close

Вход

Забыли?

вход по аккаунту

?

Effects of defrosting period on mold adhesion force of epoxy molding compound.

код для вставкиСкачать
ASIA-PACIFIC JOURNAL OF CHEMICAL ENGINEERING
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
Published online 13 October 2008 in Wiley InterScience
(www.interscience.wiley.com) DOI:10.1002/apj.186
Special Theme Research Article
Effects of defrosting period on mold adhesion force of
epoxy molding compound
Hwe-Zhong Chen,1 Wen-Hung Lee,1 Huei-Huang Lee,1 Durn-Yuan Huang,2 Shyang-Jye Chang3 and
Sheng-Jye Hwang3 *
1
Department of Engineering Science, National Cheng Kung University, Tainan, Taiwan
Department of Environmental and Safety Engineering, Chung Hwa College of Medical Technology, Tainan, Taiwan
3
Department of Mechanical Engineering, National Cheng Kung University, Tainan, Taiwan
2
Received 30 May 2008; Revised 9 June 2008; Accepted 8 July 2008
ABSTRACT: In integrated circuit (IC) packaging, when epoxy-molding compound (EMC) is filled in the mold cavity
and cured in the mold, adhesion occurs in the interface between EMC and the mold surface. Too large an adhesion
force can cause many problems. For example, too large an adhesion force may damage an IC during ejection and
cause the package to fail and thus lower the yield rate. To resolve mold adhesion problems, improving the mold
design and applying suitable surface treatments, such as mold surface coating, are the common approaches. Applying
suitable surface coating is a more popular and practical approach. Defrosting is a process to increase the frozen EMC
temperature to room temperature, and to retain it at room temperature for some period before molding. It is a common
practice to put EMC under required atmospheric environment during defrosting. It has been found by molding engineers
that increased defrosting period will increase the frequency of mold cleaning. But there is no quantitative description
on how much the adhesion force increases during the defrosting process.
This paper describes the use of a semiautomatic EMC adhesion force test instrument to measure the normal adhesion
force between the mold surface and EMC. By measuring the adhesion force, one can quantify how much adhesion
force exists between EMC and the mold surface under different defrosting periods. The results show that it is best
to use the EMC with 24–32 h of defrosting, to prevent excessive amount of mold adhesion force and it has been
found that the adhesion force of the 24 h defrosting period will be 24% less than that of the 48 h defrosting period.
Decreasing moisture absorption will decrease the increase in adhesion force for prolonged defrosting period cases. 
2008 Curtin University of Technology and John Wiley & Sons, Ltd.
KEYWORDS: EMC; mold adhesion force; defrosting period
INTRODUCTION
Mold adhesion is a problem in the integrated circuit
(IC) packaging industry. We expect a strong adhesion between the EMC and the leadframe or the die
surface, and prefer to have the smallest amount of
adhesion between EMC and the mold surface. Adhesion between EMC and the mold surface is especially
annoying because the mold surface needs to be cleaned
every 500–600 shots depending on the recipe of the
epoxy-molding compound (EMC). This means an automolding machine needs to be shut down for 2–3 h
every day for cleaning. The productivity of an automolding machine will be reduced by an amount of
10–15% because of the mold adhesion problem. The
*Correspondence to: Professor Sheng-Jye Hwang, Department of
Mechanical Engineering, National Cheng Kung University, Tainan,
Taiwan. E-mail: jimppl@mail.ncku.edu.tw
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
mold adhesion problem gets even worse with the green
EMC, which is designed to be more sticky.
Until now, there are three approaches to reduce the
problem caused by mold adhesion effects. The first
one is the special treatment of the mold surface, such
as: surface coating on the metal mold to change mold
surface quality. The second one is to add mold release
agent into the encapsulation materials or apply mold
release agent on the surface of the mold. The third
approach is to change the ejection pin positions to
reduce the damage that mold adhesion effects bring
on IC packages during ejecting process because of the
imbalanced adhesion forces. The first approach is more
effective and is the trend.
Mold surface coating or treatment is considered an
effective approach to resolve the problems of mold
adhesion in IC packaging. To evaluate the effectiveness
of any surface treatment, the authors have designed and
fabricated an instrument to quantity the effectiveness of
the mold surface coating.[1,2]
162
H.-Z. CHEN ET AL.
Asia-Pacific Journal of Chemical Engineering
Defrosting is a process to increase the frozen EMC
temperature to room temperature, and retain it at room
temperature for some time before molding. It has been
found by molding engineers that increased defrosting
time will increase the frequency of mold cleaning.
But there has been no quantitative description on
the increase of the mold adhesion force during the
defrosting process. This paper describes the use of a
modified version of the previous test instrument to
evaluate the effects of defrosting period on the mold
adhesion force of EMC. A modified semiautomatic
EMC normal adhesion force test instrument was used
to evaluate the effects of the defrosting time on the
magnitude of adhesion force between EMC and mold
surface. The influence of moisture on the adhesion force
during defrosting is also discussed.
MOLD ADHESIVE TESTING CRITERION
As far as we know, there is no standard test procedure
for EMC mold adhesion force test in normal direction.
The only standard test procedure found is Standard
Test Method for Tensile Properties of adhesive bonds,
STD. ASTM D897-95a.[3 – 5] STD. ASTM D897-95a
regulates the adhesive strength measuring standard. The
adhesive material is placed between wood or metal
block materials in a butt-join way. The geometry of
the block is cylindrical with a circular cross section
of diameter 20.17 ± 0.13 mm for wood and 28.68 ±
0.13 mm for metal. During the test, a force is applied at
one end of the cylindrical block. The maximum amount
of force needed to separate the cylindrical blocks was
measured and is considered as the strength of the
bonding.[6]
The STD. ASTM D897-95a is suitable for measuring
the interfacial bonding strength of adhesives and is
not suitable for measuring the adhesion force between
EMC and the mold surface. However, this standard
does provide some hints for the development of our
Top Mold (Specimen)
Middle Mold Plate
Bottom Mold
Picture of the test machine.
This figure is available in colour online at
www.apjChemEng.com.
Figure 2.
Figure 3. Schematic of the EMC sample.
instrument. We also adopted the bonding area to be
a circular cross section with a diameter of 28.68 ±
0.13 mm for the sample used in our instrument.
In 2003, Masaki Yoshii et al .[7] attached a load cell
to the bottom of the ejection pin in IC packaging mold.
By ejecting the molding sample, releasing force can
be measured by the load cell at the same time. They
also found that as the number of injection increases,
releasing force would also increase. Since this approach
cannot separate the normal and shear modes of the
adhesion force, we developed a new test equipment to
accurately measure the adhesion force between EMC
and the mold surface.[1,2]
STRUCTURE OF THE TEST MACHINE AND
MOLD
Mold Base
Plunger
Figure 1. Schematic of the mold system.
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
The experiments used a newer version semiautomatic
EMC normal adhesion force test instrument that had
been developed and fabricated. The earlier version of
this instrument has been reported.[1,2] There have been
some modification in this newer version of machine.
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
Middle Mold Plate
EFFECTS OF DEFROSTING PERIOD ON EPOXY MOLDING COMPOUND
Specimen
EMC
Transfer Pot
Bottom Mold
Mold Base
(a) Step 1: Open and clean mold
(b) Step 2: Insert the EMC pellets
Mold Cavity
Gate
Runner
(c) Step 3: Heat the EMC pellets
(e) Step 5: EMC melt fills and cures
in the cavity
(d) Step 4: Close the mold
(f) Step 6: Pull the top mold of the sample and
separate the surface of the mold and trigger
the measurement
Ejection Pin
Plunger
(g) Step 7: Eject the sample
(h) Step 8: Remove the sample and
end of one cycle
Figure 4. Schematic of the test procedures. This figure is available in colour
online at www.apjChemEng.com.
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
DOI: 10.1002/apj
163
164
H.-Z. CHEN ET AL.
The main differences between this newer version
instrument and the earlier version instrument are:
1. The transfer pot of the instrument is moved to a new
position, which is different from the mold cavity.
This makes the instrument to simulate the real auto
mold better or transfer mold process.
2. The actuators of the clamping system and the injection unit were changed to servomotors to reduce the
amount of vibration induced at the moment when the
top mold is opened.
3. The mold is modified to a three-plate type mold as
shown in Fig. 1.
During the experiment, the environment should be
maintained at relative humidity of 50 ± 2% and temperature 23 ± 1 ◦ C. This condition is from STD. ASTM
D897-95a. The laboratory will need an air conditioner
to control the environment.
The instrument is able to handle different mold surface treatments and control EMC or mold temperature,
cavity pressure, charge-in time and charge-out time.
The picture of the test machine is shown in Fig. 2.
The geometry of the sample or the mold shape is also
modified. Schematic of the modified EMC sample is
shown in Fig. 3. The sample is pan shaped and is very
similar to the previous design but for four nut-shaped
columns. The reason for using pan-shaped sample with
columns is to provide a surface for the mold to clamp
the sample when the top surface is pulled and separated
from the top mold surface. The mold opening speed is
set to be less than 2 mm/s to avoid impact.
Asia-Pacific Journal of Chemical Engineering
load and it can measure the normal mold adhesion
force when the test interface is separated. After that,
the middle mold plate will be raised and the molded
part of EMC can be ejected.
EXPERIMENT OF EMC UNDER DIFFERENT
DEFROSTING TIME
The focus of this research is to study the effects of
different EMC defrosting periods on the mold adhesion
force. We used a set of optimal process parameters to
conduct this experiment. The optimal process control
factors obtained from the optimal experiment (minimize
the mold adhesion force) by using Taguchi’s method
are shown in Table 1. The type of EMC was EMEG770 made by SUMITOMO company in Japan. The
mold surface was coated with G2 material, which is a
Cr-based surface coating material and is developed by
Metal Industries Research and Development Center in
Kaohsiung, Taiwan. For each experiment, there were six
specimens coated with G2 material. Roughness of the
Load Cell
Bracket
1~2 mm
Motion Mold Platen
ENCAPSULATION PROCESS
The schematic of the test procedure and the schematic
of the test mechanism are shown in Figs 4 and 5
respectively. The test procedure is the same as the
earlier version of this machine but with offset injection
pot and a three-plate type mold. The basic cycle is very
similar to an auto mold machine. Open and clean mold.
→ Insert the pellet. → Heat the EMC pellet. → Close
the mold. → EMC melt fills the cavity. → Wait until
EMC is cured. → Pull the top mold of the sample and
separate the surface of the mold and trigger the test. →
Eject the sample. → Remove the sample. → End of
one cycle. When the top mold closes, the motion mold
plate will press the insulation and the test specimen. The
insulation and the test specimen will press the middle
mold plate and the sample of EMC. The reading of the
load cell is zero at this time. There is a 1–2 mm gap
between the connecting rod of the test specimen and the
bracket when the mold is closed. When the top mold
opens, the motion mold plate will rise and pull the test
specimen through the bracket. The measurement can be
done at this moment. The load cell is under a tension
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
Specimen
Insulation
Pot
Ejection Pin
Mold Cavity
Mold
Plunger
Schematic of the front view of the test
mechanism. This figure is available in colour online at
www.apjChemEng.com.
Figure 5.
Table 1. Control factors for the experiment.
Factor
Parameter
EMC type
Mold surface rough
Mold surface treatment
Test specimens × 6
Mold temperature
Filling pressure
Preheat time
Curing time
EME-G770
Ra 2.0
G2 coating
G2 coating
170◦ C
90 kgf/cm2
12 s
100 s
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
EFFECTS OF DEFROSTING PERIOD ON EPOXY MOLDING COMPOUND
mold surface is Ra = 2.0 µ. Mold temperature was set
at 170 ◦ C. Injection pressure was 90 kgf/cm2 . Preheating
time was 12 s. and curing time was 100 s. In this
experiment, the mold adhesion force was measured
40 times for each defrosting period and six different
defrosting periods were chosen. Thus, the total number
of experiments is 240. The time interval between
each experiment set is 8 h. The shortest defrosting
period is 24 h. Therefore, 2 days (48 h) and six sets of
measurements were needed to complete all experiments.
Figure 6. Results of the first experiment of EMC defrosted for 24 h. This figure
is available in colour online at www.apjChemEng.com.
Figure 7. Results of the first experiment of EMC defrosted for 32 h. This figure
is available in colour online at www.apjChemEng.com.
Figure 8. Results of the first experiment of EMC defrosted for 40 h. This figure
is available in colour online at www.apjChemEng.com.
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
DOI: 10.1002/apj
165
166
H.-Z. CHEN ET AL.
Asia-Pacific Journal of Chemical Engineering
The experiment process was:
• EMC pellets were taken out from the icebox The test
began only after 24 h defrosting and only 40 shots
were injected. Eight hours later, another 40 shots
were injected.
• At the beginning of each period, a new G2 specimen
was used. That is to say, defrosting period of EMC
is the only control factor specified.
• Six G2 specimens were used in this experiment and
the time interval was 8 h. The experiments were
repeated six times at different defrosting periods, that
is, EMC was molded after 24 h, 32 h, 40 h, 48 h,
56 h and 64 h defrosting before the test.
EXPERIMENT RESULTS
The experiment results are shown in Figs 6–11. The
average adhesion force and the standard deviation of
the experiments for each interval are shown in Table 2.
Figure 9. Results of the first experiment of EMC defrosted for 48 h. This
figure is available in colour online at www.apjChemEng.com.
Figure 10. Results of the first experiment of EMC defrosted for 56 h. This
figure is available in colour online at www.apjChemEng.com.
Figure 11. Results of the first experiment of EMC defrosted for 64 h. This
figure is available in colour online at www.apjChemEng.com.
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
EFFECTS OF DEFROSTING PERIOD ON EPOXY MOLDING COMPOUND
The tendency picture of average adhesion force for
all experiments is shown in Fig. 12. It was shown
that as the defrosting time period of EMC gets longer
the mold adhesion force increases. The average mold
adhesion force difference between the 1 day (24 h) and
2 days (48 h) defrosting cases of EMC is 8 kgf. The
mold adhesion force difference between the 24 h and
64 h defrosting cases of EMC is 12.45 kgf. Take the
average adhesion force of EMC defrosted for 24 h as
the base point (0%). The percentage increase of average
adhesion force for other defrosting time period cases
is shown in the fourth row of Table 2. The average
adhesion force of EMC defrosted for 64 h had already
increased by about 38%. We infer that the increase of
adhesion force after increased defrosting time period
was caused by moisture absorption in EMC. In order to
verify this inference, a set of confirmation experiments
was executed.
Table 3. Comparison of the average adhesion force for
the first and second experiments.
Defrosting
time
period (h)
Avg. of 1st
exp. (kgf)
Avg. of 2nd
exp. (kgf)
24
32
40
48
56
64
32.77 33.22 36.96 40.77 41.79 45.22
32.48 32.79 35.36 37.66 36.17 42.46
RESULTS OF CONFIRMATION EXPERIMENT
In order to verify whether the increase of adhesion force
for increased defrosting time period cases was caused
by moisture absorption in EMC, a set of confirmation
experiments was executed. This time, we took EMC
Table 2.
Average adhesion force and standard
deviation of experiments for each defrosting time
period.
Defrosting
time
period (h)
24
32
40
48
56
64
Average
32.77 33.22 36.96 40.77 41.79 45.22
adhesion
force (kgf)
Standard
3.82 3.44 3.73 3.93 4.38 5.32
deviation
(kgf)
Percentage of
0
1.37 12.79 24.41 27.53 37.99
increase (%)
Figure 12. Tendencies of average adhesion force for the
first experiments. This figure is available in colour online at
www.apjChemEng.com.
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
Comparison tendencies of average adhesion
force for the two experiments. This figure is available in
colour online at www.apjChemEng.com.
Figure 13.
pellets out from the icebox and put into a moisturecontrolled box to defrost. The relative humidity and
temperature of the moisture-controlled box were set
as 25% and 23 ± 1 ◦ C, respectively. The humidity of
the moisture-controlled box is lower than the humidity
of the environment. After 24 h defrosting, the test
began. All conditions, parameters and procedures of
the experiments were the same as the first experiments.
This paper is focused on the relationship between
the defrosting time of EMC and the mold adhesion
force. Thus, the mechanism of moisture influence is
not discussed here. The comparison data of the average
adhesion force for the two experiments are shown in
Table 3. The compared tendency of average adhesion
force for the two experiments is shown in Fig. 13.
According to the results in Fig. 13 and Table 3, we
found that the average value of adhesion force of
the second experiments is lower than those of the
first experiments. The small difference between these
two results can be taken as the experiment error. The
data of the second experiments also verify that as the
defrosting time of EMC gets longer, mold adhesion
force increases. This result shows that moisture really
has influence on the mold adhesion force. The influence
of moisture on the adhesion force during defrosting
is apparent. As the defrosting time gets longer, EMC
absorbs air moisture and thus increases the mold
adhesion force.
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
DOI: 10.1002/apj
167
168
H.-Z. CHEN ET AL.
Asia-Pacific Journal of Chemical Engineering
CONCLUSIONS
In this paper, an effort was taken to show the modification of the adhesion force test machine and the mold.
This new machine is semiautomatic and can measure the
adhesion accurately. The performance of the machine is
also very stable. In addition, the effects of the defrosting time on the mold adhesion force between EMC and
mold surface were found in the defrosting experiments
of EMC under different defrosting time periods. The
results show that the longer the defrosting time period,
the larger the mold adhesion effect. Part of the increase
of the adhesion force during defrosting is due to moisture. Therefore, it is best to use the EMC with 24–32 h
of defrosting to prevent excessive amount of mold adhesion force. In future, molding engineers can follow this
rule to proceed with molding.
Industry in Kaohsiung, Taiwan. The authors would like
to express their appreciation for the support.
REFERENCES
[1] S.-J. Chang, S.-J. Hwang. IEEE Trans. Electron. Packag.
Manuf., 2003; 26, 281–285.
[2] S.-J. Chang. Measurement and analysis of adhesion force on
IC encapsulation mold. PhD thesis, Department of Mechanical
Engineering National Cheng Kung University, Taiwan, 2004.
[3] ASTM Standard Test Method for Tensile Properties of Adhesive
Bonds, STD. ASTM standards Vol. 15.06, D897-95a.
[4] ASTM Standard Test Method for Apparent Shear Strength
of Single-Lap-Joint Adhesively Bonded Metal Specimens by
Tension Loading, STD. ASTM standards Vol. 15.06, D1002-94.
[5] ASTM Standard Test Method for Cleavage Strength of Metalto-Metal Adhesive Bonds, STD. ASTM standards Vol. 15.06,
D1062-96.
[6] J.U. Duncombe. IEEE Trans. Electron Devices, 1959; ED-11,
34–39.
[7] M. Yoshii, Y. Mizukami, H. Shoji. Hitachi Chem. Tech. Rep,
2003; 40, 13–20.
Acknowledgment
Part of this project was financially supported by Metal
Industries Research and Development Center in Kaohsiung, Taiwan and Philips Electronic Building Elements
 2008 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2009; 4: 161–168
DOI: 10.1002/apj
Документ
Категория
Без категории
Просмотров
0
Размер файла
397 Кб
Теги
effect, compounds, period, mold, defrosting, force, molding, adhesion, epoxy
1/--страниц
Пожаловаться на содержимое документа