February, 1928 INDUSTRIAL A N D ENGINEERING CHEMISTRY 199

Thinners for Nitrocellulose Lacquers J. G. Davidson and E. W. Reid

C A R B I D E A N D C A R B O N C A & Y I C A L S C O R P O R A T I O N , NEW Y O R K , N. Y.

New thinners for the ordinary type of nitrocellulose base lacquers are discussed. These thinners differ from the type formerly used in that they contain small percent- ages of solvents boiling from 135' to 160" C. which re- place larger quantities of solvents boiling from 110' to 140" C. These thinners when used with the usual type base lacquer give results comparable in every respect with old type thinners.

HE rise of the automobile-lacquer industry has elimi- nated many of the problems inherent in the application T of the old-time varnish enamels. It has, however, intro-

duced certain new arrangements in the handling of materials that require special methods. For example, nitrocellulose enamels cannot readily be shipped in the form in which they are used, because the pigments and gums tend to settle out in the bottom of the container and it is rather difficult to get them back into solution again with the assur- ance that no undissolved solid lumps or particles that would mar the finished surface will be present. For this reason automobile enamel is generally shipped as a heavy base lacquer, so thick and viscous that its constituents will not ordinarily separate through the influence of gravity. At the plant this base lacquer is diluted with a thinner, usually in the proportions of approximately 1 : 1 but some- times as high as 2:l or even 3:l .

Requirements of Lacquer Thinners

For the ordinary type of base lacquer there are certain general requirements that must be fulfilled by any thinner. It must be a solvent for nitrocellulose; otherwise the sudden addition of a large amount of thinner, or drowning as it is familiarly called, will result in precipitation of flocculent nitrocellulose that cannot readily be stirred back into solu- tion. The thinner must not contain any constituents that will be detrimental to the final lacquer, such as solvents that are likely to hydrolyze. It must not present diffi- culties from the physiological standpoint either by actual toxicity of the vapor or through an unpleasant odor due to impure materials. Lack of odor or pleasant odor are, of course, distinct advantages. It must cut the base lacquer readily and with little stirring and it must produce a mixed lacquer that will work properly in the spray gun. Finally, the solvents must be so proportioned that blushing will not occur under the conditions in which the lacquer is used.

Prevention of Blushing

In general the usual type of thinner will meet all the requirements except the last, and it is here that most diffi- culties arise in modern automobile plants. If the thinner does not contain the proper solvents in the proper proportion, moisture is liable to be condensed upon the work under humid conditions or, conversely, some of the moisture present in the original lacquer is not carried off. This brings about the familiar phenomenon of blushing. Inasmuch as the con- ditions obtaining in humid climates are normally very differ- ent than those in dry atmospheres, if the quantity of thinner used is very great, it will pay to formulate a special thinner for each locality. In general, however, this has not proved worth while and most firms content themselves with a sum- mer and winter thinner, the summer thinner having a high blush resistance to cope with the high humidity of summer

and the other having lower blush resistance for use in colder weather when the humidity is low.

The properties of blush-resisting solvents (retarders) are as follows:

1-The boiling point is above the boiling point of water. 2-They are only partially soluble in water or have the ability

to form constant-boiling mixtures with water. 3-They are good solvents for nitrocellulose.

Other things, such as solvent power for nitrocellulose, being equal, the amount of blush-resisting solvent that must be used is inversely proportional to the boiling point. Thus, much larger quantities of a solvent boiling a t 110' C. must be used than one boiling a t 190' C. The temperatures cited are, however, quite extreme and would not generally be advo- cated. A solvent boiling a t 110" C. would evaporate much too quickly unless very large quantities were used, while a solvent boiling a t 190" C. would probably evaporate too slowly and leave a soft coating that could not be sanded and polished in the time allowed. Usually forced drying could be used in the latter case, but this is objectionable if the same results can be secured without forcing.

Composition of Thinners

The general type of thinners hitherto available has had a composition varying approximately as follows: 5 to 25 per cent ethyl acetate, 20 to 40 per cent butyl acetate, 10 to 20 per cent butyl alcohol, and 20 to 60 per cent diluent.

The diluent is usually toluene, but it may be replaced in whole or part by xylene or gasoline. These thinners have been very satisfactory and have only given trouble when attempts have been made to lower the percentage of butyl acetate and butyl alcohol below the limits experience has shown to be necessary.

The development of new nitrocellulose solvents has recently made available a number of new compounds having boil- ing ranges higher than butyl acetate but still not so high that their evaporation would unduly retard the setting of the lacquer film.

The more prominent of these materials are amyl acetate, which in consequence of its synthetic production has recently become available a t prices much lower than formerly, butyl propionate, Cellosolve (ethylene glycol monoethyl ether) and Cellosolve acetate (ethylene glycol monoethyl ether acetate). In the general thinner formula given above the butyl acetate and butyl alcohol can be replaced in whole or in part by smaller quantities of any of these compounds. The relative efficiency of these retarders as blush resistors is approximately as follows: 8 parts of amyl acetate are ap- proximately equal to 6 parts of butyl propionate and 5 parts of Cellosolve acetate. Cellosolve itself presents a special case that will be discussed later.

A general formula that will have a blush resistance of 90 per cent humidity a t 90" F. when blended with the usual type of automobile base lacquers is as follows: 10 to 20 per cent ethyl acetate, 5 to 10 per cent Ansol, 10 per cent Cel- losolve acetate, and 60 to 80 per cent toluene. The Cello- solve acetate, however, seems to confer a special benefit in the form of better gloss and less orange peel.

Cellosolve itself acts as a retarder up to conditions of approximately 65 per cent humidity a t 90" F. As it will no doubt eventually be somewhat cheaper than its deriva-

200 I-VDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 20, No. 2

tive, Cellosolve acetate, it can be used in any amount in the winter time and in the summer where the humidity does not exceed the conditions outlined above Except when un- usually severe humidity pi evails half the Cellosolve acetate may be replaced by Cellosolve for summer use. In the winter from three-quarters to all the Cellosolve acetate can be replaced to advantage by the Cellosolve, as it is a very powerful solvent and has a higher dilution ratio than any of the other retarders. It will be understood, of course, that this statement does not hold for brushing lacquers, where percentages of Cellosolve up to 3040 per cent are used successfully.

How far the subject of new solvents should be investigated is somewhat problematical. Just how much the water solubility of the solvent changes the blush resistance is also an undetermined factor. For example, experiments have shown that the diethyl ether of ethylene glycol which is only partially soluble in water while Cellosolve is entirely so, is a better blush resistor than Cellosolve when used in amounts up to 5 per cent, but quantities in excess of this do not improve the results obtained with this amount. On the other hand, the butylethyl ether of ethylene glycol, boiling at 170" C., which is somewhat more soluble in water than Cello- solve acetate, has approximately the same blush resistance, and may be used in quantities up to 20 or 25 per cent. How- ever, it does not seem to give a surface that has the high gloss produced by Cellosolve acetate.

Solvent Power

The solvent power of thinners used in connection with automobile base lacquers must be rather carefully watched. If it is below a certain critical value complete solution of the nitrocellulose is not readily made. Furthermore, on stand- ing over a period as short as 10 to 12 hours the solution may gel or stiffen. The solvent power of the thinner can be calculated by either one of two methods. In the first1 method the various solvents in common use have been given index values that have been worked out from practical use. The values for some of the more commonly used mate- rials are M follows when toluene is used as a diluent:

Ethyl acetate 1 '/a Ethyl alcohol (when used in quantities not greater than 10%) 1 Butyl alcohol (when used in quantities not greater than 15%) 1 Butyl acetate 1 Cellosolve Cellosolve acetate

2 1

When the summation of the percentage of active solvent multiplied by its index value equals 35 or more, the thinner will prove acceptable for use with modern base lacquers.

The second method involves merely a setting forth of the dilution ratios of the various solvents with respect to the dilu- ents used. This system requires the fixing of an arbitrary value for the alcohols, which are not by themselves solvents for nitrocellulose but which increase the solvent power of many of the commonly used solvents. Under this system when toluene is used as the diluent the dilution ratio or the solvent power number of each solvent with respect to tolu- ene is used as the index value. If gasoline is used as the diluent the index values become the dilution ratios with respect to gasoline. The index figures for the materials shown above are set forth below, the diluent being toluene:

Ethyl acetate 3 . 7 Cellosolve 5 . 3 Ethyl alcohol 2 .0 Cellosolve acetate 2 . 7 Butyl alcohol 2 . 0 Butyl propionate 2 . 8 Butyl acetate 2 6

With this system the total of the index numbers mul- tiplied by the percentage used should equal 90 in order to

1 Private communication from H. C. Mougey, General Motors Research Laboratories.

give the same results that the total of 35 gives in the other system. Two examples will illustrate this empirical rule.

Example I INDEX

COhfPOUND COMPOSITION NUMBER Per cent

15 = z 2 . 5 -55.5

=10.0

Ethyl acetate

Ethyl alcohol i i . 0 = 5 . 0

Toluene Cellosolve

Cellosolve acetate = 2 7 . 5

= 1 3 . 5 = 5 . 0

- 100 Solvent power.. ... 4 2 . 5 106.5

Since the value of this mixture is greater than 35 in one case and 90 in the other, its solvent power will be satisfactory.

Example I1 INDEX

COMPOUND COMPOSITION NUMBER Per cent 1;:; = 7 . 5

=18 5 Ethyl acetate

- 10.

= 15.

Toluene 70 Butyl alcohol

Butyl acetate 10 x 1;:; -20 .0

15 X {i:: ~ 3 9 . 0 - 100 Solvent power.. . . . 3 2 , 5 77 5

As the value of this thinner is less than 35 by Mougey's method and less than 90 by the other, it will not have sufficient solvent power when used with the usual type of automobile base lacquer in the proportion of 1 :1.

Tests for Suitability i n Lacquers

The only test that has given useful information as to the way in which a lacquer thinner will work is a practical test in which large panels are coated under conditions of temperature and humidity that can be controlled. This test is later followed up by actual plant tests in which several automobile bodies are sprayed. It has been found that boiling point curves, specific gravity, viscosity, rate of evaporation, and amount of cooling produced by evapo- ration are all without inferential value unless applied to formulas that do not involve changes in the active solvent used. When the use of new solvents is proposed, aging tests of panels sprayed with lacquer containing the new solvent are, of course, necessary to determine whether any reaction is set up that will destroy the film sooner than usual. These tests can be quickly secured under an ultra-violet light or from a practical standpoint panels exposed for a period of 6 months will give a very definite indication of what may be expected with reference to the durability of the panel.

So far as Cellosolve and the pure ethers are concerned, there is no question of deterioration. Cellosolve acetate has been investigated very thoroughly with respect to hydrolysis. The results show that under the conditions of general lacquer use it is a t least as stable as the esters commonly used. Lacquers containing Cellosolve acetate have stood for many months without developing acidity while panels sprayed with lacquers containing Cellosolve acetate have shown no tendency to break down after having been exposed to the atmosphere over a period which now approximates 9 months.

This question of hydrolysis has been studied very fully and the results of this work will be presented later.

Acknowledgment

The authors wish to express their appreciation of assist- ance extended them during the collection of these data by H. C. Mougey, of the General Motors Research Labo- ratories, the laboratories of the du Pont Company, and several motor-car manufacturers who cooperated in the full-scale tests made in a production way.

[END OF SYMPOSIUM]