JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 22, NO. 6, DECEMBER 2013
1347
Wafer-Level Vacuum Sealing by Coining of Wire
Bonded Gold Bumps
Mikael Antelius, Andreas C. Fischer, Niclas Roxhed, Member, IEEE, Göran Stemme, Fellow, IEEE,
and Frank Niklaus, Member, IEEE
Abstract—This paper reports on the investigation of a novel
room-temperature vacuum sealing method based on compressing
wire bonded gold bumps which are placed to partially overlap
the access ports into the cavity. The bump compression, which
is done under vacuum, causes a material flow into the access
ports, thereby hermetically sealing a vacuum inside the cavities. The sealed cavity pressure was measured by residual gas
analysis to 8×10−4 mbar two weeks after sealing. The residual
gas content was found to be mainly argon, which indicates
the source as outgassing inside the cavity and no measurable
external leak. The seals are found to be mechanically robust
and easily implemented by the use of standard commercial
tools and processes.
[2013-0003]
Index Terms—Vacuum, packaging, MEMS, wire bonding, sealing, hermetic.
I. Introduction
V
Fig. 1. Principle drawing of the sealing process. (a) First, a gold bump is
wire bonded partially overlapping the access port of a cavity. (b) Second, the
bump is compressed inside a vacuum chamber, causing plastic deformation
and a flow of the bump material into the access port, thereby sealing the cavity.
Manuscript received January 3, 2013; revised March 8, 2013; accepted
March 23, 2013. Date of publication June 25, 2013; date of current version
November 25, 2013. This work was supported by the European Research
Council through the advanced grant project xMEMs (267528). Subject Editor
N. de Rooij.
The authors are with the Department of Micro and Nanosystems, KTH Royal Institute of Technology, 100 44, Stockholm Sweden
(e-mail: [email protected]; [email protected]; [email protected];
[email protected]; [email protected]).
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/JMEMS.2013.2262594
Hermetic sealing at room-temperature is challenging since
most bonding methods require elevated temperatures. Earlier
reported room-temperature sealing methods include plasma
activated direct bonding [5] and, in a sense, sealing structures
heated locally on the wafer [6]. However, these methods are
not straightforward and either impose restrictions on the choice
of materials due to surface chemistry or add significant process
complexity. Recently, plastic deformation of two types of
gold structures has been used for room-temperature packaging
applications [1], [7]–[9]. First, cold welding by deformation
of electroplated gold sealing rings was used for both cavity
formation and sealing of liquids [7] or vacuum [8]. Later,
wire bonded ball bumps were used for liquids sealing by wire
bonding gold stud bumps directly on the access ports of liquidfilled microcavities [9].
Low-temperature deformation of gold is extensively used
in a process called coining, where wire bonded gold bumps
are plastically deformed in order to both make the top surface
flat and to achieve a predetermined bump height. This process
was originally introduced in order to increase the electrical and
mechanical reliability of stud bumps used in flip chip packaging [10]. Recently it has also been used for 3-dimensional
micro structuring, where the bumps were imprinted with a
structured mold [11].
In this paper we report on a novel wafer-level roomtemperature vacuum sealing process consisting of two steps.
ACUUM packaging [1], [2], the process of enclosing a
device under vacuum, is crucial for the performance of
a wide range of MEMS devices. For example, vacuum packaging enables better thermal insulation of microbolometers in
infrared imaging sensors, absolute pressure sensors with stable
reference pressures and reduced gas damping effects in inertial
MEMS sensors. Ensuring a high vacuum environment for a
packaged device over its entire lifetime requires both a low
residual gas pressure in the cavity directly after sealing as
well as a small flux of gasses into the package during its
lifetime. This flux has three sources, leaks through cracks and
voids, internal outgassing and permeation through the sealing
materials. Sealing a high vacuum at low temperature is difficult
due to the often inherently high pressures and temperatures of
sealing processes, e.g. the deposition pressure and temperature
of a CVD sealing or the outgassing from anodic or glass-frit
bonding processes [3], [4]. Furthermore, vacuum sealing at a
low temperature, down to room-temperature, is beneficial in
order to avoid thermal expansion mismatch, thermally induced
performance degradation and thermally induced outgassing.
c 2013 IEEE
1057-7157 1348
JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 22, NO. 6, DECEMBER 2013
Fig. 2. Fabrication scheme. (a) Recess formation by KOH etching. (b) Access port formation by DRIE. (c) Wafer fusion bonding and gold sputter deposition
(d) 1. Using a wire bonder, an electrical discharge forms a ball at the end of a gold wire. 2. Ball bonding with an offset to the access ports. (e) At a chamber
pressure of 10−5 mbar a bond force of 3.2 kN is applied, forcing the bumps into the access ports and thereby hermetically sealing the vacuum cavities. (f)
Silicon diaphragm deflection induced by the external atmospheric pressure.
First, wire bonded gold bumps are placed in such a way that
they partially overlap the access port to a cavity (Fig. 1a).
Second, a wafer-level coining process is performed in a
vacuum chamber by applying a force that causes the bump
to plastically deform, flatten, and flow into the access ports,
thereby clogging the ports as shown in Fig. 1b. This sealing
process is performed using the combination of a high-speed
wire bonder and a commercial wafer bonder, ensuring a costefficient implementation with standard tools.
The placement of the bumps using a high-speed wire bonder
can be a cost-efficient solution even for a large number of
cavities since wire bonding is an extremely mature back-end
technology with very high throughput [12]. For very high volume applications the cost of wire bonding processes has been
reported to be on the order of 14 USD / 100,000 bumps [13].
The only prerequisites for using this method are (1) sufficiently
small ports into the cavity, (2) a sufficient stiffness of the
package material, and (3) a wire-bondable surface.
II. Fabrication
For evaluating the sealing method, test cavities were fabricated according to the process scheme shown in Fig. 2. The
evaluation structures consist of two bonded silicon wafers,
where one has etched cavities and ports. The cavity wafer
was a 545μm thick double-side polished 100 mm diameter
silicon substrate with a 2.2 μm thick silicon dioxide layer on
both sides, created by thermal wet oxidization at 1100°C. The
patterned silicon dioxide acted as a hard mask for a wet etching
step in KOH, which formed the cavities in the cavity wafer,
shown in Fig. 2a. The etch depth was 440 μm, which resulted
in a cavity diaphragm thickness of approximately 105 μm and
a final cavity volume of 50 μl. The silicon dioxide layer was
subsequently removed on the top side by etching in BHF.
A standard lithography on the top side of the substrate
defined the circular openings for the access ports, as indi-
cated in Fig. 2b. These ports were intentionally placed above
the tapered side-wall of the (111) crystal plane exposed by
the KOH etch. This placement had three purposes: (1) this
prevented the fragile 12 × 12 mm2 pressure sensitive silicon
diaphragms from being damaged by the pressure applied by
the wire bonding tool in a later step; (2) this ensures minimal
mechanical deformation of the sealed ports by deflection of the
pressure sensitive diaphragm; and (3) the ports are connected
to the cavity with a single etching step. However, without the
thin pressure sensitive diaphragm, made here for evaluation
purposes, this placement restriction could be avoided. The
access ports were etched into the cavities by a deep reactive
ion etch (DRIE) process.
In preparation for the following wafer fusion bonding, the
bottom protective silicon dioxide layer on the cavity wafer
was removed in 5% HF. The two wafers were thereafter
cleaned and surface activated for bonding by a 10 minute
hydroxylation in a boiling piranha etch, followed by a rinse
in deionized water and spin drying. The fusion bonding was
initiated by joining of the wafers in 1 mbar of N2 , with a
tool force of 1 kN at 300°C. The wafer bond was finally
strengthened by a 3 hour N2 -anneal at 900°C. A 100 nm TiW
+ 500 nm Au sputter deposition on the frontside provided a
bondable layer for the subsequent gold bump sealing process.
This metallization also covers the top of the side-walls of the
access ports, as shown in Fig. 2c.
The vacuum sealing was performed in two steps. First,
gold bumps were wire bonded off-center on the 30 μm
diameter cavity access ports, as illustrated in Fig. 2d. The
gold bumps were bonded at a rate of 14 bumps/s using
a fully automated wire bonder (ESEC 3100+, ESEC Ltd,
Switzerland). The bumping process was optimized to obtain
high and narrow bump shapes. The wire bonding tool is
of a thermo-compression type, where temperature, force and
ultrasonics are applied to bond the gold bumps to the gold
layer on the wafer surface. Increased ultrasonic power and
ANTELIUS et al.: WAFER-LEVEL VACUUM SEALING BY COINING OF WIRE BONDED GOLD BUMPS
1349
Fig. 3. SEM micrographs of wire bonded gold bumps. (a) A gold bump prior to vacuum sealing. The bump has a diameter of approximately 90 μm and
a height of 50 μm. It is placed with an offset of 35 μm from the center of the 30 μm diameter port. (b) A coined gold bump after the vacuum sealing. A
force of approximately 9 N per bump was used to plastically deform the bumps and seal the access ports. The bump is flattened and its height is reduced to
15 μm. (c) Cross sectional view of a sealed access port. The cross section was made by grinding and polishing. The gold has been pressed approximately
90 μm into the access port.
duration was used in order to optimize the bond process
for a low substrate temperature. Typical wire-bonding chuck
temperatures of 100–160°C could thereby be decreased to
40°C in order to maintain a low thermal budget throughout
the whole bump-sealing process.
The cavity wafer was finally transferred to a substrate
bonder (Süss CB8, Süss MicroTec AG, Germany) where the
cavities were sealed at wafer-level by compressing the bumps
in a vacuum environment. A 525 μm thick oxidized silicon
wafer was placed on top of the cavity wafer in order to ensure
a flat and evenly distributed compression of the bumps over
the entire wafer area, as indicated in Fig. 2e. The wafer stack
was placed in the bond chamber and after pumping down for
60 hours, to a vacuum level of 10−5 mbar, a force of 3.2 kN
was applied on the wafer stack by the bond tool. The wafer
contained 350 bumps which means that the corresponding
force per bump was approximately 9 N. The entire compression process was performed with a bond chuck temperature of
30°C. The plastic deformation of the bumps into the access
ports resulted in hermetic sealing of the cavities, as illustrated
in Fig. 2e and f. After removing the wafers from the vacuum
chamber, the pressure difference between the cavity volume
and the atmosphere caused a deflection of the diaphragm as
indicated in Fig. 2f. The completed devices were diced without
extra protection in a Disco DAD 320 dicing saw.
III. Results and Discussion
Fig. 3a shows an SEM image of a gold bump that has been
wire bonded with an offset of 35 μm with respect to the center
of the access port. The effect of the plastic deformation is
clearly visible when comparing the SEM images of the gold
bump before, Fig. 3a, and after compression, Fig. 3b. Gold
bumps that initially had a diameter of 90 μm and a height of
50 μm were flattened to a diameter of 160 μm and a height
of 15 μm by the coining process. The access port has been
completely covered by the coined bump and the top surface of
the bump is flat. An SEM image of a polished cross section of
a sealed access port is shown in Fig. 3c. The image shows that
the compression caused the gold from the bump to completely
fill and seal the access port to a depth of 90 μm.
Fig. 4. Comparison of different gold bump offset positions before and after
coining. The three rows show the same gold bump at an angle, from above
and from above after coining, respectively. The different columns correspond
to different offsets, in μm. The port position is indicated in white.
The used wire bonding tool has a pattern recognition
system, which is able to detect the access port and to place
the stud bump with an accuracy of 2.5 μm (3σ) on the wafer
surface. A comparison of placing gold bumps with offsets
between 35 and 50 μm with respect to the port center is
shown in Fig. 4. This figure shows that an offset of 35 μm
was sufficient for both maintaining an open port before the
compression and being able to seal the port completely by the
compression. Therefore, an offset of 35 μm was used for the
bump-sealing process.
A. Residual Gas Analysis
Two bump-sealed cavity chips were used for a residual gas
analysis (performed by SAES Getters S.p.A, Lainate, Italy).
This is a test where the sealed cavity is mechanically opened
in a vacuum chamber connected to a mass spectrometer. The
mass spectrometer measures the amount and composition of
the gas that escapes from the cavity. The measured results for
the two cavities, one with 80 ports and one with 2 ports, is
reported in Table I.
The achieved vacuum level is lower for the sample with
more ports. This difference correlates to the difference in
evacuation rate caused by the different amount of ports. The
1350
JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 22, NO. 6, DECEMBER 2013
TABLE I
Measured Residual Gas Pressures in Two Sealed Cavities with
Different Amounts of Access Ports
Gas
H2 O
Ar
N2
H2
CH4
O2
CO2
He
Ne
C2 H6
C 3 H8
CO
Kr
Total
Pressure, 2 ports
[mbar]
%
3.33 × 10−3
56.96a
1.00 × 10−3
17.17
8.52 × 10−4
14.56
5.25 × 10−4
8.97
8.75 × 10−5
1.50
2.77 × 10−5
0.47
2.18 × 10−5
0.37
0
0
0
0
0
0
0
0
0
0
0
0
5.85 × 10−3
100.00
Pressure, 80
[mbar]
0
6.15 × 10−4
0
0
0
0
2.04 × 10−4
1.89 × 10−7
0
0
0
0
0
8.19 × 10−4
ports
%
0
75.11
0
0
0
0
24.86
0.02
0
0
0
0
0
100.00
[a] The test was performed at room-temperature, and so
the content of water could be underestimated.
evacuation rate is addressed further in section III-C. The
measured cavity pressure, 8.19×10−4 mbar, is sufficient even
for very demanding applications such as microbolometers,
assuming a low leak rate [14]. The achieved cavity pressure
is well below reported values for cavity sealing using CVD
depositions [15], aluminum sputtering [16] or glass frit reflow [4], [17], which achieve pressures between 1 and 10 mbar.
A sealing method using epitaxial silicon deposition has been
reported to achieve a comparable pressure of 8×10−3 mbar,
although at a temperature of 950°C [18]. Implementing a getter
material would further reduce the cavity pressure at the cost
of an anneal at >300°C for getter activation [14], [19].
The gas analysis revealed a high concentration of argon.
Although argon is present in the ambient air, the main source
of argon in the sealed cavities is considered to be from the
sputter deposited gold, since argon used in thin-film sputter
deposition is known to be incorporated into the sputtered thinfilms during deposition processes, causing argon outgassing
in microcavities [20], [21]. If argon had been leaking into
the bump-sealed cavity from the atmosphere, there would
also have been other species from the atmosphere present in
the cavities, mainly nitrogen and oxygen. However, in the
measured samples, no traces of nitrogen or oxygen could
be measured, indicating a full hermeticity and no measurable external leak. Internal outgassing on the other hand is
inherently finite and decreasing over time. The presumably
argon outgassing gold is located near the sealed openings
of the ports, and has no function here. This means that the
argon contribution could be reduced by depositing the gold by
evaporation instead, reducing the incorporation of argon. The
other internal cavity surfaces are either pristine or previously
annealed at 900°C, thus there should be no argon outgassing
from these surfaces. However, the measured pressures of argon
does not correlate with the larger surface area of gold in the
device with 80 ports, which has a lower argon pressure. But
several other factors could also be affecting the argon level: the
two cavities had different positions on the wafer, which could
affect the pumping rate; a part of the measured argon could
come from the gold sputtered on the outside of the devices,
which is released when the diaphragm is broken during testing;
the chip to chip pressure variations are unknown since the
reported data is only from two packages.
The relatively high water pressure in the cavity with two
ports is typical of an incomplete evacuation, since water is
the most problematic species to desorb from unbaked vacuum
chambers [22]. As water is adsorbed on the chamber walls,
it is dissociated, which in turn can lead to desorption of
H2 CH4 and CO2 into the vacuum chamber [23], [24]. It
should be noted that in [23], [24] the vacuum chamber walls
were studied, which are made of different materials than the
microcavities in this work.
B. Leak Rate
The leak rate of a package is important for estimating the
lifespan of vacuum packaged devices. A short term “worst
case” estimate of the leak rate is to assume that at the time
of the residual gas analysis, two weeks after the vacuum
sealing, the cavity had leaked to the measured vacuum level,
essentially neglecting gas contributions from an incomplete
pump-down of the cavity during sealing. This conservative
assumption yields a leak rate of 2.9 × 10−14 mbarL/s for
the cavities in this work. This leak rate would indicate an
extrapolated cavity pressure of 2 × 10−2 mbar after one year.
This is considered a worst case, since it is not known what
cavity pressure was enclosed during the bump-sealing, and
hence this gives a maximal leak rate. Although the leak rate
level reported here is not as impressive as, e.g., the 8 × 10−17
mbarL/s recently reported by Santagata et al. [15], the cavity
pressure increase in mbar after one year reported in this work
is 500 times lower than [15], mainly due to the smaller cavity
volume in [15].
For a longer term leak rate evaluation, the deflections of
the diaphragms suspended over the vacuum cavities were
monitored over 30 days by white light interferometry. These
measurements yielded the deflection at the center of the
diaphragm, which is directly proportional to the ambient
pressure over the diaphragm [25]. The varying ambient pressure was compensated for using pressure data from a nearby
weather station. The reduction in diaphragm deflection over
time can in turn be used to evaluate the leak rate La using
equation 1 [26], [27],
V
W t1
(1)
La = ln
Wt2
t2 − t1
where W is the surface deformation, V the cavity volume
and t2 − t1 is the time passed between the two deformation
measurements. The measured deflection is presented in Fig.5.
This figure shows that no leak can be detected, and from the
noise level of the measurement, a limit of detection can be
estimated. The variation of the measured deflection data is
determined to be about ±0.25 μm. This variation was higher
than the limit of the measurement system, which had been
measured to less than ±0.14 μm (2σ, including the pressure
compensation) for a 24 h long continuous measurement of
the diaphragm deflections. This limit was however achieved
ANTELIUS et al.: WAFER-LEVEL VACUUM SEALING BY COINING OF WIRE BONDED GOLD BUMPS
Fig. 5. Measurement made with an optical profilometer of the diaphragm
deflection over time in order to evaluate the sealing. The data has been
normalized and compensated for varying ambient pressure. 11 out of 15
cavities are followed over a month, the remaining four were removed from
the experiment and used in other tests.
1351
Fig. 6. Calculated cavity pressures during evacuation through a single square
access port of five different dimensions. Port dimensions are in μm2 .
without unloading and loading the wafer onto the stage of
the white light interferometer, which may be the source of
the additional uncertainty. The variation in the deflection data
yields a leak rate better than 9×10−11 mbarL/s. This estimated
leak rate is lower than the typical range of the helium fine
leak test [26]. However, it is 3000 times higher than the leak
rate value from the residual gas analysis data, meaning that
this method would only detect sharp and unexpected pressure
increases over much longer time intervals. No such events have
been observed here. For comparison, the leak rate measured by
residual gas analysis after 15 days corresponds to a deflection
change of 2 nm after one year.
as gas sinks, trapping gas species with a long residence time,
when only the wafer chucks are heated in the tool.
A way to increase the evacuation rate would be to increase
the diameter of the access ports. The maximal port diameter
is limited by the available volume of gold in the bump,
and the applied pressure for deformation. The gold bump
volume, or the size of the gold ball, is limited by the wire
diameter, which in this work was fixed by the wire bonding
tool configuration to 25 μm.
The preferred way to increase the wafer throughput for the
sealing step would be to evacuate and seal more wafers at the
same time. This is possible, since the process step performed
in the wafer bonder can be run as a batch process. Several
wafers could be stacked on top of each other without any
anticipated adverse effects on pumping and sealing, thereby
drastically increasing the throughput.
C. Evacuation Rate
D. Stress Testing
The evacuation rate of the cavities is important for the
fabrication throughput. The rate of evacuation can be modeled
by treating the evacuation port as a resistor draining a fluidic
capacitor, the cavity [6], [28]. Using this model the cavity
pressure decrease during evacuation through one rectangular
port, of different sizes, was calculated and plotted in Fig. 6.
This figure shows that the model predicts sufficient pumping
of the cavity already with a single port. However, this simple
model omits the slow surface desorption of certain species,
i.e. water, meaning that the pumping can take longer time
than predicted. This slow surface desorption, or “surface
stickiness”, of water is most likely the dominating reason
for the slow evacuation of the cavity with only two ports.
This analysis corresponds to the higher pressure measured by
residual gas analysis of the chip with only two ports, where the
dominating pressure contribution was from H2 O. The situation
would have been different if the wafers had been heated in the
vacuum chamber before the sealing process. For example, the
residence time for water on a metal surface is reduced from
105 to 10−5 s when the metal surface is heated from 22 to
450°C [22]. Additionally, the cold chamber walls would act
Two vacuum sealed packages were temperature cycled 500
times between -10 and 70°C. The cycling was done using
a single-chamber setup at a rate of 1.5 cycles/h with a
5 minutes wait at the endpoint temperatures, all in accordance
to JEDEC test method A104D [29]. The hermeticity of the cycled packages were evaluated using the diaphragm deflection
described in section III-B. This evaluation showed no change
in diaphragm deflection and no leaks, with a cavity pressure
limit of detection of 3 mbar.
Two vacuum sealed cavities were subjected to additional
overpressures of compressed air to test the mechanical robustness of the seals. The cavities were put in a chamber with
overpressures of 0.5, 1.5, 3 and 5 bar, each for a duration of
15–20 h. The packages were evaluated using the diaphragm
deflection method as above and showed no measurable leaks.
IV. Conclusion
We have described and demonstrated a novel roomtemperature vacuum sealing process using coining and deformation of wire bonded gold bumps into access ports of
1352
JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 22, NO. 6, DECEMBER 2013
cavities. The measured cavity pressure was 8×10−4 mbar,
and the content of the residual gas is dominated by argon
and carbon dioxide after 60 hours of pumping, indicating
that the source is internal outgassing. The corresponding
leak rate for the evaluated cavity is less than 2.9×10−14
mbarL/s, presumably with no measurable external leak contribution. The evaluation of the cavity pressure, leak rate
and sealing robustness validates the feasibility of this vacuum sealing method, which enables uncomplicated and cost
efficient vacuum sealing at room-temperature using standard
commercial processing tools and processes on a variety of
substrates.
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
References
[27]
[1] M. Antelius, A. C. Fischer, N. Roxhed, G. Stemme, and F. Niklaus,
“Room-temperature wafer-level vacuum sealing by compression of
high-speed wire bonded gold bumps,” in Proc. Transducers, 2011,
pp. 1360–1363.
[2] M. Esashi, “Wafer level packaging of MEMS,” J. Micromech. Microeng.,
vol. 18, no. 7, p. 073001+, May 2008.
[3] H. Henmi, S. Shoji, Y. Shoji, K. Yoshimi, and M. Esashi, “Vacuum
packaging for microsensors by glass-silicon anodic bonding,” Sens.
Actuators A, vol. 43, no. 1–3, pp. 243–248, May 1994.
[4] R. Knechtel, “Glass frit bonding: An universal technology for wafer level
encapsulation and packaging,” Microsyst. Technol., vol. 12, no. 1–2, pp.
63–68, Dec. 2005.
[5] S. N. Farrens, J. R. Dekker, J. K. Smith, and B. E. Roberds, “Chemical
free room temperature wafer to wafer direct bonding,” J. Electrochem.
Soc., vol. 142, no. 11, pp. 3949–3955, 1995.
[6] Y. T. Cheng, W.-T. Hsu, K. Najafi, C. T. C. Nguyen, and L. Lin,
“Vacuum packaging technology using localized aluminum/silicon-toglass bonding,” J. Microelectromech. Syst., vol. 11, no. 5, pp. 556–565,
Oct. 2002.
[7] M. A. Lapisa, F. Niklaus, and G. Stemme, “Room-temperature waferlevel hermetic sealing for liquid reservoirs by gold ring embossing,” in
Proc. Transducers, 2009, pp. 833–836.
[8] A. Decharat, J. Yu, M. Boers, G. Stemme, and F. Niklaus, “Roomtemperature sealing of microcavities by cold metal welding,” J. Microelectromech. Syst., vol. 18, no. 6, pp. 1318–1325, Dec. 2009.
[9] M. Antelius, A. C. Fischer, F. Niklaus, G. Stemme, and N. Roxhed,
“Hermetic integration of liquids using high-speed stud bump bonding
for cavity sealing at the wafer level,” J. Micromech. Microeng., vol. 22,
no. 4, p. 045021, March 2012.
[10] L. Levine, “Ball bumping and coining operations for TAB and
Flip Chip,” in Proc. 3rd Int. Symp. Adv. Packaging Mater., 1997,
pp. 110–112.
[11] R. S. Pai, M. M. Crain, and K. M. Walsh, “Maskless shaping of gold
stud bumps as high aspect ratio microstructures,” Microelectron. Eng.,
vol. 88, no. 1, pp. 135–139, Jan. 2011.
[12] G. Harman, Wire Bonding in Microelectronics, 3rd ed. New York, NY,
USA: McGraw-Hill, 2010.
[13] V. McTaggart, L. Levine, and G. Dunn, “Stud bumping and die attach
for expanded flip chip applications,” Adv. Packag., vol. 13, no. 9, pp.
30–34, Sep. 2004.
[14] R. Ramesham and R. C. Kullberg, “Review of vacuum packaging
and maintenance of MEMS and the use of getters therein,” J. Micro/Nanolithogr. MEMS MOEMS, vol. 8, no. 3, p. 031307+, Jul. 2009.
[15] F. Santagata, J. F. Creemer, E. Iervolino, and P. M. Sarro, “Tubeshaped Pirani gauge for in situ hermeticity monitoring of SiN thin-film
encapsulation,” J. Micromech. Microeng., vol. 22, no. 10, p. 105025+,
Sep. 2012.
[16] R. N. Candler, W.-T. Park, H. Li, G. Yama, A. Partridge, M. Lutz, and
T. W. Kenny, “Single wafer encapsulation of MEMS devices,” IEEE
Trans. Adv. Packag., vol. 26, no. 3, pp. 227–232, Aug. 2003.
[17] D. Sparks, J. Trevino, S. Massoud-Ansari, and N. Najafi, “An all-glass
chip-scale MEMS package with variable cavity pressure,” J. Micromech.
Microeng., vol. 16, no. 11, p. 2488, Oct. 2006.
[18] R. N. Candler, M. A. Hopcroft, B. Kim, W.-T. Park, R. Melamud,
M. Agarwal, G. Yama, A. Partridge, M. Lutz, and T. W. Kenny,
“Long-term and accelerated life testing of a novel single-wafer vacuum
[28]
[29]
encapsulation for MEMS resonators,” J. Microelectromech. Syst., vol. 15,
no. 6, pp. 1446–1456, Dec. 2006.
B. Ferrario, “Chemical pumping in vacuum technology,” Vacuum,
vol. 47, no. 4, pp. 363–370, Apr. 1996.
H. F. Winters and E. Kay, “Gas incorporation into sputtered films,” J.
Appl. Phys., vol. 38, no. 10, pp. 3928–3934, Sep. 1967.
W. W. Y. Lee and D. Oblas, “Argon entrapment in metal films by dc
triode sputtering,” J. Appl. Phys., vol. 46, no. 4, pp. 1728–1732, Apr.
1975.
J. F. O’Hanlon, A User’s Guide to Vacuum Technology, 2nd ed. New
York, NY, USA: Wiley, 1989.
A. Klopfer, S. Garbe, and W. Schmidt, “Residual gases in vacuum
systems,” Vacuum, vol. 10, no. 1–2, pp. 7–12, Feb. 1960.
I. Marklund and H. Danielsson, “Residual gas analysis in a vacuum
evaporator,” Vacuum, vol. 16, no. 7, pp. 365–371, Jul. 1966.
M. Di Giovanni, Flat and Corrugated Diaphragm Design Handbook,
1st ed. New York, NY, USA: Marcel Decker, Jan. 1982.
A. Goswami and B. Han, “On ultra-fine leak detection of hermetic wafer
level packages,” IEEE Trans. Adv. Packag., vol. 31, no. 1, pp. 14–21,
Feb. 2008.
M. Antelius, G. Stemme, and F. Niklaus, “Small footprint wafer-level
vacuum packaging using compressible gold sealing rings,” J. Micromech.
Microeng., vol. 21, no. 8, p. 085011+, Aug. 2011.
B. H. Stark and K. Najafi, “A low-temperature thin-film electroplated
metal vacuum package,” J. Microelectromech. Syst., vol. 13, no. 2, pp.
147–157, Apr. 2004.
Temperature Cycling, JEDEC Std. JESD22-A104D, Mar. 2009.
Mikael Antelius received the M.Sc. degree in chemical engineering from Uppsala University, Uppsala,
Sweden, in 2007.
After a one year stay in the industry, he joined the
Department of Micro and Nanosystems, School of
Electrical Engineering, KTH Royal Institute of Technology, Stockholm, Sweden, where he will defend
his Ph.D. thesis in April 2013. His current research
interests include silicon photonics and wafer-level
vacuum and liquid packaging, particularly regarding
gas sensors.
Andreas C. Fischer was born in Baden-Baden, Germany. He received the Diploma degree in microsystems engineering from the University of Freiburg,
Freiburg, Germany, and the Ph.D. degree in microsystem technology from the KTH Royal Institute
of Technology, Stockholm, Sweden, in 2008 and
2013, respectively.
He is currently a Post-Doctoral Researcher at
the Department of Micro and Nanosystems, School
of Electrical Engineering, KTH Royal Institute of
Technology. His current research interests include
the development of novel integration and fabrication techniques for 3-D microand nanodevices, heterogeneous integration of MEMS, and IC technology as
well as advanced wire bonding technology.
Niclas Roxhed (M’09) was born in Stockholm,
Sweden, in 1978. He received the M.Sc. degree
in electrical engineering and the Ph.D. degree in
microsystem technology from KTH Royal Institute
of Technology, Stockholm, Sweden, in 2003 and
2007, respectively.
He is currently an Assistant Professor and a Team
Leader of Medical MEMS with the Department of
Micro and Nanosystems, School of Electrical Engineering, KTH Royal Institute of Technology. His
current research interests include sensors for medical
diagnostics and medical-aid microsystems. He is also involved in highprecision etching using DRIE for RF-MEMS switches and 3-D integration
of MEMS on ICs for infrared imagers. He has authored or coauthored more
than 50 scientific papers.
ANTELIUS et al.: WAFER-LEVEL VACUUM SEALING BY COINING OF WIRE BONDED GOLD BUMPS
Göran Stemme (F’06) received the M.Sc. degree
in electrical engineering and the Ph.D. degree in
solid-state electronics from Chalmers University of
Technology, Gothenburg, Sweden, in 1981 and 1987,
respectively.
In 1981, he joined the Department of Solid State
Electronics, Chalmers University of Technology,
where he became an Associate Professor (docent)
heading the silicon sensor research group in 1990.
Since 1991, he has been a Professor at KTH Royal
Institute of Technology, Stockholm, Sweden, where
he is the Head of the Department of Micro and Nanosystems, School of
Electrical Engineering. His research is devoted to microsystem technology
based on micromachining of silicon. He has published more than 300 research
journal and conference papers and has more than 22 patent proposals or
granted patents.
Dr. Stemme is a member of the Royal Swedish Academy of Sciences (KVA).
1353
Frank Niklaus (M’06) received the M.Sc. degree in
mechanical engineering from the Technical University of Munich, Munich, Germany, in 1998, and the
Ph.D. degree in MEMS from KTH Royal Institute
of Technology, Stockholm, Sweden, in 2002.
He was a Consultant to the industry and a Visiting
Scholar at Rensselaer Polytechnic Institute, Troy,
NY, USA, where he was with 3-D-IC technologies.
He is currently an Associate Professor at the Department of Micro and Nanosystems, School of Electrical Engineering, KTH Royal Institute of Technology,
where he is the Team Leader for all micro- and nanosystem integration
activities.