- Email: [email protected]
A structural study of silicon-based interpenetrating polymer networks by solid state 1H and 13C NMR spectroscopy
Journal of Molecular Structure 602±603 (2002) 321±333
www.elsevier.com/locate/molstruc
A structural study of silicon-based interpenetrating polymer networks by solid state 1H and 13C NMR spectroscopy q Masatoshi Kobayashi a,*, Shigeki Kuroki a, Isao Ando a, Kazuo Yamauchi b, Hideaki Kimura c, Koichi Okita c, Manabu Tsumura d, Keisuke Sogabe e a
Department of Chemistry and Materials Science, Tokyo Institute of Technology, Ookayama, Meguro-ku, Tokyo 152-8552, Japan b Bruker Japan, Ninomiya, Tsukuba, Ibaraki 305-0051, Japan c Japan Chemical Innovation Institute, 2-1-6 Sengen, Tsukuba, Ibaraki 305-0047, Japan d Kobe Research Laboratories, Kaneka Corporation, 1-2-80 Yoshida-cho, Hyogo-ku, Kobe 652-0872, Japan e Kaneka Techno Research Co., Ltd, 1-2-80, Yoshida-cho, Kobe 652-0872, Japan Received 8 March 2001; accepted 27 March 2001
Abstract High-resolution solid-state 13C CP/MAS NMR spectra and two-dimensional (2D) 13C± 13C and 1H± 1H exchange NMR spectra of cured ladder silsesquioxane oligomer (LDS)/polycarbosilane (PCS) system which forms interpenetrating polymer networks (IPN) have been measured, in order to elucidate the phase structure and miscibility. From the 13C CP/MAS NMR experimental results on cured LDS/PCS IPN samples with various mixture weight ratios, it was found that the 1 HT1r value of cured LDS/PCS IPN with mixture weight ratio 8/2[LDS/PCS(8/2)] sample was smaller than the 1 HT1r values of cured LDS/ PCS IPN samples with other mixed weight ratios. This suggests that in the cured LDS/PCS(8/2), the intermolecular dipole± dipole interaction most effectively contributes to the 1 HT1r because of the shortest intermolecular distance between LDS and PCS chains. Further, from the 2D 13C± 13C and 1H± 1H exchange NMR results, it was found that the intramolecular spin diffusion occurs in all the samples, and intermolecular spin diffusion occurs in cured LDS/PCS(8/2). This shows that the miscibility of LDS/PCS(8/2) is higher compared to that in cured LDS/PCS samples with the other mixed weight ratios, and in cured LDS/PCS(8/2) the interchain distance between LDS and PCS portions is within the order of several 10 nm. q 2002 Elsevier Science B.V. All rights reserved. Keywords: Ladder silsesquioxane oligomer; Polycarbosilane; Interpenetrating polymer networks; Solid state 13C NMR; 2D-NMR; Spin diffusion
1. Introduction Silicon-based organic±inorganic composite materials have been studied to develop new materials q
Dedicated to Professor Graham A. Webb on the occasion of his 65th birthday. * Corresponding author. Tel. : 181-3-5734-2139; fax: 181-35734-2889. E-mail address: [email protected] (M. Kobayashi).
with the properties of both organic materials, such as light weight, high ¯exibility and high moldability and those of inorganic materials such as high thermal stability and high strength [1±3]. These materials are expected to be used in various ®elds. Also, many kinds of carbon-based interpenetrating polymer networks (IPNs) have been widely studied in order to improve the thermal and mechanical properties of polymer materials [4,5]. Recently, Tsumura et al. synthesized silicon-based
0022-2860/02/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved. PII: S 0022-286 0(01)00776-1
322
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 1. The chemical structures of LDS and PCS.
IPNs consisting of stable Si±O and Si±C linkages [6,7], where a ladder silsesquioxane oligomer (LDS) and a polycarbosilane (PCS) have been used as two kinds of components for cured LDS/PCS IPN
samples. From these experimental results, it was found that the thermal and mechanical properties of cured LDS/PCS INP sample with mixture weight ratio of 8/2[LDS/PCS(8/2)] are better than those of LDS/
Fig. 2. The pulse sequences for determining 1H spin±lattice relaxation time in the rotating frame T1Hr (a) and for measuring 13C± 13C 2(b) and 1 H± 1H(c) 2D exchange NMR spectrum.
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
323
Fig. 3. 13C CP/MAS (TOSS) NMR spectra of ®ve kinds of cured LDS/PCS systems LDS=PCS 10=0; 8=2; 5=5; 2=8; 0=10 wt=wt:
324
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 4. 13C CP/MAS spectrum of cured LDS/PCS(5/5) decomposed by computer-®tting with Gaussian function.
PCS IPN samples with other mixture weight ratios. However, structural characterization of cured LDS/ PCS IPN system with various mixture ratios has been insuf®ciently performed. From such a situation, we aim to elucidate the structure and miscibility of silicon-based LDS/PCS IPN systems with various mixture ratio by using solid-state 13C, 1H CRAMPS and 1H± 1H 2D exchange NMR methods which provide very useful information about the structure and dynamics of polymers in the solid state [8±11].
2. Experimental 2.1. Materials Silicon-based LDS/PCS IPN samples with various mixture weight ratios used in this work, according to the previous paper [6] were obtained by condensation reaction and hydrosilylation reaction with LCS [commercial name: Glass Resin GR100] and PCS. Five kinds of cured LDS/PCS IPN samples with
mixture weight ratios wt=wt 10=0; 8=2; 5=5; 2=8 and 0/10 were prepared. The chemical structures of LDS and PCS are shown in Fig. 1. 2.2. NMR measurements High-resolution solid-state 13C cross polarization/ magic angle spinning (CP/MAS) NMR spectra were obtained by a JEOL GSX-270 NMR spectrometer operating at 67.4 MHz and a Chemagnetics CMX-300 NMR spectrometer operating at 75 MHz. The 1H 908 pulse width is 3.65 ms and the contact time is 1 ms. The acquisition number was 800, and the pulse delay was 5 s. Samples were contained in a cylindrical rotor and spun at 3±4 kHz. The 1H spin±lattice relaxation time in the rotating frame T1Hr was measured by the pulse sequence shown in Fig. 2(a). The spin locking time for 1 H magnetization was 1±80 ms. 13 C± 13C 2D exchange NMR spectra were obtained by a Bruker DSX-300 NMR spectrometer operating at 75 MHz. The pulse sequence as shown in Fig. 2(b) was used with the mixing time of 0.5 s and the acquisition number was 80 £ 128. 1H CRAMPS NMR
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
325
Fig. 5. Partially-relaxed 13C spectra of cured LDS/PCS(5/5) (a) and cured LDS/PCS(2/8) sample (b) by the pulse sequence for T1Hr measurement.
326
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
3. Results and discussion 3.1. Structural and dynamic elucidation by 13C CP/ MAS NMR
Fig. 6. T1Hr decay curves for cured LDS/PCS(8/2). (a) Signal intensity for the methyl, ethylene and phenyl carbons, and (b) decomposed signal intensity for LDS-C3 and PCS-C1 carbons in the phenyl region.
spectra and 1H± 1H 2D NOESY NMR spectra were obtained by a Bruker DSX-300 NMR spectrometer with the BR24 1H homonuclear decoupling pulse. The 1H 908 pulse length was 1.5 ms and the mixing time was 100 ms. The acquisition number was 16.
Observed 13C CP/MAS TOSS NMR spectra of ®ve kinds of cured LDS/PCS IPN samples with mixture weight ratios LDS=PCS 10=0; 8=2; 5=5; 2=8 and 0=10 are shown in Fig. 3. The 13C CP/MAS spectrum of unblended LDS is shown at the bottom of these spectra in Fig. 3. LDS has the phenyl carbons and the methylene carbons as seen in Fig. 1. Signals appearing at about 17 and 125±135 ppm are assigned to the methyl and phenyl carbons, respectively. In the phenyl signal, peaks at 135, 130 and 126 ppm are assigned to the C1 and C2 carbons, the C4 carbon and the C3 carbons. On the other hand, the 13C CP/ MAS spectrum of unblended PCS sample is shown at the top in Fig. 3. PCS has phenyl carbons, two types of the methylene carbons and three types of methyl carbons as seen from Fig. 1. Peaks appearing at about 25 and 10 ppm are assigned to the methylene and methyl carbons, respectively. In the phenyl signal, the down®eld peak at 140 ppm is assigned to the C1 carbon, and the upper®eld peak at 126 ppm to other C2 carbons. These assignments are based on other silicon-based polymers [7]. The 13C chemical shift values of the individual carbons in unblended LDS and PCS are useful for the signal assignments of cured blend LDS/PCS samples. The observed 13C CP/MAS spectra of cured LDS/ PCS systems with three kinds of mixture ratios are shown in Fig. 3. The methyl and phenyl carbon peaks from the LDS and PCS portions in the cured blend sample overlap with each other. Therefore, these peaks were decomposed by computer-®tting as shown in Fig. 4. The 13C CP/MAS spectrum of cured LDS/PCS(5/5) is decomposed by Gaussian function. As expected, the intensities of individual peaks correspond to the mixture composition in LDS/PCS samples. The 13C chemical shift values of the decomposed peaks are almost same as those of LDS and PCS. This shows that there are no new peaks with signi®cant intensity formed by curing LDS/PCS(5/5) sample. The peak intensities determined by such a decomposition are used in the T1Hr analysis, in order to elucidate the miscibility of blended LDS/PCS samples.
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
327
Table 1 Determined T1Hr values (experimental error is ^2 ms) of cured LDS/PCS samples with various mixture ratios determined at 258C Samples
Cured IPN LDS/PCS(8/2) Cured IPN LDS/PCS(5/5) Cured IPN LDS/PCS(2/8) Cured LDS Cured PCS a
T1Hr (ms) Methyl
Methylene
Phenyl (total)
Phenyl (LDS)
Phenyl (PCS)
34 40 46 39 38
32 42 49
35 39 47 38 37
36 36 43
29 a 38 46
38
Experimental error is larger than other values because of low signal -to-noise ratio.
Fig. 5 shows partially-relaxed 13C spectra of cured LDS/PCS(5/5)(a) and cured LDS/PCS(2/8)(b) as a function of spin locking time t by the pulse sequence for the T1Hr measurement (Fig. 2(a)), where the asterisks indicate the spinning sidebands [cured LDS/PCS(5/5) not shown]. It is seen that the intensities of individual peaks (M(t )) decay with an increase in t . As these decays follow equation M t=M 0 exp 2t=T1Hr ; where M(0) is the peak intensity at t 0: The plots of ln [M(t)/M(0)] for the methyl, methylene and phenyl carbons as obtained from the decomposition of the overlapped peaks by computer®tting, as mentioned-above against t are shown in Fig. 6 and then T1Hr values of the corresponding carbons are determined from its slope. From these T1Hr curves, it was found that the relaxation behavior for the methyl, methylene and phenyl carbons is a single exponential decay. This means that molecular motions of these carbons are a single component. The determined T1Hr values of cured LDS/PCS samples are summarized in Table 1. The T1Hr values for the methyl, ethylene and phenyl carbons of each sample is, 34 ms for LDS/PCS(8/2), 40 ms for LDS/PCS(5/5) and 47 ms for LDS/PCS(2/8). These values for the three functional groups are close to each other in the sample. This means that intramolecular 1H spin diffusion occurs in cured LDS/PCS samples. On the other hand, the T1Hr values in unblended LDS are 39 and 38 ms for the methyl and phenyl carbons, respectively, and in unblended PCS 38, 38 and 37 ms for the methyl, methylene and phenyl carbons, respectively. It was found that the T1Hr values of LDS/ PCS(8/2) were smaller than those of unblended LDS and PCS. On the other hand, the T1Hr values of LDS/PCS(5/5) are almost same as unblended LDS
and PCS, and those of LDS/PCS(2/8) are much larger than them. Dipole±dipole interactions between nuclear spins predominantly govern relaxation mechanism in polymer systems. Dipole±dipole interaction Hd is expressed by X Hd h2 =8p2 gi gj 3cos2 u 2 1 Ii Ij 2 3Iiz Ijz =rij3 1 where h is the Planck constant, rij is the internuclear distance between nuclei i and j, u is the angle between the vector i±j and the magnetic ®eld, g i and g j are the magnetogyric ratio of nuclei i and j, respectively, and I is nuclear spin operator. In general, the 1H NMR relaxation is mainly induced by dipole±dipole interaction and is greatly affected by internuclear distance of nuclear spins ¯uctuating in the magnetic ®eld. Ê However, interatomic distance more than ca. 5 A does not affect almost the 1H NMR relaxation as seen from Eq. (1) because Hd is proportional to r 23. Thus, the NMR relaxation times, in principle, provide information about interatomic distance r, if r is less Ê . From such a background, we are than ca. 5 A concerned with the miscibility between the LDS and PCS portions in blended LDS/PCS samples. Table 2 The density of cured LDS/PCS samples with various mixture ratios determined at 258C Samples
Density (g/cm 3)
Cured LDS Cured IPN LDS/PCS(8/2) Cured IPN LDS/PCS(5/5) Cured IPN LDS/PCS(2/8) Cured PCS
1.258 1.258 1.216 1.161 1.019
328
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
329
Fig. 7. (continued)
The intermolecular distance between the LDS and PCS portions in blended LDS/PCS IPN systems predominantly affects T1Hr Ð considered here, which is closely related to the miscibility. This may be supported by the experimental results on the density of the blended LDS/PCS samples as summarized in Table 2. The density of LDS/PCS(8/2) is the largest of three kinds of blended LDS/PCS samples. Therefore, it can be said that the intermolecular distance between LDS and PCS in LDS(8/2) may be the shortest one in three kinds of blended LDS/PCS IPN samples. This result suggests that intermolecular dipole±dipole interaction affects ef®ciently T1Hr in LDS/PCS(8/2).
From these experimental results, the miscibility of LDS and PCS in LDS/PCS(8/2) is predicted to be the highest. 3.2. 2D 13C± 13C and 1H± 1H exchange NMR experiments In order to evaluate the intermolecular distance between LDS and PCS in blended LDS/PCS systems, 13 C± 13C and 1H± 1H spin diffusion experiments were carried out. Fig. 7(a) shows 13C± 13C 2D exchange NMR spectrum of LDS/PCS(5/5) with a mixing time of 500 ms. In order to recognize the cross
Fig. 7. (a) 13C± 13C 2D exchange NMR spectrum of LDS/PCS(5/5) with mixing time 500 ms. Expanded spectra of the aliphatic region (b) and phenyl region (c).
330
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 8. 1H CRAMPS NMR spectra of unblended LDS and PCS, and blended LDS/PCS(8/2).
peaks as appearing due to intramolecular interaction and intermolecular interaction, the expanded spectra of the aliphatic and phenyl regions are shown in Fig. 8(b) and (c), respectively. As seen from these spectra, cross peaks were observed between the methyl carbon and ethylene carbon in the aliphatic region and between PCS-C1 and PCS-C2, PCS-C1 and LDS-C1, PCS-C1 and LDS-C2, LDS-C3 and LDS-C1, LDS-C3 and LDS-C2, and LDS-C3 and PCS-C3 in the phenyl region. The peak assignments follow those made by
computer-®tting as shown in Fig. 4. These cross peaks are assigned to intramolecular 13C± 13C spin diffusion in LDS and in PCS portions. In 13C± 13C 2D exchange NMR spectra of LDS/PCS(8/2) and LDS/PCS(2/8), intramolecular spin diffusion was clearly observed. On the other hand, no cross peak, which is induced by intermolecular 13C± 13C spin diffusion between PCS-C1 and LDS-C3 carbons in LDS/PCS(8/2) appears. As expected the miscibility is much higher, because the intensity of PCS-C1 peak is very weak.
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
331
Fig. 9. 1H± 1H 2D exchange NMR spectra of LDS/PCS(8/2) with a mixing time of 100 ms (a). (b) Expanded spectra of the cross peak region (b).
Fig. 8 shows 1H CRAMPS NMR spectra of unblended LDS and PCS, and blended LDS/PCS(8/ 2). The aliphatic proton signal and phenyl proton signal are clearly observed, of which the 1H chemical shift values are indicated on the peak top. As seen from these spectra of unblended LDS and PCS, the 1 H chemical shifts of the methyl and phenyl protons of LDS appear at upper®eld than those of PCS. On the other hand, in LDS/PCS(8/2) the peaks of the methyl and phenyl protons in LDS portion and PCS portion overlapped with each other. The apparent top position of the signal changes by changing the mixture ratio. The positions of the overlapped peaks in LDS/PCS(8/ 2) are between those in LDS and PCS. These 1H
chemical shift behaviors will be used for 1H± 1H 2D exchange NMR spectral analysis. Fig. 9(a) shows 1H± 1H 2D exchange NMR spectrum of LDS/PCS(8/2) with a mixing time of 100 ms. The cross peaks which are derived from the existence of 1H± 1H spin diffusion between the phenyl protons and CH3 protons, and the phenyl protons and the CH2 protons appear clearly. For convenience, expanded spectrum in the cross peak region is shown in Fig. 9(b). The dashed lines indicate 1H chemical shift positions of unblended LDS and PCS. These values become reference data for understanding the observed cross peaks due to 1H± 1H spin diffusion. The peak top at around 1H chemical shift position X1; X2
332
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 9. (continued)
20:8; 6:9 ppm may be assigned to intramolecular spin diffusion between the CH3 protons of LDS and the phenyl protons of LDS because of the most intense peak. The shoulder peak at 1H chemical shift position X1; X2 20:4; 7:2 ppm may be assigned to intramolecular spin diffusion between the CH3 and CH2 protons of PCS and the phenyl protons of PCS because of the next most intense peak. The remaining shoulder peak at 1H chemical shift position X1; X2 20:4; 6:9 ppm may be assigned to intermolecular spin diffusion between the CH3 and CH2 protons of PCS and the phenyl protons of LDS because of the very weak peak. The two cross peaks at X1; X2 20:4; 6:9 ppm and (20.4, 7.2 ppm) positions can be signi®cantly recognized by making careful spectral analysis of its region. In the other blend LDS/PCS samples, a corresponding shoulder peak at 1H chemical shift position X1; X2 20:4; 6:9 ppm does not appear. From these experi-
mental results, the intermolecular 1H± 1H spin diffusion occurs in LDS/PCS(8/2) only. This suggests that the interchain distance between LDS and PCS portions is almost within the order of several 10 nm. 4. Conclusions The phase structure and miscibility of silicon-based interpenetrating polymer networks composed of LDS and PCS portions were successfully elucidated by solid-state 13C CP/MAS NMR, 1H CRAMPS and 2D 1 H± 1H exchange NMR. The 13C CP/MAS NMR experimental results suggest that the intermolecular dipole±dipole interaction occurs signi®cantly in LDS/PCS(8/2) because of short intermolecular distance between LDS and PCS portions. It was suggested from the observation of intermolecular spin diffusion in the 2D 1H± 1H exchange NMR
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
spectrum in LDS/PCS(8/2) that the interchain distance between LDS and PCS portions is almost within the order of several 10 nm. References [1] B.M. Novak, C. Davies, Macromolecules 24 (1991) 5481. [2] B. Wang, G.L. Wilkes, J. Polym. Sci. Polym. Chem. 29 (1991) 905. [3] Y. Chujo, T. Saegusa, Adv. Polym. Sci. 100 (1992) 11. [4] D. Klempner, L.H. Sperling, L.A. Utracki, Interpenetrating Polymer Networks, Advances in Chemical Series 239, American Chemical Society, Washington, DC, 1994.
333
[5] L.H. Sperling, Interpenetrating Polymer Networks and Related Materials, Plenum, New York, 1981. [6] M. Tsumura, K. Ando, J. Kotani, M. Hiraishi, T. Iwahara, Macromolecules 31 (1998) 2716. [7] S. Kuroki, K. Okita, T. Kakigano, J. Ishikawa, M. Itoh, Macromolecules 31 (1998) 2804. [8] R.A. Komoroski (Ed.), High Resolution NMR Spectroscopy of Synthetic Polymers in Bulk VCH Publishers, Deer®eld Bleach, FL, 1986. [9] I. Ando, T. Yamanobe, T. Asakura, Prog. NMR Spectrosc. 22 (1990) 349. [10] K. Kurosu, S. Ando, H. Yoshimizu, I. Ando, Ann. Rept. NMR Spectrosc. 28 (1994) 189. [11] I. Ando, T. Asakura (Eds.), Solid State NMR of Polymers Elsevier Science, Amsterdam, 1998.
www.elsevier.com/locate/molstruc
A structural study of silicon-based interpenetrating polymer networks by solid state 1H and 13C NMR spectroscopy q Masatoshi Kobayashi a,*, Shigeki Kuroki a, Isao Ando a, Kazuo Yamauchi b, Hideaki Kimura c, Koichi Okita c, Manabu Tsumura d, Keisuke Sogabe e a
Department of Chemistry and Materials Science, Tokyo Institute of Technology, Ookayama, Meguro-ku, Tokyo 152-8552, Japan b Bruker Japan, Ninomiya, Tsukuba, Ibaraki 305-0051, Japan c Japan Chemical Innovation Institute, 2-1-6 Sengen, Tsukuba, Ibaraki 305-0047, Japan d Kobe Research Laboratories, Kaneka Corporation, 1-2-80 Yoshida-cho, Hyogo-ku, Kobe 652-0872, Japan e Kaneka Techno Research Co., Ltd, 1-2-80, Yoshida-cho, Kobe 652-0872, Japan Received 8 March 2001; accepted 27 March 2001
Abstract High-resolution solid-state 13C CP/MAS NMR spectra and two-dimensional (2D) 13C± 13C and 1H± 1H exchange NMR spectra of cured ladder silsesquioxane oligomer (LDS)/polycarbosilane (PCS) system which forms interpenetrating polymer networks (IPN) have been measured, in order to elucidate the phase structure and miscibility. From the 13C CP/MAS NMR experimental results on cured LDS/PCS IPN samples with various mixture weight ratios, it was found that the 1 HT1r value of cured LDS/PCS IPN with mixture weight ratio 8/2[LDS/PCS(8/2)] sample was smaller than the 1 HT1r values of cured LDS/ PCS IPN samples with other mixed weight ratios. This suggests that in the cured LDS/PCS(8/2), the intermolecular dipole± dipole interaction most effectively contributes to the 1 HT1r because of the shortest intermolecular distance between LDS and PCS chains. Further, from the 2D 13C± 13C and 1H± 1H exchange NMR results, it was found that the intramolecular spin diffusion occurs in all the samples, and intermolecular spin diffusion occurs in cured LDS/PCS(8/2). This shows that the miscibility of LDS/PCS(8/2) is higher compared to that in cured LDS/PCS samples with the other mixed weight ratios, and in cured LDS/PCS(8/2) the interchain distance between LDS and PCS portions is within the order of several 10 nm. q 2002 Elsevier Science B.V. All rights reserved. Keywords: Ladder silsesquioxane oligomer; Polycarbosilane; Interpenetrating polymer networks; Solid state 13C NMR; 2D-NMR; Spin diffusion
1. Introduction Silicon-based organic±inorganic composite materials have been studied to develop new materials q
Dedicated to Professor Graham A. Webb on the occasion of his 65th birthday. * Corresponding author. Tel. : 181-3-5734-2139; fax: 181-35734-2889. E-mail address: [email protected] (M. Kobayashi).
with the properties of both organic materials, such as light weight, high ¯exibility and high moldability and those of inorganic materials such as high thermal stability and high strength [1±3]. These materials are expected to be used in various ®elds. Also, many kinds of carbon-based interpenetrating polymer networks (IPNs) have been widely studied in order to improve the thermal and mechanical properties of polymer materials [4,5]. Recently, Tsumura et al. synthesized silicon-based
0022-2860/02/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved. PII: S 0022-286 0(01)00776-1
322
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 1. The chemical structures of LDS and PCS.
IPNs consisting of stable Si±O and Si±C linkages [6,7], where a ladder silsesquioxane oligomer (LDS) and a polycarbosilane (PCS) have been used as two kinds of components for cured LDS/PCS IPN
samples. From these experimental results, it was found that the thermal and mechanical properties of cured LDS/PCS INP sample with mixture weight ratio of 8/2[LDS/PCS(8/2)] are better than those of LDS/
Fig. 2. The pulse sequences for determining 1H spin±lattice relaxation time in the rotating frame T1Hr (a) and for measuring 13C± 13C 2(b) and 1 H± 1H(c) 2D exchange NMR spectrum.
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
323
Fig. 3. 13C CP/MAS (TOSS) NMR spectra of ®ve kinds of cured LDS/PCS systems LDS=PCS 10=0; 8=2; 5=5; 2=8; 0=10 wt=wt:
324
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 4. 13C CP/MAS spectrum of cured LDS/PCS(5/5) decomposed by computer-®tting with Gaussian function.
PCS IPN samples with other mixture weight ratios. However, structural characterization of cured LDS/ PCS IPN system with various mixture ratios has been insuf®ciently performed. From such a situation, we aim to elucidate the structure and miscibility of silicon-based LDS/PCS IPN systems with various mixture ratio by using solid-state 13C, 1H CRAMPS and 1H± 1H 2D exchange NMR methods which provide very useful information about the structure and dynamics of polymers in the solid state [8±11].
2. Experimental 2.1. Materials Silicon-based LDS/PCS IPN samples with various mixture weight ratios used in this work, according to the previous paper [6] were obtained by condensation reaction and hydrosilylation reaction with LCS [commercial name: Glass Resin GR100] and PCS. Five kinds of cured LDS/PCS IPN samples with
mixture weight ratios wt=wt 10=0; 8=2; 5=5; 2=8 and 0/10 were prepared. The chemical structures of LDS and PCS are shown in Fig. 1. 2.2. NMR measurements High-resolution solid-state 13C cross polarization/ magic angle spinning (CP/MAS) NMR spectra were obtained by a JEOL GSX-270 NMR spectrometer operating at 67.4 MHz and a Chemagnetics CMX-300 NMR spectrometer operating at 75 MHz. The 1H 908 pulse width is 3.65 ms and the contact time is 1 ms. The acquisition number was 800, and the pulse delay was 5 s. Samples were contained in a cylindrical rotor and spun at 3±4 kHz. The 1H spin±lattice relaxation time in the rotating frame T1Hr was measured by the pulse sequence shown in Fig. 2(a). The spin locking time for 1 H magnetization was 1±80 ms. 13 C± 13C 2D exchange NMR spectra were obtained by a Bruker DSX-300 NMR spectrometer operating at 75 MHz. The pulse sequence as shown in Fig. 2(b) was used with the mixing time of 0.5 s and the acquisition number was 80 £ 128. 1H CRAMPS NMR
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
325
Fig. 5. Partially-relaxed 13C spectra of cured LDS/PCS(5/5) (a) and cured LDS/PCS(2/8) sample (b) by the pulse sequence for T1Hr measurement.
326
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
3. Results and discussion 3.1. Structural and dynamic elucidation by 13C CP/ MAS NMR
Fig. 6. T1Hr decay curves for cured LDS/PCS(8/2). (a) Signal intensity for the methyl, ethylene and phenyl carbons, and (b) decomposed signal intensity for LDS-C3 and PCS-C1 carbons in the phenyl region.
spectra and 1H± 1H 2D NOESY NMR spectra were obtained by a Bruker DSX-300 NMR spectrometer with the BR24 1H homonuclear decoupling pulse. The 1H 908 pulse length was 1.5 ms and the mixing time was 100 ms. The acquisition number was 16.
Observed 13C CP/MAS TOSS NMR spectra of ®ve kinds of cured LDS/PCS IPN samples with mixture weight ratios LDS=PCS 10=0; 8=2; 5=5; 2=8 and 0=10 are shown in Fig. 3. The 13C CP/MAS spectrum of unblended LDS is shown at the bottom of these spectra in Fig. 3. LDS has the phenyl carbons and the methylene carbons as seen in Fig. 1. Signals appearing at about 17 and 125±135 ppm are assigned to the methyl and phenyl carbons, respectively. In the phenyl signal, peaks at 135, 130 and 126 ppm are assigned to the C1 and C2 carbons, the C4 carbon and the C3 carbons. On the other hand, the 13C CP/ MAS spectrum of unblended PCS sample is shown at the top in Fig. 3. PCS has phenyl carbons, two types of the methylene carbons and three types of methyl carbons as seen from Fig. 1. Peaks appearing at about 25 and 10 ppm are assigned to the methylene and methyl carbons, respectively. In the phenyl signal, the down®eld peak at 140 ppm is assigned to the C1 carbon, and the upper®eld peak at 126 ppm to other C2 carbons. These assignments are based on other silicon-based polymers [7]. The 13C chemical shift values of the individual carbons in unblended LDS and PCS are useful for the signal assignments of cured blend LDS/PCS samples. The observed 13C CP/MAS spectra of cured LDS/ PCS systems with three kinds of mixture ratios are shown in Fig. 3. The methyl and phenyl carbon peaks from the LDS and PCS portions in the cured blend sample overlap with each other. Therefore, these peaks were decomposed by computer-®tting as shown in Fig. 4. The 13C CP/MAS spectrum of cured LDS/PCS(5/5) is decomposed by Gaussian function. As expected, the intensities of individual peaks correspond to the mixture composition in LDS/PCS samples. The 13C chemical shift values of the decomposed peaks are almost same as those of LDS and PCS. This shows that there are no new peaks with signi®cant intensity formed by curing LDS/PCS(5/5) sample. The peak intensities determined by such a decomposition are used in the T1Hr analysis, in order to elucidate the miscibility of blended LDS/PCS samples.
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
327
Table 1 Determined T1Hr values (experimental error is ^2 ms) of cured LDS/PCS samples with various mixture ratios determined at 258C Samples
Cured IPN LDS/PCS(8/2) Cured IPN LDS/PCS(5/5) Cured IPN LDS/PCS(2/8) Cured LDS Cured PCS a
T1Hr (ms) Methyl
Methylene
Phenyl (total)
Phenyl (LDS)
Phenyl (PCS)
34 40 46 39 38
32 42 49
35 39 47 38 37
36 36 43
29 a 38 46
38
Experimental error is larger than other values because of low signal -to-noise ratio.
Fig. 5 shows partially-relaxed 13C spectra of cured LDS/PCS(5/5)(a) and cured LDS/PCS(2/8)(b) as a function of spin locking time t by the pulse sequence for the T1Hr measurement (Fig. 2(a)), where the asterisks indicate the spinning sidebands [cured LDS/PCS(5/5) not shown]. It is seen that the intensities of individual peaks (M(t )) decay with an increase in t . As these decays follow equation M t=M 0 exp 2t=T1Hr ; where M(0) is the peak intensity at t 0: The plots of ln [M(t)/M(0)] for the methyl, methylene and phenyl carbons as obtained from the decomposition of the overlapped peaks by computer®tting, as mentioned-above against t are shown in Fig. 6 and then T1Hr values of the corresponding carbons are determined from its slope. From these T1Hr curves, it was found that the relaxation behavior for the methyl, methylene and phenyl carbons is a single exponential decay. This means that molecular motions of these carbons are a single component. The determined T1Hr values of cured LDS/PCS samples are summarized in Table 1. The T1Hr values for the methyl, ethylene and phenyl carbons of each sample is, 34 ms for LDS/PCS(8/2), 40 ms for LDS/PCS(5/5) and 47 ms for LDS/PCS(2/8). These values for the three functional groups are close to each other in the sample. This means that intramolecular 1H spin diffusion occurs in cured LDS/PCS samples. On the other hand, the T1Hr values in unblended LDS are 39 and 38 ms for the methyl and phenyl carbons, respectively, and in unblended PCS 38, 38 and 37 ms for the methyl, methylene and phenyl carbons, respectively. It was found that the T1Hr values of LDS/ PCS(8/2) were smaller than those of unblended LDS and PCS. On the other hand, the T1Hr values of LDS/PCS(5/5) are almost same as unblended LDS
and PCS, and those of LDS/PCS(2/8) are much larger than them. Dipole±dipole interactions between nuclear spins predominantly govern relaxation mechanism in polymer systems. Dipole±dipole interaction Hd is expressed by X Hd h2 =8p2 gi gj 3cos2 u 2 1 Ii Ij 2 3Iiz Ijz =rij3 1 where h is the Planck constant, rij is the internuclear distance between nuclei i and j, u is the angle between the vector i±j and the magnetic ®eld, g i and g j are the magnetogyric ratio of nuclei i and j, respectively, and I is nuclear spin operator. In general, the 1H NMR relaxation is mainly induced by dipole±dipole interaction and is greatly affected by internuclear distance of nuclear spins ¯uctuating in the magnetic ®eld. Ê However, interatomic distance more than ca. 5 A does not affect almost the 1H NMR relaxation as seen from Eq. (1) because Hd is proportional to r 23. Thus, the NMR relaxation times, in principle, provide information about interatomic distance r, if r is less Ê . From such a background, we are than ca. 5 A concerned with the miscibility between the LDS and PCS portions in blended LDS/PCS samples. Table 2 The density of cured LDS/PCS samples with various mixture ratios determined at 258C Samples
Density (g/cm 3)
Cured LDS Cured IPN LDS/PCS(8/2) Cured IPN LDS/PCS(5/5) Cured IPN LDS/PCS(2/8) Cured PCS
1.258 1.258 1.216 1.161 1.019
328
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
329
Fig. 7. (continued)
The intermolecular distance between the LDS and PCS portions in blended LDS/PCS IPN systems predominantly affects T1Hr Ð considered here, which is closely related to the miscibility. This may be supported by the experimental results on the density of the blended LDS/PCS samples as summarized in Table 2. The density of LDS/PCS(8/2) is the largest of three kinds of blended LDS/PCS samples. Therefore, it can be said that the intermolecular distance between LDS and PCS in LDS(8/2) may be the shortest one in three kinds of blended LDS/PCS IPN samples. This result suggests that intermolecular dipole±dipole interaction affects ef®ciently T1Hr in LDS/PCS(8/2).
From these experimental results, the miscibility of LDS and PCS in LDS/PCS(8/2) is predicted to be the highest. 3.2. 2D 13C± 13C and 1H± 1H exchange NMR experiments In order to evaluate the intermolecular distance between LDS and PCS in blended LDS/PCS systems, 13 C± 13C and 1H± 1H spin diffusion experiments were carried out. Fig. 7(a) shows 13C± 13C 2D exchange NMR spectrum of LDS/PCS(5/5) with a mixing time of 500 ms. In order to recognize the cross
Fig. 7. (a) 13C± 13C 2D exchange NMR spectrum of LDS/PCS(5/5) with mixing time 500 ms. Expanded spectra of the aliphatic region (b) and phenyl region (c).
330
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 8. 1H CRAMPS NMR spectra of unblended LDS and PCS, and blended LDS/PCS(8/2).
peaks as appearing due to intramolecular interaction and intermolecular interaction, the expanded spectra of the aliphatic and phenyl regions are shown in Fig. 8(b) and (c), respectively. As seen from these spectra, cross peaks were observed between the methyl carbon and ethylene carbon in the aliphatic region and between PCS-C1 and PCS-C2, PCS-C1 and LDS-C1, PCS-C1 and LDS-C2, LDS-C3 and LDS-C1, LDS-C3 and LDS-C2, and LDS-C3 and PCS-C3 in the phenyl region. The peak assignments follow those made by
computer-®tting as shown in Fig. 4. These cross peaks are assigned to intramolecular 13C± 13C spin diffusion in LDS and in PCS portions. In 13C± 13C 2D exchange NMR spectra of LDS/PCS(8/2) and LDS/PCS(2/8), intramolecular spin diffusion was clearly observed. On the other hand, no cross peak, which is induced by intermolecular 13C± 13C spin diffusion between PCS-C1 and LDS-C3 carbons in LDS/PCS(8/2) appears. As expected the miscibility is much higher, because the intensity of PCS-C1 peak is very weak.
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
331
Fig. 9. 1H± 1H 2D exchange NMR spectra of LDS/PCS(8/2) with a mixing time of 100 ms (a). (b) Expanded spectra of the cross peak region (b).
Fig. 8 shows 1H CRAMPS NMR spectra of unblended LDS and PCS, and blended LDS/PCS(8/ 2). The aliphatic proton signal and phenyl proton signal are clearly observed, of which the 1H chemical shift values are indicated on the peak top. As seen from these spectra of unblended LDS and PCS, the 1 H chemical shifts of the methyl and phenyl protons of LDS appear at upper®eld than those of PCS. On the other hand, in LDS/PCS(8/2) the peaks of the methyl and phenyl protons in LDS portion and PCS portion overlapped with each other. The apparent top position of the signal changes by changing the mixture ratio. The positions of the overlapped peaks in LDS/PCS(8/ 2) are between those in LDS and PCS. These 1H
chemical shift behaviors will be used for 1H± 1H 2D exchange NMR spectral analysis. Fig. 9(a) shows 1H± 1H 2D exchange NMR spectrum of LDS/PCS(8/2) with a mixing time of 100 ms. The cross peaks which are derived from the existence of 1H± 1H spin diffusion between the phenyl protons and CH3 protons, and the phenyl protons and the CH2 protons appear clearly. For convenience, expanded spectrum in the cross peak region is shown in Fig. 9(b). The dashed lines indicate 1H chemical shift positions of unblended LDS and PCS. These values become reference data for understanding the observed cross peaks due to 1H± 1H spin diffusion. The peak top at around 1H chemical shift position X1; X2
332
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
Fig. 9. (continued)
20:8; 6:9 ppm may be assigned to intramolecular spin diffusion between the CH3 protons of LDS and the phenyl protons of LDS because of the most intense peak. The shoulder peak at 1H chemical shift position X1; X2 20:4; 7:2 ppm may be assigned to intramolecular spin diffusion between the CH3 and CH2 protons of PCS and the phenyl protons of PCS because of the next most intense peak. The remaining shoulder peak at 1H chemical shift position X1; X2 20:4; 6:9 ppm may be assigned to intermolecular spin diffusion between the CH3 and CH2 protons of PCS and the phenyl protons of LDS because of the very weak peak. The two cross peaks at X1; X2 20:4; 6:9 ppm and (20.4, 7.2 ppm) positions can be signi®cantly recognized by making careful spectral analysis of its region. In the other blend LDS/PCS samples, a corresponding shoulder peak at 1H chemical shift position X1; X2 20:4; 6:9 ppm does not appear. From these experi-
mental results, the intermolecular 1H± 1H spin diffusion occurs in LDS/PCS(8/2) only. This suggests that the interchain distance between LDS and PCS portions is almost within the order of several 10 nm. 4. Conclusions The phase structure and miscibility of silicon-based interpenetrating polymer networks composed of LDS and PCS portions were successfully elucidated by solid-state 13C CP/MAS NMR, 1H CRAMPS and 2D 1 H± 1H exchange NMR. The 13C CP/MAS NMR experimental results suggest that the intermolecular dipole±dipole interaction occurs signi®cantly in LDS/PCS(8/2) because of short intermolecular distance between LDS and PCS portions. It was suggested from the observation of intermolecular spin diffusion in the 2D 1H± 1H exchange NMR
M. Kobayashi et al. / Journal of Molecular Structure 602±603 (2002) 321±333
spectrum in LDS/PCS(8/2) that the interchain distance between LDS and PCS portions is almost within the order of several 10 nm. References [1] B.M. Novak, C. Davies, Macromolecules 24 (1991) 5481. [2] B. Wang, G.L. Wilkes, J. Polym. Sci. Polym. Chem. 29 (1991) 905. [3] Y. Chujo, T. Saegusa, Adv. Polym. Sci. 100 (1992) 11. [4] D. Klempner, L.H. Sperling, L.A. Utracki, Interpenetrating Polymer Networks, Advances in Chemical Series 239, American Chemical Society, Washington, DC, 1994.
333
[5] L.H. Sperling, Interpenetrating Polymer Networks and Related Materials, Plenum, New York, 1981. [6] M. Tsumura, K. Ando, J. Kotani, M. Hiraishi, T. Iwahara, Macromolecules 31 (1998) 2716. [7] S. Kuroki, K. Okita, T. Kakigano, J. Ishikawa, M. Itoh, Macromolecules 31 (1998) 2804. [8] R.A. Komoroski (Ed.), High Resolution NMR Spectroscopy of Synthetic Polymers in Bulk VCH Publishers, Deer®eld Bleach, FL, 1986. [9] I. Ando, T. Yamanobe, T. Asakura, Prog. NMR Spectrosc. 22 (1990) 349. [10] K. Kurosu, S. Ando, H. Yoshimizu, I. Ando, Ann. Rept. NMR Spectrosc. 28 (1994) 189. [11] I. Ando, T. Asakura (Eds.), Solid State NMR of Polymers Elsevier Science, Amsterdam, 1998.