T H O M A S W . B R I G N A L L , A . M . T o d d C o m p a n y , K a l a m a z o o , IV lich. official procedure of the United States Pharmacopoeia, is known to be erroneous and to give variable results in differ
ent laboratories; however, none of the modifications pro
posed thus far has been entirely satisfactory. This procedure is known to indicate as alcohols such nonalcohol constituents as aldehydes, amines, phenols, esters, and some of the more unstable terpenes which are acetylated along w ith the alcohols when th a t method is used. T he present are included during saponification. During th e course of research this m ethod has been thoroughly investigated and, as a basis for further conclusions, it is believed advisable to review briefly the possible sources of error in the order in which they occur during the analysis of an oil. These discrepancies in
clude those observed in the author’s laboratory as well as those reported in the literature.
In a report upon the analysis of oil of peppermint, Nelson (5) states th at “the proper care of the sample previous to analysis is highly important: all reagents used must be of the best quality and the sodium acetate used as a catalyst must be absolutely anhydrous.”
I t has been observed here th at acetic anhydride must be used within a short time after the container is opened; otherwise, deterioration of the anhydride, due to its hygroscopic nature,
will be evidenced by low results in alcohol analyses. Frequent standardization of solutions, especially of the alcoholic potassium hydroxide, is necessary to ensure consistent results. I t has been found necessary to use hot solutions when washing the acetylated oil to hydrolyze the excess acetic anhydride completely. No evidence of hydrolysis of the acetylated oil has been found if this procedure is used and the dried acetylated oil has always been neutral.
The author has long stressed an exact 1-hour saponification time, having observed th a t variations in this factor caused marked differences in results. Recently Baldinger (1) conducted a time study, varying the length of both acétylation time and saponifica
tion time with each sample. The interpretation of the results of this study is summarized as follows: “ I t would seem th at the time of acétylation may vary within broad limits, while the time of saponification had best be varied between 45 and 60 minutes.
While experimental evidence is not yet available to prove the contention, it is believed that résinification or polymerization of certain constituents is induced by prolonged heating with potas
sium hydroxide and th a t some base is used up, thereby leading to erroneous results.”
Redemann and Lucas (7) have observed th at more rapid hy
drolysis of esters results if diethylene glycol is substituted for ethyl alcohol as a solvent for potassium hydroxide. This has been verified by Hall, Holcomb, and Griffin (4), who applied this method to the analysis of the isomers of menthol.
In the titration of the saponified oil. variations occur due to differences in the estimation of the end point by different ana
lysts. Badly oxidized or poor quality samples are often difficult to analyze consistently, owing to discoloration during saponifica
tion. This interferes with the observation of the end point when phcnolphthalein is used as an indicator.
The method of Delaby, Sabetay, and Breugnot (2) for the determination of free alcohols in the sandalwood oils gives results from 5 to 8.5 per cent lower than those obtained by Power and Kleber. These authors traced this divergence to the fact that
M arch 15, 1941 A N A L Y T I C A L E D I T I O N 167
the santalenes present were esterified along with the alcohols by the latter method.
W ith these facts in mind, the following experimental work was performed in an effort to reduce to a minimum the method of analysis it was found to contain negligible quantities of esters and alcohols. From a 2-kg. lot of natural oil of pepper
mint the menthol was separated, using the fraction boiling above 70° C. at 2 mm. This crude menthol was purified by centrifuging and by repeated recrystallization'from benzene until a constant melting point of 43° C. was obtained for the product. This menthol was dissolved in varying proportions in the redistilled pinene to give samples having definite percentages of free men
thol. Using these samples, a time study similar to th at of Bal- dinger (1) was conducted.
Pyrex-resistant glassware, with ground-glass joints and inter
changeable connections, was used throughout. Electric heaters equipped with sand baths were used and preheated so th at sam
ples placed upon them began to boil almost immediately. Acety- lation and saponification periods were varied by half-hour intervals up to and including 2 hours. Except for the introduc
tion of these variables, the method described in the U. S. P. XI was employed.
The results of this study (Table I) substantiate th e pre
viously mentioned contentions of Baldinger regarding the reaction of alkali during saponification w ith components in the oil other than esters. The am ount of base entering into these side reactions increases sharply if the acetylated oil is perm itted to saponify over 1.5 hours. This contributing fac
tor obviously depreciates the classical method by indicating a higher than actual alcohol content.
Several methods ($, 8, 8), using an acetylant mixture of py
ridine and acetic anhydride, have been proposed for the analysis of free prim ary or secondary alcohols. Because of the objectionable odor of pyridine, an investigation was con
ducted for solvents th a t m ight be substituted in place of this reagent.
A n a ly t ic a l P r o c e d u r e
One gram of the sample, accurately weighed in a tared acetyla- tion flask, is treated with 5 ml., accurately measured by means of a Koch microburet, of a freshly prepared acetylant mixture consisting of four parts by volume of n-butyl ether and one part of acetic anhydride. A blank is prepared in an identical manner, omitting the oil. The flasks are connected to air condensers and the contents boiled gently for 1 hour on a sand bath. Without glass connections rinsed by means of a water wash bottle, allowing the rinsings to flow into the flasks. At this point, the acid strength of the blank is approximately 0.5 N . Eight or 10 drops of phenol- phthalein test solution (dissolve 1 gram of phenolphthalein in 100 ml. of alcohol) are added and the excess acid is neutralized with 0.5 N alcoholic potassium hydroxide solution. The blank is titrated to the full red color of the indicator and the oil sample matched with the blank. I t is necessary to conduct this titration on a 2-phase system, but this presents no difficulty. The differ
ence in the amount of potassium hydroxide indicates the
per-168 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 13, No. 3
term ination is made by the regular method and the following formula will effect the necessary conversion: average about 3 per cent lower than those done by the method of Power and Kleber. In Table I I I this divergence is traced to small quantities of aldehydes contained in the lower boiling fractions, which aldehydes, it is believed, are acetylated by the pharmacopoeial process, and to polymerization during saponification of the higher boiling resinous constituents, with a subsequent utilization of alkali. The d ata on redistilled and badly oxidized oil of peppermint indicate th a t the am ount of resinous m aterial affects the extent of th e divergence.
The d ata presented would indicate th a t a free alcohol con
the method herein described, would very closely approxim ate the U. S. Pharmacopoeia 50 per cent total alcohol minimum or the British Pharmacopoeia 46 per cent free alcohol mini
mum if the analysis were made by the classical procedure.
The pure alcohols included in Table V I give harmonious re
sults regardless of the method used, except th a t d-neomenthol m ust be saponified for more than 3 hours if the U. S. P. X I procedure is used (6).
Differences obtained when assaying the impure alcohols included in Table VI or the essential oils shown in Table V are explained by the presence of nonalcohol constituents which are acetylated by th e classical method b u t are not attacked by th e acetylant m ixture proposed in this paper.
Ta b l e V I. An a l y s e so f St o c k Sa m p l e so f Al c o h o l s investigated. Sources of discrepancies have been found and suggestions m ade to prevent their occurrence.
A new simplified procedure, utilizing an acetylant mixture of n-butyl ether and acetic anhydride, permits a more selec
tive esterification of alcohols th a n the pharmacopoeial method.
This procedure is apparently applicable, w ithout variations, for the analysis of free prim ary or secondary alcohols, regard
less of content, in any of the essential oils. Since saponifica and prevents the formation of an interfering precipitate.
The procedure described also seems to be more satisfactory than those using a pyridine-acetic anhydride mixture. The odor of n-butyl ether is not objectionable and this solvent possesses the added advantage over pyridine th a t its boiling point (142° C.) coincides well w ith th a t of acetic anhydride (140° C.), whereas pyridine boils a t 115° C. More rapid and complete acetylation would therefore be indicated in an ether mixture than in a pyridine-acetic anhydride mixture.
G reater dilution of the acetic anhydride is possible if n-butyl the University of N otre Dame for his assistance in preparing th e manuscript.
M arch 15, 1941 A N A L Y T I C A L E D I T I O N 169
L it e r a t u r e C ite d
(1) Baldinger, L. H ., J . A m . Pharm . Assoc., 28, 155 (1939).
(2) D elnby, R ., Sabetay, S., and Brcugnot, Y ., Perfumery Essent.
Oil Record, 26, 334 (1935).
(3) Freed, M ., and W ynne, A. M ., I n d . E n q . C h e m . , Anal. Ed.. 8, 278 (1936).
(4) H all, R . T ., Holcomb, J. H ., and Griffin, D . B ., Ib id ., 12, 187 (1940).
(5) N elson, E . K ., J . Assoc. Official Agr. Chem., 19 , 529 (1936).
(6) Power, F . B ., and Kleber, C., Pharm. Rundsch., 1 2 , 162 (1S94).
(7) Redemann, C . E ., and Lucas, H . J., I n d . Eng. C h e m . , Anal. Ed., 9 , 521 (1937).
(8) Verley. A., and Bolsing, F ., Ber., 34, 3354 (1901).