IDTs were firstly reported in 1965 by White and Voltmer [30] for generating SAWs in a piezoelectric substrate. An IDT, in its most simple version, is formed by two identical combs-like metal electrodes whose fingers are located in a periodic alternating pattern (see Figure 2). One comb is connected to the shield (ground) and the other to the center conductor of a coaxial cable where a radio frequency (rf) signal is provided.

Each transducer finger may be considered to be a discrete source for the generation of surface waves in a piezoelectric medium since the stress varies with position near each transducer finger. The spatially periodic electric field produces a corresponding periodic mechanical strain pattern by the piezoelectric effect. This gives rise to the SGAW which radiate in both directions away from the transducers orthogonally to the electrodes [29].

For an applied sinusoidal voltage, the transducer operates more efficiently when the SGAW wavelength, λ, matches the transducer periodicity, p, defined as the center-to-center distance between two consecutive fingers of one comb of the IDT. This occurs when the transducer is excited at the synchronous frequency (f_o) in which all vibrations interfere constructively; where f_o = v_o / p, with v_o the SGAW propagation velocity.

The bandwidth B (Figure 3) of an IDT will be narrower when increasing the number of finger pairs N. However, there is a limitation in the maximum N recommended, due to the fact that, in practice, when N exceeds 100, the losses associated with mass loading and the scattering from the electrodes increase. This neutralizes any additional advantage associated with the increase of the number of the finger pairs; for example, the IDT impedance is affected by this parameter [31].

Due to symmetry of the IDT in the direction of propagation, the SGAW energy is emitted in equal amounts in opposite directions. This results in an inherent minimum of a 3 dB conversion loss for the transducer at f_o [32].

The three basic electronic configurations for characterizing piezoelectric devices are shown in Figure 8.

Acoustic waves can be classified either for the particle displacement relative to the propagation direction of the wave (longitudinal or transverse) or for the particle displacement relative to the device surface (vertical or horizontal) [19]. The particle displacement of longitudinal waves (or compressional waves) is parallel to the wave propagation direction, while the particle displacement of transverse waves (or shear waves) is perpendicular to the wave propagation direction (Figure 10). The particles displacements of a vertical wave are normal to the surface of the devices, while the particles displacements of a horizontal wave are parallel to the surface. For a further understanding of particle displacement of an acoustic wave in a solid see references [31,47].

A Y’-cut is defined as a cut which has been rotated θ degrees about the original crystallographic X-axis of a crystal. In the same way, the X’-cuts are the ones rotated θ degrees about the Y-axis. As example, the most common Y’-cuts of quartz crystal are represented in Figure 11. The rotation about the Z-axis does not receive a particular name. Depending on the degrees of rotation about the three axes the elastic, dielectric and piezoelectric constants of a crystal change (see reference [48] for further understanding and rotation equations).

In SGAW literature cuts that are used to designate the direction of the normal to the major faces are often found. An X-cut has the normal to its major faces parallel to the X-axis of the crystal. In the same way, the Y- and Z cuts have their faces perpendicular to the Y- and Z-axes, respectively. In this type of cuts the elastic, dielectric and piezoelectric constants remain the same than the original crystal, because it has not been rotated (θ = 0°).

Other notation commonly used in literature to specify the crystal cut and features of a piezoelectric substrate is: degrees of rotation – cut type – wave propagation direction – substrate material. For example: “36° YX LiTaO₃” means a 36 degrees rotated (with respect to the crystallographic X-axis), Y’-cut, X propagating lithium tantalate substrate.

In order to operate efficiently in applications which require fluid immersion, the Shear Horizontal Surface Acoustic Wave (SH-SAW) sensor was developed (Figure 12). This was the first sensor which used leaky waves, where the wave is only partially confined to the surface. The leaky-SAW mode is mainly shear horizontal but not purely shear horizontal and consequently suffers extra attenuation under fluid immersion [50]. Moreover, this wave extends several wavelengths into the device and therefore has a low sensitivity to changes at the device surface.

SH-SAW devices can be used for measurements in both liquid and gas media. The parameter that determines the resonance frequency of this device is the IDT spacing and its typical operating frequency is 30–500 MHz. The most commonly used substrate for this device is 36° YX LiTaO₃, but ST- quartz [15,51], 41° YX LiNbO₃ [52,53], 64° YX LiNbO₃ [54–56], potassium niobate (KNbO₃) [57] and Langasite (LGS) [28,58] have also been utilized.

For measurements in water an additional problem arises due to the dielectric constant of water (ε_r ≈ 80) which is significantly higher than that of quartz (ε_r = 4.7). This leads to a dramatic decrease in the acoustoelectric coupling and to a significant electrical impedance mismatch which causes short-circuit of the IDTs through the water [59,60]. The later can be minimized by using substrate materials for the device with a ε_r closer to that of water; for example, LiTaO₃ (ε_r = 47).

The major drawback of commercially available SH-SAW devices is the fact that the IDTs mostly consist of low-cost aluminum, so the lifetimes of such devices in aqueous media are limited to a few hours due to corrosion [9]. Therefore, additional protection layers are required like polyimide [61], Parylene C [62] or polystyrene [51].

STW device operates with surface shear horizontal particle displacements, so it can be used for measuring in both gas and liquid media. The parameter that determines the resonance frequency of this device is the spacing of IDTs; its typical operating frequency is 30–300 MHz and the surface mass sensitivity reported in literature is 100–200 cm²/g [29,63]. The most commonly used substrate for this device is ST-cut quartz.

A surface transverse wave is originated from a surface skimming bulk wave (SSBW) that travel very close to the surface but no exactly along it. A metal strip grating located in the surface of the devices between the input and output IDTs produces a slowing effect on the wave propagation velocity and traps the energy of the wave in the surface of the device enhancing its surface mass sensitivity (Figure 13). Thus, the STWs can be defined as grating-affected SSBWs.

The difference between leaky waves and SSBW waves is the wave propagation angle. Leaky waves have a larger propagation angle than SSBW waves. Figure 14 shows a scheme exemplifying the propagation angles of Leaky, SSBW and STW. As can be seen, Leaky waves have a higher propagation angle than SSBW and STW waves.

The SSBW was first described by Milson et al. [64] and Lewis [65] in 1977, but its advantageous qualities have turned out to be largely discredited by a substantial loss from radiation into the bulk of the substrate. It was not until 1987, when Bagwell and Bray [66] proposed a two-port resonator with an unloaded high quality factor (Q) of 5,600 in which the SSBW power was trapped to the surface via the effect of a metal strip grating.

Grating waveguides have certain advantages over plates. The grating can be matched to the transducer in order to prevent acoustic reflections and provides a stronger guiding [67]. They can also provide much higher sensitivity for the same thickness of material [68]. However, despite the numerous studies and results on STW resonant structures, little has been clearly said on how to achieve satisfactory quantitative understanding and prediction of device parameters [69]. In general, the inclusion of bulk losses appears to cause tremendous analytical difficulties. However, STW have proved to outperform conventional Rayleigh SAWs in a number of parameters. They are considerably faster, have a lower propagation loss and are more sensitive to outside impacts such as mass loading from absorbed gaseous substances. These qualities make the STWs suitable for a range of applications, including devices reaching the 3 GHz range and high sensitivity sensors [69].

Love wave devices (LWs) are comprised of a substrate that primarily excites a SSBW, which is subsequently confined by a thin guiding layer located on the top of the substrate and IDTs (Figure 15). Therefore, the IDTs remain isolated from liquids. The condition for the existence of Love wave modes is that the shear velocity of the overlay material is less than that of the substrate [43]. The waveguide layer confines the wave energy keeping it near the surface and slows down the wave propagation velocity. Thus, the LWs can be defined as layered-affected SSBWs. The sensitivity of a sensor is determined by the degree of wave confinement. If the wave is trapped tightly, it will be strongly perturbed by surface changes, yielding high sensitivity.

This device operates with a surface wave with shear horizontal particle displacements. Thus, it can operate efficiently in both gas and liquid media. The parameters that determine the resonance frequency are the spacing of IDTs and the thickness of the wave guiding layer. Typical frequencies in which this device operates are 80–300 MHz and the surface mass sensitivity reported for this device in literature is 150–500 cm²/g [29,63] .

Initially, the LWs were made in ST-cut quartz [70]; however, those devices lacked temperature stability, which is essential for field application. Thus, temperature-compensated systems based on different Y-rotated quartz and LiTaO₃ plates were investigated [71]. Later, LiNbO₃ substrates, like 64° YX LiNbO₃ [72,73], were proposed for these devices, Waveguide materials such as polymers [74], silicon dioxide (SiO₂) [71] and zinc oxide (ZnO) [75, 76] can be used for guiding layers [77].

This device was introduced in 1980s. It operates with a plate wave with shear horizontal particle displacements. Thus, this device can be used for measurements in contact with both gas and liquid media. The parameter that determines the resonance frequency of this device is the spacing of IDTs and the thickness of the substrate. The typical operation frequency of this device is between 25–200 MHz and the surface mass sensitivity reported in literature is 20–50 cm²/g [29,31,63]. The most commonly used substrates for this device are ST-cut quartz and ZX-LiNbO₃. The SH-APMs have been used for measuring mass change in liquid media and also for detecting biologic molecules [21].

The advantage of this device is that the IDTs can be located on the back side of the device and are thus away from the sensing side, what insolates the IDTs from the liquid (Figure 16). Thus, corrosion problems on electrodes resulting in deterioration of the sensor response are avoided. However, the main drawback of using SH-APMs is the fact that they are difficult to operate in a standard oscillator circuit. The reason for this is that several acoustic plate modes are usually excited simultaneously, but the frequency separation between these modes is often limited; which can produce a hopping mode in an oscillator circuit [9]. Additionally, mechanical and electrical loading of the surface can affect the APM sensor response; especially if a high-coupling piezoelectric material like LiNbO₃ is used, the acoustoelectric interaction becomes important [22].

The use of APM delay lines has been hampered by the relative immaturity of the associated design techniques. The principle issue in the design of APM delay lines is to excite and detect electrically a single acoustic mode within the plate with low distortion from intermode interference or multiple waveguide reflections. The use of single-phase unidirectional transducers (SPUDT) enables the excitation and detection of a single acoustic mode, reducing the distortions that occur in conventional transducer designs [20].

Theoretically, the sensitivity of this device for an isotropic plate is given by S_m = −J / ρ·d; where J = 1/2 for the mode n = 0 and J = 1 for higher plate modes (n > 0), ρ is the plate density and d is the plate thickness [29]. Decreasing the plate thickness increases the frequencies of higher plate modes of the SH-APM device and it also increases mass sensitivity. Thus, higher-order modes appeared to be more sensitive than lower-order modes, although they have more transmission losses [78].

McHale et al. [79,80] suggested, from theoretical considerations, that a guiding layer could be used on one substrate face of a SH-APM device to create a LG-APM in a similar way to Love waves and so obtain a sensitivity approaching that of a LW. They suggested that higher order Love modes can be regarded as continuations of the layer-guided SH-APM. The most commonly used substrates for this device are ST-cut quartz and 36° YX LiTaO₃.

Evans et al. [81] demonstrated experimentally that lithium tantalate substrates could be used for a LG-APM, which exhibit an enhanced mass sensitivity compared to the traditional SH-APM. Thus, it is possible to retain the advantages of operating with liquids on the opposite face to the transducer as with the SH-APM device with a mass sensitivity enhancement.

This device operates with a plate wave with shear horizontal particle displacements. Thus, measurements in both, gas and liquid media are possible. The parameters that determine the resonance frequency are the spacing of IDTs and the substrate thickness of the guiding layer. Typical operating frequencies are 25–200 MHz and the surface mass sensitivity reported is between 20–40 cm²/g [29,63].

FPW devices are built with plates that are only a fraction of an acoustic wavelength thick (typically 2–3 mm). The confinement of acoustic energy in such a thin membranes results in a very high mass sensitivity. The plates are composite structures (Figure 17) consisting of a silicon nitride layer, an aluminum ground plane, and a sputtered zinc oxide piezoelectric layer, all of which are supported by a silicon substrate. This device operates with a plate wave with vertical particle displacements. Measurement in gas and liquid media are possible, due to the fact that FPW velocity is less than the compressional velocity of sound in liquids. Therefore, this device does not couple compressional waves in liquid and only minor energy dissipation occurs [31]. The parameters that determine the resonance frequency of this device are the spacing of IDTs and thickness of the substrate. Typical operation frequencies are 2–20 MHz and the surface mass sensitivity reported is 200–1,000 cm²/g [82]. When the device is operated in liquid, the mass sensitivity falls from 1,000 cm²/g to about 200 cm²/g, because of mass contributed by the liquid in the evanescent field region [67,83].

When the substrate thickness is less than the penetration depth, an interaction is produced between the guided modes in both substrate faces and the Lamb modes are then generated [31]. The particle displacements of a Lamb wave are similar to those of the Rayleigh wave, elliptical components both normal and parallel to the surface. In fact, a Lamb wave can be considered as being composed of two Rayleigh waves propagating on each side of a plate with a thickness of less than one wavelength. For plates thicker than a few wavelengths, two free Rayleigh waves propagate. Plate waves can propagate in symmetric or antisymmetric modes [84]. In Figure 18, the motions of groups of atoms of a Lamb wave are depicted in the cross-sectional view of the wave propagating to the right. Particle displacement directions are represented with red arrows. The wave velocity of the first antisymmetric mode (A₀) is much lower than those of the other possible modes. A₀ is the only mode that presents a phase velocity going to zero when the normalized thickness, d/λ, goes to zero; where d is the plate thickness and λ is the wavelength. This mode has flexible features; hence it is also known as Flexural Plate Wave (FPW).

The theoretical mass sensitivity of this device is given by S_m= −1/2ρd, where ρ is the plate density and d is the plate thickness [29]. The mass sensitivity increases with decreasing plate thickness (decreasing frequency for a constant wavelength). Thus, the main advantage of flexural plate wave device (FPW) is its high sensitivity to added mass at a low operating frequency, which eases the associated electronics requirements. However, FPWs are fragile due to the reduced device thickness [85].

Other advantages of FPW sensors are their on-line, real-time performance, compatibility with aqueous samples, and variable surface chemistry. A particularly interesting feature of FPW is their potential for the sentinel activities in remote or inaccessible locations [86]. In addition, the FPW can be used to measure cell concentrations and growth rates in industrial fermentors, biofilms, and wastewater treatment facilities [87].

In 1987, Moriizumi et al. described the first biosensor application of SH-SAW transducers in liquid medium [18]. Further on, in 1993, Rapp. and coworkers suggested the use of commercially available SH-SAW filters for communication applications, with frequency ranges of 150–1,000 MHz and lithium tantalate (LiTaO₃) as the substrate, to develop an immunosensor [96]. A similar approach was presented by these authors in 1995. They produced an immunosensor by covalent antibody immobilization, via the cyano-transfer technique, to a thin polyimide layer that preserved aluminium IDT’s of the transducer from corrosion when working in aqueous buffers [60].

More recently, Deobagkar and coworkers [51] developed a SH-SAW immunosensor for the detection of E. coli O157:H7 in water. In this immunosensor, polyclonal antibodies were covalently attached to polystyrene-coated active transducer surface. The authors reported a detection range of 0.4 –100 cells/μL giving, frequency shifts over 1.5– 5.8 kHz in an 87.7 MHz oscillator. This detection allowed the determination of up to 0.4 cells/μL of E. coli in water. Thus, the device was sensitive enough to detect this pathogen at concentrations which could cause human health hazards.

In a previous work, Berkenpas et al. [28] used a SH-SAW transducer fabricated on langasite (LGS) crystals to successfully detect macromolecular protein assemblies. This device demonstrated favorable temperature stability, biocompatibility, and low attenuation in liquid environments, suggesting its applicability to bacterial detection. Later on, the same authors applied and validated this previously reported LGS SH-SAW biosensor for the detection of E. coli O157:H7 [58]. They derivatized these LGS SH-SAW delay lines by attaching a biotinylated polyclonal rabbit antibody, directed against E. coli, to NeutrAvidin™ SAM functionalized gold surface.

In 2006, Länge et al. [97] presented a new approach to integrate a SH-SAW biosensor in a microfluidic polymer chip. The chip is easy to handle and its total volume is of only 0.9 μL. According to preliminary experiments with such microdevice, the authors stated that SGAW biosensing systems based on these chips promise fast response times and low sample consumption for bioanalytical sensing applications.

The first approaches employing LW for biochemical sensing were reported in 1992 by Kovacs et al. [98] and by Gizeli et al. [99], who first demonstrated the use of such devices as mass sensing biosensors in liquids. However, it was not until 1997 that Harding et al. [100] used a LW acoustic device to detect real-time antigen-antibody interactions in liquid media. In 1999, Freudenberg and coworkers built a contactless LW device in order to protect electrodes from the conductive and chemically aggressive liquids used in biosensing [101].

In 2000, Howe and Harding [102] used a dual channel LW device as a biosensor to simultaneously detect Legionella and E. coli. In this approach a novel protocol for coating bacteria on the sensor surface prior to addition of the antibody was introduced. Quantitative results were obtained for both species down to 10⁶ cells/mL, within 3 h.

In 2003, Tamarin et al. [103] designed a LW immunosensor as a model for virus or bacteria detection in liquids (drinking or bathing water, food, etc.). They grafted a monoclonal antibody (AM13 MAb) against M13 bacteriophage on the device surface (SiO₂) and sensed the M13 bacteriophage /AM13 immunoreaction. The authors suggested the potentialities of such acoustic biosensors for biological detection. The same year, Kalantar-Zadeh et al. [104] showed that mass sensitivity of LW devices with ZnO layer was larger than that of with SiO₂ guiding layers. They monitored adsorption of rat immunoglobulin G, obtaining mass sensitivities as high as 950 cm²/g. The authors pointed out that such a device was a promising candidate for immunosensing applications.

Branch and coworkers reported in 2004 [105] a LW biosensor for the detection of pathogenic spores at or below inhalational infectious levels. A monoclonal antibody with a high degree of selectivity for anthrax spores was used to capture the non-pathogenic simulant Bacillus thuringiensis B8 spores in aqueous conditions. They suggested that acoustic LW biosensors will have widespread application for whole-cell pathogen detection.

Due to the fact that direct anti- E. coli antibodies grafting onto the sensor surface did not lead to a significant detection of whole bacteria, in 2007 Moll et al. [106] developed an innovative method for the detection of E. coli employing an LW device. It consisted of grafting goat anti-mouse antibodies (GAM) onto the sensor surface and introducing E. coli bacteria mixed with anti-E. coli MAb in a second step. The sensor response time was shorter when working at 37°C, providing results in less than 1 hour with a detection threshold of 10⁶ bacteria/mL. More recently, the same authors [107] described a multipurpose LW immunosensor for the detection of bacteria, virus and proteins. They successfully detected bacteriophages and proteins down to 4 ng/mm² and E. coli bacteria up to 5.0 × 10⁵ cells in a 500 μL chamber, with good specificity and reproducibility. The authors stated that whole bacteria can be detected in less than one hour.

Taking into account that SGAW biosensors are a powerful tool for the study of biomolecular interactions, Andrä et al. [108] used a LW sensor to investigate the mode of action and the lipid specificity of human antimicrobial peptides. They analyzed the interaction of those peptides with model membranes. These membranes, when attached to the sensor surface, mimic the cytoplasmic and the outer bacterial membrane.

Finally, Bisoffi et al. [109] used a LW immunosensor to detect Coxsackie virus B4 and Sin Nombre virus (SNV), a member of the hantavirus family. They described a robust biosensor that combines the sensitivity of SGAW at a frequency of 325 MHz with the specificity provided by monoclonal and recombinant antibodies for the detection of viral agents. Rapid detection (within seconds) for increasing virus concentrations was reported. The biosensor was able to detect SNV at doses lower than the load of virus typically found in a human patient suffering from hantavirus cardiopulmonary syndrome.

White et al. and White and Wenzel first reported in 1987–1989 the use of FPW dispositives as sensors in gaseous [82] and liquid [83,110,111] environment. By the same time, Constello et al. [112] monitored the changes in frequency and viscosity of such devices over time, when proteins adsorbed onto the sensor surface.

However, it was not until 1998 that Pyun et al. [24] described an approach of a biosensor for E. coli employing an “acousto-gravimetric” FPW transducer. They covalently coupled antibodies, against E. coli K12 and E. coli J5 bacteria, to the aminosilanized monolayer modified platinum active surface, and measured changes in frequency in a flow system. A detection range of 3.0 × 10⁵ to 6.2 × 10⁷ cells/mL was obtained. To further increase the sensitivity, they used microespheres coupled with anti E. coli antibodies in a sandwich assay, amplifying the signal about five-fold.

In 1999, Cowan et al. [87] performed an on-line real-time measurement of changes in the concentration of E. coli W3110. The authors detected those changes as the cells settled onto the sensor under the influence of gravity, and reported that experimental data were in good agreement with a developed theoretical model for sensor’s response to cell settling. They postulated the use of FPW sensors as devices to measure cell concentrations and growth rates in industrial fermentors, biofilms and wastewater treatment facilities.

Recently, Kuznetsova and Coakley [113] stated, in a complete review article, that biosensors incorporating ultrasound standing wave systems (USW) can show enhanced sensitivity due to the effect of the direct radiation force (DRF) and induced acoustic streaming. DRF concentrates particles of the analytical sample and can also increase the capture rate by delivering suspended particles to immune-coated surfaces. Acoustic streaming provides rapid and homogeneous mixing in small analytical systems, otherwise difficult to achieve. Streaming induced in FPW immunosensors, and cavitation micro-streaming in the analytical cell, significantly accelerated the antigen-antibody reaction. The authors presented an example of such an ultrasound system to demonstrate E. coli K12 cell capture efficiency on the surface of immunomagnetic beads, stating that this phenomena can be implemented in FPW based sensors.