The Relationship between Processing and Microstructure of
Tape Cast Green Sheets
by
Ming Liu
B. A., Physics
Hamilton College. 1992
Submitted to the Department of
Materials Science and Engineering
in Partial Fulfillment of the Requirements for the
Degree of
MASTER OF SCIENCE
in Materials Science and Engineering
at the
Massachusetts Institute of Technology
May 1994
1994 Massachusetts Institute of Technology
All rights reserved
Signature of Author ...
........
41. A
Departmen
.........
\
...............
aterials Science and Engineering
May 6 1994
Certifiedby .......
Accepted by ................
V
.........
..............
Michael J. Cima, Associate Professor of Ceramics
Department of Materials Science and Engineering
Thesis Advisor
I
"W
...........
Carl V. Thompson H
Professor of Electronic Materials
Chair, Departmental Committee on Graduate Students
science
AUG 1 8 1994
2
The Relationship between Processing and Microstructure of Tape Cast Green
Sheets
by
Ming Liu
Submitted to the Department of
Materials Science and Engineering on May 6 1994
in Partial Fufillment of the Requirements for the
Degree of
MASTER OF SCIENCE
in Materials Science and Engineering
Abstract
Tape casting is used to produce thin layers of ceramic-loaded polymers for
electronic packaging. his project focused on two aspects of tape casting:
(1) the measurement of plasticizer concentration across tape thickness, and
(2) the examination of drying behavior and green tape densities for three
different polymer to plasticizer ratios.
A high resolution method was developed to measure the ratio of plasticizer
(Benzyl Butyl Phthalate) to binder (Poly(vinyl butyral)) as a function of
position within ceramic green tapes. Diamond microtome sections, 07 trri
thick, were taken parallel to the surface of the tape.
Each section was
analyzed using Fourier transform infrared microspetroscopy (TIR-M
measure the absorption
intensity of unique absorption
to
bands for the
plasticizer and binder. 'Me plasticizer concentration as a function of depth
from the tape surface was determined by measuring the absorption intensity
of each band. The measurements are the first high resolution composition
measurements performed on tape cast sheets. We have discovered by
using this tool, that common polymer-plasticizer combinations forrn twophase mixtures. We have also demonstrated that solubility limits correlated
with the solubility parameter for the plasticizer.
The purpose of the second study was to find an optimal way to process the
green tape, so that dewetting and pores could be controlled.
he study
emphasized the relationship between the ratio of the binder to plasticizer
and the densities of cast tapes. We have made the first in situ density
measurements of a drying tape. There was less porosity in the green tape
with the higher polymer to plasticizer ratio. Dewetting and porosity could
be eliminated tinder slower drying rate.
Thesis Supervisor:
Dr. Michael J Cima
Title: Associate Professor of Ceramics
3
4
ACKNOWLEDGMENTS
First, I would like to express my gratitude to Professor Michael Cma.
I have benefited both professionally and personally from his unwavering
enthusiasm, high intellectual caliber, as well as his keen guidance and moral
support throughout the entire course of this work.
I also would like to express my appreciation to Dr. Wendell Rhine, who
provided me with numerous helpful suggestions, critiques and comments. My
thanks are also due to Lenny Rigione for setting up the drying bag, drying box
and the tape caster for me; John Centorino for providing me with various
technical assistance. I cannot thank Dr. Lunhan Peng enough for teaching me
FrIR; and Dr., Jiang Yue enough for teaching me microtoming.
thanks go ot
My special
to Dr. Yuying Tang for teaching me tape casting and
encouraging
e throughout this project; to Dr. Yaping Liu, Dr. Tae Sung,
Man Fai Ng and Bill Rowe for doing TEM for me.
'Me successful completion of the work would not have been possible
without some of my wonderful lab mates: Resourceful Man Fai Ng, smiling
Keigo Hirakata, fun loving Satbir KhanuJa, considerate Dr. Tae Sung,
courteous
Jae Yoo, cheerful Karina Rigby, cooperative Bill Rowe, athletic
Kevin Ressler, thoughtful Bertha Chang, humorous Dr. Neville Sonnenberg,
helpful Barbara Layne and many others. They have not only created a positive
working environment, but also provided me with tremendous amount of
assistance and support that I can never forget.
Finally, I would like to express my gratitude to all my friends and
family for moral support in all forrns. In particular I would like to thank my
parents for being there with me during all these years of my education; and my
boyfriend Yethun Goh for his loving and caring.
5
6
TABLE OF CONTENTS
TITLE PAGE
1
ABSTRACT
3
ACKNOWLEDGMENTS
5
LIST OF EXHIBITS
9
CHAPTER 1. INTRODUCTION
14
CHAPTER 2 BACKGROUND
18
2.1
2.2
Tape Casting
18
2.1.1
GreenTapeComponentsandTheirFunctionl8
2.1.2
Tape Casting Procedures
19
Drying and Density Measurement
20
2.2.1
Drying
20
2.2.2
Density Measurement
21
2.3
Microtoming
22
2.4
Phase Separation
23
2.4.1
Microscopic Techniques
23
2.4.2
Staining
24
2.4.3
DSC Technique
26
CHAPTER 3 PLASTICIZER DISTRIBUTION WITHIN CERAMIC
GREEN TAPES
37
3.1
Introduction
37
3.2
Experimental Procedures
38
3.3
Observations
40
3.4
Discussion
42
3.5
Conclusion
45
7
CHAPTER 4 DRYING AND DENSITY MEASUREMENTS
66
4.1
Introduction
66
4.2
Experimental Procedures
67
4.3
Observations
68
4.4
Discussion
69
4.5
Conclusion
7
CHAPTER 5. CONCLUSION AND FUTURE WORK
85
5.1
Major conclusions
85
5.2
Industrial Applications
86
5.3
Suggestions for Future Work
87
Bibliography
88
8
LIST OF FIGURES
Page
Figure 21
Schematic of the tape casting process (Ref. 5).
30
Figure 22
Two stage drying process (Ref. 5).
31
Figure 23
Schematic of the microtorning process (Ref 6.
32
Figure 24
Microtomed section adhering together in ribbon
fon-nation while floating on a liquid surface (Ref 6. 32
Figure 25
Osmimium tetroxide reacts with the carbon-carbon
double bonds (Ref 9.
Figure 26
33
Hydrazine or hydroxylamine reacts with polymer
containing acid and ester groups (Ref. 12).
Figure 2.7a
Phase contrast of a stained cryosection through
optical microscope showing dispersed phase particles
in a matrix (Ref. 14).
Figure 2.7b
34
Phase contrast through TEM of a stained
cryosection showing dispersed phase particles
in a matrix (Ref. 14).
Figure 28
33
34
DSC traces of polyisoprene/polybutadiene and
polybutadiene/polybutadiene immiscible blends exhibiting
two Tg's (Ref. 14).
Figure 31
35
Binder burnout TGA curves for thin sections cut from the
top and button of the tape (ref. 5).
48
Figure 32
The chemical structure of PVB.
49
Figure 33
'Me chemical structure of BBP.
49
Figure 34
Sample selection for nkrotoming.
50
Figure 35
IR spectra representative of PVB-BBP tape.
51
Figure 36
(calibration curve relating plasticizer concentration
and band ratio.
52
9
Page
Figure 37
IR spectra representative of Al,)03-PVB-BBP
53
tape.
Figure 3.8a
Plasticizer composition profile for
=
feet/min, with
54
Al')03-
Figure 3.8b
Plasticizer composition profile for
=
feet/min, with
54
Al')03-
Figure 39
Plasticizer composition profile for
= 200 feet/ min, with
55
A'203Figure 3 10
Plasticizer composition profile for green tape with binder B55
79.
Figure 311
Plasticizer composition profile for 40% (Vol.) BBP, without
56
A'203Figure 312
Plasticizer composition profile for 35% (Vol.) BBP, without
56
All03Figure 313
Plasticizer composition profile for 30% (Vol.) BBP, without
57
A'203Figure 314
TEM print Of
S04
stained PVB-BBP cross section.56
Figure 3.15a Plasticizer composition profile for Dioctyl
Phthalate.
58
Figure 3.15b Plasticizer composition profile for BBP.
59
Figure 3.15c Plasticizer composition profile for Dibutyl
Phthalate.
59
Figure 3.15d Plasticizer composition profile for Dimethyl
Phthalate.
60
Figure 3.16
Phase separation as a function of 8.
61
Figure 3.17
Phase separation during dying.
62
10
I 'age
Figure 3.18a
Phase separation during dying,
ft/min.
63
= 200 ft/min.
614
=
Figure 3.18b Phase separation during drying,
Figure 41
The "glass rod table" for tape casting (ref 3.
Figure 42
A schematic of in situ density measurement
(ref 4.
Figure 43
F4
A representative plot of green tape weight loss
during its drying for sample B.
Figure 44
77
In situ density measurement for samples B and C
from the "glass rod table".
Figure 47
78
Pore volume fraction as a function of solvent
remaining for sample C.
Figure 48
r6
A plot of thickness reduction vs. weight loss for
samples B and C.
Figure 46
F5
A schematic of the wet tape on the microscopic glass
substrate 10 minutes after casting.
Figure 45
F3
79
In situ density measurements for samples A, B and
C from the tape caster.
30
Figure 49
In situ density measurements for sample C.
Figure 4 10
The variation of the tensile strength of an Alumina
31
green sheet as a function of plasticizer concentration.
(ref. 8)
32
I 1
LIST OF TABLES
Page#
Table 21
Listing of microscopy techniques (Ref 9.
36
Table 31
The composition of green tape for plasticizer gradient
measurement.
Table 32
65
Solubility parameters for plasticizers used in this
experiment.
Table 41
65
The composition of green tape for insitu density
measurements.
83
Table 42
PVB-BBP tape densities.
83
Table 43
Final densities and drying times for samples A, B
and C
Table 44
84.
A summary of density and dewetting for sample
C under different drying rates.
12
84
13
CHAPTERI
INTRODUCTION
Tape casting has been used to produce thin layers of ceramic-loaded
polymers for multilayer ceramic (MLQ
packaging
1.
structures, such as electronic
Organic binder systems used in the electronic packaging are
critical to the final products, and they have to be completely removed prior to
the sintering.
The binder system of the ceramic tape has to provide the tape
with high tensile and yield strength, sufficient flexibility for handling and
machining, dimensional stability over time, and a uniform and reproducible
shrinkage during binder removal and sintering
2
A binder system which
satisfies these requirements should include a long chain polymer to provide
strength to the film, a plasticizer to increase the flexibility, and a deflocculant
to ensure a uniform dispersion of the ceramic powder. Proper selection of the
binder and plasticizer is very important for successful multilayer fabrication.
Many different types of binders have been used, but the most common ones for
electronic applications are acrylic and polyvinyl butryal resins.
Most tape-casting processes start with a milling procedure to
thoroughly mix the plasticizer, binders and ceramic powder.
The mixed slurry
is ready to be conditioned for the actual tape-casting process.
casting processes
Most tape
are based upon continuous casting machines where the
doctor blade is stationary and the casting surface is moving.
Dried tapes are
then cut to size, punched with via holes, screened with metallic paste and
laminated together to form multilayer greenware.[21
A major problem in the electronic packaging is the difficulty for precise
dimensional
control,
which is critical for
14
MLC manufacturing.
Most
projections
suggest
that dimensional
requirements
complexity and the size of the package increase.
will increase
Shrinkage variations within
0.2% linear and camber within 0. 0 I "fin. are often specified 3
is complicated
as the
by the use of many processing steps.
This variation
Other problems in
electronic packaging include retention of carbon in fired products, delamination
and defects such as voids and cracks.
Structural inhornogenity can also exist in tape cast ceramic sheets, such
as a density gradient across the tape thickness 4
The inhomogeneity in the
green sheets tends to be amplified during the binder removal and the firing
processes 5.
he resulting Inhornogeneous shrinkage creates stresses which
result in defect formation.
Ile origin of the above problems must be related to the microstructure
of the green tape. Few experiments have been reported to our knowledge that
describe the microstructure of the tape and its relationship to the processing
conditions.
Therefore, in this project, we will study the relationship between
processing and microstructure of tape cast green sheets. We will focus on two
problems: density of the green tape, and the composition gradient in the green
tape.
The first problem rises because most ceramic packaging is produced so
that the final ceramic structure is near full density
.
Thus, variation in
shrinkage is largely caused by variation in the component
green density.
Similarly, camber can be caused by variation in packing density through the
thickness of the green sheet. This study will enhance control of dimension and
camber.
The eason for the second study is that structural homogeneity
requires that a
the phases be uniformly distributed throughout the tape. We
discovered that blends of binder and plasticizer may be immiscible and exhibit a
two phase morphology. As a result, plasticizer distributes unevenly across the
15
green
tape
thickness.
Such
a phase
separation
will result
different
microstructure across the green tape thickness, and thus, complicate the MLC
manufacturing. The study of plasticizer gradient across the green tape win
help us to understand the cause of camber and delamination. This will help in
the development of new ceramic fori-ning operations and new binder systems.
This thesis starts with a literature review (Chapter 2 which includes
tape casting, drying, microton-ling and phase separation. Plasticizer distribution
across ceramic green sheets are described in Chapter 3
In this study, we win
introduce a method for measuring composition gradients on a scale that has
never be done before for green tapes.
We will show the discovery that
common polymer - plasticizer combinations fon-n two-phase mixtures. We will
also demonstrate the solubility limits correlate with the solubility parameter for
the plasticizer.
Chapter 4 is devoted to the relationship between the binder to
plasticizer ratio and the density of cast tapes. This is also the first time insitu
density measurements of a drying tape is made. The major conclusions of this
project are given in Chapter
5, followed
suggestions for future work.
16
by industrial
application
and
References
1.
R. Mistler, "Tape Casting: The Basic Process for Meeting the Needs of
the Electronic Industry", Ceramic Bulletin, Vol. 69, NO. 6 1990.
2.
Y. Y. Tang, Ph.D. Thesis, Department of Materials Science and
Engineering, Massachusetts Institute of Technology, 1994.
3.
M. J. Cirna, Y. Y. Tang and M. Liu, "Inhomogeneity and anisotropy of
Tape Cast Ceramic Films For Multilayer Structures", Ceramics
Processing Laboratory, Massachusetts Institute of Technology.
4.
Roosen, "Basic Requirement for Tape Casting of Ceramic Powders",
PP. 675-92 in Ceramic Transactions, Vol. 1, Ceramic Powder Science I
1, B/ Edited by G. L. Messing, E. R. Fuller and H. Hausner. American
Ceramic Society, Westerville, OH, 1988.
5.
T. Ueyarna and N. Kaneko, "Effect of Agglomerated Particles on
Properties of Ceramic Green Sheets", pp. 1451-58 in High Tech
Ceramics. Edited by P. Vincenzini. Elsevier, Amsterdam, Netherlands,
1987.
17
CHAPTER 2
BACKGROUND
In this chapter, we will review other studies on tape casting and drying.
Some common techniques used for analyzing the microstructures of polymer
blends will also be described.
2.1 Tape Casting
Tape casting has been used to produce thin layers of ceramic-loaded
polymers which can be stacked and aminated into multilayered structures.
Today, cast tapes find wide use in making multilayer chip capacitors and in
other electronic components, such as then-nistors, ferrites, inductors, and
piezoelectrics
1.1.
2.1.1 Green Tape Components and Their Functions
The organic
additives
in the tape-casting
dispersing agent, binder and plasticizer.
systems
are:
solvent,
Their functions are summarized as
follows:
Solvent: the solvent dissolves the other organic materials and
distributes
them unifon-nly throughout
the slurry.
It carries the ceramic
particles in an ordered dispersion. As solvent evaporates, the green sheet made
up of organic-ceramic composite on the substrate become dense.
9 Dispersing agent: The dispersant coats the ceramic particles
and keeps them in a stable suspension in the slurry due to steric hindrance and
electric repulsion.
With aluminum oxide and most other oxides it has been
found that Menhaden fish oil and glyceral trioleate are effective dispersants.
Hyatt
2
concluded that non-nally, as much as 3 n-d of dispersant may be used
for each 100 g of inorganics.
18
* Binder: the binder dissolves in the solvent and enhances its
viscosity to fon-n a slurry.
As the solvent evaporates,
the binder-coated
ceramic particles bond together to fon-n a strong dense green tape.
The
temporary binder is used to fon-n an adhering film around the inorganic
particles. Yan
3
suggested that Polyvinyl butyral (PVB) and various acrylic
polymers are used commonly for this purpose.
*Plasticizer: the plasticizer structurally expands the binder and
improves the distribution of the binder in the slurry. Properties of the binders
may be changed by selected use of a plasticizer.
It also makes the green tape
flexible. The common plasticizers include butyl benzyl phthalate (BBP), and
polyethylene glycol (PEG)
Roosen et al
3].
suggested that the non-optimized use of these organic
4
additives can have disadvantages, e.g. density gradients, low green strength,
poor shrinkage reproducibility, crack fon-nation during sintering, and physical
and chemical inhornogenities.
2.1.2 Tape Casting Procedures
Most tape-casting processes start with a milling procedure.
The first
stage of milling produces a low viscosity slurry by breaking down powder
agglomerates and uniformly distributing a dispersing agent in the solvent. In
the
second
stage,
the plasticizer
and
binder
are dissolved
into
the
solvent/ceramic slurry.
After the slurry has been thoroughly
n-dxed, it is ready to be
conditioned for the actual tape-casting process. The conditioning procedure
consists of de-airing by a vacuum. Chong et al
[51
concluded that standard
vacuums which are used are in the range of 635 to 710 mmHg. The viscosity
of the slurry usually range from 1000 to 5000 mPa for standard tape
formulation
6].
19
Most tape casting processes are based upon continuous casting
machines where, the doctor blade is stationary and the casting surface is
moving. For laboratory operations, batch casting with a mobile doctor blade
and stationary casting surface can be used.
Thickness control is a function of several parameters which can be
adjusted, including slurry viscosity, casting carrier speed, doctor-blade gap
setting, and reservoir depth behind the doctor blade. Chou et al
7
suggested
that the ratio of blade gap setting to final dried green tape is approximately 2 .
'Me tape can be either rolled onto a spool after drying for use in a rollto-roll process or stripped and cut into integral lengths for use in a subsequent
procedure. Figure 21 is a schematic of the tape casting process
81.
2.2 Drying and Density Measurement
2.2.1 Drying
The layer of slip dries slowly while it is carried through the machine on
the substrate.
Filtered air is blown through the machine in a direction opposite
to that of the moving slip. This arrangement allows dry air to come in contact
with dried tape at the exit end. Moreover, air passes through the length of cast
tape and gets saturated with solvent vapor before coming in contact with
freshly cast tape at the other end. The difference in solvent content between
the slip and the air controls the rate of solvent evaporation,
allowing the
remaining solvent in the slip layer to redistribute itself with only a small vertical
gradient of concentration.
This small gradient of concentration tends to
minimize the cracking of the slip as it dries.
A two-stage drying process of cast tapes is observed by Mistler et al [8].
The first stage of drying proceeds at a constant rate, and the second stage at a
gradually decreasing rate as shown in Figure 22.
In either stage of drying, the
solvent must move through three consecutive steps of transport: (1) solvent
20
flows vertically through the slip to the surface, 2) solvent then evaporates at
the surface, and 3) solvent vapor is swept away at the surface. In Step I the
slip is still fluid and solvent is easily transported through it by liquid diffusion
or capillary action.
The drying rate in Step 2 is limited by the inflow of heat
required to supply the latent heat of vaporization.
Step 3 can also be slow due
to the accumulation of nearly saturated vapor at the surface. A constant rate is
usually observed in Step .
2.2.2 Density Measurement
The density of unfired tape (often called the
green" density) is an
important parameter for characterizing a tape-casting process. The green tape
density measurement can be useful in detecting poor packing of the powder or
in detecting excessive polymeric binder content.
Mistler et al
(81
have developed a method to determine the bulk green
density. The volume measurement could be made with a flat-faced micrometer
caliper, provided about 15 thickness measurements were taken and averaged.
The same volume could be obtained by using mercury porosimeter.
Typicalbulk densities for the Alun-finatapes produced were 25 to 26
g/cc. The use of porosimeter reported some degree of permanent penetration
of the tape by the mercury, proving that there was some open porosity.
However, the weight gain due to Hg absorption could not be used as a
quantitative separation of porosity from compressibility because some mercury
flows back out of elastic porous materials when the applied pressure was
removed.
The weight gain is only a qualitative test, but nevertheless it is a
definite indication of unoccupied space in the original tape.
Karas et al 9 determined green density of cast BaT403tape from
geometrical measurements using a thickness gage accurate to 00025 mm on a
21
carefully cut 1-cm2 section. They also concluded that there was little effect of
binder to plasticizer ratio on the powder packing density.
2.3 Microtome
Urtramicrotomy is a technique of producing sections of materials thin
enough for examination under the electron microscope. It has been used
successfully to study all types of plant and animal tissue and other materials
such as metals, plastics, synthetic fibers and even samples from the moon.
Ile
production
of sections by ultran-licrotomy involves a type of
cutting action similar to that used in the machining of metals.
A prepared
specimen is moved past a cutting edge so that a thin layer of the material is
sliced off. Considerable friction is encountered as the section moves down the
sectioning facet of the blade, but this becomes almost negligible if water is
present at the knife edge. The slice then floats easily onto the liquid surface,
and will straighten out if it is allowed to float freely on the surface. In the ideal
situation, the section would be exactly the same size and shape as the face of
the specimen block from which it is cut. However, in practice the sliced piece
is not an entire cross section of the specimen block.
schematic of the microtoming process.
Figure 23
As shown in Figure 24
101 is a
101, as the
slices are produced they will adhere to one another along one edge forming
"ribbons".
Goodhew
111 suggested that in ultramicrotomy, as in both machining
and normal microtomy, variations of the knife angle and clearance angle win
affect the amount of stress introduced into the material and thus vary the
quality of the results. Other factors influencing the quality of the sections are
the speed of the cut, the temperature, and the sharpness of the cutting edge.
The production
of good sections is dependent
conditions for the material under observation.
22
on finding the optimum
2.4 Phase Separation
Blends of two or more polymers are often used to combine the desired
physical properties of each polymer to obtain an improved product.
polymers, when blended, may be thermodynamically
The
miscible, exhibiting a
single phase morphology, or the polymers may be immiscible and exhibit a two
phase morphology.
The most common instruments used to study the
multiphase polymer structures are: microscopic instruments and differential
scanning calorimeter (DSQ
121.
McBrierty
131 has
conducted studies by using
nuclear magnetic resonance (NMR) to test the multiphase behavior. In the
following sections, we will introduce these two most conu-non techniques.
2.4.1 Microscopic Techniques
There
are ranges
of general techniques
information relating to polymer microstructures.
commonly employed optical and electron
Table 21
141 with
used to provide
useful
A listing of the more
icroscopy techniques is shown in
type of features that are commonly imaged, the size range of
the structures and typical magnifications. Optical microscopy and transmission
electron microscopy (TEM) will be introduced in the following sections.
OpticalMicroscopy
Optical microscopy techniques are useful in providing a rapid overview
of structures in relation to the whole specimen. Phase contrast and Nomarski
techniques
provide
differentiation
in multiphase
polymers
where
small
differences in refractive index can provide information regarding the dispersed
phase size and distribution.
Sawyer et al
141
summarized the advantages of this technique: Optical
microscopy is shown to provide a rapid and informative overview.
Sample
preparation is minimum. Samples for optical study are not placed in a vacuum,
as with electron microscopy, and thus volatiles are not removed.
23
Beam
damage, which is common in electron
in an optical microscope.
icroscopy of polymers, does not occur
It also provides less chance of artifacts in the
preparation of a thin section for optical study than in the preparation of an
ultrathin section for TEM.
The limitations of optical
magnification
range
and
microscopy
a decreasing
depth
are the limited resolution,
of field with
increasing
magnification. Tadokoro [151 suggested that the resolution limit is on the order
of 02 gm; and the magnification limit is 2000x.
TEM Techniques
Fundamental
polymer
characterizations
generally
involve
the
application of TEM techniques. They offer the best image resolution of a of
the microscopy techniques and structural infon-nation relating to crystallinity
and crystallite sizes. Dispersed phase structures can often be observed and
quantified over size range from less than 10 nm up to I tm.
The disadvantages associated with the use of TEM are: high capital
expenditure,
time consuming in specimen preparation, and two dimensional
image interpretation. There are two reasons for the time consuming nature of
the specimen preparation for TEM: the specimens must be extremely thin (50
nm thick), and extra steps are often required to increase the contrast in
polymer specimens (section 24.2).
Most of the methods for producing
ultrathin sections are tedious such as ultramicrotomes with diamond knives for
sectioning (section 23). Image interpretation is difficult due to variation of the
specimen in the vacuum and under the electron beam, and artifacts caused by
sample preparations.
2.4.2 Staining
Image contrast in TEM is the result of variations in electron density
among the structures present.
Since most polymers are composed of low
24
atomic number elements, they exhibit little variation in electron density. As a
result, TEM micrographs of multiphase polymers often do not provide enough
contrast to image the phase clearly. One of the most common methods which
has proven useful in contrast enhancement is staining, by the addition of heavy
atoms to specific structures.
Staining involves the incorporation of electron dense atoms into the
polymer, in order to increase the density and thus enhance contrast.
Staining
of polymers can. be conducted either before or after microton-ling. The sample
is immersed in the stain solution or exposed to the vapor.
The most common staining method
staining, introduced by Andrews and Stubbs
synthetic rubbers.
is osr-nium tetroxide
161,
(S04)
who stained unsaturated
Such technique reveals the nature of the dispersed phase
domains of multiphase polymers. Osmium tetroxide reacts with the carboncarbon double bonds (Figure 25) in unsaturated rubber phases enhancing the
contrast in TEM by the increased electron scattering of the heavy metal in the
rubber compared to the unstained matrix. The reaction is slow, often taking
days to weeks. The high vapor pressure
Of
S04
is beneficial, making vapor
staining of sections viable; however, this vapor pressure, combines with the
toxicity of the stain, makes it very dangerous to use.
Several two step reactions have extended the range of OsO4 staining to
materials that cannot be stained directly.
For example, alkaline saponification
at boiling temperature followed by reaction of the hych-oxylgroups with
was used to study poly(vinyl chloride)/ethylene-vinyl acetate systems.
[171 has
S04
Kanig
developed a staining method for butyl acrylate rubber by treatment with
hydrazine or hydrozylanime and post staining with
SO4,
which works for
polymers containing acid and ester groups. The Os is deposited after the ester
is reduced by the hydrazine (Figure 26).
25
Exhibit 28 shows an example of microscopy techniques and
methods for polymers. Phase contrast optical
and TEM of a stained cryosection (Figure 27)
sO4
stain
icroscopy (Figure 2.7a)
141
141
all show dispersed phase
particles in a matrix.
2.4.3 DSC Technique:
The glass transaction temperature is the main characteristic temperature
of the amorphous solid and liquid states. A liquid becomes a solid on cooling
through the glass transition temperature.
The microscopic process involved is
the freezing of large-scale molecular motion without change in structure. The
glass-liquid transition occurs at a recognizable "transition temperature"
because of a rather large temperature dependence of the relaxation time for
large-scale molecular motion. The reported Tg is the temperature measured at
the onset of the glass transition. A blend with a single Tg is defined as miscible
and one with two Tg's was defined as being immiscible.
DSC is the most common instrument used to measure Tg of polymers.
Two types of DSC instrument have been widely used: the heat-flux DSC (e.g.
DuPont 910 DSQ and the power-compensational DSC (Perkin-Elmer and
Setarurn
11)
temperature
[181.
loop.
The theory of DSC can be explained with the differential
The signals representing
temperatures are fed to the differential-temperature
the sample and reference
amplifier via a comparator
circuit, which determines whether the reference or the sample temperature is
greater.
he differential-temperature amplifier output then supplies power to
the reference or sample heater as necessary to correct any temperature
difference between them.
A signal proportional to this differential power is
also transmitted to the pen of the recorder, giving a curve of differential power
versus time or temperature.
The area under a peak
is then directly
proportional to the thermal energy absorbed or liberated in the transition.
26
As an example of the above application, Massie et al 191conducted
DSC studies as shown in Figure 28.
and polybutadiene/Polybutadiene
Traces of polyisoprene/polybutadiene
immiscible blends exhibiting two Tg's.
27
References
R. Mistler, "Tape Casting: The Basic Process for Meeting the Needs of
the Electronic Industry", Am. Ceramic Soc. Bulletin, Vol. 69, NO. 6,
1990.
2.
E. Hyatt, "Making Thin, Flat Cerarnics-A Review", Am. Ceramic
Bulletin, Vol. 65, NO. 41986.
3.
M. F. Yan, "Microstructural Control in the Processing of Electronic
Ceramics,"
Mater. Sci. Eng., 48, 53-72 1981).
4.
A. Roosen, F. Hessel, H. Fischer, F. Aldinger, "Interaction of
Polyvinylbutyral with Alumina", Ceramic Powder Science, 1990.
5.
J. S. Chong, E. B. Christiansen, and A. D. Baer, "Rheology of
Concentrated Suspensions," Journal of Concentrated Suspensions,
15,2007-2021,
6.
1971).
E. S. Tormey, R. L. Pober, H. K. Bowen, and P. D. Calvert, "Tape
Casting--Future Development," pp 140-49 in advances in Ceramics,
Vol. 9 Forming of Ceramics, Edited by J. A. Mangels, and D. L.
Messing, American Ceramic Society, Columbus, Ohio, 1984.
7.
Y. T. Chou, Y.T. Ko, and M. F. Yan, "Fluid Flow Model for Ceramic
Tape Casting", J. Am. Ceram. Soc., 70 [10], C-280-C-282, 1987.
8.
R. Mistler, D. Shanefield, and R. Runk, "Tape Casting of Ceramics,"
pp 411-418, in Ceramic Processing Before Firing. Edited by G. Y.
Onoda and L. L. Hench. Wiley, New York, 1978.
9.
A. Karas, T. Kumagai, and R. Cannon, "Casting Behavior and Tensile
Strength of Cast BaTi4O3Tape", Advanced Ceramic materials, 34]
374-77 1988).
10.
Reid, "Ultramicrotomy", pp 217-223, 1980.
28
11.
Goodhew, P. F. 1972) Specimen preparation in materials science, in:
Practical methods in electron microscopy, A. M. Glauert, ed. NorthHolland, Amsterdam).
12.
T. K. Sherwood, "The Drying of Solids", Ind. Eng. Chem., 21 (1),
ppl2-16, 1929.
13.
V. J. McBrierty, "Heterogeneity in Polymers as Studied by Nuclear
Magnetic Resonance", Faraday Discussions of the Chemical Society,
NO. 68, 1979.
14.
L. Sawyer, D. T. Grubb, "Polymer Microscopy", Chapman and Hall,
London, New York.
15.
H. Tadokoro, "Structure of Crystalline Polymers", Wiley-Interscience,
New York, 1979.
16.
E. H. Andrews and J. M. Stubbs, J. R. Microsc. Soc. 82 pp 221, 1964.
17.
G. Kanig, Proc. Colloid Polym. Sci. 57 pp 176, 1975.
18.
Thermal Characterization of Polymeric Materials, edited by E. Turi,
Academic Press Inc.
19.
J. M. Massie, A. F. Halasa, R. N. Thudium, and C. W. Burkhart,
"Miscibility and Phase Behavior of Polyisoprene/polybutadiene and
Polybutadiene/Polybutadiene Blends", ANTEC'92.
29
SLIP
CARRIER
I FILM
IA
Figure 21
Schematic of the tape casting process (Ref. 5).
'A0
0
RUN
2 L PM
AIR FLOW
0
zW
-.1
0(A
U. 0
0
z0
P
U
4
(z
U.
0z (,
Z-
4
2W
Et
0
5
10
Is
DRYING TIME
Figure 22
20
25
30
MINUTES
Two stage drying process (Ref. 5).
31
'.14ier,81detacnea
4*
block
I
II
Act.al
.
"
-, r
,
secto.
size
, ,
Theoretcai section
e
LIQUID
Figure 23
Schematic of the microton-iing process (Ref 6.
spe
Wm
of ections
Figure 24
Microtomed section adhering together in ribbon formation while floating
on a liquid surface (Ref. 6).
32
H
H
I
I
-C C-C
=-I
H
+ 0 S 04
C-
I0
0\
I
H
0S
0/
_CC_
I
I
I
H
I
H
H
I
H
I
I
I
U\
/U
Os
0/ \0
1
I
H
Figure 25
1
-C
-C
I
H
Osmimium tetroxide reacts with the carbon-carbon double bonds (Re 9.
__T_
+ H2N-NH2
Pc\
0
0 DEC,\N
H1-11N H2
O-CH3
C30H
Figure 26
Hydrazine or hydroxylamine reacts with polymer containing acid and ester
groups (Ref. 12).
33
Figure 2.7a
Phase contrast of a stained cryosection through optical
microscope showing dispersed phase particles in a matrix
(Ref 14).
Figure 2.7b
Phase contrast through TEM of a stained cryosection.
showing dispersed phase particles in a matrix (Ref. 14).
34
I
01.
:i
Figure 28
DSC traces of polyisoprene/polybutadiene and
polybutadiene/polybutadiene
two Tg's (Ref. 14).
35
immiscible blends exhibiting
Type
Features
Size range
Mastnification
Macro-microstructures, color,
homogeneity
Spherulitic textures
Phase variants, refractive
index differences
I cm - 02 gm
IX-100OX
I cm - 0.2gm
I cm 0.2 gm
5OX-1200x
30%-1200x
I cm -5 nm
100 g.
0 nm
0.1 mm 03 nm
lOx-50000x
lox-1000OX
3000x-5000000x
10 gm -Inm
300x-300000x
Optical
Bright field
Polarized light
phase contrast
Conventional
electron
Scanning (SET)
Backscatter
Transmission
STEM
Tablell
Surface topography
Atomic number contrast
Internal morphology, lamellar
I Structures and crystallinity
Internal morphology and
I
crystallinity
Usting of microscopy techniques (ref 9)
36
CHAPTER 3
PLASTICIZER DISTRIBUTION WITHIN CERAMIC GREEN
TAPES
3.1. Introduction
Tape-casting is a well-known ceramic forming technology and has been
widely applied to the production
of thin ceramic sheets.
The primary
components ofthe ceramic sheets are polymer binder, organic plasticizer and
ceramic powder.
Blends of binder and plasticizer are often used to obtain
desired physical properties.
Low molecular weight diluents are often blended
with polymers to increase flexibility by shifting the glass transition to a lower
temperature.
'Me mixture, when blended, may be thermodynamically miscible,
exhibiting a single phase morphology, or the mixture may be immiscible and
exhibit a two phase morphology.
Differential Scanning Calorimeter Ill and
Nuclear Magnetic Resonance relaxation measurements
[21
have been used to
study the phase diagram of several blends which exhibited phase separation
behavior.
One related
study
3
in this field concerned
inhomogeneity
and
anisotropy of tape cast ceramic films for multilayer structures. A significant
result of this study was that shrinkage and camber during annealing was higher
in the tape casting direction than in the transverse direction. Another related
study
4
was on binder thermolysis on the diffusivities of dialkyl phthalates
(DAP) in plasticized binder. The results of this study demonstrated that the
distribution of binder and the kinetics of the removal process were intimately
coupled. The development of a penetrating porosity during the initial stages of
binder removal as a result of capillary redistribution could reduce the path
37
length over which species must diffuse to a length scale on the order of the
particle size.
Mistler et al. [51 showed binder burnout TGA curves for thin sections
cut from the top and bottom of the tape (Figure 3 ). They proposed that the
difference between the curves was due to segregation during
iring.
They
suggested that if segregation of the solids and organics took place as a result of
settling, there would be large difference in binder content from top to bottom.
Thus, they believed that the uniformity observed was a good measure of the
degree of deflocculation, even during the drying state.
No detailed study has been done on the inhomogeneity of composition
across the green tape thickness.
herefore, we have developed a method to
measure the ratio of plasticizer to binder as a function of position in green tape.
Diamond microtome sections, 07 [im thick, were taken parallel to the surface
of the tape.
Each section was analyzed using Fourier transform infrared
microspetroscopy (FI'IR-M) to measure the absorption intensity of unique
absorption bands for the plasticizer and binder. The plasticizer concentration
as a function of depth from the tape surface was determined by measuring the
absorption
intensity of each band.
The measurements
are the first high
resolution composition measurements performed on tape cast sheets.
Composition inhomogeneity across the tape thickness is evidence of polymer
plasticizer immiscibility for PVB-BBP systems.
3.2 Experimental Procedure
The binder (PVB) used in this study has a molecular weight range of
90,000 to 120,000g/mol (B-76
6
The chemical structure of PVB is shown
in Figure 32: The plasticizer (BBP) used in the study has the chemical
structure shown in Figure 33.
We chose
03
as the ceramic powder,
because A'203 has no absorption peak in the wavenumber region under study.
38
A doctor blade was used to cast A203-PVB-BBP
then dried.
flow rate =
tape, which was
Two drying rates used in this experiment were
feet/min), and
=
(linear air
= 200 (linear air flow rate = 200 feet/min). The
dried tape thickness varied from 130 trn to 150 tm.
The composition of the
green tape can be summarized in Table 3 .
A piece of green sample was first epoxied with Epofix "' for eight
hours. It was then polished to 02 mm in diameter for
icrotoming
A
diamond knife was used to microtome the sample parallel to the surface of the
tape from the top to the bottom (sample A). A second sample was taken from
the tape immediately next to sample A and was n-ticrotomed bottom to top
(sample B). The third sample was obtained 10 cm from the previous two and
was also microtomed (sample
thick.
(Figure 34).
Each thin slice was 07 tm
hese thin slices were collected on round Pyrex silver mirrors and
carefully recorded so that the measurements of each slide could be indexed
with position in the tape.
The sliced thin fUmswere examined at room temperature by MR
181
The silver mirror underneath the sample enhanced the signal to noise ratio of
the reflected IR beam.
he IR spectra of the films were collected in the
reflectance mode for 30 scans at a resolution of 4 cm-
The aperture was set
to a circular area of 10 [tm in diameter. These conditions remained the same
for the rest of the experiments. Plots of peak intensity vs. wavenumber were
obtained.
he relative band ratio of C=O (at wavenumber 1720 cm- I) to C-H
band (at wavenumber 2870- 1) was measured. Plasticizer concentration as a
function of position in green tape was plotted.
Binder PVB-B-76 was replaced by B-79 which has a lower molecular
weight (50,000 g/mol to 80,000 g.mol)
6].
According to the composition ratio
as shown in Table 3 1, Green tape with B-79, BBP and
39
All-03
was cast.
Several PVB-BBP films (without A1203) were cast at different
polymer to plasticizer ratios on silver mirrors for calibration purposes.
These
thin films were first spin dried and then placed in vacuum at the room
temperature to remove any remaining solvent. The final thickness of the PVB-
BBP films was approximately gm. These films were also examined at room
temperature by FTIR, and relative peak ratios of C=O/C-H were obtained. A
calibration curve relating the peak ratio and the plasticizer concentration was
plotted.
Three PVB-BBP films (with 40 vol. % BBP, 35% vol. BBP and 30%
vol. BBP) were also cast to thickness 80
m by using a doctor blade. These
films were also microtomed to 07 gm thin slices and were examined by FITR40 vol. % PVB-BBP film was cross section microtomed to 07
thick.
m
he thin slices were collected on the copper grids and the samples were
stained in
s0,j vapor for half hour.
plasticizer reacted with the
S04
The ester functional group in the
as shown in Figure 26.
As a result,
plasticizer density was increased and its contrast relative to the binder was
enhanced.
TEM micrographs
were obtained in order to study the top to
bottom microstructure variation of the sample.
Dioctyl Phthalate,
Dibutyl Phthalate
and Dimethyl Phthalate
with
different solubility parameters (defined in section 33) were used to replace
BBP in the green tapes.
These green tapes were also microtomed
and
examined under TIR. Plots of C=O/C-H ratio as functions of tape depth
were obtained.
3.3 Observations
Figure 35 is an IR spectra representative of the PVB-BBP tape. The
pertinent IR frequency range for those films is between 3000 to 1700 cm
The following absorbence bands are observed in this range: CH3 at 2960 cm-
40
CH2 at 2870 cm- I ,and C=O at 1720 cm -I . The band at 1720 cm - I is
present
only in films which contain BBP since the ester content of PVB is quite small.
The intensity of the 1720 m- I band is measured in the absorbency units and is
linear with respect to the concentration of the plasticizer. The band at 2870
cm-1 has contributions from both plasticizer and binder since both contain C-H
bonds. The relative band ratios of C=O (at wavenumber 1720 cm- I) to C-H
band (at wavenumber 2870 cm- I) represents the ratio of plasticizer to binder.
A calibration curve relating the plasticizer concentration
and band ratio is
shown in Figure 36.
An IR spectra representative of the thin sliced A203-PVB-BBP
shown in Figure 37.
The relative band ratios of CO
tape is
(at wavenumber 1720
cm- 1) to C-H band (at wavenumber 2870 cm- I) were measured again.
The
band ratios were converted to the plasticizer concentration according to the
calibration curve. Figure 38 a
b (
= ) show a substantial plasticizer
gradient giving evidence that the top surface of the tape was plasticizer rich.
Figure 39 (S =200) displays a random plasticizer distribution pattern through
the depth of the green tape. The data are reproducible from both "top to
bottom rnicrotorning" and "bottom to top microtoming" from samples A and
B. A larger degree of plasticizer distribution deviation was observed between
samples A and C. Plasticizer-rich regions (55 vol. %) and plasticizer - poor
regions 32 vol. %) were observed in all cases.
Figure 3 10 displays the plasticizer composition profile for binder PVB-
B-79 under the slower drying rate ( = ). Plasticizer gradient with the same
upper and lower boundaries was observed in the sample.
Figures 311
tapes without A203
concentration.
312 show plasticizer composition profile of PVB-BBP
powder at 40% and 35% by volume plasticizer
Plasticizer gradient with the upper (55 vol.%) and lower
41
32
vol.%) boundaries was also observed in these samples. In the 35% sample, the
plasticizer richregion was smaller than in the 40% sample. Figure 313 shows
the plasticizer composition profile of PVB-BBP tape with 30% by volume
plasticizer concentration. There is no plasticizer composition gradient across
the tape. The correct plasticizer composition level was also verified.
Figure 314 is a TEM micrograph of the cross section of PVB-BBP
film with 40% by volume plasticizer concentration. It shows larger plasticizer
domains at the top of the tape; and fine grained plasticizer at the bottom of the
tape. The dimensions of the domains varied from 0.8
m to 0 I gm.
Table 32 lists 8's for four plasticizers used in this experiment. Figures
3.15a, b, c and d are the plots of C=O/C-H ratio as functions of tape depth for
Dioctyl Phthalate, BBP, Dibutyl Phthalate and Dimethyl Phthalate green tapes
respectively.
Plasticizer gradient is also observed in these samples. As shown
in Figure 316, a larger plasticizer solubility parameter
leads to a smaller
difference in C=O/C-H ratios.
3.4 Discussion
One explanation for the plasticizer gradient is based on two component
diffusion and convection during drying 3
by evaporation.
The solvent leaves the top surface
The plasticizer flux through the top surface is negligible since
its vapor pressure is so low. The molar flux of plasticizer at any point, z, in the
tape can be written
i = XP (Js + J) -cDP ayaz)
where
Equation 31
p and J, are the molar flux of the plasticizer solvent, respectively,
P
and DP are the mole fraction and diffusivity of the plasticizer, and c is the molar
concentration of solvent and plasticizer.
'Me second term on the right
represents simple diffusion of the plasticizer toward regions of lower
concentration. The first term on the right, however, represents the convective
42
flux toward the surface of the tape which always has positive contribution of
solvent flux toward the surface. It is clear from Equation 32 that the solvent
flux convectively carries the plasticizer to the surface where diffusion must
carry the plasticizer back in to the interior of the tape. The diffusion is unable
to overcome the relatively rapid convective flux of solvent because Lewis et al.
141
have measured plasticizer diffusivities in polymers and shown them to be
rather small. Thus, the concentration of plasticizer must be highest at the
drying surface.
Due to the convective flow of solvent model, the concentration profile
should arise
such as a higher drying rate leads to higher plasticizer
concentration at the drying surface. However, Figure 39 (S =200) displays a
random plasticizer distribution pattern through the depth of the green tape.
This observation does not fit the convective flow of solvent model.
The concentration profiles shown in Figures 38-3.13 can thus only be
explained by te
existence of an immiscibility gap in the BBP-PVB phase
diagram. The 55 vol. % and 32 vol. % regions are the upper and lower phase
boundaries, respectively. Other plasticizer-polymer systems are known to have
a maximum solubility of plasticizer.
Studies of Poly (vinyl chloride) polymer
/di-isodecylphthalate (plasticizer) blends with nuclear magnetic resonance
(NMR) techniques show that plasticizer molecules can be m icroscopically
distributed into regions of either high plasticizer concentration or high polymer
concentration units
[21
. This technique was based on two characteristics of
NMR: the short-range nature of the dipole-dipole interaction and the transport
of spin energy.
The manufacturer
[61
of PVB suggests that the maximum
recommended BBP content is 40% by volume.
The BBP-PVB-solvent
solution is of course, initially single phase.
Immiscibility must occur at some point during drying. The situation is depicted
43
in Figure
317
where
the immiscibility gap
extends
to
an unknown
concentration of solvent.
Coarsening could be used to explain the variation in second phase size
across the tape thickness observed in Figure 314. Its driving force is the lower
surface energy of the large plasticizer domains compared with the high surface
energy of smaller plasticizer domains. 91The particle size grows at the expense
of smaller second phase regions. The top surface of the tape dries earlier than
the bottom surface. Therefore, phase separation occurs first at the top and the
plasticizers
there has more time to coarsen.
As a result, fine-grained
plasticizers increase in average size earlier at the top of the tape and the
plasticizer domain there becomes larger. The second explanation for Figure
3.14 is that larger plasticizer particles with lower density raise to the top under
buoyancy force (Figure 3.18a). Under faster drying condition (S=200), the
time difference in phase separation between the top and bottom of the tape is
minimized. Therefore, the morphology of the second phase under the fast
drying rate is depicted in Figure 3.18b.
The explanations above fit our observation such as under slower drying
rate (S=O), the top surface of the green tape was plasticizer rich (Figures 38);
and under faster drying rate, plasticizer-rich regions distributed more or less
uniformly throughout the tape (Figure 39).
The solubility parameter
= AE/V
is defined as:
Equation 32
12
where AE is the energy of vaporization to a gas at zero pressure, and V is the
molar volume. The dimensions of are (cal/cM3)1/2.
The relationship between
8 and phase separation (Figure 316) can be explained by using the theories of
thermodynamics of polymer solutions. 'Me n-fixingof plasticizer and binder
can be understood
as the interaction between the solvent molecules and
44
polymer monomers. The process of dissolving a polymer in a solvent is
governed by the free energy equation:
AF =A H -T AS
Equation 33
where AF = the change in Gibb's free energy, AH = the heat of mixing, T = the
absolute temperature and AS = the entropy of
ixing. A negative AF predicts
that a process wl occur spontaneously. Since the dissolution of a polymer is
always connected with a larger increase in entropy, the magnitude of the heat
term AH is the deciding factor in determining the sign of the free energy
change. Hilebrand and Scott 101proposed that
AH =
where V
M151-5212 VIV,
Equation 34
= total volume of the mixture;
1 and 8 are the solubility parameter
of the solvent and polymer; V, and V2 are the volume fraction of solvent and
polymer in the mixture. It may thus be seen that the unit heat of mixing of two
substances is dependent on [51-8,)]2. If the heat of mixing is not so large as to
prevent mixing, then [81-82]2 has to be relatively small. In fact, f 5 1-52]2 = ,
solution is assured by the entropy factor. This is mathematically equivalent to
saying that if the
values of two substances are nearly equal, the substances
will be miscible.
We were not able to calculate the solubility parameter
for PVB
monomer. However, the observation in Figure 316 shows that (3PVB
i larger
than
of the plasticizer.
As a result, least phase separation was observed in
the Dimethyl Phthalate (with the largest 5) green tape. The
correlation again
supports the idea that phase separation is occurring in the green tape.
3.5 Conclusion:
Inhomogeneous composition across the tape thickness was studied in
this chapter. For the first time, green tapes are found to be plasticizer-rich near
the top surface under slow drying conditions.
45
Plasticizer-rich regions are
distributed more uniformly across the tape thickness at faster drying rates.
Molecular weight of the polymer did not affect the
iscibility of binder and
plasticizer. This type of phase separation was also observed in other Phthalate
plasticizers.
We have discovered that solubility limits correlated
solubility parameter for the plasticizer. A larger plasticizer
degree of phase separation.
46
with the
leads to a smaller
References:
1.
J. Massie, A. Halasa, R. Thudium, and C Burkhart, "Miscibility and
Phase Behavior of Polyisoprene/Polybutadiene and
Polybutadiene/Polybutadiene Blends", Antec 92.
2.
V. J. McBrierty, "Heterogeneity in Polymers as Studied by Nuclear
Magnetic Resonance", Faraday Discussions of the Chemical Society,
No. 68, 1979.
3.
M.J. Cima, Y. Tang and M Liu, "Inhomogeneity and Anisotropy of
Tape Cast Ceramic films for Multilayer structures", Massachusetts
Institute of Technology.
4
J.A. Lewis and M.J. Cima, "Diffusivities of Dialkyl Phthalates in
Plasticized Poly(vinyl butyral) Impact on Binder Thermolysis" J. Am.
Ceram. Soc.,73 9] 2702-2707 1990).
5
R. E. Mistler, D. J. Shanefield, and R. B. Runk, "Tape Casting of
Ceramics"; pp.411-48 in Ceramic Processing Before Firing. Edited by
G. Y. Onoda and L. L. Hench. Wiley, New York, 1978.
6
Monsanto Plastics, St. Louis, MO.
7
Struers Company, Pennsylvania.
8
IBM System 9000.
9
Kingery, Bowen and Uhlmann, "Introduction to Ceramics", Wiley
Interscience, P 448-P 515
10
J. Hildebrand, R. Scott, "The solubility of Non-electrolytes",
Reinhold Publishing Crop., N. Y., 1949.
47
3rd ed.,
100 -
%%1307 TOM
'<Il
x
90
I
(D
w
3:
so
I---
0
100
----
I
- W--
200
TEMPERATURE,
Figure 3.1
I
I
1
300
400
50C
C
Binder burnout TGA curves for thin sections cut from the
top and button of the tape (ref. 5).
48
H
-Cll
(
H
-C
C
0
0
C
H
Figure 32
CA
The chemical structure of PVB.
-O(CH)3CI-I3
-- OCHi--
0
Figure 33
The chemical structure of BBP.
49
tb
.E
C
p
U
.E
49
0
u
1)
V
cn
.2Cl.
r:s
cn
Cri
2
-p
8
0
50
-:t
C2
C)
a
11
z
x
x
i
5
Tr,
N
cic
w
CD m
i
C2
zw
>
It W m
<
am
at
I
I
I
x
I
11 I
9C2
f
C3
ts
C;
1: CNI
C;
0
a
33NMIOSSY
51
a
0
C3,
I
-1
1-
Cl)
"I
X:"
m
"I
0
U.
0.
0
11-10.
0
m
6.
4-
a)
Q
0.
0
U
(I.)
.N 0.
5
C4
Ct
n-
0.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
The Ratio of C=O to C-H peaks
Figure 36
Calibration curve relating plasticizer concentration and band ratio.
2
di
V
0.
m
Q
C3
C3
4
k
0ll
,L)041
VI
I
.Z
5
4)
ll
E
ea
1=
ud.)
Q
CD
FM
5
-*
TS
I
P
4
bi
t-
rn
a
1.
I
E
u
CY
w
R
A
I
ga ai
1:
9
K
:D
zw
iSt
V,.IC,
- - - - WI9
mZ;
at:
IL
-C 1%
r_ 4)
C
bf
j
1-
a
ICI
A0
CD
C)
C2
FC.)
X:
-i
F- (1)
X: (14
1-1 (1)
ile
3ONYSaOSSV
X "
U
X:
W
C)
Cl
w
11
U.(/-I
I-
- -70 -
10
C
.0.5
C
0.4-
6-
6a IL
-... v2
a a
-I
K
NO a
a I-W x
PIN
>
I.-I
3
-
Al
It.,R..
E
a
I
-WW
a
:= 0. 'IN
U
U
-
0
(j
$-.
U
N
0.2
'U
CA 0.
.2
1
C-,
I
.
0-
I
'
'
.
20
0
t
I
i
.
I
A-i
.
.
.
i
,
I
I-T
-
I
$
i
.
I
I
I
I
40 60 80 100 IZO 140 160
Tape Depth (Micron)
_
U-07I
I.-I
I
0
C-'
--0; 0. 5
5 0.4-
or
n
U
Q 0.3-
0
C.)
U. 0.2
(U
I
.N
Top to Bottom Cutting, Point A.
a
Cn*0.1-
I- Bottom to Top Cutting, Point C.
E:
0-
iI
0
.
I
I
20
i
.
40
I
I
I
60
I
I
I
I
I
I
I
I
-
80 100 120 140 160
Tape Depth (Micron)
Figure 3.8a
Plasticizer composition profile for
=
feet/min, with
A1203-
Figure 3.8b
Plasticizer composition profile for
A120354
= feet/min, with
0.
U
7 -7750,
0.5-
:M
::X
PI
I
a 0.4-
a a7-
5
U
U
Z
._%WMQL-7-:
ZKKAOU
Wn-
t)
Q 0.3-
a
WE
0.2,
a
RON
-
Top to Bottom Cuttin2, Point A.
:: Bottom to Top Cutting, Point CI
2a. 0.1
Cn
I
- 7 .111,7 1- 7 - 1
0
0
20
40
60
80
i''.1,
i
I
100 120 140 160
Tam Devth (Micron)
0
0
m0
2- L IF so as
or
0
CZ
Ce.i.5 -
0
0
a
m
0 1-
a
11
U
0.50- I I I
0
20
Figure 39
I I
0
ME
a a
m
80% lipa
a
0
NJ
a
-I
i
A-, -,
v
ft
I I
on
OmIn
no
a
I I I I I
40 60
100 120 140 160
Tape Depth (Micron)
Plasticizer composition profile for
= 200 feet/ min, with
A1203-
Figure 3 0
Plasticizer composition profile for green tape with binder 79.
55
Ml
I-
C
Ca
a)
Q
a
U
6(U
ZN
C4
.E2
a-,
20
0
40
60
80
100
i2O
Tape Depth (Micron)
01
-a
C
0
U-
C
0U
0
U
U.
0
.5N
-O.W
M
C-
0
Figure 311
10
20 30 40 50 60 70 80 90 100
Tape Depth (Micron)
Plasticizer composition profile for 40% (Vol.) BBP, without
A1203-
Figure 312
Plasticizer composition profile for 35% (Vol.) BBP, without
Al')0356
0.5" 0.
-
ON
"0 0.45- I-I
-
I-
3 0.4Z
55
U
-
0. 3 5
0
-
-
L-M
e no
a.)
N
-
.5
Cn
CZ
-
0.25-
n-
x
0 no
5
0
0
N
M
0
0
0
N
0
a
a
-
-
0.2
I .
0
Figure 313
I
10
I I
I I
20
T
Ii ,
30 40
I
I
I I I I I I
50
60
I
I I
70
.1111.1.1
8
90
I
100
Plasticizer composition profile for 30% (Vol.) BBP, without
Al')03-
57
0
I
0
t%
P.
2
E
.1
's
0
'*O
I
eni
L
91
P.
9
d
Ak
A
A
t; 9
11
114
58
-7
w &O v
1.
0
-I
h
am
0
0
1.6 - 0 0 Ip so
Now O a
N on a on
1.4no &
a
CZ
CIO-,
1.2-
a so
I
a
0
O 0.811
U
0.6.
lisp
no
N
0.4-
a
WAS
0.27
0
I
0-.
0
I ' --
20
i -1
I
-' -'-F-7- i
d
I
M
i
,
I
-
a
-
P
I
I I
120 140 160
40 60
Tape Depth (Micron)
100
2.5
2-
0 I-on
OR mm oil
L
a
a
0
0
v
a
a
0
0
4k %
N N
W% v
E
ae. 1.5-
a
a II
N
a
0
0
0
1
0
0
on
11
N
wI
160
U
an 0
mmAN
. fto
0.5. . . I. .-
0
20
I
I
40
60
I I I I I I.
.. .I
I
.
80 10
120
Tape Depth (Micron)
.
.
I
.
140
I
160
Figure 3.15a Plasticizer composition profile for Dioctyl Phthalate.
Figure 3.15b
Plasticizer composition profile for BBP.
9
II
0.9
E
0
0.87
0E
0
0 0.7- - 50 ON so
0
P
CF" 06
a
onU
-
2
a
U
son
a
C 0.57
0
0
a
a
N
C: 04
E
0a
11
II noun
41
IN
U 0.37
of &Mal
0
v OMEN
0.20.1
I I
D
I
I
20
I
I
I I I I
40
I I
60
80
I I I
100
120
Tape Depth (Micron)
0
CZ
01-1
C
11
U
0
10
20 30
40
50
60
70
80 90 IC9
Tape Depth (Micron)
Figure 3.15c Plasticizer composition profile for Dibutyl Phthalate.
Figure 3.15d
Plasticizer composition profile for Dimethyl Phthalate.
60
1.1
Q
Q
I.,
12 L .
IU
a0
M 0.
C-,
*1 0.
U
Ii
U
0.
0.
0.02 0025
003
0035
004
0.0,15 005
Plasticizer Solubility Parameter
Figure 316
Phase separation as a function of .
61
Cj
z
Cfu
.
F
Ei-.
rl
e
11-1
4
p
a
E
O
4-
0
C
W
u
aE
(1)
0
C.(
q
d
914
CLO
0
8
pq
pq
10
.
V)
4-4
0
.40 0rN
tR e
Q
0144,
AO
tR
a0
1-0
C:
0
I=
a
0
A4 94
P
7; Q P
,4-.4zft
e%)
tT n
62
,I=
u
"-o
T*
la
0
"-4
"a
...I.
/5
j
.1
0
n
I
.9
ba
vW2
0
V4
*.b
CA
M
U
co
F--o
04
cn
0
m
t"4)
;g
q4
,= n
4-6
Om$
W
W-4
I
9
I
a
at
co
1-4
0114 9LO
(Vi
L
X
63
i
A
ITZ
i
i
4u
.
6
(1)
v
M
_-4-.4
iI
;1.0
\\",
,4
9L4
914
10
(1)
...
P--o
"-O
el
5
C14
A
93
A
1=
.9
W
0
Q
A
*.b
116.4
=
Cn V.-.4
CZ
>. a)
=4
.,.4
. .o n
,0-4
-4
0=
CL-4
4
9
la
0
am
4)
U *n
.1
op"
rA
I
2
co
el-;
L
X
64
Binder
Plasticizer
Powder
Dispersant,
Solvent
PVB
BBP
. A120
Fish Oil
Toluene
Poly vinyl butyral)
Benzyl Butyl Phthalate
Particle size: 02-0.5
Ethyl Alcohol
Table 31
7.110 g
4.309mi
43.12g
0.04g
5 ml
I
11MI
The composition of green tape for plasticizer gradient measurement.
Plasticizer
Solubility Parameter
(cal/cc)112
DiocIvI Phthalate
0.021
Butyl Benzyl Phthalate
Dibutyl Phthalate
Dimethvi Phthalate
0.027
0.032
0.049
Table 32 Solubility parameters for plasticizers used in this experiment
65
CHAPTER 4
DRYING AND DENSITY MEASUREMENTS
4.1 Introduction
Ceramic green tape is generally considered a composite of ceramic
powder, polymer and plasticizer. In reality, however, gas filled pores exist in
the green tape. The production of multi-layer ceramic devices requires specific
structural properties of the green tapes, such as no density gradient.
To
achieve such an uniformity, the optimization of slurry composition is crucial.
Fiori and DePortu III suggested some rules for the preparation of a
tape-casting slurry. These rules are: (1) the plasticizer to binder ratio must be
less than 2 2) the amount of dispersant must be chosen from within the range
where the adsorption is constant,
3 the ratio between organic components
and ceramic powder must be as low as possible, and 4) the amount of solvent
must be fixed at the minimum to ensure a good dissolution of organic
components and a good homogenization
for the slurry. An optimized slurry
leads to tapes which satisfy the following criteria: no cracking during drying, a
good cohesion to allow the manipulation of the dried sheets, and good
microstructure.
The drying behaviors of aluminum nitride tape-cast sheets was recently
studied.
2
A high apparent density, a low volume of pores, and a high
resistance to cracking require a low ratio of binder to plasticizer.
The purpose of the current study is to find an optimal way to process
the green tape, so that dewetting and pores could be controlled. The study
emphasized on the relationship between the ratio of binder to plasticizer and
the green tape density.
Air flow rates, drying time and tape dewetting were
also examined.
66
4.2 Experimental Procedures
'Me composition
summarized in Table 4 .
of the slurry used in this experiment
A203-PMMA-BBP
drying. The "glass rod table" (Figure 4 )
can be
slurry was cast followed by
E31 was
piece of microscope glass served as the substrate.
used for tape casting. A
The thickness of the tape
could be adjusted by the height between the top and bottom plates with the
height adjustment screws. For more accuracy, a tape caster was then used for
tape casting. A six by eight inch glass plate was the substrate.
Samples A,
,
and C were produced by using these two techniques.
Immediately after casting, the wet tape was placed under a microscope,
so that its thickness was measured through the drying. We chose the uniform
central square region for this thickness measurement,
due to the small
deviation in thickness (5 trn difference across the square region). At the same
time, a balance was used to weigh the tape during its drying (Figure 42
""
As a result, the intermediate densities of the tape, its thickness reduction and
its pore volume fraction were calculated from its final density, thickness and
weight during drying. This whole setting was enclosed in a Polyethylene bag,
where N, flew through.
We have developed three methods to measure the final green tape
densities.
(1 a piece
the center region.
measured.
Of
_CM2
square shaped green tape sample was cut from
Its thickness (with a micrometer) and its weight were
2 to obtain a more accurate area measurement, this piece of green
tape sample was photographed and then its print was enlarged. We calculated
the area of the green tape sample using the enlargement ratio and the area of
the larger print.
3 Mercury porosimeter
tape densities.
67
[51
was also used to verify the green
The theoretical density of the green tape was calculated assuming that
there was no porosity remaining:
Densitymeoy= WtPwdr + WtBinder
+ WtP1atiied/(VPowder
+ Vp1asticized
+ VBinder
Equation 41
Three different drying rates 6 S =0 feet/niin; S =100 feet/min; and
=200 feet/min) were used for sample C. Its final densities under these rates
were also measured.
4.3 Observations
Figure 43 is a representative plot of green tape weight loss during its
drying for sample B. 80% of the solvent weight loss occurred during the first
quarter of the drying time. The last 7% of the solvent was removed during the
second half of the drying time.
Figure 44 is a schematic of the wet tape on the
substrate 10
icroscopic glass
inutes after casting. Dewetting occurred when 50% of the
solvent was removed.
Figure 45 is a plot of thickness reduction vs. weight loss for samples B
and C. 'Me initial sample thickness was 500 tm.
Thickness reduction was
larger for sample B than that of sample C for the same amount of solvent
weight loss. The data were reliable due to the high degree of reproducibility.
Table 42 includes the density measurements of PVB-BBP tape with
35% (vol.) and 30% (vol.) plasticizer concentration.
There is 06% difference
in their experimental densities.
Table 43 summarizes the final densities and drying times for these
three samples.
was the shortest.
Sample C contained the most plasticizer, and its drying time
The density trend was consistent with Figure 45 such that
the higher the polymer concentration
of the sample, the less porosity it had.
The density difference between sample C and sample A is 66%.
68
Table 44 summarizes the density and dewetting results for sample C.
The three density measurement techniques provided a consistent trend such
that a faster drying rate resulted in less density in the final green tape. Under
= 200 feet/min, final green tape density was 8% less than that under stagnant
air. On the other hand, the faster the drying rate, the more dewetting occurred
in the green tape.
Figure 46 is a plot of tape densities during its drying. The densities of
sample
matched well with the theory. The densities of sample C started to
deviate from the theory when 30% (weight) of the solvent remained. The
results were reproducible,
for three repeated experiments.
As 30% of the
solvent remained in the tape, the volume fraction of the organic component
was 72.4%; the composition of the organic phase (by weight) was: PMMA
23.5%, BBP =21.5% and solvent = 55%.
Figure 47 shows the pore volume fraction as a function of the solvent
remaining for sample C. Again, when 30% of the solvent remained, porosity
increased to 22% by volume and stayed at this level through drying.
Dewetting still occurred at the edge of the tape as we switched to use
the tape caster.
However, the area of the glass plate was big enough so that
the central square region was assumed unaffected. Figure 48 summarizes the
in situ density measurements for samples A,
and C. This again demonstrated
that sample A correlated the most with the theory; sample C deviated most
from the theory.
Figure 49 summarizes three repeated in situ density
measurements for sample C. The reproducibility also verified the above result.
4.4 Discussion
Dewetting occurred 10 minutes after casting.
This is an indication of
the slurry particles' movement during the early stage of the drying. After 50%
of the solvent is removed, slurry particles are stationed, due to their high
69
viscosity. Because of the small area of the microscope glass, the central square
region could be affected by such a particle movement. As a result, the sample
thickness and weight data are only considered after the removal of 50% of the
solvent.
High surface tension is evidence of dewetting.
The surface energy of
the substrate must be greater than the sum of the surface tension of the slurry
and the substrate for a slurry to spread 9
If the surface tension of the slurry
was too high, the slurry would not spread evenly. A high drying rate caused an
inhornogeneity in solvent distribution, which resulted surface tension gradient
across the tape length. The retraction of the tape (or dewetting) occurred,
when it experienced such a surface tension gradient in different directions.
The observations in Figures 45-4.9
indicate that a low polymer to
plasticizer ratio resulted in low green tape density. Such a reduction in green
tape density 66%) is not caused by the addition of plasticizer (Density
reduced 06% for 5% vol. plasticizer concentration increase.). The effect of
plasticizer on binder was the cause for the decrease in density.
molecular-weight
The lower-
organic specie plasticizer was an additive that intimately
mixed with the binder as a single material. The plasticizer disrupted the close
aligning and bonding of the binder molecules, thereby increasing the flexibility
of the material. The plasticizer tended to reduce the strength of binder while
softening it 7]. Figure 410
shows the variation of the tensile strength of an
Alumina green sheet as a function of plasticizer concentration, which proves
that the plasticizer additions significantly decreased the tape strength.
The vaporization
of the solvent left vacancies
inside the slurry.
Ceramic particles which embedded in the binder matrix, rearranged themselves
to fffl up these vacancies.
A binder matrix with excessive plasticizer (sample
Q could not provide the particles enough strength. As a result, the particle
70
rearrangement was slower than the solvent vaporization and porosity were
created.
Under a faster drying rate, the particle rearrangement was slow
enough, so that porosity could be created as well.
4.5. Conclusion
We have made the first in situ density measurements of a drying tape.
The drying behavior and densities of tapes with three different polymer to
plasticizer ratios were studied. Green tapes with different drying rates were
also examined.
80% of the solvent weight loss occurred during the first quarter of the
drying time. The "glass rod table" casting method was not an effective way for
tape casting, because the microscope glass area was small enough so that
dewetting at the tape edge affected the central region.
Figures 45 to 49
indicated that the higher polymer to plasticizer ratio resulted in high green tape
density.
0.02)
Green tape A PMMA:BBP = 18:1) had a final density = 2.57 ±
g/CM3,
which was almost consistent with the theory 2.56
g/CM3).
Its
densities during drying also demonstrated the most correlation with the theory.
Green tape C (PMMA :BBP = 11) deviated most from the theory, with 22%
by volume porosity remaining as it dried. Its final density was 2.40 ± 002)
g/CM3.
For the same initial thickness, the drying time for sample C was 16%
less than that of sample A. Dewetting and low final density occurred when a
tape was dried under the faster drying rate (S = 200 feet/min).
71
References
1.
C. Fiori and G. DePortu, Br. Ceram. Proc. 38 1986) 213.
2.
E. Streicher, T. Chartier, "Study of cracking and microstructural.
evolution during drying of tape-cast aluminum nitride sheets",
Chapman and Hall Ltd. 1991.
3.
Manufactured and drawn by Bill Rowe.
4.
Drawn by Sam Gido.
5.
Micromeritics Autopore 9220.
6.
Tape caster at BP Research.
7.
G. Y. Onoda, Jr., "The Rheology of Organic Binder Solutions" in
Ceramic Processing Before Firing, Wiley, New York.
8.
R. Moreno, "The Role of Slip Additives in Tape Casting Technology,
Part 2-Binders and Plasticizers", Ceramic Bulletin, Vol. 71, No. I ,
Nov. 1992.
9.
"Avoid Coating and Drying Defects", Chemical Engineering Progress,
Jan. 1993.
72
ht
atment
";,V's
T-T-1.
Adjustment
Screws
Figure 41
'Me "glass rod table" for tape casting (ref 3.
73
Camera
Optical
1croscope
Balance
Figure 42
A schematic of in situ density measurement (ref 4.
74
0.)
C.
E
'rCl.
LI-C
0
tb
-E
22
l
lc
in
t1i
0Slo
r-
c
.2
W.
IRT
c
c .r_
c-4
E
W
c
E
v
u
zEa
ICZ
c00 c?,
c\C
.p
01.)
t-b
z
C-1
u
ZZ:
C
,,I-t,
cu
V.
IU
Cl.
u
c
SSO-1
IPZA
75
te
'5
9
17
-S
r
tr
r.4
U
.5:-
5
4
2
C
u
E
t
d
tri
I
d
d
V.
<
"IT
u
EL
76
u
1=
I'll,
,C
d"
0
4U
Irl
0
P4
U)
EA
4)
AC-)
10
20
30
40
50
Weight "ss
Figure 45
60
70
(%)
A plot of thickness reduction vs. weight loss for samples
and C.
77
80
90
100
-3 T530jurT-3)8pmPMNMIJBP--l.5:1
i
f
2.8 -
T =529pa Tf--3W pm P MMA.BB P--1.51
T;=53'7pmT
2.6-1
209pmPMMA:BBP--131
T530 pm Tf=;248pm PMMA:BBP=1I
2.4-
T.=533pmTf7-239pm P"M:BBP-11
It
I--,
M
E
_6 2.2-
*0 0
T,=535pmTf=;241pm PMNIA.BBP=lI
b
i--
.;;
z0
-'nwcre&:a1 Densiy
U
2-
-
v0. 1.8M
5,
U
C)
"M
0
0
C3qm
a i:D
-
0
1.6Z 1")C
A,
*A
L
F.] C-Z.,
1.41.2I
1
0
Figure 4.6
I
.
I
I
0.1
.
.
.
.
I
,
.
.
I
I
I
I
.
I
.
I
0.2
0.3
0.4
Wt Fraction of Solvent Remaining
In situ density measurement for samples B and C from the "glass rod
table".
78
I
0.5
I
0.3N T(Mpm"r=346pm
I
f
* T530 pm T =244pm
I
It
0.25- 0
0 0
ILA
IA
IA A, A
=
ky ...-
A
A
)
A
A
0
4Lft
won
I
=420 pm T =205 pm
I
f
0
0 0
A I
f
0
*
An
*A
0
of
MA
A
A,
0
U
CIj
4
U
g 0.15-6
A
a
M
a
0
U
0
C"
A,
A
t I
0
0.05-
I
.1
7
0
Figure 47
.
I .
I I .
0.1
I I --A
.
I I -
- - - I ,
0.2
0.3
0.4
Wt Fraction of Solvent Remaining
- - - I
0. 5
Pore volume fraction as a function of solvent remaining for sample C.
IPMMA:BBP=I:l
9
0 PMMA:BBP =1.5:1
4
2.5 -
A PMMA:BBP =.8:1
A
$b
A
----Theoretical Density
i
A
V
M
a
rn
A
a
a
A
A
0
-u
:h
1.
a we
-
A
A
0
A
a
-
A
--,- A..
e
M
K
It
A,
n
p
0
u
r')
I
-. A,...
8
II
It
-
0. -
0I-
0
RQUre 48
I
0.1
I
0.2
I ,
-1 - - - I- 1
, , 1
7
;
;
,
I
I
0.5
0.6
07
0.8
0.3
0.4
Wt Fraction of Solvent Remaining
In situ density measurements for samples A,
80
;
1
0.9
I
I
and C from the tape caster.
I
N
PMMA:BBP=I:ITrv
I
* PMMA:BBP=I:ITry2
2.5 j
A
*
.- Theoretical Density
---
.4
"
n%
'nI-
PMMA:BBP=I:ITry3
n--
A
A
IA
r
0
-b
-
It
A,
WE
Hn
1. -
11
V)
N
0
A
V
I
0
A
I
0
0
0
Figure 49
V
I
0.1
0.2
I
I
I I I
:
- -T-
----- i
I ;
I
0.3
0.4
0.5
0.6
0.7
0.8
Wt Fraction of Solvent Remaining
In situ density measurements for sample C.
81
A
1
0.9
3
15
T
i
3
a.
M
G
cm
a
10
2
5D
'A
I?
6
7
.2
G
'D
5
1
1
I
1.0
Figure 410
I
1.4
1.6
Dibutyl phthalate YA/6)
The variation of the tensile strength of an Alumina green
sheet as a function of plasticizer concentration. (ref. 8)
82
Binder:
PNINIA
Polv(methvi methacrviate)
Plasticizer:
BBP
BenzvIButvlPhthalate
Powder:
A1101
(Particle size: 02-0.5 um)
Dispersant:
Fish Oil
Solvent:
Toluene
Solvent:powder:oryanic
=
6.8: 1 I
ol.)
-
PNINIA:BBP =
1.8:1 (ol.)
Sample A
PNINIA:BBP =
1.5:1 (ol.)
Sample
I PNIMA:BBP
=
1 I (Vol.)
Table 41 Composition
BBP Concentration
30%
,Sample
C
of green tape for insitu density measurements.
(Vol.%)
Theoretical Density (g,,/cm3)
1.168
Experimental Density igcm 3)
1.150 ± 0029
1.163
1.143 ± 0032
35%
Table 42
-
PVB-BBP tape densities.
83
-
- --
Binder
Plasticizer Vol.)
1.8:
1.5:1
1:1
Table 43
2.54
2.40
=
= 100
= 0
DrvinL Time iNfin.)
200
0.03
1.02
180
169
Final densities and dyingy times for samples A.
Drying Rate (Linear
Air Velocitv: ft/min)
Table 44
Green TaDe Densitv ivcm3)
2.57 0.02
and C.
Gren Tape
Densitv Pcm3)
Square
2-55 009
2.41 ± 0.08
2.39 ± 003
Dewetting
2.47
2.45
2.44
Hg
0.03
0.02
0.04
Photo
2.53 ± 003
2.42 ± 002
2.33 ± 003
No
Yes (slightly)
Yes (severeiv)
A summary of density and dewetting for sample C under different drying
rates.
84
CHAPTER
CONCLUSION AND FUTURE WORK
5.1 Major Conclusions
This project investigated
the relationship
between
processing
and
microstructure of tape cast green sheets. Plasticizer distribution across the
tape thickness was measured.
The drying behavior and densities of tapes with
three different polymer to plasticizer ratios were studied. Green tapes with
different drying rates were also examined.
The major conclusions and
accomplishments of this research are summarized below.
1. We have made the first in situ density measurements of the drying
tape. During drying, a microscope was used to measure the tape thickness;
and a balance was used to record its weight.
As a result, the inten-nediate
densities of the tape, its thickness reduction and its pore volume fraction were
calculated from its final density, thickness and weight during drying. We found
that the higher the polymer to plasticizer ratio, the less porosity existed in the
dried tape. Green tape A PMMA:BBP = 18:1) had a final density = 257 ±
0.02)
g/CM3,
which was almost consistent with the theory 256
g/CM3)
ts
densities during drying also demonstrated the most correlation with the theory.
Green tape C (PMMA :BBP = 11) deviated most from the theory, with 22%
by volume porosity remaining as it dried. Its final density was 240 ± 002)
g/CM3.
Dewetting and low final density occurred when a tape was dried under
a faster drying rate (S = 200 feet/min).
2. We developed method for measuring composition gradients on the
0.7 gm scale, this has never be done before for green tapes. Each microtomed
thin section was analyzed using FrIR-M to measure the absorption intensity of
unique absorption bands for the plasticizer and binder.
85
The plasticizer
concentration as a function of depth from the tape surface was determined by
measuring the absorption intensity of each band.
3. We found evidence for the first time for phase separation in PVB-
BBP system. Green tapes were found to be plasticizer-rich near the top
surface under slow drying conditions. Plasticizer-rich regions were distributed
more uniformly across the tape thickness at faster drying rates. The upper
boundary
was 55% by volume plasticizer concentration,
and the lower
boundary was 32% by volume plasticizer concentration.
4. Phase separation occurs when other plasticizers are used such as
Dioctyl
Phthalate,
Dibutyl Phthalate
and Dirnethyl Phthalate.
We also
demonstrated for the first time that the solubility lmits between the binder and
plasticizer correlated with the solubility parameter for the plasticizer.
Plasticizer with larger solubility parameter resulted smaller degree of phase
separation in the PVB green tape.
5.2 Industrial.Applications
The plasticizer and polymer phase separation phenomena has not been
reported to our knowledge in the context of tape-casting formulations. The
implications for devising tape casting formulations are that truly homogeneous
tapes can not be formed if the binder phase actually is composed
of two
phases. This situation is particularly serious if the phases completely separate.
In this case, the top of the tape differs significantly in composition and physical
properties from the bottom.
manufacture
he resulting anisotropy considerably complicates
of multilayer devices.
To
inimize the above problem,
we
suggest a plasticizer that is more soluble with the binder. In other words, the
plasticizer should have a solubility parameter closest to that of the binder. For
86
example, in the PVB binder system, Dimethyl Phthalate is more recommended
than BBP.
To reduce the porosity in the green tape, moderate high polymer to
plasticizer ratio and slow drying rate are recommend.
For example, in the
PMMA-BBP system, the binder to plasticizer ratio should not be lower than
1.5 :1 by volume.
5.3 Suggestions for future work
1. More details of the ternary phase diagram of the solvent-plasticizerbinder system should be investigated.
2. The solubility parameter of PVB monomer should be estimated.
3. Other techniques such as NMR should be used to verify the
plasticizer-binder phase separation.
87
Bibliography
1.
R. Mistler, "Tape Casting: The Basic Process for Meeting the Needs of
the Electronic Industry", Ceramic Bulletin, Vol. 69, NO. 6 1990.
2.
Y. Y. Tang, Ph.D. Thesis, Department of Materials Science and
Engineering, Massachusetts Institute of Technology, 1994.
3.
M. J. Cima, Y. Y. Tang and M. Liu, "Inhomogeneity and anisotropy of
4.
Tape Cast Ceramic Films For Multilayer Structures", Ceramics
Processing Laboratory, Massachusetts Institute of Technology, 1993.
Roosen, "Basic Requirement for Tape Casting of Ceramic Powders",
PP. 675-92 in Ceramic Transactions, Vol. 1, Ceramic Powder Science I
I, B/ Edited by G. L. Messing, E. R. Fuller and H. Hausner. American
Ceramic Society, Westerville, OH, 1988.
5.
T. Ueyama and N. Kaneko, "Effect of Agglomerated Particles on
Properties of Ceramic Green Sheets", pp. 1451-58 in High Tech
Ceramics. Edited by P. Vincenzini. Elsevier, Amsterdam, Netherlands,
6.
1987.
E. Hyatt, "Making Thin, Flat Ceramics-A Review", Ceramic Bulletin,
7.
Vol. 65 NO. 41986.
M. F. Yan, "Microstructural Control in the Processing of Electronic
Ceramics," Mater. Sci. Eng., 48, 53-72
1981).
5.
A. Roosen, F. Hessel, H. Fischer, F. Aldinger, "Interaction of
6.
Polyvinylbutyral with Alumina", Ceramic Powder Science, 1990.
J. S. Chong, E. B. Christiansen, and A. D. Baer, "Rheology of
Concentrated Suspensions," Journal of Concentrated Suspensions,
7.
8.
9.
15,2007-2021, 1971).
E. S. Tormey, R. L. Pober, H. K. Bowen, and P. D. Calvert, "Tape
Casting--Future Development," pp 140-49 in Advances in Ceramics,
Vol. 9 Forming of Ceramics, Edited by J. A. Mangels, and D. L.
Messing, American Ceramic Society, Columbus, Ohio, 1984.
Y. T. Cou, Y.T. Ko, and M. F. Yan, "Fluid Flow Model for Ceramic
Tape Casting", J. Am. Ceram. Soc., 70 101,C-280-C-282,1987.
R. Mistler, D. Shanefield, and R. Runk, "Tape Casting of Ceramics,"
pp 411418, in Ceramic Processing Before Firing. Edited by G. Y.
Onoda and L. L. Hench. Wiley, New York, 1978.
10.
11.
12.
13.
A. Karas, T. Kumagai, and R. Cannon, "Casting Behavior and Tensile
Strength of Cast BaTi4O3 Tape", Advanced Ceramic materials, 34]
374-77 1988).
Reid, Utramicrotomy", pp 217-223,Wiley, New York, 1980.
Goodhew, P. F. 1972) Specimen preparation in materials science in:
Practical methods in electron microscopy, A. M. Glauert, ed. NorthHolland, Amsterdam).
T. K. Sherwood, "The Drying of Solids", Ind. Eng. Chem., 21 (1),
ppl2-16, 1929.
88
1.4.
15.
16.
V. J. McBrierty, "Heterogeneity in Polymers as Studied by Nuclear
Magnetic Resonance", Faraday Discussions of the Chemical Society,
NO. 68,1979.
L. Sawyer, D. T. Grubb, "Polymer Microscopy", pp 213-18 Chapman
and Hall, London, New York, 1965.
H. Tadokoro, "Structure of Crystalline Polymers",pp 543-567, WileyInterscience, New York, 1979.
17.
18.
E. H. Andrews and J. M. Stubbs, J. R. Microsc. Soc. 82 pp 221, 1964.
G. Kani g, Proc. Colloid Polyrn. Sci. 57 pp 176, 197 5.
19.
Thermal Characterization of Polymeric Materials, edited by E. Turi,
20.
Academic Press Inc.
J. M. Massie, A. F. Halasa, R. N. Thudium, and C. W. Burkhart,
"Miscibility and Phase Behavior of Polyisoprene/polybutadiene and
Polybutadiene/Polybutadiene Blends", ANTEC'92.
21
J.A. Lewis and M.J. Cirna, "Diffusivities of Dialkyl Phthalates in
22
23
Plasticized Poly(vinyl butyral) Impact on BinderThermolysis" J. Am.
Ceram. Soc.,73 91 2702-2707 1990).
Kingery, Bowen and Uhlmann, "Introduction to Ceramics", Wiley
Interscience, P 448-P 515,1985.
J. Hildebrand, R. Scott, "The solubility of Non-electrolytes", 3rd ed.,
24
Reinhold Publishing Crop., N. Y., 1949.
C. Fiori and G. DePortu, Br. Ceram. Proc. 38 1986) 213.
25
26.
27.
E. Streicher, T. Chartier, "Study of cracking and microstructural
evolution during drying of tape-cast aluminum nitride sheets",
Chapman and Hall Ltd. 1991.
G. Y. Onoda, Jr., "The Rheology of Organic Binder Solutions" in
Ceramic Processing Before Firing, Wiley, New York.
R. Moreno, "The Role of Slip Additives in Tape Casting Technology,
Part 2-Binders and Plasticizers", Ceramic Bulletin, Vol. 7 , No. 11,
Nov. 1992.
28.
"Avoid Coating and Drying Defects", Chemical Engineering Progress, J
an. 1993.
89