Mehmet R. Dokmeci

A hermetic glass-silicon micropackage with high-density on-chip feedthroughs for sensors and actuators

This paper describes the development of a hermetic micropackage with high-density on-chip feedthroughs for sensor and actuator applications. The packaging technique uses low-temperature (320/spl deg/C) electrostatic bonding of a custom-made glass capsule (Corning 7740, 2/spl times/2/spl times/8 mm/sup 3/) to fine grain polysilicon in order to form a hermetically sealed cavity. High-density on-chip multiple polysilicon feedthroughs (200 per millimeter) are used for connecting external sensors and actuators to the electronic circuitry inside the package. A high degree of planarity over feedthrough areas is obtained by using grid-shaped polysilicon feedthrough lines that are covered with phosphosilicate glass (PSG), which is subsequently reflown at 1100/spl deg/C in steam for 2 h. Saline and DI water soak tests at elevated temperatures (85 and 95/spl deg/C) were performed to determine the reliability of the package. Preliminary results have shown a mean time to failure (MTTF) of 284 days and 118 days at 85 and 95/spl deg/C, respectively, in DI water. An Arrhenius diffusion model for moisture penetration yields an expected lifetime of 116 years at body temperature (37/spl deg/C) for these packages. In vivo tests in guinea pigs and rats for periods ranging from one to two months have shown no sign of infection, inflammation, or tissue abnormality around the implanted package.

A high-sensitivity polyimide capacitive relative humidity sensor for monitoring anodically bonded hermetic micropackages

This paper presents the design, fabrication and complete characterization of a high-sensitivity polyimide-based humidity sensor for monitoring internal humidity level in anodically bonded hermetic micropackages. This capacitive sensor is 1 mm on a side and utilizes CU1512 polyimide film with a thickness in the range from 300 /spl Aring/ to 1200 /spl Aring/ sandwiched between two metal electrodes to sense moisture. The measured sensitivity for a sensor with a 1200-/spl Aring/-thick film is 0.86 pF/%RH, and for a 300-/spl Aring/-thick sensor is 3.4 pF/%RH. The sensor has been exposed to and survived a one-hour test at 400/spl deg/C, which is the temperature typically used to perform anodic bonding. Measurements show a drift of less than 1% RH at 50% RH and 37/spl deg/C for 48 h, and a hysteresis of <2% RH over a range from 30 to 70% RH for a 1200-/spl Aring/-thick polyimide film sensor. The measured breakdown voltage of the sensor (1200 /spl Aring/ thick) exceeds 20 V and agrees well with other results.

A passive humidity monitoring system for in situ remote wireless testing of micropackages

Reports a small passive wireless humidity monitoring system (HMS) for continuous remote monitoring of humidity changes inside miniature hermetic packages, presents its application in determining hermeticity of an implantable biomedical package, and presents long-term performance results obtained from packages implanted in guinea pigs. This 7/spl times/1.2/spl times/1.5 mm/sup 3/ system consists of a high-sensitivity capacitive humidity sensor that forms an LC tank circuit together with a hybrid coil wound around a ferrite substrate. The resonant frequency of the circuit changes when the humidity sensor capacitance changes in response to changes in humidity. The HMS can resolve humidity changes of /spl plusmn/2.5% RH over a 2-cm range. The resolution is sufficient enough to monitor internal package humidity for either in vitro or in vivo testing.

A passive humidity monitoring system for in-situ remote wireless testing of micropackages

This paper reports a small passive wireless humidity monitoring system (HMS) for continuous monitoring of humidity changes inside miniature hermetic packages, presents its application in determining hermeticity of an implantable biomedical package, and presents long-term performance results obtained from packages implanted in guinea pigs. This 7/spl times/1.2/spl times/1.5 mm/sup 3/ system consists of a high-sensitivity capacitive humidity sensor that forms an LC tank circuit together with a hybrid coil wound around a ferrite substrate. The resonant frequency of the circuit changes when the humidity sensor capacitance changes in response to changes in humidity. The HMS can resolve humidity changes of /spl plusmn/2.5%RH over a 2 cm range. The resolution is sufficient enough to monitor internal package humidity for either in in-vitro or in-vivo testing.

The biocompatibility, integrity, and positional stability of an injectable microstimulator for reanimation of the paralyzed larynx

The biocompatibility, integrity, positional stability, and potential use of hermetically sealed injectable wireless micromachined microstimulators were investigated for reanimation of the paralyzed larynx. The device, consisting of silicon and glass, has been tested and proven to be biocompatible with no evidence of pathological tissue reaction or rejection up to one-year implantation in the rat dorsum and canine larynx. By one month, each unit was encapsulated by a thin membrane, which thickened to form a fibrous layer of less than 500 /spl mu/m at 6-12 months. The microstimulators demonstrated long-term in vivo durability: the hermetic seal of only one in ten devices was breached. Once implanted, migration of the device varied from 0 to 3 cm, depending upon the extent of surgical dissection. No discernable migration was noted when the tissue dissection was minimal. Studies utilizing a modified device equipped with electrodes indicated that migration was nominal with sufficient positional stability to ensure activation of target muscles for glottis opening. This study supported the feasibility of using a microstimulator for reanimation of paralyzed laryngeal muscles that open the airway during breathing. This innovative approach to treatment would alleviate the need for a tracheotomy or surgical resection of the vocal fold.

Accelerated testing of anodically bonded glass-silicon packages in salt water

This paper reports long-term accelerated test results of glass-silicon packages in aqueous solutions including phosphate buffered saline (PBS) and de-ionized water. An analysis of the dominant failure mode and the long term biocompatibility of the packages is also performed. Using the accelerated data at 85 and 95/spl deg/C in saline with an Arrhenius model we calculate an activation energy of 1.26 eV and an expected lifetime of 177 years at 37/spl deg/C in saline. After an analysis of the failed samples, we find that the main failure mechanism is the dissolution of the polysilicon layer at elevated temperatures in saline, which causes premature failure of the package. In animal models the package is consistently found to be biocompatible and robust after implantation periods of up to a year.

Modeling and characterization of sacrificial polysilicon etching using vapor-phase xenon difluoride

Xenon Difluoride is an isotropic dry etch which is increasingly being used to release structures made of polysilicon. By using a vapor-phase XeF/sub 2/ pulse etching system we have investigated the effects of aperture size and thickness of polysilicon films versus etch rates. Decreasing the aperture size resulted in reduced etch rates. For a 100 /spl mu/m wide structure the etch rates varied from 1.38 /spl mu/m/min (1.25 /spl mu/m thick polysilicon) down to 0.41 /spl mu/m/min (0.1 /spl mu/m thick polysilicon). Rate constants are obtained for a modified, aperture-dependent Deal-Grove model from the experimental data.

Hermeticity Testing Of Glass-silicon Packages With On Chip Feed Throughs

This paper reports on hermeticity test results performed on a silicon-glass hermetic package in aqueous environments. The package utilizes anodic bonding of glass to silicon at temperatures below 320/spl deg/C to form a hermetically sealed cavity large enough to contain hybrid components. Up to 200 sealed polysilicon feedthrough lines can be obtained per mm at the edge of the package. Hermeticity is maintained by planarizing these feedthrough lines, and then depositing a smooth top poly silicon surface for anodic bonding of the glass capsule. Accelerated long term tests indicate a lifetime of at least many years. In-vivo tests in guinea pigs for two months have shown the package to be biocompatible and rugged.

Long-term hermeticity and biological performance of anodically bonded glass-silicon implantable packages

This paper reviews long-term test results obtained from a series of tests on glass-silicon (Si) hermetically sealed packages. Results are presented from 1) a 9.9-year ongoing room temperature phosphate-buffered saline (PBS) soak test of four packages; 2) accelerated soak tests in high temperature saline of 28 samples resulting in an extrapolated mean-time-to-failure (MTTF) at 37/spl deg/C of 177 years; 3) a 2.7-year in vitro 97/spl deg/C PBS soak test of a single package; and 4) in situ hermeticity and biocompatibility tests from 12 packages implanted in four guinea pigs-three packages implanted in two guinea pigs (each) for 1 month and another two guinea pigs for 20 and 22 months. All of the packages remained hermetically sealed over the lifetime of the implant. A detailed histological report of the implants is provided suggesting that they elicit no profound adverse reaction from the body.

A high-sensitivity polyimide humidity sensor for monitoring hermetic micropackages

This paper presents the fabrication and characterization of a high-sensitivity polyimide-based humidity sensor for monitoring internal humidity level in anodically-bonded hermetic micropackages. This capacitive sensor is 1 mm on a side, and utilizes CU1512 polyimide film with a thickness in the range of 300-1200 /spl Aring/ sandwiched between 2 metal electrodes to sense moisture. The measured sensitivity for a sensor with a 1200 /spl Aring/-thick film is 0.86 pF/%RH, and for a 300 /spl Aring/-thick sensor is 3.4 pF/%RH. The sensor has been exposed to and survived a one-hour test at 400/spl deg/C, which is the temperature typically used to perform anodic bonding. Measurements show an offset drift of less than 1% RH at 50% RH and 37/spl deg/C, and a hysteresis of <2% RH over a range of 30-70% RH for a 1200 /spl Aring/-thick polyimide film sensor.