of both components of the mixture (D and S5). Because infrared spectra in Figure 5 show that the mixture and the ligitonin absorpthe postulated
digitonides as chemical compounds, whereas infrared spectrophotometry does not distinguish between a simple mixture and a corresponding digitonide. ACKNOWLEDGMENT
This work was supported in part by a grant from the Michigan Heart Association. tometry is of tification of Czr from biological a n be easily difTraction powder nides cannot be frared or x-ray ’he use of x-ray stablishes sterol
LITERATURE CITED
( 1 ) Dobriner, K., Katzenellenbogen, E.
R., Jones, R. N., “Infrared Absorption Spectra of Steroids,” Interscience, New York, 1053.
RECEIVEDfor review August 5, 1956. Accepted March 18, 1057.
nation of Vanillin
No difficulties :ffects were ensupplementing ,f Morton and tion to the alter,let absorption erahydroxyalde!n their acid sore made basic. iunced change. h e n applied by ysis of flavoring tntages resulted 5 solutions in a n absorption peak ped in a longer intensity of this ‘e was less inter:ous materials. .) have extended iures of vanillin stances. !effect of adding .Ikali upon the nillin solutions, the actual p H led. The peak rption was eviwas established, of 0.2.v potasml. final volume ‘as necessary for The absorption ditions had a n , I n acid solu-
tion no absorption was observed a t this wave length. The change in absorption with change in p H has been found to respond to pH effects in the same way as ordinary acid-base indicators. It is usually more convenient to work with aqueous than alcoholic solutions of extracts. Furthermore, lead acetate used as a defecating agent, and other substances, usually exercise buffer action, so that addition of a specific amount of alkali hydroxide may not yield the same pH a,t all times, or give maximum intensity for the basic form of vanillin. Where such a state of affairs exists, significant errors result. In this laboratory, some apparently anomalous results under the. conditions indicated were found to be caused by p H differences. I n following in the initial steps of the usual procedure, 2 ml. of a genuine extract was introduced into a 200-ml. volumetric flask and 2 ml. of the lead reagent was added. After the solution was made to volume and filtered, the recommended 1 0 d . portion of the filtrate was introduced into a 100-ml. flask and 2 ml. of 0.117’ sodium hydroxide was added. When the solution was made up.to 100 ml., a pI-1 of 11 was observed. For 25and 50-ml. portions of the filtrate, similarly treated, the p H values were, respectively, 10.6 and 8. The absorbancy of the last was markedly different from a value that would have been indicated if the pH had remained constant. VOL. 29, NO. 8. AUGUST 1957
-
1151
I n this instance the reduction in pH mas caused mainly by increase in the amount of the vanillin in the portion taken, but in commercial extracts the acidity of the resins and other normal constituents of the extract may have an unpredictable effect. Accordingly, 3 study was undertaken of absorption values for vanillin in a series of aqueous solutions, in which the p H was varied, so that more exact ronditions for the analysis could be qpecified.
Table I.
Equipment. X Cary recording spectrophotometer, Model 11, \\as used in preparation of t h e absorption curves. A Beckman Model 31 p H meter n i t h glass electrodes for both the near neutral and high alkaline ranges n-as used for p H measurements. Reagents. Standard Vanillin. -4 solution containing 0.1 mg. per ml. 15-3s prepared by dissolving 100 mg. of vanillin in 250 ml. of 95% ethyl alco1101 and diluting to 1 liter. Portions of this stock solution were used in t h e experiments. The final dilution was so great t h a t the amount of alcohol present was negligible and t h e absorbancy was t h a t of a n aqueous solution. Sodium hydroxide, 0 . 2 M . Potassium dihydrogen phosphate, 0.2M.
(Proportions of acid and basic forms and pK values) Constituents, afl./loo Ml. Ohrrved Valiies Calculated 0 231 0 2.11 -4 "*; Basic % Acidic KH?POa XaOH pH 347 m p form form 2 91 0 015 0 00 100 90.6 0 170 9 40 25 3 b 31 28 5 71 5 0 360 2-5 1U 6 72 36 0 64 0 0 610 25 15 7 07 35 4 64 6 0 600 25 15 7 OT 44 7 0 930 55 3 25 20 i 41 25 20 7 41 0 920 55 0 45 0 2R 7 72 1 ~ 20.5 . . 71 9 28 10 _. 23 70 76 3 22 7 79 1 277 25 16 30 1 400 83 7 ?5 23 7 00 9 80 1 510 90 2 25 24 6 37 .. .. 10 O3* 1.BTU 100.0 0.00 t
Espt. 1
2
3 4a b
5. b
EXPERIMENTAL
Ultraviolet Absorption of Vanillin ( 1 Mg. per 100 MI.) as Affected by pH Changes
A m
i9
10
~
PK
7 30 7 12 7 32 7 33 7 32 7.32 7.31 7 28 7 28 7 41 .
#
7.30 Std. dev. 0.23 ;ir.
pH adjustcld \ n t h HCI. SnOH.
* pH adjusted with
with sodium carbonate and a l l o ~ e dto stand in a partixlly filled stoppered glass bottle. 4 t intervals the bottle was opened and samples were withdrawn for ultraviolet analysis (Table 11). The solutions are completely stable for over 3 month and s h o ~almost negligible change after nearly 6 month..
DISCUSSION
The absorption curves for vanillin under acid and basic conditions are very siniilar to those observed by Lemon for the alcoholic solution. The absorption maximum for the basic form is shifted from 353 mp for the alcoholic solution to 347 mp for the aqueous so-
PROCEDURE
Solutions of pure vanillin, 1 mg. per 100 ml., were prepared with graduated increases in alkalinity, and the absorbancies were evaluated against a blank containing no vanillin as a reference solution. At p H above 10.5 an absorbancy (1-em. cell) of 1.67 a t 347 mp was the maximuin for the alkaline form. At p H below 5 , thc absorbancy for the acid form was a t a minimum (0.015). These values may be designated as those for the full basic and full acid forms a t 347 mp. Representative curves in which the p H was varied are shown in Figure 1 and pertinent data in Table I. If 5 is specified as the decimal fraction of full basic form and (1 - X) for the full acid form, the proportion of each may be calculated from the equation 1 . 6 7 ~ (1 - 2 ) 0.015 = A
2'o
+
\\here A is the absorbancy noted for 3 mixture of the two forms. If vanillin responds to pH change as an acid-base indication, over the transition interval basic form JJH = pK log acid form
+
where pK is the negative logarithm for the dissociation constant of the vanillin. The nature of the solutions employed and characteristic data are shown in Table I. It has been demonstrated ( 2 ) that the acid form of vanillin is susceptible to slow oxidation to vanillic acid by atmospheric oxygen. To study the stability of the alkaline form, a solution of vanillin containing about 10 mg. per liter was made basic to p H 10.75
1 152
ANALYTICAL CHEMISTRY
WAVELENGTH
mp
Figure 1. Effect of pH upon absorbancy of vanillin sohtions of a concentration of 1 mg. per 100 ml.
1. pH = 2.91
2. pH 3. pH
= =
7.0i i.41
4. pH = 7.99 5. pH = 10.93
lution. The EAE. for this maximum, calculated to be 1980 for a n alcohol solution, was 1670 for the aqueous solution. The p K value found for vanillin is 7 . 3 . Hence. it has a transition interval in a p H range slightly lower than that of phenolphthalein. It resembles the latter in being essentially a onecolor t!-pe indicator, as the acid form has almost a negligible absorption a t X 347 mp, where the basic form is a t its maximum. Three well-defined isosbestic points are evident at wave lengths 238, 259, and 316 mp. The fact that the absorbancies at the isosbestic points are not affected by p H changes offers some interesting possibilities in the examination of multicomponent systems. However, measurements a t these points require careful setting and checking of the wave length adjustment, as the absorption is markedly different a t wave lengths only slightly different from the proper values,
Table II. Stability of Vanillin Solution a t pH 10.75, Room Temperature
Date of Examination April 26, 1956 March 4, 1956 June 1, 1956 July 5, 1956 Oct. 15, 1956
Absorbance in 1-Cm. Cell 247 mp 347 mp 0.60 0.60 0.60 0.61 0.61
1.68
1.68
1.68 1.67 1.67
The observed’ differences in absorption may be concerned with a n equilibrium involving a change from a benzenoid to a resonance hybrid, to which the quinoid structure makes an important contribution, and analogous to the conduct of p-nitrophenol (6). H
O
0-
H
H
O
‘C’
0
nature and amount, use of larger TOTumes may cause trouble, and p H specification and check on its value is desirable. Other parahydroxyaldehydes may have dissociation constants of different magnitudes, and selection of specific p H conditions may be even more necessary for analysis of mixtures of these compounds. After this article had been prepared for publication, attention was called to a paper by Robinson and Kiang (9). in which the ionization constants of vanillin and two of itb isomers had been reported. The pK d u e reported for vanillin was 7.496.
The proportion of the two forms may be estimated for any pH value in the transition range from the equation pH
=
fill1 basic form i‘3 + log [full acid form
1
hssuming that one part of one form is detectable in the presence of 200 of the other, 9.6 and 5.0 represent the p H values a t the limits of the transition interval. Hence, p H values slightly greater than 9.6 and less than 5.0, respectively, are necessary for quantitative evaluations of the basic and acid forms, in order to secure theoretical conversion of 99.5%. values in excess of these figures may be advantageous to allow a margin of assurance for complete conversion. The use of a 10-ml. portion of the filtrate specified gives a p H of 11. If the extraneous materials in all genuine extracts are not of the same
LITERATURE CITED
(1) Englis, D. T., Hanahan, D. J., IND. ENG. CHEM., ANAL. ED. 16. 505 (1944).
(2) Englis, ’D. T., Manchester, Merle, ANAL.CHEIII.21, 591 (1949). (3) Englis, D. T., Wollermann, L. A . , Food Research 20, 567 (1955). (4) Ensminger, L. G., j.Assoc. O$c. Agr. Chemists 36. 679 (1953). Kolthoff, I. hi., Rosenblum, Charles, “Acid-Base Indicators,” p. 152, Macmillan, New York, 1937. Lemon, H. K., ANAL. CHEY.19, 84G ( 1947) * 17) Lemon. H. W,, J . Am. C h e w SOC.69, ’ 29O8’(1947): (8) Morton, R. A., Stubbs, A. L., J . Cheni. SOC.(London)1940, 1937. (9) Robinson, R. A., Kiang, A. K., Trans. Faraday SOC.51, 1398 (1955). RECEIVED for review March 31, t!M. .4ccepted February 7, 1957.
Analyses of a Chromatographic Fraction of Organic Extracts of Soils W. G. MElNSCHElN and G. S. KENNY Magnolia Pefroleom Co.,Field Research Laboratory, Dallas, Tex.
,An investigation of the chemical composition of a chromatographic fraction of the benzene-methanol extracts of soils i s described. The benzene eluates of the soil extracts from silica gel chromatographic columns contain high concentration of waxes of the types appearing in beeswax. Qualitative and semiquantitative analyses of the soil waxes are accornplished by use of chromatographic, infrared, and mass spectrometric methods in conjunction with hydrogenation studies. The types of acids and alcohols that form the wax esters are determined by converting the esters
to saturated hydrocarbons which are analyzed mass spectrometrically. The principal constituents of the waxes are normal aliphatic acids, normal primary aliphatic alcohols, and sterols. All the soil extracts contain pentacyclic and hexacyclic compounds which may b e triterpenes.
T
of the organic extracts of sediments are of interest because these extracts may be a source of petroleum, but the chemical nature of such extracts has not been extensively investigated because of their complexity. However, in recent years HE COMPOSITIOX
significant advances have been made in the analysis of organic mixtures. One of the more important factors contributing to these advances has been the modification of a commercial niass spectrometer to the analysis of high niolecular weight compounds by O’Seal and JJ7ier (6). The comprehensive review articles hy Dibeler (6) may be consulted for the accomplishments in the field of high mass spectrometry. This investigation of a chromatographic fraction of the organic extracts of soils illustrates how high mass spectrometric analyses may be used in conjunction with other study methods to deVOL. 2 9 , NO. 8 , AUGUST 1957
* 1153