IN SITU STRUCTURAL ANALYSIS OF BPDA-PPD POLYIMIDE THIN FILM USING TWO-DIMENSITIONAL GRAZING INCIDENCE X-RAY DIFFRACTION

Size: px
Start display at page:

Download "IN SITU STRUCTURAL ANALYSIS OF BPDA-PPD POLYIMIDE THIN FILM USING TWO-DIMENSITIONAL GRAZING INCIDENCE X-RAY DIFFRACTION"

Transcription

1 IN SITU STRUCTURAL ANALYSIS OF BPDA-PPD POLYIMIDE THIN FILM USING TWO-DIMENSITIONAL GRAZING INCIDENCE X-RAY DIFFRACTION 150 J. Kikuma, 1 T. Nayuki, 1 T. Ishikawa, 1 G. Asano, 2 and S. Matsuno 1 1 Analysis and Simulation Center, Asahi-KASEI Corporation, Shizuoka, Japan 2 Marketing Center, FPC/FPD Materials, Asahi-KASEI Corporation, Shizuoka, Japan ABSTRACT Structural development of BPDA-PPD polyimide thin film has been investigated by in situ grazing incidence X-ray diffraction (GIXD) at the BL24XU beamline of the SPring-8. Optimizing the sample shape, two-dimensional images were measured successfully without sacrificing angle resolution. It has been clearly shown that the crystallization first begins in the in-plane direction, at the curing temperature of 180 C, in which the periodic structure of the molecular chain axis (c axis) is developed. The crystallization in the surface normal (out-of-plane) direction is observed later, at the curing temperature above 300 C. A slight increase of the d-spacing of the c axis during heating process has been observed, suggesting the stretching of the contracted molecular chain in accordance with the curing process. In the cooling process, the decrease of the d-spacings for a and b axes was considerable, which indicates thermal expansion of the crystals at high temperatures. The increases in the peak intensities during the cooling process have been observed, which indicate the d-spacing of each axis becomes close to the equilibrium value to produce higher periodicity. INTRODUCTION Chemical and structural changes during the curing processes of polyimides have been of great interest for decades. Although a lot of studies have been carried out for crystal structure and growth along with imidizations, there are only a few studies that investigate the curing process of polyimides in situ [1,2]. It is well known that a two-dimensional (2D) detector is a powerful tool for pursuing fast changes of the materials under any kind of processing. However, it is difficult to apply a 2D detector to a GIXD measurement [3 5] because the incident beam is spread in a large area on the sample, and this causes a severe loss of angular resolution. On the other hand, a combination of the GIXD and a 2D detector has a great advantage of being able to acquire in-plane and out-of-plane diffraction patterns simultaneously. It is also important for in situ analysis that the entire 2θ region can be measured without any time lag, compared to the counter-scanning method. We have investigated the possibility of applying the 2D GIXD method to the in situ measurement of polyimide curing process, and it turns out that, by optimizing the sample shape, it is possible to obtain diffraction patterns without sacrificing angle resolution. Using this method, the structural evolution of BPDA-PPD poly(amic acid) film has been investigated.

2 This document was presented at the Denver X-ray Conference (DXC) on Applications of X-ray Analysis. Sponsored by the International Centre for Diffraction Data (ICDD). This document is provided by ICDD in cooperation with the authors and presenters of the DXC for the express purpose of educating the scientific community. All copyrights for the document are retained by ICDD. Usage is restricted for the purposes of education and scientific research. DXC Website ICDD Website -

3 EXPERIMENTAL 151 Materials and sample preparation Biphenyltetracarboxylic dianhydride (BPDA) and p-phenylene diamine (PPD) were purchased from Mitsubishi Chemical Corporation (Tokyo, Japan) and Seiko Chemical Company (Tokyo, Japan), respectively. Poly(p-phenylene biphenyltetracarboxamic acid) (BPDA-PPD PAA) was prepared by adding BPDA to the PPD solution in N-methyl-2-pyrrolidone (NMP). The solutions were kept at 70 C for 4 h for polymerization. The film samples were prepared by spin-coating the BPDA-PPD PAA solutions onto silicon wafers, to obtain approximately 10 µm thick cured films. The wafer sample was pre-cured in an oven at 100 C for 30 min prior to the XRD measurement. GIXD measurements GIXD measurements were carried out at the BL24XU beamline of the SPring-8, using an X-ray wavelength of nm. A hot stage with a plate heater (Anton Paar, DHS900) was set on the goniometer stage, and a helium purge box was set around the hot stage in order to minimize the air scattered backgrounds. The X-ray beam size was set to be 1 mm wide and 0.3 mm high. An imaging plate was placed downstream at the sample-to-detector distance of mm, which was determined by measuring a poly-si film on a Si substrate. The pre-cured film sample was cut into 7 mm width rectangular shape, which is the optimum sample width in terms of a balance of the sufficient signal intensity and the angle resolution. GIXD measurements were carried out during both heating and cooling processes under a helium atmosphere with an incident angle of 0.5, which probes the entire thickness of the sample. During the heating process, the sample temperature was elevated at the ramping rate of 5 C/min. Temperature was kept constant for 60 min at 140 C, and for 30 min at 180 C, 220 C, 300 C, 350 C, and 400 C. The cooling rate was 10 C/min without temperature holding time. GIXD images were obtained either one or two times in each curing step and at several points in the cooling process, with an exposure time of 60 s. The background image was obtained by measuring the Si substrate without the polyimide film, and subtracted from each image data after absorption correction. It should be emphasized that both in-plane and out-of-plane GIXD patterns can be obtained simultaneously from a single image. RESULTS AND DISCUSSION Figure 1 shows a series of 2D GIXD images of the film prepared from NMP solution. The structural development during the curing process is clearly recognized from these images. Before heating, only an amorphous halo was observed, as expected for the precursor film. In the heating process, a pair of small arcs, corresponding to (004) reflection (see Figure 2 for assignment), was first observed in in-plane direction at 180 C. At 300 C, broad structures in the out-of-plane direction started to be observed. At 400 C, these structures became several clear arcs, and other arcs started to be observed in the in-plane direction as well. It should be noted that all arcs became clearer and sharper in the cooling process, and this will be discussed later in detail.

4 152 Figure 1. A series of in situ 2D GIXD images of the BPDA-PDA polyimide film, obtained from NMP solution. From images in Figure 1, series of both in-plane and out-of-plane GIXD patterns were extracted. For in-plane patterns, sector averaged data of 82 to 84 and -82 to -84 were averaged again to make a single pattern. For out-of-plane patterns, sector averaged data of -2 to 2 were taken. The results are shown in Figures 2 and 3, respectively. In these figures, diffraction peaks were indexed in accordance with the ex situ measurement in the literature [6]. Figure 2. In-plane GIXD pattern derived from 2D images shown in Figure 1. Figure 3. Out-of-plane GIXD pattern derived from 2D images shown in Figure 1. For the in-plane patterns, as is also seen from the image data, (004) peak started to be observed at 180 C, and its intensity increased with elevating temperature. Other (00l) peaks such as (006) and (0010) were observed at higher temperatures. Because the in-plane pattern is dominated by (00l) peaks and the out-of-plane pattern is dominated by (hk0) peaks (Figure 3), molecular chains are

5 strongly oriented, namely, lying on the substrate. At 140 C, there is a small peak around 8, even though the imidization has not started at this temperature. This peak would be assigned as a periodic structure of BPDA-PPD PAA precursor, the length of which should be slightly larger than the c axis of the polyimide. Thus, it is suggested that the polymer molecule is oriented largely even before imidization. The shift of the (00l) peaks is not clearly visible in this figure but a slight shift is detected in the detailed analysis, and will be described later. 153 For out-of-plane patterns, diffraction peaks started to be observed at 300 C, and increased with increasing temperature. In the cooling process, the peaks became sharper and clearly shifted to higher angles, indicating that the thermal contraction in the out-of-plane direction is much more considerable than that in the in-plane direction. At this point, it should be mentioned that the angle resolution of these patterns is comparable to those measured with a scintillation counter [6]. Generally, crystal peaks of a polymer sample are intrinsically broad, so a 2D detector is able to give a sufficient angle resolution even in grazing incidence geometry. As a parameter of the degree of crystalinity for in-plane and out-of-plane directions, the (004) peak area and (210) peak area were evaluated, respectively. The results are shown in Figures 4 and 5. Other out-of-plane peaks showed similar trends to the data in Figure 5. As mentioned above, the peak areas increase not only in the heating process but also in the cooling process. This phenomenon is more considerable for the out-of-plane direction. Figure 4. The change of (004) peak area during the curing and cooling process. Peak area increases during the cooling process as well. Figure 5. The change of (210) peak area during the curing and cooling process. Peak area increases mainly during the cooling process. The changes of the d-spacings for (004) and (210) reflections are plotted in Figures 6 and 7, respectively. At an early stage of the heating process, the d-spacing for the (004) reflection is slightly smaller than the final value. It increases during the heating process, and becomes unchanged above 300 C and in the cooling process. On the other hand, d-spacing of the (210) reflection [and also those of other (hk0) reflections] is constant above 350 C but decreases in the cooling process.

6 154 Figure 6. The change of (004) d-spacings during the curing and cooling process. Figure 7. The change of (210) d-spacings during the curing and cooling process. These phenomena suggest the following mechanism for the structural development. At the early stage of the heating process, a periodic structure starts to be formed in the direction of the polymer chain axis (c axis) with a slightly compressed d-spacing, but the periodic structures in a and b axes direction are not developed firmly. At higher temperatures above 300 C, the polymer chain axis becomes stretched to be closer to an equilibrium d-spacing, and the crystal structures in a and b axes direction start to be developed with thermally expanded d-spacings. In the cooling process, the crystallites expanded in the a and b axes direction and become closely packed with the polymer chain axis that was stretched. The cause of the intensity increase in the cooling process is that the d-spacing of each axis becomes close to the equilibrium value to produce higher periodicity. Finally, we note that the 2D GIXD is a powerful tool for the in situ structural analysis of polymer thin films. Currently, an imaging plate has been used as a detector, however use of a CCD detector would enable us to perform faster measurement of the structural development of polymer films. ACKNOWLEDGMENTS The authors are grateful to Dr. T. Kawamura of NTT Basic Research Laboratories for his valuable comments and suggestions. This study was performed with the approval of JASRI (Proposal Nos. 2006B3217 and 2007A3217). REFERENCES [1] Ree, M.; Woo, S. H.; Shin, T. J.; Kim, K.; Chang, H.; Zin, W. C.; Lee, K.-B.; Park, Y. J. Macromol. Symp. 1997, 118, [2] Shin, T. J.; Lee, B.; Youn, H. S.; Lee, K.-B.; Ree, M. Langmuir 2001, 17, [3] Takagi, Y.; Kimura, M. J. Synchrotron Radiat. 1998, 5, [4] Kodjamanova, P.; Fietzek, H.; Juez-Lorenzo, M.; Kolarik, V.; Hattendorf, H. Mater. Sci. Forum 2006, , [5] Sasaki, S.; Masunaga, H.; Tajiri, H.; Inoue, K.; Okuda, H.; Noma, H.; Honda, K.; Takahara, A.; Tanaka, M. J. Appl. Crystallogr. 2007, 40, s642 s644. [6] Ree, M.; Kim, K.; Woo, S. H.; Chang, H. J. Appl. Phys. 1997, 81,