UNIVERSITY OF CALIFORNIA.*
CHEMICAL REAGENTS FOR PURITY.
DR. C. KRAUCH,
CHEMIST TO THE FIRM OF E. MERCK, DARMSTADT.
Authorised Translation of the Third Edition
J. A. WILLIAMSON, F.C.5., AND L. W. DUFRE,
WITH ADDITIONS AND EMENDATIONS BY THE AUTHOR.
/AACLAREN & SONS, 37 AND 38, SHOE LANE,
HPHE necessity of obtaining the highest degree of purity in
those reagents which are most largely used in chemical
analysis and research, demands a means of speedily recognising
the impurities most often encountered. This means of recog-
nition, together with much general information regarding the
various reagents, is afforded in a very comprehensive form by
Dr. Krauch's valuable work, which we present for the first time
in English, the translation being that of the third German
edition. We have practically adhered to the general arrange-
ment and treatment adopted by the author.
We would particularly call the attention of the reader to the
fact that, in most instances, reference has been made to English
abstracts from the original communications. We have also
endeavoured to enhance the value of the book by substituting
for the numerous references to German text books on chemical
analysis the corresponding English works.
We have to acknowledge the valuable aid given by the author
himself in supplying us with additions and corrections, the result
of his experience since the third German edition was published.
The Latin titles of the reagents are those of the Pharma-
copcea Germanica and have been retained, as they may be of
service for the recognition of certain chemicals in price lists.
In conducting the various tests, the author emphasises rigid
adherence to the degree of concentration, the amount of reagent
used and the manner in which the test is to be carried out.
Temperatures are, in all cases, given in degrees Centigrade.
Finally, our thanks are due to those friends who have assisted
us in revising the proofs of this translation.
LONDON, October, 1902.
li UNIVERSITY }
Acid, acetic, cone, puriss. (C 2 H 4 2 ). Mol. Wt., 59'86. Clear,
colourless liquid, with a sour pungent smell, solidifying about
10C. Sp. gr. T064, containing about 96 per cent, acetic acid.
Note. A weaker acid (about 90 per cent.) shows a sp. gr. of 1*071.
Tests for Impurities.
Residue. On volatilisation of 10 gms. of the acid, no weigh -
able residue must remain.
Note, When larger quantities of the pure strong acid are volatilised,
traces of a partially combustible organic residue remain ; there ought
not, however, to be more than one milligram of residue in 50 c.o.
* Heavy Metals and Earths. On diluting 10 gms. to 100 c.c.,
adding excess of ammonia, and subsequently ammonium sulphide
and ammonium oxalate, no change should take place even after
standing for some time, and slightly warming. On diluting
20 gms. to 100 c.c. and adding freshly prepared H 2 S water, no
brown colour should appear.
Sulphuric Acid. 10 gms. in 100 c.c. of water are heated to
boiling, and barium chloride added ; after standing for several
hours no reaction should take place.
Hydrochloric Acid. On treating 5 gms. in 50 c.c. of water with
nitric acid and silver nitrate, no cloudiness should appear.
Empyreumatic Substances. On adding 50 c.c. ^5 permanganate
solution to 5 gms. of the acid diluted with 15 c.c. water, the pink
colour should remain after standing 15 minutes.
A few grams are diluted with water, phenolphthalein added,
and the acid titrated with normal alkali (free from carbonate).
1 c.c. normal alkali = 0'05986 gms. C 2 H 4 2 . The purity of the
acid may be gauged from the temperature at which it assumes
* In this and other parts of the work, the author has used for the sake
of brevity the term " heavy metals." The metals indicated are at once
recognised from the description of the test (e.g., addition of ammonium
sulphide), thus saving elaborate detail.
the glacial state, the stronger the acid the higher the temperature
at which it solidifies. An acid containing 0'5 per cent, of water
becomes glacial at 15'65, with 2'91 per cent, at 11'95, and with
G'54 per cent, at 7'1. When the acid contains 49*4 per cent,
of water, the solidifying point is reduced to - 19*8. The specific
gravity affords an easy means of ascertaining the strength.
SPECIFIC GKAVITY OF ACETIC ACID AT 15 (OUDEMANS).
Note. The specific gravities above 1*0553 in each case represent
two liquids of very different strength. In order to ascertain whether
the acid exceeds the maximum density (78 per cent.) or the reserve, it
suffices to add a little water ; in the case of a stronger acid the specific
gravity increases, and decreases if the acid is weaker.
Uses and Storage.
For analytical purposes acetic acid is chiefly used as a solvent
and neutralising agent. For use as a reagent the puriss. con-
centrated acid is diluted with three times its weight of water.
For some purposes the strong acid is used e.g., Goldenberg's
Diethod for the estimation of tartaric acid (J.S.c'.L, 1888, p. 349)
ACETIC ANHYDRIDE. 7
and Adamkiewicz's test for albumen (J.C.S., 1875, pp. 172, 919).
It is advisable to use a glacial acetic acid of known strength in
making up the dilute acid (1| per cent., or even less) employed in
microscopy. The acid must be kept in well-stoppered glass
bottles, and as the glacial acid solidifies in the winter time if kept
in a cold room, in order to avoid accident it is necessary to thaw
it very carefully, particularly if stored in large bottles.
In addition to the above acid, generally used for analytical
work, the following kinds of acetic acid are placed on the
market : (1) Glacial acetic acid, a very concentrated acid (about
99 per cent.), dissolving oil of lemons in all proportions, and
standing all the tests for purity, with the exception of that for
empyreumatic substances ; it is stronger than the puriss. (2)
Glacial acetic acid, a less concentrated acid, dissolving oil of
Note. The solubility of certain essential oils in acetic acid presents
a rough method of determining the strength. Oil of lemons only
dissolves in an acid containing not more than 2 per cent, of water ; oil
of cloves will dissolve in acids containing up to 10 per cent.
Acid acetic, pure (1*06), acid acetic, puriss. (1'06), acid acetic,
dilut., pure (1'04), and acid acetic, dilut., puriss. (1'04), contain
respectively 50 per cent. (1'06) and 30 per cent. (1*04) of the strong
acid ; those marked puriss. stand all the tests prescribed, but the
pure acids generally contain traces of empyreumatic substances.
Acid, acetic, anhydric. ((C 2 H 3 0) 2 0). Mol. Wt, 10176.
Colourless liquid, boiling at 138, and smelling of acetic acid.
Sp. gr., 1*0799 at 15'2.
When acetic anhydride is poured into water it at first sinks to
the bottom of the vessel, and then slowly combines with the water
to form acetic acid. It must be carefully preserved in well-
stoppered vessels. The solution (1 : 50) must give no reaction on
adding nitric acid and silver nitrate, and its strength is
estimated volumetrically with normal alkali. 1 c.c. normal
alkali = 0'05088 gm. acetic anhydride. It is not often used in
analysis, but is employed, inter alia, in the estimation of lanoline
(B. Fischer, die neueren Arzneimittel, 1889, <p. 79), and in the
acetine method for the estimation of glycerine (Sutton, Yol. Anal.,
7th Ed., p. 365).
Hirschsohn, in determining the essential oils of the coniferse,
employed a so-called acetic acid reagent consisting of 10 gms.
acetic anhydride and 5 drops pure concentrated sulphuric acid
(Pharm. Ztg., 1891, p. 725).
Acetonum puriss. (C 3 H 6 0). Mol. Wt., 57'87.
A clear, colourless, mobile liquid, having a smell resembling
peppermint. Sp. gr., 0797 0798 ; B.p., 56 57.
Tests for Impurities.
Volatile Matter. 30 gms. must leave no residue on evaporation.
Solubility and General Appearance. Pure acetone is perfectly
clear and colourless, and miscible with water in all proportions,
without showing any turbidity.
Action of Permanganate. On the addition of one drop of potas-
sium permanganate solution (1 : 1000) to 10 c.c. of the acetone, the
pink colour must not be completely discharged at the end of 15
minutes. This test serves to detect the presence of aldehyde. An
acetone which contains 0'5 per cent, by volume of aldehyde dis-
charges the colour in five minutes, and 0*25 in 10 minutes. Free
mineral acid must not be present.
Water. Schweitzer and Lungwitz (Chem.-Ztg., 1895, p. 1384 ;
J. S.C.I., 1895, p. 1069) detect the presence of water in acetone by
adding to 50 c.c. of the sample 50 c.c. of petroleum spirit (b.p.
40 60). In the event of water being present, two layers are
formed, which does not occur when pure acetone is similarly
treated. It is a remarkable fact that if 2 c.c. of water are added
to 50 c.c. of chemically pure acetone, instead of the lower layer
occupying a volume of 2 c.c. it is observed to occupy 5 to 7 c.c.
Pure acetone must, therefore, show no alteration on shaking with
Boiling Point (see above).' Concerning the estimation of the
b.p. and sp. gr. of acetone see Geehmugden, Zeit. f. anal. Chem.,
1896, pp. 503, 504 ; also J.C.S., 1896, pt. 2, p. 679.
Note. It is a well-known fact that in boiling-point determinations
varying results are obtained 1 depending on the material and nature of
the vessels employed for that purpose. To obtain concordant results
the author 'always uses a small copper still, as suggested by B>annow in
ALCOHOL, AMYL. 9
his Report on the Commission on Analytical Methods for Fractional
Distillation, published in the Chomische Industrie, 1886, page 328.
Acidity. Acetone must not redden blue litmus paper.
Kramer's iodoform method may be employed (conversion into
iodoform by means of excess of iodine solution, in presence of a
solution of sodium carbonate). On the quantitative estimation of
acetone, see also under " Methyl Alcohol." A new method of esti-
mation of acetone with phenylhydrazine has been proposed by
Strache, and is said to be better than the iodoform method (Zeit.
f. anal. Chem., 1892, p. 573, et seq., and J.S.C.L, 1893, p. 185).
Note. The above tests are mentioned in part by Guttmann
(Dingler's Polytechn. Journal, 1894, pp 96 111) as being chiefly
used in the- manufacture of smokeless powders, where a specially pure
a,oatone is demanded. Concerning this test see Klar, J.S.C.L, 1897,
p. 722, et seq.
Uses and Storage.
Acetone is sometimes used as a solvent in chemical investiga-
tions. It should be kept in well-stoppered bottles in a cool place,
as it is easily inflammable.
In addition to pure acetone, the quality used for technical
purposes (manufacture of smokeless powders) can also be obtained
in a very pure state in the market. Guttmann, loc. cit., found it
to have a sp. gr. of 0'7965, and 98 per cent, distilled over between
56*2 and 56'4. Good commercial qualities showed a perman-
ganate test of several minutes, and he found the acidity of those
samples which were perfectly soluble to be not higher than
0*00225 per cent. (See Klar, loc. cit.)
Alcohol amylic.pur. Amyl alcohol (C 5 H (2 0). Mol. Wt, 87*81,
B.p., 131132 ; Sp. gr., 0'814. Clear, colourless liquid, without
action on litmus paper.
Note. Crude amyl alcohol, or fusel oil, is easily distinguished from
the pure by reason of its having a very different specific gravity and
10 ALCOHOL, ETHYL.
Tests for Impurities.
Volatile Matter. On evaporating 10 gms. on the water bath no
residue must remain.
Furfurol. 5 c.c. amyl alcohol, mixed with 5 c.c. concentrated
sulphuric acid, must show only a slight yellow or reddish colour.
Boiling Point, etc. See above.
Note. The ordinary commercial Siamples of amyl alcohol are coloured
brown to blackish-brown by sulphuric acid. Amylic alcohol, which
gives a colourless liquid on mixing with sulphuric acid, can only be
obtained according to L. v. Udranzky (see that author on the furfurol
reaction, J.S.C.I., 188"9, p. 309) by repeated and troublesome treat-
ment of the commercial preparation with concentrated sulphuric acid,
etc. Perfectly pure amyl alcohol can only be obtained by the decom-
position of pure potassium amyl sulphate. For most analytical pur-
poses, a carefully-rectified preparation, with the correct boiling point
and specific gravity, and conforming with the above tests, is quite good
enough, but there are certain processes in which it is necessary to have
a preparation giving no furfurol reaction.
Estimation of the boiling point is sufficient.
Amyl alcohol is used in forensic analysis for the extraction of
alkaloids, and is also employed in the analysis of foods and drugs.
See also H. Droop Richmond's u Dairy Chemistry " regarding its
use in milk testing.
Fusel oils of different degrees of purity are found on the
market. Ordinary commercial fusel oils often contain only 30
per cent, of pure amyl alcohol. The Chemiker-Zeitung, 1889, p.
1062, reports a commercial sample containing pyridine. (See also
J.S.C.I., 1889, p. 734.)
Alcohol absolut, pur. (C 2 H 6 0). Mol. Wt., 45'90. Clear, colour-
less liquid. The sp. gr. at 15'5 is 0'796, equal to 99'6 per cent,
by weight of alcohol. Alcohol mixes with water without produc-
ing any turbidity, and does not affect litmus paper. It must be
ALCOHOL, ETHYL. 11
free from smell, other than that peculiar to alcohol. B.p. of the
100 per cent, alcohol, 78'4C.
Note. On the sp. gr. of pure alcohol and its mixtures with water,
see a paper by Squibb (Zeit. f. anal. Chem., 1887, p. 94; also alcohol
tables in Allen's Comm. Organ. Anal., Vol. I., p. 95, et seq.).
According to that author, the present figures for the specific gravity
are supposed to be too high. It has been pointed out by him that
the last traces of water are very difficult to eliminate. Alcohol is
Pictet assumes that alcohol puriss. purified by crystallisation, is
perfectly pure and contains 100 per cent.
Tests for Impurities.
Residue. 50 gms. must leave no residue on slow evaporation.
Odour, Colour, etc. (Fusel Oil). 10 c.c. of alcohol are mixed
with 30 c.c. of water (good spring water is to be preferred) in a
slightly conical glass flask, and this mixture is tested at once for
smell, taste, colour, and transparency. It must not show any
colour, and no smell other than that peculiar to ethyl alcohol, and
must have only a slightly burning taste (neutral).
Note. An important part of the examination for the purity of an
alcohol is the test regarding its odour. The method of testing as
mentioned above, is very satisfactory (see report of the annual
meeting of the Association of Swiss Analytical Cheimd<sts, in Chem.-
Ztg., 1893, No. 84). It as very easy in that way for tho'Se with
some experience to distinguish between inferior and good alcohol.
Many technical men te<&>b for fusel oil by rubbing a few drops between
the hands. It may be remarked here that the strength of the so-called
fusel oil smell is not dependent on the quantity of the fusel oil present,
but on the presence of certain odorous parts of the same.
For quantitative estimation of fusel oil, Rose's method (described
inter alia, in Bockmann, C'hem.-techn. Untersuchungsmethoden,
Springer, Berlin, 1893 ; J.S.C.I., 1886, pp. 393, 498), is the best
and the mo>st generally adopted, and is used with various modifications.
For the quantitative estimation of small traces of fusel oil in rectified
spirit and absolute alcohol by Rose's separation method the modifica-
tions introduced by Glasenapp may be advantageously employed. The
method is fully described in Zeit. f. angew. Chem., 1895, No. 22,
p. 657, et seq., and J.S.C.I., 1896, p. 140. The presence of a large
quantity of fusel oil in commercial spirits of wine can be detected at
once by the turbidity formed on diluting with water. Moreover, if, on
dilution with a large quantity of water, drops should separate out, it
must be observed, according to Borntrager (J.S.C.I., 1889, p. 64),
whether a smell of acrolein (acetal) is evolved on addition of an equal
12 ALCOHOL, ETHYL.
quantity of concentrated sulphuric acid and caustic potash solution,
and if, on adding 3 drops of concentrated hydrochloric acid and 10
drops of colourless aniline oil, a fine raspberry colour appears (amyl
Aldehyde, etc. (a) 10 c.c. of alcohol are put into a small test
tube, and 0*5 c.c. of water, and 1 c.c. of a freshly-prepared 10 per
cent, aqueous solution of meta-phenylene-diamine hydrochloride
are added. After one hour's standing there should be scarcely
any coloration. (See also note.)
(b) 10 c.c. of absolute alcohol are mixed with 1 c.c. of water, and
5 drops of silver nitrate solution added; this solution must
become neither turbid nor coloured on heating.
Note. The test with meta,-phenylene-diamine hydrochloride has
been proposed by Windisch, and is also used in the Swiss Excise
Department as an important test for impurities peculiar to the first
running's of the distillation, and for aceitaldehyde (J.S.C.I., 1887,
p. 388). The test is based on the fact that. meta-phenyl'ener-diamine
hydrochloride immediately colours, an alcoholic solution containing
aldehyde yellow to yellowish-red; then a fine greenish fluorescence
appears, which /slowly darkens. On comparing the colour reaction
so produced with the reactions of solutions containing known quantities
of aldehyde the test can be made approximately a quantitative one.
The spirits to be tested are made up at- the above-mentioned Excise
laboratory to 95 per cent, by volume, and it is required', for example,
that the rectified spirit contain not more than 0*1 part by volume of
aldehyde per 1,000 of alcohol. On long standing even pure absolute
alcohol shows a yellow colour, in consequence of formation of aldehyde
by oxidation with the air. The meta-phenylene-diamine used for testing
must be pure (see " Meta-phenylenet-diamine").
The test with silver nitrate solution is required by the German
Pharmacopoeia. Aldehyde must also be tested for by means of
ammoniacal silver solution; or with ammoniacal potassium perman-
ganate solution (see Zeit. f. anal. Chem., 1891, p. 208, and Deutscne
Medicinische Woehenschrift, 1893, p. 941) ; further, by means of the
fuchsine reaction of Gayon (J.S.C'.L, 1888, p. 238; also Allen's
Comm. Organ, Anal., Vol. I., p. 159), or with Nessler's reagent.
Caustic potash is also used as & test for aldehyde, a yellow colour
being produced. It must also be remarked that spirits which have
been stored in casks made of wood, containing tannin, show a colour
with caustic potash. For the production of an alcohol free from alde-
hyde see Paul, Zeit, f. anal. Chem., 1896, p. 648.
Foreign Organic Impurities in General. On mixing 10 c.c. of
absolute alcohol with 1 c.c. of water and 1 c.c. of potassium per-
ALCOHOL, ETHYL. 13
manganate solution, the red liquid must not become yellow before
20 minutes have elapsed.
N te. It must be remarked regarding this test, that a partial re-
duction of the permanganate takes place after 20 minutes, even with
the purest alcohol.
Cazeneuve (J.S.C.I., 1889, p. 924), observed even after five minutes
a somewhat yellowish-pink coloration. Simoe the ye>air 1882 he has
taken for comparison! chemically pure alcohol of 93 and <a potassium
permanganate solution of 1 in 1,000 ; 10 c.c. of the pure alcohol require,
according to Cazeneuve, five minutes at the ordinary temperature to
give >a rather yellowish pink colour with 1 c.c. of the permanganate' solu-
tion, which shows that the reduction is not quite complete. If an
alcohol of 93 shows a quicker reduction, this points unmistak-
ably to the presence of impurities. The Pharm. Ztg. (1889, p. 481)
points out that the test with permanganate not only shows the presence
of fusel oils, but of every trace of organic matter which originate with
the barrels and cork bungs, and is unavoidably present in every alco-
hol. A slight reduction therefore gives no occasion for rejection.
Lang (Annual meeting of the Association of Swiss Analytical
Chemists, Chein.-Ztg., 1893, No. 84) obtained the following results
with the permanganate test. The action of the permanganate test is
dependent on the temperature and strength of the spirits, on the more
or less complete elimination of the first runnings (these first runnings
have the strongest action on the permanganate), on the time of storing,
on the material of the storage vessels (samples of spirit which have
been stored for some time in wooden barrels must be distilled before
testing with permanganate), on the action of sunlight and various
other factors. It is necessary to use caution, and only the follow-
ing general conclusions can be drawn : inferior spirits show a short
oxidation test with permanganate:, and the better the sample the
longer the duration of the test.
Besides the test with permanganate there is also the sulphuric acid
test : On mixing with equal pairts of sulphuric acid, no yellow colora-
tion must take place this is used to detect fusel oils, aldehydes, and
other organic compounds; but according to Glasenapp (J.S.C.I., 1895,
p. 174), the test is useless for estimating the quality of spirits.
Spirits which show no brown colour with sulphuric acid may, under
some circumstance, contain more fusel oil than others turning brown.
The best quantitative test for fusel oils is, as already remarked, Rose's
separation method, as modified by GLasienapp (see p<age 11). This
method detects even infinitesimal quantities of fusel oils in the best
Free Acid. The alcohol must not affect litmus paper.
14 ALCOHOL, ETHYL.
Note. Sehweis&inger (Rep. d. Chem.-Zeit., 1887, p. 174) observes
that traces of free acid exist in alcohol. According to him, 0'4 c.c. T n o
soda, solution were required to neutralise the free acid in 100 c.c. of a
good alcohol. Notwithstanding this, good alcohol must not affect
litmus paper. For the estimation of free acid in the Swiss Excise
Department (Chem.-Ztg., 1893, No. 84) 50 c.c. were titrated with
53- soda solution, using phenolphthalein as indicator. In 1,070
analyses it was found that, in order to neutralise the acid in 100 c.c. of
alcohol the maximum of -.fy- soda solution used was 3*2 c.c. and the
minimum 0'5 c.c.
Furfural. To 10 c.c. of alcohol 10 drops of aniline oil and 2
or 3 drops of hydrochloric acid are added ; in presence of furf urol
a more or less pinkish-red colour is produced.