I. V. Gebler and V. G. Ignatyuk

A Rapid Method

for Determining Tannides in Bergenia

Tomsk
1928.

I. V. Gebler and V. G. Ignatiuk.

A Rapid Method for Determining Tannides in Bergenia

At the present time, an extremely large number of methods is known for the analysis of plant tanning materials (according to Grasser, over 73¹); of these the most widely used are the Löwenthal method, the American method, the Procter filtration method, and the shaking method, adopted by the International Union of Chemists as the official method. For the leather industry of the USSR, the VEM—All-Russian Unified Method—has developed detailed instructions that precisely set out all the details of work according to the official method².

The Löwenthal method is volumetric and is based on the assumption that, during oxidation, only tannides are destroyed. Potassium permanganate (chameleon) is used as the oxidizing agent. All the other methods listed above are based on the adsorption of tannide by hide powder, and work with them is carried out gravimetrically. Schematically, the course of analysis in cases where hide powder is used is as follows. From the tanning material to be investigated, an average sample is carefully selected, ground to powder in a special mill; the resulting grind is mixed, and from it a 15–20 gram portion is taken for analysis. The weighed portion of ground material is extracted first with cold and then with hot water, and in a part of the resulting aqueous extract the total quantity of tanning and non-tanning substances is determined gravimetrically. Another part of the extract is subjected to treatment with hide powder until the tanning substances are completely absorbed, and in the solution thus detannized the quantity of non-tanning substances is determined. The difference between the total quantity of extracted substances and the non-tanning substances, expressed in percent of the portion taken, gives the desired percentage content of tanning substances in the material tested.

The results of analyses obtained by these methods may, for production and practical purposes, be considered quite satisfactory, but they are all extremely troublesome, require for their execution a well-equipped laboratory and great care on the part of the chemist conducting the analysis. For this reason, the use of the indicated methods for evaluating plant tanning materials at the places where they are purchased and accepted is sometimes associated with considerable difficulties, and often is altogether impossible.

In such cases, the simpler method for evaluating plant tanning materials proposed by Professor Schroeder would be convenient. His method is based on determining the density (specific gravity) of aqueous extracts from the tanning materials under investigation³.

According to Schroeder, 50–100 grams of finely ground tanning material are taken and infused for 24 hours in a liter of water, having—

¹ Grasser, Manual for Chemical Laboratories of Leather Production.

² V. E. M. Moscow, 1925.

³ A. A. Davydov, A Simple Method for Determining Tanning Substances. Moscow, 1923.
A translation from the book: Dr. Schreede, Einfache Methode zur Bewertung der Gerbmaterialien. Freiberg i. Sa., 1922.

temperature of 15°C. The resulting solution is filtered to remove turbidity; the density of the juice is determined with an areometer or barkometer, and from tables, separate for each type of vegetable tanning material, the quantity of tanning substances corresponding to the observed density is found. Professor Schroeder divided each of the tanning materials into three categories according to its content of tanning substances: the poorest, medium, and best grades; determined in them the content of tanning substances by the gravimetric method, by the Loewenthal method, and by specific gravity, by cold infusion; and, comparing these data, compiled tables for certain tanning materials, such as oak bark, spruce bark, valonea, myrobalans, mimosa, divi-divi, algarobilla, and sumac.

Later Appelius and Merkesh compiled analogous tables for willow and birch bark.

The balatan industry now developing among us, and the associated procurement of large quantities of badan, create an urgent need for a simple method, readily accessible to everyone, for evaluating badan at procurement points. We believe that ease and speed in evaluating badan will be achieved to a considerable degree if, for determining the total quantity of extractive substances, the areometric method is used (as in Professor Schroeder), and, for determining tannides, the azo-coupling method proposed by Prof. I. V. Gebler.

For this purpose we have compiled the tables appended at the end of the article.

The basis for these tables was provided by data from 150 analyses carried out by us in 1927 for the Siberian Badan Bureau.

The analyses were conducted by the American method, with the shaking (of the extract with hide powder in determining non-tannides) replaced by filtration through Procter filters.

Leaving aside the results of this investigation, since in the present case this is not our aim, we cite only those data and observations that we used in compiling the tables.

A sufficiently large number of badan samples from various places where it grows (the samples came from expeditions working in the regions of Altai, the Sayan Mountains, the Yenisei, and Lake Baikal), and the data of investigations by Prof. I. V. Gebler¹), engineer Yakimov²), and engineer V. P. Markov³), make it possible to state that badan with a tanning-substance content below 13% and above 31% is a rare exception. These figures were adopted by us as the limiting, extreme values for the tables.

In the process of compiling the tables it was necessary to abandon Prof. Schroeder’s method and to replace the cold infusion proposed by him with extraction by water at a temperature of 90–100°, until complete extraction of the tannides from the weighed portion of roots was achieved.

The following considerations served as the basis for such a substitution. When an extract is obtained by cold infusion, chiefly readily soluble non-tannides and small quantities of tannides are leached out first, while the total amount of extract—

1) I. V. Gebler, “Some Data on Badan.” Separate offprint from Proceedings of the Tomsk Technological Institute, vol. 45, no. 3, 1924.
2) Eng. Yakimov, “The Technical Plant Badan.” Novosibirsk, 1927.
3) V. P. Markov, “Investigation of Tanning Plants of Central Altai.” Proceedings of T. T. I., vol. 43, no. 1.

…of the substances extracted is strongly dependent on such factors as the temperature at which the infusion is made, the degree of comminution of the root, and the frequency and intensity with which the infusion is stirred. The experiments carried out confirmed that, when these conditions are not observed, contradictory results are obtained. Thus, four infusions from one and the same average sample gave:

As is evident from the table presented, the total amount of extractive substances is subject to rather strong fluctuations, and only 30–40% of the extractive substances is extracted, while 70–60% of them remains in the root.

The specific gravities of the extracts, however, depend mainly on the total amount of substances present in the solution and, apparently, a change in the content of tannides in the solution has no noticeable effect on its specific gravity if the total amount of dissolved substances remains the same. The determinations of specific gravity made for each of 150 analyses confirm this assumption. As an illustration we give the specific gravities of three extracts and the analytical data for them, obtained by the American method.

Comparing the first and second extracts, we see that an insignificant difference in the content of extractive substances corresponds to an almost imperceptible change in specific gravity, despite the fact that the ratio between the contents of tannides and non-tannides is different in the two cases. Comparing the second and third extracts, we find that the tannide contents in them are almost equal, but nevertheless the specific gravity of the latter extract differs by seven units in the fourth decimal place.

As for the procedure for compiling the tables themselves, it should be said that the above-mentioned 150 determinations of the specific gravity of the extracts

were accepted by us as approximate data, while for compiling the tables themselves 10 samples of bergenia were selected with such a content of extractive substances that the extracts obtained from them made it possible to determine, for Table 1, ten points: 0.42, 0.48, 0.56, 0.60, 0.66, 0.72, 0.78, 0.84, 0.90, and 0.96. These table points were taken by us as the principal ones, while the remaining, intermediate points could have been found by interpolation; however, the availability of prepared extracts enabled us, without resorting to interpolation, to find all the intermediate points experimentally as well. For this purpose the corresponding extracts were diluted with water to the required strength.

To prepare the extracts, we took weighed portions of bergenia ground in an “Excelsior” mill, exactly corresponding to 30 grams of dry substance; they were infused in 600 cu. cm. of distilled water for 24 hours at room temperature and were then subjected to extraction in a Procter extractor at a temperature of 90–100°C. Extraction was continued until the entire quantity of tannides contained in the weighed portions had been completely removed. The test for completeness of extraction was the reaction with gelatin.

The addition of water during extraction was regulated so that the volume of the extract obtained was somewhat less than a liter. After completion of extraction, the extract, heated until complete clarification (to 80–90°C), in order to dissolve the phlobaphenes that had precipitated, was filtered from the turbidity suspended in it, cooled to 17.5°C, and made up with water in a volumetric flask exactly to one liter.

For compiling Table No. 1, the strength of the extract was determined with a carefully checked areometer, and then, for each extract, a check analysis was carried out by the American method. The method of compiling Table No. 2 for the quantitative determination of tannides in bergenia by the azo-combination method is given in the chapter “Determination of the Quantity of Tannides.”

TECHNIQUE FOR THE EVALUATION OF BERGENIA

All work on the evaluation of bergenia by means of the proposed tables may be subdivided into the following operations, easily accessible to everyone:

1. Selection of an average sample.
2. Determination of moisture.
3. Preparation of the extract.
4. Determination of the density of the extract.
5. Determination of the quantity of tannides.

1. Selection of an average sample.

The proper execution of this operation is of indisputably great importance. Only with a careful and conscientious attitude toward the selection of the average sample can one give a guarantee that the analytical data correspond to the actual average content of substances in the bergenia being evaluated. Thus, for example, in Grasser we find: “The proper selection of a sample of tanning materials requires precise knowledge of the tannide content in the individual parts of the tanning material. It often happens that different chemists, having received for analysis several samples of one and the same tanning material, obtain—

different results in their analysis, and the cause of this phenomenon may lie not in the inaccuracy of the research methods applied, but in the improper taking of samples for analysis.”¹

He further gives figures illustrating the diversity in the composition of individual samples of one and the same tanning material, which may occur within a single lot.

For bergenia such data are not yet available in the literature, but on the basis of the observations we have made it can already be asserted that bergenia, in the present case as well, is no exception, and that the diversity in the composition of its individual samples is just as great as for other vegetable tanning materials.

Thus, for example, samples of bergenia that had been poorly dried (dark-brown fracture) showed only from 9 to 14% tannides, while samples covered with mold showed from 15 to 18%, whereas normal samples taken from the same average sample contained from 23 to 25% tannides. In addition, it was observed that the leaves always contain more tannides than the roots; in the petioles, by contrast, the tannide content falls considerably.

All this must, of course, be taken into account when selecting an average sample.

To avoid improper selection of the sample, the following standards have been established for this purpose.¹

I. For root material in bundles.

A minimum of 3% of the total number of bundles is selected; from the middle of the bundles, by means of a saw or a sharp axe (and preferably on a root-cutter), pieces 2–4 cm long are cut from the middle. All the pieces are collected and mixed. If the sample obtained is too large, it is thoroughly mixed and an average sample is taken, which in no case should weigh less than 1 kilogram.

II. Ground root and small tanning materials.

A minimum of 5% of the total number of sacks is selected; they are emptied in layers onto a smooth, clean floor into one pile, and then thoroughly mixed. From the pile, perpendicular to its surface and penetrating to the floor through all its contents, samples are taken in several places and thoroughly mixed. If the sample obtained is too large, it is passed through a mill for tanning materials in order to obtain a homogeneous grind, and a minimum of 0.5 kilogram is taken as the average sample.

Grinding large quantities of bergenia, especially when it has been poorly dried, is usually associated with a great expenditure of time. To avoid this unproductive expenditure of time, one may proceed as follows. From each root taken into the average sample, chop off with an axe (or cut on a root-cutter) pieces 2–3 centimeters long, so that in total they amount to not less than 0.5 kg; then such a quantity of bergenia can be ground without difficulty.

For grinding bergenia one may recommend the “Excelsior” mill, shown in Figure No. 1, and a percussion mill (Fig. 2). For taking the average sample it is convenient to use a manual root-cutter (Fig. 3).

Having finished selecting the average sample and grinding it, the ground bergenia should be thoroughly mixed, and 50–100 grams should be weighed out, as accurately as possible, on hand apothecary scales with horn pans at least 12 cm in diameter, for the determination of moisture.

The remaining quantity of the ground sample, in order to protect it from drying out, should be poured into a jar with a tightly closing stopper.

2. Determination of Moisture.

The use, under the conditions of receiving stations, of the usual method for determining moisture by drying a definite weighed portion of the substance to constant weight in a drying oven at a temperature of 100–150°C, besides being tedious, presents a number of other difficulties. These include: (a) the necessity of frequent weighings on accurate chemical balances in order to catch the end of drying, and (b) the necessity of strict regulation of the temperature within the limits of 100–105°C, since at a lower temperature there will be no guarantee of the complete removal of all moisture from the weighed portion, and slow drying will lead to oxidation of the tannides in the weighed portion of bergenia, which, taken together, will distort the figure obtained for moisture in the direction of diminution; while raising the temperature above 105°C will cause the removal not only of moisture but also of some volatile substances contained in bergenia. In the latter case the moisture determined will be considerably higher than the actual value.

All these conditions, because of one circumstance or another, cannot always be fulfilled at procurement stations. More convenient, rapid, and simple in such cases is the well-known Hoffmann method, which has wide application in the practice of determining moisture in grain. We carried out experiments in determining moisture in bergenia by the Hoffmann method, obtaining the following results:

From the examples given it is evident that the maximum discrepancy of the results of determinations by these methods does not exceed 0.13%; consequently, the Hoffmann method may be considered applicable and entirely reliable for determining moisture in bergenia. The essence of the Hoffmann method, as is known, is as follows. A weighed portion of the substance under examination is heated to 200°C with the aid of a liquid that does not mix with water. The water is distilled off from the substance, collected in a receiver, and measured. To transfer heat to the weighed portion, lubricating oil is used, and to facilitate the distillation of the water, a cinchona–toluene mixture is used.

Determination of moisture in the Hoffmann apparatus (Fig. 4) is carried out as follows¹. Into cube “A,” standing on a tripod, are poured 200 cu. cm of oil and 10 cu. cm of a turpentine–toluene mixture (19 parts turpentine and 1 part toluene); 50–100 grams of the ground material under examination are added; the contents of the cube are shaken; into the neck of the cube is inserted the distillation, elbow-shaped tube “B,” to which the reflux condenser “C” is screwed, and to it is attached the receiver “D,” which is a glass graduated tube with a pear-shaped widening at the top. A thermometer “E” is inserted into the lid of the cube so that its mercury bulb is immersed in the liquid in the cube. Into the funnel “F,” inserted in the lid of the cube, 50 cu. cm of the turpentine–toluene mixture are poured (the tap “G” being closed). The cube is then heated with a gas burner or a kerosene Primus stove to 180°C, and this temperature is maintained for 4–5 minutes, after which tap “G” is opened, the liquid is quickly run out of the funnel, and the tap is closed again. The cube is heated as rapidly as possible to 200°C, and, after removing the burner, the cube is allowed to cool to 180°C. The receiver is removed and rotated about its long axis between the fingers, first in one direction and then in the other, to facilitate the sliding of water droplets from the glass, and the boundary between the turpentine–toluene layer and the water that has settled beneath it is read off.

Since the oil and turpentine dissolve a certain amount of water, 0.2 cu. cm is added to the reading as a correction. If, for example, the reading gives 16.2 cu. cm of water, then, consequently, the badan contains 16.4 cu. cm of it; and if 100 grams of ground badan were taken, this will amount to 16.4% of the weighed portion.

After each experiment, the boiler, distillation tube, and receiver are carefully cleaned and rinsed with toluene.

The entire determination of moisture by Hoffmann’s method takes 16–20 minutes, whereas the ordinary method for air-dry badan requires about two hours, and for raw badan even more.

3. Preparation of the Extract.

The proposed table is calculated for 30 grams of absolutely dry badan; consequently, for preparing an extract of moist or air-dry badan, a somewhat larger quantity will be required, which, on the basis of the moisture already determined, can be found with sufficient accuracy. Suppose that the badan tested showed 10% moisture; then, if 30 grams of this badan are taken for extraction, it will turn out that the absolutely dry substance in it will be only

\[ 30 - \frac{30 \times 10}{100} = 27 \text{ grams.} \]

According to the table, however, 30 grams of absolutely dry substance are required. When the missing 3 g are added, there will be introduced also

\[ \frac{3 \times 10}{100} = 0.3 \text{ gram of moisture.} \]

Consequently, for the extract it is necessary to take \(30 + 3 + 0.3 = 33.3\) grams of badan with a moisture content equal to 10%.

Having thus calculated the quantity of moist badan corresponding to 30 grams of dry substance, a weighed portion is taken on hand apothecary’s scales from the same grinding from which, earlier, before the determination of moisture, it was

poured into the bath and soaked in 400 cu. cm of water at room temperature. The soaking should be carried out in a laboratory glass beaker with a capacity of not less than 600–700 cu. cm. Soaking lasts 12 hours. The water for soaking and for subsequent extraction must be distilled, and only in extreme cases may soft river or rain water be used, but it must first be boiled and filtered through cotton wool.

After 12 hours have elapsed, the infusion obtained and the entire contents of the beaker in which the infusion was made are transferred to Procter’s extractor.

A. Yu. 11 October 1928.

and extraction is begun at a temperature of 90–100°C.

Procter’s extractor (Fig. 5) consists of an iron or copper bath “A,” in the sockets of which ordinary laboratory beakers “B” (glass, or preferably porcelain) are inserted. In each of the beakers there is a bent siphon tube “C,” on the short arm of which is attached a porcelain funnel “D,” filled with glass wool and plugged with tow. The glass wool prevents fine particles of badan from entering the extract being siphoned off, while the tow retains the glass wool. On the long arm of the siphon tube is fitted a rubber tube with a pinchcock “E,” ending, for convenience of draining, in a glass tip. The siphon tubes are set in the required position by clamps “F,” fastened to stands. One of the stands is shown in the figure. Liter volumetric flasks “G” serve as receivers for the extract.

For heating the bath to 100°, a single “Primus” kerosene stove is quite sufficient.

Work with the extractor is carried out as follows. Having filled funnel “D” with glass wool and plugged it with tow, it is lowered to the bottom of the beaker and fastened with a clamp; then the infusion prepared by the method described above is introduced into beaker “B” and heated to 80°C. It is kept at this temperature for not less than an hour, the resulting hot extract is drained off by means of the siphon tube, and, closing clamp “E,” 200–300 cu. cm of distilled water is again poured into beaker “B.” It is heated to boiling and, after being kept for not less than an hour, drained into the volumetric flask. This operation is repeated until complete extraction of the entire amount of tannides contained in the weighed sample is achieved, with the individual portions of fresh water regulated so that, at the end of extraction, the quantity of extract obtained does not exceed one liter; otherwise it will be necessary to evaporate the solution in order to bring its volume to one liter. When adding water, it is sufficient to adhere to the above-stated norm of 200–300 cu. cm for each separate portion.

At the end of extraction the liquid flowing through the siphon tube becomes completely transparent and only weakly colored (rarely colorless); but these external signs are insufficient for judging the completion of extraction, and a test for completeness of extraction is always necessary. This test is performed as follows. From the siphon tube, 3–5 cu. cm of liquid are collected in a test tube, and then 3–4 drops of a 10% gelatin solution are carefully added. If a white, voluminous precipitate appears, this indicates that the extraction is not yet complete. An excess of gelatin dissolves the precipitate formed; therefore, to carry out the reaction one should use a dropper and add the gelatin solution carefully, drop by drop. In order to become proficient in this…

with the reaction and to acquire the skill of recognizing the precipitate of tannides with gelatin, it is recommended to carry out this reaction by taking, for the experiment, 2–3 cubic cm of solution from the volumetric flask, where the solution is of higher concentration and will give a clearer reaction.

The gelatin solution, in a closed jar, may be kept for a long time without spoiling; it is therefore convenient to prepare it at once in sufficient quantity. For this purpose, 100 grams of pure table salt and 10 grams of gelatin are dissolved in a liter of hot water, and the resulting clear solution is poured for storage into a jar with a tightly closing stopper. When stored in an open vessel, mold quickly appears in the solution, owing to the fact that gelatin serves as a good substrate for various kinds of fungi. After the extraction is completed, the extracts collected in the volumetric flasks usually contain a considerable amount of turbidity that has passed through the glass wool. To remove it, the extract is heated to 80–90 degrees (to dissolve the phlobaphenes precipitated on cooling—the oxidation products of tannides) and filtered through filter paper. To reduce the time spent on filtration, the paper may be replaced by hygroscopic cotton wool. For this purpose, the necks of the funnels should be plugged with as small an amount of cotton wool as possible. The filtered extract is cooled by immersion in cold water to a temperature of 16–17°C, and, after making up to the mark on the neck of the flask with pure water, one proceeds to the final operation—the determination of the density of the extract.

4. Determination of the density of the extract.

For this operation it is necessary to have a glass cylinder with a capacity of not less than one liter, a thermometer, a wire copper stirrer, and an areometer (Fig. 6). Fig. 7 separately shows the Baumé areometer, on the neck of which divisions from 0 to 2 degrees Baumé have been marked. The scale of this areometer is divided in such a way that it makes it possible to take readings with an accuracy of up to 0.02° Baumé; and by estimating by eye the distances between the small divisions of the scale, readings can also be obtained with an accuracy of up to one hundredth of a degree Baumé. The bold numerals on the left, 0, 1, and 2, indicate whole degrees, while the numerals on the right—0.00, 10, 20, etc.—correspond to tenths of a degree Baumé. Each tenth of a degree is divided into 5 parts, so that one division corresponds to two hundredths of a degree.

Since the readings of the areometer are highly dependent on the temperature of the extract, it is always necessary to know the temperature for which the areometer is calibrated and to adhere strictly to this temperature. This temperature is in most cases indicated on the neck of the areometer. Usually, areometers are graduated for temperatures of 15 and 17.5°C and are supplied with the corresponding marks either on the neck or on the body of the areometer itself. In compiling the table we used an areometer calculated for 17.5°C; consequently, when determining the density of the extract with the aid of our table, it is necessary to ensure that the temperature of the extract is exactly 17.5°C. For this purpose, after placing the thermometer in the flask with the extract, it is immersed for several minutes in cold or hot water, shaken, and the required temperature is obtained. In the case of small deviations of the temperature from 17.5°C, there is no need to spend time cooling or heating the extract; instead, by making the measurement at any temperature, one may then introduce temperature corrections into the resulting reading for the density of the extract according to the table appended below.

To determine the density of the extract, proceed as follows. Pour the extract from the volumetric flask into a glass cylinder, stir it with a stirrer, remove the foam and air bubbles that have formed on the surface of the extract; then, holding the hydrometer by the end of the neck, carefully immerse it in the cylinder and, after waiting for the moment when the hydrometer comes to equilibrium, take the reading on the scale. Owing to the adhesion of the liquid to the glass, around the neck of the hydrometer there forms a small mound of liquid, called the meniscus. The reading should be taken along the upper edge of this meniscus, and the eye of the person taking the reading must be at the same horizontal level as the upper edge of the meniscus, as shown by the arrow in Fig. 6.

After taking the reading, record the temperature of the extract and the hydrometer reading. If the temperature of the extract was exactly 17.5°C and a hydrometer with the mark on the neck “17.5°C” was used, then one immediately turns to Table No. 1 and, in the column “Density of the extract,” finds the corresponding content of extractive substances in the weighed sample of badan.

If, however, the temperature of the extract differs from 17.5°C, it is first necessary to introduce a temperature correction according to the correction tables, and to find the probable content of substances for the corrected density of the extract; otherwise a gross error is unavoidable.

Temperature-correction table for a hydrometer calculated
at 17.5°C.

The use of this table of corrections may be explained by an example. Suppose that the temperature of the extract proved to be equal to 16°C, while on the neck of the areometer there is the mark “17.5°C.” The reading of the areometer was then 0.88°Bé; then, before turning to Table No. 1 for determining the total amount of extractive substances, we introduce a correction: according to the table of corrections it is equal to 0.03; then \(0.88 + 0.03 = 0.85^\circ\) Bé, and this will be the actual density of the extract.

In exactly the same way, when working with an areometer whose scale is calculated for 15°C, there is no need for any other corrections, and one may use directly the table for determining extractive substances if the temperature of the extract is exactly 15°C. In cases of deviation from this temperature, it is necessary either to bring the temperature of the extract to 15°C by heating or cooling, or else to introduce into the calculation a temperature correction according to a table analogous to the one given above.

Table of temperature corrections for an areometer calculated for 15°C.¹)

The manner of using this table is exactly the same as for the preceding one.

5. Determination of the quantity of tannides.

For this last and most important operation in the evaluation of bergenia, we propose to make use of the azo-coupling method, proposed by Prof. I. V. Gebler and developed in the diploma work of the student of Tomsk State University Maslov²). The essence of this method

¹) The table is borrowed from the book: Davydov, A “Simple” Method for Determining Tanning Substances. Moscow, 1923, p. 14.

²) Maslov. Diploma work on the topic: “On the Methodology of the Quantitative Determination of Tannides.” Reported at the Tomsk Section of the Russian Physico-Chemical Society.

is based on the circumstance that tannides, owing to their phenol-like structure, have the ability to enter into azo-coupling reactions with diazotized amines, thereby forming insoluble oxyazo dyes. In this process the entire tannide is titrated by a quite definite quantity of diazo solution, and, from the number of cubic centimeters of diazonium solution consumed, it becomes possible to judge with a sufficient degree of accuracy the content of tannides in the extract, if one first establishes experimentally the necessary quantity of diazonium solution of a definite titer required for precipitating, say, one percent of tannides from a certain volume of extract, the tannide content of which has been determined by any of the known methods. Such experiments were carried out by student Maslov in performing the work mentioned above; it turned out that, besides sufficient reliability (the deviations in determinations by the azo-coupling method and by Procter’s method do not exceed 0.4%), this method deserves attention also for the relative simplicity of the technique of its execution and, what is especially important, for the negligibly small expenditure of time. If a ready titrated diazonium solution is available, the entire determination takes 5–10 minutes, whereas the time required for determining tannides by the methods of Procter, V.E.M., and others is measured in hours. Maslov’s work was carried out in parallel with our investigation of bergenia for the assignment of the Siberian Bergenia Bureau, which made it possible for Maslov, using ready extracts analyzed by us by the American method, to perform more than 80 experiments and to establish a certain coefficient, relative to the American method, which, when the weight quantity of diazo compound consumed for titrating a definite volume of extract is divided by it, gives the desired tannide content in bergenia, corresponding to the tannide content that would have been obtained in analyzing the extract by the American method. The value of this coefficient was found to be 0.00596.

Knowing that this coefficient was obtained when titrating an extract with a tannide content of 22% (by the American method) with a diazonium solution whose titer is 0.0144, and that 9.2 cc of diazo solution were consumed per 25 cc of extract, it is easy to calculate an analogous coefficient for a diazo solution of any titer.

For ease of use of the proposed method, and to eliminate possible errors in calculations, we have compiled a ready table for a \(1/20\) N solution of para-nitrophenyldiazonium chloride. Using this table, without any calculations, one can at once, from the number of cubic centimeters of diazo solution consumed in titrating 20 cc of extract, find the tannide content in bergenia. (When using this table it must be borne in mind that it has been compiled for extracts from 30 grams of absolutely dry bergenia; consequently, for extracts from quantities of bergenia greater or less than 30 grams of absolutely dry substance it is, of course, unsuitable.)

In the mass production of determinations, which is inevitable at procurement stations, it is always necessary to have in advance a sufficient quantity of prepared, titrated diazo solution.

As is known, most diazo solutions are unstable and more or less rapidly undergo decomposition under the influence of such factors as direct light, etc. To eliminate this undesirable phenomenon, experiments were carried out to increase the stability of diazo solutions...

titrated para-nitroaniline. For this purpose, the directions found in the German patents were used: 85387, 28/X—1894; 89437, 21/I—1896, and 28874, 24/X—1893. These patents propose the preparation of stable diazo compounds in the form of so-called azophores by evaporating diazo solutions to dryness in the presence of a small excess of mineral acid and of some salt: sulfate ($\mathrm{Na_2SO_4}$), zinc chloride ($\mathrm{ZnCl_2}$), or aluminum sulfate ($\mathrm{Al_2(SO_4)_3}$).

Modification and combination of the conditions indicated in the above-cited patents made it possible to establish that a solution of diazotized para-nitroaniline, when sulfate and a certain excess of hydrochloric acid are added to it, can be kept in a dark place for 10–15 days without appreciable change in titer, and only after this interval does it become necessary to check the titer of the diazo solution against β-naphthol. The amounts of sulfate and hydrochloric acid necessary for imparting stability to the diazo solution will be given below, together with directions for preparing the titrated solution of para-nitroaniline.

The technique for determining tannides in badan by the azo-coupling method:

a) preparation of the titrated solution of para-nitrophenyl diazonium chloride, and

b) titration of the extract with a solution of para-nitrophenyl diazonium chloride.

Preparation of the titrated solution of para-nitrophenyl diazonium chloride.

The table is calculated for a $1/20$ N diazo solution. To prepare a liter of such a solution one must have:

(The excess of hydrochloric acid and sulfate is necessary, as indicated above, to impart stability to the diazo solution.) The method of preparing the diazo solution is as follows. Weigh exactly 6.9 g of para-nitroaniline, pour it into a small glass beaker, add 20 cu. cm of hydrochloric acid, and warm slightly on an alcohol lamp for complete dissolution of the weighed portion of para-nitroaniline. Pour the resulting solution onto finely crushed ice placed in a beaker (preferably porcelain) of not less than 600–700 cu. cm capacity and, while stirring with a glass rod, cool the solution to 4–5°C. After cooling, add to the solution (with the precipitate of the hydrochloride salt of para-nitroaniline that has formed) 4½ grams of dry sodium nitrite ($\mathrm{NaNO_2}$). The addition should be carried out quickly, with vigorous stirring, and the temperature must not be allowed to rise above 5°C. After the reaction is complete, as indicated by the clarification of the solution and the disappearance of the precipitate, add the remaining 20 cu. cm of hydrochloric acid and 28 g of sulfate; after the latter has dissolved, filter the solution from turbidity and unreacted para-nitroaniline. Transfer the filtered solution to a one-liter volumetric flask and make up with distilled water to the mark on the neck of the flask.

The solution should be kept in a vessel of dark glass, in a place protected from direct light, since light is one of the important factors causing the decomposition of diazo compounds.

The solution of para-nitrophenyldiazonium chloride prepared in this way, owing to the insufficient purity of the starting products, inaccuracies in the course of the work, and incompleteness of the diazotization reaction, will always contain a weight amount of diazo compound different from that theoretically calculated for a \(1/20\) N solution. It is therefore necessary to establish the titer of the solution, i.e., to determine the weight amount of para-nitrophenyldiazonium chloride actually present in one cubic centimeter of the prepared solution.

The titer of the diazo solution is established against chemically pure \(\beta\)-naphthol with a melting point of \(122^\circ\)C. Weigh exactly 7.2 g of \(\beta\)-naphthol, dissolve it in 300 cc of distilled water to which 6 g of caustic soda has been added, and make up with water in a one-liter volumetric flask to the mark. If the \(\beta\)-naphthol is accurately weighed and the water is measured at the temperature indicated on the flask, the titer of the \(\beta\)-naphthol solution prepared in this way may be taken as equal to 0.0072, i.e., one cubic centimeter of the solution contains 0.0072 g of \(\beta\)-naphthol.

To establish the titer of the diazo solution, proceed as follows. Take with a pipette 20 cc of the \(\beta\)-naphthol solution, dilute it in a beaker to 300–400 cc with distilled water, and acidify with dilute acetic acid in order to neutralize the free alkali. The acidification must be carried out carefully, drop by drop, so that the last, excess drop causes the formation of turbidity. Then 3 g of sodium acetate are added to the solution, and the diazo solution is carefully run in from a burette until all the \(\beta\)-naphthol contained in the 20 cc of solution taken has been converted into the azo dye.

The end of the reaction is recognized by the “run-off test.” To perform the test, a drop of the solution from the beaker, together with the precipitate of azo dye suspended in it, is taken with a glass rod and placed on filter paper. As this drop spreads over the paper, it forms in the middle a red spot of insoluble dye, around which a sufficiently broad halo is visible—the run-off of a colorless or faintly colored solution. Now take, with a clean glass rod, a drop of an alkaline solution of salt R (the sodium salt of 3,6-disulfo acid of \(\beta\)-naphthol) of arbitrary concentration and place it on the filter paper near the first run-off. As it spreads, this drop will reach the first run-off, and if there was an excess of diazonium there, the salt, reacting with it, will give a red streak at the place where the run-offs merge; this indicates that the solution has been overtitrated, i.e., that an excess of diazo solution has been added from the burette. If, however, no red streak is obtained, then there is consequently an excess of \(\beta\)-naphthol in the solution. To make certain of this, take a drop of fresh diazo solution (only not from the burette) and place it near the first run-off; if, at the place where the first and the new run-offs merge, coloration appears, then the excess of \(\beta\)-naphthol is evident and the titration is not finished. If even in this case there is no colored streak, then the titration may be considered complete.

By repeated performance of the run-off test during the second half of the titration, it is possible to determine precisely the end of the azo-coupling reaction between \(\beta\)-naphthol and para-nitrophenyldiazonium chloride.

Having observed the end of the titration, one reads the burette scale and records the number of cubic centimeters of diazo solution used for titrating 20 cubic centimeters of the β-naphthol solution. The reading is made with an accuracy to tenths of a cubic centimeter.

Let us suppose that 22.05 cubic cm of diazo solution were used for titrating 20 cubic cm of the β-naphthol solution. Then the titer of the diazo solution may be found by the following reasoning.

The titer of the β-naphthol solution, as indicated above, is equal to 1β-naphthol = 0.0072, and therefore 20 cubic cm of it contain: \(0.0072 \times 20 = 0.144\) g of β-naphthol.

The interaction between β-naphthol and chloro-para-nitrophenyldiazonium is expressed by the equation:

\[ \begin{array}{c} \text{[[structural formula of β-naphthol]]} + \text{[[structural formula of chloro-para-nitrophenyldiazonium]]} = \mathrm{HCl} + \text{[[structural formula of azo dye]]} \end{array} \]

which shows that one molecule of chloro-para-nitrophenyldiazonium is consumed per one molecule of β-naphthol.

The molecular weight of β-naphthol is 144.00, and the molecular weight of chloro-para-nitrophenyldiazonium is 185.492. Knowing the molecular weights and the reaction equation, one may set up the following proportion:

\[ \frac{185.492}{144} = \frac{x}{0.144}, \quad \text{whence} \]

\[ x = \frac{185.492 \times 0.144}{144} = 0.18547\ \text{g}. \]

Here “\(X\)” is the weight quantity of chloro-para-nitrophenyldiazonium required to convert into an azo dye 0.144 g of β-naphthol. As stated above, 22.05 cubic cm of diazo solution were used for the titration; therefore one cubic cm of our diazo solution contains:

\[ \frac{0.18547}{22.05} = 0.00925\ \text{g of chloro-para-nitrophenyldiazonium.} \]

Consequently, the titer of this diazo solution \(= 0.00925\).

With exact observance of all the standards and conditions indicated above in preparing the solutions and carrying out the titration, there is no need to repeat all the calculations given by us; one may use the ready value 0.18547, which indicates the weight quantity of chloro-para-nitrophenyldiazonium required to convert into a dye 0.144 gram of β-naphthol contained in 20 cubic cm of solution, and, by dividing this value by the number of cubic cm of diazo solution used for the titration, obtain the required titer of the diazo solution. Thus, for determining the titer of the diazo solution one may use the formula:

\[ T_{\text{diazo}}=\frac{0.18547}{a}, \]
where

“\(a\)” is the burette reading, the number of cubic centimeters of diazo solution used for titrating 20 cubic centimeters of a \(\beta\)-naphthol solution of titer 0.0072.

Having thus found the titer of the diazo solution, it is written on the label of the bottle in which the solution is kept, and one proceeds to determine the tannides in the extract by titration.

Titration of the extract with a solution of chlorinated para-nitrophenyl-diazonium.
For this purpose one should take, with a pipette, 25 cubic centimeters of thoroughly mixed extract into a beaker of 400–500 cubic centimeters capacity, dilute with water to 100–150 cubic centimeters, add 3 g of sodium acetate, and titrate with a solution of titrated chlorinated para-nitrophenyl-diazonium until a test on the extract gives no positive results. The test on the extract is carried out exactly as in determining the titer of the diazo solution, and differs from it only in that, at the point where the extracts meet, not a red but an orange-brown coloration appears.

Having established the end of the titration, the number of cubic centimeters of diazo solution expended on titrating the extract is recorded, and the percentage content of tannides in badan is found from the table. It must, however, be borne in mind that the table is calculated for a \(1/20\) normal solution of chlorinated para-nitrophenyl-diazonium having a titer of 0.00927, and if the titer of the diazo solution used for titrating the extract differs from this value in the third or fourth decimal place, then before consulting the table it is necessary to introduce a correction into the reading.

Let us suppose that, for titrating 25 cubic centimeters of extract, 20.5 cubic centimeters of diazo solution of titer 0.00900 were expended, whereas a \(1/20 N\) diazo solution has \(T=0.00927\). Then, in order to reduce the reading to a \(1/20 N\) solution, the following proportion may be set up:

\[ \frac{x}{20.5}\times\frac{0.00900}{0.00927}, \]
whence

\[ x=\frac{0.00900}{0.00927}\times 20.5=0.970\times 20.5=19.885\ \text{cubic cm}. \]

Here “\(X\)” is the number of cubic centimeters of \(1/20\) diazo solution corresponding to 20.5 cubic centimeters of a solution of titer 0.00900.

Consequently, in the table it is necessary to look up the percentage content of tannides corresponding not to 20.5 cubic centimeters, but to 19.885. The nearest figure in the column “number of cubic centimeters of diazo solution used for titration” will be 20.0 cubic centimeters, and for the present example the sought content of tannides in badan according to the table is equal to 15.56%. In the proportion analyzed by us, the ratio

\[ \frac{0.00900}{0.00927}=0.970 \]

for one and the same titer is a constant value and may serve as a coefficient for reducing the reading in titration to a \(1/20 N\) diazo solution.

Thus, in cases where the established titer of the diazo solution differs from the theoretical \(1/20 N\) titer by more than 5 units in the fourth decimal place in the direction of decrease, the correction coefficient (we shall denote it by the letter “\(K\)”) should be taken as the ratio:

\[ K=\frac{T}{0.00927}, \]

where “\(T\)” is the titer of the diazo solution determined by \(\beta\)-naphthol; and each time, before consulting the table, the titration reading should be multiplied by this coefficient “\(K\).”

Table No. 1.

for determining the total amount of extractive substances in bergenia by the areometric method.

Table No. 2

for the quantitative determination of tannides in bergenia by the azo-coupling method.

From one liter of extract prepared from 30 grams of bergenia, take exactly 25 cu. cm with a pipette and titrate with a \(1/20\) N solution of diazotized para-nitroaniline.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Specific gravity meter.

Figure 7.

1. “Commodity Science,” ed. by Nikitinsky and Petrov, Moscow, 1923, vol. IV, p. 61. ↩↩↩

2. Grasser, Manual for Laboratories of Leather Production. Moscow, p. 337. ↩