166 Analyst, March, 1974, Vol. 99, $+. 166-167 A Comparison of Two Procedures for the Determination of Organobromine by the Schoniger Oxygen-flask Method BY RONALD C. DENNEY AND PHILIP A. SMITH (School of Chemistry, Thames Polytechnic, London, SE18 6PF) A statistical comparison of the argentimetric and mercurimetric methods of finish in the oxygen-flask determination of organobromine has been carried out. While no statistical difference was found to exist, the mercurimetric method is recommended for use in laboratories that lack the facilities required for the potentiometric determination. THE oxygen-flask method for the determination of organically bound bromine still follows the basic combustion method developed for micro-scale work by Schoniger.lS2 In all analyses, assuming that total combustion takes place, accuracy is dependent upon conversion of the organobromine into a suitable form for determination by a well established method of finish.Various methods of finish have been recommended for the determination of bromine in organic compounds by the oxygen-flask m e t h ~ d . ~ , ~ Iodimetric, argentimetric and mercuri- metric titrations have all been extensively used but controversy remains over their relative merits. There is also disagreement over the need to add hydrogen peroxide to the absorbing solution.3~5 Childs, Meyers Cheng, Laframboise and Balodis,6 for example, demonstrated that for bromine determinations absorption of combustion products is rapid if a solution of either hydrazinium sulphate or hydrogen peroxide in sodium hydroxide solution is used.We have, therefore, carried out a statistical comparison of the argentimetric and mercurimetric titra- tions, both with and without the use of hydrogen peroxide. EXPERIMENTAL APPARATUS- The oxygen-flask combustion unit manufactured by Thomas and Co., Philadelphia, USA., was used because of the safety features it incorporates. We have found the method involving ignition with an infrared lamp by remote control to be of particular value in intro- ducing inexperienced workers to oxygen-flask procedures. ABSORPTION SOLUTION- For each determination in the statistical study, the absorption solution consisted of 10 ml of approximately 0.1 M sodium hydroxide solution containing 0.3 ml of 100-volume hydrogen peroxide. After ignition, the flask was left to cool for 10 minutes before being shaken in order to assist absorption of the combustion products in the absorption solution.METHODS OF FINISH- Argentimetric jnish-Titration of the neutralised solution with 0.01 M silver nitrate solution was carried out potentiometrically by using essentially the procedure recommended for chlorine determinations by the Analytical Methods Committee.' For our determinations, the absorption solution was transferred from the combustion flask into a 100-ml beaker with 40 ml of distilled water, The pH of the solution was adjusted to between 6.0 and 6.2 measured with a glass - calomel pH electrode. The titration was carried out by using a glass - silver electrode system with the potential measured on a millivoltmeter.Mercwimetric jnish-The traditional method of titration of bromide ions with standard mercury( 11) nitrate, using diphenylcarbazone with bromophenol blue as a screen, gives poor results in aqueous solution. The modified procedure developed by ChengS and applied extensively by White: which involves the use of an 80 per cent. ethanolic solution, was used for our determinations. @ SAC and the authors.DENNEY AND SMITH 167 Following the combustion procedure the flask was washed with 10 ml of distilled water and acidified with 0.1 M nitric acid to the yellow colour of bromophenol blue; 100 ml of absolute ethanol and 0.5 ml of 0-1 M nitric acid were added and the pH was adjusted to 3-6. After addition of diphenylcarbazone solution (0.5ml of a 0.1 per cent. ethanolic solution), the solutions were titrated with 0.01 M mercury(I1) nitrate solution to the first appearance of a purple colour.RESULTS AND DISCUSSION All determinations were carried out on micro-analytical reagent grade 3-bromobenzoic acid (Br content 39.75 per cent.) with 10 to 20-mg samples. Corrections for blanks were applied to all results. An initial set of results obtained in the absence of hydrogen peroxide in the absorption solution were discarded as they were found to be irregular and valueless from an analytical point of view. This finding was in sharp contrast with Steyermark’s views,5 but confirmed those expressed by Schoniger,lV2 Macdonald3 and Childs et aZ.6 Argentimetric finish 39.68, 39-35, 39.78, 39.85, 39-57, 39.57, 39-86, 39.36, 39.13, 39.56 Mean .. . . 39.57 Standard deviation . . . . 0.237 Mercurimetric finish 40.00, 39.71, 39.56, 39-84, 39.75, 39-51 Mean .. . . 39.71 Standard deviation . . . . 0.202 The results obtained for bromine, per cent., were as follows: (ten determinations) (six determinations) STATISTICAL COMPARISON OF RESULTS- Student’s &testlo was used and for the above calculated standard deviations t was found to be 1.21. As the value of t obtained lies between + 1.76 and -1.76, there is no statistically significant difference between the two means. At this level, P = 0.2 to 0.3. From a comparison of precision by using the F distribution with the above values, F is 1.38. At the 5 per cent. level, Fo.05 is given as 4.77 and as the value found for F is less than 4.77, there is no significant difference in the precision of the means obtained by the two methods.CONCLUSIONS This investigation has shown that for argentimetric and mercurimetric finishes it is essential to include hydrogen peroxide in the absorbent solution. Our initial results were wholly unrelated to each other in the absence of hydrogen peroxide. The statistical analysis of the results supports Steyemark’s contention5 that there is no significant difference between the results obtained by these two methods. It is, therefore, apparent that even a poorly equipped laboratory that lacks the facilities for potentiometric titrations can obtain reproducible results for the determination of bromine by using the less sophisticated mercurimetric finish. Of the two methods it has, in any event, the great ad- vantage of being more rapidly carried out. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Schoniger, W., Mikrochim. Acta, 1955, 123, -- , Ibid., 1956, 869. Macdonald, A. M. G., in Reilley, C. N., Editor, “Advances in Analytical Chemistry and Instrumenta- tion,” Volume 4, Interscience Publishers, New York, London, Sydney and Toronto, 1965, p. 75. Belcher, R., Gawargious, Y. A,, Gouverneur, P., and Macdonald, A. M. G., J . Chem. Soc., 1964,3560. Steyermark, A., J . Ass. 08. Agric. Chem., 1965, 48, 709. Childs, C. E., Meyers, E. E., Cheng, J., Laframboise, E., and Balodis, R. B., Microchem. J., 1963, Analytical Methods Committee, Analyst, 1963, 88, 415. Cheng, F. W., Microchem. J., 1959, 3, 537. White, D. C., Mikrochim. Acta, 1961, 449. Spiegel, M. R., “Theory and Problems of Statistics,” Schaum’s Outline Series, McGraw-Hill Book Received July 29th. 1971 Amended October l s f , 1973 Accepted October 4th, 1973 7, 266. Co., New York, Toronto, London, Sydney and San Francisco, 1961.