- Email: [email protected]
Characterization of fluorinated polyimide morphology by transition mechanical analysis
Polymer 59 (2015) 200e206
Contents lists available at ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
Characterization of fluorinated polyimide morphology by transition mechanical analysis Youngsuk Jung a, *, Yooseong Yang b, Seungyeon Lee a, Sunjung Byun a, Hyunjeong Jeon c, Myung Dong Cho b a b c
Analytical Science Group, Samsung Advanced Institute of Technology (SAIT), Suwon, Gyeonggi 443-803, Republic of Korea Energy Material Laboratory, Samsung Advanced Institute of Technology (SAIT), Suwon, Gyeonggi 443-803, Republic of Korea Film Material Laboratory, Samsung Advanced Institute of Technology (SAIT), Suwon, Gyeonggi 443-803, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 October 2014 Received in revised form 31 December 2014 Accepted 3 January 2015 Available online 9 January 2015
We report the thermal transition and concomitant morphology change of polyimide films, investigated through dynamic mechanical analyses, thermo-mechanical analyses, and X-ray diffraction spectroscopy. For this study, highly transparent polyimide films composed of 3,4,30 ,40 -biphenyltetracarboxylic dianhydride and fluorinated diamines, 2,20 -bis(trifluoromethyl)-4,40 -diaminobiphenyl, are prepared through a solvent-casting method. The films are annealed at different temperatures, and two kinds of reaction solvents (n-methyl-2-pyrrolidone and N,N-dimethylacetamide) for the polyimide precursors are compared. The mechanical analyses during thermal transition along with spectroscopic measurements reveal unusual transition behaviors and morphology differences between the polyimide films at various annealing temperatures. The results can be applied to the quantitative analysis of the imidization ratio in highly imidized polyimides when the films are imidized thermally at temperatures above 300 C. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Fluorinated polyimide Transition mechanical analysis Glass transition
1. Introduction Lightness and better portability are the most desired features in current mobile display devices [1e5]. However, the most widely used glass substrates in such devices lack lightness, flexibility, and thinness, which are necessary features for display technologies in the near future. Furthermore, the fragile nature of glass substrates means they have low durability, despite recent improvements in the impact strength of tempered glass [4,5]. Therefore, it is necessary to develop flexible plastic substrates with thermal and optical properties as good as those found in inorganic substrates such as glass or silicon wafer, in order to utilize the other various advantages they can provide [6e8]. For application in electronic devices, a high-temperature fabrication process is generally required to obtain high reliability in the manufacturing process. Such process conditions mean that common transparent plastics such as polyethylene naphthalates, poly(ethylene terephthalate), poly(ether sulfones), and polycarbonates are less suitable for these devices [7,8]. In contrast, the
* Corresponding author. Tel.: þ82 312808448; fax: þ82 312806739. E-mail address: [email protected] (Y. Jung). http://dx.doi.org/10.1016/j.polymer.2015.01.007 0032-3861/© 2015 Elsevier Ltd. All rights reserved.
thermal stability of polyimide films probably makes them the most promising candidate substrates for the future electronic devices [7e10]. In particular, the recent marked decrease in the problem of their yellowish color as well as their well-known properties such as light weight, flexibility, and chemical inertness further increases their applicability [11e17]. The well-known charge-transfer complex (CTC) formations are lessened markedly in polyimides with fluorinated functional groups, which weaken their intra- or inter-molecular interactions. In 2005, The DuPont company patented fluorinated polyimides composed of two dianhydrides, 3,4,30 ,40 -biphenyltetracarboxylic dianhydride (BPDA) and 2,20 -bis(3,4-dicarboxyphenyl)-hexafluoropropane dianhydride (6FDA), and a diamine, 2,20 -bis(trifluoromethyl)-4,40 -diaminobiphenyl (TFDB) [11]. Here, 6FDA, with its bulky trifluoro side groups, enhances transparency but reduces thermal stability [11,12]. Also, the fluorinated side groups in 6FDA and TFDB give the materials an unusual thermo-mechanical behavior that originates from moderate CTC formation and imidization [12e17]. These internal molecular re-organizations are known to make quantitative analyses difficult. In particular, when the heating temperature is higher than z300 C, estimation of the degree of imidization is difficult [18e23]. The application of organic materials or polymers to electronic devices needs comparable
Y. Jung et al. / Polymer 59 (2015) 200e206
thermal and chemical stabilities equivalent to those of inorganic materials, especially in terms of impurities generated through device operation. This requirement is related to the device's reliability and lifetime. In this regard, study of the dependence of polymer morphology on process conditions, such as thermal transition behavior and degree of imidization (DoI) in polyimides, is critical for determining the appropriate process conditions for organic materials. Herein, we report the thermal transition behavior, structural change during imidization, and CTC formation in a fluorinated polyimide (FPI), the chemical structure of which is shown in Fig. 1a, with different solvents and with various curing temperatures. Along with X-ray diffraction measurements, the dependences of the thermal behaviors on the change in polymer morphology are revealed through transition mechanical analyses, which denote dynamic and mechanical analyses during thermal transition [24e27]. The change in molecular packing structure correlated to resistance strengths obtained through transition mechanical analyses provides distinguishable information on CTC formation and imidization, even in a highly imidized film. Also, the amount of residual solvent according to the annealing temperature is measured for two solvents commonly used for polyimide film fabrication, that is, N-methyl-2-pyrrolidone (NMP) and N,Ndimethylacetamide (DMAc), to determine the solvent effect on the thermal behavior of polyimide films. 2. Experimental 2.1. Polymer solution preparation Polyimide precursor solutions (15 wt-%) were prepared from BPDA (Fig. 1b) and TFDB (Fig. 1c) in NMP (Fig. 1d) or DMAc (Fig. 1e). The solvents and monomers were purchased from Sigma Aldrich Co. BPDA (z5.2 g) (z4.7 g for DMAc) was agitated in a 250 mL round-bottomed flask in the presence of NMP or DMAc (60 mL) for about 1 h at 0 C and for a further 5 h at 25 C. Subsequently, TFDB
201
(z5.7 g) (z5.2 g for DMAc) was added, and the solution was stirred for 17 h. All the reactions were controlled by the reflux method under a regulated N2 purge to maintain the concentration of the reaction solution. 2.2. Film fabrication As introduced in a previous report [13], a flow-coating technique was employed to prepare two series of five films of homogeneous thickness. One series was fabricated from NMP solution and the other from DMAc solution. The poly(amic acid) (PAA) polymer solutions were applied to glass substrates temporarily adhered to a custom-built, motorized translation stage. The film thickness was controlled by adjusting the lip gap of a metal blade to z300 mm. After casting of the solution, the heating stage was set to 50 C for 1 h for leveling, and 80 C for 1 h to remove the solvent under a N2 flow. The films on the glass substrates were released with initiation of the removal at one edge of the film with a razor blade, and then fixed into a square-shaped metal frame to expose both sides of the film. Finally, the films in tenter frames were thermally heated up to 300, 350, 380, 400, 430, and 450 C in each series of solvent at a ramping rate of 2 C/min under a N2 purge. The thermally cured film thickness was in the range 45e55 mm. 2.3. Thermal and chemical analyses The amount of residual solvent and the evaporation rate of the solvent were measured with a TA Instruments TGA Q5000IR. The amount of water generated during imidization and the amount of residual solvent according to heating temperature were also measured with an Agilent GCeMS 6890/5973 system with a pyrolyzer and UA-1 Column. The sample was heated from 80 C to 500 C at a heating rate of 10 C/min under a He flow. A TA Instruments DSC Q2000 was used to trace the thermal curing of the films, and the heating and cooling cycles were repeated to test for reversibility. Both the TGA and DSC measurements were performed with ramping rate of 10 C/min under a N2 purge. Chemical curing was also analyzed with a Bio Rad FTS-6000 Fourier-transform IR (FTIR) spectrometer. 2.4. Spectroscopic methods Optical properties such as the yellow index (YI) and UV-Vis absorption intensity (AI) were measured with a CM3600D spectrophotometer manufactured by KONICA MINOLTA Inc. XRD measurements were conducted using a laboratory-scale wide-angle Xray scattering instrument (Philips Xpert) with Cu Ka radiation in the conventional pinhole geometry. The samples were mounted in the specular geometry. The Incidence angle of the 40 KeV/40 mA powered X-ray was fixed, and a diffraction pattern was obtained through a detector (2q) scan. The optical retardation was derived from the difference between the in-plane and out-of-plane refractive index of the film, as measured with a METRICON Prism coupler 200P-1 equipped with a HeeNe laser source with a wavelength of 632.8 nm. 2.5. Thermo-dynamic mechanical analyses
Fig. 1. (a) Fluorinated polyimide, (b) 3,4,30 ,40 -biphenyltetracarboxylic dianhydride (BPDA), (c) 2,20 -bis(trifluoromethyl)-4,40 -diaminobiphenyl (TFDB), and (d) N-methyl2-pyrolidone (NMP), (e) N,N-dimethylacetamide (DMAc).
Films cut into rectangular shapes of around 5 mm 20 mm were used for DMA and TMA measurements. TA Instruments DMA Q800 and TMA Q400 were used for the investigation of the transition behavior and the evaluation of thermal strain, respectively. In DMA measurements, a strain of 0.1% and 1 Hz oscillation were used for measuring the storage and loss modulus change as a function of temperature. Samples were scanned up to 400 C at a heating rate
202
Y. Jung et al. / Polymer 59 (2015) 200e206
of 5 C/min under a N2 purge. For TMA measurements, the samples were heated up to 400 C or until the maximum dimension change of the instruments, z2.3 mm. 3. Results and discussion 3.1. Thermal imidization and residual solvents Appropriate thermal conditions are required in solvent-casting methods to obtain fluorinated polyimide films with high transparency and a high DoI. Here, two common solvents, NMP and DMAc, were used selectively [28e32]. Fig. 2a shows the DSC thermogram of a film made from the solution with DMAc and thermally cured up to 300 C. At the first heating scan, a broad endotherm was measured around 350 C. This indicates that a certain amount of heat is supplied to the films to remove the residual solvent and to finish the thermal imidization that releases H2O as a byproduct. However, upon the second heating after the first scan up to 400 C, no distinguishable endotherms are observed. It can be supposed that thermal imidization was completed after heating at around 400 C. The amount of residual solvent, monitored by TGA, also shows that the thermal imidization could be finished at around 400 C (Fig. 2b). Despite the more volatile nature of DMAc than NMP, as shown in Fig. 2c, a larger amount of DMAc remains in the heated films, as shown by the initial amount of solvent in Fig. 2b. This is consistent with a previous report by Ebisawa et al. [30]. In this report, the C]O stretching band of hydrogen-bonded DMAc with the COOH groups of polymer chains is activated by lower energy infrared rays. This indicates stronger hydrogen bonding of DMAc to the polymer chains than the case of NMP. However, as the heating temperature approaches 400 C, the amount of residual solvents converges to 0 wt-%. According to EGA-MS result in Fig. 2d, a relatively small amount of solvent and H2O comes out from the films until the temperature slightly higher than 400 C. H2O is the byproduct of imidization and the solvent was hydrogen-bonded to
polymer chains prior to imidization. Considering the diffusion time of the trapped H2O and the organic solvent, the imidization seems to be completed at around 400 C in BTDA-TFDB PAA. On the basis of the results from DSC, TGA, and EGA-MS, a series of films heated up to 300, 350, 380, 400, 430, and 450 C were prepared to explore the change in polymer morphology in highly imidized polyimide films and to evaluate their correct imidization ratio. 3.2. Color change during imidizaiton and chargeetransfer complex formation The DoI is usually measured through one of several methods including FT-IR [18e23] and the density measurement method [33]. However, the DoI is rarely quantified by conventional methods when it is higher than z95%, as mentioned above. In FT-IR, the DoI can be calculated through the comparison of the stationary peak and changeable peak area of the imide rings with imidization ratio. Fig. S1 in the supplementary data shows the FT-IR spectra for BTDATFDB polyimide films with heating temperatures of 300, 380, and 450 C. Here, the DoI can be estimated through a comparison of the CeNeC peak area or peak intensity with the fixed CeF stretching mode as the standard stationary peak. In contrast to the quantitative results of TGA, DSC, and EGA-MS, FT-IR does show almost identical spectra for the films heated at 300 C and 450 C. In this regard, more ideal methods are necessary to obtain the DoI in highly imidized polyimides. The film color or extinction coefficient can be an index for the DoI, although they are much more affected by the amount of CTC formation [13,34]. The transparency (Tr), yellow index (YI), and film-thickness-normalized absorption intensity (AI) are summarized in Table 1. Tr and YI (AI) show opposite trends. YI is a metric in the display industry, and it represents how yellow a polyimide film is. However, this index does not increase linearly and shows saturation behavior, as detailed in our previous report [13]. More conventionally, the extinction coefficient at yellow wavelengths is
Fig. 2. (a) DSC thermograms for an FPI film heat treated up to 300 C. (b) TGA thermograms for remaining solvent amounts as a function of temperature for FPI films heat treated up to 300 C. (c) Remaining solvent amounts and solvent evaporation speed as a function of time at 80 C. (d) EGA-MS profile for the DMAc solvent and water in an FPI film released from a glass substrate at 80 C.
Y. Jung et al. / Polymer 59 (2015) 200e206
203
Table 1 Tr, YI, and AI as a function of ANN temperature.
Heating temperature ( C)
From NMP solution Tra
300 350 380 400 430 450 a b
88.4 87.1 84.5 82.2 81.5 80.9
From DMAc solution AIb
YI ± ± ± ± ± ±
0.2 0.1 0.2 0.2 0.4 0.3
2.2 3.7 12.7 15.8 18.8 21.5
± ± ± ± ± ±
0.2 0.4 0.8 0.7 0.6 1.1
Tra
0.074 0.084 0.163 0.196 0.337 0.366
± ± ± ± ± ±
0.006 0.005 0.021 0.032 0.030 0.027
88.5 87.1 85.8 82.6 82.1 80.7
AIb
YI ± ± ± ± ± ±
0.1 0.3 0.2 0.4 0.4 0.2
2.1 2.8 5.6 12.6 17.6 19.9
± ± ± ± ± ±
0.2 0.3 0.5 0.7 0.8 0.6
0.073 0.075 0.117 0.164 0.351 0.389
± ± ± ± ± ±
0.008 0.009 0.020 0.017 0.019 0.020
cf. Tr for slide glass z92.0%. Film thickness normalized absorption intensity at 450 nm wavelength.
used to express the extent of color. We have used the filmthickness-normalized AI, which is consistent with the extinction coefficient in the BeereLambert law [35]. The yellowness is proportional to CTC formation, which can be intra-chain or inter-chain in nature. Besides, the CTC formation is relevant to imidization. The majority of imidization may be intra-molecular owing to the low mobility of the polymer chains in the film when a small amount of solvent is left, as shown in Fig. 1b. In Table 1, Both YI and AI increase as the heating temperature is raised, and there are marked increases at z380 C and z430 C. Here, the contributions from imidization, intramolecular CTC formation, and intermolecular CTC formation need to be distinguished.
marked increases in YI and AI. This could be used to make a distinction between imidization and intermolecular CTC formation. Thus, it is presumed that the a-transition may be affected by intermolecular CTC formation after completion of the imidization.
3.3. Transition behavior of fluorinated polyimides To examine the molecular reconfiguration such as imidization and CTC formation, we employed DMA, which can explore the thermal transition of FPI. Fig. 3a shows a DMA thermogram of a film heated at 300 C. As the temperature increases, the storage modulus (E0 ) decreases gradually but distinctly from 150 C to 300 C, and drops markedly at z320 C. The increase in viscoelasticity leads to broad peaks of the loss modulus (E00 ) at z250 C and z320 C, which are seen more clearly in the tand curve. The lower-temperature bump and higher-temperature peak are known transitions due to the rotation about the CeC bond in benzidine (btransition) and the segmental motion of the backbone (a-transition), respectively [15,16,36,37]. The distinct decrease in E0 during the b-transition renders the film fragile in the thermal drying process. Concerning the a-transition, we found interesting transition behavior unlike that reported previously for the glass transition of polyimide films [36,37]. Fig. 3b shows tand curves for films made from NMP solution at various heating temperatures. All the films show similar b-transition temperatures. However, their atransitions shift to lower-temperatures and the peak intensity, dynamic loss ratio, decrease with increasing heating temperature, as marked with the arrows. Identical trends in the change in E00 are shown in Fig. S2. This indicates that the segmental motion of the polymer backbones observed in the films decreases after heating at higher-temperatures. More importantly, the a-transition may not be the well-known inherent glass transition of polyimide. It is supposed to be a transition that could be affected by changes in the molecular structures or that of the PAA precursors. DMA frequency sweep results showing that the transition does not depend on frequency also support the conclusion that the second transition is not an inherent property of polyimide, as shown in Fig. S3. In this regard, between the broad b-transition temperature and the atransition, intramolecular imidization and consequent intramolecular CTC formation are being completed. In addition, local rotation in the TFDB units is not affected by CTC formation. Films obtained from DMAc solution exhibited almost identical trends, as shown in Fig. 3c. According to the color-change results of the films, the temperature range after the a-transition corresponds to the
Fig. 3. (a) DMA thermograms for an FPI film heat treated up to 300 C. (b) Loss tand traces according to heat treatment temperature in FPI films from NMP solution. (c) Loss tand traces for FPI films from DMAc solution.
204
Y. Jung et al. / Polymer 59 (2015) 200e206
3.4. Morphology change in thermal transition region To correlate the changes in polymer morphology of the films and their optical properties during the thermal transition, X-ray diffraction patterns were measured, as shown in Fig. 4a and b. The FPI X-ray diffraction patterns are represented as a function of scattering vector, q ¼ (4p/l)sinq (l ¼ 1.54 Å). The d-spacing of a peak is 2p/q [38]. Fig. 4a and b correspond to the data for films from NMP solution and DMAc solution, respectively. For both series of films, four distinguishable features are observed, as designated in the figure from #1 to #4. All the films show small crystalline peaks, #2 (014) and #4 (006) [39,40]. Peak #3 is a background peak originating from the carbonate paste between the films and the sample stage. The background peak also contributes slightly to peak #4 because of the film thinness. Between q ¼ 0.8 and 2.0, the films heated up to 300 C and 350 C only show broad amorphous halos. However, the small bump 1a and peak 1b, which can be assigned to
(003) and (004), respectively [36,37,39,40], at around q ¼ 1.2 develop in the films heated above 380 C. This reveals the increase in intra-chain regularity. However, in the films heated at 450 C, the crystalline peaks from the regularity of intramolecular chains change to show a broad amorphous region, as indicated by peak #1. Therefore, the increase in the crystalline region at around 380e400 C and also the a-transition in the DMA measurements may come from the intramolecular imidization process. The square symbols in Fig. 4c show the change in the ratio of the crystalline region including both the intramolecular and the rare intermolecular chain regularity as a function of the heating temperature. Fig. S4 in the supporting materials shows the peak deconvolution for the crystallinity calculation. The circles show the variation in AI at a wavelength of 450 nm, as shown in Table 1. Here, the solid and empty symbols indicate films from NMP and DMAc solutions, respectively. The AI value increases with increasing heating temperature, and the amount of the crystalline region represents an inflection point between 400 C and 450 C. The AI that is relevant to the CTC formation reveals that intramolecular CTC formation develops markedly between 300 C and 400 C, as the imidization process in the polymer chain occurs. Furthermore, the noticeable AI increase between 400 C and 430 C may originate from intermolecular CTC formation, despite the reduction in the total crystalline region in the same temperature range. Thus, the physical property that changes upon intramolecular CTC formation between 300 C and 400 C can provide useful information about the DoI.
3.5. Evaluation of degree of imidization by transition mechanical analysis Thermo-mechanical analysis is known to have exceptional resolution compared with conventional calorimetry or spectroscopic methods in the observation of the thermal transition [41e43]. Here, we used TMA, which is one example method with superior resolution for the measurement of dimension changes. The expansion behavior in a film depends on the molecular orientation in the film, particularly the out-of-plane anisotropy [44]. In Fig. S5, the Rth (outof-plane retardation) values of the FPI films shows that the molecular anisotropy is consistent according to the heating temperature. Thus, the dimension change of the films may be affected mainly by molecular interactions. As an example, 300 C heated film is shown in Fig. 5a. Slope “a” can be related to the coefficient of thermal expansion, and slope “b” is similar to the heat deflection process in thermoplastics [45,46]. By using a proper extension force, thermal expansion higher than the transition temperature can be measured. Therefore, we used an external force of 0.05 N and obtained the thermal expansion behavior as a function of temperature in each solvent series of films, as shown in Fig. S6. We can define the resistance strength (RS) as the repulsive modulus against the external stress, similarly to the tensile modulus, which is expressed as follows:
E¼
s DL
(1)
Here, the external force is z0.050 N, and the film width and thickness are 5 mm and 50 mm, respectively. This indicates the stress is about 10 MPa. Thermal stress can also be expressed as follows:
s ¼ k$a$E$DT Fig. 4. (a) XRD patterns according to heat treatment temperature in FPI films from NMP solution. (b) XRD patterns for FPI films from DMAc solution. (c) Amount of crystalline region and AI according to heat treatment temperature.
E¼
s s ¼ k$a$DT k$Dl
(2) (3)
Y. Jung et al. / Polymer 59 (2015) 200e206
205
from the external stress and thermal strain of samples. Through mechanical analysis during the thermal transition of polyimide, the subtly different effects of the solvents, DMAc and NMP, have been distinguished clearly. It is confirmed that DMAc forms stronger hydrogen bonds than NMP with PAA molecules. This may result in a higher degree of molecular orientation and larger resistance strength in films from NMP solution than in those from DMAc solution. These results could provide a useful guideline for the determination of the imidization ratio, which is a decisive parameter for physical properties in the substrates of electronic devices. Acknowledgment Y.J. thanks Dr. Sangmo Kim at Organic Material Lab, SAIT, for useful discussion. Y.J. also thanks Daeun Yu at the Analytical Science Group, SAIT, for experimental support. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.polymer.2015.01.007. References [1] [2] [3] [4] [5] [6] [7] [8] [9] Fig. 5. (a) TMA thermogram for FPI film heat treated at 300 C. Slopes “a” and “b” are the dimension changes before and during the a-transition, respectively. (b) RS according to heat-treatment temperature. RS for FPI film treated at 450 C corresponds to the thermal strain before the a-transition, i.e., slope “a” is equal to slope “b”.
[10] [11]
[12] DL DT .
In the above equations, a denotes If aT is the expansion coefficient during the transition, as represented by slope “b” in DLT Fig. 5a,aT is DT , and then RS can be expressed as: T
Er ¼
se se $DTT ¼ k$aT $DLT k$DLT
(4)
Here, Er can be calculated from the constant external force or external stress (se Þ. Finally, the RS values calculated from Eq. (4) are shown in Fig. 5b. As explained in the previous section, the increase in RS up to 400 C can be attributed to imidization or intramolecular CTC formations. Accordingly, the DoI could be obtained through the relative RS by assuming that the DoI of the film heated at 300 C is z95%. Also, the DoI is substantially higher in the films made from NMP solution than in those from DMAc solution. 4. Conclusions Through quantitative chemical, thermal, and mechanical analyses, we have predicted that the imidization process is completed at around 400 C. In addition, we have found that the glass transition of polyimide is not an inherent property. It is apparent that the glass transition and change in modulus during the thermal transition depend critically on the film's heat-treatment conditions. Relevant imidization and CTC formation make the polyimide show variable transition behaviors. Thus, we have suggested a new evaluation standard of resistance strength, which could be defined
[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36]
Lewis J. Mater Today 2006;9:38e45. Hellemans A. IEEE Spectr 2000;37:18e21. Ong B. Laser Focus World 2004;40:85e8. McAlpine MC, Friedman RS, Lin SJK-h, Wang WU, Lieber CM. Nano Lett 2003;3:1531e5. Jazi AY, Yoon S, Liu J. J Ind Eng Chem Res 2012;51:10887e94. Nysaether JB, Lundstrom P, Liu J. IEEE Trans Packag Manuf Technol Part A 1998;21:281e7. Jang J. Mater Today 2006;9:46e52. MacDonald WA. J Mater Chem 2004;14:4e10. Bessonov MI, Zubkov VA. Polyamic acids and polyimides: synthesis, transformations, and structure. Boca Raton, Florida: CRC Press; 1993 [Chapter 1e3]. Wilson D, Stenzenberger HD, Hergenrother PM. Polyimides. New York: Chapman and Hall; 1990 [Chapter 3]. Simone CD, Auman BC, Carcia PF, Wessel RA. Low color polyimide compositions useful in optical type applications and methods and compositions relating thereto. U. S patent 7,550,194, Aug 3; 2005. Jung Y, Song S, Kim S, Yang Y, Chae J, Park T, et al. Appl Phys Lett 2013;102: 031602. Jung Y, Yang Y, Kim S, Kim H-S, Park T, Yoo BW. Eur Polym J 2013;49: 3642e50. Ando SJ. J Photo Sci Tech 2004;17:219e32. Feiring AE, Auman BC, Wonchoba ER. Macromolecules 1993;26:2779e84. Hougham G, Tesoro JG, Shaw J. Macromolecules 1994;27:3642e9. Xie K, Liu JG, Zhou HW, Zhang SY, He MH, Yang SY. Polymer 2001;42: 7267e74. Anthamatten M, Letts SA, Day K, Cook RC, Gies AP, Hamilton TP, et al. J Polym Sci Part A Polym Chem 2004;42:5999e6010. Saeed MB, Zhan M-S. Eur Polym J 2006;42:1844e54. Snyder RW, Thomson B, Bartges B, Czerniawski D, Painter PC. Macromolecules 1989;22:4166e72. Mustoa P, Ragostaa G, Scarinzia G, Mascia L. Polymer 2004;45:1697e706. Ree M, Yoon DY, Volksen W. J Polym Sci Part B Polym Phys 1991;29:1203e13. Shin TJ, Ree M. Langmuir 2005;21:6081e5. Jones DS. Int J Pharm 1999;179:167e78. Jung Y. Surface characteristics of polyfluorenes and oligofluorenes, and bulk characteristics of novel polyolefins. Ph D Thesis:Seoul, Korea. 2005. Ferry JD. Macromolecules 1991;24:5237e45. Han B, Guo Y. J Electron Packag 1995;117:185e91. Hasegawa M, Ishigami T, Ishii J, Sugiura T, Fujii M. Eur Polym J 2013;49: 3657e72. Hasegawa M, Matano T, Shindo Y, Sugimura T. Macromolecules 1996;29: 7897e909. Ebisawa S, Ishii J, Sato M, Vladimirov L, Hasegawa M. Eur Polym J 2010;46: 283e97. Hsiao S-H, Guo W, Chung C-L, Chen W-T. Eur Polym J 2010;46:1878e90. Yang C-P, Chen W-T. Macromolecules 1993;26:4865e71. Han SH, Misdan N, Kim S, Doherty CM, Hill AJ, Lee YM. Macromolecules 2010;43:7657e67. Hasegawa M, Horie K. Prog Polym Sci 2001;26:259e335. Gill SC, von Hippel PH. Anal Biochem 1989;182:319e26. Eashoo M, Shen D, Wu Z, Lee CJ, Harris FW, Cheng SZD. Polymer 1993;34: 3209e15.
206
Y. Jung et al. / Polymer 59 (2015) 200e206
[37] Bas C, Tamagna C, Pascal T, Alberola D. Polym Eng Sci 2003;43:344e55. [38] Chabinyc ML, Toney MF, Kline RJ, McCulloch I, Heeney M. J Am Chem Soc 2007;129:3226e37. [39] Cheng SZD, Wu Z, Eashoo M, Hsu SLC, Harris FW. Polymer 1991;32:1803e10. [40] Cheng SZD, Arnold Jr FE, Zhang A, Hsu SLC, Harris FW. Macromolecules 1991;24:5856e62. [41] Liu J, Cheng SZD, Harris FW, Hsiao BS, Gardner KH. Macromolecules 1994;27: 989e96.
[42] Zhao J-H, Ryan T, Ho PS, McKerrow AJ, Shih W-Y. J Appl Phys 1999;85: 6421e4. [43] Rhee S-H, Du Y, Ho PS. J Appl Phys 2003;93:3926e33. [44] Pottiger MT, Coburn JC, Edman JR. J Polym Sci Part B Polym Phys 1994;32: 825e37. [45] Factor BJ, Russell TP, Toney MF. Macromolecules 1993;26:2847e59. [46] Baeurle SA, Hotta A, Gusev AA. Polymer 2006;47:6243e53.
Contents lists available at ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
Characterization of fluorinated polyimide morphology by transition mechanical analysis Youngsuk Jung a, *, Yooseong Yang b, Seungyeon Lee a, Sunjung Byun a, Hyunjeong Jeon c, Myung Dong Cho b a b c
Analytical Science Group, Samsung Advanced Institute of Technology (SAIT), Suwon, Gyeonggi 443-803, Republic of Korea Energy Material Laboratory, Samsung Advanced Institute of Technology (SAIT), Suwon, Gyeonggi 443-803, Republic of Korea Film Material Laboratory, Samsung Advanced Institute of Technology (SAIT), Suwon, Gyeonggi 443-803, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 October 2014 Received in revised form 31 December 2014 Accepted 3 January 2015 Available online 9 January 2015
We report the thermal transition and concomitant morphology change of polyimide films, investigated through dynamic mechanical analyses, thermo-mechanical analyses, and X-ray diffraction spectroscopy. For this study, highly transparent polyimide films composed of 3,4,30 ,40 -biphenyltetracarboxylic dianhydride and fluorinated diamines, 2,20 -bis(trifluoromethyl)-4,40 -diaminobiphenyl, are prepared through a solvent-casting method. The films are annealed at different temperatures, and two kinds of reaction solvents (n-methyl-2-pyrrolidone and N,N-dimethylacetamide) for the polyimide precursors are compared. The mechanical analyses during thermal transition along with spectroscopic measurements reveal unusual transition behaviors and morphology differences between the polyimide films at various annealing temperatures. The results can be applied to the quantitative analysis of the imidization ratio in highly imidized polyimides when the films are imidized thermally at temperatures above 300 C. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Fluorinated polyimide Transition mechanical analysis Glass transition
1. Introduction Lightness and better portability are the most desired features in current mobile display devices [1e5]. However, the most widely used glass substrates in such devices lack lightness, flexibility, and thinness, which are necessary features for display technologies in the near future. Furthermore, the fragile nature of glass substrates means they have low durability, despite recent improvements in the impact strength of tempered glass [4,5]. Therefore, it is necessary to develop flexible plastic substrates with thermal and optical properties as good as those found in inorganic substrates such as glass or silicon wafer, in order to utilize the other various advantages they can provide [6e8]. For application in electronic devices, a high-temperature fabrication process is generally required to obtain high reliability in the manufacturing process. Such process conditions mean that common transparent plastics such as polyethylene naphthalates, poly(ethylene terephthalate), poly(ether sulfones), and polycarbonates are less suitable for these devices [7,8]. In contrast, the
* Corresponding author. Tel.: þ82 312808448; fax: þ82 312806739. E-mail address: [email protected] (Y. Jung). http://dx.doi.org/10.1016/j.polymer.2015.01.007 0032-3861/© 2015 Elsevier Ltd. All rights reserved.
thermal stability of polyimide films probably makes them the most promising candidate substrates for the future electronic devices [7e10]. In particular, the recent marked decrease in the problem of their yellowish color as well as their well-known properties such as light weight, flexibility, and chemical inertness further increases their applicability [11e17]. The well-known charge-transfer complex (CTC) formations are lessened markedly in polyimides with fluorinated functional groups, which weaken their intra- or inter-molecular interactions. In 2005, The DuPont company patented fluorinated polyimides composed of two dianhydrides, 3,4,30 ,40 -biphenyltetracarboxylic dianhydride (BPDA) and 2,20 -bis(3,4-dicarboxyphenyl)-hexafluoropropane dianhydride (6FDA), and a diamine, 2,20 -bis(trifluoromethyl)-4,40 -diaminobiphenyl (TFDB) [11]. Here, 6FDA, with its bulky trifluoro side groups, enhances transparency but reduces thermal stability [11,12]. Also, the fluorinated side groups in 6FDA and TFDB give the materials an unusual thermo-mechanical behavior that originates from moderate CTC formation and imidization [12e17]. These internal molecular re-organizations are known to make quantitative analyses difficult. In particular, when the heating temperature is higher than z300 C, estimation of the degree of imidization is difficult [18e23]. The application of organic materials or polymers to electronic devices needs comparable
Y. Jung et al. / Polymer 59 (2015) 200e206
thermal and chemical stabilities equivalent to those of inorganic materials, especially in terms of impurities generated through device operation. This requirement is related to the device's reliability and lifetime. In this regard, study of the dependence of polymer morphology on process conditions, such as thermal transition behavior and degree of imidization (DoI) in polyimides, is critical for determining the appropriate process conditions for organic materials. Herein, we report the thermal transition behavior, structural change during imidization, and CTC formation in a fluorinated polyimide (FPI), the chemical structure of which is shown in Fig. 1a, with different solvents and with various curing temperatures. Along with X-ray diffraction measurements, the dependences of the thermal behaviors on the change in polymer morphology are revealed through transition mechanical analyses, which denote dynamic and mechanical analyses during thermal transition [24e27]. The change in molecular packing structure correlated to resistance strengths obtained through transition mechanical analyses provides distinguishable information on CTC formation and imidization, even in a highly imidized film. Also, the amount of residual solvent according to the annealing temperature is measured for two solvents commonly used for polyimide film fabrication, that is, N-methyl-2-pyrrolidone (NMP) and N,Ndimethylacetamide (DMAc), to determine the solvent effect on the thermal behavior of polyimide films. 2. Experimental 2.1. Polymer solution preparation Polyimide precursor solutions (15 wt-%) were prepared from BPDA (Fig. 1b) and TFDB (Fig. 1c) in NMP (Fig. 1d) or DMAc (Fig. 1e). The solvents and monomers were purchased from Sigma Aldrich Co. BPDA (z5.2 g) (z4.7 g for DMAc) was agitated in a 250 mL round-bottomed flask in the presence of NMP or DMAc (60 mL) for about 1 h at 0 C and for a further 5 h at 25 C. Subsequently, TFDB
201
(z5.7 g) (z5.2 g for DMAc) was added, and the solution was stirred for 17 h. All the reactions were controlled by the reflux method under a regulated N2 purge to maintain the concentration of the reaction solution. 2.2. Film fabrication As introduced in a previous report [13], a flow-coating technique was employed to prepare two series of five films of homogeneous thickness. One series was fabricated from NMP solution and the other from DMAc solution. The poly(amic acid) (PAA) polymer solutions were applied to glass substrates temporarily adhered to a custom-built, motorized translation stage. The film thickness was controlled by adjusting the lip gap of a metal blade to z300 mm. After casting of the solution, the heating stage was set to 50 C for 1 h for leveling, and 80 C for 1 h to remove the solvent under a N2 flow. The films on the glass substrates were released with initiation of the removal at one edge of the film with a razor blade, and then fixed into a square-shaped metal frame to expose both sides of the film. Finally, the films in tenter frames were thermally heated up to 300, 350, 380, 400, 430, and 450 C in each series of solvent at a ramping rate of 2 C/min under a N2 purge. The thermally cured film thickness was in the range 45e55 mm. 2.3. Thermal and chemical analyses The amount of residual solvent and the evaporation rate of the solvent were measured with a TA Instruments TGA Q5000IR. The amount of water generated during imidization and the amount of residual solvent according to heating temperature were also measured with an Agilent GCeMS 6890/5973 system with a pyrolyzer and UA-1 Column. The sample was heated from 80 C to 500 C at a heating rate of 10 C/min under a He flow. A TA Instruments DSC Q2000 was used to trace the thermal curing of the films, and the heating and cooling cycles were repeated to test for reversibility. Both the TGA and DSC measurements were performed with ramping rate of 10 C/min under a N2 purge. Chemical curing was also analyzed with a Bio Rad FTS-6000 Fourier-transform IR (FTIR) spectrometer. 2.4. Spectroscopic methods Optical properties such as the yellow index (YI) and UV-Vis absorption intensity (AI) were measured with a CM3600D spectrophotometer manufactured by KONICA MINOLTA Inc. XRD measurements were conducted using a laboratory-scale wide-angle Xray scattering instrument (Philips Xpert) with Cu Ka radiation in the conventional pinhole geometry. The samples were mounted in the specular geometry. The Incidence angle of the 40 KeV/40 mA powered X-ray was fixed, and a diffraction pattern was obtained through a detector (2q) scan. The optical retardation was derived from the difference between the in-plane and out-of-plane refractive index of the film, as measured with a METRICON Prism coupler 200P-1 equipped with a HeeNe laser source with a wavelength of 632.8 nm. 2.5. Thermo-dynamic mechanical analyses
Fig. 1. (a) Fluorinated polyimide, (b) 3,4,30 ,40 -biphenyltetracarboxylic dianhydride (BPDA), (c) 2,20 -bis(trifluoromethyl)-4,40 -diaminobiphenyl (TFDB), and (d) N-methyl2-pyrolidone (NMP), (e) N,N-dimethylacetamide (DMAc).
Films cut into rectangular shapes of around 5 mm 20 mm were used for DMA and TMA measurements. TA Instruments DMA Q800 and TMA Q400 were used for the investigation of the transition behavior and the evaluation of thermal strain, respectively. In DMA measurements, a strain of 0.1% and 1 Hz oscillation were used for measuring the storage and loss modulus change as a function of temperature. Samples were scanned up to 400 C at a heating rate
202
Y. Jung et al. / Polymer 59 (2015) 200e206
of 5 C/min under a N2 purge. For TMA measurements, the samples were heated up to 400 C or until the maximum dimension change of the instruments, z2.3 mm. 3. Results and discussion 3.1. Thermal imidization and residual solvents Appropriate thermal conditions are required in solvent-casting methods to obtain fluorinated polyimide films with high transparency and a high DoI. Here, two common solvents, NMP and DMAc, were used selectively [28e32]. Fig. 2a shows the DSC thermogram of a film made from the solution with DMAc and thermally cured up to 300 C. At the first heating scan, a broad endotherm was measured around 350 C. This indicates that a certain amount of heat is supplied to the films to remove the residual solvent and to finish the thermal imidization that releases H2O as a byproduct. However, upon the second heating after the first scan up to 400 C, no distinguishable endotherms are observed. It can be supposed that thermal imidization was completed after heating at around 400 C. The amount of residual solvent, monitored by TGA, also shows that the thermal imidization could be finished at around 400 C (Fig. 2b). Despite the more volatile nature of DMAc than NMP, as shown in Fig. 2c, a larger amount of DMAc remains in the heated films, as shown by the initial amount of solvent in Fig. 2b. This is consistent with a previous report by Ebisawa et al. [30]. In this report, the C]O stretching band of hydrogen-bonded DMAc with the COOH groups of polymer chains is activated by lower energy infrared rays. This indicates stronger hydrogen bonding of DMAc to the polymer chains than the case of NMP. However, as the heating temperature approaches 400 C, the amount of residual solvents converges to 0 wt-%. According to EGA-MS result in Fig. 2d, a relatively small amount of solvent and H2O comes out from the films until the temperature slightly higher than 400 C. H2O is the byproduct of imidization and the solvent was hydrogen-bonded to
polymer chains prior to imidization. Considering the diffusion time of the trapped H2O and the organic solvent, the imidization seems to be completed at around 400 C in BTDA-TFDB PAA. On the basis of the results from DSC, TGA, and EGA-MS, a series of films heated up to 300, 350, 380, 400, 430, and 450 C were prepared to explore the change in polymer morphology in highly imidized polyimide films and to evaluate their correct imidization ratio. 3.2. Color change during imidizaiton and chargeetransfer complex formation The DoI is usually measured through one of several methods including FT-IR [18e23] and the density measurement method [33]. However, the DoI is rarely quantified by conventional methods when it is higher than z95%, as mentioned above. In FT-IR, the DoI can be calculated through the comparison of the stationary peak and changeable peak area of the imide rings with imidization ratio. Fig. S1 in the supplementary data shows the FT-IR spectra for BTDATFDB polyimide films with heating temperatures of 300, 380, and 450 C. Here, the DoI can be estimated through a comparison of the CeNeC peak area or peak intensity with the fixed CeF stretching mode as the standard stationary peak. In contrast to the quantitative results of TGA, DSC, and EGA-MS, FT-IR does show almost identical spectra for the films heated at 300 C and 450 C. In this regard, more ideal methods are necessary to obtain the DoI in highly imidized polyimides. The film color or extinction coefficient can be an index for the DoI, although they are much more affected by the amount of CTC formation [13,34]. The transparency (Tr), yellow index (YI), and film-thickness-normalized absorption intensity (AI) are summarized in Table 1. Tr and YI (AI) show opposite trends. YI is a metric in the display industry, and it represents how yellow a polyimide film is. However, this index does not increase linearly and shows saturation behavior, as detailed in our previous report [13]. More conventionally, the extinction coefficient at yellow wavelengths is
Fig. 2. (a) DSC thermograms for an FPI film heat treated up to 300 C. (b) TGA thermograms for remaining solvent amounts as a function of temperature for FPI films heat treated up to 300 C. (c) Remaining solvent amounts and solvent evaporation speed as a function of time at 80 C. (d) EGA-MS profile for the DMAc solvent and water in an FPI film released from a glass substrate at 80 C.
Y. Jung et al. / Polymer 59 (2015) 200e206
203
Table 1 Tr, YI, and AI as a function of ANN temperature.
Heating temperature ( C)
From NMP solution Tra
300 350 380 400 430 450 a b
88.4 87.1 84.5 82.2 81.5 80.9
From DMAc solution AIb
YI ± ± ± ± ± ±
0.2 0.1 0.2 0.2 0.4 0.3
2.2 3.7 12.7 15.8 18.8 21.5
± ± ± ± ± ±
0.2 0.4 0.8 0.7 0.6 1.1
Tra
0.074 0.084 0.163 0.196 0.337 0.366
± ± ± ± ± ±
0.006 0.005 0.021 0.032 0.030 0.027
88.5 87.1 85.8 82.6 82.1 80.7
AIb
YI ± ± ± ± ± ±
0.1 0.3 0.2 0.4 0.4 0.2
2.1 2.8 5.6 12.6 17.6 19.9
± ± ± ± ± ±
0.2 0.3 0.5 0.7 0.8 0.6
0.073 0.075 0.117 0.164 0.351 0.389
± ± ± ± ± ±
0.008 0.009 0.020 0.017 0.019 0.020
cf. Tr for slide glass z92.0%. Film thickness normalized absorption intensity at 450 nm wavelength.
used to express the extent of color. We have used the filmthickness-normalized AI, which is consistent with the extinction coefficient in the BeereLambert law [35]. The yellowness is proportional to CTC formation, which can be intra-chain or inter-chain in nature. Besides, the CTC formation is relevant to imidization. The majority of imidization may be intra-molecular owing to the low mobility of the polymer chains in the film when a small amount of solvent is left, as shown in Fig. 1b. In Table 1, Both YI and AI increase as the heating temperature is raised, and there are marked increases at z380 C and z430 C. Here, the contributions from imidization, intramolecular CTC formation, and intermolecular CTC formation need to be distinguished.
marked increases in YI and AI. This could be used to make a distinction between imidization and intermolecular CTC formation. Thus, it is presumed that the a-transition may be affected by intermolecular CTC formation after completion of the imidization.
3.3. Transition behavior of fluorinated polyimides To examine the molecular reconfiguration such as imidization and CTC formation, we employed DMA, which can explore the thermal transition of FPI. Fig. 3a shows a DMA thermogram of a film heated at 300 C. As the temperature increases, the storage modulus (E0 ) decreases gradually but distinctly from 150 C to 300 C, and drops markedly at z320 C. The increase in viscoelasticity leads to broad peaks of the loss modulus (E00 ) at z250 C and z320 C, which are seen more clearly in the tand curve. The lower-temperature bump and higher-temperature peak are known transitions due to the rotation about the CeC bond in benzidine (btransition) and the segmental motion of the backbone (a-transition), respectively [15,16,36,37]. The distinct decrease in E0 during the b-transition renders the film fragile in the thermal drying process. Concerning the a-transition, we found interesting transition behavior unlike that reported previously for the glass transition of polyimide films [36,37]. Fig. 3b shows tand curves for films made from NMP solution at various heating temperatures. All the films show similar b-transition temperatures. However, their atransitions shift to lower-temperatures and the peak intensity, dynamic loss ratio, decrease with increasing heating temperature, as marked with the arrows. Identical trends in the change in E00 are shown in Fig. S2. This indicates that the segmental motion of the polymer backbones observed in the films decreases after heating at higher-temperatures. More importantly, the a-transition may not be the well-known inherent glass transition of polyimide. It is supposed to be a transition that could be affected by changes in the molecular structures or that of the PAA precursors. DMA frequency sweep results showing that the transition does not depend on frequency also support the conclusion that the second transition is not an inherent property of polyimide, as shown in Fig. S3. In this regard, between the broad b-transition temperature and the atransition, intramolecular imidization and consequent intramolecular CTC formation are being completed. In addition, local rotation in the TFDB units is not affected by CTC formation. Films obtained from DMAc solution exhibited almost identical trends, as shown in Fig. 3c. According to the color-change results of the films, the temperature range after the a-transition corresponds to the
Fig. 3. (a) DMA thermograms for an FPI film heat treated up to 300 C. (b) Loss tand traces according to heat treatment temperature in FPI films from NMP solution. (c) Loss tand traces for FPI films from DMAc solution.
204
Y. Jung et al. / Polymer 59 (2015) 200e206
3.4. Morphology change in thermal transition region To correlate the changes in polymer morphology of the films and their optical properties during the thermal transition, X-ray diffraction patterns were measured, as shown in Fig. 4a and b. The FPI X-ray diffraction patterns are represented as a function of scattering vector, q ¼ (4p/l)sinq (l ¼ 1.54 Å). The d-spacing of a peak is 2p/q [38]. Fig. 4a and b correspond to the data for films from NMP solution and DMAc solution, respectively. For both series of films, four distinguishable features are observed, as designated in the figure from #1 to #4. All the films show small crystalline peaks, #2 (014) and #4 (006) [39,40]. Peak #3 is a background peak originating from the carbonate paste between the films and the sample stage. The background peak also contributes slightly to peak #4 because of the film thinness. Between q ¼ 0.8 and 2.0, the films heated up to 300 C and 350 C only show broad amorphous halos. However, the small bump 1a and peak 1b, which can be assigned to
(003) and (004), respectively [36,37,39,40], at around q ¼ 1.2 develop in the films heated above 380 C. This reveals the increase in intra-chain regularity. However, in the films heated at 450 C, the crystalline peaks from the regularity of intramolecular chains change to show a broad amorphous region, as indicated by peak #1. Therefore, the increase in the crystalline region at around 380e400 C and also the a-transition in the DMA measurements may come from the intramolecular imidization process. The square symbols in Fig. 4c show the change in the ratio of the crystalline region including both the intramolecular and the rare intermolecular chain regularity as a function of the heating temperature. Fig. S4 in the supporting materials shows the peak deconvolution for the crystallinity calculation. The circles show the variation in AI at a wavelength of 450 nm, as shown in Table 1. Here, the solid and empty symbols indicate films from NMP and DMAc solutions, respectively. The AI value increases with increasing heating temperature, and the amount of the crystalline region represents an inflection point between 400 C and 450 C. The AI that is relevant to the CTC formation reveals that intramolecular CTC formation develops markedly between 300 C and 400 C, as the imidization process in the polymer chain occurs. Furthermore, the noticeable AI increase between 400 C and 430 C may originate from intermolecular CTC formation, despite the reduction in the total crystalline region in the same temperature range. Thus, the physical property that changes upon intramolecular CTC formation between 300 C and 400 C can provide useful information about the DoI.
3.5. Evaluation of degree of imidization by transition mechanical analysis Thermo-mechanical analysis is known to have exceptional resolution compared with conventional calorimetry or spectroscopic methods in the observation of the thermal transition [41e43]. Here, we used TMA, which is one example method with superior resolution for the measurement of dimension changes. The expansion behavior in a film depends on the molecular orientation in the film, particularly the out-of-plane anisotropy [44]. In Fig. S5, the Rth (outof-plane retardation) values of the FPI films shows that the molecular anisotropy is consistent according to the heating temperature. Thus, the dimension change of the films may be affected mainly by molecular interactions. As an example, 300 C heated film is shown in Fig. 5a. Slope “a” can be related to the coefficient of thermal expansion, and slope “b” is similar to the heat deflection process in thermoplastics [45,46]. By using a proper extension force, thermal expansion higher than the transition temperature can be measured. Therefore, we used an external force of 0.05 N and obtained the thermal expansion behavior as a function of temperature in each solvent series of films, as shown in Fig. S6. We can define the resistance strength (RS) as the repulsive modulus against the external stress, similarly to the tensile modulus, which is expressed as follows:
E¼
s DL
(1)
Here, the external force is z0.050 N, and the film width and thickness are 5 mm and 50 mm, respectively. This indicates the stress is about 10 MPa. Thermal stress can also be expressed as follows:
s ¼ k$a$E$DT Fig. 4. (a) XRD patterns according to heat treatment temperature in FPI films from NMP solution. (b) XRD patterns for FPI films from DMAc solution. (c) Amount of crystalline region and AI according to heat treatment temperature.
E¼
s s ¼ k$a$DT k$Dl
(2) (3)
Y. Jung et al. / Polymer 59 (2015) 200e206
205
from the external stress and thermal strain of samples. Through mechanical analysis during the thermal transition of polyimide, the subtly different effects of the solvents, DMAc and NMP, have been distinguished clearly. It is confirmed that DMAc forms stronger hydrogen bonds than NMP with PAA molecules. This may result in a higher degree of molecular orientation and larger resistance strength in films from NMP solution than in those from DMAc solution. These results could provide a useful guideline for the determination of the imidization ratio, which is a decisive parameter for physical properties in the substrates of electronic devices. Acknowledgment Y.J. thanks Dr. Sangmo Kim at Organic Material Lab, SAIT, for useful discussion. Y.J. also thanks Daeun Yu at the Analytical Science Group, SAIT, for experimental support. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.polymer.2015.01.007. References [1] [2] [3] [4] [5] [6] [7] [8] [9] Fig. 5. (a) TMA thermogram for FPI film heat treated at 300 C. Slopes “a” and “b” are the dimension changes before and during the a-transition, respectively. (b) RS according to heat-treatment temperature. RS for FPI film treated at 450 C corresponds to the thermal strain before the a-transition, i.e., slope “a” is equal to slope “b”.
[10] [11]
[12] DL DT .
In the above equations, a denotes If aT is the expansion coefficient during the transition, as represented by slope “b” in DLT Fig. 5a,aT is DT , and then RS can be expressed as: T
Er ¼
se se $DTT ¼ k$aT $DLT k$DLT
(4)
Here, Er can be calculated from the constant external force or external stress (se Þ. Finally, the RS values calculated from Eq. (4) are shown in Fig. 5b. As explained in the previous section, the increase in RS up to 400 C can be attributed to imidization or intramolecular CTC formations. Accordingly, the DoI could be obtained through the relative RS by assuming that the DoI of the film heated at 300 C is z95%. Also, the DoI is substantially higher in the films made from NMP solution than in those from DMAc solution. 4. Conclusions Through quantitative chemical, thermal, and mechanical analyses, we have predicted that the imidization process is completed at around 400 C. In addition, we have found that the glass transition of polyimide is not an inherent property. It is apparent that the glass transition and change in modulus during the thermal transition depend critically on the film's heat-treatment conditions. Relevant imidization and CTC formation make the polyimide show variable transition behaviors. Thus, we have suggested a new evaluation standard of resistance strength, which could be defined
[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36]
Lewis J. Mater Today 2006;9:38e45. Hellemans A. IEEE Spectr 2000;37:18e21. Ong B. Laser Focus World 2004;40:85e8. McAlpine MC, Friedman RS, Lin SJK-h, Wang WU, Lieber CM. Nano Lett 2003;3:1531e5. Jazi AY, Yoon S, Liu J. J Ind Eng Chem Res 2012;51:10887e94. Nysaether JB, Lundstrom P, Liu J. IEEE Trans Packag Manuf Technol Part A 1998;21:281e7. Jang J. Mater Today 2006;9:46e52. MacDonald WA. J Mater Chem 2004;14:4e10. Bessonov MI, Zubkov VA. Polyamic acids and polyimides: synthesis, transformations, and structure. Boca Raton, Florida: CRC Press; 1993 [Chapter 1e3]. Wilson D, Stenzenberger HD, Hergenrother PM. Polyimides. New York: Chapman and Hall; 1990 [Chapter 3]. Simone CD, Auman BC, Carcia PF, Wessel RA. Low color polyimide compositions useful in optical type applications and methods and compositions relating thereto. U. S patent 7,550,194, Aug 3; 2005. Jung Y, Song S, Kim S, Yang Y, Chae J, Park T, et al. Appl Phys Lett 2013;102: 031602. Jung Y, Yang Y, Kim S, Kim H-S, Park T, Yoo BW. Eur Polym J 2013;49: 3642e50. Ando SJ. J Photo Sci Tech 2004;17:219e32. Feiring AE, Auman BC, Wonchoba ER. Macromolecules 1993;26:2779e84. Hougham G, Tesoro JG, Shaw J. Macromolecules 1994;27:3642e9. Xie K, Liu JG, Zhou HW, Zhang SY, He MH, Yang SY. Polymer 2001;42: 7267e74. Anthamatten M, Letts SA, Day K, Cook RC, Gies AP, Hamilton TP, et al. J Polym Sci Part A Polym Chem 2004;42:5999e6010. Saeed MB, Zhan M-S. Eur Polym J 2006;42:1844e54. Snyder RW, Thomson B, Bartges B, Czerniawski D, Painter PC. Macromolecules 1989;22:4166e72. Mustoa P, Ragostaa G, Scarinzia G, Mascia L. Polymer 2004;45:1697e706. Ree M, Yoon DY, Volksen W. J Polym Sci Part B Polym Phys 1991;29:1203e13. Shin TJ, Ree M. Langmuir 2005;21:6081e5. Jones DS. Int J Pharm 1999;179:167e78. Jung Y. Surface characteristics of polyfluorenes and oligofluorenes, and bulk characteristics of novel polyolefins. Ph D Thesis:Seoul, Korea. 2005. Ferry JD. Macromolecules 1991;24:5237e45. Han B, Guo Y. J Electron Packag 1995;117:185e91. Hasegawa M, Ishigami T, Ishii J, Sugiura T, Fujii M. Eur Polym J 2013;49: 3657e72. Hasegawa M, Matano T, Shindo Y, Sugimura T. Macromolecules 1996;29: 7897e909. Ebisawa S, Ishii J, Sato M, Vladimirov L, Hasegawa M. Eur Polym J 2010;46: 283e97. Hsiao S-H, Guo W, Chung C-L, Chen W-T. Eur Polym J 2010;46:1878e90. Yang C-P, Chen W-T. Macromolecules 1993;26:4865e71. Han SH, Misdan N, Kim S, Doherty CM, Hill AJ, Lee YM. Macromolecules 2010;43:7657e67. Hasegawa M, Horie K. Prog Polym Sci 2001;26:259e335. Gill SC, von Hippel PH. Anal Biochem 1989;182:319e26. Eashoo M, Shen D, Wu Z, Lee CJ, Harris FW, Cheng SZD. Polymer 1993;34: 3209e15.
206
Y. Jung et al. / Polymer 59 (2015) 200e206
[37] Bas C, Tamagna C, Pascal T, Alberola D. Polym Eng Sci 2003;43:344e55. [38] Chabinyc ML, Toney MF, Kline RJ, McCulloch I, Heeney M. J Am Chem Soc 2007;129:3226e37. [39] Cheng SZD, Wu Z, Eashoo M, Hsu SLC, Harris FW. Polymer 1991;32:1803e10. [40] Cheng SZD, Arnold Jr FE, Zhang A, Hsu SLC, Harris FW. Macromolecules 1991;24:5856e62. [41] Liu J, Cheng SZD, Harris FW, Hsiao BS, Gardner KH. Macromolecules 1994;27: 989e96.
[42] Zhao J-H, Ryan T, Ho PS, McKerrow AJ, Shih W-Y. J Appl Phys 1999;85: 6421e4. [43] Rhee S-H, Du Y, Ho PS. J Appl Phys 2003;93:3926e33. [44] Pottiger MT, Coburn JC, Edman JR. J Polym Sci Part B Polym Phys 1994;32: 825e37. [45] Factor BJ, Russell TP, Toney MF. Macromolecules 1993;26:2847e59. [46] Baeurle SA, Hotta A, Gusev AA. Polymer 2006;47:6243e53.