Mikromechanische Ultraschallwandler aus Silizium [Elektronische Ressource] = Silicon micromachined ultrasonic transducers / vorgelegt von: Chenping Jia
125 Pages
English
Downloading requires you to have access to the YouScribe library
Learn all about the services we offer

Mikromechanische Ultraschallwandler aus Silizium [Elektronische Ressource] = Silicon micromachined ultrasonic transducers / vorgelegt von: Chenping Jia

-

Downloading requires you to have access to the YouScribe library
Learn all about the services we offer
125 Pages
English

Description

Mikromechanische Ultraschallwandler aus Silizium (Silicon Micromachined Ultrasonic Transducers) von der Fakultät für Maschinenbau der Technischen Universität Chemnitz Genehmigte Dissertationsschrift zur Erlangung des akademischen Grades des Doktor der Ingenieurwissenschaften (Dr.-Ing.) vorgelegt von: M. Eng., Chenping Jia Geboren am 13.05.1972 in Huoxian County, P. R. China eingereicht am 14.01.2005 Gutachter: Prof. Dr.-Ing. Arved Carl Hübler Prof. Dr.-Ing. habil. Thomas Gessner Prof. Dr.-Ing. habil. Bernd Michel Prof. Dr.-Ing. Hans Heinrich Gatzen Chemnitz, den 14.01.2005 Abstract Bibliographische Beschreibung Jia, Chenping Mikromechanische Ultraschallwandler aus Silizium – Silicon Micromachined Ultrasonic Transducers Dissertation an der Fakultät für Maschinenbau und Verfahrenstechnik der Technischen Universität Chemnitz, Institut für Print- und Medientechnik Chemnitz, 14.01.2005 Dissertation A S.: 125 Abb.: 112 Tab.: 5 Lit.: 88 Referat This paper discusses basic issues of micromachined ultrasonic transducers, including their design and fabrication. First, the acoustic fundamentals of ultrasonic transducers are introduced, and relevant simulation methods are illustrated. Following these topics, important aspects of silicon micromachining are presented. Based on this knowledge, two distinctive micromachining processes for transducer fabrication are proposed.

Subjects

Informations

Published by
Published 01 January 2005
Reads 13
Language English
Document size 17 MB

Exrait



Mikromechanische
Ultraschallwandler aus Silizium

(Silicon Micromachined
Ultrasonic Transducers)

von der Fakultät für Maschinenbau
der Technischen Universität Chemnitz

Genehmigte

Dissertationsschrift
zur Erlangung des akademischen Grades des
Doktor der Ingenieurwissenschaften
(Dr.-Ing.)

vorgelegt
von:
M. Eng., Chenping Jia
Geboren am 13.05.1972 in Huoxian County, P. R. China
eingereicht am 14.01.2005
Gutachter:
Prof. Dr.-Ing. Arved Carl Hübler
Prof. Dr.-Ing. habil. Thomas Gessner
Prof. Dr.-Ing. habil. Bernd Michel
Prof. Dr.-Ing. Hans Heinrich Gatzen

Chemnitz, den 14.01.2005
Abstract

Bibliographische Beschreibung

Jia, Chenping
Mikromechanische Ultraschallwandler aus Silizium –
Silicon Micromachined Ultrasonic Transducers

Dissertation an der Fakultät für Maschinenbau und Verfahrenstechnik
der Technischen Universität Chemnitz, Institut für Print- und Medientechnik

Chemnitz, 14.01.2005

Dissertation A
S.: 125 Abb.: 112 Tab.: 5 Lit.: 88

Referat
This paper discusses basic issues of micromachined ultrasonic transducers, including their
design and fabrication. First, the acoustic fundamentals of ultrasonic transducers are
introduced, and relevant simulation methods are illustrated. Following these topics, important
aspects of silicon micromachining are presented. Based on this knowledge, two distinctive
micromachining processes for transducer fabrication are proposed. One of them, the bulk
process, has been proved to be successful, whereas for the second one, a surface process,
some improvements are still needed. Besides these works, an innovative direct bonding
technology is also developed. This technology constitutes the basis of the bulk process. Of
course, it can also be used for the packaging of other MEMS devices.

Schlagwörter
Ultrasonic transducer, capacitive transducer, MEMS, micromachining, bulk micromachining,
surface micromachining, bonding, direct bonding, FEM analysis

2 Contents

CONTENTS

Bibliographische Beschreibung.............................................................................................. 2
Glossary of Symbols .............................................................................................................. 5
Foreword ................................................................................................................................6
1. Introduction ........................................................................................................................ 7
1.1 Ultrasonic Transducer – An Overview......................................................................... 7
1.2 Micro System Technology and cMUTs ....................................................................... 8
1.3 Current Research State................................................................................................. 9
1.4 Document Organization ............................................................................................. 10
2. Acoustic Fundamentals .................................................................................................... 11
2.1 Capacitive Ultrasonic Transducer .............................................................................. 11
2.2 Vibration Mode and Resonance Frequency ............................................................... 11
2.3 Mechanical Impedance of a Circular membrane........................................................ 13
2.4 Piston Source.............................................................................................................. 16
2.5 Array Line Source ...................................................................................................... 18
2.6 Reflection and Transmission...................................................................................... 20
2.7 Absorption and Attenuation in Medium..................................................................... 22
2.8 Canonical Equations and Equivalent Circuit ............................................................. 24
3. Transducer Performance Simulation................................................................................ 29
3.1 Finite Element Method............................................................................................... 29
3.2 Structure Static Analysis ............................................................................................ 30
3.3 Modal Analysis .......................................................................................................... 33
3.4 Electrical-Mechanical Couple Analysis..................................................................... 35
3.5 Acoustic Field Simulation.......................................................................................... 40
4. Silicon Micromachining................................................................................................... 45
4.1 Silicon Crystallography.............................................................................................. 45
4.2 Wet Etching of (100) Silicon Wafer .......................................................................... 47
4.3 Wet etching of (110) Wafer ....................................................................................... 49
4.4 Bulk Micromachining of (111) Silicon ...................................................................... 50
4.5 Dry Etching ................................................................................................................ 59
5. Wafer Bonding Techniques 63
5.1 Wafer Bonding and Device Packaging 63
5.2 Seal Glass Bonding .................................................................................................... 65
3 Contents
5.3 Low Temperature Direct Bonding ............................................................................. 74
5.4 Direct Bonding of Silicon and Glass.......................................................................... 83
5.5 Void Prevention Schemes .......................................................................................... 86
6. Bulk Micromachined Ultrasonic Transducer ................................................................... 87
6.1 Process Overview....................................................................................................... 87
6.2 Cavity Wafer Fabrication........................................................................................... 89
6.3 Membrane Wafer Preparation .................................................................................... 90
6.4 Bonding and Membrane Releasing ............................................................................ 94
6.5 Sample Quality Evaluation......................................................................................... 95
7. Fabrication of cMUTs by Surface Micromachining ........................................................ 99
7.1 Process Overview....................................................................................................... 99
7.2 Experiment Investigation 100
7.3 Failure Mechanism Analysis and Possible Solutions............................................... 106
8. Conclusion...................................................................................................................... 109
Reference............................................................................................................................ 110
List of Figures .................................................................................................................... 115
Selbständigkeitserklärung .................................................................................................. 119
THESES ............................................................................................................................. 123
Lebenslauf .......................................................................................................................... 125

4 Glossary of Symbols
Glossary of Symbols

a acceleration; absorption coefficient; v speed; velocity
radius of circular membrane V volume; voltage; effective voltage
b(θ,ϕ) beam pattern amplitude
w bandwidth c speed of sound
x independent variable C electrical capacitance; heat capacity
X electrical reactance d, d electrode gap; initial plate gap 0
X mechanical reactance mE Young’s modular
Xr radiation reactance f instantaneous force; frequency (Hz)
y transverse displacement F peak force amplitude
Z impedance (acoustic, electrical, H(θ,ϕ) directional factor
mechanical)
j imagine unit
Z mechanical impedance m
J Bessel function of order m m
Z radiation r
k wave number; spring stiffness
α angle variable
L inductance
β one-half of the damping coefficient
m mass
γ ratio of heat capacities
M microphone sensitivity
δ boundary layer thickness
p acoustic pressure
η coefficient of shear viscosity;
P peak acoustic pressure amplitude
efficiency
Pr Prandtl number
θ incidence angle; phase angle
q charge quantity
κ thermal conductivity
Q quality factor
λ wavelength
r characteristic acoustic impedance;
3ρ density, kg/mradius
ρ equilibrium density 0R resistance (acoustic, electrical,
mechanical); reflection coefficient 2ρ surface density, kg/ms
R mechanical resistance m σ Poisson’s ratio
R radiation resistance r τ relaxation time; pulse duration
s spring constant, condensation
Φ transformation factor; phase angle
S surface area
ω angular frequency, rad/s
t time
ω natural angular frequency 0
T period of motion; temperature;
Ψ solution of Helmholtz function; tension; transmission coefficient
displacement
u speed; velocity
5 Foreword
Foreword
This work is accomplished during my study at Zentrum für Mikrotechnologie (ZfM) in TU
Chemnitz; at the same time, I also take part in research works in the Micro Device and
Equipment Department of Fraunhofer Institute IZM. The excellent research conditions here
greatly facilitate the carrying out of this project, it is my pleasure to work in these two
institutes.
After three years of hard work, this manuscript is ready for print. At this moment, I want to
express my appreciations to Prof. Dr. -Ing. Arved Carl Hübler, and Prof. Dr. -Ing. Thomas
Gessner. As outstanding scientists and supervisors, they give me a lot of valuable instructions;
their encouragement and assistance constitute the basis of this research work. I really
appreciate their help.
I thank also Prof. Dr. -Ing. Bernd Michel (Fraunhofer Institute für Zuverlässigkeit und
Mikrointegration, FhG-IZM, Berlin), and Prof. Dr. -Ing. Han-Heinrich Gatzen (Institut für
Mikrotechnologie, Universität Hannover) for their evaluation of this research work.
Dr. Maik Wiemer is my colleague, throughout this investigation, he coordinates the
cooperation with other colleagues, and contributes a lot of professional advises to this project,
I appreciate his help very much.
During the experiment period, many colleagues took part in this project. It is just their
proficient skill and enthusiastic help, that makes this project so successful, I would like to
express my special thanks to:
• Dipl. -Ing. Norbert Zichner and Dipl. -Ing. Reinhard Müller for project management.
• Dipl. -Ing. Dagmar Zeiß, Dipl.-Chem. Isa Streiter, Dipl. -Ing. Helga Hesse (retired),
Dipl. -Ing. Ina Schubert, Dipl. -Ing. Ingrid Hempel, Dipl. -Ing. Gunther Schwenzer,
Dipl. -Ing. Jürgen Grunert, Dipl. -Ing. Thomas Werner, Dipl. -Ing. Rene Reich,
Dipl. -Ing. Mattias Küchler, Mr. Sven Uhlig, Mr. Matthias Plänitz, Mr. Lars Garbart,
and Mr. Patrick Schwarz for wafer processing.
• Dipl. -Ing. Iris Höbert and Dipl. -Ing. Monika Henker for SEM inspection.
• Dr. -Ing. Wolfgang Bräuer, Dr. -Ing. Jürgen Bräuer, and Dr. -Ing. Steffen Kurth for
mask preparation and sample evaluation.
• Dr. -Ing. habil. Karla Hiller, Dr.-Ing. Christian Kaufmann, Dipl. -Ing. Matthias
Hoffman, and Dipl. -Ing. Jörg Frömmel for helpful discussions.
• Prof. Dr. -Ing. habil. Thomas Otto, Dr. -Ing. Hermann Wolf, Dr. -Ing. Wolfgang
Seckel, Dr. -Ing. Reinhard Streiter, Dr. rer.nat. Erben Jens, Dr. -Ing. Knut Gottfried,
Dipl. wirt.-Ing. Mario Baum, and Dipl. -Ing. Jörg Nestler for kind help and supports.
As my persistent supporters, my parents taught me a lot of important knowledge. With their
diligent efforts, I get the opportunity to study in a formal school. Moreover, whenever I was
in difficulty, they lend me always their hands with no hesitate. I would like to express my
cordial thanks to them.
Finally, I want to thank those who had helped me, but did not give me theirs names. I would
never forget them, because their kindness originates from their warm heart. In my future life, I
would also give my hand to those who need help, but has nobody to rely upon, just as those
kind peoples had done. In this way, I hope our world will be more harmonic and happy.

6 1.1 Ultrasonic Transducer – An Overview
1. Introduction
1.1 Ultrasonic Transducer – An Overview
For quite a long time, people had wondered about bat, because this able creature can capture
moving insects in pitch darkness, while flying at full speed. At last, scientist found that bats
rely hardly on light. Instead, they emit ultrasound and detect signals reflected from objects.
Their biological pulse-echo location system is so elaborately developed that, they can
distinguish a moth from a falling leaf.
Ultrasound is not only useful to bat, in many industry and research fields, such as Non-
Destructive Evaluation (NDE), medical imaging, obstacle identification, acoustic microscopy
and so on, it plays also an important roll [1-2]. One attractive feature of ultrasound is that it
can provide inside information of the object without contact on it. Although optical remote
sensors provide similar capabilities even at much greater distance, their effectiveness depends
heavily on the optical property (transparency, roughness, orientation etc.) of the reflecting
surface [2]. Moreover, in harsh industrial environment, dust and vapor may also present
influences. On the other hand, optical measurement based on triangulation or laser Time-of-
Flight (TOF) analysis is also too expensive.
For these applications, ultrasonic sensors provide a low-cost whereas accurate enough
solution. As opposed to light, sound can penetrate almost any material, no matter it is
transparent or opaque. Depending on the frequency of the emitted wave, detection ranges can
span from some millimeters to several meters, within this distance, dust or other suspensions
have nearly no influence on sound propagation. By using transducer array, lateral details of
the object can be detected, without moving of the transducer (Phase Array Operation). What
-1 -1is more, due to the relative low sound velocity in medium (343 m⋅s in air, 1481 m⋅s in
water, 20 °C), long distance resolution can be achieved by simple electronic circuits.
According to the underlying physical principles, ultrasonic transducers can be classified as
electrostatic (capacitive), piezoelectric, magnetostrictive or pneumatic type. Currently,
piezoelectric and electrostatic transducers dominate the market. Transducers based on other
principles do not possess satisfactory dynamic properties, therefore are seldom utilized.
Piezoelectric devices are frequently used in modern ultrasonic transducer. In fact, it is just the
discovery of piezoelectric effect that leads to the invention of the first ultrasonic transducer
[3]. Piezoelectric transducers possess the advantages of compact, ruggedness and high
efficiency. For flaw detection in opaque object, they are the most favorable choice.
However, piezoelectric transducer has also drawbacks. First, piezoelectric materials’ acoustic
6 2impedance is in the order of 30×10 kg/m ⋅s, this magnitude is many orders larger than that of
2air (400 kg/m ⋅s), large impedance mismatch implies lower energy transform efficiency,
therefore, piezoelectric materials are inherently unsuitable for air-couple ultrasound
application. Besides impedance mismatch, there are still some other deficiencies. Because a
piezoelectric block’s operating frequency is determined by its geometry, the size and
frequency requirement of a certain system may not converge to a realizable configuration. On
the other hand, piezoelectric device’s geometry defines also its electrical impedance, so that
sensitive receiving electronics may be forced to operate under loads that worsen their noise
performance, this conflict is particular problematic for two-dimensional transducer matrices.
-4In addition, piezoceramic is limited to strain level of approximately 10 , which translates to a
surface displacement limit of about 0.5 µm in the low MHz range, for fluid applications, such
amplitude was proved to be insufficient [4]. Finally, most commercial available piezoelectric
7 1. Introduction
ceramics depole at relatively low temperature (e.g. 80°C), this prevents them from being used
in severe environments. Specialized ceramics that depole at higher temperature have lower
coupling constants and are in general very expensive.
The idea of capacitive ultrasonic transducer is as old as its piezoelectric counterpart. Compare
with piezoelectric transducers, capacitive devices have many prominent advantages.
Generally speaking, capacitive transducers have much wider bandwidth, a wide bandwidth
transducer does not simply increase the resolution, but also enables the design of new image
modalities and analysis principles [5]. Secondly, capacitive transducers comprising thin
vibrating membrane have or can be adjusted to have approximate acoustic impedance as fluid
medium. This latter property makes the realization of high efficient airborne or underwater
transducer possible.
However, till 1990s’, capacitive ultrasonic transducers had almost no commercial influence,
the reasons that impeded them from succeeding lies in their structure. Capacitive transducers,
as their name indicates, usually consist of two isolated parts: one membrane electrode for
ultrasound emission and a recessed substrate to act as counter electrode. In order to deform
4 2this thin membrane, area loads as large as kilogram per square centimeter (10 kg/m ) or
8electric field strengths on the order of million volts per centimeter (10 V/m) are required [4].
Before the development of silicon micromachining technology, both the fabrication of a stable
thin membrane and the realization of so intensive electric field were difficult, therefore,
piezoelectric transducers were much preferred than capacitive type.
1.2 Micro System Technology and cMUTs
Since 1970s’, people has attempted to fabricate mechanical structures through technologies
originally developed for semiconductor industry [6-8], these attempts were finally proved to
be successful. Today, silicon micromachining has evolved into a significant branch of modern
manufacturing industry. Micromachining technology has many unique advantages, for
example, devices fabricated through this method are usually no more than several millimeters
long, whereas their performance is no less than their traditional competitor. Decreased
dimension not only reduces cost and power consumption, but also improves the portability of
the complete system. Moreover, because the structures are fabricated through similar
procedures as microelectronic circuits, it is possible to integrate the mechanical and electronic
parts of a device in one chip, the benefits of monolithic integration to the reduction of
parasitic effect and the elimination of external disturbance are obviously apparent.
Silicon micromachining is particularly suitable for the fabrication of Capacitive
(Micromachined) Ultrasonic Transducers (cMUTs). The main features of this technology and
the essential requirements of this transducer conform so close to each other, that it looks like
it is especially designed for them: cMUTs are usually very small, which is determined by the
resonance frequency of the vibrating membrane, this geometry smallness complies with the
basic property of silicon micromachining. cMUTs requires a thin membrane no more than
several micrometers thick, for traditional mechanical processing, such dimension is almost
unimaginable, whereas for silicon micromachining, that is just the very common dimension
for most semiconductor materials. cMUTs prefers cluster operation to increase its radiation
power and to realize phase array control, the batch duplicate capability and individual
interconnecting potential of photolithography process provide such possibility. Not to
mention the dimension precision and shape accuracy supplied by this powerful technology,
consequently, cMUTs is considered as an attractive alternative to conventional piezoelectric
transducer.
8 1.3 Current Research State
1.3 Current Research State
During the past decades, many micromachining processes have been proposed for the
fabrication of cMUTs [9-19]. These processes can be divided into two catalogs: surface
process and bulk process. The first cMUTs was developed by Haller in 1996 [9], it was
fabricated through surface process. Although its performance is still unsatisfactory, this
transducer comprises all the essential components of a cMUTs: deposited Si N membrane, 3 4
thermal oxide insulation layer and conductive silicon substrate. In order to eliminate damping
effect caused by trapped air, and avoid the influence of dust and vapor, in an improved
version, etching holes were sealed through LP-CVD oxide or nitride [4]. If tight membrane
thickness tolerance is also required, a buffered lateral sealing process can prevent the sealing
species from entering the cavity [10]. Besides, accumulated sealing materials on the
membrane can also be removed through an additional lithography step.
Meanwhile, a different surface process was developed in Sandia National Laboratories [11-
15]. The main features of this process are double sacrifice layer design for quick membrane
releasing, and additional Chemical Mechanical Polishing (CMP) step for surface
planarization. Smooth surface finish improves not only lithography resolution, but also the
clamping state of the pre-stressed membrane. With this method, the stability of the membrane
can be enhanced.
Another surface process originally developed for the production of micro tubes can also be
utilized to fabricate cMUTs [16]. Benefit from the permeability of thin poly silicon layer,
sacrifice material beneath this porous skeleton can be removed without preparation of any
access holes. After sacrifice layer etching, tiny holes inside the skeleton are sealed through
LP-CVD nitride, so that the cavities under nitride membrane are under negative pressure.
Compare with other schemes, this method is much simpler. However, poly silicon’s porosity
must be carefully controlled so that both releasing and sealing process can be realized. In
addition, the selectivity between porous material and sacrifice layer should also be reasonably
selected to protect the porous material from over etching.
Besides these works, different membrane, sacrifice layer and substrate materials have also
been tested to examine the possibilities of cMUTs with various compositions [17-19]. In
general, surface micromachined cMUTs has the advantages of small electrodes gap and
bonding-free packaging procedure, therefore only low drive voltage is required. In surface
process, it is also possible to place all the electrodes on the same side of the substrate.
However, surface micromachined cMUTs cannot be made large, otherwise the time required
for sacrifice layer etching will be extremely long. Moreover, the poor uniformity of the
transducer cells may also influence its performance and introduce problems to array
operation.
As opposed to surface process, bulk micromachined devices have more accurate dimension
control. In most cases, they are also simpler and cheaper. However, as far as we know, there
is only a few bulk micromachined cMUTs. Some special requirements of cMUTs may lead to
the failure of this attractive process. First, the lateral dimension of ultrasonic element is
usually only tens or hundreds of micrometer, the vertical gap should also be less than 10
micron, within this distance, effective electrical connection for each element must be
established. For traditional bulk micromachining, handling of such dimension and
arrangement is still troublesome. Second, advantages brought by independent fabrication of
membrane and cavity components can only be realized with the assistance of reliable
bonding, yet existing bonding process can seldom fulfill these requirements. Finally, cMUTs
prefers operation with vacuum back cavity, whereas the importance of hermetic packaging
was always ignored. A bulk micromachined cMUTs using Silicon on Insulation (SOI) wafer
91. Introduction
may illustrate these contradiction [20], this process uses silicon fusion bonding (>1100 °C) to
join the wafers, and stock removal of a whole substrate to release the membrane, therefore,
the fabrication cost is greatly increased.
Normally, silicon (100) wafers will be used in a bulk process. Consequently, bulk
micromachined structures are always rectangular. For many other applications, such as micro
accelerometer or micro mirror, rectangular structure is preferred, because the analysis and
layout of such structures are much easier than other shapes. However, for cMUTs application,
the opposite is true. In this case, round or hexagonal structures are much desirable, for
structures with sharp corners induce stress concentration. However, in bulk micromachining,
by no means should a round structure be introduced, because such structure can never be
accurately fabricated through wet etching.
Fortunately, a (111) silicon bulk micromachining process provides opportunity of hexagon
fabrication [21-23]. In this process, dry and wet etching processes are smartly combined to
release thick freestanding structures. Benefits from the relative relation of the slow etching
planes in (111) silicon, after wet etching, the finally formed cavities are always hexagonal.
Although the shape and dimension of these hexagonal cavities are all satisfactory, it consists
only one half of the cMUTs, the realization of the final transducer still requires a stable
membrane and a reliable method to combine the two components, these requirements
motivate the implementation of this research work.
1.4 Document Organization
In this dissertation, two silicon micromachining processes are proposed for the fabrication of
cMUTs. One of them, the bulk process, has been proved successful, whereas for the second
one, a surface process, there is still some work to do. Besides, a newly developed low
temperature direct bonding technology is also presented, this technology is critical for the
realization of the first process. Of course it can also be used for the hermetic packaging of
many other Micro Electro and Mechanical (MEMS) devices.
In chapter 2, theoretic fundamentals of ultrasonic transducer are introduced. Particularly, the
working principles of array operation are illustrated in detail, because most cMUTs are
operated in that way.
Chapter 3 presents the simulated results of a hexagonal cMUTs, including its static, dynamic,
electro-mechanical couple and acoustic properties, for transducers that involve more than one
physical field, these results are invaluable.
Chapter 4 analysis the crystallography of silicon and its application in micromachining,
accompanied are some innovative structures fabricated in silicon.
As the herald of the bulk process, a low temperature direct bonding technology was developed
and evaluated. These works are described in chapter 5.
Fabrication results of the bulk and surface micromachined cMUTs are illustrated in chapter 6
and 7 respectively, for bulk process, preliminary evaluation results are presented; for surface
process, some improved methods are proposed.
Main conclusions of this work are listed in chapter 8.

10