T H E TINTING STRENGTH O F PIGMENTS BY
T. R . BRIGGS
In a paper on t h e opacity and hiding power of pigments, G. W. Thompson1 has considered briefly the question of the color strength or tinting power. I quote from the original: “In many laboratories, tests for opacity have been conducted on the assumption that wh.at is known as the strength or tinting strength of a pigment is a measure of its opacity. From numerous tests which we have made, we have come to the conclusion that strength is only an indication of opacity and that, working on pigments of the same composition, it is not safe to assume that the strength of a pigment is a measure of its opacity. By stre?zgth or tixti?zg power we mean here: ‘‘ ‘The relative power of coloring a given quantity of paint or pigment selected as standard for comparison,’ which is the definition agreed upon by Committee D1 of the American Society for Testing Materials. Much heated discussion has appeared in the Farbex-Zeitung during the last year or more as to whether strengh is proportional to opacity . . . As far as the discussions have gone, it would appear that they have not led to any definite conclusion.” In an earlier paper, Thompson put forward somewhat different views. “There has been a great deal of discussion, especially in a German periodical (Farben-Zeitung) recently on the relation between strength or tinting power and opacity, some claiming that there is a direct relation and others that there is an inverse relation. Most of these discussions are academic and not based on practical tests. There are reasons for believing that the relation between strength and opacity is a direct relation. The reasons are as follows: Strength increases as opacity does with the fineness of the pigment. I n t h e case of white lead, for instance, its strength and opacity are both increased by grinding it finer . . . . . We hope with Jour. Ind. Eng. Chem., 5 , I Z I (1913). Eighth Internat. Cong. App. Chem., 2 5 , 802 (1912).
T h e Ti?zti?zgStrevgth os Pigments
217
the opacity measuring apparatus referred to and the disc machine to determine the relation between strength and hiding power and opacity . . . .” I have quoted from these two papers by Thompson because they show the interest which the practical paint manufacturer and chemist have in the physical properties of pigments and, beside this, they indicate how little is really known about one of these properties-the coloring strength or tinting power. The guess that there should be a direct relation between tinting strength and opacity is an obvious one; that this relation may be masked by other and disturbing factors is indicated clearly by the contradictory results obtained when the hypothesis was tested. Tinting power is determined’ usually by mixing the particular colored pigment with a white-ordinarily oxide of zinc. Equal weights of the pigments to be tested are mixed in linseed oil with one gram of a special zinc oxide, of uniform composition and kept solely Jor the purpose. The colored pigment should not exceed five percent by weight of the zinc oxide. By comparing the color depth of the different mixtures with that of a standard mixture, some idea of the relative strength of different samples of the same color may be obtained. Or one may follow Ostwald’s suggestion2 and determine how much white one must add to a certain weight of colored pigment before the color of the latter just disappears. It is possible, of course, to measure the tinting power of white pigments in an analogous manner. I n a recent paper, Ostwald3 discusses the problem in some detail. He points out that Prussian blue seems to have one hundred times the coloring power of permanent green (chromium oxide) and that the coloring power is a function of the size of particles. “It is a fact,” he writes, “familiar to every painter who knows his materials, that colors become Cf. Walker: Chem. News, 102, 81 (1910); Holley: “Analysis of Paints and VarDish Products,” 126 (191;). * Zeit. Kolloidchemie, 16, 3 (191j). Ibid., 16, I (1915).
T.R.Briggs
218
ever stronger as they are ground increasingly fine.” Further than this, Ostwald does not commit himself, though he returns to the subject in a second paper1 on lakes and lake pigments precipitated on an inert base. The following passages are worth quoting: “For this reason a new type of precipitated color has been developed, consisting of a colorless base (Trager) on which there has been produced . . . an insoluble precipitate of the color lake. Such pigments, accordingly, exhibit under a powerful microscope an appearance which is quite different from that which the true lakes possess. The material of the base appears completely unaffected by the color and retains th.e same form and color peculiarities that it has in the pure state. Adhering, however, to the base are tiny masses of the insoluble precipitate, in amounts, indeed, th.at appear astonishingly small in comparison to the strength of the coloration. For these quantities, according to the usual authorities, amount as a matter of fact seldom to more than ten percent of the bases and often to very much less.” Ostwald proceeded to study the effect of precipitating a methyl violet-tannin-tartar emetic lake on a lithopone base. So little as one part of lake on 6400 parts of base reduced the brightness of the white from 1000 to 656, while one part on IOO of base reduced it from 1000 to 164. He next studied the influence of the base, using one part methyl violet lake to 400 parts of the base. The data refer t o brightness, pure white being equal to 1000. Each mixture was made into a tempera with 4 percent of glue.
TABLEI
-
Barytes, ground CaC03, coarse ppt. Gypsum, ground Gypsum, burnt C a C 0 3 , fine ppt. White lead Blanc fixe (moist) 1
121 99
143
;I:
11 I
213 216
Zeit. Kolloidchemie, 17, 6 j (1915).
Kaolin Chalk Blanc fixe (dried) Lithopone Zinc white Magnesia
225 226 239
322 330 349
The Tinting Strength o j Pigments
219
Prom the fact that heavy spar requires but l/&h the quantity of lake that magnesium carbonate does to give an equally dark compound pigment (Fiillfarbstoff), it does not at all follow that, with the products obtained in this way, having an equal depth of color, one can cover equal surfaces. While a gram of the pigment compounded with magnesia is sufficient t o cover 500 square centimeters, a gram of the barytes pigment suffices for only 40 square centimeters. These data are only round numbers but they indicate very clearly that the marked differences described above are bound up with a property of the carrier, which ordinarily is designated as couering power.” Ostwald concludes, in brief, that the tinting power of the white pigments varies directly as their hiding power and, inasmuch as hiding power is a function of the size of the particles, the finer the particles, under most circumstances, the greater the tinting power of a given white pigment. He makes the further assumption that as subdivision increases, the hiding power and strength pass through a maximum. It seems almost certain, however, that this maximum can never be reached with pigments, the particles of which are visible under the microscope. Throughout the discussion, Ostwald has overlooked the possibility that the fineness of the methyl violet lake itself may not be constant throughout his experiments. The importance of this point will become increasingly apparent as we proceed. There have been reported from time to time various data on the colors of pigment mixtures. In nearly every case the author appears to be somewhat mystified by his results. According to Gardnerl an equal weight of ground barytes may be mixed with red oxide of iron without diminishing appreciably the intensity of the red. The case of chrome green (lead chromate-Prussian blue) is even more striking, for a mixture containing 80 percent of barytes shows its green color practically unimpaired. No explanation for these phenomena is offered. Proc. Am. Inst. Mining Eng., 50, 983 (191j).
T.R. Briggs
220
Sabin1 mentions another case. “Red lead has a very brilliant color, -but little staining power ; one part of dry lampblack in 500 of red lead changes it to a coffee color, and I percent to a dark chocolate.” A statement by Holley2is worth repeating: ‘ T w o chrome yellows may appear to possess identically the same tint and tone, yet when reduced with equal quantities of a white pigment, e. g., white lead, one may be found to possess 2 5 percent more strength than the other and will still maintain a clear yellow tone while the other will appear dirty or may pass over into another tone or color altogether. Chemically these yellows may be the same and the difference is therefore due t o the method of manipulation.’’ When two differently colored powders are mixed, the shade or tint of the mixture is not dependent solely upon the percentage composition. This point is beautifully illustrated by soils colored red and yellow by iron oxide.3 TABLE I1 “Acid Digestion of Soils of Alabama and Arkansas” Color
Dark red Dark red Deep red Deep red
Percent Fez03 removed
10.7 7 . 2
4.9 2.5
1 ,
Color
Percent Fen08 removed
Red Red Red Red
9.3 23.6
5.0 2.6
It is apparent that the iron oxide content of these soils bears no direct relation t o the depth of color. This fact led Robinson and McCaughey to make the following very interesting remarks : “It is unfortunate that the mechanical analyses of these soils are not available, so that the depth of film of ferruginous material on the soil grains from the several samples might be compared with the color. It is evident that a film of coloring 2
“Technology of Paint and Varnish,” 297 (1917). “The Lead and Zinc Pigments,” 2 I 7 (1909). Robinson and McCaughey: Bur. Soils Bull., 79, 2 0 (1911).
The Tintixg Strength o j Pigments
221
matter of a certain depth over large1soil particles would contribute much less to the total percentage of the coloring matter than if the film of the same thickness was spread over the same weight of exceedingly fine particles ; and since the film is postulated to be of the same thickness in each case, the color must be alike in the two instances. The following laboratory experiment will serve to illustrate this point: Samples of 2 0 grams each of coarse, white quartz sand, averaging 2 mm in diameter, and white quartz flour were treated with ferrous sulphate solutions of such a concentration that the dried material should contain from 0.1percent Fez03to 2.5 percent of Fe203,with five intermediate percentages. The material was evaporated to a paste and enough ammonia added to precipitate the iron present. Hydrogen peroxide was then added to oxidize the iron and the material was heated to a dull redness to drive off the ammonium sulphate and dehydrate the iron oxide. I n this manner the iron was deposited upon the surface of the quartz grains. The quartz flour containing 2.5 percent Fez03was very nearly matched in color by the coarse quartz sand containing 0.25 percent FezOa.” We have now reached a point where it is necessary to consider the general case of a mixture of two powdered substances. The simplest case is where one mixes together a white and a colored pigment. If the particles of one pigment are small compared with the particles of the other, there will be a tendency for the smaller particles (if dry) to coat or “dust over” the larger ones. This tendency will be more marked as the difference in size increases. Anything which increases the adhesion will also favor the coating of the large particles. It is obvious, therefore, that the color of the mixture is determined very largely by this coating effect; the more completely the large particles of one pigment are covered by the smaller particles of the second and the greater the hiding power or opacity of the second pigment, the more will the mixed pigments approximate the color of the small particles. If the colored pigment is coarse and the white is fine, a mixture of [The italics are mine T.R. B.]
T . R.,Briggs
222
equal proportions of the two may be very nearly white; if the white is coarse and the colored pigment is fine, the mixture may be nearly indistinguishable from the pure colored pigment. The tinting strength of a pigment, besides depending upon the absolute size of the particles, is determined t o a very great extent by the relative size of the particles compared with those of the second pigment. This,’ though perhaps obvious enough, has been overlooked almost entirely in previous work. Since color is only one of many physical properties, we should expect to find other properties of mixed powders influenced largely by the relation between the size of particles. The electrical conductivity is such a property. This case has been discussed by Fink,l using mixtures of thoria and tungsten powders. He remarks that “the electrical conductivity of mixtures of finely divided substances is a function of the relative size of the components.,” His results also show how the appearance of the mixtures is affected in the same way, as the following table, prepared from Fink’s data, indicates : T A B ~I11 E
I
Fineness
Mixture
T h o z (white)
Tungsten (black)
Appearance
0.720 0.720
0.113 0.350 0.235 0.113 0,330 0.577
White White White White Nearly black Kearly black
0.576 0,305 0.305 0.238
Specific resistance
-
173 ohms -
0.108 ohm
Equal parts by weight of thoria and tungsten were present in each mixture. The fineness was determined by a so-called “tap test,” not a very satisfactory way. The conductivity data were obtained with rods fired at 1600’ C for three hours. Since the resistance of the pure thoria was greater than 1 0 ohms as compared with 0.016 for the tungsten, it may seem a 1
Jour. Phys. Chem.,
21,
,
38 (1917).
~ ~
The Tinting Strength o j Pigments
223
bit surprising that a value so low as 173 ohms should have been obtained for the first mixture, as the color would indicate that the coarser tungsten particles (black) were very effectively coated over by the finer thoria (white). Sintering occurred during the firing of the rods, however, and this must have changed greatly the orientation and relative size of the particles, for the shade of all the mixtures became much more nearly uniform after firing. As color is determined largely by the small particles coating the larger ones, so is the conductivity of the mixture very nearly that of the smaller particles, in case the coating is sufficiently complete. We have as an extreme case, metal gels1 and deposits of finely divided metals containing an adsorbed substance of low electrical conductivity. Benedick’s2“acetate copper” formed by electrolysis of cupric acetate owes its strikingly low conductivity to the fact that the copper crystals are coated with copper oxide. Faraday’s mirrors3 of colloidal gold were practically non-conducting as were similar mirrors of Lea’s ~ i l v e r . ~Fink; also points out that the low conductivity of copper containing sulphur, etc., may be due to sulphide films coating the copper crystals. When the finely divided component constitutes only a small part of the mixture the question of adhesion becomes important. We do not know why powders adhere t o surfaces6 when powder and surface both are presumably dry, nor do we know to what extent adhesion occurs. We have carried out in the Cornel1 laboratory some preliminary experiments along this line and hope t o be able to report progress before very long. It seems, however, that adhesion to the large particles increases as the small ones become still smaller and microscopic Cf. Briggs: Jour. Phys. Chem., 17,281 (1913). Metallurgie, 4,j , 33 (1907). Phil. Mag., (4)14,402, j 1 2 (1857). Cf. Barus and Schneider: Zeit. phys. Chem., 8, 278 (1891); Liidtke: Wied. Ann., 50, 678 (1893). LOC.cit. Cf. Tamrnann: Drude’s .4nn., 18, 857 (1905); Biltz: Ibid., 31, 1050 (1 910). 1
224
,
T . R. Briggs
examination indicates that the topography of the large surfaces is of great influence; the rougher the surface the more the tiny particles are caught and held. If the large crystals are covered with a thin film of moisture or oil, the adhesion will be greatly increased, of course, in case both powders are wetted by the liquid. On the other hand if the moisture or oil is present in sufficient amount, the apparent tinting strength of the fine pigment may become less, owing to the falling off in hiding power when one changes the medium surrounding the fine particles from air to liquid. Even when the oil is present in large excess, the same general relations will hold between size and orientation, for the small particles, especially if they form readily a suspension in the oil, will tend to surround and hide the larger ones. The high tinting power of the lakes precipitated on coarse bases such as barytes is due to the small size of the color particles in comparison with the base, and to the completeness with which the lake adheres t o the base by adsorption, or mechanical adhesion. It is possible to summarize the theory as follows : Tinting strength depends upon at least three f actors-the relative size of the particles, the extent t o which the smaller particles adhere to the larger ones and the hiding power at constant size of the small particles. Strength can be used as a measure of hiding power in case the pigments under examination actually coat over or hide something, and if the size relation and adhesion (coating) factor are kept constant throughout. Since little or no attention has ever been paid t o these points, there is little wonder that nobody has been able to decide just what relation the data on strength and opacity (hiding power) indicate. If strength tests are made to gain some idea of the hiding power of a pigment, it would seem that oxide of zinc ismost unfortunately chosen as the standard white color. Few, if any, pigments consist of particles of smaller average size, so that the tendency will be to coat the other pigment with the zinc or else neither pigment will coat the other to any appreciable extent. As a result, the test may tell something about the
The Tintivtg Strength o j Pignzenls
.
225
coloring and hiding strength of zinc oxide, but does not necessarily indicate anything regarding the colored pigment. The white color should be relatively coarse ; ground barytes ought t o be preferable to zinc oxide. On the other hand, we have seen that Ostwald interpreted his data with methyl violet lake on different white bases as measuring the hiding power of the white base, although the latter was coated with the violet lake. It seems to me, however, that the data indicate merely that the tinting power of a given percentage of lake grows smaller the greater the total surface to be covered and the smaller the particles of the base. In so far as the fineness of the base is a measure of its hiding power, Ostwald’s conclusion is a just one; but fineness of subdivision is only one of the factors determining hiding power. As a matter of fact, Ostwald was measuring the extent to which a given quantity of a colored lake was able to hide surfaces whose total area differed within wide limits, and he was assuming that the data measured the hiding power of substances which were actually hidden more or less completely. Gardner’s statement about red oxide of iron and chrome green containing barytes is now perfectly simple. Red oxide of iron and the green are much finer than barytes, hence the latter are coated over and hidden more or less completely. The greater effect with chrome green indicates that the particular chrome green was finer th.an the red oxide, though adhesion or biding power may have been greater as well. Sabin’s data are easily explained by the theory, for red lead is reported to be rsther coarsely crystalline (except in the form of orange mineral) while lampblack is extremely fine.l Holley’s remarks concerning the tints of chrome yellow indicate merely that different samples have particles of quite different average sizewhich certainly is neither unexpected nor mysterious. Experimental During the summer of 1916, Mr. Tears carried out, as part of his senior research, some experiments of a preliminary Cf. Kuhn Zeit angew. Chem,
28,
126 (1915).
T . R. Rriggs
226
nature to test the hypothesis outlined above. Lacking at that time a white pigment of a suitable nature he used dolomite that had been analyzed mechanically by means of a set of standardized sieves. With dolomite as the white, the tinting power of various samples of “soluble” Prussian blue, lampblack, chromium oxide, and iron oxide was determined. Ten grams of dolomite were taken for each test and the particular color added from a weighing bottle in minute quantities at a time until the mixture matched exactly a standard shade. Great care was taken to have the powdered substances dry and to see that the mixture was uniform and complete. Much trouble had been given in early experiments by the presence of lumps of the very fine coloring material; it was found necessary to rub up the mixture with a spatula or t o grind it very gently until no further deepening of the shade occurred. The experiments were tedious and it was not easy to obtain good checks. The data follow: TABLE IV Size of dolomite variable. Dolomite
(IO
Weight CrzOs
grams)
passed through
Size of colored pigments constant
~
Weight Prussian blue
l-
Weight lampblack
-
I
200
100-200 do- -I 0- -0 7-
retained by 40
0.02j
1
0.032
0.20
0.08
0.019 0.008
Although these data are striking enough in all three cases that were studied, they are particularly so in the case of lampblack. On column of .data it will be~ seen ._.. _ _ _ dancing_..__ ~ ~at -the ~ _- third ~~-~~ _ ~~ . ~ ~ _ ~ _ ~ that the finest dolomite required eighty times as much lampblack as the coarsest dolomite did to produce the same shade of grey. Someone may object that it is very hard to judge accurately the appearance of mixtures composed of very coarse and very fine dolomite, but even if one considers only the first two horizontal rows of data, the numbers are sufficient t o compel attention. Thus the finer dolomite required ten times as ~~~
~~
~
~~
~~~~
~~
~~
~~~
~
~
~
~~~
The Tinting Strew,& o j Pigments
227
much ch.romium oxide, eight times as much Prussian blue and twenty-six times as much lampblack as did the coarser dolomite (100-200 mesh). These numbers should be regarded as rough approximations only, but they do show that the results are in accord with the general theory. A few months ago we were fortunate enought to receive from the National Lead Company, through the courtesy and cooperation of Messrs. G. W. Thompson and R. I,.Hallett, samples of white lead and red lead, graded carefully according to size by means of the Thompson c1assifier.l Tests were made with these pigments by Mr. A. W. Ward, as part of his senior research, but the work was interrupted rather prematurely by Mr. Ward leaving to enter upon service in connection with the war. A microscopic examination of the pigments was made for the purpose of measuring approximately the average size of the particles in the different samples. The necessary equipment was placed at Mr. Ward’s disposal through the courtesy of Professor Chamot. The micrometer eyepiece was calibrated in the usual manner and the size of the particles was ascertained by averaging IOO readings. The data.follow:
TABLE V Microscopical examination of graded pigments. Size determined by averaging IOO readings. Ocular micrometer method Average size in microns
Thompson classifier Cone number
White lead
1 I -
I
32 25
Red lead
-
1 34 23
‘7 7
I2
4
5
9
Preliminary experiments were carried out with dolomite and red lead. Tear’s procedure was followed, except that one gram of dolomite was used in each run. Greater care was 1
Jour. Ind. Eng. Chem.,
2, 90 (1910).
T . R. Briggs
228
taken in adding the red lead and in making the mixture uniform. The data obtained confirm the previous results.
TABLE VI Experiments with dolomite and red lead. Dolomite I gram. Cone No. 5 red lead added until each mixture had about the same shade Dolomite
Read lead added (in grams)
~
Passed 350 mesh sieve 200-350 100-200
I
0.126 0.079 0.033 0.016
80'100
Next, mixtures of white and red leads were prepared. Owing to the limited supply of classified pigments, only one gram of the white was taken for each determination. The red was added from a weighing bottle slowly and with extremely careful intermixing, until the appearance and color of all the mixtures were as nearly as possible the same. The color matching was done with the eye and the pigments were mixed dry. The loss in weight of the weighing bottle containing the red lead was taken to be the amount of red added. Each experiment was run in duplicate.
TABLE VI -
White Red lead
Experiment
Original
number -I
I
2
2
3 4
3 4
5
5 5
6
Red lead added (gram)
*
Percent red lead in mixture Duplicate (Average) (Average)
5 5 5 5 5
0.0345 0.0345 0.0345 0.0478 0.0450 0.0464 0.0553 0.0510 o.oj32 0.0636 0.0658 0.0647 0.124
0.110
0.117
I
0.887
0.905
0.896
3.33 4.43 5.05 6.08 10.47 47.31
Experimmt ?--The mixture prepared in Experiment 6 above was ground in an agate mortar. The shade darkened
.The Tinting Strength o j Pigments
229
to a very great extent, showing the effect of pulverizing the coarse red lead. Experinzeqzt 8.-The mixture resulting from Experiment I was likewise ground. The effect was striking, the shade becoming much lighter and the light red of the original turning to a very pale brown. The data confirm the theory in a very satisfactory way. The finer the red and the coarser the white, the less is the amount of red required to produce a given shade. Experiments one and six represent the most extreme cases. I n one there is present more than fourteen times as much red lead, and yet the two mixtures have almost the same color and appearance. The results of this paper may be summarized as follows: ( I ) The ability of one pigment to impart its color to a mixture containing a second is greatly increased if the first pigment coats over and hides the particles of the second one. ( 2 ) The smaller the particles of the first pigment relatively to those of the second, the less is required to coat the particles and the greater becomes the tinting strength of the first pigment. (3) Anything which favors the adhesion of the fine particles to the coarse ones and thus promotes the coating-over effect, increases the tinting strength of the first pigment. (4) #or particles of the same size, and equal coating tendency (adhesion, etc.) the greater the hiding power the higher is the tinting strength. (5) With proper control of the other factors, i t ought t o be possible to estimate hiding from tinting strength, in case the pigment to be tested is very fine compared with the one with which it is mixed and if it actually coats over and hides the latter. (6) Other physical properties of mixed powders besides color are dependent very largely upon relative size of particles and the coating-over effect . (7) Relative size of particles has an important bearing on the question of a mixture of two pigments, one of which is “inert.” A larger amount of inert pigment may be added
230
T . R.Briggs
without appreciably affecting the shade of the mixture, the coarser the inert pigment is in comparison with the other one. This observation helps to account for the fact that ground barytes, silex and asbestine are used so generally as inert fillers, for all three, in the form in which the paint manufacturer uses them, consist of particles which are far larger than is the case with the majority of the “true” pigments. The question of how far one should carry the addition of an inert pigment is of course another story. Cornell University
