A study on the characteristics of antiplasticized polycarbonates and their optical disk substrates.
Recently, storage media with large capacities for digital information have become necessary, with the development of satellite digital broadcasting or broadband communication of the Internet. Particularly, optical disks such as CD-R (Recordable Compact Disk), MO (Magneto Optical Disk), or DVD (Digital Versatile Disk) could be considered widely accepted items in the consumer market because of their excellent reliability for long-term storage, ease of handling, and low cost. Optical disk substrates are generally produced by injection molding of a transparent polymer using specially designed injection machines and molds. The precise replication of the geometry of submicron-sized "pits" or "grooves." originally formed on a stamper, onto a molded substrate is necessary to obtain better signal quality as an optical disk final product. Until now, PC, PMMA (poly(methyl methacrylate)), or cyclic polyolefin has been employed as a raw material for optical disk substrates (1). Of these, PC has been recognized as the most balanced material from the point of view of transparency, mechanical strength, heat capability, dimensional stability, cost, and material availability. Especially, bisphenol A backboned PC has been widely used in the optical disk production industry. The fluidity of the PC designed for optical disks is well adjusted to maintain excellent replication of pits or grooves and to reduce internal stress causing birefringence in the substrate so that the current PC is sufficient for the present CD-R, MO, or DVDs.
However, for the next generation of higher storage density optical media, more precise injection molding techniques are required because of smaller pit lengths and narrower track pitch lengths. In addition, little birefringence and better flatness of substrates should also be achieved as required by specification for the next generation of optical disks. Furthermore, in the case of the axial asymmetry structure of the disk (2), little water absorption, which is preferable for preventing warpage due to uneven water swell from each surface, is necessary for a polymer to form a suitable substrate. If a polymer is used for a thin transparent cover layer as part of the first surface optical disk (2), higher surface hardness of the polymer is also required to avoid scratches from accidental contact with pickup lenses. Therefore, an innovative polymer to meet the requirements is expected as well as the development of injection molding technology. Over the past decades, a considerable number of studies have been conducted to improve the properties of polymers for use as optical disk substrates concerning their optical, physical, or processing properties (1, 3, 4). Although great efforts have been made to improve the properties, it has been more likely that other disadvantages have been introduced when the polymers were modified in some way. For instance, decreasing the molecular weight of the polymer to obtain better fluidity causes brittleness of the substrate or poor heat resistance of the polymer. Innovative molecular architecture to reduce the optical stress constant may sacrifice processability or raise the cost of the material. Therefore, an effective method to balance all properties of polymers has never been available for higher density optical disks.
Recently, the author tried to apply an "antiplasticized" PC to optical disk substrates (5-9). It was found that the addition of small amounts of "antiplasticizer" to a general PC resulted in better qualities of optical disk substrates while maintaining its transparency and low cost. Jackson and Caldwell originally performed studies on antiplasticizers and their composites in the 1960s (10-12). They defined "antiplasticization" as the decrease of glass transition temperature, simultaneous mechanical stiffening, and embrittlement by the addition of particular substances to PC or some other glassy polymers, which are rigid and have polar structures. Afterwards, some groups conducted studies of several antiplasticizers and their polymer composites concerning polymeric antiplasticization of PC with polycaprolactone (13), physical properties with mobility of antiplasticizer (14), slightly plasticized poly(vinyl chloride) (15), secondary loss transition behavior (16, 17), and the discussion of local intermolecular structure (18). Jackson and Caldwell also summarized requirements regarding the polarity, glass transition temperature, rigidity, and dimensions of antiplasticizers (11, 12). They interpreted the action of antiplasticization as the lowering of the free volume of the polymer matrix and consequently by restriction of local non-cooperative in-chain molecular motions. However, nobody has applied the antiplasticized polymer to an optical disk substrate. In previous work (5-9), the author tried to apply the characteristic features of the antiplasticized blends to a raw material for optical disk substrates; that is, decreasing glass transition temperature and increasing stiffness of the blends simultaneously. It was thought that the decreasing glass transition temperature of the blends corresponded to an equivalent effect to increase mold temperature, which allows obtaining of better microreplication of pits or grooves and reducing internal stress causing birefringence. Similarly, it was expected that the increased stiffness resisted thermal deformation of substrate driven by the temperature difference between mold and room temperatures so that flat substrates were obtained by injection molding. These great advantages are achieved by only small addition of antiplasticizer to an existing polymer, which is significantly more cost-effective than modifying monomer molecules. The most important characteristic of the antiplasticizer is that it acts like a common "plasticizer" at the melt state and as a "reinforcing filler" below glass transition temperature. These features predict a great contribution to simultaneously produce efficient microreplication, low birefringence, flat substrate, and little water absorption, which are all important properties for the next generation of optical disks.
M-tPh was chosen as an antiplasticizer for the PC in the present study. Strictly speaking, m-tPh is not an antiplasticizer as defined by the previous researchers (10-12), since a polar structure is absent in the molecule of m-tPh. It causes: however, decreasing glass transition temperature and increasing stiffness in the PC. It also has excellent compatibility with PC and is easily available as an industrial product. Therefore, the author proposed the possibility of antiplasticized PC with the addition of m-tPh to optical disk substrates. The main purpose of this paper is to study the characteristics of antiplasticized PCs and their optical disk substrates. The relationship between properties of polymers and their optical disk substrates, especially the ability to microreplicate, birefringence, or flatness of substrate, has been an important issue for scientific and industrial interest. Only a few attempts; however, have so far been made in these areas. For instance, Yoshii et al. (19) proposed an equivalent thickness of the vitrified layer concerning transcription of grooves of optical disk substrates. However, no one has conducted systematic studies of microreplication, birefringence, and flatness of molded substrates of optical disks in terms with the properties of polymers. From the present experimental results, it was suggested that viscoelastic property behavior of the polymer at the melt state and during cooling from the melt to room temperature highly reflected the observed microreplication, birefringence, and the flatness of the molded substrates. The author proposed that the ideal viscoelastic property model, that is, exhibiting lower modulus above glass transition temperature and higher modulus below glass transition temperature for the polymer of interest, led to better microreplication, lower birefringence, and better flatness of the substrates. It was demonstrated that the viscoelastic property behavior of the present antiplasticized PC accorded well with the proposed model.
The molecular structures of bisphenol A backboned PC and m-tPh are shown in Fig. 1. The PC for optical use, prepared by conventional interfacial polymerization, was obtained as a powder from Mitsubishi Engineering-Plastics Corp.; its viscosity-averaged molecular weight was 16,000, and it is a typical PC used for optical disk substrates. The m-tPh was obtained from Tokyo-Kasei Corp. as a reagent with a purity of 99%. Zero to five percent by weight of m-tPh was mixed independently in the PC by melt mixing using an extruder (Isuzu Corp.) with 40-mm [phi] single screw at 270[degrees]C, followed by cooling in water and then pelletizing. The pellets thus obtained were dried for 6 hours at 120[degrees]C and then were subjected to injection molding of a substrate with 120 mm diameter and 0.6 mm thickness for a DVD (Digital Versatile Disk). The transmittances for the PC composites with respect to wavelengths of the molded substrate at 0.6 mm thickness were measured using a UV spectrometer (UV-3100PC, Shimazu Corp.). In order to avoid extra diffraction from the formed grooves on a substrate; the incident light source was applied on a mirror portion of the disks. The glass transition temperatures were determined by differential scanning calorimetry (SSC-5100, SII) of the pellets on heating at 10[degrees]C/min immediately after rapid cooling at 50[degrees]C/min from the melt at 260[degrees]C. The flexural modulus and tensile modulus with elongation ratio as specially designed molded specimens were obtained by the general methods of ISO 178 and ISO 527-1, respectively. Apparent viscosities at 320[degrees]C with apparent shear rates were collected using a capillary viscometer (CAPIROGRAPH 1C, TOYO SEIKI CORP.). A capillary orifice with 0.5 mm diameter and 15 mm length was used. The measurement temperature of 320[degrees]C was chosen since it corresponded to a practical setting cylinder temperature for an injection molding machine of optical disk substrates. The injection molding machine for a DVD substrate was SD40 (Sumitomo Heavy Industries Corp.) with a hydraulic pressure system. A stamper of DVD-R (Recordable DVD; track pitch = 0.74 [micro]m, groove depth = 160 nm) was used and was attached onto a mirror part of a mold on the injection molding machine. The molding parameters are indicated in the text. Basically, mold temperatures were changed so as to investigate the properties of the substrates. The microreplication was determined by the measurements of groove depth of the molded substrates using an atomic force microscope (AFM, NV2100. Olympus Corp.). The replication was obtained as the groove depth; that is, the height between the top and the bottom of the grooves on the substrates. The maximum birefringences with a single pass were measured as in-plane mode by a vertical incident source of 632.8 nm using a birefringence tester (ADR-130N, ORC). The in-plane birefringence reflects the difference in refractive indices in mutually perpendicular directions in plane of the substrate multiplied by the thickness of the substrate. The laser beam was applied to four spots along the circumference at distances of 25 mm, 37 mm. 49 mm, and 57 mm in the radial direction from the center of the substrate. The maximum absolute value was chosen from them. The flatness of the molded substrate was determined as a radial tilt angle of the substrate using an automatic tilt angle tester with a laser beam at 780 nm (S3DH-13N, Admon Science Corp.). Radial tilt angle reflects radial warpage deformation of a substrate. The laser beam was applied to the four spots along the circumference at distances of 25 mm, 37 mm, 49 mm, and 57 mm in the radial direction from the center of the substrate, measuring the tilt angle along the radial direction. The maximum absolute value of the radial tilt was chosen among them. The water absorptions of the substrates were determined as the weight change of substrates when dipped into water by the method of ASTM D-570. The surface hardness was obtained using a micro Vickers hardness tester (HM 124, AKASHI CORP.) on the mirror portion of the molded substrates with 9.8N load in 10 sec. A conventional PMMA of a well-known high surface hardness polymer was also measured as a reference.
[FIGURE 1 OMITTED]
The temperature dependencies of tan [delta] at low temperatures. representing local molecular interaction, were observed using a dynamic viscoelastometer (Rheogel E-4000, UBM Corp.). The specimen was sliced from the molded substrate and the measurements were in tensile mode at 10Hz with a heating rate of 2[degrees]C/min. For the purpose of relating the viscoelastic property behavior at melting and during cooling from the melt to the properties of substrates, one should observe the change of modulus over the entire range from the melt state down to room temperature. Therefore, a slit shear modulus measurement (same equipment as tensile mode) was employed to detect shear stress between the melting resin and the slit plate. However, in the vicinity of glass transition and below it, there was uncertainty in the shear stress data because slippage between the almost solidified resin and the slit plate occurs. Hence, to obtain a more precise modulus below glass transition temperature, a general tensile modulus at 10Hz with heating rate of 2[degrees]C/min was employed, instead of the shear modulus on cooling. As a result, the viscoelastic behavior during cooling from the melt state was discontinuous above and below the glass transition temperature because of the difference in the testing modes.
[FIGURE 2 OMITTED]
RESULTS AND DISCUSSION
Figure 2 illustrates the flexural moduli and the glass transition temperatures of the blends with the content of m-tPh. The flexural modulus increased and the glass transition temperature decreased simultaneously with an increasing m-tPh content. This looks unusual for a conventional plasticizer, but is prominent evidence of the antiplasticizing effect of m-tPh on PC. Figure 3 shows a tensile stress-strain curve for the composites, indicating that the addition of m-tPh induced the embrittleness of the PC. This also has good agreement with the previous work that the antiplasticization led to the embrittleness of the matrix polymer (16). The effect of the increase of modulus is interpreted as a restriction of molecular motion induced by m-tPh's filling of the free volume of PC. The previous researchers proposed two mechanisms of antiplasticization, that is, the polar interaction between antiplasticizer and polymer, and the antiplasticizer's filling of the free volume of a polymer (10). It can be reasonably concluded that the latter mechanism is applicable for the present system since the m-tPh has no polar molecular structure. Filling the free volume with m-tPh may be the sole cause of antiplasticization in the present system. It would be expected that the blends became stiffer by antiplasticization so that the molded substrate would resist thermal deformation taking place during cooling from mold temperature down to room temperature after being taken out of the mold. In contrast, the m-tPh as antiplasticizer also affected the relaxation of glass transition of the blends. The lower glass transition temperatures of the composites were obtained for the higher contents of m-tPh, shown in Fig. 2. Previous work suggested that an antiplasticizer might not decrease glass transition temperature as appreciably as a common plasticizer (10). The m-tPh as varied from 0% to 5%; however, the glass transition temperature for the antiplasticization fell rapidly. Better microreplication in injection molding of optical disk substrates could be predicted for the blends with decreasing glass transition temperatures as demonstrated in the following section. Because decreasing the glass transition temperature is equivalent to increasing the mold temperature, this might delay the development of solidification of the molten resin at the very surface of a substrate during filling on a stamper.
[FIGURE 3 OMITTED]
As shown in Fig. 4, viscosities at 320[degrees]C decreased with the content of m-tPh all over the regions of apparent shear rates of interest. This suggests that the antiplasticized PCs with m-tPh exhibit improved processability for injection molding of optical disk substrates. The m-tPh as an antiplasticizer acts as a common plasticizer at the melt state, since the molecular interaction between m-tPh and PC polymer chains is weaker than those of other polar antiplasticizers, owing to the absence of a polar structure for the m-tPh. The forces between m-tPh and PC are easily broken by thermal energy, so the m-tPh acts as a common plasticizer, which lubricates segmental movement of polymer chains of PC at the melt state. Also, at higher contents of m-tPh, the shear rate dependency is weaker than at low contents, suggesting a uniform flow property for all areas of optical disks. This is an appreciable advantage for the adjustment of process parameters to produce optical disks.
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
The transmittance in 0.6 mm thickness with wavelengths for 0%, 2%, and 5% additions of m-tPh is shown in Fig. 5. Even at 5% addition, the transmittance is well above 80% at practical regions of wavelength (above 350 nm) for optical disks, suggesting maintenance of the transparency of neat PC. This result suggests a good compatibility of m-tPh with the PC.
Figure 6 represents AFM images for comparison of microreplication of grooves for the 0% and 5% additions of m-tPh at the mold temperature of 108[degrees]C. Obviously, microreplication with the 5% addition of m-tPh is remarkably increased even at the same mold temperature. The decreased glass transition temperature of the antiplasticized PC inhibited development of a vitrified layer at the very surface of the disks during filling into the cavity. The observed groove depths of molded substrates at the mold temperature of 108[degrees]C with respect to the radial position of the disks are shown in Fig. 7. Better microreplication was obtained for the PC with a higher content of m-tPh. reflecting lower glass transition temperatures for the blends. It was particularly manifested for the outer radius position, which generally showed poor microreplication. This may be caused by both the decreased glass transition temperature and the improved flow properties. The mold temperature dependencies of microreplication for the blends are given in Fig. 8. This result indicates that the mold temperature could be substantially decreased for the blends to achieve a certain magnitude of microreplication for the neat PC, which is preferable for obtaining a flat substrate, as described in the following section.
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
The radial tilt angles of the substrates, representing the flatness of the substrate, were compared at the resultant groove depths, resulting from changing mold temperatures, shown in Fig. 9. The increased tilt angles with increasing groove depths indicate thermal deformation during cooling from the applied mold temperature, which achieves better microreplication, to room temperature after being taken out from the mold. Since the thermal deformation has been considered to be caused by the thermal stress generated by the temperature difference between mold and room temperature for the optical disk (20), a lower mold temperature is better for obtaining a flat substrate. As a consequence, the observed improvement of the flatness for the antiplasticized PC was mainly due to the decreased mold temperature. It should be noted that the stiffness (modulus) became higher at regions between the mold temperature and room temperature for the antiplasticized PC, as described later. Probably, the increased stiffness of the blends also contributed to resisting thermal deformation.
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
In-plane birefringences of the molded substrates for the composites are shown with respect to mold temperatures in Fig. 10. Generally, birefringence is strongly affected by mold temperature, since birefringence is caused by internal stress of the substrate consisting of molecular orientation stress, which occurs during filling into a cavity, and by which occurs thermal stress during cooling from mold temperature to room temperature (20). It is apparent that lower birefringences were obtained for higher contents of m-tPh. This may be attributed to the lower viscosities for the higher contents of m-tPh, as described above. Because the molecular orientation stress during filling governing birefringence is mainly induced by the high viscosity of the polymer. The lower viscosities are attributed to the plasticizing action of m-tPh at the melt state and hence they resulted in the lower molecular orientation stress causing lower birefringence. This result is also interpreted as that the thermal stress during cooling is more relaxed by the reduction of glass transition temperature with the presence of m-tPh.
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
Lower water absorption was attained with the increasing m-tPh content, shown in Fig. 11. This is in good agreement with a reported result of antiplasticization (17).
The surface hardness of the molded substrates was measured for the composites, given in Fig. 12. The surface hardness increased with the content of m-tPh approaching PMMA, which is known to have excellent surface hardness as a transparent polymer. This result may be explained by the contribution of the increased stiffness of PC by addition of m-tPh.
In antiplasticization systems, it has been demonstrated that the modulus of the composite was increased while ultimate elongation and impact strength were decreased (17). According to other works (14), depression of [beta]-relaxation transition, which was expressed by tan [delta] at low temperatures of dynamic viscoelastic measurements, was observed for the antiplasticized polymers. They demonstrated that the loss of ductility of antiplasticized blends was associated with a reduction in subsidiary relaxation processes or their elimination. However, this depression of [beta]-relaxation transition is evidence of the restriction of local non-cooperative molecular motions between antiplasticizer and polymer, which leads to an increase of modulus above room temperature. To ensure the mechanism of the increased modulus in the present system, the temperature dependencies of tan [delta] for the blends were observed at temperatures between--150[degrees]C and 0[degrees]C in the dynamic viscoelastic measurements, shown in Fig. 13. Generally, the [beta]-relaxation transition is located below room temperature for bisphenol A PC at some frequencies (16). The peaks of tan [delta] were depressed and the peak positions slightly shifted to lower temperatures with increasing m-tPh content, which is in good agreement with a previous report (14). The results indicate the suppression of local molecular mobility with the presence of m-tPh, causing increased modulus at and above room temperature.
[FIGURE 12 OMITTED]
The shear storage modulus (G') for the PC composites with addition of m-tPh at the melt and during cooling from the melt to glass transition temperature is shown in Fig. 14. Obviously, G' decreased with the content of m-tPh at the temperature regions from 260[degrees]C to 150[degrees]C. As discussed in the section on flow properties, lower viscosities for the PC with the addition of m-tPh exhibited lower birefringence. Therefore, the observed birefringences are controlled by G' as well as the viscosity for the blends at the melt state. Similarly, the microreplication of grooves for substrates could be determined during cooling from the melt state to glass transition temperature. This is because in injection molding, the molten resin is cooled rapidly when filled into the cavity and is solidified until the resin reaches the glass transition temperature at which point the polymer melt ceases to be soft. Apparently, G' for the blends at the regions from the melt to 150[degrees]C also decreased with the content of m-tPh. It suggests better microreplication for the PC with the addition of m-tPh. From these results, it can be deduced that the better microreplication for the PC with the addition of m-tPh is caused by both the lowering the glass transition temperatures and the lowering of G' during filling of the cavity with molten resin.
[FIGURE 13 OMITTED]
[FIGURE 14 OMITTED]
Similarly, tensile modulus at the regions from room temperature to the glass transition temperature for the blends is shown in Fig. 15. Usually for the injection molding of optical disks, the mold temperatures are adjusted to below the glass transition temperature of the employed polymer in order to avoid softening of the substrate immediately after being taken out from the mold. The deformation of the substrate, such as warpage or distortion, might be caused by the thermal stress induced during cooling from mold down to room temperature. If the modulus of the polymer becomes higher during cooling from the mold temperature to room temperature, it would resist thermal deformation and eventually the substrate becomes flat. The fact that the PC with the addition of m-tPh showed higher modulus at the temperature regions between mold and room temperature would result in the better substrate in flatness, supporting the results of flatness as described above.
[FIGURE 15 OMITTED]
[FIGURE 16 OMITTED]
Finally, a general formulation was attempted to depict a schematic representation of viscoelastic property model during cooling for a polymer to achieve better microreplication, lower birefringence, and better flatness of a molded substrate. As given in Fig. 16, the dotted circles denote the temperature regions of controlling birefringence, microreplication, and flatness, respectively. The ideal and the conventional behavior are drawn by bold and thin curves, respectively. The ideal polymer exhibits lower modulus above glass transition temperature and higher modulus below glass transition temperature compared with that of a conventional polymer, which exhibits better microreplication with lower birefringence and flatter substrate, respectively. The viscoelastic properties of the present composites are in good agreement with the proposed model as described above.
PC blends with the addition of small amounts of m-tPh showed decreased glass transition temperature and increased stiffness. The blends exhibited better flow properties, better microreplication, lower birefringence, flatter substrate, little water absorption, and higher surface hardness compared with those of the neat PC, which are all important properties for the next generation of high-density optical disks. It was found that the increased stiffness was associated with the elimination of secondary mechanical loss transition at low temperatures. The ideal viscoelastic property behavior of a polymer for an optical disk substrate to achieve better microreplication, better flatness, and low birefringence was proposed; that is, exhibiting lower modulus above glass transition temperature and higher modulus below glass transition temperature for the polymer of interest.
The author acknowledges his debt to Mitsubishi Engineering-Plastics Corp. for providing an opportunity to report the present work and to all the members of the Optical Disk Team at the technical center of MEP for their stimulating discussions.
1. A. Tohdo and H. Kajiura. Plastics, 38(41), 7 (1987).
2. Ichimura, F. Maeda, K. Osato, K. Yamamoto, and Y. Kasami, Proceedings of SPIE, 3864. 228 (1999).
3. S. Ushio, Plastic Age, Nov. 137 (1985).
4. R. T. Woodhams. Polym. Eng. Sci., 25, 446 (1985).
5. M. Ueda, Proceedings of SPIE. Optical Data Storage 2001. 4342, 39 (2001).
6. M. Ueda. Technical Digest. Joint International Symposium on Optical Memory and Optical Data Storage Topical Meeting 2002, 33 (2002).
7. M. Ueda, Polymer Preprint, Japan Polymer Society, 49(14), 4381 (2000).
8. M. Ueda, Technical Paper, Annual Meeting of Japan Society of Polymer Processing, 175 (2001).
9. M. Ueda, Japan Patent 2000-195099 and 2000-239513.
10. W. J. Jackson, Jr., and J. R. Caldwell, Adv. Chem. Ser., 48, 185 (1965).
11. W. J. Jackson, Jr., and J. R. Caldwell, J. Appl. Polym. Sci., 11. 211 (1967).
12. W. J. Jackson, Jr., and J. R. Caldwell, J. Appl. Polym. Sci., 11. 227 (1967).
13. M. Shuster and M. Narkis, Polym. Eng. Sci., 34, 1613 (1994).
14. S. E. B. Petrie, R. S. Moore, and J. R. Flick, J. Appl. Phys., 43(11). 4318 (1972).
15. N. Kinjo and T. Nakagawa, Polymer J., 4(2), 143 (1973).
16. L. M. Robeson and J. A. Faucher, Polym. Lett., 7, 35 (1969).
17. L. M. Robeson, Polym. Eng. Sci., 9, 277 (1969).
18. A. K. Roy, P. T. Inglefield, J. H. Shibata, and A. A. Jones, Macromolecules, 20, 1434 (1987).
19. M. Yoshii, H. Kuramoto, and K. Kato, Polym. Eng. Sci., 34, 1211 (1994).
20. H. Kikuchi and K. Koyama, Polym. Eng. Sci., 36, 1309 (1996).
Corporate Planning Department Mitsubishi Engineering-Plastics Corp. Yaesu-Daibiru Building, 1-1. Kyobashi 1 chome. Chuo-ku, Tokyo 104-0031, Japan
Presented at the joint program of ISOM (International Symposium of Optical Memory) and ODS (Topical Meeting of Optical Data Storage 2002), USA in July 2002, at the ODS 2001. USA in April 2001, at the Annual Meeting of the Society of Polymer Science, Japan in September 2000, and at the Annual Meeting of Japan Society of Polymer Processing, in June 2001.