Prof. Ya. I. Mikhailenko and T. M. Demchenko

Complex Salts of Organic Dyes with Silver

Organic dyes containing an “amino” group or an “imino” group are substituted ammonias or ammoniums; and therefore, by analogy with the latter, it may be assumed that they can form complexes with silver.

Compounds of this kind have not been described in the literature. It may be that, upon detailed searches, only a few fragmentary data could be found.

Obtaining these compounds is of interest not only for supplementing our knowledge concerning the composition of complex silver compounds. One may think that these compounds will not be indifferent to light and therefore may find application in the development of photography in natural colors. In addition, one may think that they may also be of significance as pharmaceutical preparations¹.

The experiments were carried out with auramine, Congo red, phenosafranine, and chrysoidine.

For all the above-mentioned dyes, complex salts with silver have not been described in the literature.

\[ \text{Auramine}\quad \mathrm{HN}=\mathrm{C}<_{\mathrm{C_6H_4N(CH_3)_2}}^{\mathrm{C_6H_4N(CH_3)_2}} \]

Auramine, which was used for the experiments, was partly obtained in the laboratory and partly was a commercial preparation of auramine hydrochloride with m.p. \(264^\circ\).

Action of \(\mathrm{AgNO_3}\) on HCl—auramine. To a heated alcoholic solution of HCl—auramine there was added an excess of an aqueous solution of \(\mathrm{AgNO_3}\). An abundant precipitate of yellowish color at once separated out, apparently finely crystalline. The precipitate was filtered off. The filtrate, slightly colored yellow, contained an excess of \(\mathrm{AgNO_3}\)—it did not deposit silver in the residue, i.e. on standing it gave no precipitate.

The dried precipitate was extracted with ether on heating (to the boiling of the alcohol). From the filtrate, while still warm, there separated out a crystalline substance of yellow color with an orange tint. Under the microscope it appeared as well-formed, transparent, yellow-greenish crystals in the form of prisms.

In the residue, after extraction of this substance, an amorphous-appearing substance of gray-yellowish color was obtained.

In different experiments the crystalline substance was obtained sometimes at once completely homogeneous, sometimes clearly contaminated by an admixture of another substance, under the microscope not crystalline or only slightly crystalline.

Such a contaminated substance, on ignition in a crucible, gave a small bead of silver and was obtained in the case when a hydroalcoholic solution of HCl—auramine was taken and heated longer.

In the case of contamination by this foreign, silver-containing substance, the crystalline substance was purified by repeated crystallization from alcohol, since the contaminating substance dissolved in alcohol with more difficulty.

The solutions of the crystalline substance have a bitter taste. The crystals melt at \(200^\circ\)—\(210^\circ\). The m.p. \(210^\circ\) is apparently already that of a homogeneous

the substance upon further crystallization is lowered to \(200^\circ\), while the color changes somewhat, passing from orange to lemon-yellow.

With a solution of diphenylamine in sulfuric acid a dark-blue coloration is obtained.

On combustion of the air-dried substance with CuO the following results were obtained:

1. A sample of 0.1690 g gave 0.3444 g of CO\(_2\) and 0.1271 g of H\(_2\)O
2. A sample of 0.1711 g gave 0.3485 g of CO\(_2\) and 0.1187 g of H\(_2\)O
3. A sample of 0.1170 g gave 17.5 c.c. of nitrogen, measured at \(25^\circ\) C. and 754.9 mm pressure.

Thus, the substance obtained apparently is a hydrate of auramine nitric-acid salt, which is obtained by double decomposition of HCl—auramine with AgNO\(_3\), according to the equation:

\[ (C_{17}H_{21}N_3)'\,Cl' + Ag'(NO_3)' = (C_{17}H_{21}N_3)'(NO_3)' + AgCl \]

Preparation of HNO\(_3\)—auramine by the action of HNO\(_3\) on auramine. For verification, HNO\(_3\)—auramine was obtained by the action of HNO\(_3\) on the free base of auramine.

The free base of auramine (auramine) was obtained by the action of caustic soda on HCl—auramine. Recrystallized from alcohol, it consisted of lemon-yellow leaflets with m.p. \(125^\circ\)—\(128^\circ\).

To an alcoholic solution of auramine a weak solution of HNO\(_3\) was added; at once orange-yellow crystals of HNO\(_3\)—auramine precipitated. Recrystallized from alcohol, the salt acquired a lemon-yellow color and had m.p. \(200^\circ\). Under the microscope—the same transparent yellow crystals as in the first experiment, only shorter and thicker.

Solutions of HNO\(_3\)—auramine have a bitter taste. With a solution of diphenylamine in sulfuric acid—a dark-blue coloration.

Thus the substance obtained by the action of AgNO\(_3\) on HCl—auramine is, in its properties, identical with the synthetically obtained nitric-acid salt of auramine. Nitric-acid auramine is not described in the literature. There is only an indication that hydrochloric auramine with AgNO\(_3\) does not give a precipitate of silver chloride.\(^1\)

Silver salt of auramine from HCl—auramine and AgNO\(_3\). In the experiments described above on the action of AgNO\(_3\) on hydrochloric auramine, it was indicated that under certain experimental conditions nitric-acid auramine is obtained contaminated with an admixture of another substance containing silver. Namely, if one takes an aqueous-alcoholic solution of HCl—auramine and allows it to stand for several days, then on addition of AgNO\(_3\) in different experiments a precipitate is obtained containing a greater or lesser amount of the silver-containing substance; in certain experiments alcohol extracted this substance alone.

The experiment was carried out as follows: to a hot aqueous-alcoholic solution of HCl—auramine an aqueous solution of AgNO\(_3\) was added—immediately a precipitate fell out. The precipitate was filtered, dried, and extracted with alcohol at boiling. The residue after boiling with alcohol was an amorphous substance of gray-yellowish color (AgCl and an admixture of the silver salt of auramine).

From the hot alcoholic filtrate a yellow substance separated in the form of flakes (more complete precipitation was effected with water). Under the microscope the flakes appeared as thin hairs gathered into fan-shaped bundles and

\(^1\) Chem. Centralbl. 1911. II. 140.

optics. An attempt to crystallize this substance showed that each time one obtains needles of a gray-yellow color, insoluble in alcohol upon boiling with it; on boiling with alcohol some alteration of the substance occurs (analyses of this residue showed that it contains somewhat more silver than the original). Under the microscope this residue, insoluble in alcohol, has the same appearance as the portion that had dissolved. The same [[unclear: word]].

The experiment is modified further in this way: HCl—auramine and \(\mathrm{AgNO_3}\) are poured together under conditions identical with those described above, but then the precipitate is treated with abundant \(\mathrm{NH_4OH}\), with the aim of dissolving the \(\mathrm{AgCl}\) formed. The insoluble part of the precipitate, when examined under the microscope, consists of fine, short hairs such as those described above. When dried, it crystallizes out of alcohol, and, as in the preceding experiments, part of it, on boiling with alcohol, changes into a substance insoluble in alcohol.

The substance obtained in both the one and the other case dissolves much better in methyl alcohol and acetone.

It melts at \(170^\circ\)—\(175^\circ\) with decomposition, which begins at \(150^\circ\)—\(155^\circ\); at the beginning of the decomposition the substance in external appearance becomes similar to that modified part which is obtained on boiling with alcohol. It gives no Beilstein reaction for halogen. With a solution of diphenylamine in sulfuric acid a dark-blue coloration is obtained.

Determinations of silver gave the following results:

1. Sample of substance dried at \(100^\circ\mathrm{C}\) . . . . . . . . . . \(0.1871\) g
   Residue after ignition . . . . . . . . . . . . . . . . . . . . . . \(0.0287\) »
2. Sample of substance dried at \(100^\circ\) . . . . . . . . . . . . \(0.1437\) »
   Residue after ignition . . . . . . . . . . . . . . . . . . . . . . \(0.0220\) »

On combustion of the air-dried substance with \(\mathrm{CuO}\), it was found:

1. Sample \(0.1530\) g gave \(\mathrm{CO_2}\)—\(0.3115\) g and \(\mathrm{H_2O}\)—\(0.0800\) g.
2. Sample \(0.1498\) g gave \(\mathrm{CO_2}\)—\(0.3063\) g and \(\mathrm{H_2O}\)—\(0.0789\) g.

Thus, under the known conditions of the experiment, the action of \(\mathrm{AgNO_3}\) on auramine hydrochloride yields a complex silver salt of the composition:

\[ [(\mathrm{C}_{17}\mathrm{H}_{21}\mathrm{N}_3)_2\mathrm{Ag}]\cdot[\mathrm{NO_3}]. \]

The composition of the complex silver salt dictates the taking of auramine (free base) in \(2:1\). The experiment was carried out as follows: to an alcoholic solution of auramine, the proper quantity of \(\mathrm{AgNO_3}\) was added. Immediately a precipitate of a silver salt formed; after a certain time the precipitate was filtered off, washed free of \(\mathrm{AgNO_3}\), and dried. It was a yellow substance, paler in color than the silver salt described above, but, as was shown by analyses, identical with the latter. Melting point \(170^\circ\)—\(175^\circ\). With a solution of diphenylamine in sulfuric acid it gives a dark-blue coloration.

Determinations of the silver content gave the following results:

1. Sample of substance dried at \(100^\circ\) . . . . . . . . . . . . \(0.1533\) g
   Residue after ignition . . . . . . . . . . . . . . . . . . . . . . \(0.0234\) »
2. Sample of substance dried at \(100^\circ\) . . . . . . . . . . . . \(0.1398\) »
   Residue after ignition . . . . . . . . . . . . . . . . . . . . . . \(0.0218\) »

Upon crystallization from alcohol, this salt gave the same picture as with the silver salts obtained by the action of HCl-auramine on \(AgNO_3\), i.e., in this case too the salt partly dissolved in boiling alcohol, then, upon cooling of the solution, precipitated in the form of fine hairs gathered into bundles; another part underwent some change, passing into an insoluble state and acquiring a gray-yellow color.

Thus, auramine gives a complex silver salt of the composition

\[ [(C_{17}H_{21}N_3)_2Ag]\cdot (NO_3). \]

By the action of an ammoniacal solution of \(AgNO_3\) on hydrochloric auramine, it once proved possible to obtain a silver salt, of the analytical composition

\[ [(C_{17}H_{21}N_3)Ag]\cdot (NO_3). \]

But this observation still requires confirmation.

\[ \begin{array}{c} \textbf{Congo red}\\[2mm] \begin{array}{ccccc} C_6H_4 & - & N=N & - & C_{10}H_5\left\langle\begin{array}{l}SO_3Na\\ NH_2\end{array}\right.\\ \big| &&&&\\ C_6H_4 & - & N=N & - & C_{10}H_5\left\langle\begin{array}{l}SO_3Na\\ NH_2\end{array}\right. \end{array} \end{array} \]

The preparation of Congo red used for the experiments was prepared in the laboratory.

An aqueous or aqueous-alcoholic solution of Congo red was taken and mixed with an aqueous solution of \(AgNO_3\) (1 mol. of dye to 3 mol. of \(AgNO_3\)); immediately a precipitate fell out (a jelly), which settled very slowly. The filtrate, containing an excess of \(AgNO_3\), was colorless. The precipitate was washed first by decantation, then on the filter until \(AgNO_3\) had been completely removed, and was dried in air. It is dark-brown in color, with a very high melting point (at \(280^\circ\) the substance still does not melt). In water [[unclear: phrase obscured]], apparently, as a colloidal solution. It dissolves little in alcohol.

Determinations of silver gave the following results:

1. Sample of substance dried at \(100^\circ\) . . . . . . . . \(0.3761\) g
   Residue after ignition . . . . . . . . . . . . . . . . . . \(0.1258\) »

2. Sample of substance dried at \(100^\circ\) . . . . . . . . \(0.2635\) »
   Residue after ignition . . . . . . . . . . . . . . . . . . \(0.0915\) »

\[ \begin{array}{ccc} \text{Found:} && \text{Calculated for:}\\ 1 & 2 & C_{32}H_{22}N_6S_2O_6Ag_2;\; C_{32}H_{22}N_6S_2O_6Ag_2\cdot AgNO_3\\ Ag—33.4 & 33.72 & 24.95 \qquad 31.21\% \end{array} \]

Apparently the composition of the salt obtained corresponds to the formula:

\[ C_{32}H_{22}N_6S_2O_6Ag_2\cdot AgNO_3. \]

The somewhat larger amount of \(Ag\) found is explained by the fact that the starting preparation could not be completely freed from \(NaCl\), which was used for salting out the dye.

The experiment for obtaining the silver salt was modified further as follows: the precipitation of the dye by silver nitrate was not carried out at once with the whole quantity of the \(AgNO_3\) required for this purpose, but in portions, \(1/4\) of the total \(AgNO_3\) solution needed for the experiment. By this procedure it was hoped to purify the dye from the impurity \(NaCl\), if the latter was present in the dye.

The experiment was conducted as follows: to an aqueous hot solution of the dye was added the first \(1/4\) of the \(AgNO_3\) solution—the mixture was heated to boiling and filtered.—On the filter there remained a small residue (\(AgCl\)). To the hot filtrate was added the second \(1/4\) of the \(AgNO_3\) solution; the mixture was again filtered, etc., until the entire \(AgNO_3\) solution had been added.

From the hot mixtures, upon addition of the 2nd, 3rd, and 4th portions of \(AgNO_3\), no precipitate fell out, and after standing for twenty-four hours the mixture had to be evaporated; the precipitate from evaporation was washed with water by decantation and with hot alcohol on the filter.

Determination of silver gave the following results:

1. Weight of substance, dried at \(100^\circ\) . . . . . . . . . . . . . . . . . . . . \(0.1902\) g
   Residue after ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . \(0.0621\) "

2. Weight of substance, dried at \(100^\circ\) . . . . . . . . . . . . . . . . . . . . \(0.1626\) "
   Residue after ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . \(0.0506\) "

Found:

Calculated for the formula:

\[ \mathrm{C}_{32}\mathrm{H}_{22}\mathrm{N}_6\mathrm{S}_2\mathrm{O}_6\mathrm{Ag}_2 \cdot \mathrm{AgNO}_3 \]

\[ 31.27\%. \]

Finally, one preparation of the silver salt was varied further in the following manner. The dye solution was first precipitated with a small portion of \(\mathrm{AgNO}_3\), and then the heated mixture was filtered. On the next day the cold filtrate was completely precipitated with a solution of \(\mathrm{AgNO}_3\). The precipitate that fell immediately was washed and dried.

Determination of silver gave the following results:

Weight of substance, dried at \(100^\circ\) . . . . . . . . . . . . . . . . . . . . . \(0.1705\) g
Residue after ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . \(0.0521\) "

Found:

Calculated by the formula:

\[ \mathrm{C}_{32}\mathrm{H}_{22}\mathrm{N}_6\mathrm{S}_2\mathrm{O}_6\mathrm{Ag}_2 \cdot \mathrm{AgNO}_3 \]

\[ 31.27\%. \]

From the data of the analyses given above it follows that the silver precipitate of acid red congo apparently has the following composition:

\[ [(\mathrm{C}_{32}\mathrm{H}_{22}\mathrm{N}_6\mathrm{S}_2\mathrm{O}_6\mathrm{Ag}_2)\cdot\mathrm{Ag}]'|(\mathrm{NO}_3)'. \]

That is, at first, by exchange decomposition, the silver salt of the two sulfonic groups is formed, and then—a complex at the expense of one of the “azine” groups.

Phenosafranine.

The starting material for the experiments was the hydrochloride salt of phenosafranine:

\[ \mathrm{H}_2\mathrm{N}\cdot\mathrm{C}_6\mathrm{H}_3 \begin{matrix} & \mathrm{N} \\ \langle & \rangle \\ & \mathrm{N} \end{matrix} \mathrm{C}_6\mathrm{H}_3\cdot\mathrm{NH}_2 \]

\[ \mathrm{C}_6\mathrm{H}_5\mathrm{Cl} \]

Action of \(\mathrm{AgNO}_3\) on phenosafranine hydrochloride. 1.5 g of the phenosafranine salt was dissolved in alcohol with heating, and to the warm solution was added an aqueous solution of \(\mathrm{AgNO}_3\) (for 1 mole of dye, 2 moles of \(\mathrm{AgNO}_3\) were taken). A precipitate appeared at once. The filtered, washed, and dried precipitate was extracted with alcohol while heating. The substance extracted by alcohol separated from the hot solution in the form of a gelatinous precipitate, which under the microscope proved to be an interlacing of very thin and delicate hairs. On ignition the substance gave traces of silver, and the residue, recrystallized once more from alcohol after ignition, gave none. The solution of the diphenyldiamine in sulfuric acid gave a dark-blue coloration.

Apparently, it was acridinium phenosafranine, formed by a reaction of exchange decomposition:

\[ (\mathrm{C}_{18}\mathrm{H}_{15}\mathrm{N}_4)'\mathrm{Cl}' + \mathrm{Ag}'|(\mathrm{NO}_3)' = (\mathrm{C}_{18}\mathrm{H}_{15}\mathrm{N}_4)'|(\mathrm{NO}_3)' + \mathrm{AgCl}. \]

In addition to the nitric-acid phenosafranine, readily soluble in alcohol, in this reaction there are also obtained: 2) a substance containing silver, dissolving in alcohol with more difficulty than \(\mathrm{HNO}_3\)-safranine; 3) a residue not extracted by alcohol—\(\mathrm{AgCl}\).

Action of \(\mathrm{AgNO}_3\) on the free base of phenosafranine. To an alcoholic solution of the free base of phenosafranine was added an aqueous solution of \(\mathrm{AgNO}_3\). The precipitate that fell immediately was filtered off, washed, and dried. It was not possible to recrystallize it from alcohol, i.e. it turned out that on heating with alcohol it partially decomposes. Namely, on boiling with ...

with alcohol part of the substance dissolved and, upon cooling of the alcoholic solution, separated in the form of a gelatinous precipitate not containing silver; the residue, insoluble in alcohol, was a complex silver salt of safranin.

The determination of silver gave the following results:

Sample of substance . . . . . . . . . . . . . . . . . . . . . . . . . . 0.1819 g.
Residue after ignition . . . . . . . . . . . . . . . . . . . . . . . . 0.0523 „

Found:

\[ \mathrm{Ag} \;—\; 28.75 \]

Calculated according to the formula

\[ (\mathrm{C}_{18}\mathrm{H}_{14}\mathrm{N}_{4}\mathrm{Ag})'(\mathrm{NO}_{3})' \]

\[ 28.7\%. \]

The excess over the theory of silver is explained by decomposition of the complex salt during boiling with alcohol.

Thus, under the action of \(\mathrm{AgNO}_{3}\) on the free base of phenosafranin, apparently a complex salt of the composition is formed:

\[ [(\mathrm{C}_{18}\mathrm{H}_{14}\mathrm{N}_{4})\mathrm{Ag}]'(\mathrm{NO}_{3})'. \]

Chrysoidine \(\mathrm{H}_{5}\mathrm{C}_{6} — \mathrm{N}=\mathrm{N} — \mathrm{C}_{6}\mathrm{H}_{3}(\mathrm{NH}_{2})_{2}\).

Action of \(\mathrm{AgNO}_{3}\) on hydrochloric-acid chrysoidine. An aqueous solution of hydrochloric-acid chrysoidine was mixed with an aqueous solution of \(\mathrm{AgNO}_{3}\) (2 molecules of \(\mathrm{AgNO}_{3}\) were taken per 1 molecule of dye). A precipitate was obtained at once. After filtering off, washing, and drying, it was extracted with alcohol under reflux. From the hot filtrate there separated a gelatinous substance which, under the microscope, appeared as a delicate network of thin long hairs. From its outward appearance alone it was already clear that it was not homogeneous. Together with the gelatinous clot there was also a dense, dark substance which had settled on the bottom of the crystallizing dish. The substance [[unclear: damaged line]] in analyses gave about 7% [[unclear: continuation obscured]]; crystallized several times from alcohol, it contained no silver at all. Its m.p. was \(90^\circ—92^\circ\) C.; with a solution of diphenylamine in sulfuric acid it gives a blue coloration.

Obviously, this substance is nitric-acid chrysoidine, which is formed by a reaction of double decomposition:

\[ (\mathrm{C}_{12}\mathrm{H}_{13}\mathrm{N}_{4})'\mathrm{Cl}' + \mathrm{Ag}'(\mathrm{NO}_{3})' = (\mathrm{C}_{12}\mathrm{H}_{13}\mathrm{N}_{4})'(\mathrm{NO}_{3})' + \mathrm{AgCl}. \]

The experiment was also modified as follows: after precipitation of the HCl–chrysoidine with \(\mathrm{AgNO}_{3}\), the mixture was treated with \(\mathrm{NH}_{4}\mathrm{OH}\) (to dissolve \(\mathrm{AgCl}\)); the precipitate obtained was recrystallized from alcohol, and the same picture was obtained as in the first experiments: part of the precipitate dissolved in alcohol and easily separated from it in the form of a gelatinous clot; the other part of the precipitate dissolved in alcohol with more difficulty and, evidently, decomposed upon boiling. This latter contained silver.

Action of \(\mathrm{AgNO}_{3}\) on the free base of chrysoidine. An alcoholic solution of the base of chrysoidine was mixed with an aqueous solution of \(\mathrm{AgNO}_{3}\). The precipitate which separated at once was washed until traces of \(\mathrm{AgNO}_{3}\) had been removed, and was dried.

The determination of silver gave the following results:

1. Sample of dry substance . . . . . . . . . . . . . . . . . . . . . 0.1940 g.
   Residue from ignition . . . . . . . . . . . . . . . . . . . . . . . . 0.0553

Found:

\[ \mathrm{Ag} \;—\; 28.50 \]

Calculated according to the formula:

\[ \mathrm{C}_{12}\mathrm{H}_{12}\mathrm{N}_{4}\cdot \mathrm{AgNO}_{3} \]

\[ 28.5\%. \]

Consequently, chrysoidine, like the other dyes investigated, gives with \(\mathrm{AgNO}_{3}\) a complex salt of the composition:

\[ [(\mathrm{C}_{12}\mathrm{H}_{12}\mathrm{N}_{4})\mathrm{Ag}]'(\mathrm{NO}_{3})'. \]

Thus, the investigations we have carried out show that dyes containing in the molecule an “amino” group are capable of forming compounds with silver. The type of complexes is the same as for ammonia.

As is known, ammonia forms with silver the following complexes:

\[ (\mathrm{NH_3Ag})\cdot[(\mathrm{NH_3})_2\mathrm{Ag}] \quad \text{and others.} \]

Of these the most stable, firmly established complex is \([(\mathrm{NH_3})_2\mathrm{Ag}]\).

Of the dyes investigated, only one—auramine—gave a complex salt of this type, namely:

\[ [(\mathrm{C}_{17}\mathrm{H}_{21}\mathrm{N}_3)_2\mathrm{Ag}]\cdot(\mathrm{NO}_3). \]

The others form complexes in which one molecule of dye falls to one atom of silver. Such are the complex salts: \([(\mathrm{C}_{32}\mathrm{H}_{22}\mathrm{N}_6\mathrm{S}_2\mathrm{O}_6\mathrm{Ag}_2)\mathrm{Ag}]\cdot(\mathrm{NO}_3)\) (Congo red), \([(\mathrm{C}_{12}\mathrm{H}_{12}\mathrm{N}_4)\mathrm{Ag}]\cdot(\mathrm{NO}_3)\) (chrysoidine), \([(\mathrm{C}_{18}\mathrm{H}_{14}\mathrm{N}_4)\mathrm{Ag}]\cdot(\mathrm{NO}_3)\) (phenosafranine).

In the formation of the complex, as also in the formation of a stable salt, apparently one and the same ammonia residue takes part. Thus, for example, in the molecule of chrysoidine,

\[ \mathrm{C}_6\mathrm{H}_5-\mathrm{N}=\mathrm{N}-\mathrm{C}_6\mathrm{H}_3 \begin{matrix} \diagup \mathrm{NH}_2\\[-2pt] \diagdown \mathrm{NH}_2 \end{matrix} \]

there are two “amino” groups; only one participates in formation of the complex with silver:

\[ \mathrm{H}_5\mathrm{C}_6-\mathrm{N}=\mathrm{N}-\mathrm{C}_6\mathrm{H}_3 \begin{matrix} \diagup \mathrm{NH}_2\\[-2pt] \diagdown \mathrm{NH}_2 \end{matrix} +\mathrm{Ag}\cdot(\mathrm{NO}_3)' = \]

\[ = \left[ \mathrm{C}_6\mathrm{H}_5-\mathrm{N}=\mathrm{N}-\mathrm{C}_6\mathrm{H}_3 \begin{matrix} \diagup \mathrm{NH}_2\mathrm{Ag}\\[-2pt] \diagdown \mathrm{NH}_2 \end{matrix} \right]\cdot(\mathrm{NO}_3)'. \]

The same applies to auramine, which contains one “imino” group, and to Congo red, which contains two “amino” groups. As for phenosafranine, two possible structural formulas are ascribed to the so-called free base of this dye:

\[ \mathrm{HN}=\mathrm{C}_6\mathrm{H}_3 \begin{matrix} \diagup \mathrm{N}\\[-2pt] \diagdown \mathrm{N} \end{matrix} \mathrm{C}_6\mathrm{H}_3\cdot\mathrm{NH}_2 \quad \text{and} \quad \mathrm{H}_2\mathrm{N}-\mathrm{C}_6\mathrm{H}_3 \begin{matrix} \diagup \mathrm{N}\\[-2pt] \diagdown \mathrm{N} \end{matrix} \mathrm{C}_6\mathrm{H}_3\cdot\mathrm{NH}_2 \]

\[ \begin{matrix} & & \mathrm{C}_6\mathrm{H}_5\\ & & \mathrm{C}_6\mathrm{H}_5\mathrm{OH} \end{matrix} \]

According to the first formula, phenosafranine is a functioning imino-amine, and its complex salt with silver must have the composition:

\[ \left[ \mathrm{H}_2\mathrm{N}\cdot\mathrm{C}_6\mathrm{H}_3 \begin{matrix} \diagup \mathrm{N}\\[-2pt] \diagdown \mathrm{N} \end{matrix} \mathrm{C}_6\mathrm{H}_5=\mathrm{NHAg} \right]\cdot(\mathrm{NO}_3)'. \]

\[ \begin{matrix} & \mathrm{C}_6\mathrm{H}_5 \end{matrix} \]

According to the second formula, phenosafranine is the hydroxide of a quaternary ammonium base, and its complex salt should contain oxygen:

\[ \left[ \mathrm{NH}_2\cdot\mathrm{C}_6\mathrm{H}_3 \begin{matrix} \diagup \mathrm{N}\\[-2pt] \diagdown \mathrm{N} \end{matrix} \mathrm{C}_6\mathrm{H}_3\cdot\mathrm{NH}_2\mathrm{Ag} \right]\cdot(\mathrm{NO}_3)'. \]

\[ \begin{matrix} & \mathrm{C}_6\mathrm{H}_5\mathrm{OH} \end{matrix} \]

Our very unsatisfactory analysis of the silver salt does not make it possible to resolve this question. We hope to return once more to this subject and to investigate in detail the silver salt obtained by us.

Tomsk, November 10, 1923.

1. Chem. Centralbl. 1922 II, 1143. ↩