Oct., 19511 BELCHER AND TATLOW 593 The Determination of Fluorine in Organic Compounds BY R. BELCHER AND J. C. TATLOW A method is described for the determination of fluorine in organic compounds. Decomposition is effected with sodium in a nickel bomb of special design and then the fluoride ion is determined gravimetrically as lead chlorofluoride, THE accurate determination of fluorine in organic compounds is complicated both by the difficulty of decomposing the sample effectively and by the lack of convenient or reliable methods for determining the fluoride ion. Whilst certain fluorine-containing compounds can be decomposed readily by simple hydrolysis, the method is restricted mainly to acyl fluorides and compounds in which fluorine atoms are activated by other groupings present in the molecule; where the fluorine is bound more firmly much more drastic treatment is necessary.Fluorocarbons or mixed halofluorocarbons are particularly resistant to decom- position. Elving and Ligett,l who have very comprehensively reviewed the methods of n 1‘ Fig. 1. Bomb for combustion A, nickel cup ; B, nickel lid ; C, hole to take bar for opening; D, steel nut; E, copper gasket decomposition and determination, recommend decomposition with sodium or potassium in a sealed tube or nickel bomb and determination of fluoride ion by titration with thorium nitrate or precipitation as lead chlorofluoride. From our own experience the most convenient and effective method of decomposition is with sodium in a nickel bomb. Decom- position with sodium peroxide in a Parr bomb2 is effective for certain compounds but does not completely decompose some of the fluorocarbons we have encountered.The design of the bomb we use is shown in Fig. 1. It is easy to manipulate and sufficiently compact to enable several bombs to be placed in a small muffle furnace at the same time. After destruction of the organic matter and conversion of the fluorine to sodium fluoride, there are several possible methods for carrying the determination to completion. Because of its rapidity, considerable attention was devoted to the thorium nitrate method, but for our purpose it did not prove satisfactory. The many modifications of this method, including that described in the Analytical Methods Committee’s “Determination of Fluorine in food^,"^ We have not found it necessary to use potassium.594 BELCHER AND TATLOW: THE DETERMINATION OF [Vol.76 to which we propose to return at a later date, have not been examined. As results were urgently required, a modification of the well-known lead chlorofluoride method was used, the working details of which were supplied to us in a private communication by Mr. H. S. Stretch of the Ministry of Supply. This procedure required about 80 mg of fluorine to be present, which necessitated more sample than could normally be afforded. Accordingly, the method was further modified to make it effective for samples of limited size that contained from 30 to 60mg of fluorine. When the amounts of sample were very limited, sometimes only 20 mg or even 10 mg of fluorine have been determined, but the error is then slightly larger.In general, with 20mg of fluorine the results are about 0.5 per cent. lower than theoretical, and with 10mg about 1 per cent. lower. The final gravimetric estimation of the fluoride ion is preferred because of the favoura.ble conversion factor. By this procedure, satisfactory results, some of which are shown in Table I, have been obtained, for a wide variety of compounds. TABLE I RESULTS OF A TYPICAL SERIES OF ROUTINE DETERMINATIONS BY THE RE COMMENDED METHOD (NOT SPECIALLY SELECTED) Fluorine content Compound m-Trifluoromethylbenzoic acid . . .. .. 2-Nitro-5-acetaminobenzotrifluoride . . .. m-Acetaminobenzotrifluoride . . . . . . 3 : 3'-Bis(trifluoromethyl)azobenzene . . . . N-Trifluoroacetyl-2 : 5-dichloroaniline .. .. Y-Trifluoroacetylbenzylamine . . .. .. Perfluorodimethyldecalin . . . . .. .. Perfluorodicyclohexylethane . . .. . . N-Trifluoroacetylaniline . . .. . I * . Heptafluorobutyric acid . . .. .. . . Perfluorodicyclohexylpentane . . .. .. Benzotrifluoride . . .. .. . . .. 1 : l-Bis(~%fluorophenyl)-2 : 2 : 2-trichloroethane Trifluoroacetamide . . .. .. . . 4 : 6-Benzylidene trifluoroacetyl-a-methylglucoside .. .. .. .. .. .. .. .. .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Found, 30.0 22.8 27.7 35.6 22-1 28-2 39.1 11.4 74.0 74.9 49.9 15.0 30.0 61.8 74.7 % Calculated, 30.0 23.0 28.1 35.8 22-1 28.1 39.0 11.8 74.4 74.6 50.4 15.1 30.1 62.1 74.9 % So far compounds containing only carbon, hydrogen, oxygen, chlorine and nitrogen as well as fluorine have been encountered.The method might require some modification for other types of compound and we hope to address ourselves to this problem in the near future. We have found that m-trifluoromethylbenzoic acid is a suitable standard substance, for it is readily purified and is non-hygroscopic. METHOD THE BOMB- The bomb, shown in Fig. 1, has already been described in detail.4 The cup, A, is made of nickel, the upper portion being threaded to take the nut, D, and the lower portion being hexagonal to fit a spanner. The thread on the body of the bomb should be relieved to a very slight taper to avoid binding during the unscrewing process. The lid of the bomb, B, is also of nickel, a hole, C, being drilled in the top to enable a bar to be inserted to open the bomb.As the nut, D, is unscrewed, the bar impinges on the top of this nut and the cup and lid of the bomb are separated. The nut, D, is made of stainless steel"; brass has also been used successfully but is not as satisfactory as steel, particularly at high temperatures. The sealing ring, E, is made of copper, and is a hollow rolled gasket similar to a sparking-plug washer (there is no asbestos packing). Before use each gasket is cleaned and is then softened by being heated in a flame to redness and cooled in ethanol. A new gasket is necessary for each determination. It is convenient to construct a bench jig by welding or brazing the hexagonal section of a 5/16-inch box spanner to a steel plate that can be mounted on the bench. The cup, A, can rest on this while the bomb is being assembled or taken down.* Firth Vickers H.R. Crown Max.Oct., 19511 FLUORINE I N ORGANIC COMPOUKDS 595 After each determination, the flanges of the bomb and lid should be cleaned by means The bomb components may be mounted in a lathe chuck for rotation of fine emery paper. during the cleaning procedure. REAGENTS- Nitric acid-5 N. Acetic acid-A 30 per cent. v/v solution. Lead chloride-A saturated solution. Lead chlorojuoride-A saturated solution. *4 bsolute alcohol. A cetone. PROCEDWRE- Weigh sufficient sample from a weighing tube to yield between 30 and 60 mg of fluorine and transfer to the cup of the bomb. Add sodium in the form of small pellets, taking about 2 to 3 times the weight of sample, place the treated copper washer in position and screw the nut down tightly with a spanner.Place the sealed bomb for at least 1 hour in a furnace maintained at 550" to 600" C. Some fluoro- carbons may require a temperature of 650" to 700" C. After heating for the required time, remove the bomb and allow it to cool. Remove the lid of the bomb and wash any material adhering to the underside into a 100-ml beaker with a fine jet of water, not exceeding a total of 10 ml. Place the cup in the beaker and very carefully add 1 ml of absolute alcohol to destroy the residual sodium. Allow to stand for 10 minutes, then stir with a nickel rod and leave for a further 3 minutes. Add 2 or 3 drops of water at intervals to complete the destruction of the sodium. Empty the contents of the bomb into the beaker, wash the bomb thoroughly inside and out and remove it from the beaker.Filter through a Pregl sintered-glass filter of No. 1 porosity. Transfer the filtrate to a 100-ml beaker and wash the filter tube with water. Add 8 drops of a 0.1 per cent. solution of methyl red and add 5 N nitric acid dropwise until the indicator just turns pink. Add 0.25 ml of the acetic acid solution and heat to 60" to 70" C. Meanwhile, heat 200 ml of saturated lead chloride solution, containing 0-5 ml of the acetic acid solution, nearly to boiling in a 400-ml beaker. Add the hot fluoride solution to the lead chloride solution and wash the 100-ml beaker twice with 15-ml portions of saturated lead chloride solution. Place a clock glass over the 400-ml beaker and heat gently to boiling. Allow to stand for at least 4 hours or, preferably, overnight.Wash the beaker twice with 15-ml portions of lead chloride solution taking care to remove all traces of precipitate. Finally, wash the beaker with a jet from the wash-bottle containing the lead chloride solution. Wash the precipitate thrice with 15-ml portions of lead chlorofluoride solution, then thrice with 15-ml portions of acetone. Suck the crucible dry, wipe the outside and then place in an oven at 110" C for 20 to 30 minutes. Cool in a desiccator for 20 minutes, place in the balance case and weigh after a further 10 minutes. Weigh liquids in a gelatin capsule, size No. l.* Filter the precipitate through a clean, dry, weighed No. 4 sintered-glass crucible. Weight of fluorine = weight of precipitate x 0.07263 The authors express their gratitude to the Department of Scientific and Industrial Research for a grant in support of this work; to Professor M. Stacey, F.R.S., for his interest and help; to Mr. W. Massingham for the construction of the bombs; and to Mr. B. S. Noyes, who has been responsible for the majority of the routine determinations. REFERENCES 1. 2. 3. 4. Eking, P. J., and Ligett, W. B., Ind. Eng. Chem., Anal. Ed., 1942, 14, 449. Nicholls, M. L., and Olsen, J. S., Ibid., 1943, 15, 342. Analytical Methods Committee, "Determination of Fluorine in Foods," A nalysl, 1944, 69, 243. Stacey, M., Tatlow, J. C., and Massirlgham, W. E., British Patent Application 3631/51, dated February 14th, 1951. DEPARTMENT OF CHEMISTRY THE UNIVERSITY BIRMINGHAM, 15 March, 1951 * Manufactured by Eli Lilly and Co., Indianapolis, U.S.A.