T H E PHYSICAL CHEMISTRY OF COLOR LAKE FORMATION. I V R E D CONGO ACID AND COXGO R E D LAKES BY HARRY B. WEISER AND ROBERT S. RADCLIFFE
Congo red, the disodium salt of diphenyl-disazo-bis-naphthylene-4-sulfonic acid is a red dye substantive to cotton. From osmotic pressure measurements on the dye solution using a parchment membrane, Bayliss' obtained a molecular weight 88 to 97 percent of the theoretical value, 696. Similar observations by Biltz and Vegesack* and by Donnan and Harris3 would indicate that the salt dissolves in the form of undissociated molecules. Conductivity measurements on the solutions indicate, however, that the salt is practically completely dissociated; but if this were the case, the osmotic pressure should be twice as large and the molecular weight half as large as the values actually obtained. To account for the conflicting observations, Biltz assumed that the particles have a molecular weight which is a multiple of the simplest value, the experimental value between 600 and 700 resulting from dissociation. I n line with this, Herzog and l'olotzky' found the diffusion constant of Congo red to be 0.126 X IO-^ a t 6.9" which is considerably smaller than would be expected if the molecular weight were but 696. For example, erythrosin B with a molecular weight of 800 and with 37 atoms instead of 7 0 for Congo red, has a diffusion constant of 0.265 X IO+ a t 6.6'. An apparent objection to assuming that the dye anions are associated is that the high conductivity of the salt would require large ions to have a relatively large migration velocity, contrary to what one might expect, a priori. This difficulty disappears in the light of McBain's5 observations on soap solutions. The anions in such solutions were found to be colloidal ions or ionic micelles (aggregates of simpler ions) possessing a mobility of the order of magnitude of that for the potassium ion. The slow migration velocity of large ions such as palmitate is due to its one electron being insufficient to charge up the ion to the surface density necessary for the average mobility. On the other hand, with the aggregates of ions, the ratio of the number of charges to the size, is such that the particles may possess a mobility even larger than the average. The colloidal nature of the Congo red anion is indicated by its very small diffusion into gelatin6 and its failure to dialyze? through parchment or col'Proc. Roy. S O C . ,81 B, 269 (1909); 84,229 (1912). Z.physik. Chem., 73, 481 (1910). J. Chem. Soc., 99, 1554(1911). Z.physik. Chem., 87,449 (1914). J. SOC.Chem. Ind., 37, 249 T (1918); McBain, Laing, and Titley: J. Chem. Soc., 115, I279 (1919);McBain and Salmon: J. Am. Chem. SOC., 42, 426 (1920);Proc. Roy. SOC., 97 A, 44 (1920). Herzog and Polotzky: Z. physik. Chem., 87,449 (1914). Teague and Buxton: Z. physik. Chem., 60,479(1907); Vignon: Compt. rend., 150,619 (1910).
'
'
1876
HARRY B. WEISER AND ROBERT S. RADCLIFFE
lodion. This is further evidenced by the following ultrafiltration experiment: Aqueous solutions of Congo red containing 0.025, 0.05,and I gram, respectively, of the purified salt per liter were filtered through a cellophane membrane ultrafilter under approximately 7 atmospheres pressure. The filtrate was perfectly clear in every case, but was slightly alkaline owing to partial hydrolysis of the salt forming the non-filterable blue colloidal acid, and alkali which will pass the membrane. These observations taken together with the dialysis experiments are quite indicative of the colloidal nature of the anion. The only other way of accounting for the failure of the anion to pass the membrane is to assume that the cellophane is acting as a semipermeable membrane and not as a sieve. This would mean that the membrane must show sufficiently strong negative adsorption for the Congo red anion to fill the pores with water. Actually the dye is adsorbed by the membrane. A further possibility is that the dye-cellophane adsorption complex shows strong negative adsorption for Congo red just as the copper ferrocyanide-potassium ferrocyanide adsorption complex exhibits negative adsorption for ferrocyanide and so prevents its passage under certain conditions.’ This possibility is ruled out, however, on the ground that the adsorption of an ion must be practically irreversible or the adsorption complex cannot act as a semipermeable membrane for the ion in question. This is approximately true for the adsorption of ferrocyanide by copper ferrocyanide from solutions below I normal, but is by no means true for the adsorption of Congo red by cellophane. There is, therefore, little doubt but that the red anions of Congo red are colloidal micelles or aggregates of single ions, too large to pass a fine ultrafilter and too hydrous and fine to be visible in the ultramicroscope. The solution of the red sodium salt becomes violet a t a pH value of 4 and blue at a pH value of 3.2 The blue color is due to a quite insoluble blue acid formed by replacing the sodium of the salt with hydrogen. This blue acid can form a stable negative sol containing particles sufficiently large and anhydrous to be visible in the ultramicroscope.
The Red Congo Acid I n addition to the blue Congo acid Schaposchnikoff and Bogojawlensk? claim to hava prepared a red acid by dissolving the solid blue acid in pyridine, evaporating, and heating the resulting salt a t 120’ to drive off the pyridine. Hantzsch4 points out that the acid prepared in this way is not red but is reddish brown and, in compact mass, a dark brown-violet color. The blue acid is reported to be quite soluble in alcohol and acetone giving stable solutions of the red acid. Hantzsch reports also that heating a dilute blue sol in a platinum or quartz dish sometimes causes it to turn to a red which reverts to the original blue on cooling. Weiser: “The Colloidal Salts,” 283 (1928). Salm: 2. physik. Chem., 57, 471 (1906). 8 J. Russ. Phys. Chem. SOC.,44, 1813 (1913). Ber., 48, I j8 (1915). 1
2
I877
PHYSICAL CHEMISTRY OF COLOR LAKE FORMATION
The effect of the nature of the solvent on the color change was shown by Hantzschl in the following way: Thirty cubic centimeters of 0.0004 molar salt wm diluted with 2 5 0 cc of liquid. If water was used the solution was turned to blue by 6 cc of X/IO alcoholic HCl whereas if alcohol was used the solution was red after the addition of 15 cc of N/IOalcoholic HC1 and did not change to blue for 3 to 5 minutes. As indicated above, the first transformation in color on adding dilute acid to Congo red is from red to purple or violet. The equation for the reaction may be represented as follows, R”standing for the Congo red anion. H2R 2 Na‘ z C1’ z Na’ R” 2 H’ 2 C1’ red blue
+ +
+
+
+
Hantzsch observed, however, that when equivalent amounts cf salt and acid are brought together in dilute solution, the color is purple and not blue. This purple color is believed by Hantzsch to be a mixture of red and blue acid; but this does not follow necessarily since the reaction indicated above is probably reversible to a certain extent. Indeed, Michaelis and Rona2found that neutral salts change the violet-colored solution to red. Similarly, Kolthoff3 showed that an N/so solution of sulfuric acid saturated with sodium chloride is colored red by Congo red and not blue as it is in the absence of salt. Furthermore, a one percent solution of sodium chloride renders Congo red useless as an indicator in acidimetry. Wedekind and Rheinboldt4observed that a violet solution prepared as described above turns red on heating and becomes violet again on cooling. Here also, one cannot say to what extent this is due to the formation of red acid and to what extent to reversal of the reaction with the formation of more of the red anion derived from the salt. Hantzsch is of the opinion that the color changes of all indicators are due to change in structure. He concludes therefore that the red and blue acids are the azoid and chinoid forms, respectively, of the following formulas: o < 0 2 s > ~ o ~ 5*
O
O < ~ ~ ~ > I O :HN ~. N H * C6H4 . CsH4. N H : CioHs %H* (chinoid acid, blue)
The only evidence for this difference in structure is a variation in the visible portion of the absorption spectrum of solutions obtained when equivalent amounts of Congo red and sulfuric acid are mixed ( I ) in aqueous solution and (2) in alcoholic solution. Ber., 48, 158 (1915 ) ; cf., also, IColthoff: Rec. Trav. chim., 42, 251 (1923). 2. Elektrochemie, 14, 2 j I (1908). Chem. Weekblad, 13, 284 (1916); Rec. Trav. chim., 42, 251 (1923). Ber., 52, 1013 (1919).
1878
HARRY B. WEISER AND ROBERT 6 . RADCLIFFE
Since the color of metallic sols such as gold is determined in part by the size of the particles, Wo. Ostwald has proposed a theory of color change in indicators based on change in the degree of dispersion of colloidal particles. Thus Ostwald' showed that the color change from red to blue in Congo rubin sol is produced by almost all electrolytes and that the effect of ions of different valency on the change is similar to that on typical hydrophobic sols. From these observations, he concludes that the color changes of Congo red are likewise due to differences in dispersion. M70.Ostwald's theory is not accepted* generally although i t is recognized that degree of dispersion may influence the color, in certain instances. But with indicators such as Congo red, litmus, phenolphthalein, etc., where the color is so closely related t o the hydrogen ion concentration and is so little dependent on other ions, it seems a very remote possibility that change in dispersion is the only factor coming in. Indeed, HallerS is of the opinion that the change of color of Congo rubin with acids is not the same in nature as that with salts.
Experimental Since the alkali dissolved from ordinary glass at room temperature is sufficient to change the color of the dilute Congo acid from blue to red it was necessary to carry out all observations in Pyrex or quartz vessels and to use particular care in the purification and storing of all reagents Preparation of the Blue Sol. A good grade of commercial Congo red was purified by recrystallization from alcohol according to the method of H ~ n t e r . ~ A 5-gram sample of the salt was dissolved in 400 cc of water and a sufficient excess of hydrochloric acid was added to give a blue curd. This curd was thoroughly washed in 2 5 0 cc bottles by the aid of the centrifuge, the supernatant liquid after each washing being discarded and replaced by water. At the outset of the washings, a small portion of the blue gel was repeptized but the bulk of the acid was thrown down during the centrifuging. After several repetitions of the washing process, there came a time when the whole mass of the gel waa peptized, none being thrown down by the centrifuge. A sample of this relatively pure, concentrated sol was placed in a Pyrex balloon flask, diluted, and dialyzed in a Neidle dialyzer with a cellophane membrane. After the first day the dialysate gave no test for chlorides but the operation was continued for two weeks in order to get as pure a product as possible. The sol used in the following experiments contained 1.04 gram of the acid per liter. The Acid from Pyridine. A a-cubic centimeter sample of the sol was evaporated to dryness on a water bath, in a transparent quartz dish and the residue dissolved in pure pyridine. The deep red solution of the pyridine salt Kolloidchem. Beihefte, 10, 179 (1919); cf. Schulemann: Biochem. Z., 80, I (1917). ZHantasch: Ber., 46, 1537 (19x3); 48, I 58 (1915 ) ; Kolloid-Z., 15, 79 (1914); Pihlblad: 2. physik. Chem., 81, 417 (1913); Voigt: Kolloid-Z., 15, 84 (1914); Kruyt andKolthoff: 21, 22 (1917). ZKolloid-Z., 27, I88 (1920);Liiers: 26, 1 5 (1920). 4 Biochem. J., 19, 42 (192 5 ) . 1
PHYSICAL CHEMISTRY O F COLOR LAKE FORM.4TION
I879
was taken to dryness on the water bath. As the evaporation proceeded, and a film of the solid appeared on the wall of the vessel, the crust assumed a reddish-brown appearance. After the liquid evaporated, the deposit was heated in a hot air oven to 120' to remove the pyridine completely. The film was blue by transmitted light and reddish brown by reflected light. Since the surface color of the solid blue acid is reddish brown, it is altogether probable that Schaposchnikoff and Hantzsch who worked with porcelain dishes mistook the brown surface color of the blue acid for the solid red acid. Solubility of the Blue Acid in Alcohol. Since the blue acid is reported to be fairly soluble in alcohol and acetone it was thought that evaporation of one of these solutions a t low temperature might yield the solid red acid. Accordingly, I O cc of the sol was placed in a 500 cc transparent quartz distilling flask connected to a vacuum pump through a suction flask surrounded by a freezing mixture. Evaporation was hastened by rotating the distilling flask a t intervals thus coating the walls with liquid. By this means a film of the acid was deposited which was blue by transmitted light but distinctly reddish brown by reflected light. To the flask was added jo cc of freshly prepared absolute alcohol. Contrary to what was expected, the acid proved to be almost insoluble giving but a faint brownish-pink solution. Thinking that the evaporation had influenced the rate of solution of the parhcles, some of the pure sol was added to the absolute alcohol. The addition of 0.05 cc of sol, containing I gram per liter, to I O cc of alcohol gave the brownish-pink color re~ gave a brown; 0.15 cc a lavender; and 0.25 cc a blue. ferred to above; o . cc On standing several hours, the excess blue sol coagulated giving the weak brownish-pink solution. This very slight solubility in alcohol did not agree with the observations reported by others or with our preliminary observations with some stock 95 percent alcohol. With a certain sample of the latter reagent, 3 cc of the sol gave a bright red solution, and 4 cc a brown. This solubility, more than 60 times that in the absolute alcohol, was traced to the preseme of alkali dissolved from the soft-glass bottle in which the alcohol was stored for a year or more. A freshly distilled sample possessed the weak solvent action which characterized the absolute alcohol. The brownish-pink solution in absolute alcohol yields a film of the blue acid on evaporating in a quartz dish. If an extremely dilute pale pink solution is evaporated, the very thin film appears brown rather than blue. It is impossible to say whether this brownish film contains some solid red acid or whether the red present is due to salt formed by interaction with the alcohol or with a minute trace of some impurity therein. The latter seems the more probable in view of the fact that evaporation of a pink aqueous solution gives a film of the blue acid. Solubility in Water. I n some preliminary observations with dilute sols of the blue acid in water, the sol was observed to change from blue to red on heating and to return to blue once more on cooling. I n order to study this phenomenon quantitatively, z cc of the original sol was diluted to 100 cc with conductivity water. Sols of the concentrations shown in Table I were made
I880
HARRY B. WEISER AND ROBERT S. RADCLIFFE
TABLE I Concentration of Dye Sol used in Experiments recorded in Table I1 Solution
Concentration of dye acid Grams per liter Milliequivalents per liter
Ccs diluted sol in I O cc
No. I
0.5
0.0OIO~
2
0.75
O.OOIj8
0.0048
3 4 5 6 7
1.00
0.00210
0.0064
1.25
0.00263
0.0080
I . 50
0.00315
0.0024
1.75
0.00368
0.0028
2.00
0.00420
0.0032
8
2. 50
0.0052j
0.0040
9
3.00
0.00630
0.0048
IO
4.00
0.00840
0.0064
I1
5.00
O.OIOj0
0.0180
0.0032
up in transparent quartz test tubes that had been thoroughly cleaned. Over the tubes was inverted a short test tube which served to prevent undue evaporation during heating. The tubes were placed in a thermostatically controlled electric oven and heated for 24 hours at each of the temperatures noted in Table 11. The level of the liquid in the tubes was marked and the slight loss in water by evaporation was replenished at intervals.
TABLE I1 Effect of heating Dilute Sols of the Blue Congo Acid Solution
No.
40°
50"
Color after heating a t 60'
rooo
pink
P*
pink
pink
pink
2
almost colorless blue
3
blue
almost colorless blue
4
blue
blue
5 6
7 8
blue blue blue blue
9 IO
I
75"
light red red
light red red
blue blue blue blue
very light blue light blue blue blue blue blue
lavender lavender blue blue
blue
blue
blue
blue
blue
blue
blue
blue
red red red brown red brown red lavender
PHYSICAL CHEMISTRY O F COLOR LAKE FORMATION
I881
The colors of the solutions after heating a t the several temperatures are given in Table 11. At room temperature all of the solutions appeared blue, differing only in intensity, and remained so indefinitely. On raising the temperature to 40' enough red acid was formed in solution I to give a mixture of red and blue that appeared colorless. At 50' most of the blue acid of solution I had dissolved, the solution becoming pink or light red. At this temperature, solution z became colorless and a t 60°, it turned pink. At the same time, red was clearly distinguished in solutions 3, 4,and 5, while a t the higher concentration the red was completely masked by the blue. At IOO', the most concentrated solution in the series contained enough red to make it appear distinctly lavender. If the red solutions are allowed to stand at room temperature they return to the original blue color, the dilute ones passing through the colorless stage. The rate of this change is slower the more dilute the solutions and the freer they are from the blue sol, the presence of which prevents supersaturation. The rate of transformation of a red solution to blue can be hastened by seeding with some of the blue sol. If this is a case of solution, the amount of red in presence of blue should be the same irrespective of the excess of blue present. Since the solutions are so dilute it was found impossible to determine, by absorption measurements] the amount of red in contact -4th blue. Moreover, the insoluble blue acid may adsorb some of the red anion thus reducing the concentration of the latter in proportion to the amount of the blue sol present. In spite of this latter source of error, the following observation gives quite conclusive evidence that the color change is the result of a solution process: Solutions numbers 3, 5, and 6 were heated to 70' until 3 became pink with a tinge of blue, 5 light lavender, and 6 lavender. All three solutions were then made the same strength by adding blue sol to solutions 3 and 5 and water to 6. As nearly as could be determined by colorimetric observations all of the sols were now of the same color. I n view of the similarity in color of aqueous solutions of Congo red and of the red acid, it is altogether probable that the color of the latter like the former is due to the red colloidal anion. The blue acid is relatively insoluble but to the extent that it dissolves, it yields a red colloidal anion. From this point of view, the blue color is that of the non-dissociated acid while the red color is that of the ion. Since evidence of the existence of a solid red acid is lacking, it is impossible to say whether there are two structurally different acids of Congo red. It seems unlikely that a change in temperature from 30' to 40° would cause the molecule to change from the azoid to the quinoid structure which Hantzsch assumes; but there is no definite proof either way. I n any case, there is no need of assuming such a transformation to account for the observations herein recorded, Congo Red Lakes The hydrous oxides of iron, chromium, and aluminum adsorb Congo red forming stable color lakes. From the standpoint of the theory of the lake formation process, the most interesting Congo red lakes are those obtained
IS82
HARRY B. WEISER AND ROBERT S. RADCLIFFE
from the blue acid. Bayliss' found that hydrous alumina adsorbs the blue acid from its colloidal solution. If this adsorption complex is washed, suspended in water, and heated, the color changes from blue to red. Since Congo red salts are red, Bayliss attributed this change in color to the formation of an aluminum salt. The experiments were extended to the precipitates obtained by mixing the blue negative sol with the positive sols of the hydrous oxides of aluminum, zirconium, and thorium. The blue adsorption complex becomes red on heating in every case, provided the hydrous oxide sols are dialyzed until practically free from acid. A small amount of acid is sufficient to prevent the color change. Assuming that the color change is due to the formation of a Congo red salt of aluminum or zirconium, there is no obvious reason why a trace of acid should prevent the change provided there is an excess of hydrous oxide with which the Congo red acid can react.* Blucher and Farnau3 attempted to get around this difficulty by assuming that the hydrous oxide adsorbs and stabilizes the free red Congo acid which they erroneously believed t o be instable in aqueous solution. Wedekind and Rheinboldt4confirmed Bayliss' observation and suggested, but did not prove, that aqueous sols of the blue acid contain more or less red acid which is changed to a blue isomer by acids. The red lake was believed to be a salt of indefinite composition formed by reaction of alumina with the red acid.
Experimental The Congo red sol used in the subsequent experiments was prepared as previously described. It contained 0.7 grams of the acid per liter. Since alumina gives a typical lake, the alumina sol was used as the starting point in the preparation of all lakes. It was prepared by precipitating a solution of aluminum chloride with ammonia, washing the gel by the aid of the centrifuge, suspending in water and peptizing completely with a small amount of hydrochloric acid. The sol was purified by dialyzing boiling hot in a Neidle dialyzer with a continuous flow of water a t the rate of two liters per hour for approximately IOO hours. The sol employed in the following experiments contained 1.7 grams of alumina per liter. Both sols were prepared and stored in Pyrex apparatus and the experiments were carried out in Pyrex tubes steamed before use. Color of Alumina Lakes. The first series of experiments was carried out to show the connection between the lake color and the relative amounts of dye and alumina. I n all cases, the lakes were prepared by the aid of a mixerJ which was known to give rapid and uniform mucing of the two constituents. The mixture was poured immediately into Pyrex test tubes which were placed in a water bath a t 100' for two hours. Thereafter, the tubes were allowed to 'Proc. Roy. Soc., 84 B, 881 (1911). "ancroft: J. Phys. Chem., 19, 57 (1915). J. Phys. Chem., 18, 634 (1914). Ber., 52, 1013 (1919). 6 Weber and Middleton: J. Phys. Chem., 24, 48 (1920).
1883
PHYSIC.4L CHEMISTRY O F COLOR LAKE FORMATION
TABLE I11 Color of Congo Red-Alumina Lakes Sols mixed (Total volume z j cc) In ccs In milliequivalents A120a Dye A1203 DY e 9
I
8 7 6
2
0.9 0.8
5 4 3
7 8 9
2
I
0.j
0.4
0.0130
0.6
Supernatant liquid
Red Red Red Red Purple Purple Purple Blue Blue
Pink Pink Colorless Colorless Red Red Red Red Blue
0.0022
0.0043 0.006j 0.0086 0.0108
0.7
3 4 5 6
Color ai‘ter heating Precipitate
0.3
0.OIjI
0.2
0.0173
0.I
0.0194
stand for twelve hours and the observations noted in Table I11 were made. Although there is present at all times a large excess of alumina, the lakes vary in color from red through purple to blue as the concentration of the dye sol is increased. It will be noted that, as usual, the range of complete mutual precipitation of oppositely charged sols, is narrow. E$ect of Acid. The effect of acid on the color of a given lake is shown in Table IV. It will be seen that under the conditions of the experiment o.oooo j normal hydrochloric acid prevents the formation of a pure red lake. Heating in the presence of acid causes complete coagulation of the sols.
TABLE IV Effect of Acid on Color of Congo Red-Alumina Lakes Ingredients mixed in cubic centimeters (Total volume z cc) d C lN) AlzOs Dye (o,025 5
2
0.0
5
2
0.5
5
2
1.0
5
2
2 .O
j j
2
z
4.0 6.0
Without heating after two days Precipitate Liquid
Red Blue Blue
f: Precipitate
Color After heating Precipitate
Liquid
Red Purple Purple Purple Purple Blue
Colorless Colorless Colorless Colorless Colorless Colorless
Colorless Red Pale blue Light blue Blue Blue
Mechanism of the Lake Formation Process. In the light of the above observations and those recorded in the previous section, the mechanism of the formation of the lakes appears to be as follows: I n the blue sol the following equilibria are set up:
nHzR % nHzR % 2nH‘ blue sol solution
+ nR” red
1884
HARRY B. WEISER AND ROBERT S. RADCLIFFE
where R" is the anion of Congo red. Rise in temperature displaces the equilibria to the right. On mixing the hydrous oxide sol with the blue dye, mutual coagulation of oppositely charged particles results in the formation of a blue lake. Following this process the alumina adsorbs the red colloidal ion shif. ting the equilibrium gradually to the right. The color of the lake ultimately formed depends on the hydrogen ion concentration and the amount of the blue dye. If the amount of dye is not in excess of the adsorption capacity of alumina for the red anion, the lake will be red. If there is an excess of the blue dye, the blue color blends with the red giving purple or if the excess of blue is sufficiently great the red color is masked completely. Rise in temperature increases the rate a t which the adsorption equilibrium is set up. The addition of even a small amount of acid cuts down still further the low solubility of the acid as well as the degree of ionization of the acid, and blue lakes only are formed. From these considerations it follows that if the excess of blue dye above the adsorption capacity of alumina for the red anion, were such that it would dissolve and become red on heating, the lake should be red when hot and purple to blue when cold. This proved to be the case: To z j cc of a blue sol containing 0.00067 gram of dye was added I cc of the alumina sol. The blue lake was changed to red on heating but on standing for some time at room temperature, it became purplish blue. The observations indicate that the red lake like the blue is an adsorption complex and not an aluminum salt. If the alumina sol is free from cations other than hydrogen, then the red lake consists ultimately of an adsorption complex of alumina and the red acid. T o the extent that alkali or ammonium cations are present, the dye may be held as the red alkali or ammonium salt. Bancroft' believes that some ammonium must be present in the alumina and chromic oxide sols used by Weiser and Porter2in preparing their alizarin lakes, otherwise the lakes would not be red, since alizarin i s yellow and not red. This contention is valid unless it should turn out that a thin film of alizarin adsorbed on a hydrous oxide is red. The mechanism of the dyeing of cotton red from the sol of the blue acid is similar to that of the formation of the red color lakes. summary
Aqueous solutions of Congo red contain a red colloidal anion which is too small and hydrous to be visible in the ultramicroscope but which will not pass a membrane permeable to ions in true solution. Replacing the sodium in Congo red with hydrogen gives a blue col2. R" (red) zH' loidal acid. A typical equation for the reaction is: z Na' 2 C1' % H2R (blue) 4- z Na' z Cl', where R" stands for the anion of the dye. Since this reaction is partly reversible Congo red is not suitable for an indicator in acidimetry, in the presence of salts. I.
+
1
Private communication. 1824 (1927).
* J. Phys. Chem., 31,
+
+
+
PHYSICAL CHEMISTRY O F COLOR LAKE FOR>fATION
188j
3. The blue acid is very slightly soluble in water yielding the red colloidal anion. The solubility increases with rising temperature. In the blue sol the following equilibria exist: nH2R (blue) nH2R (in solution) kj 2nH' nR' (red). 4. The positively charged hydrous oxide mordants, such as alumina, adsorb the blue colloidal acid giving blue lakes. If the amount of blue acid is small the blue lakes change to red very slowly a t ordinary temperatures and more rapidly a t higher temperatures. The process consists in the adsorption of the red anion by the hydrous oxide displacing the reaction given in (3) to the right until all the blue acid disappears. If the hydrogen ion concentration of the bath is too high or the amount of blue acid is in excess of the adsorption capacity of the hydrous oxide for the red anion, the lake remains purple to blue in color. 5 . The blue lake is an adsorption complex of the blue acid and the hydrous oxide. The red lake is an adsorption complex of hydrous oxide and the red acid (or alkali salt). The red alumina lake is not an aluminum salt of Congo red as assumed by Bayliss. 6. The mechanism of the dyeing of cotton red from a sol of the blue acid is similar to that of the formation of the red hydrous oxide lakes, 7 . Definite assurance of the existence of the red Congo acid in the solid state is lacking. The product formed by decomposing the pyridine salt is blue by transmitted light and reddish brown by reflected light. It is probable that Schaposchnikoff and Hantzsch mistook the surface color of the blue acid for the alleged red isomer.
+
The Rice In&%.de, Houston. Texas.
