General Qualitative Test for Epoxides RICHARD FUCHS, RUSSELL C. WATERS, AND CALVIN A. VANDERWERF Unioersity of Kansas, Lawrence, Kan. importance of epoxides, both in industrial use THandE groning in research, underscores the need for a quick, general qualitative test for the epoxide linkage. This paper reports such a test, using acidified periodic acid. This reagent effects hydrolysis of the epoxide to the corresponding glycol, which is then oxidized by the periodic acid, and the resulting iodic acid is detected by precipitation of silver iodate upon addition of silver nitrate. Periodic acid has been used successfully for some years in this laboratory for proof of structure of unknown epoxides (3)and, rather recently, in the quantitative estimation of ethylene oxide (1). The periodic acid test is negative for simple aldehydes and ketones, and therefore serves to distinguish these compounds sharply from epoxides, a differentiation which cannot be made by analysis and which is not clear cut from physical constants nor by use of the familiar carbonyl reagents. Styrene oxide, for example, gives a positive test with 2,4-dinitrophenylhydrazine reagent and with fuchsin-aldehyde reagent, probably because of an acid-catalyzed isomerization to phenylacetaldehyde. Other compounds that are oxidized b y periodic acid give positive tests; thus 1,2-gIycols, a-hydroxyaldehydes and ketones, l,a-diketones, and or-hydroxy acids may interfere. I n practice, however, these may be differentiated from the corresponding epoxides by use of other chemical tests and by the relatively sharp differences in physical properties. The test was positive for the following representative epoxides
tested: ethylene oxide, propylene oxide, epichlorohydrin, 1,2epoxybutane, 2,3-epoxybutane, l-bromo-2,3-epoxybutane, butadiene monoxide, cyclopentene oxide, cyclohexene oxide, styrene oxide, p-chlorostyrene oxide, and safrole oxide. Very slightly water-soluble epoxides, such as safrole oxide, give the test rather slowly in water solution, but rapidly in 50% aqueous dioxane. PROCEDURE
The procedure is a modification of the qualitative glycol test of Shriner and Fuson (2). Exactly 2 drops of roncentrated nitric acid is added to 2 ml. of a 0.5% solution of periodic acid, and 1 or 2 drops of the unknown is added. Water-insoluble unknowns should first be dissolved in 2 ml. of dioxane or acetic acid, The mixture is shaken and I or 2 drops of 5% silver nitrate is added a t room temperature. A positive test is the appearance of a 1vhit.e precipitate of silver iodate which usually forms immediately, but may require up to 5 minutes for complete precipitation. I n all cases a blank should be run for comparison. The test is negative for simple aldchydes, ket,ones, and alcohols. LITERATURE CITED
(1) Eastham, A. ?VI.,and Latremouille, G. A., Can. J. Research, B28, 264 (1962). (2) Shriner, R. L., and Fuson, R. C., “Identification of Organic Compounds,” 3rd ed., pp. 115-16, New York, John Wley & Sons, 1948. (3) Waters, R. C., Ph.D. thesis, University of Kansas, 1962. R E ~ E I v E Dfor
review March 14, 1952.
Accepted J u n e 26, 1952.
Flame Photometric Determination of Sodium in Salts of Organic Acids SAMUEL B. KNIGHT AND M. H. PETERSON‘ Venable Chemical Laboratory, University of North Carolina, Chapel Hill,N . C . organic acids may often be characterized by analysis BECAUSE of their sodium salts, an accurate and rapid method for the
determination of sodium in salts of organic acids is of considerable interest. A number of surface active sodium salts were available in this laboratory from the studies of Bost and coworkers (4). It was felt that these surface active salts of high molecular weight, would represent extreme cases of the applicability of a flame photometric procedure for the determination of sodium in salts of organic acids. The results from the flame photometric procedure developed for the analysis of the sodium salts reported in this paper are compared with the thcorctical amount of sodium, assuming the salts to be pure, and v i t h the generally used metal residue method of Xiederl and Xederl (6) for the determination of sodium. hIetal residue determinations were available from the work of Bost and coworkers on some of the sodium salts and enough eample was available on most of the remaining salts to complete both a flame determination and a metal residue determination. On some few samples, however, not enough material -ms available for both flame and metal residue determinations. Therefore, a few metal residue comparisons are unavailable. FLAME PHOTOMETRIC MEASUREMENTS
The instrument used for this study is a Perkin-Elmer Model 52.4 flame photometer, which may be used for readings by eithcr the direct (or absolute) method or the internal standard method. Both methods are well knon-n. 1 Present address, U. S. Naval Ordnance Laboratory, White Oak, Silver Spring, Md.
I n the direct method, the sample is introduced into the flame, and the amount of emitted light is measured relative to concentration through the use of standards. The direct method assumes that the delivery of mist from the atomizer is constant, that the rate a t which the sample is atomized is constant, and that the flame characteristics do not vary from sample t o sample. This is not strictly true and may introduce considerable error in some instruments. It is also knon-n that considerable error may be introduced through the presence of salts, acids, ions, and organic materials ( 2 , 3, ‘7). Considerable effort was made to determine sodium in these salts by direct comparison with curves prepared from known concentrations of pure sodium chloridc. The direct method gave percentages n-hich were of the order of 50% of the known value. It was to bc expected that the direct method would give erroneous results on the instrument available, as the work of Dunker and Passaw ( 5 )indicated that emission is affected by a change in the surface tension of a solution. Because the salts under examination produce a marked lowering of surface tension, the authors expected to find high sodium values instead of the low values actually recorded. They have no eqlanation whatever for this phenomenon. Further work showed that if 0.1% of a nonionic surface active agcnt were added t o both standard and unknown solution, the results were improved somen-hat but still unacceptable. Efforts a t determining sodium by the direct method on the instrument available were therefore abandoned. Thc internal standard method (1, 2) employing lithium a t the same concentration in known and unknown samples n-as used exclusively in the analyses reported in this paper. 1514
V O L U M E 24, NO. 9, S E P T E M B E R 1 9 5 2 Table I.
1515
Flame Photometric Determination of Sodium i n Surface Active Salts of Organic Acids
z*&Ga0
Sample, Wt. of Z d Gram ii?" (Mean) (M.R.) (Calcd.) 0.1956 1 3 . 2 4 1 3 . 2 0 . . . 13.21 0.2000 1 3 , l 5
Compound Sodium o-methoxybenzoate Sodium o-ethoxybenzoate
0.1930 '12.33 12.42 0,2000 1 2 . 3 0 0 0472 1 2 . 5 3 0.0479 13.52
12.20
Sodium o-butoxybeuz~iate
0.2062 0.2000
10.77 10 70
10.73
1 0 . 2 5 10.64
Sodium m-ethoxybenzoate
0.2143 0.0543 0.0606 0.0523 0.0534
1 2 . 1 3 12.20 12 27 12.20 12.26 12.13
1 2 . 1 8 12.22
0.1925 0.0324 0.0466
12.20
0.043G
1 2 . 3 6 12.42 12.47 12 41 12.46 10.0: 10.08 10.10 11.19 1 1 . 1 4 11.05 11.20
...
9.99
Sodium phenosy-n-butyrate
0.0988 0.0583 0.2279 0.0490 0.0430
11.16
11.37
Sodium phenosyaretate
0.2065 0.0518
13.22 13.21 13.21
...
13.21
Sodium Huorene-2-bulfonate
0.1836 0.2000 0.1589 0.0713 0.0524
8.51 8.50 8.46 8.46 8.45
8.50
8.53
8.57
0.3534 0.0597
8,04
8.07
7.87
8.03
Benzylidene fluorene-2-sodium- 0,0907 sulfonate 0.0676 HeptylideneHuorene - 2 - sodium 0.0686 sulfonate 0.0647 0,0501
6.39 6.39 5.94 6.05 5.99
6,39
6.25
6.45
5.99
6.01
6.31
- Diiiiethyleiniiiobenzylidene- 0.0762 Huorene-2-sodium sulfonate 0,0830 0.0768
5.91 6.05 5.99
5.98
5.61
5.76
5.30
5.30
5.47
5.52
Sodium p-ethoxyhenzoate
Sodium p-n-pentoxybenzoate
Sodium dibenzofurane-2-sulfo nate Sodium dibenaothiophene-2-sulfonate
p
2,4 - niinethoxybenzylidenefluorene-2-sodium sulfonate
0,0773
8.11
8.46
12.22
W t . of Sample, Gram 0.0553 0.0492
Sodium m-(2-methoxyethoxy-) benzoate
0.0458 0.0458
10.29 10.29
Sodium p - (2 benzoate
0.0481 0.2000
9.74 9.75
9.74
9.79
9.90
0.0600 0.0654 0.0529
8.33 8.65 8.57
8.51
8.38
8.32
0.0691 0.0554
8.76 8.84
8.80
8.94
8.84
0.0766 0.0398
10.61 10.53
1 0 . 3 7 10.47
10.54
0.0471 0.0562
9.92 9.99
9.95
9.91
9.90
0.0510 0.0539
8.96 8.91
8.94
8.96
8.84
0.0549 0.0538
8.99 9.05
9.02
8.98
8.84
Sodium 0-(2'-ethoxy-2-ethoxyethoxy-) benzoate
0.0536 0.0571 0.0593
8.52 8.58 8.61
8.67
8.32
8.33
Sodium m-(2'-ethoxy-2-etboxyethoxy-) benzoate
0.0536 0.0554 0.0624
7.89: 7.94
7.95
8.25
8.33
Sodium p-(2 -ethoxy-2-ethoxyethoxy-) benzoate
0,0571 0.0534
8.27 8.35
8.31
8.38
8.33
Sodium m-(2-phenoxyethoxy-) benzoate
0,0520 0.0546
8.22 8.06
8.14
.. ,
8.21
Sodium m-(2'-butoxy-2-ethoxyethoxy-) benzoate
0.0782 0.0853
7.54 7.52
7.53
7.50
7.56
Sodium 0-(2'-butoxy-2-ethoxyethoxy-) benzoate
0.0552 0.0598
7 . 1 6 e 7.20 7.24e
7.41
7.56
-
ethoxyethoxy-)
- butoxyethoxy-)
Sodium-o-(Z-methoxyethosy-) benzoate
-
- ethoxyethoxy-)
-
- butoxyethoxy-)
Sodium o (2 benzoate
8.40
-
Sodium p-(2'-ethoxy-Z-ethoxyethoxy-) benzoate Sodium p (2 benzoate
12.22
Sodium o (2 benzoate Sodium m benzoate
- (2 - butoxyethoxy-)
8.51
EXPERIMENTAL
Reagent grade sodium chloride and lithium nitrate yere used in the preparation of standard solutions. Standards for the concentration range from 20 to 30 p.p.m. of sodium, and for the concentration range of 5 t o 15 p.p.m. were prepared. All the standards contained 100 p.p.m. of lithium. The concentration range used in the determination must be chosen with some care. Bills ( 3 )has shon-n that dilute solutions show less enhancement and depressant effects than do most concentrated solutions. With the Perkin-Elmer 52A instrument, the calibration curve is more nearly linear using dilute standards, and the instrument is more sensitive in the dilute ranges. Concentrations of salts higher than necessary for the determination may tend to clog the burner. When very dilute solutions are used, the signal-to-noise ratio becomes objectionable, so, in practice, a compromise between opposing factors must be reached. At first, a concentration range of 20 to 30 p.p.m. of sodium was used. -4s the work progressed, it became evident t h a t greater sensitivity was obtained by using solutions of the order of 10 p.p.m. of sodium, so a standard working curve of from 5 to 15p.p.m. of sodium was used. In the preparation of working curves, the highest standard and yater were alternated until equilibrium was reached, and reproducible internal standard dial readings were obtained. Then the highest standard and the nest highest standard n-ere alternated in the same manner, and the process was repeated until the complete curve was obtained. The calibration curve was checked a t each working period, since there vias considerable vertical displacement of the curve from day to day. Because unknown concentrations were determined by a difference reading from the
% Av. 76 % Kab 3 ' , NaC Xad i i a a (Mean) (M.R.) (Calcd.) 10.64 10.71 10.52 10.54 10.78
Compound Sodium p-(2-methoxyethoxy-) benzoate
a
10.28 10.58 10.54
8.016
Found b y flame photometer measurement. Mean of per cent sodium found by dame photometer. Mean of per cent sodium found by metal residue measurements. Per cent sodium calculated assuming compounds to be pure. Sample was a low melting wax, a n d could not be dried in the oven.
nearest standard, it was not necessary to adjust the curve to coincide precisely Kith the plotted curve, but the gain was manipulated so that there was never a large displacement from the plotted curve. The difference between the reading for the sample and the nearest standard was determined, and this difference was applied to the calibration curve counting from the nearest standard. The samples were dried overnight a t 110" C. and cooled in a vacuum desiccator over calcium sulfate. Tn-o samples vere low melting waxes which could not be heated; these are so marked in the table of data. The dried samples were weighed, transferred to a volumetric flask, and diluted to the mark. A proper aliquot was further diluted to the chosen sodium concentration range and lithium was added in an amount that would give 100 p.p.m. in the diluted aliquot. The weight of sample taken, the aliquot used, and the results obtained are given in Table I, together with the calculated per cent sodium, and the results of the metal residue analyses. The per cent sodium found in each aliquot is an average of a t least three readings. These readings usually agreed to i 0.05%. DISCUSSION OF RESULTS
Metal residue determinations were obtained on 26 of the 30 sodium salts vihich were analyzed with the flame photometer. Good agreement (&0.20'-% actual error) \vas obtained in 19 of these cases. In 12 of these 19 cases, the results agreed t o better than 4~0.10%actual error. I n no case did the flame photometer results prove unreliable, if the results are weighed against both the theoretical percentage and the metal residue method. Excellent agreement was obtained with the calculated percentage of
ANALYTICAL CHEMISTRY
1516
ACKNOWLEDGMENT
sodium iri the four cases whwc maid rcriduc determination e i n t l d not he medc. A thorough ~-xaminationof Tnhlr I shom that tho flanie photometric determinations compare favorably with tho rcsultfi 011taincd by the accepted mctal residue method. It cannot bc mid that the results obtainod with the flame photometcr prove that this method is applicable t o all sodium salts, but i t seems reasonable to assume that the surface active salts chosen for this iT-ork do provide a good test of the method. The data 011 nll sodium salts tested are supplied in this paper; no sample i n s discsrdvrl because the method wan innpplica.bl.hlo. Finally, the flame photometric dctarmination of sodium in salts of organic acids m5ll be of interest to organic chemists because oi the rapidity of flame phot,omet,ryand tbesmall amount of samplc consumed. Much larger samples than were necessary were used in order t o obtain stock solutions for use in det.ermining optimum experimental conditions.
:,., .-
. .
...
..,
.
.
The authors' thanks mi!due to the Atomic Energy Commission for financial support in pursuing this project. LITERATURE CITED
R. C'., S . .-lli.ii~mJ . Med. S c i . . 14, 103-iO (1949). ( 2 ) Iiervy. J. \I-., (!hnppell. D. C., a i d ILwncs. I(. H.. IND.EXG. C n e ~ . AN.,,.. . En.,18, 19-24 (1940). (:?) Hills. C . E., Mc.Donald. F. G . . Niederniiei.. W..and Rohwarta. 3'1.C . , bar.. C ~ m r . 21, , 10i6-80 (1949). (4) Rost. R. W., arrd cmr-orkem. University of North Csroliim, (I)Beinstein.
1942. (7) Parks, T.11.. .Johnmi. H. U.. and Lykken, S22-5 (1948).
L.,ANA?..C x e x . , 20,
. . .
,
. .
'9'
.
60. Pseudotropine ('ontrihuted by ROBERT L. CLARKE, Sterling-Winthrop Researoh Institute, Rensselaer, N. Y., AND JOHN KRC, JR., Armour Research Foundation of lllinoir Institute o f Toohnnlngy, Chicago 16, Ill.
H,C---C I
I
FI---ClT?
SCHI
I
CHOH
Structural F o r m u l a for I'seudotropine
~copeslide from acetone, ethyl alcohol, ether, water, and chloroform either result in very poorly formed crystals or yield amorphous pseudotropine. Figure 3 is an orthographic projection of a typical tablet of pseudotropine from Skelly solvent C. No polymorphs were observed in these studies.
T X C E L L E N T crystals of pseudotropine can be obbnincd by on a iwnero scale from hydrocarbon solvents such as Skelly solvent C or IPI Xo. 0-421) (International Printing Ink Solvent). Tropine and pscudotropine a ~ s t e r e o isomem. TVit,h careful tcehniqua, good crystals can also bi obt.ained on n microscope slide from theae same solvents (Fig ure 1). The he8t crystals for optical and x-ray analysis, however BE obtained by sublimation from microscope slide t.a cover glss (Figure 2). .ittempts t,o recrystallize pscudotropine on a micro
E, rccrystnllization
Crystals of Pseudotropine from P I Solvent No. 0-429
CRYSTAL M O R F H O W G Y Crystal System. Orthorhombic. Form and Habit. Tablets, rods, or needles lying on prism I 110) showing macropinamid IIOO], bipyramid I111 1, maerodome l1011. and hraebvdome lO111. Arid1 R&o. a:h:e = 0.888 :I: 0:760. Interfacial Angles PO^). n0-A inn = 41'35'; 110ai1o = 0 6 ~ 5 0 ' i; o i ~ i n = i 860;n i i ~ n i = i 74°30'. Cleavage. Excellent parallel to I110); good parallel to I100); U O O uarellel ~ to inini. S-R~Y DIFFRACTION DATA Cell Dimensions. a = 0.36 A,; b = 10.55 A,; c = 8.020 A. Formula Weights per Cell. 4 (3.W calculated from x-ray data). Formula Weight. 141.2. DenRity. 1.175 (flotation in srlene and ethylene dibromide); 1.183 (x-ray),
