566 HARRISON AND HEAYSMAN : METHOD 5 FOR DETECTION, DETERMINATION [Vd. 86 Methods for the Detection, Determination and Identification of Quinizarin in Hydrocarbon Oil BY R. B. HARRISON AND L. T. HEAYSMAN (Department of Scientific and Industrial Research, Laboratory of the Government Chemist, Clement’s Inn Passage, Strand, London, W.C.2) A colorimetric method for detecting quinizarin and a spectrophotometric method for its identification and determination in concentrations from 0.2 to 2.0 p.p.m. in hydrocarbon oil are described. IN connection with the marking of heavy oils to distinguish between duty-paid oil for road vehicles (DERV) and duty-free oil for other purposes, quinizarin was chosen as one of the markers, as it fulfilled the necessary conditions stated by Harrison, Palframan and Rose in their paper on furfura1dehyde.l DETECTION OF QUINIZARIN Quinizarin is 1,4-dihydroxyanthraquinone ; a search of the literature2 s3 14 produced the information that it is soluble in ethanol, ether, benzene, alkali or sulphuric acid and gives a blue colour with aqueous solutions of alkali, alkali carbonate or ammonia, but no method of determination was found.A rapid test for quinizarin in solution can be carried out by dipping the end of a chromato- graphic column of alumina into the solution. Quinizarin is adsorbed in a band well defined by the formation of a red lake. With a solution in gas oil, however, the combined effects of the colour of the oil and the low concentration of quinizarin made this technique of little value. The method described below for the roadside detection of quinizarin depends on the colour produced in alkaline solution.When a gas oil containing 2 p.p.m. of quinizarin is shaken with an aqueous solution of sodium hydroxide of any concentration between 1 and 30 per cent. w/v, quinizarin is extracted from the oil and produces a blue-violet colour in the aqueous phase. Stability of the colour increases with increasing concentration of sodium hydroxide, and is sufficient for a visual test for up to 30 minutes when a 1 per cent. w/v solution of alkali is used. METHOD REAGENT- Sodium hydroxide solution, 1 per cent. w/:v, aqueous. PROCEDURE- Pour 15 ml of the gas oil into a stoppered test-tube (6 inches x 0.75 inch), and shake for approximately 15 seconds with 2 @ of the sodium hydroxide solution. Allow the dis- persion to separate into two layers (this may take 2 or 3 minutes), and observe the colour of the lower aqueous phase; a blue-violet colour indicates the presence of quinizarin.The test works well for the chosen concentration of marker (2 p.p.m.) and for a dilution of one part of marked oil with one part of unmarked oil. For a dilution of one part of marked oil with four parts of unmarked oil, the colour is not easy to see in a pale oil and impossible to see in a dark oil. The sensitivity of this test. can be increased by using a larger volume of oil, but the same volume of sodium hydroxide solution. DETERMINATION OF QUINIZARIN An attempt was made to utilise the blue-violet colour produced with alkali as the basis of a quantitative spectrophotometric method, but several difficulties were encountered.Straightforward extraction of quinizarin from a marked gas oil by an aqueous solution of sodium hydroxide, with n-butyl alcohol to prevent emulsification, resulted in a solution that was not optically clear, owing to entrained droplets of oil. Acidification of this extract, subsequent extraction with diethyl ether, chloroform or light petroleum and then re-extractionSeptember, 1961 J AND IDENTIFICATION OF QUINIZARIN IN HYDROCARBON OIL 567 from the organic solvent with aqueous sodium hydroxide did not solve the problem. Centri- fugation of the separated aqueous layer gave an optically clear solution, which, when spectro- photometrically examined, was found to have an absorption maximum at 560 mp, but the colour was not completely stable; further, different gas oils gave different colours varying from blue to blue-violet.During work on the problem of positive identification, the ultra-violet - visible spectrum of a 0.001 per cent. w/v solution of quinizarin in cyclohexane was plotted with an automatic recording spectrophotometer. A characteristic curve was produced having five peaks a t 463, 476, 487, 508 and 521 mp; those peaks at 508 and 521 mp were extremely well defined. It was found that this method could be applied to the determination of quinizarin in marked samples of gas oil, thereby avoiding the difficulties mentioned above and giving results of greater accuracy than those obtained when the coloured aqueous solution was examined spect rophot ome t rically .METHOD REAGENTS- Sodium hydroxide solution, 5 per cent. wlv, apeow. Hydrochloric acid, concentrated. Cyclohexane, spectroscopically pure. n-Butyl alcohol. PROCEDURE- Shake 50.0 ml of the gas oil suspected to contain quinizarin with 5 ml each of the sodium hydroxide solution and n-butyl alcohol in a 100-ml separating funnel for 45 seconds. When the two layers have separated (this may take several minutes), run the aqueous phase into a second 100-ml separating funnel, and wash it through with about 2 ml of distilled water. Extract the gas oil with a further 5ml of the sodium hydroxide solution by shaking for 45 seconds, allow the layers to separate, run the aqueous layer into the second separating funnel, and wash through with water as before.Repeat with a further 5 ml of the sodium hydroxide solution, and, after separation, add the aqueous layer to the previous extracts. Quinizarin forms a blue-violet colour with sodium hydroxide, and the third extract should be colourless; if it is not, extract with further 5-ml portions of the sodium hydroxide solution until a colourless extract is obtained. Acidify the combined extracts by adding 1 ml of concentrated hydrochloric acid for each 5 ml of sodium hydroxide solution used. When the solution has cooled, add, by pipette, 10.0 ml of spectroscopically pure cyclohexane, and shake for 30 seconds. After separation of the two layers, run the aqueous layer to waste, and use a clean dry pipette to transfer the amber-coloured cyclohexane solution to a stoppered parallel-sided 1-cm cell of an Optica automatic recording spectrophotometer.With pure cyclohexane as blank, plot the absorption spectrum from 420 to 540mp; use the peak at 521 mp for determining quinizarin, as this is the most pronounced peak and is also furthest away from any possible interference from components of the gas oil. Prepare a standard curve by using a 0.001 per cent. w/v solution of recrystallised quinizarin in spectroscopically pure cyclohexane. DISCUSSION OF THE METHOD Complete absorption due to components of the gas oil carried over in the separations usually occurs at wavelengths shorter than 420mp, but there is no interference with the five peaks produced by quinizarin. Since the quinizarin used to mark gas oils is of the quality normally available on the market and not specially purified, a more realistic result is obtained if an average absorption figure is taken as standard. Accordingly, 0.01 per cent.w/v solutions of quinizarin from different sources were prepared in cyclohexane and used to mark a gas oil at the level of 2.0 p.p.m. The 0.01 per cent. w/v solutions of quinizarin were also diluted to 0.001 per cent. w/v, and these solutions were used directly in the spectrophotometer. The aliquots of marked gas oil were extracted by the proposed procedure, and the optical densities of the cyclohexane extracts at 521 mp were compared with those of the 0-001 per cent. w/v solutions of quinizarin; the results were- Quinizarin sample No. . . .. .. .. .. 1 2 3 4 Optical density of 0.001 yo w/v solution in cyclohexane .. 0-244 0.240 0.235 0-253 Optical density of quinizarin extracted from gas oil . . 0.243 0.240 0.235 0.253568 HARRISON AND HEAYSMAN METHODS FOR DETECTION, DETERMINATION [Vol. 86 from which it can be seen that the extraction process is satisfactory. The mean optical density of the 0.001 per cent. w/v solutions of quinizarin is 0.243, and there is a Linear relationship between optical density and concentration of quinizarin. In Table I are Listed the results obtained when the proposed method was applied to different oils marked at the level of 2.0 p.p.m. of quinizarin; the mean figure of 0.243 for the optical density at 521 mp was taken as being equivalent to 2.0 p.p.m. of quinizarin. A pale oil has a colour equal to or less than NPA 1, and a dark oil has a colour equivalent to NPA 3.5 TABLE I RECOVERY OF 2 P.P.m.OF QlJINIZARIN FROM VARIOUS OILS Oil No. Colour of oil 1 Pale 2 Dark 3 Dark 4 Pale 5 Pale 6 Dark 7 Dark 8 Dark 9 Dark 10 Dark 11 Pale Optical density at 521 m p 0.237 0.253 0.258 0.237 0.237 0.248 0.253 0,253 0-257 0.257 0.237 Quinizarin content found, p.p.m. 1.95 2-08 2.12 1-95 1-95 2-04 2.08 2.08 2.1 1 2.1 1 1.95 DISCUSSION OF RESULTS The repeatability of the method is within 1 per cent. The result of an extraction from a pale oil is within 2 per cent. of the correct amount, but the result of an extraction from a dark oil may be high by up to 8 per cent. This was found to be caused by absorption from components of the gas oil carried over in the separations. All gas oils absorb slightly in the range 460 to 540mp, and the darker the oil, the greater is the absorption. In the laboratory, where the unmarked oils were available for blank determinations, a correction could be applied, and this brought all the results to within 2 per cent.of the correct amounts. However, in practice, unmarked oil is not available, and no method of producing a satis- factory blank from the marked oil has been found. When this is taken into account, together with the fact that samples of quinizarin from all the manufacturers in the United Kingdom varied in purity by 4 per cent. from the mean value, the proposed method will in the worst circumstances give a result within 10 per cent. of the amount present; for a pale-coloured oil, the result will be well within 5 per cent. of the amount present.During the extraction, the colour in the sodium hydroxide solution may change from blue-violet to blue or fade completely, but this does not affect the determination. The method is satisfactory for concentrations of from 0.2 to 2.0 p.p.m. of quinizarin in gas oil and therefore can be used for a mixture of one part of marked oil diluted with nine parts of unmarked oil. The stability of quinizarin at a concentration of 2 p.p.m. in gas oil is good; determinations at the time of preparation of a dilution and 3 months later gave identical results. IDENTIFICATION OF QUINIZARIN The five peaks at 463, 476, 487, 508 and 521 mp produced by the spectrophotometric analysis are used to identify quinizarin. Other dihydroxyanthraquinones absorb in this region, but they can be easily distinguished from one another.The 1,2- and 1,5- isomers have only two peaks in this region, and these occur between 400 and 440 mp. The absorption spectrum of the 1,8- isomer is similar to that of quinizarin; it has five peaks, each occurring at a wavelength about 50mp shorter than the corresponding peak for quinizarin, but the peaks near 508 and 521 mp are not at all well pronounced, as they are for quinizarin. Meek and Watson3 reported on the spectrophotometric curves of six other polyhydroxyanthra- quinones, none of which can be confused with that of quinizarin. It is possible to see the typical structure of the spectrophotometric curve even when the extraction is made from a gas oil containing 0.2 p.p.m. of quinizarin, i.e., a dilution of one part of marked oil with nine parts of unmarked oil.September, 19611 AND IDENTIFICATION OF QUINIZARIN IN HYDROCARBON OIL 569 This paper is published with the permission of the Government Chemist, Department of Scientific and Industrial Research. REFERENCES 1. 2. 3. 4. 5. A.S.T.M., Tentative Method D155-45T. Harrison, R. B., Palframan, J. F., and Rose, €3. A., Analyst, 1961, 86, 561. “The Merck Index of Chemicals and Drugs,” Seventh Edition, Merck & Co. Inc., Rahway, New Meek, D. B., and Watson, E. R., J. Chem. Soc., 1916, 109, 544. Radt, F., Editor, “Elsevier’s Encyclopaedia of Organic Chemistry,” Series 3, Elsevier Publishing Received April 18th, 1961 Jersey, 1960, p. 893. Co., Amsterdam, 1946, Volume XIII, p. 527.