sensors Article
Surface Acoustic Wave Sensor with Pd/ZnO Bilayer Structure for Room Temperature Hydrogen Detection Cristian Viespe
and Dana Miu *
National Institute for Laser, Plasma and Radiation Physics, Laser Department, Atomistilor # 409, 077125 Bucharest-Magurele, Romania; [email protected] * Correspondence: [email protected]; Tel.: +40-21-457-4027 Received: 6 June 2017; Accepted: 28 June 2017; Published: 29 June 2017
Abstract: A Surface Acoustic Wave (SAW) hydrogen sensor with a Pd/ZnO bilayer structure for room temperature sensing operation has been obtained by Pulsed Laser Deposition (PLD). The sensor structure combines a Pd layer with optimized porosity for maximizing mass effects, with the large acoustoelectric effect at the Pd/ZnO interface. The large acoustoelectric effect is due to the fact that ZnO has a surface conductivity which is highly sensitive to chemisorbed gases. The sensitivity of the sensor was determined for hydrogen concentrations between 0.2% and 2%. The limit of detection (LOD) of the bilayer sensor was about 4.5 times better than the single ZnO films and almost twice better than single Pd films. Keywords: SAW sensor; hydrogen sensor; bilayer; pulsed laser deposition; nanoporous film; Palladium; ZnO
1. Introduction Hydrogen has many important applications such as alternative, clean energy sources, propulsion systems, and biomedical devices [1–3]. However, serious problems are implied by hydrogen use: in concentrations over 4.6% it is highly explosive [4], and over 4% it is extremely flammable [5]. The additional fact that hydrogen is colorless and odorless makes the need for sensors capable of detecting concentrations of 2% or less very important. Sensors based on resistance, MOS-field effect transistors, or optical devices have been used for hydrogen detection, each having its advantages and disadvantages [6]. Surface acoustic wave (SAW) sensors have been widely used as hydrogen sensors due to advantages such as high sensitivity, fast response, reliability, and low cost [6–8]. Detection limits for H2 gas of 1250 to 2000 ppm (operating temperatures 80–120 ◦ C) [9], 600 ppm (at 80 ◦ C) [10] or as low as 200 ppm (at 175 ◦ C) [11] using SAW sensors have been reported. In SAW sensors, the sensing mechanism is based on the perturbation of the surface acoustic wave propagation by mechanical and/or electrical effects. From a practical standpoint, the relevant effects for SAW sensing are changes in mass density and in electrical conductivity. In certain conditions, the changes in electrical conductivity can lead to a much stronger SAW sensor response than the changes in mass density [12]. This is because the acoustoelectric interactions between the electrical potential associated with surface acoustic wave propagation on a piezoelectric crystal and the mobile electric charges in the sensing film can lead to large modifications of the SAW attenuation and propagation velocity [13]. Bilayer SAW structures have been proven to increase the sensor sensitivity compared to single layer ones, and to decrease the detection limit, even in the case of room temperature operation [14,15]. By using bilayer films, the absolute value of the multilayer film conductivity can be adjusted to operate in a region where the acoustoelectric effect is efficient. Various material combinations such as Pd/WO3 , ZnO/GaN, AlN/ZnO, or ZnO/SiO2 have been explored for SAW gas sensors based on this
Sensors 2017, 17, 1529; doi:10.3390/s17071529
www.mdpi.com/journal/sensors
Sensors 2017, 17, 1529
2 of 9
effect [9,16–18]. In the case of hydrogen sensing, an efficient acoustoelectric effect is desirable, since the mass effect for hydrogen is smaller than for heavier gases [9]. We propose a SAW bilayer structure for the detection of hydrogen consisting in Pd/ZnO, which, to our best knowledge, has not been investigated. Pd has excellent hydrogen absorption properties, having the highest hydrogen solubility of any element at atmospheric pressure as well as thermodynamically favored H2 dissociation [19]. The absorbing properties are improved even more in the case of nanoporous Pd layers, as we have shown in previous experiments [20]. ZnO is a wide band-gap semiconductor which is simple, reliable, low-cost, and easily mass-produced [21]. It is a piezoelectric material with a high coupling coefficient K, which is a measure of the conversion efficiency between electrical and acoustic energy [22]. When a semiconducting ZnO layer is deposited on a piezoelectric substrate, its mobile carriers couple to the electric field associated with SAW propagation along the piezoelectric substrate, leading to SAW attenuation and velocity change [16]. An important characteristic of ZnO for our case is the strong sensitivity of its surface conductivity to adsorbed species [23,24], which will increase the overall sensitivity of the bilayer sensor to hydrogen. It is worth mentioning that ST quartz substrates, which are used in our case, have a low temperature coefficient in comparison to other materials used for SAW substrates such as LiNbO3 or LiTaO3 , and are therefore less sensitive to temperature variations [25]. We have therefore considered a bilayer structure consisting in a combination of a Pd layer with optimized porosity for maximizing mass effects, and a ZnO layer having an electron conductivity which is very sensitive to the exposure of the surface to hydrogen. Thus, both changes in mass density and in electrical conductivity contribute to the sensor response, leading to a SAW sensor with large hydrogen sensitivity. 2. Materials and Methods 2.1. Film Deposition and Characterization The sensitive layers of the SAW sensors were deposited onto ST-X quartz substrates by pulsed laser deposition (PLD). PLD was used because it presents a series of advantages in thin film deposition in general [17,26], and in particular has been proven to allow good control of sensitive film porosity, which is essential for SAW gas sensors [20]. The beam of a Nd:YVO4 laser is focused onto the surface of the target, leading to ablation of the target material and deposition onto a substrate placed at a distance of 35 mm in front of the target and parallel to its surface (Figure 1). The laser generates pulses of 10 ps duration, at a 10 kHz repetition rate, and a wavelength of 532 nm. The targets and substrate are placed in a vacuum chamber equipped with a system for the control of the gas pressure and flow. Mass flow meters are placed on the gas bottles and controlled by a mass flow controller, combined with a throttle valve placed on a rotary vane vacuum pump controlled by a pressure controller maintain the desired pressure in the ablation chamber and a controlled gas flow. Targets are placed on computer-controlled x-y tables which ensure continuous target movement during deposition, avoiding target erosion which leads to large film roughness. The system also permits deposition from multiple targets for multilayers. The laser energy density on the target can be modified by changing the distance between the focusing lens and the target.
Sensors 2017, 17, 1529 Sensors Sensors 2017, 2017, 17, 17, 1529 1529
3 of 9 33 of of 99
Figure 1. Experimental layer deposition. deposition. Figure 1. Experimental setup setup of of sensitive sensitive layer
In In order order to to compare compare the the sensing sensing properties properties of of single single layer layer and and bilayer bilayer SAW SAW sensors, sensors, we we have have In order to compare the sensing properties of single layer and bilayer SAW sensors, we have deposited aa series of films all having aa thickness of 330 nm (Figure 2), as measured using aa QUANTA deposited series of films all having thickness of 330 nm (Figure 2), as measured using QUANTA deposited a series of films all having a thickness of 330 nm (Figure 2), as measured using a QUANTA EE == 50 keV Electron Microscope (SEM) and a Tokyo Seimitsu Surfcom 130 A profilometer. keV Scanning Scanning E = 50 50 keV Scanning Electron Electron Microscope Microscope (SEM) (SEM) and and aa Tokyo Tokyo Seimitsu Seimitsu Surfcom Surfcom 130 130 A A profilometer. profilometer. The bilayer film has aa thinner metal layer (Pd) on top of the semiconducting layer (ZnO) since this is The bilayer film has thinner metal layer (Pd) on top of the semiconducting layer (ZnO) since this is The bilayer film has a thinner metal layer (Pd) on top of the semiconducting layer (ZnO) since this known to lead to aa higher sensitivity [13,15]. In addition, the very good H 22 absorption properties of known to lead to higher sensitivity [13,15]. In addition, the very good H absorption properties of is known to lead to a higher sensitivity [13,15]. In addition, the very good H2 absorption properties Pd ensure penetration of gas species to ZnO layer. The Pd, and thin were PdPd ensure penetration of the the gasgas species to the the ZnOZnO layer. The The Pd, ZnO, ZnO, and Pd/ZnO Pd/ZnO thin films films were of ensure penetration of the species to the layer. Pd, ZnO, and Pd/ZnO thin films directly deposited onto the quartz substrates at an average laser power of 0.7 W in aa gas flow of 0.5 sccm. directly deposited onto the quartz substrates at an average laser power of 0.7 W in gas flow of 0.5 sccm. were directly deposited onto the quartz substrates at an average laser power of 0.7 W in a gas flow of Before Before deposition, deposition, the the chamber chamber was was evacuated evacuated to to aa base base pressure pressure of of 10 10-5-5 Torr. Torr. In In the the case case of of Pd, Pd, films films 0.5 sccm. Before deposition, the chamber was evacuated to a base pressure of 10−5 Torr. In the case of were were obtained obtained by by ablation ablation of of aa pure pure Pd Pd target target (99.95%) (99.95%) in in an an Ar Ar atmosphere atmosphere of of 400 400 mTorr. mTorr. The The ZnO ZnO Pd, films were obtained by ablation of a pure Pd target (99.95%) in an Ar atmosphere of 400 mTorr. films films were were obtained obtained by by ablation ablation of of aa ZnO ZnO target target (99.95%) (99.95%) in in O O22 at at 400 400 mTorr. mTorr. All All depositions depositions were were The ZnO films were obtained by ablation of a ZnO target (99.95%) in O2 at 400 mTorr. All depositions made made at at room room temperature, temperature, and and no no post-deposition post-deposition thermal thermal treatments treatments were were used. used. were made at room temperature, and no post-deposition thermal treatments were used.
Figure 2. Schematic diagram of film structure. Figure Figure 2. 2. Schematic Schematic diagram diagram of of film film structure. structure.
The The morphology morphology of of the the sensing sensing films films was was studied studied by by SEM. SEM. The The crystalline crystalline structure structure of of the the films films and the average size of the crystallites was determined by XRD, with a Panalytical X-ray and the theaverage average size of the crystallites was determined by XRD, with a Panalytical X-ray size of the crystallites was determined by XRD, with a Panalytical X-ray diffractometer diffractometer for diffractometer for films films and and powders. powders. for films and powders.
Sensors 2017, 17, 1529 Sensors 2017, 17, 1529
4 of 9 4 of 9
Sensors 2017, 17, 1529
4 of 9
2.1. Sensor Sensor Structure Structure and and Testing Testing 2.2. 2.1.The Sensor finalStructure value of ofand theTesting central frequency frequency of of the the SAW SAW device device is is circa circa 69.5 69.5 MHz; MHz; this this is is aa decrease decrease of of The final value the central about 300 kHz from the initial value, before the film deposition. The SAW sensors (delay line type), The final value of the central frequency of the SAW device is circa 69.5 MHz; this is a decrease of about 300 kHz from the initial value, before the film deposition. The SAW sensors (delay line type), about 300 the initial value, before the filmpairs, deposition. SAW sensors type), consisted of two port electrodes with a The periodicity of 11 µm 3). The consisted of kHz twofrom portresonators resonatorswith with50 50 electrodes pairs, with a periodicity of(delay 11 (Figure µmline (Figure 3). consisted of two port resonators withand 50 electrodesµm pairs, with a periodicity of 11 The µm (Figure 3). The periodicity IDTs waswas 45 45 µm aperture. center-to-center The periodicitythe of the IDTs µm anda a2500 2500 µm wide wide acoustic acoustic aperture. The center-to-center periodicity of theIDTs IDTswas was10 45 mm. µm and a 2500 were µm wide acoustic aperture. The photolithographic center-to-center distance between The obtained using standard distance between IDTs was 10 mm. The IDTs IDTs were obtained using standard photolithographic distance between IDTs was 10 mm. The IDTs were obtained using standard photolithographic techniques, and and consist consist in 150-nm thick thick gold of aa 10 10 nm nm chromium chromium layer, layer, the the latter latter techniques, in aa 150-nm gold layer layer on on top top of techniques, and consistto inthe a 150-nm thick gold [27–30]. layer onThe top 10 of ×a 38 10 mm nm 22chromium layer, the latter ensuring good adhesion quartz substrate quartz substrate was cut at ensuring good adhesion to the quartz substrate [27–30]. The 10 × 38 mm2 quartz substrate was cut at ensuring good adhesion toreduce the quartz substrate [27–30]. The SAW 10 × 38reflection mm quartzfrom substrate was cut atthe a 45° angle, in order to the effect of spurious the edge of ◦ angle, in order to reduce the effect of spurious SAW reflection from the edge of the piezoelectric a 45 a 45° angle, in order to reduce the effect of spurious SAW reflection from the edge of the piezoelectric [31]. substrate [31].substrate piezoelectric substrate [31].
Figure 3. Scheme and dimension of SAW delay line. Figure3.3.Scheme Scheme and and dimension dimension of Figure ofSAW SAWdelay delayline. line.
The loss sensors was amplified (DHPVA-200 FEMTO The losssignal signalin theoscillation oscillation circuit circuit of FEMTO The loss signal ininthe the oscillation circuit of the the sensors sensors was wasamplified amplified(DHPVA-200 (DHPVA-200 FEMTO amplifier), and the system frequency shift was analyzed (CNT-91 Pendulum counter analyzer and amplifier), and the system frequency shift was analyzed (CNT-91 Pendulum counter analyzer and amplifier), and the system frequency shift was analyzed (CNT-91 Pendulum counter analyzer and Time View III software). A network /spectrum /impedance analyzer (Agilent 4396B) which includes Time View III software). A network /spectrum /impedance analyzer (Agilent 4396B) which includes Time View III software). A network /spectrum /impedance analyzer (Agilent 4396B) which includes a transmission/reflectiontest testkit kit(Agilent (Agilent 87512 87512 A/B) A/B) was values and aa transmission/reflection test kit (Agilent 87512 wasused usedtoto tooptimize optimizethe theinductor inductor values and transmission/reflection A/B) was used optimize the inductor values and measure the signal attenuation and phase. measure the signal attenuation and phase. measure the signal attenuation and phase. The performances of the sensors were determined at various hydrogen concentrations. Mass The were determined at various hydrogen concentrations. Mass flow The performances performancesofofthe thesensors sensors were determined at various hydrogen concentrations. Mass flow controllers (Figure 4) were used to combine the hydrogen gas mixture (2% H 2 /98% synthetic controllers (Figure 4) were used to combine the hydrogen gas mixture (2% H /98% synthetic air) with 2 flow controllers (Figure 4) were used to combine the hydrogen gas mixture (2% H2/98% synthetic air)synthetic with pure synthetic air, thus obtaining different gas concentrations. Theflow totalrate gas was flowmaintained rate was pure thus obtaining concentrations. The total gas air) with pure air, synthetic air, thus different obtaininggas different gas concentrations. The total gas flow rate was at a of constant value all of cases. 0.5 L/min in all cases. All sensorsatwere measured at room at amaintained constant at value 0.5 L/min All sensors were measured maintained a constant valueinof 0.5 L/min in all cases. All sensors room were temperature. measured at room temperature. temperature.
Figure 4. Experimental setup for SAW-sensor frequency shift measurements for hydrogen (H2) Figure 4. Experimental setup for SAW-sensor frequency shift measurements for hydrogen (H2 ) detection. detection. Figure 4. Experimental setup for SAW-sensor frequency shift measurements for hydrogen (H2) detection.
Sensors 2017, 17, 1529
5 of 9
Sensors 2017, 17, 1529 3. Results Sensors 2017, 17, 1529
5 of 9 5 of 9
3. Results 3.1. Film Morphology and Structure 3. Results The Morphology 400 mTorr gas 3.1. Film andpressure Structureused for laser ablation in our case leads to nanoporous films both in the case of ZnO and of Pd, as can be seen in the SEM images presented in Figure 5. Such relatively 3.1. Film Morphology andpressure Structureused for laser ablation in our case leads to nanoporous films both in mTorr gas high The laser400 deposition pressures are known to lead to films having a larger porosity [20]. This is due the case of ZnO and of Pd, as can be seen in the SEM images presented to in Figure 5. Such relatively The 400 of mTorr gas pressure used forthe laser ablation ingas, ourwhich case leads in to collisions the ablated species with background lead tonanoporous lowering of films their both kinetic high laser deposition pressures are knownin tothe lead to films having a larger porosity [20]. This is due the case and of ZnO of Pd, as can be seen SEM Figure 5. Such relatively energy to and cluster nucleation, producing filmsimages with apresented porous in morphology [32,33]. Such to collisions of the ablated species with the background gas, which lead to lowering of their kinetic nanoporous films lead to superior sensingtoproperties duehaving to rapid gas in/out diffusion andislarge high laser deposition pressures are known lead to films a larger porosity [20]. This due energy and to cluster nucleation, producing films with a porous morphology [32,33]. Such surface area.of the ablated species with the background gas, which lead to lowering of their kinetic to collisions nanoporous films lead to superior sensing properties due to rapid gas in/out diffusion and large energy and to cluster nucleation, producing films with a porous morphology [32,33]. Such nanoporous surface area. films lead to superior sensing properties due to rapid gas in/out diffusion and large surface area.
Figure 5. Scanning electron microscopy (SEM) images of the ZnO and Pd nanoporous films grown on the SAW sensors. Figure 5. and PdPd nanoporous films grown on Figure 5. Scanning Scanning electron electronmicroscopy microscopy(SEM) (SEM)images imagesofofthe theZnO ZnO and nanoporous films grown the SAW sensors. on the SAW sensors.the XRD results for the ZnO and Pd films. The patterns reveal the formation of Figure 6 presents
pure polycrystalline compounds, a stoichiometric wurtzite ZnO phase, and a fcc Pd phase, Figure 6 presents the XRD results for the ZnO and Pd films. The patterns reveal the formation of respectively. For both thin films, the mean crystallite size values were in the nanometer range: 34 nm pure polycrystalline a stoichiometric wurtzite ZnO phase, a fccand Pd phase, polycrystallinecompounds, compounds, a stoichiometric wurtzite ZnO and phase, a fcc respectively. Pd phase, for ZnO and 13 nm for Pd. The deposition conditions which were used circumvent the known respectively. both films, the mean sizein values were in the nanometer 34 nm For both thin For films, thethin mean crystallite sizecrystallite values were the nanometer range: 34 nmrange: for ZnO problem of potentially substantial Zn enrichment of the ablated target, which leadsand to for ZnO and 13 nm for Pd. The deposition conditions which were used circumvent the known 13 nm for Pd. The deposition conditions which were used circumvent the known problem of potentially nonstoichiometric PLD-deposited ZnO films [34], so that in our case the ZnO films are problem ofZnpotentially substantial enrichment of tothe ablated target, which leadsZnO to substantial enrichment of the ablatedZn target, which leads nonstoichiometric PLD-deposited stoichiometric. nonstoichiometric [34], so that in our case the ZnO films are films [34], so that inPLD-deposited our case the ZnOZnO filmsfilms are stoichiometric. stoichiometric.
Figure 6. XRD spectra spectra of of the the of of ZnO ZnO and and Pd Pd films Figure 6. XRD films on on quartz quartz substrates. substrates.
Figure 6. XRD spectra of the of ZnO and Pd films on quartz substrates.
Sensors 2017, 17, 1529 Sensors 2017, 17, 1529
6 of 9 6 of 9
3.2. Sensor Sensor Properties Properties 3.2. The SAW SAW sensor sensor response response was was measured measured for for the the Pd, Pd, ZnO, ZnO, and andPd/ZnO Pd/ZnO sensitive sensitive layers layers for for The hydrogen concentrations up to 2% (half of the safety limit for this gas). Figure 7 shows the response hydrogen concentrations up to 2% (half of the safety limit for this gas). Figure 7 shows the response of of the SAW sensor at different concentrations, at room temperature; experimental point is the SAW sensor at different gas gas concentrations, at room temperature; eacheach experimental point is the the average result of five measurements. It was observed that the frequency shift Δf is proportional average result of five measurements. It was observed that the frequency shift ∆f is proportional to the to the concentration for all for sensors a gas concentration between 2000 and These concentration for all sensors a gasfor concentration between 2000 and 20000 ppm.20000 Theseppm. results are results are similar to those reported by Phan and Chung in the same concentration domain, for room similar to those reported by Phan and Chung in the same concentration domain, for room temperature temperature sensor operation [35]. sensor operation [35].
Figure Frequency shift Figure 7. 7. Frequency shift dependence dependence of of hydrogen hydrogen concentration concentration at at RT. RT.
Table 11 presents presents the the sensitivities sensitivities and and detection detection limits limits for for the the sensitive sensitive layers layers of of Pd, Pd, ZnO, ZnO, and and Table Pd/ZnO. The limit limit ofofdetection detection(LOD) (LOD)depends depends defined 3× Pd/ZnO. on on the the noisenoise level,level, beingbeing defined as 3× as noise noise level/sensitivity. noise level estimated around theallallthe thefilms. films. The The noise noise level/sensitivity. TheThe noise level waswas estimated at at around 3030 HzHz forfor the assessment was performed the frequency fluctuation over 10 assessment performed in inair air(without (withoutanalyte) analyte)bybymeasuring measuring the frequency fluctuation over min; it it represents (best-fit line). line). The The 10 min; representsthe themaximum maximumfrequency frequencydeviation deviationfrom from the the trend line (best-fit sensitivity, defined as the frequency shift in Hz per unit analyte concentration in ppm, was sensitivity, defined as the frequency shift in Hz per unit analyte concentration in ppm, was determined determined from sensitivity an averagevalue sensitivity for a gas concentration between 0.2%The andresponse 2%. The from an average for a value gas concentration between 0.2% and 2%. response time (to 90% of maximum signal) for the tested sensorswas tested was between for time (to reach 90%reach of maximum signal) for the sensors between 12 and 12 16 and s for16s these these hydrogen concentrations. hydrogen concentrations. that thethe LOD of Pd/ZnO filmsfilms was was ~4.5 times betterbetter than the single The results in in table Table1 1indicate indicate that LOD of Pd/ZnO ~4.5 times than the ZnO films almost 2-times 2-times better than thethan single Pdsingle layers.Pd Inlayers. most cases, SAW sensors aresensors heated single ZnOand films and almost better the In most cases, SAW in order to in improve theimprove hydrogen [10,36]. Our[10,36]. results are for obtained room temperature are heated order to theresponse hydrogen response Ourobtained results are for room operation, and are better than comparable to those reported byreported other groups in such conditions temperature operation, and areor better than or comparable to those by other groups in such [11,12,35,37]. conditions [11,12,35,37]. Table1.1.Sensitivity Sensitivityand andlimit limit detection (LOD) of the sensors towards hydrogen = frequency Table of of detection (LOD) of the sensors towards hydrogen (∆f =(Δf frequency shift; cshift; = hydrogen concentration; n = noise level).level). c = hydrogen concentration; n = noise
Coating Material Sensitivity Δf/c (Hz/ppm) Coating Material Sensitivity ∆f/c (Hz/ppm) Pd 0.29 Pd 0.29 ZnO 0.15 ZnO 0.15 Pd/ ZnO 0.51 Pd/ZnO 0.51
LOD (3*n)/(Δf/c) (ppm) LOD (3*n)/(∆f/c) (ppm) 105 105 261 261 59 59
Sensors 2017, 17, 1529
7 of 9
4. Discussion and Conclusions When a bilayer (metal/semiconductor) SAW sensor is exposed to air, oxygen is adsorbed and dissociates at the metal surface. The resulting atomic oxygen diffuses to the metal/semiconductor interface, where it captures electrons and increases the depletion layer width in the semiconductor. Subsequent exposure of the bilayer sensor to H2 similarly results in the formation of atomic hydrogen which diffuses to the interface, reacting with the previously generated oxygen species, and releasing electrons into the conduction band of the semiconductor. The increase in sheet electron conductivity will lead to a decrease in acoustic velocity, since the SAW wave is slowed down by the presence of carriers through the acoustoelectric effect [38,39]. Surface acoustic wave propagation is very sensitive to surface perturbations such as small changes in electrical conductivity. The acoustoelectric effect in the case of a single sensitive layer leads to a relative change of wave propagation velocity ∆ν given by ∆f ∆ν ∼ K2 σs2 = = 2 f0 ν0 2 σs + ν0 Cs2
(1)
where ∆f is the frequency shift, f0 the central frequency of the SAW device, ν0 the unperturbed SAW velocity, K the electromechanical coefficient, Cs = ε0 + εp is the sum of the permittivity of the region above the film and the substrate, and σs is the surface conductivity of the sensing film [11,18]. In the case of bilayers, important additional parameters which determine velocity changes are the electrical conductivities of both layers and their thicknesses, since both film layers influence the SAW propagation through interaction between their mobile charges and the travelling electron potential ϕ of the SAW [17]. These mechanisms explain the improvement of the sensor properties in the case of the Pd/ZnO bilayer compared to the single Pd and ZnO layers. The hydrogen absorbtion and diffusion properties of Pd, combined with the excellent diffusion properties of the nanoporous Pd film, cause increased amounts of hydrogen to reach the Pd/ZnO interface. At this interface, the large sensitivity of ZnO surface conductivity to adsorbed species brings about increased hydrogen detection sensitivity. The sensor response of the bilayer Pd/ZnO thin film proposed by us thus combines both mass and acoustoelectric effects, leading to improved sensitivity to hydrogen. The mass effect of the gas on the sensing layer is known to always lead to a decrease of the center frequency of the SAW sensor. The variation of the electrical conductivity, on the other hand, can lead both to an increase and to a decrease of the center frequency [40]. In order for the combination of the mass and acoustoelectric effects to be efficient, the latter must lead to the same result, i.e., a decrease of the SAW center frequency. Otherwise, the two effects counteract each other and the combined response is smaller. As outlined above, this is the case for our sensor structure, since hydrogen is a reducing gas which reacts with the adsorbed oxygen species, releasing electrons into ZnO, increasing the surface conductivity and decreasing the center frequency. We consider that further optimization of the deposition conditions and sensor structure can lead to further improvement of the hydrogen sensing properties, and perhaps to applications for other gases. In conclusion, a SAW bilayer structure consisting in porous Pd and ZnO layers has been obtained by PLD. The bilayer combines the excellent hydrogen absorbtion and dissociation properties with the sensitivity of surface electron conductivity of ZnO to adsorbed species. The SAW sensor based on this bilayer structure has an improved sensor response due to a combination of the mass and acoustoelectric effects appearing in the presence of hydrogen. The LOD of Pd/ZnO films is about 4.5 times better than the single ZnO films and almost twice better than the single Pd layers, reaching a value of 59 ppm. To our knowledge, it is for the first time that this material combination has been used for bilayers in hydrogen sensors based on surface acoustic waves. Optimization of the bilayer structure can lead to further improvement of the bilayer Pd/ZnO sensor properties, and future research will be addressed in this direction.
Sensors 2017, 17, 1529
8 of 9
Acknowledgments: This work was supported by a grant of the Romanian National Authority for Scientific Research and Innovation, CNCS-UEFISCDI, projects number PNII-RU-TE-2014-4-0342 and project NUCLEU 16470104. Author Contributions: Cristian Viespe and Dana Miu conceived and designed the experiments and analyzed the data; Cristian Viespe performed the experiments; Dana Miu wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4.
5. 6. 7. 8. 9. 10. 11.
12.
13. 14. 15. 16. 17.
18.
19.
Jacobson, M.Z.; Colella, W.G.; Golden, D.M. Clearing the Air and Improving Health with Hydrogen Fuel-Cell Vehicles. Science 2005, 308, 1901–1905. [CrossRef] [PubMed] Bassam, A.M.; Phillips, A.B.; Turnock, S.R.; Wilson, P.A. Development of a multi-scheme energy management strategy for a hybrid fuel cell driven passenger ship. Int. J. Hydrog. Energy 2017, 42, 623–635. [CrossRef] Grimes, C.A.; Ong, K.G.; Varghese, O.K.; Yang, X.; Mor, G.; Paulose, M.; Dickey, E.C.; Ruan, C.; Pishko, M.V.; Kendig, J.W.; et al. A Sentinel Sensor Network for Hydrogen Sensing. Sensors 2003, 3, 69–82. [CrossRef] Boon-Brett, L.; Bousk, J.; Black, G.; Moretto, P.; Castello, P.; Hübert, T.; Banach, U. Identifying performance gaps in hydrogen safety sensor technology for automotive and stationary applications. Int. J. Hydrog. Energy 2010, 35, 373–384. [CrossRef] Buttner, W.J.; Post, M.B.; Burgess, R.; Rivkin, C. An overview of hydrogen safety sensors and requirements. Int. J. Hydrog. Energy 2011, 36, 2462–2470. [CrossRef] Hubert, T.; Boon-Brett, L.; Black, G.; Banach, U. Hydrogen sensors—A review. Sens. Actuators B Chem. 2011, 157, 329–352. [CrossRef] Cole, M.; Spulber, I.; Gardner, J. Surface acoustic wave electronic tongue for robust analysis of sensory components. Sens. Actuators B Chem. 2015, 207, 1147–1153. [CrossRef] Bhasker, R.V.; Nimal, A.T.; Parmar, Y.; Sharma, M.U.; Sreenivas, K.; Gupta, V. Cross sensitivity and selectivity studies on ZnO surface acoustic wave ammonia sensor. Sens. Actuators B Chem. 2010, 147, 517–524. Jakubik, W. Hydrogen gas-sensing with bilayer structures of WO3 and Pd in SAW and electric systems. Thin Solid Film 2009, 517, 6188–6191. [CrossRef] Ippolito, S.J.; Kandasamy, S.; Kalentar-Zadeh, K.; Wlodarski, W. Layered SAW hydrogen sensor with modified tungsten trioxide selective layer. Sens. Actuators B Chem. 2005, 108, 553–557. [CrossRef] Yang, L.; Yin, C.; Zhang, Z.; Zhou, J.; Xu, H. The investigation of hydrogen gas sensing properties of SAW gas sensor based on palladium surface modified SnO2 thin film. Mater. Sci. Semicond. Process. 2017, 60, 16–28. [CrossRef] Jakubik, W.; Urbanczyk, M.; Kochowski, S.; Bodzenta, J. Palladium and phthalocynine bilayer films for hydrogen detection in a surface acoustic wave sensor system. Sens. Actuators B Chem. 2003, 96, 321–328. [CrossRef] Jakubik, W. Elementary theory of SAW gas sensor based on electrical conductivity changes in bi-layer nanostructures. Sens. Actuators B Chem. 2014, 203, 511–516. [CrossRef] Jakubik, W.; Krzywiecki, M.; Maciak, E.; Urbanczyk, ´ M. Bi-layer nanostructures of CuPc and Pd for resistance-type and SAW-type hydrogen gas sensors. Sens. Actuators B Chem. 2012, 175, 255–262. [CrossRef] Jakubik, W.; Powroznik, P.; Wrotniak, J.; Krzywiecki, M. Theoretical analysis of acoustoelectrical sensitivity in SAW gas sensors with single and bi-layer structures. Sens. Actuators B Chem. 2016, 236, 1069–1074. [CrossRef] Li, R.; Reyes, P.I.; Ragavendiran, S.; Shen, H.; Lu, Y. Tunable surface acoustic wave device based on acoustoelectric interaction in ZnO/GaN heterostructures. Appl. Phys. Lett. 2015, 107, 073504. [CrossRef] Dutheil, P.J.; Orlianges, C.; Crunteanu, A.; Catherinot, A.; Champeaux, C. AlN, ZnO thin films and AlN/ZnO or ZnO/AlN mulatilayer structures deposited by PLD for surface acoustic wave applications. Phys. Status Solidi A 2015, 212, 817–825. [CrossRef] Wang, S.Y.; Ma, J.Y.; Li, Z.J.; Su, H.Q.; Alkurd, N.R.; Zhou, W.L.; Wang, L.; Du, B.; Tang, Y.L.; Ao, D.Y.; et al. Surface acoustic wave ammonia sensor based on ZnO/SiO2 composite film. J. Hazard. Mater. 2015, 285, 368–374. [CrossRef] [PubMed] Noh, J.S.; Lee, J.M.; Lee, W. Low-dimensional Palladium nanostructures for fast and reliable Hydrogen gas detection. Sensors 2011, 11, 825–851. [CrossRef] [PubMed]
Sensors 2017, 17, 1529
20. 21. 22. 23. 24.
25.
26. 27. 28. 29. 30. 31. 32.
33. 34. 35. 36. 37. 38. 39. 40.
9 of 9
Viespe, C.; Grigoriu, C. SAW sensor based on highly sensitive nanoporous palladium thin film for hydrogen detection. Microelectron. Eng. 2013, 108, 218–221. [CrossRef] Janotti, A.; Van de Walle, C.G. Fundamentals of zinc oxide as a semiconductor. Rep. Prog. Phys. 2009, 72, 126501. [CrossRef] Wu, T.T.; Wang, W.S. An experimental study on the ZnO/sapphire layered surface acoustic wave device. J. Appl. Phys. 2004, 96, 5249. [CrossRef] Schmidt, O.; Kiesel, P.; Van de Walle, C.G.; Johnson, N.M.; Nause, J.; Döhler, G.H. Effects of an electrically conducting layer on the zinc oxide surface. Jpn. J. Appl. Phys. 2005, 44, 7271–7274. [CrossRef] Schmidt, O.; Geis, A.; Kiesel, P.; Van de Walle, C.G.; Johnson, N.M.; Bakin, A.; Waag, A.; Döhler, G.H. Analysis of a conducting channel at the native zinc oxide surface. Superlattice. Microstruct. 2006, 39, 8–16. [CrossRef] Ballantine, D.S.; White, R.M.; Martin, S.I.; Ricco, A.J.; Zellers, E.T.; Frye, G.C.; Wohltjen, H. Acoustic Wave Sensors, Theory, Design and Physico-Chemical Applications; Academic Press: San Diego, CA, USA, 1997; pp. 337–339. Chrisey, D.B.; Hubler, G.K. Pulsed Laser Deposition of Thin Films; Wiley: New York, NY, USA, 1994. Marcu, A.; Viespe, C. Laser-grown ZnO Nanowires for Room-temperature SAW-sensor Applications. Sens. Actuators B Chem. 2015, 208, 1–6. [CrossRef] Viespe, C. Surface Acoustic Wave Sensors based on Nanoporous Films for Hydrogen Detection. Mater. Appl. Sens. Transducer. Key Eng. Mater. 2014, 605, 331–334. [CrossRef] Marcu, A.; Viespe, C. Active surface geometrical control of noise in nanowire-SAW sensors. Sens. Actuators B Chem. 2016, 231, 469–473. [CrossRef] Marcu, A.; Viespe, C. Surface Acoustic Wave Sensors for Hydrogen and Deuterium Detection. Sensors 2017, 17, 1417. [CrossRef] [PubMed] Campbell, C. Surface Acoustic Wave Devices and THEIR Signal Processing Applications; Academic Press: San Diego, CA, USA, 1989; p. 26. Di Fonzo, F.; Tonini, D.; Li Bessi, A.; Casari, C.S.; Beghi, M.G.; Bottani, C.E.; Gastaldi, D.; Vena, P.; Contro, R. Growth regimes in pulsed laser deposition of aluminum oxide thin films. Appl. Phys. A 2008, 93, 765–769. [CrossRef] Riabinina, D.; Irissou, E.; Le Drogoff, B.; Chaker, M.; Guay, D. Influence of pressure on the Pt nanoparticle growth modes during pulsed laser ablation. J. Appl. Phys. 2010, 108, 034322. [CrossRef] Claeyssens, F.; Cheesman, A.; Henley, S.J.; Ashfold, M.N. Studies of the plume accompanying pulsed ultraviolet laser ablation of zinc oxide. J. Appl. Phys. 2002, 92, 6886–6894. [CrossRef] Phan, D.; Chung, G. Surface acoustic wave hydrogen sensors based on ZnO nanoparticles incorporated with a Pt catalyst. Sens. Actuators B Chem. 2012, 161, 341–348. [CrossRef] Rane, S.; Arbuj, S.; Rane, S.; Gosavi, S. Hydrogen sensing characteristics of Pt-SnO2 nano-structured composite thin films. J. Mater. Sci. Mater. Electron. 2015, 26, 3707–3716. [CrossRef] Jakubik, W.P.; Urbanczyk, M.; Mariak, E.; Pustelny, T. Bilayer structures of NiOx and Pd in Surface Acoustic Wave and electrical gas sensor systems. Acta Phys. Polonica A 2009, 116, 315–320. [CrossRef] Ricco, A.J.; Martin, S.J.; Zipperian, T.E. Surface acoustic wave gas sensor based on film conductivity changes. Sens. Actuators 1985, 8, 319–333. [CrossRef] Fischer, B.H.; Malocha, D.C. Study of the acoustoelectric effect for SAW sensors. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2010, 57, 698–706. [CrossRef] [PubMed] Fan, H.; Ge, H.; Zhang, S.; Zhang, H.; Zhu, J. Optimization of sensitivity induced by surface conductivity and sorbed mass in SAW gas sensors. Sens. Actuators B Chem. 2012, 161, 114–123. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Surface Acoustic Wave Sensor with Pd/ZnO Bilayer Structure for Room Temperature Hydrogen Detection Cristian Viespe
and Dana Miu *
National Institute for Laser, Plasma and Radiation Physics, Laser Department, Atomistilor # 409, 077125 Bucharest-Magurele, Romania; [email protected] * Correspondence: [email protected]; Tel.: +40-21-457-4027 Received: 6 June 2017; Accepted: 28 June 2017; Published: 29 June 2017
Abstract: A Surface Acoustic Wave (SAW) hydrogen sensor with a Pd/ZnO bilayer structure for room temperature sensing operation has been obtained by Pulsed Laser Deposition (PLD). The sensor structure combines a Pd layer with optimized porosity for maximizing mass effects, with the large acoustoelectric effect at the Pd/ZnO interface. The large acoustoelectric effect is due to the fact that ZnO has a surface conductivity which is highly sensitive to chemisorbed gases. The sensitivity of the sensor was determined for hydrogen concentrations between 0.2% and 2%. The limit of detection (LOD) of the bilayer sensor was about 4.5 times better than the single ZnO films and almost twice better than single Pd films. Keywords: SAW sensor; hydrogen sensor; bilayer; pulsed laser deposition; nanoporous film; Palladium; ZnO
1. Introduction Hydrogen has many important applications such as alternative, clean energy sources, propulsion systems, and biomedical devices [1–3]. However, serious problems are implied by hydrogen use: in concentrations over 4.6% it is highly explosive [4], and over 4% it is extremely flammable [5]. The additional fact that hydrogen is colorless and odorless makes the need for sensors capable of detecting concentrations of 2% or less very important. Sensors based on resistance, MOS-field effect transistors, or optical devices have been used for hydrogen detection, each having its advantages and disadvantages [6]. Surface acoustic wave (SAW) sensors have been widely used as hydrogen sensors due to advantages such as high sensitivity, fast response, reliability, and low cost [6–8]. Detection limits for H2 gas of 1250 to 2000 ppm (operating temperatures 80–120 ◦ C) [9], 600 ppm (at 80 ◦ C) [10] or as low as 200 ppm (at 175 ◦ C) [11] using SAW sensors have been reported. In SAW sensors, the sensing mechanism is based on the perturbation of the surface acoustic wave propagation by mechanical and/or electrical effects. From a practical standpoint, the relevant effects for SAW sensing are changes in mass density and in electrical conductivity. In certain conditions, the changes in electrical conductivity can lead to a much stronger SAW sensor response than the changes in mass density [12]. This is because the acoustoelectric interactions between the electrical potential associated with surface acoustic wave propagation on a piezoelectric crystal and the mobile electric charges in the sensing film can lead to large modifications of the SAW attenuation and propagation velocity [13]. Bilayer SAW structures have been proven to increase the sensor sensitivity compared to single layer ones, and to decrease the detection limit, even in the case of room temperature operation [14,15]. By using bilayer films, the absolute value of the multilayer film conductivity can be adjusted to operate in a region where the acoustoelectric effect is efficient. Various material combinations such as Pd/WO3 , ZnO/GaN, AlN/ZnO, or ZnO/SiO2 have been explored for SAW gas sensors based on this
Sensors 2017, 17, 1529; doi:10.3390/s17071529
www.mdpi.com/journal/sensors
Sensors 2017, 17, 1529
2 of 9
effect [9,16–18]. In the case of hydrogen sensing, an efficient acoustoelectric effect is desirable, since the mass effect for hydrogen is smaller than for heavier gases [9]. We propose a SAW bilayer structure for the detection of hydrogen consisting in Pd/ZnO, which, to our best knowledge, has not been investigated. Pd has excellent hydrogen absorption properties, having the highest hydrogen solubility of any element at atmospheric pressure as well as thermodynamically favored H2 dissociation [19]. The absorbing properties are improved even more in the case of nanoporous Pd layers, as we have shown in previous experiments [20]. ZnO is a wide band-gap semiconductor which is simple, reliable, low-cost, and easily mass-produced [21]. It is a piezoelectric material with a high coupling coefficient K, which is a measure of the conversion efficiency between electrical and acoustic energy [22]. When a semiconducting ZnO layer is deposited on a piezoelectric substrate, its mobile carriers couple to the electric field associated with SAW propagation along the piezoelectric substrate, leading to SAW attenuation and velocity change [16]. An important characteristic of ZnO for our case is the strong sensitivity of its surface conductivity to adsorbed species [23,24], which will increase the overall sensitivity of the bilayer sensor to hydrogen. It is worth mentioning that ST quartz substrates, which are used in our case, have a low temperature coefficient in comparison to other materials used for SAW substrates such as LiNbO3 or LiTaO3 , and are therefore less sensitive to temperature variations [25]. We have therefore considered a bilayer structure consisting in a combination of a Pd layer with optimized porosity for maximizing mass effects, and a ZnO layer having an electron conductivity which is very sensitive to the exposure of the surface to hydrogen. Thus, both changes in mass density and in electrical conductivity contribute to the sensor response, leading to a SAW sensor with large hydrogen sensitivity. 2. Materials and Methods 2.1. Film Deposition and Characterization The sensitive layers of the SAW sensors were deposited onto ST-X quartz substrates by pulsed laser deposition (PLD). PLD was used because it presents a series of advantages in thin film deposition in general [17,26], and in particular has been proven to allow good control of sensitive film porosity, which is essential for SAW gas sensors [20]. The beam of a Nd:YVO4 laser is focused onto the surface of the target, leading to ablation of the target material and deposition onto a substrate placed at a distance of 35 mm in front of the target and parallel to its surface (Figure 1). The laser generates pulses of 10 ps duration, at a 10 kHz repetition rate, and a wavelength of 532 nm. The targets and substrate are placed in a vacuum chamber equipped with a system for the control of the gas pressure and flow. Mass flow meters are placed on the gas bottles and controlled by a mass flow controller, combined with a throttle valve placed on a rotary vane vacuum pump controlled by a pressure controller maintain the desired pressure in the ablation chamber and a controlled gas flow. Targets are placed on computer-controlled x-y tables which ensure continuous target movement during deposition, avoiding target erosion which leads to large film roughness. The system also permits deposition from multiple targets for multilayers. The laser energy density on the target can be modified by changing the distance between the focusing lens and the target.
Sensors 2017, 17, 1529 Sensors Sensors 2017, 2017, 17, 17, 1529 1529
3 of 9 33 of of 99
Figure 1. Experimental layer deposition. deposition. Figure 1. Experimental setup setup of of sensitive sensitive layer
In In order order to to compare compare the the sensing sensing properties properties of of single single layer layer and and bilayer bilayer SAW SAW sensors, sensors, we we have have In order to compare the sensing properties of single layer and bilayer SAW sensors, we have deposited aa series of films all having aa thickness of 330 nm (Figure 2), as measured using aa QUANTA deposited series of films all having thickness of 330 nm (Figure 2), as measured using QUANTA deposited a series of films all having a thickness of 330 nm (Figure 2), as measured using a QUANTA EE == 50 keV Electron Microscope (SEM) and a Tokyo Seimitsu Surfcom 130 A profilometer. keV Scanning Scanning E = 50 50 keV Scanning Electron Electron Microscope Microscope (SEM) (SEM) and and aa Tokyo Tokyo Seimitsu Seimitsu Surfcom Surfcom 130 130 A A profilometer. profilometer. The bilayer film has aa thinner metal layer (Pd) on top of the semiconducting layer (ZnO) since this is The bilayer film has thinner metal layer (Pd) on top of the semiconducting layer (ZnO) since this is The bilayer film has a thinner metal layer (Pd) on top of the semiconducting layer (ZnO) since this known to lead to aa higher sensitivity [13,15]. In addition, the very good H 22 absorption properties of known to lead to higher sensitivity [13,15]. In addition, the very good H absorption properties of is known to lead to a higher sensitivity [13,15]. In addition, the very good H2 absorption properties Pd ensure penetration of gas species to ZnO layer. The Pd, and thin were PdPd ensure penetration of the the gasgas species to the the ZnOZnO layer. The The Pd, ZnO, ZnO, and Pd/ZnO Pd/ZnO thin films films were of ensure penetration of the species to the layer. Pd, ZnO, and Pd/ZnO thin films directly deposited onto the quartz substrates at an average laser power of 0.7 W in aa gas flow of 0.5 sccm. directly deposited onto the quartz substrates at an average laser power of 0.7 W in gas flow of 0.5 sccm. were directly deposited onto the quartz substrates at an average laser power of 0.7 W in a gas flow of Before Before deposition, deposition, the the chamber chamber was was evacuated evacuated to to aa base base pressure pressure of of 10 10-5-5 Torr. Torr. In In the the case case of of Pd, Pd, films films 0.5 sccm. Before deposition, the chamber was evacuated to a base pressure of 10−5 Torr. In the case of were were obtained obtained by by ablation ablation of of aa pure pure Pd Pd target target (99.95%) (99.95%) in in an an Ar Ar atmosphere atmosphere of of 400 400 mTorr. mTorr. The The ZnO ZnO Pd, films were obtained by ablation of a pure Pd target (99.95%) in an Ar atmosphere of 400 mTorr. films films were were obtained obtained by by ablation ablation of of aa ZnO ZnO target target (99.95%) (99.95%) in in O O22 at at 400 400 mTorr. mTorr. All All depositions depositions were were The ZnO films were obtained by ablation of a ZnO target (99.95%) in O2 at 400 mTorr. All depositions made made at at room room temperature, temperature, and and no no post-deposition post-deposition thermal thermal treatments treatments were were used. used. were made at room temperature, and no post-deposition thermal treatments were used.
Figure 2. Schematic diagram of film structure. Figure Figure 2. 2. Schematic Schematic diagram diagram of of film film structure. structure.
The The morphology morphology of of the the sensing sensing films films was was studied studied by by SEM. SEM. The The crystalline crystalline structure structure of of the the films films and the average size of the crystallites was determined by XRD, with a Panalytical X-ray and the theaverage average size of the crystallites was determined by XRD, with a Panalytical X-ray size of the crystallites was determined by XRD, with a Panalytical X-ray diffractometer diffractometer for diffractometer for films films and and powders. powders. for films and powders.
Sensors 2017, 17, 1529 Sensors 2017, 17, 1529
4 of 9 4 of 9
Sensors 2017, 17, 1529
4 of 9
2.1. Sensor Sensor Structure Structure and and Testing Testing 2.2. 2.1.The Sensor finalStructure value of ofand theTesting central frequency frequency of of the the SAW SAW device device is is circa circa 69.5 69.5 MHz; MHz; this this is is aa decrease decrease of of The final value the central about 300 kHz from the initial value, before the film deposition. The SAW sensors (delay line type), The final value of the central frequency of the SAW device is circa 69.5 MHz; this is a decrease of about 300 kHz from the initial value, before the film deposition. The SAW sensors (delay line type), about 300 the initial value, before the filmpairs, deposition. SAW sensors type), consisted of two port electrodes with a The periodicity of 11 µm 3). The consisted of kHz twofrom portresonators resonatorswith with50 50 electrodes pairs, with a periodicity of(delay 11 (Figure µmline (Figure 3). consisted of two port resonators withand 50 electrodesµm pairs, with a periodicity of 11 The µm (Figure 3). The periodicity IDTs waswas 45 45 µm aperture. center-to-center The periodicitythe of the IDTs µm anda a2500 2500 µm wide wide acoustic acoustic aperture. The center-to-center periodicity of theIDTs IDTswas was10 45 mm. µm and a 2500 were µm wide acoustic aperture. The photolithographic center-to-center distance between The obtained using standard distance between IDTs was 10 mm. The IDTs IDTs were obtained using standard photolithographic distance between IDTs was 10 mm. The IDTs were obtained using standard photolithographic techniques, and and consist consist in 150-nm thick thick gold of aa 10 10 nm nm chromium chromium layer, layer, the the latter latter techniques, in aa 150-nm gold layer layer on on top top of techniques, and consistto inthe a 150-nm thick gold [27–30]. layer onThe top 10 of ×a 38 10 mm nm 22chromium layer, the latter ensuring good adhesion quartz substrate quartz substrate was cut at ensuring good adhesion to the quartz substrate [27–30]. The 10 × 38 mm2 quartz substrate was cut at ensuring good adhesion toreduce the quartz substrate [27–30]. The SAW 10 × 38reflection mm quartzfrom substrate was cut atthe a 45° angle, in order to the effect of spurious the edge of ◦ angle, in order to reduce the effect of spurious SAW reflection from the edge of the piezoelectric a 45 a 45° angle, in order to reduce the effect of spurious SAW reflection from the edge of the piezoelectric [31]. substrate [31].substrate piezoelectric substrate [31].
Figure 3. Scheme and dimension of SAW delay line. Figure3.3.Scheme Scheme and and dimension dimension of Figure ofSAW SAWdelay delayline. line.
The loss sensors was amplified (DHPVA-200 FEMTO The losssignal signalin theoscillation oscillation circuit circuit of FEMTO The loss signal ininthe the oscillation circuit of the the sensors sensors was wasamplified amplified(DHPVA-200 (DHPVA-200 FEMTO amplifier), and the system frequency shift was analyzed (CNT-91 Pendulum counter analyzer and amplifier), and the system frequency shift was analyzed (CNT-91 Pendulum counter analyzer and amplifier), and the system frequency shift was analyzed (CNT-91 Pendulum counter analyzer and Time View III software). A network /spectrum /impedance analyzer (Agilent 4396B) which includes Time View III software). A network /spectrum /impedance analyzer (Agilent 4396B) which includes Time View III software). A network /spectrum /impedance analyzer (Agilent 4396B) which includes a transmission/reflectiontest testkit kit(Agilent (Agilent 87512 87512 A/B) A/B) was values and aa transmission/reflection test kit (Agilent 87512 wasused usedtoto tooptimize optimizethe theinductor inductor values and transmission/reflection A/B) was used optimize the inductor values and measure the signal attenuation and phase. measure the signal attenuation and phase. measure the signal attenuation and phase. The performances of the sensors were determined at various hydrogen concentrations. Mass The were determined at various hydrogen concentrations. Mass flow The performances performancesofofthe thesensors sensors were determined at various hydrogen concentrations. Mass flow controllers (Figure 4) were used to combine the hydrogen gas mixture (2% H 2 /98% synthetic controllers (Figure 4) were used to combine the hydrogen gas mixture (2% H /98% synthetic air) with 2 flow controllers (Figure 4) were used to combine the hydrogen gas mixture (2% H2/98% synthetic air)synthetic with pure synthetic air, thus obtaining different gas concentrations. Theflow totalrate gas was flowmaintained rate was pure thus obtaining concentrations. The total gas air) with pure air, synthetic air, thus different obtaininggas different gas concentrations. The total gas flow rate was at a of constant value all of cases. 0.5 L/min in all cases. All sensorsatwere measured at room at amaintained constant at value 0.5 L/min All sensors were measured maintained a constant valueinof 0.5 L/min in all cases. All sensors room were temperature. measured at room temperature. temperature.
Figure 4. Experimental setup for SAW-sensor frequency shift measurements for hydrogen (H2) Figure 4. Experimental setup for SAW-sensor frequency shift measurements for hydrogen (H2 ) detection. detection. Figure 4. Experimental setup for SAW-sensor frequency shift measurements for hydrogen (H2) detection.
Sensors 2017, 17, 1529
5 of 9
Sensors 2017, 17, 1529 3. Results Sensors 2017, 17, 1529
5 of 9 5 of 9
3. Results 3.1. Film Morphology and Structure 3. Results The Morphology 400 mTorr gas 3.1. Film andpressure Structureused for laser ablation in our case leads to nanoporous films both in the case of ZnO and of Pd, as can be seen in the SEM images presented in Figure 5. Such relatively 3.1. Film Morphology andpressure Structureused for laser ablation in our case leads to nanoporous films both in mTorr gas high The laser400 deposition pressures are known to lead to films having a larger porosity [20]. This is due the case of ZnO and of Pd, as can be seen in the SEM images presented to in Figure 5. Such relatively The 400 of mTorr gas pressure used forthe laser ablation ingas, ourwhich case leads in to collisions the ablated species with background lead tonanoporous lowering of films their both kinetic high laser deposition pressures are knownin tothe lead to films having a larger porosity [20]. This is due the case and of ZnO of Pd, as can be seen SEM Figure 5. Such relatively energy to and cluster nucleation, producing filmsimages with apresented porous in morphology [32,33]. Such to collisions of the ablated species with the background gas, which lead to lowering of their kinetic nanoporous films lead to superior sensingtoproperties duehaving to rapid gas in/out diffusion andislarge high laser deposition pressures are known lead to films a larger porosity [20]. This due energy and to cluster nucleation, producing films with a porous morphology [32,33]. Such surface area.of the ablated species with the background gas, which lead to lowering of their kinetic to collisions nanoporous films lead to superior sensing properties due to rapid gas in/out diffusion and large energy and to cluster nucleation, producing films with a porous morphology [32,33]. Such nanoporous surface area. films lead to superior sensing properties due to rapid gas in/out diffusion and large surface area.
Figure 5. Scanning electron microscopy (SEM) images of the ZnO and Pd nanoporous films grown on the SAW sensors. Figure 5. and PdPd nanoporous films grown on Figure 5. Scanning Scanning electron electronmicroscopy microscopy(SEM) (SEM)images imagesofofthe theZnO ZnO and nanoporous films grown the SAW sensors. on the SAW sensors.the XRD results for the ZnO and Pd films. The patterns reveal the formation of Figure 6 presents
pure polycrystalline compounds, a stoichiometric wurtzite ZnO phase, and a fcc Pd phase, Figure 6 presents the XRD results for the ZnO and Pd films. The patterns reveal the formation of respectively. For both thin films, the mean crystallite size values were in the nanometer range: 34 nm pure polycrystalline a stoichiometric wurtzite ZnO phase, a fccand Pd phase, polycrystallinecompounds, compounds, a stoichiometric wurtzite ZnO and phase, a fcc respectively. Pd phase, for ZnO and 13 nm for Pd. The deposition conditions which were used circumvent the known respectively. both films, the mean sizein values were in the nanometer 34 nm For both thin For films, thethin mean crystallite sizecrystallite values were the nanometer range: 34 nmrange: for ZnO problem of potentially substantial Zn enrichment of the ablated target, which leadsand to for ZnO and 13 nm for Pd. The deposition conditions which were used circumvent the known 13 nm for Pd. The deposition conditions which were used circumvent the known problem of potentially nonstoichiometric PLD-deposited ZnO films [34], so that in our case the ZnO films are problem ofZnpotentially substantial enrichment of tothe ablated target, which leadsZnO to substantial enrichment of the ablatedZn target, which leads nonstoichiometric PLD-deposited stoichiometric. nonstoichiometric [34], so that in our case the ZnO films are films [34], so that inPLD-deposited our case the ZnOZnO filmsfilms are stoichiometric. stoichiometric.
Figure 6. XRD spectra spectra of of the the of of ZnO ZnO and and Pd Pd films Figure 6. XRD films on on quartz quartz substrates. substrates.
Figure 6. XRD spectra of the of ZnO and Pd films on quartz substrates.
Sensors 2017, 17, 1529 Sensors 2017, 17, 1529
6 of 9 6 of 9
3.2. Sensor Sensor Properties Properties 3.2. The SAW SAW sensor sensor response response was was measured measured for for the the Pd, Pd, ZnO, ZnO, and andPd/ZnO Pd/ZnO sensitive sensitive layers layers for for The hydrogen concentrations up to 2% (half of the safety limit for this gas). Figure 7 shows the response hydrogen concentrations up to 2% (half of the safety limit for this gas). Figure 7 shows the response of of the SAW sensor at different concentrations, at room temperature; experimental point is the SAW sensor at different gas gas concentrations, at room temperature; eacheach experimental point is the the average result of five measurements. It was observed that the frequency shift Δf is proportional average result of five measurements. It was observed that the frequency shift ∆f is proportional to the to the concentration for all for sensors a gas concentration between 2000 and These concentration for all sensors a gasfor concentration between 2000 and 20000 ppm.20000 Theseppm. results are results are similar to those reported by Phan and Chung in the same concentration domain, for room similar to those reported by Phan and Chung in the same concentration domain, for room temperature temperature sensor operation [35]. sensor operation [35].
Figure Frequency shift Figure 7. 7. Frequency shift dependence dependence of of hydrogen hydrogen concentration concentration at at RT. RT.
Table 11 presents presents the the sensitivities sensitivities and and detection detection limits limits for for the the sensitive sensitive layers layers of of Pd, Pd, ZnO, ZnO, and and Table Pd/ZnO. The limit limit ofofdetection detection(LOD) (LOD)depends depends defined 3× Pd/ZnO. on on the the noisenoise level,level, beingbeing defined as 3× as noise noise level/sensitivity. noise level estimated around theallallthe thefilms. films. The The noise noise level/sensitivity. TheThe noise level waswas estimated at at around 3030 HzHz forfor the assessment was performed the frequency fluctuation over 10 assessment performed in inair air(without (withoutanalyte) analyte)bybymeasuring measuring the frequency fluctuation over min; it it represents (best-fit line). line). The The 10 min; representsthe themaximum maximumfrequency frequencydeviation deviationfrom from the the trend line (best-fit sensitivity, defined as the frequency shift in Hz per unit analyte concentration in ppm, was sensitivity, defined as the frequency shift in Hz per unit analyte concentration in ppm, was determined determined from sensitivity an averagevalue sensitivity for a gas concentration between 0.2%The andresponse 2%. The from an average for a value gas concentration between 0.2% and 2%. response time (to 90% of maximum signal) for the tested sensorswas tested was between for time (to reach 90%reach of maximum signal) for the sensors between 12 and 12 16 and s for16s these these hydrogen concentrations. hydrogen concentrations. that thethe LOD of Pd/ZnO filmsfilms was was ~4.5 times betterbetter than the single The results in in table Table1 1indicate indicate that LOD of Pd/ZnO ~4.5 times than the ZnO films almost 2-times 2-times better than thethan single Pdsingle layers.Pd Inlayers. most cases, SAW sensors aresensors heated single ZnOand films and almost better the In most cases, SAW in order to in improve theimprove hydrogen [10,36]. Our[10,36]. results are for obtained room temperature are heated order to theresponse hydrogen response Ourobtained results are for room operation, and are better than comparable to those reported byreported other groups in such conditions temperature operation, and areor better than or comparable to those by other groups in such [11,12,35,37]. conditions [11,12,35,37]. Table1.1.Sensitivity Sensitivityand andlimit limit detection (LOD) of the sensors towards hydrogen = frequency Table of of detection (LOD) of the sensors towards hydrogen (∆f =(Δf frequency shift; cshift; = hydrogen concentration; n = noise level).level). c = hydrogen concentration; n = noise
Coating Material Sensitivity Δf/c (Hz/ppm) Coating Material Sensitivity ∆f/c (Hz/ppm) Pd 0.29 Pd 0.29 ZnO 0.15 ZnO 0.15 Pd/ ZnO 0.51 Pd/ZnO 0.51
LOD (3*n)/(Δf/c) (ppm) LOD (3*n)/(∆f/c) (ppm) 105 105 261 261 59 59
Sensors 2017, 17, 1529
7 of 9
4. Discussion and Conclusions When a bilayer (metal/semiconductor) SAW sensor is exposed to air, oxygen is adsorbed and dissociates at the metal surface. The resulting atomic oxygen diffuses to the metal/semiconductor interface, where it captures electrons and increases the depletion layer width in the semiconductor. Subsequent exposure of the bilayer sensor to H2 similarly results in the formation of atomic hydrogen which diffuses to the interface, reacting with the previously generated oxygen species, and releasing electrons into the conduction band of the semiconductor. The increase in sheet electron conductivity will lead to a decrease in acoustic velocity, since the SAW wave is slowed down by the presence of carriers through the acoustoelectric effect [38,39]. Surface acoustic wave propagation is very sensitive to surface perturbations such as small changes in electrical conductivity. The acoustoelectric effect in the case of a single sensitive layer leads to a relative change of wave propagation velocity ∆ν given by ∆f ∆ν ∼ K2 σs2 = = 2 f0 ν0 2 σs + ν0 Cs2
(1)
where ∆f is the frequency shift, f0 the central frequency of the SAW device, ν0 the unperturbed SAW velocity, K the electromechanical coefficient, Cs = ε0 + εp is the sum of the permittivity of the region above the film and the substrate, and σs is the surface conductivity of the sensing film [11,18]. In the case of bilayers, important additional parameters which determine velocity changes are the electrical conductivities of both layers and their thicknesses, since both film layers influence the SAW propagation through interaction between their mobile charges and the travelling electron potential ϕ of the SAW [17]. These mechanisms explain the improvement of the sensor properties in the case of the Pd/ZnO bilayer compared to the single Pd and ZnO layers. The hydrogen absorbtion and diffusion properties of Pd, combined with the excellent diffusion properties of the nanoporous Pd film, cause increased amounts of hydrogen to reach the Pd/ZnO interface. At this interface, the large sensitivity of ZnO surface conductivity to adsorbed species brings about increased hydrogen detection sensitivity. The sensor response of the bilayer Pd/ZnO thin film proposed by us thus combines both mass and acoustoelectric effects, leading to improved sensitivity to hydrogen. The mass effect of the gas on the sensing layer is known to always lead to a decrease of the center frequency of the SAW sensor. The variation of the electrical conductivity, on the other hand, can lead both to an increase and to a decrease of the center frequency [40]. In order for the combination of the mass and acoustoelectric effects to be efficient, the latter must lead to the same result, i.e., a decrease of the SAW center frequency. Otherwise, the two effects counteract each other and the combined response is smaller. As outlined above, this is the case for our sensor structure, since hydrogen is a reducing gas which reacts with the adsorbed oxygen species, releasing electrons into ZnO, increasing the surface conductivity and decreasing the center frequency. We consider that further optimization of the deposition conditions and sensor structure can lead to further improvement of the hydrogen sensing properties, and perhaps to applications for other gases. In conclusion, a SAW bilayer structure consisting in porous Pd and ZnO layers has been obtained by PLD. The bilayer combines the excellent hydrogen absorbtion and dissociation properties with the sensitivity of surface electron conductivity of ZnO to adsorbed species. The SAW sensor based on this bilayer structure has an improved sensor response due to a combination of the mass and acoustoelectric effects appearing in the presence of hydrogen. The LOD of Pd/ZnO films is about 4.5 times better than the single ZnO films and almost twice better than the single Pd layers, reaching a value of 59 ppm. To our knowledge, it is for the first time that this material combination has been used for bilayers in hydrogen sensors based on surface acoustic waves. Optimization of the bilayer structure can lead to further improvement of the bilayer Pd/ZnO sensor properties, and future research will be addressed in this direction.
Sensors 2017, 17, 1529
8 of 9
Acknowledgments: This work was supported by a grant of the Romanian National Authority for Scientific Research and Innovation, CNCS-UEFISCDI, projects number PNII-RU-TE-2014-4-0342 and project NUCLEU 16470104. Author Contributions: Cristian Viespe and Dana Miu conceived and designed the experiments and analyzed the data; Cristian Viespe performed the experiments; Dana Miu wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4.
5. 6. 7. 8. 9. 10. 11.
12.
13. 14. 15. 16. 17.
18.
19.
Jacobson, M.Z.; Colella, W.G.; Golden, D.M. Clearing the Air and Improving Health with Hydrogen Fuel-Cell Vehicles. Science 2005, 308, 1901–1905. [CrossRef] [PubMed] Bassam, A.M.; Phillips, A.B.; Turnock, S.R.; Wilson, P.A. Development of a multi-scheme energy management strategy for a hybrid fuel cell driven passenger ship. Int. J. Hydrog. Energy 2017, 42, 623–635. [CrossRef] Grimes, C.A.; Ong, K.G.; Varghese, O.K.; Yang, X.; Mor, G.; Paulose, M.; Dickey, E.C.; Ruan, C.; Pishko, M.V.; Kendig, J.W.; et al. A Sentinel Sensor Network for Hydrogen Sensing. Sensors 2003, 3, 69–82. [CrossRef] Boon-Brett, L.; Bousk, J.; Black, G.; Moretto, P.; Castello, P.; Hübert, T.; Banach, U. Identifying performance gaps in hydrogen safety sensor technology for automotive and stationary applications. Int. J. Hydrog. Energy 2010, 35, 373–384. [CrossRef] Buttner, W.J.; Post, M.B.; Burgess, R.; Rivkin, C. An overview of hydrogen safety sensors and requirements. Int. J. Hydrog. Energy 2011, 36, 2462–2470. [CrossRef] Hubert, T.; Boon-Brett, L.; Black, G.; Banach, U. Hydrogen sensors—A review. Sens. Actuators B Chem. 2011, 157, 329–352. [CrossRef] Cole, M.; Spulber, I.; Gardner, J. Surface acoustic wave electronic tongue for robust analysis of sensory components. Sens. Actuators B Chem. 2015, 207, 1147–1153. [CrossRef] Bhasker, R.V.; Nimal, A.T.; Parmar, Y.; Sharma, M.U.; Sreenivas, K.; Gupta, V. Cross sensitivity and selectivity studies on ZnO surface acoustic wave ammonia sensor. Sens. Actuators B Chem. 2010, 147, 517–524. Jakubik, W. Hydrogen gas-sensing with bilayer structures of WO3 and Pd in SAW and electric systems. Thin Solid Film 2009, 517, 6188–6191. [CrossRef] Ippolito, S.J.; Kandasamy, S.; Kalentar-Zadeh, K.; Wlodarski, W. Layered SAW hydrogen sensor with modified tungsten trioxide selective layer. Sens. Actuators B Chem. 2005, 108, 553–557. [CrossRef] Yang, L.; Yin, C.; Zhang, Z.; Zhou, J.; Xu, H. The investigation of hydrogen gas sensing properties of SAW gas sensor based on palladium surface modified SnO2 thin film. Mater. Sci. Semicond. Process. 2017, 60, 16–28. [CrossRef] Jakubik, W.; Urbanczyk, M.; Kochowski, S.; Bodzenta, J. Palladium and phthalocynine bilayer films for hydrogen detection in a surface acoustic wave sensor system. Sens. Actuators B Chem. 2003, 96, 321–328. [CrossRef] Jakubik, W. Elementary theory of SAW gas sensor based on electrical conductivity changes in bi-layer nanostructures. Sens. Actuators B Chem. 2014, 203, 511–516. [CrossRef] Jakubik, W.; Krzywiecki, M.; Maciak, E.; Urbanczyk, ´ M. Bi-layer nanostructures of CuPc and Pd for resistance-type and SAW-type hydrogen gas sensors. Sens. Actuators B Chem. 2012, 175, 255–262. [CrossRef] Jakubik, W.; Powroznik, P.; Wrotniak, J.; Krzywiecki, M. Theoretical analysis of acoustoelectrical sensitivity in SAW gas sensors with single and bi-layer structures. Sens. Actuators B Chem. 2016, 236, 1069–1074. [CrossRef] Li, R.; Reyes, P.I.; Ragavendiran, S.; Shen, H.; Lu, Y. Tunable surface acoustic wave device based on acoustoelectric interaction in ZnO/GaN heterostructures. Appl. Phys. Lett. 2015, 107, 073504. [CrossRef] Dutheil, P.J.; Orlianges, C.; Crunteanu, A.; Catherinot, A.; Champeaux, C. AlN, ZnO thin films and AlN/ZnO or ZnO/AlN mulatilayer structures deposited by PLD for surface acoustic wave applications. Phys. Status Solidi A 2015, 212, 817–825. [CrossRef] Wang, S.Y.; Ma, J.Y.; Li, Z.J.; Su, H.Q.; Alkurd, N.R.; Zhou, W.L.; Wang, L.; Du, B.; Tang, Y.L.; Ao, D.Y.; et al. Surface acoustic wave ammonia sensor based on ZnO/SiO2 composite film. J. Hazard. Mater. 2015, 285, 368–374. [CrossRef] [PubMed] Noh, J.S.; Lee, J.M.; Lee, W. Low-dimensional Palladium nanostructures for fast and reliable Hydrogen gas detection. Sensors 2011, 11, 825–851. [CrossRef] [PubMed]
Sensors 2017, 17, 1529
20. 21. 22. 23. 24.
25.
26. 27. 28. 29. 30. 31. 32.
33. 34. 35. 36. 37. 38. 39. 40.
9 of 9
Viespe, C.; Grigoriu, C. SAW sensor based on highly sensitive nanoporous palladium thin film for hydrogen detection. Microelectron. Eng. 2013, 108, 218–221. [CrossRef] Janotti, A.; Van de Walle, C.G. Fundamentals of zinc oxide as a semiconductor. Rep. Prog. Phys. 2009, 72, 126501. [CrossRef] Wu, T.T.; Wang, W.S. An experimental study on the ZnO/sapphire layered surface acoustic wave device. J. Appl. Phys. 2004, 96, 5249. [CrossRef] Schmidt, O.; Kiesel, P.; Van de Walle, C.G.; Johnson, N.M.; Nause, J.; Döhler, G.H. Effects of an electrically conducting layer on the zinc oxide surface. Jpn. J. Appl. Phys. 2005, 44, 7271–7274. [CrossRef] Schmidt, O.; Geis, A.; Kiesel, P.; Van de Walle, C.G.; Johnson, N.M.; Bakin, A.; Waag, A.; Döhler, G.H. Analysis of a conducting channel at the native zinc oxide surface. Superlattice. Microstruct. 2006, 39, 8–16. [CrossRef] Ballantine, D.S.; White, R.M.; Martin, S.I.; Ricco, A.J.; Zellers, E.T.; Frye, G.C.; Wohltjen, H. Acoustic Wave Sensors, Theory, Design and Physico-Chemical Applications; Academic Press: San Diego, CA, USA, 1997; pp. 337–339. Chrisey, D.B.; Hubler, G.K. Pulsed Laser Deposition of Thin Films; Wiley: New York, NY, USA, 1994. Marcu, A.; Viespe, C. Laser-grown ZnO Nanowires for Room-temperature SAW-sensor Applications. Sens. Actuators B Chem. 2015, 208, 1–6. [CrossRef] Viespe, C. Surface Acoustic Wave Sensors based on Nanoporous Films for Hydrogen Detection. Mater. Appl. Sens. Transducer. Key Eng. Mater. 2014, 605, 331–334. [CrossRef] Marcu, A.; Viespe, C. Active surface geometrical control of noise in nanowire-SAW sensors. Sens. Actuators B Chem. 2016, 231, 469–473. [CrossRef] Marcu, A.; Viespe, C. Surface Acoustic Wave Sensors for Hydrogen and Deuterium Detection. Sensors 2017, 17, 1417. [CrossRef] [PubMed] Campbell, C. Surface Acoustic Wave Devices and THEIR Signal Processing Applications; Academic Press: San Diego, CA, USA, 1989; p. 26. Di Fonzo, F.; Tonini, D.; Li Bessi, A.; Casari, C.S.; Beghi, M.G.; Bottani, C.E.; Gastaldi, D.; Vena, P.; Contro, R. Growth regimes in pulsed laser deposition of aluminum oxide thin films. Appl. Phys. A 2008, 93, 765–769. [CrossRef] Riabinina, D.; Irissou, E.; Le Drogoff, B.; Chaker, M.; Guay, D. Influence of pressure on the Pt nanoparticle growth modes during pulsed laser ablation. J. Appl. Phys. 2010, 108, 034322. [CrossRef] Claeyssens, F.; Cheesman, A.; Henley, S.J.; Ashfold, M.N. Studies of the plume accompanying pulsed ultraviolet laser ablation of zinc oxide. J. Appl. Phys. 2002, 92, 6886–6894. [CrossRef] Phan, D.; Chung, G. Surface acoustic wave hydrogen sensors based on ZnO nanoparticles incorporated with a Pt catalyst. Sens. Actuators B Chem. 2012, 161, 341–348. [CrossRef] Rane, S.; Arbuj, S.; Rane, S.; Gosavi, S. Hydrogen sensing characteristics of Pt-SnO2 nano-structured composite thin films. J. Mater. Sci. Mater. Electron. 2015, 26, 3707–3716. [CrossRef] Jakubik, W.P.; Urbanczyk, M.; Mariak, E.; Pustelny, T. Bilayer structures of NiOx and Pd in Surface Acoustic Wave and electrical gas sensor systems. Acta Phys. Polonica A 2009, 116, 315–320. [CrossRef] Ricco, A.J.; Martin, S.J.; Zipperian, T.E. Surface acoustic wave gas sensor based on film conductivity changes. Sens. Actuators 1985, 8, 319–333. [CrossRef] Fischer, B.H.; Malocha, D.C. Study of the acoustoelectric effect for SAW sensors. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2010, 57, 698–706. [CrossRef] [PubMed] Fan, H.; Ge, H.; Zhang, S.; Zhang, H.; Zhu, J. Optimization of sensitivity induced by surface conductivity and sorbed mass in SAW gas sensors. Sens. Actuators B Chem. 2012, 161, 114–123. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
