Colloidal Changes during Rubber Vulcanization - Industrial

Note: In lieu of an abstract, this is the article's first page. Click to increase image size Free first page. View: PDF | PDF w/ Links. Related Conten...
1 downloads 0 Views 581KB Size
1190

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 26, No. 11

ments and rubber is the key to the problem. Rather, vis- such as fatty acids, and various surface treatments upon the cosity considerations, such as that advanced by Park (8), wetting of pigment materials by rubber. deserve attention. He postulated that reenforcement may LITERATURE CITED be related to the flow of rubber through the capillary spaces (1) Am. SOC.Testing Materials, Triennial Standards E 11-26, Pt. between pigment particles that are surrounded by oriented 11, p. 1244 (1933). rubber. Another explanation that should be kept in mind (2) Bartell, F. E., and Osterhof, H. J., J. Phys. Chem., 37, 5 4 3 is that the pigment particles and the rubber oriented around (1933). them may act as obstacles (4)in the path of a tear and thereby (3) Bartell, F. E., and Walton, C. W., Jr., Ibid., 38,503 (1934) (4) Deoew. H. A , . Rubber Aae (N. Y.), 24. 378 (1929). make tearing more difficult. i5j Enhres; H. A.; IND. ENQ; cam.,is, 1 1 4 s ( i 9 2 4 ) . Irrespective of the explanation, these data show that with (6) Green, H., Ibid., 13, 1029 (1921). relatively coarse pigment materials the rubber will pull away (7) Hock, L., and Schmidt, H., Rubber Chem. Tech., 7, 462 (1934) ; with a force of adhesion that depends on the nature of the pigtr. from Kautschuk, 10,33 (1934). (8) Park, C. R., IND.ENG.CHEM.,News Ed., 11, 343 (1933). ment material. Additional tests and correlation with the ENQ.CHEM.,12, 3 3 (1920). (9) Schippel, H. F., IND. work of Bartell and his co-workers may make it possible to (10) Steele, F. A . , New Jersey Zinc Go., private communication. express the force of adhesion in standard units. This paper ~ C ~ I VSeptember ED 15, 1934. Presented before the Division of Rubber is only an introduction to this subject and i t is intended to R Chemistry at the 88th Meeting of the American Chemical Society, Cleveland. use the described method to determine the effect of agents, Ohio, September 10 to 14, 1934.

Colloidal Changes during Rubber Vulcanization IRAWILLIAMS,E. I. du Pont de Nemours & Company, Wilrnington, Del.

R

I t i s t h e P u r p o s e of t h e UBBER is colloidal in Rubber which has been lightly vulcanized with n a t u r e a n d colloidal sulfur can be di,ysolved in benzene with the help of P r e s e n t P a p e r to show that rubber which has combined with phenomena s h o u l d b e peptizing agents. The I ’ d b e r may be sulfur will disaggregate under expected to play a part in the the influence of accelerators and Peptized the adion Of accelerators in the absence decrease of plasticity which is recognized as v u l c a n i z a t i o n . of solvent. These materials are eficient peptizto present data which indicate ing agents only after rubber is attacked chemically. that this disaggregated product The elasticity of vulcanized and The action of solub[e zinc c o m p o u n ~on pepcan gel e i t h e r because of its unvulcanized r u b b e r i s s u b concentration or under the instantially the same (W), and i t tized rubber sulfur causes a gelling action with fluence of gelling agents to prois p r o b a b l e t h a t elasticity is a function of the molecule while great increase in tensile strength and modulus. duce vulcanization. plasticity is a function of the PEPTIZATION OF LIGHTLY VULCANIZED RUBBER state of aggregation. Many authors have suggested that either disaggregation or depolymerization is one step in vu1Vulcanization may be carried out in solution to produce a canization (1, 8, 16, 16, 19). Other authors believe that ag- gel which can be readily peptized. Cements containing sulgregation or polymerization is a factor (2, 4, 6, 9, 10, 13, 14, fur, zinc oxide, and piperidinium pentamethylene dithio18, 19, 21, 22). I n many cases it is probable that the term carbamate will gel in a comparatively short time a t room “polymerization” or “depolymerization” is used loosely and temperature. If the resulting gel is treated with more of the is not intended to involve a change in unsaturation. Ostro- accelerator, the gel will be reduced to a mobile liquid. I n mislensky ( l a ) ,Stevens (17), and Twiss (20) have suggested certain instances the resulting liquid will again gel and may that vulcanized rubber contains the reaction product of rub- again be peptized by the addition of more accelerator. This ber and sulfur dispersed through the rubber as h e particles process, however, cannot be repeated indefinitely since a gel or threads. Feuchter (5) suggests a similar although some- will h a l l y result which can no longer be peptized. The inwhat more complicated system which includes accelerator. ability to peptize a mlcanizate which is in a more advanced Whitby ($1) concludes from his experiments with drying oils state of gelation indicates that the peptizing action of the that accelerators act on rubber containing combined sulfur added accelerator was not due only to LL reversal of the preto produce gelation or polymerization. Stevens and Stevens vious gelling process. (18) have milled zinc oxide and accelerator into acetoneThe gelling and peptizing action are illustrated in the folextracted vulcanized rubber and find that heating produces lowing experiments: further vulcanization. The base stock consisted of smoked sheet rubber containing Vulcanization may result primarily from colloidal phe- 2 per cent of sulfur and 2 per cent of zinc oxide. Seventy grams nomena, but there is no doubt that combination of sulfur is of this stock were made into a cement with 930 grams of benzene. Part of this cement was treated with 1 per cent of piperidinium essential in hot vulcanization and the presence of an accelera- pentamethylene dithiocarbamate accelerator based on the tor is beneficial. While combined sulfur is necessary, no re- weight of the rubber, and 10-cc. portions were placed in test lationship has been found between the extent of combination tubes. The time of gelation at room temperature varied among of sulfur and the physical properties of the rubber when vul- the different tubes from 9 t o 11 days. As soon as the cement canized under different conditions. The great variation in in one of the tubes had gelled, it was treated with 0.2 gram of the accelerator on the surface of the gel. After standing for physical properties obtainable with the same amount of com- 5 hours, the gel had been reduced to a cement more mobile than bined sulfur indicates that the chemical combination of sulfur the original. Similar experiments were conducted with the other tubes at increasing periods of time after gelation. After is not the primary cause of vulcanization.

November, 1934

INDUSTRIAL AND ENGINEERING CHEMISTRY

the cement had been gelled for about 5 days, it was necessary to increase the accelerator to 0.3 gram and insert it into the gel with a glass rod t o produce peptization. After 15 days the gel softened but did not produce a smooth cement. After 2 months the gel could not be peptized. The peptizing action can also be observed by following the time required for gelation in the presence of increasing amounts of the peptizing agent. Portions of the above cement were used, and increasing amounts of piperidinium pentamethylene dithiocarbamate were added. The cement was considered to be gelled when the test tube could be inverted without the cement's flowing. The time required for gelation increased with the amount of peptizing agent as follows: ACCEL~RATOR BASED ON RUBBER

% 0.5 1.0 2.0 3.0

GELATION TIM^ Days 7

%

Days

4.0

28 50 55

15

21 23

This phenomenon may be extended under proper conditions to cements gelled with other accelerators or with sulfur monochloride and to peptizing agents other than those used to produce the original gel. Certain accelerators appear to act as peptizing agents only a t high temperatures, and other materials which act as accelerators of vulcanization appear to have little action as peptizing agents. No peptizing agents have been found which are not accelerators of vulcanization. To illustrate the action of other materials, a base stock consisting of smoked sheet containing 1 per cent of sulfur and 5 per cent of zinc oxide was made up to a 9 per cent cement with xylene. Various conditions of gelation and aftertreatment were employed with the results shown in Table I.

TABLEI. EFFECTOF PEPTIZING) AGENTSON CEMENTS GELLEDBY DIFFERENTMETHODS MATERIAL PRESENT DCRINQ TEMP.OF PEPTIZINQPEPTIEING GELATION GELATION AQENT TEMP. RESULTS 0 c. 0 c. -. a-e th yl- 8100 Piperidinium penta- 30 Thin cement in 4 2%ropylacroleinmethylenel dihr. aniline reaction thiocarbamate product a-Ethyl+prq yl30 Not peptized in aor olein-am{ ne 120 hr. reaction product SO Thin in 2 hr.

1% S&lr added as 10% benzene

30

soh.

+

HIS SO2 each added i n benzene soln.

30

Diphenylguanidine

30

Piperidinium pentamethylene dithiocarbamate None Diethyl ammonium diethyl dithiocarbamate

30

TABLE

30 30

11.

Not peptized in 120 hr. Extremely thin in 18 hr. Thin in 168 hr. Thin in 6 hr.

After 7 days

After 15 days

A 100

ComDound Smoked sheets Sulf UT Zinc oxide Vulcanization at 140° C . , min.

B 100 4 5 60

6

0 75

The results of the action of various peptizing agents are shown in Table 11; these results may be influenced by factors other than the peptizing action of the added material. Certain of the materials are only slightly soluble in benzene. In none of the cases is the partition of the material between benzene and rubber known. Solubility phenomena may explain why benzothiazyl disulfide affects the surface without causing a great deal of swelling of the rubber. The composition of acidic material or of amines can be changed by combination with zinc, and the dithiocarbamates will be changed to amine salts in the presence of sufficient acid. I n general, the strongest accelerators appear to be the strongest peptizing agents, and in many cases the action is sufficiently great to change the rubber from an insoluble condition into a condition which produces a thin solution in a relatively short time. The difference in speed a t which compounds A and B were peptized was due to differences in the physical state of cure. Many of these accelerators have an influence on rubber containing no combined sulfur. The effects are not great enough to be easily observed in the rate of solution of unmilled smoked sheet but can be detected by measuring the viscosity of cements. A benzene cement containing 10 per cent of smoked sheet rubber was divided, and portions were treated with the accelerators used in the previous experiments. Three per cent of material based on the rubber content of the cement was used. The relative viscosity was determined after 24 hours by observing the time required for a '/s-inch (0.318-cm.) steel ball to fall through a IO-cm. column of cements.

PEPTIZIPiG ACTION O F

COMPOUND A-

PEPTIZING AGENT

Rubber may be peptized after being gelled under the ordinary conditions of vulcanization. I n this case successful peptizing depends on treating the rubber during the early stage of vulcanization. I n conducting these experiments, the compounds were vulcanized just sufficiently to permit the removal of the vulcanized slab from the mold. Strips approximately 1.5 X 4 X 15 mm. were cut from the slab and placed in a test tube containing 15 cc. of a 3 per cent benzene solution of the peptizing agent. The extent of the action of the peptizing agent can be followed roughly before solution of the rubber takes place by the degree of swelling and the disappearance of the sharp edges of the strip. Two stocks which were used in this type of experiment are as follows:

ACCEJLERATOR BAfiED GELATION ON RUBBER TIM= 6.0 10.0

1191

MATERIALS ON COMPOUNDS

.-

Approx. time for complete soln.

A

O

M

P

O

U

N

D

B

After 9 days

After 2 days

Approx. time for complete soh.

Days

None (control)

As first obsvd.

Insol.

As first obsvd.

Like control Considerably softer than control Softer and more swollen than control

Insol. Over 40

As first obsvd. Less firm than control

As first obsvd.

n-Butanol Diphenylguanidine

Swollen but firm and edges sharp Like control Like control

Mercaptobenzothiazole

Like control

Benzothiazyl disulfide

Appears to be disaolving from surface; corners rounded Corners becoming rounded A soft jelly

More than half dissolved: remainder a firm jelly

22

Much swollen to a soft jelly In soln.

28

Much softer than control

Practically all dissolved

15

Dissolved

Dissolved

1

Almost jn s o h . Almost in soln.

In s o h . In s o h .

13 13

Dissolved Almost all into soln.

Dissolved Dissolved

2 3

Dissolved

Dissolved

Tetramethylthiuram monosulfide Piperidine salt of mercaptobenzothiazole Piperidine Butyraldehyde-butylamine reaction product Piperidinium pentamethylene dithiocarbamate Lead oleate

Over 40

Almost in soln.

In s o h

12

Swollen more than control

Almost dissolved; small amount of gel remainIng

18

As first obsvd. Much swollen and very soft; edges indistinct Swollen somewhat more uite soft and transthan control ZnO dieparent but edges still solving from'the rubber distinct Surface appears to be die- Entirely dissolved solving

....

'

....

Insol. Insol. 20

35 9 10

.. ..

1192

INDUSTRIAL AND ENGINEERING CHEMISTRY

The results of this series of experiments, given in Table

111, show that all of the materials have a measurable effect on the viscosity. The magnitude of the action is not as great as might be indicated by the effects produced on lightly vulcanized rubber, and the order of activity is somewhat different. TABLE111. EFFECTOF VARIOUSMATERIALS ON VISCOSITY OF RUBBERCEMEXTS RELATIVE VISCOSITY

MATERIAL

See.

Benaothiaayl disulfide Control, no addition Mercaptobenzothiazole n-Butanol Tetramethylthiouram monosulfide Diphenylguanidine Piperidine salt of mercaptobenaothiaaole Piperidinium pentamethylene dlthiocarbamate Butyraldehyde-butylamine reaction product Piperidine

17.6 17.1 14.4 13.3 13.0 10.5 10.3 10.1 8.8 8.7

The extent to which lightly vulcanized rubber will be peptized depends not only on combined sulfur and on the efficiency of the peptizing agent but also to a very great extent upon the physical properties of the vulcanizate. Rubber vulcanized with an efficient accelerator to produce high tensile strength with low combined sulfur cannot be peptized. On the other hand, a n unaccelerated rubber-sulfur mixture of low tensile strength may contain considerable combined sulfur and be easily brought into solution. The peptizing action can be observed in the absence of solvents. Fifty grams of compound A after vulcanization were acetone-extracted and placed on the rubber mill. The compound was sufficiently vulcanized so that a continuous sheet could not be obtained, and the rubber assumed the form of small pieces of crepe. This product was treated with 2 grams of piperidinium pentamethylene dithiocarbamate, and the milling was continued. The rubber almost immediately began to adhere and in a few minutes had assumed the form of a smooth sheet resembling unvulcanized rubber except that the sheet had considerable tack and tended to cling to the mill rolls. The same peptizing agent added to unvulcanized rubber on the mill produces no noticeable effect.

strips were cut from each slab and acetone-extracted for 10 days, the position of the strips being frequently changed to insure complete extraction. After removal from the extractor the strips were dipped in a 5 per cent acetone solution of phenyl-P-naphthylamine and were then dried in a vacuum desiccator over calcium chloride. Each strip was then treated by rubbing the various materials on the surface until the approximate required weight was present. The strips were then closely wrapped in tinfoil for heat treatment. Since the acetone extraction removed the resins and fatty acids, artificial resins were prepared for treating the rubber. Resin of the following different compositions was used for each compound: Zinc propionate Zinc caproate Stearic acid Oleic acid Rosin oil Pine tar Cottonseed oil Phenyl-@-naphthylamine

Compound Pale crepe rubber Smoked sheet rubber Sulfur

C 100

... 6

D

...

100 6

Slabs of compound C were vulcanized for 90 and 120 minutes and of D for 60 and 75 minutes a t 145" C. The slabs of rubber were approximately 3 mm. thick. Dumb-bell test

USEDWITH COMPOUND C 50

USEDWITH COMPOUND D

..

.. ..

50

..

12

10 40

7 12 7 12

..

... .

The various strips were treated in the following manner: 1. Extracted sample not further treated. 2. Four per cent resin; held 24 hours at 70" C. 3. One per cent piperidinium pentamethylene dithiocarbamate and 4 per cent resin; held 24 hours at 70" C. 4. One per cent piperidinium pentamethylene dithiocarbamate; held 24 hours at 70" C. 5. Four per cent zinc propionate; held 24 hours at 70" C. 6. Placed in refluxing saturated acetone solution of zinc chloride for 6 hours and dried under reduced pressure.

Combined sulfur was determined on the 75-minute cure of compound D after the various treatments and the physica1 properties were determined in all cases. The results are shown in Table IV. TABLEIV. PHYSICAL PROPERTIES OF EXTENDED RUBBER AFTER TREATMENT TREATMENT

GELLIXGOF PEPTIZED RUBBERSULFUR Rubber containing combined sulfur after being peptized with accelerator is in a condition to gel to an insoluble form. Smoked sheet rubber containing 10 per cent of its weight of sulfurwas vulcanizedfor 60 minutes at 140" C., creped on a rubber mill, and acetone-extracted for 9 days. The combined sulfur was found to be 1.02 per cent. After extraction the material was treated on the mill with 1 per cent of phenyl-@naphthylamine to prevent oxidation and 2 per cent of piperidinium pentamethylene dithiocarbamate. The resulting material resembled unvulcanized rubber and dissolved readily in benzene. After standing a t room temperature for 25 days the material would no longer dissolve in benzene without the addition of more accelerator. After 3 months it was no longer possible to get the rubber into solution. The combined sulfur a t this time was found t o be 1.05 per cent, which checks the original within the error of measurement. The extent of the gelation can be easily shown by its effect on the physical properties of the rubber. Two lightly vulcanized, pure gum rubber compounds of the following composition were used for this purpose:

Vol. 26, No. 11

LOADAT ELONGATION OF: 600% 700% 800% -Grams

1 2 3 4 5

...

42 122 '

53

TENSILE STRENGTHELONQA-

COMBINED AT TION BREAK A T B R E A K SULFUR

per sq. mm.-

%

%

1000

960 900 980 900

... ... ... ... ...

900 900 880

... ...

COMPOUND C , 90-MINCTh C U R E

53 77 245 43 94

87 122 540 77 154

420 335 940 245 290

COMPOUND C , 120-MINUTE C U R E

1 2 3

38 102 210

1 2 3 4

24 10 92 52 74

91 160 330

151 273 595

238 470 840

COMPOUND D, &MINUTE

6

38 22 171 94 123

68 35 272 171 190

CURE

273 215 735 660

390

1120 1140

1000

1150 960

... ...

... ... ... ...

COMPOCND D, 75-MINUTE CURE

Considerable change in the physical character of the rubber has been produced by the various treatments. Resin C in each case when used alone had a slight stiffening effect while resin D had a definite softening action. Treatments 5 and 6, which consisted in the action of zinc salts but omitted the softening agents contained in the resins, in each case produced a definite increase in modulus. Treatment with accelerator alone has produced a small but variable effect. Treatment with accelerator and resins containing zinc has produced a marked increase in stiffness of the rubber. No appreciable change in the amount of combined sulfur has resulted.

Kovember, 1934

INDUSTRIAL AND ENGINEERING CHEMISTRY

The gelling action in the presence of accelerators and zinc is not always so easily shown. It can be demonstrated with many of the more rapid accelerators, such as zinc dimethyl dithiocarbamate, cadmium pentamethylene dithiocarbamate, and the piperidine salt of mercaptobenzothiazole. On the other hand, such accelerators as butylamine-butyraldehyde condensation products or diphenylguanidine will usually have a softening effect. I n such cases the gelling action can be demonstrated by placing small pieces of the treated rubber in benzene solutions of piperidinium pentamethylene dithiocarbamate and comparing the rate of solution with that of the untreated rubber. The rubber will be found either not to disperse or to disperse much more slowly than before the treatment. The gelling action of soluble zinc compounds has been shown by their retarding action on the rate of solution of lightly vulcanized rubber. The solvent used was a 4 per cent benzene solution of the various peptizing agents, or, if less soluble, a saturated solution was used. Two test tubes were half-filled with each solution, and to one of the tubes 2 per cent of zinc propionate was added. Strips of compound A, 1.5 X 4 X 20 mm., were added to each test tube and the swelling was observed. The observations after 24 hours are recorded in Table V. TABLE V.

RETARDING ACTIONOF SOLUBLE ZINC SALTS ON SWELLING OF RUBBER

PRESENCE OF ZINC ABSENCEOF ZINC Increase Increase In in PEPTIZING AGENT Condition length Condition length % % Butyraldehyde-butylamine Practically all ,. Edges sharp 200 dissolved and firm Piperidine Soft jelly edges . . Edges sharp 200 indistinct and firm Pi eridine salt of mercapto- Firm and edges Eeneothiazole sharp 175 Firm 130 Tetramethylthiuram mono108 108 Firm sulfide Firm 115 135 Firm Mercaptobenzothiazole Firm 100 115 Firm Diphenylguanidine Firm ,, Appenre to dis.. Lead oleate In xylene a t Dissolved solve from the 80‘ C. surface; halfdissolved remainder a soft gel

.

.

.

.

STEPS INVOLVED IN ~TULCAKIZATIOS

The experiments which have been described lead to the conclusion that vulcanization is the result of several actions which take place to a varying extent under different conditions. The properties of the resulting vulcanizate will vary accordingly.

.1

R+s=Rs

accelerator+- XR’

c

I

+ R“s

I

Zinc j. XR‘n n

FIGURE1 Some of the essential reactions which appear t o be involved in vulcanization are illustrated diagrammatically in Figure 1 in which R, R’, and R” are different units of rubber and s is sulfur or other substance which is capable of combining with rubber. The first step consists in chemical action of the sulfur and rubber to form the unit Rs. The speed of this action can be influenced by materials such as accelerators which modify the size and activity of R and by the activity of s. The second step consists in the breaking up of chemical product Rs into X smaller units, R‘, m-hich are more

1193

active chemically and colloidally, and a smaller chemical unit, R”s. The value of X and the size of R” is largely a function of the degree of peptizing action of the accelerator. The third step consists in the combination of units R’into a larger unit less active chemically and more stable in the presence of peptizing agents. This action is increased in the presence of soluble zinc. If sufficient rubber is changed to this type of gel the solubility and plasticity of the entire mass is greatly decreased. Various accompanying reactions can be written with most parts of Figure 1 as starting points. For example, some accelerators react directly on unit R t o produce new units. Unit R” can be further reduced by more efficient accelerators or can combine with sulfur to approach ebonite in composition. The unit (X/n)Rn may also combine with sulfur but at a reduced rate. The presence of zinc should not favor the combination of sulfur because of the decrease in concentration of R’. This agrees with the known fact that zinc oxide does not favor the production of ebonite. Beadle and Stevens (4) have shown that, while zinc oxide greatly increases the tensile strength of a rubber-sulfur compound, the rate of combination of sulfur is not increased. Kelly (7) has shown that, after correction for sulfur combined with zinc and other nonrubber constituents, the amount of sulfur combined with rubber containing only natural accelerators is decreased by the presence of zinc oxide. The course of vulcanization is undoubtedly altered in many cases by change in the chemical nature of the accelerator during the process and by activation of sulfur by certain accelerators. Sulfur may be replaced by other agents, in which case different peptizing agents may be necessary or the rubber may disaggregate without peptizing agents. Other gelling agents than zinc are conceivable. Vulcanization appears to consist of a chemical reaction that is accompanied by changes of a colloidal nature. The physical properties of the resulting vulcanized rubber do not depend on the amount of chemical action but on the degree to which the units Rs have been resolved into units XR’ and in the degree of combination of these units.

LITERATURE CITED Axelrod, Gummi-Ztg., 24, 352 (1909). B a r y a n d Fleurent, Compt. rend., 184,947 (1927). B a r y a n d Hauser, Kautschuk, 4, 96 (1928). Beadle and Stevens, J . Soc. Chem. Ind., 30, 1421 (1911). Feuchter, Kolloidchem. Beihefte, 20, 136 (1924). Hauser, Inst. Rubber Ind. Trans , 2, 243 (1926). Kelly, W.J., ISD. ENG.CHEM.,14, 197 (1922). Kindscher in Memmler’s H a n d b u c h der Kautschukwissenschaft, pp. 329-30, S. Hirzel, Leipeig, 1930. Kirchhof, Kautschuk, 3 , 29 (1927). Kirchhof, Kolloid-Z., 1 4 , 4 3 (1914). Ibid., 26, 168 (1920). Ostromislensky, India Rubber J . , 52, 467 (1916). Porritt, Ihid., 60, 1161 (1920). Pummerer, Kautschuk, 3, 234 (1927). Seidl, Gummi-Ztg , 34, 797 (1920). Staudinger, Kolloid-Z., 54, 139 (1931). Stevens, J . Soc. Chem. Ind., 47, 37T (1928) Stevens a n d Stevens, Ibid., 51 44T (1932). Toyabe, Rubber Chem. Tech., 3, 3S4 (1930). Twiss, J . SOC.Chem. Ind., 44, 106T (1923); Inst. Rubber Ind. Trans., 3, 396 (1928). W h i t b y , Inst. RubherInd. Trans., 6 , 61 (1930). EXG.CHEM.,17, 931 (1925). W h i t b y a n d Simmons, IND. Williams, Ibzd., 21, 872 (1929). RECEIVED September 15, 1934. Presented before the Division of Rubber Chemistry a t the 88th Meeting of the American Chemical Society, Cleveland, Ohio, September 10 t o 14, 1934. This paper is Contribution 30 of the Jackson Laboratory, E. I . du Pont de Nemours & Company.