INDUSTRIAL AND ENGINEERING CHEMISTRY
1186
the basis for differentiation between stabilities, rather than the actual rise, as in the present work. The actual maximum rise has been used because it varies in a definite manner between nitrocottons of known stabilities. All the data secured show that the point of incipient decomposition is too indefinite and does not vary between wide enough limits to judge stability sharply. However, with apparatus free from heat losses, as suggested below, this point may become marked enough t o furnish an additional check on the stability. The test is still under investigation and, as mentioned be-
Vol. 16, No. 11
fore, is susceptible to far more precise methods. A few experiments have been carried on and others are planned, using Dewar flasks and resistance thermometers. The Dewar flask prevents heat leaks to the bath and causes more sharply defined temperature changes. These well-marked temperature inflections offer very interesting possibilities in the decomposition study of the cellulose esters. The vacuum tube also allows examination of finished powders and other explosive materials, which because of their structure cannot be used in the test-tube apparatus.
Effect of Hydrogen-Ion Concentration on Compound Formation and Adsorption of Dyes by Mordants‘ By R. E. Marker and N. E. Gordon UNIVERSITY OB MARYLAND, COLLEGE PARK, MD.
scientific methods and pure materials has led to contradictory results. For instance, some results point to the fact that the adsorption of dyes by mordants is more or less physical in nature, while other investigators have been willing to believe that such adsorption is of a chemical nature. All these investigations have a commercial value, for it is the technical preparation with which the industries have to deal. On the other hand, if we are to know the truth, careful quantitative work is necessary where the purity of the material used and of the intermediate products is taken into account.
MATERIAL USED The mordants (inorganic gels) used were silica, alumina, and ferric oxides, chosen because they were believed to be chemically representative, one having an acid character, one a basic, and one an amphoteric nature. The commercial silica gel was ground to pass through a 100-mesh sieve and then washed until free of all electrolytes. The ferric oxide and alumina gels were prepared and purified by the usual methods. The two basic dyes used were crystal violet and methylene blue, and orange I1 and metanil yellow were chosen as representative of the acid dyes.3 The dyes were purified until on analysis they gave the following results: Crystal violet Methylene blue Orange I1 Metanil yellow
Per cent 99.98 99.97 99.97 99.93
EXPERIMENTAL One-half per cent dye sdlutions were used. The different hydrogen-ion concentrations were obtained by adding sul1 Presented by White, Marker, and Gordon under the title “Effect of Hydrogen-Ion Concentration on the Adsorption of Dyes by Gels” before the Division of Dye Chemistry a t the 67th Meeting of the American Chemical Society, Washington, D . C., April 21 t o 26, 1924. * THIS JOURNAL, 16, 518 (1923). a Furnished through the courtesy of E. I. du Pont de Nemours & Company.
the exact weight of dry gel used was calculated. I n introducing the dye solutions into the bottle, both the quantity of gel and its water content were taken into account, so that the ratio of dye to gel in each bottle was exactly the same. Preliminary experiments showed that a t loooC. the dye in the solution reached an equilibrium with the gel by constant shaking for one hour. After the equilibrium was established and the gel allowed to settle, a n aliquot of the dye solution was titrated with titanium trichloride under a n atmosphere of carbon dioxide, as recommended by Knecht.4 The amount of dye adsorbed was calculated by the difference in titration before and after adsorption. The hydrogen-ion concentration of the various dye solutions was taken by means of a Bailey electrode. Many types of electrodes were tried, but the Bailey was found the most satisfactory. With any electrode the readings had to be taken rapidly because of the adsorption of the dyes by the electrodes. After each operation the electrodes were washed with chromic acid to destroy any adsorbed dye. I n each case several readings were taken and the mean was used in the final result. KOreading was far from this mean.
FERRIC OXIDE GEL The experiments with ferric oxide gel and the acid and basic dyes were carried out as described above. The results are given in Table I. TABLEI-ADSORPTIONOF ACIDAND BASICDYES BY FERRIC OXIDEGEL ,---BASICDYES-------ACID DYES-MEIHYLENE BLUE CRYSTAL VIOLET ORANGE I1 MSTANILYELLOW Mg. adMg. adMg. adMg. adsorbed sorbed sorbed sorbed per gram per gram per gram per gram PH gel PH gel pH gel pH gel 1.96 27.6 2.06 23.2 2.3 429 1.92 361 2.3 340 2.23 29.0 2.94 33.0 3.2 75 5.95 30.0 5.02 42.3 5.27 70 3.38 255 0.85 32.1 9.01 50.6 10.14 52 7.46 211 11.12 33.8 10.95 56.1 11.02 50 11.60 80.7 12.00 131.0
It will be noted from Table I that the higher the pH the more of a basic dye is adsorbed, whereas the reverse is true 4
B e y . , 36, 1552 (1903); 40, 3819 (1907).
acid and basic dyes always occurs a t approximately the same pH. One reason for this will be brought out in the following discussion. INDICATIONS O F COMPOUND
FORMATION
I n the foregoing experiments certain phenomena were
pH 1 96 2 23 5 95 9 85 11 12 1 2 00
per gram dry gel 65 6 66 1 67 5 77 0 82 7 279 0
pH 1 50 R 44 9 18
io io
11 12
per gram dry gel 3 8 45 282 413
pH 2 30 3 20 5 27
in
14
11 0 2
per gram dry gel 452 186 179 162 136
pH 1 92 2 30 7 46 9 67 11 60
_--
per gram dry gel 703 460 276 226 115
The adsorption of both the acid and basic dyes shows a behavior with the alumina very similar to that with the ferric oxide gel. The marked change in adsorption takes place a t about the same pH, but the ferric oxide gels have the higher power of adsorption. This is more clearly shown by Fig. 2. The sharp breaks in these curves will be discussed later. SILICAGEL The inactivity of silica gel did not warrant the anticipation of any great chemical adsorption. The same acid and basic dyes were used as with the other gels, and results obtained as given in Table 111. TABLE 111-ADSORPTIONon ACIDA N D BASICDYESB Y SILICAGEL ,-----BASIC DYES----,-------ACID DYES---METHYLGNE BLUE CYRSTAL VIOLET ORANGE I1 METANIL YELLOW Mg. adMg adMg. adMg. adsorbed sorbed sorbed sorbed per gram per gram per gram per gram PH gel pH gel pH gel PH gel 1.96 9.4 0.068 0 . 3 0.00 1.92 73.3 2.3 2.23 10.7 1.31 1.4 3.2 0.00 2.30 67.7 5.95 10.9 2.8 0.00 2.94 5.37 3.38 2 2 . 9 9.8.5 11.7 5.02 4.3 10.14 0.00 7 46 18.5 '11.12 12.6 9.01 4.9 11.02 0.00 9.97 14.6 12.00 23.9 10.95 5.7 12.00 0.00 11.60 3.9
Although the adsorption with the silica gel is not so marked as that of iron and alumina gel, the adsorption follows the
FIG. 2-ALUMINA GEL aa-Methylene blue cc-Orange I 1 bb-Crystal violet dd-Metanil yellow
beautiful crystals was obtained in each case. The alumina gel gave needle-shaped crystals, while the ferric oxide gel gave crystals having a twisted, threadlike appearance. When subjected to analysis the purified crystals gave the following percentage composition: Calculated Found
Fe(SOa.CaH4N:N.CloHn.OH)a Fe Dye radical 5.4 94.6 5.2 94.4
AI(S03.CsHa.N:N.CioHa.OH)~, Calcu!ated Found
AI 2.7 2.6
Dye radical 97.3 97.7
Total 100 99.8 Total 100 100.3
This work showed conclusively that the foregoing assumed equilibrium took place and that the adsorption was chemical
INDUSTRIAL A N D ENGINEERING CHEMISTRY
1188
in nature. The hydrogen-ion concentration was the principal factor which determined the completeness of the reaction. As an auxiliary test for compound formations, solutions of the same pH but varying concentrations of the acid dye were
Mg8. dye
40
20
ndsorbed par grm of dry e e l . 60
FIG. %--SILICA GEL aa-Methylene blue bb-Crystal violet cc-Metanil yellow
Vol. 16, No. 11
- ++H
Acid X M z - L X H a X
The increase of acid moves the equilibrium to the right, and since the same acid is produced in all cases all metallic salts should produce the same color q t the same pH provided the hydrogen-ion concentration is great enough for the color of the acid to predominate over the color of the un-ionized saltwhich proved to be the case. The fastness on the wool decreased with the increase in pH. This might reasonably be anticipated from the foregoing equilibrium. When the pH is increased the equilibrium moves to the left. I n other words, we have the dye in the form which acts less readily with the fiber. Work is now being done to find out if orange I1 does not form a definite chemical compound with the fiber just as it has been shown to form a definite compound with the mordant. If this proves to be the case, it should not be too much to assume that such a compound as alumina has mordant properties, because one or two of its valencesare satisfied by the dye while its remaining valence is satisfied by the fiber. This might be graphically represented thus:
//" *l\
/"
Or
*I\\
'Y
prepared and shaken with both the ferric oxide and alumina gels. After equilibrium was reached the dyes were titrated with the results shown in Table IV.
kY
where X i s the dye radicals and Y represents the fiber. There are many reasons for believing that the fiber has amphoteric properties.
TABLE IV-ADSORPTIONO F DYES BY ALUMINA AND FERRIC OXfDES GEL AT THE SAMEpH BUT VARYING DYE CONCENTRATIONS
Original dye in solution Per cent
0.5 0.75 1.16
Dye in 100 c c . solution over the alumina gel after adsorption Mg.
40 40 40
Dye in 100 cc. solution over ferric oxide gel after adsorption Mg.
14.7 14.7 15.4
This acted as a confirmatory test to show that there must have been a reaction between the dye and the gel. The equal amounts of dye left in the solutions resulted from the solubility of the respective iron and aluminium dyes. The aluminium salt is nearly three times as soluble as the iron salt. F I G . &-BASIC
DYES
aa-Alumina gel and crystal violet bb-Alumina gel and methylene blue cc-Ferric oxide gel and methylene blue dd-Ferric oxide gel and crystal violet ee-Silica gel and methylene blue f f 4 i l i c a gel and crystal violet
The alumina and ferric oxide products of metanil yellow and crystal violet have been sufficiently investigated to show that the metanil yellow adsorption is due to the precipitation of the dye in the acid medium, whereas the adsorption of'the crystal violet promises to be chemical. DYES cc-Alumina gel and metanil yellow a a S i l i c a gel and metanil yellow bb-Ferric oxide gel and metanil yellow dd-Ferric oxide gel and orange I1 ee-Alumina gel and orange I1 F I G . &-ACID
CONCLUSIONS
1-The lower the p H of an acid dye solution the more dye is adsorbed by inorganic gels. 2-The higher the pH of a basic dye solution the more dye Manganese, nickel, lead, strontium, mercury, calcium, and is adsorbed by inorganic gels. 3-The pH has a marked effect on the color. magnesium salts of orange I1 were also prepared. Their 4-The pH helps to govern the fastness of the dye. physical and chemical properties will be reported in a later 5-Adsorption of certain dyes by inorganic gels is chemical, paper. With these salts almost any desired shade of orange from a bright red to a pale yellow could be obtained. The a t least under the conditions stated above. 6-The power of the mordant to increase the fastness of lower the pH the nearer the color approached red. This' means simply that more and more of the salt is converted the dye may be due to the fact that it shares its valences between the dye and the fiber. into the acid as the hydrogen ion is increased.
