Vol. 15, No. 4
INDUSTRIAL AND ENGINEERING CHEMISTRY
256 r
formation, since only the latter compound produced a color TABLE IV. COMPARISON OF GRAVIMETRIC ALOCHOL-INSOLUBLE approaching the characteristic iodo-blue. Improved characMETHOD WITH PHOTOMETRIC METHOD teristics of this curve could have been obtained by using a more suitable wave length for examination of the test soluMoisture From alcohol-Active From hotelSample (Xylene Alcohol inaoluble method ometer ?I filter, tions. This method of analysis may be suitable for other No. Distillation) Insoluble b y difference sample B alkyl aryl sulfonates. % % % % 1 2
3
5.0 4.0 6.0
59.25 59.07 60.05
35.75 36.93 83.95
41.0 39.6 38.6
centage of active ingredient for each of the samples was calculated from the concentration of the sample, and reference to the photelometer curve for sample B, Figure 1. This information is shown in Table IV. It is obvious from this that the photometric method gives more nearly correct results than does the alcohol-insoluble method. There was a possibility that certain other alkyl aryl sulfonates might be amenable to this method of estimation. Tests were made with two alkylhydroxydiphenyl sulfonates and wbutyldiphenyl sodium sulfonate (Table 111). Apparently the hydroxy group interferes with this particular color
Acknowledgment The constructive suggestions made by Carroll A. Hochwalt and Ross W. Moshier have been of great assistance in the preparation of this paper.
Literature Cited (1) Am. SOC.Testing Materials,
“Sampling and Chemical Analysis of Soaps and Soap Products”, A. S. T. M. D-460-41,Section 11 (1941). (2) Assoc. Official Agr. Chem., Official and Tentative Methods of I
Analysis, p. 73 (1940). (3) Biffen, F. M.,and Snell, F. D., IND.E m . CHEM.,ANAL.ED., 7, 234 (1935). (4) Hart, Ralph, Ibid., 5, 413 (1933). (5)Ibid., 11, 33-4 (1939). (6) Scales, F. M., and Kemp, Muriel, Assoc. Bull. (Intern. ASEOC. Milk Dealers), 31, 187-208 (1939).
Identification of Carbon Black by Surface Area Measurements F. H. AMON, W. R. SMITH, AND F. S. THORNHILL, Godfrey L. Cabot, Inc., Boston, Mass.
*P
-
ARKINSON (6) has suggested that the particle size of lampblack recovered from a tire tread stock by means of nitric acid is unchanged from that of the original supply. His conclusions were drawn from sedimentation experiments. The authors have found (8) that the low-temperature nitrogen-adsorption method of Emmett and his co-workers (1, 2, 3) for determining surface areas of finely divided substances is a comparatively rapid and accurate method of measuring the total surface of commercial carbon blacks. The object of the present investigation was to expand Parkinson’s observation to a variety of commercial carbon blacks, employing the nitrogen-adsorption technique. Since the surface area of most of the standard carbon blacks has been previously determined (3, 8 ) , a method of identifying the carbon black present in an unknown rubber stock can be readily devised if it is first established that the carbon black can be recovered from a rubber stock with unchanged surface area. The present paper reports the surface area values determined by the nitrogen-adsorption method for a number of commercial carbon blacks recovered from vulcanized tread stocks. Experimental Details The technique and apparatus for determining the surface area of carbon black by the low-temperature nitrogen-adsorption isotherm have been described in a previous publication (8). An oxygen thermometer is now used t o measure the temperature of the liquid nitrogen bath during a run. From this temperature, the a propriate saturation pressure, pp, of the adsorbate nitrogen is ogtained and employed in calculating the final surface area value (5)*
$The carbon blacks studied were compounded in the following recipe: Ingredient Smoked sheet Zinc oxide Sulfur
Pine tar
Parts by Weight 100 5
3 3
Ingredient Captax Stearic acid Agerite Hypar Black
Parts by Weight 0.9 4.0 1.0
45
One stock was made from Grade 6 carbon black in Buna S: Buna S Zinc oxide Sulfur Pine tar B-L-E (antioxidant) B-J-F (accelerator) Laurex Grade 6 black
100 5 2 2 1.0 1.5 2 50
The smoked sheet stocks were cured for 30, 60, and 90 minutes at 134’ C., and the Buna S stock at 144’ C. The 60-minute cure was nearly optimum in the majority of cases, and was the only cure selected for study. The following commercial carbon blacks were selected for study: GASTEX,a nonimpingement-type black manufactured from . natural gas. A semireinforcing black of the “soft” type widely
used in rubber goods. ACETYLENEBLACK(Shawinigan), pre ared by thermal decomposition of acetylene. A soft-type blac! which does not possess the marked rubber-reinforcing properties of channel blacks. Its electrical conductance in rubber stocks is very high. SPHERON, grades 9 through 1, a series of rubber-reinforcing carbon blacks prepared by impinging natural gas flames on metal surfaces. Their essential difference is particle size, grade 9 being the coarsest and grade 1 the finest of these typical rubber blacks. SPHERON N, a channel black of fine particle size with superior electrical conducting properties when compounded in rubber stocks.
The free carbon in the vulcanized rubber stocks was separated by means of nitric acid. The procedure followed was essentially that of Oldham and Harrison ( 5 ) . In the authors’ laboratory it is the practice t o em loy a 0.75t o 1.00-gram sample of rubber stock instead o r the 0.15gram sample suggested by Oldham and Harrison (6). While larger samples increase the time required for washing and filtering, it is the authors’ experience that greater accuracy and reproducibility are attained thereby. Jacob Gabry, to whom the
A p d 15, 1943
ANALYTICAL EDITION
257
ported are the average of a t least two independent recoveries. The deviaArea after Recovery Carbon in Rubber Area after tion in all cases was not more than 3 from Rubber "01 Cresol Carbon Original Treating per cent. Black Area with HNOa B y "01 B y cresol method method Calculated The data indicate that carbon black Sq. m./g. Sq. m . / g . Sq. m./g. Sq. m./o. %b % % undergoes no appreciable alteration in Gastex 26O ... 22 21 28.4 27.7 27.8 .~ 70 Acetvlene 64 67 63 Acetylene surface area during incorporation in, Cabot 99 ... 100 99 22718 7:s 2;:2 27:8 C a b d grade 9 27:2 Cabot grade 6 109 112 110 107 27.7 27.4 27.8 or vulcanization of, a rubber stock. Cabot grade 4 143 181 175 125 27.9 27.7 27.8 With but a single exception, the surCabot grade 1 210 . . . 203 156 28.1 27.7 27.8 330 316 326 204c .. Spheron N face areas of the carbon blacks studied Grade 6 in Buna S 109 ... 113 ... 29:9 .. 30: 3 are substantially the same before inA previous ublication (8)re orted 40 sq. m. per gram as the surface of standard Gastex. C. W. corporation and after removal from Snow, General l t l a s Co., has kinsly stated that this material was actually CS-3 carbon. The present value is that of standard Gastex. the rubber stock by means of nitric b Ash-free basis. C N o t completely aoluble. acid. It is not possible to offer any ready explanation for the 25 per cent increase in surface of the grade 4 authors are indebted for the analytical results reported here, also sample. This is not due to any peculiar effect of the rubber points out that it is essential to control the time of digestion of the on the carbon black, since the black alone in nitric acid rubber stock and nitric acid t o 3 t o 4 hours at the specified temshowed an appreciable increase in surface. There was nothing perature of 60" t o 70" C. While in the case of channel blacks, peculiar in the method of manufacture or properties of this the weight of carbon black recovered by this method is generally reported (6)to be 105 per cent of its original weight, under the particular sample to account for its apparent increased conditions reported here a value of 108 per cent has been found activity toward nitric acid. I n any case, the alteration in more general. Consequently, a correction factor of 1.08 is emthe surface area of the recovered carbon is still not sufficient ployed instead of the value 1.05 recommended by Oldham and to confuse its identity. Harrison (6). One to 0.75 gram of the vulcanized stocks was digested for one I n general, both the nitric acid and cresol methods gave hour with 15 cc. of concentrated acid. After the addition of excellent results for the amount of free carbon in the stocks. another 35 cc. of acid, digestion was continued for 2 hours more. The surface areas of the blacks recovered by the cresol method In order to avoid contamination of the black for the subsequent show satisfactory agreement with the original supply in the surface area determination, a Selas sintered filter crucible FS20-100 was used in place of the usual asbestos Gooch. The case of the coarser blacks. However, with the finer particle crucible and black were washed according t o the standard method, blacks, the surface is considerably reduced. Thus, grade 1 then dried at 110" for 2 hours. The per cent carbon in the rubber black has an original surface of 210 square meters per gram. stock was determined directly without ignition. The surface After recovery from the tread stock, by cresol extraction area of the dried sample of recovered black was then determined by low-temperature nitrogen adsorption. The ash was deterthis value was only 156. After treating a 0.5-gram sample mined on a duplicate sample by ignition. In all compounds of this type black with cresol according to the standard studied, it amounted t o about 0.4per cent by weight of the whole procedure, the surface was 165 square meters per gram. rubber stock. Thus the recovered black as weighed out for the Evidently a small amount of material is retained by the surface area determination contained 1.44 per cent ash. The surface values in column 4 of Table I are accordingly 1.44 per carbon in spite of the extensive washings and heating decent lower than if they had been calculated on ash-free basis. scribed in the procedure. The amount is not sufficient to However, in the absence of data on the surface area of ash itself, alter the weight of the sample appreciably, but it is sufficient such a simple correction is not justified. The correction in any to block off certain interspaces in the black that were formerly case would not amount t o more than 1 or 2 square meters per gram, a value too small to cause any confusion in the identification accessible to the nitrogen molecule. Such an effect should of the black. become more pronounced the finer the particle size of the It was necessary t o establish that no appreciable change in surblack. Accordingly, the cresol method of recovery is not face area of the carbon black was brought about by the nitric acid satisfactory for separating carbon black from "unknown" during the separation of the black from the rubber stock. In order t o establish this, "blanks" of 0.5-gram samples of carbon rubber stocks for identification by surface area measureblack alone were subjected to the same treatment employed with ments. the compounded rubber stock and the surface areas of the blacks The Buna S stock showed little solubility in cresol and were then determined. quantitative recovery of carbon was not possible. It was possible, however, to get a satisfactory recovery of the carbon The use of cresol for the separation of carbon black from from the Buna S stock with nitric acid. I n order to obtain rubber stocks has been described by Roberts ( 7 ) . I n natural complete solution, i t was necessary to digest this stock with rubber stocks, this method gives excellent results for free nitric acid somewhat longer than the specified 3 hours. carbon. It was felt that there would be less opportunity This more drastic procedure is reflected in the slight increase for alteration of the carbon surface during separation by this in surface area of the recovered carbon. method than with concentrated nitric acid. The carbon black was separated from the stocks by cresol according to Conclusions the method of Roberts ( 7 ) . The surface area of the recovered carbon was then measured. I n general then, we may conclude that carbon black can .4 single determination of the area of a grade 6 carbon be recovered from rubber stocks with unchanged surface area, black recovered from a Buna S stock with nitric acid was The nitric acid technique is the most effective method of carried out. As it was not possible to get complete solution effecting the separation. The digestion temperature must of this stock in cresol, the method had to be abandoned in be controlled to between 60" and 70" C. for a total of not more this instance. than 3 to 4 hours. This method of identifying the carbon black in an unknown rubber stock is directly applicable only Experimental Results in the presence of a single type of carbon black. If blends of blacks are employed in the material under investigation, Table I presents a summary of the data collected. With some secondary identification is also required. The nonthe exception of the Gastex and grade 9 channel black, which impingement-type blacks, for example, are readily identified are values for single recoveries, the surface area values reTABLE I. EXPERIMENTAL RESULTS
258
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 15, No. 4
by microscopic observation. From the known surface areas of these materials and the total per cent carbon present in the stock, a fairly positive identification of the blend can be made. The fact that carbon black can be recovered quantitatively and with unchanged surface area from vulcanized rubber stocks appears to lend impetus to a physical concept of carbon black reinforcement. This point of view implies that any chemical combination between the ingredients of the rubber stock and the carbon black would be evidenced by some alteration in the surface of the latter. A few experiments were performed in an attempt t o establish to what extent this concept was valid.
While unaltered surface area need not necessarily be interpreted as evidence of complete lack of surface reactions, it is the authors' opinion that the extent of chemical combination at the carbon black surface is very small. This interpretation is in accord with the views expressed in a previous publication (9), where it was suggested that the chief role of carbon black in rubber reinforcement may rest on its ability to orient the chains of rubber molecules (4) and thus alter the extent and type of rubber-sulfur bonds normally formed in nonreinforced rubber stocks.
One hundred grams of grade 6 carbon black were intimately mixed with 6.6 grams of sulfur. This is about the ratio in which they are present in a standard rubber batch. Samples of this mixture were subjected t o the standard curing temperature of 134' C. for 30, 60, and 90 minutes. The free sulfur was then extracted for 40 hours with acetone and the combined sulfur on the carbon was determined. Values of 0.16, 0.21, and 0.4 per cent combined sulfur were obtained. The original sample of grade 6 carbon black had a surface area of 108 square meters per gram. The extracted sample of black containing 0.4 per cent of combined sulfur had a surface area of 109 square meters per gram. These values are identical within experimental error.
Emmett, P. H., and Brunauer, S.,Ibid., 56, 36 (1934). Emmett, P. H., and DeWitt, T.,ISD. ESQ. C H E x . , ASAL. ED.,
Literature Cited Brunauer, S., Emmett, P. H., and Teller, E., J . Am. Chem. Soc., 60, 309 (1938).
13, 28 (1941).
Gehman, S. D., and Field, J. F., ISD.EXG.CHEY.. 32, 1401 (1940).
Oldham, E. W., and Harrison, J. G., Jr., ISD.E m . CHEX.,ANAL.
ED.,9, 278 (1937). Parkinson, D., Trans.Inst. Rubber Ind., 16, 94 (1940). Roberts, J. B., Jr., Rubber Cham. Tech., 14, 241 (1941). Smith, W. R., Thornhill, F. S., and Bray, R. I., IND. ENQ.CHmr., 33, 1303 (1941).
Thornhill, F. S.,and Smith, W. R., Ibid., 34, 218 (1942).
Rapid Iodine Number Determinations FRANK A. NORRIS AND ROBERT J. BUSWELL General Mills, Inc., Research Laboratories, Minneapolis, Minn.
0
F T H E numerous iodine number methods described in the literature for determining the unsaturation of fats and oils, the Wijs and Hanus methods are used most extensively (d), although the Rosenmund-Kuhnhenn method, especially in its micromodification, is used to a considerable extent in biological work. The Wijs reagent, official with the American Oil Chemists Society, gives theoretical values for pure nonconjugated unsaturated fatty acids, but it is sensitive t o light and is not recommended for use when more than 30 days old. The Hanus reagent, on the other hand, usually gives results about 2 to 4 per cent lower than those obtained with the Wijs method, but it is stable and when protected from light will remain in a satisfactory condition for a year or longer. The Rosenmund-Kuhnhenn reagent i s also stable but gives results appreciably low for all oils of iodine number greater than 100 (a). The reaction time for all three methods is usually 0 5 to 1 hour, depending upon the degree of unsaturation of the sample. Some time ago, Hoffman and Green (3) suggested the addition of mercuric acetate to the Wijs reagent in order to cut the reaction time to 3 minutes without any change in the iodine number. This convenient rapid method has apparently escaped much attention. One of the present authors ( 6 ) , however, has used this modified procedure on oils containing conjugated double bonds. Results obtained on tung oil are shown in Figure 1. Here the standard Wijs, rapid Wijs, and standard Rosenmund-Kuhnhenn methods are compared. Results obtained with the standard Hanus method are not shown, since they are known to be erratic when the reagent is employed in its normal concentration (0.2 N ) . [Theoretical iodine numbers on pure conjugated fatty acids have been reported by von Mikusch (6),using an
approximately double strength Hanus reagent. However, normal strength Hanus reagent usually gives high and erratic results (I).] Standard (slow) Wijs values are largely dependent upon the excess of reagent employed, the iodine number varying 43 units in an excess range of 25 to 272 per cent. Rapid Wijs values, however, vary only insignificantly in the reagent excess range of 30 to 225 per cent, permitting much greater latitude in sample weights. The 55 * 3 per cent excess re-1
EFFECT OF REAGENT EXCESS
ON IODINE VALUE OF TUNG
I 50
I
OIL
I
100 150 200 THEORETICAL % EXCESS REAGENT
FIGURE 1
I a60
300
