374 CROMPTON : IODIMETRIC DETERMINATION OF [Analyst, 1701. 91 Iodimet ric Determination of Organo-aluminium Compounds BY T. R. CROMPTON (Carrington Plastics Laboratory, Shell Chemical Company Limited, Carringtow, Cheshire) The alkyl groups in various types of organo-aluminium compounds have been shown to react with iodine in hydrocarbon solution, and the stoicheio- metry has been determined of the reactions occurring between iodine and trialkylaluminium, dialkylaluminium chloride and dialkylaluminium alkoxide compounds. Based on these reactions a reasonably rapid and accurate iodimetric method has been devised for the determination of low- concentrations of organo-aluminium compounds in various hydrocarbon solvents. The method is applicable to the analysis of the hydrocarbon solutions of organo- aluminium catalysts used for the polymerisation of ethylene and propene.Good agreement is obtained between the iodimetric procedure and by a procedure based on conductiometric titration with a standard solution of isoquinoline for trialkylaluminium compounds and dialkylaluminium chlorides. The iodimetric procedure is also applicable to dialkylaluminium alkoxide compounds, which cannot be determined by isoquinoline titration. DILUTE hydrocarbon solutions of various types of organo-aluminium compounds are used as co-catalysts with Group IV and VI halides, in processes for the polymerisation of ethylene and propene to polyolefin polymers. Three of the principal types of organo- aluminium catalysts that are used in these processes are AlR,, AlR,(OR’) and A1R2C1 (where R and R’ are C, to C, alkyl groups).A rapid and simple method was required for determining these types of catalysts in the hydrocarbon solutions used for ethylene and propene poly- merisation, in amounts down to 10 millimoles per litre. Various methods, based on gasometric principles, have been described for determining methyl to butyl alkyl groups, and hydride groups in organo-aluminium compounds : Bonitzl and Ziegler, used the reaction between the organo-aluminium sample and 2-ethylhexanol, and determined the amounts of alkane and hydrogen gases produced by mass-spectrometric analysis- >AIC,H,n+, + ROH = >Al(OR) + CnH2n+2 >AlH + ROH = >Al(OR) 1- H2 Crompton and Reid3 and Dijkstra and Dahmen4 used the reaction between organo- aluminium compounds and hexanol, followed by water, and lauric acid, respectively, and analysed the resulting mixture of liberated alkane and hydrogen gases by gas chromatography.Neumann5 developed a gasometric method for determining aluminium-bound hydride groups, based on a reaction with N-methylaniline at low temperatures- Methods of analysis based on these principles, although capable of giving excellent information on the composition of the sample, were too lengthy and complex for routine control testing of plant streams. Bonitzl described a method based on a conductiometric or potentiometric titration with isoquinoline for determining organo-aluminium compounds. He described the colourless 1 to 1 complexes formed between isoquinoline and dialkylaluminium hydrides and trialkyl- aluminium compounds, and the strongly red coloured 2 to 1 complex formed between iso- quinoline and dialkylaluminium hydrides.These complexes were further studied by Neu- mann.5 Farina et ~ 1 . ~ and Sebbia and Pagani’ have reported modified potentiometric- titration procedures for determining organo-aluminium compounds.June, 19661 ORGANO-ALUMINIUM COMPOUNDS 375 Mitchens extended the studies of the red coloured 2 to 1 complex formed between iso- quinoline and dialkylaluminium hydrides and, based on his observations, devised a method for the simultaneous spectrophotometric determination of trialkylaluminium compounds and dialkylaluminium hydrides in mixtures. Wadeling utilised this colour-forming reaction to devise a photometric-titration method for determining the total isoquinoline-reactable organo- aluminium compounds.Razuvaev and GraevskiilO and Hagen and Leslief1 have devised methods involving the use of visual indicators for titrating organo-aluminium compounds with standard solutions of bases and ethers. Methods based on a conductiometricl q 5 or potentiometricl v 5 y 6 y 7 titration with organic bases, or the spectrophotometric method described, in which these reagents are useds v 9 are excellent for determining trialkyl-aluminium compounds and dialkyl aluminium chlorides (in both, only low concentrations of aluminium-bound hydride groups are assumed to be present). They cannot, however, be applied to the determination of dialkylaluminium alkoxide compounds, which do not co-ordinate with bases.In addition, these methods did not have the required sensitivity. Similarly, visual-indicator titration procedureslO could not be applied to the determination of dialkylaluminium alkoxide compounds as no reaction occurs between these compounds and organic bases or ethers. Bartkiewicz and Robinson1, have shown that a hexane solution of triethylaluminium consumes iodine according to the following equation, and have used this reaction for deter- mining the reducing capacity of triethyl aluminium- A1(C2H,), + 31, = AlI, + 3C2H5I It seemed that this reaction might offer the basis for a sensitive and rapid method for the analysis of organo-aluminium compounds. Based on this observation, the method discussed below has been developed for the determination of trialkylaluminium compounds and other types of organo-aluminium compounds in hydrocarbon solutions.EXPERIMENTAL The preliminary iodination experiments were carried out with a 5 per cent. solution of diethylaluminium chloride in anhydrous toluene. Several volumes of this solution were transferred by pipette into dry, nitrogen-purged reaction flasks (see Fig. 1) by using the tech- nique described by Crompton.13 A fixed volume (50 ml) of iodine reagent (0.4 N) was then added to each solution, and the mixture was left for 5 minutes to allow the reaction to proceed. C A = Burette with polyethylene stop- C = Glass stirrer B = B24-to-B24 adaptors with Gaco cock D = 250-ml, 3-neck (B24) flask seal Fig. 1. Apparatus for determining the iodine number Aqueous acetic acid was then added to each reaction mixture.The excess of iodine remaining was determined by titration with sodium thiosulphate solution and the amount of iodine consumed by the various sample volumes was calculated.376 CROMPTON IODIMETRIC DETERMINATION OF [Analyst, 1701. 91 A plot of iodine consumption against the volume of diethylaluminium chloride solution taken (Fig. 2, Procedure A) shows that the iodine consumption is not proportional to the volume of sample taken, and that the line drawn through the experimental points intersects the sample-volume axis at a positive value, indicating that low iodine consumptions are being obtained in these determinations. This suggested that the iodine reagent contained a small amount of an impurity that reacted rapidly with the alkyl groups in diethylaluminium chloride. Volume of sample solutioti, ml Fig.2. Iodine consumption of dilute diethylaluminium chloride solutions : graph A, 50 ml of iodine reagent added to sample; graph R, sample added to 50 ml of iodine reagent 6 taken, g Fig. 3. Influence of excess of iodine and reac- tion time on the iodine consumption: graph -1, triethylalumininm ; graph B, tripropylaluminium ; graph C, diethylaluminium chloride I t can be seen in Fig. 2 (Procedure B) that higher iodine consumptions are obtained when the order of mixing the diethylaluminium chloride solution and iodine is reversed, i.e., when the iodine is transferred by pipette into the reaction flask first. These conditions are, presumably, less favourable for the occurrence of the side reaction.However, the presence of the impurity still affects the determination of iodine consumption to some extent. The effect of the impurity was overcome by using a “double titration” procedure, described below. The iodine consumptions, I , g and I 2 g, of two different volumes, I/, ml and V , ml, of the sample solution are determined. The same volume of iodine reagent is used in each determination. The correct iodine consumption of Vl - I/, ml of sample solution is, therefore, equal to I , - 12.g of iodine. Several iodine-number determinations (grams of iodine consumed per 100 grams of sample) were carried out on a solution of diethylaluminium chloride by the “double titration” procedure, and these results are compared, in Table I, with the results obtained by the “single titration” procedure.The “double titration” results are consistently higher and do not vary appreciably with the amount of sample taken for analysis. TABLE I DIETHYLALUMINIUM CHLORIDE (Ti PER CENT. IN TOLUENE) : EFFECT OF DISSOLVED IMPUKITY IN THE IODINE REAGENT Iodine number, g of iodine consumed per 100 g of sample volume tincorrected Corrected A Sample - 7 taken, ml (“single titration”) (“double t i tration”) 2.5 3.0 5.0 7.5 22.7 26.2 26-7 - 28.6 28-5 28.1 28.4 The effect of excess of iodine on the iodine-number determination was examined. Varying volumes of dilute solutions of triethylaluminium, diethylaluminium chloride and tripropylaluminium were added to a fixed volume (50 ml) of the iodine reagent and theJune, 19661 ORGANO-ALUMINIUM COMPOUXDS 377 reaction was allowed to proceed for 10 minutes.It can be seen from Fig. 3 that when iodine consumption is plotted against sample size the iodine consumption is not affected, unless more than 80 per cent. of the iodine present is consumed in the reaction. These experiments were then repeated with dilute solutions of diethylaluminium ethoxide and dipropylaluminium isopropoxide. Fig. 4 shows that the iodine consumption of this less reactive type of organo- aluminium compound is more dependent upon the molar excess of iodine present. \Vith a 10-minute reaction period the iodine consumption is affected if more than 50 to 60 per cent. of the iodine is consumed. Extension of the reaction time to 20 minutes, however, circum- vents this effect. Complete iodination of all the types of organo-aluminium compounds examined is obtained in 20 minutes, even in the presence of only a 20 to 30 per cent.excess of iodine reagent. 5 Fig. 4. Influence of excess of iodine and reaction time on the iodine consumption : graph A, diethylaluminium ethoxide, 20 minutes’ reaction ; graph B, diethylaluminium ethoxide, 10 minutcs’ reaction ; graph C, dipropylaluminium propoxide, 5 minutes’ reaction In the experiments described so far, dilute aqueous acetic acid (2 N) has been added to the mixture of iodine reagent and sample before back-titrating the excess iodine with standard sodium thiosulphate solution. It was observed, however, that considerably higher iodine consumptions occurred for dialkylaluminium alkoxides if distilled water only was added at this stage.I t did not occur with trialkyl- aluminium and dialkylaluminium chloride compounds, and secmed to be connected in some way with the presence of alkoxide groups in the molecule. I t was decided to use aqueous acetic acid in the final analytical procedure. Xo explanation was found for this effect. STO I C HE I 0 $1 E TK Y 0 F THE I 0 D I SAT1 0 N 0 F 0 RG AN 0- AL U M I N I U 31 C OM PO V N L) S- The stoicheiometry of the reactions that occur during the iodination of iso-octane solu- tions of the purest available specimens of various organo-aluminium compounds was then examined. Ethyl, propyl, butyl, hydride and alkoxide groups in these samples were deter- mined by the procedure described by Crompton and Reid.3J3 Alkyl groups higher than butyl were determined by a procedure in which a cold toluene solution of the organo- aluminium compound was decomposed at -60” C by the addition of a dilute solution of glacial acetic acid in toluene.Aqueous sodium hydroxide was then added to the solution. Liquid paraffins, produced by the hydrolysis of the higher alkyl groups, were then determined in the separated toluene phase by gas chromatography. In Table I1 are shown the values of the determined iodine consumptions of iso-octane solutions of various organo-aluminium compounds. Each alkyl group in trialkylaluminium compounds consumes one mole of iodine. These results confirm the conclusions reached by Bartkiewicz and Robinson.lZ Similarly, each alkyl group in diethylaluminium chloride consumes one mole of iodine.The reaction between iodine and dialkylaluminium alkoxides follows a different course, however, as only 1.25 moles of iodine are consumed per mole of this type of compound, i.e., 0.625 moles of iodine are consumed per alkyl group.378 CROMPTON IODIMETRIC DETERMINATION OF [Analyst, VOl. 91 TABLE I1 STOICHEIOMETRY OF THE IODINATION OF THE ORGANO-ALUMINIUM COMPOUND Sample description Triethylaluminium .. . . . . Tripropylaluminium . . .. Diethylaluminium chloride .. .. Dipropylaluminium isopropoxide . . Composition of sample, w/w per cent. A1(C2H5)3 A1(C2H5)2H (C,H5)2(C,H,) A1(C2H5)2(0C2H5) 4.3 1.4 2-2 J 93.0 3.0 4.0 } Iodine consumption, moles of iodine consumed per mole of organo-aluminium compound 2.97 3.09 2.00 1.25 REPRODUCIBILITY OF THE PROCEDUKE- The reproducibility of the method for the determination of the iodine consumption was determined by statistical analysis.Iodine-consumption determinations were made with 6 different sample volumes of dilute solutions of 3 typical organo-aluminium compounds. The mean iodine number of samples, its standard error and its 95 per cent. confidence limits are shown in Table 111. I t can be seen that, for the 3 organo-aluminium compounds examined, the standard errors of the iodine-number determinations are acceptably low. TABLE I11 REPRODUCIBILITY OF IODINE-NUMBER DETERMIXATIONS Mean iodine number, g of iodine 95 per cent. Kumber of consumed per Standard confidence Sample description determinations 100 g of sample error limits Triethylaluminium . . . . 6 613.8 2.6 613.8 & 7.3 Diethylaluminium ethoxide .. 6 158.6 1.9 158.6 & 5.3 156.6 & 3.1 Dibu t ylaluminium 6 156.5 1.1 ethoxide . . . . , . 6 42.9 0.3 42.9 & 0.9 31 E TH o D APPARATUS- inlet sidearm and stopcock above the graduation mark. DiZutio7z jasks-These are 100-ml stoppered Pyrex-glass calibrated flasks with nitrogen Safety Pipettes-These are “Exe1o”-type plunger pipettes of 1, 2, 5, 10 and 25-ml capacity. T-Pieces ; glass, 3 iitches. Graduated cylinder, 50 ml. Burette, 50 mZ-This is preferably fitted with an E-MIL polythene stopcock (obtainable Reactio~zjask, 250 ml-This is a B24, three-neck flask with glass stirrer and nitrogen and from H. J. Elliott Limited, Treforest Industrial Estate, Pontypridd, Glamorgan). burette inlets (see Fig. 1).REAGENTS- able from British Drug Houses Ltd.) for 2 weeks. than 25 p.p.m. Iso-octane-Dry iso-octane by standing it over 50 g of molecular sieve, type 4A (obtain- Nitrogen-Dry by passing through a molecular-sieve packed tower, oxygen content less Swirl the bottle daily.June, 19661 ORGANO-ALUMINIUM COMPOUNDS 379 Iodine reagent (0-4 N)-To 2-5 litres of toluene in a dry bottle add 130 g of analytical- reagent grade iodine and shake the contents to dissolve the iodine. Add to the solution 50 g of freshly heated (at 120” C) 4A molecular sieves, stopper the flask, and leave it for several days, occasionally swirling it. Sodium thiosulphate (0-25 N) aqueous, standardised. Acetic acid (4 N) aqueous-Dilute 250 ml of glacial acetic acid to 1 litre with distilled water.SAMPLING- If the sample contains more than 20 per cent. of organo-aluminium compound it is necessary to dilute the solution with iso-octane in the following way. Transfer by pipette 20 ml of dry iso-octane into a dry 100-ml calibrated flask with a nitrogen inlet side-arm, and purge the solvent with nitrogen for 30 seconds. Connect a nitrogen line to the side-arm of the calibrated flask, open the stopcock and apply a gentle nitrogen purge. Transfer sufficient of the sample into the calibrated flask, by means of a safety pipette, to give a concentration of approximately 20 per cent. of the organo-aluminium compound in the diluted solution. Purge the exterior of the tip of the safety pipette with dry nitrogen during the transfer, as described by Crompton.13 Make up the volume to 100 ml with dry iso-octane, stopper the flask and mix the contents thoroughly .PROCEDURE- With a pipette fitted with a rubber suction bulb, transfer 50 ml of the same batch of iodine reagent into two dry 250-ml reaction flasks. Apply a gentle purge of nitrogen to displace the air from the flasks. Switch on the stirrers and adjust the speed to approximately 1 revolution per second. Transfer a different volume of the sample solution into each flask by means of a safety pipette. Observe the precautions described above to prevent decompo- sition of the sample during transfer. Stopper the reaction flasks immediately after the sample delivery. Suitable pairs of sample volumes required for the analysis of a 200 millimole per litre solution, of various types of organo-aluminium compounds, are shown in Table VIII.Corres- pondingly larger or smaller volumes should be taken if necessary. Maintain the gentle nitrogen purge during the subsequent reaction. TABLE VIII OPTIMUM PAIRS OF SAMPLE VOLUMES REQUIRED FOR THE ANALYSIS OF A 200 MILLIMOLES PER LITRE SOLUTION OF VARIOUS TYPES OF ORGANO-ALUMINIUM COMPOUNDS Sample volumes required* , 1 Type of organo-aluminium “A” “B” Trialkylaluminium compounds . . . . 6 12 compound analysed nil ml Dialkylaluminiuni chlorides . . . . 9 18 Diallq-laluminium allioxides . . . . 12-5 25 * The following relationships are used to calculate the sample volumes required :- 1 mole of trialkylaluminium compound = 6 x 126-9 g of iodine 1 mole of dialkylaluminium chloride = 4 x 126-9 g of iodine 1 mole of dialkylaluminium alkoxide = 2.5 x 126.9 g of iodine Let the reaction proceed for 20 minutes, then remove the nitrogen supply and introduce 40 ml of 2 N acetic acid into each reaction flask.Increase the stirrer speed until the aqueous and toluene phases are thoroughly mixed and then titrate the solution with 0.25 N sodium thiosulphate solution. Continue the titration until the solution becomes pale brown in colour. Commence drop-wise titration and stop the stirrer between each addition of titrant. Continue the titration until the pink colour completely disappears from the toluene phase and the toluene becomes pale yellow.380 CROMPTON IODIMETRIC DETERMINATION OF [,4?lnlyst, LT0l. 91 CALCULATIONS- Iodine consumption (g of iodine consumed per litre of sample) (TI - T,) x f x 126.9 x 1000 - - (V, - V,) x 1000 Trialkylaluminium compounds (millimoles of trialkylaluminium compound per litre of sample) Dialkylaluminium chlorides (millimoles of dialkylaluminium chloride per litre of sample) - (T, - T,) x f x 1000 Y 2 - Vl) x 4 Dialkylaluminium alkoxides (millimoles of dialkylaluminium alkoxide per litre of sample) ( T , - 7-2) x f x 1000 - - ( V , - V1) x 2.5 Where V , = Volume of sample solution taken (smaller volume), ml.V , = Volume of sample solution taken (larger volume), ml. T , = Back-titration of sodium thiosulphate obtained with smaller sample volume, T, = Back-titration of sodium thiosulphate obtained with larger sample volume, ml. f ml. = Normality of sodium thiosulphate solution. DISCUSSION OF RESULTS ANALYSIS OF DILUTE HYDROCARBON SOLUTIONS OF TRIALKYLALUMINIUM COMPOLXDS- The catalyst contents of dilute hydrocarbon solutions of various trialkylaluminium compounds was determined iodimetrically, by the conductiometric titration with isoquinoline and by the determination of aluminium.It can be seen in Table IV that reasonably good agreement is obtained between the iodimetric and the isoquinoline methods of analysis. The iodimetric method can be applied to solutions containing as little as 20 millimoles per litre of catalyst. TABLE IV DILUTE TKIALKYLALUMINIUM SAMPLES Sample description Triethylaluminium in iso-octane . . . . Triethylaluminium in iso-octane . . . . Tripropylaluminium in iso-octane . . . . Tripropylaluminium in iso-octanc . . . . Tripropylaluminium in iso-octanc .. . . Tripropylaluminium in iso-octane . . . . r Trialkylaluminium content, millimoles per litre Rased on Based on Based on determination consumption consumption 715 687 684, 673 1353 1314 1305 314 294 309 105 -* 104 A 7 aluminium isoquinoline iodine 62-8 -* 61.3 20.1 -* 19.6 * Isoquinoline method is not applicable because of the low electrical conductivity of the test solution. Commercial trialkylaluminium usually contains a small amount of dialkylaluminium alkoxide as an impurity, which is produced bv oxygen contamination during the manu- facture. An isoquinoline titration determines only the “active” trialkylaluminium content of the sample ; aluminium determinations include both “active” trialkylaluminium and “inactive” dialkylaluminium alkoxide.Higher results are therefore expected, and indeed found, in the latter method of analysis. The presence of small amounts of dialkylaluminium alkoxide in trialkylaluminium compounds causes little interference in the iodimetric method. Thus, the determined iodine number of a solution known to contain 180 millimoles of trialkylaluminium compound per litre and 20 millimoles of dialkylaluminium alkoxide per litre (i.e., total organo-aluminium content of sample contains 10 per cent. of the alkoxide derivative), indicates a trialkylaluminium content of 187 millimoles per litre, which is about 4 per cent. higher than the added amount.June, 19661 ORGANO-ALUMINIUM COMPOUNDS 381 ANALYSIS OF DILUTE HYDROCARBON SOLUTIONS OF DIALKYLALUMINIUM CHLORIDES- It can be seen from Table V that good agreement is obtained between the iodimetric and the isoquinoline methods of analysis. Diethylaluminium chloride usually contains a maximum total of 5 per cent. of ethylaluminium chloro-ethoxide and triethylaluminium or ethylaluminium dichloride as impurity.The presence of these contaminants at this level of concentration does not interfere appreciably in the iodimetric determination of diethyl- aluminium chloride. TABLE V DILUTE DIETHYLALUMINIUM CHLORIDE SAMPLES Diethylaluminium chloride content, millimoles per litre n I v Based on aluminium Based on isoquinoline Based on iodine determination consumption consumption 218 198 197 192 172 177. 179 ANALYSIS OF DILUTE HYDROCARBON SOLUTIONS OF DIALKYLALUMINIUM ALKOXIDES- Depending upon the method of manufacture used, dialkylaluminium alkoxide catalysts might contain small amounts of either trialkylaluminium or alkylaluminium dialkoxide as impurity.The dialkylaluminium alkoxide content of several dilute hydrocarbon solutions was determined iodimetrically . These values are compared in Table VI with results obtained by aluminium determinations. Good agreement was obtained between the two methods when the total organo-aluminium content of the test solution contained less than 5 per cent. of the previously mentioned impurities (sample A). As expected, poorer agreement was obtained for a solution of dipropylaluminium isopropoxide which contained an appreciable amount of propylaluminium di-isopropoxide as impurity (sample B) . Commercial preparations of dialkylaluminium alkoxides usually contain less than 5 per cent.of their total organo- aluminium content in the form of alkylaluminium dialkoxide or trialkylaluminium impurity and no serious interference from these impurities is therefore to be anticipated. TABLE VI DILUTE DIALKYLALUMINIUM ALKOXIDE SAMPLES Calculated as dialkylaluminium alkoxide, millimoles per litre 7- -- A t Based on Based on aluminium iodimetric Differences, Sample description determination determination per cent. Sample -4 : AlR,(OR) containing less than 5 per cent. of AlR, or -\lR(OR), impurity Diethylaluminium ethoxide in iso-octane . . 795 795 Nil Sample B: AlR,(OR) containing 25 per cent. Dipropylaluminium isopropoxide in iso- of =ZlR(OR), impurity octanc . . .. . . .. . . 272 273 238 - 12-5 The iodimetric method of analysis also presents a method for detecting whether a change in the composition of stocks of hydrocarbon solutions of organo-aluminium compounds has occurred during storage. This could arise from the contamination of the material with extraneous oxygen and/or water, thus leading to a reduction in the iodine number because of alkyl group decomposition as follows- 2 > A1R + 0, - - - 2 > Al(0R) > A1R + H,O - - - > Al(0H) + RH Regular iodimetric determinations present a method, therefore, of detecting whether contamination of the organo-aluminium compound occurs to any extent during storage.Such information could not be obtained from aluminium determinations alone, as these remain virtually unchanged even when the sample has become heavily contaminated.383 CROMPTON [Analyst, 1701. 91 APPLICATION OF THE IODIMETRIC METHOD TO HIGHER MOLECULAK WEIGHT ORGAXO-ALUMINIUM The iodimetric method was applied to a solution of impure trihexadecylaluminium in The usual stoicheiometry was assumed in the reaction of the components of this It is seen in Table VII that reasonable agreement is obtained between COMPOUNDS- toluene. sample with iodine. the expected and the found iodine consumptions of this substance. TABLE VII APPLICATION OF THE IODIMETRIC METHOD TO TRIHEXADECYLALUMINIUM Iodine consumption, g of iodine per 100 g of sample Composition of sample, r--7 w/w per cent. Expected Found 16H33) 47-8 4.0 4.5 9.2 1.6 32.0 55.7 54.0 - Total 99.1 - The author thanks the Ilirectors of Shell Chemical Company Limited for permission to publish this paper. 1. 2 . 3. 4. 5. 6. 7 . 8. 9. 10. 11. 12. 13. RE FE RE 5 c E s Bonitz, E., Chem. Ber., 1955, 88, 742. Ziegler, I<., Justus Liebags Annln Chem., 1954, 589, 91. Cronipton, T. R., and Reid, V. W., Analyst, 1963, 88, 713. Dijkstra, R., and Dahmen, E. ,4. M., 2. anal~rt. Chem., 1961, 181, 399. Neuniann, W, P., Justus Liebzgs Annln Chem., 1960, 629, 23. Farina, M., Donati, M., and liagazzini, RI., Aiznalz Chim., 1958, 48, 501. Nebbia, L., and Pagani, R., Chzmaca Ind., iikfzlano, 1962, 44, 383. Illitchen, J . H., Analyt. Chem., 1961, 33, 1331. \\'adelin, C. W., Talanta, 1963, 10, 917. Razuvaev, G. A., and Graevskii, X. I . , Dokl. Akad. iYauk SSSR, 1959, 128, 309. Hagen, D. F., and Leslie, W. D., AnalJit. Chem., 1963, 35, 814. Bartkiewicz, S. -4., and Robinson, J . W'., Analytzca Chiin. A d a , 1969, 20, 326. Crompton, T. R., Analyst, 1961, 86, 65%. Received Jirly 14th, 1965