You are on page 1of 7

Thin Solid Films 331 (1998) 222228

Structure and orientation control of organopolysilanes and their application to electronic devices
Shoji Furukawa
Department of Computer Science and Electronics, Kyushu Institute of Technology, 680-4 Kawazu, Iizuka-shi, Fukuoka-ken 820-8502, Japan

Abstract Molecular structure and packing of various organopolysilanes having symmetrical side-chains of alkyl substituents, such as poly(dimethyl silane), poly(di-ethyl silane), poly(di-propyl silane), poly(di-butyl silane), poly(di-pentyl silane), and poly(di-hexyl silane), are examined, and the method of orientation control of the polysilane thin lm is proposed. The application of the evaporated lm to the electro-luminescent device is also demonstrated. q 1998 Elsevier Science S.A. All rights reserved. Keywords: Polysilane; Organopolysilane; Poly(di-methyl silane); Poly(di-hexyl silane); Molecular structure; Electro-luminescent device

1. Introduction Polysilanes consist of a Si backbone and organic substituents. In the Si backbone, the bonded sigma-electrons are well delocalized along the chain direction, so the structure is considered to be the quasi one-dimensional electronic system. It has also been pointed out that they have a band structure similar to direct-gap-semiconductor [1], and show an efcient photo-luminescence [2]. Therefore, they are expected to be used for future optical and electronic devices. In this paper, molecular structure and packing of various organopolysilanes having symmetrical side-chains of alkyl substituents are examined, and the method of orientation control of the polysilane thin lm is proposed. The application of the evaporated lm to the electro-luminescent device is also demonstrated. Conventionally, GaAs-based semiconductors were used as an active layer in the light-emitting diodes (LEDs). These materials are considered to be not good for the human body as well as earth-scale environment. Therefore, it is better for us to use other materials, such as the present polymer, instead of GaAs-based semiconductors. 2. Theoretical calculation and experimental procedure Polysilanes investigated in this study were poly(dimethyl silane), poly(di-ethyl silane), poly(di-n-propyl silane), poly(di-n-butyl silane), poly(di-n-pentyl silane), and poly(di-n-hexyl silane). They were synthesized by Wurtz-type reductive coupling of corresponding dichlorosilanes. In the X-ray diffraction measurement, the wavelength

used was 0.154 nm, and its power was 40 kV 80 mA. The experimental X-ray diffraction pattern was compared with the theoretical pattern calculated using the computer program written in C language. The packing of the polymer chains was also checked using the computer graphic display. Poly(di-methyl silane) thin lms were obtained by evaporating the source powder material in the vacuum chamber [3]. The electro-luminescent (EL) device having a structure of aluminum/evaporated poly(di-methyl silane)/ ITO/glass substrate has been fabricated, and the EL spectrum was measured by changing the applied voltage. 3. Results and discussion Fig. 1ac shows the X-ray diffraction patterns for poly(dimethyl silane) [Si(CH3)2]n powder, source-melted lm, and theoretical calculations based on the reasonable molecular structure and the packing, respectively [4]. Fig. 2 shows the schematic diagram of the obtained crystal structure of poly(di-methyl silane). a, b, and c denote the crystallographic primitive translation vectors of the sub-cell. It has a monoclinic unit cell (sub cell) with a 0:745 nm, b 0:724 nm, c 0:389 nm, and g 67:18. In Fig. 2, the Si atoms are not illustrated, but the Si backbones are shown by the zig-zag lines (all-trans conformation). The lled circles indicate the center of the CH3 group, and the envelope shows the van der Waals radius of the CH3 group. The most important point is that the molecules are closely packed in space, as shown in Fig. 2. The second important point is that the van der Waals radius of CH3 groups (about 0.2 nm) is almost the same as that of the half of the period

0040-6090/98/$ - see front matter q 1998 Elsevier Science S.A. All rights reserved. PII S0 040-6090(98)009 23-7

S. Furukawa / Thin Solid Films 331 (1998) 222228

223

Fig. 3. Experimental (a) and theoretical (b) X-ray diffraction patterns for the powdered poly(di-ethyl silane) crystal.

Fig. 1. X-ray diffraction patterns for poly(di-methyl silane) powder (a), source-melted lm (b) and theoretical calculation based on the reasonable molecular structure and the packing (c).

(0.389 nm) along the c-axis. Therefore, each CH3 group is tightly bonded to the CH3 groups attached to the second nearest-neighbor Si atoms by van der Waals force. This causes well-dened all-trans backbone conformation. Fig. 3a,b shows the experimental [5] and theoretical Xray diffraction patterns for the powdered poly(di-ethyl silane) [Si(C2H5)2]n crystal. Fig. 3b was calculated on the assumption that the crystal has a monoclinic unit cell with a 0:820 nm, b 0:810 nm, c 0:399 nm, and g 84:88

Fig. 2. Schematic diagram of the crystal structure of poly(di-methyl silane). In the gure, the Si atoms are not illustrated, but the Si backbones are shown by the zig-zag lines (all-trans conformation). The lled circles indicate the center of the CH3 group and the envelope shows the van der Waals radius of the CH3 group. a, b and c denote the primitive translation vectors of the sub-cell.

Fig. 4. Schematic diagram of the crystal structure of poly(di-ethyl silane). In the gure, the large circles marked by a cross, the medium circles and the small lled circles indicate Si, C and H atoms, respectively. The envelope shows the van der Waals radius of the CH2 and CH3 groups. a, b and c denote the primitive translation vectors of the sub-cell. The conformation of the Si backbone is all-trans.

224

S. Furukawa / Thin Solid Films 331 (1998) 222228

Fig. 5. Experimental (a) and theoretical (b) X-ray diffraction patterns for the powdered poly(di-n-propyl silane) crystal.

(see Fig. 4). In Fig. 4, the large circles marked by a cross, the medium circles, and the small lled circles indicate Si, C, and H stoms, respectively. The envelope shows the van der Waals radius of the CH2 and CH3 groups. The conformation of the Si backbone is all-trans. It should be noted that the two C2H5 groups stretch asymmetrically due to the intramolecular steric hindrance of the CH3 groups bonded to the same Si atom, and each C2H5 group attaches to the side-chains bonded to the second nearest-neighbor Si atoms by van der Waals interaction.

Fig. 6. Schematic diagram of the crystal structure of poly(di-n-propyl silane). The notations are the same as those of Fig. 4. The conformation of the Si backbone is all-trans, which is the same as that of poly(di-methyl silane) and poly(di-ethyl silane).

The theoretical calculation for the powdered poly(diethyl silane) crystal was performed using the conguration shown in Fig. 4. As a result, the diffraction peaks indexed by (hk0) were well explained by the model. However, the peaks indexed by (hkl), el 0, were not explained. The broad peak appearing near 258 in Fig. 3a suggests that the unit cell is not well dened. One of the plausible origins is the phase shift of the nearest-neighbor polymer chain along the c-axis. Fig. 3b was calculated on the assumption that there were 12 congurations, in which the shift of the nearestneighbor chain along the c-axis was different. The theoretical pattern thus obtained is in good agreement with the experimental pattern shown in Fig. 3a. The peak near 258 in Fig. 3b corresponds to (111) diffraction. However, it contains the ambiguity of the phase shift along the c-axis. Therefore, the Miller index is not indicated for the peak. Fig. 5a,b shows the experimental [5] and theoretical Xray diffraction patterns for the powdered poly(di-n-propyl silane) [Si(C3H7)2]n crystal. Fig. 5b was calculated on the assumption that the crystal has a tetragonal unit cell with a b 0:980 nm, c 0:399 nm (see Fig. 6). Similar to Fig. 3b, Fig. 5b was obtained by superimposing 12 diffraction patterns, which were calculated by changing the phase shift of the nearest-neighbor chain along the c-axis. Therefore, the half-width of the diffraction peaks near 24 and 298 is large in Fig. 5b. In the structural model shown in Fig. 6, the notations are the same as those of Fig. 4. The conformation of the Si backbone is all-trans, which is the same as that of poly(di-methyl silane) and poly(di-ethyl silane). Similar to poly(di-ethyl silane), the two C3H7 groups stretch asymmetrically due to the intra-molecular steric hindrance of the second CH2 group in the side-chains bonded to the same Si atom. In Fig. 5b, the intensity of the (200) and (020) reection is very weak. The diffraction intensity is determined by the crystallographic structure factor, which is deduced from the diffraction index as well as the positions of atoms in a unit cell. As shown in Fig. 6, the atoms are widely distributed between both the (200) and (020) planes. This indicates the increase of out-of-phase reection, causing the reduction of the diffraction peak intensity. Fig. 7a,b shows the experimental [6] and theoretical Xray diffraction patterns for the powdered poly(di-n-butyl silane) [Si(C4H9)2]n crystal. The dotted and solid curves in Fig. 7a are the original pattern and the pattern, in which the background is eliminated, respectively. Fig. 7b was calculated on the assumption that the crystal has an orthorhombic unit cell with a 1:284 nm, b 2:224 nm, c 1:390 nm (see Fig. 8). Fig. 8a,b are views from directions parallel and perpendicular to the Si chains, respectively. a, b, and c denote the primitive translation vectors of the sub-cell. The large open circle and small lled circle indicate Si and C atoms, respectively. H atoms are not illustrated in Fig. 8. The envelope shows the van der Waals radius of the CH2 and CH3 groups. The conformation of the Si backbone is 7/3 helix, which is different from that of poly(di-

S. Furukawa / Thin Solid Films 331 (1998) 222228

225

Fig. 9. Molecular structure of poly(di-n-butyl silane) proposed by Walsh et al. [7]. These are not correct. The details are explained in the text.

Fig. 7. Experimental (a) and theoretical (b) X-ray diffraction patterns for the powdered poly(di-n-butyl silane) crystal. The dotted and solid curves in (a) are the original pattern and the pattern, in which the background is eliminated, respectively.

reported in Ref. [7]. The two side-chains stretch symmetrically in both Fig. 9a,b. These models are not correct, because the theoretical diffraction pattern based on these models is not in agreement with the experimental pattern. Fig. 10a,b show the experimental [6] and theoretical Xray diffraction patterns for the powdered poly(di-n-pentyl silane) [Si(C5H11)2]n crystal. The dotted and solid curves in Fig. 10a are the original pattern and the pattern, in which the background is eliminated, respectively. Fig. 10b was calculated on the assumption that the crystal has a monoclinic unit cell with a 1:385 nm, b 2:380 nm, c 1:390 nm, and g 85:08 (see Fig. 11). Figs. 11a,b are views from directions parallel and perpendicular to the Si chains,

methyl silane), poly(di-ethyl silane), and poly(di-n-propyl silane). In this case, the two C4H9 groups also stretch asymmetrically due to the steric hindrance of the second CH2 group in the side-chains bonded to the same Si atom. Each C4H9 group attaches the side-chain bonded to the rst nearest-neighbor Si atom by van der Waals interaction, which is different from that of the polysilane crystal having all-trans conformation. Fig. 9 shows the schematic diagram of molecular structure of poly(di-n-butyl silane), which was

Fig. 8. Schematic diagram of the crystal structure of poly(di-n-butyl silane). (a) and (b) are views from directions parallel and perpendicular to the Si chains, respectively. The large open circle and small lled circle indicate Si and C atoms, respectively. H atoms are not illustrated in this gure. The envelope shows the van der Waals radius of the CH2 and CH3 groups. a, b and c denote the primitive translation vectors of the sub-cell. The conformation of the Si backbone is 7/3 helix.

Fig. 10. Experimental (a) and theoretical (b) X-ray diffraction patterns for the powdered poly(di-n-pentyl silane) crystal. The dotted and solid curves in (a) are the original pattern and the pattern, in which the background is eliminated, respectively.

226

S. Furukawa / Thin Solid Films 331 (1998) 222228

Fig. 11. Schematic diagram of the crystal structure of poly(di-n-pentyl silane). The notations are the same as those of Fig. 8. The conformation of the Si backbone is 7/3 helix.

respectively. Similar to Fig. 8, the large open circle and small lled circle indicate Si and C atoms, respectively, and H atoms are not illustrated in Fig. 11. The envelope shows the van der Waals radius of the CH2 and CH3 groups. The conformation of the Si backbone is 7/3 helix, which is the same as that of poly(di-n-butyl silane). The two sidechains attached to one Si atom stretch asymmetrically due to the intra-molecular steric hindrance. The important point in the poly(di-n-pentyl silane) crystal is that the fth C atom occupies a gauche-like position in order to increase the van der Waals interaction between the intra-molecular sidechains. Fig. 12 shows the experimental and theoretical X-ray diffraction patterns for the powdered poly(di-n-hexyl silane) [Si(C6H13)2]n crystal. The theoretical pattern was calculated on the assumption that the crystal has an orthorhombic unit cell with a 1:384 nm, b 2:398 nm, c 0:400 nm (see Fig. 13). There are two polymers in a unit cell. The conformation of the Si backbone is all-trans, which is the same as

Fig. 12. Experimental (lower) and theoretical (upper) X-ray diffraction patterns for the powdered poly(di-n-hexyl silane) crystal.

that of poly(di-methyl silane), poly(di-ethyl silane), and poly(di-n-propyl silane). Fig. 14 shows the summary of the molecular structure and the packing of polysilanes having all-trans conformation. The conformation of the Si backbone of poly(di-n-butyl silane) (PDBS) and poly(di-n-pentyl silane) (PDPeS) after treatment of a high pressure of 1500 MPa became all-trans [8]. In those cases, the two side-chains stretch asymmetrically as shown in the right two gures of Fig. 14, which is similar to those of poly(di-methyl silane), poly(di-ethyl silane), and poly(di-n-propyl silane). As shown in Figs. 113, the conformation of the Si backbone of poly(di-methyl silane), poly(di-ethyl silane), poly(di-n-propyl silane), and poly(di-n-hexyl silane) is alltrans, whereas that of poly(di-n-butyl silane) and poly(di-npentyl silane) is 7/3 helix at room temperature and atmospheric pressure. How does the chain length of n-alkyl substituents inuence the conformation of the backbone? The details are not clear at the present stage. However, the conformation of the Si backbone of poly(di-n-butyl silane) and poly(di-n-pentyl silane) after treatment of a high pressure of 1500 MPa became all-trans as shown in Fig. 14. The density of the crystal having all-trans conformation is larger than that of the crystal having 7/3 helical conformation. Therefore, it is strongly suggested that the conformation of the Si backbone of poly(di-n-butyl silane) and poly(di-n-pentyl silane) becomes all-trans at very low temperature, because the lattice vibrations are suppressed at low temperature, causing the structure with high density. Fig. 15 shows the X-ray diffraction patterns of poly(dimethyl silane) thin lms prepared by means of a vacuum evaporation technique. The thickness of these lms is several microns, and the substrates used are single-crystal Si(100) wafers, whose diffraction occurs more than 708 (Si(400) reection occurs at 2u t 708). As shown in Fig. 15, only the diffraction peaks indexed by (hk0) are observed for the lms obtained under high deposition rate, indicating that the Si backbone is parallel to the substrate surface. On the other hand, only the diffraction peak indexed by (002) is observed for the lms obtained under low deposition rate. In the latter case, the direction of the Si backbone is perpendicular to the substrate surface. The orientation of poly(dimethyl silane) chains is also controlled by changing substrate temperature during the evaporation [9]. Similar results were obtained for other substrates, such as fused silica and slide-glass. The lms shown in Fig. 15 were taken off from the substrate, and the powdered specimens were made from them. Then, their crystal structure was examined by the X-ray diffraction technique. As a result, it was conrmed that the crystal structure of the lm was the same as that of the source powder material. However, the molecular weight of poly(di-methyl silane) in the evaporated lm was very much reduced, and estimated to be [Si(CH3)2]n (n 10 t 20), which was deduced from the half-width of the (002) peak in the X-ray diffraction pattern. The reduc-

S. Furukawa / Thin Solid Films 331 (1998) 222228

227

Fig. 13. Schematic diagram of the crystal structure of poly(di-n-hexyl silane). There are two polymers in a unit cell.

tion of the average molecular weight in the lm was also conrmed by the blue shift of the optical absorption spectrum [9]. The dependence of the molecular orientation on the growth rate and substrate temperature during the evaporation [8] can be discussed as follows. In the recrystallization process on the substrate surface, the evaporated molecules having a low molecular weight move to their stable positions, and rotate their Si backbones to the equilibrium direction, when the evaporation speed is low and the substrate

Fig. 15. X-ray diffraction patterns of poly(di-methyl silane) thin lms prepared by means of a vacuum evaporation technique.

temperature is high. Therefore, it is considered that the equilibrium direction is perpendicular to the substrate surface. When the growth rate is high, the evaporated molecules cannot move sufcient distance on the growing surface, because of the high concentration of the molecules. This causes three-dimensional growth, and the Si backbones become parallel to the substrate surface. Using the evaporated lm of poly(di-methyl silane), the EL device was fabricated as stated in the Section 2. The structure of the EL device fabricated is shown in Fig. 16. The dimension of the device is 2 10 mm 2, and the thickness of the evaporated poly(di-methyl silane) is 0.7 mm. In this case, the polymer chains are perpendicular to the

Fig. 14. Molecular structure and packing of polysilanes having all-trans conformation. PDES, PDPrS and PDHS indicate poly(di-ethyl silane), poly(di-n-propyl silane) and poly(di-n-hexyl silane), respectively. The conformation of the Si backbone of poly(di-n-butyl silane) (PDBS) and poly(di-n-pentyl silane) (PDPeS) after treatment of a high pressure of 1500 MPa becomes also all-trans [7].

Fig. 16. Structure of fabricated electro-luminescent device.

228

S. Furukawa / Thin Solid Films 331 (1998) 222228

4. Summary and conclusion Molecular structure and packing of various polysilanes have been claried, and the method of orientation control of the polysilane thin lm has been proposed. The application of the evaporated lm to the electro-luminescent device has also been demonstrated. Acknowledgements The author would like to thank Professor Norihiko Kamata of Saitama University, Professor Reiji Hattori of Kyushu University, and Professor Fumiya Shoji of Kyushu Kyoritsu University for their helpful discussion. References
[1] J.W. Mintmire, Phys. Rev. B 39 (1989) 13350. [2] S. Furukawa, N. Kamata, R. Hattori, F. Shoji, Technical Report of The Institute of Electronics, Information and Communication Engineering of Japan, OME 97-1, 1997, p. 1 (in Japanese). [3] S. Furukawa, M. Nagatomo, H. Kokuhata, K. Takeuchi, H. Fujishiro, M. Tamura, J. Phys. Condensed Matter 4 (1992) 8357. [4] S. Furukawa, K. Takeuchi, Solid State Commun. 87 (1993) 931. [5] A.J. Lovinger, D.D. Davis, F.C. Schilling, F.A. Bovey, J.M. Zeigler, Polym. Commun. 30 (1989) 356. [6] F.C. Schilling, F.A. Bovey, D.D. Davis, A.J. Lovinger, R.B. Macgregor Jr., Macromolecules 22 (1989) 4648. [7] C.A. Walsh, F.C. Schilling, A.J. Lovinger, D.D. Davis, F.A. Bovey, J.M. Zeigler, Macromolecules 23 (1990) 1742. [8] S. Furukawa, K. Takeuchi, M. Shimana, J. Phys. Condensed Matter 6 (1994) 11007. [9] K. Takeuchi, M. Mizoguchi, M. Kira, M. Shimana, S. Furukawa, M. Tamura, J. Phys. Condensed Matter 6 (1994) 10705. [10] H. Suzuki, S. Hoshino, C. Yuan, M. Fujiki, S. Toyoda, N. Matsumoto, Thin Solid Films 331 (1998) 64.

Fig. 17. Effect of applied voltage on electro-luminescence spectrum.

substrate surface. Fig. 17 shows the EL spectrum and its dependence on the applied voltage. The intensity of the luminescence increases with an increase in applied voltage. Other devices, in which the Si backbones were parallel to the substrate surface, were fabricated, and their EL spectra were also measured. The observed spectra showed an increase of the component below 400 nm [10]. However, much more experiments are needed in order to investigate the effect of orientation on the EL spectrum. Further investigations are also needed in order to clarify the mechanism as well as to control wavelength and band-width of the spectrum.

You might also like