Crystallization in Unstretched MICROSCOPIC STUDY IN POLARIZED LIGHT E. N. Campbell and M. D. Allen U. S. Rubber Ce., Paeeaic, N. d.
T
HERE is considerable
Since many elastomers indicate, by their physical propmoisture condensing on the erties, that crystallization occurs even when such crystalinterest in the crystalback lens surface of the objeclization is not detectable by x-ray diffraction, a direct lization of rubbers both from tive. Consequently the obmicroscopic study in polarized light was undertaken. the practical and the theojective had to be removed This study confirms the presence of such crystallinity retical points of view. In the and its rear surface dried off not only in rubber but also in polybutadiene and some low first place, crystallization of about once an hour during raw r u b b e r c a u s e s i n c o n temperature copolymers. Furthermore, the same crystal continuous observations. venience in rubber processpattern is reproduced on melting and refreezing provided Results the intermediate melting temperature is not too high. ing and crystallization in the finished p r o d u c t s e v e r e 1y This indicates that x-ray diffraction is not a very sensiFigures 1, 2, and 3 show tive method for detecting small amounts of crystallinity the progress of crystallization limits its usefulness at low t e m p e r a t u r e s Secondly, in high polymers. The reproductionof the crystal pattern with time. The sample is a knowledge of degree and rate on refreezing shows that the molecular segments have film of raw rubber 0.5 to 1 of crystallization zts a function limited mobility even at room temperature; this may remicron thick prepared by dipof molecular structure is of quire revision of current theories of the origin of retracping a microscope slide in great importance in designing diluted Hevea latex and drytive forces in elastomers. synthetic elastomers and in ing. Photographs of the same developing field were taken after 24, - - theorim of their stress-strain relationships and other physical properties. 48, and 96 hours a t -25" C. The chief characteristic is the Most studies of the crystallization of rubbers have been made formation of spherulites due to needlelike crystals growing out by observing changes in volume by dilatometer methods (8, 13) from nuclei. These spherulites are considerably flattened since or by x-ray diffraction. Some study has been made of rubber the sample thickness was less than their diameter. All nuclei crystals obtained from solution (11,11). are not formed at once but new ones keep appearing as time The x-ray diffraction method is not very sensitive in detecting progresses. T h e rate of crystallization is thus strongly affected small amounb of crystallinity, chiefly because the crystals by the rate of nucleus formation as well as by the rate of crystal are so imperfect. Some elastomers which show considerable growth. I n Figure 3, where crystallization ha8 proceeded pracvolume change in the dilatometer (9) show just perceptible tically to completion, the spherulites seem to fill all the space x-ray effects (3, 8). Consequently it was considered advisable available, and the sample has the superficial appearance of being to study crystallization by direct microscopic examination. completely crystallized. Figures 4,5, and 6 show the effects of melting and refreezing a Procedure sample of rubber. Figure 4 is a fresh sample frozen 3 days a t The method consisted of freezing thin films of elastomers a t -25" C. Figure 5 is the same field after the sample had been about -25' C. for varying lengths of time and then photographmelted a t room temperature and refrozen for 3 days a t -25 C., ing them with a polarizing microscope. The cold stage consisted and Figure 6 is the same field after subsequent melting essentially of a glass vessel 3.5 inches in diameter by 1.5 inches a t 40" C. and refreezing at -25" for 4 days. I n Figure 5 the deep, with a double evacuated bottom. This vessel contained an antifreeze mixture and a cooling coil consisting of about ten pattern of Figure 4 is repeated. Not only are the same clusters turns of S/le-inch copper tubing. The samples were in the form of reformed, but the detailed structure of individual clusters is thin films laid down on microscope slides, covered with regular reproduced with striking fidelity. Thue, even though the cryscover glasses, and waterproofed by cementing their edges. These tals melt well below room temperature, the motion of the molecwere laid flat on the bottom of the vessel. The cooling coil was supported independently of the microscope to allow rotation of ular segments is so restricted a t room temperature that on rethe stage, vessel, and sample together. The temperature was freezing the same crystal pattern is renewed. Figure 6 shows the regulated by pumping cold acetone through the coil from a comsame pattern on the left part of the field but a different pattern mercial refrigerating unit, so that temperatures down to -20" C. on the right, indicating that some changes have taken place on could be maintained indefinitely. The microscope objective was a Zeiss 40X, 0.75NA, 4.3 mm. heating to 40" C. achromat corrected for water immersion. The antifreeze soluFigures 7 , 8, and 9 show the same effect in a sample of polytion was a mixture of water, 55; ethylene glycol, 30; and alcobutadiene. This is a sample polymerized at 41 'F. and presumed hol, 15. This mixture has a low enough freezing point; it disfrom dilatometer studies to crystallize readily although only solved ice crystals condensed from the atmosphere; it was sufficiently mobile; and its refractive index was not far enough from slight evidence of crystallinity was observed by x-ray diffraction. that of water to affect serious1 the performance of the water [Itis comparable to material studied by Lucas (9) and Beu (S).] immersion objective. One of t i e old type objectives corrected Here also the crystal pattern is reproduced after melting a t for glycerol immersion would be more suitable for this work, 12" C. but not after remelting a t 20" C. The crystal clusters are but such lenses seem to be no longer available. The microscope objective was thermally insulated from the body tube of the much smaller than in rubber and individual spherulites are microscope by means of an extension dollar, about an inch long, difficult t o identify. made of machined plastic. Although, in general, melted and refrozen samples tended to This arrangement permitted observations t o be made over reproduce the same crystal pattern, a number of inconsistent indefinite periods of time. Difficulty was experienced due to results were obtained indicating that the previous history of the
.
413
414
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 43, No. 2
Disenssion These results indicate that direct microscopic examination is a more positive qualitative test for crystallinity in high molecular weight polymers than is x-ray diffraction Materials such as rubber and polychloroprene have long ago bwn shown to be crystalline by x-ray diffraction. However, in most of the newly developed cold rubbers, appreciable amounts of crystallinity have not been detectable by w a y diflraction except in Figlrre 1. 24 Fig,lre 2. 4R Hours Figure s. 96 Hours low temperature polybutadiene (4). X-ray diffraction has not Rubber Frozen at -25" C. (750X) shown crystallinity in polybutadienes polymerized a t higher temperatures nor in copolymers of styrene and butadiene of the . . Glt-S type even when polymerized a t low temperatures. On the other hand, dilatometer studies show first order transitions in t,hese materials, large change8 for low temperature polybutadiene, and small changes for some low temperature copolymers (3). Figure 4 shows a considerable amount of cryutallization in a polybutadiene polymerized a t 41 F. Photographs have been taken of other polymers: regular GI1-S showed no evidence of crystallization even after weeks a t -30" C. T w o other polyFigure 4. Original Figure 5. Refro- Figure 6. Refroniers-ne a polybutadiene polymerized a t 122' F. and the other a Crystallization, zen after Melting zen after Melting 9O:lO butadiene-styrene copolymer polymerized at 41 F.72 hours at at 40" C. a t 20' C. -25' C. showed small amounts of crystallinity. Neither of these Yamplea showcd any evidence of cryst,allinity by x-ray diffraction a t Rubber Refrozen at -25" C. (75OX) temperatures down to -50" C. but did shorn small first order transitions in the dilatometer. Since the microscopic method ia not quantit,ative, it would seem that the dilatometer offers the most promising method of making crystallinity studies of elantomers. It is possible that the effects seen with the niicroscope arc not due to true crystallization. Local orientation and alignment of long chains would produce regions showing birefringence. Ilowever, such high degree of orientation mould have the same practical effect on the physical properties a8 true crystallization. CrystaIlizaGon zen after Melting Z e n after ~ e ~ t i n g Furthermore, such orientation is essentially imperfect crystallizaat 20" C. at 12" C. tion. All high polymers form imperfect crystals at best, and Polybutadiene Refrozen at -25" C. (750X) undoubtedly the inability of x-ray diffraction to detect crystallizasample had a.n effect on its crystallization. Consequently tion in these samples is due to the a sample w m submitted to repeated cycles of melting a t 20" C. higher degree of imperfection in and freezing a t -25" C. Figures 10, 11, 12, and 13 show part these materials than in less imA 0" of this series. These represent the same field in the sample perfect crystals such as natural after the original freezing and after the first, second, and fifth rubher and polyamides. cycle. This is the same sample as Figures 1, 2, and 3, which arc It is quite certain that even the sixth cycle of this series. Only pictures after 24 hours are under the best possible conditione shown a8 they accentuate the differences. There is a considercrystallization of a high polymer B 22 = able change in detail of the pattern on the first cycle but the is not complete. There has been differences tend to disappear with subsequent cycles. Exactly some difference of opinion in the similar results were obtained on samples of cured rubber except literature as to the degree of crysthat the cured samples froze a t a slower rate. Repeated remelting tallization obtainable in rubber. a t a temperature of 60" C. resulted in poor reproduction of However, it is probable that in the patterns. rubber the crystallinity is not c 4.53 Figure 14 is a large spherulite in polychloroprene showing the much greater than 26 to 30% (6, change in appearance on rotating the Nicol prisms. This sample '?, IO). In Figure 3 the spheruwas thick relative to the diameter of the spherulite, so it was lites apparently fill all of the probably not flattened. The general appearance is the same as available s p a c e . H e n c e , t h e that observed with convergent polarized light in a single biaxial s p h e r u l i t e B themselves m u s t D 67.3 crystal and known to petrographers as the "directions image," actually contain more amorphous "conoscopic view," or "interference pattern" (6j. It semis reasonuhle to expect effccts, in radiating E 90' crystnls i n parallel light, similar t o t l i o i c : olxerved with Figure 14. Neoprene (1000 X) ;i single crystal in Figure 10. First Figure 11. First Figure 12. Second Figure 13. Fifth converging light Freezing Refreezing Refreezing Refreezing Single spherulite showing affect of roRefrozen Rubber, 24-Honr Periods a t -25" C. (750X) (4). tating Nicol prisms O
I
February 1951
INDUSTRIAL AND‘ENGINEERING CHEMISTRY
material than crystalline. Although the crystallites are not resolved there can be no doubt from their appearance in polarized light that the spherulites consist of needlelike crystals radiating from a nucleus. Comparison of the sign of the birefringence with t h a t of crystals formed by stretching rubber show that the rubber chains are oriented perpendicularly to the axes of the crystals-that is, the molecular chains are arranged tangentially in the spherulite. The picture of the spherulite therefore is somewhat like a ball of string where there are radial regions of similar orientation which form crystals and regions between these crystals where the chains are not parallel-that is, amorphous regions. It does not appear feasible to make quantitative measurements of birefringence under these circumstances, although Brenschede ( 4 ) has tried i t with some polyamides. The magnitude of the retardation is very low first because of the fact that the cluster is only partly orystalline and,secondly, in these pictures, the clusters are not complete spheres. The photographs of the neoprene cluster show a n interesting effect. If the appearance can be justifiably interpreted as a biavial directions image, some deductions as t o crystal orientation are possible. T h e crystal is known from x-ray diffraction to be orthorhombic. The B axis is the direction of the molecular chains and the greatest refractive index. The sign of the double refraction in the cluster shows this axis to be oriented tangentially. The definite biaxial appearance of the dark bands indicates orientation of some higher order, probably with the B axis constrained to be parallel to a plane. In the spherulite on the left in Figure 14 the vertical dark bar in the zero orientation picture is the trace of thie plane on the sphere. T h e distance between the hyperbolas in Figure 14C is a measure of the optic angle. A comparison of neoprene with rubber shows either that the rubber crystal must be nearly uniaxial ( I I ) , or there is random orientation of the B axis in the spherulites. When the cluster is in or near the 45’ position the two small quadrants would not be very prominent if the cluster were small and the retardation low. This would account for the general appearance of the polybutadiene pictures. Uniaxial spherulites have been shown in polyamides ( 4 ) and in medium molecular weight polyesters in this laboratory. It seems, therefore, that all linear high polymers tend t o crystallize in a similar way-spherulitic clusters of radiating needlelike crystals with the molecular chains in a tangential direction in the spherulite The higher order orientation has so far only been observed in pol ychloroprene.
415
The reproducibility of the crystal pattern on refreezing illustrated in Figures 10 t o 13 is of considerable interest in connection with theories of retractive force in stretched elastomers. I n view of these results the following picture of the mobility of polymer chains is probably justified: In a sample of polymer (at room temperature), which has come
to an essentially steady state, the chains do not move or migrate
to any extent, compared to the range of forces involved in crystaHization since recrystallization brings the same molecular segmenta back into essentially the same crystal configuration. However, when a sample is first formed the molecular segments are not necessarily in their most favored configuration. Thus on the fmt cycles, minor chan es of configuration take place before they settle down to the preferred state; after this, practically no further changes occur on subsequent cycles. This would explain the changes in crystal pattern observed on the first refreezing and such well-known effects as the higher modulus of rubber on its first cycle of elongation. Current theories ( 1 ) as to the mechanism of the retractive force usually postulate considerable freedom of the molecular segments. The observations reported here indicate severe restrictions of motion a t room temperature. These restrictions are reduced at higher temperatures and may not be important above about 60” C. in rubber. It has not been possible by these experimenta to estimate the maximum motion permitted.
Literature Cited Alfrey, T., “Mechanical Behavior of High Polymers,” pp. 236-67, New York, Interscience Publishers, Inc., 1948.
Bekkedahl, N., J . Research Natl. Bur. Standards,43,145 (1949). Beu, K.E.,et al., J. Polymer Sci., 3,465 (1948). Brenschede, Von W., Kolloid Zeit., 114,35 (1949). Bunn, C. W.,“Chemical Crystallography,” p. 77, London, Oxford Press, 1945. Field, J. E., J.Applied Phys., 12,23 (1941). Goppel, J. M., Rubber Chem. and Technol., 21,773 (1948). Hanson, E. E., and Halverson, G., J. Am. C h . Soc., 70, 779 (1 948). Lucas, V. E., et al., IND.ENCI. CREM.,41,1623 (1949). Lukin, B. V., and Kasatochin, V. I., Rubber Chem. and Technol., 21,621 (1948). Smith, W. H., and Saylor, C. P., J . Research Natl. Bur. Standards, 13,453 (1934). Ibid., 21,257(1938). Wood, L. A., and Bekkedahl, N., Ibid., 43,145 (1949). R E C E K VSeptember ~D 28, 1950. This work wae carried out under the sponsorship of the Chemicals and Plastics Section, Research and Development Branch, Offioe of the Quartermaster General.
Influence of Soaps on Gelling Characteristics of Hevea Latices a
D. J. McKeand Dunlbp Rubber Co., Ltd., BZrmZngham, England T h e work was initiated to elucidate further the mechanism of gelling and to study the influence of soaps on the gelling characteristics of natural rubber latices. From a study of the surface and interfacial tensionpH relationships of fatty acid soaps it was presumed that soap can play a major part in gelation. Addition of a synthetic long-chain anionic or nonionic detergent whose surface tension only changes slightly with pH, inhibits gelling with sodium silicofluoride. On the other hand, small amounts of cationic soaps, which by virtue of the positive charge carried destabilize the rubber-serum system, are found to sensitize the latex to gelling. This
is especially marked with latex preserved with sodium pentachlorophenate. The influence of some of these soaps on the mechanical stability of latices has also been investigated. Thus, by treatment of a latex with different types of soap, it is possible to influence the course of gelling.
I
N A paper t o the Institution of the Rubber Industry, Madge
and Pounder ( 2 ) described investigations intb the mechanism of the gelling of latex b y sodium silicofluoride. They suggested that because of its ability t o gel latex a t an alkaline pH, sodium sili-
