452 Analyst April 1983 VoL. 108 pp. 452-456 Improvements to the Oxygen Flask Combustion Procedure for Assay of Halogenated Organic Compounds Duncan Thorburn Burns and Binod K. Maitin Department of Analytical Chemistry The Queen’s University of Belfast Belfast BT9 5AG An improved oxygen flask procedure is described for the accurate precise and rapid determination of chlorine bromine and iodine in halogenated organic compounds. Quantitative recoveries were achieved by using neutral hydrazine hydrate in the absorbing solution which may be titrated potentio-metrically directly with silver nitrate solution. Hydrazine hydrate does not interfere under acidic conditions. Accurate and precise results were obtained when the equivalence points were determined by Gran’s method.Keywords Halogen determination ; oxygen $ask combustion ; hydrazine hydrate ; potentiometric titration ; Gran’s method Several procedures for the complete decomposition of organic compounds have been described of which the oxygen flask combustion method is the most widely used. This method was devised by Hempell who applied it to the determination of sulphur in coal and organic materials. As Macdonald has noted293 few further applications appeared until SchOniger4s5 adapted the method to the micro-scale. Subsequently the method has been applied widely and a very large number of modifications and applications have been described and d i s ~ u s s e d . ~ ~ 9 ~ - ~ ~ In oxygen flask combustion problems may arise from the incomplete combustion of samples. Although filter-paper is the most common material used as sample carrier for solid samples, other materials such as cigarette paper polythene and poly(methy1 methacrylate) sheets have also been u ~ e d .~ ~ + Many workers add some auxiliary combustible material to assist complete decomposition. Glucose sucrose potassium or sodium nitrate paraffin wax naphthalene, phthalic acid benzoic acid and dodecanol have each been used as combustion aids.3*7,12J3 To overcome problems due to sublimation or volatilisation of certain highly halogenated aromatic compounds Awad et aL.12 recommended repeated folding of the paper carrier in all directions, so that several layers surround the sample. Complete absorption of the combustion products is an important factor in these determina-tions.In order to convert the liberated halogens into an ionic form it is normally necessary to use a reducing absorption medium. A large variety of reducing agents have been used for example neutral and alkaline hydrogen peroxide alkaline sulphur dioxide saturated aqueous hydrogen sulphite alkaline hydrogen sulphite alkaline sodium arsenite acidic sodium nitrite, alkaline sodium tetrahydroborate( 111) dilute ammonia solution ammoniacal peroxide and alkaline hydrazine ~ulphate.~$l~ Of these alkaline hydrogen peroxide was commonly used for the determination of chlorine and bromine but not for iodine.13 Alkaline hydrazine sulphate is now more favoured than alkaline peroxide and can be used for determination of chlorine, bromine and iodine.15-19 Detailed collaborative studies organised by Lalancette and co-workers showed that satisfactory results for the three halogens could be obtained using alkaline hydrazine sulphate.20,21 Mazzeo-Farina and MazzeoZ2 have suggested the use of alkaline hydrazine hydrate instead of alkaline hydrazine sulphate.In this study quantitative recoveries of the halide ions have been achieved using neutral hydrazine hydrate. Several organic halogen compounds have been analysed by potentiometric titration of the solution after oxygen flask combustion with silver nitrate titrant the equivalence points were deter-mined by Gran’s method.23 Experimental Apparatus The following apparatus was used an Orion 901 Microprocessor Ionalyser to read 0.1 mV; a calibrated ALGA Micrometer Syringe burette of 0.5 ml capacity; a Pye silver billet indicator electrode; an EIL mercury - mercury(1) sulphate reference electrode; a magnetic stirrer; and a Cahn 21 automatic electrobalance THORBURN BURNS AND MAITIN 453 Reagents and Solutions Sodium chloride 90.99*yo.Potassium bromide 99.9*yo. Potassium iodide 99.75*y0. Silver nitrate solutiopz 0.2 M. Cellulose powder. Paperfor m e as samele carrier. Nitric acid 4 M. Aluminium jhotassiztm sulfihate solution 0.05 M. Ammonium vanadate solution 1% m/V. Bromophenol blue indicator solution in ethanol 0.05% mlV. BDH Chemicals Ltd. Hydraxine hydrate 99-100yo. BDH Chemicals Ltd. Doubly distilled water was used throughout the work for making up the solutions dilutions and rinsing of glassware etc. All items of calibrated glassware used were certified “A” grade.Procedure Accurately weigh by difference an amount of the sample to contain 0.04 & 0.002 mmol of halide ion in the sample mass. Place it on a filter-paper cut into an L shape (3 x 3 cm with a 1 x 3 cm tail at one edge). Add 30-40 mg of dry cellulose powder and make a parcel (1 x 1 cm) by folding the paper containing the sample in such a way so that the sample is covered with several layers of paper in all directions. Place an additional strip of paper (1 x 6 cm) in between the folds fold it once over the parcel and then wrap the parcel completely with the tail of the original paper. The unfolded portion of the additional strip is used as a fuse for ignition of the parcel. Clamp the parcel into a platinum gauze (1 x 1.5 cm) attached to a thin tungsten rod sealed into a ground-glass stopper (B24/29).Ignite the fuse and combust the sample in a 500-ml Erlenmeyer flask (neck B24/29) that has been flushed previously with oxygen and contains 10 ml of water and 5 drops of hydrazine hydrate. After the combustion is complete shake the flask gently on a mechanical shaker for 15 min and then allow it to stand for another 15 min. Carefully transfer the contents into a 100-ml beaker. Wash the stopper platinum gauze and walls of the flask into the beaker using small portions of water taking care to ensure that the total volume does not exceed 60 & 5 ml. Whilst stirring add 3 drops of bromophenol blue indicator followed by addition of 4 M nitric acid dropwise until the yellow colour of the indicator is just restored then add a few drops in excess.Add 4 drops of ammonium vanadate solution heat the solution gently on a hot-plate for a few minutes until it almost reaches boiling-point remove and stir for 30 s. Add 2 ml of aluminium potassium sulphate solution and dilute to 75 ml. Place the beaker in a brown-glass container and titrate potentiometrically with 0.2 M silver nitrate solution as follows. Add 0.05-ml aliquots at the beginning of the titration and 0.01-ml aliquots in the vicinity of the equivalence point in 8-10 steps. Note the potential readings, allowing a 1-min waiting time after each addition of the titrant. Determine the equivalence points by Gran’s method using data after the equivalence point. Standardise the silver nitrate solution by titrating a set of solutions containing standard inorganic halides equivalent to the halide content to be determined in the sample (i-e.about 0.04 mmol). Titrate another set of solutions containing known amounts of inorganic halides as above in the solutions obtained after combustion of a similar amount of paper cellulose powder and all other reagents used in the combustion of organic samples. From the titre values obtained in these titrations and those obtained in the standardisations above calculate the blank values. Separate standardisation and blank determinations are required for each halide. Calculate the percentage of halogen(s) present in the samples after deducting the blank values. AnalaR BDH Chemicals Ltd. AnalaR BDH Chemicals Ltd. AnalaR BDH Chemicals Ltd.AnalaR BDH Chemicals Ltd. Whatman CCL41 dried at 105 “C. AnalaR Hopkin and Williams. Cut from Whatman No. 1 filter-paper. AnalaR Hopkin and Williams. May and Baker. Dissolved using a few millilitres of 2.5 M sulphuric acid. Place a magnetic-stirring bar in the beaker. Allow to cool to room temperature. The reagent blanks are determined as follows. Results and Discussion In order to achieve rapid and complete combustion of the halogenated organic compounds, sucrose dodecanol glucose starch and cellulose powder were examined as combustion aids. * Percentage purity from gravimetric assay 454 THORBURN BURNS AND MAITIN IMPROVEMENTS TO OXYGEN Anahst VOJ. 108 Cellulose powder was found to be the best for complete and uniform burning of the samples. Sucrose and glucose "sparked" whilst burning and loss of sample was possible.Starch burned very slowly and often gave black residues. Dodecanol combusted quickly and the paper was burnt before the complete combustion of the sample which then caused soot formation and only partial combustion of the sample. Because of high vapour pressure many halogenated materials tend to distil out of the flame leading to low results owing to partial combustion. Partial combustion and soot formation also resulted when samples were not carefully and tightly wrapped. To avoid such problems the sample wrapping procedure as suggested by Awad et a1.12 was adopted. The size of the platinum gauze was also found to be critical; it should be small enough so that it remains entirely inside the flame during the burning process otherwise soot formation and distillation are unavoidable.After studies of the combustion of the compounds attention was given to the absorption stage of the method. Neutral hydrogen peroxide was found to give satisfactory results for compounds containing chlorine and bromine. Iodine was however absorbed and reduced in alkaline peroxide but as noted earlier by Lalancette et al.,13 the absorbing solutions occasionally remained yellow indicating the presence of free iodine. In addition hydrogen peroxide must be completely eliminated before titration of iodide. The decomposition of hydrogen peroxide is difficult free iodine was found to appear on acidification even after boiling the solution after combustion for several minutes. Similar problems arose when alkaline sodium tetrahydro-borate(II1) was used as reducing agent.Alkaline hydrazine sulphate has been considered a better reducing agent and used by many workers to give satisfactory results for chloride, bromide and iodide. In this study hydrazine hydrate has been shown to be more convenient than hydrazine sulphate alkaline conditions were not required as had been suggested by Mazzeo-Farina and Mazzeo. Further it was found that removal of any excess of hydrazine was not necessary provided the solution is kept acidic during the titration. Earlier workers have removed excess of hydrazine by using hydrogen peroxide and then eliminated the excess of hydrogen peroxide by boiling. Although details were not given it would appear that TABLE I ASSAY OF ORGANIC CHLORINE COMPOUNDS Compound l-Chloro-2,4-dinitrobenzene* C1C6H,(N0,), S-Benzylthiuronium chloride C6H,CH,SC( NH)NH,Clt p-Chlorobenzoic acid .. ClC,H,COOH Hexachlorobenzenes . . c6c16 Sample mass/mg . . 8.621 8.607 8.610 8.617 8.593 . . 8.616 8.647 8.649 8.645 8.647 8.639 . . 6.927 6.907 6.888 6.902 6.907 6.903 . . 2.089 2.092 2.093 2.083 Calculated yo Found yo Mean % 17.50 17.54 17.55 17.59 17.52 17.56 17.54 17.49 17.54 17.50 17.53 17.50 17.54 17.42 17.45 22.65 22.61 22.60 22.56 22.55 22.64 22.63 22.59 74.70 74.39 74.32 (74.28)y 74.25 74.15 74.50 Standard Apparent deviation, +0.05 f0.03 error yo % +O.Ol f0.05 -0.05 f0.04 -0.38 A0.15 (+ 0.04) TI * Microanalytical-reagent grade ; relative molecular mass 202.56 ; and melting-point 50-51 "C.BDH t Microanalytical-reagent grade; relative molecular mass 202.72 ; and melting-point 176-177 "C. Microanalytical-reagent grade; relative molecular mass 156.57 ; and melting-point 240-241 "C. 5 Microanalytical-reagent grade purity expected Q 99.5% ; relative molecular mass 284.8 1 ; and 7 The values in parentheses are based on the purity found by GLC. Chemicals Ltd. BDH Chemicals Ltd. BDH Chemicals Ltd. melting-point 228-230 "C. BDH Chemicals Ltd. Purity found by GLC 99.44% April 1983 FLASK COMBUSTION FOR HALOGENATED ORGANICS 455 Childs et al.15 have also titrated directly without the removal of the excess of hydrazine. Recently Chengl* also titrated the final solution directly.Many workers7J4~24-27 have indicated that bromate or iodate may be formed after the com-bustion. In order to overcome this problem a few drops of ammonium vanadate were added to the acidified solution after combustion and the solution was heated for a few minutes. This procedure catalyses the conversion of any bromate or iodate if formed after combustion into bromide or iodide in the presence of hydrazine hydrate. For the analysis of compounds containing chlorine alone such treatment is not required but as it did not show any adverse effects it was also followed for compounds containing chlorine to maintain a single procedure for all sample types. During studies of the potentiometric titration of halides using silver nitrate solution a con-centrated titrant was found to give sharp potential breaks therefore avoiding dilution during the titration.28 The use of a concentrated titrant also improved the stability of the potential responses.A calibrated microburette was used for accurate and precise delivery of the small volumes of the titrant. It was also found advisable to use similar amounts of halides in samples and standards. The titrations were monitored using a silver billet indicator electrode and a mercury - mercury(1) sulphate reference electrode. In a study of electrode response i t was noticed that it was necessary to allow 1 min after each addition of the titrant in order to obtain reproducible potential readings. Aluminium potassium sulphate was added to avoid the problems that may occur due to ad~orption.~~ Equivalence points were evaluated by Gran's method using 8-10 points after the equivalence point.These results were found to be better than those calculated by other graphical or numerical methods. The paper used as sample carrier was found to give a blank value. The blank values were different for each halide and varied with the different methods of equivalence point evaluations examined. It is essential to calculate the blank separately for each halide titration. It was found that the blanks were more reproducible when measured by difference as in the procedure herein than when measured directly. Several standard compounds and other organic compounds containing single halogens were assayed by using the final procedure described above and results obtained are given in Tables 1-111.It is seen that these are satisfactory except for hexachlorobenzene which was exam-ined by gas - liquid chromatography but when the results are corrected for over-all purity they are then in good agreement. Compound Bromobenzoic acid* BrC,H,COOH N-Bromosuccinimidet C,H,O,NBr Bromoacetanilide: CH,CONHC,H,Br TABLE I1 ASSAY OF ORGANIC BROMINE COMPOUNDS Standard Sample Calculated Apparent deviation, masslmg yo Found yo Mean yo error % % . . 8.552 39.75 39.89 39.82 +0.07 f0.09 8.522 39.73 8.567 39.71 8.541 39.87 8.532 39.78 8.525 39.93 7.538 45.03 7.591 45.14 7.596 44.87 7.568 44.83 . . 7.601 44.89 45.16 45.01 +0.12 f0.15 . . 9.068 37.33 37.23 37.35 +0.02 fO.08 9.135 37.37 9.111 37.38 9.121 37.32 9.115 37.43 * Microanalytical-reagent grade; relative molecular mass 201.02 ; and melting-point 256-257 "C.t Relative molecular mass 177.99 melting-point 180 O C ; Aldrich Chemical Co. 99% material recrystal-$ Relative molecular mass 214.07 ; melting-point 169 "C; BDH Chemicals Ltd. 99-101% material BDH Chemicals Ltd. lised five times from benzene. recrystallised five times from ethanol 456 THORBURN BURNS AND MAITIN TABLE I11 ASSAY OF ORGANIC IODINE COMPOUNDS Compound p-Iodonitrobenzene* C6H4N021 N-Iodosuccinimidet CH,CONICOCH, o-Iodobenzoic acid$ IC,H,COOH Sample masslmg . . . . 10.983 10.982 10.951 10.982 10.987 . . 9.930 9.930 9.930 9.912 9.914 9.924 . . 10.953 10.934 10.924 10.931 10.942 Standard Calculated Apparent deviation, % Found % Mean yo error % % 50.96 50.92 50.86 -0.10 f0.20 50.88 51.11 50.83 50.55 56.19 56.19 56.52 56.64 56.10 51.40 51.15 51.37 51.18 56.40 56.42 56.34 -0.06 f0.22 51.16 51.48 51.32 +0.16 f0.14 * Relative molecular mass 249.01 ; melting-point 172-173 “C ; Aldrich material recrystallised five times t Relative molecular mass 224.99 g ; melting-point 203-204 “C; Hopkin and Williams Ltd.laboratory-$ Microanalytical-reagent grade; relative molecular mass 248.03 ; and melting-point 162-163 “C. from ethanol. reagent grade material recrystallised five times from dioxan - carbon tetrachloride. BDH Chemicals Ltd. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.22. 23. 24. 25. 26. 27. 28. 29. References Hempel W. 2. Angew Chem. 1892 13 393. Macdonald A. M. G. Analyst 1961 86 3. Macdonald A. M. G. in Reilley C. N. Editor “Advances in Analytical Chemistry and Instru-Schoniger W. Mikrochim. Acta 1955 123. Schoniger W. Mikrochim. Acta 1955 869. Patai S. “The Chemistry of Carbon - Halogen Bond Part 1,” John Wiley New York 1973. MAzor L. “Analytical Chemistry of Organic Halogen Compounds,” Pergamon Press Oxford 1975. Ma T. S. and Rittner R. C. “Modern Organic Elemental Analysis,” Marcel1 Dekker New York 1979. Ingram G. “Methods of Organic Elemental Microanalysis,” Chapman and Hall London 1962. Levy R. Pure APPl. Chem. 1972 29 629. Griepink B. and van Sandwijk A. Mikrochim. Acta 1969 1014. Awad W. I.Gawargious Y. A. Hassan S. S. M. and Milad N. E. Anal. Chim. Acta 1966 36, Lalancette R. A. Lukaszewski D. M. and Steyermark A. Microchem. J. 1972 17 665. Celon E. Mikrochim. Acta 1969 592. Childs C E. Meyers E. E. Cheng J. Laframboise E. and Balodis R. B. Microchem. J. 1963 7, Nara A. Kobayashi N. and Honba K. Microchem. J. 1975 20 200. Rittner R. C. and Ma T. S. Mikrochim. Acta 1976 243. Cheng F. W. Microchem. J. 1980 25 86. Campiglio A. and Traverso G. Mikrochim. Acta 1980 I 485. Lalancette R. A. Steyermark A. Lukaszewski D. M. and Kostrazewski P. L. J . Assoc. Ofl. Lalancette R. A. and Steyermark A. J . Assoc. Of. Anal. Chem. 1974 57 26. Mazzeo-Farina A. and Mazzeo P. Microchem. J . 1978 23 137. Gran G. Analyst 1952 77 661. Dixon J. P. “Modern Methods in Organic Microanalysis,” Van Nostrand New York 1968. Olson E. C. in Kolthoff J. M. and Elving P. J. Editors “Treatise on Analytical Chemistry Part MAzor L. PApay K. M. and Klatsmanyi P. Talanta 1963 10 557. Belcher R. and Fildes J. E. Anal. Chim. Ada 1961 25 34. Thorburn Burns D. and Maitin B. K. Anal. Proc. 1982 19 324. Lingane J. J . “Electroanalytical Chemistry,” Second Edition Interscience New York 1958. mentation,” Volume 4 John Wiley New York 1965 p. 75. 339. 266. Anal. Chem. 1973 56 888. 11,” Volume 14 John Wiley New York 1971 p. 1. Received August 6th 1982 Accepted November l l t h 198