Subsecond-response SAW humidity sensor with porphyrin nanostructure deposited on bare and metallised piezoelectric substrate R. Rimeika, D. C ˇ iplys, V. Poderys, R. Rotomskis, S. Balakauskas and M.S. Shur The surface acoustic wave (SAW) humidity sensor based on the Meso-tetra (4 sulfonatophenyl) porphyrin (TPPS 4 ) nanostrip structure deposited on a bare or metallised surface of the piezoelectric YZ LiNbO 3 substrate has been investigated. The SAW attenuation, which greatly increases with humidity in sensors with TPPS 4 layers on bare substrates, is almost humidity-independent in devices with TPPS 4 layers on metallised substrates. For both types of sensors, the SAW velocity exhibits close-to-linear decrease with growing humidity. The sensor exhibited fast response and recovery times of from 0.8 to 0.5 s in both amplitude and phase of the output signal. The real-time monitoring of human breath with the help of the sensor was demon- strated. Introduction: Surface acoustic wave (SAW) devices are attractive for use in humidity sensors owing to their small size, low cost, high sensitivity and reliability [1]. There is growing demand for fast- response humidity sensors, in particular, for medical applications [2, 3]. Recently, we have demonstrated the strong impact of ambient humidity on the SAW velocity and attenuation in structures consisting of bio-organic Meso-tetra (4 sulfonatophenyl) porphyrin (TPPS 4 ) layers deposited on piezoelectric YZ LiNbO 3 substrates [4, 5]. In this Letter, we compare the performance of SAW humidity sensors with a TPPS 4 nanostrip structure deposited on bare and metallised surfaces of YZ LiNbO 3 substrates, and demonstrate a fast subsecond response of these sensors to abrupt humidity changes. Sample preparation: Two-port SAW delay lines were fabricated by depositing interdigital transducers (IDTs) on YZ LiNbO 3 substrates using standard photolithography technique. The IDT centre frequen- cies were 87, 108, and 144 MHz at SAW wavelengths L of 40, 32, and 24 mm, respectively. The layers of self-assembled TPPS 4 J-aggregate nanostructures with thickness of about 0.5 mm were formed on the SAW propagation path between IDTs by casting a drop of acidic aqueous solution of TPPS 4 (from Frontier Scientific, Inc.) and drying it in ambient air. Typically, in our samples, the centre-to-centre IDT spacing L was from 10 to 15 mm, and the TPPS 4 layer length along the SAW propagation was 4 to 6 mm. More details of the preparation and properties of the TPPS 4 layers can be found elsewhere [4–6]. In our experiments, we used samples with TPPS 4 layers deposited on a bare surface of LiNbO 3 substrates as well as on the LiNbO 3 surfaces coated with thin Au or Ag films. Steady-state response: A sample was placed into the closed chamber, in which the relative humidity (RH) was slowly changed with the rate of approximately 3%=min and monitored using a hygrometer RH411 (from Omega Engineering, Inc.). Changes in delay-line SAW trans- mission loss were measured in a pulse mode by exciting the input IDT from an external RF pulse generator and monitoring the signal received at the output IDT with an oscilloscope. A change in the SAW attenuation constant a (normalised with respect to the wave number k) was determined from the variation of the transmission loss DA (in dB) as: Da k ¼ L 2pl D A 8:68 ð1Þ where l is the length of the TPPS 4 layer along SAW propagation. Changes in the SAW velocity were measured using the SAW oscillator technique: the sample under test was connected into the feedback loop of a broadband RF amplifier, and the frequency f of resulting CW oscillations was measured by a frequency meter connected to the output IDT. A relative change in the SAW velocity, DV=V 0 , was determined from the oscillator frequency variation Df as: DV V 0 ¼ L l Df f 0 ð2Þ where the zero subscripts denote initial values of f and V , and L is the centre-to-centre spacing between the IDTs. The measured dependen- cies DA (RH) and Df (RH) as well the variations in SAW attenuation and velocity extracted using (1), (2) are shown in Figs. 1 and 2, respectively. In the sample with the TPPS 4 layer on a bare surface, the near-exponential growth of the SAW attenuation (Fig. 1, curve (i)) and the linear decrease in the SAW oscillator frequency (Fig. 2, curve (i)) are observed with increasing RH. The presence of the metal film drastically reduced the humidity-induced attenuation change (Fig. 1, curve (ii)) but had no significant effect on the velocity variation slope (Fig. 2, curve (ii)). This implies that different physical mechanisms are responsible for the SAW velocity and attenuation changes [7]. The velocity decrease with growing RH can be attributed to the purely mechanical effect of mass loading. The attenuation behaviour is strongly influenced by the SAW electric field, which penetrates into the film on the bare surface and is short-circuited on the metallised surface of the piezoelectric crystal. The loss reduction by substrate metallisation extends the measurable RH range of the SAW oscillator- based sensor. In the absence of a TPPS 4 layer, both the SAW attenuation and velocity remained constant, within measurement error of a few per cent (Figs. 1 and 2, curves (iii)). This confirms the substantial role of the TPPS 4 layer in the variation of SAW parameters and excludes the effect of temperature on the SAW velocity in the LiNbO 3 substrate. Fig. 1 Changes in SAW delay-line transmission loss and normalised attenuation against relative humidity L ¼ 24 mm, l ¼ 4 mm Fig. 2 SAW oscillator frequency and relative velocity change against relative humidity L ¼ 32 mm, l ¼ 5 mm, L ¼ 12 mm Transient response: The response of the TPPS 4 -based SAW sensor to abrupt humidity changes was examined as follows. The humidity in the measurement chamber initially was set at a value different from that of the air in the room. The abrupt change of RH was performed by fast removal of the chamber cover and exposing the sample to the ambient atmosphere. Examples of RH increase and decrease are shown in Figs. 3a and b, respectively. The time-dependent changes in the SAW sensor transmission loss DA(t) and in the input-output phase shift DF(t) were measured with a network analyser E5062A (Agilent Technologies, Inc.). The phase variation is related to the change in the SAW velocity as: ELECTRONICS LETTERS 13th September 2007 Vol. 43 No. 19