Characterization of epoxy resins by photochemical hole burning (PHB) spectroscopy.
Photochemical hole burning (PHB) has long been recognized as both a magnificent tool for ultra-high-density optical memory (1) and a fruitful spectroscopic method for detecting properties of matrix polymers (2). Various kinds of materials have been used for establishing the relationship between PHB phenomen and properties of materials. In particular, dye molecules in amorphous polymer systems have attracted attention because of their high quantum yield and good processability. In dealing with these materials, the two-level system model, in which inhomogeneity or free volume plays a good part, has shown its importance (2, 3). An energy shift of the pseudophonon side band to zerophonon band, [E.sub.s], has also been assigned to stand for the action of matrix to guest material (chromophore) through the so-called electron-phonon interaction based on the discovery (4) that the phonon frequency, [E.sub.s], depends only on the matrix and not on the guest in the usual weak interaction systems. Although som results have been obtained for several kinds of materials (4-7), a study of Es of the same component but with different microstructures has been found only in undrawn and drawn samples of poly(ethylene terephthalate) (8) and in annealed samples of polyimides (9).
Epoxy resin is one of the materials that have been studied extensively. Oleynik (10) and others (11, 12) have shown that epoxy resin of the same components displays different thermal properties when its thermal histories are different. Simha, et al. (13), have also shown that even the totally cured epoxy resins possess some different dynamic properties if their thermal treatment conditions vary. These microstructure-related changes have also been verified by fluorescence measurement (14) and by means of a chemical reaction probe (15), which have shown convincing evidence that free volume or free volume distribution changes under a variety of treatment. Positronium annihilation experiments (16, 17) have been done in order to detect the free volume directly and to establish its correlations with the results of thermal measurements.
Photochemical hole burning of a dye molecule in epoxy resin was first studied b Furusawa, et al. (18); they discussed high temperature PHB performance and the result of [E.sub.s] in relation to crosslinking structure and the hydrogen bond in the resin. Rosenberg (19) has also proved, with inelastic neutron scattering the relationship between the microstructure and a low energy excitation mode, which showed the same energy as Es in PHB measurement (18). However, thermal treatment of the epoxy resin has not yet been studied by PHB. Moreover, the study of [E.sub.s] and other PHB phenomena depending on the crosslinking condition of the epoxy resin will be helpful in finding the relationship betwee microstructure and the low frequency excitation mode. In the present paper the phonon frequency, [E.sub.s], of epoxy resins with various cure conditions is detected by PHB measurements with free-base tetraphenylporphin (TPP). Cycle annealing experiments are also carried out. Crosslinking extents and glass transition temperatures, [T.sub.g], are measured and related to the phonon frequencies of the systems.
Epicote 828 (diglycidyl ether of bisphenol-A, DGEBA), obtained from Konishi Co. Ltd., was dried at 80 [degrees] C under vacuum. Diglycidyl ether of resorcinol (DGER) was obtained from Nagase Chemicals Ltd. and used without further purification. Free-base tetraphenylporphin (TPP), obtained from Wako Pure Chemical Industries, was used without further purification as a photoreactive guest molecule whose hole burning occurs as a phototautomeric reaction of internal protons. The crosslinking agent, m-phenylenediamine (m-PDA), obtained from Tokyo Chemical Industry Co. Ltd., was distilled before use. TPP-doped epox resin films were prepared by dissolving TPP and m-PDA into an epoxy resin at first and then curing them between thin glass plate at various conditions. Concentration of TPP was [approximately] [10.sup.-3] mol/1, and film thickness was about 0.1 mm.
Samples were set in a cryostat with a cryogenic refrigerator (Sumitomo, SRD204) and irradiated by an Argon-ion-laser-pumped single-mode continuous wave dye laser (Coherent, 699-01) with 0.44 mW/[cm.sup.2] laser power at about 640 nm corresponding to the absorption band of the lowest 0-0 transition (Q band) of TPP. Holes were detected by changes of transmittance with a 1-m-long monochromator (Jasco CT-100), a photomultiplier, and a lock-in amplifier (8). Curing extent test were made with infrared spectrophotometer (IR). The [T.sub.g measurements were made with a differential scanning calorimeter (DSC).
RESULTS AND DISCUSSION
Low Energy Excitation Modes in Polymers
Thermal properties, e.g., heat capacity, of amorphous materials are quite different from those of crystalline materials because of the existence of the nonthermally equivalent state. Compared to the heat capacity of Crystalline states, amorphous polymers have excess heat capacity. which is caused by the existence within them of low energy excitation modes originating from nonthermally equivalent states. The low energy excitation modes are extramodes compared to Debye modes in crystal. These extramodes have been studied and elucidated with heat capacity measurement (20), neutron inelastic scattering measurement (19, 21, 22), as well as PHB measurement (4-7).
Generally, a PHB hole consists of three parts (6), that is, a zero-phonon hole, which is a sharp hole emerging at the irradiation frequency and which is create by a photoreactive guest molecule whose zero-phonon line exists at the laser frequency; a phonon side hole, which possesses a broad hole emerging at the higher energy side of zero-phonon hole and which is created by the phonon side band of the same guest molecule that produces the zero-phonon hole; and finally a pseudo-phonon hole, which is a broad hole emerging at the lower energy side o the zero-phonon hole and which is created by molecules whose phonon side band exists at the laser frequency. At very low temperatures, a phonon creation process is dominant so that the phonon side band of guest molecules exists mainly at the higher energy side of their zero-phonon hole. Moreover, when interaction between the guest molecules and the matrix is weak, the phonon side band absorption is weak, and the Debye-Waller factor is close to unity. Because of this, the phonon side hole is usually small. On the other hand, the pseudo-phonon side hole is relatively large because of the broadness of side band absorption, which causes a large number of guest molecules to be selected by the laser. Thus, in general, the pseudo-phonon side hole is usually selected to be measured rather than the phonon side hole.
In the weak interaction system, one phonon process is dominant in the phonon side band absorption so that the energy difference between the zero-phonon hole and its pseudo-phonon side hole reflects the energy of a phonon mode, i.e., phonon frequency.
In the energy level of a guest probe dispersed in the matrix material, there ar the electronic level, the vibronic level, and the electron-phonon interaction energy level. The electronic level is in accordance with the laser energy used in PHB. The vibronic energy level can be detected by the energy difference between the zero-phonon hole and the satellite hole (23). The electron-phonon interaction energy level, [E.sub.s], determined by the energy difference betwee the zero-phonon hole and the pseudo-phonon side hole, measures nothing but the properties of the matrix in the case of weak interaction, based on the fact tha electron-phonon interaction is a mutual action of the electron of the guest and the phonon of the matrix and that the values of [E.sub.s] in usual polymers do not depend on the nature of the guest materials (4).
DSC and IR Measurement
Figure 1 shows typical IR charts of a DGEBA/m-PDA system. The extent of cure, [X.sub.cure], was calculated based on the change in absorbance of epoxide group at 930 [cm.sup.-1] divided by the absorbance of aromatic ring at 1600 [cm.sup.-1] as an internal standard. Results for both DGEBA/m-PDA and DGER/m-PD systems are shown in Table 1. It can be said from these results that the extent of cure is almost equal in all cases if the curing temperature is higher than 130 [degrees] C and curing time is longer than 4 h.
Figure 2, DSC charts of a DGER/m-PDA system, shows that the thermal behaviors o the samples with different cure conditions or thermal treatment conditions are different. Profound enthalpy relaxation is observed for the annealed samples. This coincides to some extent with the results given by Oleynik (10). The influence of curing conditions on several mechanical properties has also been elucidated (11). These facts show clearly that the microstructure of the polyme system varies when the curing condition changes. The values of [T.sub.g] were evaluated from the transition points of the DSC charts, and are listed in Table 1. The [T.sub.g] does not show large dependence on thermal treatment conditions
Figure 3 shows the absorption spectra of TPP in DGER at 20 K and room temperature. Holes were TABULAR DATA OMITTED burned around 645 nm in the lowest 0-0 transition of the Q-band with an inhomogeneous line width, [Delta][[Omega].sub.i], of approximately 550 [cm.sup.-1]. The absorbance of the band increases with a decrease in temperature, while the line width becomes narrower. Figure 4 shows the typical photochemical holes of TPP in epoxy resin films burned with a 0.44 mW/[cm.sup.2] ring dye laser at 20 K. Figure 5 is the change in hole depth and hole width during laser irradiation. The increment of hole depth. [Delta]A/[A.sub.o], where [Delta]A is the difference in absorbance produced by hole formation and that before irradiation, does not obey the first-order rule and shifts away during irradiation. The quantum efficiency of hole formation, [Phi], is calculated from the slope in Fig. 5 (6) by using Eq 1
[Phi] = [[d(A/[A.sub.o])dt].sub.t=o][A.sub.o]/[[10.sup.3][I.sub.o]](1 - [10.sup.-[A.sub.o]])[Epsilon]R] (1)
where [I.sub.0] is the laser intensity with the dimension of einstein/[cm.sup.2]/s, [Epsilon] is the molar extinction coefficient for the inhomogeneous line profile at the hole burning wave length and burning temperature, and R = [Delta][[Omega].sub.i]/[Delta][[Omega].sub.h] is the reciprocal initial fraction of TPP molecules within a homogeneous line width, [Delta][[Omega].sub.h], at the laser frequency. Table 2 shows the quantum efficiency for hole formation of TPP in epoxy resin with various curing conditions. It can be seen from these Tables and Figures that although the plot of hole depth or width vs. irradiation energy show some changes with different curing conditions, the quantum yield does not seem to be affected greatly. This is in accordance with our previous results (24) that [Phi] for TPP/polymer systems depends little on the matrix polymers as long as the samples were prepared in optimum conditions, that is, with sufficiently low concentration of TPP, effective evaporation of the solvent of solvent-casting sample and hot pressing.
Thermal Stability of Holes Burned at 20 K
Figure 6 show typical changes in hole profiles during cyclic annealing of TPP i the epoxy resin. The holes burned at 20 K become wider and shallower with an increase in cyclic annealing temperature, but TABULAR DATA OMITTED the mark of the holes is still observable when the temperature rises to about 80 K. It is known (8) that three factors, i.e., the decrease in Debye-Waller factor, the broadening of homogeneous width resulting from dephasing of the electronically excited state by electron-phonon interaction, and the spectral diffusion proces induced by local structural relaxation of the matrix polymer, are supposed to widen and shallow the hole profiles with increasing temperature. While the firs two factors are analyzed to be reversible processes with the change in temperature, the third one, spectral diffusion, is believed to be an irreversible process. Thus, by analyzing hole recovery observed when the sample is cooled down from each annealing temperature to 20 K, information on the hole broadening and hole filling caused by structural relaxation can be obtained. Figure 7 shows the changes in hole depth and hole area during cycle annealing. It can be deduced from these uniform curves with various curing conditions that the curing condition does not affect the temperature dependence behavior, mainl struciural relaxation, of the epoxy resin.
Dependence of [E.sub.s] on Cure Condition
In Figure 8 are shown typical PHB spectra a DGEBA/m-PDA system treated in different conditions. From these spectra, phonon frequencies, [E.sub.s], are calculated. Results are listed in Table 1. The [E.sub.s], the energy difference between the zero-phonon hole peak and the pseudo-phonon side hole peak, changes with the conditions of thermal treatment. In other words, the [E.sub.s] becomes larger when curing temperature is higher, and vice versa. Nevertheless, even at the same curing temperature, [E.sub.s] becomes higher when the sample, soon after being cured, was annealed at a temperature 20 [degrees] lower than the curing temperature, and becomes lower instead when the sample is quenched into liquid nitrogen just after being cured. The phonon frequency for the samples with the same component but different equivalent ratio [Mathematical Expression Omitted] display the same tendency as mentioned above. Moreover, some factors that affect electron-phonon interaction seem to be present, because it is noted that even when the epoxy component is changed from bisphenol-A (DGEBA) to resorcinol (DGER), the tendency as to the effect of thermal treatment seems unaffected. One of them is probably a free volume. For example, at the same curing temperature, the annealing samples have time enough to reach the equilibrium state, the arrangement of chains is fairly good, and therefore, distribution of free volume is narrower and the mean value is smaller (13), The value of [E.sub.s] becomes larger in this occasion and the phonon frequency becomes higher. On the other hand, because the samples quenched into liquid nitrogen have no time to reach their equilibrium state, arrangement of chains should be bad, and thus distribution of free volume is wide while the excess free volume over the equilibrium free volume is high. Needless to say, [E.sub.s becomes smaller and the phonon frequency becomes lower. When the curing temperature is high (150 [degrees] C in our experiment), free volume becomes smaller, in view of the fact that the samples are more easily rearranged during curing reaction than samples cured at low temperature.
Several studies (8, 9) have attributed the change in [E.sub.s] to a change in the ordered structure of the matrix. Considering the general interaction betwee guest and matrix, it is possible that both the distance between guest and matri (affecting the distance of action) and the ordered structure of matrix (affecting the density of state) have an effect. There is not enough evidence t distinguish these two factors at present, but it seems sound to attribute the change in [E.sub.s] in this study to the change in free volume, if we take into account that when the system changes from DGEBA to DGER, the [E.sub.s] becomes somewhat larger, and the main change in the matrix is concerned more with the change in the length of the chain than with the change in its orderliness.
Epoxy resin films were synthesized to carry out PHB experiments. Although quantum efficiency for hole formation and temperature cycling experiments did not show changes for samples under different curing conditions or annealing conditions after curing, the energy shift [E.sub.s] corresponding to the phonon frequency of the matrix polymers appeared to become large when curing temperature is high, and vice versa. At the same curing temperature, [E.sub.s] is large when the sample is annealed, and [E.sub.s] is small when the sample is quenched. The results of DSC measurement are in accordance with the [E.sub.s] results. The extent of cure measured by IR did not show large changes for different thermal treatment conditions.
One of the authors (Y.D.) gratefully acknowledges the invaluable assistance of Dr. T. Torii, Dr. Q. Jin, and Dr. R. Yokota. Thanks are due to people in the Horie research group for their assistance. He also expresses thanks to Mrs. Suxin Wang for reading the manuscript.
1. W. E. J. Moerner, Mol. Electron, 1, 55 (1985).
2. W. E. J. Moerner, ed., Persistent Spectral Hole Burning; Science and Applications, Springer, Berlin (1988).
3. K. Horie and I. Mita, Adv. Poly. Sci., 88, 77 (1989).
4. A. Furusawa, K. Horie, and I. Mita, Chem. Phys. Lett., 161, 227 (1989).
5. S. Saikan, T. Kishida, Y. Kanematsu, H. Aota, A. Harada, and M. Kamachi, Chem. Phys. Lett., 166, 358 (1990).
6. K. Horie and A. Furusawa, Progr. Photochem. Photophys., 5, 49 (1992).
7. B. Sauter, Th. Basche, and C. Brauchle, J. Opt. Soc. Amer., B, 9, 804 (1992)
8. K. Horie, K. Kuroki, I. Mita, H. Ono, S. Okumura, and A. Furusawa, Polymer, 32, 851 (1991).
9. A. Furusawa, K. Kuroki, K. Horie, and I. Mita, Polymer, 32, 2167 (1991).
10. E. F. Oleynik, Adv. Polym. Sci., 80, 50 (1986).
11. V. B. Gupta, L. T. Drzal, and C. Y-C. Lee, Polym. Eng. Sci., 25, 812 (1985)
12. I. C. Choy and J. P. Donald, J. Polym. Sci., Part B: Polym. Phys., 24, 1303 (1986).
13. M. Shmorhum, A. M. Jamieson, and R. Simha, Polymer, 31, 812 (1990).
14. L. Vladimirov, M. Hasegawa, T. Yamashita, and K. Horie, Polym. Prepr. Jpn., 40, 1202 (1991).
15. S. Tahara, T. Yamashita, and K. Horie, Polym. Prepr. Jpn., 41, 817 (1992).
16. Y. C. Jean, T. C. Sandreczki, and D. P. Ames, J. Polym. Sci., Part B: Polym Phys., 24, 1247 (1986).
17. Y. Kobayashi, W. Zheng, E. F. Meyer, J. D. McGervey, A. M. Jamieson, and R. Simha, Macromolecules, 22, 2302 (1989).
18. A. Furusawa, K. Horie, T. Suzuki, S. Machida, and I. Mita, Appl. Phys. Lett., 57, 141 (1990).
19. H. M. Rosenberg, Phys. Rev. Lett., 54, 704 (1985).
20. W. Rease, J. Macromol. Sci., A, 3, 1257 (1969).
21. T. Kanaya, K. Kaji, S. Ikeda, and K. Inoue, Chem. Phys. Lett., 150, 334 (1988).
22. K. Inoue, T. Kanaya, S. Ikeda, K. Kaji, K. Shibata, M. Misawa, and Y. Kiyanagi, J. Chem. Phys., 95, 5332 (1991).
23. K. Horie, M. Ikemoto, T. Suzuki, S. Machida, T. Yamashita, and N. Murase, Chem. Phys. Lett., 195, 563 (1992).
24. K. Horie, A. Furusawa, K. Kuroki, and I. Mita, in Polymers for Microelectronics, p. 519, Y. Tabata, et al., eds., Kodansya-VCH (1990).
|Printer friendly Cite/link Email Feedback|
|Author:||Du, Yingjun; Horie, Kazuyuki; Ikemoto, Makoto|
|Publication:||Polymer Engineering and Science|
|Date:||Sep 15, 1994|
|Previous Article:||Degradation of unstabilized medium-density polyethylene pipes in hot-water applications.|
|Next Article:||A theoretical analysis of the strength of composite gels with rigid filler particles.|