Packaging of Shellac Varnish - American Chemical Society

plate cans have been used widely, therefore, for shellac varnishes, and experience has shown them to be superior to any other metal-lined package whic...
0 downloads 0 Views 537KB Size
INDUSTRIAL AND ENGINEERING CHEMISTRY

1178

The necessity of having an elevated temperature, excess calcium salt, and low water content in order to obtain a good conversion to soluble strontium salt is explained on the basis of this theory. CONDITION OF CALCIUM SULFATE. The calcium sulfate formed by the reactions between strontium sulfate and calcium chloride and between strontium sulfate and calcium nitrate was precipitated a t a temperature above 150"C. from a n aqueous solution containing a high concentration of soluble calcium salt. The transition temperature of CaS04.2HzO into CaS04is 60" C. which indicates that the stable phase is Cas04 above 60" C. The high concentration of calcium ions in the aqueous reaction media also tended to precipitate the calcium sulfate without water of hydration. These conditions of the aqueous reaction media indicate that the Cas04 formed in the above reactitm does not contain water of hydration. The extraction solvent for the reaction between strontium sulfate and calcium chloride was pure methanol. After the calcium sulfate precipitate from the reaction mass was washed with methanol according to the procedure, i t was dried a t 50" C. to constant weight. The precipitate was then heated a t 200" C. for 2 hours. The vapors were condensed and analyzed for methanol and water. The liquid that came off a t 50" C. was found to be pure methanol by determining the density with a pycrometer. As the temperature was raised to 200" C., the cake did not lose any more weight. From these observations it is concluded that the calcium sulfate was precipitated without water of hydration and that it stayed in that state throughout the operations involved. In subsequent experiments the extraction solvent for the second reaction was a mixture of methanol and water. The extraction was carried out above 60" C. The calcium sulfate precipitate was treated as before. It lost no more weight

(e),

VOL. 31, NO. 9

after being dried a t 50" C. for 3 hours, and the condensed vapors from this initial drying were pure methanol. T h e same conclusion may be drawn as in the case of the chloride reaction; that is, the calcium sulfate does not contain any chemically bound water.

Acknowledgment It is a pleasure t o credit the Mallinckrodt Chemical Works. of St. Louis for substantial help in supplying chemicals and apparatus used in this study.

Literature Cited Browning, J. C., thesis, Purdue Univ., 1937 [cf. Shreve, Watkins, and Browning, IND.ENQ.CHEM.,Anal. Ed., 11, 215 (1939)l. Farr, H. V., U. S. Patent 1,752,244 (March 25, 1930). Landolt-Bornstein, Physikalisch-chemische Tabellen, Band I, p. 683 (1923). Ibid., p. 644. Linder, F., German Patent 460,572 (1924). Mactear, J., British Patent 1916 (1886). Partrjdge, E. P., and White, A. H., J. Am. Chem. SOC.,51, 360 (1929). Seidell, A., "Solubilities of Inorganic and Organic Compounds," 2nd ed., Vol. 1, p. 215, New York, D. Van Nostrand Co., 1919. Shreve, R. N., and Pritchard, W. N., IND.ENG.CHEM.,27, 1488 (1985). Shreve, R. N., Pritchard, W. N., Farr, H. V., Still, A. J., and Crosby, J. D., U. S. Patent 2,030,659 (Feb. 11, 1936). Uhde, R., German Patent 412,698 (1923). Virck, Chem. Zentr., 1862, 402; International Critical Tables, Vol. VII, p. 343 (1930). Williams, P. E., and Briscoe, H. T., Chem. News, 145, 177 (1932).

Wright, J. S., thesis, Purdue Univ., 1936. P R E S ~ N Tbefore E D the Division of Industrial and Engineering Chemistry a t the 96th Meeting of the American Chemical Society, Milwaukee, Wis. This paper is a portion of a thesis submitted by Charles H. Watkins i n partial fulfillment of the requirements for the degree of dootor of philosophy a t Purdue University.

Packaging of Shellac Varnish PAUL F. BRUINS AND WM. HOWLETT GARDNER Polytechnic Institute of Brooklyn, Brooklyn, N. Y.

A

METAL container is one of the most serviceable types of package, if it does not contaminate the varnish. Terneplate cans have been used widely, therefore, for shellac varnishes, and experience has shown them to be superior to any other metal-lined package which might be used for this purpose. Eight-pound terne plate is the most common type employed, but terne-plate manufacturers cannot guarantee freedom from defects such as pinholes. Some consumers are now using 15-pound terne-plate cans in an effort to correct this difficulty. Even these cans are subject t o certain defects.

Cause of Contamination Examination of a large number of terne-plated cans of different manufacture revealed that many of them had several small defects, besides pinholes, which induced corrosion between the metal and the varnish. The number of defective containers was surprisingly high. Some cans even showed signs of iron rust before any varnish had been placed in them. This was said to be due to water which had been left in the cans after they had been tested for possible leaks. Even distilled water was found to produce a rapid deteriora-

tion of the coating in several instances where defects were present. Corrosion was most prevalent a t the point where the spout is soldered onto the top of the can, Figure 1, A , shows. This is the weakest point of the structure of many types of terneplate containers as far as corrosion is concerned. It is extremely difficult in the manufacture of these cans to cover this bare iron edge completely with solder. Rapid corrosion was not confined entirely to this one weak point. Many cans showed that corrosion took place first where the protective lining had been either weakened by oxidation during soldering or by forming operations which produce a higher tension at certain points than in the surrounding metal. Other cans showed rust starting as small specks on the sides of the cans, which was taken as evidence that the plate had contained pinholes. Still other cans showed corrosion in long streaks where the soft terne plate had been scratched during fabrication. The gradual breakdown of the terne-plate coatings of various cans filled with shellac varnish was clearly due to electrolytic reactions similar to those common to corrosion in general. Corrosion proceeds at a much faster rate when

SEPTEMBER, 1939

INDUSTRIAL AND ENGINEERING CHEMISTRY

the two metals, iron and terne-plate alloy, are exposed to the varnish than when iron alone is placed in the resin solution. However, the terne plate is also corroded by shellac varnish a n d forms insoluble lead and tin salts which settle to the bottom of the can, This may be due in part to the condition of the terne plate where mechanical stresses have produced local areas of a different physical state, so that the coating may act as both electrodes in an electrolytic cell. These corrosion reactions probably involve a partial deterioration of the resin present in the varnish, as well as the formation of heavy metal soaps. Judging from the strong acidic odor of the varnish in cans where excessive disintegration of the container has taken place, oxidation of the alcohol which was used as solvent may have occurred also. The same varnishes packaged in glass did not undergo these changes. Freshly cut bleached shellac varnishes showed no signs of containing any mineral acid even when special tests were used for detecting their presence (1). Marked evidence of corrosion of the can occurred with these varnishes before any detectable change was observed in the same types of varnishes stored in glass. When exposed to iron, the bleached shellac varnishes darkened rapidly. The ferric salts formed first produced a gel which was rapidly peptized.

Effect of Contamination When present in a varnish, these contaminating ferric salts ,can react with tannic acid and similar compounds present in hardwoods to produce black compounds similar to those used in the manufacture of some inks. The occasional blackening or graying of floors when a shellac varnish is applied to bare wood such as oak can invariably be traced to the use of a varnish which has been taken from a corroded container. Considerable expense is involved when this occurs since such floors have to be refinished completely. The black compounds formed tend to penetrate the wood, and complete resanding a n d scraping are necessary. The large use of maple flooring probably accounts for the fact that there are not more instances of the above type. Complaints are most apt to come from sections of the country where humidity conditions are such that cans may show signs of iron rust before the varnish is placed in them. Cans should alFIGURE1. POINTSOF MOST PREVALENT CORROSION ways be carefully inspected just before packaging for this A . Soldered work B. Pin holes reason. The presence of iron C. Scratohes D . Seams in a shellac varnish is always obiectionable. even when it is not applied to a bare wood sirface because a dark color is imparted to the varnish.

Corrosion Resistance of Various Coatings The standard weight of coating of terne plate which is employed for shellac varnish containers is 8 pounds per base box. Fifteen-pound terne-plate coatings have been found to be more resistant to corrosion but have not been generally adopted because the degree of improvement did not warrant the increased cost. Furthermore, the heavier coatings were easily damaged during fabrication and were not so suitable a

1179

The presence of soluble iron salts in shellac varnishes is the cause for the occasional graying or blackening of floors when these varnishes are applied to bare hardwoods, such as oak. The contamination of the shellac varnish has been repeatedly traced to imperfections in the terne coatings of the metal containers used for packaging it. Standard metal packages containing shellac varnishes were tested for their corrosion resistance. Contamination with iron could be invariably detected after 2 to 5 months. Protective coatings of various other materials besides terne plate were investigated. The most satisfactory packages were those which had received a coating of modified wax after complete fabrication of the can. This extra coating was readily applied at very little extra expense. Double protection was provided against contamination of varnish when the wax coating was applied over terne plates.

base for lithographing the exterior of the containers. The exposed iron surfaces a t the nozzle of the can were not eliminated, which proved to be as serious a source of contamination as with the lighter coatings. The use of organic coatings of various types over terne plate or over bare iron sheeting was investigated also. Coating the flat sheets before fabrication was found to be entirely unsatisfactory for this purpose. I n like manner, coating the containers after fabrication with various types of lacquers and enamels was not practical without a complete change in the design of the different types of containers. Tests were made of a number of lacquers and baked finishes, which are commonly employed for cans of various uses, but they were all rapidly attacked by the shellac varnish. Aluminum containers were tested also but were found to react rapidly with the shellac to form a gel. The only satisfactory type of containers were those coated with certain modified mineral waxes. It was found that these materials could be applied readily in the molten state to the container after it was fabricated. The metal surfaces did not require a special cleaning, and no difficulty w&s encountered in obtaining fairly uniform, continuous films of wax for cans of various designs.

Wax-Lined Cans

A number of modified mineral waxes were investigated for this purpose. The most satisfactory was one manufactured by the Quaker Chemical Products Company and sold as number 351. It was completely insoluble in shellac varnishes, in addition to imparting excellent corrosion resistance to the containers. Coatings remained intact over a temperature range from 50" to -20" C. and showed no tendency to crack, chip, or peel, even when the cans were subjected to such distortions as might occur during severe use. An extensive series of tests was undertaken to compare the behavior of standard 8-pound terne-plate and tin-plate

INDUSTRIAL AND ENGINEERING CHEMISTRY

1180

VOL. 31, NO. 9

cans were corroded by regular 5-pound shellac varnish within 5 months. and in less time if the Crams terne plate was defective. Under identical concan Of WBX Test for Iron io Varnish,&Months cansine N ~ .~ ~ ~ it ~ 2 ~ 3. n 4 5 6 7 s 9 7 ditions the use of a thin wax coating had preVarnish vented any corrosion from taking piace for 8 0 0 0 0 Pint months and, during the life of the test period, 0 0 0 0 o o o x prevented the varnish from becoming contami+ + + + nated with a dangerous amount of iron salts. o x x a Quart o x a x Since bleached shcl1a.c varnishes usually develop o a x r 0 r r r poor drying properties, this degree of protection O X X X was considered completely adequate in this $ $ ? $ period of time. Similar protection was imparted. to the tin plate, as Table I1 shows. + + + + o o o o 0.5~~1. The tests made with French varnish indicated ++0 ++0 ++0 i +0 - that it was more corrosive than regular bleached shellac. However, the presence of a reducing o o o o 1 gal. 0 0 0 r reagent, oxalic acid, caused the iron content of ++ ++ ++ ++ the varnish to vary in definite cycles as the corrosion of the can progressed. Wax-coated cam for o x o o Pint French varnish prevented any t,race of corrosion X X 0 0 0 0 0 0 for 6 months, with the exception of one cono x o o tainer which showed slight signs of corrosion. 0 0 Qw.rt 0 0 : : The varnish in the uncoated standard terne-plate 0 o + cans, witli one exception, became badly eon8 0 %r %o o0 00 laminated with iron salts during this period. 0 O f 0 Many of the uncoated cans showed signs of 0 0 0 0 corrosion even &s early as 2 months. Similar re0 0 x 0 6.6 o o o o o o o 0 o o o sults were obtained with the tin-plated eono.s %SI 35 % $$ !? " ! " " ! ?, ! tainers (Tables I and 11). "' ...-. " " " 3 8 N . C . O O O O O O O + x a + The cans were cut open for inspection at the 3912.5 0 0 0 0 0 0 0 0 0 0 0 s gsl. 40 s0.9 o o o o o o o 0 o 0 conclusion of each test (Figure 2). The terneor ! ! ! ! tin plate of the uncoated cans was badly etched in all instances; the wax coatings on the coated 0 N.C. = not coated with wax. b 0 = no test for iron BBLtsi x = trnne of iron salta preaenti + = PoBitiVe teat for iron. cans appeared to he in its original condition except in a few instances where certain spots liad received a very thin coating. The metal plating cans with the same cans lined with a coating of this wax. beneath the wax coating was always bright and undected.. Tests were made under conditions duplicating those of actual Production Methods for Applying the W a x service. A study was made in cooperation with the American Can The cans were filled with a standard £ cut of regular Company Of XWiOUS methods Nhich might be used in tho com-. bleached in one series of tests and with 5- ound French mercial production of wax-coated cans. Available mechanivarnish (dewaxed bleached shellac) in another.eh! ' cans were sealed and stored at room temperature, which fluctuated between 15' and 35" C. The wax coating was applied to cans in a molten condition. It was heated to 120' C. and then poured into the cans. The cans were rotated and then inverted, to allow the excess to drain out. In this manner all of the surfaces, the crevices. and the nozzle were coated. Each can was wkighd before and after coating to determine the amount of wax applied. The cans were impected once a month after being filled, and the varnish was tested for the presence of iron salts by the addition of throe drops of a. 10 per cent solution of ammonium polysulfide to a I-ml. sample of the varnish. This test detected the presence of dissolved iron salts in the ratio of 1- part to 20,000 parts of varnish. Black sulfides are instantly formed, if iron salts are oresent. The ~rosence of this amount of iron n a s an aecurate'indiextion of the beginnin of corrosion of the container. Although &is amount of contamination did not cause bare oak to darken. the iron content of the varnishes inrrettsed rapidlv after snch corroTABLE I. WAX-COATED &POUND TERNE-PLATE CANS

+ + z +

I

2; E:$:

:

:

L1

: :

:

jj

0"

: : I

:

solved ferric iron in a ratio of 1 part to 500 parts of varnish; when this occurs, the varnish cannot he applied safely to bare wood unless a reducing or precipitating agent has been added (e).

Table I shows the results of a series of tests for a period of 11 months. The data show that all 8-pound terneplate

LI

+"

cal devices could be used in coating the fabricated cans by tile flushing method such as was employed in the laboratory for coating the above containers. This method had the advantage that all corners and crevices were thoroughly coated. It w&san economical method of application.

SEPTEMBER, 1939

INDUSTRIAL AND ENGINEERING CHEMISTRY

TABLE11. WAX-COATED TINPLATE CANS Can Size

Quart

0.5gal.

Quart

O5gal.

Grams Can of Wax No, Coating5

-Test of Iron in Varnish,b Months 1 2 3 4 5 6 7 8 9 1 Containing 5 Pounds of Regular Bleached Shellac Varnish 43 6,8 44 6 . 1 0 0 0 0 0 0 0 0 x 45 7 . 7 0 0 0 0 0 0 0 0 0 46 N . C . 0 0 O + + + + + + + t 47 4.8 0 0 0 0 0 0 0 0 x 48 5 0 0 0 0 0 0 0 0 O x 49 6 . 1 0 0 0 0 0 0 0 0 0 5 0 N . C . O O O O + f + + f + Containing 5 Pounds of French Varnish 0 0 0 0 0 0 0 0 0 51 4.8 52 5 . 0 0 0 0 0 0 0 0 0 0 53 6.1 0 0 0 0 0 0 0 0 5 4 N . C . O O O O + x $ + + $ 55 4.3 0 0 0 0 0 0 0 0 0

,

5 .6 7 0 0 0 0 0 N.C. 0 0 0 0 0 N.C. = not coated with wax. 0 = negative test for iron: x = trace of iron present; 56 57 58

b

0 0

0 0

0 0

0 1 1 x 0

0 0

0 0 0 +

x 0

0 0

0

0

+ 0

0

0

0

0

+ = positive test for iron.

Cans were also coated by being sprayed with either the molten wax or a solution of it in a suitable hydrocarbon solvent. These methods gave some trouble because of the difficulty of obtaining uniform, smooth, adherent coatings, but this was c ~ ~ r e c t eby d Passing the coated cans through a dryer maintained slightly above 70" c.7 the melting ternPerature of the wax. Cans coated in this mmner Protected

0

0

0 -

t

the shellac varnish equally as well as those coated by the flushing method. It was not considered practical or advisable to coat the flat terne-plate sheet with wax before fabrication into cans, since the coating would be too easily damaged during the manufacturing process. Soldering would also destroy the wax film. The cost of the wax used in coating the different sized cans was small, as Table I11 shows. These costs were based upon a price of 10 cents per pound for the wax. A low application cost for coating these cans will obviously depend upon a quantity demand for this type of container. Other products besides shellac varnish could probably be packaged to advantage in this type of coated can. Waxlined containers for beer are already commonly used. Double protection of the varnish is as-. sured by applying the wax to terne-plated containers. The extra protective coating can be applied as well to lithographed cans. - -

Acknowledgment The authors wish to thank those connected with the American Can Company for their cooperation in making this a complete study of the problem, and to express their appreciation to the Quaker Chemical Company and to the many can manufacturers who kindly supplied samples of their products and gave many helpful suggestions.

FOR LINING CANS TABLE111. COSTOF WAXREQUIRED

Size of Can Pint Quart 0 . 5 Gal. 1 Gal.

Av. Wt. of Wax, Grams 3 4

7

12

cost, Cent 0.066 0.088 0.155 0.286

1181

Literature Cited (1) De Sylva, O., thesis, Polytechnic Inat. of Brooklyn, 1932. (2) Verman, L. C., and Bhattacharya, R., London Shellac Research Bur., Tech. Paper 8 (1936).

PRESENTED before t h e Division of Paint and Varnish Chemistry a t the 96th Meeting of the American Chemical Society, Milwaukee, Wis.

SEPARATION PROCESSES Fractionation of Partially Miscible Liquids' MERLE RANDALL AND BRUCE LONGTIN University of Caliiornia, Berkeley, Calif.

The distillation of binary mixtures of the minimum boiling and the eutectic vaporization types is discussed. A design method is given for a particular type of equipment for such a distillation. The design method may easily be generalized to other types of equipment.

l-

N PREVIOUS papers (3, 4) general methods of analysis

of separation processes were presented. These methods are free from the restrictions of the usual simplifying assumptions. I n this paper an important application of these methods is discussed.

The molal heat content vs. mole fraction diagram for a twocomponent system which forms two partially miscible liquid phases a t the boiling point deviates widely from the conditions necessary for the application of the McCabe and Thiele methods of graphical analysis. Gay (1) gave a design method for a particular restricted case of distillation of such mixtures. This paper presents the introduction to a general analysis of all such separation processes.

Phase Diagrams for Partially Miscible Liquids The temperature 8s. composition diagram for vaporization of partially miscible liquids is quite familiar. As the upper portion of Figure 1 shows, it is analogous to the melting point diagram for partially miscible solids, which form a eutectic mixture (2). The analysis of the case that the eutectic temperature lies between the two boiling points will be treated in a later paper. 1 This is the aixth paper in this series. The first five appeared in September, October, November, 1938, and in February and July, 1939. respectively.