Low-CTE photosensitive polyimide based on semialicyclic poly(amic acid) and photobase generator

Article abstract of DOI:10.1002/pola.23892

A negative-type photosensitive polyimide (PSPI) based on semialicyclic poly(amic acid) (PAA), poly( frans-1, 4-cyclohexylenediphenylene amic acid), and {[(4, 5-dimethoxy-2-nitrobenzyl)oxy]carbonyl]} 2, 6-dimethylpiperidlne (DNCDP) as a photobase generator has been developed as a next-generation buffer coat material. The semialicyclic PAA was synthesized from 3, 3', 4, 4'-biphenyltetracarboxylic dianhydride and frans1, 4-cyclohexyldiamine in the presence of acetic acid, and the PAA polymerization solution was directly used for PSPI formulation. This PSPI, consisting of PAA (80 wt %) and DNCDP (20 wt %), showed high sensitivity of 70 mJ/cm² and high contrast of 10.3, when it was exposed to a 365-nm line (i-line), postexposure baked at 190 °C for 5 min, and developed with 2, 38 wt % tetramethylammonium hydroxide aqueous solution containing 20 wt % isopropanol at 25 °C. A clear negative image of 6-μm line and space pattern was printed on a film, which was exposed to 500 mJ/cm² of i-line by a contact printing mode and fully converted to poly(trans-1, 4-cyclohexylenebiphenylene imide) pattern upon heating at 250 °C for 1 h. The PSPI film had a low coefficient of thermal expansion of 16 ppm/K compared to typical PIs, such as prepared from 3, 3', 4, 4'-biphenyltetracarboxylic dianhyariae ana 4, 4'-oxydianiline

Full text of DOI:10.1002/pola.23892

Low-CTE Photosensitive Polyimide Based on Semialicyclic
Poly(amic acid) and Photobase Generator
TOMOHITO OGURA, TOMOYA HIGASHIHARA, MITSURU UEDA
Department of Organic and Polymeric Materials, Graduate School of Science and Engineering, Tokyo Institute of Technology,
2-12-1 H120 O-okayama, Meguro-ku, Tokyo 152-8552, Japan
Received 10 November 2009; accepted 14 December 2009
DOI: 10.1002/pola.23892
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: A negative-type photosensitive polyimide (PSPI)
based on semialicyclic poly(amic acid) (PAA), poly(trans-1,4-
cyclohexylenediphenylene amic acid), and {[(4,5-dimethoxy-2-
nitrobenzyl)oxy]carbonyl} 2,6-dimethylpiperidine (DNCDP) as a
photobase generator has been developed as a next-generation
buffer coat material. The semialicyclic PAA was synthesized
line and space pattern was printed on a film, which was exposed
2
to 500 mJ/cm of i-line by a contact printing mode and fully con-
verted to poly(trans-1,4-cyclohexylenebiphenylene imide) pat-
ꢀ
tern upon heating at 250 C for 1 h. The PSPI film had a low
coefficient of thermal expansion of 16 ppm/K compared to typi-
0
0
cal PIs, such as prepared from 3,3 ,4,4 -biphenyltetracarboxylic
0
0
0
from 3,3 ,4,4 -biphenyltetracarboxylic dianhydride and trans-
,4-cyclohexyldiamine in the presence of acetic acid, and the
dianhydride and 4,4 -oxydianiline. VC 2010 Wiley Periodicals, Inc.
1
J Polym Sci Part A: Polym Chem 48: 1317–1323, 2010
PAA polymerization solution was directly used for PSPI formu-
lation. This PSPI, consisting of PAA (80 wt %) and DNCDP (20
KEYWORDS: high-temperature materials; lithography; low-tem-
perature imidization; negative type; negative-type photosensi-
tive polyimide; photobase generator; photoresists; photo-
sensitive polymer; poly(amic acid); polyimides; refractive
index; resists
2
wt %), showed high sensitivity of 70 mJ/cm and high contrast
of 10.3, when it was exposed to a 365-nm line (i-line), postex-
ꢀ
posure baked at 190 C for 5 min, and developed with 2.38 wt %
tetramethylammonium hydroxide aqueous solution containing
ꢀ
2
0 wt % isopropanol at 25 C. A clear negative image of 6-lm
INTRODUCTION Polyimides (PIs) have been widely studied
wafers. In previous works, we have developed several PSPIs,
1
,2
3,4
25,26
as insulating materials, optical applications, and piezo-
which can be imidized at low temperature.
However, the
5
electric materials. Moreover, photosensitive PIs (PSPIs) have
matrix polymer of these PSPIs is a PAA prepared from
0
0
0
been used in the semiconductor industry as interlevel insula-
tors, buffer coat materials, a-ray shielding layers, and so on,
because of their excellent thermal stability, high mechanical
properties, and relatively low dielectric constants. Further-
more, they simplify processing and avoid the use of photore-
3,3 ,4,4 -biphenyltetracarboxylic dianhydride (BPDA) and
4,4 -oxydianiline (ODA), which has a relatively high coeffi-
cient of thermal expansion (CTE). Therefore, these PSPIs
have limitations in reducing thermal strain even at low cur-
ing temperature, and PSPIs using low-CTE polyimides (PIs)
are highly demanded. Although low-CTE PIs can be obtained
from monomers having rigid structures, PI films prepared
from BPDA and 1,4-phenylenediamine, for example, are brit-
tle because of their rigidity, arising a difficulty to apply them
to PSPIs.
6
–23
sist in the microelectric industry.
The commercially avail-
able PSPIs usually consist of poly(amic acid) (PAA)
derivatives with methacryloyl pendant groups as crosslinking
6
,7
sites. However, PSPIs are further expected to improve their
performance with the increase developments in the semicon-
ductor industry. For example, a chemically amplified system
Hasegawa and Koyanaka reported the properties of poly
1
6
in PSPI has been introduced to show high sensitivity. Kim
and coworkers reported a positive-working PSPI, which can
(trans-1,4-cyclohexylenebiphenylene imide) prepared from
BPDA and trans-1,4-cyclohexyldiamine (CHDA), which
showed high thermal stability, a low dielectric constant, high
transparency, and low CTE because of its rigid and semialicy-
2
2
be developed in alkaline aqueous solution.
Recently, silicon substrates are stacked in a memory chip to
achieve high-density packaging, and those in multichip pack-
ages are thinner than the usual ones due to fitting them in
2
7,28
clic structure.
Moreover, the film formed from a highly
viscous poly(trans-1,4-cyclohexylenebiphenylene amic acid)
solution was flexible, which is suitable for a PSPI system.
Accordingly, PAA from BPDA and CHDA is excellently suitable
to a matrix polymer of PSPI, which can resolve the warpage
2
4
the traditional packaging size. Future PSPIs as buffer coat
materials require a lower imidization temperature because
high-temperature curing causes warpage of thinner silicon
Correspondence to: M. Ueda (E-mail: ueda.m.ad@m.titech.ac.jp)
Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 48, 1317–1323 (2010) V
2010 Wiley Periodicals, Inc.
C
LOW-CTE PHOTOSENSITIVE POLYIMIDE, OGURA, HIGASHIHARA, AND UEDA
1317

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
SCHEME 1 Patterning process of low-CTE PSPI.
of thinner silicon wafers. As PAAs can be imidized at low
dried in vacuo at room temperature. BPDA purchased from
2
5
ꢀ
temperature in the presence of a base catalyst, a PSPI
could be constructed using the polymerization solution of
PAA from BPDA and CHDA with a photobase generator
TCI was dried in vacuo at 180 C for 12 h before use.
9
DNCDP was prepared according to the reported procedure.
TMAHaq purchased from AZ Electronic Materials (JAPAN)
K. K. iPrOH purchased from GODO and was used in
unrefined.
(
PBG). In this article, we report a low-CTE PSPI based on
semialicyclic PAA, prepared from BPDA and CHDA, as a ma-
trix polymer and {[(4,5-dimethoxy-2-nitrobenzyl)oxy]car-
bonyl} 2,6-dimethylpiperidine (DNCDP) as a PBG. The pat-
terning process of this PSPI is shown in Scheme 1. The PSPI
resist solution is spin-coated on a silicon wafer and pre-
baked. The film is exposed to a 365-nm line (i-line) through
a photomask to produce 2,6-dimethylpiperidine (DMP) from
DNCDP. Upon PEB treatment of the PSPI film, DMP catalyzes
the imidization of PAA. The dissolution rate of the exposed
area to 2.38 wt % tetramethylammonium hydroxide aqueous
solution (TMAHaq) containing 20 wt % isopropanol (iPrOH)
decreases and a negative pattern is formed. Finally, partial PI
film is converted to complete PI film by thermal treatment
Synthesis of Poly(amic acid) from BPDA and CHDA
CHDA (0.114 g, 1.00 mmol) was placed in a 5-mL vial pot and
dissolved in DMAc to adjust the concentration of 15 wt %.
Acetic acid (0.132 g, 2.20 mmol) was slowly added to the dia-
mine solution. After 10 min, BPDA (0.294 g, 1.00 mmol) was
added into the solution, and the solution was stirred at room
temperature for 12 h. The yield was quantitative. The inherent
viscosity of this PAA was 0.24 dL/g at a concentration of
ꢀ
0.5 dL/g in DMAc at 30 C.
Degree of Imidization
ꢀ
The polymer solution containing acetic acid was diluted with
DMAc to a concentration of 15 wt %, followed by the addi-
tion of DNCDP. This solution was spin-coated on a silicon
at 250 C. This pattern formation method provides a versa-
tile process, such as direct resist formulation from a varnish
of the PAA, base-catalyzed low-temperature imidization, and
low-CTE PI film preparation for buffer coat materials.
ꢀ
wafer and baked at 100 C for 20 min in air. The film thick-
ness was about 1.0–1.2 lm. Then, the films were baked on a
ꢀ
hot plate at each temperature (110–250 C) for 20 min. A
EXPERIMENTAL
ꢀ
reference PI film was prepared by heating at 350 C for 1 h
Materials
under nitrogen. Absorption intensities on a Fourier-trans-
form infrared spectroscopy (FTIR) spectrum at 1365 cm
ꢁ
1
N,N-Dimethylacetamide (DMAc) was purified by vacuum
distillation. CHDA purchased from Tokyo Chemical Industry
(A1365) assignable to the CAN stretching of an imide group
ꢁ
1
(
TCI) was recrystallized from n-hexanes under nitrogen and
and at 609 cm (A609) assignable to a silicon wafer were
1318
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
SCHEME 2 Structures of BPDA/CHDA PAA and DNCDP.
measured, and the degrees of imidization (DIs) were deter-
mined using the following equation,
wide, and 26 lm thick) was measured as an average with
around 100–200 C by thermomechanical analysis (TMA) on
ꢀ
a Seiko TMA/SS 6100. Film thickness on silicon wafer was
measured by a Veeco Instrument Dektak surface profiler.
Imidizationð%Þ ¼ ðA1365=A609ðsampÞ ꢁ A1365=A609ðinitÞÞ=
ðA1365=A609ðimidÞ ꢁ A1365=A609ðinitÞÞ ꢂ 100:
3
The scanning electron microscope (SEM) photos were taken
with a Technex Lab Tiny SEM 1540 scanning electron micro-
scope with 15 kV accelerating voltage for imaging. Refractive
indices of PI and the PSPI films formed on quartz substrates
were measured at a wavelength of 1320 nm at room temper-
ature with a Metricon model PC-2000 prism coupler. The in-
plane (nTE) and out-of-plane (nTM) refractive indices of PI
films were measured with a prism coupler equipped with a
half-waveplate in the light path and a HeNe laser light
source (wavelength: 633 mm). nTE/nTM and birefringence
(1)
Subscripts between parentheses followed A1365/A609 in the
equation indicate the states of the polymer films; for exam-
ple, (samp) indicates the polymer sampled at each heating
ꢀ
temperature level (110–250 C); (init) indicates initially pre-
ꢀ
baked PAA at 100 C; (imide) indicates the fully cured at
ꢀ
3
50 C for 1 h in nitrogen.
Dissolution Rate
(
Dn) were calculated as a difference between nTE and nTM
.
DNCDP was dissolved in a PAA polymerization solution to
formulate a photosensitive polymer. The polymer film was
spin-casted from the solution on a silicon wafer and pre-
The average refractive index was calculated according to
2
2
1/2
equation: nAV ¼ [(2n þ n )/3] . The dielectric constant
TE
TM
at 10 MHz frequency was calculated from the following em-
pirical equation, e ¼ 1.1 n , where nAV is an average refrac-
ꢀ
baked at 100 C for 5 min, then exposed to a filtered ultra-
2
AV
high-pressure mercury lamp at i-line, followed by PEB at
tive index.
ꢀ
ꢀ
1
50–200 C for 5 min or PEB at 190 C for 1–7 min. The
exposed film was developed with 2.38 wt % TMAHaq con-
taining 20 wt % iPrOH at 25 C and then measured the
change of film thickness before and after the development to
determine the dissolution rate (Å/s).
RESULTS AND DISCUSSION
ꢀ
Synthesis of PAA from BPDA/CHDA
Generally, it is rather difficult to prepare semialicyclic PAAs
from aromatic dianhydrides and alicyclic diamines. The car-
boxylic acids of PAAs easily generate salts or gels with alicy-
clic diamines because the basicity of alicyclic diamines (pKa
of the ammonium ion: 10.7) is much higher than that of aro-
Photosensitivity
A 1.7-lm-thick photosensitive polymer film was prepared
from a varnish consisting of 80 wt % PAA and 20 wt %
DNCDP, followed by spin-coating on a silicon wafer, and then
matic diamines (pK of the ammonium ion: 4.6). This ionic
a
ꢀ
prebaked at 100 C for 5 min. This film was exposed to i-
interaction prevents the synthesis of PAAs. We previously
line using a filtered super high-pressure mercury lamp, fol-
reported a facile method for the synthesis of PAA from BPDA
and organic salts consisting of CHDA and acetic acid. This
ꢀ
29
lowed by PEB at 190 C for 5 min, then developed with 2.38
ꢀ
wt % TMAHaq containing 20 wt % iPrOH at 25 C for 15 s,
polymerization proceeded in homogeneous state without gel
formation and yielded PAA with high molecular weight. The
synthetic scheme of PAA from BPDA and CHDA and the
structure of DNCDP are shown in Scheme 2. The inherent
viscosity of PAA was intentionally set as low as 0.24 dL/g so
that high dissolution contrast (DC) was expected. The PAA
polymerization solution containing acetic acid was directly
used for the PSPI formulation.
and subsequently rinsed with distilled water. The character-
istic photosensitivity curve was obtained by plotting a nor-
malized film thickness as a function of exposure dose (unit:
2
mJ/cm ). Image-wise exposure was carried out in a contact
printing mode.
Measurements
The FTIR spectra were obtained on a Horiba FT-210 spectro-
1
13
photometer. The H and C nuclear magnetic resonance
NMR) spectra were recorded on a BRUKER GPX300 (300
MHz) spectrometer. Dynamic mechanical thermal analysis
Lithographic Evaluation
(
PSPI was formulated by mixing the PAA polymerization solu-
tion with DNCDP, as shown in Scheme 1. To clarify the disso-
lution rate between the exposed and unexposed areas to
2.38 wt % TMAHaq containing 20 wt % iPrOH, the effects of
DNCDP loading, PEB temperature, PEB time, and exposure
dose were investigated. The PSPI films were obtained by
(
DMA) was performed on a Seiko DMS 6300 at a heating
ꢀ
rate of 2 C/min with a load frequency of 1 Hz in the air
using the PI or thermally cured PSPI films (10 mm long, 5
mm wide, and 25 lm thick). The glass transition tempera-
ꢀ
tures (T s) were determined as the peak temperature of tan
spin-casting on a silicon wafer and then prebaking at 100 C
g
d plots. CTE of the PI and PSPI films (10 mm long, 4 mm
for 5 min in air. These PSPI films were irradiated with UV
LOW-CTE PHOTOSENSITIVE POLYIMIDE, OGURA, HIGASHIHARA, AND UEDA
1319

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
FIGURE 1 Effect of DNCDP loading to PAA on the dissolution
rate for the resist system under exposed (^) and unexposed
area (&). The prebake, the i-line exposure, and PEB conditions
FIGURE 2 Effect of PEB temperature on the dissolution rate for
the PAA/DNCDP (80/20 wt/wt) resist system under exposed (^)
and unexposed area (&). The prebake, the i-line exposure, and
ꢀ
2
ꢀ
2
were fixed to 100 C for 5 min, 500 mJ/cm , and at 190 C for
min, respectively.
ꢀ
PEB time were fixed to 100 C for 5 min, 500 mJ/cm , and for
min, respectively.
5
5
In a similar way, the effect of PEB time on the dissolution
rate was investigated under the same condition (Fig. 3). The
dissolution rate at the exposed area gradually decreased
with increasing the PEB time and became almost zero under
condition of 5 min PEB time. Table 2 shows the DI at each
PEB time. It indicates that a small difference in the DI
strongly affects the dissolution rate.
light at i-line using a filtered super high-pressure mercury
lamp, PEB at a set temperature for a prescribed period of
time, and developed with 2.38 wt % TMAHaq containing 20
ꢀ
wt % iPrOH at 25 C. The partially imidized PAA by the PEB
process at the exposed area was slightly soluble in 2.38 wt %
TMAHaq. To reduce the solubility of the partially imidized
PAA, 20 wt % iPrOH was added to 2.38 wt % TMAHaq solu-
tion. The dissolution rate was estimated from the change in
the film thickness before and after development.
Based on these preliminary optimization studies, PSPI con-
taining PAA (80 wt %) and DNCDP (20 wt %) was formu-
lated. The photosensitivity curve of the resist film with 1.7-
lm-thick film is shown in Figure 4. This resist has excellent
First, the effect of DNCDP loading on the dissolution rate of
the film was investigated (Fig. 1), where each film was pre-
2
ꢀ
2
sensitivity (D ) of 70 mJ/cm and high contrast (c ) of 10.3
0
0
baked at 100 C for 5 min, exposed to 500 mJ/cm , and then
ꢀ
with i-line exposure. These findings indicate that a photogen-
erated DMP effectively catalyzes the imidization of PAA by
PEB treatment, producing a large DC between the exposed
and unexposed areas.
PEB at 190 C for 5 min. The dissolution rate of the exposed
area becomes almost zero in the case of 20 wt % DNCDP
loading, and the DC between the exposed and unexposed
areas in 2.38 wt % TMAHaq containing 20 wt % iPrOH
reaches 3000 times. It should be mentioned that the low-
temperature imidization at the exposed area was not affected
by acetic acid, because most acetic acids might be evapo-
rated in the prebaking process. Moreover, DNCDP did not
decompose by the PEB process, because it was stable up to
Low-Temperature Imidization of PSPI
To conduct the complete imidization of this PSPI, PSPI solu-
tion containing PAA (80 wt %) and DNCDP (20 wt %) was
spin-coated on a silicon wafer, prebaked on a hot plate at
ꢀ
9
ꢀ
2
50 C.
100 C for 20 min, and then postexposure baked at each
ꢀ
temperature (110–250 C) for 20 min after exposure to 500
The PEB temperature is a very important parameter for this
PSPI system, because the DI depends strongly on the curing
temperature. PSPI consisting of PAA (80 wt %) and DNCDP
2
mJ/cm of i-line. A reference PI film was also prepared by
a
TABLE 1 The Degree of Imidization at Each PEB Temperature
(
20 wt %) was formulated, and then the effect of PEB tem-
perature on the dissolution rate was investigated (Fig. 2). A
large DC between the exposed and unexposed areas is
observed and reaches 3000 times at 190 C. The DI at each
ꢀ
Temperature ( C)
150
160
170
180
190
200
ꢀ
Exposed area
2
0
3
0
8
2
12
6
34
13
54
24
temperature is summarized in Table 1. PAA film with over
Unexposed area
3
4% imidization does not dissolve in 2.38 wt % TMAHaq
a
Conditions for preparation of PSPI films: the resist films (PAA/DNCDP,
containing 20 wt % iPrOH. The DI at the unexposed area
also increases with increasing the PEB temperature, and the
dissolution rate at the unexposed area decreases.
ꢀ
8
0/20 (w/w)) was prebaked at 100 C, exposed with the i-line at 500 mJ/
2
cm , postexposure baked at the set temperature for 5 min. The degree
of imidization was calculated by eq 1.
1320
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
FIGURE 3 Effect of PEB time on the dissolution rate for the
PAA/DNCDP (80/20 wt/wt) resist system under exposed (^)
and unexposed area (&). The prebake, the i-line exposure, and
FIGURE 4 Characteristic photosensitive curve for the PAA/
DNCDP (80/20 wt/wt) resist system in 1.7-lm film thickness.
ꢀ
The prebake and PEB were fixed to 100 C for 5 min and at 190
ꢀ
2
PEB temperature were fixed to 100 C for 5 min, 500 mJ/cm ,
ꢀ
C for 5 min, respectively.
ꢀ
and at 190 C, respectively.
2
ꢀ
5
00 mJ/cm of i-line, postbaking at 190 C for 5 min, and
ꢀ
heating at 350 C for 1 h under nitrogen. Figure 5 shows the
developing to 2.38 wt % TMAHaq containing 20 wt %
iPrOH. A clear negative pattern with a 6-lm feature could be
obtained when a 1.4-lm-thick film was used. The printed
PAA pattern was converted to the PI pattern by heating at
DI of PSPI film with DNCDP and the pristine PAA film. In
ꢁ
1
this study, two bands at 1365 and 609 cm due to the CAN
stretching of an imide group and a silicon wafer, respectively,
were used for the estimation of the DI. The imidization of
PSPI film is completed at 250 C because of the presence of
photogenerated DMP from DNCDP. In contrast, additive-free
PAA film reaches 80% of imidization at 250 C. The cleavage
percentage of DNCDP with exposure to 500 mJ/cm was
around 17%, estimated by data in the previous report of
photosensitive poly(benzoxazole) using PBG. It is indicated
that low-temperature imidization of this PSPI completed in
the presence of a small amount of DMP. In a previous work,
the PAA film from BPDA and ODA containing a base catalyst
could enforce imidization at 200 C. This PSPI film, how-
ever, required high thermal treatment at 250 C to complete
ꢀ
an elevated temperature up to 250 C for 1 h under nitrogen
ꢀ
[
Fig. 6(b)]. The film thickness decreased from 1.4 to 1.0 lm,
and no damage of the formed patterns was found. The for-
mation of PI was confirmed by FTIR spectroscopy, in which
ꢀ
2
the peak assignable to the CAN stretching of imide groups
ꢁ1
at 1365 cm
appeared and the absorption of carboxylic
3
0
ꢁ1
acid at 1716 cm disappeared.
Refractive Index and Dielectric Constant
A dielectric constant (e) is one of the important properties of
PSPIs as insulator materials. The e of polymer film at 10
ꢀ
25
ꢀ
the imidization. It is assumed that DMP generated from
ꢀ
DNCDP as a base catalyst vaporizes at around 200 C, judg-
ing from the very slow imidization rate at these tempera-
tures. Therefore, a nonvolatile base from PBG should be
developed to conduct lower temperature imidization in PSPI.
Image Formation of PSPI
The SEM image in Figure 6(a) reveals the result of a negative
PAA pattern after a series of treatments, such as exposure to
a
TABLE 2 The Degree of Imidization at Each PEB Time
Time (min)
1
2
3
4
5
7
Exposed area
8
3
18
5
23
9
29
11
34
13
37
17
Unexposed area
a
Conditions for preparation of PSPI films: the resist films (PAA/DNCDP,
FIGURE 5 The degree of imidization of the PAA film containing
DNCDP (20 wt %) (^) and no additive (&) estimated from
FTIR.
ꢀ
8
0/20 (w/w)) was prebaked at 100 C, exposed with the i-line at 500 mJ/
2
ꢀ
cm , postexposure baked at 190 C for the set time. The degree of imid-
ization was calculated by eq 1.
LOW-CTE PHOTOSENSITIVE POLYIMIDE, OGURA, HIGASHIHARA, AND UEDA
1321

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
FIGURE 6 SEM images of negative patterns: (a) a 1.4-lm-thick PAA film based on the PAA/DNCDP (80/20 wt/wt) resist system.
ꢀ
2
ꢀ
The prebake, the i-line exposure, and PEB were fixed to 100 C for 5 min, 500 mJ/cm , and at 190 C for 5 min, respectively, (b)
ꢀ
a 1.0-lm-thick PI film cured at 250 C for 1 h under nitrogen.
ꢀ
MHz is calculated from the average refractive index (nAV) of
the film according to a modification of Maxwell’s equation,
plots, is observed at 359 C, whereas the Tg of PSPI is
ꢀ
observed at 329 C. It is assumed that the Tg of PSPI is
2
31
e ¼ 1.1 n
.
The PAA and PSPI solutions were spin-casted
slightly lower than that of PI probably because of the resi-
dues of DNCDP in the PSPI film. It is noteworthy that the
PSPI film has a mechanical property similar to the pristine
PI film. The CTE values of PI and PSPI films measured by
TMA were 10 and 16 ppm/K, respectively. Because of low-
temperature curing and the residue of the PBG, the CTE
value of the PSPI film is slightly higher than that of the pris-
tine PI film. This CTE value, however, is enough low for the
next-generation PSPIs.
AV
on a quartz plate, and the films were imidized by heating.
Table 3 shows the refractive indices and the optically esti-
mated e values of the PI and PSPI films. Generally, the e
value of PSPIs becomes higher than the matrix PI because
photoactive compounds have high polar character. The aver-
age refractive indices (nAV) of both films are determined as
1.657 and 1.647, respectively, which are translated into e val-
ues of 3.03 and 2.99, respectively. These results suggest that
the incorporation of DNCDP in PSPI does not affect the e
value very much. A high Dn value of PI film is observed
when compared with that of PSPI film because of the orien-
tation of PI by high-temperature curing. This result suggests
that the intermolecular orientation of PSPI film is suppressed
by low-temperature imidization.
CONCLUSIONS
In this article, low-CTE semialicyclic PI prepared from BPDA
and CHDA could be formed as PSPI, which is suitable as a
buffer coat in multichip packages. This negative-type PSPI
consisted of a PAA from BPDA and CHDA with DNCDP as a
PBG. This PSPI system has the following advantageous points
Thermal and Mechanical Properties of PSPI
The thermal stability and mechanical properties of PI and
PSPI were investigated by DMA. The DMA curves of PI film
ꢀ
cured at 350 C for 1 h and PSPI film exposed to 500 mJ/
2
ꢀ
cm and cured at 250 C for 1 h are shown in Figure 7. The
initial storage moduli (E ) of PI and cured PSPI at about 50
C are 1.4 and 1.1 GPa, and their loss moduli (E ) at the
same temperature are 133 and 124 MPa, respectively. The T
0
ꢀ
00
g
of PI, determined from the peak temperature of the tan d
TABLE 3 Refractive Indices and Dielectric Constants of Polymer
Films
a
b
Film
n
TE
n
TM
n
AV
Dn
e
PI
1.716
1.670
1.539
1.601
1.659
1.648
0.177
0.069
3.03
2.99
PSPI
a
Conditions for preparation of PSPI films: the resist films (PAA/DNCDP,
ꢀ
8
0/20 (w/w)) was prebaked at 100 C, exposed with the i-line at 500 mJ/
2
ꢀ
cm , postexposure baked at 190 C for 5 min. Second, the resist films
ꢀ
were cured at 250 C for 1 h under nitrogen. Conditions preparation of
0
ꢀ
FIGURE 7 DMA curves (storage modulus E and loss modulus
PI films: the films cured at 350 C for 1 h under nitrogen.
b
00 ꢀ
Birefringence: Dn ¼ nTE ꢁ nTM
.
E , 1 Hz, 2 C/min, in air) of PSPI and cured PI films.
1322
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
and satisfactory properties: (i) direct formulation of PSPI
13 Mochizuki, A.; Ueda, M. J Photopolym Sci Technol 2001, 14,
from the PAA polymerization solution, (ii) high sensitivity of
677–687.
2
7
0 mJ/cm and high contrast of 10.3 with i-line exposure,
1
4 Fukushima, T.; Oyama, T.; Iijima, T.; Tomoi, M.; Itatani, H. J
ꢀ
(
iii) low-temperature imidization at 250 C, and (iv) the low-
Polym Sci Part A: Polym Chem 2001, 39, 3451–3463.
CTE film of 16 ppm/K. Furthermore, this PSPI exhibits high
thermal stability and good mechanical properties and is suit-
able for the next-generation PSPIs.
1
5 Zhang, A.; Li, X.; Nah, C.; Hwang, K.; Lee, M. H. J Polym Sci
Part A: Polym Chem 2003, 41, 22–29.
1
6 Watanabe, Y.; Fukukawa, K.; Shibasaki, Y.; Ueda, M.
J Polym Sci Part A: Polym Chem 2005, 43, 593–599.
REFERENCES AND NOTES
17 Watanabe, Y.; Shibasaki, Y.; Ando, S.; Ueda, M. Polym J
2
005, 37, 270–276.
1
Ye, Y. S.; Yen, Y. C.; Chen, W. Y.; Cheng, C. C.; Chang, F. C. J
Polym Sci Part A: Polym Chem 2008, 46, 6296–6304.
18 Pyo, S. H.; Lee, M. Y.; Jeon, J. J.; Lee, J. H.; Yi, M. H.; Kim,
J. S. Adv Funct Mater 2005, 15, 619–626.
2
Wahab, M. A.; Mya, K. Y.; He, C. J Polym Sci Part A: Polym
Chem 2008, 46, 887–5896.
19 Sakayori, K.; Shibasaki, Y.; Ueda, M. J Polym Sci Part A:
Polym Chem 2006, 44, 6385–6393.
3
You, N. H.; Fukuzaki, N.; Suzuki, Y.; Nakamura, Y.; Higashi-
hara, T.; Ando, S.; Ueda, M. J Polym Sci Part A: Polym Chem
2
0 Shin, G. J.; Jung, J. C.; Chi, J. H.; Oh, T. H.; Kim, J. B.
2009, 47, 4428–4434.
J Polym Sci Part A: Polym Chem 2007, 45, 776–788.
4
You, N. H.; Suzuki, Y.; Nakamura, Y.; Higashihara, T.; Ando,
2
1 Fukukawa, K.; Ueda, M. Polym J 2008, 40, 281–296.
S.; Ueda, M. J Polym Sci Part A: Polym Chem 2009, 47,
2
2 Li, T.; Hsu, S. L. J Polym Sci Part A: Polym Chem 2009, 47,
4886–4894.
1
575–1583.
5
Gonzalo, B.; Vilas, J. L.; Breczewski, T.; P e´ rez-Jubindo, M. A.;
Fuente, M. R. D. L.; Robriguez, M.; Le o´ n, L. M. J Polym Sci Part
A: Polym Chem 2009, 47, 722–730.
23 Saito, Y.; Mizoguchi, K.; Higashihara, T.; Ueda, M. J Appl
Polym Sci 2009, 113, 3605–3611.
6
Rubner, R.; Ahne, H.; K u¨ hn, E.; Koloddieg, G. Photogr Sci
24 Tsukui, T. J Jpn Inst Electron Packag 2008, 11, 317–325.
Eng 1979, 23, 303–309.
2
5 Fukukawa, K.; Shibasaki, Y.; Ueda, M. Polym Adv Technol
7
Yoda, N.; Hiramoto, H. J Macromol Sci Chem 1984, 21, 1641–
2006, 17, 131–136.
1663.
2
6 Ogura, T.; Higashihara, T.; Ueda, M. J Polym Sci Part A:
8
Mochizuki, A.; Teranishi, T.; Ueda, M.; Matsushita, K. Poly-
Polym Chem 2009, 47, 3362–3369.
mer 1995, 36, 2153–2158.
2
7 Hasegawa, M. High Perform Polym 2001, 13, 93–106.
9
Mochizuki, A.; Teranishi, T.; Ueda, M. Macromolecules 1995,
2
8 Hasegawa, M.; Koyanaka, M. High Perform Polym 2003, 15,
28, 365–369.
4
7–64.
10 Ueda, M.; Nakayama, T. Macromolecules 1996, 29, 6427–
2
9 Ogura, T.; Ueda, M. Macromolecules 2007, 40, 3527–3529.
6431.
11 Nakayama, T.; Mochizuki, A.; Ueda, M. React Funct Polym
30 Mizoguchi, K.; Ueda, M. J Polym Sci Part A: Polym Chem
1996, 30, 109–115.
2008, 46, 4949–4958.
12 Nakano, T.; Iwasawa, H.; Miyagawa, N.; Takahara, S.;
31 Mizoguchi, K.; Higashihara, T.; Ueda, M. Macromolecules
Yamaoka, T. J Photopolym Sci Technol 2000, 13, 715–718.
2009, 42, 3780–3787.
LOW-CTE PHOTOSENSITIVE POLYIMIDE, OGURA, HIGASHIHARA, AND UEDA
1323