Karl Grosh

Scandium-doped aluminum nitride (ScAlN) is a promising replacement for undoped aluminum nitride in MEMS vibration and acoustic sensors due to its higher piezoelectric coefficients, and for RF MEMS due to its enhanced piezoelectric response and ferroelectric switching capability. However, poor process conditions often lead to degraded film performance. In this work, we optimized the growth conditions of ScAlN thin films deposited by reactive pulsed-DC magnetron sputtering system by studying the impact of N₂ flow rate, target–substrate distance, substrate temperature, and substrate bias on film stress, crystallinity, and surface morphology. Based on stress measurements, XRD rocking curves along the c-axis (002), and roughness with AOG formation probability extracted from AFM and SEM images, an optimized deposition recipe was developed that balances stress, crystallinity, and AOG density. With this optimized recipe, samples were fabricated for dielectric, ferroelectric, and piezoelectric coefficient (d33,f and d31,f) measurements. To verify scalability, d33,f, εr, and tan(δ) were measured on 100, 150, and 200 mm substrates. Dual beam laser interferometry results showed d33,f values of around 18 pm/V, εr of 18, and lowest tan(δ) of 0.4%. Cantilever-based d31,f measurements yielded a value of −6.22 pC/N. The optimized ScAlN films also exhibited remnant polarization, Pr = 130 μC/cm², and coercive field, Ec = 3.5 MV/cm.

Mechanics of Hearing Discussion Moderated by Anthony Peng, Renata Sisto, Sebastiaan Meenderink, Jong-Hoon Nam and related to sessions 4-7: Hair bundle mechanics II, Nonlinear dynamics of the cochlea, Cochlear processing, and Cochlear mechanics: methods and results.

Mechanics of Hearing Discussion Moderated by Ernst Dalhoff, Dáibhid Ó Maoiléidigh and Susan Voss and related to sessions 1, 2 and 3: External and middle ear, Hair bundle I and Otoacoustic emissions.

Mechanics of Hearing Discussion Moderated by Anthony Peng, Renata Sisto, Sebastiaan Meenderink, Jong-Hoon Nam and related to sessions 4-7: Hair bundle mechanics II, Nonlinear dynamics of the cochlea, Cochlear processing, and Cochlear mechanics: methods and results.

Mechanics of Hearing Discussion Moderated by Ernst Dalhoff, Dáibhid Ó Maoiléidigh and Susan Voss and related to sessions 1, 2 and 3: External and middle ear, Hair bundle I and Otoacoustic emissions.

Hearing aids and cochlear implants help patients treat their hearing loss, but have limitations impacting their use rates. Completely implantable auditory prostheses would expand the range of activities a prosthesis user could engage in and enable 24/7 use. However, the lack of a completely implantable microphone that is robust, lightweight, and low noise prevents the wide adoption of implantable devices. Current implantable sensors struggle to meet or exceed the performance necessary for this application. This work develops a discretized and exhaustive design optimization approach to identify multi-bandwidth transducers that meet the 20-phon noise floor over 100 Hz-8 kHz. The design procedure is based on an experimentally validated analytical model that simulates the response of miniature piezoelectric microelectromechanical systems (MEMS) accelerometers. A four-bandwidth accelerometer with constrained proof mass thicknesses is selected as the design that best balances area minimization with ease of manufacturability. The estimated MEMS die dimensions are 825 μm × 575 μm, which is a 23% MEMS die area reduction compared to the previously published dual-bandwidth sensor [A. E. Hake, P. Kitsopoulos, and K. Grosh, IEEE Sens. J., 23(13), 13957-13965 (2023)].

Abstract Mammalian outer hair cells (OHCs) enhance sound amplification and frequency tuning through stereociliary hair bundles (HBs), which convert mechanical motion into electrical signals via mechano-electric transducer (MET) channels. Experiments show that the HB displacement creeps, and the MET current evinces dual timescales of adaptation in response to mechanical stimulus. Understanding these mechanisms is crucial for elucidating normal auditory function and disorders, yet their origins remain unclear. To address this, we developed a mathematical model of the OHC HB that incorporates three rows of stereocilia with distinct nonlinear adaptive gating mechanisms, nonlinear kinematics, and viscoelastic mechanics. Our model accurately replicates experimental responses to fluid-jet stimulation, predicting simultaneous mechanical creep and slow adaptation of the MET current. Using stiff probe stimulation, the same model reveals even faster adaptation, aligning with experimental observations and emphasizing the stimulus-dependence of the response. The model provides new insights into the functional importance of the three-row stereocilia configuration, offering a mechanistic explanation for its ubiquity in mammalian HBs and its role in facilitating the complex timescales of adaptation.

Although diversity, equity, and inclusion (DEI) interventions in science, engineering, and higher education are often discussed as being led at the individual or institutional level, departments can be an effective academic entity for creating meaningful culture change. One way a department can embark on this work is through strategic planning, which can help a diverse group of stakeholders come together to identify a set of goals and pathways for achieving those goals over a sustained amount of time. In this piece, we present an overview of the University of Michigan Department of Mechanical Engineering’s three-phase DEI strategic planning process, which involved proposing strategic planning, creating the strategic plan, and preparing for implementation of the plan. Guiding questions and lessons learned from this process are provided to help other departments create their own locally relevant strategic plans in DEI.

Sound perception arises from oscillations of stereocilia, hair-like structures that convert mechanical motion into electrical signals via ion channels. These signals, depending on the stereocilia's location, drive neural activation and nonlinear sound amplification, crucial for normal hearing. Mathematical models of isolated hair bundles (HBs), composed of stereocilia, provide a mechanistic understanding of these processes. A key parameter in these models is geometric gain, which estimates how mechanical deflections impact gating spring extension, modulating ion channel activity. Traditionally, this gain is approximated as the ratio of horizontal spacing between adjacent stereocilia pivots to their average height. However, this simplified ratio can lead to inaccuracies in predictions of HB sensitivity and stiffness. We analyzed gain at five cochlear locations in adult mice, comparing two definitions: the approximated “stick geometric gain,” based on two infinitesimally thin stereocilia, and the “true geometric gain” derived using our nonlinear two-row isolated HB model with complete morphology. We found that the stick model overpredicts gain by a factor of 1.5–2, resulting in overestimated changes in gating spring tension, sensitivity, stiffness, and narrower activation curves. We discuss the implications of gain-overprediction on the interpretation of experimental results and theoretical models. [Work supported by NIH-R01-NIDCD04084].

The ear relies on the transduction of sound input to mechano-electric transducer (MET) current by hair bundles (HBs) for normal hearing. These HBs, composed of hair-like stereocilia, have ion channels at the tip of all shorter stereocilia that enable transduction. Experiments have shown that HBs stimulated by a step-like fluid-jet force display a displacement creep post initial ascent, termed mechanical rise, and the MET current adapts with a single slow time constant. Under a step-like probe-actuated displacement stimulus, the creep disappears, and the current decays with two (fast and slow) time constants. Since adaptation plays a possible role in precluding cell damage and restoring HB sensitivity upon exposure to loud sounds, it is important to understand the underlying mechanisms for these seemingly disparate responses. We developed a single nonlinear model of an isolated mammalian HB capable of representing the mechanical and electrical response under both the slow fluid jet and much faster stiff probe stimulus. Linearizing the model enabled us to identify the three underlying system time constants (mechanical rise, fast, and slow adaptation), along with a mechanistic explanation of how these three different behaviors arise.

In this study, a method for determining the optimal location and orientation of an implantable piezoelectric accelerometer on the short process of the incus is presented. The accelerometer is intended to be used as a replacement for an external microphone to enable totally implantable auditory prostheses. The optimal orientation of the sensor and the best attachment point are determined based on two criteria-maximum pressure sensitivity sum and minimum loudness level sum. The best location is determined to be near the incudomalleolar joint. We find that the angular orientation of the sensor is critical and provide guidelines on that orientation. The method described in this paper can be used to further optimize the design and performance of the accelerometer.

Abstract In this study a method for determining the best placement and optimal orientation of the implantable piezoelectric accelerometer attached to the short process of the incus is developed. The accelerometer is intended to be used as a replacement for an external microphone to enable totally implantable cochlea implants or as part of a totally implantable hearing aid. The best location is determined to be near the incudomalleolar joint. The optimal orientation of the sensor at the determined location is obtained using two criteria - maximum voltage sum and minimum loudness level sum. The method described in the paper can be used to further optimize the design and the performance of the accelerometer.

The outer hair cells (OHCs) of the mammalian cochlea are the mediators of an active, nonlinear electromechanical process necessary for sensitive, frequency specific hearing. The membrane protein prestin conveys to the OHC a piezoelectric-like behavior hypothesized to actuate a high frequency, cycle-by-cycle conversion of electrical to mechanical energy to boost cochlear responses to low-level sound. This hypothesis has been debated for decades, with two key remaining issues: the influence of the rate dependence of conformal changes in prestin and the OHC transmembrane impedance. In this paper, we mainly focus on the rate dependence of the conformal change in prestin. A theoretical electromechanical model of the OHC that explicitly includes rate dependence of conformal transitions, viscoelasticity, and piezoelectricity. Using this theory, we show the influence of rate dependence and viscoelasticity on electromechanical force generation and transmembrane impedance. Further, we stress the importance of using the correct mechanical boundary conditions when estimating the transmembrane capacitance. Finally, a set of experiments is described to uniquely estimate the constitutive properties of the OHC from whole-cell measurements.

Active metamaterials promise advanced wave control beyond what is achievable with passive structures. Practical bulk devices have yet to be realized, though, for lack of a method to assess the stability of interacting cells. The authors address this obstacle by developing a general stability analysis that requires only the frequency response of an isolated unit cell to determine the stability of metamaterials of many such cells arranged in arbitrary geometries. This analysis is used to accurately predict the stability bounds of an experimental active acoustic metamaterial, and to reveal key constraints that must be respected when designing e.g. waveguides, cloaks, or noise absorbers.

External vibrations cause the eardrum to oscillate, resulting in the excitation of the sensory structures of the cochlea. In this talk, we focus on the hair bundles (HBs) of the outer hair cells situated inside the cochlea. These bundles, protruding from the cells, convert mechanical motion into electrical currents. Experimental observations reveal that the calcium concentration inside the stereocilia (hair-like structures that comprise the HB) influences current adaptation. This regulation impacts shifts of the curve, sensitivity, and active range of bundles. We aim to mechanistically understand the intracellular calcium effects by adjusting the adaptation complex in our HB model. We modified model parameters to match experimental data on current, bundle displacement, and shift trends at different calcium concentrations. This involved increasing the adaptation stiffness, stall force, stereocilia pivot, and gating stiffness. A stiffer adaptation spring reduces ion-channel reclosure, affecting the steady-state response. A higher stall force lessens the effective force on the adaptation complex, replicating experimental observations. Unlike the other properties, pivot, and gating stiffness likely do not depend on calcium concentration. Therefore, we will conduct error minimization analyses to identify adaptation complex properties while maintaining constant pivot and gating stiffness across calcium concentrations. Work supported by NIH grant NIH-NIDCD-R01 04084.

In our current research, we focus on predicting the stationary nonlinear response of a cochlear model that extends from the base to the apex when subjected to harmonic input, taking into account the tapering of the cochlear scalae along with electromechanical coupling of outer hair cells and the microstructures of the organ of Corti. We are interested in explaining the paradoxical phenomenon of hypercompression in the motion of the reticular lamina (RL), whereby the response decreases with an increase in the acoustic excitation at the stapes. We derive frequency–response curves for both the basilar membrane and RL at various locations, exploring a wide range of excitation frequencies and amplitudes. In accordance with experimental data, we find that the RL exhibits hypercompression at the base of the cochlea. This behavior is found to arise from the interaction of saturating nonlinearity arising from the active process and the linear response identified as the passive response to acoustic stimulus. These two responses are not synchronized in phase. As the excitation level increases, the two effects tend to partially cancel, giving rise to lower responses with increasing sound pressure levels. [Work supported by NIH-NIDCD R01 04084.]

The goal of this work is to develop a completely implantable ossicular vibration sensor utilizing microelectromechanical systems (MEMS) technology in combination with integrated-circuit approaches as part of a larger effort to enhance auditory prostheses by eliminating their external components. To develop a completely implantable device, sensors are required to replace the external microphones used in traditional systems (e.g., cochlear implants and hearing aids). Present-day implantable sensors do not meet the stringent requirements for acoustic performance or size to fit in the middle ear; however, in our lab, a dual-resonance design has been developed that holds the potential to achieve these results. In this paper, we show the analytical model used to predict and understand the behavior of these sensors. We discuss preliminary cadaveric temporal bone results, show how a fabricated MEMS proof-of-concept prototype mounted on a printed circuit board (PCB) along with the amplifying electronics and covered with a high-resolution 3D printed lid results in a packaged size small enough to fit in the middle ear, and outline future testing protocols.

Research using the model organism nematode <italic>C. elegans</italic> has greatly facilitated our understanding of sensory biology, including touch, olfaction, taste, vision and proprioception. While hearing had long been considered to be restricted to vertebrates and some arthropods, we recently discovered that <italic>C. elegans</italic> is capable of sensing and responding to airborne sound in a frequency and sound source-size-dependent manner. <italic>C. elegans</italic> auditory sensation occurs when airborne sound physically vibrates their external cuticle (skin) to activate the sound-sensitive mechanosensory FLP/PVD neurons via nicotinic acetylcholine receptors (nAChRs), triggering aversive phonotaxis behavior. Here, we report stepwise methods to characterize these three features of <italic>C. elegans</italic> auditory sensation, including sound-evoked skin vibration, neuronal activation, and behavior. This approach provides an accessible platform to investigate the cellular and molecular mechanisms underlying auditory sensation and mechanotransduction mechanisms in <italic>C. elegans</italic>.

In this paper, the relation between the current and displacement responses due to an external mechanical stimulus on the outer hair cell (OHC) hair bundle (HB) is studied using a theoretical approach. We seek to understand the interplay between the time constants of the external loading and those intrinsic to the HB. To incorporate HB adaptation and channel gating, we used the model of an isolated HB, denoted as the TMJ model. We solved the nonlinear equations for the bundle dynamic response due to an externally applied force that consisted of an exponential temporal rise to a constant value. We determined the dependence of the bundle displacement over which the current continued to increase (the apparent operating range (OR)) on the rise time (τF ) of the applied force. In addition, we developed a model linearized about the resting open probabilities for a given static, biasing load to provide closed-form approximations of the dependence on the stimulus rise time, τF, and bundle adaptation time constants. Finally, we wanted to determine if the inclusion of more precise kinematics of the tip link motion relative to the stereociliary rotation influenced model predictions of the channel opening and bundle stiffness. Hence, we developed geometrical relations between the two rows of stereocilia to establish coupled kinematic relations for inclusion in our HB kinetic model. We predict an OR of 30−50 nm for small τF and an overestimation by a factor of 10 in the OR for τF higher than the slow adaptation time constant. Finally, with accurate bundle kinematics, lower HB displacement, current, and adaptation motor displacement were predicted in contrast to the TMJ model. We are exploring the implications of this model on nanoscale mechano-electrical transduction.

A detailed model of the physical processes occurring within the organ of Corti can be compared with experimental data and can give an indication of the mechanisms of its active behaviour. On the other hand, a model that involves wave propagation in the cochlea can give more insight into its coupled response. In principle many kinds of waves can propagate in the cochlea, but it is a single travelling wave that is generally assumed to mainly determines its coupled response, which can be characterized by the frequency variation of its complex wavenumber. In the case of a cochlear model using a locally reacting basilar membrane (BM) and 1D fluid coupling, this wavenumber can be calculated explicitly. When longitudinal coupling, due to the mechano-electrical structure of the Organ of Corti or other forms of fluid coupling are introduced however, additional wave types are possible, and it is not so straightforward to calculate the wavenumber of the main travelling wave. This paper is the second of a series, in which a method is presented of deriving the wavenumber distribution associated to different wave types, based on an elemental model of the cochlea. This allows an investigation of the effect of different forms of longitudinal coupling on the wave motion. In general, the main travelling wave dominates the BM response, and when the model is active, the imaginary part of its wavenumber is positive in a frequency region just before the characteristic frequency. The extent of this active region depends on the form of longitudinal coupling assumed in the organ of Corti.

The effect of different forms of longitudinal coupling on the active response of the cochlea are analysed using an elemental approach, based on a previous finite element model of the guinea pig cochlea that has three mechanical degrees of freedom. The overall basilar membrane (BM) admittance can be readily calculated using the elemental method, and since this is the only aspect of the organ of Corti dynamics that couples into the fluid, it is a useful indicator of its overall behaviour. As has been shown in previous studies, mechanical longitudinal coupling in the TM, together with 3D fluid coupling, are most important in obtaining a coupled BM frequency response that is both tall and broad, as observed in experimental data. An intuitive and efficient method is used for representing the longitudinal coupling in the fluid, by decomposing the 3D fluid coupling in the wavenumber domain into a local near-field mass loading of the organ of Corti and the analytically simpler 1D fluid coupling, representing long range fluid-structure coupling. The 3D fluid coupling is also necessary to obtain a phase variation that is consistent with experimental measurements. The model of the organ of Corti can be represented as a mechanical three degree of freedom system, and when it is locally reacting all the mechanical elements, and hence the BM admittance, are independent of the wavenumber. When longitudinal coupling is introduced in the TM, the stiffness and damping associated with the TM shear motion then depends on the wavenumber. Similarly, when the near field fluid coupling is associated with the BM mass in the mechanical model, this too is now a function of wavenumber. The model can then be used to calculate a wavenumber-dependent BM admittance, which can be combined with the simple analytical equation for 1D fluid coupling to give the dispersion equation for the waves in the coupled cochlea.

The outer hair cell (OHC) of the mammalian cochlea is the nexus of the active processes giving rise to the nonlinear, biologically vulnerable, acoustic response. We present a model for the behavior of the OHC in view of its mechanical and electrical properties, and the external loading of the cell. Because of the low-pass electrical membrane impedance and rate dependent processes, there is a continuing debate on the mechanism of the amplification process at high frequencies. We will focus on the electrical-to-mechanical energy conversion at the cellular level, and show how we must consider the external mechanical loading of the cell to interpret the power transfer. In addition, we show that simple models can be used to fit in vitro data from experiments, but subtle model changes in the parameters change the predictions of power deposition by the OHCs.

The outer hair cells (OHCs) of the mammalian cochlea are the mediators of an active, nonlinear electromechanical process necessary for sensitive, frequency specific hearing. The membrane protein prestin conveys to the OHC a piezoelectric-like behavior hypothesized to actuate a high frequency, cycle-by-cycle conversion of electrical to mechanical energy to boost cochlear responses to low-level sound. This hypothesis has been debated for decades, and we address two key remaining issues: the influence of the rate dependence of conformal changes in prestin and the OHC transmembrane impedance. We develop a theoretical electromechanical model of the OHC that explicitly includes rate dependence of conformal transitions, viscoelasticity, and piezoelectricity. Using this theory, we show the influence of rate dependence and viscoelasticity on electromechanical force generation. Further, we stress the importance of using the correct mechanical boundary conditions when estimating the transmembrane capacitance. Finally, a set of experiments is described to uniquely estimate the constitutive properties of the OHC from whole-cell measurements.

Summary Despite lacking ears, the nematode C. elegans senses airborne sound and engages in phonotaxis behavior, enabling it to locate and avoid sound sources1. How worms sense sound, however, is not well understood. Here, we report an interesting observation that worms respond only to sounds emitted by small but not large speakers, indicating that they preferentially respond to localized sound sources. Notably, sounds emitted by small speakers form a sharp sound pressure gradient across the worm body, while sounds from large speakers do not, suggesting that worms sense sound pressure gradients rather than absolute sound pressure. Analysis of phonotaxis behavior, sound-evoked skin vibration and sound-sensitive neuron activities further support this model. We suggest that the ability to sense sound pressure gradients provides a potential mechanism for worms to distinguish sounds generated by their predators, which are typically small animals, from those produced by large animals or background noise. As vertebrate cochlea and some insect ears can also detect sound pressure gradients, our results reveal that sensing of sound pressure gradients may represent a common mechanism in auditory sensation across animal phyla.

According to the Center for Disease Control, hearing loss is the third most common chronic physical condition in the US, and it is more prevalent than diabetes or cancer. Most cases of the permanent hearing loss results from damage to the sensory cells of the inner ear. These sensory hair cells (called outer hair cells) provide us with sensitive hearing through a delicate process of amplification of the sound-evoked motions of the inner ear epithelia.

For the last 20 years, researchers have developed accelerometers to function as ossicular vibration sensors in order to eliminate the external components of hearing aid and cochlear implant systems. To date, no accelerometer has met all of the stringent performance requirements necessary to function in this capacity. In this work, we present an accelerometer design with an equivalent noise floor less than 20 phon equal-loudness-level over a 0.1-8 kHz bandwidth in a package small enough to be implanted in the middle ear. Our approach uses a dual-bandwidth (two sensing elements) microelectromechanical systems piezoelectric accelerometer, sized using an area-minimization process based on an experimentally-validated analytical model of the sensor. The resulting bandwidth of the low-frequency sensing element is 0.1-1.25 kHz and that of the high-frequency sensing element is 1.25-8 kHz. These sensing elements fit within a silicon frame that is 795 μm × 778 μm, which can reasonably be housed along with a required integrated circuit in a 2.2 mm × 2.7 mm × 1 mm package. The estimated total mass of the packaged system is approximately 14 mg. This dual-bandwidth MEMS sensor fills a technological gap in current completely implantable auditory prosthesis research and development by enabling a device capable of meeting physical and performance specifications needed for use in the middle ear.

The mammalian cochlea is responsible for transforming incoming acoustic energy into neural signals. Efficient modeling of the cochlear response is extremely challenging, because of the length scales, which vary from the sub-micron to centimeters, and time scales, which vary from microseconds to seconds, that must be resolved. In the current work, we predict the stationary nonlinear response of a base-to-apex cochlear model to a harmonic input, considering the taper of the cochlear scalae. We seek to understand the influence of bulk fluid viscosity and geometric fluid-duct tapering on the nonlinear response of the system. Coupled equations are derived from a kinematically constrained Langrangian dynamics formulation. To solve these equations, we have used an iterative algorithm, the alternating frequency-time method. The algorithm swaps between frequency and time domains using the Fourier and inverse Fourier transforms, and is based on a fixed-point iteration. We analyzed the solutions for nonlinear models where the frequencies and stimulus levels are varied from 6 kHz to 20 kHz, and 10 dB SPL to 90 dB SPL, respectively. We show that our model predicts previously unexplained responses in the cochlea, including so-called hypercompression (when increasing sound levels produce a reduced cochlear response).

The electromotility of the outer hair cell (OHC) contributes to the sensitivity of the mammalian cochlea by amplifying traveling waves through electrical-to-mechanical energy conversion realized at the molecular level by an electromotile protein called prestin. Rate-dependent effects, including viscous damping, transmembrane electrical impedance, and state-dependent conformal transitions, hold the potential to attenuate OHC-mediated active processes at high frequencies thereby rendering prestin ineffective in high frequency cycle-by-cycle amplification. Determining the upper frequency limit of prestin remains a central challenge in cochlear biophysics. In this study, we will build a simplified OHC model to explore the influence of the rate dependence on active force generation and power deposition. Through numerical investigations, the proposed model can be used to model charge and electromotility data from in vitro experiments, but subtle variations in the rate parameters significantly change the predictions of electromechanical force as well as the power deposition. Based on these theoretical considerations, we propose an experimental approach to consistently determine the rate dependence and other OHC response parameters by characterizing the electrical and mechanical behavior about a resting position. This approach holds the potential to conclusively determine the upper frequency limit of prestin under physiological resting conditions.

When designing an engineered electroacoustic sensor, a key question is “what is the lowest level sound that can be sensed?” To answer this, we design to achieve a desired input referred noise. In the cochlea, there are many sources of internal noise, such as Johnson noise arising from membrane conductances, clatter noise in channels, and thermoviscous damping that will conspire to cause vibrational responses in the absence of external stimulus. Noise can dramatically affect our ability to sense desired sounds. Experimental and theoretical cochlear mechanics has focused on determining the sensitivity of the cochlea. However, to our knowledge, there is only one set of published measurements of the displacement response of the cochlea in quiet (the levels are low, below 1 atto-meter2 per Hz), and no simulations in a global cochlear model of the noise response. We introduce a method for predicting the global response to noise from electromotile outer hair cells in the cochlea for small fluctuations about equilibrium using our finite-element based numerical model. Furthermore, we show the relative contribution of different noise sources (in particular, channel noise versus conductance noise) and the spatial distribution of the response to these sources.

Moderate to severe hearing loss is a debilitating condition that affects over 5% of the world’s population. Hearing aids and cochlear implants positively impact the lives of those who suffer from sensorineural hearing loss. However, both display numerous limitations that affect their adoption and use rates, notably including those associated with the external elements of these devices (e.g., microphone and signal processor). These external components impact device safety, appearance, acoustic performance, and ease-of-use. A totally implantable auditory prosthesis would help to address these issues by eliminating external components. A major barrier to progress toward this goal is the lack of a completely implantable acoustic sensor capable of matching or exceeding the performance of commercial external microphones. Our previous studies have indicated that piezoelectric microelectromechanical systems (MEMS) accelerometers have the potential to function as implantable sensors within the middle ear meeting a 20-phon noise floor over a 100–8 kHz range. In this paper, we describe a process to design and fabricate dual and tri-resonance devices that are comprised of piezoelectric cantilever bimorph beams tip-loaded by a proof mass, can meet or exceed the 20-phon noise floor, and can produce a measurable voltage output when deflected by sound-induced ossicular vibrations.

Sound waves vibrating the eardrum excite the ossicles in the middle ear ultimately driving waves in the cochlea. Cochlear vibrations are processed by inner hair cells and outer hair cells (OHCs). Our focus is on the OHCs that nonlinearly amplify the sound converting a time-varying motion of its apically adorned hair bundle (HB) to an alternating current. The OHC HB consists of roughly three rows of stereocilia arranged according to their heights. Understanding how the bundle stiffness, sensitivity, and transduction current depend on the physiology and anatomy of the stereocilia is crucial and open question. Therefore, we are developing a three-row model of an isolated HB to quantify each row’s contribution to the passive and active mechanics of the HB. The derived equations of motion include the nonlinear kinematics, viscoelastic HB mechanics, and the nonlinear response of the mechano-electric transducer channels coupled to an adaptation mechanism. We also linearize the model to conduct stability analysis and determine the dependence of the responses on the rate constants. Our preliminary results show a higher current influx through the middle row than the shortest row and a more significant stiffness contribution from the middle-to-tallest row pair compared to the shortest-to-middle row pair connection.