SAW device with antenna and (inset) a detail of the SAW device. The SAW device was placed in a furnace (STF55433C-1, Lindberg, Riverside, MI). For wireless measurements, the front sheet metal cover was removed and replaced with a sheet aluminum box of dimensions 52 × 40 × 52 cm. A quarter wavelength antenna was inserted 50 cm from the SAW device and connected to the interrogation electronics. Fig. 6 shows a wireless measurement of the reflection magnitude as a function of delay time at 700°C for device T5 with Euler angle (0, 138.5, 27). The exciting signal was at 335 MHz at a power level of 17 dBm. One large and two smaller reflections at 1.12, 2.22, and 2.66 μs are clearly visible at this temperature. These are the first three reflections expected for the SAW device and correspond to a propagation velocity of 2680 m/s. Phase differences between peaks 1–2 and 1–3 were recorded as a function of furnace setpoint temperature. Using the known propagation distance, these phase differences were converted to effective propagation velocities. Fig. 7 summarizes wired measurements of the effective velocity change for device T1 (0, 138.5, 117) and devices T2 and T3 (0, 138.5, 27). Fig. 8 summarizes wireless measurement of the effective velocity change for devices T4 (0, 138.5, 117) and T5 (0, 138.5, 27). There is good agreement between wired and wireless measurements. As discussed in the following section, the reflection amplitude

Fig. 6. Amplitude of reflections as a function of time measured for device T5 at 700°C. The measurement was performed in wireless mode using the pulse (radar) method.

Fig. 7. Wired measurements of velocity change as a function of temperature for SAW device with Euler angle (0, 138.5, 117) (device T1) and (0, 138.5, 27) (devices T2 and T3). 

zheng et al.: langasite surface acoustic wave sensors: fabrication and testing

Fig. 8. Wireless measurements of velocity change as a function of temperature for SAW device with Euler angle (0, 138.5, 117) (device T5) and (0, 138.5, 27)  (device T4). 

decreases with temperature, limiting the maximum measurement temperature to 900°C in wired measurements. Because of the additional link loss in the wireless measurements, measurements were only possible up to 700°C. The surface acoustic wave velocity decrease at 700°C of the wireless langasite (0, 138.5, 27) sensor was 45 m/s, whereas for the wired langasite sensor with the same Euler angle, it was 34 m/s. This difference could result from the use of two different wafers for these devices with possibly different misalignment. The measurement result for the langasite (0, 138.5, 27) sensor is in fair agreement with previous work on langasite temperature sensors, in which the surface acoustic wave velocity decrease at 700°C was about 50 m/s for langasite (0, 138.5, 26.6) in [23], and 45 m/s for langasite with an unspecified Euler angle in [24]. The (0, 138.5, 27) SAW device is suitable for temperature measurements at 200°C and above, for which the temperature dependence is strong and monotonic. The resolution of the temperature measurement was estimated from the RMS deviation from the average of many sequential temperature measurements. This deviation was calculated from 5 min of data in which each measurement is obtained from the average of 1000 samples. Temperature measurement resolution better than ±0.5°C was achieved from 200°C to 600°C. V. Effect of Overlayers on SAW Propagation We now consider the effect of overlayers on the SAW characteristics. Overlayers can cause a change in surface

299

Fig. 9. Surface acoustic wave attenuation of different SAW devices as a function of temperature. Square: bare langasite; triangle: sputter-deposited ZnO/langasite; circle: spin-coated SiO2/langasite; diamond: spincoated ZnO/SiO2/langasite. 

wave velocity and attenuation resulting from the electroacoustic effect (when the electrical conductivity is within the sensing range) or the mechanical influence of the overlayer. Because langasite itself has increasing acoustic loss with increasing temperature [23], additional losses resulting from overlayers are of particular concern because they may cause reflections to become too weak to be detected. Langasite SAW sensors were placed in a tube furnace for high-temperature measurement in an air ambient atmosphere. The attenuation of pulses relative to the amplitude at room temperature was measured and scaled to a 1-μs propagation time. Table II shows the various surface layers that were investigated. ZnO layers were deposited either by sputtering or spin coating (details in appendix). For all SAW devices examined, strong reflections were observed at room temperature, and the reflection magnitudes decreased with increasing temperature. Fig. 9 shows the comparison of surface acoustic wave attenuation per microsecond propagation time as a function of temperature. For the spin-coated ZnO/SiO2/langasite SAW device, the attenuation increased strongly from 100°C to 250°C. Above 250°C, the reflections are too small to detect with our measurement system. A device with only spin-coated SiO2 showed nearly the same attenuation as a bare langsite SAW device, suggesting that the sol-gel ZnO was the major source of attenuation. The sputtered ZnO SAW device showed somewhat more attenuation than the uncoated device, although reflections were still observable at 600°C.

TABLE II. Overlayers and Fabrication Methods. Device ID S1 G1 G2 G3

Layered structure 100 nm 140 nm 100 nm 100 nm 100 nm

SiO2/ langasite ZnO/ SiO2/ langasite ZnO/ langasite ZnO/ langasite

ZnO fabrication method

SiO2 fabrication method

— Spin coating

Spin coating Spin coating

RF sputtering RF sputtering

— —

300

IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control ,

vol. 59, no. 2,

February

2012

Fig. 10. Oxygen concentration as a function of time.

The increasing attenuation at higher temperature appears to be caused by acoustic losses in the overlayer materials and not the electro-acoustic effect. The electroacoustic effect causes a loss that peaks when the overlayer conductivity is in a particular range [13], [25]. ZnO is known to have a thermally activated conductivity; if the electro-acoustic effect was the major factor, we would observe a loss that first increases and then decreases. Furthermore, the losses for sol-gel and sputtered ZnO are very different despite the fact that these are the same material and should have similar conductivity. It is interesting to note that a similarly increasing loss with temperature has been previously reported for metallic overlayers [26]. We attribute the higher losses in the sol-gel ZnO to the layer microstructure. It is known that sol-gel ZnO layer has skeletal wrinkles structure [27], whereas the sputtered ZnO layers are very smooth. We conclude that the sputtered ZnO is suitable for high-temperature sensing application. VI. SAW Oxygen Sensor Based on the measurements in Section V, the surface wave attenuation resulting from sputtered ZnO is low enough for operation as a gas sensor in the 500°C to 700°C temperature range. In this section, we report the observation of oxygen sensing in ZnO-coated SAW devices. Oxygen sensing is achieved using the electro-acoustic effect. Simulations presented elsewhere [28] indicate that the ZnO conductivity is in the range useful for sensing in this temperature range. Gas-sensing experiments were performed in the wired mode as described previously. An oxygen and nitrogen mixture was injected into an alumina tube with inside diameter of 71 mm which was inserted into the furnace. Computer-controlled mass flow controllers were used to perform step changes in the composition of the gas entering the tube. The total mass flow rate was kept constant at 1000 sccm by varying both the oxygen and nitrogen flow rates. Because of the relatively large volume of the tube, rapid changes in the gas composition were not pos-

Fig. 11. Phase change per microsecond of propagation time as a function of gas testing time. From top to bottom: bare langasite SAW device at 600°C; ZnO/langasite SAW device at 500°C, 600°C, and 650°C. 

sible. The oxygen concentration was varied from 0.25% to 80% as shown in Fig. 10. The experiments reported here were performed using device G2 from Table I. Fig. 11 shows the measured phase difference between the first reflection and the coupled transmitting pulse as a function of time at four different temperatures. At low temperatures, the ZnO electrical conductivity is too low to have a measurable effect on the phase difference. The ZnO conductivity increases with temperature, giving larger phase change at higher temperature. These results are consistent with the expected behavior. However other possible physical mechanisms could cause a change in phase difference, including temperature variations caused by the changing gas flow and interaction between the ambient atmosphere and the langasite substrate. Two additional experiments were performed to rule out temperature variations as an explanation for these results. First, a thermocouple was placed in close proximity to the SAW device. Temperature changes during the measurements were less than ±1°C, which would cause a phase change of about ±0.11 rad, smaller than the phase changes observed at 600°C and 650°C [29], [30]. Second, measurements were performed with two SAW devices connected in parallel, one with a ZnO sensing layer and one without [31]. The reflector positions in these two SAW devices were different so that reflections could be distinguished. These measurements also showed almost no phase change in the bare SAW device and large phase change in the ZnO-coated device. This experiment also rules out the possibility that ambient-atmosphere-induced changes in the langasite substrate are significant at these temperatures. According to the theory of the electro-acoustic effect, the oxygen-induced phase change should first increase with increasing temperature and then decrease at high temperatures. To look for this effect, we performed three series of measurements with the same sequence of oxygen flows, increasing the maximum temperature each time.

zheng et al.: langasite surface acoustic wave sensors: fabrication and testing

301

cating that there are no additional changes to the sensor, provided exposure to temperatures above 650°C is avoided. VII. Summary

Fig. 12. Measured phase change from 0.25% to 80% O2 concentration at different temperatures in three series of tests of the ZnO/langasite device. 

We have reported on the development of a SAW harshenvironment oxygen and temperature sensor. This study has been performed using pulsed (radar-mode) excitation which is expected to have potential for increased sensing range. Wired temperature sensing at temperatures up to 900°C and wireless temperature sensing up to 700°C have been demonstrated. Oxygen sensing has been demonstrated using a sputtered ZnO sensing layer. The sensing layer causes additional attenuation beyond the acoustic attenuation previously observed in langasite. Even so, detectable reflections were observed up to at least 700°C when a sputtered ZnO sensing layer is used. Oxygen sensing was demonstrated in wired mode at temperatures up to about 700°C. The measurements are consistent with the electro-acoustic effect resulting from oxygen- and temperature-induced changes to the ZnO conductivity. Significant changes in the sensing behavior were observed when the sensor was cycled to 700°C and 750°C. However subsequent measurements showed good reproducibility, provided the temperature was not increased above 650°C. These results show the potential of langasite SAW devices for harsh-environment sensing. Important issues that must be addressed in future studies include extending the temperature range for sensing and determining and possibly reducing the effects of high-temperature cycling. Appendix ZnO Layer Deposition and Characterization

Fig. 13. Total phase change between 0.25% and 40% oxygen in three separate experiments. 

These observed phase changes were not reproducible between runs, but did show a consistent trend (Fig. 12). A peak in phase response is observed after the two highertemperature anneals, and the peak shifts to lower temperature with increasing anneal temperature. These results could be explained by an increase in ZnO grain size after each anneal, which would have the effect of increasing the conductivity at fixed temperature. It is not possible to rule out other effects, however, including reactions between the ZnO and the langasite substrate or reactions involving the metal electrodes. A final series of measurements was performed to determine the reproducibility of the SAW device after it had been annealed at 750°C. The oxygen response was measured between 500 and a maximum temperature of 650°C. Three complete sequences of measurements were performed. The results are shown in Fig. 13 for oxygen concantrations between 0.25% and 40%. The consistency of these three separate measurements is very good, indi-

ZnO sensing layers were deposited by two different techniques. Spin-coated ZnO was deposited on a SiO2 spacer layer. The langasite SAW device was spin coated with polysiloxanes (IC1–200 and DC4–500, Futurrex Inc., Franklin, NJ) at 2000 rpm. It was then annealed in air at 400°C for 30 min to form a 200-nm SiO2 layer. A ZnO layer, ~140 nm thick, was then deposited onto the SiO2/langasite SAW device using the sol-gel spin coating method described in [27]. Briefly, the SAW device was spin-coated using a sol-gel solution containing zinc acetate (1.3 M Zn++ concentration) followed by air drying and an air anneal at 700°C. The ZnO and SiO2 layers on the emitting IDT region were partially removed with acetone using a cotton tip to expose the terminals before each annealing step. A SAW device with only 200 nm SiO2 was also prepared for comparison. Similarly deposited ZnO layers on stainless steel substrates were characterized by X-ray diffraction using a PANalytical X’Pert ProMPD powder diffractometer (Almelo, The Netherlands) with a Cu Xray source [30]. These measurements show diffraction features characteristic of the ZnO wurtzite structure: (100) at 31.7°, (002) at 34.4°, and (101) at 36.3°, with a strong preference for growth in the (002) (c-axis) direction.

302

IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control ,

Fig. 14. Model for ZnO growth rate determined from full-factorial experiment. The shaded areas indicate films with resistance less than 104 Ω. 

ZnO was also prepared by reactive RF sputtering from a 2.5-cm-diameter Zn target. Sputter conditions were determined after performing a two-level full-factorial experiment in which the total pressure, RF power, and oxygen partial pressure were varied. Both the layer thickness resulting from a 1-h deposition and the resistance measured between two point probes separated by about 2.5 mm were determined. The data were used to construct models [32] for the logarithm of the layer thickness and layer resistance. Fig. 14 shows contours of layer thickness as a function of RF power and oxygen partial pressure for two different total deposition pressures. The growth rate decreases with oxygen partial pressure; however, only layers deposited at higher oxygen partial pressure were smooth, transparent, and non-conductive. Based on these results, ZnO layers for the SAW devices were fabricated at a pressure of 4 mT and RF power of 50 W in 25% O2/75% Ar. These conditions resulted in a layer approximately 200 nm in thickness in 60 min. Sputtered ZnO films were also characterized by X-ray diffraction after annealing at 600°C and 750°C [30]. The same X-ray diffraction features were observed as for the spin-on ZnO and there was a similar preference for growth in the (002) (c-axis) direction. In both spin-on and sputtered ZnO, the average size of crystalline domains, determined by Scherrer analysis of the (002) feature, is on the order of 30 nm for all three samples. Acknowledgments We thank J. Miller and T. Ashok for providing ZnO spin-coating and X-ray diffraction, L. Cao for fabrication of Pt/Ti electrodes, and W. Wu for construction and programming of the gas flow system. References [1] J. Hornsteiner, E. Born, G. Fischerauer, and E. Riha, “Surface acoustic wave sensors for high temperature applications,” in IEEE Int. Freq. Control Symp., 1998, pp. 615–619.

vol. 59, no. 2,

February

2012

[2] J. W. Mrosk, C. Ettl, L. Berger, P. Dabali, H. J. Fecht, G. Fischerauer, J. Hornsteiner, K. Riek, E. Riha, E. Born, M. Werner, A. Dommann, J. Auersperg, E. Kieselstein, B. Michel, and A. Mucha, “SAW sensors for high temperature applications,” in Proc. 24th Annu. Conf. Industrial Electronics Society IEEE, 1998, pp. 2386–2390. [3] U. Wolff, F. L. Dickert, G. K. Fischerauer, W. Greibl, and C. C. W. Ruppel, “SAW sensors for harsh environments,” IEEE Sens. J., vol. 1, no. 1, pp. 4–13, 2001. [4] R. Fachberger, G. Bruckner, G. Knoll, R. Hauser, J. Biniasch, and L. Reindl, “Applicability of LiNbO3, langasite and GaPO4 in high temperature SAW sensors operating at radio frequencies,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 51, no. 11, pp. 1427– 1431, 2004. [5] I. Shrena, D. Eisele, E. Mayer, L. M. Reindl, J. Bardong, and M. Schmitt, “SAW-relevant material properties of langasite in the temperature range from 25 to 750°C: New experimental results,” in 2008 IEEE Int. Ultrasonics Symp. Proc., pp. 209–212. [6] J. A. Thiele and M. Pereira da Cunha, “Platinum and palladium high-temperature transducers on langasite,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 52, no. 4, pp. 545–549, 2005. [7] M. Pereira da Cunha, R. J. Lad, T. Moonlight, G. Bernhardt, and D. J. Frankel, “High temperature stability of langasite surface acoustic wave devices,” in Proc. 2008 IEEE Int. Ultrasonics Symp., pp. 205–208. [8] S.-Q. Wang, J. Harada, and S. Uda, “A wireless surface acoustic wave temperature sensor using langasite as substrate material for high-temperature applications,” Jpn. J. Appl. Phys., vol. 42, no. 9B, pt. 1, pp. 6124–6127, Sep. 2003. [9] G. Tortissier, L. Blanca, A. Tetelina, J.-L. Lachauda, M. Benoit, V. Conédéra, C. Dejousa, and D. Rebièrea, “Langasite based surface acoustic wave sensors for high temperature chemical detection in harsh environment: Design of the transducers and packaging,” Sens. Actuators B, vol. 156, no. 2, pp. 510–516, 2011. [10] J. A. Thiele and M. Pereira da Cunha, “High temperature LGS SAW devices with Pt/WO3 and Pd sensing films,” in IEEE Ultrasonics Symp., 2003, pp. 1750–1753. [11] J. A. Thiele and M. Pereira da Cunha, “High temperature SAW gas sensor on langasite,” in Proc. IEEE Sensors, 2003, pp. 769–772. [12] A. Canabal, P. M. Davulis, E. Dudzik, and M. Pereira da Cunha, “CDMA and FSCW surface acoustic wave temperature sensors for wireless operation at high temperatures,” in 2009 IEEE Int. Ultrasonics Symp. Proc., pp. 807–810. [13] J. J. Caron, T. D. Kenny, L. J. LeGore, and D. G. Libby, C. J. Freeman, and J. F. Vetelino, “A surface acoustic wave nitric oxide sensor,” in 1997 IEEE Int. Frequency Control Symp., pp. 156–162. [14] V. Kalinin, “Passive wireless strain and temperature sensors based on SAW devices,” in Proc. 2004 IEEE Radio and Wireless Conf., pp. 187–190. [15] J. H. Kuypers, L. M. Reindl, S. Tanaka, and M. Esashi, “Maximum accuracy evaluation scheme for wireless SAW delay-line sensors,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 55, no. 7, pp. 1640–1652, 2008. [16] J. H. Kuypers, S. Tanaka, M. Esashi, D. A. Eisele, and L. M. Reindl, “Passive 2.45 GHz TDMA based multi-sensor wireless temperature monitoring system: Results and design considerations,” in IEEE Int. Ultrasonics Symp. Proc., 2006, pp. 1453–1458. [17] J. Hornsteiner, E. Born, G. Fischerauer, and E. Riha, “Surface acoustic wave sensors for high temperature applications,” in 1998 IEEE Int. Frequency Control Symp., pp. 615–620. [18] D. C. Malocha, J. Pavlina, D. Gallagher, N. Kozlovski, B. Fisher, N. Saldanha, and D. Puccio, “Orthogonal frequency coded SAW sensors and RFID design principles,” in IEEE Frequency Control Symp., 2008, pp. 278–283. [19] A. Pohl, F. Seifert, L. Reindl, G. Schol, T. Ostertag, and W. Pietsch, “Radio signals for SAW ID tags and sensors in strong electromagnetic interference,” in 1994 IEEE Ultrasonics Symp., pp. 195–198. [20] L. Reindl, G. Scholl, T. Ostertag, H. Scherr, U. Wolf, and F. Schmidt, “Theory and application of passive SAW radio transponders as sensors,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 45, no. 5, pp. 1281–1292, 1998. [21] A. Pohl, “A review of wireless SAW sensors,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 47, no. 2, pp. 317–332, 2000. [22] T. L. Chin, D. P. Zheng, W. Greve, L. Cao, and I. J. Oppenheim, “Flexible instrumentation for wireless SAW,” in 2010 Proc. IEEE Int. Ultrasonics Symp., pp. 261–264. [23] J. Bardong, M. Schulz, M. Schmitt, I. Shrena, D. Eisele, E. Mayer, L. M. Reindl, and H. Fritze, “Precise measurements of BAW and

zheng et al.: langasite surface acoustic wave sensors: fabrication and testing SAW properties of langasite in the temperature range from 25°C to 1000°C,” in IEEE Int. Frequency Control Symp., 2008, pp. 326–301. [24] S. Q. Wang, J. Harada, and S. Uda, “A wireless surface acoustic wave temperature sensor using langasite as substrate material for high-temperature applications,” Jpn. J. Appl. Phys., vol. 42, no. 9B, pp. 6124–6127, 2003. [25] D. S. Ballantine, Acoustic Wave Sensors—Theory, Design, and Physico-Chemical Applications. Amsterdam, The Netherlands: Elsevier, 1997. [26] I. Shrena, D. Eisele, E. Mayer, L. M. Reindl, and J. Bardong, “SAWproperties of langasite at high temperatures: Measurement and analysis,” in Proc. 2009 Int. Conf. Signals, Circuits and Systems, pp. 1–4. [27] B. J. Miller, H.-J. Hsieh, B. H. Howard, and E. Broitman, “Microstructural evolution of sol–gel derived ZnO thin films,” Thin Solid Films, vol. 518, no. 23, pp. 6792–6798, 2010. [28] P. Zheng, “High temperature langasite surface acoustic wave sensors,” unpublished Ph. D. thesis, Dept. of Physics, Carnegie Mellon University, Pittsburgh, PA, Apr. 2011. [29] P. Zheng, T. L. Chin, D. W. Greve, I. J. Oppenheim, V. Malone, T. Ashok, J. Miller, and L. Cao, “Langasite SAW device with gassensitive layer,” in 2010 Proc. IEEE Int. Ultrasonic Symp., pp. 1462– 1465. [30] P. Zheng, T. L. Chin, D. W. Greve, I. J. Oppenheim, V. Malone, and L. Cao, “High temperature langasite SAW oxygen sensor,” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 58, no. 8, pp. 1538–1540, 2011. [31] P. Zheng, D. W. Greve, I. J. Oppenheim, and V. Malone, “Langasite SAW temperature and oxygen multi-sensor,” in Proc. IEEE Frequency Control Symp., 2011. [32] National Institute of Standards and Technology. (2011, Aug.) “Full factorial example,” in NIST/SEMATECH e-Handbook of Statistical Methods, [Online]. Available: http://www.itl.nist.gov/div898/handbook/pri/section3/pri3332.htm

Peng Zheng was born in China in 1983 and came to United States for graduate study in 2005. His educational background is in physics and electrical engineering with a B.S. degree in physics from Nanjing University (2005), M.S. degree in physics (2007), M.S. degree in electrical and computer engineering (2010), and Ph.D. degree in applied physics (2011) from Carnegie Mellon University. His research activities have included non-destructive evaluation using ultrasonic sensors, SAW devices for wireless and high-temperature sensing, finite element analysis of SAW devices, and 3-D imaging with polarization cameras. Dr. Zheng is now a Senior R&D Engineer at Covidien, Boulder, CO, working on sensors for medical device applications. He is an IEEE member and APS member.

David W. Greve received the Ph.D. degree in electrical engineering from Lehigh University in 1980. He was with Philips Research Laboratories, Sunnyvale, before joining Carnegie Mellon University in 1982, where he is now Professor of Electrical and Computer Engineering. His research is in the general area of device physics and fabrication technology, and presently concerns various applications of ultrasonics.

303

Irving J. Oppenheim has taught at Carnegie Mellon since 1972, and his educational background is in structural engineering with a B.E. degree from The Cooper Union for the Advancement of Science and Art (1968), an M.S. degree from Lehigh University (1970), and a Ph.D. degree from Cambridge University (1972). His past research activities include dynamics, force-cognitive robotic manipulation, control of dynamically stable robots, and vibration of tensegrity structures. His recent research activities include MEMS for ultrasonics and vibration sensing, inductively coupled Lamb wave transducers, change detection from scattering in cylindrical waveguides, and SAW devices for wireless strain sensing and chemical sensing.

Tao-Lun (Darren) Chin was born in West Lafayette, IN, in 1980. His educational background is in electrical engineering, with a B.S. degree from Chung Hua University in Taiwan and an M.S. degree from Syracuse University. He is presently a Ph.D. student at Carnegie Mellon University. His recent research studies include SAW sensors and antenna designs.

Vanessa Malone was born in St. Thomas, U.S. Virgin Islands in 1989. She has a B.S. degree in chemistry from the University of Virgin Islands, St. Thomas, US Virgin Islands (2009) and an M.S. degree in computational mechanics from Carnegie Mellon University, Pittsburgh, PA (2011). Working primarily as a chemist, over the past five years she has worked in a range of disciplines ranging from chemical and biomedical analysis to environmental engineering. She’s an alumna of the Emerging Caribbean Scientists (ECS) program in which, for two years, she worked as an experimental organic chemist in the isolation of active secondary bio-metabolites to retard solid tumor growth. Adding to her chemistry career, she also part took in the National Oceanic and Atmospheric Administration (NOAA) Educational Partnership Program (EPP) and, consequently, has worked as an oceanic and atmospheric chemist within the NESDIS and NIST branches dealing with sea-surface carbon dioxide levels and halocarbon emissions, respectively. She has also worked within the engineering field as a graduate researcher in the Carnegie Mellon University Green Roof Project and, most recently, as a piezoelectric researcher in the Carnegie Mellon University Crystal Sensing Project. Currently, she works as an optical coatings chemist within the PPG, Inc. R&D Division at the Monroeville Chemicals Center: a recent change from her former position as a Probe MS and GC/MS analytical chemist within the same department and facility. In her academic and professional pursuits, she has been awarded several notable scholarships and has partaken in several relevant programs including, but not limited to: Early Admissions Scholarship (2007–2009), the HBCU-UP ECS Scholarship (2007–2009), the NOAA EPP Program (2007–2008), The University of Iowa VIGRE-REU Program (2009), the Carnegie Mellon University Department of Civil Engineering Service Assistantship (2009–2010), and the HBCU-STEM Fellowship (2009–2011). She is also the 1st place Oral Chemistry presentation award recipient for her works at the 2008 National HBCU Conference.