Aids in Vulcanization of LigninNatural Rubber Co LEAD, COPPER, AND BISMUTH OXIDES T. R. GRIFFITH
D. W. MAcGREGOR
National Research Council, Ottawa, Ontario, Can.
Howard S m i t h Paper Mills, Cornwall, Ontario, Can.
4'
HE first formulation tried in this compounding study of acid was counteracted by the addition of sodium hydroxide during the oxidation. This solution was then added to the lignin-natural rubber coprecipitates is shown in Table I. latex for coprecipitation. Other methods of preparation of This formulation, No. 1 in the table, would not cure. Lengthenoxidized lignin are possible as shown by Raff and Tomlinson (11). ing the time of cure or raising the temperature did not help. The Since it is known that lignin =-hen heated with sulfur reacts acceleration was of the conventional type and consisted of a readily to form hydrogen sulfide, it was thought that this reaccombination of benzothiazyl disulfide (1.0 part per 100 parts of tivity might explain the delayed vulcanization rate, rubber) and diphenylguanidine (0.4 part per 100 parts of rubber). This behavior was considered surprising, since this accelerator Booth and Beaver ($2) showed in 1940 t h a t rubber dissolves 1% ' of hydrogen sulfide a t room temperature and t h a t this amount cornbination is relatively powerful and under certain circumof gas is sufficient to retard the rate of vulcanization of all types stances, scorchy. of accelerators. A considerable number of formulations not specifically recorded in this paper were then tried, with little success. InThe formation of hydrogen sulfide during the vulcanization of rubber with sulfur has been observed by many investigators creasing accelerator dosage helped but little. Test slabs were including Jaworonok (8) and Okita ( I O ) . uncured or undercured, logy, often tacky, with blisters, and could be removed from the mold only with difficulty. IncreasFisher (7) proposed a theory in 1939 to the effect that hydrogen sulfide is a necessary intermediate in vulcanization. His posing the sulfur dosage did not improve matters appreciably, and tulated mechanism of vulcanization is given in Figure 1. This such a procedure was of doubtful utility because of the known diagram shows hydrogen sulfide, formed by reaction of sulfur poor aging characteristics of compounds with high sulfur. with the hydrocarbon, combining with rubber a t the double Unpublished data on hand give evidence of similar difficulty bonds after splitting up into hydrogen and sulfhydryl. Further experienced by other experimenters in this field. reaction with sulfur with evolution of hydrogen sulfide leads t o Keilen, Dougherty, and Cook (9) reported results obtained the cross-linking effect necessary for vulcanization. Since with coprecipitates of sulfate lignin with natural, nitrile, and hydrogen sulfide is evolved, it is possible that a high concentraneoprene rubbers. Relatively low sulfur with high accelerator tion of hydrogen sulfide in the medium might slow down the vuldosage ( 3 parts of mercaptobeneothiazole plus 1.5 parts of tetracanization reaction through the application of the law of mass methylthiuram monosulfide per 100 parts of rubber) was used action. in the natural rubber formulations. These investigators obtained The chemistry of vulcanization has, of course, progressed cona good tensile and modulus, and stated t h a t a more intensive siderably since 1939. Notably, Farmer (6) brought forth evisearch and testing of different curing agents could result in general dence in favor of the theory that sulfur reacts a t the a-methylimprovement of properties. enic carbon atom in rubber rather than a t the double bond. No delay in the vulcanization rate was observed 11 hen lignin SinceFarmer's theory does not explain the saturation of the double was added to rubber as a dry powder, but as is well known, this bonds occurring during vulcanization, Bloomfield ( I ) showed method of addition results in relatively low tensiles. It was when how this saturation could occur through various possible radical lignin was incorporated into rubber by the coprecipitation process reactions, inchding the addition of hydrogen sulfide to the double t h a t the delayed vulcanization rate was observed. bond. The coprecipitation process consisted in first mixing together More recent work on the chemistry of vulcanization by Craig, ammonia-preserved natural rubber latex and an alkaline aqueous solution of lignin of p H about 11 in the desired proportion. This Juve, and Davidson (5),Craig, Davidson, Juve, and Geib ( d ) , and Craig, Davidson, and Juve (S) brings out the point that mixture was then added to a coagulant consisting of a 50-50 hydrogen sulfide is a necessary intermediate in vulcanization but mixture of sodium chloride and sulfuric acid in aqueous solution. The mixture of latex and coagulant was maintained at a temperature of 90" C. during the addition. Further GH3 1 CH. coagulant was added at the same time as the latex to maintain the pH at approximately 2. The coprecipitate thus obtained was filtered, washed, and dried. Oxidized alkali lignin was used throughout. The oxidized lignin for these experiments was prepared by heating a solution of the sodium lignin CH3 CH3 J CH3 salt in water, and bubbling air through this solution. The tendency to turn Figure 1. Fisher's Postulated Mechanism of Sulfur Vulcanization 380
1
INDUSTRIAL AND ENGINEERING CHEMISTRY
February 1953
381
,5000
4000
9
4 X ICI
z 300C
Iv)
W
=I
in
W z
c 2000
Figure 2. Tensile Strength Bar Graphs Showing Effect of Lead, Copper, and Bismuth Oxides in Lignin-Natural Rubber Coprecipitate Formulation 6: 27.8 volumes of lignin, 3 parts of zinc oxide and 7 parts of other oxide per 100 parts of rubber
that when present in excessive amounts i t delays vulcanization by acting as a chain terminator. They also regard zinc oxide as a buffer that Berves t o keep the concentration of hydrogen sulfide down t o a level a t which vulcanization is not delayed. However, zinc oxide is ineffective in the presence of coprecipitated lignin. The present work calls attention to the existence of metallic oxides-namely, the oxides of lead, copper, and bismuth-which are the only ones that can control in the proper manner the concentration of hydrogen sulfide during the vulcanization of ligninnatural rubber coprecipitates. COMPOUNDING EXPERIMENTS
TESTMETHODS. ASTM procedure in the preparation of the rubber samples and in the carrying out of the physical tests was followed. FORMULATIONS. Table I shows results obtained with a few of the formulations tried in the development of a lignin-natural rubber compound which would cure satisfactorily. Formulation 1 has already been mentioned. It contains conventional acceleration, which did not cure in the presence of coprecipitated lignin. TABLE I.
The simple addition of 7 parts of litharge per 100 parts of rubber to formulation 1 brought about a remarkable improvement in the result (formulation 2). As shown by the modulus a t 300% elongation, a satisfactory cure was obtained. I n formulation 3 it is shown t h a t 7 parts of litharge per 100 parts of rubber without zinc oxide cured equally well, although the tensile strength was lower by about 400 pounds per square inch a t the longest cure (75 minutes a t 292' F.) than the formulation with both zinc oxide and litharge. I n formulation 4 the 7 parts of litharge per 100 parts of rubber used in formulation 3 was increased to 10 parts. The principal effect was a slight increase in modulus and tensile strength a t the 75-minute cure. I n formulation 5 the acceleration was changed to zinc dimethyldithiocarbamate. While this did not increase the modulus, both tensile strength and elongation a t break were considerably improved. Although all possible acceleration combinations were not tried, i t is evident that the thiocarbamate type of acceleration has a specific beneficial effect. After much experimentation it was decided t o retain the zinc oxide and thiazole types of acceleration in the formulation in addition to litharge and thiocarbamate acceleration, as shown in formulation 6. Apparently zinc oxide
FORMULATIONS FOR INITIAL EXPERIMENTS AND RESULTS OBTAINED Formulation, Partd
Natural rubber Lignin, 38.5 vols. (coprecipitated) Stearic acid Zinc oxide Lead oxide Benzothioayl disulfide Zinc dimethyldithiocarbamate Diphenylguanidine Sulfur
Press cures a t 292' F. 25 min. 45 min. 75 min.
2
1 100 50 2 3
i .'o 0.4
No cure No cure No cure
Stress at 300%
..
7 1.0
Tensile
960 1010 1040
2050 1740 1660
3' 6
510 420 370
Stress at
300% 1170 1080 1170
Tensile
2190 1870 1265
io
1.0
0:4 2.0 Elongation,
5 100 50 2
4 100 50 ..2 10
50 2
0.4 2.0
2.0
.
3 100
100 60 2 3 7 1.0
1:o
0:i
2.0 Elongation,
% 480 390 320
Stress
322% 1075 1280 1475
2:0
ElanTen- gation, sile % 2070 430
1690 1690
370 320
Stress at 300%
Tensile
Elongation,
1175 1125 1050
4170 3920
% 630 570
3550
510
382
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 45, No. 2
The other oxides capable of bringing about rapid vulcanization were copper oxide, two other oxides of lead (Pb,O, and PbO,), and bismuth trioxide (BizO?). Lead peroxide (PbQ2)was undercured in 10 minutes a t 282" F., but P the longer cures were Satisfactory. 1-ARYING THE QUANTITY OF OXIDE. Figure 3 shows the effect of varying the quantity of oxide. The tensile results for 1, 3 , 5 , 7, and 9 parts of oxide per 100 parts of rubber are shown. Forniulation 7 , Table 111, which is the mine AS formulation 6 so far as zinc oxide, accelerator, and sulfur dosage are concerned, was used for this comparison. Even 1 part of speLi:il oxide per 100 parts of ruhber was ert'wtive. Copper Figure 3. Effect on Tensile a t Best Cure of Varying Quantities of Oxide in oxide was effective in the lowest conLignin-Natural Rubber Coprecipitate centration, since with only 1 part per Formulation 7: 27.8 volumes of lignin 100 pa1ts of rubber a tensile strength A l l formulations contain 3 parts of zinc oxide per 100 parts of rubber of 4100 pounds per square inch na3 rcnched. The highest ;ensile strcmgth carries out its normal function in bringing out the best physical (4900 pounds per square inch) T a s reached with 7 parts of characteristics of the rubber. litharge per 100 parts of rubber. Xot much seemed to be gained by the use of quantities of oxide over 7 parts per 100 SEARCHFOR OTHEROXIDES. The metals whose oxides were parts of rubber. effective in bringing about rapid vulcanization of lignin-natural Certain organic accelerators, such as the lead, copper, and bisrubber coprecipitates were, as stated, lead, copper, and bismuth. muth salts of dithiocarbamic acid derivatives, would provide It is thought t h a t these three metals are distinguishable from zinc in the solubility of their sulfides. These three metals have a source of the three elements shown in the present study t o be desirable for the vulcanization of lignin-natural rubber coprecipsulfides which are insoluble in dilute acids and in sodium and itates. These accelerators are available cornmorcially, and when ammonium polysulfide. Zinc sulfide is soluble in dilute acid present in rubber in normal dosage-i.e., 1 part per 100 parts and belongs in a different group in the qualitative analysis scheme. of rubber-would provide the equivalent of about 0.4 part, of hisThis suggests that the three metals effective in the vulcanization mutb oxide and about 0.2 part of copper or zinc oxide per IO0 of lignin-natural rubber coprecipitates may be characterized parts of rubber. These quantities are, however, Ion-er thnn thc by the somewhat greater insolubility of their aulfides. optimum amount. for satisfiictorj. vulcanization. Many other oxides were tried without success, including those of calcium, magnesium, barium, cadmium, cobalt, niaganese, nickel, iron, and others. The fact that such oxides as those of calcium and magnesium are ineffective shows t h a t the delay in TABLE 111. FORMULATIOS 7, FOR DETERMINIKG THE ],I 11,C'l' Or the rate of vulcanization of lignin coprecipitates is not due t o VARYING QU.ISTITIES O F OXIDES (PI R T S ) residual acidity from the coprecipitation, since such oxides v-odd h-atural rubber 72.2 J.ignin, 27.8 vol. (eoprecipitatcd) 36. 1 be expected t o neutralize acidity. Smokcd sheets 27. 8 COMPARISON OF OXIDES. Figure 2 shows a comparison of the 2 Stearic acid Zinc oxide 3 oxides found satisfactory t o date. The compounds compared Zinc dimethyldithiorarhaniate 1 N-cyclohexyl-2-benzothiazolesulfenamide 0 5 were made up according to formulation 6, Table 11, and all Sulfur 2 contained zinc oxide. The acceleration was a combination Other metallic oxides ( 8 9 ahown in bar graphs) 1.3,>,7,0 a .2dded t o mako rubber equal 100. of I\'-cyclohexyl-2-benzothiazole sulfenamide and zinc dimethyldithiocarbamate. Special oxide, added for rapid vulcanization, was used in the proportion of 7 parts per 100 parts of rubber. AGIR G
TABLE 11. FORMULATION 6, FOR COMPARISON OF OXIDES ( P A R T S )
a
Natural rubber Lignin, 27.8 vol. (coprecipitated) Smoked sheeta Stearic acid Zinc oxide Zinc dimethyldithiocarbamate A~-cyclohexy1-2-beneothiazole sulfenamide Sulfur Other metallic oxides Added t o make rubber equal 100.
72.2 36.1 27.8 2 3 1 0.5 2
7
Zinc oxide, used alone as a control, showed a relatively low tensile strength of around 2000 pounds pel\ square inch. T h e modulus, not shov-n in the figure, indicated a very poor cure. The highest tensile figure shown in this comparison was 4800 pounds per square inch given by litharge.
The results of oven aging ( 2 days a t 90" C.) are given in Figurp 4. T h e rubber compounds were prepared according t o formulations 8 to 12, inclusive, Table IV, and all contained 3 parts of zinc oxide per 100 parts of rubber with the exception of formulation 9, the lignin control, which contained 5 parts per 100 parts of rubber. The three special oxides tested in the lignin compounds were lead oxide, cupric oxide, and bismuth trioxide. The lignin control compound, just mentioned, contained only zinc oxide ( 5 parts per 100 parts of rubber). For purposes of comparison n channel black tread formulation was placed in this series and had conventional acceleration-i.e., A~-cyclohexgl-2-benzothiazole sulfenamide-with 2.5 parts of sulfur per 100 parts of rubber. Thc lignin compounds had the same thiazole and dithiocarbamate acceleration as previous formulations, with 2 parts of sulfur and 7 parts of the various oxides per 100 parts of rubber as shown. The cures mere 10, 20, 30, and 40 minutes at 282" F. Since
INDUSTRIAL AND ENGINEERING CHEMISTRY
February 1953
TABLEIV.
Natural rubber Lignin, 27.8 vol. (coprecipitated) Smoked sheet M P C black, 27.8 vol. Stearia acid sym-di-@-Naphthyl-p-phenylenediamine Pine tar Zinc oxide Lead oxide Cupric oxide Bismuth trioxide Zinc dimethyldithiocarbamate N-cyclohexyl-2-benzothiazole sulfcnamide Sulfur
7l/z min. 10 min. 20 min. 30 min. 40 min.
FORMULATIONS FOR AGINGEXPERIMENTS AND RESULTS
-
8
_1Z-
I1
...
72.2
... 100
36.1 27.w
36.1 27.80
36.1 27. Sa
2
2
'2'
2
2
2
2
...
3 7
3
...
...
2 4 3
...5
...
... ...
...
... ...
2380 4200 4190 3990
...
..
540 640 560 520
'io
65 68 68
30s 465 510 575
720 710 650
0.5 2
52 55 57
595 1035 1210 1226
3480 4630 4670 4440
36.1 2 7 . 8a
... 2
..
3
... 1
1
if40 $9'0 bi
2235 2540 2520
72.2
.7'
...
0.5 2
72.2
...
3.-
1
0.9 2
..goo
__
Formulation, Parts 10 72.2
9
50 2
1635 1860 1840
383
'
1
0.5 2
0.5 2
f40
680 670 660
55 65
66
67
i'
405 705 765 810
2335 4340 4240 3976
680 740 720 700
50 57 59 60
970 io70 1085 1070
3725 3910 3580 3440
640 640 620 600
60 61
. . . . .
735 850 855 850
4000 4105 3780 3710
740 690 670 680
64 66 68 68
iiio
3420 3500 3560 3400
iio
6s 69 70 70
After Aging 2 Days in Oven at 90' C. 71/2 min. 1725 3775 10 min. 2210 3780 20 min. 2560 3520 30 min. 2305 3185 40 min. a Added t o make rubber equal 100
570 510 450 380
66
72 74 74
7% 740 710 680
2320 2120 1910 1870
6'30 560 550 550
6s
59 62 61
the cupric oxide compound was faster in curing than the others, a shorter cure of 7.5 minutes was added and the 40-minute cure eliminated. All the compounds contained 2 parts of symmetrical dip-naphthyl-p-phenylenediamine antioxidant per 100 parts of rubber, since it was felt that an antioxidant is necessary for satisfactory aging. The lignin compound, with zinc oxide only, had a relatively low tensile strength of around 2400 pounds per square inch before aging, which dropped to around 1900 pounds per square inch for thp longer cures, after aging. A t thc short& cure all the compounds, except the lignin compound containing bismuth trioxide, increased in tensile strength after aging. The litharge compound reached the highest tensile strength before and after aging. The unaged figure a t the 30-minute cure of the litharge compound, for example, was around 4700 pounds per square inch, and the aged figure at the same cure was 4100 pounds per square inch. The best unaged and aged tensile figures for the channel black compound (20-minute cure) were 4200 and 3500 pounds per square inch, respectively. From the point of view of drop in tensile strength as well as actual tensile strength after aging, the litharge compound aged better than both the carbon black and the copper compound. The bismuth compound had a somewhat lower unaged tensile strength than the channel black compound and the other lignin compounds containing the special oxides. However, with bismuth oxide in the formulation, the tensile strength was maintained a t a high figure after aging, since the aged tensile a t the 30- and 40-minute cure of the bismuth compound was better than t h a t for the channel black compound a t the same cures. Upon the basis of percentage drop in tensile strength upon aging, at t h e longer cures, the bismuth compound appears t o have shown the best aging characteristics of all the compounds tried. The longest three cures of the cupric oxide compound showed noticeably better aged figures than the longest three channel black cures, and the cupric oxide unaged figures were either better than or equal t o the corresponding channel black figures. The
lis5 1745 1865 1890
4i60 4470 4095 3890
&io 66
560 520 500
69
70 71
..
58 62
1235 1260 1275
570 580 570
maximum drop in tensile strength of the cupric oxide compound upon aging was around 1501, while t h a t of the channel black compound was around 20%. Other physical characteristics of the copper compound, such as elongation a t break, stress a t 300y0 elongation, and hardness, showed equally good aging (Table IV). It is realized t h a t copper is generally considered t o be harmful to the aging of rubber, particularly natural rubber. I n view of this, the aging experiments described above, were repeated, The repetition confirmed the results obtained. Villain ( l a ) found that among those materials t h a t protect rubber against rapid aging because of copper are the accelerator and the antioxidant used in the lignin formulations in the present aging study. He also found t h a t a typical copper inhibitor, disalycylalethylenediarnine, needs an antioxidant, not necessarily a copper inhibitor, to bring about the maximum protective effect. Lignin-natural rubber masterbatch requires an antioxidant in the presence of, and in the absence of copper, but this antioxidant may be aldol a-naphthylamine which, according to Villain, does not have a specific anticopper effect. For example, formulation 11,Table IV, used in the aging tests, containing copper and dithiocarbamate, would have aged just as well if aldol anaphthylamine had been substituted for the copper-inhibiting 8 ymdi-P-naphthyl-p-phenylenediamine. Villain used the equivalent of 0.02 and 0.0670 of copper, calculated on the rubber, and observed serious aging in the absence of protection against copper, whereas in the aging tests described herein, the copper equivalent of the copper oxide was 5.6y0 on the rubber, or about 100 times the quantity used by Villain. The high copper content and the satisfactory aging of the lignin formulations indicate strongly t h a t lignin is a copper inhibitor. However, a study of the possible anticopper properties of lignin does not come within the scope of the present investigation. I n view of the limited number of aging tests performed, it is probably not wise t o recommend indiscriminate use of copper in lignin coprecipitate formulations. Nevertheless, in circumstances where contamination of rubber with copper during manufacture or in service has to be guarded against, it is thought t h a t a lignin-coprecipitate formulation might be given consideration.
9
384
INDUSTRIAL AND ENGINEERING CHEMISTRY
PbO
W A G E D , AGE0
4 500
-4
-
Vol. 45, No. 2
t ZnO CuOfZnO
ZnO
ONLY
3500
P
J2 c a
z w
E
o t-
Y
Z
w z
ZnO
ONLY
2500
b-
I500
J R E T I M E , MINUTES F I L L E R , 27.8 VOLS
1
LIGNIN
I
I
I
I
I
Figure 4. Comparison of Oven-Aged Lignin-Natural Rubber Coprecipitate with Channel Black Tread Compound Formulations 8 through 12 A l l formulations contain zinc oxide; other oxides where present are in amounts of 7 parts per IO0 parts of rubber
COMPARISON WITH CHANNEL BLACK
TENSILE STRENGTH. Figure 5 shows the result of a comparison of channel black with coprecipitated lignin a t various loadings on the basis of tensile strength a t best tensile cure. Formulations 13 to 20, inclusive, Table V, were used. The channel black compounds had conventional thiazole acceleration. The special oxide in the lignin formulations in addition t o zinc oxide was litharge, which was added in the proportion of 5, 7 , 10.5, and 17.5 parts by weight for lignin loadings of 19.2, 38.5,
57.7, and 77.0 volumes, respectively. The lignin formulations contained both thiocarbamate and thiazole acceleration. The thiocarbamate acceleration was increased for the higher Ioadingsi.e., from 1 part, which was the dosage at the lower loadings-to 1.5 parts per 100 parts of rubber for the 57.7-volume loading and to 2 parts per 100 parts of rubber for the 77-volume loading. The tensile strength of the lignin compound was considerably greater than that of the channel black compound a t all loadings. The greatest difference was a t the 57.7-volume loading where the tensile strength of the lignin compound was 4000 pounds per square inch and t h a t of the channel black compound was 2400 pounds per square inch.
5000 TABLE
V.
FORMULATIONS FOR COMPARISON
O F COPRECIPITATED
LIGSINWITH CHAXNEL BLACK AT VARIOUSLOADISGS 13
;
4000
r 0
W
+-v) W
1
3000 3
c W
3
Smoked sheet M P C black Stearic Pine taraoid Zinc oxide N-cyclohexyl-2-benzothiazole sulfenamide Sulfur Filler volume loading Best tensile cure, min. a t 282'
100
34.6
?
5
F.
0
-
E 0
6
x -
2000
0
Figure 5. Comparison of Tensile at Best Cure of Lignin and Channel Black i n Natural Rubber at Equal Volume Loadings Formulations 13 through 20
0.7 2 19.2 40
Natuial rubber Lignin (ooprecipitated) Stearic acid Zinc oxide Lead oxide Zinc dimethyldithiocarbamate N-ovclohexvl-2-benzothiaaole sdfenamide Sulfur Filler volume loading Best tensile cure, min. a t 282' F.
17 100 25 2 3
5
1
0.5 2 19.2 20
Formulation, Parts 14 15 100 69.3 loo 104 2 2 5 5 5
0 7 2 38.5 40
5
0.7 2 57.7 60
Fsrmulation, Parts 18 19 100 100 50 75 2 3
7
1
0.5 2 38.5
30
2
__ 16 100 138.5
2
?
0 7 2 77.0 40
~-
20
100 100
3 10.5 1.5
2 3 17.5 2
0.5
0.5
57.7
77.0 20
2
20
2
INDUSTRIAL AND ENGINEERING CHEMISTRY
February 1953
385
7 00
90
s z E ! t4
z
(3
3
7 0
300
W
50
I
2000
BASHORE RESILIENCE
.-
ia
v)
50
v) W
a Iv)
IO00 I
30
t
I
VOLS. 19.2
VOLS. 19.2
38.5
57.7
77.0
Figure 6. Shore Hardness and Bashore Resilience of Lignin and Channel Blaolr Compounds a t Best Tensile Cure a t Various Loadings
I
I
I
38.5
57.7
77.0
I
Figure 7. Elongation and Modulus of Lignin and Channel Black Compounds a t Best Tensile Cure Compared at Equal Loadings Formulations 13 through 20 Lignin formulations contain lead oxide
Lignin formulations contain lead oxide Formulations 13 through 20
HARDNESS AND RESILIENCE. Figure 6 shows a comparison of lignin and channel black on the basis of Shore hardness and Bashore resilience a t the best tensile cure. The formulations were those just described-i.e., 13 t o 20, inclusive, Table V-and the tests were carried out on the same slabs as were used for the determination of tensile strength. For equal volume loadings the channel black compound is harder than the lignin compound, and the lignin compound is more resilient than the channel black compound. The curves show t h a t lignin is more resilient than channel black at equal hardness as well as a t equal volume loading. At a hardness of 70, for example, the resilience of lignin is about 38, whereas at this hardness channel black has a resilience of only 28. ELONGATION AND MODULUS AT 3oOy0. Elongation a t break and modulus at 3oOy0 were also observed for compounds made from formulations 13 to 20, inclusive. Figures obtained a t the best tensile cure are shown in Figure 7. The lignin compounds had a considerably higher elongation at break than the channel black compounds a t equal loadings and the channel black a considerably higher modulus at 300% elongation.
The compound containing litharge was the most scorchy, but if considered necessary, this could be corrected by adjustment of accelerator or special oxide.
TABLEVI.
MOONEYSCORCH VALUESON LIGNIN-NATURAL RUBBERCOPRECIPITATE (FORMULATION 6) Metallic Oxide ZnO only ZnO ZnO ZnO
+ PCuO bO + + BisOs
Mooney'Scorch, Min. to 4-Point Rise, M S Units, 2.50" F. 31 16
35 35
COST OF OXIDES
The prohibitive price of bismuth oxide at around $4.50 per pound should be noted. Litharge costs around 22 cents per pound, zinc oxide around 20 cents. Copper oxide is still competitive at around 40 cents per pound because it is effective in lower dosages. These are rough quotations taken from the current literature. CONCLUSION
MOONEY SCORCH
Table VI shows the Mooney scorch figures for the various oxides compounded according to formulation 6. These compounds had 3 parts of zinc oxide plus 7 parts of additional special oxide per 100 parts of rubber, with the exception of the control, which had only 3 parts of zinc oxide per 100 parts of rubber. It is interesting t o note t h a t the control, which did not cure too well, was slightly more scorchy b y a matter of a few minutes than the compounds containing zinc oxide plus cupric oxide, and zinc oxide plus bismuth oxide.
Although the experiments recorded herein are strictly of a compounding nature, nevertheless i t is believed that they have made contributions t o the knowledge of the chemistry of vulcanization. The most probable explanation of the delaying effect of lignin upon vulcanization is that it reacts too readily with sulfur during vulcanization and thus builds up a n excessive hydrogen sulfide concentration. There are metals whose oxides can control t h e hydrogen sulfide concentration t o the proper degree during the vulcanization of lignin-natural rubber copredpitates; such oxides are those of lead, copper, and bismuth.
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
386
Vol. 45, No. 2
ACKNOWLEDGibIENT
(3) Craig, D.. Daviclsoii, IT-.L., and Juve, -4.. E., J . Folj/me/, S c i . , 6,
The authors are indebted to G. H.Tomlinson 11, director of research, Howard Smith Paper Mills, Cornwall, Ontario, for valuable suggestions and for permission to publish this paper. The help of C. H. Rimmer and L. A. Baker with experimental work, including the preparation of coprecipitate, carried out a t Cornwall, is gratefully acknowledged. The authors wish to thank C. M. Barker for assistance in compounding and testing in the National Research Council rubber laboratory, Ottawa, Ontario.
(4) Craig, D., Davidson, W. L , Juve, -4. E., and Geib, I. G., Ibid.. 6 , 1, 7 (1951). ( 5 ) Craig. D., Juve, A. E , and Davidson. W. L.. Ihid.. 5 , 709 (1950) ( 6 ) Farmer, E. H., T r a n s . Faraday Soc., 38, 340-8 (1942).
177 (1951).
(7)Fisher. H. L.. IND.ENG.CHEM..31. 1381-9 (1939) i s ) Jaworonok, S. G., J. A p p l i e d Chem. (Lr.S.S.R.), 9, S o . 7 , 1290-8 11936).
(9) Keilen, J. J.,D'ougherty, W, K., and Cook, W. It., IND. EXG. CHEV.,44, 163-7 (1952). (10) Okita, Tadao, J . SOC.Rubhe? I d . Jtrpan, 13, No. 9, 731-8 (1940). (11)
Raff, R. A. V.. and Tonilinson 11, G. H., C a n . ,I. R e s e c o c h , F27, 399-418 (1949).
LITERATURE CITED
(12) Villain, Henri, Rubber Chern. h Technol., 23, 352.-61 (1950).
(1) Bloomfield, 6. F., India-Rubber J., 111, Nos. 9, 10 (1946). ( 2 ) Booth, E. W., and Beaver, D. J., IND. ENG.CHEM.,32, 1008-8
(1940).
RECEIVED for review June 3, 1952. ACCEPTED September 8. 19>2. Presented before the Dirision of Ri1bh.r Cheinistrj. of t h e A M E R I C .C~~~r i i h i ICAL
SOCIETY, Cincinnati, Ohio. 19.52.
roducts of Oxidation of an Olefin Structurally Related to GR-S G. R . YLITCHELL' AND J. REID SHELTON Case I n s t i t u t e of Technology, Cleveland 6 , Ohio
T""
fields of rubber, plastics, surface coatings, and petroleum include a considerable portion of the industrial activity of the world, and in each of these industries oxidation is an important factor in product processing and in the properties of the final product. Information regarding the nature of the oxidation reactions observed in any of these fields can be applied in part to the others. Because of the complexity of the materials involved, i t is frequently desirable to study pattern molecules of comparable structure and to seek to apply the knowledge obtained in this way to the more complex systems. This investigation is a part of a research program on the nature of the oxidation of natural and synthetic rubber and related materials. It is confined to the study of an olefin, 5-phenyl-Bpentene, which represents one of the possible repeating structural units of GR-S rubber. There is considerable information in the literatuie regarding the probable mechanism of oxidation of hydrocarbons, both saturated and unsaturated. The oxidation of olefins, for example, has been studied extensively by vorkers in the laboratory of the British Rubber Producers' Research Association ( 1 , 3, 4) and the folloming sequence of reactions for either thermal or photo-oxidation with molecular oxygen was proposed : 3
2ROOH ---+ R . , Ron. R* 0 2 ---+ Ron. R H +ROOH R. ROL. 2RO2. --+ stable products
++
+
Tobolsky, Metz, and Mesrobian (16)have shown that reaction kinetics deduced from the above mechanism fit the data for the oxidation of dimethylcyclohexane, a saturated hydrocarbon, with molecular oxygen at 75' C. and no added catalyst. I n addition t o these primary reactions, there are a host of secondary reactions leading t o cross linking, chain scission, and the formation of various oxygenated functional groups. Some of the ultimate products formed have been characterized for specific compounds and possible chemical mechanisms have been proposed (5), b u t much more information will be required t o establish the true nature of these reactions. 1
Present address, Olin Industries, Inc , New Haven, C o n n
The object of the present study n a s to determine some of the products of the oxidation of a n olefin structurdly related t o GR-S and to seek t o interpret the data in terms of possible reactions involveti. This paper is the third in a sequence of btudics on this general subject ( 6 , l . d ) . E X P E R I M E N T l L PROCEDURES
MATERIALS.The 5-phenyl-2-pentene was prepared by the method of Lawrence and Shelton (6) p i n g y-plienylprop,vl bromide as the starting material. The final step involved thr dehydrohalogenation of 5-phenyl-2-bromopentane. The olefin was purified by fractionation and it analyzed 97 to 9970 pure, based on unsaturation present. However, infrared data obtained in subsequent analysis of initial and oxidized samples (from a preliminary run a t 90" C. on a sample of 90% purity) showed the presence of some terminal unsaturation. Thus the isomer, 5-phenyl-1-pentene was also present, and the relative amount, based on the infrared data, may have been as much as 207,. (While this complicates the interpietation of the data, the mivture actually corresponds more closely to GR-S, since the isomer contains a terminal double bond like t h a t of the sidechain vinyl groups formed to the extent of about 20% by 1,2polgnierization of butadiene.) Oxidations were carried out a t two temperatures on samples of the following purities, based on unsaturation: Oxidat,kon C.
Temp.,
80 100
Piienylpenteiie P u r i t y , 70 97,s 99.0
The apparatus and procedure APPARATUS A N D PROCEDURE. were the same as used by Warner and Shelton ( 1 4 ) . Oxygen w-as recycled through the liquid olefin a t constant temperature (f0.3' C.) and atmospheric pressure. Water and carbon dioxide were absorbed continuously, and the volume of oxygen absorbed was measured periodically. Samples Were removed for analysis at various stages of oxidation and stored under nitrogen a t 0" C. until the analyses were completed. ANALYTICAL METHODS. Peroxide was determined by the method of Wagner, Smith, and Peters ( 1 3 ) . The mcthod is based on the liberation of bromine from potassium bromide and titration with sodium t h i o d f a t c .
