JULY 15, 1936

-4NALYTICAL EDITION

tilling p ridine which had been dried by refluxing over barium

oxide. %he 20 per cent reagent is preferred for sugars, whereas
either reagent may be used with the other compounds analyzed in
the present work.
PROCEDURE.
Exactly 2 cc. of reagent are pipetted into a
clean dry test tube containing an accurately weighed portion of
the substance under analysis, care being taken to wash down with
reagent any particles adhering to the side. The weight of the
sample varies with the hydroxyl content of the compound-for
example, 50 mg. of glucose are sufficient. The mixture is carefully heated over an open flame until the solution boils and for 1
minute longer. After cooling it to room temperature, the solution is diluted with 5 cc. of carbon dioxide-free water and transferred t o a small flask; three washin 8 of the tube are made,
two with 10-cc. portions of water, andone with 10 cc. of ethyl
alcohol. The acid is titrated with 0.1 N sodium hydroxide to the
cresolphthalein end point. It is necessary, of course, to run
blanks for the reagent alone.
In the case of substances containing appreciable proportions
of fatty acids which may form difficultly decomposable mixed
anhydrides with the reagent it is desirable, after the dilution of
the reaction mixture with 5 cc. of carbon dioxide-free water,
to boil the solution for 1 minute before finally cooling and titrating. Furthermore, in the analysis of lipoidal substances the final
titration is carried out in the presence of sufficient ethyl alcohol to
provide a concentration of approximately 50 per cent after the
addition of the alkali. Since only one-half of the titratable acid is
available for reaction as acetic anhydride, it is desirable, in order
to ensure an excess of the anhydride, that three-quarters of the
original titratable acidity remain after the completion of the
acetylation reaction.
I n the analysis of the compounds listed in the second part of
Table I the aqueous solution of the reactants was boiled,
then cooled, and alcohol was added to the flask before titration. Practically theoretical results were obtained for all
compounds listed except the sugar alcohols, the results for

279

which were about 2.5 per cent low, and the single tertiary
alcohol, which did not react. The results for pregnandiol,
obtained by the method described, were low; allowing the reaction mixture, after the preliminary heating, to stand
overnight (13 hours) a t room temperature the results obtained were much closer to the theoretical value. The sample
of triolein analyzed was obviously impure, having a deep red
color. The method of recording the results tends to emphasize slight errors.

Summary
1. The acetylation method of Verley and Bolsing has been
modified to make possible the analysis of small samples of organic compounds.
2. The new method is rapid and convenient and does not
require the use of condensers or ground-glass stoppers. It
has been applied successfully to representative organic compounds.
3. Tertiary alcohols cannot be estimated by this method.

Acknowledgment
The writers are indebted to G. F. Marrian of this department for the samples of oestrone and pregnandiol used in the
investigation.

Literature Cited
(1) Peterson,V. L., and West, E. S., J . Bid. Chem., 74,379 (1927).
(2) Verley, A., and Bolsing, F., Bey., 34,3354 (1901).
(3) West, E. S., Hoagland, C. L., and Curtis, G. H., J. Biol. Cham.,
104, 627 (1934).
RECEIWD
February 13, 1936.

Estimation and Identification of the Glucoside

Salicin
MORRIS B. JACOBS, Department of Health, 125 Worth St., New York, N. Y.
NICHOLAS T. FARINACCI, 336 West Twelfth St., New York, N. Y.

A qualitative and quantitative method

for the determination of salicin, a glucoside, is presented. The quantitative
method depends on the production of a
polymerized cleavage product, obtained by
acid hydrolysis, which is estimated gravimetrically. A qualitative test which is outlined depends on the solubility of the polymer in a solution of sodium hydroxide with
the production of a violet color, and is
modified so as to be used as an approximate
quantitative colorimetric method. Another qualitative method based on a
coupling reaction with p-diazobenzenesulfonic acid anhydride is detailed.

HE identification of salicin and its estimation have been

studied over a number of years, but the methods obtained by these studies have been long and not easily adapted
for accurate work. The importance of glucosides as a class
necessitates the development of short and accurate means for

their estimation. The methods in use a t present rely on the

hydrolysis of salicin either by enzymatic action (7) or by the
action of acids ( I S ) and the subsequent estimation of the glucose part of the molecule. None of the methods available
are dependent upon the specific chemical nature of salicin,
itself, or of the cleavage product, saligenin.
It is well known that when salicin (saligenin-/3-glucoside)
is hydrolyzed by enzymes or acids, the products are glucose
and saligenin. These reactions are noted in Beilstein (S),
where it is also noted that saligenin polymerizes on boiling
with dilute sulfuric acid to saliretin.
* Much work was done by Piria ( 1 4 , who noted that both salicin
and saligenin boiled with dilute hydrochloric acid yielded saliretin. He concluded that saliretin prepared from salicin did not
have a definite composition, but prepared from saligenin its formula was GHsO.
Kraut (10) took powdered salicin and ten parts of hydrochloric
acid (sp. gr. 1.125), warmed until dissolved, heated to 80' C.,
and noted the yellowish red color of the resultant saliretin.
He concluded that it was saligenosaligenin, HOCeH&H*.O.CeHcCH20H. On the other hand, von Gerhardt ( 6 ) and Beilstein and Seelheim (8) found higher polymers. Kraut also noted
that saliretin was soluble in alkalies.
Voswinkel (16)did not obtain the same product from salicin
plus hydrochloric acid as with sulfuric acid.
Wischo (17) treated salicin with nitric acid, producing picric
acid. He boiled an aqueous solution of salicin with dilute hydrochloric acid, forming saliretin, dissolved the saliretin in potas-

280

INDUSTRIAL AND ENGINEERING CHEMISTRY

sium hydroxide, and added this solution to the picric acid solution, producing picramic acid and a bluish red color.
Dott (5) gives the following test to identify salicin: To an
aqueous solution of the substance to be investigated add twotenths its volume of hydrochloric acid and warm. The solution
develops an aromatic odor and precipitates saliretin. Extract
this with ethyl ether and evaporate to dryness. If salicin is
present the amorphous residue will be red.
Jackson and Dehn (8) present in tabular form various reactions
for the identification not only of salicin but of many other glucosides.
Ware (16) notes that salicin gives a color when warmed with
resorcinol and phosphoric or sulfuric acid.

It occurred to the authors that salicin, which has the nature

of a phenolic ether, should under the proper conditions couple
( I , 11, 12) with a diazonium compound. Salicin does couple
with p-diazobenzenesulfonic acid anhydride, but only with
an alkaline solution prepared from the dry compound.

mean is 0.006. The method may be applied in the presence

of small amounts of lactose, sucrose, inulin, maltose, arabic,
tragacanth, karaya, agar, Irish moss, locust kernel, ghatti,
starch, and the glucoside amygdalin. These substances do
not interfere, for they do not produce a precipitate on evaporation with concentrated hydrochloric acid. Diluted normal
urine also does not interfere. On the other hand, the glucosides digitonin and saponin do interfere, for they yield an
acid-insoluble precipitate when treated as directed. These
precipitates do not have a red color and are insoluble in 10
per cent sodium hydroxide solution. B y means of the method
given below, salicin may be determined in the presence of
these glucosides.
TABLE
I. CONVERSION
OF SALICIN
TO SALIRETIN
A:id
Precipitate

Salicin

Qualitative Tests
TEST1. Dissolve 15 mg. of p-diazobenzenesulfonic acid anhydride (4) in 2 cc. of 10 per cent sodium hydroxide solution.
Add this to 5 cc. of salicin solution, mix, and place in a warm
water bath (approximately 80' C.) for 1 minute. At the end of
this time, if salicin is present, a deep red color will develop.
Upon the addition of acid the color changes to orange. A very
deep red color will be produced by 1 mg. of salicin per cc., and
will also develop in the cold, in the case of 10 mg. per cc. at the
end of 30 minutes, and in the case of 1mg. per cc. at the end of 2
hours. It is best to run a blank with the diazonium reagent, in
order to be sure that the colors are not due to decomposition of
the reagent itself.

This method for the identification of salicin is very simple

except for the preparation and keeping qualities of the
diazonium reagent. However, at best, it can only be indicative of the presence of this glucoside because of the numerous
other substances which will couple with this reagent under
exactly the same conditions. The color produced by the
reaction between salicin and the diazonium reagent is apparently proportional to the concentration of the former, for
the depth of color increases with the concentration of salicin.
TEST2. Proceed as directed under quantitative colorimetric
method, evaporating almost to dryness. A red precipitate soluble
in alkali with the production of a violet color, which on dilution
becomes salmon-colored, shows the presence of salicin. This color
will be produced with as little as 0.2 mg. per cc.

Quantitative Gravimetric Method

The regularity of the changes produced in salicin during
evaporation with concentrated hydrochloric acid led to the
belief that the production of the precipitate was entirely
quantitative.
Place 25 to 50 cc. of the solution to be analyzed containing 1 to
5 mg. of salicin per cc. in a tall-form 100-cc. beaker and add 25
cc. of concentrated hydrochloric acid. Cover with a watch glass
on glass hooks and evaporate very slowly on a hot plate carefully
regulated to 80' C. or less, down to a volume of not more than 10
cc. Prepare a Gooch crucible or a porous glass crucible with a
thin pad of acid-washed asbestos in the usual manner and dry in a
constant-temperature oven to constant weight at 100' C. Cool
in a desiccator and weigh. Filter the precipitate obtained from
the salicin through the Gooch crucible. Carefully transfer all the
precipitate to the crucible by means of a rubber policeman and
wash bottle. Wash thoroughly with distilled water, suck dry,
place in a constant-temperature oven at 100" C., and dry to constant weight. Multiply the weight of the precipitate by the factor 2.524 to obtain the quantity of salicin present in the original
aliquot taken for analysis.
Table I gives the results of this method on known quantities of recrystallized salicin.
Acid hydrolysis of salicin conducted as outlined yields results, based on an average recovery of saliretin, which are at
least 98.5 per cent accurate. The average deviation from the

VOL. 8 , NO. 4

MO./d5-60

39
41
50
50
50
62.5
62.6
75
75
100
100
125
125
250
260

CC.

Ratio of
Column 2 to
Column 1

Mo.
15.5
16.0
19.0
19.8
20.0
24.2
24.6
29.2
29.3
39.6
39.8
52.9
49.7
101.2
100.6

0.397
0.390
0.380
0.396
0.400
0.387
0.394
0.389
0.391
0.396
0.398
0.423
0.398
0.405
0.402
Mean 0.396

Assuming that salicin splits quantitatively, yielding saligenin and glucose

C1aHisO.l

+ Ha0 +CEHI~OE
+ OHCBH~CH~OH

and that saligenin goes completely to saliretin

2GHsO2 +Cl4H1408

+ HzO

then two moles of salicin are equivalent to one mole of saliretin-that is, 572 grams of salicin are equivalent to 230 grams
of saliretin. The theoretical yield, if the above holds true,
would be 230/572 or 40.21 per cent of salicin used. Table I
demonstrates that we recover, expressing the ratio of
columns 2 to 1 in percentage, 39.6 per cent of the salicin
used, or 98.5 per cent of the theoretical yield. These results
indicate that only one polymer of saligenin is formed when
conditions of acid hydrolysis are rigidly controlled. Kraut
( I O ) , it may be noted, felt that the production of saliretin is
quantitative. The factor used in the method is obtained from
the reciprocal of the average recovery of saliretin and will, of
course, yield the quantity of salicin sought.

Quantitative Colorimetric Method

Place 5 cc. of a solution of salicin containing ap roximately 1
to 2 mg. per cc. in a 10- to 15-00. beaker with a mar[ at 2 cc., add
5 cc. of concentrated hydrochloric acid, and evaporate slowly on a
hot plate kept at 80" C. or less to 2 cc. Allow to cool and filter
through a small filter. Wash both beaker and filter well with cold
water, and replace the receiver of wash water with a tube having a
graduation at 10 cc. or with a 10-cc. volumetric flask. Wash
beaker and filter paper with 4 successive I-cc. portions of 10 per
cent sodium hydroxide, wash filter once more with 1cc. of 10 per
cent sodium hydroxide, and again wash beaker and filter faper
with four successive 1-cc. portions of distilled water. Ma e receiver tube or volumetric flask up to volume, mix thoroughly, and
read in a colorimeter against standard control treated in exactly
the same way. The ratio of fourteen determinations comparing
unknown to standard had a minimum of 0.925, a maximum of
1.17, and a mean of 1.02 as the variation from unity.

It is obvious that this method has not been completely

developed to yield high accuracy; however, the results are of
the correct order and are better than some quoted in the

JULY 15, 1936

ANALYTICAL EDITION

literature (9). Using standard of concentration close to that

of unknown would aid in obtaining higher accuracy. This
method is far more rapid than others and where interfering
substances are present in the gravimetric method, it is better
than none. It is a t present being studied for better development.
As the hydrolysis of salicin proceeds, the following changes
take place: Shortly after the solution reaches a temperature
of 80 C., a white turbidity is noticeable. This turbidity
changes slowly in an hour or so to a pink precipitate which
grows deeper in color as the time of heating and evaporation
proceeds. For quantitative conversion, it is essential to
evaporate very slowly and a t a temperature never higher than
80 c.
The length of time necessary for accurate results indicates
that the formation of saliretin proceeds with a definite
velocity. The rate of this reaction has never been studied.
The concentration of hydrochloric acid is of great importance
for, when it falls below a certain minimum, the reaction does
not yield quantitative results within a reasonable time.
This work was Derformed in the Chemical Laboratorv of
the Bureau of Food and Drugs, Department of Health,
City of New York.

281

Literature Cited
Anwers, K., and Michaelis, F., Ber., 47,1275 (1914).
Beilstein, F.,and Seelheim, F., Ann., 117,83 (1861).
Beilstein, 3rd ed., Vol. 111, p. 608 (1897); Vol. 11, p. 1109
(1896); Vol. 11,supplement, p. 680 (1903).
Cumming, W. M., Hopper, I. V., and Wheeler, T. S., Systematic Organic Chemistry, 2nd ed., p. 441, New York,
D. Van Nostrand Co., 1931.
Dott, H. B., Pharm. J.,119, 691 (1927).
Gerhardt, von, Ann. Chim., [317,215 (1843).
Hudson, C. S., and Paine, H. S., J. Am. Chem. Soc., 31, 1242
(1909).
Jackson, K. E., and Dehn, W. M., IXD.ENO.CHEM.,Anal. Ed.,
6,382 (1934).
Kerstan, G., Planta (Abt. E., 2. wiss. Biol.), 21,657 (1934).
Kraut, K.,
Ann., 156,123 (1870).
Meyer, K. H.,Irschick, A., and Schlosser, H., Ber., 47, 1741
(1914).
Meyer, K. H., and Lenhardt, S., Ann., 398,66 (1913).
Moelwyn-Hughes, E. A., Trans. Faraday SOC.,24,309 (1928).
Piria, R.,Ann., 56,35 (1845).
Voswinkel, A., Chem. Zentr., (1900) I, 771.
Ware, A. H.,Chemist and Druggist, 120,74 (1934).
Wischo, P.M. F., Pharm. Post, 28.42 (1895).
RECEIVED
February 4, 1936.

Battery-Type Stand Assembly for Distilling Equipment

*

ROY L. MOBLEY, P. 0. Box 1021, Baton Rouge, La.

H E efficiency of distillation has in many ways been

lowered by the equipment available. I n the older types
of apparatus vapors and liquids were brought in contact with
metals, and with cork and rubber joints, stoppers often leaked,
and in some instances the units making up the apparatus
were inefficient.
The advent of all-glass interchangeable-joint apparatus and
improved condensers has brought about much greater
efficiency and more accurate results. However, the complete
change of type and general form of equipment has made
older accessories obsolete, and it is extremely time- and accessory-consuming to set up two or more of the newer types a t a
time.
I n an attempt to overcome these handicaps, the author
made the necessary calculations and arrangements, diagrammatic sketches, etc., and developed a working model of a
battery-type stand assembly which has proved very satis-

factory during the past 3 years. The photographs show front

and diagonal end views of the assembly as built for 500-cc.
distilling flasks and indicate arrangement of the parts.
The apparatus was built and used in the laboratory of
William L. Owen, Baton Rouge, La.
RBCEIVED
May 14,1936.

CORRECTION.
In the article on ILDeterminationof Nitric Oxide
in Coke-Oven Gas [IND.
ENG.CHEM.,Anal. Ed., 8, 164 (1936)]
an error was made in Figure 2. The glass balls should be of 6or 6-mm. diameter, as the 8-mm. balls specified would be entirely
too large for this apparatus.
J. A. SHAW