Method for Operation of Resonator
1. A method of sensing a target analyte in a liquid sample including the steps:
- providing a bulk acoustic wave resonator device;
placing a liquid sample on the bulk acoustic wave resonator device;
operating the bulk acoustic wave resonator device to generate bulk acoustic waves; and
measuring a shift in the fundamental resonant frequency of the bulk acoustic wave resonator device,wherein the bulk acoustic wave resonator device comprises a resonator structure and an acoustic decoupling layer, the resonator structure comprising;
a piezoelectric material layer;
electrodes arranged to apply a driving signal to the piezoelectric material layer to generate bulk acoustic waves; and
a resonator structure surface,wherein;
the resonator structure has a resonator structure acoustic impedance and the liquid sample has a liquid sample acoustic impedance;
the acoustic decoupling layer is formed over the resonator structure surface; and
the acoustic decoupling layer has an acoustic decoupling layer acoustic impedance, the acoustic decoupling layer acoustic impedance being;
up to ⅕
times or not less than 5 times the resonator structure acoustic impedance, andup to ⅕
times or not less than 5 times the liquid sample acoustic impedance.
Disclosed is a method of sensing a target analyte in a liquid sample using a bulk acoustic wave resonator device. The liquid sample is placed on the bulk acoustic wave resonator device which is operated to generate bulk acoustic waves. A shift in the fundamental resonant frequency of the bulk acoustic wave resonator device is measured. The bulk acoustic wave resonator device comprises a resonator structure and an acoustic decoupling layer. The resonator structure comprises: a piezoelectric material layer; electrodes arranged to apply a driving signal to the piezoelectric material layer to generate bulk acoustic waves; and a resonator structure surface. The acoustic decoupling layer is formed over the resonator structure surface. The acoustic decoupling layer acoustic impedance is: up to ⅕ times or not less than 5 times the resonator structure acoustic impedance, and up to ⅕ times or not less than 5 times the liquid sample acoustic impedance.
- 1. A method of sensing a target analyte in a liquid sample including the steps:
providing a bulk acoustic wave resonator device; placing a liquid sample on the bulk acoustic wave resonator device; operating the bulk acoustic wave resonator device to generate bulk acoustic waves; and measuring a shift in the fundamental resonant frequency of the bulk acoustic wave resonator device, wherein the bulk acoustic wave resonator device comprises a resonator structure and an acoustic decoupling layer, the resonator structure comprising; a piezoelectric material layer; electrodes arranged to apply a driving signal to the piezoelectric material layer to generate bulk acoustic waves; and a resonator structure surface, wherein; the resonator structure has a resonator structure acoustic impedance and the liquid sample has a liquid sample acoustic impedance; the acoustic decoupling layer is formed over the resonator structure surface; and the acoustic decoupling layer has an acoustic decoupling layer acoustic impedance, the acoustic decoupling layer acoustic impedance being; up to ⅕
times or not less than 5 times the resonator structure acoustic impedance, and
up to ⅕
times or not less than 5 times the liquid sample acoustic impedance.
- View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13)
The project leading to this application has received funding from the European Union'"'"'s Horizon 2020 Research and Innovation programme under grant agreement No 636820.
The present invention relates to a method of operation of a resonator. The invention has particular, but not exclusive, applicability to a bulk acoustic wave resonator (BAWR), such as a thin film bulk acoustic wave resonator (FBAR) or a solidly mounted resonator (SMR). Such resonators are of use in sensor applications, such as in gravimetric-based sensing.
Acoustic resonators can be used for sensing of various chemical and biological agents in air. Such resonators are typically small and highly sensitive to mass changes, and have a relatively low manufacturing cost. For a typical acoustic mass sensor, the resonant frequency of the acoustic-wave resonator changes in response to the mass change on the resonator surface. Typically, such sensors are not suitable for sensing applications in liquid, because the thickness longitudinal mode (TLM) experiences a substantial reduction in Q factor. This reduction is typically large, usually more than 90%. The reduction of the Q factor is considered to be due to damping caused by loss of acoustic energy into the liquid.
Zhang and Kim (2005) disclose using an FBAR sensor in longitudinal mode. The sensor has a Q factor of 200-300 at about 1 GHz in air, allowing a vapour mass change of 10−9 g/cm2 at its surface. In air, the sensor is operated in fundamental resonance mode. However, when the sensor is operated in liquid, the Q factor is reduced by several orders of magnitude, due to damping. Zhang and Kim (2005) propose instead to operate the sensor in liquid at its second harmonic resonance, at about 2 GHz, and report a Q factor of about 40. Therefore operating at the second harmonic resonance reduces the effects of damping by the liquid, but there is still more than 88% reduction in Q factor. Pottigari and Kwon (2009) form an FBAR device using a piezoelectric ZnO film sandwiched between Al electrodes as a transducer. A layer of parylene is supported over the uppermost electrode on supporting posts, with a vacuum gap between the parylene layer and the uppermost electrode. This approach was used in order to investigate the sensitivity of the FBAR device to sensing in liquid environments, with the parylene layer constituting a passive sensing diaphragm. In effect, the supporting posts guide the acoustic waves from the transducer into the passive sensing diaphragm. An effect is shown which is that the Q factor of the driving resonator is reduced by only 9% when sensing in water compared with air. The posts are essential to support the parylene layer and to withstand the pressure differential between the vacuum and the external atmosphere and they are the only way to transmit acoustic waves from the resonator to the sensible area. Additionally, the device structure proposed by Pottigari and Kwon (2009) is fragile and not robust enough for practical applications.
Xu et al (2011) discloses an FBAR sensor used in conjunction with microfluidic confinement in order to limit the available damping provided by the liquid. The thickness of the microfluidic channel is critical to this reduced damping and substantially limits the utility of the sensor to sample a suitable volume of liquid.
The present inventors have realised that it would be advantageous to provide a resonator device that is operable in liquid and which provides useful effects compared with prior art resonator devices, and propose a method of sensing a target analyte in a liquid sample using the device.
Accordingly, in a first preferred aspect, the present invention provides a method of sensing a target analyte in a liquid sample including the steps: (i) providing a bulk acoustic wave resonator device, (ii) placing a liquid sample on the bulk acoustic wave resonator device, (iii) operating the bulk acoustic wave resonator device to generate bulk acoustic waves, (iv) measuring a shift in a resonant frequency of the bulk acoustic wave resonator device.
The bulk acoustic wave resonator device comprises a resonator structure and an acoustic decoupling layer, the resonator structure comprising a piezoelectric material layer, electrodes arranged to apply a driving signal to the piezoelectric material layer to generate bulk acoustic waves, and a resonator structure surface. The resonator structure has a resonator structure acoustic impedance and the liquid sample has a liquid sample acoustic impedance. The acoustic decoupling layer is formed over the resonator structure surface. The acoustic decoupling layer has an acoustic decoupling layer acoustic impedance, the acoustic decoupling layer acoustic impedance being up to ⅕ times or not less than 5 times the resonator structure acoustic impedance, and up to ⅕ times or not less than 5 times the liquid sample acoustic impedance. In other words, the acoustic impedance of this layer is, by a factor of at least 5, smaller or larger than the respective acoustic impedances of the resonator structure and of the liquid sample.
Where the resonator structure includes a series of layers, the relevant resonator structure acoustic impedance may be taken as the acoustic impedance of the uppermost layer of the resonator structure.
By providing an acoustic decoupling layer with such an acoustic mismatch with respect to the acoustic impedance of the liquid sample, and of the resonator, the layer can act as an acoustic isolation layer. The acoustic impedance of the acoustic decoupling layer may be a factor of more than 10 times smaller or larger than the resonator structure acoustic impedance, and of the liquid sample acoustic impedance. However, the acoustic mismatch is preferably not so high as to prevent propagation of the wave through the acoustic decoupling layer for sensing. This is confirmed by the expression R=(Z/Z0−1)/(Z/Z0+1), where R is the reflection coefficient. If Z/Z0 is too large, the intensity of the wave passing through the interface is too low and can be vanished in the layer due to acoustic losses. A mismatch of 10 (i.e. 10 times larger or smaller) gives a transmission of 19% of the intensity, which is considered to be sufficient transmission to allow for useful sensing.
The acoustic impedance of the acoustic decoupling layer, resonator, and liquid sample may be measured in a manner well-known to the skilled person. Generally, to measure acoustic impedance, the sound velocity of the selected mode in the material is determined, and multiplied by mass density of the material. As the skilled person will be aware, the acoustic impedance of a material strongly depends on the propagation mode. In the present invention, therefore the acoustic impedance of the different elements is assessed with respect to the propagation mode used for the bulk acoustic wave resonator device. The acoustic impedance of a material typically does not vary with frequency in the range of operation of devices discussed herein. The bulk acoustic wave resonator device proposed herein may be an SMR type device. Alternatively, the device may be an FBAR type device. The fundamental resonant frequency of the bulk acoustic wave resonator device may be in the range 1-5 GHz.
The following references set out suitable methods for determining the acoustic impedance as used in this disclosure: DeMiguel-Ramos et al (2015); Iborra et al (2013).
Further optional features of the invention are set out below. These may be applied singly or in any suitable combination.
Preferably, the thickness of the acoustic decoupling layer is between 1/8 λ and 50 λ. λ is the wavelength in the decoupling layer at a fundamental resonant frequency of the resonator. More preferably, the upper limit of the thickness range of the acoustic decoupling layer may be selected to be e.g. 20 λ, 10 λ, or 2 λ. The lower limit of the thickness range may be e.g. 1/8 λ, 1/4 λor ½λ. Preferably, the thickness of the acoustic decoupling layer is a multiple of ½λ.
The acoustic decoupling layer may be formed from a number of materials, given that they fulfil the acoustic impedance requirements discussed above. Preferably, the acoustic decoupling layer comprises carbon nanotubes. Other examples of possible materials suitable for use as an acoustic decoupling layer may include, but are not limited to, polydimethylsiloxane (PDMS), parylene, and SU-8. Typical acoustic impedance values for different materials are set out further below.
As the skilled person will readily understand, the density of the acoustic impedance layer will have an effect on the sound velocity of the selected mode in the layer. With carbon nanotubes, the density of the layer will be affected by the number density of carbon nanotubes. Accordingly, where the acoustic impedance layer comprises carbon nanotubes, the density of carbon nanotubes in the layer may be at least 0.3×1010 cm−2, alternatively may be at least 0.5×1010 cm−2, at least 1×1010 cm−2, or at least 2×1010 cm−2.
The acoustic decoupling layer may include binding sites specific for the analyte in the liquid sample. For example, the acoustic decoupling layer may be functionalised to promote attachment of specific analytes.
The liquid sample may be an aqueous liquid, e.g. deionised (DI) water or a biological buffer. The mass density of typical biological buffers varies in the range 1 to 1.3 times of that of DI-water. Sound velocity between such liquids typically varies less than 3%. Therefore the acoustic impedance of a typical liquid sample is in the range of 1.48 MRayl (corresponding to pure water) to 2 MRayl for a highly concentrated buffer. These values are for typical buffers used in biological detection. Blood, for instance, has an acoustic impedance of 1.66 MRayl. The liquid sample may partially penetrate the acoustic decoupling layer, although this is not necessary.
Embodiments of the invention will now be described by way of example with reference to the accompanying drawings in which:
In recent years the need for miniature, low cost, label-free and ultra-sensitive chemical and biological sensors has grown significantly. Gravimetric sensors based on acoustic resonators may be beneficial in achieving these objectives. Quartz crystal microbalances (QCMs) are the most well-established gravimetric sensors; these operate generally in the shear mode [Lu, 1972; Benes, 1984]. Mass accretion on the sensor surface (which can be functionalised to only permit certain molecules to attach) decreases the resonant frequency of the sensor, fr; this is the basis for sensing. This frequency shift, Δfr, and hence the mass sensitivities, Sm, are proportional to f2 according to the Sauerbrey equation [Sauerbrey, 1959]. Better sensitivities may therefore be achievable by operating at higher frequencies. Thin film technology has led to emergence of thin film bulk acoustic wave (BAW) resonators operating in the range of 1-5 GHz.
BAW resonators generally consist of a thin piezoelectric slab of aluminium nitride (AlN) or zinc oxide (ZnO) “sandwiched” between two metallic electrodes [Flewitt, 2015]. The application of an AC signal across the electrodes generates a standing wave in the piezoelectric layer, which resonates at a frequency that is dependent on the film thickness [Rosenbaum, 1945]. There are two general types of BAW resonators: film bulk acoustic resonators (FBARs) and solidly mounted resonators (SMRs). FBARs and SMRs are different in the way the acoustic wave is confined. Typically, FBARs use air for acoustic isolation. Typically, SMRs use an acoustic reflector consisting of layers alternating between low acoustic impedance and high acoustic impedance [Aigner, 2003; García-Gancedo, 2013]. High frequency thin film BAW resonators normally operate in the thickness longitudinal mode (TLM) which is excited in piezoelectric films with a c-axis oriented parallel to the surface normal. Thus the application of an AC signal causes the mechanical deformation to be parallel to the direction of wave propagation [Auld, 1990]. Biosensing with such longitudinal mode thin film BAW resonators is usually performed in dry conditions, because the TLM couples efficiently into liquids, which may lead to severe damping (>90% drop in quality-factors at resonance, Q). This generally prevents TLM resonators from being used in liquid environments, which is a major hindrance for both chemical and biological sensing as it requires surfaces to be dry when measurement takes place. The thickness shear mode (TSM) is instead preferred because it does not compress in liquids and hence has lower attenuation [Wingqvist, 2010].
In-liquid TSM resonators in the GHz regime have already been shown in real-time measurements of antibody-antigen binding [Weber, 2006], study of DNA hybridization [Gabl, 2004; Nirschl, 2009], and organophosphorus pesticide detection [Chen, 2013]. TSM resonances can be excited by using a piezoelectric layer with a c-axis inclined between 20° and 40° with respect to the surface normal. This may be achieved by an off-axis growth of the film [Wingqvist, 2007; DeMiguel-Ramos, 2015; DeMiguel-Ramos, 2017; Clement, 2014]. Both the shear and the longitudinal modes are excited (quasi-shear mode) in such piezoelectric films, with the shear mode typically resonating at a lower frequency compared with the longitudinal mode. Nonetheless, damping may still occur when the shear resonance propagates through more viscous liquids [Rughoobur, 2016; Qin, 2011]. Additionally, the shear mode thin film BAW resonators compromise their sensitivity to be only around one-third that of a typical longitudinal mode device with similar dimensions, due to its lower operating frequency [Weber, 2006].
The scalability of fabricating such inclined c-axis films may be limited compared to the highly oriented c-axis films due to significant and expensive modifications necessary for the depositions as reported in previous works [Yanagitani, 2007]. It would therefore be preferably to find a means of using the TLM for in-liquid sensing. Different methods have been attempted in several previous works to reduce damping in liquids such as confining the liquid using microchannels [Xu, 2009; Xu, 2011], however improper channel thicknesses can lead to exponential decay of the TLM. In-liquid sensing using the second harmonic of the TLM has also been reported, although this can lead to a lower Q of around 40, which compromises measurement of the frequency shift and hence sensitivity [Zhang, 2005]. Pottigari and Kwon [Pottigari, 2009] designed FBARs using a vacuum gap isolation created by fabricating parylene microposts for resonators with lower damping with Q of about 140 in liquid. Nonetheless vacuum gapped FBARs have not been explored fully yet. Furthermore, parylene is fragile and hence may be prone to collapse. Therefore a robust interface layer that can prevent liquids from direct contact with the active area, while increasing the sensitivity of the biosensors is desirable. The present inventors have developed a new approach to using TLM devices for mass sensing in liquid environments. An interface layer is added between the resonator and the liquid. This is designed so that the resonant acoustic wave preferably penetrates only partially into the interface layer, and preferably does not penetrate into the bulk liquid. The interface layer itself may be functionalised to allow mass attachment (from molecules in the liquid) which is detected by the resonant acoustic wave. In this way, the interface layer can decouple the resonant device from the process of mass attachment from the liquid whilst retaining mass sensitivity. The dual-purpose acoustic decoupling and mass attachment layer may conveniently be a carbon nanotube (CNT) layer, although the inventors consider that a number of other materials may be suitable for use as the acoustic decoupling layer. Other examples of possible materials suitable for use as an acoustic decoupling layer may include, but are not limited to, polydimethylsiloxane (PDMS), parylene, and SU-8.
For reference, the acoustic impedance of various materials is set out in the table below.
Owing to their excellent electronic and mechanical properties (Stiffness coefficient of about 1 TPa, acoustic velocity of about 62000 m/s) [Treacy, 1996; Iborra, 2013], CNTs have great potential for use in applications ranging from molecular electronics to ultrasensitive biosensors. Their biological compatibility and affinity provides them with specific chemical handles that would make several of these applications possible [Wang, 2003]. The larger surface area to volume ratio of CNTs can make them an attractive alternative to metallic electrodes for improving the sensitivities and detection limits of such resonators. Indeed CNT integration on the surface can increase the binding area without substantially increasing the size of the devices. Consequently, for the same concentration of biological samples, a larger number of targeted molecules can be bound onto the resonator surface compared to devices with metal electrodes; both ZnO [Garcia-Gancedo, Al-Naimi, 2011] and AlN [García-Gancedo, Zhu, 2011] resonators with higher sensitivities due to the presence of a CNT layer have already been demonstrated in air using the TLM. In addition, CNTs having lower density (1.3 g/cm3) [Mann, 2010] which, depending on their morphology, can reduce mass loading and their acoustic impedance, can be designed to ensure maximum confinement of the acoustic wave in the piezoelectric material [Iborra, 2013].
When the CNTs are grown directly on the substrates, they can form different morphologies depending on their number density. With number densities as high as about 1011-1013 cm−2, the growth of vertically-aligned (VA) CNT forests may be desirable in certain applications [Zhong, 2012]. Although in this work only the forest number density will be considered, the low mass densities of CNT forests ranging from 0.03 g/cm3 [Fubata, 2006] to 1.6 g/cm3 [Sugime, 2014] can reduce mass loading on the resonators compared with the use of metals. Depending on their morphology, CNTs can have different acoustic impedances, which can be tuned to ensure a sufficient mismatch with the acoustic impedance of the piezoelectric material to yield appropriate confinement of the acoustic wave in the piezoelectric resonator [Iborra, 2013; Mann, 2010]. CNTs offer an additional benefit compared to metals in that it is possible to functionalize CNTs for direct covalent bonding to molecules, or with different types of polymers [Joddar, 2016] which could be used for non-covalent binding. This can prevent non-specific attachment or the need for additional binding layers. Ultimately, CNTs have a relatively simple chemical composition and atomic bonding configuration, being entirely composed of carbon atoms, which provide a natural match to organic molecules. Hence, the potential exists to use chemically-modified CNTs which would function as both the electrode and the sensing layer [Garcia-Gancedo, Iborra, 2011]. However, electrodes made only with CNTs may introduce parasitic resistances, which can deteriorate the Q factor at resonance, Qr, in air [Esconjauregui, 2015].
In particular, in the following discussion of examples, the inventors propose either quasi-shear mode or thickness longitudinal mode AlN based SMRs, and a method of sensing an analyte in a liquid using such SMRs. Such devices are shown schematically in
SMRs as shown schematically in
- 1. Form a layer of n-type (100)-oriented Si
- 2. Thermal growth of SiO2, Sputter Mo, SiO2, Mo and SiO2, perform chemical mechanical polishing
- 3. Deposit Ir electrodes and pattern using Mo hardmask
- 4. Deposit AlN seed, then piezoelectric AlN layer
- 5. Deposit Mo top electrode, pattern then etch
- 6. Pattern and deposit Al/Fe catalyst
- 7. Grow CNTs using CVD process
The SMRs are fabricated on (100)-oriented, n-type, 500 μm thick Si wafers, which have been thermally oxidized (wet-oxidation) to obtain about 620 nm of SiO2. Reflector layers consisting of Mo (about 650 nm) and SiO2 (about 620 nm) are then sputtered sequentially in a Leybold Z500 sputtering system to achieve a reflection centred around a frequency of 2.2 GHz for the longitudinal mode. The top reflector layer is polished mechanically using alumina slurry to reduce the roughness to less than 2 nm. Ir (about 150 nm) is deposited as bottom electrode by electron beam evaporation and patterned by Ar ion-milling using a hard mask of Mo. AlN sputtering is performed in an in-house ultra-high-vacuum system, pumped to a base pressure below 8×10−7 Pa. A pulsed DC source (MKS ENI 235, Andover, Mass., USA) operating at 250 kHz powers a high purity (99.999%), 150 mm diameter Al target. An AlN seed layer is grown prior to the piezoelectric AlN, at high pressure (0.66 Pa) and 600 W without intentionally heating or biasing the substrate. With these conditions seed layers with mainly (103) orientation are deposited, which is beneficial for promoting the growth of the subsequent AlN films with inclined grains. AlN films with the wurtzite c-axis inclined up to 24° with respect to the surface normal in the substrates placed between 2 and 5 cm from the target axis.
Multi-walled CNTs are grown on the devices by chemical vapor deposition (CVD) method in a cold wall chamber (Black Magic, Aixtron SE, Herzogenrath, Germany). The substrates are loaded in the chamber, which is then evacuated to a base pressure of 6.0 Pa with a rotary pump. Ammonia (NH3) is introduced at a rate of 100 sccm in the chamber to raise the pressure to 16 Pa and stabilized for 30 s. NH3 is chosen as a reducing gas since it is typically more efficient than H2 to reduce catalyst particles to its metallic state especially at low temperature [Hofmann, 2009]. The substrate is subsequently heated up from room temperature to the growth temperature (ranging from 450° C. to 650° C.) at a rate of 3° C./s and stabilized for 300 s. This ramping rate is chosen as a compromise to prevent delamination of the layers while avoiding Oswald ripening of the catalyst. Upon heating, the Fe catalyst is reduced and dewets to form nanoislands. By varying the thickness of Fe (from 0.5 nm to 4.5 nm), different sizes of nanoislands are formed. Different sizes of nanoislands can lead to CNTs with different densities and thicknesses. After the annealing step, the carbon feedstock acetylene (C2H2) is added to commence growth. The ratio of NH3 and C2H2 is 1:1 to prevent the deposition of amorphous carbon during the growth, which is carried out at a pressure of 32 Pa. After the growth time has elapsed, the heater is turned off, C2H2 and NH3 are switched off before flowing Ar (200 sccm) until the system cooled down to room temperature. Attempts are also made with ZnO as a piezoelectric material, however the ZnO film reacts with NH3 and is reduced to Zn. A passivation layer that remains unreacted in the annealing step is therefore advantageous to protect ZnO for CNT growth on the active area.
CNTs are characterized using field emission FE-SEM (LEO 1530VP scanning electron microscope, LEO electron microscopy Inc, New York, USA) using an acceleration voltage ranging from 0.5 keV to 5 keV. Cross-sections of cleaved samples, onto which the CNT forests have been grown, are analyzed to assess the height and forest density of the CNT layer in more detail. The CNT structures are also assessed by Raman spectroscopy (inVia confocal Raman microscope, Renishaw plc, Gloucestershire, UK) using a 532 nm wavelength laser. The scan range is chosen to be from 1000 cm−1 to 3500 cm−1 as there were no peaks at lower wavenumbers to suggest the presence of SWCNTs.
Scanning electron microscope (SEM) images shown in
With 4.5 nm Fe, the tubes grow in a random “spaghetti”-like morphology. The growth rate may be difficult to estimate with this kind of CNTs. Therefore higher magnification images are obtained to estimate the thickness of the layer. The growth rate does not change significantly as the tubes are randomly oriented and are not vertically aligned. Estimation of the number of tubes in the forest is carried out using the SEM images. Five sets of 2 μm length regions are chosen and visible tubes are counted manually as measurements M1, M2, M3, M4 and M5 respectively. An average of the number of tubes in each set is shown in Table 2. This method gives a comparison of the number of CNTs in a region.
Optical microscope (OM) images shown in
The Raman spectra displayed in
Similarly with 4.5 nm Fe the IG/ID is further reduced to 0.84, thereby demonstrating lower crystallinity in the CNTs grown with a thicker Fe layer.
The devices are electro-acoustically characterized on a coplanar probe station using ground-signal-ground (G-S-G) probes with 150 μm pitch (Picoprobes, GGB industries Inc., Naples, Fla., USA). A network analyzer (Model E5062A, Keysight Technologies, Santa Rosa, Calif., USA) is used to measure the frequency response of the devices from 0.5 to 3.0 GHz. The frequency corresponding to the maximum of the real part of the electrical admittance, Y, is considered as fr. The frequency corresponding to the maximum of the real part of the electrical impedance, Z is considered as fa. To assess the resonator performance, Qr and effective electromechanical coupling, keff2, coefficient are determined from standard definitions [Iborra, 2013]:
where ΦY is the phase of Y.
50 μL of deionised (DI) water is added by a micropipette on top of the resonator to determine the performance of the device with CNTs in liquids. For in-liquid mass sensing, BSA dissolved in water with a concentration of 20 mg/mL (B86657-5 mL) is purchased from Sigma Aldrich Ltd. This solution which is then diluted with DI water into 3 different concentrations: 50 μg/mL (1:399 BSA:DI water), 1000 μg/mL (1:19 BSA:DI water) and 2500 μg/mL (1:7 BSA:DI water). Pure DI water is initially used for each device tested to obtain the unloaded fr. The device is subsequently dried. 50 μL of one of the three concentrations of BSA solution is dropped by means of a micropipette on the SMRs covering the GSG probe as well and the Y spectra are measured after 5 mins to allow the BSA to bind to the CNT. The SMR responses for the remaining BSA concentrations are then measured similarly.
The electrical admittance (Y(S)) spectra measured in the frequency range from 0.5 GHz to 3.0 GHz of the SMRs fabricated with and without CNTs are shown in
The average of the TSM and TLM (
where Qair is the unloaded Q in air, and Qwater is the Q in a liquid environment.
Table 3 shows a quantitative analysis of the effect of different CNT forest densities on the average shear mode Qair, Qwater, and the dissipation factor D. The standard deviation from the mean value is used to estimate the variability in the values.
In Table 3, D is calculated for the thickness shear mode (TSM) and it is observed that with the presence of medium and high density CNTs on the active area, D improves from (0.75±0.26)% to (0.63±0.20)% and (0.39±0.23)% respectively. Nonetheless with low density CNTs a higher acoustic leakage is observed as D increases to 1.33. This may be caused by the lower initial Qair of the devices.
Table 4 shows a quantitative analysis of the effect of different CNT forest densities on the average longitudinal mode Qair, Qwater, and the dissipation factor D. The standard deviation from the mean value is used to estimate the variability in the values.
In Table 4 the corresponding values of D are calculated for the TLM and a remarkable reduction in acoustic leakage in DI water is observed (D decreases from 30% without CNTs to less than 1.5% by the presence of CNTs). Indeed without the CNTs, the TLM suffers from large acoustic damping in DI water, having Q values of only about 3-5, whereas with LD CNTs this Q value increases to about 64, and with HD CNTs, an average Q value of about 160, and D around about 0.45% is achieved in DI water. Using taller HD CNT forests of length 30 μm, devices with TLM Qwater above 200 have been measured indicating that the TLM resonance is significantly isolated from the DI water droplet (see
A possible concern might be that the acoustic wave is so confined within the resonator that there is no longer any mass sensing ability of the device. Therefore, the ability of the CNT layer to both acoustically isolate the piezoelectric resonator from the liquid and allow detectable mass attachment is assessed using bovine serum albumin (BSA).
With the addition of BSA solutions (even at the saturation concentration of about 2.5 mg/mL) on top of taller CNT forests (about 30 μm), there are no significant changes to the fr and Q as illustrated by the electro-acoustic responses in
The effect of this partial isolation on the SMRs mass sensitivity is assessed using bovine serum albumin (BSA). The effect of BSA solutions of concentrations ranging from 0.05 mg/mL to 2.5 mg/mL on both the TSM and TLM resonances with devices having about 15 μm tall CNT layers are shown in
Although a trend is evident as time progresses after dropping the 2.5 mg/mL BSA solution, the Δfr are almost two orders (10−6 for taller CNTs compared to 10−4 for 15 μm heights) of magnitude lower than in the case for a lower Qr (about 150 with about 15 μm CNTs) resonator in liquid. The positive Δfr in the case of the longitudinal mode may be caused by temperature drift due to evaporation of the BSA solution from the surface of the resonator, leading to surface cooling of the resonator. Nonetheless both the TSM and TLM shift by only a few kHz, which is just above the noise floor for frequency regimes higher than 1 GHz. In addition the concentration of the BSA solution is high enough to saturate the CNT forest of about 15 μm as demonstrated previously.
It is deduced that thicker CNT forests can improve the Q values in liquid by isolating the device completely from the liquid, yet this can reduce the mass sensitivity significantly to the extent that the SMR may no longer be suitable for gravimetric sensing. Hence a trade-off in the selection of the forest height and density may be necessary to achieve both a reasonable Q and mass sensitivities for the devices.
Nevertheless it is necessary to identify some preferred features when fabricating CNT forest to isolate the TLM resonance in liquid for biological detection. Firstly, it is possible for the forest to be so tall that there is no resonance at the liquid CNT interface. Secondly, the CNT forests may be densified when used in liquids that have low vapour pressure such as isopropyl alcohol (IPA) and acetone. This type of densification can consequently lead to voids, or can expose the resonator surface to the liquid, thereby leading to a significant dissipation.
Another possible limitation is that the high temperature required for growing CNTs via CVD may render them unsuitable for growth on plastic substrates. As an alternative, CNTs can be grown on an independent substrate and transferred into a liquid solution via an ultrasonic spinner. The solution containing debundled CNTs can then be deposited over the devices with a pipette, eliminating the need for high temperature processing. Although in this case dense and VA-CNTs would not be feasible and alternative methods such as indium assisted transfer VA-CNTs reported by Barako [Barako, 2014] would be necessary. Another substitute can be graphene layers, but the acoustic leakage could be significant where the graphene layer is thin.
It is possible to simulate the effect of varying the acoustic impedance of the acoustic decoupling layer on the reflection coefficient provided by the interface between the top of the resonator structure and the acoustic decoupling layer and between the acoustic decoupling layer and the liquid. This is shown in
In this work, the growth of CNTs of different morphologies on the active area of SMRs is achieved by CVD at 600° C. varying the catalyst (Fe) thickness. The purpose of the CNT layer is to both isolate the resonator from liquids such as DI water to reduce acoustic damping by confining the acoustic wave in the SMR, in other words, to act as an acoustic isolation layer, or acoustic decoupling layer, and to permit mass attachment so as to give a measureable resonant frequency shift. AlN SMRs with both a TSM (1.1 GHz) and a TLM (1.9 GHz) resonance are fabricated to assess and compare the performance of CNT coated devices to non-coated SMRs in DI water. In this work, it is demonstrated that SMRs with highly packed CNT forests achieve the lowest TLM dissipation factors of 0.45% compared to 30% in uncoated SMRs. Additionally TSM resonances also suffer from a lower acoustic wave leakage (D decreases from 0.75% to 0.39%) into the deionised water droplet due to the presence of a highly packed CNT forest on the active area. Despite the presence of a tall CNT forest of up to about 15 μm heights, the SMRs are sensitive to mass loading caused by BSA solutions. The result in this work is significant in that the TLM can be made to operate in deionised water by using such an acoustic isolation layer. The present invention can therefore potentially lead to the successful scaling up of FBAR based biosensors for in-liquid applications.
While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.
All references referred to above and/or listed below are hereby incorporated by reference.
Lu, C. -S.; Lewis, O. Investigation of film-thickness determination by oscillating quartz resonators with large mass load. Journal of Applied Physics 1972, 43, 4385.
Benes, E. Improved quartz crystal microbalance technique. Journal of Applied Physics 1984, 56, 608.
Sauerbrey, G. Verwendung von Schwingquarzen zur Wägung dünner Schichten und zur Mikrowägung. Zeitschrift für Physik 1959, 155, 206.
Ruby, R. Review and Comparison of Bulk Acoustic Wave FBAR, SMR Technology. Ultrasonics Symposium, 2007. IEEE. 2007; pp 1029-1040.
Rosenbaum, J. F. Bulk Acoustic Wave Theory and Devices; Artech House: New York, 1945.
Aigner, R. MEMS in RF Filter Applications: Thin-film Bulk Acoustic Wave Technology. Sensors Update 2003, 12, 175-210.
García-Gancedo, L.; Pedrós, J.; Iborra, E.; Clement, M.; Zhao, X. B.; Olivares, J.; Capilla, J.; Luo, J. K.; Lu, J. R.; Milne, W. I. et al. Direct comparison of the gravimetric responsivities of ZnO-based FBARs and SMRs. Sensors and Actuators B: Chemical 2013, 183, 136-143.
Auld, B. A. Acoustic Fields and Waves in Solids, Vol. 2; Krieger Pub Co, 1990.
Wingqvist, G. AlN-based sputter-deposited shear mode thin film bulk acoustic resonator
(FBAR) for biosensor applications—A review. Surface and Coatings Technology 2010, 205, 1279.
Weber, J.; Albers, W. M.; Tuppurainen, J.; Link, M.; Gabl, R.; Wersing, W.; Schreiter, M. Shear mode FBARs as highly sensitive liquid biosensors. Sensors and Actuators A: Physical 2006, 128, 84-88.
Gabl, R.; Feucht, H. D.; Zeininger, H.; Eckstein, G.; Schreiter, M.; Primig, R.; Pitzer, D.; Wersing, W. First results on label-free detection of DNA and protein molecules using a novel integrated sensor technology based on gravimetric detection principles. Biosensors and Bio-electronics 2004, 19, 615.
Nirschl, M.; Blither, A.; Erler, C.; Katzschner, B.; Vikholm-Lundin, I.; Auer, S.; Vörös, J.; Pompe, W.; Schreiter, M.; Mertig, M. Film bulk acoustic resonators for DNA and protein detection and investigation of in vitro bacterial S-layer formation. Sensors and Actuators A: Physical 2009, 156, 180-184.
Chen, D.; Wang, J.; Xu, Y.; Li, D.; Zhang, L.; Li, Z. Highly sensitive detection of organophosphorus pesticides by acetylcholinesterase-coated thin film bulk acoustic resonator mass-loading sensor. Biosensors and Bioelectronics 2013, 41, 163-167.
Wingqvist, G.; Bjurstrom, J.; Liljeholm, L.; Yantchev, V.; Katardjiev, I. Shear mode AlN thin film electro-acoustic resonant sensor operation in viscous media. Sensors and Actuators B: Chemical 2007, 123, 466-473.
DeMiguel-Ramos, M.; Mirea, T.; Clement, M.; Olivares, J.; Sangrador, J.; Iborra, E. Optimized tilted c-axis AlN films for improved operation of shear mode resonators. Thin Solid Films 2015, 590, 219-223.
DeMiguel-Ramos, M.; Diaz-Duran, B.; Escolano, J. -M.; Barba, M.; Mirea, T.; Olivares, J.; Clement, M.; Iborra, E. Gravimetric biosensor based on a 1.3 GHz AlN shear-mode solidly mounted resonator. Sensors and Actuators B: Chemical 2017, 239,1282-1288.
Clement, M.; Iborra, E.; Olivares, J.; DeMiguel-Ramos, M.; Mirea, T.; Sangrador, J. On the effectiveness of lateral excitation of shear modes in AlN layered resonators. Ultrasonics 2014, 54, 1504-1508.
Rughoobur, G.; DeMiguel-Ramos, M.; Mirea, T.; Clement, M.; Olivares, J.; Diaz-Duran, B.; Sangrador, J.; Milne, W. I.; Iborra, E.; Flewitt, A. J. Room temperature sputtering of inclined c-axis ZnO for shear mode solidly mounted resonators. Applied Physics Letters 2016, 108, 034103.
Qin, L.; Chen, Q.; Cheng, H.; Chen, Q.; Li, J. -F.; Wang, Q. -M. Viscosity sensor using ZnO and AlN thin film bulk acoustic resonators with tilted polar c-axis orientations. Journal of Applied Physics 2011, 110, 094511.
Yanagitani, T.; Kiuchi, M.; Matsukawa, M.; Watanabe, Y. Characteristics of Pure-shear Mode BAW Resonators Consisting of (112 0) Textured ZnO Films. Ultrasonics, Ferroelectrics, and Frequency Control, IEEE Transactions on 2007, 54, 1680-1686.
Xu, W.; Zhang, X.; Yu, H.; Abbaspour-Tamijani, A.; Chae, J. In-Liquid Quality Factor Improvement for Film Bulk Acoustic Resonators by Integration of Microfluidic Channels. Electron Device Letters, IEEE 2009, 30, 647-649.
Xu, W.; Zhang, X.; Choi, S.; Chae, J. A High-Quality-Factor Film Bulk Acoustic Resonator in Liquid for Biosensing Applications. Microelectromechanical Systems, Journal of 2011, 20, 213-220.
Zhang, H.; Kim, E. S. Micromachined Acoustic Resonant Mass Sensor. Microelectromechanical Systems, Journal of 2005, 14, 699-706.
Pottigari, S. S.; Kwon, J. W. Vacuum-gapped film bulk acoustic resonator for low-loss mass sensing in liquid. Proc. Actuators and Microsystems Conf. TRANSDUCERS 2009-2009 Int. Solid-State Sensors. 2009; pp 156-159.
Treacy, M. M. J.; Ebbesen, T. W.; Gibson, J. M. Exceptionally high Young'"'"'s modulus observed for individual carbon nanotubes. Nature 1996, 381, 678-680.
Iborra, E.; Sangrador, J.; Clement, M.; Mirea, T.; DeMiguel-Ramos, M.; Olivares, J.; Capilla, J.; García-Gancedo, L.; Esconjauregui, S.; Flewitt, A. J. et al. Acoustic properties of carbon nanotube electrodes in BAW resonators. European Frequency and Time Forum International Frequency Control Symposium (EFTF/IFC), 2013 Joint. 2013; pp 984-987.
Wang, S.; Humphreys, E. S.; Chung, S. -Y.; Delduco, D. F.; Lustig, S. R.; Wang, H.; Parker, K. N.; Rizzo, N. W.; Subramoney, S.; Chiang, Y. -M. et al. Peptides with selective affinity for carbon nanotubes. Nature Materials 2003, 2,196-200.
García-Gancedo, L.; Al-Naimi, F.; Flewitt, A. J.; Milne, W. I.; Ashley, G. M.; Luo, J. K.; Zhao, X.; Lu, J. R. ZnO-Based FBAR resonators with carbon nanotube electrodes. IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control 2011, 58, 2438-2445.
García-Gancedo, L.; Zhu, Z.; Iborra, E.; Clement, M.; Olivares, J.; Flewitt, A. J.; Milne, W. I.; Ashley, G. M.; Luo, J. K.; Zhao, X. B. et al. AlN-based BAW resonators with CNT electrodes for gravimetric biosensing. Sensors and Actuators B: Chemical 2011, 160, 1386-1393.
Mann, M.; Zhang, Y.; Teo, K.; Wells, T.; El Gomati, M.; Milne, W. Controlling the growth of carbon nanotubes for electronic devices. Microelectronic Engineering 2010, 87, 1491-1493.
Zhong, G.; Warner, J. H.; Fouquet, M.; Robertson, A. W.; Chen, B.; Robertson, J. Growth of Ultrahigh Density Single-Walled Carbon Nanotube Forests by Improved Catalyst Design. ACS Nano 2012, 6, 2893-2903.
Futaba, D. N.; Hata, K.; Yamada, T.; Hiraoka, T.; Hayamizu, Y.; Kakudate, Y.; Tanaike, 0.; Hatori, H.; Yumura, M.; lijima, S. Shape-engineerable and highly densely packed single-walled carbon nanotubes and their application as super-capacitor electrodes. Nature materials 2006, 5, 987-994.
Sugime, H.; Esconjauregui, S.; D'"'"'Arsié, L.; Yang, J.; Makaryan, T.; Robertson, J. Growth Kinetics and Growth Mechanism of Ultrahigh Mass Density Carbon Nanotube Forests on Conductive Ti/Cu Supports. ACS Applied Materials & Interfaces 2014, 6, 1544-15447.
Joddar, B.; Garcia, E.; Casas, A.; Stewart, C. M. Development of functionalized multi-walled carbon-nanotube-based alginate hydrogels for enabling biomimetic technologies. Scientific Reports 2016, 6, 32456.
García-Gancedo, L.; Iborra, E.; Clement, M.; Olivares, J.; Zhu, Z.; Flewitt, A. J.; Milne, W. I.; Ashley, G. M.; Luo, J. K.; Zhao, X. B. et al. Solidly mounted resonators with carbon nanotube electrodes for biosensing applications. Frequency Control and the European Frequency and Time Forum (FCS), 2011 Joint Conference of the IEEE International. 2011; pp 1-5.
Esconjauregui, S.; Makaryan, T.; Mirea, T.; DeMiguel-Ramos, M.; Olivares, J.; Guo, Y.; Sug-ime, H.; D'"'"'Arsié, L.; Yang, J.; Bhardwaj, S. et al. Carbon nanotube forests as top electrode in electroacoustic resonators. Applied Physics Letters 2015, 107, 133106.
Robertson, J.; Hofmann, S.; Cantoro, M.; Parvez, A.; Ducati, C.; Zhong, G.; Sharma, R.; Mattevi, C. Controlling the Catalyst During Carbon Nanotube Growth. Journal of Nanoscience and Nanotechnology 2008, 8, 6105-6111.
DeMiguel-Ramos, M.; Clement, M.; Olivares, J.; Capilla, J.; Sangrador, J.; Iborra, E. Induced surface roughness to promote the growth of tilted-AlN films for shear mode resonators. European Frequency and Time Forum International Frequency Control Symposium (EFTF/IFC), 2013 Joint. 2013.
Yokoyama, T.; Iwazaki, Y.; Nishihara, T.; Ueda, M. Analysis on electromechanical coupling coefficients in AlN-based bulk acoustic wave resonators based on first-principle calculations. 2012 IEEE International Ultrasonics Symposium. 2012.
Barako, M. T.; Gao, Y.; Won, Y.; Marconnet, A. M.; Asheghi, M.; Goodson, K. E. Reactive Metal Bonding of Carbon Nanotube Arrays for Thermal Interface Applications. IEEE Trans-actions on Components, Packaging and Manufacturing Technology 2014, 4, 1906-1913.
Mirea, T.; Olivares, J.; Clement, M.; DeMiguel-Ramos, M.; de Frutos, J.; Sangrador, J.; Iborra, E. Carbon nanotube growth on piezoelectric AlN films: influence of catalyst under-layers. RSC Adv. 2015, 5, 80682-80687.
Olivares, J.; Mirea, T.; Diaz-Duran, B.; Clement, M.; DeMiguel-Ramos, M.; Sangrador, J.; de Frutos, J.; Iborra, E. Growth of carbon nanotube forests on metallic thin films. Carbon 2015, 90, 9-15.
Hofmann, S.; Blume, R.; Wirth, C. T.; Cantoro, M.; Sharma, R.; Ducati, C.; Hävecker, M.; Zafeiratos, S.; Schnoerch, P.; Oestereich, A. et al. State of transition metal catalysts during carbon nanotube growth. The Journal of Physical Chemistry C 2009, 113, 1648-1656.
Lakin, K. M. Acoustic bulk wave composite resonators. Applied Physics Letters 1981, 38, 125.
Flewitt, A. J.; Luo, J. K.; Fu, Y. Q.; Garcia-Gancedo, L; Du, X. Y. ; Lu, J. R.; Zhao, X. B.; Iborra, E; Ramos, M; Milne, W. I. ZnO based SAW and FBAR devices for bio-sensing applications. J. Non-Newtonian Fluid Mech. 2015, 222, 209-216.
DeMiguel-Ramos, M. et al. Assessment of the Shear Acoustic Velocities in the Different Materials Composing a High Frequency Solidly Mounted Resonator. Ultrasonics, 2015, 62 p195-199.