Reactions of Accelerators during Vulcanization. III-Carbo-Sulfhydryl

Industrial & Engineering Chemistry · Advanced Search .... III-Carbo-Sulfhydryl Accelerators and the Action of Zinc Oxide. C. W. Bedford, L. B. Sebrell...
10 downloads 0 Views 752KB Size
T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

1034

and are unable to tell how to avoid them. Chloroform, which is a good solvent for the tetrabromide, was tried, but although no precipitate came down, emulsions were formed. &Two errors in the original articles are pointed out, relative to the amount of potassium iodide used. 5-The results of several analyses of purified rubber are given, which show differences from 0.15 to 24.27 per c m t . 6-If vulcanized samples are treated with alcoholic sodium

VOl. 13,No. 11

hydroxide in addition to the acetone extraction, it is found that they do not dissolve in the tetrachloroethane within 3 days’ heating. By adding lime to the solvent and rubber they can be dissolved in about 12 hrs. 7-Titration of the hydrogen bromide with alkali has no special advantages over the iodate method described in the original article. 8-The method requires further study and elaboration before it can be used with accuracy.

Reactions of Accelerators during Vulcanization. 111-Carbo-Sulfhydryl Accelerators and the Action of Zinc Oxide’ By C. W. Bedford and L. B. Sebrell GOODYEAR TIRE& RUBBERCO., AKRON,OHIO

There seems to be a. slowly but steadily increasing interest in the chemical reactions which organic accelerators undergo during the vulcanization of rubber. The chemistry of the sulfur reactions of organic nitrogen compounds has previously found its main application in the sulfur dye industry, and even after years of world-wide research it is still a desert, with here and there an oasis of definite knowledge regarding chemical composition or mechanism of reaction. Sldfur dyes are exhaustively sulfurated at high temperatures as compnred with the Use of the Same or similar nitrogen Cornpounds as accelerators in the vulcanization of rubber. Here we find a lower temperature and a lower quantity of eulfur as well as the use of shorter time, so that the chemistry of the first reactions of nitrogen compounds with sulfur may, be applied to accelerators. REACTIONS OB ANILINE WITH SULFUR Merz and WeithZ were the first to study the sulfuration of Pure OrffaniC comPounds. As early as 1869 they heated aniline with sulfur and studied the reaction products. According to L a ~ ~ gthe e , ~first step in this reaction is the formation of a carbo-sulfhydryl group in ortho Position to the amino group, giving o-aminothiophenol as the first intermediate product.

p

+

s

*

@ ;;

MECHAKISM OF VULCANIZATION WITH AMMoNruhf SULFIDE Kratz, Flower and Shapirol have found that) ammonium hydrosulfide (T\TH4SH) will vulcanize rubber tvithout the use of free sulfur. They used a milled stock mrhich, as is WTel known, is honeycombed with air bubh!es, and heated it in a bomb tube containing ammonium hydrosulfide and air. The sulfur available for vulcanization conies from t h e atmospheric oxidation of the sulfhydryl groups to disulfides and the further decomposition of the disulfide3 to monosulfides and free sulfur or polysulfide sulfur, as in the above sulfur reactions of aniline. This is the same reaction which causes a colorless solution of ammonium sulfide to turn yellow or red on standing on the laboratory shelf in contact with air.

+

II

S

SULFURO N WZ-DIAMINES There are many dithio compounds, similar to the above dithioaniline, and containing the grouping E C4-s-c E, which have been described, and we attach great importance to their property of picking up extra sulfur to form polythio compounds. KalleZ found that m-diamines react readily with sulfur, some a t as low a temperature as 80’ C . Schultz and BeyschlagS working with m-tolylenediamine, proved that carbo-sulfhydryl groups are first, formed, which further oxidize to disulfides and that these disulfides have the power of forming polysulfides similar to ammonium polysulfide.

Further reaction with sulfur oxidizes two moles of the mercaptan to one mole of the disulfide, forming o-dithioaniline. At a higher temperature the disulfide loses one sulfur and rhanges to the monoeulfide.

“2s 2

H:c-QNH* HzN

1 Presented before the Rubber Division at the Blst Meeting of the American Chemical Society, Rochester, N.Y., April 26 to 29, 1921. 2 Ber., 2 (l809),341; 8 (;870), 978;4 (18711,384. 8 “Die Schwefelfarbstoffe,” 18.

;;;e /

HpNHrC1C)--S -NHz

The above are typical equations which apply to practically all the reactions of sulfur with the benzene nucleus of accelerators, and while they may not take place with sufficient rapidity a t curing temperatures to be considered as a part of the mechanism of vulcanization by aniline, they are instructive as a basis for comparison with the following sulfur Beactions of other accelerators.

+

+

N H r S H 3.0 HS-NHd NHrSS-NHd HzO NH~S-S-NHI +NHrS-NHd --+ (NH4)9 f S

HZC-@HnN-

So---

-HEN

+H?S

(SX) NHz HzN

These investigators found this disulfide-polysulfide to be soluble in hydrochloric acid without loss of the polysulfide sulfur except after long stranding. They also found this 1 2

*

THISJOURNAL, 18 (1921), 67. D. R. P. 86,096. B e y . , 42, 743, 753.

Nov., 1921

THE JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY

super-sulfur to be available for further sulfuration of amines or for further sulfuration in the nucleus of the disulfide itself. The reactions of amines with sulfur frequently appear to he autocatalytic, the disulfides, once formed, producing polysulfide sulfur which is more active than the original elemental sulfur. I n many cases and for the same reason, we are able to produce greater curing power or entirely different physical properties in a rubber stock by causing the accelerator to react with sulfur before compounding. This is especially true with accelerators which react slowly with sulfur or which lose none of their basic nitrogen during the sulfur reaction. The curing power of m-tolylenediamine is evidently not due to the liberation of ammonia, as will later be shown to be the case with parydiamines. No ammonia is lost on heating this meta-diamine with sulfur up to 160" C., which is a t least 18" C. above the curing temperature of 40 lhs. steam, while the reaction to form disulfide-polysulfides may be completed a t 80" C. At temperatures of 180" to 200" C. ammonia is lost, as well as more hydrogen sulfide The curing power of m-tolylenediamine or its disulfide is therefore attributed to the rapid formation of the disulfide-polysulfide without the loss of ammonia, and to the activity of this polysulfide sulfur as differentiated from any form of elemental sulfur. SULFUR O N P-DIAMINES The meta position of the amino group s e e m to be necessary for the formation of stable disulfides and the retention of all the nitrogen. p-Phenylenediamine, when heated with sulfur a t curing temperatures, liberates ammonia and hydrogen sulfide, forming thionine or Lauth's Vio1et.l I n writing the equations for this reaction we first assume a mechanism similar to that of the meta-diamines.

"7F12

HxN-

0 -NH2

+3s+

HzN -CL~::~Q-NH~

+

HnS

This para-diamine disulfide is evidently unstable a t the temperature necessary for a sulfur reaction with the amine and breaks down by heat into the monosulfide, thereby bringing the two amino groups into such close proximity that one of the nitrogens is liberated as ammonia and further oxidation by sulfur gives the chromophore grouping. H%N-(J

"(H'

Ha-

S

-S -

--3

HIN-

="+

3-sH2N-

"8-k

HnS

This purple dye shows about the same curing power as the corresponding amount of aniline. The curing power of p-phenylenediamine, however, is so much greater than the corresponding amount of aniline or of the purple dye, especially in the curing of hard rubber, that we are forced to consider its action as due chiefly to the liberation of ammonia and hydrogen sulfide with the subsequent formation of ammonium polysulfide.2 Lanae, "Die Schwefelfarbstoffe," 40. If 9-phenylenediamine be heated with sulfur under reflux, the condenser tube frequently clogs up with the alkaline compounds of ammonia and hydrogen sulfide described in Watt's Dictionary, Vol. I, 204. 1

SULFURO N PHENOLATES The sodium phenolates used as accelerators by Porritt' have been considered simply as a means of obtaining a perfect distribution of caustic in a rubber mix. Molau and Seyde2 and Haitinger13 however, have shown that a t 100" to 115" C. there is an easy reaction with sulfur to form thiophenols, disulfides, and polysulfides in the same manner as with aniline or the meta-diarnine~.~

o-oNa+s - o+ @ :: SH

-..----?-O-@(j

+ s + H js-

- - - - - - - I -

@O -O ;:r-

+

6s

-

ONa NaO-

Has

- s-

p

a

N a O a

- s-(Ss)-S

Here we have the complication of two or more possible mechanisms for the formation of polysulfide sulfur. The group 5 C-S-S-C E easily forms polysulfides, the sodium atom may pass from the hydroxy group to the sulfhydryl group which is the stronger acid, and the sodium salt of the mercaptan may then form polysulfides or the sodium may be removed by hydrogen sulfide and form sodium polysulfides. Perhaps future research will decide whether one or all of these reactions function during vulcanization. The presence of alkali materially lowers the temperature necessary for a rapid reaction of sulfur with phenol, as is the case with many of the compounds used in the manufacture of sulfur dyes. The inorganic polysulfides of sodium or potassium furnish active sulfur which will react with organic materials a t temperatures lower than that required by elemental sulfur. Organic polysulfides seem to act in the same manner and the sulfuration of many organic compounds shows evidence of autocatalysis. It is this same principle which is responsible for the polysulfide theory in the vulcanization of rubber. Lange6 quotes Molau and Seyde and states that most of the sulfur dyes are considered as aromatic derivatives of hydrogen polysulfide. Very few of the sulfur dyes, however, are good accelerators after the removal of the alkali. The temperatures used in their manufacture are so high and the amounts of sulfur used are so large that sulfuration is evidently carricd too far or the disulfides are changed to monosulfides, for it is our experience that the best of the mercapto or disulfide acceleratoys lose much of their curing power by overheating. SULFURO N ALDEHYDE AMMONIA There are some investigators who do not look with favor on the theory that most accelerators enter into chemical reaction with sulfur before they activate the balance of the sulfur. Twiss and Braaier,B evidently discussing our British patent,' ktate that "this view may be correct in certain cases, but evidently cannot be accepted generally for all vulcanization catalysts." They refer t o their work showing aldehyde ammonia to function as a powerful accelerator a t as low a temperature as 98" C. and comment that "this observation militates against the belief of some investigators in this field

(Jys~z~O-NH2

a:>o

-NHp

1035

Brit. Patent 129.798 (1919). Chem.-Z'tg., 1907, 937; 2. fihys. Chem., 54, 274. 8 Ai'onatsh, 4, 163. 4 Lange, "Die Schwefellarbstoffe," p. 96, gives several other importad references. 6 Ibid., p. 97. 8 J . SOG. Chem. I n d , 39 (1920). 125t. 1 Brit. Patent 130,857 (1919). 1

T H E JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY

1036

that vulcanization catalysts are not themselves able to expedite vulcanization, but that during early stages of the process they combine with sulfur, giving rise to substances which possess the desired activity.” We fail to follow their logic, eepecially in the case of aldehyde ammonia, which is one of the most striking examples of an accelerator which violently reacts with sulfur. Using temperatures much lower than the 98” C. of Twiss and Brazier, we find that aldehyde ammonia reacts vigorously with sulfur in boiling alcohol with the loss of ammonia and hydrogen sulfide, giving a reddish brown resinous product of comparatively low curing power. By far the largest portion of the curing power of aldehyde ammonia is lost in the two gases which form ammonium po!ysulfide during vulcanization. A control test, made by boiling aldehyde ammonia in alcohol without sulfur, gave only a trace of ammonia and the solution was only slightly colored after 5 hrs.’ boiling, the aldehyde ammonia being recovered almost unchanged. This accelerator affords another very good means of compounding ammonia and hydrogen sulfide, being similar in this respect to p-phenylenediamine. Stevens’ has pointed out that our theory of sulfur reaction of accelerators must necessarily include an efficiency factor depending upon the rate of reaction between accelerator and sulfur. I n the British patent above mentioned it is stated that “the temperatures used in the vulcanization of caoutchouc are only occasionally and by the merest coincidence the temperatures most suitable for a reaction between sulfur and a nitrogen accelerator.” It is certainly true that, we unconsciously choose accelerators on this very factor and that, we discard many accelerators as worthless, such as carbanilide after testing it a t 40 lbs. steam pressure, whereas a t 60 lbs. pressure it is a rapid accelerator. There are still other factors than temperature, such as the action of secondary or inorganic accelerators and the ratios of sulfur and accelerators to each other and to t)he rubber. NITROSOACCELERATORS Stevens2also gives a very pertinent criticism of our theories regarding the action of nitroso accelerators. He states that since p-nitrosophenol is an accelerator and p-aminophenol is not, he cannot support the idea that nitroso accelerators first react with hydrogen sulfide to produce amidoaccelerators. A recent Canadian patent3 has announced the hydrogen sulfide reactiop product of p-nitrosodimethylaniline rts a vulcanization accelerator. A similar reaction takes place between p-nitrosophenol and hydrogen sulfide, but the reduction product is not an accelerator, on account of the acidic action of the hydroxy group. Nitrosophenol, nevertheless, is a strong oxidizing agent and will speed any organic reaction wherein hydrogen sulfide is formed by oxidizing and removing the same. An example of this action is the almost instantaneous formation of thiocarbrtnilide from carbon disulfide and aniline by the addition of nitroso compound^.^ We attribute the curing power of nitroso accelerators first to their oxidizing power, which will greatly speed the reaction of the sulfur with accelerator, rubber resin or rubber protein, thereby bringing the nitrogen in the rubber into quicker action as an accelerator. I n this sense, therefore, nitroso accelerators are, first of all, secondary accelerators acting in the same manner as litharge. Nitrosophenol has this type of accelerating power and the fact that its reduction product is not a primary accelerator has no bearing on its action as a secondary accelerator. 1

I n d i a Rubber J . , 62 (1020),719.

2

LOC.

3

Bedford and Sibley, Can. Patent 207,982 (1921). Bedford and Sihley, Can. Patent 207,953 (1921).

4

Cit.

Vol. 13, No. 11

I n the latter portion of this paper it will be shown that a rubber-sulfur-zinc oxide cement will not air-cure or gel by the addition of p-nitroeophenol and carbon disulfide, but that if the nitrosophenol be first reduced with hydrogen sulfide the cement will gel or air-cure within a few hours. The acidic properties of p-aminophenol evidently prevent its action as a primary accelerator but do not prevent its reaction with carbon disulfide to form dithiocarbamates whose zinc salts will vulcanize a rubber cement a t room temperature. It may also be noted that the corresponding thiourea, p,p-dihydroxydiphenylthiourea,is a fair accelerator for vulcanizing rubber.

EXPERIMENTAL PART The carbo-sulfhydryl accelerators constitute the most important class of vulcanization aids known to-day. Thiocarbanilide is probably used in larger tonnage than the gross weight of s!l other accelerators put together. Other than thiourea derivatives, this class includes the thiurams, dithiocarbamates, mercaptans, mercaptides, disulfides and all accelerators which produce these or similar compounds during the vulcanization process. I n considering the class as a whole, one cannot overlook the great effect that zinc oxide has on the curing power of these accelerators, and i t soon becomes evident that the action of the secondary accelerator must be known before much can be done to explain the action of the primary accelerator. The dithiocarbamates and their ability to cure rubber compounds a t ordinary temperatures, seemed to the authors to offer a suitable basis for research on the action of the metallic oxide and as a result the data in Table I were obtained. The zinc oxide cement used in these experiments was prepared from a milled stock consisting of pale crepe 100.0, ZnO 10.0, and sulfur 6.0. This compound was made into a benzene cement and distributed in wide-mouth glass bottles, so that each bottle contained 118.0 g. cement, or 10.0 g. compound and 108.0 g. benzene. The weights given in Table I indicate the amounts of chemical added to each bottle, I n many cases the weight of accelerator is equal to or greater than the entire weight of rubber per bottle. This was done in order to shorten the time required for the test, as well as to show up weak curing action that might otherwise be overlooked. The values for combined sulfur are calculated as parts per 100 parts rubber (&) and are qualitative only, since the ages of the different samples were not necessarily the same a t the time of analysis. I n several of the bottles the zinc oxide entirely disappeared and the cement took on the appearance of R, pure gum cement with none of the milkiness remaining, and on longer standing there appeared large concretionary crystals containing zinc, which the authors believe to be the zinc salt of the dithiocarbamate. This action is indicated in the righthand column as “Clear” and “Cryst.”l A repetition of all of the above experiments in’ absence of zinc oxide gave negative results in every case with the exception of the three zinc salts, namely: zinc ethyl xanthate, zinc thiophenol, and zinc mercaptobenzothiazol. It is to be noted that two of these compounds are nitrogen-free and without basicity other than that due to the presence of the zinc. The xanthate and thiophenol mere prepared without the use of nitrogen in any form and were washed free from caustic which even if present would not cause the cement to gel. All three of these accelerators, and the mercaptans of the last two, function in regular heat cures. 11x1 many of the tests there also appeared a cloudy precipitate. A similar precipitate forms after a few hours on dissolving zinc thiopbenol and sulfur in cold aniline or at once on heating. The composition of these precipitates have not yet been studied.

T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

Nov., 1921

TABLEI-ZINC Weight Grams 10.0

CHEMICAL Aniline H2S

+

Aniline (2 mol.) CSn (1 mol.)

OXIDECRMENT

Cement Set Time Sc on R No 2 mo. N0 2 mo.

+

Aniline (2 mol.) -I- CSa (2 mol.)

14 days 36 hrs. 24hrs. 20 hrs.

+ csz + csz

#-Phenylenediamine Thiocarbanilide NaOH

-+ + PbO 4- CeHsNHz -tcsz

pNitrosodimet hylAniline

5.4 3.8 11.4

1.12 0.53

14 days 14 days1 8 days 10 days

Yes

20hrs.

1.04

Yes

18 hrs.

1.06

48hrs.

6days

0.88

22hrs.

6days

I

No

2 mo.

No

2 mo.

No No

2 mo. 2 fIl0. 2 mo. 2 mo. 10 dags

No Yes

6.8

No

No No Yes 6.2

+ Has .t CSz

N O

No NO

Yes

Ammonia gas HIS CB HzS CSz Ammonium polysulfide, aq. Sodium sulfide, aq. Caustic soda, aq.

Nor No 2 hTo2

2 mo. 2 mo, 2 mo.

Yes

16 days { 30 days

NO2

Sodium-ethyl xanthate Zinc ethyl xanthate

Yes

24hrs.

Thiophenol Zinc thiophenol

5.5

No Yes

2 mo. 48hrs.

Piperidine H2S thiophenol CSa

3.0

No No No8 Yes

2mo. 2mo. 2mo. 24 hrs. or less

++ +

i.1

Mercaptobenzothiazol

Yes

Zinc mercaptohenzothiazol

Yes

I-Thiorarhanilide ( I mole) Aniline ( I mole OY excess). Dissolved in acetone or benzene-On standing, large lemon-yellow crysatls are formed, often weighing nearly 2 g. Several years ago these crystals were analyzed (unpublished data by pi. Scott) and found to consist of exactly one mole each of aniline and thiocarbanilide. The same crystals are obtained on crystallizing thiocarbanilide from aniline. We have previously mentioned the crystailine product formed by mixing one mole each of the two liquids, piperidine and thiophenol, which we believe to be salt of an organic acid and base and to contain pentavalent nitrogen as in ammonium sulfide. In the aniline-thiocarbanilide crystals here described, we believe the thiocarbanilide to be in the mercapto form and the crystals to be the aniline salt of the mercaptan. The crystals are stable only in contact with the mother liquors, and easily decrepitate into aniline and flat crystals of thiocarbanilide. The best proof of the mercapto form of thiocarbanilide in contact with aniline is its extraordinary reactivity in dissolvingzinc oxide, which power it does not have in absence of a basic amine. 2-Thiocnrbanilide and Z i n c Oxide in Benzene-No zinc oxide was dissolved, as indicated by filtering, evaporating, and ashing.

2 mo. 2 mo. 2 mo. G days

NO?

No2

6 d a y s None

2 mo. Cement becomes thinly liquid 2 mo. 2 mo. 5 days

2 mo. 2 mo. 2 mo. 2 mo.

No2

++ +

3 - mo

3days

4.7 4.7

+ + HZS HIS -t CSn

++ HzS csz

8 days 14 hrs. 10 hrs.

NO

-?- CSa

p-Nitrosophenol

13 days 13 days 2days 8days 20hrs. Bdays

Yes

light in a dark room and found to be as free from colloidal particles RS could br: expected in solutions prepared in the open laboratory. These tests were confined to the thioureas.

+

1. O B 0.66 1.04

1.o Yes 2.5 Yes 5.0 Yes 8.5(1/~omol.)Yes

Dimethylaniline

-RemarksClear Cryst.

1037

3-Aniline

and Z i n c Oxide in Benzene-No

zinc soluble.

4--Thiocarbanilide, A nilzne, and Z i n c Oxide in Benzene-On standing a t room temperature with frequent shaking for 18 to 24 hrs., large amounts of zinc are found in the clear filtered solution. 5-The above tests were repeated with zinc hydrate with the same results. The amount of zinc dissolved in Test 4 was higher than for the oxide.

6-In a few cases the residual zinc oxide, as filtered off and washed, was found to contain small amounts of sulfide, but usually there was no test for hydrogen sulfide on treating with acid. Thin -el Thickhgel

Zn salt does not form

7-To date, the organic zinc compound which is soluble in benzene, acetone, or aniline has not been isolated or its properties determined. Assuming that it is the zinc salt of the carbosulfhydryl form of thiocarbanilide (R-S-Zn-S-R) the following quantitative data were calculated to the percentage of the thiourea transformed to its zinc salt. TABLB I1

7 days Zn salt forms slowly or less 48hrs.

1 Large crystals dispersed in perfectly clear cement. 2 Alkali. whenever used. causes a thickeninn of the cement.

Total organic

TIME Hrs.

due to the &ect o f t h e alkali on the iubber. This is no&eable a t once for NaOH, and after a few days for NHaOH. The cement thickened a t once owing to the mass of crystals of the piperidine-thiophenol salt, but quickly thinned out with more benzene

16 40 84

ZINC MERCAPTIDES-In a number of cases, such as with the toluidines, the acetone extract of the solvent-free, aircured vulcanizate was found to contain large amounts of zinc in solution as shown by zinc oxide in the ash. This fact, together with the disappearance of the zinc oxide in the cements, indicated soluble zinc compound, probably the zinc salt of the dithiocarbamate. On the same basis, the action of thiocarbanilide of air curing in the presence of aniline and zinc oxide was supposed to function through the zinc salt of the tautomeric or mercapto form of the thiourea. The following tests tend to corroborate this assumption. Several of the cement tests were repeated, leaving out the rubber and sulfur. The soluble zinc was decermined by filtering, evaporating the solvent, and ashing the residue. The solutions,yafter filtering, were exposed to a beam of

16 40 84

*

Tfmp. C. Room Room Room

315.8

Thiourea Transformed to Zinc Salt Total Zn in S o h . (asZnO) Weight Gram Gram Per cent 0.04 0.09 0.16

0.225 0.507 0.903

0.99 2.22 3.95

Di-o-tolylthiourea 25.6 g. o-Toluidine 107.0 g . Benzene 200.0 g.

45min. 105min. 105min. 24

Room Room Room

0.075 0.160 0.211

1.85 3.95 5.23

1.340

(Same as above without the benzene) 100 0.242 1.510 100 0.324 2.045 100 Room] 0.383 2.425

Di-o-tolylthiourea 64.0 g. 16 64

0.474 1.013

ZnO 10.0 g.

100 100

o-Toluidine 321.0 g .

0.723 1.391

4.70 9.05

5.9 8.0 9.5

ZnO 10.0

g.

7.3 14.2

We find, therefore, .an exact parallel between the solubility of zinc oxide in a benzene solution of thiocarbanilide and the air-curing of a similar rubber-sulfur-zinc oxide cement. Thiocarbanilide or aniline alone will neither vulcanize the cement nor dissolve the zinc oxide, while the combination of the two causes both reactions to take place.

1038

T H E JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY

The action of zinc oxide as an activator for thiocarbanilide and a retarder for aniline have been the subject of much recent discussion. It now appears that the first mechanism of its action with carbo-sulfhydryl accelerators is to form a zinc salt or mercaptide of the general formula R-S-ZnS-R, which in some manner is able to activate the sulfur and hand it on to the rubber. By preparing these mercaptides before compounding the zinc may be used in much lower quantities and pure gum cements may be made to cure a t ordinary temperatures.

8UMMARY 1-The vulcanization of rubber by ammonium hydrosulfide has been explained by oxidation and the liberation of free sulfur from the disulfide, The loss of S, from polysulfides has previously been proposed as the mechrtnism for vulcanization. 2-Meta-diamines are differentiated from para-diamines by their sulfur reactions. m-Tolylene-diamine forms stable disulfide-polysulfides to which its curing power is attributed. &-Sodium phenolates form disulfide-polysulfides similar to the meta-diamines and aniline. $--Aldehyde ammonia is very rapid in its reaction with



Vol. 13, No. 11

sulfur a t or below curing temperatures, ”and forms ammonium polysulfide during vulcanization. 5-p-Nitrosophenol is believed to function only FLE a secondary accelerator, acting in this manner similarly to litharge. 6-Zinc oxide or zinc mercaptides have been found neceseary in all rubber cements which cure a t room temperature. 7--Weak bases, such as aniline, will vulcanize a zinc oxidecarbon disulfide cement a t room temperature, just as piperidine or dimethylamine. 8-In cements containing amine and carbon disulfide, zinc oxide is dissolved and the cement may take on the appearance of a pure gum cement. 9-A mixture of aniline and thiocarbanilide will dissolve zinc oxide a t ordinary temperatures. 10-A mixture of aniline and thiocarbanilide will vulcanize a cement containing zinc oxide a t room temperature, while either alone will not. 11-Zinc thiophenol and zinc ethylxanthate are given as two accelerators which are free from nitrogen or alkali and which function either in heat cures or in curing pure gum cements a t room temperature. Zinc mercapto-bensothiazole acts similarly. 12-The ultimate mechanism of vulcanization by mercaptides and sulfur has not been discussed.

The Preparation and Instability of Tricalcium Arsenate’ By J. H.Reedy and I. L. Haag DEPARTMENT on CHFMISTRY, UNIVERSITY OA ILLINOIS, URBANA, ILLINOIS

Of late attention has been sharply directed to the instability of tricalcium arsenate, which is extensively used as an insecticide for certain crops. During storage this product undergoes some change in composition which results in a considerable increase in the amount of water-soluble arsenate. The latter substance is considered responsible for the “burning” of the foliage of the plants to which it is applied. At the time of preparation, according to the manufacturer’s analysis, the soluble arsenate content is well within the limit (0.75 per cent AszOs), but upon reaching the consumer the value may he considerably in excess of this. Contact with moisture and carbon dioxide of the air was thought a t first to be the cause of this deterioration, but the use of airtight containers has not been sufficient to prevent it. Furthermore, manufacturers have not been able to control their processes so as to give products of uniform composition. The material from one run may be thoroughly satisfactory, and that from the next may have to he rejected on account of excess of water-soluble arsenate.

EXPERIMENTAL PREPARATION OF SALTS FOR comRoLs-Three well-defined arsenattes of calcium2 are known: the normal or tricalcium arsenate, Ca3(As0&; the secondary or calcium hydrogen arsenate, CaHAsOc; and the primary or monocalcium arsenate, Ca(HzAsO&. The last of these is decidedly unstable, and its presence in the commercial product is unlikely. Samples of the normal and secondary s d t R were prepared by the methods of Robinson.8 For the former, calcium chloride and disodium arsenate were allowed to react in slightly acid solution. The latter was made by the action of arsenic acid solution on an excess of saturated calcium hydroxide solution. The precipitates were washed, dried, and analyzed, indicating the respective formulas, CaHAsOJhO and Ca3(AsO4)2.31120. Robinson determined the solubility of these salts in 100 Received June 13, 1921. Abegg, “Handbuch der Anorganischen Chemie,” 111 (1907), 641. a Oregon Agricultural Experiment Station, Bulletin 131 (1918).

g. of water at 25” as: CaHAs04, 0.3308 g.; Ca3(As04)2,0.014 g. These pure salts were used as controh in the study of the products obtained by methods available for large-scale production. INDUSTRIAL PREPARATION METHODS-while Several methods suggest themselves for the manufacture of tricalcium arsenate, only one has been found in practice to give a satisfactory product a t a sufficiently low cost. This involves the action of arsenic acid on a paste of slaked lime?

3 Ca(OH)2 f 2 HsAs04

+Cas(AsO4)a f

3 H20 (1) In the present work, a high-grade lime was slaked with three or four times its weight of boiling water, and a solution of arsenic acid was slowly added, with constant stirring, until the mixture was only faintly alkaline to phenolphthalein. This process involves approximately equal parts of lime and arsenic acid (estimated as As206),in the molecular proportion of 4 CaO: As206. The final product was assumed to be the tricalcium arsenate with an excess of calcium hydroxide. The reaction between calcium chloride and trisodium arsenate gave a product that was unsatisfactory on account of the high water-soluble arsenate content. A batch of the material was made as follows: A solution of disodium arsenate was converted into the trisodium salt by adding a molecular amount of sodium hydroxide, and this wm added to an excess of calcium chloride solution;

The by-product of Reaction 1 is water only, while the mother liquor of Reaction 2 contains sodium chloride and the excess of calcium chloride. These were removed from the precipitate by washing. However, the washings contained soluble arsenates, and the residue still remained high in soluble arsenates. The part played by these soluble salts in increasing the soluble arsenate content will be discxlssed later.

1 I)

I

Cf. Haywood and Smith, U S. Department of Agricultrre. B&&n

760 (1918).