Cochlear implant (CI) is a surgically implanted electronic device that provides a sense of sound to a person who suffers from profound hearing loss or deafness. The present generation of hearing systems bypass the normal hearing process. Outside the skin, it consists of a microphone, a speech processor, and a transmitter, which transmits signals through an internal receiver to an array of electrodes placed in the cochlea. Though, in the last three decades the technology has undergone an impressive improvement, there are still some challenges to be addressed for higher wearing comfort [1
]. Because of the external units, the system is visible, making the patients stigmatized. It also limits the activities that can be undertaken while wearing the device. Broken wires, cables, and speech processors can cause derangement, too. This could be minimized by having a fully implantable cochlear implant (FICI, sometimes also referred to as totally implantable cochlear implant, TICI) which functions round-the-clock while sleeping, showering, swimming, and during many types of vigorous physical activities [2
]. FICI is supposed to be composed of an internal microphone or a piezoelectric acoustic sensor, an electronic device which transform the signal for the hearing nerves, a battery and/or an energy harvesting unit, as well as a multielectrode array inserted in the cochlea.
Several solutions have been proposed to mimic the frequency selectivity (tonotopy) of the cochlea. The topologically more faithful physical model is an elastic membrane having varying widths along its length [3
]. The other approach is to apply an array of micro-electromechanical system (MEMS) cantilevers having varying length, and thus, varying natural resonance frequency. Because of the smaller size and more reliable fabrication procedure, the latter approach seems to have higher potential for FICI. In 1998 Harada et al. demonstrated a fishbone structured acoustic sensor using piezoresistive read-out elements [8
]. Later, Xu et al. reported polymeric micro-cantilever array to mimic the mammalian cochlea [9
]. Recently, Jang et al. used an aluminum nitride (AlN) coated array as an artificial basilar membrane [10
]. Though the rectangular unimorph cantilevers showed excellent frequency selectivity, they covered only the upper half of the human hearing range (2.9–12.6 kHz) [11
], since the natural resonance frequency at fixed cantilever thickness scales up with decreasing length. To obtain sensitive cantilevers in the range of 300–700 Hz is, however, more challenging.
Low frequency, spiral, and spiral-like cantilevers were proposed and theoretically evaluated by Choi et al. in 2006 [12
], and recently, also experimentally demonstrated. Zhang et al. [13
], and Lu et al. [14
] reported an S-shaped Lead Zirconate Titanate (PZT) coated flexure suspended MEMS device on a chip size of 6 mm × 6 mm, for vibration sensing and energy harvesting. Though the covered frequency range is attractive for the proposed FICI concept, a smaller footprint, higher Q-factor (>100), and bio- and metal-oxide-semiconductor (CMOS) compatible piezoelectric material, like AlN [15
], are needed for the device to be implantable in the human middle ear.
Voice detection in CIs is done directly by perceiving the modulation of air or fluid pressure using a microphone, even if the functions of the tympanic membrane and of the middle ear are intact. An alternative approach, also an aim in this paper, is to measure the vibration of the ossicles with a miniaturized MEMS based implant. As a guideline for the design, we referred to the work of Gan et al. [17
] on implantable middle ear hearing devices (IMEHDs). In their work, a small magnet was mounted between the malleus and stapes, and driven electromagnetically by a coil placed under the ear canal bony wall. The diameter and length of the cylinder-shaped magnet is 1.5 and 2 mm, respectively, and it weighs 26 mg, which is comparable to the size and mass of a 3D packaged Si multicantilever system. Beker et al. demonstrated a rectangular cantilever with Si seismic mass to achieve resonance in this lower voice frequency range (474 Hz) [18
]. However, the size (6 mm × 6 mm) and weight of the chip, especially in packaged multichannel form, is too large to be fixed onto one of the middle ear bones. Moreover, the applied technique to bond bulk piezoceramic PZT dices to the substrate, and the grinding, make the wafer scale processing highly demanding.
Here, we demonstrate an array of spiral cantilevers with Si seismic mass at their ends, which is optimized to achieve a small footprint, compactness, bio- and CMOS compatibility, low resonance frequency (300–700 Hz), high Q-factor in air (117–254), and high robustness with limited internal stress in the Si cantilever. These 2 mm × 2 mm cantilever chips are small enough to pack them in a compact multichannel device which fits into the middle auditory system, and can provide a new solution for next generation FICI systems.