INDUSTRIAL A N D ENGINEERING CHEMISTRY
There is a tendency on the part of the stocks containing some retarders to give a greater percentage decrease in temperature coefficient of scorching over the range studied than does the control stock. With other retarding materials this does not hold. A scorch-retarding material is best measured by its effect upon index numbers of a control stock rather than by its effect upon the temperature coefficient of scorching. Acknowledgment
The author wishes to thank A. J. Crawford for work done on measurement of rate of temperature rise and fall and calcula-
Vol. 23, No. 12
tion of theoretical temperature coefficients of scorching. Thanks are also due L. B. Sebrell and R. P. Dinsmore for permission to publish this paper. Literature Cited (1) Cadwell, S. M., U. S. Patents 1,778,707-709, incl. (Oct. 4. 1930). (2) DeFrance. M. J.. and Krantz. W. J.. I N D . ENG CHBM.28, 524 (1931). (3) Morse H . B , U. S. Patents 1,734 633-640, incl. (Nov. S, 1929) (4) Park, C. R.. Paper presented before 82nd Meeting of American Chemical Society Buffalo N. Y.. Aug. 3 1 to Sept 4 1931 (5) Somerville. A. A., U. S. Patent 1.791.876 (Peb. 10, 1930). (6) Thies, H. R.. INO. ENG.CIIBM., 40, 1223 (1928).
Compatibility Relationships of Aroclors in Nitrocellulose Lacquers' Russell L. Jenkins and Robert N. Foster SWANNRESEARCH, INC..ANNISTON. ALA.
N THIS paper are pre-
This paper deals with certain physical properties of preciable h a r d e n i n g upon the chlorinated diphenyl products, known commersented preliminary data heating to moderately high cially as aroclors. The aroclors are described, and a obtained on the aroclors temperatures. discussion is presented of their stability to light, to or polychlorodiphenyls, esIf the series of aroclors of hydrolytic influences, and to heat. Data are given pecially as rega.rds their comprogressively increasing visshowing the relative permanency or vaporization rates patibility with nitrocellulose cosity is considered, the folof the aroclors and the more common plasticizers and and with systems containing lowing p r o p e r t i e s increase softeners. A table is given of the physical properties nitrocellulose and resins or when ascending the series: of the aroclors used in this investigation. density, r e s i s t a n c e to displasticizers. The main part of the paper deals with the compaticoloration by sunlight, softenThe nitrocellulose-lacquer bility relationships of the aroclors with common ining point, tendency to crysfield lends itself more readily gredients of lacquers. The relative effectiveness of tallize, flash and fire-points, to scientific approach t h a n different plasticizers in increasing the limiting comand lack of odor. the other protective-coating patible ratios of aroclor to nitrocellulose is shown On the other hand, in defields, because p r a c t i c a l l y graphically. Trilinear diagrams are given for a numscending the series from the its e n t i r e development has ber of three-component systems containing 1/2-second more viscous aroclors to the occurred within recent times. nitrocellulose, an Aroclor, and certain typical resins. less viscous ones, the following It is less t r a m m e l e d with propertiesincrease: volatility, tradition and the art of comDoundina. and it has already been systematically invee- ccmpatitility with nitrocellulose, and solubility in solvents. I nthe lacquer field the choice of suitable aroclors for investitigated It is fortunate that this is-the case when new materials like the aroclors appear on the market because it is gation depends on a balance of some of the above-named possible to study the properties of the new substances with, properties. The maximum compatibility with nitrocellulose respect to the lacquer field and to determine with some is desired; the volatility must be low enough to give perdegree of certainty the best methods for the formulation of manency to the aroclor in the lacquer film; good solubility in lacquer solvents is necessary, and presumably the resistance to lacquers from the new substances. It is already well known that the aroclors are chlorinated discoloration or change by ultra-violet light should be as high diphenyl products, varying from water-white mobile oils as possible. I n certain cases a low density may be desirable through viscous oils and soft waxes to pale amber brittle for economic reasons. Taking these and other factors into account, aroclors 1254 resins and opaque crystalline solids. The aroclors are nondrying. They are not polymerized or condensed products in and 1262 have been chosen as the most suitable for investigathe ordinary sense of this term, nor do they show any ap- tion. Where less permanence is not objectionable, aroclor 1242 is of interest because of its greater compatibility with Presented before the Division of Paint 1 Received August IS, 1931. nitrocellulose. Aroclor 1819 is regarded as too volatile for and Varnish Chemistry at the S2nd Meeting of the American Chemical lacquer work except in special cases. Society, Buffalo, N. Y.,August 3 1 to September 4, 1931.
Table I-Physical Color (Lovibond yellow: 1 inch depth of liquid) pergram Maximum acidity (mg. NaOH per gram of aroclor) Viscosity, seconds Saybolt at 100' C. Distilling range (cor.) (cor.), C. Pour point (A. S. T. M.), pour M . ) , 0' C. Freezing point, e C. Specific gravity at 2eo/25O C. g Coefficient of expansion, cc./cc./" C. Coef >d, e C. Flash poi point, Cleveland open-cup method, Fire point n&-v Fire C. Refractive' index (n D at 27' C.) Refr 0 At 99' C. (210' F.). b At 90"/90° C. --'-I
Data on Aroclors Used in This Investigation
1819 0.2 0.005 30 278-295
14 1.15 0,00079 138 177 1,6125
1242 0.3 0.004 34a 320-360 Minus 16 or lower
Non-crystalline at 0
1.38 0.00073 Above 165 (indistinct) Above 300 1,6248
0.5 0.01 46 366-399
0.9 0.02 96 388-431
Not higher than 10 Non-crystalline at 0
N o~... ne .
INDUSTRIAL AND ENGINEERING CHEMISTRY
Aroclors 1819 and 1242 are water-white mobile oils, aroclor 1254 is a pale yellow viscous oil, and aroclor 1262 is a soft sticky wax a t ordinary temperatures. Typical data covering the physical properties of these aroclors are given in Table I. Chemical Stability
The aroclors are very stable chemically. They are not subject to hydrolysis under any ordinary conditions, nor do they become rancid. The aroclors are very resistant to oxidation, and they are strictly non-drying in their characA NITROCEUULOS.
Compatibility of Aroclors with Nitrocellulose
The aroclors are not solvents for nitrocellulose, but they are compatible with it up to certain definite limits. The compatibility limit of each aroclor was determined by the usual method. Solutions were made of equal concentrations of the aroclor and of '/*-second R. S. nitrocellulose. These two solutions m-ere mixed in the proper proportions to give a series of lacquers containing increasing ratios of the aroclor and nitrocellulose. Samples of each lacquer were dried slowly on glass or bright tin sheet. The appearance of the series of dried films indicated a t once the limiting compatible concentration of the aroclor, the films containing less than the limiting concentration being clear and those containing slightly more than this limit being opalescent by reflected light and brownish by transmitted light. The series of films have been inspected after aging for 4 months a t room temperature. No difference in the results has appeared. The limiting concentrations of the aroclors were found to be as follows: AROCLORS
Figure I - C o m p a t i b i l i t y of Aroclors w i t h Resins a n d l/t-Second Nitrocellulose Resin Aroclor Solvent I Dammar 1262 1 I1 Dammar 1254 1 I11 Dammar 1284 2 IV Rezyl-12 1254 1 Rezyl-12 1254 3 V 1262 1 VI Kopol-A 1254 1 VI1 Kopol-A VI11 Glyptal-1202 1262 1 IX Glyptal-1202 1254 1
1242 1254 1262
50 to 60 40 32
NITROCELLULOW Ratio 50 to 40 About 1 1 eo 2 3 e8 8: 17
For the determination of the compatibility limits, a single efficient slow-evaporating solvent for both the aroclor and the nitrocellulose was chosen. This choice was made after it ,became evident that mixed solvents containing different types of high-boiling solvents gave radically different results. Thus a mixed solvent, containing cellosolve as the high-boiling solvent, gave the compatability limit of aroclor 1254 as 12 per cent, whlle with a high-boiling acetate (amyl or butyl acetate) the limit was 40 per cent or slightly above. A
teristics. Unlike many other synthetic materials that have similar uses, they are not subject to polymerization. They may be heated indefinitely at, €60' C. and a t higher temperatures, a t least for short periods, without undergoing any apparent change. Arocloh 1254 and 1262 are non-flammable and retard the rate of burning of nitrocellulose films. Some data covering their chemical stability are given in Table 11. Tests of Certain Aroclors INCREASE IN FREE CHLORIDES AFTERHEATING 4 IN EXPOSED IN QUARTZ DAYSAT 204' C. NaOH 48 HOURSTO (400' F.) SOLUMERCURY ARC Colora Acid No. TIONb Colora Acid No.
BEFORE TEST ARO- Col- Acid CLOB Ora NO.
% 1819 1242 1254
0.0037 1/z 0.0150 0 1/4 0.0043 0:0033 1/z 0.0096 1/4 ,0.0087 Slightlyless 0.0112 0.0061 Slightlyless 0.0113 than 1/z than 1/1 1 5 2 v//( 0.0030 Slightlyless 0.0047 0.0029 a/, 0.0075 than 1 5 Relative color was determined by comparison with N. P. A. Standards. Q The aroclor was boiled in each case with 8 times its weight of 10 per cent sodium hydroxide solution for 8 hours. The chloride content was determined by a turbidity method.
Solubility of Aroclors in Various Solvents
The aroclors are very soluble in practically all of the commonly-used organic solvents, except those that contain some water or those that contain more than one hydroxy group, as glycerol for example. While the aroclors have a limited solubility in ordinary ethyl alcohol, which contains 5 per cent of water, they are readily soluble in absolute alcohol and in the balanced-solvent mixtures that are in everyday use in-the lacquer industry.
Figure 2-Compatibility of Aroclors w l t h Resins and '/:-Second Nitrocellulose Resin Aroclor Solvent X Ester gum 1262 1 XI Ester gum 1254 1 XI1 Ester gum 1254 4 XI11 Amberol 200-B 1254 1
It seems reasonable to suppose that the limit observed, when a mixed solvent is employed, depends mainly on the type of the high-boiling solvent that is last to evaporate from the film. It would follow then that the same compatibility limit would be obtained with any properly balanced mixed solvent in which the last volatile solvent to leave the film is a high-boiling acetate. This conclusion appears to be amply supported by the experimental results given in Figures 1and 2. In these figures all of the curves converge to the same point on the AC line in those cases where the high-boiling component of the solvents was butyl or amyl acetate.
Vol. 23, No. 12
INDUSTRIAL A N D ENGINEERING CHEMlSTRY
Effect of Plasticizers and Other Lacquer Constituents on Compatibility Limit
method was employed as in the simpler binary system of nitrocellulose and an aroclor. Most of the test films, on which the The limiting compatible ratio of an aroclor to nitrocellulose compatibility curves are based, have been inspected after may be increased by the addition of plasticizers or of certain aging for 3 months or more at room temperature, and no resins. The increase in the limiting ratio is conveniently change in the results has been observed. The curves are shown in Figures 1 and 2. In these curves shown in Figure 3. Lindol (tricresyl phosphate), unblown the areas A, to the left and above each curve, include all those compositions of the three non-volatile components that give I.0 clear unblushed films. The areas B , to the right and down, B give blushed films or films that sweat. In the systems in1.6 vestigated non-compatibility almost always appeared as an i? g 1.0 opalescence or a white blush, rarely as a sweating out of the film. The curves may be classified into two distinct groupsw M w those of the dammar or glyptal type (Figure I), and those of 9mm.w p w mu - m u m s MuEMu w A, pnFRochWrnosF, urn *RtCm. the ester-gum type (Figure 2). The first type of curve drops Figure 3-Relative Effectiveness of Lindol, Castor Oil, a n d Dibutyl Phthalate in Itlcreasing immediately from the AC line and provides a wide compatible Limiting Compatible Ratio of Aroclor 1262 t o area. The second type rises from the AC line before it drops Nitroc e 11u 1o8e (Films aged 2 days) and consequently gives a comparatively narrow compatible field. The resins which give rise to the two types are as castor oil, dibutyl phthalate, and dammar all increase the follows: limiting ratio, while ester gum decreases it.
The relative volatility of the aroclors and of three common lacquer constituents as determined by the A. S. T. M. Standard Method of Test for Loss on Heating of Oil and Asphaltic Compounds (D6-27) is as follows: SUBSTANCE
Aroclor 1242 Dibutyl phthalate Aroclor 1254 Aroclor 1262 Castor oil Lindol
6.2 3.98 1.70 0.88 0.60 0.31
TYPE1 Dammar (dewaxed) Rezyl 12 Kopol A Glyptal 1202
Ester gum Amberol 200-B
The first class of resin provides an ample area for the preparation of most types of lacquer. Especially is this- true , when one of the standard plasticizers or softeners is also added in small amounts to the mixture. On the other hand, the ester-gum class of resin apparently requires the addition of one of the common plasticizers in order to obtain an appreciable aroclor-nitrocellulose ratio. I
The above results were obtained on the pure substances. The relative vaporization rates from films containing 30 per cent of the substance under test and 70 per cent of llrsecond nitrocellulose were also determined and are shown graphically in Figure 4. The results given in Figure 4 are the average of check determinations. The method consisted in weighing onto a bright tin panel equal amounts of the test solutions. The weighings were made by difference from weighing bottles provided with medicinal droppers. The test solutions consisted of a mixture of nitrocellulose and the aroclor or other substance under test dissolved in butyl acetate. The solutions were spread on the panels so as to cover equal areas as nearly as possible. These films were air-dried overnight and then subjected to drying for 21.5 hours in a forced draft at room temperature (about 21' C.). From this time until the end of the tt-st (59 hours) the films were evenly subjected at 34" C. to a blast of air. At the end of the test the areas of the films were measured with a planimeter, and the changes in weight corrected for area.
The curves shown in Figure 4 indicate a greater volatility from the lacquer film for aroclor 1242 and seem to indicate a lower volatility for aroclors 1254 and 1262 than for dibutyl phthalate, castor oil, or lindol. However, the results are complicated by an apparent retention of solvent, particularly by aroclors 1262 and 1254. The films containing these substances did not attain, during the time of the test, the theoretical weight of the dry film. The results of the standard A. S. T. M. test for volatility and of the test for volatilization from nitrocellulose films indicate that aroclors 1254 and 1262 are as permanent as, or more so, than dibutyl phthalate. Compatibility in Three-Component Systems The compatibility of aroclors 1254 and 1262 in several three-component systems-comprising an aroclor, a resin, and I//rsecond R. S. nitrocellulose-was investigated. The same
Figure &Loss of Weight on Accelerated Drying of Films Containing 70 Per Cent of Nitrocellulose a n d 30 Per Cent of Material Indicated B-Theoretical weight of films, assuming complete evaporation of solvent and no loss of nonvolatiles
The solvents used in this investigation are indicated by a solvent number in Figures 1 and 2. The composition of these solvents is as folloys: SOLVENT COMPOSITION 1 Technical amyl acetate 2 50 parts toluene 6.4 parts alcohol (specially denatured, formula 3-A), 6.4 paris butyl alcohol, 6.4 parts ethyl acetate, 30.8 parts butyl acetate 3 70 parts toluene 10 parts alcohol (specially denatured, formula 3-A), 10 parts'ethyl acetate, 10 parts cellosolve 4 65 parts toluene, 10 parts butyl alcohol, 25 parts butyl acetate
The effect of different types of solvents on the compatible area is clearly indicated by the curves. Curve V (Figure l), as compared t o curve IV, shows a radical change in the compatible area caused by the change from a balanced solvent (No. 3), containing cellosolve as the high-boiling solvent to an acetate type of solvent (No. 1). On the other hand, curve I11 for dammar and aroclor 1254 in a balanced solvent, where butyl acetate is the high-boiling solvent, is identical with curve I1 for the same, where amyl
acetate alrme was used, axing to tiic fact that tlic last srilvent to leave the film in each case was a high-boiling ucetat,e. In tlie special case of the ester gurn--arocIor eiirve, it WLS hopxi that tbe URL' tif a niixi:d solwnt containing a rdativeiy liigh-l.~oilingalcohol along Tvitli butyl acetat,e ~roiildiiiaterially increase the conipatilJle area. ILmever, curve 5 I1 (Icig.ure 2) provides only a slightly grcat,er useful area than tlie corresponding curve SI ahere a siinple ai&& a h i c as cniployed as t h e solvent. The formulation of lacquers, based on tlie data here given, and also extended esposure and other tests on finiiiiiedlacquers are uridcr way, but any report. on this work would be preinstiire at this t,irnr. Tiie a r d o r s are not regarded as exact substitutes for any of the present constitueuts of Iacqiicrs. They have some uf the properties of a softencr, and at least aroclor 12fi2 has certain properties of n resin. The fundaniental data presented in this papcr arc intended to answer sonic of t,he questions frequently put by investiga-
tors. The cuioput~ibilitydiagrams are useful iir avoiding forinulntion of lacquers containing an amount of 8x1 aroclor beyond the cornpatit,le limit. In working with pigmented lacquers, such a condit,iiiri may readily arise witliout the knowledge of the formuliitiir, arid the ftiilore of tlie film may he ihed to the aroclur, whercas the real difiiculty lies in the use of too mu& of it. Tlie data in this p p e r also provide the starting p i n t for e. systematir iiwest.igation of the aroclors as lacquer constituents. Acknowledgment
Tlie writers are indebted to H. W. Klinger arid J. K. Speicher of t,lic Ilerculns Powder Company for advice and
suggestions during tlie ciiurse of this work. Literature Cited W. x , and ~ d d E. , (1928);ai, e55 (iezvj.
From New Research Tools Come the Tools of Industry
Couribsy of Boidwin-Southwork Corpowion
Southwark-Emery Spring-Teetlng Machlne. Standard Steal Worlis, Buroham, Pa.
Additional articles presented at the Symposium on New Research Tools appear in INDUSTRIAL AND CHEMISTRY, 23, 1223-1246 (1931).