technical review - American Chemical Society

---

TECHNICAL REVIEW

Sulfur Vulcanization of Hydrocarbon Diene Elastomers Michael M. Coleman,*' J. Reid Shelton, and Jack L. Koenig Division O f Macromolecular Science,

C a m Western Reserve University,

his whole lorge Sweat to work Out Achilles his armour" Sir Thomas brow^ne 11505-1682)

Michael M. Coleman receiued his B.Sc. in chemical technology at Borough Polytechnic, London, En. gland, in 1968 and his M.S. and Ph.D. in macromolecular science from Case Western Reserue lJniversit2 in 1971 and 1973, respectivi?ly. He joined the Elastomer Chemicals Departmenl : of the Du Pont Company as a research chemist ii% September, 1973. Prior to obtaining his undergradLlate degree, he worked as an analytical chemist for nine years in Zambia and England Dr Coleman has seueral Dublications in the fiela on the ay. I

J. Reid Shelton receiued his B.S. in Chemistry in 1933, his M.S. in 1934, and his Ph.D. in 1936, all from the State Uniuersity of Iowa at Iowa City. He joined the faculty of Case Institute of Technoloev in 1936 and has been a ill1 professor since 1949. He is author and coauthor of more than 80 research publications mainly in the areas of oxidation and mechanism of inhibitor action, mechanisms in rubber uulcanization, chemistry o f free radicals, and reactions of peroxides with organic sulfur compounds. He is a consultant in organic and polymer chemistry with special interests in the chemistry of natural and synthetic rubbers. He is presently Professor of Organic Chemistry and Macromolecular Science, Department of Chemistry, Case Western Reserve Uniuersity.

154

Ind. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3,1974

Jack L. Koenig receiued his B.S. degree from Yankton College in 1956 and his Ph.D. at the Uniuersity of Nebraska in 1959. He joined the Plastics Department of the Du Pont Company as a research chemist. In 1963, he was appointed as an assistant professor at Case Institute of Technology and has been a member of the faculty since that time. He is presently Professor of Macromolecular Science in the School of Engineering at Case Western Reserve Uniuersity. Dr. Koenig's research interests include the characterization of polymeric systems using spectroscopic methods including infrared, Raman, and nuclear magnetic resonance. He is a Fellow and member of the executive committee o f the Polymer Physics Division of the American Physical Society. He is a member of the Polymer Chemistry Division of the American Chemical Society. Dr. Koenig is on the editorial board of the reuiew journal entitled Macromolecular Reviews and is a member of the editorial board of the Journal of Macromolecular Science. He served on the staff of the Diuision of Materials Research at the National Science Foundation as Program Director of Solid State Chemistry and Polymer Science in 1972-1973.

'

Address correspondence to this author at E. I. du Pont de Nemours and Co.. Elastomer Chemical Department, Experimental Station, Wilmington. Del. 19898.

Sulfur vulcanization has been extensively studied for over a century since the initial discoveries of Charles Goodyear and Thomas Hancock. Nevertheless, t h e mechanism of unaccelerated and the more industrially significant accelerated sulfur vulcanization is still not completely understood. Some of t h e v e r y basic questions still have to be elucidated fully. A major debate concerning t h e mechanism has centered around t h e question of w h e t h e r t h e predominant reaction is of a free radical or ionic nature. During t h e 1950’s, free radicals were in vogue and m a n y researchers advanced free radical mechanisms for sulfur vulcanization. In t h e late 1950’s a n d early 1960’s, however, t h e British NRPRA group suggested a polar mechanism which received wide acceptance. In recent years, there has been a tendency to view sulfur vulcanization in terms of a mixed free radical and polar mechanism. There is reason to believe that both free radical a n d ionic reactions are taking place simultaneously with one or t h e other predominating at different stages of t h e overall vulcanization reaction depending on t h e vulcanizing system. In this review the free radical and polar mechanisms are discussed together with recent spectroscopic evidence that supports t h e mixed free radical and ionic mechanism for accelerated SUIf u r vulcanization.

Introduction Historical Development of Vulcanization. Vulcanization, a word derived from the mythical Roman God of fire and metal working, is used to describe the process whereby a rubber reacts with sulfur to produce a network of cross-linked polymer chains. Rubber has been known for over 450 years but it was only in the early 1800’s that the material gained universal acceptance following the discovery of the vulcanization process. Before the advent of vulcanization, rubber was used for such purposes as waterproofing, footwear, etc. However, i t unfortunately had a tendency to melt in the summer, freeze hard in the winter, and develop offensive odors over a relatively short period of time ( I ) . The discovery of vulcanization is usually attributed jointly to Charles Goodyear in the United States and Thomas Hancock in England. Both were awarded patents in the early 1840’s ( 2 ) . The vulcanized (or “changed”) rubber showed a remarkable improvement in the chemical and physical properties compared to the unvulcanized material. It no longer melted a t elevated temperatures nor stiffened on contact with ice, and furthermore, was much more chemically resistant. Hancock is credited with first observing that sulfur alone would vulcanize rubber. By introducing a rubber sheet into molten sulfur, he observed that the sulfur migrated slowly through the sheet. From these studies, Hancock discovered the technically important material-ebonite. On the other hand, Goodyear approached the subject in a somewhat different manner. He dissolved the rubber in turpentine and mixed it together with a suspension of sulfur and white lead in the same solvent. A composite material was prepared using layers of cotton batting which was then exposed to a “high degree of heat.” Goodyear, without fully realizing the significance, had used the first inorganic accelerator (basic lead carbonate) in his formulation. Although the rubber materials developed from the formulation of Goodyear and Hancock were far superior in many ways when compared to the unvulcanized rubber, they were still far from optimum. Large amounts of sulfur and relatively long curing times were necessary. Overvulcanization, which leads to a marked deterioration of physical properties, was a serious problem. The vulcanizates were highly colored, prone to sulfur blooming (diffusion of sulfur to the surface), and exhibited poor age resistance. It is now known that the network derived from unaccelerated sulfur vulcanization (or in the presence of inorganic accelerators) is very complicated. Apart from the many different sulfidic cross-links the vulcanizate contains a large proportion of main chain modifications such as cyclic sulfides, conjugated unsaturation, and cis/trans isomeriza-

tion of the double bond. Unaccelerated sulfur vulcanization is thus a very inefficient process. The next major breakthrough in the subject of vulcanization chemistry came with the discovery of organic accelerators in the early 1900’s (38). Apart from the increased rate of vulcanization, implied in the name accelerator, it was found that there were many other advantages to accelerated sulfur vulcanization. The use of organic accelerators allowed the vulcanization temperatures to be lowered and the cure times to be reduced. Consequently, the rubber was not subjected to such drastic conditions which in turn minimized the possibility of thermal and oxidative degradation. Furthermore, the level of sulfur could be reduced and still retain optimum physical properties of the vulcanizate. This resulted in a reduction of sulfur blooming and far superior aging properties. The possibility of overvulcanization was also lessened due to the “plateau” effect observed in the graph of sheer strength us. cure time resulting from the lower concentration of.sulfur employed. It was also found that transparent and virtually colorless vulcanizates could be prepared from accelerated sulfur systems. Carbon black and other fillers which retard vulcanization could also be incorporated in the rubber mix because of the enormous increase of vulcanization rate using organic accelerators. This was found to be an extremely important factor due to the synergistic effect of carbon black on the physical properties of the final product. In terms of network features, the network derived from accelerated sulfur vulcanization is found to be far simpler than obtained if only sulfur is used to vulcanize the rubber. There is considerably less main chain modification with accelerated sulfur vulcanization and so it is called an “efficient” process. The first patent for organic accelerators was awarded to the German firm of Bayer and Co., and it covered the use of aliphatic, cycloaliphatic, and heterocyclic amines. In common with many important discoveries, the chemists Wolfgang and Walter Oswald were actually screening organic bases for possible applications as antioxidants when they observed the acceleration effect of piperidine. It was then observed that the addition of strong organic bases greatly accelerated the vulcanization. Independent of the German chemists, Oenslager of the Diamond Rubber Co. had already discovered organic accelerators in 1906 but his findings were not published until after the First World War. Oenslager had initially used aniline, but due to its toxicity, he substituted thiocarbanilide and other compounds which are less toxic. There was tremendous activity in the field of accelerator chemistry after the First World War. Three major classes of accelerators were developed that are still used Ind. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3, 1974

155

extensively today. These are as follows: (1) accelerators based on guanidines, (2) accelerators based on dithiocarbamic acid, and (3) accelerators based on Z-mercaptobenzothiazole. Only three guanidine type accelerators are commercially available and by far the most important is diphenylguanidine (DPG) (see Table I). Although for some time this type of accelerator was very popular, they are now used mainly as secondary accelerators in conjunction with primary accelerators such as 2-mercaptobenzothiazole (MBT) and its derivatives. The dithiocarbamates, which are classed as ultraaccelerators due to their extremely rapid acceleration properties, came into prominence around 1918. In fact, accelerated sulfur vulcanization using the ammonium salt of dialkyldithiocarbamic acid (ADADC) is so rapid that it causes serious “scorch” (premature vulcanization during compounding) problems. Attempts to moderate the extraordinarily rapid acceleration properties of ADADC resulted in the discovery that the metallic salts, especially the zinc salt (ZDADC), cured a t a much slower rate. However, ZDADC-accelerated sulfur vulcanization is still too rapid for most purposes and prone to scorch difficulties. The thiuram sulfides, e.g., tetramethylthiuram disulfide (TMTD), tetramethylthiuram monosulfide (TMTM), and dipentamethylenethiuram tetrasulfide (DPMTT), were the outcome of further experimentation designed to substitute the mercaptan sulfur of dithiocarbamic acid in order to obtain a more satisfactory cure rate. Thiuram sulfides have become very important accelerators. The significance of zinc oxide and fatty acids as activators discovered in the early 1920’s (1) for the thiuram-based accelerator systems has made them very versatile vulcanizing systems for both natural and synthetic rubbers. Accelerators based on 2-mercaptobenzothiazole (MBT) and its derivatives are without doubt the most important class of accelerators used industrially. The development of MBT-based accelerators is attributed jointly to C. W. Bedford and L. B. Sebrell in the United States and simultaneously to G. Bruni and E. Romani in Italy during the early 1920’s. These chemists discovered that MBT, the disulfide derivative (MBTS), and its metal salts (e.g. the zinc salt; ZMBT) were very effective accelerators with some very important advantages over those based on dithiocarbamic acid. The major advantage was that the MBT type accelerator offered outstanding scorch delay properties, which is an exceptionally important factor in the industrial manufacture of rubber goods. Efforts to optimize the scorch delay properties of the MBT type accelerators led to the development, in 1932, of a n important new subgroup of accelerators. These accelerators are called sulphenamides. The mercaptan sulfur of MBT is substituted by an amine (see, e.g., DABS and CBS in Table I) and this substitution results in a retardation of the vulcanization. These materials proved to be very valuable and they are probably a t present the most widely used accelerators. A typical natural rubber stock which illustrates the complexity of the formulations currently used industrially is shown in Table II. The Mechanism of Sulfur Vulcanization

In spite of the fact that it is now over 130 years since Goodyear and Hancock independently discovered sulfur vulcanization, the mechanism is still not completely understood. It may seem rather surprising that considering the technical importance of rubber, the voluminous literature, and the numerous chemists that have been con156

Ind. Eng. Chem.,‘Prod. Res. Develop., Vol. 13, No. 3,1974

Table I NH DPG

S $N-

1.

c -s-NH,+

ADADC

S ($N-

Me2N-C Me,N-C

II

C -S-),Zn2+

ZDADC

S

S

-S-SS S

C -NMe,

1I

II

II

II

TMTD TMTM

-S-C-NMe,

DPMTT MBT

MBTS

ZMBT

DABS

fJ>-s-NH

‘0

CBS

Table I1

Stock Smoked sheet (natural rubber) HAF Black (carbon black) Zinc oxide (activator) Stearic acid (activator) Sulfur (vulcanizing agent) Phenyl-p-naphthylamine (antioxidant) N-Nitrosodiphenylamine (retarder) N-Cyclohexylbenzothiazole-2-sulfenamide(CBS)

Phr 100 50 5 .O 3 .O 2.5 1.o 1.o 0 .5

cerned with the mechanism of sulfur vulcanization, there are still many basic questions that have to be elucidated fully. One of the major problems has been the chemical characterization of insoluble cross-linked networks that result from the vulcanization process. As a consequence, most of the work in this field has been accomplished using indirect methods. Low molecular weight analogs of natural and synthetic rubbers have been reacted with vulcanization chemicals to give compounds that are amenable to study by conventional analytical techniques (6-8, 36, 58, 70, 71). Another indirect approach to elucidate the structure of the vulcanizate has been the application of chemical probes. These chemical probes are designed to selectively cleave specific types of sulfidic cross-links and have widely added to our knowledge of the type and quantity of specific cross-links in the vulcanizate (16, 18, 20, 35, 49, 50, 60, 66). Instrumental techniques, in the main, have not been very helpful for the chemical characterization of rubber vulcanizates, but there have been some advances

recently using Raman spectroscopy (21, 23, 40, 63). Carbon-13 nmr spectroscopy also appears to offer potential for the future (14). A further problem that has contributed to the difficulties encountered by those probing the mechanism of sulfur vulcanization has been the complexity of the multicomponent formulations employed. Apart from the many materials originally introduced (see Table 11), the reaction of these materials with one another produces a multitude of intermediates. Ironically, the simplest formulation for sulfur vulcanization developed by Hancock, in which the rubber is reacted with sulfur alone, produces the most complex network. The incorporation of organic accelerators and activators, while increasing the complexity of the formulation, reduces the complexity of the resultant network. Significantly, there has been considerable interest in recent years in the so-called “efficient vulcanization” (EV) systems where the goal is to produce only monosulfidic cross-links and a minimum of other cross-links and main-chain modifications. The EV systems are based on rather complicated formulations and contain more than one organic accelerator (59, 64). The major debate concerning the mechanism of sulfur vulcanization has centered around the question of whether the predominant reaction is of a free radical or ionic nature. This is due to the fact that the eight-membered sulfur ring has the potential of undergoing homolytic or heterolytic fission as follows homolytic ’

s-s,-s.

heterolytic +

s-ss-s-

In the following two subsections the mechanism of unaccelerated and accelerated sulfur vulcanization will be briefly reviewed prior to the forthcoming discussion on the results of the Raman and electron spin resonance (esr) spectroscopic studies. Unaccelerated Sulfur Vulcanization. Our knowledge of the mechanism of unaccelerated sulfur vulcanization comes almost exclusively from studies of the sulfuration of model olefins (6-8, 36, 58, 70, 71). Farmer and Shipley (33) and Armstrong, et al. ( 3 ) , studied the reaction of sulfur on olefins such as cyclohexene and isobutylene. These researchers found that the predominant products were alkyl alkenyl polysulfides. On the basis of these results and those obtained from studies of the reaction of hydrogen sulfide and tetrasulfides with olefins (57) a mechanism was advanced by Farmer and Shipley (33) and Bloomfield and Naylor (13) for the overall reaction of sulfur with olefins. The mechanism is illustrated in Reaction Scheme 1, using isobutylene as an example. In essence the mechanism predicts that sulfur forms a sulfenyl radical which abstracts an allylic hydrogen from the olefin forming a carbon radical. The latter attacks sulfur producing a polysulfenyl radical which in turn abstracts a hydrogen from a second olefin molecule. The hydrogen alkyl polysulfide, RSxH, then adds ionically to a third olefin molecule according to the Markovnikov direction. One difficulty with the mechanism is that the hydrogen alkyl polysulfides which are postulated as intermediates have not been detected (57). More serious objections to the mixed free-radical and ionic mechanism have been put forward by Bateman and his coworkers (20) a t the NRPRA Laboratories in England. These researchers carried out detailed studies of

Reaction Scheme 1

YS;

-

+ CH,-C=CH,

YS,H

+ .cH,--C=CH,

i“

CH3

H-S,

I -CH2-C=CH2 ionic

the sulfuration of a number of olefins including oct-1-ene, cyclohexene, hept-2-ene, 2-methylpent-2-ene, and 2,6dimethyl-2,6-octadiene (6, 7, 8, 58). The latter is a low molecular weight model for the isoprenoid structure of natural rubber. The main product of the sulfuration consists of alkyl alkenyl polysulfides as previously reported by Farmer and Shipley but in addition, small amounts of thioepoxides and considerable amounts (15-3570 depending on the olefin) of cyclic sulfides were found. On the basis of these results and a kinetic study by Ross (58), a new ionic mechanism was proposed by the NRPRA group (7, 20). The overall reaction mechanism is illustrated in Reaction Scheme 2 using the hypothetical olefin R-H. Reaction Scheme 2 Initiation:

polysulfide P

--+

TS.+

+ TS;

Propagation:

TS,R

Termination:

+ RH2+

TS,RH2

+ R+

RH^+

TS,RH+

The essential features of the mechanism are as follows. Initiation involves heterolytic scission of an S-S bond in initially formed polysulfides or molecular sulfur to yield persulfenium ions TS,+. Persulfenyl anions TSb- act as terminators. TSa+ adds to the double bond to give the cyclic persulfonium ion (I) which may undergo a number of competitive reactions depending on its structure.

So

I

T (1)

The ability of the polar mechanism to predict the precise alkyl alkenyl polysulfides found from oct-1-ene and 2-methylpent-2-ene is compelling evidence for this mechaInd. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3, 1974

157

nism. An example is shown in Reaction Scheme 3 for 2methylpent-2-ene. A similar reaction scheme, but involving H + transfer instead of H- transfer, is postulated for the sulfuration of oct-1-ene. This is detailed in the book edited by Bateman (20), and predicts the formation of the alkyl alkenyl polyshown below sulfide

Reaction Scheme 4

(m)

CH,-CH=-CH+,H,

I

r

CH3-CH-CH,-C,Hii

(111) Reaction Scheme 3

TS,+

+ Me,C=CH-CH,Me

-

Me,C=CH---CH,Me '\*

+ ,/

,, So

I

T Me,C==CH-

+

CH-Me

Me,C=CH+H-Me

b H' transfer

Me,C;--/CH+H,Me '\+I \ I

Me&-

Sa

CH2- CH,Me

I

I

I

Reaction Scheme 5

7';

T

T

% Me,C=CH---CH-Me

Me&=CH-:H-Me

I

I

S,

+

ma-1

P

Figure 1 shows the Raman spectra of unvulcanized cis1,4-polybutadiene (CB) and sulfur vulcanized CB. The Raman line a t 1664 cm-1 is indicative of the trans isomer. A feasible ionic mechanism for the cis/trans isomerism based on the cyclic persulfenium ion is shown below.

Me,C=CH--CH-Me

1

H- transfer

/

k C 4 H Me,&CH,-CH,Me

ma+

(Cis)

\

HC-=CH '\ \ +I I(

/

>C-CH

I

/

+

(11)

Bateman and his coworkers (20) point out that if the mechanism were indeed free radical then the reaction of sulfur with oct-1-ene and 2-methylpent-2-ene, respectively, would give predominantly (IV)and (V). CH?

CH-CH-C,Hil

CH,-CH=CH

I

s.

+

CHi-CHZ-CH,+,Hl1 Me2C=CH-CH

Me,CH-

I

s.

CH-CH,Me

I

s,

I CHZ-CH2-CHZ+BH11

CH2- CMe=CH-

+

I I Me,CH-CH-CH,-Me S*

-C,Hll

(W CH,Me

(V)

This is contrary to experimental fact. Further confidence for the polar mechanism comes from the ability to predict the cyclic sulfidic structures and conjugated species that are found on the sulfuration of 1:5 dienes. Examples are shown in Reaction Schemes 4 and 5. It is also well known that unaccelerated sulfur vulcanization leads to considerable cis/trans isomerization of the double bond which is in contrast to the case of accelerated sulfur vulcanization. We have obtained some direct evidence for this isomerization using Raman spectroscopy. 158

Ind. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3, 1974

T

T

Jt TS,'

+ \HC-CH

\ (trans)

&C"+H

I

\

T

In addition to the wealth of data from model compound studies that supports the ionic mechanism there are two other pieces of important supporting evidence. Esr studies of natural rubber-sulfur mixes during vulcanization do not indicate the presence of free radicals (IO), and freeradical acceptors (hydroquinone, benzoquinone, etc.) do not inhibit the cure of a styrene butadiene rubber-sulfurzinc oxide-stearic acid mix (62). A critical review of the conclusions of Ross (58) and Bateman, et al. (7, 20), has been published by Prior (57). He notes that both the mechanism of Farmer and Shipley (33) and the Bateman group postulate ionic addition steps but that they differ mainly in the dissociation step which may involve either free radicals or ions. The NRPRA group favored ionic dissociation because, a t that time, no examples of thermally induced homolysis of alkyl disul-

1RANS

1700

A

1650

MOO

cm-1

Figure 1. Raman spectra: cis/trans isomerism in sulfur-cured CB vulcanizates.

fides had been reported a t temperatures below 140°C. However, as Prior points out, there is no evidence of thermally induced heterolysis of polysulfide bonds either and the presence of polysulfenyl radicals cannot be eliminated because the possibility of homolytic scission of complex polysulfides present in sulfur-olefin reactions is high even a t temperatures below 140°C. Prior also questions other pieces of evidence based on the effect of additives and polar solvents that the NRPRA group use to support their totally ionic mechanism. More recently, Wolfe and his coworkers (71) have employed gas-liquid chromatography and time-of-flight mass spectrometry to study the sulfuration of cyclohexene. Their results are accounted for by a radical chain mechanism in combination with a secondary polar reaction which is very similar to that proposed.by Farmer and Shipley (33) and subsequently rejected by Ross (58) and Bateman, et al. (7, 2 0 ) . Thus we turn full circle. Wolfe and his coworkers do not completely rule out the possibility of polar reactions but they consider the free-radical chain mechanism more likely in view of the tendency of sulfur and polysulfides to form persulfenyl radicals by thermally induced homolytic cleavage (34, 39, 54, 67, 68) and the autocatalytic nature of the reaction. They also suggest that the polar reaction may be significant in the very early stages before much polysulfide has formed as the autocatalytic nature of the reaction indicates that the thermal homolysis is more easily undertaken by linear polysulfides than the cyclic SSmolecule. Significantly, in the presence of zinc oxide, Wolfe, et al., consider the reaction of sulfur with cyclohexene to be of an ionic nature. Zinc oxide is postulated to initiate an ionic chain reaction, which due to its greater velocity, dominates the radical chain reaction. On balance, the mechanism of unaccelerated sulfur vulcanization appears to be predominantly ionic in nature but the possibility of the involvement of persulfenyl radicals cannot be entirely eliminated. Accelerated Sulfur Vulcanization. Over the past two decades there has been considerable debate concerning the mechanism of accelerated sulfur vulcanization. During the 1950’s researchers such as Scheele, Craig, Bevilacqua, Dogadkin, Blokh, and their respective coworkers (9, 12, 27, 29, 61) advanced free radical mechanisms for the accelerated sulfur vulcanization of unsaturated elastomers. In the late 1950’s and early 1960’s, however, the British NRPRA group suggested a polar mechanism for accelerated sulfur vulcanization as an extension of their work on unaccelerated sulfur vulcanization (20, 56). In discussing the mechanism of accelerated sulfur vulcanization it is immediately apparent that it is difficult to generalize. There are many unanswered questions and there is reason

to believe that both free radical and ionic reactions are taking place simultaneously with one or the other predominating at different stages of the overall vulcanization reaction depending on the vulcanizing system (44, 62, 70). In this brief review of accelerated sulfur vulcanization we will initially consider the polar mechanisms attributed to Krebs and the NRPRA group, followed by the free radical mechanisms advanced prior to 1960, and finally mixed free radical and polar mechanisms that are currently gaining acceptance. As mentioned previously, the first organic accelerators to be exploited were amines (e.g., DPG). Krebs (42) studied the reaction of sulfur with amines and the sulfide ion. He concluded that the Sg ring undergoes rapid ionic cleavage at relatively low temperatures in the presence of amines or sulfide ion.

s, + NR,

Room temp

R,N+

+

q-S-

Krebs noted that the accelerating properties of amines were a function of the base strength and suggested that amines accelerate sulfur vulcanization by ring-opening of sulfur. This conclusion is consistent with the polar mechanism advanced by the NRPRA group for unaccelerated sulfur vulcanization. It was also Krebs, et al. (43), who considered zinc oxide to participate in the formation of complexes and postulated that complexes such as zinc dimethyldithiocarbamate (ZDMDC) and zinc benzothiazolyl mercaptide (ZMBT) react with amines and sulfur in the following manner.

+

NR3 RP

= X-S-Zn-S-X

X-S-Zn-S-X

I

I

S +NR, (x)

Complexes of type IX have been isolated (26, 37, 46, 47) which increase the solubilization of the zinc salt in the rubber. The postulated complex X is of considerable mechanistic importance and will be referred to later, The most widely accepted mechanism for accelerated sulfur vulcanization is due to the British NRPRA group (20, 56) and represents an extension of their studies of unaccelerated sulfur vulcanization in conjunction with the ideas advanced by Krebs. A summary of the overall course of accelerated sulfur vulcanization as viewed by the NRPRA group is given in Table III. In essence, the mechanism may be summarized as follows. (i) Reaction of Accelerator a n d Activator

XSZnSX

RCOOH

Ligand

L ~

1

XSZnSX

t

(VIII) S

Ind. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3, 1974

159

The reaction of the accelerator with the activators (ZnO, fatty acids, amines) leads to complex IX which is in accord with the suggestion of Krebs (43). (ii) Reaction of the Zinc Accelerator Complex with Sulfur XSZnSX

+ SB

XSS,ZnS,SX

(XI)

This represents a departure from the suggestion of Krebs, who postulated complex X. There is no direct evidence for this step in the mechanism. The persulfurated zinc salt (XI) has not been isolated but it is thought by the NRPRA group to be formed in a series of equilibrium reactions which lie well on the side of the mercaptide complex (IX) and free sulfur (56). Bateman, et al., suggest the mechanism for the formation of the persulfurated zinc complex (XI) is as follows

6-

62+

JNR3 6-

canization. Chelate ring expansion has been found to occur when specific metal dithiolates react with sulfur or polysulfides as depicted below.

Compounds of type XI1 have been isolated such as nickel(I1) and zinc(I1) dithioarylates and although the zinc(I1) analog of DMDC has not been isolated it is feasible that a rapid equilibrium exists between ZDMDC, sulfur, and Me*NC(S)SZZn. It appears that insertion of sulfur into the chelate is limited to one atom and there is no evidence to suggest that greater ring expansion can occur. Fackler and his coworkers suggest the following mechanism for sulfur exchange.

+s-s,

6-

S/s-s-

'z:s;s 6+

(XI)

As Bateman and his coworkers point out, the formation of the persulfurated zinc complex (XI) results in a considerable loss of chelating energy but they suggest there is considerable stabilization from the formation of persulfenyl anions. The persulfurated zinc complex (XI) is considered to be the active sulfurating agent. In the TMTD-ZnO vulcanizing system, where there is initially no free sulfur, it is suggested that the persulfurated zinc complex (XI) is formed by interchange between thiuram polysulfides (TMTP) and ZDMDC which are formed in a rapid ionic reaction shown in Reaction Scheme 6.

/

as=c

-

/s

?-

,s\

zn?s/c' XI11

This mechanism offers an alternative scheme to that proposed by Krebs (42) for the promotion of the ionic cleavage of sulfur by ZDMDC. (iii) Formation of the Rubber-Bound Intermediate. It is now generally accepted that the precursor to the formation of cross-links in accelerated sulfur vulcanization is the rubber-bound intermediate XIV.

Reaction Scheme 6

S

S

I NMe,

I NMe,

I

(XW

S

S

II

C-S-S3

I

NMe2

-

-

I1

S-S-C-NMe,

I c=s I

NMe, S

S

1I

C-S-S-S-C

II + -S-C-NMe2

I

I

N Me,

NMe,

etc.

The above sequence leads to the formal stoichiometry nTMTD

+ ( n - 1)ZnO = Me,NC(S)SS,_,SC(S)NMe, + i(n - I)[Me,NC(S)O+Zn + k(n - ')[Me,NC(S)S+Zn

The interchange reaction is viewed as Me2NC(S)S,(S)CNMe,

+ ZDMDC == TMTD + Me,NC(S)S,-ZnSC(S)NMe, (XI)

The recent investigations of Fackler and his coworkers (32) are significant to the questions concerning the possible formation of the persulfurated zinc complex (XI) and the mechanism by which ZDMDC accelerates sulfur vul160

X is an accelerator

X

Ind. Eng. Che,m., Prod. Res. Develop., Vol. 13, No. 3, 1974

fragment

Evidence for XIV is most definitive for the TMTD-ZnO vulcanizing system. Moore and Watson (51) have demonstrated in the vulcanization of cis-1,4-polyisoprene by the TMTD-ZnO system that high amounts of nitrogen and sulfur are combined with the rubber after short cure times but that these are both reduced to limiting values as cross-linking reaches a maximum. The nitrogen and sulfur removed from the network is seen as ZDMDC which attains an approximate final yield of 72% the initial TMTD. On the basis of kinetic studies, Scheele (61) has demonstrated that the rate of consumption of TMTD is faster than the rate of formation of ZDMDC and cross-links. He proposes that intermediate compounds are formed by the reaction of sulfur and accelerators which then transfer sulfur to the rubber chain. Scheele also observed the similarity of the kinetics of accelerated sulfur vulcanization with the kinetics of many enzyme reactions which are catalyzed by intermediate products. Additional evidence for the rubber-bound intermediate comes from the fact that when an undercured vulcanizate (which has had all the extra-network materials except ZnO and ZnS removed) is reheated, new cross-links are formed. Each new cross-link is accompanied with the formation of one molecule of ZDMDC (31, 51). Model compound studies of the reaction of 2-methylpent-2-ene with TMTD a t 140°C confirm the chemical

Table 111. Summary of Overall Course of Accelerated Sulfur Vulcanization (NRPRA Group) Vulcanizing ingredients (accelerator

c Active sulfurating agent

1

R-H (R-H H

= =

+ sulfur + activators)

rubber hydrocarbon a-methylenic or a-methylic)

Rubber bound intermediate (R-S,-X)

.1J Network maturing reactions

Initial polysulfide cross-links (R-S,-R)

(1) Cross-link shortening with additional cross-linking (2) Cross-link destruction (3) Main chain modifications (dehydrogenation and cyclic sulfide formation)

+

Final vulcanizate network Service

4.

“Aged” vulcanizate network

&

structure of the rubber-bound intermediate (56). The following structures were identified

B,

02

Not all these structures led to cross-links and it was confirmed that XV is the precursor to cross-link formation (56)which occurs in the following manner. 2RS,-C(S)-NMe, RS,-C(S)-NMe,

+ R-H

-

RS,R

XS,SR

m,X

+

RSZR

A2

ZnS

; Zn---S k,6+

lZnO

RSSC(S)NMe, (XV) RSC(S)NMe, RSC(O)NMe, [RC(S)NMe,] inferred

A,

6-

+ ZDMDC + H,O

Chemical probes in conjunction with conventional analytical techniques have been employed by several chemists to detect and measure pendent side groups in vulcanizates. Campbell and Wise (19) used sodium borohydride and ultraviolet spectroscopy to estimate the number of mercaptobenzothiazole capped pendent groups. Sulfur analyses before and after treatment with chemical probes have suggested that considerable amounts of sulfur may be present in the form of pendent side groups (50, 65). Radiotracer techniques have been employed by Campbell (17) for determining the amount of pendent accelerator groups in the initial stages of vulcanization and confirms that these groups are the precursors to the formation of cross-links. They further note that larger amounts of attached accelerator fragments were found for “efficient” and so-called “sulfurless” systems than for conventional accelerator-sulfur systems especially a t short cure times. More recently, Parks and his associates (52, 53) have successfully introduced the rubber-bound intermediate onto the rubber backbone in the absence of zinc salts. These results cast doubts on the zinc perthiomercaptide complex being the active sulfurating species leading t o the formation of the rubber-bound intermediate. We will refer back to this point later in this discussion. According to the NRPRA mechanism, the zinc perthiomercaptides (XI) are considered to be the actual sulfurating species. The mechanism by which the rubberbound intermediate XIV is formed from the reaction of XI with the rubber is not known for certain but the NRPRA group favor a polar mechanism uia a cyclic transition state which is illustrated as follows.

XS,ZnS,X H,O The essential features of the mechanism are a nucleophilic attack of the terminal perthiomercaptide sulfur atom on a n a-methylenic or a-methylic carbon atom of the rubber together with concomitant displacement of hydrogen as an “incipient hydride ion” and the formation of zinc sulfide. Bateman and his coworkers (20) find supporting evidence for this mechanism based on the rate and course of the reaction in terms of whether a-methylic substitution occurs with 2-methylpent-2-ene ( i e . , A or B type substitution). They argue that coordination of electron donating ligands (such as amines or fatty acids) to the zinc atom will weaken the Zn-S bond increasing the charge on the terminal sulfur atom making it more nucleophilic. Substitution at a-methyl groups will be promoted over a-methylene groups. This effect is observed for the TMTD-ZnO system in the presence of ligands (20),

The possibility of homolysis of the zinc perthiomercaptide complex XI to yield persulfenyl radicals which could then react with the rubber to form the rubber-bound intermediate (XIV) shown in Reaction Scheme 7 , was considered by the NRPRA group but emphatically rejected for several reasons (20). Reaction Scheme 7 XSS,ZnS,.SX

R.

-XS,.

-P

XS; -t RH

-

R. + XSS,. + XSS,ZnS,SX

R.

+ ZnS t XS,. + XS,H

RSBX FS,X ZnS

+

+ XSS;

In the words of Bateman, et al. (ZO), “It is irrational that the introduction of such obviously polar or polarizable auxiliary vulcanizing ingredients as metal oxides, zinc soaps and accelerators of the amine and complex thiol or disulfide type should change the course of sulfuration from the polar processes obtaining in unaccelerated systems.’’ This statement may be too general, however, and we will take issue with this conclusion later in this review. The NRPRA group also point out that if persulfenyl radicals were formed they would be expected to undergo additive sulfuration of the olefin in competition with substitutive sulfuration which is contrary to the known disubstitutive sulfuration pattern for “efficient” accelerated sulfuration. Furthermore, if the mechanism were free radical, Ind. Eng. Chem., Prod. Res.

Develop., Vol. 13, No. 3,1974

161

one would expect mesomeric alkenyl radicals which would lead to substantial amounts of A2 type products.

* Me$-CH=CH-Me

Me2C=CH-CHMe

(A,.)

In addition, the free radical mechanism offers no explanation for the marked changes in the B and A type groups in cross-linked products from different recipes. (iv) Initial Polysulfide Cross-Links, Cross-Link Shortening, and Competing Reactions. Sulfuration of 2-methylpent-2-ene by accelerated sulfur systems leads to allylic cross-linked polysulfides (56) 2

&+&

L-

++s++

s+

The average length of the polysulfides decreases from about x = 4 to approximately x = 1 as the reaction proceeds. This cross-link shortening is accompanied by a change in the sulfur attachment. Initially A1 and B1 type groups predominate but A2 and B2 groups appear as the reaction time is increased. Furthermore, it has been demonstrated by Watson (69) that the cross-link shortening is catalyzed by ZDMDC and ZMBT and is accompanied by the formation of additional cross-links and ZnS. In summary FS,R

XSZnSX __*

RS2R

XSZnSX

RSR

+ FSR’+ XSSJnS,SX

1”-H

Additional cross-links R = A, or B, groups

R’ =

or B, groups

According to Porter (561, the pathway to the formation of cyclic species is unknown but it is known that cyclic sulfurated groups accompany cross-link destruction. The NRPRA group conclude that if desulfuration proceeds rapidly the final network will be highly cross-linked (predominantly monosulfidic). On the other hand, if desulfuration is slow, thermal decomposition ensues and a much more complex network, in terms of main chain modifications, results. Porter reports that in the sulfuration of 2methylpent-2-ene by a ZDMDC-ZnO-Ss vulcanizing system (an inefficient system), a proponderance of the more slowly desulfurated A-type polysulfides is formed. However, if a CBS-ZnO-Ss-propionic acid vulcanizing system is employed (an efficient system), the polysulfides formed are predominantly of the B type which are more rapidly desulfurated. The NRPRA group consider that zinc mercaptide complexes (IX) play a major role in determining vulcanizate structure as these complexes are involved both in the formation of rubber-bound intermediates (XIV) and in the desulfuration of the initial polysulfidic cross-links. The concentration and solubility of the zinc mercaptide complexes are major factors in the relative efficiency of vulcanization (5, 48). Many free radical mechanisms have been advanced for accelerated sulfur vulcanization (particularly those involving TMTD, MBT, MBTS, and CBS) and the “sulfurless” TMTD-ZnO system (9, 12, 27, 29, 61) prior to the development of the NRPRA polar mechanism. An excellent review has been published by Scheele (61) which covers the literature up to 1960. These free radical schemes are basically very similar and the initial stages are shown in Reaction Scheme 9. A marked similarity between this scheme and that presented in Reaction Scheme 7 for the possible homolysis of the zinc perthiomercaptide complex is also noted. Reaction Scheme 9

The mechanism is thought by the NRPRA group to proceed as depicted in Reaction Scheme 8 and is consistent with the above experimental observations. Reaction Scheme 8

xssx

XS.

R.

+ XSSX

X=Me,NC(=S)--;

+ XSSZnSX I

sx

s

I

SR XSZnSX

There are actually two major competing reactions to which the initially formed polysulfidic cross-links are subjected. They may either be desulfurated to form stable monosulfidic cross-links and the extruded sulfur used for further cross-link formation or they may decompose thermally to yield cyclic sulfides and conjugated unsaturation. The latter may be formed in the following manner

& - k+

RS,H

S,R

162

Ind. Eng. Chem.. Prod. Res. Develop., Vol. 13, No. 3, 1974

----t

+ RH R. + S8

as.

XSH RSB. RSX

+R

+ XS.

a > c

etc.

Bevilacqua (9) postulated the formation of a rubberbound intermediate, RSC(=S)NMe2, for the TMTD-ZnO vulcanizing system arising from the above scheme. He further postulated basic hydrolysis by zinc oxide or basic zinc salt and subsequent oxidation of the mercaptide by TMTD leading to RSSR and ZDMDC. This mechanism was seriously criticized by Bateman, et al. (20) who pointed out that (a) prior interaction of TMTD with ZnO was ignored, (b) hydrolysis of RSC(=S)NMe2 by ZnO is not an easy process, (c) it requires that TMTD be present for cross-linking when it is known that cross-links and ZDMDC are formed even when all the TMTD has reacted, (d) does not account for the formation of polysulfides in early stage of cure, and (e) ignores the presence of RS,C(=S)NMe2 which are thought to be precursors to cross-links. Previous esr studies have failed to produce convincing evidence for the presence of radicals in accelerated sulfur vulcanization. The esr spectroscopic studies of the vulcanization of rubber by TMTD-based accelerator systems have been plagued by the presence of minor amounts of copper present as an impurity which forms the copper complex of dimethyldithiocarbamic acid (CuDMDC) a t

vulcanizing temperatures (28). CuDMDC has a very strong esr signal and has led to misinterpretations in the past (11, 15, 29). A recent study on the mechanism of the action of retarders concludes, from esr spectroscopy, that benzothiazole sulfenamides do not undergo thermal dissociation into radicals ( 4 ) . The experiments were carried out a t 135°C and the authors point out that the involvement of short-lived radicals cannot be dismissed. Over the past few years there has been a tendency to view accelerated sulfur vulcanization in terms of a mixed free radical and polar mechanism. As mentioned previously, in a complex mixture of rubber, ZnO, sulfur, fatty acids, and accelerators it is conceivable that both free radical and polar reactions are occurring simultaneously. The dominant mechanism will be determined by a large number of variables such as the type and concentration of the accelerator, the concentration and stability of zinc complexes, etc. Studies of Shelton and McDonel (62) support this view. These authors reported the results of experiments in which free radical scavengers (benzoquinone, hydroquinone, and 1,l-diphenyl 2-picrylhydrazyl) were used in an attempt to decide between the free radical and ionic mechanisms for a number of well known vulcanizing systems. Their conclusions are summarized in Table IV. While there appears to be no ambiguity in the mechanism of either the unaccelerated and DPG-accelerated sulfur systems (polar) or the peroxide and radiation cure (free radical), the other vulcanizing systems indicate mixed polar and free radical reactions. It is important to note that zinc oxide was present in each of the systems described above with the exception of the peroxide and radiation cures. In the case of the TMTD-ZnO system the data from Shelton and McDonel’s work did in fact show that radical traps retard the vulcanization but since the possibility of direct reaction of the radical trap with TMTD had to be considered, the result could not be taken as proof of freeradical participation. The data are, however, consistent with a free radical mechanism. Similar results were reported more recently by Manik and Banerjee (44) based on kinetic studies of dicumyl peroxide (DCP) decomposition in the presence of TMTD alone or in conjunction with ZnO and/or sulfur during the vulcanization of natural rubber. The conclusions are summarized in Table V. Manik and Banerjee ( 4 5 ) , using similar techniques, conclude that the CBS-sulfur vulcanizing system acts predominantly by a free-radical mechanism. However, when ZnO and stearic acid are added the results are interpreted in terms of a mixed free radical and ionic mechanism. Wolfe (70) used the dependence of the stereochemistry on the polar or nonpolar nature of the reaction of sulfur on cyclohexene in the presence of organic accelerators based on derivatives of dithiocarbamic acid to determine between free radical and ionic mechanisms. He concludes that the ZDMDC-accelerated sulfur reaction proceeds uia an ionic mechanism while the TMTD and TMTM-accelerated sulfur reaction proceeds uia a free radical reaction. In the presence of ZnO, the latter two accelerators will form ZDMDC as the reaction proceeds causing a mixed free radical and ionic reaction. The effect of thiourea on the TMTD-ZnO vulcanizing system has been investigated by Duchacek (30). He concurs with the NRPRA mechanism as shown in Reaction Scheme 6 and explains the acceleration effect of added thiourea by postulating a thioanion disulfide interchange to give a further source of MeZNC(=S)SS-, Thiourea also acts as a radical scavenger and Duchacek suggests that the conversion of the rubber-bound intermediate into

Table IV

System Unaccelerated sulfur DPG-accelerated sulfur TMTD-accelerated sulfur ZDEDC-accelerated sulfur MBT-accelerated sulfur MBTS-accelerated sulfur CBS-accelerated sulfur T M T D alone Peroxide cure Radiation cure

Conclusion Polar Predominantly polar Mixed polar and free radical Inconclusive Free radical

Table V

System

Conclusion

TMTD TMTD-ZnO TMTD-sulfur TMTD-sulfur-ZnO

Free radical Polar Free radical Predominantly polar

cross-links is by a free radical process in accord with the theories of Scheele (61) and Bevilacqua (9). This would explain the retardation effects of thiourea in the later stage of vulcanization. From this brief review of accelerated sulfur vulcanization it is evident that the mechanism is still open to question. Formidable as the evidence for the NRPRA mechanism is, there are however, still some unanswered questions centered around whether the zinc perthiomercaptide complex (XI) is the actual sulfurating agent leading to the rubber-bound intermediate. Complexes of type XI have not been isolated and their presence is based on circumstantial evidence. The observations of Parks and his coworkers (53) casts serious doubts on the necessity for a zinc perthiomercaptide complex. The results of Shelton and McDonel (62), Manik and Banerjee (44, 451, Wolfe (70) and Duchacek (30) suggest that free radicals cannot be ignored. The presence and participation of persulfenyl radicals cannot be ruled out in the formation of the rubber-bound intermediate. Recent Raman a n d Esr Spectroscopic Studies of Accelerator Systems a n d Their Significance to the Mechanism of Accelerated Sulfur Vulcanization During the past few years we have been studying accelerated sulfur vulcanization with two spectroscopic toolsRaman and esr (22, 24, 25, 41). Laser Raman spectroscopy, with its unique sensitivity to highly polarizable bonds such as the sulfur-sulfur bond, offers considerable potential to the study of accelerated sulfur vulcanization. Employed in conjunction with esr spectroscopy, which yields information regarding the possible involvement of free radicals, it is feasible to identify intermediate products that are formed in the thermal decomposition of accelerator systems. In our view, these results are important to the mechanism of sulfuration and shed some light on the actual sulfurating agent leading to the formation of the rubber-bound intermediate. A summary of our major conclusions follows. Formation of t h e Rubber-Bound Intermediate from TMTD Based Systems. The results of our Raman and esr studies of the TMTD based accelerator systems are considered to be compelling evidence for a free radical mechanism for the formation of the rubber-bound intermediate (XIV) which is applicable in the presence and absence of zinc salts and which has as its common denominator thiuram persulfenyl radicals. In the case of the TMTD alone and TMTD-sulfur systems the whole mechanism is viewed as free radical. ThiInd. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3,1974

163

uram persulfenyl radicals are postulated and observed by the formation of T M T P as intermediate products during thermal decomposition. Supporting evidence for thiuram and thiuram persulfenyl radicals has been obtained from esr spectroscopy. Furthermore, it is significant that the esr spectrum of the samples quenched from 145°C shows the presence of thiuram persulfenyl and in the case ofLhe TMTD-Ss system the cyclic carbon radical Me2N-CLSx that are necessary for the formation of persulfurated rubber-bound intermediate. In the presence of zinc salts, a mixed ionic and free radical mechanism is postulated. A rapid ionic reaction, proposed by the NRPRA group, occurs initially yielding TMTP which is confirmed by the Raman studies. These polysulfides are susceptible to thermal cleavage at temperatures below that of the disulfide to yield thiuram persulfenyl radicals. These radicals are known to be reasonably stable (39, 54, 55) and thiuram, thiuram persulfenyl, and the cyclic carbon radicals have been detected by esr spectroscopy for the TMTD-ZnO system. The overall scheme proposed is shown in summary in Reaction Scheme 10. The free radical reaction leading to the formation of the rubber-bound intermediate is quite straightforward and closely resembles that advanced by Bevilacqua (9) in his initial step (except that i t allows for persulfurated rubberbound intermediates necessary for polysulfidic cross-links a t a later stage of the reaction) and that postulated by the NRPRA group (20) for the homolysis of the zinc persulfurated complex and detailed in Reaction Scheme 7. Although reaction step iii is reversible and it is known to be displaced mainly toward the thiyl radical, any R - formed has the possibility of being trapped by reaction with XS,X. Furthermore, in the presence of ZnO, displacement of reaction step iii toward the formation of Re would be enhanced. This mechanism (Reaction Scheme 10) is consistent with the observations of Shelton and McDonel (62) and Manik and Banerjee (44), where mixed polar and free radical reactions were indicated for TMTD-accelerated sulfur vulcanization. The mechanism is also consistent with the conclusions of Wolfe (70) and Parks, et a2. (53) where the rubber-bound intermediates were formed in the absence of zinc salts. In addition, the results of Duchacek (30) may be explained on the basis of Reaction Scheme 10.

Reaction Scheme 10

XSSX

+ ZnO

XS,X

Thermolysis

xs.+ xs,

R. + XS,H(XSH) +ZDMDC R. + XS,X RS,X + XS,-; RS,X(RSX) R. + XS;(XS.) XS;(XS.)

+ R-H

ZnO

L

-+

+

There was no evidence from the Raman studies for the zinc perthiomercaptide complex (XI) proposed by the NRPRA group and this result together with the observations of Parks and his coworkers (53) casts serious doubts on its presence and hence its participation in the formation of the rubber-bound intermediate. 164

Ind. Eng. Cheh., Prod. Res.

Develop., Vol. 13, No. 3, 1974

O n the Acceleration Properties of Thiuram Sulfides. The Raman and esr spectroscopic studies of the TMTMSg and TMTD-Ss vulcanizing systems have proved most interesting. The results of these studies have been interpreted as evidence for the interaction of the thiuram sulfide with sulfur, followed by breakdown into thiuram, thiuram persulfenyl, and cyclic carbon radicals at temperatures below that of significant homolysis of the parent thiuram sulfide. This offers an explanation of the acceleration properties of the thiuram sulfides. In general, the mechanism is viewed as

Me,N -C Z S 9

(XVI)

11 2

Me,N-C

‘Sea

+ X. (XVII)

+ x.

For TMTM X. = Me2NC(=S)S. and for TMTD X. = Me2NC(=S) SS.. The homolysis of TMTM, which is not a vulcanizing agent, a t 125°C is extremely slow and is probably due mainly to the stability of the C-S bond. However, if interaction of TMTM with sulfur occurs as depicted in XVI and evidenced in the Raman spectral studies homolysis may occur more readily. The driving force for this homolysis is the formation of the relatively stable thiuram and cyclic carbon radicals (XVII). Esr results of the TMTMSSsystems are consistent with the above mechanism. A similar mechanism is envisaged for the initial stages of the decomposition of the TMTD-Ss system but there are several conflicting reactions that must be considered (24). ZDMDC-Accelerated Sulfur Vulcanization. If it is assumed that the zinc perthiomercaptide complex (XI) is not formed from the reaction of ZDMDC and sulfur, as suggested by the NRPRA group (20), the question arises, “How does ZDMDC accelerate sulfur vulcanization and why is it much more rapid than the thiuram accelerated systems?” According to the NRPRA scheme the active sulfurating species leading to the formation of the rubber-bound intermediate is the persulfurated zinc complex XI. If the reaction of ZDMDC with sulfur leads to an appreciable concentration of XI a t vulcanizing temperatures we would expect to see significant changes in the Raman spectra due to new S-S bonds as observed for the TMTP. We do not observe any changes in the Raman spectrum of ZDMDC-S (1:2 M) after heating at temperatures of 125 and 145°C. The Raman spectra were recorded at these temperatures and after quenching in ice water. It should also be noted that in the final Raman spectrum of the TMTD-ZnO (15 M) decomposition studies only the presence of ZDMDC with sulfur is observed (24). These results are consistent with those of Craig and his coworkers who report that they could observe no reaction between ZDMDC and a radioactive isotope of sulfur up to temperatures of 150°C. If the zinc perthiomercaptide complex (XI) is rejected as the active sulfurating agent the complex (X) suggested by Krebs deserves reconsideration. ZDMDC is considered to act in much the same way as amines (such as DPG) in that it promotes the formation of ionic cleavage of sulfur. An ionic mechanism similar to that proposed by Bateman and his coworkers for unaccelerated sulfur vulcanization could then ensue. A possible intermediate is as follows.

Bateman, L., Moore, C. G . , Porter, M., J. Chem. SOC., 2866

(1958). Bateman, L., Glazebrook, R. W., Moore, C. G., Rubber Chem. Techno/., 35, 633 (1962). Bevilacqua, E. M . , RubberChem. Techno/.. 32. 721 (1959). Blokh, G. A., Dokl. Akad. Nauk, 129. 361 (1959). Blokh, G. A., RubberChem. Techno/., 33, 1005 (1960). Blokh. G. A., "Organic Accelerators in the Vulcanization of Rubber" Israel Program for Scientific Translations Ltd., 1968. Bloomfield, G. F.. Naylor, R. F.. Xlth Int. Congr. Pure Appl. Chem..

2, 7 (1947).

Alternatively, a similar type of intermediate could be constructed where the carbon atom of ZDMDC is considered to be the acceptor in the manner discussed by Fackler and coworkers (32). This type of ionic mechanism is consistent with the observation that ZDMDC-accelerated sulfur vulcanization is more rapid than thiuram-accelerated sulfur vulcanization which occurs by a free radical mechanism. It is also documented that the ZDMDC-accelerated sulfur system is an inefficient vulcanizing system (56) which is suggestive of a n ionic mechanism that involves the double bond. Furthermore, Wolfe (70) concludes from his studies that ZDMDC-accelerated sulfuration occurs uia an ionic mechanism. A final comment on vulcanizing systems based on derivatives of dithiocarbamic acid concerns the common TMTD-ZnO-Sg vulcanizing system. This is a very complex system where both free radical reactions and a polar reaction considered in this section may occur, assuming that sufficient sulfur is available when ZDMDC is formed. Consequently, this would predict an intermediate sulfuration efficiency depending on the concentration of sulfur and ZDMDC when the latter is formed from the reaction of TMTD and ZnO. It is interesting that the Monsanto rheograph of a CB-ZnO-TMTD-S8 system shows two distinct cure rates as a function of time consistent with two mechanisms occurring in this system. Accelerators Based on 2-Mercaptobenzothiazole. Currently, the Raman studies have been extended to accelerator systems based on 2-mercaptobenzothiazole and its derivatives ( 4 1 ) . Although there are some significant differences in the type of products found in thermal decomposition studies, (when compared to the thiuram based systems), specifically, the additional formation of the monosulfide, there is no conflict with the major conclusions outlined in Reaction Scheme 10. Reaction in the Presence of Rubber. A major critism of model compound studies is that the results may bear little relation to the actual reactions occurring in the presence of rubber. However, employing the methods of Parks, et al. (53), where accelerators etc. are imbibed into a previously lightly cross-linked rubber and subsequently raised to vulcanizing temperatures, the presence of persulfurated accelerator species (XS,X) has been detected by Raman spectroscopy ( 4 1 ) . This is good evidence that the mechanism suggested by Reaction Scheme 10 is valid in the presence of the rubber.

Literature Cited Alliger, G . ,Sjothun. I . J . , "Vulcanization of Elastomers." Chapter I , Reinhold, New York, N. Y., 1964. American Chemicai Society, "A Centennial Volume of the Writings of Charles Goodyear and Thomas Hancock," 1939. Armstrong. R . T., Little, J. R . , Doak. K. W.. Ind. €ng. Chem., 36.

Bovey, F. A., "High Resolution NMR of Macromolecuies," Acadernic Press, New York, N . Y.. 1972. Bresler. S. E., Dogadkin. B. A,. Kazbekov, E. N.. Saminskii. E. M., Shershnev, V. A,, Rubber Chem. Technol., 34, 318 (1961). Campbell, D . S.,J. Appl. Polym. Sci.. 13, 1201 (1969). Campbell, D . S..J . Appl. Polym. Sci., 14, 1409 (1970). Campbell, D. S..Saville, 8.. Proceedings of the Institute of Rubber Conference, Brighton. England, May, 1970. Campbell. R. H., Wise, R. W., Rubber Chem. Techno/., 37, 635

(1964). "The Chemistry and Physics of Rubberlike Substances," Chapter

15,L. Bateman, Ed., MacLaren, London, 1963. Coleman, M. M., Masters Thesis, Division of Macromolecular Science, Case Western Reserve University, 1971 Coleman. M. M.. Ph.D. Thesis, Division of Macromolecular Science, Case Western Reserve University, 1973. Coleman, M. M., Shelton, J. R . , Koenig, J. L., Rubber Chem. Techno/., 45, 173 (1972). Coleman. M . M., Shelton. J. R., Koenig, J. L., Rubber Chem. Techno/., 46. 4.938,957 (1973). Coleman, M. M., Shelton, J. R., Koenig, J. L.. J . Poiym. Sci., in press. Coran. A. Y . , RubberChem. Techno/., 37, 679 (1964). Craig, D . , Rubber Chem. Techno/.. 29, 994 (1956). Degtyarev, L. S.,Ganyuk, L. N., Vysokomoi. Soedin.. 1. 31 (1964). Dogadkin, B. A., Shershev, V. A , , Rubber Chem. Techno/., 33, 401

(1960). Duchacek, V., Angew. Makromoi. Chem.. 23, 21 (1972). Dunn, J. R., Scanlan, J., J. Appl. Poiym. Sci., 1, 84 (1959). Fackler, J. P., Jr., Fetchin. J . A,, Fries. D. C . . J . Amer. Chem. SOC..94. 7323 (1972),and-references cited therein. Farmer, E. H.. ShiDlev, F. W.. J. Polym. Sci.. 293 (1946);J. Chem. S O C ,1519 (1947) Gardner D M Fraenkel G K J Amer Chem SOC 78. 3279

(1956). Gregg, E. C.,RubberChem. Techno/., 42. 1136 (1969). Gregg, E. C., Katrenick. S. E., Rubber Chem. Techno/.. 43. 549

(1970). Higgins. G. M. C . , Saville. B., J. Chem. SOC..2812 (1963). Hofmann. W., "Vulcanization and Vulcanizing.~ Agents." MacLaren. London, 1967. Kende, I . , Pickering, T. L.. Tobolsky, A. V.. J. Amer. Chem. Soc.. 87. 5582 11965). Ko'enig, J. L . , Coleman, M. M.. Shelton. J. R., Starmer. P. H.,RubberChem. Techno/.. 44, 71 (1971). Koenig, J. L.,Shelton. J. R., Kapur, R., Rubber Chem. Techno/.. to be published Krebs H Rubber Chem Techno/ 30,962 (1957) Krebs H , Fassbender M , Jorgens F , Chem Ber 90, 425

11957) Manik,

S.

P.. Banerjee, S.. Rubber Chem. Techno/., 43, 1294

(1970). Manik. S. P., Banerjee, S., Rubber Chem. Techno/., 43. 1311

(1970). Milligan, B., J. Chem. Soc.. 34 (1966). Milligan. B., Rubber Chem. Techno/.. 39, 1115 (1966) Moore, C. G . . Porter, M., J. Appl. Polym. Sci., 11. 2227 (1967). Moore, C. G., Trego. B. R., J , Appi. Polym. Sci.. 1 5 , 299 (1961) Moore, C . G., Trego, 6.R.. J . Appi. Polym. Sci.. 8 . 1957 (1964) Moore. C. G., Watson, A. A., J. Appi. Polym. Sci.. 8. 581 (1964) Parks, C. R.. Parker, D . K., Chapman, D . A,, Cox, W. L., Rubber Chem. Techno/.. 43, 572 (1970). Parks, C . R.. Parker, D.K . . Chapman, D . A,, Rubber Chem. Techn o / . , 45, 467 (1972). Pickering. T. L.. Saunders, K. J.. Tobolsky, A. V., J. Amer. Chem. SOC.,89. 2364 (1967). Pickering. T. L., Saunders, K. J . , Tobolsky, A. V., in "The Chemistry of Sulfides," A. V. Tobolsky, Ed., Interscience, New York, N. Y.. 1968. Porter, M . , "The Chemistry of Sulfur Vulcanization of Natural Rubber," in "The Chemistry of Sulfides," A. Tobolsky, Ed., Interscience, New York. N. Y., 1968. Prior, W., "Mechanisms of Sulfur Reactions," McGraw-Hill. 1962. Ross, G . W.. J . Chem. SOC.,2856 (1958) Russell, R . M., Skinner, T. D., Watson, A. A,, Rubber Age, 99 ( l l ) ,

69 (1967). Saville, B., Watson, A. A , , Rubber Chem. Techno/., 40 (1). 100

(1967). Scheele. W., Rubber Chem. Technol., 34, 1306 (1961). Shelton, J. R., McDonel. E. T., Rubber Chem. Techno/.. 33, 342

(1960).

2647 (1972).

Shelton, J. R., Koeniq, J. L., Coleman, M. M., Rubber Chem. Techno/.. 44, 904 (1971). Skinner,T. D., Watson, A.A., RubberAge. 9 9 ( l l ) , 76 (1967). Skinner. T. D . . Watson. A. A,. Rubber Chem. Techno/.. 42. 404

Barton. 8. C . , Hart, E. J..lnd. Eng. Chem.. 44, 2444 (1952) Bateman, L., eta/., J. Chem. SOC.,2838 (1958).

Studebaker, M L , Rubber Chem Techno/ 43. 624 (1970)

628 (1944). Ballard, D. G., Myatt, J . , Richter, J. F. P., J. Appl. Poiym. Sci.. 1 6 ,

(1969)

Ind. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3,1974

165

(67) Tobolsky. A. V., Eisenberg, A , . J. Amer. Chem. Soc., 81. 780 (1959). (68) Trivette, C . D., Coran, A . Y . , J. Org. Chem., 31, 100 (1966). (69) Watson. A . A . , Ph.D. Thesis. University of London, (1965), cited in ref 56. (70) Wolfe, J. R..Rubber Chem. Techno/..41, 1339 (1968).

(71) Wolfe. J. R.. Pugh, T . L., Killian, A . S.,Rubber Chem. Techno/..41, 1329 (1968).

Receiuedfor review April 23, 1974 Accepted May 28,1974

Deactivation of a Platinum Monolithic Carbon Monoxide/Hydrocarbon Oxidation Catalyst David Liederman,* Sterling E. Voitz, and Paul W . Snyder Mobil Research and Development Corporation, Research Department Paulsboro, New Jersey 08066

T h e effects of aging on the CO/hydrocarbon oxidation activity and important physical properties of a platinum monolithic catalyst were determined. Catalyst samples were aged to 25,000 equivalent miles at a maximum temperature of 145OOF with unleaded fuel on an engine dynamometer. The activity for t h e oxidation of CO, C3ti6, and engine exhaust gas hydrocarbons decreased rapidly during t h e first 1000-5000 miles with very little additional deactivation during further aging. A large decrease in C3l-l~ oxidation activity continued throughout t h e 25,000 miles of aging. T h e growth of very large platinum crystallites appeared to b e a primary cause of catalyst deactivation. No significant changes in surface area or solid-state phase composition were noted. Small amounts of lead were detected on aged catalyst samples.

Introduction Catalyst durability is a critical problem for catalytic converters used in the control of CO/hydrocarbon emissions from automotive vehicles. Oxidation catalysts can be deactivated during vehicle operation by exposure to high temperatures which can result from a variety of causes, for example, ignition failure. The activity of these catalysts can also be affected by certain fuel and lubricant constituents such as lead, phosphorus, and sulfur. This paper describes the effects of aging on both the oxidation activity and key physical properties of Engelhard FTX(R)-A catalyst, a commercial platinum monolithic catalyst. Samples of catalyst were aged on an engine dynamometer for as long as 25,000 equivalent miles, at a maximum temperature of about 1450°F. The aged catalysts were characterized by CO and hydrocarbon oxidation activity tests and by physical techniques which included surface area, CO and Hz chemisorption, X-ray line broadening, X-ray diffraction analysis, and scanning electron microscopy. Aykan, et al. (1972, 1973a,b,c), have previously reported some effects of thermal treatments and aging on the oxidation activity of PTX(R)-A catalyst. Thermal treatments at temperatures up to 1800°F significantly reduced the activity for the oxidation of CO and ethylene. In contrast, aging to 1233 miles on an engine dynamometer had a relatively small effect. The activity for propane oxida166

Ind. Eng. Chem., Prod. Res. Develop., Vol. 13, No. 3, 1974

tion was substantially reduced by either thermal treatment or aging. Some effects of both temperature and exposure time on the oxidation activity for CO and ethylene were also determined. It should be noted that a more thermally stable noble metal monolithic oxidation catalyst, PTX(R)-IIB, has been developed (by Engelhard) and will be used on certain 1975 motor vehicles. No PTX(R)-A catalysts will be used for that purpose. Other studies have been made in the past several years on the deactivation of noble metal oxidation catalysts (Briggs and Graham, 1973; Gallopoulos, et al., 1973; Giacomazzi and Homfeld, 1973; Hetrick and Hills, 1973; Hunter, 1972; Mooi, et al., 1973; Neal, et al., 1973; Roth and Gambell, 1973; Shelef, et al., 1973). The activities and physical properties of CO/HC oxidation catalysts have been determined after thermal deactivation in laboratory furnaces, aging on engine dynamometers, and aging on test vehicles. Considerable insight on deactivation mechanisms of thermal deactivation and poisoning by certain constituents of fuels and lubricants has been obtained.

Experimental Procedures The laboratory test used to measure the low-temperature oxidation activity has been previously described (Liederman, et al., 1973; Snyder, et al., 1972). The apparatus, shown in Figure 1, consisted of three main parts: