1227
V O L U M E 26, NO. 7, J U L Y 1 9 5 4 An interference study was made employing concentrations of diverse ions 100 times that of bismuth. Oxalate and phosphate prevented the normal response of bismuth to the reagent and cadmium, zinc, mercury, silver, and tin interfered by color masking of the bismuth-dithizone reaction. Attempts to sequester interferences through the use of conventional masking agents such as cyanide, thiocyanate, fluoride, and chloride met with some success. Cyanide ion and chloride ion effectively masked the interferences of mercury and silver, respectively, without an appreciable reduction in the sensitivity of the test.
Table I .
Cation
-
IIg S Zn T - -' T '
co*Sb++31n'ce+-+"go+- +
+: ;
+ Fe++Ph++ Cd++
Reactions of Dithizone-Naphthalene Solution with Various Cations Color Developed Yellow-orange Dull red Pink Red Orange-red Pale pink Red-brown I'eiiolv-orange Yellow-brown Brown-violet Orange-red Dull red Red
Time Required, Seconds
< 10 < 10 < 10
Color of Dithiaonate in Chloroform Yellow-orange Purple-red
....
> 80
Violet
> 60 >eo < 10
Vio1et:b;own None Yellow None Violet
60 >60 > 60 > 60
Red Red
< 10
....
the bright red coloration does not develop in 10 seconds, a negative test for bismuth should be reported. If mercury is known to be present, the dry sample may be taken up in a small quantity of water and a few milligrams of sodium cyanide added to the solution, before evaporation, in order to obtain intimate mixing and to prevent the interference of that ion. If silver is present, sodium chloride may be utilized similarly. REMARKS
The reagent as used in the manner suggested here offers a very sensitive means of detecting bismuth, definitely more sensitive than the chloroform-dithizone extraction procedure. During the study of the reactions of molten 8-quinolinol with metal salts Feigl and Baumfeld ( 1 ) also noted greater sensitivity than when the reagent was employed in an aqueous reaction. It is pointed out in Table I that several cations react with the reagent, some of them quickly and with the formation of strong colors. Among these the most sensitive are: cadmium, 0.027; mercury, 0.017; tin, 0.004-y; and silver, 0.027. Where these ions are known not to exist in cornhination, or n-here separations' can conveniently be made, the dithixone-naphthalene reagent may be useful in their detection. ACKNOWLEDGiMENT
This research was supported i n xhole or in part by The United States Air Force under contract number AF lS(600)-960, maintained by the Ofice of Scientific Research, Air Research and Development Command. The authors wish to express their appreciation for this support.
REC0311lEhDED TEST PROCEDURE
Iiito the depression of a white spot plate which has been previously heated to about 90" C. on a hot plate, a small amount of the solid sample is dusted (if the sample is in solution, a drop of the test solution is added and evaporated) and a few crystals of the dithizone-naphthalene mixture are added. I n the presence of bismuth the melt turns a brilliant red immediately: if
LITERATURE CITED
(1) Feigl, F., and Baumfeld, L., Anal. Chim. Acta, 3, 15 (1949). (2) Fischer, H., 2. angew. Chem., 42, 1025 (1929). (3) West, P. W., and Granatelli, L., -4~~1.. CHEW,24, 870 (1962).
RECEIT-ED for review Sovember 27, 1953. Accepted April 8, 1951.
Determination of Fumaric, Malic, and Succinic Acids in Fermentation Broths NIKOLAJ LEMJAKOV P i e d m o n t College, Demorest,
Ga.
Editor's Note: Professor Lenijakov Fas in the laboratories of Heinrich Knobloch, Institute of Technology in Prague, during 1943-45, as head of a research group. Results of his work on fumaric acid fermentation mere to he the basis of his doctoral dissertation,
but he was forced to leave Czechoslovakia before the thesis could be completed. What is published here is a brief summary of a portion of Lemjakov's work, reconstructed from memory.
S
T.41JDARD European methods for the determination of acids i n fumaric acid fermentation samples are inconvenient. For example, the quantitative determination of fumaric acid usually necessitates an initial 3-day extraction with ether, folloxed by application of the mercurous fumarate method. As i t was not feasible to use such elaborate and lengthy proceduies for large scale tests, attempts were made to simplify the annlytical process. I t was found possible to obtain reliable analytical results for fumaric acid by applying the mercurous fumarate method directly to the fermented substrate without extraction. The differences in the results of the analysis of samples of a substrate and of other samples of the same substrate to which \?ere added liriown quantities of fumaric acid corresponded to the amounts
of added fumaric acid. Large scale practical application of this technique Tms satisfactory. The results were further checked by polarographic determination of the fumaric acid; concordant results were obtained by both methods. The polarographic method appears to be the more accurate as well as the more rapid method, and requires less sample. The mercurous nitrate method requires 10 ml., whereas 2 ml. are sufficient for the polarographic method. Preliminary experiments showed fumaric acid to be destroyed quickly by permanganate in cold acid solution, but malic and sucrinic acids were not. Malic acid is often present in a fumaric acid fermentrtion. Recalling the titration of oxalic acid with permanganate and the building of catalytically active intermediary products, the author conducted an experiment with a mixture of fumaric and malic acids; complete oxidation of both acids occurred. On the basis of these observations a new method for the determination of succinic acid TVBS devised, which eliminated the extraction necessary in the usual methods. SUCCINIC ACID DETERMINATION
Transfer 10 ml. of fermented substrate to a porcelain dish. Add 2 Inl. (or an excess) of concentrated nitric acid (nitric acid is
ANALYTICAL CHEMISTRY
1228 used because of its volatility). Slowly add a concentrated solution of potassium permanganate with a pipet, stirring constantly, until the color disappears approximately 1 minute after the drop is added. When the oxidation is completed, add an excess of pure calcium carbonate to neutralize the acid and place the dish on a steam bath. (Complete neutralization and thorough drying are essential.) Add to the dry precipitate on the steam bath 10 ml. of distilled water, 10 ml. of standard 0.1N silver nitrate solution, and 15 ml. of 50y0 ethyl (or methyl) alcohol. Cool and filter the contents of the dish through a paper filter. Wash the precipitate several times with 50% alcohol, adding the washings to the filtrate. Add ferric sulfate indicator to the latter and titrate the excess silver ion with standard 0.l.V ammonium thiocyanate solution. The volume of silver nitrate consumed is used to calculate the quantity of succinic acid. The results of control tests in TThich various amounts of fumaric and malic acids were added to 40 mg. of succinic acid Ehowed that these acids did not interfere. I n each case the determination indicated the presence of approximately 40 mg. of succinic acid; the greatest deviation was never more than & I mg. and was probably due to incomplete drying and washing of the precipitate. DETERMINATION OF MALIC ACID
Transfer 10 ml. of the fermented substrate to a porcelain dish on a steam bath, neutralize with an excess of calcium carbonate, and evaporate to dryness. Add 10 ml. of distilled water, 10 ml. of standard 0.1N silver nitrate solution, and 15 ml. of 50% alcohol. After cooling, filter and wash the precipitate with 50% alcohol. Add ferric ion indicator and titrate with standard 0.1.V thiocyanate solution. The titration gives the sum of all three acids (malic, fumaric, and succinic); calculation is made on the basis of fumaric acid, since the molecular weights of the three acids are similar and fumaric acid occurs in far greater quantities in the mixture. The
sum of the fumaric and succinic acids as previously determined is subtracted to obtain the malic acid. PRESENCE OF RIBOFLAVIN I N FUMARIC ACID FERMENTATION
As indicated, the direct determination of fumaric acid in fermented substrate gives satisfactory results. However, crystals of mercurous fumarate obtained from pure fumaric acid solution are colorless, while those from the fermented substrate range in color from yellow to orange. The supernatant liquid is decolorized by the crystallization of mercurous fumarate, indicating that some dye is absorbed by the crystals of mercurous fumarate. The technical glucose in the tablets (Tveighing 6 to 8 pounds) used for these experiments is brownish, but of a diffprent shade from that of the fermented substrate. Precipitation from nonfermented substrate after the addition of pure fumaric acid gave colorless crystals. Thus, i t seems likely that some dye is produced in this fermentation. Crystallization of the mercurous fumarate occurs i n a solution containing 10% nitric acid; this concentration of acid a t 100" C. (steam bath temperature) destroys most natural dyes except riboflavin. Analytical tests on numerous samples of fermented substrates showed riboflavin to be present in quantities from 4 to 8 y %, the amount corresponding to the quantity of fumaric acid produced and to the deepening of the color of the fermented substrate. ACKNOWLEDGMENT
The author wishes to express his gratitude to Heinrich Knobloch, in whose laboratory a t the Institute of Technology in Prague the work described was carried out in 1943-46, and to Elizabeth Deseive for the help given by the determination of the riboflavin. RECEIVED for review December 21, 1953.
Accepted April 13, 1954.
Chromatographic Separation of p-Phenylazophenacyl Esters on Silicic Acid ROBERT M. IKEDA, A. DINSMOOR WEBB, snd R. E. KEPNER Department of Chemistry and Department of Viticulture, University of California, Davis, Calif.
T
H E separation and identification of small amounts of fatty acids either in the free state or from the hydrolysis of esters have been necessary in investigations of the identity of flavor materials in grapes and wines which are under way in this laboratory, The chromatographic separation of the acids as pphenylphenacyl esters often leaves much to be desired, as the bands are invisible under ordinary light and must be viewed in a dark room under ultraviolet light, \There they exhibit a pale blue fluoresence. The bands are often difficult to distinguish and to separate, particularly LTith the compounds of higher molecular weight. Highly colored p-phenylazophenacyl esters have been reported by Masuyama ( 2 ) and by Sugiyama ( 4 ) . The chromatographic separation of a number of these bright orange colored p-phenylazophenacyl esters on silicic acid is described in the present paper. Other attempts in this laboratory to separate these derivatives on alumina columns were not successful. MATERIALS
p-Phenylazophenacyl Bromide. p-Phenylazoacetophenone was prepared according to the method of Angeli ( 1 ) by the condensation of equimolar quantities of p-aminoacetophenone and nitrosobenzene in glacial acetic acid a t room temperature. Recrystallization of the crude product from ethyl alcohol gave orange crystals of p-phenylazoacetophenone, melting point
114.5' to 116' C. An equimolar quantity of bromine was added dropwise over a period of an hour to a stirred solution of p-phenylazoacetophenone in glacial acetic acid maintained below 20". The reaction mixture was poured into ice xvater and the precipitate was filtered, washed with water, and dried over calcium chloride in a vacuum desiccator. The p-phenylazophenacyl bromide was purified by chromatography on silicic acid using 50% benzene in Skellysolve B as the developing solvent by the technique described. Three small bands and one major band appeared as the column s a s developed. The major band was eluted to give p-phenylazophenacyl bromide, melting point 104-105' C. ilnalysis: calculated for C11HllhT20Br: carbon, 55.46; hydrogen, 3.66; found: carbon, 55.12; hydrogen, 3.68. The yield in the bromination step was 15%. Various attempts to brominate p-phenylazoacetophenone with Shrommuccinimide using a variety of conditions were unsuccessfJ1. ldsorbent. Silicic acid (Mallinckrodt, analytical reagent, S o . 2847, 100 mesh) was activated by pumping in a vacuum desiccator over phosphorus pentoxide for 8 minutes a t full capacity of a Cenco Hyvac pump. The activated adsorbent was stored in a tightly capped bottle until used. Developing solvents. Thiophene-free benzene and Skellysolve B were used without further purification. p-Phenylazophenacyl esters. The derivatives were prepared from the various acids and p-phenylazophenacyl bromide using the method of Shriner and Fuson ( 3 ) for the preparation of pphenylphenacyl esters. Each derivative was purified by chromatography on silicic acid as described.
