Determination of Fluorine and Other Halogens in Organic Compounds


Determination of Fluorine and Other Halogens in Organic Compounds...

3 downloads 134 Views 2MB Size

INDUSTRIAL

and

ENGINEERING CHEMISTRY

ANALYTICAL PUBLISHED

BY

THE

AMERICAN

EDITION SOCIETY

CHEMICAL



HARRISON

E.

HOWE,

EDITOR

Determination of Fluorine and Other Halogens in Organic Compounds PHILIP J. ELVING

AND W. B.

LIGETT, Purdue University, Lafayette, Ind. known accuracy. However, if fluorine is present, difficulties introduced due to the great stability of the carbon to fluorine bond in aromatic compounds and in highly fluorinated aliphatic compounds. The most widely used method for the recovery of halide ion from organic compounds, the Carius method, cannot be used for fluorine since the hydrogen fluoride or hydrofluoric acid formed attacks the glass of the reaction tube. In speaking of the analysis of certain chlorofluoropropanes, Henne and Renoll {19) assert that the chlorine in these substances could be determined only by the Carius method, which required a whole week of continuous heating at 250° to 300° C. to yield quantitative results. All other methods gave an incomplete decomposition. As an indication of the stability of the carbon to fluorine bond, Meyer {28) asserts that the other halogens in halofluoro compounds can be determined by the Carius method or by lime fusion in glass without splitting out the fluorine. The methods suggested in the literature for the decomposition of organic fluorine compounds are classified in Table I. Most of the fluorine compounds for which analyses are reported are of comparatively low fluorine content, commonly having only a single fluorine atom in a molecule of high molecular weight. This choice of compounds makes the methods appear better than they actually are for two reasons: (1) a relatively large error on the basis of fluorine present appears as a small percentage error; and (2) applicability of the method to the more stable compounds containing two or three fluorine atoms on a single carbon atom is not tested. In the present investigation, however, many of the compounds analyzed were chosen because they were supposed to be particularly stable and could not be analyzed by the methods described in the literature.

A method for the analysis of organic fluoro compounds is presented. The procedure depends upon decomposition of the compound by heating with an alkali metal in an evacuated sealed tube at moderately elevated temperature, and determination of the resulting alkali fluoride by standard for the methods. The same technique serves analysis of chloro, bromo, and iodo compounds and has several advantages over the procedures comApplimonly employed in these determinations. cation of the method to various types of fluoro compounds discloses no compound too stable to be decomposed by the conditions described. A simulof fluorine and chlorine taneous determination in chlorofluoro compounds may be made. The decomposition conditions leave the halide ions in an

environment

suitable for determination

without

are

a

preliminary separation. The method is applicable and rapid, to solids, liquids, and gases, is accurate and requires

no

special apparatus

or

reagents.

on and use of organic compounds fluorine has called attention to the desirability of a simple method for determining this element, especially in the very stable compounds containing more than one fluorine atom on the same carbon atom. The methods at present described in the literature either do not work for the more stable compounds or are described in insufficient detail and without sufficient experimental verification to permit their use. The standard references on organic analysis fail in most cases even to mention the topic {27, 32, 37). The determination of fluorine and other halogens in organic compounds resolves itself into two parts: (1) destruction or decomposition of the compound to convert the halogen into an ionizable form, and (2) determination of the halide ion. In the case of organic compounds containing halogens other than fluorine, the halide can usually be obtained by any of several recognized methods—e. g., the Carius method of heating with nitric acid in a closed tube, oxidation by sodium peroxide in a Parr bomb or by fuming sulfuric acid, combustion in oxygen, or hydrogenation in the presence of a suitable catalyst. These methods are of varying degrees of convenience as regards simplicity of apparatus, time required, and skill necessary in the operator. The separation and determination of chloride, bromide, or iodide ion are well taken care of by standard methods of

increasing research

THE containing

Table I.

Decomposition ing

I. II.

III.

IV. V.

449

Organic Compounds ContainFluorine

of

Oxidation Methods A. Combustion in oxygen {5, 25, 31, 34) B. Fusion with sodium peroxide {12, 14, 22, 35, 42) C. Alkaline oxidation (4) Reduction Methods A. Combustion in hydrogen {8, 46) B. Treatment with sodium in liquid ammonia {47) C. Alkali metal fusion {2, 23, 36, 39, 43)

D. Treatment with alkali metal in organic solvent {48, 46) Methods Involving Alkaline Fusion A. Fusion with calcium oxide (3, 10, 11, 20, 24, 29, 40, 4^> 44) B. Fusion with sodium carbonate {6) Methods Involving Reaction with Silicon Dioxide A. Corrosive action on glass (7, 17, 33) B. Combustion over silicon dioxide using oxygen and hydrogen (9, 18, 19, 21)

Hydrolytic Methods {16, 26,

30, 45)

INDUSTRIAL

450

Figure

1.

Furnace

for

AND

ENGINEERING

Halogen Determination

The objective of the research was the development of a simple, rapid, and complete procedure for converting all the halogens present in organic compounds to halide ions in forms and environment permitting their determination with a minimum of further separation. Principal requirements were that the method should serve for the recovery of fluoride ion from stable organic fluoro compounds and should permit the determination of fluorine and other halogens in the same sample.

Principle of Proposed Method The results of extensive preliminary experiments with the promising methods described in the literature indicate that the most satisfactory method is based upon fusion with an alkali metal. The compound to be analyzed is heated with metallic sodium or potassium to a moderately elevated temperature in an evacuated sealed tube, yielding alkali halides in addition to the pyrolysis products, chiefly carbon, of the organic residue. The halide ion is then determined o© by standard procedures. The method described in detail below was evolved after many analyses of stable chlorofluoro compounds under a variety of conditions.

more

CHEMISTRY

Vol. 14, No. 6

routine analysis of halogen compounds in this laboratory, a suitable furnace was designed. The furnace together with one of the reaction tubes is shown in Figure 1; an outline drawing is given in Figure 2. It may be constructed at a cost of about three dollars for material and the expenditure of 8 man-hours of labor. Construction of Furnace. The furnace tubes are four iron pipes, 2.5 cm. (1 inch) in diameter and 30 cm. (12 inches) long, supported horizontally by the two 22.5-cm. (9-inch) paint-can lids which constitute the ends of the furnace. The tubes are arranged symmetrically on a circle of 3.25-cm. (1.5-inch) radius, with 2.5 cm. (1 inch) protruding at each end. The ends are threaded to take lock nuts and pipe couplings, into which are screwed the cast-iron closure plugs. Holes are drilled in the plugs to provide for release of pressure. A 9.375-mm. (0.375inch) pipe through the center and paralleling the furnace tubes serves as a thermocouple well. The five tubes are wrapped as a unit with asbestos paper, and the unit is wound with 525 cm. (17.5 feet) of No. 23 Chromel wire having a resistance of 1.25 ohms per foot. The ends of the resistance wire terminate in a motor plug base set in one end of the furnace near the circumference. The resistance winding is covered with several more layers of asbestos paper held in place with asbestos cord. The jacket of the furnace is a 25-cm. (10-inch) section of galvanized furnace pipe. Sil-O-Cel is used as insulation between the jacket and the asbestos-wrapped unit of five tubes. No insulation was used between the furnace tubes themselves. The furnace reaches 400° C. on the 110-volt line in 75 minutes, and is held at this temperature by an external resistance of 2 ohms in series with the furnace. The resistance, made by winding approximately 120 cm. (4 feet) of No. 19 Chromel wire on a cone heater, is mounted on a switch box in parallel with a fuse, so that it may be shunted out simply by screwing in the fuse plug. The furnace and external resistance are mounted on a board, the furnace resting in a cradle made of 12.5 X 3 mm. (0.5 X 0.12 inch) strap iron.

Procedure for Analysis Most of the compounds analyzed were liquids and the prowas first developed for material in this state.

cedure

Analysis of Liquids, Ampoules for sampling liquids are blown from 3-mm. Pyrex tubing, slightly drawn out. The bulbs are 6 to 8 mm. in diameter, with stems 5 to 6 cm. long; the successive steps in their preparation are shown in Figure 3. A 0.10to 0.15-gram sample of the liquid compound is drawn into the previously weighed ampoule, and the ampoule is sealed and weighed again. Care is taken to pyrolyze none of the compound during sealing. The ampoule containing the sample is placed in the reaction tube and 5 ml. of ether are added. The sodium or potassium metal is introduced in the form of very small pieces, the amount depending upon the size of sample taken and its halogen content (the optimum amount is usually 0.3 to 0.5 gram, giving several hundred per cent excess), and is weighed roughly or simply estimated. The reaction tube is then connected to an

FJRNA CE

3

rfiH-

rtf”QU.-

-

j

CONS TRUC Tl ON

A.

FUSE

B.

CONE RESISTANCE FURNACE TUBES

C. D.

(SHUNT)

THERMOCOUPLE \NELL

Reagents and Apparatus A saturated solution of lead ehlorofluoride was prepared as described by Scott (13). All reagents were of tbe best obtainable grade and tested for tbe presence of the halide ion were being determined. It is particularly important to check the sodium and potassium metals for tbe halide ion being determined, and to use anhydrous ether. The reaction tubes were made by drawing out the open ends of Pyrex ignition test tubes, 20 X 150 mm., and sealing on 12-cm. lengths of 10-mm. tubing. These reaction tubes may be used repeatedly, since there is little attack on the glass. The ampoules prepared for sampling solids, liquids, and gases are described below. Since a large number of determinations were to be run and the method was to be used for the

10

H

INCHES

Figure 2.

Furnace

for

Halogen Determination

ANALYTICAL

June 15, 1942

aspirator and all the ether is drawn off. The evacuation of the reaction tube in this manner serves to remove oxygen and water vapor completely from the tube and to give a low pressure. Removal of the oxygen and water vapor maintains the surface of the alkali metal clean, and therefore more reactive, and at the same time prevents attack of the reaction tube. Reduction of the pressure decreases the possibility of explosion of the tube when heated, and increases the vaporization of the compound and alkali metal, thus speeding up the reaction. The stem of the reaction tube is sealed off with a hand torch while the tube remains connected to the water aspirator. The reaction tube is most conveniently sealed off when supported in a vertical position by means of the rubber tubing from the aspirator. It is obviously advisable to have a trap in the line to prevent water being drawn into the reaction tube in case of a drop in the water pressure. After the reaction tube is sealed, the ampoule is broken by shaking and the reaction tube placed in the furnace at 400° C. After 15 to 30 minutes, depending upon the compound being analyzed, the tube is removed and allowed to cool. It is then opened by breaking off the stem close to the original seal and the excess alkali metal is decomposed by the cautious addition of ethyl alcohol. The contents of the tube are transferred with several portions of wash water to a 100-ml. beaker and the ampoule, including stem, is crushed. The broken glass, carbon, and any silica are then filtered off through a No. 4 Jena glass filtering funnel or crucible or equivalent Pyrex filter. If the filter becomes clogged with carbon, it can be cleaned by immersion overnight in warm chromic acid-sulfuric acid solution. The filtrate is neutralized with nitric acid and the halide determined in the filtrate by standard methods. In this investigation, chloride, bromide, and iodide were determined gravimetrically by precipitation as the silver salts, and fluoride was determined by precipitation and weighing as lead chlorofluoride, using essentially the precipitation technique of Hawley (15). If compounds containing nitrogen are analyzed, the halide procedure used must avoid interference with cyanide ion; the latter does not interfere with the fluoride determination. If the compound being analyzed contains fluorine and another halogen, both may be determined on the same sample, or a separate sample may be used for each determination, the choice depending upon the relative importance of rapidity and accuracy. To determine both halogens from a single sample, aliquot parts of the fusion filtrate are taken. Analysis of Low Melting Solids. Solids which have an appreciable vapor pressure at room temperature must be introduced into the reaction tube in sealed ampoules to avoid loss during evacuation of the tube. The ampoules for sampling solids are made from 10-cm. lengths of 6-mm. Pyrex tubing, wffth a fragile bulb blown in one end (see Figure 3). After the ampoule has been weighed and the sample introduced, it is evacuated and sealed off near the middle with a torch, care being taken to pyrolyze none of the compound. The two sections of the glass are weighed to give by difference the weight of sample contained in the sealed portion. The buoyancy effect of air on the weight of the ampoule is appreciable and a correction is calculated on the basis of the measured inside diameter and length. The correction is given by the formula, C rVM, where r is the internal radius of the ampoule in centimeters, l is the length of the sealed ampoule in centimeters, and d is the density of air in grams per cubic centimeter. The correction, C, is simply added to the apparent weight of the sample. This correction assumes that the ampoule is completely evacuated; this, of course, is not true, but the false assumption simplifies calculation and introduces an error of considerably less than 0.1 mg. The sealed ampoule containing the sample is placed in the reaction tube, fragile end down, and the addition of ether and alkali metal and the evacuation and sealing-off are carried out as for liquids. The fragile end of the ampoule is then easily broken by shaking and the solid intimately mixed with the small pieces of alkali metal by rotating the reaction tube. The rest of the procedure is the same as for liquids, except that in removing the contents of the reaction tube the ampoule is not crushed. It is easily and thoroughly cleaned by adding 5 ml. of distilled water to the reaction tube, bringing the water to boiling, and momentarily removing the tube from the flame to allow the water to cool, thus drawing water into the ampoule. The water is again boiled, forcing it out of the ampoule, and poured into the beaker containing the previous washings of the reaction tube. This proc=

is repeated several times. Analysis of Gases. Gases are cooled below their boiling points and the resulting liquids poured into ampoules made from 10-cm. lengths of 5-mm. Pyrex tubing. With the lower half of the ampoule immersed in the cooling bath, the open end is drawn out to a fragile tip and sealed. The ampoule is placed in the reaction tube with the fragile tip downward, facilitating release of the sample when desired. Correction is made for the buoyancy ess

EDITION

Figure 3.

451

Steps

in

Construction

of

Ampoules

for

Sampling

effect of air. The analysis is carried out as for liquids, and the inside of the ampoule is rinsed as in the analysis of solids. The lowest boiling substance sampled by this method was dichlorodifluoromethane, boiling point —30° C.

Choice of Decomposition Conditions Many organic halogen compounds were analyzed using decomposition temperatures from 150° to 600° C., varying the time of heating from 15 minutes to 2 hours, and using either sodium or potassium metal. On the basis of these preliminary investigations, the standard procedure adopted for compounds containing only chlorine, bromine, or iodine is fusion with sodium at 400° C. for 15 minutes. Occasional low results were obtained with organic fluoro compounds with these decomposition conditions. As there is no apparent relation between the type of fluoro compound and the discordant results, the standard procedure adopted for fluoro compounds is decomposition with potassium metal at 400° C. for 30 minutes. It has been observed that potassium metal flows to coat the reaction tube, giving a clean surface of considerable area, whereas sodium does not. The higher cost of potassium amounts to only about one cent per determination. 6-Chlorocoumarin was the only compound containing only chlorine, bromine, or iodine for which conditions other than heating to 400° C. for 15 minutes with metallic sodium had to be applied. The fusion filtrate was always strongly colored, and the colored matter precipitated upon acidification of the fusion filtrate preparatory to precipitation of the chloride as silver chloride. Under the conditions adopted as standard for fluorine compounds, although the temperature was inadvertently run to 450° C., good results were obtained.

Presentation of Data Table II gives the data obtained in the analysis of 24 organic halogen compounds using the method evolved in this investigation. All results obtained on every compound whose analysis was attempted by means of the proposed method are reported. The mean value and the average deviation for the halogen content of each of the compounds analyzed are given, together with the number of determinations upon which these values were based. Since work was completed on the development of the procedure, a large number of organic halogen compounds have been successfully analyzed by the method. Since the procedure was developed primarily to fill the need for a rapid and accurate method for determining fluorine in organic compounds being prepared in this laboratory, the majority of compounds analyzed were fluoro compounds. Several compounds containing only chlorine, bromine, or iodine were analyzed because it appeared that the method had certain advantages over the precedures commonly era-

INDUSTRIAL

452

ENGINEERING

AND

CHEMISTRY

Vol. 14, No. 6

the first precipitate formed was white. Results were most often high. The No. of Halogen poorest results were obtained with Source and Purity of Sample Detns. Found Theory Compounds Analyzed two of the three gases analyzed, di% % chloromonofluoromethane (Freon 21) E. K. 2793, rectified, 84.5-84.6° C./ F 19.77 19.66 ±0.05 6 Fluorobenzene, CeH&F and 754 mm. 1,2-dichloro-l, 1,2,2-tetrafluoroE, K. 3141, rectified, 156.63 F 15.06 15.18 ± 0.04 ; -Fluoroanisole, ethane (Freon 114). With these com156.7° C./748 mm. F.C6H4.OCH3 E. K. 3203, rectified, 187.03 F 17.10 17.03 ±0.04 p-Flucroaniline, were pounds, the negative errors 187.1° C./753 mm. F.CcH4.NH2 associated with sample sizes much b. du Pont Freon 8.9° C. An6 F 18.46 18.72 ± 0.17 Dichloromonofluoromethane, 21, p. CHChF Cl 68.89 68.78 ± 0.19 alyzed as received higher than optimum and it seems du Pont Freon 12, b. p. —29.8° C. 3 F 31.43 31.57 ± 0.11 Dichlcrodifluoromethane, CC12F2 Cl 58.64 58.72 ±0.07 Analyzed as received likely that the most careful analysis ± du Pont Freon b. 3.8° C. F 6 44.46 44.47 0.34 1,2-Dichloro-l ,1,2,2-tetra114, p. would have given consistently high Cl 41.49 41.62 ± 0.25 fluc,roethane, CC1F2CC1F2 Analyzed as received Research compound, rectified, 88.03 F 18.64 18.79 ±0.03 l,l,2,2-Tetrachloro-l,2-divalues. Slightly less than average 88.5° C./750 mm. Cl 69.57 69.87 ±0.08 fluoroethane, CC12FCC12F 9 F 33.18 33.27 ±0.18 -Chloro-2,2-difluoropropane, Research compound, rectified, 55.1precision might have been expected 55.4° C./750 mm. CH2CICF2CH3 Cl 30.96 31.10 ± 0.08 with gases, owing to sampling errors, E. K, 3036, rectified, 134.0° C./750 3 F 14.55 14.73 ±0.06