Thanks to visit codestin.com
Credit goes to www.scribd.com

0% found this document useful (0 votes)
20 views5 pages

Mems

Uploaded by

sfar aida
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
20 views5 pages

Mems

Uploaded by

sfar aida
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Silicon Wafer Bonding for Micromechanical Devices

Martin A. Schmidt
Microsystems Technology Laboratories
Department of Electrical Engineering and Computer Science
Massachusetts Institute of Technology
Cambridge, MA 02139

ABSTRACT conducted in a cleanroom environment. The surface preparation


step involves cleaning the mirror-smooth, flat surfaces of two
The process for bonding silicon wafers together at high temperature wafers to form a hydrated surface. Following this preparation, the
is reviewed. Specific details of the bonding process as they pertain wafers are contacted in a clean environment by gently pressing the
to the formation of micromechanical devices are described. Methods two surfaces together at one point. The surfaces come into contact
of characterization of the bonded wafers are discussed. at this point and are bound by a surface attraction of the two
hydrated surfaces. A contact wave is initiated at this point and
sweeps across the wafer surfaces, bringing them into intimate
INTRODUCTION contact over the entire surface. The exact origin of the attractive
force is not universally agreed upon [3], and depends to a certain
Micromachining encompasses a broad range of technologies extent on whether the bond is Si-Si or Si-SiO2. The most common
anchored in the core technology of microlithographic pattern assumption is that a bond is formed between -OH groups on the
transfer. A large fraction of the micromachining technologies are opposing surfaces. The final step in the bonding process is a high
specific to the silicon material system principally due to the origins temperature anneal of the contacted pair at temperatures between
of the field, namely the silicon integrated circuit industry. In the 800-1200C. While the room temperature contacted samples are well
silicon micromachining field, there have been two dominant adhered, this anneal generally increases the bond strength by more
fabrication methods, broadly classified as bulk micromachining than an order of magnitude. Measurements of the bond strength as a
(etching deep features into a wafer) and surface micromachining function of anneal temperature indicate three distinct regions. The
(depositing, patterning, and selective etching of films on a wafer). first region, for anneal temperatures less than 300C, exhibits a
Fundamentally, both of these techniques rely on some form of relatively constant bond strength equal to the bond strength of the
etching or material removal. More recently, techniques have wafers prior to. anneal. At temperatures greater than 300C, the bond
emerged for bonding or fusing of silicon wafers. As the bulk or strength increases and then levels out. It is presumed that a Si-O-Si
surface micromachining technologies might be compared to material bridging bond is formed between the surfaces and a water molecule
removal processes in conventional machining (end-milling, drilling), is liberated. At temperatures greater than 800C, the bond strength
the wafer bonding processes are analogous to welding processes in begins to increase again. In this third region it has been suggested
conventional machining. that surfaces can more easily deform (oxide flow) and trapped water
may oxidize surfaces bringing them into better contact. At
There is some potential for confusion in discussing wafer bonding temperatures of lOOOC or greater, the bond strength is in the range
processes since a wide range of processes exist. For example, the of the strength of the silicon crystal itself.
silicon-glass anodic bond is routinely used in many commercially
available sensors. Additionally, low-temperature metal eutectic
bonds are often used. While these bonds are quite useful in low WAFER BONDING CHARACTERIZATION
temperature, back-end processes, they generally are not applicable
as high temperature stable bonds. Thus the distinguishing feature Several non-destructive and destructive techniques exist for
between these methods and the bonding that we will discuss in this mechanical characterization of the bonding process. These
paper will be the ability of the bond to withstand high temperature techniques are bond imaging, cross-sectional analysis, and bond
processing without the need for intermediate layers, externally­ strength measurement. The imaging methods are non-destructive
applied pressure, or electrostatic fields to assist the bond. This bond and can be used as in-process monitors, while the cross-sectional
can be performed between nearly any smooth surfaces [1-3], but analysis and bond strength measurements are destructive and require
generally, we will restrict our discussion to bonding of silicon control wafers for characterization.
wafers with or without silicon dioxide.
The three dominant methods for imaging a bonded pair of silicon
Interest in the process of wafer bonding as defined in the preceding wafers are infrared transmission, ultrasonic, and X-ray topography.
paragraph was generated by the publications in 1985-86 of Lasky Examples of the images obtained by these methods for a poorly
[4] and Shimbo [5]. Subsequently, a number of investigators have bonded 4" silicon wafer pair are shown in Figure 1. A simplified
explored the use of this process for fabrication of electronic devices schematic of an infrared imaging system is shown in Figure 2. It
(SOI MOSFETs, Power Devices) [l-6]. Several commercially consists of an IR source (typically an incandescent light bulb), and
available electronic products exist today which employ silicon wafer an IR-sensitive camera. A silicon CCD camera has sufficient
bonding. Additionally, a number of investigators considered the sensitivity in the near-IR range that it can be used when outfitted
application of wafer bonding to sensor and actuator structures. Two with a filter for visible light. The bonded wafer pair is located
of the first sensors to be fabricated by wafer bonding were reported between the source and camera. Any imperfection in the bond
at the 1988 Solid-State Sensor and Actuator Workshop in Hilton shows up as changes in contrast in the IR image. Large un-bonded
Head [7 ,8]. These included a pressure sensor and an accelerometer. regions ('voids') appear with a characteristic 'Newton's Rings'
Since that time, a significant number of sensors and actuators have pattern. This imaging method generally can not image voids with a
been reported which employ silicon wafer bonding. Several separation of surfaces less than one quarter of the wavelength of the
micromachined sensor products which use silicon wafer bonding are IR source. Based on a typical particle void, this translates to a
now available commercially. spatial resolution of several millimeter. Figure 1 clearly illustrates
voids not present in the IR image which do show up in the other
methods. Also, this technique works for silicon wafers of moderate
THE BONDING PROCESS doping level with smooth surfaces. Highly doped layers, IR
absorbing films, or rough surfaces (backside of wafer), can limit the
Extensive review articles have been written on the wafer bonding image quality. In spite of this resolution limit, the IR method has
process, particularly as it pertains to electronic device fabrication (1- the advantage of being simple, fast and inexpensive. It can be used
3 ). This section will simply summarize the major points of the directly in the cleanroom to image the wafers before and after
process for the bonding of silicon or silicon dioxide surfaces. The anneal. The other two imaging methods offer higher resolution at
silicon wafer bonding process consists of three basic steps: surface the expense of speed, cost and incompatibility with cleanroom
preparation, contacting, and annealing. All of the process steps are T)rocessing.

0-9640024-0-X/hh1994/$20©1994TRF 127 Solid-State Sensors, Actuators, and Microsystems Workshop


DOI 10.31438/trf.hh1994.30 Hilton Head Island, South Carolina, June 12-16, 1994
X-RAY TOPOGRAPH INFRARED IMAGE ULTRASONIC IMAGE

Figure 1. Images of a Bonded Silicon Wafer Pair by Several


Methods. (Courtesy of Dr. T. Abe - SEH)

Cross-sectional analysis can be performed at the bonded interface by IR


SENSITIVE
cleaving the sample. SEM and TEM techniques have been used to CAMERA

image the bonded interface at a submicron scale. These studies have


helped to understand the composition of the bonded interface [3].
Additionally, it is possible to gain a great deal of information about
the bonded interface by simply defect etching the cross-sectioned
sample. Several groups have demonstrated the benefit of this
approach, particularly for visualization of voids on the order of tens
of microns ('microvoids') [9].
The bond strength has been characterized by a number of
techniques. Figure 3 highlights the most common techniques.
Pressure burst tests can often yield a number which has engineering
significance in the design of sensors, but yields little information
about the detailed nature of the bond due to the complicated loading
of the interface. A tensile/shear test sample gives better information
on the bonded interface, but is often limited by difficulties in loading
and sample handling [10]. The knife-edge technique has the IR
advantage of creating a very well defined loading on the bonded SOURCE

interface. A blade of defined thickness is inserted between the


bonded pair in a region where a crack has been initiated. Using IR
imaging methods, the length of the crack is measured, from which Figure 2. IR Imaging System
the surface energy can be inferred through a knowledge of the
sample and blade thicknesses and the elastic properties of the wafer
[11]. This method has been used with very good success.
Unfortunately, the surface energy is forth power dependent on the
crack length, and thus uncertainties in the crack length produce large
uncertainties in the extracted surface energy. Other methods based
on patterned samples have been proposed to eliminate this problem
[12].

DETAILED PROCESS ISSUES TENSILE OR


SHEAR LOAD

A number of factors contribute to the success of the wafer bonding


process. Some of the more important details as they pertain to
application in micromechanics are summarized in this section.

Starting Material and Surface Preparation

The large scale and small scale roughness of the wafer surface is
very important to the success of a wafer bonding process. It is
difficult to establish strict requirements on the needed wafer bow
and microscale surface roughness since the surface preparation and
contacting methods play a large roll. Generally, people have found
that VLSI grade wafers tend to be acceptable for bonding if their
microscale roughness is less than 5 A (measured optically or by
STM) and the bow is of the order of 5 µm (on a 4" wafer). Abe and
Masara have done careful studies of the influence of roughness
[13,14]. Protrusions and particles are sources of problems.
Protrusions might be present in deposited films (epi-spikes) or be
formed by processes such as oxidation of etched cavities prior to Figure 3. Examples of three bond strength measurements.
bonding [15]. Wafer polishing can be employed to minimize this
problem.

128
As Contacted 400C - 1 Hour 600C - 1 Hour

Figure 4. IR image of intrinsic voids formed in a silicon-silicon


wafer bonded pair as a function of anneal temperature.

Generally, it is believed that the surface preparation should include a Annealing


vigorous hydration of the surface to make it hydrophilic, followed
by a de-ionized water rinse and spin dry. Processes ranging from 1In general, the highest possible anneal temperature should be used to,
simple water rinses to hot acid dips have been used for the get the best quality bond. There is very little evidence to suggest
hydration. However, people have also reported that HF dips, which that the ambient used during the anneal has an impact on the bond
produce a hydrophobic surface, can also be used. While the quality. When processing wafers with cavities, it is important to
hydrophobic wafers do not contact as easily as hydrophilic wafers, slowly ramp the annealing temperature. Rapid temperature rises can
there is some evidence that the final bond is better [3]. A major cause the gas in the cavity to expand, building up a pressure which
concern in surface preparation is hydrocarbon contamination. can separate the wafers before the bond has time to anneal. All
Exposure to plastic wafer holders has been shown to cause evidence suggest that the bonding is generally complete within
hydrocarbon contamination of the surfaces. These contaminants are minutes of reaching the anneal temperature, although there is
not readily removed by standard wafer cleans, and can cause evidence that some incremental increase in bonding occurs over
complications in the bonding as will be discussed later. Various much longer times.
groups have tried methods such as high temperature bakes and
oxidation followed by oxide strips prior to the hydration step to Structure of the Bonded Interface
minimize this problem.
When bonding silicon dioxide surfaces either to silicon or silicon
Contacting dioxide, the bonded interface appears, based on TEM, to be nearly
identical to a bulk silicon dioxide film or a thermally-grown
The contacting is performed immediately after the surface silicon/silicon dioxide interface. Occasionally, there are small
preparation to minimize contamination. This is done in a cleanroom, (micron-scale) voids or occlusions present. In contrast, the
although some work has demonstrated the ability to do the bonding structure of the silicon-silicon bonded interface is very'Sensitive to
in a 'micro-cleanroom' [16] and under water [17]. The contact surface preparation (hydrophobic/hydrophilic), the crystal alignment
should be initiated at one point and allowed to propagate across the of the wafer, and the type of crystal (czochralski or float-zone).
wafer surfaces. Contacting at multiple points can cause air pockets Bengtsson provides a detailed summary of the observations of this
to be trapped between the wafers. As discussed later, in some interface [3]. For hydrophilic surfaces, an interfacial SiOx layer is
instances it is desirable to contact the wafers in an ambient other than observed (5-40 A thick), but this layer can breakdown or form
air, such as oxygen or vacuum. There is considerable interest in spheroids of SiOx under conditions of perfect alignment of crystal
aligned bonding of wafers, and a variety of schemes have been planes. Preliminary evidence suggests that the hydrophobic
proposed. These include the use of specialized alignment tools or surfaces do not have as much of an interfacial layer, which would be
mechanical fixtures which align to previously etched features in the consistent with the removal of the native oxide layer during the HF
wafer. Two common methods of mechanical alignment are to use dip used to create the hydrophobic surface.
optical fibers in v-grooves etched in the edge of the wafers, or to use
reference flats on the wafers.

129
Voids
When contacting and annealing a wafer pair, voids are sometimes
observed upon inspection. These voids are generally lumped in two
categories; extrinsic and intrinsic. The extrinsic voids are those
created by particles, protrusions on the wafers surface, or trapped
air. These voids are usually observed on contact and do not change
significantly during annealing. Figure 1 shows a wafer with various
forms of extrinsic voids. Intrinsic voids are voids which are

C
generated during the anneal cycle. Figure 4 is a series of IR images CONTACT
of a silicon-silicon bonded pair annealed at increasing temperatures.
After contact, the wafer pair appears void-free. As the anneal \ P=1atm /

temperature is increased, voids begin to appear above 400C, and


subsequently disappear above 900C. The voids are usually only
seen when bonding silicon to silicon without an intermediate oxide
and they are often attributed to hydrocarbon contamination, although HIGH TEMPERATURE ANNEAL
there is not a consensus on their origin. It has been observed that
cavities in the wafers can serve to 'getter' these microvoids, thus
minimizing the problem [18].

GENERALIZED BONDING
While we have exclusively discussed bonding of silicon wafers, the
same basic process steps can be applied to bonding a variety of
materials. Quartz wafers can be bonded by this method [19].
\ P• O.Batm /
THIN

'
Examples of bonding of dissimilar materials include the bonding of
GaAs to Si [20] and Si to glass [21]. Provided the surfaces are
mirror smooth and can be hydrated, the bonding proceeds in a
fashion identical to Si-Si bonding. When bonding dissimilar Figure 5. Sealed Cavity Formation.
materials, the major complication is stresses generated during the
high temperature anneal due to differences in thermal coefficient of thickness control between 0.25-1.0 µm. Chemical etch-stops such
expansion of the two materials [21]. as the p ++ etch-stop have also been used with good success.
However, several complications result in using this process. First,
Bonding has also been demonstrated for samples with deposited the surface roughness of a p++ layer is sometimes too great to
films. Examples of bonding to silicon wafers with deposited achieve good bonding. Second, the residual stress in the layer can
polysilicon or silicon nitride have been reported [22-24]. Polishing produce a large wafer bow which prevents bonding. Lastly, the p++
is often needed to establish the necessary level of roughness. layer is formed prior to bonding, and thus the bonding temperature
and time must be minimized to eliminate diffusion of the layer.
Counter doping the p++ with Ge to reduce the stress in the material
BONDING WITH SEALED CAVITIES can eliminate some of these problems. Electrochemical etch-stop
methods have been used successfully in a number of wafer bonding
Nearly all micromechanical applications of silicon wafer bonding applications [27]. One disadvantage of this etch-stop is that it
require bonding of wafers with cavities etched in one or the other requires complicated fixturing.
wafer, thus forming sealed cavities in the wafer after the bond. The
nature of gases that exist in the cavities can be very important
particularly in subsequent high temperature bonding. It has been CONCLUSIONS
shown that when wafers are contacted in air, and subsequently
annealed at high temperature, the oxygen in the cavity can react with The wafer bonding process is an extremely powerful process for
the silicon surface and create a partial vacuum (Figure 5) [25]. micromechanical devices. It complements other micromachining
When the oxygen is completely consumed (for shallow cavities), the techniques by permitting the 'welding' of silicon wafers. This paper
resultant pressure inside the cavity is 0.8 atm, consistent with the reviewed the methods for silicon wafer bonding. A detailed
consumption of the 20% oxygen in air. These results indicate that description of the process of wafer bonding has been described.
the bonding process forms rapidly enough that it can trap gasses in Issues which impact the success of the wafer bonding process were
the cavities. Under high temperatures, the residual gases trapped in identified. While a better understanding of the process is still
the cavities can induce plastic deformation in thin silicon membranes possible, the bonding process is readily applicable in a broad range
as the gases expand [25]. This problem can be reduced or of devices.
eliminated by controlling the ambient under which the wafers are
contacted. We have demonstrated that the pressure inside the cavity
can be reduced by bonding the wafers in an oxygen rich ambient ACKNOWLEDGMENTS
[26]. Alternatively, it is possible to bond the wafers in a vacuum.
The wafer bonding work at MIT has been sponsored by a number of
organizations including SRC (Contract 93-SC-309), Delco
THINNING METHODS Electronics, 3M, Bosch, ARPA (Contract J-FBI-92-196), Draper,
and the NSF Presidential Young Investigator Award. Many of the
Controlled wafer thinning is often a necessary process following the ideas expressed in this paper are the result of extensive discussions
wafer bonding. This is particularly true in microelectronic with students and staff at MIT who worked on wafer bonding.
applications such as SOI, but also in micromechanics. The two Those persons include Errol Arkilic, Chris Bang, Mousumi Bhat,
methods usually applied are precision grinding/polishing and . John Gilbert, Howard Goldberg, Peter Gravesen, Charles Hsu,
chemical etching wi� etch-stops. Most of these techniques have Michael Huff, Vincent McNeil, Kay-Yip Ng, Mitch Novack, Lalitha
been very effectively summarized in review articles on wafer Parameswaran, Javad Shajii, Daniel Sobek, Simon S. Wang, and
bonding [1,2]. The precision grinding and polishing yield absolute Albert Young.

130
REFERENCES [15] M.A. Huff, and M.A. Schmidt, "Fabrication, Packaging, and
Testing of a Wafer-Bonded Microvalve," IEEE Solid-State
[1] J. Haisma, G.A.C.M. Spierings, U.K.P. Biermann, and J.A. Sensor and Actuator Workshop, Technical Digest, Hilton
Pals, "Silicon-on-Insulator Wafer Bonding-Wafer Thinning Head Island, S.C., p. 194, 1992.
Technological Evaluations," Japanese Journal of Applied [16] R. Stengl, K.-Y. Ahn, and U. Gosele, "Bubble-Free Silicon
Physics, Vol. 28, No. 8, p.1426, August 1989. Wafer Bonding in a Non-Cleanroom Environment," Japanese
[2] W.P. Maszara, "Silicon-on-Insulator by Wafer Bonding: A Journal of Applied Physics, Vol. 27, No. 12, p.L2364, 1988.
Review," Journal of the Electrochemical Society, Vol. 138, [17] C. Parkes, E. Murray, H.S. Gamble, B.M. Armstrong,
No. 1, p.341, January 1991. S.J.N. Mitchell, and G.A. Armstrong, "Characterization of
[3] S. Bengtsson, "Semiconductor Wafer Bonding: A Review of Electronic Devices Employing Silicon Bonding Technology,"
Interfacial Properties and Applications," Journal of Electronic Proceedings of the First International Symposia on
Materials, Vol. 21, No. 8, p.841, August 1992. Semiconductor Wafer Bonding: Science, Technology, and
[4] J.B. Lasky, S.R. Stiffler, F.R. White, and J.R. Abemathey, Applications, Vol. 92-7, p.321, The Electrochemical Society,
"Silicon-on-Insulator (SOI) by Bonding and Etch-Back," 1992.
Technical Digest, 1985 IEEE International Electron Devices [18] W. Kissinger and G. Kissinger, "Microstrucutures for Perfect
Meeting, p.684, December 1985. Wafer Bonding in Different Temperature Ranges,"
[5] M. Shimbo, K. Furukawa, K. Fukuda, and T. Tanzawa, Proceedings of the First International Symposia on
"Silicon-to-Silicon Direct Bonding Method," Journal of Semiconductor Wafer Bonding: Science, Technology, and
Applied Physics, Vol. 60, No. 8, p.2987, October 1986. Applications, Vol. 92-7, p.73, The Electrochemical Society,
[6] see Proceedings of the First and Second International 1992.
Symposia on Semiconductor Wafer Bonding: Science, [19] D. Sobek, S.D. Senturia, and M.L. Gray, "Microfabricated
Technology, and Applications, The Electrochemical Society. Fused Silica Flow Chambers for Flow Cytometry," Solid­
[7] K. Petersen, P. Barth, J. Poydock, J. Mallon, and J. Bryzek, State Sensor and Actuator Workshop, Technical Digest, Hilton
"Silicon Fusion Bonding for Pressure Sensors," Technical Head Island, S.C., 1994.
Digest, IEEE Solid�State Sensor and Actuator Workshop, [20] G.G. Goetz, "Generalized Reaction Bonding," Proceedings of
Hilton Head, S.C., p. 144, 1988. the First International Symposia on Semiconductor Wafer
[8] P. Barth, F. Pourahmadi, R. Mayer, J. Poydock, and K. Bonding: Science, TecJ:mology, and Applications, Vol. 92-7,
Petersen, "A Monolithic Silicon Accelerometer with Integral p.65, The Electrochemical Society, 1992.
Air Damping and Over-range Protection," Technical pigest, [21] T. Abe, K. Ohki, K. Sunakawa, K. Yoshizawa, S. Tanaka,
IEEE Solid-State Sensor and Actuator Workshop, Hilton and Y. Nakazato, "Bonded SOI Wafers with Various
Head, S.C., p. 35, 1988. Substrates for IC Fabrication," Proceedings of the Second
[9] R.D. Homing, and R.R. Martin, "Wafer-to-Wafer Bond International Symposia on Semiconductor Wafer Bonding:
Characterization by Defect Decoration Etching," Proceedings Science, Technology, and Applications, Vol. 93-29, p.32,
of the Second International Symposia on Semiconductor The Electrochemical Society, 1993.
Wafer Bonding: Science, Technology, and Applications, Vol. [22] W.G. Easter, G.T. Jones, R.H. Shanaman, and C.A.
93-29, p.199, The Electrochemical Society, 1993. Goodwin, "Polysilicon to Silicon Bonding in Laminated
[10] S.N. Farens, B.E. Roberds, J.K. Smith, and C.E. Hunt, Dielectrically Isolated Wafers," Proceedings of the First
"Analysis of Bond Characteristics in Si Direct-Bonded International Symposia on Semiconductor Wafer Bonding:
Materials," Proceedings of the Second International Symposia Science, Technology, and Applications, Vol. 92-7, p.223,
on Semiconductor Wafer Bonding: Science, Technology, and The Electrochemical Society, 1992.
Applications, Vol. 93-29, p.81, The Electrochemical Society, [23] V.M. McNeil, M.J. Novack, and M.A. Schmidt, "Design and
1993. Fabrication of Thin Film Microaccelerometers using Wafer­
[11] W.P. Maszara, G. Goetz, A. Caviglia, and J.B. McKitterick, Bonding", Technical Digest, 7th International Conference on
"Bonding of Silicon Wafers for Silicon-on-Insulator," Journal Solid-State Sensors and Actuators, Yokohama, Japan, pp.
of Applied Physics, Vol. 64, No. 10, p.4943, Nov. 1988. 822-825, June 1993.
[12] R. D. Homing, D.W. Bums, and A.I. Akinwande, "A Test [24] C.A. Bang, J.P. Rice, M.I. Flik, D.A. Rudman, and M.A.
Structure for Bond Strength Measurement and Process Schmidt, "A Silicon Wafer Bonding Approach for High­
Diagnostics," Proceedings of the First International Symposia Temperature Superconducting Bolometers", IEEE Journal of
on Semiconductor Wafer Bonding: Science, Technology, and Microelectromechanical Systems, December 1993.
Applications, Vol. 92-7, p.386, The Electrochemical Society, [25] M.A. Huff, A.D. Nikolich, and M.A. Schmidt, "Design of
1992. Sealed Cavity Microstructures Formed by Silicon Wafer
[13] T. Abe, M. Nakano, and T. Itoh, "Silicon Wafer Bonding Bonding", IEEE Journal of Microelectromechanical Systems,
Process Technology for SOI Structures," Proceedings of the Vol. 2, No. 2, 74-81, June 1993.
4th International Symposium on Silicon-on-Insulator [26] L. Parameswaran, V.M. McNeil, M.A. Huff, and M.A.
Technology and Devices, p.61, The Electrochemical Society, Schmidt, "Sealed-Cavity Microstructure using Wafer Bonding
1990. Technology", Technical Digest, Transducers '93, 7th
[14] W. Maszara, B-L. Jiang, A. Yamada, G.A. Rozgonyi, H. International Conference on Solid-State Sensors and
Baumgart, and A.J.R. de Kock, "Role of Surface Morphology Actuators, Yokohama, Japan, pp.274-277, June 1993.
in Wafer Bonding," Journal of Applied Physics, Vol. 69, No. [27] V.M. McNeil, S. S. Wang, and M. A. Schmidt, "Issues
1, p. 257, January 1991. Regarding the Application of the Electrochemical Etch-Stop
Technique to Fabricate Microstructures using Wafer
Bonding," Proceedings of the First International Symposia on
Semiconductor Wafer Bonding: Science, Technology, and
Applications, Vol. 92-7, p.180, The Electrochemical Society,
1992.

131

You might also like