1. Introduction
Polyimides have many applications due to their excellent thermal, electrical and mechanical properties. In electronic packaging with multi-level interconnections, such as wafer-level chip scale packaging, polyimides are widely used as dielectric materials. The related processes include the first polyimide layer curing, copper trace electroplating and the second polyimide layer curing. However, one concern in the application of polyimides is stability, because polyimides have to go through a cure or reflow process several times at relatively high temperatures in the fabrication process of multi-level interconnections.
The curing process is critical for polyimides. Russell studied the effects of curing temperature on crosslinking by measuring the dynamic mechanical analysis (DMA) properties, and indicated that crosslinking occurs above 300 ℃ for the polyimide: AFRIOOB. The samples were heated to isotherms from 275 to 400 ℃ at a rate of 2 ℃/min over a duration from 1 to 16 h. It was found that crosslinking via the reverse Diels-Alder reaction occurs up to 350 ℃[1].
Rich studied the effects of curing on the removal of solvents and the development of chemical crosslinks[2]. Sasaki performed a detailed cure study on oligomers by staged heating to final cure temperatures of 370, 400, 420, and 450 ℃, and DMA showed an increase with temperature and a decrease in the drop of storage modulus (E) with increasing post-cure temperature[3]. Many researchers have focused on the stress measurements of the polyimide during the cure process.
The stability of polyimides can be examined by dynamic mechanical analysis. The stability of electron beam irradiated polyimides has been examined by measuring the glass relaxations of polyimides through dynamical mechanical analysis[4]. The stability of proton beam irradiated polyimides has also been reported[5].
However, very few efforts have been made on the effects of curing temperature history on the stability of polyimides. In general, the polyimide cure process compromises many factors. In electronic packaging with multi-level interconnections, the stability is especially of concern. To our knowledge, there are no published reports on the influence of curing temperature on the above mentioned properties of PMDA-ODA film. These properties will help understand the effects of the curing process, such as the ramp rate of the temperature, curing time etc, on the stability of the polyimides, and thus the results will help fabricate electronic packaging with highly reliable multi-level interconnections.
In this paper, the effects of the curing process on the stability of the polyimide film (PMDA-ODA) were studied using DMA. DMA is the most accurate technique for the study of the glass relaxation and relaxation processes in polyimides. The tests focus on the shift in the glass relaxation peak due to different curing processes. The goal of this research is to determine the effects of the curing process on the stability of polyimide films.
2. Experimental
2.1 Materials
To prepare the polyimide films, polyimidic solution (HD4100, HD MicroSystems) was coated onto 4 inch glass substrates.The spin speed was 2000 rpms and the spin time was 30 s. Then the coated substrate was soft baked on hot plate at 90 ℃ for 100 s and 100 ℃ for 100 s. After that, the coated substrate was cured under 11 different curing conditions in nitrogen atmosphere in a reflow oven (SRO-702/704), as shown in Table 1.
![]() |
These may be divided into three groups, namely A, D, G, H and I, which differ with the heat treatment time at 375 ℃ (30, 60, 90, 120 and 150 min), then A, E, F, J and K, which differ by the time at 200 ℃ (again 30, 60, 90, 120 and 150 min), and finally A, B, and C, which differ in the rate of heating (2, 5-7 and 10 ℃/min). The evaluation of the effect of changing curing times at one temperature was carried out having fixed the time in the other curing step.
After curing, the polyimide and glass substrates were diced into 35
2.2 Measurement
The dynamic mechanical analysis of the polyimide film was performed with film: tension modeled by a DMA Q 800 dynamic mechanical analyzer (TA Instruments, Inc., USA). The temperature range was from 40 to 400 ℃, and the polyimide film samples were heated at a ramp rate of 3 ℃/min in air at a frequency of 1 Hz. The glass transition temperature is defined as the high temperature peak in the tan
3. Results and discussion
The typical DMA results are shown with three curves: storage modulus, loss modulus, and tan
Generally, there are three relaxation processes: the low temperature and medium temperature relaxation processes are defined as the
As shown in Fig. 2, the tan
The low temperature peak may originate either from the contribution of adsorbed water molecules or
The effects of the cure temperature history on the
3.1 Polyimide synthesization
Polyimide is synthesized by a two-step reaction, as shown in Fig. 3. In this work, PMDA-ODA polyamic acid (PAA) solution is coated onto the substrate. The second step reaction occurs during the cure process.
Chemical structure changes occur during the cure process, which affect the
3.2 The effects on glass transition temperature
3.2.1 The effect of temperature ramp rate
The effect of temperature ramp rate on
The ramp rate related properties may be attributed to an incomplete reaction, or residual solvent or photoproducts, which could act as plasticizers. A higher ramp rate means a shorter curing time, which will lead to an incomplete reaction, which means a lower cure degree and lower glass transition temperature.
Meanwhile, the solvent may not evaporate completely at a low ramp rate for relatively thick films. For relatively thin films, the solvent could evaporate completely at a low ramp rate, as shown in Fig. 5(a), but for relatively thick films, a low ramp rate may form a cured film at the top first, which prevents the evaporation of the solution at the bottom. Thus, the solvent may not evaporate completely at a low ramp rate for relatively thick films, as shown in Fig. 5(b). The incomplete cured films result in a lower cure degree and lower glass transition temperature. Thus, a proper temperature ramp rate is beneficial for high glass transition temperature.
3.2.2 The effect of constant temperature time at 200 and 375 ℃
The effects of time at 200 ℃ and 375 ℃ on the glass transition temperature are shown in Fig. 6.
According to Fig. 6, for different times at 200 ℃, 90 min at 200 ℃ (Process F) results in the highest glass transition temperature. For different times at 375 ℃, 120 min at 375 ℃ (Process H) results in the highest glass transition temperature. 90 min at 375 ℃ (Process G) results in a very close glass transition temperature. Besides, it also reveals that the glass transition temperature is more sensitive to constant temperature time at 375 ℃ than that at 200 ℃.
The results can be explained by the relationship of cure degree and
But after complete reaction, the longer time at 200 ℃ or 375 ℃ results in a
TG(MN)=T∞G−KMN, |
(1) |
where
4. Conclusions
The effects of cure temperature history on the stability of hinged structure poly (4, 4-oxydiphenylene pyromellitimide) (PMDA-ODA) polyimide were studied by dynamic mechanical analysis. The polyimide films were cured under different curing conditions and peeled off by substrate etching. It was found that a proper cure time and temperature ramp rate improves the stability, in terms of higher glass transition temperature. 90 min at 375 ℃ or 200 ℃ is beneficial for high glass relaxation. The temperature ramp rate should be between 2 ℃/min and 10 ℃/min, which is neither too high nor too low.
Acknowledgements: The authors would like to extend their heartfelt gratitude to Dr. Cheng Yuanrong, Dr. Yang Jun and Dr. Pan Qilin of the Department of Materials Science, Fudan University, Shanghai, China for the DMA analysis and useful discussions. The authors would also like to express sincere gratitude to M. A Kehui Guo for English editing.