A T H E R T O N SEIDELI,
IOSS
[ C O S T R I B U T I O S F R O M THP: 1)RI:G 1,:SHELI
~ , h H O R A T O R Y .III-RE.\U
OF CI3I:lf I S T R Y .
]'VI;-
BY P I ~ R 3 I I S S I O SO F THE: SECi l a y ,
2 , I,'I:.
T h e experiinents described iii this paper ]\'ere uiiclertaken for thc purpose of obtaining the necessary data iipoii \vliicli to liase a satisfactor), method for the separatioii of home of the ct!iisiittieiits of lieadache poivders. T h e object vas to fiiitl solvciits i!i whicli the solubilit!- of oiic or more of the above iiained substaiiccs differed iiiost from that of the others uiider t h e saiiie coiiditioiis. T h e tleterniiiiatioiis \Yere tliereforc not made with tlie care \vhicii s1io:ild be devoted to result:; siibiiiitted priniarilj- a s solubilit!, constaiits, siiice it \vas recognized that onl!- ivitlc differences iti the actual :iiiioLiiit,s of :lie se:-era1 suhitaiices clissolvetl b>2 single solvent, could be utilized for the cluaiititati\~e q i a r a t i o i i of mixtures of these conipouiicls. A search of the literature indicated that comparative!y few solubility results ivliich coulcl be used for the purpose. were available. T h e usual reference books upon piiai-maceutical and organic clieiiiistrJ- give approsimate deteriiiiiiatioiis for each of these substances in the more coii~iiioii solvents such as water, alcohol, ether. cliloroforiii. ctc. 111 adclitioii to data of this character there were found results upoii the solubility of acetanilide in methyl alcohol, ethyl alcoliol, aiicl chloroform ~ > ySpe>-ers' and i n iiiistures of etlij-1 alcohol and water 11~.Holletiiaii and Aliituscli.' T h e material used for the deteriiiiiiatioiis here recorded was in all cases that sold uiider tlie designation c'. 1'. or 7:. S. P. The esperiiiieiits ivci-e made a t room temperature arid therefore the temperature..; recorded are probably accurate in iiiost cases to within a degree o r tlvo. T h e method of deteriiiiiiatioii was as follows : .\bout 30 cc. of tlie 501vent and an excess of the solid i i i each case irere placed iii glass stoppcreil cylinders ivliich were then attached t o ail axle ivhicli \vas rotated b y iiieaiis of a water motor. A11 deteriiiiiiatioiis iii any oiic soli-eiit Ivere made simultaneously and hence are comparable in so far as temperature, time of shaking etc. are concerned. T h e time of rotation was at least six hours i n all cases, usuallJ. iiiuch longer. T h e removal of a clear portion of the solution was effected by drawing it up with the aid of suction iiito a pipette provided with a filter of ahsorbciit cotton a t the opeiiiiig through which tlie liquid eiitered. T h e solution was then transferred to a pycnometer and after being weighed was washed (usually with alcohol) into a tared platinum dish. T h e solveiit was then removed by evaporatioii at rooiii temperature with tlie aid of ail electric Spcyers, . h i . J. Sci. ( 4 1 . 14, 294, ( 1932). Ho:le:iin:i and A n t u i c l i , Rec. t r a v . chin:,, 13,
293, I
iS94).
1089
SOLCBILITY O F ACETAKILIDE, ETC.
fan. T h e residue remaining in the dish was alternately weighed and dried at 60' until no further loss occurred. I n the case of benzaldehyde, the benzoic acid formed during the evaporation was determined by titration with standard alkali and its amount deducted from the total weight of residue found. With salol in benzaldehyde, however, a determination could not be made since no solid separated even on continued evaporation of the solution by means of the current of air. T h e first series of determinations were made in aqueous ethyl alcohol solutions and the results are presented in Table I together with those of Holleman and Antusch upon similar aqueous alcoholic solutions, determined at 25'. Eight saturated solutions were prepared and rotated at the temperature of the room which held closely a t 30' for several hours previous to the withdrawal of portions for the analysis.
_
T.4BLE I.
SOLUBILITT O F ACETANILIDE I N 3IIXTURES O F ETHYL ALCOHOL A N D WATER.
Weight Percent Alcohol i n Mixtures 0 IO
-___-
Results at zjOl
0.54
0.93
20
1.2s
30
2.30 4.Sj S.S7 14.17 19.84 25.17 26.93 27.6j 26.S~ 24.77
40 50 60 io
so 85 90
95 I00
Results at S p gr.
of
Solution
0.997 0.995 0.973 0.962 0.950 0.939 0.92s 0.918
30° 7
_---A__-
h
Gms. C,;HaXHOCsH, per roo gins Sat. Solution
Gins. CBH5NHOC2H3 Sp. pr. per IM) gms. of Sat. Solution Solution
0.69
1.000
1.00
0.984 0.970 0.9j6 0.945 0.934 0.926 0.917
2.20
4.80 9.40 15.40 22.00
2 7.60
0.90j
31.20
0.W
0.899 0.890 0 Si4
31.70 31.60 33.s3 29.00
0.900 0.893 0.SSj
0.SjI
0.876
T h e weight percents of alcohol in the eight water alcohol mixtures used were calculated from their specific gravities determined by the pycnometer method. T h e dissolved acetanilide was determined by weighing the residue left after the evaporation of the solvent from weighed portions of the saturated solutions. T h e interpolated values shown in the table were read from t h e curve drawn through the eight points plotted on cross section paper. For this curve, the weight percents of alcohol in the alcohol-water mixtures were taken as abscissas and the grams of acetanilide per IOO grams of solution as ordinates. T h e curve plotted in the same manner from the 25' results of Holleman and Antusch lies somewhat below the 30' curve but in all other respects the two are entirely similar. An examination of these curves shows that t h e amount of dis'Holleman and Antusch.
1090
X TH E R T0 S S E IDEL I.
solvecl acetanilide increases at first slo\vl?- n i t h iiicrease i i i coiiceiitratioii of alcohol, and then quite rapitll!. 111) to the inasiina. lvliicli iii t h e caic of tlie 2 j ocurve is at 90 \\-eight percent alcohol aiid i i i the case of tile 30' curl-e is a t 8 j percent alcohol. *iboi.e these coiiceiitratioiis of alcohol the solubility diminishes in each case. A satisfactory esplaiiatioii for this increased solubilit!, of acctaiiilide i i i alcohol coiitniniiig I O ~ w r cent water (and eve11 inore at a liiglier temperature) is :it present iiot appare11t. Tlic solubility of acetanilide in methyl alcohol, etlij.1 alcohol aiid i l l i i i chloroform as determined by Speyers' is reported i i i terms of gram iiiolecales of acetanilide per I O O #rani molecules of the sol\.eiits at irregular temperatures, and in addition the weights of tlie saturated soliitioiis are given for other temperatures than those at which the solubility determinations were made. Xltliough 110 new results for these solvents li31.e been determiiied by the present writer it was tlionght of interest to recalculate those of Spej-ers to tlie basis selected for presenting all the other solubility determiiiatioiis included herewith. T h i s has been done aiicl the results given in Table 11. T.IBLE 11. SUI.LlBIl,ITY
01' ACET.INII,IDE
IS 3IETHY1,
AI.COHOI.,
CHLOROIXIK~I ( SPEYERS). Iii E t h y l Alcohol
111 lfetlivl Alcohol _____ Temper. Gnis. C,,H;.UHOC,H3 U't of G m s . C , ; H ; S H O C ? H . I Wt. of per IOO gms. I cc. per 100 pms. I cc. aSoiiltion Solution Soliiticn Solution titre. IS.j 0.b6o 12,s 0.842 0
ETHYI,
ALCOHOL
__-_
I . j03
I . 39s
0. 895
14.5 1S.7 23.7
0.920
29. I
1.272
16. j
0.844
29. I
21.>?
O.SjO
3"
.:j.I
0.S92
26.5
0.S60
40
42.9 51.7
u.911
32.9
o.s;4
0.912
39.4
59.2
0.957
46.4
20
50 60
G i n s C , . H ; S H O C ? H 3 \Yt of per loo gms. I cc. Sclutiotl Sollition
3.53 j.24 10.7
0.864 0.87j
23.1
IX
I n Chlorcform
__-__A
IO
,491)
1.475
1.440
1.254 1.314
I n Table I11 is shown the comparative solubilities of acetanilide, phenacetine, caffeine and salol iii twelve different solvents. Although this number is not as large as would be desirable in order to fully accoinplish the purpose mentioned in the first part of this paper, it is believed that a fair degree of success can be attained in the separation of mixtures of at least three of the above named substances on the basis of the results shown in the table. T h u s it may be expected that with proper nianipulation either toluene, benzene or xylene will remove salol when mixed with the other three compounds, and amyl alcohol or acetate will effect a separation of caffeine from a mixture of this substance with acetanilide and phenacetine. I t is seen that with none of the solvents is there a wide difference between the solubility of acetanilide and phenacetine.
' !oc. cit.
ACETANILIDE I N HEADACHE POWDERS
1091
TABLE 111. SOLCBILITY OF ACETAXILIDE, PHENACETINE, CAFFEINE A N D SALOL I N ORGANIC SOLVESTS. Solvent.
sp. gl.. of solvent. t')
Grains per
IOO
grams saturated solution :
_ A
Acelanilide.
Phenacetine.
Caffeine.
Salol.
0.997 25 0.56 0.11 2.14 insoluble 0.716 '' 7.7 1.56 0.27 1.476 11.0 16.6 4.76 2.18 (0.E32) 90.99 0.527 30-3I 31.75 (0.902) 10.68 1.22 ( 0 . S 7 j ) 8S.jj (1.145) 0.S 7 2 2.46 (0.S75) 0.65 (0.573) C' ? I.0jj 18.83 (1.06s) 5.44 (1.063) 11.62 (1.087) 0.S60 10.46 (o.SS2) 2.42 (0.865) 0.72 (0.862) S5.29 (1.136) 1.02 19.38 (1.034) 9.46 (r.025) 22.89 (1.080) very soluble 0 SI4 0.49 (0.810) 20.44 (0.569) 25 14.00 3.5r (0.519) 1.055 21. j 33.21 13.65 (1.064) 2.44 63.24 (1.143) 0.547 32.5 1.6j (0.847) 1.25 (0.847) 1.11 (0.S47) s7.14Toluene 0.S62 25.0 0.jo ( 0 . S 6 2 ) 0.30 (0.863) o 57 (0.861) 83.62 (1.12s) Figures i n parentheses are the specific gravities of the saturated solutions.
'Water 'Ether 'Chloroform Acetone Benzene Benzaldehyde Amylacetate Aniline Aniylalcohol Acetic acidY Xylene
Fairly good results have been obtained by the author in separating mixtures of different amounts of salol, caffeine, and acetanilide by digesting first in toluene for the removal of salol and then in amyl alcohol for the removal of the acetanilide. As is shown by the table, none of the four compounds are completely insoluble in any of the solvents, and consequently it cannot be expected that a very sharp separation can be made. T h i s is especially true for mixtures containing relatively small amounts of either of the substances. I t is also to be mentioned that the influence of the presence of one substance upcn the solubility of the others in the mixture has not been investigated and therefore conclusions based upon the solubility of single substances in a given solvent may be considerably in error when applied to mixtures of two or more substances acted upon by the same solvent. T h e author regrets that circumstances have prevented the extension of these determinations to a larger number of solvents and the application of the results to a more accurate method of separating and determining the constituents of headache powders. [CONTRIBUTION FROM THE DRUGLABORATORY, BVREAUOF CHEMISTRY, LISHED BY PERMISSION O F THE SECRETARY O F AGRICULTURE].
PUB-
A RAPID METHOD FOR THE QUANTITATIVE DETERMINATION OF ACETANILIDE IN HEADACHE POWDERS. BY ATHERTON S E I D E L L .
Received May
2. 1902.
\\.'bile continuing the investigations described in the previous paper a suggestion was obtained which has led to the following method for deterResults from U. 99.5 %
s. P., 8th Revision.
