Determination of Thiophene by Oxidative Decomposition with Nitric Acid LEONARD S. LEVITT
EDGAR IIOW'ARD, JR.
AND
Temple Unicersity, Philadelphia 22, Pa.
HE purpose of this investigation has been to devclop B neiv 'method for the quantitative determination of thiophene, particularly in the presence of aromatic sulfides. It was felt that such a method should prove rather useful as a supplement to the widely used procedure of Ball ( Z ) , which, in the determination of various type sulfur constituents of petroleum products, makes no differentiation between aromatic sulfides and thiophenes, but piecipitates them together as the mercuric nitrate derivatives. Furthermore, the present method should be applicable regardless of the color or turbidity of the solution which contains the thiophene. This offers a distinct advantage over the various colorimetric methods which have from time t o time been proposed for the est,imation of thiophene in benzene, such as the isatin reaction (16), and Liebermann's test (fO),which have been found t o be unreliable for use on petroleum products ( 2 ) . The analytical procedures involving the precipitation of thiophene with mercuric salts ( 2 , S, 5 , I d ) are subject to numerous objections, among which are the fact that the precipitates are of indefinite composition ( 2 ) ; that mercuric salls react also with olefins ( 1 8 ) and with organic sulfides ( 2 ) likely t o be present; and that some organic compounds present in petroleum may react very violently with mercuric salts (13). The present method depends entirely on the exceptional instability of the thiophene ring to nitric acid, which has long been known to decompose thiophene and oxidize the sulfur to the trioxide (1, 8 ) . I t was thought by the authors that this reaction should proceed quantitatively in the presence of an excess of nitric acid t o the formation of sulfuric acid, and, indeed, it has been found t o do so. Thus 1 mole of thiophene gives rise to 1 of sulfuric acid, which may then be precipitated as barium sulfate in the usual manner and weighed as such. It would be of considerable interest t o know if all thiophene derivatives are decomposed and oxidized to sulfate by nibric acid. I n this connection it should be noted that 3,4-dibromothiophene, tetrahromothiophene, and 2-methyI-3,4,5-tribromothiophene (1), and 2,5-dihydrosythiophene and tetrahydroxythiophene (S) have all been observed t o decompose and give sulfuric acid on contact with concentrated nitric acid. I n similar qualitative tests, the authors have found that 2-tert-butylthiophene and 2,5-di-tert-wtylthiophene also exhibit this instability toward nitric acid, with resultant formation of large quantities of sulfate. The foregoing facts strongly suggest that the present method might very well be applicable to the quantitative determination in petroleum distillates of thiophene and its derivatives as a single group. On the other hand, those exceptional thiophene derivatives such as 3,4-diphenyl- and tetraphenylthiophene ( 7 ) , thianaphthene ( 9 ) , and dibenzothiophene ( 4 ) , which on oxidation form stable sulfones, might be expected in any complete analytical procedure t o fall into a different group. REAGENTS
Sitric acid, 8 N , shelf reagent diluted 1 to 1 M-ith water. Barium chloride solution, 0.1 to 0.5 M , Concentrated hydrochloric acid solution. Concentrated ammonia solution.
or the oxygen bornl) mrtliod on thc, originul sample, arid again after each of the following steps ~vhirhremoves a single sulfur type:
S -+Hg RSSR
HgS
Zn z PRSH HOAc *=OH E
2R&
(-2)
The laat step could probably be omitted without cauciiip anv interference in the subsequent analysis for thiophene, suice d i phatic as well as aromatic sulfides are oxidized by nitric acid only SO far as the suIfone3, which are very resistant toward furthrr oxidation (14). In fact, aliphatic sulfones may be prepared by oxidizing sulfides at elevated temperatures with fuming nitric acid (15). ANA LYTIC 4 L PROCEDURE
I n an all-glass apparatus consisting of a round-bottomed flask and water-cooled reflux condenser is placed about 100 nil. of 8 S nitric acid, followed by a quantity of sample large enough to give about 0.6 gram of barium sulfate. The resulting two-phase system is heated t o reflux temperature for 3 t o 6 hours during nhich time the aqueous nitric acid phase should go through the following color changes: colorless, yellow, orange, clear light yellow. More highly concentrated nitric acid, even the concentlated shelf reagent (16 N ) , may be used for low-thiophene samples, and in the case of very low- thiophene content, where closer intermixing might be desirable to speed the reaction, dioxane or acetic acid serves as an efficient mutual solvent for the aqueous phase and the thiophene-containing phase. The authors have found that dioxane apparently is stable even toward hot concentrated nitric acid. The organic layer remaining after refluxing, if of a light volatile nature, may be removed by heating on a steam bath, Higher boiling organic material may have to be separated froin the aqueous phase by means of a separatory funnel, the oil portion then being washed with distilled water and the washings added t o the water phase. The acidic aqueous solution is then cautiously neutralized n-ith concentrated ammonia, whereupon the yellow solution, conveniently serving as its own indicator, turns deep orange. Then concentrated hydrochloric acid is added until the yellow color returns, after which 1 or 2 ml. more of hydrochloric acid is added to ensure the proper acidity for the sulfate precipitation. The solution is then adjusted t o a volume of approximately 300 ml., heated to near boiling, and an excess of a hot solution of 0.1 to 0.5 M barium chloride added slowly to it. T h e precipitated barium sulfate is then digested on a steam bath and allowed t o stand until the precipitate becomes more coarse, after which i t may be filtered by means of a Gooch crucible. Finally the p ~ e cipitate is xmshed on the filter with small portions of boiling water until free of chloride ion, and then dried, ignited, and weighed. RESULTS AND DISCUSSION
PRELIMINARY TREATMENT OF SAMPLE
Stock solutions of thiophene in benzene and thiophene in cyclohexane were used as the samples for analysis. These solutions were analyzed by the oxygen bomb method and found t o give 1.20 grams of barium sulfate per 10 ml. of sample, which is equivalent t o a sulfur content of 1.85% by weight (calculated as S).
The sample, before analysis for thiophene, must be freed of all other type sulfur constituents except aromatic and aliphatic sulfides. These other sulfur compounds may be removed one a t a time and determined quantitatively by a modified Ball procedure, which consists of determination of total sulfur by the lamp method
196
V O L U M E 25, NO. 1, J A N U A R Y 1 9 5 3 Table I.
Analysis of Thiophene Solutions Containing 1.85% S
Saniple SO.
1 2
3 4
5
6 7 8 9 10
a
197
HNO,, N 4 8 8 8 8 8 8 8 8 8
BaSOa, Grams 1.20 1.18 1.20 1.13 1.28 1.24 1.25 1 22 L24 1.28
S Found,
%
1.85 1.82 1.85 1.74 1.97 1.91 1.93 1.88 1.91 1.97 1.89 1.84 1.97 1.89 1 86 1.93 1.70 1.97 1 93 1.89
During the course of this investigation it became necessary t o substantiate the fact that organic sulfides would not be oxidized t o sulfate by nitric acid. Accordingly, diphenyl sulfide and diethyl sulfide were each refluxed separately with 12 N nitric acid for 3 hours, after which time no sulfate could be detected in either case. Furthermore, these two sulfides vere each added t o separate samples of solutions of thiophene in cyclohexane (Table 11, samples 21 and 22) and the mivture heated at reflux tempeiature for 3 hours with 8 N nitric acid, hut in neither cme was there observed any additional sulfatc ascribable t o the oxidative decomposition of the added sulfide. ACKNOULEDG\IENTS
The many helpful suggestions of Hazel Tonilirison are gratefully acknowledged. Thanks are due also t o the Socony-Vacuum Oil Co. for its generosity in supplving thiophene and derivatives.
1 gram oxalic acid added. LITERATURE CITED
Table 11. Analysis of Low Sulfur Samples ”Os, Thiophene S Total S SO. N Present, % Found, %
Sample 210
8 22b 8 23 12 24 12 1 grain of diphenyl sulfide added.
0.37 0.36 0.37 0.38 0.18 0.16 0.18 0.17 0 . 5 gram of diethyl sulfide added.
I n a total of twenty analyses using these solutions the percentages of sulfur actually found ale given in Table I. Also included in this table are the normality of the nitric acid used and the actual weight of the barium sulfate obtained from 10 ml. of sample. A statistical analysis of these twenty deteiminstions reveals that the standard deviation is 0.049 gram, from which it may be calculated that the probable error is 0.033 gram. The arithmetic mean of the weight of the precipitates is 1.22 grams, differing by 1.7% from the value of 1.20 grams obtained using the oxygen bomb method. It may be concluded that the probable value for any single determination is 1.22 =k 0.03 gram, corresponding in per cent of sulfur t o 1.88 2 0.04%. [-1 recent report (11) on a colorimetric method for the determination of thiophene revealed a 15 t o 20% error using the photoelectric colorimeter and up t o a 30% error for the visual colorimeter.] Although it is generally considered desirable t o rid the solution of nitrate prior t o the precipitation of sulfate, no significant differences in results could be noted when the ienioval of nitrate by evaporation with small portions of concentrated hydrochloric acid was omitted. The error caused by such coprecipitation is smaller than the probable error of this method. I n this connection, it was recently found ( 6 ) , in the course of numerous determinations of sulfuric acid in the presence of nitric acid, that the high results due t o coprecipitation of barium nitrate did not exceed 1yo. It was deemed advisable t o test the present method in the analysis of samples of much lower sulfur content. Bccordingly, samples of thiophene in cyclohexane were prepared containing 0 . 3 i % and 0.18% sulfur by weight. The results obtained are given in Table 11. The oxidation of thiophene has been shown t o give rise t o some oxalic acid ( l 7 ) ,and in order t o preclude the possibility of the precipitate’s containing any barium oxalate which would subsequently be ignited t o the oxide and weighed along with the sulfate, 1 gram of oxalic acid Jvas added t o each of two runs, but, as may lie seen from Table I (samples 11 and la), no significant differenre in the weight of the baiium sulfate precipitate was observed.
Angeli, A , , and Ciamician, G., Ber., 24, 74, 1347 (1891). Ball, J. S.,U. S. Bur. M i n e s , R e p t . Invest., 3591, 50 (1942). Claxton, G., and Hoffert, TT’. H., J . SOC.Chem. Ind. ( L o n d o n ) , 6 5 , 333 (1946).
Cullinane, N. M., and Davics, C . G., Rec. trau. chim., 5 5 , 881 (1936).
Deniges, G., Bull. soc. chim., Fiance, [3] 15, 1064 (1896). Goddard, D. R., Hughes, E. D., and Ingold, C. K., J. Chem. SOC.,1950,2575.
Hinsberg, O., Ber., 48, 1611 (1915). Lanfrey, M., Compt. rend., 153, 73 (1911). Ibid., 154, 519 (1912). Liebermann, C., Ber., 20, 3231 (1887). McKee, H. C., Herndon, L. K., and Withrow, J. R., AXLL. CHEM.,20, 301 (1948). Meyer, A., Compt. rend., 169, 1402 (1919). Mixer, R. Y., Chem. Eng. News, 26, 2434 (1948). Noller, C. R., “Chemistry of Organic Compounds,” p. 2G8, Philadelphia, W. B. Saunders Co., 1951. Ibid., p. 274, Schwalbe, C., J. SOC.Chem. Ind. ( L o n d o n ) ,24,988 (1905). Sernagiotto, E., Atti accad. nazl. Lincei, 28, I, 432 (1919). Whitmore, F. C., “Organic Chemistry,” p. 33, New York, D. Van Nostrand Co.. 1951. RECEIVED for review September 13, 1951. Accepted September 26,
1952.
Abstracted from t h e thesis t o be siihmitted by Leonard 9. Levitt t o t h e Graduate Council of Temple Univeisity in partial fulfillment of the requirenients for the Ph.D. degree.
CORRECTIONS Food I n the review on “Food” [Natchett, J. R.. and von Loesecke, -4~7.4~. CHEM., 24, 242 (1952)l reference (61) should read: (61) Food Processing, 12 (IO), 46 (1981).
H. If7.,
Estimation of Sodium Triphosphate I n the article on “Estimation of Sodium Triphosphate” [Bell, Wreath, and Curless, ~ ~ N A LCHEU., . 24, 1998 (1952)] the last line in the first column under Procedure should read “betn-een 2G and 28 ml. of sodium hydroxide.”
Determination of Free Acid in the Presence of Hydrolyzable Ions I n the article on “Determination of Free Acid in the Presence of HydrolyzableIons” [Pepkowitz, Sabol, and Dutina, ANAL.CHEhr., 24, 1956 (1952)l the titles of Figures 2 and 4 have been interchanged. They should read : Figure 2. Conductometric Titration Cell. and, Figure 4. Complexing .-lction of Sodium Fluoride.