GOODYEAR MEDAL ADDRESS Reflections on Rubber Research

---

’Ulr

Medal Award Address

t ons on r Resear ch G. STAFFORD WHITBY University of Akron, Akron, Ohio

~

~~

~~

~

~~~

~

~~

~

~~

0

N LOOKING back over the years since I first became

that progress-progress that has so greatly raised the quality of rubber goods-are, of course, the use of organic accelerators and antioxidants (especially the former), of carbon black, and of cord fabric. As illustrating the relatively primitive state of rubber science a t the time to which I glance back, there comes to my mind the very inadequate, almost worthless, character of the tests conducted by one of the largest rubber manufacturing firms in this country and by the largest rubber manufacturing firm on the continent of Europe when I submitted to them for evaluation batches of experimental samples of raw rubber that I had p r e pared on a plantation. The test stocks consisted of 100 parts of rubber, 100 parts of zinc oxide in one laboratory and 150 parts in the other, and 8 parts of sulfur. The stocks were given one cure only, under fixed conditions of time and temperature, and the only property measured on the cured stocks was the breaking strength! T o what extent the physical testing and evaluation of rubber has been refined and elaborated since can readily be judged by glancing a t the careful survey of the subject that is to be found in the chapter by A. E. Juve in the book, “Synthetic Rubber,” just published under the auspices of the Division of Rubber Chemistry of the AMERICAN CHEMICAL SOCIETY.I n making a favorable comparison between the practice of rubber testing now and its state in 1910, I would guard against any suggestion that perfection has yet been reached or even that it is in sight. I n fact, a further improvement in testing methods that will bring their results into closer parallelism with the performance of rubber goods in service still remains very much of a desideratum. (The showing of GR-S in laboratory tests hardly prepared us for the good service that the rubber has actually given in tires!) I n this connection, an intensified study of the physics of rubber is something to be encouraged.

associated with the rubber industry, the dominant reflection that comes to me is that, with the incoming of synthetic rubber, the subject of rubber has entered the main stream of chemistry in a way and to an extent that did not prevail before, when the raw rubber used by the industry was restricted to the natural product. The change has made rubber a field, for both the laboratory researcher and the technologist, of greater and broader interest than ever before; it has quickened the pace of its scientific and technological march and expanded the scope and novelty of its future possibilities of development. When I say that rubber is now more fully than before in the main stream of chemical advance, I think of the fact that, with its present wide variety of synthetic rubbers and the still wider variety that will undoubtedly be developed in the future, rubber has now become an integral and substantial part ot the chemical manufacturing industry, and still more I think of the fact that the fundamental scientific issues that manufactured rubbers have raised are problems of polymers-their formation, structure, and properties-and that in consequence rubber as a field of scientific research is an essential part of the field of polymer science. Research on rubber now makes important contributions to this rapidly growing branch of science and in turn itself benefits b y the accumulation of knowledge in that branch. Polymer chemistry as a field of intensive study is still in its early days; it will undoubtedly in the years immediately ahead attain to much new insight into macromolecules and many new techniques for their controlled preparation and their study-to all of which future rubber research may be expected to contribute and from which reap benefits. I n thus speaking of the interesting and challenging future that rubber offers to the researcher and the technologist, it is far from me to reflect unfavorably on the extent of the advance registered by rubber science and technology in the period before the incoming of synthetic rubber caused the changes I have mentioned. I have only to look back to 1910-the year of my first connection with the industry-to realize the substantial progress that was in fact achieved during that period. The outstanding landmarks of

RUBBER AND PLASTICS

One effect of the advent of synthetic rubber has been to bring the rubber industry into close relation to the plastics industry.

806

The

Goodyear Medal was first awarded i n 1939 on

the centennial of Charles Goodyear’s discovery of vulcanization, and is a n annual award by the ACS Division of Rubber Chemistry. G. Stafford Whitby, the 1954 recipient, delivered the address published here a t the 126th Meeting of the AMERICANCHEMICAL SOCIETY in New York City.

The ACS monograph Dr. Whitby holds i n his hand is just a sample of the writing he has been engaged in during the last 45 years. These writings have been published in journals all over the world, and Dr. Whitby himself, English by birth, worked in the East Indies, Canada, and England before finally moving to the University of Akron in 1942 as professor a f rubber chemistry and director of rubber research. Dr. Whitby has worked actively in almost every phase of rubber science and technology and has been most active in the government synthetic rubber program. He is now retired, holding the title of professor emeritus of rubber chemistry, and *is a consultant to the University of Akron’s department of rubber research.

Strictly speaking, the manufacture of rubber goods has always been a plastics industry, since it normally involves passage of the material through a plastic condition (in which i t is shaped by molding, extrusion, etc.) into a thermoset (vulcanized) condition. Rubber manufacting has not usually been classed as a branch of the plastics industry because of ( a ) the historical circumstance that the industry was already well established before the modern plastics industry, as conventionally regarded, came into existence, ( b ) the fact that rubber products, unlike the early synthetic plastics, are (excepting hard rubber) elastic, not rigid, and (c) the fact that no chemical manufacturing operation need be applied to natural rubber in order to render i t plastic, whereas most of the raw materials of the conventional plastics industry are synthetic products and whereas cellulose needs to be converted to one of its chemical derivatives to make it plastic. The fact that in rubber manufacturing synthetic rubbers are now used to an important extent is not the only factor tending ,to bring the rubber industry and the plastics industry closer together. There is also the fact that many of the products of the plastics industry are now not hard and rigid but soft, flexible, and even somewhat extensible, as witness polyethylene and plasticized polyvinyl chloride. Further, some of the synthetic rubbers--e.g., nitrile rubber of high nitrile content, G R S and Butyl of high Mooney viscosity-require the incorporation of rather large proportions of plasticizers for the display of satisfactory rubberlike properties. And again, there is a growing tendency to develop materials in which on the one hand the properties of rubber are modified by blending with i t synthetic resins, which have a stiffening action, and on the other hand the properties of synthetic resins and plastics are modified by the incorporation of synthetic rubbers, which confer a n element of flexibility that improves impact strength. I n these and other developments, we see it becoming more and more difficult to draw any clearly defined line between rubber and plastics. As further witnessing to the close association, if not the merging, of the fields of rubber and of plastics, it was the rubber industry that developed (on the basis of GR-S techniques) and now manufactures the so-called high styrene resins, used in large tonnage

in shoe soles and elsewhere. Also, the copolymer used in the very successful latex paints has a ratio of butadiene to styrene such that it is difficult tg say offhand whether i t is a rubber or a plastic. I n contemplating, then, the present state and future prospects of rubber research, my dominant feeling is one of satisfaction that rubber, thanks especially to its association with polymer science, has entered the main march of chemical progress and that i t offers now more than ever before intellectual adventure to the researcher and variety and scope of challenge to the technologist. I do not, however, propose to devote the rest of this lecture to elaborating and exemplifying this theme. I propose, rather, to use the rest of my time in offering reflections on a small number of specific aspects of rubber research. These aspects are selected primarily because of my personal interest in them, but some will serve incidentally to illustrate the theme with which I started. BIOGENESIS O F RUBBER

Saturally occurring compounds the molecules of which are composed of or comprise isoprenoid units are so numerous and include so many substances of physiological and economic importance-such as terpenes, rubber, rosin acids, carotenoids, vitamin A, phytol, vitamin K, steroids, including vitamins D and E, the sex hormones, and cortisone-that the question of the manner in which they arise in animals and plants naturally excites great curiosity. A variety of schemes according to which isoprenoid compounds might arise in nature have been put forward from time to time. Some of them are no more than speculative exercises on paper. A scheme suggested by Collie (49) takes the branched chain five-carbon sugar apiose as the starting material. About the only positive fact worthy of mention in connection with such a scheme is that apiose occurs, as a glucoside, in celery and that the sesquiterpene hydrocarbon selinene occura in oil of celery. A scheme elaborated by Hall (23) based on saccharic acid and metasaccharonic acid (derived from carbohydrates) as the starting materials is also completely speculative.

807

INDUSTRIAL AND ENGINEERING CHEMISTRY

808

Other schemes are mostly speculative in essence but have a alight element of experimental support in that it has been demonstrated by experiments, in the chemical laboratory but not in the plant, that isoprenoids can actually be prepared from one of the intermediates figuring in the scheme. Favorsky and LBbBd6va (19)take as a starting point the branched chain sixcarbon amino acid, leucine, derived from protein. One of the intermediates in their scheme, [‘koprene alcohol” [( CH3)2C(OH)CH:CHz], was shown experimentally by them to form some isoprene and CIOterpenes when treated with cold 20% sulfuric acid. Fischer (90)suggested p,fi-dimethylacrolein [&methylcrotonaldehyde, (CHs)gC:CHCHO] as the precursor of isoprenoids on the strength of his finding that it could be condensed in the laboratory t o form a series of unsaturated isoprenoid aldehydes ranging from dihydrocitral to higher members having a relationship t o the carotenoids. An early speculativ? scheme is that of Aschan ( I ) , who suggested that isoprene arises in plants by the aldol condensation of products of carbohydrate breakdown such as acetone, dihydroxyacetone, and pyruvic acid, followed by reduction to glycols and elimination of water. Thus, for example: (CB3)zCO

+ CHICHO

(CHa)zC( OH)CH&HO + (CHa)gC(OH)CHzCHzOH + CHz :C( CHa)CH:CHz

Hesse (26)found in the latex of a species of Caalotropis (a genus of milkweeds) pyroterebic acid [( CHa)&:CHCH&OOH] in the form of esters with triterpene alcohols. He suggested that the precursor of this and various isoprenoid-type structures might be “acetonepyruvic acid,” [( CHI)&( OH)CH2COCOOH\, a substance that he was able to obtain (as its lactone) in the laboratory by the condensation of acetone and pyruvic acid. Johnson ( 8 7 ) has offered interesting comments on some of the foregoing schemes, especially on Hesse’s suggestions. Work of Bonner. Only recently haa there become available from actual experiments on rubber-producing plants, direct evidence relative to the course of the biosynthesis of rubber. I n noteworthy studies on the formation of rubber in guayule (Parthenium argentaturn), Bonner (6) has shown that when seedling plants or isolated stem segments of guayule are grown in solution cultures the addition of acetate, acetone, or p-methylcrotonic acid causes an increase in rubber formation. Further, experiments in which guayule plants were grown in quartz sand t o which nutrient solution containing isotopically labeled acetate was supplied indicated that the increased formation of rubber for which the acetate was responsible was (to judge by the specific radioactivity of the rubber) entirely derived from the acetate. From the indicated conclusion that acetate can serve as the sole source of carbon for rubber synthesis and the observation that the branched fivecarbon compound, &methylcrotonic acid, will support rubber formation, Bonner proceeded to in vitro experiments with enzymes on the question of the nature of the steps from acetate to 6-methylcrotonic acid ( 7 ) . (That there is in fact a route in the plant from acetate to P-methylcrotonate is shown by the fact that, when fed C14 labeled acetate, guayule synthesized some C14 labeled 8-methylcrotonate.) As a result of these experiments, i t was concluded that acetate is first activated by conversion into acetyl-coenzyme A and that subsequent steps are as shown in the following scheme. The reactions depend on the presence, not only of the necessary plant enzymes (which apparently are not yet individually identified), but also of coenzyme A, coenzyme I (diphosphopyridine nucleotide) and ATP (adenosine triphosphate). I n these experiments, steps 1 to 4 were brought about by an enzyme system prepared from flax seedlings.

+

CHsCOOH coA $ CH3COcoA (Acetyl-coA) 2CHaCOcoA e CH3COCH2COcoA4(Acetoacetyl-coil) coA

+

(1)

(2)

+ +

+

Vol. 47, No. 4

+

CHaCOCHzCOcoA H2O CHaCOCHzCOOH COA CII~COCOA CHaCOCHzCOOH CHzCOOH ,,,&>C< (P-Hydroxy-8-methylglutaryl-coA) HO CHzCOcoA CHs>hacrylateproduced 41 times its weight of additional popcorn polymer under similar conditions. 3. Popcorn polyisoprene (discussed later) in 100 times its weight of styrene a t 55" C. grew t o 73.2 times its original weight, and its limit of growth had not been reached, since material from the first generation growth would, in a second generation, produce a further quantity of popcorn polymer 2.7 times its own weight (Figure 4).

The extent to which the growth of popcorn polymer will proceed is determined broadly by the extent of unsaturation in the seed polymer. So that, for example, the extent to which popcorn derived from styrene containing butadiene will grow in styrene depends essentially on the proportion of butadiene present. I n the growth process new polymer chains, derived from the substrate monomer, attach themselves to the unsaturated centers in the seed, and lead to a highly cross-linked product in which vast new chains are grafted onto the insoluble nucleus or seed first introduced. It is unlikely that this grafting process is a mere branching reaction involving chain transfer by the abstraction of hydrogen atoms from the seed, because popcorn growth has been found to occur with many chemically W e r e n t types of seed--e.g., not only with seeds in which the unsaturated elements are butadiene units, b u t also with seeds consisting of crosa-

814

INDUSTRIAL AND ENGINEERING CHEMISTRY

linked alkyd resins in which the unsaturated elements are units derived from maleic anhydride, itaconic acid, and aconitic acid. It also appears that the growth does not depend on the catalytic action of peroxides on the seed, because the phenomenon takes place as well in nitrogen as in air and because it is observable with seeds differing widely in chemical character. POPCORN POLYMERIZATION OF DIENES

The phenomena of popcorn polymerization have been found t o display themselves in the bulk polymerization under suitable conditions of a wide variety of monomers. Popcorn polymerization can be obtained from single monomers, without the inclusion of a second cross-linking monomer, when the single monomer is a conjugated diene. Such monomers (as we know but too well from experience with GR-S production) are self-cross-linking agents, capable of cross linking their own primary, linear polymer chains. It has long been known that when dimethylbutadiene is allowed to undergo spontaneous polymerization the product is insoluble “cauliflower” polymer-popcorn polymer as we should now call it. And in Germany during World War I advantage was taken of the ability of this popcorn polymer t o grow by adding some of it, as “seed,” to the monomer (dimethylbutadiene) from which methyl rubber H was made. It has also been known that under some conditions chloroprene gives rise to a peculiar type of polymer that we may now call popcorn polymer. Carothers (18)called it w-polymer, perhaps because it seemed to be “the last word” among the many different types of polymer he had encountered in his pioneering work! It is interesting to recall Carothers’ remarks about it: The conditions favoring its formation are not very clearly understood, since it occasionally appears under the most diverse conditions. It, seems certain, however, that its formation is autocatalytic. When a speck of this polymer appears in a sample of chloroprene during the early stages of its polymerization, the granular growth continues t o spread through the whole sample. Because of its cell-like structure it occupies more volume than the same amount of p-polymer (rubberlike), and if the growth begins t o spread laterally through a sample it may burst the walls of a heavy Pyrex container.

I have found that a simple procedure for obtaining popcorn polymer from chloroprene is to allow a quantity of the monomer containing a polymerization inhibitor-e.g., plenty of hydroquinone-to stand in a corked flask. After a few weeks it will be observed that specks of polymer have formed on the cork and thereafter rapidly grow downward in the vapor space, consuming the monomer and forming a suspended mass of popcorn polymer capable of being grown in suitable monomer substrates-e.g., in styrene. (In passing, it may be mentioned that, curiously enough, a similar vapor-phase formation of popcorn polymer from methyl vinyl ketone has been encountered. This suggests that the ketone is a self-cross-linking agent, although as normally formulated it is merely a vinyl compound as far as polymerizing ability is concerned.) Butadiene too readily undergoes the popcorn type of polymerization. Starting with some experiments made in 1947, I have found that when butadiene (without diluent or modifier) is placed in sealed tubes and allowed t o stand a t room temperature it invariably polymerizes practically exclusively in the popcorn manner. After an interval of 31/$ months or longer, specks of insoluble polymer are seen t o have separated. (The liquid a t this time is still quite limpid.) Thereafter popcorn polymerization proceeds fairly rapidly, after a total of 7 t o 12 months the whole had become typical popcorn polymer, insoluble, nonelastic, friable, and expanded (Figure 5). Isoprene seems to show much less tendency than either butadiene or dimethylbutadiene to form popcorn polymer. This is perhaps a reflection of a smaller tendency toward cross linking. I n most of the batches of isoprene that we have held in sealed

Vol. 47, No. 4

vessels a t room temperature, no popcorn has appeared; the monophase character has been retained, and the isoprene has simply become more and more viscous as in normal bulk polymerization, It seems certain, however, that under some conditions (not yet clear) isoprene allowed to polymerize spontaneously a t room temperature undergoes the popcorn type of reaction. The late H. Hibbert left in his laboratory a sealed bottle labeled, “Isoprene rubber from spontaneous polymerization (15-year period) of liquid isoprene.” The contents were typical popcorn polymer (Figure 6) and showed high capacity for growth in a variety of vinyl monomers. Further, a sample of petroleumderived isoprene stored in our laboratory showed the separation of polymer specks after 5’/2 years, when the liquid phase was already quite viscous. The sample has now been kept for 10 years, but, unexpectedly, the popcorn specks have failed to show much growth, although the liquid phase has in the meantime become so viscous that it is no longer pourable. At this point a note may be injected regarding the quality of the elastic nonpopcorn polyisoprene formed by the spontaneous polymerization of isoprene. After 41/2 years the last-mentioned sample of isoprene, which was a t that time free from popcorn polymer, contained 7.26% of precipitable polymer, soluble in benzene. It also contained 4.8% of oily dimer. The polymer was subjected t o vulcanization tests in a gum stock, with the following results. The data for each cure are the average of tensile tests on four test pieces. Teat stock: polymer, 100; stearic acid, 3; zinc oxide, 5 ; sulfur, 3; mereaptbenzothiazole, 1.5 Min. a t Tensile Strength, Elongation, 2 9 2 O F. Lb./Sq. Inoh % 15 30

215

200

255 235

The material is, clearly, not a t all like natural rubber and in fact seems inferior in gum properties t o emulsion polyisoprene. SCOPE O F POPCORN POLYMERIZATION

Aside from the special case, mentioned above, of methyl vinyl ketone, the formation of popcorn polymer from vinyl monomers such as styrene requires the presence in them of a cross-linking (bifunctional) monomer, in a proportion depending on the efficiency of the latter as a cross-linking agent. Some cross-linking agents we have found to approach divinylbenzene in their efficacy in producing popForn polymer are as follows: Allyl methacrylate 2-Chloroallyl acrylate 2-Chloroallyl methacrylate Di-2-chloroallyl maleate From the study of these cross-linking agents in popcorn polymerization, it was found t h a t they are capable of replacing, albeit a t higher concentrations, the divinylbenzene used in the production, by emulsion polymerization of easy-processing GR-S (GR-S 1008). The following are vinyl monomers that, in the presence of small proportions of cross-linking comonomers, have given popcorn polymer in our experiments: Styrene, chloroptyrene, a-methylstyrene Methyl methacrylate, methyl acrylate (Figure 5), butyl acrvlate Methacrylonitrile 2- and 4-Vinylpyridines, 2-vinj.l-5-methylpyridine A variety of monomer in addition to styrene have been found capable of serving as substrates for the growth of popcorn polymer. It is not necessary that there shall be any chemical similarity, aside from ability t o polymerize, between the substrate monomer and the monomer from which the popcorn seed was derived (68). And, further, a succession of different monomers can be grafted onto the same seed, provided of course that a t the conclusion of each step sufficient unsaturation is left in the seed for more growth. For example, we have grown popcorn

April 1955

INDUSTRIAL AND ENGINEERING CHEMISTRY

polyisoprene in methyl methacrylate and then grown the resulting polymer in styrene. GRAFT POLYMERIZATION

Graft polymers are formed by a two-step polymerization process. I n the second or grafting step, polymer branches are chemically attached to or grafted onto the primary polymer chains derived from the first step. The process, because of its twostage character and because the branches or grafts may be derived from a monomer different from that forming the primary chains, is distinct froin the branching that (it has been supposed but not proved) may occur during the one-step polymerization of a diene. Several techniques for the production of graft polymers have been described lately. One line of procedure is t o treat the primary or stock polymer in such a way as to generate on it active points where it can readily give rise t o free radicals, which serve to initiate polymerization when, in the second step, the stock polymer is introduced to a monomer. New polymer chains derived from the second monomer grow onto the primary chains a t the active points. Melville (68)has described such a technique. He partially brominates polystyrene, then dissolves it in styrene monomer, and irradiates the solution with ultraviolet light of such a wave length that bromine atoms are removed. In this way a proportion of the monomer polymerizes onto the original polystyrene, thus forming grafted branches. A more practicable and interesting technique has been described by Mark (34). The chain elements of polystyrene have a distinct resemblance to cumene and like the latter comprise active tertiary hydrogen, which, when the material is treated with oxygen, is capable of giving rise t o hydroperoxide groups, thus:

H

0

H

0

815

ing. Experiments (in conjunction with H. L. Stephens) were accordingly carried out in which various vinyl monomers were caused to polymerize in the presence of previously prepared, gelfree diene polymers and copolymers (designated as stock polymers), the latter being in the form of latex. The results indicate that under such conditions new polymer chains derived from the vinyl monomer do in fact become attached to the preexisting polymer, the result being a graft polymer. The influence on the properties of the stock polymer produced by grafted branches depends, as might be expected, on the nature of the vinyl monomer. If the latter is such that its homopolymers are hard and resinous, the grafts stiffen the stock polymer and may in fact produce a degree of reinforcement comparable with that produced by loading the stock polymer with carbon black. If the vinyl monomer is such-e.g., ethyl acrylate-that its homopolymers are soft and elastic, the grafts have little stiffening action. Polymerization techniques used in this work were mostly as follows: 1. Stock polymer was prepared from butadiene or from a mixture of butadiene and styrene in the following recipe: (16 hours a t 5O C.)

Parts Monomers Water Potassium laurate Potassium chloride tert-Dodecyl mercaptan (Sulfole) Cumene hydroperoxide Triethylenetetramine

100

180 5

0.8 0.25

0.08 0.10

There was then added as a shortstop 0.2 part sodium dimethyldithiocarbamate; unreacted butadiene was vented, and unreacted styrene was removed by stripping. (The dithiocarbamate shortstop or the thiuram disulfide, into which it presumably becomes converted, did not prevent the subsequent graft polymerization. ) 2. A vinyl monomer and additional quantities of peroxamine reagents were now introduced to the stripped latex, and the second (graft) polymerization step was conducted, the recipe being as follows: (4 hours a t 5O C . )

.,

By oxidizing polystyrene t o a small degree, in order t o introduce some hydroperoxide groups, and then adding a monomer together with ferrous ions (adapted to decompose the hydroperoxy groups and thus generate free radicals on the polystyrene), the monomer polymerizes onto the polystyrene. If the monomer is vinyl acetate and the acetate groups are then hydrolyzed, a very interesting product is obtained-a product in which the stock (polystyrene) is oil- “soluble” and the grafts (polyvinyl alcohol) are water-“soluble.” Elastomeric graft polymers having an “amphoteric” character such as that possessed by the graft polymer just cited might well be of interest. Mark (34) has indicated a number of other devices by which graft chains can be attached t o a stock polymer, but none falls in the field of elastomers. And, as is shown in the following section, grafts can be made onto diene polymers and other unsaturated polymers by proceeding in a much simpler and more direct way than the techniques mentioned in this section. GRAFT ELASTOMERS

Popcorn polymerization is a special case of graft polymerization, and the first case of it t o be studied. I t s study clearly indicated that, when a vinyl monomer is caused to polymerize in the presence of a pre-existing polymer containing double bonds, polymer chains attach themselves to the double bonds-that is, chains of vinyl polymer become grafted onto the already formed polymer. I n popcorn polymerization, the pre-existing polymer is cross linked and insoluble and so is the final product. It was of interest t o ascertain whether grafting would occur when the pre-existing polymer was a soluble material, free from cross link-

Stock polymer (as latex) Graft monomer tert-Dodeoyl mercaptan Cumene hydroperoxide Triethylenetetramine

.

Parts 100 Various ( mostly 2 5 ) Various (mostly 0 . 0 ) 0.2

- 0.3 0.3

At the end of the polymerization period the antioxidant phenyl-@naphthylamine was added, the latex was coagulated by methanol containing sulfuric acid, and the graft polymer was isolated. The content of grafted vinyl polymer in the product was calculated from the weight of the latter. The vinyl monomer most generally used in such experimenta was methyl methacrylate, as a given weight of this monomer, when grafted onto a diene stock polymer or copolymer, produced a greater degree of reinforcement than did the same weight of styrene, which in turn produced more reinforcement than either mono- or dichlorostyrene. That the result of the second polymerization was not merely a mechanical mixture of the stock polymer with homopolymer of the vinyl monomer, but that polymethyl methacrylate actually became chemically united with the stock polymer (grafted onto it) was shown in several ways, especially by extraction and fractional precipitation. The behavior was compared of ( a ) stock polymer, ( b ) graft polymer, (c) polymethyl methacrylate, and ( d ) mixtures of stock polymer and polymethyl methacrylate. The polymethyl methacrylate used in these experiments was prepared in a peroxamine recipe similar to that used for the preparation of the stock polymer. The mixture was made by blending latexes of the two polymers in question. When placed in acetone a t room temperature, the polymethyl methacrylate sample went into solution within 1 hour. Samples of graft polymer and of the mixture ( d ) , when exhaustively ex-

INDUSTRIAL A N D ENGINEERING CHEMISTRY

816

tracted with acetone for 96 hours a t room temperature, gave the following results in triplicate runs: Cold Acetone Extraotion of Graft Polymer and of Polymer Mixture Soluble in Acetone (6) (b) Graft polymer (methyl methacrylate grafted onto polybutadiene; 15.2% 1.64 1.20 1.76 polymethyl methacrylate content) (a) Mixture (polybutadiepe,andpolymethyl methacrylate containing 26.6% polymethylmethacrylate) 24.3 24.3 24.2

It is clear that, whereas most of the polymethyl methacrylate ia readily extractable from a mechanical mixture with the stock polymer, little is extractable from the graft polymer.

Vol. 47, Na. 4

onto 75/25 butadiene-styrene stock polymer and calculated to contain 18.55% polystyrene gave 15.33 and 15.25% (duplicate experiments) of polystyrene when i t was oxidized b y tert-butyl hydroperoxide and osmium tetraoxide on the lines of a procedure described by Kolthoff (SO). The stock polymer when similarly oxidized went entirely into solution. A sample of polystyrene (prepared in the peroxamine recipe already given) by this wet oxidation method gave recoveries of 85.43 and 84.19% (duplicate experiments). These results indicate that polystyrene side chains in graft diene polymers survive the oxidation of the unsaturated stock polymer to which they are attached. The grafting of polyvinyl side chains onto diene polymers has, somewhat unexpectedly, little tendency to cross link the primary chains; it is not difficult to conduct the process in such a way that the graft polymeric product is free from gel. By way of contrast, in the preparation of the so-called polyesters, by the polymerization of styrene in the presence of an unsaturated alkyd, cross linking occum t o such an extent t h a t the final product is not only insoluble b u t also shows almost no swelling liquid aromatic hydrocarbons. Presumably this is because the unsaturated centers in diene polymers are much less reactive than those (typically maleic units) in the “polyesters.” The reinforcing action of polymethyl methacrylate graft chains attached to a stock of polybutadiene or poly(butadiene-styrene) is such that vulcanizates of the graft polymers in gum stocks may show figures for modulus and tensile strength approaching those shown by the stock polymers in black compounds. To give some idea of the magnitude of the reinforcement produced by grafting, the following may be mentioned as typical results in gum stocks. Graft polymers in which the stock polymer was 90/10 or 75/25 butadiene-styrene and in which the content of graft polymethyl methacrylate was 20% have shown figures of the following order a t optimum cure: Modulus at 300%, lb./sq. inch Tensile stren th, Ib./sq. inch Elongation, $b

900-1300 1500-2000 400

For comparison, the stock polymers themselves give in gum stocks figures of the following order only: 40 60 METHANOL I%) IN LlOUlD

20

00

Figure 7. Fractional precipitation of (a) stock polymer (polybutadiene); ( b ) graft polymer containing 19.6% grafted polymethyl methacrylate; ( c ) polymethyl methacrylate; ( d ) mixture of stock polymer and polymethyl methacrylate containing 20.270 of latter

Figures 7 and 8 show the precipitation curves for samples of the types ( a )( d ) . Graft polymers, when subjected t o fractional precipitation (by adding methanol to benzene solutions, followed b y sedimenting and centrifugating) behave quite differently from the corresponding stock polymers and from mechanical mixtures of stock polymer and polymethyl methacrylate. When the stock polymer consisted of polybutadiene (Figure 7), the graft polymer precipitated at the same methanol concentration as polymethyl methacrylate; when the stock polymer contained some styrene (Figure 8) the precipitation point of the graft polymer came a t a somewhat lower concentration of methanol. I n experiments on a graft polymer prepared by grafting polystyrene onto a 75/25 butadiene-styrene stock polymer, the precipitation characteristics of the stock and graft polymers were not sufficiently different to be of evidential value when the precipitant was methanol. However, by using acetone as the precipitant, the ciirves showed conclusively that polystyrene was actually grafted onto the stock polymer, although the precipitation curves for stock and graft polymer were not as widely separatebtas corresponding curves in Figures 7 and 8. A polymer (100% soluble) consisting of polystyrene grafted

Modulus at 300%, lb./sq. inch Tensile strength, lb./sq. inch Elongation, %

200-250 250-350 350-400

Vulcanizates of graft polymers have not yet been fully evaluated in all respects, but limited tests indicate that their gum stocks compare favorably with GR-S black stocks (tread type) in resistance to abrasion and to flex cut growth. It may well prove t h a t in the future development of synthetic rubbers, less reliance will be placed on the incorporation of reinforcing powders, such as carbon black, t o secure the desired levels of modulus. It would seem logical t o obtain the desired modulus characteristics by building a suitable structure into the molecules of the polymer itself, rather than by particulate reinforcement. That it is possible to have a high modulus in gum stocks is shown by examples of graft elastomers and Vulcollane-type elastomers. Worthy of mention is the fact that gum stocks of polymers reinforced by grafting show the so-called Mullins Effect (softening on prestretching) to a t least as great a degree as polymers reinforced by black. This finding provides food for thought. In black stocks, the softening effect of prestretching may be thought of as due to detachment of the rubber matrix from the surface of the reinforcing particles of pigment. If this picture is correct, it might perhaps have been expected that graft polymers would not show the effect, since their gum stocks contain no particulate reinforcement but only the reinforcement due to the grafted polyvinyl side chains, which are chemically attached to the elastomer chains and form an integral part of the rubber phase. The following table, referring t o optimum cures of 90/10 butadiene-styrene onto which polymethyl methacrylate has been grafted, and 90/10 butadiene-styrene reinforced by carbon black (40 parts H A F black per 100 parts rubber) shows the softening

April 1955

INDUSTRIAL AND ENGINEERING CHEMISTRY

effect of prestretching. The samples were stretched four times in succession (with an interval of 30 seconds between stretchmgs) to a predetermined length short of rupture and were then subjected t o a fifth stretch t o the breaking point. Cures: 90 minutes a t 292O F. with zinc oxide, 5, sulfur, 1.5, and benzothiazolyl disulfide, 1.75 (Prestretch.. 30070) Breaking Breaking No. of Modulus at Strength Elon ation, Stretch 100% 200% 300% Lb./Sq. d c h

k

Graft polymer containing 26.7% grafted polymethyl methacrylate-gum stock 1 645 955 1400 2 40 175 880 3 40 155 880 4 40 155 780 5 40 155 780 1890 375

817

Mark ( 3 4 ) has indicated methods of a more practical character for the production of block polymers, involving the preparation separately of blocks of each of the units, the polymerizations being carried out in such a way that the blocks terminate in reactive groups. By then treating a mixture of different blocks with a suitable bifunctional reagent, the blocks can be joined together. Thus, polystyrene prepared by an initiating system adapted t o leave terminal hydroxyl groups on the polymer molecules can be united with polymethyl methacrylate similarly prepared, by treatment with a diisocyanate, to yield a block polymer.

Stock polymer-black stock

The graft polymer gum stock suffers softening in the early part of the stress-strain curve more severely than the stock polymerblack stock does. The softening of rubber vulcanizates by prestretching has been mentioned in the rubber literature before recent papers by Mullins appeared. Holt (26) not only reports the phenomenon but, in the following words, underscores its practical bearings. The data given show the elusive character of the stress-strain curve of rubber. The initial-stretch curve which is ordinarily used in evaluating rubber compounds is possibly the most defibut it is the curve least definite in character. If the nite stress-strain curve is to be used in designing rubber compounds for specific purposes, it is obvious that the conventional curve may not give the proper stress-strain relations. A study of the phenomena encountered in repeated stressing . . should throw light on the structure of rubber compounds and on the behavior of different compounding ingredients. . . The lower part of the stress-strain curve, which is seldom accurately determined, may have an important bearing on the real properties of a compound.

...

.

.

Bloomfield and co-workers (6) have described graft elastomers prepared from natural rubber by the polymerization of vinyl monomers in Hevea latex. Here the stock polymer, unlike the synthetic stock polymers with which we have worked, have, of course, good strength in gum stocks. BLOCK POLYMERIZATION

r

What have been called block polymers are composed of two or more different monomer units that, instead of being intermingled as in copolymers ( . . .ABABABAB. . . ), occur in alternating blocks, each block being composed of a considerable number of units of the same kind [ . . AA(A),ABB(B),B . . . ] . The properties of block polymers will, clearly, be different from those of copolymers, especially if the units involved are markedly different in chemical character. Melville (18) has described two procedures by which block polymers may be prepared, of which the second may be outlined. The method depends on rapidly generating free radicals in high concentration and thus initiating polymerization in a monomer, A , and then, during the short period that the propagation reaction occupies and in which free radicals persist, quickly mixing the monomer with a second suitable monomer, B, the polymerization of which becomes initiated by the still-active polymer radicals of A , and thus leads to the growth of chains of B units onto the chains of A units. This procedure has been successfully applied t o the preparation of block polymers from butyl acrylate and styrene and from acrylonitrile and styrene. Polybutyl acrylate free radicals have a relatively long life (a few seconds). By running butyl acrylate containing a photosensitizer a t a rapid rate through a capillary irradiated b y ultraviolet light and then bringing the liquid into a vessel containing styrene, block polymers were secured.

Figure 8. Fractional precipitation of (a) stock polymer (90/10 butadiene-styrene); ( b ) graft polymer containing 18.4% grafted polymethyl methacrylate; ( c ) polymethyl methacrylate; ( d ) mixture of stock polymer and polymethyl methacrylate containing 19.75% of latter

The application of such a device to the preparation of elastomers of the diene type has not hitherto been reported. The device is readily applicable to the Vulcollane type of condensation polymers. The preparation of polymers of this type depends essentially on the union, by means of reaction with diisocyanates, of polyesters having terminal hydroxyl groups. Coffey and Meyrick (IS) have recently described the preparation of urethane-linked polyesters in which polyester blocks of two different kinds are united alternately. They indicate that by properly choosing the chemical composition of the two blocks, special effects are obtainable-e.g., outstanding impact strength, internal plasticization, and lowered melting point. The effect on the melting point of the block arrangement in polyesters may be of practical significance in the field of urethane-polyester elastomers. One of the weaknesses of the original Vulcollane (Vulcollane N ) is a tendency t o crystallize and harden on storage. NONSULFUR VULCANIZATION

Before the advent of synthetic rubbers, when we thought of vulcanization we thought inevitably of heating rubber with sulfur. And we still rely mainly on sulfur as the best and most

INDUSTRIAL AND ENGINEERING CHEMISTRY

818

practical vulcanizing agent for natural rubber and the diene hydrocarbon elastomers-GR-S, nitrile rubber, and Butyl rubber. For nonhydrocarbon synthetic rubbers, however, other types of vulcanizing agents are used. Agents other than sulfur able to vulcanize diene polymers are continually coming to light in laboratory studies. A good many are substances capable of giving rise t o free radicals, and their mode of action presumably depends on attack (abstraction of hydrogen) by such radicals on the polymer chain a t the methylene group alpha t o the double bond. Benzoyl peroxide and diazoaminobenzene .are representative of such agents that have long been known. Recently the use of organic peroxides other than benzoyl peroxide has been described (IO), and I have found diazoethers to have the ability t o vulcanize natural rubber and GR-S. These compounds, RN:NSR' (where R and R ' are aryl groups), decompose on heating to generate the free radicals Re and .SR. From a practical viewpoint, they, like diazoaminobenzene, have the disadvantage of causing microporosity owing t o the liberation of nitrogen that accompanies their decomposition. An interesting class of vulcanizing agents reported by Sturgis, Baum, and Trepagnier (60) are certain polyhalogen compounds, which apparently function by generating free radicals. Examples of such agents, which act best in the presence of metallic oxides euch as litharge, are benzal chloride, benzotrichloride, and 1,1,1,3-tetrachloropropane. Such agents may be as effective as conventional vulcanization by sulfur and accelerators in regard to the quality of the vulcanizates produced. Thus, for example, o-chlorobenzotrichloride (2.5 parts) along with litharge (10 parts) produced from GR-S loaded with EPC black (50 parts) a vulcanizate with a 3ooyO modulus of 1710 pounds per square Inch, tensile strength of 3140 pounds per square inch, and elongation of 43oa/,. Such a vulcanizate ages much better than GR-S vulcanized by sulfur; it does not show on aging the increase in modulus and drop in elongation characteristic of sulfur-cured GR-S . VULCANIZATION BY BISTHIOL ACIDS

Of some interest is a new type of vulcanizing agents, the bisthiol acids. Whereas yulcanization by sulfur depends essentially, according to current views, on attack a t the a-methylene group of the polymer, vulcanization by the bisthiol acids depends on direct addition of the reagent a t the double bond. Such addition produces cross linking, thanks t o the bifunctionality of the reagent. Two diene units in different polymer chains become united, thus:

. CH2

/

\

'CH &H

+

HS.CO(CHz),COSH

+

CH,.

..

CH

I

AH \

/

Vol. 47, No. 4

that it produces cross linking and causes the rubber to crumble while it is being incorporated in the rubber on a mill. However, by introducing the reagent in the form of a solution in benzene, which strips of rubber contained in evacuated sealed tubes were allowed to imbibe, and by then applying heat, vulcanized test pieces were obtained. They show tensile strengths up to about 2500 pounds per square inch. Marvel (35) has made use of the ability of thiol acids to add to unsaturated hydrocarbons by preparing high polymers by reaction between the bifunctional reagents, a bisthiol acid, and a diolefin, such as diallyl. In work with F. S. Shannon, I have found bisthioladipic acid to be an active cross-linking agent in the synthetic rubbers, GR-S, and nitrile rubber. It rapidly vulcanizes these rubbers in the course of incorporating it in them on a mill, as is shown by swelling measurements on the products. When 1% of bisthioladipic acid was milled with GR-S, the product showed a swelling capacity of 4.83 grams of benzene per gram of rubber. When the material was then heated for 1 hour a t 292" F., the swelling capacity fell only to 4.58, thus indicating that the cross-linking reaction had almost completed itself during the milling. Swelling figures of this order indicate that 1% of the reagent (which, if fully utilized in making cross links, corresponds to only 0.35% of sulfur in the form of disulfide links) is sufficient to vulcanize GR-S fully. The degree of swelling is of the same order as that shown by sulfur vulcanizates of GR-S a t optimum cure. More interesting than the fact that bisthioladipic acid will vulcanize butadiene copolymers is the finding that it will vulcanize Neoprene W. Considerable vulcanization is produced in Neoprene during the milling of bisthiol acid into it, although, relatively speaking not so much as in the corresponding treatment of GR-S. Owing to the prevulcanization, satisfactory, smooth test sheets were not secured when heating in molds was applied t o the milled rubber, and accordingly the tensile results shown are poor : Bisthioladipic Acid,

% 1 2

5.18 3.93

3.47 2.78

Tensile Properties after Heating Modulus Tensile E1onga.strength, tion, Harda t 300%, lb./sq. inch lb./sq. inch % ne88 425 67 452 895 500 688 350 65

That thiol acids are in fact capable of adding t o chlorovinyl groups readily (especially in the presence of a little pyridine) wa8 shown in experiments with model compounds. Thiolacetic acid and vinyl chloride gave an addition product boiling a t 49' to 52" C./5 to 6 mm., and showing ng 1.4973 and C1, 25.9%; S 24.0% (calcd. for C4H70SC1: C1, 25.6%; S, 23.1%). 2Chloro-2-butene gave an adduct with thiolacetic acid boiling a t 56O to 57' C./5 mm. and showing ng 1.4859, density a t 22' C., 1.08; C1, 21.3%, S, 19.6% (calcd. for C6H11OSC1: C1, 21.3%; S, 19.270). FURTHER STUDY O F NONSULFUR VULCANIZATION

/ . . .CH2 '

C ' H,

.

It has been known for some time that simple mercaptans (RSH) and also thioglycolic acid (HSCH&OOH) show some ability to add to the double bonds of the diene polymers, especially in the presence of peroxides. Cunneen has found that addition of thiol acids (RCOSH) is more vigorous than that of simple mercaptans or of mercapto acids ( 1 6 ) and that bisthiol adipic acid [HS.CO( CH2)aCOSHI is capable of vulcanizing natural rubber (15). Its action on natural rubber is in fact so vigorous

The subject of vulcanization by agents other than sulfur would perhaps repay fuller exploration, especially in view of the fact, noted in the section on copolymerization, that a wide variety of copolymers specially adapted to cross linking by many different bifunctional reagents can readily be prepared. It is conceivable that it may be possible by properly chosen nonsulfur vulcanization to obtain tighter cures and hence better physical properties than by sulfur, Further, it may be possible to obtain better aging than that displayed by sulfur vulcanizates. The aging of diene rubbers cured by normal sulfur vulcanization leaves a good deal to be desired. It is inferior to the aging of unvulcanized diene rubbers, and this fact may perhaps be regarded as implying that the sulfur links in the cured network are an element of weakness in respect of resistance to aging.

April 1955

INDUSTRIAL AND ENGINEERING CHEMISTRY

It is not inconceivable that cross-linking agents that are also antioxidants could be developed and antioxidant elements thus actually built into the vulcanized network. NEOPRENE VULCANIZATES

Neoprene vulcanized by zinc oxide has good resistance to aging, although, in view of our present limited knowledge of the aerial oxidation of Neoprene, it would be going too far to say that the absence of sulfur links between the chains is the only or even the main factor in the good aging displayed by cured Neoprene. In illustration of the excellent age resistance that polychloroprene is capable of exhibiting, two samples of vulcanized polychloroprene that were prepared early in 1931 and have since been in my possession show no signs whatever of deterioration; they are completely free from checking and are apparently as strong and snappy as they were originally. It should perhaps be added that these samples were made a t a time when the commercial polychloroprene was called DuPrene and was prepared presumably by the bulk polymerization of chloroprene, without modifier, to a low conversion. Whether polychloroprene prepared by polymerization in emulsion, with sulfur as a modifier, possesses such a high degree of long-time resistance to aging, I cannot say. According to Mochel and Peterson (SY), sulfur used as a modifier in the polymerization of chloroprene actually becomes incorporated into the polymer chains. I know of no evidence as t o whether such sulfur links influence the aging of sulfur-modified Neoprene. The mechanism of the vulcanization of Neoprene by metallic oxides is by no means fully understood. The trend of present opinion seems to be that the vulcanization is dependent on the prescnce of a small proportion of reactive (nonvinyl) chlorine in Neoprene, the presence of which is indicated by the fact that about 1.5% of the chlorine is removable by organic bases ( S I , 45). ATTEMPTED REACTION OF NEOPRENE AND METAL ALKYLS

How unreactive is the vinyl chlorine of Keoprene (it arises from 1,4-addition of the monomer, -CH&Cl:CHCHZ-) may be illustrated b y the negative results of certain experiments by G . P. Wheeler in our laboratory. The Staudingers (48) found zinc ethyl to react readily, even at low temperatures, with rubber and balata hydrochlorides, to replace the chlorine (nonvinyl) by ethyl and yield ethyl hydrorubber and ethyl hydrobalata. They stated briefly that zinc dimethyl and zinc diethyl react with polychloroprene, but that they had had to postpone investigation of the products. If indeed the chlorine atoms of polychloroprene could be reacted with zinc dimethyl and thus replaced b y methyl groups, it would be of great theoretical interest, since the product would be expected to be trans-polyisoprene. We found, however, that even when the metal alkyl was in excess and the temperature was raised to 95” C. no evidence of reaction between Neoprene and either zinc or cadmium dimethyl could be secured. And, further, the model compound, 2-chloro-2-butene, containing vinyl chlorine, gave no evidence of replacement of the chlorine b y methyl when it was treated with zinc dimethyl a t the boiling point of toluene. SUMMARY

1. Since the introduction of synthetic rubbers, the subject of rubber has entered into the main stream of chemistry in a way and t o a n extent that did not prevail before, when the raw rubber used by industry was restricted to the natural product. Rubber as a field of research is now a n integral and substantial part of the new realm of polymer science. 2. The biogenesis of rubber is discussed, and i t is concluded that a n understanding, when attained, of the way in which plants synthesize rubber is unlikely to point the way to a practical manufacturing process for the production of synthetic cis-polyisoprene. 3. Attention is drawn to the meagerness of our knowledge

819

of the chemical constitution of the natural antioxidant of Hevea rubber. 4. The desirability of studying variation among Hevea trees in respect of the quality of the latex and rubber is suggested. 5. New types of polymerization discussed with reference to their significance in the development of new and improved elastomers are ( a ) popcorn polymerization, ( b ) graft polymerization, and ( c ) block polymerization. The preparation of graft synthetic elastomers is described. It is shown, by fractional precipitation and otherwise, that, when methyl methacrylate is polymerized in the presence of polybutadiene or poly-(butadiene-styrene), branches of polymethyl methacrylate become grafted onto the stock polymer. The grafts have a marked reinforcing effect. 6. Some aspects of nonsulfur vulcanization are considered. It is reported that bisthiol acids are capable of vulcanizin Neoprene. Also, it is reported that the chlorine of Neoprene fails to react with zinc dimethyl. ACKNOWLEDGMENT

Appreciation is expressed to the Office of Synthetic Rubber, Reconstruction Finance Corporation, for permission to quote work carried out under its sponsorship. LITERATURE CITED

Aschan, O., Chem.-Ztg., 49, 689 (1925). Bloch, K., and Rittenberg, D., J . Biol. Chem., 143, 297 (1942); 145, 625 (1942); 159, 45 (1945); 160, 417 (1945). Bloomfield, G. F., J. Rubber Research Inst. Malaya, 13, Communication 271 (1951). Bloomfield, G. F., Rubber Chem. and Technol., 24, 737 (1961). Bloomfield, G. F., Merrett, F. M., Popham, F. J., and Bwift, P. McL., I n d i a Rubber World, 130, 670 (1954); Rubber Age, N. Y . , 75, 529 (1954). Bonner, J., and Arreguin, B., Arch. Biochem., 21, 109 (1949); Arreguin, B., and Bonner, J., Ibid., 26, 178 (1950); Arreguin, B., Bonner, J., and Wood, B. J., Ibid., 31, 234 (1951). Bonner, J., Parker, M. W., and Montermosso, J. C., Science, 120, 549 (1954). Bougault, J., and Bourdier, L., Compt. rend., 147, 134 (1908); J. pharm. chim., 29 (6), 561 (1909); 30 (6), 10 (1909); Bougault, J., Compt. rend., 150, 874 (1910); J. pharm. chim., 1 (7), 425 (1910); 3 (7), 101 (1911); Bougault, J., end Cettelain, E., Compt. rend., 186, 1746 (1928). Boundy, R. H., Boyer, R. F., and Stoesser, S. M., “Styrene,” pp. 729-31, Reinhold, New York, 1952. Braden, M., Fletcher, W. P., and McSweeney, Trans. Inat. Rubber Ind., 30, 44 (1954). ENQ.CHEM., Bruson, H. A., Sebrell, L. B., and Vogt, W. W., IND. 19, 1187 (1927). Carothers, W. H., Williams, I., Collins, A. M., and Kirby, J. E., J . Am. Chem. SOC.,53, 4203 (1931). Coffey,D. H., and Meyrick, T. J., I n d i a Rubber World, 130, 671 (1954). Compagnon, P., and Tixier, P., Rev. gen. caoutchouc, 27, 691, 663 (1950); India-Rubber J., 119, 810 (1950). Cunneen, J., J. A p p l . Chem., 2, 353 (1952); Rubber Chem. and Technol., 26, 370 (1953). Cunneen, J., J. Chem. SOC.,1947, 134. Dunbrook, R. F., I n d i a Rubber World, 117, 745 (1948). Dunn, A. S., and Melville, H. W., Nature, 169, 699 (1962); Hicks, J A., and Melville, H. W., J. Polymer Sci., 12, 461 (1954). Favorsky, A. E., and LQbBdBva,A. I., Bull. soc. chim., 6, No, 6, 1347 (1939); J. Gen. Chem. (U.S.S.R.), 8, 879 (1938). Fischer, F. G., and Lowenberg, K., Ann., 494, 263 (1932); Fischer, F. G., and Hultssch, K., Ber., 68, 1726 (1936); Fischer, F. G., and Schulse, H., Ibid., 75, 1467 (1942). Freeman, R., I n d i a Rubber World, 130, 669 (1954). Guenther, E., Soap, 19 (October and November 1943). Hall, J. A., Chem. Revs., 20, 305 (1937). Heilbron, I. M., Jones, E. R. H., Roberts, K. C., and Wilkinson, P. A., J . Chem. SOC.,1941, 344; Rubber Chem. and Technol., 15, 91 (1942). Hesse, G., Eilbracht, E., and Reicheneder, Ann., 546, 233 (1941). Holt, W. L., IND. ENG.CHEM.,23, 1471 (1931). Johnson, J. R., “Record of Chem. Progress” (Kresge-Hooker Scientific Library), Summer 1950. Jones, M. H., Melville, H. W., and Robertson, W. G. P., Nature, 174,79 (1954). Kemp, A. R., and Peters, H., I n d i a Rubber World, 110. 639 (1944).

INDUSTRIAL AND ENGINEERING CHEMISTRY

820

Kolthoff,I. M., Lee, T. S., and Cam, C. W., J . Polymer Sci., 1, 429 (1946).

Kovacio, P., Rubber Age, N . Y., 75, 700 (1954). Langdon, R. G., and Bloch, K., J. Biol. Chem., 200, 129 (1953). Leeper, H. M., and Schlesinger, W., Science, 120, 185 (1954). Mark, H., Textile Research J., 23, 294 (1953). Marvel, C. S., and Kraiman, E. A., J . Org. Chem., 18, 707 (1953).

Mayo, F. R., and Walling, C., Chem. Reva., 46, 191 (1950); Alfrey, T., Bohrer, J. J., and Mark, H., “Copolymerization,” Interscience, New York, 1952. Mochel, W. E., and Peterson, J. H., J . Am. Chem. Soc., 71, 1426 (1 949)

I

Morton, A. A., Rubber A g e , N . Y., 72, 473 (1953). Peters, R., Endeavour, 13, 147 (1954). Potter, V. R., and Heidelberger, C., Nature, 164, 180 (1949); Rudney, H., Lorber, B., Utter, M. F., and Cook, M., Federation Proc., 9 , 179 (1950). Resing, W. L., India Rubber World, 130, 670 (1954). Rittenberg, D., and Bloch, K., J . Biol. Chem., 154, 311 (1944). Ritter, F. J., India-Rubber J., 126, 55, 70 (1954). Robinson, R., I X Congr. intern. qutm. pura 2/ aplicada ( M a d r i d ) , 5, 17-38 (1934). Salomon, G., and Koningsberger, C., Rec. trav. chim., 69, 711 (1950).

Schlesinger, W., and Leeper, H. M., Science, 112, 51 (1950); IND.ENG.CHEM.,43, 398 (1951).

Direct hydrogenation of char provides

. , , an

insight into the mechanism of methane formation

. . .a potential method for producing high 8. t. u. 9-

Vol. 47, No. 4

(47) Schopfer, W. H., and Grob, E. C., Experientia, 8 , 140 (1952). (48) Staudinger, H., and Staudinger, Hj., J . prakt. Chem., 162, 148 (1943); Rubber Chem. and Technol., 17, 15 (1944). (49) Stewart, A. W., “Recent Advances in Organic Chemistry,” 5th ed., Vol. I, p. 292, Longmans, London, 1931.

(50) Sturgis, B. M., Baum, A. A., and Trepagnier, J. H., IND.ENG.

CHEM.,39, 64 (1947). (51) Weinstein, L. H., Robbins, W. R., and Perkins, H. F., Science, 120, 41 (1954). (52) Whitby, G. S., Kolloid-Z., 12, 147 (1913). (53) Whitby, G. S I “Plantation Problems of the Next Decade,” Weltervreden, Java, 1914; Ann. Botany (London), 31, 313 (1919); “Plantation Rubber and the Testing of Rubber,” pp. 5-7, Longmans, London, 1920. (54) Whitby, G. S., Dolid, J., and Yorston, F. H., J . Chem. SOC., 1926, p. 1448. (55) Whitby, G. S., and Greenberg, H., Biochem. J . , 35, 640 (1941); Rubber Chem. and Technol., 1 5 , 9 6 (1942). (56) Whitby, G. S., and Greenberg, H., IND.ENG.CHEM.,18, 1168 (1926). (57) Whitby, G. S., Gross, M. D., Miller, J. R., and Costanza, A. J., Chem. Eng. N e w , 29, 3952 (1951) ; J . Polymer Sci., in press, 1955; of. Willis, J. M., IND.ENG.CHEM.,41, 2276 (1949). (58) Wibaut, J. P., Rec. trav. chim., 62, 205 (1943). RECEIVED for review October 7,1954.

ACCEPTED January 5. 1955.

Kinetics of Carbon Gasification INTERACTION OF HYDROGEN WITH LOW TEMPERATURE CHAR AT 1500”TO 1700” F. C. W. ZIELKE AND EVERETT GORIN Pittsburgh Consolidation Coal Co., Library, P a .

N EXPERIMENTAL program has been in progress in these laboratories for some time on the kinetics of the gasification of low temperature char by hydrogen steam mixtures. Relatively large quantities of methane were obtained in these experiments ( 6 , 7 ) ,particularly at high pressures and high hydrogen-sfeam ratios. I n order to provide a better insight into the mechanism of methane formation, it was desirable to study the rate of gasification of char in pure hydrogen. This study is also of potential practical importance in connection with the production of high B.t.u. gas by the hydrogenation of coal or char. Some data have already been reported ( 7 ) on the hydrogenation of char a t 1600” F. These data have been extended to cover a wider pressure range and also two other temperatures. It has been long known that methane can be synthesized by the direct hydrogenation of carbon. The only extensive work that has been done previously in this field, however, is that of Dent and coworkers (2-4). These workers obtained integral rate data by heating various British coals and cokes in a batch bed to final temperatures of 800” to 900” C. under hydrogen pressures up to 50 atmospheres. When the final temperature was reached, i t was maintained constant while the hydrogenation was continued. It is only during this period that a comparison between their data and our data is possible.

EXPERIMENTAL TECHNIQUE

Procedure. The equipment and experimental procedure employed in this investigation were substantially the same as described previously for the char-steam kinetics work (6, 6). The inlet gas in the majority of the runs was pure hydrogen which was passed over a nickel catalyst for the removal of oxygen and through silica gel and magnesium perchlorate drying tubes before entering the reaotor. The feed char in all runs was the prepared 65- to 100-mesh Disco which was used in the previous char-steam kinetics studies. The char in each experiment was pretreated for one hour at the reaction temperature by fluidizing in purified nitrogen before the hydrogen was introduced. A fluidizing velocity of 0.44 foot per second was employed. Exit gas rates and gas samples were taken throughout the runs using pure hydrogen as the inlet gas while the temperature, pressure, and fluidizing velocity were held constant. The weight of feed char used in most runs was 0.4 pound. The reactor was the same ll/&ch diameter Uniloy reactor previously described (6). Besides the pure hydrogen runs, a series of experiments was conducted t o determine the effect of methane in suppressing the reaction rate b y adding methane to the inlet hydrogen. It was