IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 66, NO.
5, MAY 2019 3835
Development of Six-Degree-of-Freedom Inertial
Sensors With an 8-in Advanced MEMS
Fabrication Platform
Guoqiang Wu , Beibei Han, Daw Don Cheam, Leong Ching Wai, Peter Hyun Kee Chang,
Navab Singh, and Yuandong Gu
Abstract—This paper presents the development of an
8-in six-degree-of-freedom (DOF) inertial sensor based on
an advanced capacitive inertial sensor fabrication plat-
form. The platform integrates three-axis gyroscopes and
three-axis accelerometers on the same chip. The fabricated
sensors are vacuum packaged at the wafer level with a
polysilicon-based through-silicon interposer (TSI) using the
aluminum–germanium eutectic bonding approach. Wafer-
level measurement results indicate that a fabrication yield
of greater than 92% and a vacuum level of ∼100 mtorr
have been achieved. The fabricated inertial sensors and the
customized application-specific integrated circuits are en-
capsulated in a 5 mm × 5 mm × 1.3 mm quad-flat no-leads
package using the plastic molding technology. The system-
level characterization of the developed six-DOF inertial
sensors have been implemented. Several reliability tests
conducted according to the relevant JEDEC standards
prove that the packaged sensors are highly reliable and ro-
bust for a wide range of operating environments.
Index Terms—Accelerometer, aluminum–germanium (Al–
Ge) bonding, gyroscope, inertial sensors, mircoelectrome-
chanical systems (MEMS), through-silicon interposers
(TSIs), wafer-level vacuum packaging.
I. INTRODUCTION
ICROELECTROMECHANICAL system (MEMS)
M inertial sensors are ubiquitous in the consumer and in-
dustry fields due to their compact size, low power, and good per-
formance [1]–[5]. Recently, the demands for a single accelero-
meter and a single gyroscope have declined in the consumer
and automotive markets. Meanwhile the demands for combo
Manuscript received February 2, 2018; revised April 24, 2018 and
May 31, 2018; accepted June 13, 2018. Date of publication July 9,
2018; date of current version December 28, 2018. This work was
funded by Mubadala Development Company-Abu Dhabi, Economic De-
velopment Board-Singapore and GLOBALFOUNDRIESSingapore un-
der the framework of ‘Twinlab’ project with participation of A*STAR
Institute of Microelectronics-Singapore, Masdar Institute of Science
and Technology-Abu Dhabi and GLOBALFOUNDRIES-Singapore.
(Guoqiang Wu, Beibei Han, and Daw Don Cheam contributed equally to
this work.) (Corresponding author: Guoqiang Wu.)
The authors are with the Institute of Microelectronics, Agency for
Science, Technology, and Research, Singapore 138634 (e-mail: wugq@
ime.a-star.edu.sg; hanb@ime.a-star.edu.sg; cheamdd@ime.a-star.
edu.sg; wailc@ime.a-star.edu.sg; changhk@ime.a-star.edu.sg; navab@
ime.a-star.edu.sg; guyd@ime.a-star.edu.sg).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TIE.2018.2851946
0278-0046 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
solutions grow rapidly, especially for the inertial measurement
units (IMUs), which usually include three-axis accelerometers
and three-axis gyroscopes [6]–[8]. IMUs are widely used for
many applications, such as GPS-aiding, offshore navigation,
drilling, and platform stabilization [9], [10]. However, more
than 75% of the MEMS IMU market is dominated by the
major players, such as STMicroelectronics, Bosch Sensortec,
InvenSense, Freescale, etc. [10]–[13]. Besides the revenue, the
main barrier of entry for new players is the availability of the
mass-producible process platform for six-degree-of-freedom
(DOF) or nine-DOF inertial sensors, which requires the
platform to have the following features:
1) both in-plane and out-of-plane sensing capabilities [5];
2) wafer-level encapsulation for protection of movable mi-
cromechanical structures;
3) wafer-level vacuum packaging for improving the perfor-
mance of IMU sensors [14], [15].
In this paper, we report the development and evaluation of six-
DOF inertial sensors based on an 8-in advanced inertial sensor
fabrication. In Section II, the details of inertial sensor fabrica-
tion platform are presented. The design and simulation of the
reported six-DOF inertial sensors are described in Section III.
Then, the fabrication of the inertial sensors is presented in
Section IV. Section V illustrates the measurement results of the
fabricated inertial sensors. The details of assembly and pack-
aging of the fabricated sensors are elaborated in Section VI.
Section VII reported the reliability evaluations for the packaged
sensors. Finally, Section VIII concludes this paper.
II. PLATFORM DESIGN
Fig. 1 shows the important features of the reported 8-in
advanced six-DOF capacitive inertial MEMS platform. This
platform consists of two parts: a top passive cap wafer with a
through-silicon interposer (TSI) and a bottom cavity-silicon-
on-insulator (CSOI) MEMS device wafer. The cap wafer
and the MEMS wafer are bonded together in high vacuum
at the wafer level using the aluminum–germanium (Al–Ge)
eutectic bonding approach [13]. It not only protects the
microstructure from mechanical damage, but also reduces the
air damping for an improved performance. The cap wafer with
a getter is used to create and maintain the vacuum level of the
platform.
3836
TABLE I
COMPARISON OF THE REPORTED INERTIAL SENSOR PLATFORM WITH THE EXISTING PLATFORMS
Fig. 1. Overview of a six-DOF capacitive inertial MEMS sensor fabri-
cation platform.
Table I illustrates the key feature and differences of the re-
ported platform with the existing platforms. The key feature of
this platform is the adoption of the highly doped polysilicon,
which serves as the out-of-plane electrodes, redistribution lines
(RDLs), standoffs, and the TSI. The polysilicon achieves the
electrical interconnections and the out-of-plane sensing capa-
bility in this platform. Compared to existing commercial pro-
cesses, such as Bosch or Fairchild’s process, which require
RDLs and wire bonds to an application-specific integrated cir-
cuit (ASIC), shorter RDL traces and no bond wires are re-
quired here with the vertical integration, reducing parasitics
and allowing for a high density of interconnects required by
inertial systems today with numerous tuning electrodes. Com-
pared to mCube and InvenSense’s processes, there are no out-
gassing films that can cause sensor drift. The epi-seal process
has been successful for high-stability timing, but the base vari-
ant that is commercialized has limitations on the gap size and
has no out-of-plane electrodes, which limits its use for inertial
systems [19]–[21].
The six-DOF inertial sensors are fabricated in the CSOI wafer.
A thick device layer of single-crystal silicon (SCS) can be used
in this platform in order to reduce mechanical noise and improve
the performance. The vertical direction integration has shrunk
the size of the fabricated sensors. Moreover, it allows 2.5-D
integration of fabricated sensors with the help of the top RDLs.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 66, NO. 5, MAY 2019
These top RDLs can be interfaced to any other sensor, ASIC, or
printed circuit board via flip chip or wire-bonding techniques.
III. SIX-DOF SENSOR DESIGN
All the six-DOF inertial sensors presented here work by the
principle of capacitive sensing. The accelerometers estimate
the acceleration by detecting the change of the capacitance
[22]–[24]. The X/Y -axis accelerometer consists of an iner-
tial proof mass with a set of springs. The lateral movements in
the X-axis or Y -axis are detected by the fixed sensing comb
fingers with a variable-gap configuration. The Z-axis ac-
celerometer is based on an asymmetric seesaw structure. The
differential capacitance changes caused by the Z-axis acceler-
ations are detected by the out-of-plane sensing electrodes. The
gyroscopes used in this platform are the Coriolis-force-based
sensors [4], [25]–[28]. Fully decoupled designs are employed to
reduce the mechanical cross coupling between the drive mode
and the sense mode. The X/Y -axis gyroscope employs the
torsional-vibration structures. The Coriolis-force-induced out-
of-plane movement is measured using the capacitive sensing
electrodes on the cap wafer. The Z-axis gyroscope is based
on the linear vibration structures. The coriolis-force-induced
movement is detected by the in-plane sensing electrodes. Fig. 2
illustrates the simulated mode shapes of the six-DOF sen-
sors, which are obtained by the finite-element method using
CoventorWare.
IV. FABRICATION
The reported six-DOF inertial sensors are fabricated based on
an 8-in advanced capacitive inertial MEMS platform. The main
fabrication process consists of the MEMS wafer fabrication, the
cap wafer fabrication, and the wafer-level vacuum bonding, as
shown in Fig. 3.
Fig. 3(a) illustrates the main feature of the MEMS wafer. The
fabrication process starts with an 8-in customized CSOI wafer
with a 30-µm-thick device layer. The resistivity of the device
layer is approximately 0.01 Ω·cm. A 0.7-µm-thick Al layer is
sputtered and patterned on top of the device layer to form the
bonding pads and the sealing ring in the device wafer. Then,
the inertial sensor structures are patterned and released using
the deep reactive-ion etching (DRIE).
The main feature of the cap wafer is shown in Fig. 3(b).
It starts with through-silicon via (TSV) etching using DRIE.
After a 2-µm-thick linear oxide growth as the insulation
layer, the TSVs are filled with highly doped polysilicon using
low-pressure chemical vapor deposition. Then, the polysilicon
WU et al.: DEVELOPMENT OF SIX-DOF INERTIAL SENSORS WITH AN 8-IN ADVANCED MEMS FABRICATION PLATFORM
Fig. 2. Simulated mode shapes of the six-DOF inertial sensors.
(a) X/Y -axis accelerometer. (b) Z -axis accelerometer. (c) Drive mode
of the X/Y -axis gyroscope. (d) Sense mode of the X/Y -axis gyro-
scope. (e) Drive mode of the Z -axis gyroscope. (f) Sense mode of the
Z -axis gyroscope. The color scheme indicates the magnitude of modal
displacement across the structure. Color in red indicates the largest dis-
placement occurred in the vibrating modes.
Fig. 3. Cross-sectional views of the main steps of the six-DOF inertial
sensor platform. (a) CSOI-based MEMS inertial sensors. (b) Cap wafer
with a polysilicon-based TSI and getter. (c) Wafer-level vacuum pack-
aging of the six-DOF MEMS inertial sensors using the Al–Ge eutectic
bonding approach.
is polished to the desired thickness and a Ge layer is deposited.
It follows that the standoff is patterned and formed by partially
etching the Ge film and polysilicon. The remaining polysilicon
is patterned and etched stopping at the 2-µm-thick linear oxide
layer to form the bottom RDLs on the cap wafer. Next, the cav-
ity on the cap wafer is formed. In theory, a deeper cavity would
allow for a higher vacuum level to be achieved. The Titanium
(Ti)/Nickel (Ni) getter is formed at the bottom of the cap wafer
cavity with an integrated shadow mask method. The getter only
covers the nondevice area in the cavity. It is used to achieve a
high vacuum and to stabilize at the required vacuum level.
3837
Fig. 4. SEM images of the fabricated six-DOF inertial sensors before
wafer-level vacuum bonding. (a) Three-axis accelerometers. (b) Z -axis
gyroscope. (c) X/Y -axis gyroscope. (d) Zoom-in of the comb finger
structure of the X/Y -axis gyroscope. (e) Cross-sectional view of the
sealing ring. (f) Cross-sectional view of the TSI cap wafer bonded to the
MEMS wafer.
The wafer-level vacuum packaging process is illustrated in
Fig. 3(c). The MEMS wafer with fabricated six-DOF inertial
sensors and the cap wafer with polysilicon TSI are bonded
together in a vacuum chamber using the Al–Ge eutectic bonding
approach. Then, the bonded wafer is grinded on the TSI side to
expose the highly doped TSV in the cap wafer. Following oxide
deposition and TSV hole opening, an Al layer is deposited and
patterned to form the top RDL and the wire-bonding pads. Then,
a oxide layer is deposited as the passivation layer, followed
by oxide etching to expose the wire-bonding pads. Finally, the
bonded wafer is thinned down to 0.5 mm by grinding the MEMS
wafer.
Fig. 4(a)–(d) shows the scanning electron microscope (SEM)
images of the fabricated inertial sensors after DRIE release (be-
fore bonding with the cap wafer). Fig. 4(e) and (f) shows the
cross-sectional views of the sealing ring and the TSI in the cap
wafer after bonding. The Al and Ge layers interdiffuse into each
other to achieve a robust eutectic bonding, and also to realize
the electrical connections between the MEMS wafer and the
cap wafer. The polysilicon completely fills the vias. No obvious
voids or defects are found.
The device singulation is done using standard saw dicing.
The measured average shear strength of 20 singulated packaged
samples is 54.5 MPa, which is much greater than the die shear
strength criteria of 11.68 MPa according to the MIL-STD-883G
Method 2019. Fig. 5 shows an optical image of the diced six-
DOF sensors. The size of the fabricated six-DOF sensors is
2 × 2 × 0.5 mm3 .
V. MEASUREMENT RESULTS
A. Platform Evaluation
The fabrication yield and vacuum level achieved are evaluated
by measuring the dynamic response and the relative resistance
change of a Pirani gauge. Fig. 6 shows the measured capacitance
versus dc bias voltage (C–V) curves of one fabricated inertial
3838 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 66, NO. 5, MAY 2019

Fig. 5. Photograph of the diced wafer-level vacuum-packaged six-DOF Fig. 7. Measured relative resistance changes of the Pirani gauge in an
inertial sensors. 8-in wafer. According to the reference curve, a vacuum level of less than
100 mtorr has been achieved.

Fig. 6. Wafer-level capacitance–voltage curves of one fabricated iner-

Fig. 8. Residual gas analysis for 14 packaged samples with HR-IVA.
tial sensor. A total of 64 in 69 voltage sweeps show highly parabolic
The most abundant residual gas that appears consistently across all
curves, which indicates that a yield of greater than 92% has been
samples is argon.
achieved.

across all samples. It is the main outgassing species from the

sensor at the wafer level. A total of 64 in 69 voltage sweeps show
argon-sputtered physical vapor deposition films of the Al and
good parabolic curves. The average static capacitances of the
Ge. The second largest component in residual gas is hydrogen,
64 devices across the wafer is 1.11 pF with a standard deviation
which is most likely produced from the epi-polysilicon depo-
of 0.023 pF (2.07%). The measured results indicate that a yield
sition step. The residual hydrogen gas may be eliminated by
of greater than 92% has been achieved.
high-temperature annealing [29]. Moisture is detected on two
The vacuum level of the inertial platform is reflected with a
samples. Typically, the moisture is brought into the package
Pirani vacuum indicator. The relative resistance changes of the
cavity through adsorbing on the walls of the package interior.
Pirani gauge in one fabricated 8-in wafer are shown in Fig. 7. It
Finally, carbon dioxide is also traced on two samples, which
shows that more than 62% of the tested chips achieve a vacuum
usually comes from the polymer residue in the sidewall of the
level of less than 100 mtorr, and more than 72% of the tested
cavity or the structures. The residual polymer may be completely
chips have a vacuum level of less than 400 mtorr. This level
removed by proper cleaning.
of vacuum has been repeated with similar consistency across
several wafers. The vacuum level achieved in this work meets
the requirement for most of the inertial sensors [7]. B. System-Level Measurement
A high-resolution internal vapor analysis (HR-IVA) is con- The fabricated MEMS inertial sensors are hybrid connected
ducted to identify and quantity the residual gases sealed in the to the customized ASICs for the accelerometers and the gy-
cavity in order to further improve the vacuum level. Fig. 8 shows roscopes in order to obtain the acceleration responses of the
the HR-IVA results for 14 packaged samples. We observe that MEMS accelerometers and the angular rate responses of the
argon is the most abundant residual gas that appears consistently MEMS gyroscopes, respectively. The wire bonding is used for
WU et al.: DEVELOPMENT OF SIX-DOF INERTIAL SENSORS WITH AN 8-IN ADVANCED MEMS FABRICATION PLATFORM
Fig. 9. Measured scale factors of the six-DOF inertial sensors.
(a) Three-axis accelerometer. (b) Three-axis gyroscope.
the electrical connections between the inertial sensors and the
ASIC chips. Fig. 9(a) and (b) shows the measured scale factors
of the three-axis accelerometers and three-axis gyroscopes, re-
spectively. The measured scale-factor nonlinearity is less than
1% for the three-axis accelerometers and less than 0.5% for the
three-axis gyroscopes.
The bias instability (BI) and noise density of the inertial
sensors are obtained by evaluating the Allan variance data. BI,
or in-run BI, is the minimum point in the Allan variance curve.
It represents the best bias stability that could be achieved over
a specified period of time at constant temperature. The random
walk due to the white noise can be represented by fitting a line
with slope −1/2 to the region of the Allan variance curve and
then be obtained by reading its value at average time of 1 s. The
measured Allan variance curves of the three-axis accelerometers
and the three-axis gyroscopes are illustrated in Fig. 10(a) and
(b), respectively. The measured BI of the accelerometer is lower
than 2.8 mg, and the√measured velocity random walk (VRW)
is lower than 48 µg/ Hz for all the axes, respectively. For the
gyroscope, the measured BI is better than 0.07 ◦ /s for all the
3839
Fig. 10. Measured Allan variance curves of the six-DOF inertial sen-
sors. (a) Three-axis accelerometer. (b) Three-axis gyroscope.
axes. The
√ angular random walk (ARW) is measured less than
0.17 ◦ / s in all axes.
The performance comparison of the developed six-DOF in-
ertial sensors with the state-of-the-art commercial available in-
ertial sensors is illustrated in Table II. The reported three-axis
accelerometers are shown to be a competitive high-performance
device in the consumer electronics space. The performance of
the three-axis gyroscopes would be improved by optimizing the
sensor and readout circuit designs, as well as using the advanced
correction algorithms.
VI. ASSEMBLY AND PACKAGING
The fabricated three-axis accelerometer and the X/Y -axis
and Z-axis gyroscope are packaged in a 5 mm × 5 mm ×
1.3 mm quad-flat no-leads (QFN) package. The MEMS die is
stacked on top of the ASIC die. The ASIC die is attached to a
QFN substrate. Wire bonding is used to achieve the electrical
connections between the chips and the substrate. Then, the chips
and the substrate are packaged using the plastic mold compound,
3840
TABLE II
PERFORMANCE COMPARISON OF THE DEVELOPED SIX-DOF INERTIAL SENSORS WITH THE COMMERCIAL AVAILABLE INERTIAL SENSORS
Fig. 11. Assembly and packaging of the inertial sensors. (a) Packaging
concept. (b) Photograph of a packaged inertial sensors. (c) X-ray image
of the packaged three-axis accelerometer.
as shown in Fig. 11(a). The photograph of the packaged inertial
chips is shown in Fig. 11(b). The size of the packaged chip is
5 mm × 5 mm × 1.3 mm. Fig. 11(c) shows an X-ray image
of the packaged three-axis accelerometer chip. The lead frame
and the wire-bonding connections between the MEMS and the
ASIC can be clearly seen.
VII. RELIABILITY EVALUATION
The fabricated inertial sensor is packaged with a dummy
ASIC die to simulate the real packaging situations. The pack-
aged devices are evaluated by the mechanical shock (MS), tem-
perature cycling (TC), high-temperature storage (HTS), and
highly accelerated stress test (HAST) in order to assess the
reliability of the packaged devices. Fifteen packaged chips
(five three-axis accelerometers, five X/Y -axis gyroscopes, and
five Z-axis gyroscopes) are evaluated for each test. We tested the
C–V responses for the three-axis accelerometers and frequency
response for the X/Y -axis and Z-axis gyroscope.
A. MS Test
The MS test is implemented to investigate the fragility of
products and to evaluate protective packaging. In this work,
the shock test condition is for a total of three shock pulses,
each at a peak of 1500 G. After the MS test, the packaged
chips are measured at the room temperature, and the measured
results are compared with the ones before the MS test. The
measurement results of the MS test are shown in Fig. 12(a). The
resonant frequency variations for the gyroscopes are less than
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 66, NO. 5, MAY 2019
±0.3%, and the static capacitance change for the accelerometer
is less than 4%. The static capacitance changes may be from the
measurement error and the stress-induced error.
B. TC Test
In this work, the TC reliability tests and analysis are con-
ducted to verify the reliability and robust of the wafer-level
vacuum bonding quality. Five hundred thermal cycles are im-
plemented to verify the long-term stability of the packaged de-
vices placed under outdoor ambient conditions. The temperature
condition ranges from −40 to 125 ◦ C with a ramp rate of approx-
imately 15 ◦ C/min and a dwell time of 15 min at the two extreme
temperatures. The details can be referred to the JESD22-A104D
standard. After 500 thermal cycles, the packaged chips are tested
at room temperature. The measurement results of the TC test
are shown in Fig. 12(b).
C. HTS Test
The HTS test is used to check the effect of time and temper-
ature under the unbiased storage conditions in order to inves-
tigate the thermally activated failure mechanisms. The storage
temperature is +150 ◦ C for 500 h, according to JESD22-A103D
standard. After 500 h, the packaged chips are tested at the room
temperature. The measurement results of the HTS test after
500 h are shown in Fig. 12(c).
D. HAST Evaluation
A HAST is performed to evaluate the reliability of packaged
devices under a controlled humidity and temperature environ-
ment. In this work, the HAST is performed at a temperature of
130 ◦ C and a relative humidity of 85% for 96 h as per JEDEC
standard JESD22-A110-B. After the HAST, the measured reso-
nant frequency drifts of the three-axis gyroscopes are less than
±0.5%, as illustrated in Fig. 12(d). It indicates that all the tested
devices did not experience any moisture ingress or delamination
even after 96 h of HAST.
In summary, all the tested devices have passed the MS, TC,
HTS, and HAST, as shown in Fig. 12. No failure mechanism
is observed. These measured reliability evaluation results verify
the reliability and robustness of the fabricated sensors and the
packaging.
WU et al.: DEVELOPMENT OF SIX-DOF INERTIAL SENSORS WITH AN 8-IN ADVANCED MEMS FABRICATION PLATFORM
Fig. 12. Resonant frequency variations of the three-axis gyroscope and the static capacitance changes of one pair of electrodes of the three-axis
accelerometers after the four reliability evaluations: (a) MS; (b) TC; (c) HTS; and (d) HAST, respectively.
VIII. CONCLUSION
The development of the six-DOF inertial sensors with an
advanced capacitive inertial sensor fabrication platform was re-
ported in this paper. The platform featured compact size, good
performance, and high yield. The wafer-level and system-level
measurement results verified the robustness of the sensor. The
reliability evaluations, such as MS, TC, HAST, and HTC, il-
lustrated the robustness of the sensor and the packaging. The
reported inertial sensors exhibited competitive performance in
some critical parameters such as BI, noise density, and scale-
factor nonlinearity. The technology reported in this paper paves
a way for the mass production of the inertial sensor for the huge
industry and consumer markets.
REFERENCES
[1] D. K. Shaeffer, “MEMS inertial sensors: A tutorial overview,” IEEE Com-
mun. Mag., vol. 51, no. 4, pp. 100–109, Apr. 2013.
[2] R. N. Dean et al., “A characterization of the performance of a MEMS gy-
roscope in acoustically harsh environments,” IEEE Trans. Ind. Electron.,
vol. 58, no. 7, pp. 2591–2596, Jul. 2010.
[3] C. W. Tsai, K. Chen, C. K. Shen, and J. Tsai, “A MEMS doubly decoupled
gyroscope with wide driving frequency range,” IEEE Trans. Ind. Electron.,
vol. 59, no. 12, pp. 4921–4929, Dec. 2012.
[4] G. Q. Wu, G. L. Chua, and Y. D. Gu, “A dual-mass fully decoupled MEMS
gyroscope with wide bandwidth and high linearity,” Sens. Actuators A:
Phys., vol. 259, pp. 50–56, Jun. 2017.
[5] C. Acar, “High-performance 6-axis MEMS inertial sensor based on
through-silicon via technology,” in Proc. IEEE Int. Symp. Inertial Sens.
Syst., Laguna Beach, CA, USA, Feb. 22–25, 2016, pp. 62–65.
3841
[6] H. Sun et al., “A monolithic inertial measurement unit fabricated with
improved DRIE post-CMOS process,” in Proc. IEEE Sens. Conf., 2010,
pp. 1198–1202.
[7] G. Langfelder, C. Buffa, A. Frangi, A. Tocchio, E. Lasalandra, and A.
Longoni, “Z-axis magnetometers for MEMS inertial measurement units
using an industrial process,” IEEE Trans. Ind. Electron., vol. 60, no. 9,
pp. 3983–3990, Sep. 2013.
[8] I. E. Ocak et al., “A monolithic 9 degree of freedom (DOF) capacitive
inertial MEMS platform,” in Proc. IEEE Int. Electron Devices Meeting,
San Francisco, CA, USA, 2014, pp. 22.6.1–22.6.4.
[9] S. Zihajehzadeh, T. J. Lee, J. K. Lee, R. Hoskinson, and E. J. Park,
“Integration of MEMS inertial and pressure sensors for vertical trajectory
determination,” IEEE Trans. Instrum. Meas., vol. 64, no. 3, pp. 804–814,
Mar. 2015.
[10] Status of the MEMS Industry 2014, Yole Développement, Lyon, France,
2014.
[11] R. Neul et al., “Micromachined angular rate sensors for automotive
applications,” IEEE Sens. J., vol. 7, no. 2, pp. 302–309, Feb. 2007.
[12] J. Seeger, M. Lim, and S. Nasiri, “Development of high-performance
high-volume consumer MEMS gyroscopes,” in Proc. Tech. Dig. Solid-
State Sens. Actuators Microsyst. Workshop, Jun. 2010, pp. 61–64.
[13] A. D. Oliver et al., “A new three axis low power MEMS gyroscope for
consumer and industrial application,” in Proc. 18th Int. Conf. Solid-State
Sens., Actuators Microsyst., Anchorage, AK, USA, Jun. 2015, pp. 31–34.
[14] G. Q. Wu, D. H. Xu, B. Xiong, Y. C. Wang, Y. L. Wang, and Y. L. Ma,
“Wafer-level vacuum packaging for MEMS resonators using glass frit
bonding,” J. Microelectromech. Syst., vol. 21, no. 6, pp. 1484–1491, Dec.
2012.
[15] G. Q. Wu et al., “Wafer-level vacuum-packaged high performance AlN-
on-SOI piezoelectric resonator for sub-100 MHz oscillator applications,”
IEEE Trans. Ind. Electron., vol. 65, no. 4, pp. 3576–3584, Apr. 2018.
[16] Bosch Sensortec BMX055 9-Axis MEMS IMU—Reverse Costing Analysis
System Plus Consulting, Nantes, France, Dec. 2013.
[17] mCube, “The advantages of integrated MEMS to enable the Internet of
moving things,” White paper, San Jose, CA, USA, Jun. 2014.
3842
[18] A. C. Fischer, F. Forsberg, M. Lapisa, S. J. Bleiker, G. Stemme, and
N. Roxhed, “Integrating MEMS and ICs,” Microsyst. Nanoeng., vol. 1,
no. 15, 2015, Art. no. 15005.
[19] SiTime, “SiTime’s MEMS first and EpiSeal processes,” Mar. 2018. [On-
line]. Available: https://www.sitime.com/api/gated/AN20001-MEMS-
First-and-EpiSeal-Processes.pdf
[20] E. J. Ng, V. A. Hong, Y. Yang, C. H. Ahn, C. L. M. Everhart, and T.
W. Kenny, “Temperature dependence of the elastic constants of doped
silicon,” J. Microelectromech. Syst., vol. 24, no. 3, pp. 730–741, Jun.
2015.
[21] R. N. Candler et al., “Long term and accelerated life testing of a novel
single-wafer vacuum encapsulation for MEMS resonators,” J. Microelec-
tromech. Syst., vol. 15, no. 6, pp. 1446–1456, Dec. 2006.
[22] M. H. Bao, Handbook of Sensors and Actuators: Micro Mechanical
Transducers, Pressure Sensors, Accelerometers and Gyroscopes, vol. 8,
S. Middelhoek, Ed. New York, NY, USA: Elsevier, 2000, pp. 362–365.
[23] O. Aydin and T. Akin, “A bulk-micromachined fully differential MEMS
accelerometer with split interdigitated fingers,” IEEE Sens. J., vol. 13,
no. 8, pp. 2914–2921, Aug. 2013.
[24] R. Abdolvand, B. V. Amini, and F. Ayazi, “Sub-micro-gravity in-plane
accelerometers with reduced capacitive gaps and extra seismic mass,”
J. Microelectromech. Syst., vol. 16, no. 5, pp. 1036–1043, Oct. 2007.
[25] S. E. Alper, Y. Temiz, and T. Akin, “A compact angular rate sensor
system using a fully decoupled silicon-on-glass MEMS gyroscope,” J.
Microelectromech. Syst., vol. 17, no. 6, pp. 1418–1429, Dec. 2008.
[26] S. Sonmezoglu et al., “Single-structure micromachined 3-axis gyroscope
with reduced drive-force coupling,” IEEE Electron Device Lett., vol. 36,
no. 9, pp. 953–956, Sep. 2015.
[27] D. E. Serrano et al., “Substrate-decoupled, bulk-acoustic wave gyro-
scopes: Design and evaluation of next-generation environmentally robust
devices,” Microsyst. Nanoeng., vol. 2, Feb. 2016, Art. no. 16015.
[28] S. A. Zotov, A. A. Trusov, and A. M. Shkel, “High-range angular rate
sensor based on mechanical frequency modulation,” J. Microelectromech.
Syst., vol. 21, no. 2, pp. 398–405, Apr. 2012.
[29] B. Kim et al., “Hermeticity and diffusion investigation in polysilicon
film encapsulation for microelectromechanical systems,” J. Appl. Phys.,
vol. 105, Jan. 2009, Art. no. 013514.
[30] BMI160 Small, Low Power Inertial Measurement Unit, Datasheet, Bosch
SensorTec, Reutlingen, Germany, Feb. 2015.
[31] FIS1100 6D Inertial Measurement Unit with Motion Co-Processor and
Sensor Fusion Library, Fairchild, Sunnyvale, CA, USA, Jul. 2016.
Guoqiang Wu received the B.Eng. degree in
electrical science and technology from Xidian
University, Xi’an, China, in 2008, and the Ph.D.
degree in microelectronics and solid-state elec-
tronics from the Shanghai Institute of Mi-
crosystem and Information Technology, Chinese
Academy of Sciences, Shanghai, China, in
2013.
He is currently a Research Scientist with
the Institute of Microelectronics, Agency for Sci-
ence, Technology and Research, Singapore. His
research interests include microelectromechanical systems (MEMS) res-
onators, radio frequency MEMS, inertial MEMS, MEMS integration, and
packaging technologies.
Dr. Wu was the recipient of the Best Dissertation Award of Shanghai
in 2014.
Beibei Han received the B.S. degree in elec-
tronic science and technology from Shandong
University, Jinan, China, in 2006, the M.S.
degree in mechanical engineering from the
Korea University of Technology and Education,
Cheonan, South Korea, in 2008, and the Ph.D.
degree in microelectronics from Nanyang Tech-
nological University, Singapore, in 2015.
From 2008 to 2010, she was a Research Sci-
entist working on cognitive robotics and simu-
lation with the Korea Institute of Science and
Technology and the Korea Advanced Institute of Science and Technol-
ogy. Since 2015, she has been a Scientist with the Institute of Micro-
electronics, Agency for Science, Technology and Research, Singapore.
Her research interests include physical microelectromechanical system
sensors and acoustic sensor systems.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 66, NO. 5, MAY 2019
Daw Don Cheam received the B.Eng., M.S.,
and Ph.D. degrees in electrical engineering from
Michigan Technological University, Houghton,
MI, USA, in 2001, 2004, and 2009, respectively.
He is currently an Industry Development
Manager with the Institute of Microelectronics,
Agency for Science Technology and Research,
Singapore.
Leong Ching Wai received the bachelor’s de-
gree in materials engineering from Nanyang
Technology University, Singapore, in 2003.
After graduation, she worked in the semicon-
ductor packaging industry for two years. Since
then, she has been with the Institute of Micro-
electronics, Agency for Science Technology and
Research, Singapore, where she is currently the
Senior Research Engineer of the Interconnect
and Packaging Platform.
Peter Hyun Kee Chang received the B.S., M.S.,
and Ph.D. degrees in electrical engineering and
computer science from Seoul National Univer-
sity, Seoul, South Korea, in 1997, 1999, and
2010, respectively.
He founded Intellimicrons in 2000 and pro-
vided sensor solutions for defense and auto-
motive applications. He is currently the Deputy
Head of the Sensors, Actuators, and Memo-
ries Program at the Institute of Microelectron-
ics (IME), Agency for Science Technology and
Research, Singapore. He has been leading various developments of mi-
croelectromechanical system sensor and packaging technologies since
he joined IME in 2012.
Navab Singh received the M.Tech. degree in
solid state materials from the Indian Institute of
Technology, Delhi, India, in 1995, and the Ph.D.
degree in electrical and computer engineer-
ing from the National University of Singapore,
Singapore, in 2008.
He is currently a Senior Scientist and the Di-
rector of Sensors, Actuators, and Memories Pro-
gram at the Institute of Microelectronics, Agency
for Science, Technology, and Research, Singa-
pore. He has authored or coauthored more than
200 technical papers.
Dr. Singh was a recipient of the George E. Smith Award in 2007,
the Singapore National Technology Award in 2008, and the Agency for
Science Technology, and Research Talent Award in 2010.
Yuandong Gu received the M.E.E. degree in
electrical engineering and the Ph.D. degree in
pharmaceutics from the University of Minnesota,
Minneapolis, MN, USA, in 2001 and 2003,
respectively.
He is currently the Deputy Executive Direc-
tor with the Institute of Microelectronics (IME),
Agency for Science, Technology, and Research,
Singapore. Before joining IME, he was a Prin-
cipal Research Scientist with the Honeywell
Sensors and Wireless Lab for ten years.