Polymer Breakdown in Oil-Extended GR-S Masterbatches - Industrial

Polymer Breakdown in Oil-Extended GR-S Masterbatches. W. K. Taft, June Duke, A. D. Snyder, M. Feldon, R. W. Laundrie. Ind. Eng. Chem. , 1953, 45 (5), ...
0 downloads 0 Views 1MB Size
May 1953

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY

used commercially. I n the group of oils, A l , S1, and S4, which have practically identical viscosities at 210' F., there are differences in incorporation and processing which can only be attributed to chemical structure. T h e compatibility of all of the oils in the uncured and cured stocks is good except for the high molecular weight, relatively paraffinic oil, S-5. This oil blooms from the cured stocks a t room temperature, indicating a n upper limit of paraffinicity for rubber process oils. T h e S-5 oil is definitely free from paraffin wax so t h a t the blooming tendency cannot be attributed t o it. The low temperature properties of the cured stocks, as judged by temperature retraction tests, showed that the relatively aromatic oil gave poorer low temperature properties than either the relatively naphthenic or relatively paraffinic oils. Thie was evidenced at both molecular weight levels. If oils of equal viscosity are compared in regard to volatility, aromatic oils will be found t o be more volatile than either relatively naphthenic or relatively paraffinic oils. The relatively paraffinic oil will be slightly less volatile than the relatively naphthenic. Oils with a molecular weight below 350 may be too volatile for some applications. As was expected from the earlier laboratory work, for equal volume loadings the d a t a on tensile strength, 300% modulus, elongation, tear, Shore hardness, and rebound of the factorymixed compounds showed only small differences, both at room and elevated temperatures for the various oils tested. The differences were so small t h a t it is difFicult t o relate them to the chemical structure or physical properties of the oils. Tire test results also indicate t h a t the composition of the oils had very little effect on the abrasion resistance and serviceability of oil-enriched rubbers. The data presented in this paper indicate t h a t petroleum plasticizers for high Mooney GR-S polymers may have a wide

1043

spread in chemical composition and physical properties and, following the usual rubber compounding practices, a n oil should be selected t o give the best over-all balance of properties desired in the final vulcanizate. Compatibility, processability, and low temperature properties are the most important factors t o consider in selecting the petroleum oil for high viscosity rubbers. ACKNOWLEDGMENT

The authors wish to express their thanks t o The General Tire & Rubber Co., Polymer Corp., and Sun Oil Co. for their permission t o publish this paper; t o E. J. Buckler, L. A. McLeod and J. H. Curran for the analysis of the physical test data; t o S. S. Kurtz and C. C. Martin for the analysis of the petroleum oils; t o A. B. Hoe1 for his help in selecting the oils; and t o C. M. Hofmann for arranging the factory evaluation and tire tests. REFERENCES

.

(1) Am. SOC. Testing Materials, Method D 925-51T. (2) D'Ianni, J. D., Hoesly, J. J., and Greer, P. S., Rubber A g e , N . Y., 69,No. 3,317 (1951). (3) Harrington, H. D., Weinstock, K. V., Legge, N. R., and Storey, E. B., India Rubber Workd, 124, No. 5 , 571-5 (1951). (4)Kurtz, S. S . , Jr., Harvey, W. T., and Lipkin, NI. R., Ind. Eng. Chem., Anal. Ed., 11,476-83 (1939). (5) Kurtz, S. S., Jr., and Martin, C. C., India Rubber World, 126, NO. 4, 495-9 (1952). (6) McBain, J. W., J.Phys. Chem., 30,239-47 (1926). (7) Off. Dig. Federation of Paint & Varnish Production Clubs, 7-19 11940). (8) SvtklikT'J. F., and Sperberg, L. R., India Rubber World, 124, NO. 2,182-7 (1951). (9) Swart, G. H., Pfau, E. S., and Weinstock, K. V., Ibid., 124, No. 3, 309-19 (1951). REICEWED for review November 4, 1962. ACCEPTEDMarch 11, 1953. Presented a t the meeting of the Division of Rubber Chemistry. AMERICAN CHEMICAL SOCIETY, Buffalo, N. Y., October 1952.

Polymer Breakdown in OilJ

Extended GR-S Masterbatches W. K. TAFT, JUNE DUKE, A. D. SNYDER, M. FELDON, AND R. W. LAUNDRIE Government Laboratories, University of Akron, Akron, Ohio

T

HE use of oils as extenders for natural rubber is not new.

A British patent (W) granted in 1843 described methods of combining asphalt with caoutchouc as well as the treatment of the latter with sulfur. I n 1935, Rostler and Wilson (8), in Austrian Patent 158,486, describe, as a rubber extender, slightly unsaturated resinous hydrocarbons obtained in a boiling range of 200" t o 280" C. from the acid tars which occur as by-products of petroleum refining. I n 1941, the same authors (11) describe the results obtained by mixing these oils with solid natural rubber or latex. Ludwig et al. (6)describe the effect of mineral oils and unsaturated petroleum derivatives as plasticizers on the processing and physical properties of rubber stocks. Many of the British investigators have reported, in the Journal of Rubber Research, the effects of oils on the various physical properties of compounded stocks. Piper and Scott ( 6 )found that incorporation of small amounts of mineral oil into GR-S without heating or mechanical working had a small softening or plasticizing action, and that large amounts of oil (31.5%) were required for marked softening. The reports on German practices (3) state that oil treatment had been investigated as an alternative t o the usual

heat treating method for softening Buna S. This seems to be related t o the practice shown by the Rostler and Wilson patent (8). Swart et al. ( I d ) remarked that variations in results are obtained by the oil-masterbatching technique. No exact specifications for suitable oils could be stated, and they also cautioned that care must be exercised in drying the oil-rubber mixtures to avoid heat plastication. Data presented by D'Ianni et al. (1) indicate t h a t Circosol2XH yields a n oil-polymer of much higher Mooney viscosity than does Sundex-53 when the oils are incorporated in equal amounts in a high viscosity GR-S latex prepared at 41 O F. The same trend is shown by dilute solution viscosity (DSV) determinations of the dried masterbatches. A recalculation of the data presented by these authors on work done at the Government Laboratories showed that a masterbatch of a high viscosity polymer with 36 parts of Circosol-2XH had broken down to a small degree, as is indicated in Table I. Rostler and Pardew (Q), in discussing commercial petroleum products for use as rubber plasticizers or extenders, describe their stability or instability as the important chemical characteristic.

INDUSTRIAL AND ENGINEERING CHEMISTRY

1044

TABLE I. MOLECULAR WEIGHTDISTRIBUTION AS REFLECTED BY DILUTE SOLUTION VISCOSITY Dilute Solution Viscosity 0

LO

1

1 to 2 2 to 3 3 to 4

4 to 5 5 to

-

Original polymer 15 17 13 11 15 29 Total = 100

-

Fractionated Polymer, % Masterbatch Calculated Determined 40 13 11

37.5 1 % .5 10 8

n

9 18

11

21-

Total = 100

Total =

E

They continued by pointing out that the chemical composition of an oil is of primary importance and influences the properties of the oil compounds. In the present paper, the authors have examined various oils or other plasticizers in masterbatches with high Rlooney viscosity GR-S t o determine how different oils from various sources affect the polymer structure, and whether the same fractions of oils, as defined by Rostler's classification, from different sources behave similarly in their effect on polymer structure. The molecular size of the polymer has been characterized generally by gel and dilute solution viscosity measurements. PROCEDURE

Pilot-plant batches were made by masterbatching, in the latex stage, high viscosity, 41' F. GR-S with a variety of oils. Circosol-2XH and Dutrex 20 v, hich are two commercially available oils differing substantially in chemical composition were chosen as extremes in the range of composition. A series of oils, described by the supplier (Socony-Vacuum Oil Co.) as varying in composition from "aromatic" through "naphthenic" to "paraffinic" (differentiation based mainly on source), were masterbatched with the same high viscosity latex and then tested. Other experimental masterbatches made from the same type of latex and with various other commercial oils were used to complement the foregoing series. The analytical data for these oils are shown in Table 11. It has been reported (1) that each fraction of an oil, as distinguished by the Rostler method (IO) of analysis, imparts distinctly different physic+] properties to the masterbatches and their vulcanixates. Tests were made to determine if these oil fractions also affect the breakdown of polymer differently with respect to fraction type. The parent oil and various fractions

Vol. 45, No. 5

prepared a t the Government Laboratories from another type or source of oil as well as fractions received from the Golden Bear Oil Co. were masterbatched with the same polymer from a gelfree latex made by the GR-S-1500 formula. Circosol-2XH and Dutrex 20, two distinctly different types of oils, were each separated into three fractions. An additional sample of nitrogen bases was prepared from Califlux TT. Since Circosol-2XH and Dutrex 20 did not contain asphaltenes, an asphaltene fraction was prepared from Paraflux. The asphaltenes were prepared by precipitating and filtering the pentane-insoluble material from an equal volume of Paraflux and benzene mixed with 10 times ita volume of normal-pentane. Re-solution of the dried, insoluble residue in benzene and precipitation with pentane were repeated twice, and the final product was dried a t 140" F. in a hot-air oven for 24 hours. The Rostler analyses of the Paraflux and the batch of asphaltenes are: Gravity, OAPI 2 8 -8 1

Paraflux Asphaltenes

Asphaltenes,

Sitropen Bases,

228 97 0

6 3

%

56

1st Acidaffins,

%

2nd Acldaffins, Paraffins,

%

%

c-71.9-*

3 0

c

L

About 19 gallons of Circosol-ZXH, dissolved in an equal volume of normal hexane, were successively extracted a t room temperature with 20% by volume of 25, 50, 70, and 85% sulfuric acid. The acid sludge or extract from this treatment was neutralized with a 10% solution of sodium hydroxide, and the p H was adjusted to about 10. The nitrogen bases and other materials were extracted from this solution with three portions of 2% benzene by volume. The benzene and n-ater were separated from the extract by distillation A 50% by volume solution in benzene of the hard, tarry residue from the distillation was poured slowly with agitation into 10 times its volume of normal pentane. After the solution had stood overnight, the insoluble matter was removed by filtering on a Buchner funnel. The filter cake consisted mainly of sodium sulfate, sodium organo-sulfonates soluble in benzene, and mateiial similar to asphaltenes which comprised about 20% by weight of the tarry residue. The bulk of the pentane was removed from the solution by distillation and the remainder of the volatile material by evaporation on a hot plate The final residue, the so-called nitrogen bases, was a black, tarry ma89 that was not fluid a t room temperature, but which became fluid at 212' F. Six hundred and fifty grams of this product were obtained, which, according t o the Rostler analysis, contained 64.3% nitrogen bases and represented a recovery of about 25% of the nitrogen bases in the original,Circosol-2XH. A similar method mas used to prepare a product, containing 77.1% nitrogen bases, from Dutrex 20. A fraction consisting predominantly of first acidaffins was not

TABLE 11. ANALYTICAL DATAFOR OILS= Rostler Analysis, 70

Saybolt Universal

-Viscosity, Seconds at

looo F.

2,045

!9)

3,195 16,647 78.4 429.4 11,474 103 881 b Too heavy 270.8 103.6 T o o heavy 3,902 362.4d 128.3 504 8,448

210° F. 83.4

86.6 99.8 38.4 45.9 178'~ 38 50 b 3 174 44.4 39.7 132.9 74.8 38.24 44.1 71.4 96.5

"fgf"' a t 60" F. 18.1 11.7 4.2 3.2 12.2b 9 6 25.8 0.96 -0.4 10.3 32.1 9.0 3.2 25.6d 16.9 19.7

7.2

~~~~~i~~ No.

Asphaltenes

7.2

0

17.2 41.8 34.4 23.7 23.5

0

...

37.4 57.0 52.7 2.3 31.5 34.5 4.5d 17.6

...

33.3

0

0 0 1.0 3.8

...

3.6 16.5 10.4 0 0 1.4 0 0 0 0

N

bases 1.0 1.6 6.4 21.5 2.0 5.2 15.6

...

7.9 11.4 12.5 0.0 21.2 13.0 1.9 1.6 0.6 17.7

1st acidaffins

2nd acidaffins

Paraffins

5.8 4.7 11.3 21.6 29.3 18.2 16.4

41.5 43.2 60.5 51.1 64.7 35.1 44.7

51.7 60.5 21.8 5.8 4.0 40.5 19.5

30.5 27.6 29.4 0.4 17.9 34.8 3.4 10.4 7.6 13.8

5.51 43.1 44.2 19.9 48.3 40.2 35.4 63.8 61.2 56.3

i :9

...

...

1.4 3 ,6 79.7 12.6 10.6 59.3 24.2 30.6 12.2

a The sources of the oils were: Dutrex 20, Shell Oil Carp Emeryville Calif ' Golden Bear Light Oil Califlux T T and Califlux G P Golden Bear Oil Co Oil&le Calif.; Sundex-53 and Circosol-ZXH, Sun Oil Co., Marcus Hood, Pa.;"Indonex C-1, special simple from S6andard Oil Co (indiana) Chicago s$ 111.; SovLloid C, Sovaloid N, Sovaloid W, Process Oil 908, PD-1003-A, Prorex Oil D (special oil samples) and 8330 Proc. Oil and 908, Sbcony-Vac;um Oil Co., New York 4 N Y. b Da'ta supplied b y the Socony-Vacuum Oil Co. c prooess Oil 908 is a laboratory sample of the 908 used commercially as rubber processing oil. Data for this material were furnished by the Socony Vacuum Oil Co d Determined on RM-2640.

May 1953

INDUSTRIAL AND ENGINEERING CHEMISTRY

1045

The various masterbatches, compared with oil-free controls of th,e base polyTABLE111. PROPERTIES O F CIRCOSOL-2XH AND ITSFRACTIONS mer, were heated at 140 , 200", and 300" F. or were treated in the Banbury Circosol-2XH Nitrogen 2nd Aliquot (RM-2473) Bases Acidaffins Paraffins Composite for various times and with a standard, oil-free, low-temperature pol -7.1 7.9 24.7 19.3 Gravity "API at 60' F. 18.1 mer. The progressive changes in tXk1.0150 0.9058 0.9383 0.9459 1.1373 Specific 'gravity at 60" F./60° F. Viscosity SUS at 100' F 2045 ... 876 1671 polymer were measured by the changes Viscosity' sus at 2100 F: 83 606' ' 2 a7 70 76 in dilute solution viscosity, the gel conBromine 'No. 7.2 43.3 11.2 0.34 4.5 tent (if any), and Mooney viscosity. 0.24 ... ... ... ... Sulfur (Parr bomb), YG Nitrogen (Kjeldahl), % ... 1.37 ... ... Attempts were made t o correlate the inViscosity index -28 ... 5i' ' -31 herent effects that each oil or fraction Rostler analysis thereof has on a polymer with the Asphaltenes % 0 3.4 0 0 0 physical and chemical properties of the 0 0.6 Nitrogen bakes, % 1.7 64.3 oils. LO} 0 2.7 1st acidaffins, % 4.6 10.1 42.8 18.6 89.9 20.5 42.1 The samples for oven-aging were passed 2nd acidaffins, yo Paraffins, % 50.9 3.6 9.1 79.5 54.6 through hand-tight mill rolls, and the sheets were cut into narrow strips. The strips were heated for various times, removed from the oven, and tested for gel content and dilute solution viscosity by the modified Baker (6) method, and also were tested in a prepared from Circosol-2XH, as the content of this material Mooney viscometer at 212" F. An equilibration period of 24 in the original oil was less than 5%. hours was used for the gel and DSV tests in the case of soluble samples, and 3 days were allowed in the case of gelled samples. About 700 grams of a fraction consisting of about 90% second To determine the amount of soluble material plus the content of were prepared by 16 extractions and cross extractions additive in the polymer, all samples were extracted in a Soxhlet the small amount of of Circosol-~XHwith furfural. To apparatus with acetone for l5 hours. The gel and DSV the first acidaffins present, the extracts were diluted with five were On the acetone-inso1ub1e basis. times their volume of normal hexane, and the resultant solution Loadings Of 'Oo0 and l4Oo grams Of a polymer and the was extracted three times with 10% by volume of 96% sulfuric masterbatches were masticated at a rotor speed of 116 and 80 acid. The final raffinate,diluted with hexane, was contacted with r.P.m.9 respectively, in a si2e-B Banbury for 5 and 10 minutes about 5% of Filtrol at room temperature for 12 hours to with 5 gallons per minute of cooling water of about 70" F. being the oil-soluble sulfonates. used* Gel content, DSV, and Mooney viscosity values were The acidic oil remaining after the nitrogen bases were extracted from the Dutrex 20 was neutralized, washed with water, ~ ~ $ ~ 1,5~ ML-4) ~ o and distilled, and the moisture-free oil was diluted with hexane. and 83% of tight gel (.swelling index of 12) was latex-masterBy the method of cross-extracting with furfural, two fractions batched with 81 and 85 parts Of Circosol-2XH and Sundex-53, were made which were rich in second acidaffins. oneof these respectively. The base Polymer Was prepared in a 500-gallon contained an appreciable amount of first acidaffins and the other reactor by the GR-S-1500-type formula. The polymerization a n a preciable amount of araffins (see Table IV). was stopped at the end of 33 hours when the conversion was 93%; TEe acidic solution of &rcosol-2XH, from which the nitrogen l.5% Of phenyl 8-naphthy1amine Was added t o the latex Prior to bases had been extracted, was diluted with hexane and successively extracted with three and four portions of 5% by volume of storage. The heat-treated two andas describedOfpreviously. the polymer were and tested gs% and 20% fuming sulfuric acid, respectively. The oilsoluble hydrocarbon sulfonates were extracted with a 2-propanolwater mixture. The final purification of the paraffinic white oil RESULTS AND DISCUSSION was obtained by contacting it with Filtrol a t 150" C. and subsequently percolating it through Attapulgus clay. I n some of the first pilot plant batches made with Circosol-2XH , using 45 Parts of oil Per 100 Parts of Polymer, it was found t h a t A 1000-gram batch of an aliquot composite of the oil fratthe Mooney viscosity values for various batches dried a t 140' F. tions was prepared to demonstrate that the properties of the in a forced-air oven were relatively uniform, whereas the viscosCircosolto those of the composite were ities of similar batches made with Dutrex 20 were variable. As 2XH, and that the properties of the individual fractions, with certain lots of a Dutrex 20 masterbatch had varied in viscosity respect to the properties exhibited by the original oil, were not

...

d e ~ ~ ~ , e " , $ ~ (:@ ~ $$9%' ;:$ ~ ~

altered greatly. Since the first acidaffins were not collected and were removed as thoroughly as possible from the other fractions, their effect on the properties of the composite wa8 assumed to be small. The first acidaffin content of the original Circosol-2XH was less than 5%, and it probably did not contribute very much to the over-all properties of the original oil. The quantities of the several fractions of Circosol-2XH used to prepare the composite were: Ingredients Nitrogen bases 2nd acidaffins Paraffins

% 3.0 31.0 66.0

-

Weight, Grams 30 310 660

Table I11 shows the analytical data obtained for the Circosol2XH, the three fractions, and the aomposite of the fractions. Table IV shows the analytical data obtained for the original Dutrex 20 and its fractions. The analyses of the nitrogen bases from Califlux TT (prepared as described earlier) as well as the nitrogen bases, acidaffins made from Califlux GP, second acidaffins, and naphthenic saturated hydrocarbons are shown in Table V. Masterbatches were made from one GR-S-1500-type latex stabilized with about 1.5% phen 1 B naphthylamine with the above-mentioned oil fractions, F g x o i TOF (tri-2-ethyl hexyl phosphate), and some of the other oils described in Table 11.

from to by much 2o to 50 ML-4 units, made with each Oil were dried and left in the Oven for days. The change in viscosity with time of heating is shown in Table VI. The series of oils (analytical data shown in Table II), comprising Sovaloid C, PD-1003-A, Sovaloid N, and W, were described by the Socony-Vacuum Oil CO. as "aromatic oils" t h a t decrease in "aromaticity" in the order stated, Process Oil 908 was described

TABLE IV. PROPERTIES OF DUTREX20

,

~

Dutrex Nitrogen Bases 20 Gravity 'API at 60° F. 4 . 2 0.9 Sp gr )BOe/600 F. 1.0428 1.0687 Viicos?ty, SUS at 100' F. 16,647 T o o heavy Vicosit SUS at 100 1,529 B$2Pn!ho. 42 Sulfur (Parrbomb). % 0 . 9 3

... ...

Nitrogen % (K'eldahl)' Rostler analysis Asphaltenes, YG

~

~

~

2nd acidaffin;, % Paraffins, %

... 0

ig:~ 47.7 5.6

2.10 2.0

~7 87 .. 13 , 11.2 1.4

AND

ITSFRACTIONS

1st and 2nd Acidaflns -0.7 1.0818 35,720

1592

112 .

2nd Acidaffins and Paraffins 14.9 0.9665

72 I

.

.

... ...

0 1.1 3 2". 7 65.4 0.8

...

...

... 0

~

0~. 6 1.2 76.1 22.1

%

'

,

1046

Vol. 45, No. 5

INDUSTRIAL AND ENGINEERING CHEMISTRY

gel is 10 or more, then these values seem t o indicate a definite breakdown h-itrogen Bcidaffins Saturated in the soluble polymer chain. This Bases from Nitrogen from 2 nd Hydrocarbons, Califlux TT Basesa Califlux GPa Aoidaffinsa Kaphthenioa breakdown apparently increases with Viscosity, SUS aromaticity of the oil. At loo0 F. Too heavy T o o heavy . . . 2910 350.4 A t 210° F. 1403 158.2 ... 119.2 83.5 The fact that contrary conclusions can Gravity "API at 60" F. 1.2 3.4 ... 9.2 27.1 Bromini No. 47.2 43.8 26.0 13.0 0.0 be drawn from the two series of data for Rostler analysis the effect of variation in degree of Asphaltenes % 3.8 Trace 0 0 0 aromaticity of the oil on the polymer Nitrogen bakes % 66.8 89.1 2.6 1.3 0 1 s t acidaffins, 9 0 8.9 5.0 2 2.7 0 probably mean8 t'hat this particular 2nd acidaffins, % 17.2 5.1 72 2 .. 1 2 90.6 5 .. 23 Paraffins, % 3.3 0.8 3.1 5,4 94.5 variable may not be a critical one, but a Furnished by F. 9.Rostler of the Golden Bear Oil Co. that other fact'ors also are involved. The data of Table VI11 appear to demonstrate that the various oils contain TABLEVI. MOONEYVISCOSITYOF XASTERBATCHES~ AFTER some material or materials that effect, \TARIOUS T I l f E S O F HEATIXG .4T 140' F. the breakdown of polymer on heating. Viscosity, ML-4 To determine quantitatively this breakdown due to heating Days Heated Circosol-PXH Dutrex 20 and Banbury mixing, the procedures already described were used. 60 70 1 (just dry) I n making comparisons, the dilute solution viscosity values of the 52 55 2 60 47 3 masterbatches were calculated for the contained polymer, based 55 36 4 57 27 5 on the assumptions that the amount of acetone-insoluble polymer 55 21 6 is not changed by the treatments used and that the DSV values a >lade with 45 palts of Circosol-2x11 or Dutrex 20 per 100 parts of polymer. of the oil or plasticizer and polymer are additive functions, and t h a t the DSV of the contained oil or plasticizer did not change during the treatment of the masterbatch, and was negligible in comparison to that of the polymer. Some substantiation of the as naphthenic, and Prorex Oil D was described as mainly parafvalidity of these assumptions has been arrived at (data not finic. Summarized viscoeity. gel content, and dilute solution shown here) by demonstrating that the DSV of benzene soluviscosity values for masterbatches of the same latex with 36 tions of known oil-polymer mixtures are the same as those calcuparts of each of these oils, compared with similar masterbatches lated from these assumptions. containing Circosol-2XH and Sundex-53 (both naphthenics), The amounts of rubber hydrocarbons were estimated by aceas well as X-624, a cold rubber made without oil, are shown in tone extraction of the masterbatch and the base polymer. In Table VII. using the acetone extraction method for evaluating the gel and DSV data, certain assumpPOLYMER MASTERRATCHED WITH 36 PARTS O F OILSO F TABLEVII. GR-S-~~OO-TYPE tions were made with respect to DIFFERENT TYPES Sov. C PD-1003-9 Sov. N Sov. W Pro. 908 Prorex D Ciroosol-2XH Sundex-53 X-624 the magnitude of the errors Viscosity, ML-4 63 57 56 65 56 50 60 62 48 that would be introduced. It Gel, % 4 14 9 7 5 3 6 2 .. was assumed that the quantiDSV 2.42 2.51 2.31 2.36 2.40 2.36 ,2.37 2.33 .. ties of organic acids, soaps, inorganic salts, shortstops, and antioxidants in the contained polymer remained essentially constant for all of the masterbatches These data do not permit any conclusions with respect to the as well as for the base polymer; the polymer solubilized by the effect of drying a t 140' F. on the rate of breakdown by the oils, oil or plasticizer and retained in the acetone extract was insignifiso that any possible effect of the oil type (naphthenic or paraffinic) cant in amount; and the acetone-insoluble material was the is not apparent. rubber hydrocarbons, and that any error caused by incomplete Another series of masterbatches was made in 500-700-pound extraction could be ignored. lots using higher viscosity polymer that contained gel and 45 parts of several oils varying in degree of aromaticity. I n order that masterbatches of comparable viscosity might be obtained, the raw polymer for the Dutrex 20 masterbatch was made to a higher viscosity than that used with the other oils. The data in Table VI11 indicate generally that the breakdown of the polymers has increased with an increase in aromaticity of the oil. TABLE

Tr.

AXALYSES O F FRACTIONS O F TTARIOUS OILS

B 1 5 E POLIKER

TABLEVIII.

C ~ A N GI NE GEL ASD DSV XITH ISCREME IN AROMATICITY OF ADDEDOIL

(Oil-extended masterbatches compared to raw polymer masterbatch) Oils in Order of Decreasing Aromaticity (Kaphthenio to Paraffinic) Dutrex Califlux Sundex- Circosol20 TT 8330 53 2XH 908 Loss in gel, % 9 10 14 3 14 units 18 Change in DSV, -0.08 -0.23 0.38 0.34 -0.02 units -0.16

ClflCOSOCZXH O C L C L N ECAR L l G H l OIL

F L F X l L TOF

CALIFLUX GP OUTREX 20

I

1.00

0 12

24

48

72

IZO

910

336

ASlNG TIME, HOURS

These data indicate that there is actually a solubilization of the gel. Assuming that the dilute solution viscosity of the solubilieed

Figure 1. Change in DSV of Masterbatches with Time o f Heating at 140" F. 30 p a r t s of plasticizer

INDUSTRIAL AND ENGINEERING CHEMISTRY

May 1953

1047

similarly to the base polymer, but the rate of breakdown was higher initially, and condensation occurred somewhat faster as indicated by the higher DSV value when heated for 72 hours. The masterbatches containing the Golden Bear Light Oil, up to 12 hours of heating, broke down similarly, or a t only a slightly higher rate, than did those containing the Circosol-2XH. The abrupt change in the apparent rate of breakdown at 12 hours may have been caused by volatilization of the oil, The polymer masterbatched with Indonex C-1 broke down a t this temperature almost as \ \\ INDONY PLASTICIZER C l fast as did the polymer with Califlux GP, but then condensed at a more rapid rate. Both the Indonex and Sundex samples increased in DSV after about 12 hours of aging, indicating t h a t the condensation reaction had become predominant. Dutrex 20 and the Califlux G P caused rapid breakdown LOO' 01 3 6 12 24 48 72 of the polymer. The sudden change in the rate AGING TIME, HOURS of breakdown with Dutrex 20 between 3 and 6 Figure 2. Change in DSV of Masterbatches with Time of Heating at hours, and the equally abrupt change in rate for 200" F. the Califlux GP-polymer a t 6 hours, may indicate 30 parts of plasticizer two possibilities: first, depletion of reactive groups in a material or materials that caused the breakdown, or second, the competing reactions of breakThis method corrects for volatilization of the additive during down and condensation were in equilibrium. Again, the data the drying process and the heat and Banbury treatments; however, means were not available for correcting the Mooney viscosity values for any oil lost from the masterbatch. The error introduced by variations in the time of drying the polymers, which causes irregular breakdown of the polymer chain, was recognized. I n computing the gel content and DSV values of the contained polymer, some errors might be introduced, but the results were interpreted by comparison, so errors probably were mutually compensating. The comparisons should be valid. The change in dilute solution viscosity with time of heating at 140' F. of masterbatches containing commercial oils described in Table I1 and containing Flexol TOF and the base polymer is provided in Figure 1. Aging of gel-free base polymer a t 140" F. caused relatively little breakdown as judged by change in DSV. Differences in the dried polymers at zero time of breakdown may be due partly t o experimental error or to the variable breakdown obtained during drying of the masterbatches. Circosol-2XH and Golden Bear Light Oil caused faster and more prolonged breakdown of the oil-polymer, as compared to the base polymer. Stocks with Califlux G P and Dutrex 20 were generally alike; they broke down at a high initial rate, with a marked decrease in rate of breakSUNDEX !NOONEX PLASTIClZER C1 down after 120 hours. The data for Flexol TOF (Figures 1 and RlLlMER 2) indicate that this phosphate caused breakdown and that i t 2.00 probably inhibited condensation a t 140' and 200' F. Sundex-53 FLEXOL TOF and Indonex C-1 occasioned rather rapid initial breakdown. I.BOI o 025 as0 0.75 1.0s 1.50 2.00 The data for changes in DSV of masterbatches with time of AGING TIME, HOURS heating at 200" F. are shown in Figure 2. The rate of breakdown Figure 3. Change in DSV of Masterbatches with Time of Heating at 300" F. of the base polymer a t 200" F. was considerably greater than a t 140" F. The polymer masterbatched with Circosol-2XH behaved 30 parts of plasticizer

0A

TABLE I X . EFFECT ON GEL" AND ACETONE EXTRACTS OF HEATING POLYMERS AT 300" F. ti^^ Base Polymer Circosol-2XH G.B. Light Oil Califlux GP Dutrex 20 Flexol TOF Sundex-53 Acetone Time, Acetone Acetone Acetone Acetone Acetone Acetone Hr. Gel, % ext., % Gel, % ext., % Gel, % ext., % Gel, % ext., % Gel, % ext., % Gel, % ext., % Gel, % ext., % 0 0.25 0.50 1.0 2.0 4.0 4.75 a

5 1 1 1 15 35 45

7.6

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

8.0

4 1 1 3 1 14 27

29.3

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

...

28.8

8 8 3 6 19 13 31

Gel calculated t o rtoetone-insoluble basis,

29.0 25.9 23.1 23.3 20.4 19.3 19.6

3 1 1 4 3 21 15

28.3

'

... ... ... ... ... 27.2

4 3 1 4 1 10 20

31.1 31.0 29.0 29.9 30.5 28.8 28.4.

3

1 1 4 4

-18

16

29.4 29.2 28.9 27.1

22.g

x7.v 26.8

4 6 1 1 9

... ...

Indonex C-1 Acetone ext., %

Gel, %

30.9

10

... 29.8 ... ...

3 1 11

... ...

3

... ...

31.0 ..I

...

...

29.3

... ...

01

0 12 24

48

72

I20

240

AGING TIME.

gLn

8

336

HOURS

AGING TIWE, HOURS

Figure 4. Change in Mooney Viscosity of Masterbatches with Time of Heating at 140' F.

Figure 5. Change in Mooney Viscosity of Masterbatches with Time of Heating a t 200' F.

30 parts of plasticizer

30 parts of plasticizer

10 60

(n

>

Vol. 45, No. 5

INDUSTRIAL AND ENGINEERING CHEMISTRY

1048

BHT OIL

50 40

LUX GP

30

F L E X O L TO?

ClQC3SOL.ZXH

20

DUTREX 20

cc GOLOEN B E A R L h m T OIL

0

0.25 050

Figure 6.

I.M

2.00 AGING TIME, HOLRS

4.00

Change in Mooney Viscosity of Masterbatches with Time of Heating a t 300" F. 30 parts of plasticizer

35

I

-0

5 BANBURY - , Y E , MINJTES

3

Figure 7. Effect of Banbury Mixing Time on the Mooney Viscosity of Masterbatches 30 parts of plasticizer, 1000-gram loading

.% FLEXDL

-OF

C . R C O S O L - ZXH COLOEN BEAR LIGHTOIL

I

CA. cLUX GP

2 20 BANBURY TIME, MINUTES

Figure 9. Effect of Banbury Mixing Time on the DSV of Masterbatches 30 parts of plasticizer, 1000-gram loading

1049

INDUSTRIAL AND E N G I N E E R I N G CHEMISTRY

May 1953

1 NlTROGEIiBASES

HER

GOLOEN BEAR LIGHT OIL TOF

O LXEL$/ /

.,

zoo

m

2.20 5 10 BANBURY TIME, MINUTES

0

Figure 10. Effect of Banbury Mixing Time on the DSV of Masterbatches 30 parts of plasticizer, 1000-gram loading

90

M 9 E POLYMER c i a w s a ~ . B n(ai PI

BO

SUNOEX 53 185 P I

I

70

* 50

0

I G E D AT I4O.F '

6

'

12

24

Rostler Analysis of Oil Sources for Nitrogen Bases Circosol2XH Dutrex 20 Golden B e a r Califlux TT Trace 3.8 Asphaltenes, % 3.4 2.0 89.1 66.8 Nitrogen bases, 90' 64.3 77.1 5.0 8.9 1st acidaffins, Yo 10.1 8.3 5.1 17.2 2nd acida5ns. % 18.6 11.2 .. Paraffins, $6 3.6 1.4 0.8 3.3

72

48

61%

POLYMER

Rostler Analysis of Oil Sources for Acidaffins CircosolAcidaffins, 2XH Dutrex 20 Califlux G P Golden Bear

SUNOEX 53

AGED AT ZOOOF

Nitrogen bases lstacida5ns 2 n d aoidaffins Para5ns

'DL

1.1 0.6 32.7 1.2 65.4 76.1 0.8 22.1

2.6 22.1 72.2 3.1

1.3 2.7 90.6 5.4

101 P I R T S I

40

"

"O 89.9 9.1

O ,. K AGED 4 1 300.F 0.20

0.50

O.&

ILO

2.w

AGING TIME, HOURS

Figure 11. Change in Gel of Masterbatches with Time of Heating at 140°,200", and 300' F.

for masterbatches with Flexol TOF indicated a gradual breakdown. The approach to the break-point with the Circosol2XH masterbatch was not as rapid as with the Sundex-53 and Indonex C-1 masterbatches, but there seemed t o be an increase in DSV with all oils, except the Light Oil and Flexol TOF. When heated a t 300" F. (Table IX), the base polymer and that masterbatched with Golden Bear Light Oil gelled between the first and second hour. The Sundex-53 and Indonex C-1 polymers showed some indication of gel a t 2 hours, whereas the other polymers developed 10 to 20% gel after 4 hours of heating. Some of the oils a t 300' F. (Figure 3 ) effected a more gradual change in predominance of chain scission over to that of iondensation as compared with other oils. The data for Flexol TOF show that some condensation occurred after heating for 1 hour. The rates of breakdown of the polymers masterbatched with the various oils do not differ so much for short times of heating a t 300" F. as at the other temperatures. All of the oils seem to have

GOLOEN BEAR OIL GO CIROOSOL.2XH OUTREX 20 8bSE POLYMER CALlFLYX TT

, 0

d dblO4FfIHS FROM CIRCOSOL'ZXW

0.50

2nd bCICdFfINS FROM O M E N BFAR OIL M

2w

1.50 0 1

3

e

10

PI

AGING TIME, HOURS

Figure 13. Changes in DSV with Time of of Heating a t 200' F. of Masterbatches Containing 30 Parts of Nitrogen Bases or Acidasns

1050

INDUSTRIAL AND ENGINEERING CHEMISTRY

L

3 50

20

8

50

0

025

054

0.75

bGINi'?IME, HOURS

2.0

Figure 14. Changes in DSV with Time of Heating at 300" F. of Blasterbatches Containing 30 Parts of Nitrogen Bases or Acidaffins

Vol. 45, No. 5

the oil had no effect on the breakdown, greater slope would be expected because of the greater amount of work and heat developed with the stiffer polymers. At 1400-gram loadings, as shown in Figure 8, the differences caused by the oils are smaller. The DSV-breakdown data, shown in Figure 9 for 1000-gram loadings of the Banbury, in general confirm the trends shown by the Mooney viscosity data. The DSV pattern (Figure 10) a t loadings of 1400 grams is not comparable to that at 1000-gram loadings, but differences in DSV data caused by the various oils are pronounced. Sundex-53, incorporated into a very high viscosity polymer containing a large amount of tight (relatively unswollen in benzene) gel, solubilized and loosened the gel structure to a greater extent than did Circosol-2XH a t 140" and 200" F., as shown by Figure 11. The rate of solubilization with both oils was much higher when the temperature was increased to 300' F. Although the initial rate a t 200" F. with Sundex-53 was not so high as that at 300" F., the solubilization was more sustained t o a Ion-er gel level. The effect of Circosol-2XH was considerably less. The final data show a lower gel content for Sundex-53 and a higher gel content for Circosol-2XH after treatment a t 200" F. than at 300" F. The preceding data indicate that there were differences in oils relative to their effects on the heat breakdown, solubilization of gel, and probably on the processing characteristics of the masterbatches. Masterbatches made with the oil fractions, distinguished by the Rostler method as shown in Tables 111, ISr, and V, were heat treated as previously described. The plots of the DSV (corrected to the rubber hydrocarbon basis) versus time of heating are shown in Figures 12, 13, and 14 for temperatures of 140', 200", and 300" F., respectively. The variation in the DSV of the unheated masterbatches shown in these figures is probably caused by lack of uniform drying of the masterbatches after coagulation. Comparison of the rates of breakdown or condensation of the polymer for the masterbatches made from the nitrogen bases of the various oils with those of the raw polymer is shown in Table X. according to this comparison, it is evident that nitrogen bases from different sources do not behave alike. The balance

inhibited the rate of the cross-linking reaction, as shown by the gel results in Table IX. I n general, Dutrex 20, Califlux GP, and Indonex C-1 contributed to more rapid breakdown of polymer than did Sundex-53, whereas Circosol-2XH and Golden Bear Light Oil yielded the least breakdown. The change in breakdown rate of the polymers with time of heating varied considerably for the various oils, which suggests, in some cases, that the material causing the breakdown had been depleted of reactive groups. Another possible explanation is that there were some materials present that accelerated and some that inhibited condensation. The activity of these assumed breakdown and condensation catalysts varied with the oil, and their effects on the rate of scission and condensation seemed to be influenced by temperature. Figures 4, 5, and 6 show the ;\looney viscosity after heating the masterbatches prepared with the various oils and plasticizer a t 140", 200°, and 300' F. The Mooney viscosity breakdown a t the > % 3.00 three temperatures generally correlated with the 0 ? effect shown by dilute solution viscosity-i.e., t? some oils caused a relatively high rate of break2.50 down (Califlux GP, Dutrex 20, Indonex C-1, and Sundex-63) while with another (Circosol-2XH), the breakdown was relatively small. 2.001 48 72 24 0 8 I2 These data indicate that the viscosity of dried AGING TIME, HOURS masterbatches made from the same latex with Figure 15. Changes in DSV with Time of Heating at 140' F. of different oils will be variable. Such variability of Masterbatches Containing Circosol-2XH and Its Fractions viscosity may result in lack of plant control of the 30 parts of plasticizer final viscosity, when certain oils are used in the Rostler Analysis of Circosol-2XH and Its Concentrated Fractions masterbatch, and of breakdown during Banbury Composite Oil Nitrogen 2nd of treatment and probably during milling. The difStocks Bases Acidaffins Paraffins Fractions ference in Banbury breakdown trend is shown in Asphaltenes, % 3.4 Figure 7 where 1000 grams of masterbatch were Nitrogen treated for various times in the size-B Banbury. bases, Yo 1.6 64.3 0.6 1st acidaffins, The lesser slope for the higher viscosity original e?4.7 10.1 1.0 2.7 Znd"acidaffins, masterbatches confirms that there is a difference 43.2 18.6 89.9 20.5 42.1 % 3.6 9.1 79.5 54.6 in Mooney breakdown with the various oils. If Paraffins, % 50.5

..

..

.. .. ..

~

..

May 1953

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY

tions, or in the impurities themselves, may be the predominating factor. It is probable that the oil source, reflecting the differences in chemical composition, is the important variable. Assuming that there are one or more active groups which will accelerate breakdown or condensation, fractionation based upon the Rostler classification will not necessarily separate the active constituents from the inactive ones. A comparigon of the effects of the several oil fractions was made to show the contribution of each fraction to the properties of the original oil. The aliquot blend of the fractions of Circosol-2XH was tested to reveal any deviations in composition which could be ascribed to the method of fractionating the original oil. Figures 15, 16, and 17 show the data obtained for the fractions of Circosol-2XH; Table X summarizes the data. It is evident from the data that the nitrogen bases of Circosol2XH contain the material that causes the fastest breakdown a t 140' and 200' F. From the results obtained a t 300' F., it appears that the materials responsible for breakdown or condensation react at different rates and that these rates are greatly affected by temperature. The general agreement between the data obtained with the original Circosol-2XH and the aliquot composite of the fractions apparently demonstrates that the method used for preparing the fractions is sound.

2nd AOIWFFIHS >-NlTROOEN

8ASfS

awosirf

Powwns

AGING TIME, HOURS

Figure 16. Changes in DSV with Time of Heating at 200" F. of Masterbatches Containing Circosol-2XH and Its Fractions

1051

.

30 parts of plasticizer

between breakdown and condensation is different for each of the nitrogen-base fractions, and the effect of temperature on this balance is not the same for each of the fractions. Since the nitrogen bases tested in this investigation differ both in molecular weight (as evidenced by different viscosities) and in their content of elementary nitrogen (see Tables I11 and IV), the difference in behavior of the products might be due to differences in structure and in basicity. The existence of two different types of nitrogen bases has been pointed out earlier (IO).

TABLEX. CHANGES IN POLYMER STRUCTURE CAUSED BY HEATING MASTERBATCHES MADEWITH NITROGEN BASES,ACIDAFFINS, AND CIRCOSOL-2XH A N D ITSFRACTIONS Nitrogen bases from Circosol-2XH Dutrex 20 Golden Bear Oil Co. Califlux TT

2.M COMPOSITE

Temperature of Heating, F. 140 200 300

A 0 A C A: 0

A

O C

0' c A: C

0

025

050

OK

I O

bGlNG TIME, HOURS

20

Figure 17. Changes in DSV with Time of Heating at 300" F. of Masterbatches Containing Circosol-2XH and Its Fractions 30 p a r t s of plasticizer

Circosol-ZXH and its fractions Circosol-2XH (original) c 0 A O,A a'A Nitrogen bases A A, c d A 2nd acidaffins C a, C Paraffins C C, a O'a A Aliquot composite 0 0, a A: d, A Compared with base polymer A = accelerated breakdown. a = slightly accelerated breakdown. C = condensation predominates. c = slight eondensation. 0 = no change in rate of breakdown or condensation. a Furnished by F. S. Rostler of the Golden Bear Oil Co.

Table X represents, in a similar manner, the test data for the masterbatches containing acidaffins from the various oils. T h e same general conclusions may be noted for the acidaffins as for the nitrogen bases; however, the variattons in the purity of the frac-

Table X also indicates that the second acidaffins and paraffins are relatively inert a t 140' or 200' F., and either are free of materials that cause breakdown and condensation or contain them i o balance. However, a t 300" F., the materials that cause breakdown or the reaction products of such materials remain to prevent gelation. Figure 1 shows that the breakdown a t 140' F. is faster for t h e polymer in the masterbatch with Circosol-2XH than for the raw polymer, while, in Figure 15, the reverse is true. Different batches of Circosol-2XH were used to make up the two masterbatches, but the latex was the same, except for time of storage; this indicates that there is a difference in effect caused by differences in the Circosol samples. At 200' F. (Figures 2 and le), the rates of change in the DSV of the masterbatched polymer and that of the raw polymer are similar. The data shown in Figures

INDUSTRIAL A N D ENGINEERING CHEMISTRY

1052

BASE

0

12

6

24

48

POLYMEh

72

AGING TIME, HOURS

Figure 18. Changes i n DSV with Time of Heating at 140' F. of Masterbatches Containing Dutrex 20 and Its Fractions 30 parts of plasticizer

BASE 'POLYMER

i 131 a 2 n d

AGIDA~NS

I

. a

Vol. 45, No. 5

down, condensation, or Mooney viscosity change. The two other fractions from Dutrex 20 had no counterparts among the Circosol-2XH fractions, hence any comparison of the former with the Circosol-PXH fractions would have no significance. The sample of asphaltenes was Iatex-masterbatched, and the masterbatch was heated in a manner comparable t o t h a t for heating the other oil-masterbatches. Since the asphaltenes were not removed quantitatively from masterbatches by acetone extraction, the DSV values cannot be relied upon to correlate with the breakdown of the polymer. The plasticizing action of the asphaltenes either was negligible or a slight stiffening effect was produced; the base polymer had a viscosity of 131 ML-4 whereas the asphaltene masterbatch had a viscosity of 134 ML-4. The breakdown a t 140" and 200" F., as indicated by the Mooney viscosity (data not shown) was comparable to t h a t of the base polymer with indications that the asphaltenes inhibited breakdown slightly. At 300' F., the asphaltenes effected a slower breakdown of the masterbatch than that of the base polymer, and the retardation of polymer cross-linking in the masterbatch was significant; however, a t the end of the 2-hour heating period, 20 to 30% gel was formed which had no apparent effect on the Mooney viscosity. SU-MRIARY

0

It has been shown that various commercial oils used in masterbatching high Mooney GR-S, as well as fractions of these oils, differ in their Figure 19. Changes in DSV with Time of Heating a t 200' F. of effect on rate of polymer breakdown or conMasterbatches Containing Dutrex 20 and Its Fractions densation, or on the change in Mooney viscosity. 30 parts of plasticizer The differences in the function of oils in high Mooney viscosity GR-S masterbatches are caused bv differences in chemical comDosition of the oils. 3 and 17 a t 300' F. are similar but probably not quantitatively which are not fully defined by the analysis determining per cent alike. composition in terms of asphaltenes, nitrogen bases, acidaffins, The changes in Mooney viscosity of the masterbatches prepared with the fractions of Circosol-2XH and with the original oil and aliquot composite of the fractions vary as expected from the changes in the DSV (data not shown). From tests a t 140" F., it appears that the nitrogen bases contain the materials that cause most of the breakdown. However, a t 200' and 300" F., 350 no real dzerence appeared in Mooney viscosity breakdown caused by the oil fractions. The decrease in the DSV of the polymer masterbatched with 0ASE P M l H E R Dutrex 20 and its several fractions, plotted against the time of P heating at the three temperatures, is shown in Figures 18, 19, G E N BASES and 20. At 140' and 200' F., the polymers containing ( a ) the first and second acidaffins and ( b ) the second acidaffins and paraffins broke down in a manner comparable to the breakdown o 025 am 075 io 2D nTith the original oil, but the nitrogen bases caused slower breakA G I N G T I M E , HOURS down than did the original oil. At 300" F. the masterbatch Figure 20. Changes in DSV with Time of Heating a t containing the nitrogen bases and also that containing the first 300' F. of Masterbatches Containing Dutrex 20 and Its and second acidaffins broke down less in the first half hour but Fractions did not condense as much after this time as did the masterbatch 30 parts of plasticizer with the original Dutrex 20. The second acidaffin and paraffin did not cause as rapid breakdown as did the original Dutrex 20, but the breakdown was more sustained with less indication of condensation. and paraffins. Attempted correlation of the various effects obtained wit,h physical constants such as gravity, viscosity, viscosity The effect of the nitrogen-base fractions, compared t o the efindex, or bromine number of the oils does not show any obvious fect of the original oils, was not the same for fractions prepared from Circosol-2XH and Dutrex 20 with respect t o polymer breakinterrelationships. ~~~~~

~

I

3

6

I2

A G I N G TIME, HOURS

24

INDUSTRIAL AND ENGINEERING CHEMISTRY

1953

LITERATURE CITED

D'Ianni, J. D., Hoesly, J. J., and Greer, P. S., Rubber Age ( N . Y.), 69, 317-21 (1951). Hancock, Thomas., Brit. Patent 9952 (1843). KixMiller, R. W., and Weidlein, E. R., Jr., Dept. of Commerce, Washington 25, D. C., OTS PB Rept. 13340, April 21, iLnVA c 7.V.

Ludwig, L. E., Sarbach, D. V., Garvey, B. S., Jr., and Juve, A. E., IndiaRubber World, 111,55 (1944). Mullin, J. W., and Baker, W. D., private communications. ( 6 ) Piper, G. H., and Scott, J. R., J . Rubber Research, 17, 135-44

1053

S.,and Mehner (Wilson). Vilma. Austrian aatent 158,486 (1935). (9) Rostler, F. S., and Pardew, M. B., Ibid., 63, 317-26 (1948). (10) Rostler, F. S., and Sternberg, H. W., IND.ENG.CHEM.,41, 598-608 (1949). (11) Rostler, F. s., and Wilson, V. Mehner, India Rubber World, 104, 47-51 (1941). (12) Swart, G:H., PfaU, E. s., and Weinstock. K. V., Ibid., 124, 30919 (1951). ( 8 ) Rostler. F.

I

(1948). (7) Rostler, F. S., Rubber Age ( N . Y.),69, 559-78 (1951).

RECEIVED for review November 4, 1952. ACCEPTED February 26, 1953. Work sponsored by the O 5 c e of Synthetic Rubber, Reconstruction Finance Corp. in connection with the government synthetic rubber program.

Properties of GR-S Extended with Rosin-Type Acids L. H. HOWLAND, J. A. REYNOLDS, AND R. L. PROVOST Naugatuck Chemical Division, United States Rubber Co., Naugatuck, Conn.

S

I

I N C E the early days of the government synthetic rubber program, experimental work has been carried out by many investigators on polymerization in the presence of large amounts of soap. While the objectives were varied, most of this early work was conducted to obtain fundamental information. Although there was some reason to hope for quality advantages in the polymers so prepared, the fact t h a t evaluation techniques were not fully satisfactory made observation of any inherent advantages difficult. I n view of the development of oil-extended polymers, (3, 4)it became desirable to employ the approach used in the case of these products in the evaluation of rubbers polymerized in the presence of large amounts of soap. However, in order fully to evaluate the effects of the several variables involved in this procedure, preliminary work included the addition of soaps to normal latices containfng high Mooney viscosity polymers in order t o use them for controls so that the true effect of polymerizing in the presence of high amounts of emulsifier could be determined. The technique employed in the work with latices originally containing normal (about 5 parts per 100 parts of charged monomers) amounts of soap was more or less analogous t o that used in the preparation of polymers extended with petroleum oils. The oil-extended products utilize a high Mooney viscosity cold GR-S t o which is added a cheap petroleum oil in quantities sufficient t o soften the rubber t o the extent t h a t the final viscosity of the extended stock is within the usable range. I n practice the oil is emulsified in water with conventional soaps, and the emulsion is added to the synthetic rubber latex just prior t o the coagulation step. T h e addition of salt and acid during the coagulation step then destroys the latex and oil emulsions simultaneously giving a fairly homogeneous dispersion of oil in rubber which is treated in subsequent processing, compounding, and curing operations as if it were 100% rubber hydrocarbon. The oil-extended polymers have been shown t o compare favorably with their all-rubber counterparts from the standpoint of -resistance t o abrasive wear in tires and have given vulcanizates with lower heat build-up a s measured in laboratory tests. The economic implications of the process have been discussed b y Rostler (6). Addition of extra soap t o latices already containing the normal "amount resulted, upon coagulation, in incorporation of relatively large amounts of the corresponding organic acids. The major

portion of the work reported herein was done with rosin-type soaps since these are commonly employed in a number of high quality general purpose synthetic rubbers. While polymerization in the presence of large quantities of soap is also reported, the major portion of the work conducted so far has been on addition of extra soap t o completed normal latices. I n these investigations, it has been found t h a t the incorporation of relatively large amounts of rosin acids into high molecular weight GR-S polymers offers interesting possibilities from the standpoint of enhancement of several desirable polymer characteristics. Acids receiving particular attention are ones whose watersoluble salts are commonly employed as emulsifiers in GR-S polymerization, or related crude products normally relatively inactive in polymerization but whose addition subsequent to polymerization is not objectionable. These fall into the general classification of rosin products which include such materials as wood rosin, disproportionated rosin, abietic acid, and rosin dimer. F a t t y acids have been included in some of the tests for comparative purposes. Water soluble soaps of these materials are converted t o the corresponding acids during the normal coagulation procedure so t h a t problems of handling, mixing, and retention in the polymer do not occur. Also, since small amounts have always been present in many types of GR-S, existing analytical methods are adequate for testing of the extended rubbers for extender content. LATEX-MASTERBATCHED PRODUCTS

For the purpose of this discussion, rosin-extended polymers prepared by addition of the soap t o GR-S latex will be termed latex masterbatches because of the analogy t o the process employed for addition of the petroleum oils, and carbon blacks. The addition of the soap to latex which has already been polymerized and stripped of residual monomers is the most convenient way of preparing the polymer-acid mixtures for study. Products made by polymerization in the presence of large amounts of soap, which are discussed later, are still masterbatches in the general sense, but should be distinguished from the stocks mentioned immediately below.

Method of Preparation. GR-S latices were polymerized a t 41" F. with different emulsifiers and at varying &tooney viscosity levels. Typical polymerization recipes are listed as Recipes I and I1 in Table I. The latices were shortstopped a t approxi-