OBSERVATIONS ON AJIEBICA-19 MENTHOL. BY J. H. LONG.
As is well known, the menthol from the American oil of pepperment differs mncli in its physical appearance from t h a t obtained from the English or Japanese oil. The common Japanese menthol appears in commerce in the form of small prismatic crystals, while the American products appear its long, slender needlee. This difference is evidently not accidental, depending on the method of distillation and purification followed by the different prodiicers, as i t is observed in the products which have been recrystallized by the same person under the same general conditions. A distinct difference in the odor of the two products may also be noted, even after several ciytaliizations and distillation in vacuo. T h e American mcnthol is produced chiefly from the oil distilled in New.To1.k and llichigan, b u t owing to its relatively high cost is not commonly found in the market a t the present time. A grade of very great purity is produced by the well known distiller of essential oils, d l b e r t M. Todd, Esq., of Xottawa, Mich., to whom I am indebted for samples used in this investigation. The peculiaritiefi of this menthol mere explained in the Bnzericnn J O ? l l . 7 d qf Phnrmccy for June, 1884, and the name pij)rneuthoZ then suggested. A fuller description of the product was given by Prof. Henry Trimble in the August, 1884, number of the same journal. T o obtain the menthol from the oil the latter is placed in a vessel with double walls, the space betwceii being packed with a mixture of ice and salt. After a time crystallization takes place, from 40 to 40 per cent. of the menthol present separating out. The solvent oil is thenallowed to drain out through a tubeleading from the bottom of the inner vessel. By exposure to the air the crystals soon dry and are practically pure, as shown by tests of numerous samples made a t different times. From oil very rich in the
camphor crystds scp,ir:itc o i i t .tt .t tcniperatuw near 0 C'. specimen scrnied I I I this \iay dricd by siniply passing betweeii folds o f filter papcr mhen I t Tvtth foulid t o I J C IlearlJ pule. 111a preliminary ( miiiation of a iiiinibci of samples obtained from sevcral s o i i r t v , s ~ l u t i o ~\+ i s('re niatic 1). tlissol\ ing 10 grms. of the comnii~rci:ilI ) i m l u c t i i i : t l ~ o l i o i o f 95 p i ' c w t . stlciigtli anii making u p to 100 e(>. These soi~itioiiawere polai i ~ e d111 a 300 nim. tube of the large I ~ i i d o l at i ' p m i t n s \ v i t l i the followiii- result: so. a, 1-- 14. S(!3. '!-14. ;11. :L--14*;45, 4-14.660. 3-- 14 .(j90. (&--14. 7 60. Samples 4 and 5 were quite moist. After drying liiglier results vere obtaiiied, eacli showing
an= - 14.73, From this it appears that the commercial prodnct is nearly constant 111 composition, if the polarization can be taken as a criterion. I had a t my disposal a large amount of sample No. 6, which was purified as follows: About 500 grms. was melted and poured into warm water 111 a beaker on a water bath. T1.e two liquids were repeatedly mixed by stirring and kept a t a temperature of 60-70" C half a n hour. I n this way a slight amount of some insoluble body was separated from the menthoi, aiid collected at t h e bottom of the beaker. T h e menthol layer was then poured off aiid allowed to stand until it crystallized, the crystals being afterwards dried between porous plates in the air. An attempt was made to distil these dried crystals under the ordinary atmospheric pressfire, h u t without perfect success, as a slight decomposition took place. This was indicated by a change of color in the boiling liquid, it becoming yellowish brown, and also by a lowered rotation in the distillate. A solution of 10 grms. made to 100 cc. with alcohol, as before, gave now an= - 14."51.
OBSERVATIONS O N A M E R I C A N M E S T H O L .
151
However, distilling under reduced pressure gave a much better result. Some specimens were distilled several times under R pressure of 50-100 mm. and when pohrized under the same conditions as before gave uniformly A,= - 14."520. The whole sample was purified in this way and nsed for a number of experiments chiefly in t h e line of polarization phenomena. DENSITY,
MELTING POINT A X D BOILIXTO roIwr.
The dcnsity of the solid menthol in compact form was found by weighing in water to be 20 d - 1 .SOO. 4
This is the value given by Moriya (JOUY.Chem. SOC.sxxis, 7 ' ; ) for the crystals a t lj", For the density of the melted menthol I find 44.G d-- .PSlO, 4
which is somewhat higher than the value given by Gladstone and Dale for the Japanese. JIy own determination for the Japanese is 47 d= .573. 4
The melting point of the distilled American menthol I find as 42."3, while that of the single Japanese siimple was a trifle higher, vie : 4 2 O . G . These are somewhat higher than the values usually given, which were probably obta'ined from the undistilled product. With crystals dried in the air or between paper I find lower values. Moriya (loc. cit.) gives 30°, Atkinsou & Yoshida (Jour. Chcm. SOC. xli, 50) give low values for the air dricd crystals, but find AT?."'? as the melting point of a sample purified by distillation. Tlie boiling point of the distilled American menthol I find to be P l ? . " j (corr,) with Bar. = '742. The values given by various observers for tlic Japanese range between 210" and 213". S P E C I F I C ROTATIOX.
I find nothing in the literature on the specific rotation of the liqnid (melted) menthol.
Some results are found for the Japanese
15?
O U S E K T A T I O K S ON A U E R I C d S MENTHOL.
menthol lased 011 a n examination of alcoholic solutions, b u t the date seems to be, in sewral instances, inaccurate. The value given by Oppeiilieim (Jour. Chem. Soc. xv, ? 6 ) is [a ID = --- 59. :ti, dedncecl from a 10 per cent. solution in alcohol. Rloriya (loc. cit.) gives [illj= - 50."3 wliich corresponds approximatel? to [a],= - 4 9 . O 4 for a 10 per cent. solution. For a 10 per cent,. alcoholic solution, d r t h (Ann. de C'him. e t de_PIiys. [GI? vii, 435) gives [a],= - 50."1, while for one of 5 per cent. he gives [a],= - 49. i4. I have made R direct determination of the rotation in a sample of the melted American menthol. A 200 mm. jacketed tube was employed and a temperature of 46" was maintained during t h e time of the experiment by means of a current of warm water. This temperature is higli enough to keep the substaiice in a state of perfect fluidity, and is the temperature a t which the specific gravity was found. The mean of a number of closely agreeing readings was :ID= - 8';."7lG, and this gives, after introducing a slight correction for length of tube a t the higher temperature, [a],= -49."86. T o determine how this value agrees with those deduced from the rotatioil of solutions was the object of a lengthy series of experiments, the main details of which are here given. Solutions were made in alcoliol, benzeiii and glacial acetic acid, the menthol used being all t:tken from the same distilled sample. SOLT'TIOXS
1s- A,LCOHOL.
F o r these tests a pure redistilled alcohol of approximately 95 vol. per cent. was used. This was taken in preference to absolute alcohol where change of strength by exposure toair is unavoidable.
OBSERVATIONS O S ANERICAN MENTHOL.
153
The actual strength of the alcohol employed in all the tests is given by the specific gravity, 20 d-- .81441. 4 Eight solutions were made with approxiniately 10, 20, 30, 40, BO, 60, 70and SO per cent. of menthol, all weights being reduced to vacuo. A portion of each solution was used for a specific gravity determination in a small pyknometer, brought always to a temperature of 20" in a vessel of water. This temperature was maintained with an error of less tlian 0."1 in all the experiments. The remainder of each solution was used for polarization in a 300 mm. tube, kept a t POc by a current of water. The variation of the length of the tube from 300 is inappreciable, and can be neglected. The results of the experiments are shown in tabnlar form, as follows:
'
Per Cent. of Yrnthol.
Per ('elrt '?f
Alcohol.
From the observed values of tlie above table tlie specific rotations given below mere calcu1att:d according to the iisual formula.
Where a is the observed angle of rotation, 1 the length of tube in millimeters, c the concentration, or grms. in 100 cc. of solution, p the percentage strength of tlie solution and d its density. The Talues of [a] are not quite constant, but vary sliglitly with the concentration. decreasing as the latter increaees. I n other words [ a ] increases with q, expressing by this the percentage amount of the inactive solvent present. By plotting tlie values of tlie specific rotation as ordinates with the amounts of solvent as abscissas TVC get a curve nearly straight a t first but finally slightly concave upward, as shown below in the figure at A. The deviation from ;L straight line is so little, however, t h a t t h e values of [a] can be very well espiesscci by t h e simple interpolation formula: [a] = .!! By. cqz. From solutions 4, 6 and 8 I find [a] = 48.2163 + .0132:57 q. - .00001041 qz. From G, 7 and 8, [a] = 45.191 .014974 (1. - .00003886 q e . From 3, 63 and 8, [a] = 48.230 -b .011510 (1. 00001795 q 2 . From 2 , 5 and 8,
+
+
+
+
OBSERVATIONS
155
ON AJlERIC.4N MENTHOL.
+ +
+ +
[a] = 48.312 .006840 q. .00007501 42. The mean of the last three is, [a] = 45.247 .011108 q. .00001870 qt. I employed this last equation in the calculation of the values given in the table below, under : "[a] cal." No.
I ~
pd.
-
c.
5.3281 16.6124 25.2203 34.1845 41.6807 51.8020
G1.1320 60.6040
1
1
[a] obs.
49."5';1 49. "330 49."131
48."968 48."843 48. "727 48."GOG
48."489
[a] cal.
49."39G 40."255 40. "I 15 48."076 48. "851 48. "720 48. "598 48. "488
I
diff.
- .1$5 - .075 - .016
+ .008 + .008
- .007 - .008 - ,001
The calculated and observed values agree very closely except for the two most dilute solutions, where the specific rotations by observation is higher. In Arth's experiments, referred to above, the strength of t h e alcohol is not given, nor is the density of his solutions, but i t is evident that the value for the 10 per cent. solution is too high. This value is inconsistent when taken in connection with that of the 5 per cent. solution. From thc above observations it is seen that the specific rotation of the pure substance, as deduced from tests of solutions, must be about 48.O25 a t 20". I n order tQ compare this value with 'that obtained by observations on the melted menthol I polarized solntions a t higher temperatures. One such solution contained 17.31 per cent. of menthol and had the specific gravity : d = .798. I n the 200 mm. tube it gave a = - 21."05, from which [a] == - 50."62. That is, the specific rotation increases with the temperature. Assuming that the increase is approximately the same for all concentrations the specific rotation deduced from strong polutions a t
156
OHSERVATIONS O S A l I E K I C A X MESTHOL.
this temperature should be about - 4!I.-.i. torily with the observed value - 49.:89.
This agyees satisfac-
SOLVTIOSS I S BENZEh-E.
Several tests were made with pure recrystallized benzene as a menstruum, in which the menthol is very readily soluble. The chief results of these exl'eriments are given i n this table : _____ SO.
_____
__
___._____~_
I I
~~~~~~~
Percent. of Menthol.
~
~
Percent. of Benzene.
89.9G74 79.9837 69.5657 GO. 1223 50.GlOl 40.1201 30.3486
~
~
~~
i
I
~
~.
. .-~ .. .. ~ .
~ p~. r . of Solution.
.molG .88131 .Stl'?;G .88435 .88575 .88X8 .so025
-
Obserred Rotation.
- 12."282
- 25.O265 - 38.'928
- 51."437 - 64. "206 - 7 8 ."300 -9 1 . O 5 1 0
Calculating the specific rotations from the abore observations and plotting them as functions of the dilution, (1, we obtain a curve wliich deviates greatly from a straight line, The deviation at the points corresponding to weak solutions is so great that a simple interpolation formula will not satisfy the whole curve. For the stronger solutions this equation gives results which can be used : [a] = 49.O511 - .025G34 (1. + ,0008403 q g . - .00001102 y3. Xearly the same constant is given by the equation : [a] = 49."495 - 28.614 q. 283G.S + q. The true specific rotation bf the menthol as found from the benzene solutions appears, therefore, to be higher than that found from the solutions in alcohol. The specific rotation of the melted menthol lies between that deduced from the benzene and alcohol solutions.
157
OBSERVATIOKS O B AMERICAN MEXTHOL.
Owing to the greater volatility of the benzene it is probable that the rotations observed are a trifle high because of a slight loss of the menstruum by evaporation, even when great care was taken i n the manipulation. The values of the .specific rotation are given in the following table, while the plotted curve, showing variations in some, is given a t B iu the figure. No.
I
pd.
=
c.
1
[ti]
obe.
46. "366 47. "741 48.O299 48."619 48."922 49.O007 49."193
1
[a] cal.
4a.0055 48.O613 48. "940 49. "125 49.O199
!
tiff.
- .214 - .006
+ ,018 + ,028 + ,006
SOLUTIONS IY ACETIC ACID.
A third series of experiments was carried out, using pure recrystallized acetic acid as the solvent. The solutions were made jn all cases without the application of heat and were not allowed to go alrove 22". The table below gives the results of the individual tests : No.
1
Per Cent. of
Menthol.
Per Cent.
Sp. Gr.
Acid.
Solution.
of
80. '7512 ? 9.8802 69.8044 50.5014 50.4876 40.1409 20.9780
of
1.03009 1.01295 .09615 .98010 ,96606 ,95161 .9375$
Observed Rotation,
- 15."482
- 29."744 - 43."739 - 57."559 - 09.~201 - 82."164 - 94."454
The results of the calc.ulated specific rotations, when plotted aE before, give a very regular curve resembling that found from tho alcohol solution. This curve is shown a t C in the figure.
15s
O B S E K Y A T I O S S O X A J f E R I C d N JIENTHOL.
The points on this curve are given by equations as follows : From 20, 2 1 and 2 2 , [a] = 47.Gtj3 + .00763X q. .00006!129 q'. From 1 7 , 20 and 32, [a] = 47.G1.5 .010226 9. .0000371'? q L . From 16, 18 and 21, [a] = 47.526 .001993 q. + j- .00010754 42. The mean of these gives, [wj = 4;,;11 .OOG38G c!. j- .00007142 'f2. from wli~cli the calculated specific rotations 111 the table below were fountl. These agree very wcll with the results of direct experiment :
+ +
-+ + +
10,5665 20.3718 30.0 ;92 39.6928 4728320 5G.9625 6 5 . G50G
48. "840 4s. "GGS 4s. "471 4s. c33s 4s. 9 2 5 4s. "OS1 47. "954
4s. "859 48. - G 7 G 4s. "505 4s. "343 45."215 48. "682 47.3GCi
+ .019 + .008 -t .034 + .005 - ,010 + .OOl + .012 __
It is apparent that acetic acid is not a good menstruum in this case, although the values given are very regular. The specific rotation in the concentrated solutions falls below that observed with the liquid menthol. A t high temperaturos nientliol and acetic acid act on pach other with production of acetate, and it is possible that some of the same product isformed a t a temperature of 20" even. But Oppenhuim has shown (loc. cit.) that the acetate has a much higher specific rotation than the original menthol, which reenlt I have confirmed by recent experiments Ilowever, with the ether formed a t the ordinary temperature I shonld have found a higher rather than a lower specific rotation with t h e solutions experimented on. An observation having some bearing on %he qnestion was made during the tests in this manner. Several solu-
OBSERVATIONS ON AMERICAN MENTHOL.
159
tions, first polarized immediately after preparation, were kept ten days in well stoppered bottles and polarized again. I n the second test a higher result was always found, which would correspond to the behavior of ether produced in the interval. Apparently, the first effect of the acid with the menthol is to produce a combination characterized by slightly lower specific rotation, which later gives place to the ether with higher than normal rotation. Several experiments were tried with a single sample of Japanese menthol dissolved in alcohol and acetic acid. The results found were slightly higher than those obtained from any one of t h e American samples. This Japanese product was clean and dry and apparently a pure commercial specimen. I redistilled i t under low pressure and used this product for the test. I n the 300 mm. tube it gave, a = - 14."840 for a solution containing exactly 10 grms. in 100 cc., the same alcohol employed iii tho other test being used here. The purified American sample gave under the same coxlitions, a = - 14."820. The slight differences in physical properties between the two varieties of menthol are not important enough to point to a difference in chemical structure. It is possible that they may be due to the presence of traces of associated products in the original oil which cannot be perfectly separated by the simple methods employed in pnyification. I t appears from the investigations of Beckett and Wright (Jour. Chem. SOC.,XXIX., 1)and Trimble (loc. cit.) that the oil of peppermint contains, besides mentliol, another oxidized product in small amount, probablj mcnthonc. This is a body with high boiling point, only slightly lower than that of menthol and + rotation. A trace of it could easily remain with t h e menthol and exert an appreciable influence on the specific rotation, sufficient, probably, to account for the trifling differences noted above. CHICAGO,May, 1892.
