16-04-2013, 03:36 PM
Packaging and Non-Hermetic Encapsulation Technology for Flip Chip on Implantable MEMS Devices
[attachment=53091]
Abstract
We report here a successful demonstration of a
flip-chip packaging approach for a microelectromechanical systems
(MEMS) device with in-plane movable microelectrodes implanted
in a rodent brain. The flip-chip processes were carried out
using a custom-made apparatus that was capable of the following:
1) creating Ag epoxy microbumps for first-level interconnect;
2) aligning the die and the glass substrate; and 3) creating nonhermetic
encapsulation (NHE). The completed flip-chip package
had an assembled weight of only 0.5 g significantly less than the
previously designed wire-bonded package of 4.5 g. The resistance
of the Ag bumps was found to be negligible. The MEMS microelectrodes
were successfully tested for its mechanical movement
with microactuators generating forces of 450 μN with a displacement
resolution of 8.8 μm/step.
INTRODUCTION
PACKAGING HAS been a significant challenge in the development
of cost-effective microelectromechanical systems
(MEMS) devices. Often, packaging issues and processes
do not get sufficient attention until the MEMS front-end processes have been completely developed. MEMS packaging
costs now typically contribute more than 50% [1]–[6] of the
total material/assembly cost. The reasons for the higher packaging
cost are partly because MEMS devices typically involve
active mechanical structures, requiring stringent packaging
constraints [7] and different kinds of unconventional sealing
techniques, such as nonhermetic [for pressure sensors and biomedical
MEMS (bio-MEMS)] and hermetic (for RF MEMS).
METHODS
MEMS Testing Samples
There are three different types of MEMS testing samples
that have been used in this study for the development of flipchip
packaging technology: 1) folded beam (chevron) type of
actuator; 2) linear actuator; and 3) linear actuator with a rotating
gear. Chevron actuator mechanism, as shown in Fig. 1, has the
smallest feature size of about 1.5 μm and has the most complex
actuationmechanism that finally moves the microelectrode with
the size of 50-μm width and 4-μm thickness. Therefore, the
results from chevron actuator are reported and discussed more
frequently in the study. As the flip-chip technology reported
in this paper has been tested successfully in three different
complex actuator mechanisms (as shown in Fig. 5), we expect
the flip-chip method reported here to be applicable for other
MEMS designs.
Flip-Chip Assembly Steps
The overall steps involved in the proposed flip-chip packaging
approach for the MEMS testing sample using the custommade
machine are shown in Fig. 6. These processes can be
extended for packaging other MEMS test samples as well.
1) Kitting process to prepare the three main components:
theMEMS chip, glass substrate, and Omnetics (Omnetics
Connector Corporation) third-level interconnects (TLIs).
2) Bumping process involved dispensing Ag epoxy (E3001;
Epoxy Technology, Inc.) on each of the Al pads to create
microscale bumps to build the FLIs with a height of
approximately 100 μm (a total of 18 bumps per a testing
sample).
Black Epoxy Seal
The remaining three sides of the die which were not accessed
by the MEMS microelectrodes were sealed by black epoxy.
A commercially available epoxy [from Epoxy Technology;
mixing two parts, namely, part no. 353ND, part A (90%), and
part no. 353ND, part B (10%)] cured at 120 ◦C for 5 min was
used to generate the near-hermetic seal. The literature suggests
that the epoxy does not produce a complete hermetic seal as
per military standard but rather a near-hermetic seal [19]. At
this stage, we have not yet tested the level of hermeticity of the
black epoxy seal. A small opening was left on the back side
of the die for air ventilation. This air ventilation was necessary
to keep the pressure inside the MEMS chip enclosure close to
ambient during implantation.
RESULTS
Package Size and Weight
The flip-chip technology reduced the final package weight
significantly from 4.5 g (using wire-bonding technology) to
0.5 g, which is almost an order of magnitude reduction in the
overall assembled weight. Since the weight of the head of an
adult rat is approximately 30 g (for a rat weighing 300 g),
the aforementioned significant weight reduction will now allow
neuronal monitoring without significant impairment to mobility
and behavior which is very important in neurophysiological
studies. The weight reduction also now allows for scaling up
the number of devices and hence numbers of neurons to be
monitored in vivo.
Microelectrode Movement Test
After the MEMS packaging steps were completed, the microelectrodes
were tested for its movement as shown in Fig. 10. For
this testing sample, the electrothermal microactuators moved
a rotating gear, which, in turn, drove the MEMS microelectrode.
To move the microelectrode forward [outside the package
boundary, as shown in Figs. 9(b) and 10(a)], a current pulse
of 300-ms duration was applied to open the forward-locking
mechanism. In the middle of this 300-ms current pulse, another
100-ms pulse of current was applied tomove themicroelectrode
forward by 8.8 μm. The total current required for two sets
of pulses was 30 mA. The pulses were applied at the rate of
0.5–1 Hz. The electrode was able to move over a maximum
displacement of 5 mm [the condition at Fig. 10(b)]. To move the
electrode backward [as shown in Fig. 10© and (d)], similar set
of pulses was required for backward-locking mechanism and
the backward movement actuator.
[attachment=53091]
Abstract
We report here a successful demonstration of a
flip-chip packaging approach for a microelectromechanical systems
(MEMS) device with in-plane movable microelectrodes implanted
in a rodent brain. The flip-chip processes were carried out
using a custom-made apparatus that was capable of the following:
1) creating Ag epoxy microbumps for first-level interconnect;
2) aligning the die and the glass substrate; and 3) creating nonhermetic
encapsulation (NHE). The completed flip-chip package
had an assembled weight of only 0.5 g significantly less than the
previously designed wire-bonded package of 4.5 g. The resistance
of the Ag bumps was found to be negligible. The MEMS microelectrodes
were successfully tested for its mechanical movement
with microactuators generating forces of 450 μN with a displacement
resolution of 8.8 μm/step.
INTRODUCTION
PACKAGING HAS been a significant challenge in the development
of cost-effective microelectromechanical systems
(MEMS) devices. Often, packaging issues and processes
do not get sufficient attention until the MEMS front-end processes have been completely developed. MEMS packaging
costs now typically contribute more than 50% [1]–[6] of the
total material/assembly cost. The reasons for the higher packaging
cost are partly because MEMS devices typically involve
active mechanical structures, requiring stringent packaging
constraints [7] and different kinds of unconventional sealing
techniques, such as nonhermetic [for pressure sensors and biomedical
MEMS (bio-MEMS)] and hermetic (for RF MEMS).
METHODS
MEMS Testing Samples
There are three different types of MEMS testing samples
that have been used in this study for the development of flipchip
packaging technology: 1) folded beam (chevron) type of
actuator; 2) linear actuator; and 3) linear actuator with a rotating
gear. Chevron actuator mechanism, as shown in Fig. 1, has the
smallest feature size of about 1.5 μm and has the most complex
actuationmechanism that finally moves the microelectrode with
the size of 50-μm width and 4-μm thickness. Therefore, the
results from chevron actuator are reported and discussed more
frequently in the study. As the flip-chip technology reported
in this paper has been tested successfully in three different
complex actuator mechanisms (as shown in Fig. 5), we expect
the flip-chip method reported here to be applicable for other
MEMS designs.
Flip-Chip Assembly Steps
The overall steps involved in the proposed flip-chip packaging
approach for the MEMS testing sample using the custommade
machine are shown in Fig. 6. These processes can be
extended for packaging other MEMS test samples as well.
1) Kitting process to prepare the three main components:
theMEMS chip, glass substrate, and Omnetics (Omnetics
Connector Corporation) third-level interconnects (TLIs).
2) Bumping process involved dispensing Ag epoxy (E3001;
Epoxy Technology, Inc.) on each of the Al pads to create
microscale bumps to build the FLIs with a height of
approximately 100 μm (a total of 18 bumps per a testing
sample).
Black Epoxy Seal
The remaining three sides of the die which were not accessed
by the MEMS microelectrodes were sealed by black epoxy.
A commercially available epoxy [from Epoxy Technology;
mixing two parts, namely, part no. 353ND, part A (90%), and
part no. 353ND, part B (10%)] cured at 120 ◦C for 5 min was
used to generate the near-hermetic seal. The literature suggests
that the epoxy does not produce a complete hermetic seal as
per military standard but rather a near-hermetic seal [19]. At
this stage, we have not yet tested the level of hermeticity of the
black epoxy seal. A small opening was left on the back side
of the die for air ventilation. This air ventilation was necessary
to keep the pressure inside the MEMS chip enclosure close to
ambient during implantation.
RESULTS
Package Size and Weight
The flip-chip technology reduced the final package weight
significantly from 4.5 g (using wire-bonding technology) to
0.5 g, which is almost an order of magnitude reduction in the
overall assembled weight. Since the weight of the head of an
adult rat is approximately 30 g (for a rat weighing 300 g),
the aforementioned significant weight reduction will now allow
neuronal monitoring without significant impairment to mobility
and behavior which is very important in neurophysiological
studies. The weight reduction also now allows for scaling up
the number of devices and hence numbers of neurons to be
monitored in vivo.
Microelectrode Movement Test
After the MEMS packaging steps were completed, the microelectrodes
were tested for its movement as shown in Fig. 10. For
this testing sample, the electrothermal microactuators moved
a rotating gear, which, in turn, drove the MEMS microelectrode.
To move the microelectrode forward [outside the package
boundary, as shown in Figs. 9(b) and 10(a)], a current pulse
of 300-ms duration was applied to open the forward-locking
mechanism. In the middle of this 300-ms current pulse, another
100-ms pulse of current was applied tomove themicroelectrode
forward by 8.8 μm. The total current required for two sets
of pulses was 30 mA. The pulses were applied at the rate of
0.5–1 Hz. The electrode was able to move over a maximum
displacement of 5 mm [the condition at Fig. 10(b)]. To move the
electrode backward [as shown in Fig. 10© and (d)], similar set
of pulses was required for backward-locking mechanism and
the backward movement actuator.