96 



THE INDIA RUBBER WORLD 



November 1, 1920 



rubber after vulcanization. The action of these accelerators, per 

 sc, is another problem entirely. We are convinced, however, 

 that in many instances both the action and effect of organic ac- 

 celerators are dependent in great extent upon the presence of cer- 

 tain mineral substances in the mixture. 



EXPERIMENTAL PART" 



The experimental procedure was similar in all respects to that 

 of our former experiments. Essentially it differs from that of 

 Stevens only in the method of vulcanizing, and in the substitu- 

 tion of straight pieces for physical tests in place of the rings em- 

 ployed by him. The relative effects of vulcanization in steam and 

 in a platen press are commented upon later. 



Two samples of thin, pale, first latex crepe (Hevea) were 

 used. Sample 408 was from the lot of rubber that was used 

 in our former experiments, while Sample 444 was chosen from 

 another lot of equally good appearance, which was found to have 

 different chemical and physical characteristics whep employed 

 in factory mixtures. 



A partial analysis of the two rubbers gave the following 

 results. No great importanife, however, is attached to the figures 

 for total nitrogen. 



No. 408, No. 444, 



Per Cent Per Cent 



Total ash O-i 



Acetone extract .' 2.6j^ 



Total nitrogen "19 



0.29 

 2.99 

 0.20 



The extra light magnesia (specific gravity 3.4S) was the best 

 grade obtainable. It lost 4.00 per cent on ignition, after which 

 it contained 93.67 per cent MgO. Accelerator A was prepared 

 by the condensation of an amine with formaldehyde and was 

 C. P. grade. 



The conditions of milling and making physical tests were 

 identical with those previously employed, but, as comparisons 

 were made after the method of Stevens, the physical properties of 

 the different mixtures at break have been omitted. 



Sulphur determinations were made by our method"" in place 

 of that of Rosensiein-Davies." The coefficients represent the 

 combined sulphur of vulcanization expressed as a percentage of 

 the rubber in the mixture. 



Throughout the work all cures were made in a platen press of 

 the usual type. Stevens' samples were wrapped with cloth and 

 vulcanized in steam." 



Experiment I. This work consisted virtually of a repetition of 

 the previous work, using the sample of rubber (No. 408) pre- 

 viously employed, but substituting extra light magnesia for the 

 heavy calcined material used in our former experiments. The 



0.6 0.3 yo O? O.'f 0.6 



Per Cent Accelerator on Rut't'er 

 The Journal of Industrial and Engineering Chcmisiry 

 Fio. 1. Relation Between Sui-riivu Coefficients and Amount of Accel- 



EBATOR, AND BeT\VEEN AmOUNT OF ACCELERATOR AND LOAD 



Required to Effect a Given Extension 



results are given in detail in Table I,"" and the relation between 

 sulphur coefficients and amount of accelerator is shown graphi- 

 cally in Fig. 1. The relation between the amount of accelerator 

 and the load required to effect a given extension is also shown 

 in Fig. I. In both cases these results confirm previous ones, even 

 to the shape of the curves themselves. As has already been 



stated, if accelerators (such as our Accelerator A) are present 

 in the mixture even in small amount, it is evident that the load 

 required to effect a given extension is not a measure of the phy- 

 sical projKTties of the mixture, nor is it a reliable criterion of 

 the rate of cure of the vulcanized mixtures. 



Table I — Tests on Rubber No. 408 



Acceler- 

 ator 



Extra 

 light ■ 

 magnesia 



Accel 

 erator A- 



Accel- 



erator 



Per 



Cent 



Con- 

 trol 

 0.10 

 0.25 

 0.50 

 0.75 

 L 1.00 

 Con- 

 trol 

 0.10 

 0.25 

 0.50 

 0.75 

 1.00 



Sulphur 

 Coemcient 



I.oad in Hectograms 

 per Sq. Mm. to Effect Extension 



,9 > 



Actual Excess Actual Excess Actual Excess Actual Excess 



-1 to 7- 



-! to 8- 



-1 to < 



0.580 

 0.704 

 0.916 

 1.874 

 2.553 

 2.599 



0.580 

 1.356 

 1.987 

 2.925 

 3.309 

 3.603 



0.124 

 0.336 

 1.294 

 1.973 

 2.019 



0.776 

 1.407 

 2.345 

 2.729 

 3.023 



0.34 

 0.58 

 1.03 

 2.09 

 3.03 

 3.21 



0.34 

 0.88 

 1.70 

 2.14 

 2.16 

 2.30 



0.24 

 0.69 

 1.65 

 2.69 

 2.87 



0.54 

 1.36 

 1.80 

 1.82 

 1.96 



0.42 

 0.74 

 1.37 

 3.20 

 4.84 

 5.17 



0.42 

 1.10 

 2.25 

 2.77 

 2.72 

 2,98 



0.32 

 0.95 

 2.78 

 4.42 

 4.75 



0.68 



1.83 

 2.35 

 2.30 

 2.56 



0.50 

 0.94 

 1.98 

 5.11 

 8.17 

 8.92 



0.50 

 1.61 

 3.57 

 4.23 

 4.32 

 4.81 



0.44 

 1.48 

 4.61 



7.65 

 8.42 



1.11 

 3.07 

 3.73 

 3.82 

 4.31 



Experiment II. This consisted of a repetition of Experiment 

 I upon Sample 444. The results are given in Table II, and the 



I' 

 t 



i 

 § 



4 8 12 4 8 12 , 16 



Excess Hecfo<^rams per sq.mm. - 1 to9 Exten yon 



The Journal of Industrial and Engineering Chemistry 



Fig. 2. Relation Between Coefficient of Vulcanization and Load at a 

 Given Extension 



relation between sulphur coefficients and the amount of accelerator, 

 and between the amount of accelerator and the load required to 

 effect a given extension, are shown in Fig. 1. On the basis of 

 the sulphur coefficients. Accelerator A and extra light magnesia 

 appear to be of almost equal activity; or, conversely, Sample 

 444 vulcanizes at the same rate with either accelerator. When 

 judged by the load required to effect a given extension, however, 

 extra light magnesia appears to be much the more active. As it 

 is now almost .£»enerally conceded that sulphur coefficients 

 afford the most reliable indication of the state of cure, the load 

 required to effect a given extension is again seen to be unreliable 

 as an indication of the rate of cure of accelerated mixtures. 

 These results have strengthened our former opinion that unless 

 a complete series of stress-strain measurements are made, when 



>» In collaboration with Bernard .T. Shapiro. 



»The India Ruhhcr World, 61 C1920), 356. 



=iThe Chcmisl-An.-iIvst, 15 (1916). 4. 



~- Stevens in his first paper on this subject has mentioned the possibility 

 of a difference m the rate of cure uf accelerated mixtures when vulcanized 

 in dry heat. We have found similar mixtures which contained magnesia to 

 show increasingly laree sulphur coefficients when vulcanized, respectively, 

 in dry heat, pl.-'.tcn press, and open "itpani. This was not found to be true, 

 however, for mixtures which contained organic accelerators. With the latter, 

 results obtained in a platen press were frequently higher than thi-se ob- 

 tained in open steam. Evidently many organic accelerators are partially 

 soluble in, or volatile with steam and. consequently, results obtained with 

 these substances in .-ipen steam are apt to be low. This i>oinl is of con- 

 siderable importance, and accounts in large part for the difference in the 

 results obtained by Stevens and ourselves. We have found that most uni- 

 form results are obtained with samples encased in molds and vulcanized in 



open steam. . . . t, ,,•,..,- 



"= The results obtained and rhown in this table are somewhat higher than 

 those ariginallv finnd, and which were given in Table I of our former 

 paper. This discrepancy may be attributed to a change in the rubber 

 itself, as the sample used in 



the 



,-.- - present instance had aged for over one 



year'in"roll form in a partially broken-down condition before the experiment 

 was repeated. Also, the method previously employed for the estimation of 

 combined sulphur was found to give slightly low results for mixtures which 

 contained but small amounts of miner.al substances. 



