Open Access
Add to BibSonomy
Abstract
The analysis of thin layers deposited on various substrates is widely employed in thickness monitoring, materials research and development and quality control. Measurements are often performed based on changes to acoustic resonance frequencies of quartz micro-balance devices. The technique is extremely sensitive, but it is restricted to hundreds of MHz frequencies and requires electrical connectivity. In this work we propose and demonstrate the analysis of elastic properties of thin layers deposited on surface acoustic wave-photonic devices in standard silicon-on-insulator. The devices operate at 2.4 GHz frequency, and their interfaces are fiber-optic. The radio-frequency transfer functions of the devices are modified by sub-percent level changes to the group velocity of surface acoustic waves following deposition of layers. Layers of aluminum oxide and germanium sulfide of thickness between 10-80 nm are characterized. The analysis provides estimates for Young’s modulus of the layers.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Sensor devices based on the actuation of hypersonic surface acoustic waves (SAWs) are widely employed in precise measurements of mass loading, surface analysis and thin films characterization [1–6]. SAW-based thickness monitors are part of most standard equipment for thin layers deposition [7,8]. Most devices rely on piezo-electric actuation and detection of SAWs [1–8]. Given sufficient feedback gain between output and input voltages, the devices can be driven to oscillations [9]. The exact radio frequency of oscillations is sensitive to small changes in the SAW velocity [1]. Frequency measurements can properly identify the mass loading of monolayer deposition on top of the device surface [10]. The proper pre-functionalization of the surface may lead to sensitive and specific detection of target reagents [11]. However, the standard sensors require a piezo-electric substrate and electrical connectivity, and their acoustic frequencies are often restricted to hundreds of MHz.
SAWs may also be stimulated and detected using optical interfaces [12–14]. The absorption of modulated light in spatially periodic metallic patterns can lead to alternating thermal expansion and contraction, and the launch of SAWs [12–14]. The modulating waveform may be recovered in the optical domain based on photoelasticity [12–14]. Recently, our group has demonstrated SAW devices as part of photonic integrated circuits in standard silicon-on-insulator (SOI) substrates [15–16]. Pump light is absorbed in metallic grating elements to generate surface waves, which then modulate an optical probe wave in a standard resonator waveguide. Modulation information is thereby transferred from pump to probe, via SAWs. The layout of the devices can be designed to implement microwave-photonic filters [16]. Due to the slow acoustic velocity, the devices support long acoustic delays and ultra-narrow filter passbands of only a few MHz widths [16]. The transfer function of the filters depends on the exact SAW velocity.
In this work, we propose and demonstrate the SAW analysis of thin layers deposited on top of SOI-photonic circuits. The frequency of operation is 2.4 GHz, much higher than those of most piezoelectric devices. The deposited areas are only 0.05 mm2. The microwave frequency transfer functions of the devices are sensitive to sub-percent changes in the SAW group velocity. The principle is demonstrated through the deposition of two types of layers: aluminum oxide layers of 10-30 nm thickness and germanium sulfide layers of 50-80 nm thickness. Measurements identify the deposition of layers as thin as 10 nm, with 2 × 10−9 gr mass. The measurements provide estimates of Young’s moduli of the deposited layers.
The results represent a first proof-of-concept for new potential functionality of silicon-photonic integrated circuits, as SAW sensors. Devices can be produced at large volume and low cost and do not require piezoelectric substrates. The measurements can be used for the analysis of any deposited layer, as well as properties of the silicon device and buried oxide layers themselves following processes such as ion implantation. Due to the higher hypersonic frequencies used, the introduction of proper feedback may eventually improve the sensitivity of the proposed SAW-photonic sensors beyond those of present piezoelectric devices. Optical rather than electrical connectivity might be more suitable for hazardous environments. Preliminary results were briefly reported in a recent conference [17].
2. Principle of operation
A SAW-photonic device in the SOI platform is illustrated in Fig. 1 [16]. The principle of operation of such devices and their application as microwave-photonic filters are described at length in our recent works [15,16]. The concept is briefly described below for convenience, and the reader is referred to the earlier works for complete detail. An optical pump wave from fiber input 1 illuminates a grating of metallic stripes with spatial period Λ. The intensity of the pump wave is modulated at radio frequency Ω. The absorption of modulated light leads to periodic heating and cooling of the metallic stripes, accompanied with their expansion and contraction [12–14]. The thin metallic stripes thermalize within picoseconds [12–14]. The strain pattern induced by the grating is periodic in both time and position, with frequency Ω and wavelength Λ.
Fig. 1. Schematic illustration of a SAW – photonic device in SOI, used for the analysis of thin deposited layers. Patches of a thin layer under test are noted by yellow regions in between optical waveguide stretches.
Strain is transferred to the underling SOI layer stack. If $\mathrm{\Omega }$ and $\mathrm{\Lambda }$ match the frequency and wavelength of a surface acoustic mode of the SOI structure, SAWs are launched away from the grating region [12–16]. Figure 2 shows numerical calculations of the fundamental surface acoustic mode of the SOI layer stack, with a wavelength of 1.4 µm and 2.55 GHz frequency (see Methods for parameters used in calculations). The magnitude of the generated SAW is proportional to the intensity modulation of the incident pump wave [15–16]. We denote the radio frequency response of the SAW generation as ${H_G}(\mathrm{\Omega } )$. The response is maximal at ${\mathrm{\Omega }_{max}} = 2\pi {v_{ph}}/\mathrm{\Lambda }$, where ${v_{ph}}$ is the phase velocity of the surface acoustic mode [15–16]. The bandwidth of ${H_G}(\mathrm{\Omega } )$ is approximately given by $\mathrm{\Delta }{\mathrm{\Omega }_G} \approx ({\mathrm{\Lambda }/L} ){\mathrm{\Omega }_{max}}$ [16], where L is the metallic grating size. The thermo-elastic excitation of SAWs has been used in bulk substrates for over 15 years [12–14]. We have recently carried over this technique to silicon-photonic circuits, as an alternative to piezo-electric actuation [15–16].
Fig. 2. Calculated profile of the fundamental surface acoustic mode of the SOI layer stack, with 1.4 µm wavelength and 2.55 GHz frequency. The thickness of the silicon device layer and the buried oxide layer are 220 nm and 2 µm, respectively. (a): Normalized transverse displacement, normal to the surface. (b): Normalized longitudinal displacement, along the direction of propagation.
A resonator waveguide is defined in the silicon device layer, in proximity to the metallic grating. The resonator layout consists of N straight sections that run parallel with the grating stripes, and they are separated by equal offsets $\Delta y$ (Fig. 1). The traveling SAW front induces photoelastic perturbations to the effective index of the guided optical mode in each waveguide section. The perturbation magnitude is proportional to that of the SAW. The perturbation to waveguide section $l = 0 \ldots ({N - 1} )$ is delayed by $l\tau = l\Delta y/{v_g}$ with respect to that of the one nearest the grating ($l = 0$), where ${v_g}$ is the SAWs group velocity and $\tau \equiv \Delta y/{v_g}$. The overall index perturbation to an optical probe wave that is guided in the resonator as a function of time t is given by [16]:
In Eq. (1), $\Delta {n_0}(\mathrm{\Omega } )$ is the magnitude of photoelastic perturbations in the waveguide section nearest the metallic grating, which scales with the grating response ${H_G}(\mathrm{\Omega } )$ and the intensity modulation of the pump wave [16]. Also in the same equation, ${\alpha _{SAW}}$ is the coefficient of SAW intensity losses (in m−1), and $\{{{a_l}} \}$ are weights assigned to each waveguide section. The magnitude and phase of each tap coefficient ${a_l}$ can be arbitrarily chosen, independent of those of all others, through modifications to waveguides widths and positions [16]. Throughout this work, ${a_l} = 1$ is used in all sections for simplicity.
An optical probe wave is coupled to the resonator waveguide at input port 2. The probe wavelength is aligned to a maximum spectral slope of the resonator transfer function [15–16]. The index perturbations induced by the SAWs manifest as intensity modulation of the probe wave at the output port of the device (Fig. 1., [15–16]). The modulation of the input pump wave is therefore transferred onto the probe, via slow-moving SAWs. The transfer function between the voltage magnitude modulating the input pump wave and that of the detected output probe is given by ${H_{Tot}}(\mathrm{\Omega } )= C{H_G}(\mathrm{\Omega } ){H_W}(\mathrm{\Omega } )$, where:
3. Device fabrication and testing
Devices were fabricated in standard SOI substrates with a 220 nm-thick silicon device layer on top of a 2 µm-thick buried oxide layer. Figure 3(a) shows a top-view optical microscope image of a SAW-photonic device with $N = 12$ straight sections within the resonator waveguide. Ridge waveguides were defined in the silicon device layer using electron-beam lithography and subsequent inductively coupled plasma reactive ion-etching (see Methods for details). The partial etching depth was 70 nm, and the width of etched ridges was 700 nm. The spacing $\Delta y$ between adjacent parallel waveguide sections within the resonator was 60 µm. Gold gratings of 20 nm-thin stripes were deposited using a sputtering process (see Methods). The gratings period $\mathrm{\Lambda }$ was 1.4 µm with 50% duty cycle for all devices, and the grating dimensions were 60×60 µm2. Probe light was coupled between standard single-mode optical fibers and devices under test using vertical grating couplers. Coupling losses were 10 dB per interface. The transfer functions of the probe wave optical power through devices under test were characterized using an optical vector network analyzer with 3 pm wavelength resolution. Figure 3(b) presents an example of the optical power transfer function through a 12-sections device.
Fig. 3. (a): Top-view optical microscope image of a SAW-photonic device in SOI. The resonator waveguide consists of 12 parallel waveguide sections, implementing a radio frequency response of a 12-tap, delay-and-sum filter. (b): Measured normalized transfer function of optical power through the device of panel (a).
The experimental setup for the measurement of radio frequency transfer functions is illustrated in Fig. 4(a) [15–16]. Light from a first laser diode at 1540 nm wavelength was used as the source of input pump waveforms. The laser diode light passed through an electro-optic amplitude modulator (${V_\pi }$ = 3.5 V), driven by the output voltage of a radio-frequency vector network analyzer (VNA). The modulation voltage was scanned across a range of radio frequencies $\mathrm{\Omega },$ and its electrical power was +20 dBm. The modulated optical waveform was amplified by an erbium-doped fiber amplifier (EDFA) to an average power of 150 mW. The end facet of the EDFA output fiber was held above the metallic grating of the device under test. The vertical separation between the fiber facet and the substrate was adjusted so that the illuminating spot size matched the grating extent.
Fig. 4. (a): Schematic illustration of the experimental setup used to measure the radio frequency transfer functions of SAW-photonic devices [16]. EDFA: erbium-doped fiber amplifier; PC: polarization controller; PD: photodiode; BPF: optical bandpass filter; VNA: vector network analyzer; MZM: electro-optic Mach-Zehnder intensity modulator. (b): Measured normalized transfer function ${|{{H_{Tot}}(\mathrm{\Omega } )} |^2}$ of radio frequency electrical power between the modulation voltage of the input pump wave and that of the detected output probe wave. The transfer function is characterized by periodic passbands of 5 MHz bandwidth and a free spectral range of 65 MHz. (c): Measured time-domain impulse response $|{{h_{Tot}}(t )} |$ of the same device.
Light from a second laser diode of 1545 nm wavelength and 30 mW power was the source of the optical probe wave. The exact wavelength of the probe diode source was adjusted to a maximal slope of the optical transfer function of the device under test, using temperature and current tuning. The average power of the probe wave at the device output was amplified by another EDFA to an average power of 3 mW. An optical bandpass filter was used to suppress the amplified spontaneous emission of the EDFA. The probe wave was detected by a broadband photo-receiver (responsivity of 27 V × W−1, rise time of 15 ps), and the detector voltage was analyzed by the input port of the VNA. Figure 4(b) presents an example of the measured normalized transfer function of radio frequency electrical power ${|{{H_{Tot}}(\mathrm{\Omega } )} |^2}$ through a device with 12 waveguide sections within its resonator. The response is characterized by narrow and periodic passbands, with full widths at half maximum of 5 MHz and a free spectral range of 65 MHz. The impulse response ${h_{Tot}}(t )$ of the device under test was calculated offline through the inverse-Fourier transform of measured, complex-valued ${H_{Tot}}(\mathrm{\Omega } )$ (Fig. 4(c)).
4. Analysis of deposited thin layers
In a first set of experiments, thin layers of aluminum oxide were deposited on top of SAW-photonic circuits using atomic layers deposition (see Methods). Aluminum oxide was deposited in between waveguide sections. The exact thicknesses of the deposited layers were verified using atomic force microscopy. Figure 5(a) shows a top-view microscope image of a device following the deposition of 30 nm thick layer. The transfer functions ${|{{H_{Tot}}(\mathrm{\Omega } )} |^2}$ of each device were measured twice: before and after the aluminum oxide deposition.
Fig. 5. (a): Top-view optical microscope image of a SAW-photonic device, following the deposition of a 30 nm-thick aluminum oxide layer. (b)-(d): Measured radio-frequency electrical power transfer functions of the device shown in panel (a), before (red) and after (blue) the deposition of 30 nm of aluminum oxide. The three panels show different frequency ranges.
Panels (b) through (d) of Fig. 5 show the measured transfer functions ${|{{H_{Tot}}(\mathrm{\Omega } )} |^2}$ before and after the deposition of a 30 nm-thick aluminum oxide layer. The frequency span of five free spectral ranges increased by 2.4 ± 0.3 MHz, or 0.8 ± 0.1%, following the layer deposition. This change corresponds to an increase of the SAW group velocity by 35 m × s−1, from 3750 m × s−1 to 3785 m × s−1. Note that the patches of the deposited layer do not cover the entire lateral extent between adjacent waveguide sections, as small areas near the waveguides themselves are deliberately left uncovered. The calculation of the group velocity takes into consideration the exact actual dimensions of the deposited patches. Measurements were also performed for 20 nm and 10 nm thin layers. The observed increases in SAW group velocity were 25 m × s−1 and 15 m × s−1, respectively. The latter represents the thinnest aluminum oxide deposition that could be reliably identified using the current setup and devices. The total mass of the that layer was 2 × 10−9 gr.
The surface acoustic modes of the SOI layer stack following the deposition of aluminum oxide were calculated numerically (see Methods for the parameters used). Young’s modulus of the aluminum oxide layer was used as a fitting parameter. Calculations best reproduced the observed increase in acoustic group velocity following the layers deposition for aluminum oxide modulus of 200 ± 10 GPa. That value is higher than literature reports of 144 GPa [18]. The difference could be due to the details of the deposition process, or due to the high hypersonic acoustic frequency used. The stoichiometry of the deposited layer might also be different from that of the target itself.
In a second set of experiments, thicker layers of germanium sulfide were deposited on similar devices using thermal evaporation (see Methods for details). Figure 6(a) shows a top-view microscope image of a device following the deposition of an 80 nm-thick layer of germanium sulfide. Figure 6(b) presents the measured impulse responses of three devices, following the deposition of 52, 63 and 80 nm thick layers of germanium sulfide, alongside that of a reference device. The mass of the thinnest layer tested was 7.2 ± 0.5 × 10−9 gr. The deposition of germanium sulfide layers increases the unit delay of the impulse response: from a reference value of 15.8 ns to 16.9 ns, 17.15 ns and 18.0 ns following the deposition of 52, 63 and 80 nm thick layers, respectively. The corresponding SAW group velocities are 3800 m × s−1 for the reference device and 3500, 3440 and 3250 m × s−1 following deposition (Fig. 6(c)).
Fig. 6. (a) Top-view optical microscope image of a SAW-photonic device, following the deposition of 80 nm-thick patches of germanium sulfide layer between the straight waveguide sections. (b) Measured impulse responses of SAW-photonic devices following the deposition of germanium sulfide layers of different thicknesses (see legend), alongside that of a reference device. The deposition of layers increases the unit group delay of the devices (see legend). (c) Measured (blue squared markers) and calculated (red circular markers) group velocities of a SAWs as a function of the thickness of deposited germanium sulfide layer on top of the SOI layer stack. Young’s modulus of the germanium sulfide layer was fitted as 12 GPa (see Methods).
The group velocities of surface acoustic modes following the deposition of germanium sulfide layers of different thickness were also calculated numerically (Fig. 6(c), see Methods section 6.1 below). The values obtained for the 52 nm, 63 nm and 80 nm thick layers are 3590, 3540 and 3260 m × s−1. The density of the layers was measured as 2.9 ± 0.2 g×cm−3 using Rutherford Backscattering (RBS) analysis (see Methods). The Poisson ratio of the layer was taken as 0.14 [18], and its Young’s modulus was used as a fitting parameter. Best agreement between calculations and experiment was obtained for a modulus of 12 ± 2 GPa. This value is in general agreement with literature reports of 15 GPa [18]. The soft, thin layers of germanium sulfide induce larger changes in SAW group velocity than high-modulus layers of similar thickness, such as aluminum oxide.
5. Summary and conclusions
The analysis of thin deposited layers on top of standard SOI substrates was proposed and demonstrated using SAW-photonic integrated circuits. The devices implement narrowband microwave photonic filters, in which incoming radio frequency modulation is converted from an optical pump wave onto an optical probe wave via slow moving SAWs. Sub-percent changes to the group velocity of SAWs, on the order of 15 m × s−1, were readily observed in the measured transfer functions of devices under test. The measurements identified the deposition of aluminum oxide layers as thin as 10 nm and provided estimates of Young’s modulus of aluminum oxide and germanium sulfide layers at high hypersonic frequencies. Thin, soft layers include larger changes in SAW group velocity than stiffer layers of comparable thickness, which are closer in their elastic properties to those of silica and silicon.
The sensitivity of the measurement is limited by the minimum detectable offset in the frequency response of devices under test. That limit is currently on the order of hundreds of kHz. The sensitivity of corresponding piezo-electric sensor devices is greatly enhanced with the introduction of amplified feedback between their output and input voltage ports [9–11,19]. Feedback leads to ultra-narrow radio frequency oscillations, and the monitoring of their exact frequencies identifies ppm-level changes in acoustic group velocities [9–11,19]. In an earlier study, we were able to obtain narrowband phonon oscillations in multi-core optical fibers, using a similar scheme to the one used in this work [20]. Modulated pump light in one core of the fiber generated guided acoustic waves in a forward stimulated Brillouin scattering process, and the acoustic waves in turn induced photoelastic modulation of a probe wave in a different core [20]. Sufficient optical and/or electrical feedback gain between the detected probe modulation and that of the input pump has driven the setup into microwave frequency oscillations, with linewidths below 100 Hz [20]. Similar feedback between probe and pump may be applied to the SAW-photonic devices described in this study. The introduction of feedback is the subject of ongoing research. If achieved, SAW-photonic oscillator devices would be able to identify velocity changes that are orders of magnitude smaller than reported here.
The velocities of SAWs are not affected by possible angular misalignment between thin amorphous layers, such as those used in this work, and the underlying silicon device layer. Large angular misalignment could, in principle, refract the surface waves with respect to the optical waveguides and modify the accumulation of photoelastic phase modulation along waveguide sections. However, such misalignment can be avoided through careful, standard lithography. In case the deposition process requires an adhesion layer, that layer should be included in the pre-calibration of devices response.
The SAW-photonic devices can operate at tens of GHz frequencies [12–14], two orders of magnitude higher than those of most piezo-electric sensors. Therefore, they may eventually reach superior sensitivity. The proposed devices would extend the elastic characterization of thin layers to higher frequencies than currently available. Compared with stimulated Brillouin scattering based devices [21], SAW devices are not restricted to a discrete set of specific frequencies. The surfaces of silicon-photonic devices can be functionalized for the adsorption of specific target reagents [9–11,19]. In this manner, sensitive and selective detection of chemical and biological species may be achieved.
In conclusion, this work suggests an additional functionality for silicon photonics, in precision SAW mass sensors, surface analysis and thin films characterization. The devices may serve as thickness monitors as part of deposition equipment, with remote access optical fiber connectivity. They could also be applied to control of fabrication processes, reliability testing, and materials science and engineering characterization. The SAW-photonic devices may be produced in large volumes and at low cost. Future work would involve sensitivity enhancement through feedback operation, and the characterization of photo-sensitive layers used in photonic circuits such as various chalcogenide glasses [22,23].
6. Methods
6.1 Numerical analysis of surface acoustic modes
The surface acoustic modes of the SOI layer stack, with and without the deposition of thin layers, were calculated numerically using a commercial platform [18]. Modal profiles and frequencies $\mathrm{\Omega }$ were obtained for different surface acoustic wavelengths $\mathrm{\Lambda }$. The acoustic phase and group velocities were calculated as $\mathrm{\Omega }/K$ and $\partial \mathrm{\Omega }/\partial K$, respectively, where $K \equiv 2\pi /\mathrm{\Lambda }$. The material parameters used in calculations are summarized in Table 1.
Table 1. Materials parameters used in numerical analysis of surface acoustic modes
The density of the germanium sulfide layers was measured as 2900 ± 200 kg×m−3 using Rutherford Backscattering (RBS) analysis (see below). Young’s moduli of the aluminum oxide and germanium sulfide layers were fitted for best agreement between measured and calculated changes in SAW group velocities (see Results).
6.2 Device fabrication
Devices were fabricated in standard SOI substrates with a 220 nm-thick silicon device layer on top of a 2 µm-thick buried oxide layer. Optical waveguides were defined in the silicon device layer using electron-beam lithography, followed by inductively coupled plasma reactive-ion etching. The etching process used a mixture of SF6 and C4F8 gasses, at flow rates of 65 sccm and 10 sccm, respectively. Etching was carried out at a vacuum level of 4×10−10 bar and RF power of 100 W at a 6 nm×s−1 rate. 700 nm-wide ridge waveguides were partially etched to a depth of 70 nm. Vertical grating couplers were patterned at the ends of the bus waveguides of the resonators. The positions of metallic grating stripes were defined by electron-beam lithography, together with the optical waveguides, and partial etching to the same depth. A 5 nm-thick chromium adhesion layer and 20 nm of gold were deposited inside the etched grating patterns by a sputtering process. The sputtering rates for chromium and gold were 0.2 nm×s−1 and 0.5 nm×s−1, respectively. The vacuum level during sputtering was 5×10−6 bar, and the rotation rate was 5 rpm.
Regions of layers depositions were defined by spin coating of AZ-1518 photoresist on top of a SAW-photonic device, followed by photolithography. Atomic layer deposition of aluminum oxide (Al2O3) was carried out using trimethyl-aluminum (TMA) and water vapor at 115 °C. A purge of 40 and 60 seconds was performed following the application of TMA and water, respectively, during each deposition cycle. The deposition chamber was allowed to cool down to room temperature overnight before the removal of the sample. Lastly, the photoresist was lifted off in dimethyl sulfoxide (DMSO) at 80 °C for 10 minutes, followed by sonication in acetone and isopropanol. Germanium sulfide (Ge2S3) layers were thermally evaporated from a GeS2 source in an alumina crucible. Temperature was controlled at 760 °C and the initial chamber pressure was 2 × 10−7 Torr. The layers thicknesses were controlled by a quartz crystal microbalance and verified by atomic force microscopy. At the conclusion of germanium sulfide evaporation, the resist was lifted off in N-methyl-2-pyrrolidone (NMP) at 80 °C for 10 minutes, followed by rinsing in acetone and isopropanol.
6.3 Rutherford backscattering measurements of layer density
Samples for RBS analysis were mounted on a holder using double-side, self-adhesive carbon tape. RBS measurements were performed using a 2.024 MeV ± 1keV 4He+ beam from a 1.7 MV Pelletron accelerator, NEC [26]. The beam current was ∼8 nA, and its nominal diameter was 1.5 mm. One electron suppressor between the beam entrance and the sample holder was biased at −100 V vs. ground. A second suppressor was connected before sample and biased at −1000 V. A normal incidence beam was used in all measurements. Spectra were collected using fixed silicon drift detector (ULTRATM Silicon-Charged Particle Detector, ORTEC) with 15 keV full width at half maximum. The RBS detector scattering angle was 169° (Cornell geometry), and the solid angle was 2.7 msr. The NDF code was used to analyze the data [27]. Depth profiles were extracted automatically from RBS spectra using the Surrey IBA DataFurnace software [28].
The intensities of the peaks corresponding to S and Ge were used to estimate the ratio between the two elements in the deposited layer. The widths of the peaks were related to the areal density of the material, in atoms per unit area [29]. The areal density of the layer was converted to volume density based on atomic force microscope thickness measurements.
Funding
European Research Council (ERC) (SAW-SBS (101001069)).
Disclosures
The authors declare no conflict of interest
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. H. Wohltjen, “Mechanism of operation and design considerations for surface acoustic wave device vapour sensors,” Sens. Actuators 5(4), 307–325 (1984). [CrossRef]
2. D. S. Ballantine Jr and H. Wohltjen, “Surface acoustic wave devices for chemical analysis,” Anal. Chem. 61(11), 704A–715A (1989). [CrossRef]
3. W. D. Bowers, R. L. Chuan, and T. M. Duong, “A 200 MHz surface acoustic wave resonator mass microbalance,” Rev. Sci. Instrum. 62(6), 1624–1629 (1991). [CrossRef]
4. S. Shiokawa and J. Kondoh, “Surface acoustic wave sensors,” Jpn. J. Appl. Phys. 43(5B), 2799–2802 (2004). [CrossRef]
5. K. Länge, B. E. Rapp, and M. Rapp, “Surface acoustic wave biosensors: a review,” Anal. Bioanal. Chem. 391(5), 1509–1519 (2008). [CrossRef]
6. B. Liu, X. Chen, H. Cai, M. M. Ali, X. Tian, L. Tao, Y. Yang, and T. Ren, “Surface acoustic wave devices for sensor applications,” J. Semicond. 37(2), 021001 (2016). [CrossRef]
7. E. Benes, “Improved quartz crystal microbalance technique,” J. Appl. Phys. 56(3), 608–626 (1984). [CrossRef]
8. C. Lu and A. W. Czanderna (Eds.), Applications of piezoelectric quartz crystal microbalances. Elsevier (2012).
9. D. T. Bell and R. C. M. Li, “Surface-acoustic-wave resonators,” Proc. IEEE 64(5), 711–721 (1976). [CrossRef]
10. T. W. Schneider and D. A. Buttry, “Electrochemical quartz crystal microbalance studies of adsorption and desorption of self-assembled monolayers of alkyl thiols on gold,” J. Am. Chem. Soc. 115(26), 12391–12397 (1993). [CrossRef]
11. L. W. Moore, K. N. Sprjnger, J. X. Shi, X. Yang, B. I. Swanson, and D. Li, “Surface acoustic wave chemical microsensors based on covalently bound self-assembled host monolayers,” Adv. Mater. 7(8), 729–731 (1995). [CrossRef]
12. C. Giannetti, B. Revaz, F. Banfi, M. Montagnese, G. Ferrini, F. Cilento, S. Maccalli, P. Vavassori, G. Oliviero, E. Bontempi, L. E. Depero, V. Metlushko, and F. Parmigiani, “Thermomechanical behavior of surface acoustic waves in ordered arrays of nanodisks studied by near-infrared pump-probe diffraction experiments,” Phys. Rev. B 76(12), 125413 (2007). [CrossRef]
13. D. Nardi, M. Travagliati, M. E. Siemens, Q. Li, M. M. Murnane, H. C. Kapteyn, G. Ferrini, F. Parimgiani, and F. Banfi, “Probing thermomechanics at the nanoscale: impulsively excited pseudosurface acoustic waves in hypersonic phononic crystals,” Nano Lett. 11(10), 4126–4133 (2011). [CrossRef]
14. M. Schubert, M. Grossman, O. Ristow, M. Hettich, A. Bruchhausen, E. S. C. Barretto, E. Scheer, V. Gusev, and T. Dekorsy, “Spatial-temporally resolved high-frequency surface acoustic waves on silicon investigated by femtosecond spectroscopy,” Appl. Phys. Lett. 101(1), 013108 (2012). [CrossRef]
15. D. Munk, M. Katzman, M. Hen, M. Priel, M. Feldberg, T. Sharabani, S. Levy, A. Bergman, and A. Zadok, “Surface acoustic wave photonic devices in silicon on insulator,” Nat. Commun. 10(1), 4214 (2019). [CrossRef]
16. M. Katzman, D. Munk, M. Priel, E. Grunwald, M. Hen, N. Inbar, M. Feldberg, T. Sharabani, R. Zektzer, G. Bashan, M. Vofsi, U. Levy, and A. Zadok, “Surface acoustic microwave photonic filters in standard silicon-on-insulator,” Optica 8(5), 697–707 (2021). [CrossRef]
17. M. Hen, D. Munk, M. Katzman, M. Priel, S. Taragin, and A. Zadok, “Surface-acocstic-wave characterization of thin layer deposition on a standard silicon-photonic circuit,” paper STh1R.3 in Conference on Lasers and Electro-Optics (CLEO 2020), OSA Technical Digest (Optical Society of America, 2020).
18. COMSOL Multi-Physics Materials Library, www.comsol.com.
19. A. Mujahid and F. L. Dickert, “Surface acoustic wave (SAW) for chemical sensing applications of recognition layers,” Sensors 17(12), 2716 (2017). [CrossRef]
20. H. H. Diamandi, Y. London, G. Bashan, A. Bergman, and A. Zadok, “Highly-coherent stimulated phonon oscillations in multi-core optical fibers,” Sci. Rep. 8(1), 9514 (2018). [CrossRef]
21. B. J. Eggleton, C. G. Poulton, P. T. Rakich, M. J. Steel, and G. Bahl, “Brillouin integrated photonics,” Nat. Photonics 13(10), 664–677 (2019). [CrossRef]
22. A. E. Owen, A. P. Firth, and P. J. S. Ewen, “Photo-induced structural and physico-chemical changes in amorphous chalcogenide semiconductors,” Philos. Mag. B 52(3), 347–362 (1985). [CrossRef]
23. D. Munk, M. Katzman, Y. Kaganovskii, N. Inbar, A. Misra, M. Hen, M. Priel, M. Feldberg, M. Tkachev, A. Bergman, M. Vofsi, M. Rosenbluh, T. Schneider, and A. Zadok, “Eight-Channel Silicon-Photonic Wavelength Division Multiplexer With 17 GHz Spacing,” IEEE Select. Topics Quantum Electron. 25(5), 1–10 (2019). [CrossRef]
24. M. A. Hopcroft, W. D. Nix, and T. W. Kenny, “What is the Young's Modulus of Silicon?” J. Microelectromech. Syst. 19(2), 229–238 (2010). [CrossRef]
25. H. E. Engan, B. Y. Kim, J. N. Blake, and H. J. Shaw, “Propagation and optical interaction of guided acoustic waves in two-mode optical fibers,” J. Lightwave Technol. 6(3), 428–436 (1988). [CrossRef]
26. G. A. Norton, J. B. Schroeder, G. M. Klody, and M. D. Strathman, “3S-R10 automated RBS system,” Nucl. Instrum. Methods Phys. Res., Sect. B 40-41, 785–789 (1989). [CrossRef]
27. N. P. Barradas, C. Jeynes, and R. P. Webb, “Simulated annealing analysis of Rutherford backscattering data,” Appl. Phys. Lett. 71(2), 291–293 (1997). [CrossRef]
28. C. Jeynes, N. P. Barradas, P. K. Marriott, G. Boudreault, M. Jenkin, E. Wendler, and R. P. Webb, “Elemental thin film depth profiles by ion beam analysis using simulated annealing-a new tool,” J. Phys. D: Appl. Phys. 36(7), R97–R126 (2003). [CrossRef]
29. W. K. Chu, J. W. Mayer, and M. A. Nicolet, Backscattering spectrometry. Academic Press (1978).
