The determination of the constitution of an unknown commercial dye is one of the most interesting, but also one of the most difficult, tasks of the dye chemist. In the earlier days when only natural products were used, it was a relatively easy task, in many cases, to determine the origin of a product by examining its outward appearance. Cochineal, indigo, alizarin (madder) have such characteristic appearances that even the inexperienced could distinguish them. The only question in- volved was whether the product was suitable for the particular use in dyeing, or whether it had been either damaged or adulterated. The widely used inorganic colorants, such as Prussian blue, chrome oxide, iron oxide, vermilion,, etc., could also be identified by purely qualita- tive investigation. The situation became quite different with the ap- pearance of commercial organic dyes. Although in the early days of the dye industry, it was still relatively easy to identify, for example, fuchsin, aniline blue, methyl violet, or simple azo dyes, the difficulties increased as more new dyes appeared on the market. To be sure, attempts were made to alleviate these difficulties by compiling tables in which all known dyes were described accurately. So many new dyes were evolved as time went on, however, that it was not possible to keep pace with current developments, and the well-known dye tables of Schultz and Green became of less and less service. The question is no longer: "With which of the dyes described in the tables is this product identical?" but: "What is the composition of a dye which has not been mentioned in any scientific publication?"
The once widely used works, therefore, could no longer be used as the foundation for modern dye analysis, and new methods, independent of the older works, had to be worked out.
It would lead too far afield to try to present all the known facts in this book. On the other hand, it is of interest that the beginner should become acquainted, at least in broad outline, with the principles of modern dye analysis. To this end, references to publications in this
399
400 ANALYSIS OF COMMERCIAL DYES
field are given at the end of this chapter, and three examples are given to illustrate the procedures used in determining the constitution of an unknown dye. This subject has been dealt with in some detail in an address given by the author.*06
Inasmuch as there is a lack of the more exact bases analogous to those on which the science of analytical chemistry is built, methods must sometimes be used which are not encountered in scientific re- search. Basically, however, the methods of modern dye analysis involve the same general principles as the classical analytical procedures. First of all, one must rely upon the available literature to gain a clear opinion as to how to proceed. Today, this literature is not contained, or is contained in only exceptional cases, in the scientific publications which are eventually collected in reference books like those of Beilstein and Gmelin-Kraut. There are, however, other sources which are useful in many cases. These sources are published patents, the trade journals of dye chemistry, and technical communications, all of which may give hints about a product in question.
The product often may be mentioned in a patent and frequently the dye package is labelled "Patented." The same dye is usually "an- nounced" in the trade journals, and it may be assumed that the date of patenting is not far from the date on which the dye was placed on the market. Furthermore, the manufacturer of the dye is almost always known. Thus, it may be inferred that the dye has been patented by the manufacturer, either by means of an application made by him, or by assignment, and that the patent was granted, or at least applied for, before the first sale of the product. This type of investigation, of course, does not involve the methods of exact science, but, as it was once stated, it is more of the nature of detective work.
Still further importance is attached to the fact that a patent exists.
Most patent laws require only that the general procedure be illustrated by several characteristic examples, and not that the disclosure specify the exact compound which is to be manufactured subsequently. The Swiss patent law, on the other hand, states expressly that only those products which are accurately described in the patent shall be given patent protection, and, in addition, that each patent can be drawn to only one product. If it is known, therefore, that a dye is patented in Switzerland, it can be assumed that the unknown dye is accurately described in one particular patent. A Swiss patent, therefore, could be of the greatest use in clearing up the constitution of a dye. In many
106 Fierz-David, /. Soc. Dyers Colourists, 45, 133 (1929).
ANALYSIS OF COMMERCIAL DYES 401 cases, these methods fail completely, as when no patent has appeared, or when the dye is manufactured by another firm after the expiration of the patent, or when the name of the dye is changed. Frequently in these cases, the structure can be ascertained by comparison of the properties of the dye with those of a known product.
At one time, the first source to be consulted for information on the nature of a dye was the well-known dye tables of Arthur Green. This work is so far out of date today that it is rarely consulted. The simple color reactions are of little use in modern times, and it is necessary, therefore, to evolve new methods which, unfortunately, are frequently time consuming and still lead to no useful end.
The group to which a dye belongs must first be established. This is relatively easily done, since the diflFerent dye classes exhibit diflFerent reactions. Thus, a vat dye would be sought in that group which gave the correctly colored vat. Indigo and thioindigo dyes give yellow or color- less vats. They dissolve in concentrated sulfuric acid to give, usually, yellowish green solutions, and they are reprecipitated unchanged from the alkaline reduction mixtures by air oxidation. With anthraquinone vat dyes, it is observed that the vats are usually intensely colored, and this difference permits easy differentiation between indigo and anthra- quinone dyes. Further, many anthraquinone vat dyes yield anthracene or anthracene derivatives on distillation with zinc dust.
Many heterocyclic dyes behave analogously to ordinary indigo in that they are decolorized when treated with reducing agents and are regenerated on reoxidization. This applies to the azines, thiazines, ox- azines, and other similar dyes. In contrast to these, the triphenylmethane dyes are decolorized easily by reduction, but their reduction products are usually considerably more difficult to reoxidize by air, i.e., the leuco compounds in this series are relatively stable. However, they can be very easily and quantitatively oxidized by chloranil (see page 146).
It is possible in some cases to obtain analytically pure dyes. This is most successful with the vat dyes, which can frequently be crystal- lized from a high boiling solvent such as chlorobenzene or nitrobenzene, or from glacial acetic acid or pyridine. Tetrabromoindigo (Ciba blue 2B), for example, can easily be obtained analytically pure from dichloro- benzene, as can other vat dyes of the type of indanthrene blue. With these dyes, quantitative chemical analysis is often of great value. Other special methods of analysis can also be employed, such as the Zeisel determination of alkoxyl groups.
402 ANALYSIS OF COMMERCIAL DYES
Consideration must also be given to the spectroscopic method of Formanek which depends on the determination of the absorption maxi- mum. Formanek showed that many dye groups had characteristic ab- sorption spectra, and his method is often of use. The tables compiled by Formanek, however, are out of date and are helpful only in recogniz- ing the dyes which are listed. They are of as little use as the Schultz and Green tables with respect to dyes which have not been described.
One method which is very useful in the azo dye field is the so-called reductive-splitting reaction. Most azo dyes can be split at the —N—N—
group to produce two amines which can be separated and studied. Any nitro groups which are present are simultaneously reduced, of course.
The reaction is effected by various reducing agents: hydrosulfite, stan- nous chloride, zinc dust, and many others. No reducing agent can be applied universally/ In certain cases, hydrosulfite may bring about not only fission, but also the introduction of a sulfo group into one of the fission products. Stannous chloride, on the other hand, may cause a rearrangement (benzidine or semidine type) of the first-formed hydra- zo compound. For example, orange I, the dye from diazotized sul- f anilic acid and a-naphthol, on reduction with stannous chloride in hy- drochloric acid solution, does not yield 1,4-aminonaphthol and sulfanilic acid, but instead it forms the semidine by rearrangement and no splitting occurs. In other cases, hydrosulfite carries the reduction only as far as the hydrazo stage and no splitting ocurs. These cases, however, are the exceptions.
Since it is not within the scope of this book to go into great detail, we shall consider only the general principles of azo dye analysis.
It is first established whether the product is a single compound. This is done by dusting a small sample of the dye on moistened filter paper.
A mixture can frequently be recognized from spots of different color.
(Mixtures of water-insoluble dyes can often be detected by sprinkling a small sample onto concentrated sulfuric acid, advantageously placed in a depression in a white porcelain plate for the purpose.) Attempts are then made to purify the dye by reprecipitation until a pure material is obtained. Every effort is made to obtain the purest sample possible.
This is a general rule for all dye analyses.
Reductive splitting is carried out, after the best conditions for the reaction have been established by experiment, and the resulting solu- tion is investigated. Occasionally, part of the reduction products separ- ate directly from the warm solution in a more or less pure condition.
These are filtered off, recrystallized, and, if possible, analyzed quantita-
ANALYSIS OF COMMERCIAL DYES 403 tively. The reduction product may already be described and the identity can be established by carrying out any necessary reactions. The small book by Brunner is useful in this connection.
Solutions prepared by stannous chloride reduction can advantage- ously be detinned electrolytically. This is an easy process to carry out and it has the advantage over H2S precipitation that no foreign substances are added to the solution.
Mixtures of various dyes, as well as other compounds, can sometimes be separated by the Tswett chromotographic adsorption method. This elegant method is not widely applicable, however, succeeding only with simple dye mixtures.1 ° 7
•If a pure reduction product is obtained, it is analyzed quantitatively and then it is established whether the product found can be related to an example in a patent. A quantitative analysis is not necessary in those cases where the product is a well known compound (aniline, sulfanilic acid, H acid, etc) or where its identity can be established by reference to tables. Thus, 1-amino-y acid and 7-amino-H acid are recog- nized immediately, without further work, on the basis of accurately described color reactions. These color reactions are carried out by placing a dilute solution of the substance being investigated on filter paper and spotting it with various reagents such as metal salt solutions, acids, alkalis, oxidizing agents such as ferric chloride, hydrogen perox- ide, etc. The colorations produced lead, in very many cases, to im- mediate recognition of the fission product, thus establishing part of the structure of the dye. When enough of the fission products have been identified, and the probable structure of the dye is established, perhaps by reference to a patent, the synthesis of the probable structure is undertaken. When the dye has been synthesized successfully, the task is complete.
Three examples will now be given of dyes which are not recognized from their appearance. These examples should show how one proceeds, in any given case, to identify a dye product. The dyes in question are:
( I ) Polar brilliant red 3B (Geigy). This dye has not been described in a German patent; (2) Benzo light grey BL (By). A patent could be found by searching the literature; and (3) Brilliant sulfo flavine (I.G.).
Here, the inventor and the German patent number are identified in a communication which is available to anyone, and an analysis of the dye therefore appears unnecessary.
i07Ruggii and Jensen, Helv. Chim. Acta, 18, 624 (1935). See also Zechmeister and Cholnoky, Die chromatographische Adsorptionsmetnode. 2nd ed., Springer, Wien, 1938.
404 ANALYSIS OF COMMERCIAL DYES
Polar Brilliant Red 3B and B
These two dyes were recently added to the sample card of the acid- and milling-fast Polar dyes of the J. R. Geigy A.G. firm. Since methods for preparing milling-fast azo dyes have not been patented by this firm for several years, it is assumed that the two Polar brilliant red dyes either come under an old disclosure or are not patentable at all.
Polar Brilliant Red 3B
Reduction with Hydrosulfite. The reprecipitated dye (30 grams) is dissolved in 250 cc. water containing enough soda to make the solution distinctly alkaline to litmus. Hydrosulfite is added in small portions to the boiling solution until it becomes colorless. An excess of hydrosulfite is to be avoided or sulfur may separate.
The yellowish brown oil which separates from the yellow reduction solution solidifies on cooling. It is recrystallized twice from aqueous alcohol, using decolorizing carbon, yielding a base melting at 68°C.
The filtered reduction solution is acidified and treated with salt, and after a short time a voluminous white precipitate is formed. This is fil- tered off and washed with salt solution. From 50 grams of the commercial product, 9 to 10 grams of purified base and 15 grams of the sulfonic acid are obtained.
The base has the following properties: It dissolves in dilute hydro- chloric acid only on warming, and, on cooling the solution, a hydrochlor- ide separates which melts at 168-170°C. and which is hydrolyzed by water. If the hydrochloric acid solution is heated to about 80°, partial decomposition occurs, producing an oil volatile with steam. The base contains halogen, but not sulfur. Nitrogen and halogen determination give a 1:1 ratio of chlorine to nitrogen and a molecular weight of 219.
The acetyl derivative melts at 166°.
The salted-out sulfonic acid can be diazotized and coupled, but is very stable otherwise. Its properties suggest N-acylated H or K acid.
Reduction with Stannous Chloride. A solution of 20 grams of the purified dye in 250 to 300 cc. water is heated to boiling in a round-bot- tomed flask fitted with reflux condenser and stirrer. A solution of 40 grams of stannous chloride in 100 cc. concentrated hydrochloric acid (1.19) is added and the mixture is boiled for 3 hours. If the solution is still not completely decolorized, more of the stannous chloride solution is added.
ANALYSIS OF COMMERCIAL DYES 405 The tin is now removed electrolytically from the reduction mixture.
The liquid is placed in an acid-resistant clay cell of 350-cc. capacity and the latter is placed in a porcelain or Pyrex beaker filled with 10 per cent sulfuric acid to the same level as the liquid in the cell. The tin is plated out on a copper gauze electrode at a temperature of 80-90°, using a carbon rod as anode. At an E.M.F. of 8 volts and a current of 6 to 8 amperes, all of the tin used to reduce 20 grams of dye is removed in 4 to 5 hours. Electrolyzing is continued until hydrogen begins to be evolved. Nothing crystallizes from the detinned solution, so it is evapor- ated in vacuum to half its volume. On cooling, a light gray powder comes out. This material exhibits the reactions of 7-amino-H acid. The spec- trum of the oxidized material has the following bands: k=530 and490 xrifi (from 7-amino-H acid, X = 528 and 491 m^)
The filtrate, which smells strongly of toluenesulfonyl chloride, is evaporated to dryness. The gray residue is treated with soda solution, and the solution is shaken out with ether. From the ether extract, a substance crystallizes which has a melting point of 140°C. and which is soluble in hydrochloric acid and sodium hydroxide, but not in soda solution. Ferric chloride gives a red color with the substance in hydro- chloric acid solution. The compound contains halogen and can be diazo- tized and coupled with R salt to produce a dye. It is apparently a chloro- aminophenol.
One gram of the sulfonic acid from the hydrosulfite reduction is boiled under reflux with 20 cc. 10 per cent hydrochloric acid. After a short time, the solution can be oxidized by air or an oxidizing agent to produce the characteristic red color obtained from 7-amino deriva- tives of l-amino-8-naphtholsulfonic acids. Since it is known that the Polar dyes contain the toluenesulfonyl group, it may be assumed that the acyl residue connected to the N is the p-toluenesulfonyl group.
The identification of the base is accomplished as follows. Since an aminophenol is formed in the stannous chloride reduction, the original structure was most probably that of an ester or an ether. 8 grams of the base is boiled with 80 cc. 20 per cent hydrochloric acid in a flask fitted with a downward condenser and steam is introduced simultane- ously. A lachrymatory liquid distills. It is heavier than water and boils at 175°C. The compound, when warmed with silver nitrate, produces silver chloride. It is oxidized very rapidly by neutral permanganate solution to produce benzoic acid, m.p. 121°. The distillate is therefore benzyl chloride.
The liquid remaining in the flask is made alkaline with caustic soda
406 ANALYSIS OF COMMERCIAL DYES
and filtered. It is then acidified and treated with soda solution. In the course of a day, white plates crystallize out. They melt at 139°C.
This material is identical with the chloroaminophenol produced in the acid reduction of the dye. It is also identical with the chloroaminophenol derived from 2-nitro-4-chlorophenol. The original base is therefore 4- chloro-2-aminophenylbenzyl ether. This compound can be prepared synthetically by heating l,4-dichloro-2-nitrobenzene with 2.5 moles of 10 per cent sodium hydroxide in an autoclave for 10 hours at 150-160°
(see page 109). The resulting sodium salt of nitrochlorophenol is heated for 5 hours with benzyl chloride in alcoholic solution,108 and the nitro ether is reduced. The base obtained in this way melts at 68°, either alone or mixed with the base from the dye.
By analysis, therefore, Polar brilliant red 3B has the structure:
Synthesis of the Dye. To a solution of 0.1 mole of H acid in 200 cc. water containing 0,2 mole of soda is added with stirring, at 60-70°C, small portions of toluenesulfonyl chloride until the solution shows no reaction with nitrite. A two- to threefold excess of toluenesulfonyl chlor- ide is required since a toluenesulfonyl group goes onto the hydroxyl group also. When the reaction is complete, enough soda is added to make the mixture a 10 per cent soda solution, and the mixture is boiled for 30 minutes to hydrolyze the toluenesulfonyl group on the —OH. The solu- tion is then cooled in ice and the diazo solution is added. The reaction mixture is stirred for several hours, then warmed to 60° and salted out with a small amount of salt. The dye is purified by reprecipitation. No differences are found between the synthesized product and the original standard, either in dyeing properties or in characteristics in solution.
i°8Frische, Ann., 224, 141 (1884).