Document Yrgy4jn9gjDmoy7ZDO30k20y

566 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 18. No. 6 clay at various steps is based entirely on the behavior of the retained samples. Table II is an abstracted record of the work done by the clay after the various burnings. All the runs with clay of the same condition have been grouped together, and each column represents the total. The individual yield in barrels per ton varied only slightly in any case from the average for its group. The constants determined on the refined product for control-- A. P. I. gravity, Engler end point, and color--are indicated by giving the highest and lowest of each in each group. The table also shows the gain in weight of the clay at each use, and the loss on burning, Refining with Reburned Clay The retained samples of clay from the various burnings were compared by refining with them the gasoline from a Fleming pressure distillate which was known to give a product of the desired characteristics with the unused clay, at about 450 barrels per ton in the glass apparatus. In running these tests, 11.5 grams of the clay were used in each case, 1220 cc. of the pressure distillate, which had a gasoline content of about 75 per cent, being charged. Table III shows the results obtained, regards color,- stability, and gumming tests, and gives for comparison the methylene blue numbers and turpentine temperature reaction. From the above it is evident that, although there is a slight difference in the stability of the products from the burned clays as compared with that from the unburned, it is not large enough to be of practical importance, especially since rebumings after the first seem to leave the clay unaffected. There is, therefore, no reason to doubt that the reburning and re-use can be repeated practically indefinitely without further impairing the efficiency of the clay. Table III--Comparison of Reburned Clays for Vapor-Phase Refining ------------------------ RBBTrRNBD----------------------- s Clay used used Once Twice times times times Methylene blue number 0.10 1.0 1.5 1.0 1.0 1.5 Turpentine temperature re- action, ,F. 148 6 12 4 5 6 Vield, barrels per ton 456 461 461 460 456 467 Product, initial boiling point, 0 F. 108 110 118 104 116 110 Temperature, F.: 20% off 210 208 212 208 216 216 50% off 288 282 286 282 284 288 90% off 384 384 381 380 380 382 End point 416 410 422 414 414 426 Gum mg. per 100 cc. Copper dish Glass dish Color (Sayboit): 23.9 19.6 21.3 18.6 13.6 16.2 6689 5 7 Original After 2 days (south light) After 4 days (south light) After 7 days (south light) 30 30 29 30 28 30 29 27 26 26 27 27 26 25 25 25 25 25 25 23 24 23 24 23 a A Fleming pressure distillate was used in performing these tests. Conclusions Organic material left in clay used for vapor-phase refining can be burned out sufficiently at a temperature of 1020 F. to allow of efficient re-use of the clay. This has been found to leave the efficiency of a Georgia fuller's earth entirely unimpaired by one reburning. JReburned Florida clay, revivified at 1020 F., is only slightly less active than the unused clay. After the first reburning its utility remains unimpaired by further burnings as far as the present experiments have been carried--that is, through five reburnings. Effect of Various Carbon Pigments upon Rate of Oxidation of Linseed Oil1 '* By F. H. Rhodes and H. E. Goldsmith Co r n b ix Un iv e r s it y , It h a c a , N. Y. HE effect of the car Lampblack and carbon black markedly inhibit the T bonpigmentsupon the rate of dryingof paints oxidation of raw linseed oil. This effect is not due to the presence of an oily impurity In the pigment, but has been discussed by nui s due largely to the adsorption bn the pigment of the merous authorities. It is intermediate oxidation product which is the true cata generally agreed3-4-6 that lyst in the drying of linseed oil. When added to lin some carbon pigments, par seed oil containing drier, the carbon pigments adsorb ticularly lampblack, act as "retarders in the drying of paints and tend to inhibit some of the drier and for this reason also tend to re tard the drying. The extent to which the oxidation is inhibited varies with the nature of the drier present. the oxidation of the oil. of such pigments upon the rate of oxidation of linseed oil. The present investigation was undertaken to determine quantitatively the effects of the black carbon pigments on the rate of oxidation of linseed oil, and to establish the causes for these observed effects. The exact cause of this effect is not fully understood. Sev eral authorities'3-4-6*-6 attribute the retarding action of lamp black to the presence in the pigment; of a certain amount of oily material. In none of this previous work on the effect of black carbon pigments on the rate of drying of Materials J Pure linseed oil from North American seed was used. It showed the following analysis: paints has a quantitative study been made of the effect 1 Received January 19, 1926. a This article Is respectfully dedicated by the authors to Prof. L. M. Specific gravity at 15 C. Refractive index at 15 C. Acid number Saponification number Iodine number 0.9315 1.4820 4.51 195.3 170.5 Dennis, and will be reprinted as Article Number Two in the Louis Munroe Dennis Quarter Century Volume to be published iu commemoration of the completion by Professor Dennis of twenty-five years of service as head of The driers were linoleate paste driers, showing the foUowing the Department of Chemistry at Cornet! University. * Toch, "Chemistry and Technology of Paints," 2nd ed., 1916, p. 99. * Sabin, "Technology of Paint and Varnish," 2nd ed., 1916, p. 201. i Hurst, "Painters' Colours, Oils and Varnishes," 2nd ed., 1901, p. 250. Morrell, "Rubber, Resins, Paints and Varnishes," 19S0, p. 136. Co b a l t Pa s t s Dr ib s Cobalt (calcd. as metal) Maneanese Magnesium, copper, nickel 5.74.per cent Trace . Spectroscopic traces only June, 1926 In d u s t r ia l a n d e n g in e e r in g c h e mis t r y 567 Ma n g an e s e Pa s t e Dr ie r Manganese (calcd. as metal) Iron Aluminum, copper, sodium r .es^only De a d Pa s t e Dr ie r Lead (calcd. as metal) Iron Sodium, aluminum, arsenic, magnesium, silicon 17.59 per cent 0.0605 per cent Spectroscopic traces only In making up the vehicles for the paints studied, a calcu lated amount of the paste drier was dissolved in a known weight of warm linseed oil. The vehicles were stored in sealed bottles until used. The amounts of drier used were sufficient to give the following concentrations of active material (calculated as metal) in the vehicles: Vehicle No. 1 2 3 Me t a l Lead Manganese Cobalt Per cent 0.275 0.0732 0.0221 The carbon pigments were: Ash Moisture Soluble in benzene Carbon black Per cent 0.166 3.68 Trace Lampblack Per cent 0.2028 2.15 5.82 Vine black Per cent 8.516 5,72 None It,hah been stated that the oily impurities in lampblack are largely responsible for the action of this pigmenV.ni.,.,' ,. retarding the oxidation of linseed oil. In order to obtain " ' further information as to the effect of oil-soluble impurities in the-pigments,, samples of paints made from carbon black and from lampblack were centrifuged and the rates of oxi- ' dation of the clarified oils thus obtained were determined^ ' The linseed oil obtained from the carbon black paint was clear and of a lighter color than the original raw oil from which the paint! had been prepared; the oil from the lamp black paint was Idark red and somewhat fluorescent. Procedure The procedure was similar to that described by Rhodes and Van Wirt.7 Paints prepared from pigment and linseed oil were exposed to an atmosphere of pure oxygen at 30 C. and the rate of absorption of oxygen and the rate of evolution of volatile matter were measured. Unless otherwise specified, each paint contained one part by weight of pigment and nine parts by weight of vehicle. With each paint there were made at least two parallel determinations, giving results which agreed to within the limits of experimental error. Results The results are shown graphically by the accompanying charts, in which the amounts of oxygen absorbed and the amounts of volatile matter evolved, each expressed in terms of percentage by weight of the oil in the paint, are plotted against the lengths of time of exposure. A. single cufve is shown to represent the average results obtained in the two or more check determinations made with each sample. Linseed Oil Alone Chart I shows the rates of oxidation and the rates' of evolution of volatile matter of raw linseed oil alone and of the mixtures of linseed oil with the various driers. Theautocatalytie character of the reaction involved in the drying of linseed oil and the effects of the driers in accel erating the formation of the autocatalyst and thus increasing the initial rate of oxidation are clearly indicated. CHART I The results of the oxidation experiments made with these oils are shown on Chart II. The oils separated from the paints oxidized a little more slowly at first than did the raw linseed oil, but after a short period of induction the separated oils took up oxygen at almost the same rate as did the fresh raw oil. These experiments prove that the effect of black carbon pigments in inhibiting the oxidation is not due to the presence in the pigment of impurities that are soluble in linseed oil. The action of the blacks in de laying the drying must, therefore, be due to some specific action of the pigments themselves. That the oils separated from the paints showed more marked periods of induction than fresh linseed oil suggests that the pigments may have adsorbed and removed the small amount of the autocatalytie intermediate oxidation product which is normally present in ordinary raw linseed oil. This, in turn, suggests that the slowness of drying of paints containing carbon pigments may be due to . the fact that the pigments continuously adsorb and remove the autocatalyst as fast as it is formed. Carbon Pigments Added % Chart II shows the results obtained in a series of experi ments made to determine the rate of oxidation of paints made from raw linseed oil and carbon pigments. The addition of the carbon pigments markedly retarded the, rate of oxidation of the oil. The paints were still somewhat tacky after 400 hours' exposure. Ab a supplementary experiment, samples of carbon black paint and of linseed oil alone were painted in parallel strips on a glass plate and allowed to stand in.the air at room temperature. The film of oil became tacky in 4 days and was dry to the touch in 7 days. The paint made from carbon black became tacky only after 22 days and was dry to the touch only after 28 days, 1 Th is Jo u r n a l , 16, 1135 (1923). H0UR6 Z50 300 350 Additional evidence in favor of this hypothesis is supplied by the following experiment: A paint made with carbon black was oxidized at 100 C. until it became thick when cooled. The pigment was separated from the vehicle by centrifuging, and was then washed with raw oil and again centrifuged to remove the wash oil. The resulting pigment was made up into a paint with raw linseed oil and the rate of oxidation of this paint was determined. Oxidation took 568 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 18, No. 6 place very slowly for the first 150 hours, but after this period of induction oxygen was absorbed about as rapidly as by the raw oil alone. It will be observed that this paint made from pigment recovered from a partly oxidized paint dried much more rapidly than did paint made from the original pigment. The results indicate that during the preliminary treatment the carbon black had become almost saturated with the intermediate oxidation product, so that in the being exposed to oxygen the initial rate of oxidation was appreciably less than the initial rate of oxidation of the freshly prepared paint. This appears to be due to the gradual adsorption of some of the cobalt drier by the pigment. In order to determine to what extent this adsorption of the drier occurs, freshly prepared and week-old samples of the carbon black paint were clarified by centrifuging and the resulting dear vehicles were analyzed for cobalt CHART AT 350 oxidation of the final paint complete saturation was attained in a short time, after which the autocatalyst remained in the oil to exert its normal effect in accelerating the drying. Cobalt Drier The effect of carbon pigments on the rate of oxidation of linseed oil in the presence of cobalt drier was next in vestigated. The paints were made up to contain 10 per cent of pigment suspended in linseed oil in which had been dissolved sufficient cobalt linoleate to contain an amount of cobalt equivalent to 0.0221 per cent of the weight of the oil. The results are shown by Chart III. Curves III-B and III-C show the rates of oxidation of freshly prepared paints made from carbon black and from lampblack. These two curves are practically identical in form. In each case the presence of the pigment causes a very slight retardation of the initial rate of oxidation of the vehicle. In these experiments the effect of the pigments in inhibiting the oxidation of the oil was very much less than in the experiments made with paints prepared from linseed oil without drier. It appears that in the presence of the powerful cobalt drier the intermediate oxidation product that is the true catalyst in the oxidation reaction is formed much more rapidly than it is renioved by adsorption on the pigment. Vine black (Curve III-D) had even less effect and tested to determine the rate of absorption of oxygen. The clarified vehicle from the freshly prepared paint con tained 0.003 per cent of cobalt; that from the week-old paint contained only a trace of cobalt. The rates of oxidation of the two clarified samples are shown by Chart IV. The clarified vehicle from the freshly made paint oxidized at almost exactly the same rate as did the original vehicle from which the paint was prepared. The vehicle recovered from the paint which had been allowed to stand for a week oxidized much more slowly. This was to be expected, since the concentration of drier in this sample was extremely low. From a comparison of Curves IV-B and IV-C it will be seen that the vehicle recovered from the week-old paint oxidized even more slowly than did the paint itself. It would appear that the cobalt drier still possesses some catalytic activity even when adsorbed in the pigment, although, of course, the adsorbed drier is less effective than drier in solution. Lead Drier In studying the effect of the pigments upon the rate of drying of linseed oil in the presence of lead drier, the vehicle used contained an amount of iead linoleate equivalent to 0.275 per cent of lead. The paints were made up to . contain 10 per cent of pigment as before. The paint made upon the initial rate of oxidation than had carbon black or lampblack. This is to be expected, since the vine black is relatively dense and gives relatively less adsorbing surface. It was observed (Curve III-E) that when the carbon black paint was allowed to stand for several days before from carbon black absorbed no measurable amount of oxygen during the first 5 hours' exposure. Oxidation then began and, after a short period of induction, progressed normally. The paint made from lampblack showed about the same characteristics as did the paint made from carbon June, 1926 INDUSTRIAL AND ENGINEERING CHEMISTRY 569 black, although the length of time during which no oxygen was absorbed was somewhat longer. The. effect of the pigments in retarding the oxidation catalyst. It will be observed that in the case of the paints oxidation, when once started, took place somewhat more rapidly than in the case of the clarified vehicle separated of the vehicle may be due in part to the adsorption of the from the paint (Chart VI). This provides additional in drier by the pigment. To determine the extent to which dication that drier adsorbed on the pigment may still show such adsorption occurs, a sample of the carbon black paint some catalytic activity. was centrifuged and the resulting clarified vehicle was analyzed for lead. For comparison, an analysis was made on a portion of the original vehicle which had been similarly Paints made up to contain 10 per cent of vine black oxi dized almost as rapidly as did the original vehicle. "When the concentration of the vine black was increased to 25 centrifuged to remove any drier not actually in solution. per cent, a somewhat more pronounced effect was noted. The clarified original vehicle contained 0.097 per cent of That vine blapk is much less effective than carbon black lead, while the vehicle recovered from the paint contained or lampblack in retarding the drying of the oil is due, of only 0.0312 per cent of lead. Obviously, the carbon black course, to the smaller adsorptive power of the vine black. had adsorbed and removed over two-thirds of the drier present in actual solution in the oil. Manganese Drier The progress of the oxidation of these clarified vehicles is shown by Chart V. In spite of the fact that .the lead content of the original vehicle was reduced from 0.275 per Finally, the effects of carbon pigments upon the drying of the vehicle containing manganese drier were studied (Chart VII). The vehicle contained 0.0732 per cent of manganese. The lampblack and carbon black paints OX/DAT/ON Of RA/NTS WITH MANGANESE DRIER contained 10 per cent by weight of pigment; the vine black paint was made up to contain 25 per cent by weight of pigment. In no case did the presence of the pigment appear to have much influence upon the rate of oxidation of the linseed oil, although the paint made with carbon black showed a slightly longer period of induction than did the VTHKU ABONT s =c ap9m s ik k p*mr\ c *ia b p b ia c k p a in t ] io*piafT/rr D=VtN BLACK BMNT~Z5% P/GNCNt *Ql ABMRBT& k k . n. vehicle alone. To determine to what extent the pigment absorbs and removes the manganese drier, samples of the original vehicle and of the carbon black paint were clarified by centrifuging. The centrifuged vehicle contained 0.0262 per cent of man ....SO /OO /SO HOURS ZSO 300 360 ganese,' while the vehicle recovered from the paint contained CHART m only 0.0065 per cent. The rates of oxidation of the clarified vehicle and of the vehicle recovered from the paint were cent to 0.097 per cent by centrifuging, the centrifuged ve substantially the same as that of the original vehicle (Chart hicle oxidized at almost exactly the same rate as did the VIII). original. These results indicate that, under certain con ditions at least, only a portion of the lead linoleate added to Conclusions linseed oil actually passes into solution in the oil, and that only the comparatively small amount of drier in actual solution is effective in promoting the drying of the oil. The vehicle which was separated from the paint, and which contained only 0.0312 per cent of lead, showed a more pro nounced period of induction and an appreciably slower 1-- The addition of lampblack or carbon black to raw linseed oil very markedly inhibits the oxidation of the oil. 2-- The effect of the carbon pigments in inhibiting the oxidation of linseed oil is not due to the presence of oily or oil-soluble impurities in the pigments. rate of oxidation than did the original vehicle. This is the logical result of the lower concentration of dissolved drier. The adsorption of a large part of the lead drier by the pigment does not, however, explain all the phenomena observed in connection with the oxidation of the paints made with carbon black or lampblack. That practically no oxygen was taken up during the first few hours of exposure may be explained by the hypothesis that the pigment also adsorbed the small amount of autocatalyst which may have been present in the original oil or which may have been produced during this initial period. Only after several hours was sufficient unadsorbed autocatalyst produced to exert a marked effect in promoting the oxidation. It will 250 be observed that in the case of the paint made up to contain CHART WE 15 per cent of lampblack the initial period during which no oxidation took place was somewhat more pronounced than 3-- The effect of the carbon pigments in delaying the in the case of the paint containing only 10 per cent of lamp oxidation of raw linseed oil is due to the continuous ad black. This is in agreement with the theory as outlined sorption of the intermediate oxidation product which acts above. as the true catalyst in the drying reaction. That the effect of the pigment in inhibiting the oxidation 4-- The carbon pigments only slightly retard the oxi continued for a few hours only instead of for many days, dation of freshly prepared paints containing cobalt drier. as in, the case of the paint made from raw linseed oil, is due The inhibiting effect becomes more pronounced as the to the fact that in the present case the adsorption of the paints are allowed to stand. This is due to the gradual lead drier by the pigment very greatly decreased the ad adsorption of the cobalt drier by the pigment. sorptive capacity of the pigment for the intermediate auto 5--When added to linseed oil containing lead drier, carbon 570 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 18, No. 6 black anil lampblack almost completely inhihit. the oxi dation of the oil during the first few hours of exposure. This effect is due to the fact that the pigments adsorb not only a portion of. the lead drier but also the small amount of autocatalytic oxidation product that, is formed at the beginning of the oxidation. 6:--Carbon pigments have little effect upon the rate of oxidation of linseed oil containing manganese drier. Heterogeneous Catalysis1 HI--Hydrogenation of Cottonseed Oil with Platinum By A. S. Richardson and A. O. Snoddy Th e Pr o c t e r & Ga mb l e Co ., Iv o r y d a l e , Oh io HE glycerides of cottonseed oil are derived from linoleic T acid, oleic acid, and solid saturated acids, chiefly palmitic. Hydrogenation of such an oil may con The mere fact that the average cottonseed oil is about 75 per cent derived from liquid fatty acids has never been a serious handicap to its use in edible fat. The average edible ceivably involve any combination of the following thrfeaet should and does contain a substantial excess of liquid changes: (1) linoleic to oleic acid; (2) oleic to stearic acid; constituents. However, the fact that about 65 per cent of (3) linoleic to stearic acid. The problem of the distribution this liquid fatty acid consists of linoleic acid renders cotton of the total hydrogenation of cottonseed oil among these seed oil extremely susceptible to oxidation and resulting ran three changes belongs to a general field of investigation which cidity. Hence the possibility of converting linoleic to the has appropriately been described by various investigators as relatively stable oleic acid by selective hydrogenation was the problem of "selective hydrogenation." even more attractive to the far-sighted manufacturer than the The previous literature on the selective hydrogenation of formation of solid saturated from unsaturated acid. fatty oils has been reviewed in the earlier papers of this series.2 Burchenal's discovery was followed, in turn, by industrial The present paper deals with the selective action of platinum development and by a series of investigations which show that black in the hydrogenation hydrogenation inherently of cottonseed oil, but be tends to be selective, but fore proceeding to a discus sion of the experimental data involved it will be in order to correct a rather widespread misunderstand ing of the fundamental pur pose of hydrogenation. Practical Importance of Selective Hydrogenation Contrary to the impres The predominating importance of selective hydro genation in industry is pointed out. The hydrogena tion of cottonseed oil with platinum catalyst has been investigated at temperatures varying from 40 to 240 G. The preferential conversion of linoleic acid to the oleic acid stage and also the formation of solid unsaturated acid Is favored by increasing temperature over practi cally the whole of the range investigated. Hydrogena tion of cottonseed oil with platinum catalyst does not appear to be so selective as with nickel catalyst. not completely so. Many conditions in practice may interfere with or defeat the purpose of selective hydro genation. If this were not so, it is almost inconceivable that selective features of hydrogenation should have so long been unnoticed. The most important fac tor to be controlled is the sion which the casual reader end point of hydrogenation. may gain from a survey of For instance, at a more or the published papers on selective hydrogenation of oils, the less well-defined critical point hydrogenation of vegetable oil subject is not primarily a theoretical one. It is a subject in containing linolein ceases to involve primarily the conversion which practical development on a large scale preceded any of linolein to olein and results in substantial increase of stearin scientific literature. content. Carried beyond that point, hydrogenation becomes The pioneers3 in the field of hydrogenation of oils regarded effectively less selective. Operating conditions may also render their process merely as a means of converting liquid un the process of hydrogenation effectively nonselective. For in saturated to solid saturated fatty acids--for example, oleic stance, unhydrogenated oil may be swept through the system or linoleic to stearic acid. As long as that idea prevailed in into the finished product! This is particularly likely to occur industry, hydrogenation was making important but com in a continuous process, but may also occur in a batch process paratively slow progress. The chief impetus to industrial in which pockets of unhydrogenated oil persist throughout development was the discovery by Burchenal4 that hydro the reaction period as a result of incomplete agitation. genation could be controlled bo as to produce a product es Temperature, perhaps, is the most important factor which pecially suited for edible purposes, in which the principal has been shown to affect the inherent selective tendency chemical change is the selective conversion of the highly of hydrogenation of oils in the presence of nickel catalyst. unsaturated linoleic acid to the less unsaturated oleic acid, Moore, Richter, and Van Arsdel6 found that the selective with the formation of relatively little stearic acid. conversion of linolein to olein was favored by increasing tem 1 Presented as a part of the Symposium on Cotton and Its Products and Vegetable Oils before the joint session of the divisions of Agricultural and Food Chemistry, Biological, Cellulose, and Industrial and Engineering perature in the hydrogenation of cottonseed oil. Richardson, Knuth, and Milligan2 have made the same observations on several oils in the temperature range 150 to 200 C., but have Chemistry at the 71st Meeting of the American Chemical Society, Tulsa, obtained conflicting results above 200 C. Olda., April 5 to 9, 1926. * Richardson, Knuth, and Milligan, Th is Jo u r n a l , 16, 519 (1924); Scope of Present Work 17, 80 (1925). * Leprince and Sieveke, German Patent 141,029 (August 14, 1902); The present work was undertaken at the suggestion of H. Normatm, British Patent 1515 (January 21, 1903). . J. Morrison for the immediate purpose of extending the * TJ. S. Patent 1,135,351 (application filed November 10, 1910, patent granted April 13, 2915). * Th is Jo h r ma l , 9, 451 (1917).