Document 4v5E8JVxNM6nr2nzy4JZ5bwja

No. 9 i September, 1924 INDUSTRIAL AND ENGINEERING CHEMISTRY 957 Effect of Certain Metallic Soaps on the Drying of Raw Linseed Oil1'2 ore of water, it vpl- . tentat that. id the oduct iously led in stones in the urther pected ion of , if the at lies npurime do. : same, ethoc tend erabli ,th the, } hy 4'. SUlfSr^ SlWie OMi`4 lethol lethod^l ketonM dstilTlp By L. L. Steele Bu r b a u on St a n d a r d s , Wa s h in g t o n , D. C. EAD, manganese, and cobalt and combinations of L these metals in the form of their soaps are the com monly used driers for linseed and other drying oils. inches. The excess oil was drained by inclining the plates at an angle of approximately 45 degrees to the vertical for 15 minutes, the bottom edge wiped off, and the plates laid in a Iron and copper are two other metals which may be introho rizontal position. The drying tests were made in an evenly duced through the heat treatment of a drying oil in kettles heated laboratory, free from combustion or chemical fumes, made from one of these two metals. In addition, iron has been and lighted by northern exposure. A recording thermometer used as a drier in the manufacture of black varnishes. of the wet and dry bulb type yielded data concerning any In this paper data are presented on the comparative effects fluctuations in temperature or humidity. The average rela of different proportions of these five metals and combinations tive humidity during the drying tests was 40 per cent, with a of them on the drying time of raw Linseed oil. variation from 36 to 43 per cent, with the exception of one 24- Pr e p a r a t io n o p Me t a l l ic So a p s hour period when the relative humidity reached 49 per cent. The average day temperature during the drying period was Fused metallic resinates were selected as a convenient 23.3 C. (74 F.), with a variation from 22.2 to 23.9 C. (72 i form of the respective metals for introduction into raw lin- to 75 F.), while the average night temperaturewas 20 C. (68 i seed oil. They were prepared by customary methods,3 so F.), with a variation from 18.9 to 20.5 C. (66 to 69 F.). :: that only a brief description of their preparation is necessary. The oil films were tested at regular intervals, approximately The resinates were prepared by the fusion of water-white in the center, 2.5 cm. (1 inch) from the top. The oil was con rosin with the acetates of lead, manganese, cobalt, and copper sidered to be dry when the finger could be drawn lightly over ; and freshly precipitated ferric hydroxide made from elec- the surface at an arbitrarily chosen point without marring it. , trolytic iron containing less than 0.01 per cent manganese. Fairly definite end points could be obtained with the rapid The metallic acetates were C. P. reagents. The lead resinate was light amber in color and contained r 6.73 per cent of lead. The manganese resinate was amber ss - colored and contained 2.93 per cent of manganese. The cobalt to : resinate was ruby colored and contained 3.29 per cent of .cobalt, The copper resinate was deep emerald in color and ? contained 4.16 per cent of copper. The iron resinate was very ss so li dark brown and contained 2.76 per cent of iron. All these presinates dissolved to a clear solution in benzene, except iron, - Which showed traces of sediment. *70 * LEAD " Pr e p a r a t io n o p Sa mp l e s Co n t a in in g Dif f e r e n t Amo u n t s o f Dr ie r ? A clear sample of raw linseed oil with an iodine number of 491, which conformed to the requirements of Bureau of Stand ards {Circular 82,2nd ed., U. S. Government Specification for linseed Oil), was selected for use in all the drying tests. Stock samples of oil containing definite percentages of metal rere prepared by incorporating the necessary amount of the respective metallic resinate in a weighed quantity of oil. She- resinate and oil were heated to 150 C., thoroughly 'Shaken, and allowed to cool. In this way thorough mixing .Was obtained without an appreciable alteration of the oil . through heat treatment. Portions of these different stock Isslutions were diluted with definite amounts of raw linseed to give a series of samples with decreasing percentages of Metal. These samples were all kept in tightly corked tubes, |the air in which had been displaced by carbon dioxide to pre$hi(fe any preliminary oxidation of the oil. De t e r min a t io n o f Dr y in g Time o f Sa mp l e s fe The linseed oil samples containing definite percentages of pjnetals were flowed in streaks approximately 2 inches wide on phoroughly cleaned window-glass plates measuring 6 by 8 J Received April 14, 1924. Presented before the Section of Paint and |$ambh Chemistry at the 67th Meeting of the American Chemical Society, gtys^ington, D. C., April 21 to 26, 1924. % * Published by permission of the Director, XJ. S, Bureau of Standards. 1* Educational Bureau of the Paint Manufacturers' Association of the fiited States, Circular 120 j Seeligmann and Ziecke, "Habdbuch der LacI'Finrisindustrie," p. 70.1* G u^A corns * IRON r*..........I QZ .OS- 1 MANGANESE -- . COBALT ii .OS JO JX ./< ,, Fro. 1--Ef f e c t s o f Dif f e r e n t Pe r c en t ag e s o f De a d , Co f f e r , Ir o n Ma n g a n e s e, a n d Co b a l t o n t h e Dr y in g o p Lin s e e d Oil drying oil mixtures; with the slow drying samples the exact length of time for the film to become dry could not be de termined with a high degree of accuracy. It should be em phasized that this method for the determination of the drying time involves a considerable personal element, but the data presented are intended chiefly to show the relative catalytic drying effects of the five metals and combinations thereof. 958 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 16, No. 9 Fro. 2--Effect of Lb a p on Dr y in g o r Sa m p l es o r I.u;si;Bn On, Co n t ain in g Co b al t PaCEM- OFMANGANESEINLINSEED OIL Fro. 3--Ef f ec t o f Le a d o n Dr y in g o f Sa m p l e s o f Lin s e e d Oil Co n t a in in g Ma n g a n e s e , a n d Ef f e c t o f Sma l l Amo u n t s o f Co p pe r o n Sa mp l e s o f Lin s e e d Oil Co n t a in in g Le a d a n d Ma n g a n e s e PEA CENT OP /ffON ANO COPPER, AESPEC Tl VELYt IN LINSEED OIL Fig . 4---Ef f e c t o f Le a d o n Dr y in g o f Sa m p l e s o f Lin s e e d Oil Co n t a in in g Co p p e r a n d Ir o n Ta b u l a t io n o f Re s u l t s All the data on the drying effects of the different metals and their combinations on linseed oil are shown in graphic form in Pigs. 1 to 5, inclusive. Dis c u s s io n o f Da t a Fig . S--Er r e c t o r a Co n s t a n t Sma b z. Fs k c b n t a g b o r Ma n g a n b s b o n *h b Dr y in g o r Sa mr b b s o r Lin s b b d Oh , Co n t a in in g Dir r br b n t Amo u n t s o r LEAS From Fig. 1 it appears that neither copper nor lead resinate alone is an efficient drier for raw linseed oil. Of the five metals studied, copper is the only one that shows any marked increase in drying effectiveness with decrease in concentra tion. Iron seems to approach manganese and cobalt in effectiveness when used in comparatively high concentrations. The excessively dark color of the iron solutions would be a serious objection to their use. Cobalt appears to be somewhat more effective than manganese in concentrations ranging from 0.3 to 0.03 per cent; below this concentration these metals seem to be nearly equal in their catalytic action, which begins to diminish rapidly at this point with further decrease in concentration. There seems to be practically no advantage in the use of more than 0.05 per cent of either manganese or cobalt to dry linseed oil. At a concentration of 0.05 per cent of metal, taking cobalt as 100, the relative catalytic effects of the other metals studied Would be, roughly, manganese 46, iron 17, copper 13, and lead 6.5. From Fig. 2 it is seen that if lead is added to samples of linseed oil containing cobalt in the proportion of 3 parts of lead to 1 part of cobalt, there is very little added catalytic effect when the concentration of cobalt is above 0.02 per cent. Lead in the same proportion seems to add effectiveness when cobalt is present in concentrations approximating 0.02 per cent. A greater concentration of lead adds catalytic effect in a marked degree to small concentrations of cobalt. With September, 1924 INDUSTRIAL AND ENGINEERING CHEMISTRY 959 0.1 per cent of lead present, an amount of cobalt as small as 0.006 per cent dries linseed oil in less' than 12 hours. A more extended study of the effect of different concentrations of lead on small concentrations of- cobalt is desirable. From Fig. 3 it is seen that if lead is added to samples of linseed oil containing manganese in the proportion of 3 parts of lead to 1 part of manganese, there is a marked increase in the catalytic drying effect. A sample of linseed oil containing 0.01 per cent of manganese is changed in drying time from 85 to 24 hours by the addition of 0.03 per cent of lead. The addition of small amounts of copper to an oil containing lead and manganese appears to have no appreciable effect, upon the drying time. This point is of interest because of the possible entrance of copper into linseed oil which has been heat-treated in a copper kettle. From Fig. 4 it is seen that if lead is added to samples of linseed oil containing iron in the proportion of 3 parts of lead to 1 part of iron, there is no appreciable increase in the cata lytic drying action. In no case did samples of linseed oil con taining 0.1 per cent of lead and 0.02,0.01, and 0.005 per cent of iron, respectively, dry in less than 35 hours. This fact indicates that iron cannot be classed with manganese and co balt in catalytic drying action. The addition of lead to samples of linseed oil containing copper in the proportion of 3 parts of lead to 1 part ofcopper seemed to have a marked retarding effect upon the drying time. In a similar manner, larger amounts of lead with com paratively small amounts of copper did not produce an effec tive catalytic drying action. Fig. 5 shows in a marked way the large increase in drying effect obtained by the addition of relatively small percentages of manganese to samples of linseed oil containing different percentages of lead. There appears ,to be no advantage in using more than 0.05 per cent of lead in combination with 0.01 per cent of manga nese or more than 0.1 per cent of lead in combination with 0.005 p8r cent of manganese. In conclusion, it should be emphasized that the data shown in the paper apply to the relative drying effects of certain metals and combinations of these metals when incorporated with raw linseed oil in the form of metallic resinates. The effect of heat treatment during incorporation of the different metals in the manufacture of a true boiled linseed oil has not been considered. Such heat treatment might have a very considerable modifying effect on the catalytic action of the five metals studied. Silicon in Aluminium-Silicon Alloys1 By John D. Gat 405 Se c o n d St ., N. W., Ca n t o n , Oh io N ANALYTICAL chemist who deals with this class of A alloys is sometimes at a loss when his results do not agree, and although all probable variables have been kept an alloy silicon is present in a state which, after proper treatment, will yield 100 per cent of its graphitic variety. The fact that some silicon dioxide is always present finds its practically constant, he frequently desires considerably meoxreplanation in the oxidizing power of the factors entering a accuracy. The general rule in such a case is to attribute the routine analysis, which is sufficiently intense to convert par discrepancies to the lack of uniformity of the sample. Though ticles of silicon that are very minutely divided--as, for exam the lack of uniformity in the aluminium-silicon series is some ple, in solid solution--into silicic aoid during the compara times very pronounced, a wrong assumption seems to be re tively short time used in an ordinary analytical determina sponsible for devising the methods now in general vogue, and tion. which are therefore inherently inaccurate. If, being powerless to increase materially the concentration In textbooks of inorganic chemistry and, to the author's of the acids employed, we intensify other variables affecting knowledge, in all treatises on analysis, it is generally stated oxidization, such as time and temperature of ignition, we shah that so-called "graphitic" silicon obtained in the course of observe in a given sample a gradual transformation of the analysis cannot be oxidized to silicon dioxide by ignition and ratio of silica to graphitic silicon. The percentage of silica will remain unaffected when treated with a mixture of hydro be continuously increasing and, under proper conditions, no chloric, nitric, and sulfuric acids generally used for decompo graphitic silicon will be present. The intensity of the nec sition of the sample. During some investigations on the essary conditions is in close relation to the silicon content of microstructure of aluminium-silicon alloys undertaken in the sample, comparatively little depending on the heat this laboratory, certain lack of coordination between the struc treatment to which the alloy is subjected- ture of metal as seen under a microscope and the results of It is quite impossible to consider the amount of silicon in a analytical determinations led to a closer study of the proper given sample present in the graphitic state as an absolute ties of silicon, especially its graphitic variety, and resulted in value. If no standardization of analytical procedure is made, two statements, the accuracy of which was proved: the results obtained will be in inadmissible discord. If all 1--Graphitic silicon is oxidized to silicic acid when heated with a mixture of hydrochloric, nitric, and sulfuric acids, 2--Graphitic silicon rapidly oxidizes when subjected to the temperature and for the length of time necessary for dehydra tion of silica. The amount of oxidation is closely related to the temperature and time of ignition. In other words, the amount of graphitic silicon in a certain details are carefully standardized and closely adhered to, one may be sure of the amount of graphitic silicon, but for a dif ferent set of conditions, a different value will result. Microscopic examinations and study of cooling curves support the theory that silicon is present in alloys in one modification, graphitic silicon, which as such, instead of being an ingredient the presence of which may indicate certain alloy is not a definite quantity, but a function of the composi properties of the alloy, becomes just an undesirable complica tion and amount of acid mixture used for solution of the sample, temperature, and time of ignition. With a considera ble degree of certainty the theory may be advanced that in tion in analytical work. A method was devised whioh eliminates from considera tion graphitic silicon and, without being long or elaborate, * Received Mcurch-3, 1924 gives dependable data for routine determinations: