摘要:

AUGUST, 1968 THE ANALYST Vol. 93, No. 1109 Chromatographic Methods for the Identification of the New Hallucinogen, 4-Methyl=2,5dimethoxy-a-methyl- phenethylamine, and Related Drugs BY K. GENEST AND D. W. HUGHES (Ir'cseavch Laboratories, Food and Drug Directorate, Ottawa, Ontario, Canada) Thin-layer and gas-chromatographic methods are described for the detection of 4-methyl-2,5-dimethoxy-a-methylphenethylamine in the sub- microgram range. Cliromatographic criteria of identity for this new hallu- cinogen and for the related amines, amphetamine, methamphetamine and mescaline, and thc hallucinogens, dimethyltryptamine and bufotenine, are also reported. 1 HE ever-increasing number of hallucinogens, and their expanding illicit use, make the need for reliable methods of identification imperative.While adequate methods are available for compounds such as LSDl ,2e3 9 4 and hallucinogens derived from natural product^,^,^,^ newer compounds in this group have not been covered as thoroughly. Recently, it was reported*pa that many persons who allegedly used a material known colloquially as STP suffered severe psychotomimetic reactions. Some samples, identified as STP, were shown by the U.S. Food and Drug Administration to contain 4-met hyl-2,5-dimethoxy-c-methylphenethylamine. Chemically, this compound is closely related to the physiologically active amines, mescaline and amphetamine, and it has been proposed that it be included among the drugs controlled under the U.S. Drug Abuses Control Amendments of 1966. Establishment of identity for forensic purposes frequently requires the application of two or more different analytical techniques, and it is the purpose of this paper to report methods for the identification by thin-layer and gas chromatography of 4-methyl-2,5-dimethoxy-cc-methylphenethylamine alone, and in the presence of amphetamine, methamphetamine, mescaline and two other hallucinogens, buf o tenine and dimet li ylt rypt amine.,. EXPERIMENTAL MATERIALS- For thin-layer chromatography, solutions (1 pg per pl) of 4-methyl-2,5-dimethoxy- K-methylphenethylamine hydroclilmide (Dow Chemical Company), m-a-methylphenethyl- amine sulphate (amphetamine) (Smith, Kline and French Laboratories), N-cx-dimethyl- phencthylamine liydrochloride (methamphetamine) (Abbott Laboratories), 3,4,5-trimethoxy- plieneth ylainine sulpliate (mescaline) (California Biochem.) , NN-diniethyltryptaniine (Kocli- Light Laboratories Ltcl.) and 6-hydroxy-NN-dimethyltryptaniine (bufotenine) (California 13iochem.) were used.For gas chromatography, free bases were prepared from aqueous ammoniacal solutions of salts by exhaustive extraction with diethyl ether (benzene was used with mescaline). Solutions (1 pg per pl) of NN-dimethyltryptamine and bufotenine in methanol, and of niescaline in benzene, were used. All of the other bases were dissolved in diethyl ether. THIN-LAYER CHROMATOGRAPHY PROCF: n u IIE - Glass plates, 20 x 20 cni, were coated with 250-p thick layers made from slurries of silica gel G (E. Merck) (30g) and water (65ml) or alumina G (E. Merck) (30g) and water (55 ml), with the aid of a Desaga apparatus according to Stahl.The plates were activated before use at 110" C for 1 hour. 0 SAC and the authors. 485486 GENEST AND HUGHES CHROPIIATOG1<Al'HIC METHODS FOR [Analyst, Vol. 93 CHROMATOGRAPHIC SYSTEMS- System I. Ethyl methyl ketone - diinethylformamide - ammonia solution (sp.gr. 0-90) System 11. System 111. (13 + 1.9 + 0.1) on silica gel G. Methanol - chloroform (1 + 1) on alumina G. Chloroform - methanol - acetic acid (75 + 20 + 5) on silica gel G. DETECTION REAGENTS- Reaged A-This consists of the reagents (i) sodium acetate (10 per cent. aqueous), (ii) 2,G-dibromo-~-benzoquinone-4-cl~lorimine (B.D.H. Ltd.) (1 per cent. in ethanol) and (iii) iodine crystals (2 g distributed in two small Petri dishes placed at the bottom of a thin- layer chromatographic tank.[(i) and (ii) remain usable for 5 days if kept in a refrigerator.] The dried plates were sprayed lightly with (i), then immediately with (ii), carefully avoiding overspray. Coloured spots appeared promptly. Plates chromatographed in system I11 were aerated until the acidic background had become dispelled and then sprayed twice with (i) before application of (ii) spray and exposure to iodine vapour. They were then placed in the iodine tank. p- Dimethylamino benzaldehyde.5 Modijed Pvochdxka reageizt.1° Potassium iodoplatinate.ll GAS CHROMATOGRAPHY APPARATUS- A Varian-Aerograph Gas Chromatograph, Model 1520-B, equipped with a Varian- Aerograph Recorder, Model 20, a disc integrator and a hydrogen flame-ionisation detector, was used.It was operated both isothermally and with a matrix multi-linear programmer. Stainless-steel columns, 5 feet x inch, were packed with 5 per cent. silicone rubber S.E. 30 on 60 to 80-mesh Chromosorb W, or 3 per cent. silicone GE X.E 60 on 100 to 120-mesh Aeropak 30. The carrier gas used was nitrogen with a flow-rate of 25 ml per minute at 200" C. The injection port and detector were maintained a t temperatures of 50" and 25" C higher, respectively, than that of the column. Gas flow-rates for the flame-ionisation detector were 25 ml per minute for hydrogen and 300 ml per minute for air. The X.E. 60 column was conditioned overnight at 240" C with 10 ml per minute gas flow. The S.E. 30 column was conditioned overnight at 250" C with no flow, then at 240" C with 25 ml per minute gas flow until a stable base-line was obtained.PROCEDURE- Samples (0.5 to 4 4 ) were injected as free bases in diethyl ether, benzene or methanol solution. Salts of the bases and other solvents, such as chloroform, could also be used, but more frequent cleaning of the flame-ionisation detector was necessary with these. Retention times reported are averages (+Om1 minute) from six chromatograms of each compound. REACTION WITH CARBON DISULPHIDE- Ten to twenty micrograms of free base in diethyl ether (10 to 20 pl) (benzene was used for mescaline) were transferred into a small test-tube and an equal volume of carbon di- sulphide was added. The solution was left at room temperature (26" C) and, after timed intervals, or after 5 minutes with mixtures, 1 to 4 pl were injected into the gas chromatograph. RESULTS AND DISCUSSION THIN-LAYER CHROMATOGRAPHY- 4-Methyl-2,5-dimethoxy-a-methylphenethylamine could be separated from the related compounds in all of the systems (Table I).The low sensitivity of non-phenolic phenylamines towards general alkaloidal reagents, such as Dragendorff's or potassium iodoplatinate, or other non-specific reagents is well known,10912913914,16 and identification of mescaline has been considerably improved by application of a modified Prochgzka reagent and observation under ultraviolet light .lo Strong fluorescence was also observed for the new hallucinogen withAugust, 19681 THE IDENTIFICATIOK OF THE NE\V HALLUCINOGEN, STP TABLE I R, VALUES* IN THREE THIN-LAYER CHROMATOGRAPHIC SYSTEMS AND COLOITR WITH DETECTION REAGENT A 487 System - Compound I I1 Amphetamine .. .. . . . . .. . . 0.58 0.40 Methamphetamine . . . . . . . . . . 0.14 0-74 4-Methyl-2,5-diniethoxy-a-methylplienethylamine 0.54 0.48 Mescaline . . .. . . . . . . . . 0.49 0.24 NN-Dimethyltryptaminc . . . . . . . . 0.35 0.93 Bufotenine . . . . . . .. . . . . 0.15 0.84 * Average (40.02) of six measurements. 7 .- 111 Colour 0-2 1 Violet 0.33 Yellow 0-33 Yellow 0.20 0.10 Orange - broxvii 0.05 Dark grey this reagent, but the reaction was less sensitive than for mescaline (Table 11). The formation of isoquinoline derivatives, which has been suggested in the reaction of formalin with mescaline under the conditions of the thin-layer chromatographic spray,1° could be impeded in 4-methyl- 2,5-dimethoxy-a-methylphenethylamine, either because of steric hindrance caused by the a-methyl group or because of the electronic effect of the 5-methoxy group.Detection reagent A produced different colours with tlie various compounds (Table I) and was, further, the most wnsitive detection reagent used in this investigation (Table 11). Presence of the indole TABLE I1 LIMIT OF DETECTION (pg) FOR 4-METHYL-2,5-1lIMETHOXY-dl-METHY LPHENETHYLAMINE Reagent Reagent 2 1 . , .. . . Thin-layer chromatographic system tLg .. I 0-3 I1 0.03 Modified ProchAzka reagent .. Potassium iodoplatinate . . . . I I1 I I1 0.4 0.1 2.0 1.0 derivatives, NN-dimethyltryptatnine and bufotenine, was ascertained by spraying with 9-dimethylamino1)enzaldehyde reagent. Other sprays tried, such as ninhydrin, Fast blue B, dansyl chloride, potassium permanganate, sodium tetraphenylborate - fisetin and $-nitro- aniline, showed no advantages over those reported under Detection reagents.TABLE I11 GAS-CHROMATOGRAPHIC RETENTION TIMES (MINUTES) OF IIAT.LI'CINOGENli AND RELATED BASES ON TWO COLUMNS Columns r 7 -A S.E. 30, 5 per cent. X.E. 60, 3 pcr cent. 1-7 r \ Temperature, "C .. .. 130 180 205 220 110 150 180 210 340 Compounds - - Amphetamine . . .. . . 2.05 0.85 0.7 - 3.1 1.1 - Methamphetamine . . . . 2.6 0.9 0.8 - 3.2 1.1 - - - 4-Methyl-2, 5-dimethoxy-a- methylphenethylamine . . 8-2 2.2 1.1 - >40 9.2 3.2 1.5 0.8 Mescaline .. . . . . - 6.5 2.5 - - >40 9.0 3.0 1.2 NN-Dimethyltryptamine . . - 6.7 3.4 2.1 - - 12-5 4.6 1.85 8.5 5.0 - - - 34.5 10-3 Bufotenine .. .. .. - I GAS CHROMATOGRAPHY- Results of isothermal gas chromatography of the six compounds on 5 per cent. S.E. 30 and 3 per cent. X.E. 60 are shown in Table 111. All of the compounds gave sharp peaks on both columns. For mixtures and temperature-programmed analyses the X.E. 60 column is458 GENEST AND HUGHES : CHROMATOGRAPHIC METHODS FOR [A?tdyst, Vol. 93 preferred. 4-Methyl-2,5-dimethoxy-a-methylyhenethylaniine is well separated from the other compounds (Fig. 1). Adequate separation of amphetamine and methamphetamine can be obtained on the S.E. 30 coluinn at 130" C or on the same column with lower loads of licliiid phase16J7; 0.02 pg of the new hallucinogen can be detected on both coliimns. 4 D i - 5 10 I Retention time, minutes Fig. 1. Programmed gas chromatogram of a mixture of 1 pg each of amphetamine and meth- amphetamine (A), 2 p g of 4-methyl- 2,5-dimethoxy-a-methylphenethyl- amine (B), 1 pg of mescaline (C), 2 pg of dimethyltryptamine (D) and 2 pg of bufotenine (E); 3 per cent.X.E. 60, single-column opera- tion, flash heater at 220" C, flame- ionisation detector at 260" C, range 1, attenuation 64 and flow-rate of nitrogen, 25 ml per minute. Pro- gramme : initial temperature 100' C; 2/20" C; 3/hold; 4/30' C; S / S o C; 12/15' C; 16/end (numbers before oblique represent minutes after injection, and those after represent the increase in temper- ature in degrees per minute) Additional analytical information can be obtained by gas chromatography of derivatives produced by reaction of the primary bases with carbon disulphide at room temperature.Methamphetamine, NN-dimethyltryptamine and bufotenine show no reaction, but 4-methyl- 2,5-dimethoxy-a-methylphenethylamine, mescaline and amphetamine, after brief treatment with carbon disulphide, yield a derivative generating a distinct peak with higher retention time on either column (Table IV). Formation of the derivative can also be achieved by "on-column reaction gas chromatography," i.e., injection of 1 to 2 pg of the free base, followed, within 10 to 20 seconds, by 3 to 4p1 of carbon disulphide. As retention times obtained by this method are, however, often distorted and conversions incomplete, the procedure as described under Experimental is preferred. Differences were observed in the rate of conver- sion of the three amines that form derivatives with carbon disulphide.The reaction of mescaline is rapid, being more than 99 per cent. complete after 20 seconds. The reactions of amphetamine and 4-methyl-2,5-dimetlioxy-a-methylphenetl~ylamine with carbon disulphideAugust, 19681 THE IDENTIFICATION OF THE NEW HALLUCINOGEN, STP 489 are slower. Approximate conversion rates were determined by comparing the integrated areas for the peaks of the free bases and their derivatives. In this manner, it was established that after 20 seconds, 2.5, 16 and 40 minutes, and 16 hours of contact with carbon disulphide, 51,56,71 , 87 and 98 per cent. of amphetamine and 47,53,59, 67 and 99 per cent. of 4-methyl- 2,5-dimethoxy-6c-methylphenethylamine were converted into the corresponding derivatives.TABLE IV GAS-CHROMATOGRAPHIC RETENTION TIMES (MINUTES) OF ISOTHIOCYANATES ON TWO COLUMNS Columns L r > S.E. 30, 5 per cent. X.E. 60, 3 per cent. Temperature, O C . . .. .. 13-0 il0 150 :SO 210 220 Compounds Amphetamine - carbon disul- 4-Methy1-2,5-dimethoxy-a- methylphenethylamine - car- Mescaline - carbon disulphide . . phide . . .. .. . . 8.7 2.2 1.5 - 25.5 5.4 2.2 0.8 - - - 11.2 4.0 1.8 bon disulphide . . .. - 10.7 5.0 - - - - 36.8 10.2 3.1 16.1 6.9 - The reaction of primary amines with carbon disulphide has been applied to the detection of amphetamine in urine,l* and it has been shown that the product of this reaction is the mustard oil, ac-methylphenethylis~thiocyanate,~S which is formed via the dithiocarbamic acid. Mescaline and Pmethyl-2,5-dirnethoxy-a-methylphenethylamine probably follow an analogous course of reaction, and the formation of 2,4,5-trimethoxyphenethylisothiocyanate and 4-methyl-2,5-dimethoxy-a-methylphenethylisothiocyanate is likely under the experimental conditions.We are indebted to the Dow Chemical Company for the gift of a sample, and to Mr. G. Belec for technical assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. REFERENCES Genest, K., and Farmilo, C. G.. J . Pharm. Pharmac., 1964, 16, 250. Dal Cortivo, L. A., Broich, J. R., Dihrberg, A., and Newman, B., Analyt. Chem., 1966, 38, 1969. Lerner, M., Bull. Narcot., 1967, 19, No. 3, 39. Hale, D., and Taylor, C., Perkin-Elmer Instrum. News Sci. Ind., 1967, Nov. 18, p. 6. Genest, K., J . Chromat., 1965, 19, 531. Heim, R., Genest, K., Hughes, D. W., and Belec, G., J . Forens. Sci. SOL, 1966, 6, 192. Der Marderosian, A. H., Lloydia, 1967, 30, 23. Chem. Engng News, 1967, 45, 39. Snyder, S. H., Faillace, L., and Hollister, L., Science, N.Y., 1967, 158, 669. Seiler, N., and Wiechmann, M., Hoppe-Seyler’s 2. physiol. Chem., 1964, 337, 229. Farmilo, C. G., and Genest, K., in Stewart, C . P., and Stolman, A., Editors, “Toxicology, Mechanisms and Analytical Methods,” Academic Press, New York, Volume 11, 1961, p. 576 H3. Noirfalise, A., J . Chromat., 1965, 20, 61. Morrison, J. C., and Om, J. M., J . Pharm. SC~., 1966, 55, 936. Paris, R. R., Sarsunova, M., and Semonsky, M., Annls Pharm. Fr., 1967, 25, 177. Heyndrickx, A., and De Leenheer, A., J . Pharm. Belg., 1967, 22, 109. Brochmann-Hanssen, E., and Baerheim Svendsen, A., J . Pharm. Sci., 1962, 51, 938. Kazyak, L., and Knoblock, E. C., Analyt. Chem., 1963, 35, 1448. Hall, C. R., Cordova, V., and Rieders, F., Pharmacologist, 1965, 7, 148. Brandenberger, H., and Hellbach, E., Helv. Chim. A&, 1967, 50, 958. Received December 16th, 1967

ISSN:0003-2654    

DOI:10.1039/AN9689300485

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

490 AfiaZyst, August, 1968, Vol. 93, $9. 490-493 Di-2-pyridyl Ketoxime, a New Reagent for the Rapid Gravimetric Determination of Palladium BY WILLIAM J. H O L L A N D , JOHN ROZIC AND JESSE GERARD (Department of Chemistry, University of Windsor, Windsor, Ontario, Canada) Di-2-pyridyl ketoxime is proposed for the gravimetric determination of palladium. The chelate is precipitated over a pH range of between 3 and 11 and weighed as Pd(C,1H,0N,)2. An extensive study of foreign ions and analytical conditions for precipitation indicates the procedure to be widely applicable, rapid and simple. NUMEROUS precipitating reagents have been proposed for the gravimetric determination of palladium, and have recently been excellently reviewed by Beamish.l The most important reagents for palladium contain the oxime grouping, C=N-OH.Dimethylglyoxime,2 cyclohexane-1 ,2-dione dioximeS (nioxime) and /3-furfuraldoxime4 are among the well estab- lished reagents in this class. More recently 2-selenophene aldoxime,b di-2-thienyl ketoxime6 and 8-mercapt~quinoline~ have been proposed as new gravimetric reagents for palladium. The purpose of the present work was to extend the application of di-2-pyridyl ketoxime to other analytical problems. This reagent has been investigated previously, and subsequently proposed for the spectrophotometric determinations of gold,* palladi~m,~ cobaltlo and iron.ll Further studies indicated that over a wide pH range, di-2-pyridyl ketoxime and bivalent palladium formed bis(di-2-pyridyl ketoxime) - palladium( 11) in quantitative yield- In this paper, the above reaction has been applied to the gravimetric determination of bivalent palladium in the presence of many diverse cations and anions.The palladium chelate is insoluble in hot water, while the reagent is completely soluble. Consequently, filtration from a hot solution, after a brief digestion period, results in a procedure that is rapid, with no loss of accuracy. Pd2+ + 2C11H90N, -+ Pd(C11H,ON3), + 2H+. EXPERIMENTAL PREPARATION OF THE REAGENT- Di-2-pyridyl ketoxime was synthesised by the method outlined in a previous publi~ation.~ A 1 per cent. solution in 95 per cent. ethanol, which is stable indefinitely, was used as the reagent. Conversion of di-2-pyridyl ketone (available from the Aldrich Chemical Co., Milwaukee, Wisconsin) into the oxime proceeds rapidly and gives a high yield.PREPARATION OF PALLADIUM SOLUTION- A stock solution was prepared by dissolving purified palladium(I1) chloride in concen- trated hydrochloric acid and diluting to 1 litre in a calibrated flask. The concentration of palladium(I1) was 1*002 mg per ml, as determined by the dimethylglyoxime method.12 'All other chemicals used were of analytical-reagent quality. RECOMMENDED PROCEDURE FOR THE DETERMINATION OF PALLADIUM- Transfer an aliquot of palladium solution, containing up to 50 mg of palladium(II), into a 250-ml beaker. Add about 2.5 ml of 40 per cent. sodium citrate (1 g) and adjust the pH to between 4 and 11. Heat the contents to about 80" C and add an excess of reagent solution (10 to 200 per cent. excess).Digest the hot solution for about 10 minutes, filter through a sintered-glass crucible of medium porosity, and wash the precipitate well with hot water. Dry it at 125" C to constant weight and weigh as Pd(C,lH,0N3)2. The gravimetric factor for palladium is 0.2116. 0 SAC and the authors.HOLLAND, BOZIC AND GERARD 491 EFFECT OF pH ON PRECIPITATION- The yellow chelate was precipitated from solutions of various pH values according to the procedure outlined, and dilute potassium hydroxide or hydrochloric acid used to make the pH adjustments. pH of solution . . . . 3.0 4.0 5.0 6-0 7.0 9.0 11.0 Palladium taken, mg . . 8-80 8.80 8-80 8.80 8.80 8.80 8-80 Palladium found, mg . . 8-82 8.79 8.80 8.79 8.81 8-83 8.76 The results indicate that the optimum pH conditions are between 3 and 11.However, because the precipitate was difficult to filter at pH 3, it is recommended that the pH be maintained above 4.0. EFFECT OF PALLADIUM CONCENTRATION- Analyses conducted on a series of different sample weights, as indicated below, showed that palladium could be successfully determined over the range of 4 to 50mg without co- precipitation of reagent. Because of the voluminous nature of the precipitate, the amount of palladium should not exceed 50mg. Palladium taken, mg . . . . 4.402 13.21 17.61 22.01 30.81 44.02 Palladium found, mg . . . . 4.395 13.19 17.60 22.03 30.83 44.06 Deviation, mg . . .. . . -0.007 -0.02 -0.01 +0.02 +0*02 +0*04 Each result is the average of six separate analyses. EFFECT OF REAGENT EXCESS- Various amounts of di-2-pyridyl ketoxime in excess were added to 13-21 mg of palladium.The results indicated that no co-precipitation occurs, even with 200 per cent. excess of reagent. EFFECT OF DIGESTION TIME AND TEMPERATURE- Recoveries of palladium for various digestion temperatures and times indicate that precipitation and filtration can be carried out at room temperature, or at elevated temperatures up to 100" C, with a brief digestion period of 5 to 10 minutes. DRYING CONDITIONS- A weighed amount of the chelate was isolated according to the procedure and dried at various oven temperatures and time periods. No significant weight change occurred when the chelate was dried for 1+ hours at temperatures between 118" and 210" C, or for I& hours at 125" C. Drying times of 1 hour, or less, resulted in slightly high results.COMPOSITION AND STRUCTURE OF THE PALLADIUM CHELATE- An aliquot of palladium solution, containing 25-05 mg of palladium(II), was transferred into a 250-ml beaker and the above procedure followed. Results of analyses of the precipitate for carbon, hydrogen and palladium and those calculated for the empirical formula, Pd(CllH,0N3) 2, are as follows- Palladium Carbon Hydrogen Percentage found . . .. . . 21.27 52.60 3.33 Percentage calculated . . . . 21.16 62.56 3.2 1 The palladium was determined gravimetrically by ignition to the metal, a few drops of formic acid being added to eliminate any palladium(I1) oxide formed. A probable structure for the chelate, based on infrared evidence reported previou~ly,~ is given below.492 HOLLAND et d.: DI-2-PYRIDYL KETOXIME, A NEW REAGENT [Analyst, VOl. 93 PRECISION AND ACCURACY- Pure anhydrous palladium chloride was dried over sulphuric acid for 5 hours and the amount of palladium determined twelve times according to the outlined procedure. The results obtained are listed below. They agree favourably with similar results obtained for dimethylglyoxime reported in a previous publication,6 in which di-2-thienyl ketoxime was used as a gravimetric reagent for palladium. Theoretical Palladi urn Relative mg mg per cent. palladium taken, found, standard deviation, 17-97 17.98 0.28 35-98 35.98 0.19 EFFECT OF DIVERSE IONS- Many foreign ions were studied to make the procedure as diversified as possible. Sodium citrate and EDTA were conveniently used as masking agents.In one series of test solutions 100 mg of a foreign ion were added to 15.03 mg of palladium and the outlined procedure followed. The recoveries gave a standard deviation of 0.02 mg for the following extraneous ions: Zn(II), Mn(II), Cu(II), Ni(II), Fe(II), Co(III), Al(III), Hg(II), Cd(II), Mg(II), Ca(II), Ba(II), Sr(II), Pb(II), Sb(V), Bi(III), As(III), Na, K, Ti(IV), U(VI), Mo(VI), Os(VIII), Ru(II1) , ammonium, oxalate, vanadate, chloride, pyrophosphate and dichromate. The ions of copper, nickel, cobalt and iron form soluble complexes with di-2-pyridyl ketoxime, and hence consume reagent. This effect was eliminated by masking these metals with 1 g of EDTA. In another series of test solutions, 100mg of the following foreign ions were added to 13.21 mg of palladium (the recoveries gave a standard deviation of 0.03 mg): Ag(I), Re(VII), Sn(IV), Te(IV), Ir(III), Zr(IV), Hf(IV), Ga(III), Se(IV), Ce(IV), Th(IV), In(III), Ta(V), Tl(I), iodide, thiourea, nitrite, nitrate, sulphite, sulphate, persulphate, carbonate, phosphate, tungstate, bromide, fluoride, perchlorate and thiocyanate.A silver complex was formed with ammonia at pH 10 to 11, and iridium was masked with 1 g of potassium thiocyanate at pH 8 to 9. The limit for rhodium was 50mg. There was no interference from 10mg of platinum in the regular procedure, and up to 50mg could be tolerated if the precipitation was carried out at room temperature. Gold interfered but was easily separated by reduction to the metal with sodium nitrite13 and filtration. Cyanide interfered severely in the procedure and must be absent.Cations were added as chlorides, nitrates or sulphates, and anions as sodium or potassium salts. Molybdenum, osmium, arsenic, rhenium and selenium were added as their oxides. DISCUSSION Di-2-pyridyl ketoxime is recommended for the gravimetric determination of palladium. The reagent has several advantages; it is stable indefinitely and is very water-soluble, thus eliminating co-precipitation, even up to a 200 per cent. excess of reagent. The ketone is commercially available and can readily be converted into the oxime. It is also possible to use the ketone directly and generate the oxime with hydroxylammonium chloride, thus allowing the precipitation of palladium from homogeneous solution. The chelate has a large molecular weight and hence a favourable gravimetric factor, which makes the reagent more sensitive than the classical reagent, dimethylglyoxime.Further, the di-2-pyridyl ketoxime procedure is much faster, producing comparable palladium recoveries in much less time. In the dimethylglyoxime method it is recommended that the precipitate be allowed to cool and stand overnight.2 The di-2-pyridyl ketoxime palladium precipitate is easy to handle during filtration, thus allowing up to 50 mg of palladium to be accurately determined over a wide pH range of between 3 and 11. There are no isomers to contend with because of the symmetrical structure of this oxime, and there are few interferences in the method. The applicability of this reagent to the determination of palladium was shown in a previous publi~ation,~ in which trace amounts of the chelate were extracted into chloroform, and the palladium subsequently determined by a spectrophotometric method.This work was supported by a grant from the National Research Council of Canada and a Province of Ontario Fellowship.August, 19681 FOR THE RAPID GRAVIMETRIC DETERMINATION OF PALLADIUM 493 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES Beamish, F. E., Talanta, 1966, 13, 773. Furman, N. H., Editor, “Standard Methods of Chemical Analysis,” Sixth Edition, D. Van Nostrand Company Inc., Princeton, N.J., New York, Toronto and London, 1962, Volume 1, p. 883. Voter, R. C., Banks, C. V., and Diehl, H., Analyt. Chem., 1948, 20, 652. Hayes, J. R., and Chandlee, G. C., Ind. Engng Chem. AnaZyt. Edn, 1942, 14, 491. Bark, L. S., and Griffin, D., Analyst, 1967, 92, 162. Holland, W. J., and Gerard, J., Analyt. Chem., 1966, 38, 919. Dalziel, J . A. W., and Kealey, D., Analyst, 1964, 89, 411. Holland, W. J., and Bozic, J., Analyt. Chem., 1967, 39, 109. -__ , Ibid., 1968, 40, 433. -- , Talanta, in the press. Holland, W. J., Bozic, J., and Gerard, J., Aizalytica Chim. Ada, in the press. Vogel, A. I ., “A Text-book of Quantitative Inorganic Analysis, including Elementary Instrumental Analysis,” Third Edition, Longmans, Green & Co. Ltd., London, 1961, p. 511. Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis, with Special Reference to the Analysis of Metals, Minerals and Rocks,” Second Edition, Revised by Lundell, G. E. F., Bright, H. A., and Hoffman, J. I., John Wiley and Sons Inc., New York; Chapman & Hall, Ltd., London, 1953, p. 367. Received March 15th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300490

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

494 A.nalyst, August, 1968, Vol. 93 ,$$. 494-497 Application of l=Phenyl=4=phenylamino=l,2,4=triazolium Chloride to the Determination of Cobalt(I1) BY C. CALZOLARI AND L. FAVRETTO (Istituto di Merceologia, Universitij di Trieste, Italy) A new reagent, l-phenyl-4-phenylamino-l,2,4-triazolium chloride, is proposed for the determination of cobalt (11) by extraction into 1,2-dichloro- ethane and spectrophotometric determination a t 626 nm of the ion-association compound formed in the presence of an excess of ammonium thiocyanate. The composition of the species extracted has been determined by applying the continuous variation method to the aqueous - organic phase system, and by determining the chemical composition and molecular weight of the compound precipitated from the aqueous phase.TERNARY complexes and ion-association compounds formed in the aqueous phase between a transition-metal thiocyanate and an organic base or organic cation have often been used for the extraction of the metal into an organic solvent and its spectrophotometric deter- mination. Among the organic substances proposed as reagents for cobalt( 11), tri-iso-octyl- amine,l di(4-antipyry1)methane and its derivatives,2~~ tetraphenylarsonium4.6,g and triphenyl- methylarsonium salts7 have been reported. Recently a derivative of 1 ,2,4-triazoleJ nitron, already well known as a precipitation reagent for nitrate ions,* has also been proposed for the extraction of various metals from aqueous solutions containing an excess of thiocyanate ions.9J0111J2 In the present paper, the results obtained in the study of a new 1,2,4-triazole derivative, 1-phenyl-4-phenylamino-lJ2 ,4-triazolium chloride (Cl4Hl3N,C1), as a reagent for cobalt (11) , are briefly reported.EXPERIMENTAL REAGENTS AND APPARATUS- 1-Phenyl-4-Phe~ylamino-lJ2 ,4-triazolium chloride-Prepare by the method of Runti and Nisi.ls The thiocyanate was obtained from a concentrated aqueous solution of 1-phenyl- 4-phenylamino-l,2,4-triazolium chloride by precipitation with ammonium thiocyanate. Both of the triazolium salts were purified by repeated crystallisation from ethanol - diethyl ether Standard solutions of the triazolium salts were prepared by direct weighing of the compounds previously dried at 60' C under vacuum. Stock aqueous solution of cobalt(ll), 1.00 x 10-2 F-Prepare from cobalt(I1) sulphate heptahydrate, CoS0,.7H20, and standardise complexometrically. By using this solution, prepare freshly with de-aerated water more dilute standard solutions as required.All of the solutions were adjusted to pH 3 by the addition of hydrochloric acid. Other solutions used were prepared with analytical-grade reagents [Erba (Milan Italy)]. Spectrophotometric results were obtained by using spectrophotometric-grade solvents (Erba) and a Unicam SP500 instrument. The pH of the aqueous phase was determined with a Beckman C pH meter, and the molecular weight with a Hewlett - Packard Mechrolab 302 vapour-pressure osmometer. (2 + 1, v/v). 0 SAC and the authors.CALZOLARI AND FAVRETTO PROPERTIES AND STABILITY OF THE TRIAZOLIUM SALTS- The melting-points and solubilities of the triazolium salts are given below.496 Solubility at 20" C (g per 100 g of solvent) Melting-poin t, I A -i Compound "C Water Chloroform 1,2-Dichloroethane 1 -Phenyl-4-phenylamino- l,B,Ptriazo- lium chloride .. .. . . 206, 21013 4-3 0.45 0.032 l-Phenyl-4-phenylamino-l,2,4-triazo- lium thiocyanate . . .. .. 136 0-046 1.5 0.28 The absorption spectra of the triazolium salts in chloroform solution show a maximum at wavelength 246 nm (for l-phenyl-4-phenylamino-l,2,Ptriazoliurn chloride, E = 1.15 x litres per cm per mole). In water and 1,2-&chloroethane, only an inflection is observed at 235 nm. The absorption spectrum does not change with varying pH in hydrochloric acid or Britton - Robinson buffer solutions. Buffered solutions of both salts are stable for at least 100 hours in the pH range 0.5 to 6-0.When adjusted with hydrochloric acid to pH 3, the solutions are stable for 2 weeks. Above pH 6-0, the salts are not stable and the hydrolytic oxidative degradation leads to the formation of diphenylformazane as the final product.13 SPECTROPHOTOMETRIC DETERMINATION OF COBALT( 11)- Conditions for maximzlm extraction-The absorption spectrum of the species extracted with lJ2-dichloroethane from an aqueous phase containing 1.0 x lo-* F cobalt(II), 0.6 F am- monium thiocyanate and 5.0 x F 1-phenyl-Pphenylamino-1,2,4-triazoliurn chloride is shown in Fig. 1 (broken-line curve). The spectrum is characterised by a maximum at wave- length 626 nm. Blank tests indicate that the strong ultraviolet absorption is mainly caused by the 1-phenyl-4-phenylamino-I ,2,4-triazolium thiocyanate formed with excess of thiocyanate and extracted into the organic layer.Extracts separated from the aqueous phase are stable for several days. 0 1 200 300 400 500 600 700 Wavelength, nm Fig. 1. Broken-line curve: absorption spectrum of the complex extracted with 1,2-dichloroethane (10.0 ml) from an aqueous phase (40.0 ml) containing 1.0 X 10-4~ cobalt (II), 0.5 F ammonium thio- cyanate and 5.0 x 1 0 - 3 ~ 1-phenyl-4-phenylamino- 1,2,4-triazolium chloride. Continuous-line curves : absorption spectra of the complex in 1,2-dichloroethane: curve A, at concentration 4-0 x 10-4~; curve B, at 5.0 x 10-5~; and curve C, at 2.0 x 10-6~. Optical path length 1 cm The shapes of the absorption spectra are not changed by varying the concentration of reactants in the aqueous phase, only their intensity changes.At a concentration of 1.0 x F in the aqueous phase, the extraction of cobalt(I1) is at a maximum when the concentration of the reagents reaches minimum values of 0.5 F for thiocyanate and 5-0 x F for l-phenyl-496 CALZOLARI AND FAVRETTO APPLICATION OF 1-PHENYL-4-PHENYLAMINO- [Analyst, VOl. 93 4-phenylamino-l,2,Ptriazolium chloride. Under these conditions, the variation of the pH in the range 0.3 to 4.0 does not affect the absorbance values and, at 626 nm, Beer's law is followed, at least for cobalt(I1) in the concentration range 5 x to 1 x 10"~. At 20" C the apparent specific absorption coefficient is 79.2 0.6 litres per cm per g [with an organic phase-to-aqueous phase (v/v) ratio of 4 : 13. This value has been calculated from the regression line of 12 pairs of observations. Irtterferertces-Only a few interferences have been tested as examples. With the condi- tions of maximum extraction, copper(I1) and iron(II1) give coloured extracts, which are orange - brown and red and interfere at concentrations higher than 1 x lo4 and 5 x F, respectively.Iron(I1) does not interfere. The interference of iron(II1) can be eliminated by using an excess of ammonium fl~oride.~ Manganese(I1) and nickel(I1) do not give visible absorbing extracts at levels of 1 x 10-3~, and the absorbance at 626nm caused by the cobalt( 11) is practically unaffected. Among the most common anions considered (fluorides, chlorides, bromides, iodides, sulphates and nitrates), nitrates interfere at concentrations higher than 1 x 1 0 - 3 ~ by causing the precipitation of the sparingly soluble salt, l-phenyl- 4-phen ylamino - 1,2,4- t riazolium nit rate .PROCEDURE- A slightly acidic aliquot (less than 10 ml), containing up to 40 pg per ml of cobalt(II), is transferred into a 50-ml calibrated flask, together with 5.0 ml of 4 F ammonium thiocyanate and 20.0 ml of 1.0 x F l-phenyl-Pphenylamin0-1,2,4-triazolium chloride. In the presence of up to 500 pg per ml of iron(III), 2 g of solid ammonium fluoride are added. After diluting the solution to about 38 ml, gently mixing and allowing it to stand for 5 minutes, 10-0 ml of 1,2-dichloroethane are added, the aqueous phase is diluted to the mark and the mixture shaken vigorously for not less than 3 minutes.The contents of the flask are then transferred into a 100-ml separating funnel and, after it has separated, the organic layer is run, dropwise, directly into the spectrophotometric cell. The absorbance is measured at 626 nm against the solvent and corrected for a reagent blank. For an optical path of 1 cm, the working curve lies in the concentration range of 1 to 6 pg per ml of cobalt(I1) in the aqueous phase and 1 pug of cobalt(I1) per ml is determined, with a standard error of 2 per cent. CONSTITUTION OF THE EXTRACTED SPECIES- The method of continuous variations,14 when applied to the measurement of the extracted compound in 1,2-dichloroethane at 626 nm, shows that the compound contains cobalt(I1) : thiocyanate:triazolium in the molar ratios 1:2:4, and it probably has the formula Co(SCN), (C14H13N4)2.The 600-fold excess of thiocyanate necessary for maximum extraction is about that required for the almost complete formation16 of the complex Co(SCN)i-. When a solution of cobaltothiocyanate (1 x 10" F cobalt(I1) in 2 F ammonium thio- cyanate) is added to a saturated solution of 1-phenyl-4-phenylamino-1,2,4-triazolium thio- cyanate in 2 F ammonium thiocyanate, a blue precipitate of waxy appearance is formed, which is purified by re-precipitation with carbon tetrachloride from 1,2-dichloroethane solution and dried under vacuum. Its elemental composition was found to be as follows: C , 50.22 per cent.; H, 3.53 per cent.; N, 22.35 per cent.; S, 16.65 per cent. and Co, 7-72 per cent.(total 100-40 per cent.), which agrees with that calculated for the formula Co(SCN),(CI4H,,N4),: C, 50-18 per cent.; H, 3.42 per cent.; N, 21-95 per cent.; S, 16.75 per cent.; and Co, 7.70 per cent. and, in the concentration range 0.05 to 0.3 F in 1,2-dichloroethane, the molecular weight was found to be 752 g per mole, which is in agreement with the value of 765.8 calculated for the proposed formula. The absorption spectrum in the same solvent presents two maxima, at 626 and 322 nm (see Fig. 1, continuous-line curves), and in the visible region it is identical with that observed for the organic extracts. Beer's law is obeyed, with the following molar absorption coefficients: at wavelength 626 nm, E = (1.256 -I 0.013) x lo3 litres per cm per mole (1 x 10-4 to 1 x 10-3 F) ; and at wavelength 322 nm, E = (8.82 0.26) x 103 litres per cm per mole (1 x 10-5 to 1 x 1 0 4 ~ ) (the concentration ranges used in this check are given in brackets).From the difference of absorbance at 626 nm between a 1,2-dichloroethane solution and an extract obtained from an aqueous solution containing an equivalent amount of metal, 95 per cent. of the metal is calculated to be extracted under the conditions specified by the proposed method.August, 19681 1,2,4-TRIAZOLIUM CHLORIDE TO THE DETERMINATION OF COBALT(I1) 497 CONCLUSIONS The reagent proposed for the determination of cobalt(I1) may be included among those which increase the sensitivity of the classical reaction with thiocyanate by extracting the ion-association compound formed with cobaltothiocyanate into an organic solvent.When compared with those already in use, the procedure suggested is simpler as it involves only one high-efficiency extraction. The sensitivity of the reagent is higher than that given for the tetraphenylarsonium salts6 The reagent can be used in media of uncontrolled acidity and no problems arise from instability. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Selmer-Olsen, A. R., Analytica Chim. Acta, 1964, 31, 33. Zhivopistsev, B. P., Zav. Lab., 1952, 18, 649. Sokolova, E. V., Pesis, A. C . , and Panova, N. I., Zh. Analit. Klziiiz., 1957, 12, 489. Potratz, H. A., and Rosen, S. M., Analyt. Chem., 1949, 21, 1276. Affsprung, H. E., Barnes, N. A., and Potratz, H. A., Ibid., 1951, 23, 1680. Pepkowitz, L. P., and Marley, J. L., Ibid., 1955, 27, 1330. Ellis, K. W., and Gibson, N. A., Analytica Chim. Acta, 1953, 9, 275. Busch, M., Ber. dt. chem. Ges., 1906, 38, N 861. Babenko, A. S., and Tolmachev, V. N., Ukr. Khim. Zh., 1961, 27, 732. -- , Dopov. Akad. Nauk. Ukr. SSR, 1962, 394. Tolmachev, V. N., and Babenko, A. S., Ibid., 1962, 616. Babenko, A. S., and Tolmachev, V. N., Ukr. Khim. Zh., 1962, 28, 26 and 287. Runti, C., and Nisi, C., J . Med. Chem., 1964, 7, 814. Vosburgh, C., and Cooper, G., J . Amer. Chem. SOC., 1941, 63, 437. Lehn6, M., Bull. SOC. Chim. Fr., 1961, 76. Received October 30th, 1967

ISSN:0003-2654    

DOI:10.1039/AN9689300494

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

498 Analyst, August, 1968, Vol. 93, fifi. 498-506 The Application of Polarography and Related Electroanalytical Techniques to the Determination of Sodium Diethyldithiocarbamate BY M. J. D. BRAND AND B. FLEET (Chemistry Defiartment, Imperial College of Science and Technology, London, S. W. 7 ) Cathodic stripping analysis has been shown to be an extremely sensitive technique for the determination of compounds that form insoluble mercury salts. The experimental parameters governing the application of this tech- nique have been studied and applied to the determination of sodium diethyl- dithiocarbamate. The results obtained were compared with those from con- ventional d.c. polarography, a.c. polarography and linear potential-sweep chrono-amperometry . DERIVATIVES of dithiocarbamic acid are widely used as fungicides.The compounds that have been used can be divided into three main typesl- simple substituted dithiocarbamic acid salts R S \ II N-C-S- M’+/Tz R .;’ the thiuram disulphides R S S R \ I1 I / N-c-s-s-LN R / ‘R where R = CH,, etc. R’ = H, CH,, etc. M = Na+, Fe3+, Zn2+, etc. where R= CH,, etc. and salts of ethylenebisdithiocarbamic acid S II CHZ-NH-GS- CH,-NH--C--S- i 2Mn+/.n where M = Na+, Zn2+, Mne+, etc. II S The existing analytical methods for this class of compounds are based on conversion of the sulphur into carbon disulphide, which is then determined colorirnetrically, e.g., after reaction with Viles’ reagent.2 The sensitivity of this somewhat lengthy procedure is limited. More- over, the method is only semi-specific in that it cannot distinguish between dithiocarbamates and thiuram disulphides.Polarography affords a selective and sensitive technique for the determination of a wide range of organosulphur compounds, but has so far found limited application in pesticide analysis: particularly in regard to the dithiocarbamic acid derivatives. As the dithio- carbamate ion is the fungitoxic species and is also relatively unstable, methods of analysis for this class of compound are more meaningful if the anodic wave of the dithiocarbamate anion, rather than the cathodic wave of the metal ion, is measured. Nangniot has used this approach in the determination of residues of zinc4 and irons dimethyldithiocarbamates in plants . 0 SAC and the authors.BRAND AND FLEET 499 This paper describes the study of simple substituted dithiocarbamic acid salts, sodium diethyldithiocarbamate being chosen as the model substance.Simple mono- and dialkyl- substituted dithiocarbamates show anodic waves caused by the formation of insoluble mercury salts.6 The formation of insoluble films on liquid electrodes has been studied in detail by potentiostatic and a.c. impedance measurements.' Most of the detailed polarographic and related studies have been concerned with inorganic anions, in particular the halide^.^^^ Although many organic compounds form insoluble mercury derivatives only a few of these have been studied.10 This reflects the difficulties associated with elucidation of the electrode process, e.g., determination of n, and the isolation of the metastable products of the electrode reaction.At present only a qualitative description of the electrode process for this latter group of compounds is available; this is sufficient basis, however, for the development of an analytical method. EXPERIMENTAL REAGENTS- All reagents used were of analytical-reagent grade. Sodium diethyldithiocarbamate. B u f e r solutiout--This was made from 1 M ammonia solution and 1 M ammonium nitrate. Sodium diethyldithiocarbamate stock solutions , to M, aqueous. An aliquot of the appropriate stock solution was added to 1 ml of buffer solution and diluted to 10ml. For a.c. measurements undiluted buffer solution was used as supporting electrolyte. Solutions were de-oxygenated by passage of nitrogen for 3 minutes. APPARATUS- Direct current polarograms were recorded on a polarograph, type OH-102 (Metrimpex, Hungary).A Kalousek cell with a separated saturated calomel electrode (S.C.E.) was used. The characteristics of the dropping-mercury electrode, measured at 0.0 volt versus S.C.E. in 0.1 M potassium chloride solution were t = 3-82 s, m = 2-32 mg s-1 at h 65 cm. Alternating current polarograms were measured with a Univector and general-purpose polarograph (Cambridge Limited, London), with the Metrimpex polarograph being used as the current-output recorder. The connections to the positive cell and negative meter on the general-purpose polarograph were removed and the two terminals short-circuited with a wire. This modification removed the 6-K ohm sensitivity shunt from the output, at the same time eliminating the zero and damping controls.The sensitivity, zero, and damping controls present on the recorder were subsequently used. A Kalousek cell was used in which a platinum-wire counter electrode was placed near the dropping-mercury electrode. The reference and counter electrodes were connected to a 5000-pF capacitor. Linear potential-sweep chrono-amperometric measurements with a dropping-mercury electrode were made with a KlOOO cathode-ray polarograph (Southern Analytical Limited, Surrey). Measurements were also made with a Davis differential cathode-ray polarograph (Southern Analytical Limited, Surrey). In each of these a Kalousek cell was used, as the cells supplied with the instrument have a mercury-pool anode that is unsuitable for studying compounds that react with mercury.The Davis differential polarograph was used in the single cell mode as it was impossible to use two Kalousek cells without major modification of the electrode stand. Cyclic voltammograms were obtained with the Metrimpex polarograph with a hanging mercury drop electrode in a Kalousek cell. The hanging mercury drop electrode was con- structed by cementing a glass capillary tube (0-2-mm bore) into the barrel of an Agla micro- meter syringe burette (Burroughs Wellcome Limited, London). Electrical connection was via a platinum wire sealed into the syringe barrel, which was thermally insulated with expanded polystyrene. For cathodic stripping the Metrimpex polarograph was used to supply both the applied potential for the pre-electrolysis and the voltage sweep for the stripping process.Two types of electrode system were used: a hanging mercury drop electrode and a mercury-coated platinum electrode. The mercury-coated platinum electrode consisted of a platinum wire (10 x 0-6 mm) sealed into a glass tube and coated with mercury by using the method of Joyce and Westcott.ll The reference electrode was a large current capacity S.C.E. con- nected by a salt bridge to a micro liquid junction tube (E.I.L. Limited, Surrey). The cell500 BRAND AND FLEET: APPLICATION OF POLAROGRAPHY TO THE [AndYSi!, VOl. 93 used for the hanging mercury drop electrode was a 25-ml tall-form beaker closed with a rubber bung through which electrodes and gas tubes were passed. For the mercury-coated platinum electrode the cell was made from a B45 socket reduced on to a tube of 3-cm diameter and closed with a flat end.A Perspex cover was machined to fit the ground-glass socket. Solutions were stirred magnetically, a glass-covered rotor being driven by a small magnet attached to a motor. The mercury-coated platinum electrode was mounted in a PTFE stirrer gland (Quickfit and Quartz Limited) and rotated at about 600 r.p.m. by a belt drive from a squirrel cage motor. RESULTS AND DISCUSSION DIRECT CURRENT POLAROGRAPHY- The results of a preliminary survey of the d.c. polarographic behaviour of sodium diethyldithiocarbamate were in agreement with earlier findings6 and followed the general pattern of insoluble mercury salt formation. In alkaline solution two anodic waves are observed. The wave at more negative potentials (I) is concentration independent above M, and shows a linear relationship between current and reservoir height.The wave at more positive potentials (11) begins more or less discontinuously and is only present at concentrations above ~ O - * M when it is con- centration dependent. The first wave (I) has been classified as a Brdicka absorption wave, and arises from the formation of an insoluble monolayer on the drop surface. Further oxidation is then hindered until a more positive potential is reached when the barrier to the charge-transfer reaction is overcome, resulting in a discontinuity in the current - voltage graph. ALTERNATING CURRENT POLAROGRAPHY- In alkaline solution sodium diethyldithiocarbamate showed two waves in the concen- tration range 10-3 to l o - 4 ~ (Fig.1) at potentials corresponding to the d.c. waves. The concentration dependence of these two peaks followed the pattern previously observed with d.c. polarography. Peak I was concentration independent over the range to 1 0 4 ~ , but linearly dependent in the range to 1 0 4 ~ . Peak I1 showed a rectilinear peak height - concentration relationship over the range M but disappeared below lo-* M. The use of these two peaks gave a rectilinear ip - concentration dependence over the range 10-6 to 103 M (Tables I and 11). Thus the wide concentration range attainable, together with to -0.2 4 . 3 -0.4 -0.5 -0.6 -0.7 -0.8 -0.9 Volts versus S.C.E. Fig. 1. Concentration dependence of a.c. waves of sodium diethyldithio- carbamate (I) 2 x M, (11) 4 X lo-* M, (111) 6 X M, (Iv) 8 X M, (v) 10-3 MAugust, 19681 DETERMINATION OF SODIUM DIETHYLDITHIOCARB4MATE TABLE I COMPARISON OF POLAROGRAPHIC METHODS FOR THE DETERMINATION OF SODIUM DIETHYLDITHIOCARBAMATE Currents in pA A I .50 1 Concentration of sodium diethyl- dithiocarbamate, M 2 x 10-5 4 x 10-5 8 x 10-5 1 x 10-4 2 x 10-4 4 x 10-4 8 x 10-4 1 x 10-3 6 x 6 x Direct current Alternating current polarography polarography +7 +--7 I I1 I I1 - - 0.54 - - - 1.08 - - - 1.60 - - - 2.13 - 0-29 0 2-64 0 0.29 0.34 4.54 0.64 0-29 1-42 6.30 2-05 0.29 2-50 6.43 3-46 0.29 4.08 6.40 4.84 0-29 - 6-40 6-24 * Cathodic voltage sweep. I Pre-wave. I1 Diffusion controlled wave. Linear potentiallsweep chrono-amperometry * & I I1 0.61 - 1-15 - 1-74 - 2-30 - 2-67 0 3-05 4.9 3-05 17.6 3-05 28-8 - - - - the advantages associated with the derivative type signal obtained, makes this technique ideally suited to the analysis of dithiocarbamate formulations.A concentration of 10-6 M must be taken as the lower limit attainable with the apparatus used, as at this level the instrument noise became comparable with the signal. About 1 per cent. a.c. ripple was observed on the rectified output of the Univector unit at twice the applied frequency of about 35 Hz. The recorder that was used (Metrimpex polarograph) incorporated parallel T-filters tuned to frequencies of 1.6, 0.8, 0.8 and 0.4 Hz, which are effective in filtering oscillations caused by drop growth, but do not significantly reduce the 70-H~ ripple. The limit of detection could be extended by the use of an instrument based on solid-state electronics12 or by incorporating a filter tuned to higher frequencies.At high concentrations, e.g., low3 M, when both waves are well developed it is apparent that the width of wave I at half-height, 144 mV, is considerably greater than the corresponding TABLE I1 COMPARISON OF CATHODIC STRIPPING VOLTAMMETRY WITH POLAROGRAPHIC TECHNIQUES FOR THE DETERMINATION OF SODIUM DIETHYLDITHIOCARBAMATE Currents in pA A I Concentration of sodium diethyl- dithiocarbamate, 2 x 10-8 4 x 10-8 6 x 8 x 10-8 M 1 x 10-7 2 x 10-7 4 x 10-7 s x 10-7 6 x 1 x 10-8 2 x 10-6 4 x 10-6 6 x 8 x 10-6 1 x 10-5 Linear Alternating potential- current sweep chrono- polarography amperometry" - - - - - - - - - - - - - - - - - - - - 0-03 0.05 0.07 0.11 0.10 0.16 0.14 0-22 0.17 0.27 * Cathodic voltage sweep.t 5 minutes' electrolysis. $ 2 minutes' electrolysis. 5 1 minute's electrolysis. Cathodic stripping Stirred R o t a t ; n g solution electrode - 0.377 - 0.567 - 0.787 0-77t 0*56$ 2-53? 147$ 2.78t 2*38$ 3-78? 2-88: 0.943 - 1.948 - 2-78s - 4.063 - 4.735 I - 0.207 - 0.987 1-47t l.lS$502 BRAND AND FLEET: APPLICATION OF POLAROGRAPHY TO THE [Analyst, Vol. 93 value for wave 11, 88 mV (Fig. 2). Theory predicts that the width of an a.c. wave at half- height is 90*5/n mV for low frequencies.13 However, with the apparatus used only the com- ponent of the a.c. current in phase with the applied a.c. voltage is rectified and recorded as the a.c. wave. While this results in an increase in sensitivity by eliminating purely capacita- tive currents, some distortion of the wave shape is inevitable because most faradaic processes have phase angles between 0" and 45".I t is therefore not possible to derive n values from this evidence. It is clear, however, that the two processes corresponding to waves I and I1 both show large in-phase components of the a.c. current, hence both must involve charge- transfer reactions. Potential Fig. 2. Cathode-ray' polarogram, 6 x M sodium diethyldithiocarbamate. Ano- Sensitivity 7.5 pA f.s.d. dic voltage sweep. Starting voltage -0.7 volt versus S.C.E. LINEAR POTENTIAL-SWEEP CHRONO-AMPEROMETRY AT A DROPPING-MERCURY ELECTRODE- Anodic voltage sweeps (from negative to positive potentials) showed two peaks in the concentration range 10-3 to 10-4 M (Fig.2). The height of the more negative peak (I) was concentration independent, while peak I1 showed a rectilinear concentration dependence. Thus peak I clearly corresponds to the d.c. polarographic pre-wave and peak I1 to thed.c. diffusion wave. Potential Fig. 3. Cathode-ray polarogram, 5 x 10" M sodium diethyldithiocarbamate. Cathodic voItage sweep. Sen- sitivity 30 p A f.s.d. Starting voltage -0.3 volt versus S.C.E.August, 19681 DETERMINATION OF SODIUM DIETHYLDITHIOCARBAMATE 503 Cathodic voltage sweeps (from positive to negative potentials) also showed the same two peaks, but the peak currents obtained were several times greater (Fig. 3). At concen- trations above 6 x 10" M, peak I (the concentration-independent pre-wave) was obscured by the larger peak 11, which required a reduction in the sensitivity control of the instrument to accommodate it.This increase in sensitivity was caused by pre-concentration of the insoluble mercury salt on the drop surface during the 5-second delay period. At concentrations below lo-* M peak I1 disappeared and peak I became concentration dependent (Table I) over the range lo4 to M. The consecutive use of the two peaks obtained by cathodic voltage sweeps enabled a rectilinear i, - concentration dependence to be established over the con- centration range 1 x 10-6 to 6 x lo-* M. The potential-sweep synchronisation circuit of the polarograph requires the use of a dropping-mercury electrode with a drop time of between 6.5 and 7.0 seconds. In the presence of sodium diethyldithiocarbamate the capillary was found to drop erratically, presumably because of the deposition of insoluble material in and around the capillary orifice.This problem was overcome by the use of forced-drop detachment. With the Davis differential cathode-ray polarograph the potential sweep is synchronised with mechanical detachment of the drop. At concentrations below M peak heights were not reproducible. However, modification of the instrument to allow the use of twin cells with external reference electrodes would result in a decrease in the lower concentration limit. CYCLIC VOLTAMMETRY- A cyclic voltammogram of sodium diethyldithiocarbamate in alkaline solution is shown in Fig. 4. A single oxidation- reduction peak is observed at the sweep rates used. The enhancement in peak current during the cathodic sweep compared with the anodic sweep is clearly illustrated.Whenever the electrode potential is more positive than that of the cathodic peak (0.4 volt 'uwsus S.C.E.), the insoluble mercury - diethyldithiocarbamate com- pound is formed at the electrode surface. This suggested that cathodic stripping analysis is applicable to the problem. In this technique the electrode is held for a fixed time at a potential a t which the insoluble mercury salt is formed, and then a cathodic voltage sweep is applied; the resulting current is recorded as a stripping peak, the height of which is proportional to concentration. Cyclic voltammetry, which allows the direct comparison of the oxidation and reduction processes, was used to determine the potential range available for pre-electrolysis and the optimum sweep rate.Volts versus S.C.E. Fig. 4. Cyclic voltammogram, 1 x lo-* M sodium diethyldithiocarbamate. Single cycle, initial scan anodic. Scan rate 25 mVs-1504 BRAND AND FLEET: APPLICATION OF POLAROGRAPHY TO THE [Auta&St, VOl. 93 The observed separation of anodic and cathodic peak potentials (l40mV) is greater than the expected value of about 60 mV for a reversible one-electron transfer at comparatively slow sweep rates.14 About 20 to 30mV of this separation arises through backlash in the polarograph gears. The displacement is still significantly different from the ideal case, and suggests that an activation energy is associated with the stripping process. CATHODIC STRIPPING ANALYSIS- There has been considerable work on the technique of anodic stripping15~leJ7 but relatively little work has been done on the converse technique of cathodic stripping.It is not generally realised that cathodic stripping does not suffer from many of the complicating factors encountered in the analogous anodic technique. As the electrolysis is carried out at relatively positive potentials, fewer impurities are plated out. Moreover, the cathodic stripping of an insoluble film from a mercury surface gives an improved peak definition when compared with anodic stripping in which slow diffusion from the bulk of the electrode causes tailing. Two types of electrode system were studied. Firstly a hanging mercury drop electrode was investigated and subsequently a mercury-coated platinum electrode.HANGING MERCURY DROP ELECTRODE- The stripping curve, (a), obtained with this type of electrode after pre-electrolysis in unstirred solution is shown in Fig. 5 . Peak I corresponds to the pre-wave and peak I1 to the diffusion controlled wave in Fig. 1. A random dependence of peak current versus elec- trolysis time for peak I1 was found, while the height of peak I remained constant. When the solution was stirred during the pre-electrolysis an increase in sensitivity was obtained and a rectilinear peak current - electrolysis time relationship was found. Under these conditions the pre-wave was not clearly defined, suggesting that no discrete monolayer was formed as shown in (b) in Fig. 5. The magnitude of the stripping peak showed a rectilinear concen- tration dependence over the range 5 x lo-' to 10" M.Volts versus S.C.E. Fig. 5. Stripping peaks at hanging mercury drop electrode, 2 x 1 0 - 6 ~ sodium diethyldithiocarbamate; 4 minutes' electrolysis at 0.0 volt vemysus S.C.E. Scan rate 25 mVs-1: (a) unstirred solution; (b) stirred solution MERCURY-COATED PLATINUM ELECTRODE- The two major disadvantages of the hanging mercury drop electrode, i.e., its extreme sensitivity to vibration and the difficulties associated with reproducibility of the drop size are eliminated by using a mercury-coated platinum electrode. Many of the practical difficulties experienced by early workers in the use of this electrode have recently been overcome by Joyce and Westcott.11 In the present work initial experiments were performed with a stationary electrode in a stirred solution.Results indicated that for a given depolariser concentration, in order to minimise the electrolysis time it was necessary to increase theAugust , 19681 DETERMINATION OF SODIUM DIETHYLDITHIOCARBAMATE 505 rate of stirring until just before turbulence occurred. Under these conditions well developed stripping peaks, (a) in Fig. 6, were obtained after 2 minutes’ electrolysis for concentrations down to 10-6 M. At concentrations below this level it was necessary to extend the electrolysis period to 5 minutes to allow the use of a sufficiently low current sensitivity to eliminate instrument noise; under these conditions there was a slight decrease in precision. This problem was overcome by rotating the electrode, at constant speed, during the pre-electrolysis instead of stirring the solution.The stripping peaks, (b), obtained after 2 minutes’ pre- electrolysis at a rotating electrode were better defined, as shown in Fig. 6(b), and allowed the lower concentration level to be extended to lo-’ M without loss of precision. The deter- mination of lower concentrations was possible but it was preferable to increase the electrolysis time rather than the instrument sensitivity. 2 4 t V c 2 (3 I 0 -0.2 -0.4 -0.6 0 -0.2 -0.4 -0.6 Volts versus S.C.E. Fig. 6. Stripping peaks at mercury-coated platinum electrode, 8 x lo-’ M sodium diethyl- dithiocarbamate; 2 minutes’ electrolysis at 0.0 volt versus S.C.E. Scan rate 100 mVs-l: (a) stirred solution; (b) rotated electrode CONCLUSIONS The results of the present investigation have shown that of the techniques examined, cathodic stripping proved the most suitable for the determination of trace amounts of sodium diethyldithiocarbamate.Alternating current polarography and potential-sweep chrono- amperometry were both applicable to a wide range of concentrations but in both cases the lower limit was 10-6 M. The sensitivity of cathodic stripping was limited only by instrument noise, and it is evident that with improved instrumentation by using electronic potential sweeps and oscilloscopic read-out a much lower detection limit could be achieved. A rotating mercury-coated platinum electrode was found to be the most suitable for cathodic stripping. Its sensitivity and reproducibility were far superior to the hanging mercury drop electrode; in addition it was not affected either by the rate of stirring of the solution or by vibration.Although the present work has been confined to mercury electrodes, it should be noted that certain types of solid electrode, e.g., pyrolytic graphite and carbon paste, might offer some advantages in that the electrode process will involve a true oxidation, rather than mercury salt formation. We wish to thank the Agricultural Research Council for the provision of a research assistantship to one of us (M. J. D. B.). We would also like to thank Morganite Research and Development Limited for the loan of the KlOOO cathode-ray polarograph, and Dr. D. A. Pantony, Metallurgy Department, Imperial College, for the use of the Davis differential cat hode-ray polarograph.506 BRAND AND FLEET REFERENCES Thorn, G.D., and Ludwig, R. A., “The Dithiocarbamates and Related Compounds,” Elsevier Publishing Company, Amsterdam, 1962, Chapter 10. Lowen, W. K., and Pease, H. L., in Zweig, G., Editor, “Analytical Methods for Pesticides, Plant Growth Regulators and Food Additives,” Academic Press, New York and London, Volume 111, 1963, p. 69. Gajan, R. J., in Gunther, F. A., Editor, “Residue Reviews,” Springer-Verlag, Berlin, Gottingen and Heidelberg, Volume V, 1964, p. 80; Volume VI, 1964, p. 77. Nangniot. P., Bull. Inst. Agron. Stns Rech. Gembloux, 1960, 28, 365. Halls, D. J., Townshend, A., and Zuman, P., Analyst, 1968, 93, 219. Armstrong, R. D., and Fleischman, M., J . Polurogr. Soc., 1965, 11, 31. Vlcek, A. A., Colln Czech. Chem. Commun., Engl. Edn, 1954, 19, 221. Biegler, T., J . Electroanal. Chem., 1964, 6, 357, 365 and 373. Brezina, M., and Zuman, P., “Polarography in Medicine, Biochemistry and Pharmacy,” Inter- science Publishers Inc., New York and London, 1958, p. 460. Joyce, R. J., and Westcott, C. C., “Symposium on Trace Characterisation,” National Bureau of Standards, Washington, 1966. Smith, D. E., in Bard, A. J., Editor, “Electroanalytical Chemistry,’’ Arnold Ltd., London, Volume I, 1966, p. 102. Breyer, B., and Bauer, H. H., “Alternating Current Polarography and Tensammetry,” Interscience Publishers Inc., New York, 1963, p. 50. Matsuda, H., and Ayabe, Y., 2. Electrochem., 1955, 59, 494. Shain, L., in Kolthoff, I. M., and Elving, P. J., Editors, “Treatise on Analytical Chemistry,” Interscience Publishers Inc., New York, 1963, Part I, Volume 4, p. 2533. Kemula, W., and Kublik, Z., in Reilley, C. N., Editor, “Advances in Analytical Chemistry and Instrumentation,” Interscience Publishers Inc., London, New York and Melbourne, Volume 111, 1963, p. 123. Barendrecht, E., in Bard, A. J ., Editor, “Electroanalytical Chemistry,’’ Arnold Ltd., London, Volume 11, 1967, p. 53. Received December 8th, 1967 -, Ibid., 1960,28, 313. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

ISSN:0003-2654    

DOI:10.1039/AN9689300498

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

Analyst, August, 1968, Vol. 93, $9. 507-517 507 A Field Method for the Determination of Ozone in the Presence of Nitrogen Dioxide BY I. C. COHEN, A. F. SMITH AND R. WOOD (Ministry of Technology, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, S.E. 1) A method has been developed for the determination of ozone in air in the presence of nitrogen dioxide. Samples of the test atmosphere are drawn, in parallel, through neutral, buffered potassium iodide solutions at 2 litres per minute for 20 minutes. One sample, prior to its passage through the iodide solution, is drawn through a cotton-wool plug. This removes the ozone, and the iodine liberated is a measure of the nitrogen dioxide present, expressed in terms of ozone equivalents. The iodine liberated by the other sample is a measure of the total oxidant, ozone $Zus nitrogen dioxide, also expressed as ozone equivalents.After addition of starch the iodine liberated by each sample is determined visually against prepared colour standards. The dif- ference between the two readings obtained is a measure of the ozone present. The method enables from 0.1 mg m-3 (0.05 p.p.m. v/v) of ozone in air, in the presence of up to 9 mg m-3 (5 p.p.m.) of nitrogen dioxide, to 0.8 mg m-3 (0:4 p.p.m.) of ozone in air, in the absence of nitrogen dioxide, to be deter- mined. The procedure is relatively simple and the time required for a com- plete analysis is about 30 minutes. OZONE, which has increasing applications in a variety of industries, is used, for instance, in various chemical syntheses, and as a bleaching agent for textiles.In these processes measures can be taken to minimise the contamination of the working atmosphere by ozone. However, it is less easy to control the ozone formed fortuitously, together with nitrogen dioxide, in the vicinity of high-voltage electrical discharges and during electric arc-welding operations. Ozone is a toxic gas with a current threshold limit value (T.L.V.) in the United Kingdom of only 0.2 mg m-3 (0.1 p.p.m. v/v) in air,l and it could represent a considerable hazard in an industrial atmosphere. As the T.L.V. for nitrogen dioxide in the United Kingdom is currently set at 9 mg m-3 (5 p.p.m. v/v) in air,l fifty times that of ozone, it appeared that a need existed for a simple field test for the determination of low concentrations of ozone in the presence of nitrogen dioxide.The methods already available for the determination of ozone generally either lack specificity or are unsuitable for adaptation to a field test. For example, the reaction of ozone with potassium iodide and subsequent liberation of iodine would be suitable as a basis for a field test except for the interference from oxidising and reducing gases. Bravo and Lodge2 reported a method for the determination of ozone by means of its specific reaction with 4,4’-dimethoxystilbene to yield anisaldehyde, which is then determined spectrophotometrically. Besides finding variabilities associated with the source and keeping qualities of the reagents, we considered that the corrosive nature of two of the reagents, trifluoroacetic acid and trifluoroacetic anhydride, precluded the use of this method in the field; similar views have been expressed by other worker^.^ We found a similar type of method, involving the ozonolysis of 1,2-di-(4-pyridyl)ethylene and subsequent spectrophotometric determination of the pyridyl-4-aldehyde formed: to be much more sensitive than that of Bravo and Lodge.However, under the sampling conditions of at least 2 litres per minute, that we wished to use, somewhat erratic results were obtained. Additionally, the need for a 20-minute heating period during the colour development would be disadvantageous in any field test. 0 SAC; Crown Copyright Reserved.508 COHEN, SMITH AND WOOD: A FIELD METHOD FOR THE [ArtU&St, VOl. 93 Attempts to develop a field test based on the production of stains on chemically im- pregnated papers proved unsuccessful.Di- and trihydric phenols, hydroxynaphthalenes, aminophenol, carbazole, “tetra base,” indigo carmine and fuchsin were each tried, but the passage of several litres of a 0.2 mg m-3 (0.1 p.p.m. v/v) atmosphere of ozone through the respective impregnated papers produced no visible stains. Better results were obtained from the use of the fluidised-bed te~hnique.~ Various supports, notably silica gel and activated alumina, were impregnated with each of several compounds known to change colour on exposure to ozone. Best results were obtained with fuchsin on acidic alumina and indigo carmine on silica gel. However, with both systems, insufficient colour differentiation was obtained between 0.1, 0.2 and 0-4 mg m-3 ozone levels.Recently, after we had completed the development of our method, Nash6 described a procedure for the determination of ozone based on the bleaching effect of the gas on an aqueous solution of diacetyldihydrolutidine. Although we have not investigated this method, the interference of oxides of nitrogen to the extent of 5 per cent. v/v may limit its applicability to the determination of ozone in the presence of nitrogen dioxide. Other published tests for ozone778 were found to lack the sensitivity required. In the absence of a more suitable method it was decided to attempt to develop a field test based on the reaction of ozone with neutral potassium iodide solution and incorporating a means of separating ozone from nitrogen dioxide. EXPERIMENTAL PREPARATION AND CALIBRATION OF STANDARD ATMOSPHERES- Atmospheres containing known concentrations of ozone, nitrogen dioxide and mixtures of these gases were required to assess the efficiency of sampling techniques and for the development of a field test.PREPARATION- Ozone-A dynamic atmosphere was produced by passing oxygen through a commercial ozoniser (British Oxygen Company, Mark I1 Ozoniser) at a rate sufficient to give a nominal ozone concentration of 1 per cent. w/w and diluting twice with metered streams of air to the required Concentration. Nitrogen dioxide-A dynamic atmosphere was produced by bubbling nitrogen at a few millilitres per minute through liquid nitrogen dioxide maintained at -5” C and diluting the resultant vapour twice with metered streams of air to the required concentration.Mixed atmosj5keres-These were obtained by mixing metered flows of each diluted gas in a suitable mixing chamber. CALIBRATION- Ozone-The ozone atmosphere was calibrated by sampling at 20 litres per minute for 40 minutes through three Greenburg - Smith impingers in series, each containing 60 ml of neutxa1, buffered potassium iodide solution. The total iodine liberated was determined by titration with 0.004 N sodium thiosulphate, and the concentration of ozone in the atmosphere calculated on the basis of a stoicheiometric release of iodine by ozone (0, = 12).g The use of three bubblers ensured the complete collection of the ozone and eliminated any loss of iodine by carry-over.It would have been desirable to standardise the atmosphere by another method. Although the Bravo and Lodge2 method gave results comparable with those obtained by the potassium iodide method, the difficulties experienced with the former method meant that little significance could be placed on the results obtained. However, the concentrations of various atmospheres determined by the potassium iodide method agreed closely with those calculated from the graph of ozone output supplied with the ozoniser and by using the appropriate dilution factors. For example, under one set of diluting conditions the calculated and experimentally determined concentrations of the ozone atmos- phere after the first dilution stage were 24-6 and 24.5 mg m-3, respectively, and the corre- sponding figures after the second dilution stage were 0.23 and 0-24mg m-3.Nitrogen dioxide-A precisely calibrated atmosphere of this gas was not required but the approximate concentration of nitrogen dioxide in the atmosphere was determined spectrophotometrically with a modified SaltzmanlO reaction.August, 19681 DETERMINATION OF OZONE I N THE PRESENCE OF NITROGEN DIOXIDE 509 COLLECTION OF OZONE- Saltzman and GilbertQ have reported on the various aspects of the collection of ozone in potassium iodide solution. Our choice of sampling conditions was based to a considerable extent on this work. In particular, neutral, buffered potassium iodide solution was found suitable as an absorbing solution for ozone. Provided that the solid potassium iodide was fresh, a 2 per cent.solution of this in mixed sodium and potassium hydrogen phosphate buffer was stable and usable for several months. A comparison was made of the relative merits of sintered and impinger type absorbers. The impingers were found to give more reproducible results and a higher absorption efficiency, thus confirming the previous observations of Hendricks and Larsen.3 All-glass absorbers of the impinger type shown in Fig. 2 (see page 511) were found suitable and used with 10 ml of absorbing solution. The minimum sample volume that gave a visually discernible colour with a 0.1 mg m-3 ozone atmosphere was 40 litres. With 10ml of absorbing solution in one absorber it was observed that 2 litres per minute was the maximum sampling rate that could be used without undue loss of liberated iodine.This is discussed more fully later. At higher sampling rates a loss of absorbing solution was also noted. The 20-minute sampling time required under these flow-rate conditions was considered to be compatible with the requirements of a field test. COLORIMETRIC DETERMINATION OF OZONE- Previous m e t h o d ~ ~ , ~ ~ for the determination of oxidants or ozone in air with potassium iodide solution have involved spectrophotometric measurements of the liberated iodine. In field testing, visual determination with colour standards is more convenient, especially as scientifically unskilled personnel are likely to be carrying out the test. As the colour of iodine is not suitable for easy colour differentiation, the addition of starch to give the blue starch - iodine complex was found to be the simplest means of improving on this.mg m 0.2 0.4 0.6 0.8 1.0 I 2 I .6 I .2 0.8 0.4 0 T.L.V. I 2 3 , 4 5 6 10 20 30 40 50 Ozone in 40-lltre sample,pg Fig. 1. Relationship between optical density of starch - iodine complex and pg of ozone in a 40-litre atmosphere sample : graph A, sample collected in one absorber ; graph B, sample collected in second of two absorbers in series; graph C , sample collected in two absorbers in series Several methods of producing this blue starch - iodine complex were investigated, including the linear starch reagent proposed by Lambert.12 This reagent , which contained cadmium iodide, proved unsatisfactory, giving flocculent precipitates in the presence of the phosphate buffer of the absorbing solution.Finally, a 0.25 per cent. w/v aqueous, soluble starch solution, freshly prepared each day, was found to be the most suitable reagent, producing blue colours that were stable for 30 minutes before fading.510 COHEN, SMITH AND WOOD: A FIELD METHOD FOR THE [Afia&.$t, Vol. 93 It was now necessary to relate the optical density of the colour produced from the released iodine with the weight of ozone in samples of the various standard atmospheres taken under test conditions. Forty-litre samples of standard ozone atmospheres of various concentrations were collected at 2 litres per minute in two all-glass absorbers in series, each containing 10 ml of absorbing solution. Starch was added to each solution and the respective optical densities of the solution in each of the absorbers measured separately at 540 nm with 1-cm cells.Each was plotted against the weight of ozone in the sample. Straight lines, graph A for the first absorber solution and graph B for the second (Fig. l), were obtained for each. The failure of the lines to pass through the origin of the graph was thought to be caused by a certain small uptake of iodine by the starch without a colour being produced.12 A correction for this small amount of iodine was made to the optical density of the solution in the second absorber, and the combined optical densities in both absorbers were plotted against weight of ozone in the sample. Graph C (Fig. 1) was obtained and found to be identical with that obtained when 5 ml of starch solution were added to each of a series of 10 ml of buffered potassium iodide solution containing known amounts of iodine and the optical densities of the colours produced plotted against the ozone stoicheiometrically equiva- lent to the iodine present.This suggests, assuming the 0, = I, relati~nship,~ that the two absorbers in series were collecting 100 per cent. of the ozone from the various 40-litre samples. The use of only one absorber was envisaged in the proposed test, and it can be seen from graph A (Fig. 1) that this ensures a reasonably reproducible trapping of 87 per cent. of the ozone in an atmosphere, over a 0 to 1.2 mg m-3 ozone concentration range. For the purposes of a field test this was considered to be satisfactory and no extra complication would be introduced, provided the colour standards prepared for the proposed test took this into account.Graph A was used for the calibration of all subsequent work on the development of the field test. REMOVAL OF NITROGEN DIOXIDE FROM A MIXED OZONE - NITROGEN DIOXIDE ATMOSPHERE- As previously mentioned, the reaction involving the release of iodine from potassium iodide is not specific for ozone. Nitrogen dioxide, which is likely to be found in certain indus- trial atmospheres, together with ozone, reacts in a similar manner. The problem of separating the two gases was initially tackled by seeking a means of preferentially trapping the nitrogen dioxide prior to the absorption of the ozone in the potassium iodide solution. Various solutions were tried as traps for nitrogen dioxide: chromic acid, sodium hydroxide (1 per cent.w/v), sodium carbonate (1 per cent. w/v), “nitron” (0.1 per cent. w/v), urea (2 per cent. w/v), sulphuric acid (2 per cent. w/v), alkaline permanganate and alkaline arsenite. None of these solutions completely removed the nitrogen dioxide at the sampling rate required, i.e., 2 litres per minute. Powdered manganese dioxide removed most of the gas but also com- pletely removed the ozone. It was found at this stage that cotton-wool and various man-made fibres could not be used as supports for trapping agents as they themselves removed ozone. It now seemed unlikely that a simple system could be devised for trapping nitrogen dioxide, therefore it was decided to attempt to remove the ozone from the nitrogen dioxide and determine it by difference.Two parallel sampling systems were envisaged in which one branch would indicate the total oxidant present and the other, containing an ozone trap, would indicate the total oxidant minus the ozone. The blue colours obtained in the absorbing solutions on the addition of starch would be a measure of oxidant in terms of an ozone equivalent. The fact that nitrogen dioxide did not give a stoicheiometric release of iodine from potassium iodide solution would not invalidate this, provided it released the same amount of iodine in each branch. REMOVAL OF OZONE FROM A MIXED OZONE - NITROGEN DIOXIDE ATMOSPHERE- Ozone was found to be completely removed by a solution of trans-stilbene in sym-tetra- chloroethane. However, carry-over of the solvent into the absorbing solution was a compli- cation.As the use of other solvents did not eliminate this problem, solid-state traps were considered. In particular, the previous findings in connection with attempts to trap nitrogen dioxide suggested that cotton-wool might be used. Preliminary experiments indicated that a plug of cotton-wool would completely remove ozone from a 0.6mgm-3 (0.3 p.p.m. v/v) atmosphere and allow a nitrogen dioxide atmosphere, 9 mg m-3 (5 p.p.m.), to pass unaffected. Further investigation showed that the state of the cotton-wool, untreated, washed, driedAugust, 19681 DETERMINATION OF OZONE IN THE PRESENCE OF NITROGEN DIOXIDE 511 Sampling point To Fig. 2. Field test assembly for the determination of ozone in the presence of nitrogen dioxide or damp, had no apparent effect as far as the trapping of ozone was concerned.Damp material did, however, remove some nitrogen dioxide. The trap was standardised to the use of 1 g of cotton-wool, washed with tap and then de-ionised water, dried at 105" C and packed in a glass tube (Fig. 2) to a length of about 7 cm. Such a trap absorbed about 40 pg of ozone, i.e., 200 litres of a 0.2 mg m-3 atmosphere sampled at 2 litres per minute before breakthrough occurred, as shown in Table I. TABLE I REMOVAL OF OZONE FROM ATMOSPHERES WITH COTTON-WOOL Equivalent volume of 0.2 mg m-s ozone atmosphere passed, Cotton-wool Ozone concentration litres Ozone found after trap, mg m-3 sample, passed through trap, 1 1 g mg m-3 Per run Total 1 0.465 93 93 0 0-450 90 183 0 0-445 89 272 0-265 0-425 0-445 0.440 0-735 0.235 85 85 0 89 174 0 88 262 0.250 147 147 0 94 241 0.100 Fig.2 shows the proposed field test assembly for the determination of ozone in an atmosphere that also contains nitrogen dioxide. DETERMINATION OF OZONE IN STANDARD MIXTURES OF OZONE AND NITROGEN DIOXIDE- In order to assess the validity of the proposed method, a system, represented diagram- matically in Fig. 3, was set up, in which atmospheres of ozone, nitrogen dioxide and a mixture of equal volumes of the ozone and nitrogen dioxide atmospheres could be sampled simul- taneously. The sampling rate in each branch of the system was 2 litres per minute. In Fig. 3, (OT) and (0) represent the amounts of iodine liberated in the absorbing solutions, expressed as ozone equivalents (mg m-3), by 40 litres of the ozone atmosphere with and without a cotton-wool trap in line, respectively.(MT) and (M), and (NT) and (N) represent the iodine liberated by the mixed and nitrogen dioxide atmospheres, respectively, each with and without a similar trap under similar sampling conditions. If the ozone alone is removed by the trap, then (M) minus (MT) should represent the ozone concentration in the mixed atmosphere and should be equal to (0)/2 for a mixture of equal volumes of the two gases.512 COHEN, SMITH AND WOOD: A FIELD METHOD FOR THE (OT) Ozone Ozonejnitrogen 6 dioxide Pump [Analyst, Vol. 93 Nitrogen dioxide A = Flow meter and control valve N = B = Mixing chamber C = Cotton-wool trap D = Absorbing solution 0 = Iodine liberated by ozone NT = atmosphere without cotton- M = wool trap atmosphere with cotton-wool trap OT = Iodine liberated by ozone MT = Iodine liberated by nitrogen dioxide atmosphere without cotton-wool trap Iodine liberated by nitrogen dioxide with cotton-wool trap Iodine liberated by mixed atmos- phere without cotton-wool trap Iodine liberated by mixed atmos- phere with cotton-wool trap Fig.3. Diagrammatic representation of the assembly for the analysis of mixed atmospheres of ozone and nitrogen dioxide A series of determinations was carried out with 1 + 1 v/v mixtures containing concen- trations of ozone over the range 0.1 to 0.4 mg m-3 and various concentrations of nitrogen dioxide from zero to a maximum of 11.4 mg m-3. Some typical results are shown in Table 11. TABLE I1 IODINE, EXPRESSED AS OZONE EQUIVALENTS, LIBERATED IN THE RESPECTIVE ABSORBING SOLUTIONS BY 40 LITRES OF OZONE, NITROGEN DIOXIDE AND MIXED ATMOSPHERES Actual composition of mixture, mgm-3 N i t r o g k n Ozone dioxide* (0) /2 0.118 0.120 0.1 10 0.200 0.215 0-355 0.395 0-460 0.405 0 2.9 8-5 1.7 2.9 1.7 2.9 5-6 11.4 Ozone equivalents found in respective absorbing solutions, mg m--3 A r \ Nitrogen Mixture ozone + Ozone dioxide nitrogen dioxide * Without With W-th W w t h trap trap trap trap trap trap (0) (OT) (N) (NT) (W (MT) 0.235 0 0 0 0.125 0 0.240 0 0.435 0-435 0-335 0.235 0.220 0 1.38 N.D.0.740 0.645 0-400 0 0.290 N.D. 0-335 0.150 0.430 0 0.480 0.495 0-445 0.255 0.710 0 0.305 0.305 0-475 0-170 0.790 0 0-460 0.475 0-610 0.245 0.920 0 0.970 N.D. 0.890 0.450 0-810 0 >1-5 >1*5 1-580 1.210 Ozone found in mixture, mg m--3 (MI - (MT), 0.125 0.100 0.095 0.185 0.190 0.305 0.365 0.440 0.370 * Nitrogen dioxide concentration determined by a modified SaltzmanlO method.N.D. Not determined. The results in Table I1 show that the cotton-wool trap completely removed the ozone, i.e., (OT) was zero, and also that the nitrogen dioxide concentration was unaffected by the trap, i.e., (N) was equal to (NT). The average figures for the ozone concentrations found in all the mixtures analysed by the proposed test procedure are summarised in Table 111. For ease of presentation the concentrations of ozone in the various mixtures have each been converted into the nearest 0.1 mg IT^-^. The standard deviations on the mean ozone concentrations found in each set of replicates are also given.August, 19681 DETERMINATION OF OZONE IN THE PRESENCE OF NITROGEN DIOXIDE 513 TABLE I11 OZONE CONCENTRATIONS IN PREPARED OZONE - AIR MIXTURES, WITH AND WITHOUT ADDED NITROGEN DIOXIDE, AS DETERMINED BY THE PROPOSED PROCEDURE No.of tests 2 5 8 2 11 4 24 4 2 4 3 Nitrogen dioxide concentration in mixture, mg m--3 0 2.9 5-6 8-5 0 2-9 6-6 11-4 0 2.9 5.6 Mean ozone Ozone concentration found concentration in in mixture in mixture ( 0 ) / 2 , (M) - (MT), mg m-3 mg m--3 0-1 0.109 f 0.004 0.1 0-080 f 0-005 0.1 0.087 f 0.021 0.1 0.083 f 0-018 0.2 0.203 & 0.004 0.2 0.174 & 0.009 0.2 0.180 f 0.028 0.2 0.167 f 0.007 0.4 0.407 f 0.016 0.4 0.372 & 0.028 0.4 0.396 & 0.039 Ozone found minus ozone added mg m-s - (MT)l - + 0.009 - 0.020 -0.013 -0.017 + 0.003 - 0.026 - 0.020 - 0.033 + 0.007 - 0.028 - 0.004 It can be seen from Table I11 that when nitrogen dioxide was present in the mixture the ozone found, (M) - (MT), was always slightly less than that added, (0)/2.The mean apparent loss of ozone ranged from 0.004 to 0.033 mg m-3 in the tests carried out. This would not invalidate the proposed test as it is designed to measure the ozone concentration of an atmosphere only to the nearest 0-05mgm-3 in the 0.1 to 0 ~ 2 m g m - ~ range and to the nearest 0.1 mg m-3 in the 0.2 to 0-8 mg m-3 range. This apparent loss of ozone could not have been caused by the interaction of ozone and nitrogen dioxide, as it was found in this series of determinations that within experimental v (O) 4- (N) = (M).Similarly, it could not be attributed to oxidising impurities in the error 2 cotton-wool being carried through by the nitrogen dioxide, because experimentally (N) = (NT). One explanation is that an oxidising product is formed by the interaction of ozone with nitrogen dioxide in the presence of cotton-wool resulting in an increase of (MT) and conse- quently a decrease of (M) - (MT). A possible reaction13 is 2N0, + 0, -+ N,O, + O,, initiated in this reaction by the cotton-wool, with the nitrogen pentoxide having a greater iodine- liberating power than nitrogen dioxide. COLOUR STANDARDS- Colour standards that would match the colours produced in the proposed test were required. It was found that 9mgm-3 of nitrogen dioxide was about equal, on a colour intensity basis, to an ozone equivalent of 0.7 mg m-3 of ozone (this can be calculated from the figures in Table 11).Therefore, a range of standards up to 0-8 mg m-3 ozone equivalents would enable the detection of up to 0.2 mg m-3 of ozone in the presence of up to 7.5 mg m-3 of nitrogen dioxide. It would have been desirable to prepare and use permanent inorganic colour standards. However, inorganic mixtures such as copper and cobalt solutions did not easily produce colours of suitable intensity and turbidity, therefore it was decided to prepare colour standards based on the starch - iodine complex. As the preparation and use of standard iodine solutions was considered to involve problems too exacting for field work, the liberation of iodine from potassium iodate - potassium iodide solution with standard sulphuric acid and subsequent addition of starch was investigated and found suitable.The required colours were slightly purple rather than blue. The red component of the standards was found to be dependent on the potassium iodide concentration. The use of a solution 0.00167 M with respect to potassium iodate and 4.5 per cent. w/v with respect to potassium iodide was found suitable. Aliquots of 0.002 N sulphuric acid over the range 0 to 1 ml were added to 10-ml portions of the potassium iodate - potassium iodide solution. Starch was added to each solution and a graph prepared of optical density against volume of acid used; a straight line was obtained. The optical densities corresponding to 0.1, 0.15, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7 and 0.8 mgm-3 ozone equivalents were read from the ozone calibration graph A, Fig.1. The volumes of514 COHEN, SMITH AND WOOD: A FIELD METHOD FOR THE [Analyst, VOl. 93 acid equivalent to these optical densities were obtained from the graph just prepared. Each of these volumes of acid was that required to be added to potassium iodate - potassium iodide solution to give a series of colour standards for the proposed test. By using a recording spectrophotometer a check was made on any possible absorption peak or optical density change of the coloured complexes formed in a colour standard and an ozone sample. Over a period of up to 30 minutes neither the standard nor the sample colour exhibited any wavelength shift or significant change in optical density. The validity and reliability of this series of standards were tested by collecting atmospheres of known ozone concentration in the presence of 0 to 8.5 mg m-3 of nitrogen dioxide under the conditions of the proposed test and assessing the ozone concentrations with prepared standards.The results are given in Table IV. TABLE IV COMPARISON OF COLOURS PRODUCED FROM MIXED OZONE - NITROGEN DIOXIDE ATMOSPHERES OF KNOWN OZONE CONCENTRATION WITH COLOUR STANDARDS Actual ozone concentration, Ozone concentration indicated mg m-3 .. .. . . 0.12 0.16 0.2 0-24 0.34 0.4 0.74 0.81 by standards, mg m-3 . . 0.1 to 0-15 0.15 to 0.2 0-2 0-2 t o 0.3 0.3 to 0.4 0.4 0.7 > 0.8 With some batches of potassium iodide, standards with higher optical densities than expected were obtained. This was thought to be caused by free iodine or acid, or both, in the solid potassium iodide.Accordingly, the potassium iodate - potassium iodide solution was tested before use and samples that gave a trace of blue colour on the addition of starch and 0.02 ml of 0-002 N sulphuric acid were rejected. This amount of acid is not normally sufficient to release enough iodine to produce a colour with starch. INTERFERENCES- The effects of several possible interferences likely to be encountered in practice were investigated. Sulphur dioxide-This gas would react with, and remove, any iodine that had been liberated by ozone or nitrogen dioxide. However, provided that all of the liberated iodine is not removed and that the same amount is removed in each bubbler, i.e., the sulphur dioxide is not trapped by cotton-wool, the test for ozone would still be valid, as has been found experimentally.Humidity-It was found that the method is not affected by atmospheres with humidities normally occurring in this country. Higher humidities, approaching saturation at room temperature (21" C), had no effect on the removal of ozone by cotton-wool, but some of the nitrogen dioxide was trapped when the wool became damp. Metal fume-Ozone and nitrogen dioxide are reputed to be formed during electric arc- welding of metals. Tests carried out under factory conditions showed that the presence of iron-iron oxide fume made gas testing difficult. No means of removing this fume was apparent, and any iron oxide trapped on a pre-filter was thought to remove some of the ozone.This latter point was confirmed in laboratory tests when small amounts of iron oxide were found to remove about 20 per cent. of the ozone present in 40 litres of a 0.4 mg m-3 ozone atmosphere. Oxidartts-As nitrogen dioxide is the most likely oxidant to be found in an industrial atmosphere together with ozone, the interfering effects of other oxidants on the test were not assessed. However, other oxidants should not invalidate the test, provided that they are not trapped by cotton-wool. SAMPLING PROBES- In the field the desired sampling point is often inaccessible, and a length of flexible tube is needed to enable sampling to be carried out. The suitability for this purpose of tubes made of glass and various flexible materials was assessed. Successive 0.4 mg m-3 atmospheres of ozone were passed at 2 litres per minute for 20 minutes through 60-cm lengths of each, previously unused, tube and the emergent atmosphere analysed. Table V lists the results obtained.Apart from the glass tube, which seemed to require no conditioning, PVC andAugust, 19681 DETERMINATION OF OZONE IN THE PRESENCE OF NITROGEN DIOXIDE 515 nylon appeared to be the most suitable of the flexible materials assessed. Tubes (200 cm in length) of these two latter materials were then evaluated, and from the results (Table V), nylon tubing was selected as the most suitable flexible material to use as a sampling probe. TABLE V COMPARISON OF THE SUITABILITY OF TUBES OF VARIOUS MATERIALS AS SAMPLING PROBES Internal diameter, Tube material mm Glass .. .. . . 6 PVC .. .. . . 6 Silicone rubber . . 4 Polythene . . .. 5 Nylon . . .. .. 5 Nylon . . .. .. 5 PVC . . . . . . 6 Ozone passed in successive runs with the same tube, Length, per cent. cm A f > 60 60 60 60 60 200 200 99, 101, 105, 102, 103 87, 91, 99, 96, 95 76, 83, 78, 83, 81 93, 95, 91 108, 106, 105 86, 86, 89, 92, 93 95, 97, 100, 99, 102 Altshuller and Wartburg,14 in a study of the interaction of ozone with various materials in a dynamic flow system, found that Teflon passed ozone without loss and also that glass required only a short conditioning period. These workers did not study nylon in this respect. PROPOSED FIELD METHOD FOR THE DETERMINATION OF OZONE IN THE PRESENCE OF NITROGEN DIOXIDE APPARATUS- All-glass absorbers-Two of the type are shown in Fig.2. from Glass of Mark Ltd., Jubilee Street, Northwich, Cheshire. Glass tube-This is of the type shown in Fig. 2. GEass test-tubes-Optically matched, with 13.5-mm i.d. Flow meters-Two, capable of measuring 2 litres per minute. Control valves-Two, suitable for control at a flow-rate of 2 litres per minute. Sampling $ump-Capable of drawing air through the apparatus at 4 litres per minute. Remote sampling probe-A length of nylon tubing, 5-mm i.d. REAGENTS- Unless otherwise stated, all reagents should be of analytical grade and solutions made with distilled or de-ionised water. Potassium iodide, neutral, bufered solution-Dissolve 20 g of potassium iodide, 14.2 g of disodium hydrogen phosphate and 13.6g of potassium dihydrogen phosphate in water and dilute to 1 litre.Starch solution-Dissolve 0-25 g of soluble starch in about 70 ml of boiling water, cool and dilute to 100ml. Prepare fresh daily. Cotton-wool, absorbent, B2.C.-Wash successively with tap water and distilled or de- ionised water and dry at 105" C. Store in air-tight glass bottles. Dilute sulphuric acid, 0.002 N. Potassium iodate solution, 0.0167 M-Test and reject any solution that gives a blue colour with starch solution. (Potassium iodate solution of this concentration is obtainable com- mercially.) Potassium iodate - potassium iodide solution-Dissolve 45 g of potassium iodide in water, add 100 ml of the 0.0167 M potassium iodate solution and dilute to 1 litre. Prepare this solution fresh daily and test for the absence of free iodine by adding 0.02 ml of the 0.002 N sulphuric acid and 5 ml of the starch solution to a 10-ml portion.If a blue colour is formed, reject and prepare a new solution with a fresh bottle of potassium iodide. PROCEDURE- 10 ml of neutral, buffered potassium iodide solution into each absorber. atmosphere assemble the two branches of the apparatus as shown in Fig, 2. These absorbers are obtainable These tubes are obtainable from Suitable flow meters are Tintometer Ltd. obtainable from Rotameter Manufacturing Company Ltd. Pack 1 g of cotton-wool into the glass tube to form a plug about 7 cm long. Pipette In an uncontaminated Preferably,516 COHEN, SMITH AND WOOD: A FIELD METHOD FOR THE [A%&?yst, VOl. 93 use glass ball-and-socket joints. Alternatively, ensure that all glass-to-glass connections are butted and held together with PVC tubing.To prevent water condensing in the flow meters, insert a dried cotton-wool plug between the exit of each absorber and the entry to each flow meter. In the laboratory and before proceeding to field testing, calibrate each flow meter separately on the assembled apparatus to ensure that an equal flow-rate, 2 litres per minute, is obtained in each limb. This can be done either by the use of a third flow meter, open to the atmosphere and connected in series to the flow meter being calibrated, or by any other suitable method, eg., soap film ca1ibrat0r.l~ Draw the test atmosphere from a singZe sampling point through each absorber at a rate of 2 litres per minute for 20 minutes. Remove the absorbers from the assembly in an uncontaminated atmosphere, and then the heads from the absorbers, taking care to drain any liquid in the inlet tubes into the main solutions.Pipette 5 ml of the starch solution into each absorber. Mix well and transfer each solution into the optically matched test-tubes and compare the colour produced with the colour standards within 15 minutes. PREPARATION OF COLOUR STANDARDS- To each of 9 test-tubes add 10 ml of the potassium iodate - potassium iodide solution and prepare the set of colour standards by adding the appropriate amount of the 0.002 N sulphuric acid to each tube (see Table VI), followed by 5 ml of the starch solution. Mix well and transfer each solution to the optically matched tubes. (The colour standards are stable for 30 minutes.) TABLE VI VOLUMES OF ACID REQUIRED TO PREPARE COLOUR STANDARDS Ozone equivalent, mgm-3 0.1 0.15 0-2 0.3 0.4 0.5 0.6 0.7 0.8 Sulphuric acid, 0.002 N, ml 0.08 0.14 0.20 0.30 0.41 0.52 0-63 0-74 0.85 To obviate the need to prepare a full set of colour standards the following procedure can be adopted.Add 0.1-ml increments of the sulphuric acid to a mixture of 10 ml of the potassium iodate - potassium iodide solution and 5 ml of the starch solution until an approxi- mate match with the sample colour is obtained. Note the volume of acid added and prepare standards equivalent to, and adjacent to, this volume. Match the sample colour with the freshly prepared standards. If a spectrophotometer is available, a graph of the optical densities of the prepared standards against their respective ozone equivalents can be plotted.From the measured optical densities of the samples in the two absorbers a more precise determination of the ozone concentration in the atmosphere sampled is possible. As an alternative to the preparation of the above colour standards, a series of permanent standard colours on a comparator disc available from Tintometer Ltd., can be used for this test. Calculation- Ozone in sample (mg m-3) = Ozone equivalent in absorber (M) minus ozone equivalent in absorber (MT) . APPLICATION OF METHOD- The proposed method has been assessed under field conditions at various industrial sites where ozone is known, or reputed, to occur as a contaminant in the atmosphere. These included sites where discharge tubes in certain photocopying machines were being tested, plastic sheet was being etched by using high voltage spark discharges and inert-gas shielded arc welding was practised.The field testing in the vicinity of high voltage spark discharges revealed the presence of both ozone and nitrogen dioxide, the concentrations of the latter gas being determined by a P-anisidine paper field test method.le In the welding process, as already mentioned, the presence of metal fume interfered with the determination of ozone. This fume would presumably interfere with any test that required the drawing of an air sample through an absorber. These few field tests confirmed the results of those tests in the laboratory showing that the proposed method is suitable for detecting up to 0.2 mg m-3 of ozone in the presence of up to 7.5 mg m-3 of nitrogen dioxide.In the absence of nitrogen dioxide up to 0-8mgm-3 of ozone can be determined.August, 19681 DETERMINATION OF OZONE IN THE PRESENCE OF NITROGEN DIOXIDE 517 The apparatus required for the test is portable and requires only an external electrical power supply to operate the pump. The complete assembly can be mounted in a box 8 x 12 x 15 inches. A complete determination takes about 30 minutes. As we prefer to determine nitrogen dioxide in any atmosphere by a specific field test method,lO no attempt was made in the present work to develop a method for the deter- mination of this gas on the basis of its reaction with potassium iodide and subsequent release of iodine. However, an approximate estimate of nitrogen dioxide concentration in a mixed ozone - nitrogen dioxide contaminated atmosphere, containing no other oxidant, can be made by converting the intensity of the colour in the absorber (MT), expressed as ozone equivalents, into mgm-3 of nitrogen dioxide. This is done by multiplying this ozone equivalent figure by 13, as it was observed in this work that 9 mg m-3 of nitrogen dioxide gave about the same colour intensity as 0.7 mg m-3 of ozone. This work was carried out on behalf of the Ministry of Labour Committee on Tests for Toxic Substances in Air. We are grateful to the Government Chemist for permission to publish this paper, and to H.M. Factory Inspectorate for arranging the field tests. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. REFERENCES “hlinistry of Labour, Safety, Health and Welfare, New Series No. 8, Dusts and Fumes in Factory Bravo, H. A., and Lodge, J. P., jun., Analyt. Chem., 1964, 36, 671. Hendricks, R. H., and Larsen, L. B., Amer. Ind. Hyg. Ass. J., 1966, 27, 80. Hauser, T. R., and Bradley, D. W., Analyt. Chem., 1966, 38, 1529. Lidzey, R. G., and Longmaid, F. M., Chem. & Ind., 1964, 150. Nash, T., Atmos. Envir., 1967, 1, 679. Deckert, W., 2. analyt. Chem., 1956, 153, 189. Bovee, H. H., and Robinson, R. J., Analyt. Chem., 1961. 33 1115. Saltzman, B. E., and Gilbert, N., Ibid., 1959, 31, 1914. -, Ibid., 1954, 26, 1949. A.S.T.M. Standards, Part 23, Philadelphia, 1965, p. 730. Lambert, J . L., Analyt. Chem., 1951, 23, 1247 and 1251. Johnston, H. S., and Yost, D. M., J. Chem. Phys., 1949, 17, 386. A.ltshuller, A. P., and Wartburg, A. F., Int. J . Air Wat. Pollut., 1961, 4, 70. Gooderham, W. J., J . SOC. Chem. Ind. Lond., 1944, 63, 351. “Report of the Government Chemist, 1965,” H.M. Stationery Office, London, 1966, p. 119. Atmospheres,” Third Edition, H.M. Stationery Office, London, 1966. Received January 19th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300507

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

518 Analyst, August, 1968, Vol. 93, pp. 518-521 Molecular-emission Spectroscopy in Cool Flames Part In.* The Emission Characteristics of Tin in Diffusion P b e s BY R. M. DAGNALL, K. C. THOMPSON AND T. S. WEST (Chemistry Department, Imperial College, London, S. W. 7 ) Tin can be determined in the range 3 to 3000 p.p.m. by molecular-band emission from the SnH species in a nitrogen - hydrogen diffusion flame. The emission a t 609-5 nm, which is almost line-like, gives a limit of detection of 1.5 p.p.m. of tin. Spectral interference from sodium is eliminated by use of a didymium filter. The presence of oxygen gives rise to a much broader spectrum, caused by tin(I1) oxide formation, with a visual limit of detection of 5 p.p.m. of tin. Atomic emission can be observed only in the presence of alcohols, e g ., isopropyl alcohol, but a high concentration of ground-state tin atoms exists in the diffusion flame. Mechanisms are discussed to explain the production of tin atoms, SnH and the resonance-line emission. TIN is an example of an element that is difficult to excite thermally in conventional pre-mixed flames. Herrmann and Alkemade,l for example, state that only in an oxy-cyanogen flame are the tin lines intense enough to be used analytically. In other flames, weak resonance emission can only be observed in the inner reaction zones. However, the cool diffusion flames of nitrogen and hydrogen used in earlier papers in this series for the determination of sulphur2 and phosphorus3 have been found to promote emission from molecular species not encountered in most other flames.The purpose of this paper is to report the results of a study involving tin in diffusion flames, which is of both analytical and spectroscopic interest. The diffusion flame obtained by burning a mixture of hydrogen and nitrogen in air at atmospheric pressure yields an intense red colour in its cool inner region when a tin(I1) chloride solution is sprayed into it. A thin blue outer mantle may also be observed when there is a high enough concentration of oxygen to form the tin oxide species. SPECTRAL CHARACTERISTICS BAND EMISSION- The spectrum, recorded on a Unicam SP9OOA flame spectrophotometer, consisted of a sharp peak at 6096 nm and a much weaker band with a maximum a t 621 nm (Fig. 1 (b)). Sodium is included to show the line-like nature of the Sn-H emission.The former peak is listed by Herrmann and Alkemadel and corresponds to the SnH species. Pearse and Gaydon4 also list the weaker band at 621 nm, which has only been observed previously in a tin arc operating in a hydrogen atmosphere. The emission at 609.5 nm is not very dependent on hydrogen pressure or on the position of measurement in the flame, although it reaches a maximum about 1 cm above the top of the burner head. The intensity of the emission is directly proportional to the concen- tration of tin(I1) and also to the square of the slit width (which is to be expected for a band- emission signal). Linear calibration graphs may be obtained over the range 3 to 3000 p.p.m. of tin with a detection limit of about 1.5 p.p.m.(signal-to-noise ratio = 1). The effects of extraneous cations and anions were not investigated but, unless a monochromator of fairly high resolution is used, spectral interference may be expected from sodium. However, in this instance a didymium filter interposed between the flame and monochromator slit was found to reduce the sodium emission at 589 nm by 99.5 per cent., while at the same time transmitting about 60 per cent. of the emission caused by SnH at 609.5nm. * For details of earlier parts of this series, see reference list, p. 521. 0 SAC and the authors.DAGNALL, THOMPSON AND WEST 519 Because the flame temperature is low2 (down to 280" C) there is little thermal emission, even from the most easily excited elements. The noise level caused by the flame is also low because the flame exhibits negligible background, even over the OH band region.Some matrix effects caused by aluminium and magnesium have been observed in former studies5 with this flame, and they were overcome then by introducing a small volume of oxygen through an auxiliary jet in the burner stem6 to increase the flame temperature. In this instance this would not be possible, as the presence of a small amount of oxygen replaces the red colour of the SnH species by a blue coloration. The recorded spectrum of the latter showed a continuum with superimposed bands from 360 to 540 nm, with a maximum intensity at 484nm, and corresponded to the tin(I1) oxide species,l cf. Fig. 1 (a). No analytical use can be made of the blue colour with the Unicam SP9OOA spectrophotometer because of the absence of sharp bands, but it is sensitive visually with a limit of detection of about 5 p.p.m.of tin. I I I 1 I 3 (a) (b) x v) C W U c .- - I I I I I Fig. 1. Emission spectrum of Sn-0 and Sn-H species: (a), spectrum of 3 x 1 0 - 2 ~ tin(I1) ( 0 . 5 ~ hydrochloric acid) in the nitrogen - air -hydrogen flame. Slit, 0.03 mm, gain 3,10, bandwidth 1; (b) spectrum of 3 x M tin(I1) (0.5 M hydrochloric acid) in the nitrogen - hydrogen flame. Slit, 0.03 mm, gain 3,lO. band width 1 Sodium Solutions of tin(1V) chloride produced an exactly similar spectrum [i.e., SnH and tin(I1) oxide], but the emission was about 60 per cent. more intense at 609.5 nm than that obtained with tin(I1) chloride. The increased response to tin(1V) is probably caused by the greater volatility of its chloride, which would be more in evidence in a cool flame than in a hot one.As a result of these observations it may be concluded that a substantial number of elements whose salts do not break down at the temperature of this flame (about 280" to 400" C) would cause matrix effects and thus interfere although in many instances it should be possible to produce calibration graphs of lesser gradient if the amount present were approximately known. The emission characteristics of the SnH species were also observed in the air - hydrogen separated flame.2 However, under these conditions, the emission signal was only about half of that in the diffusion flame. Further, concentrated solutions of tin caused severe staining of the quartz separator, and the flame exhibited a higher background than the diffusion flame.The SnCl and SnBr bands4 are not observed in these flames, unless the halogen concen- tration is high, e.g., 0.1 M in hydrochloric or hydrobromic acids. Under theselatter conditions a thin dark blue cone appears when chloride is present and a similar shaped green cone when bromide is present. Each cone extends about 1 cm above the top of the burner head in the coolest portion of the flame. The red SnH bands were present as before, extending much higher in the flame. It would seem that the SnCl and SnBr bands arise from the chemi- luminescent reduction of tin chloride and bromide clotlets by hydrogen. Anionic interference might be more marked, however.520 DAGNALL, THOMPSON AND WEST MOLECULAR-EMISSION [AndJJSt, VOl.93 LINE EMISSION- No thermal emission of tin at the principal resonance lines 284 and 286.3 nm was observed from aqueous solutions of tin(I1) or tin(1V) salts in any diffusion flame. However, a 1000 p.p.m. tin solution, 40 per cent. w/v in isopropyl alcohol, gave not only strong emission signals at 284 and 286.3 nm, but also many other tin lines extending far into the ultraviolet region. The major lines observed were at 286.3, 284, 270.6, 254.7, 243, 235.5, 226.9 and 224.6 nm. These lines have previously been observed by Gilbert from isopropyl alcohol solutions of tin in air - hydrogen flames with a total-consumption burner.' The chemiluminescence emission from the diffusion flame is clearly observable at 284nm even in the presence of 1 per cent.isopropyl alcohol. For emission at the 224.6-nm line, an excitation energy of 5.5 eV is required which is considerably greater than that available in the diffusion flame. Although tin atoms are produced in the diffusion flame from aqueous solutions [from atomic-absorption studies of tin in diffusion flames (R. M. Dagnall, K. C. Thompson and T. S. West, unpublished work)], no emission can be observed as excitation does not occur. It was not possible to determine whether the emission from the isopropyl alcohol solutions emanated from the relatively cool oxygen-free central flame regions or from the outer reaction zone. As a result of these observations we are led to conclude that the production of tin atoms in such diffusion flames must arise as the result of chemical reaction rather than thermal decomposition of tin(I1) oxide or tin(I1) chloride.In the presence of isopropyl alcohol the tin atoms most probably undergo excitation by decomposition products of isopropyl alcohol, e g . , CH. Gibson, Grossman and Cookeg have suggested a three-body reaction to account for this emission from isopropyl alcohol solutions in an air - hydrogen flame, viz.: Sn + CH + OH -+ Sn* + CO + H,. The emission observed in our studies extended throughout the flame and was quite intense, being as high as 4 cm above the burner head. When air was supplied to the diffusion flame through an auxiliary jet in the burner baseJ6 the emission higher up in the flame decreased with increasing air pressure, while the emission just above the burner head increased.The presence of air created an inner (ie,, primary) reaction zone from which the tin emission emanated. Ethanol, methanol and acetone gave emission signals similar in intensity to isopropyl alcohol. Although the emission increased with increasing alcohol concentration, it was not linear, the emission for a 40 per cent. w/v isopropyl alcohol - tin solution being only 50 per cent. higher than a 20 per cent. w/v solution. Glycerol gave only slight emission at 284 nm from a 1000 p.p.m. tin solution, but this could have been caused by increasedviscosity. EXPERIMENTAL APPARATUS- A Unicam SP9OOA atomic-absorption - flame-emission spectrophotometer was used in its emission mode of operation, coupled to a 0 to 10-mV Servoscribe recorder. The normal EM1 9529B photomultiplier was replaced by a more ultraviolet-sensitive EM1 9601B photomultiplier.FLAME CONDITIONS- The nitrogen-hydrogen diffusion flame was obtained with an acetylene jet in the standard burner base, with the 1.8 x 7.5-cm air - acetylene emission head supplied. The nitrogen pressure was set at 15 p s i . , and the hydrogen pressure corresponded to areading of 16cm on the dibutyl phthalate filled manometer gauge. The top of the burner head was set about 06cm below the bottom of the monochromator slit. REAGENTS- to AnalaR specifications were suitable. hydrochloric acid and dilute to 1 litre with distilled water. The reagents used were of analytical-reagent grade ; we found that reagents conforming Tirt(I1) soZution, 2000 p.p.m.-Dissolve 2 g of tin metal in 200 ml of concentratedAugust, 19681 SPECTROSCOPY IN COOL FLAMES.PART I11 521 PREPARATION OF CALIBRATION CURVE (20 TO 200 P.P.M. OF TIN)- Transfer by pipette 1 to 10ml of a 2000 p.p.m. tin(I1) solution into a series of 100-ml calibrated flasks and dilute to volume with distilled water. Nebulise into the nitrogen- hydrogen diffusion flame by using a slit width of 0.15 mm, gain 3,lO and band width 3, and measure the emission at 609.5 nm. No background correction is necessary. The readings should be taken soon after solution preparation because a decrease in response is obtained on standing (e.g., overnight). This is probably caused by the hydrolysis of the tin(I1) chloride. Other concentration ranges can be prepared by suitable dilution and appropriate adjustment of the slit and gain controls.CONCLUSION Although the SnH species, which are shown to be prevalent in diffusion flames of nitrogen and hydrogen, provide a sensitive method of analysis for tin, the mechanism of their production is perhaps of even greater interest. The cool diffusion flame (burning on a Perkin-Elmer triple-slot burner) is known to provide as good an atom reservoir for tin as an air - hydrogen pre-mixed flame for atomic-absorption studies (R. M. Dagnall, K. C. Thompson and T. S. West, unpublished work), The production of tin atoms, and subsequently SnH, presumably occurs via an exothermic reduction process involving hydrogen and tin( 11) chloride. The thermal energy of the diffusion flame is insufficient to dissociate tin(I1) oxide (dissociation energy 5.7 eV), and hence the breakdown must be caused by the highly reducing nature of the flame and the almost complete absence of molecular or atomic oxygen. We have suggested this type of breakdown in a previous paper to explain the dissociation of certain metal phosphate^.^ The air - hydrogen pre-mixed flame gives more than twice the absorbance at the major tin resonance lines of the nitrous oxide - acetylene or air - acetylene flames and this again could be caused by a lower concentration of atomic oxygen.The emission produced from alcoholic tin solutions in the &flusion flame appears to favour a reaction mechanism involving a carbonaceous species such as the CH radical, but not the CN radical as this species is not observed in the flame spectrum.Support is given to this theory because there is insufficient thermal energy available to give excitation at wavelengths as low as 224.6 nm. In addition, resonance-line emission can be observed from aqueous solutions of tin in the nitrogen - acetylene diffusion flame maintained on the same burner system as before. This flame exhibits a light blue region extending 2 to 3 cm above the burner top and is completely luminous above this. The tin emission may be observed in the lower light blue region that shows only CH and C, emission. In general, the emission intensities are only about 10 per cent. of those observed from the inner reaction zones of a slightly fuel-rich air - acetylene flame maintained on the same burner. There is negligible emission above the inner cones of the pre-mixed flame. One of us (K.C.T.) wishes to thank the Science Research Council for the award of a grant. We also thank Unicam Instruments Limited for the loan of the spectrophotometer used in this study. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. NOTE-References 2 and 3 are to Parts I and I1 of this series, respectively. Herrmann, R., and Alkemade, C . T. J., “Chemical Analysis by Flame Photometry,” Interscience Dagnall, R. M., Thompson, K. C., and West, T. S., Analyst, 1967, 92, 506. -,-,- , Ibid., 1968, 93, 72. Pearse, R. W. B., and Gaydon, A. G., “The Identification of Molecular Spectra,” Chapman and Dagnall, R. M., Thompson, K. C., and West, T. S., Talanta, 1967, 14, 1467. Mackison, R., Analyst, 1964, 89, 764. Gilbert, P. T., “Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy,’’ Gibson, J. H., Grossman, W. E. L., and Cooke, W. D., Analyt. Chem., 1963, 35, 269. Publishers Inc., New York and London, 1963. Hall Ltd., London, 1950. March, 1961. Received February 1st 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300518

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

522 Analyst, August, 1968, Vol. 93, j@. 522-527 The Determination of Beryllium by Thermal-emission and Atomic-fluorescence Spectroscopy in a Separated Nitrous Oxide = Acetylene Flame BY D. N. HINGLE, G. F. KIRKBRIGHT AND T. S. WEST (Chemistry Department, Imperial College, London, S. W.7) The determination of beryllium by thermal-emission and atomic- fluorescence spectroscopy in a separated nitrous oxide - acetylene flame has been investigated. The high temperature, reducing nature and low back- ground of this type of flame permit the determination of down to 0.25 and 0.03 p.p.m. of beryllium in aqueous solution by measurement of thermal emission and atomic fluorescence , respectively. A beryllium microwave- excited, electrodeless discharge tube was used for excitation of the atomic fluorescence.The optimum conditions for both techniques have been investigated, and the effects of other cations and anions studied. THE growing utilisation of beryllium in nuclear reactor technology, and as an alloying element, has resulted in the need for sensitive methods for the determination of this element. Thus its determination by flame-spectrophotometric procedures in hot flames has recently been investigated by several workers. Gilbert,l for example, has reported its detection by emission in an oxygen-sheathed, fuel-rich, oxy-acetylene flame burning at a total-consumption nebu- liser burner. The limit of detection at 234436nm was 0.4 p.p.m. for an aqueous solution and 0-1 p.p.m. for a methanolic solution. Beryllium has also been detected in emission down to 1 p.p.m.in an oxy-cyanogen flame.2 The determination by atomic-absorption spectroscopy in oxy-acetylene and nitrous oxide - acetylene flames has recently been described by Ramakrishna, Robinson and West.3 These authors report sensitivities of 0.09 p.p.m. for 1 per cent. absorption in oxy-acetylene and 0.02 p.p.m. in nitrous oxide - acetylene, both at 23446nm. We also observe a sensitivity of 0-024 p.p.m. at 23446nm for atomic- absorption spectroscopy in a 5-cm nitrous oxide - acetylene flame. Koirtyohann4 reports a detection limit of 0.2 p.p.m. for the atomic emission of beryllium at 234436 nm in a nitrous oxide - acetylene flame. The determination of beryllium by atomic-fluorescence spectroscopy does not appear to have been investigated. The hot, reducing pre-mixed nitrous oxide - acetylene flame provides good sensitivity in the determination of refractory elements by flame-emission spectroscopy, although it exhibits high background emission and noise levels.In an earlier communication from this laboratory, we have described the separation of this flame by using a silica tube.5 In this device, the primary and secondary reaction zones of the flame are mechanically separated, so that it is possible to view the hot, reducing interconal zone without interference from the radiation produced by the secondary “diffusion” zone. Thus the flame-background emission is severely depressed, and the separator tube prevents ingress of atmospheric oxygen to the interconal zone. The reducing nature of this zone is thereby protected, so that high popula- tions of free atoms of the refractory elements are more efficiently maintained than in the unseparated flame.This paper describes the sensitive determination of beryllium by flame- emission spectroscopy in the separated nitrous oxide - acetylene flame, and an investigation of its atomic-fluorescence spectroscopic behaviour in this flame by using a beryllium micro- wave-excited, electrodeless discharge-tube source. 0 SAC and the authors.HINGLE, KIRKBRIGHT AND WEST EXPERIMENTAL APPARATUS - 523 The experimental arrangement used is shown diagrammatically in Fig. 1. The mono- chromator, B, and photomultiplier, A, used were those of the Techtron AA4 flame spectro- photometer (Techtron Pty. Ltd., Melbourne, Australia). The monochromator was fitted with a wavelength-scanning motor (1 nm per minute), and the analytical signals were displayed on a Sargent SRL recorder.A beryllium electrodeless discharge tube, E, (transparent Vitreosil tubing, i.d. 8 mm, length 70 mm) was prepared at an argon-filler gas pressure of 0.03 mm of mercury and contained about 1 mg of pure powdered metallic beryllium and a slightly smaller amount (less than 1 mg) of elemental iodine. Other beryllium halide tubes were prepared Thermal emission Atomic fluorescence Details of experimental arrangement used (lettered parts of the apparatus are referred to in the text) Fig. 1. containing varying proportions of beryllium and chlorine or bromine, but only sources pre- pared with excess of beryllium and a small amount of iodine gave the required stability of output (k 1 per cent.per hour) and intensity. The discharge was maintained with a Micro- tron 200 microwave generator at a Model 214 L discharge cavity, F [Electro-Medical Supplies Ltd. (Greenham) London, W.11. In the Techtron AA4 flame spectrophotometer a 285-Hz amplification system is used. Consequently, for flame-emission spectroscopy measurements it was necessary to modulate the signals at this frequency by using a 285-Hz modulating chopper, D, between the flame, C, and monochromator entrance slit. Likewise for the atomic- fluorescence spectroscopy measurements, modulated fluorescence signals were obtained by placing this modulating device, D, immediately in front of the electrodeless discharge tube source, which was itself arranged so that its radiation was focused on the interconal zone of the separated flame, C, along an axis at 90" to the burner - monochromator axis.The radiation was focused on the flame with a quartz lens of 64-mm focal length attached to the modulating chopper. Burner arrangement-The burner and separator used in this study are shown in Fig. 2. The burner head is similar to that previously de~cribed,~ and consists of a stainless-steel head with a circular slot 0.75 mm wide and 11-mm internal diameter. The central steel plug of the burner head is held centrally in position by three steel grub-screws (6 B.A.) inserted at 120" intervals at the same height through the burner barrel. The base of the burner head was provided with a copper water-cooling jacket.This serves to minimise the deposition of carbon on the burner head when nitrous oxide-acetylene flames are used. A silica524 HINGLE et al. : DETERMINATION OF BERYLLIUM BY THERMAL- [Analyst, Vol. 93 separator tube of 24-mm internal diameter and length 5cm was found most suitable for the maintenance of stable separated flames with this burner. Two short silica side-arms were fused to the separator tube at the same height and at 90" to each other. These were fitted with B19 cones, to which B19 ground-glass sockets fitted with optical quality silica end- windows were attached. -100 mrn H Burner slot- Fig. 2. Burner and separator assembly Nitrous oxide and acetylene pressures, and flow-rates, were controlled by the gas control unit on the Techtron flame spectrophotometer.The indirect nebuliser of the Techtron flame spectrophotometer was used, so that sample solutions were nebulised on nitrous oxide. The separated flame was supported at an operating pressure of 15 p.s.i. of nitrous oxide (flow-rate about 7 litres per minute). A stable primary reaction zone was then obtained at the slot of the burner head. The diffusion zone, where the carbon monoxide and hydrogen produced in the primary zone burn with support from the atmosphere, is maintained at the top of the separator tube. The interconal zone of the flame, which contains the hot, reducing region of analytical interest, is then at the same height as the side-arms and observation windows. REAGENTS- Beryllium stock solzttiort-Pure beryllium metal, 0.100 g (Imperial Metal Industries Limited), was dissolved in 10 ml of aqua regia and the solution diluted to 100 ml with distilled water.This 1000 p.p.m. solution was diluted with distilled water as required to 10 or 1 p.p.m. All other reagents used were of analytical-reagent grade. RESULTS THERMAL-EMISSION CHARACTERISTICS- The background-emission spectrum of the interconal zone of the separated nitrous oxide - acetylene flame has been described in a previous paper from this laboratory.6 The OH bands and the continuum emission due to the combustion of carbon monoxide is virtually absent in the interconal zone, and the only significant background arises from the CN species, which is chiefly responsible for the reducing atmosphere of the zone. The variation of back- ground emission of CN and OH with fuel flow-rate for the separated and unseparated (conventional) flames is shown in Fig.3 (a and b). The atomic emission of beryllium introduced into the flame was examined relative to the flame background in the visible and ultraviolet region between 200 and 600nm. The only intense atomic emission observed was from the beryllium resonance line at 234.86 nm, which results from the IP, -+ IS, transition. Against the low background of the separated flame it was possible to observe emission from the unresolved non-resonance beryllium triplet at 332-1 nm (as, + 3P2,,,,) when strong beryllium samples were nebulised. The variations in the intensity of the beryllium emission at 234.86 nm, and that of the flame background, with acetylene flow-rate, are shown in Fig.4(a) for both the separated and unseparated (conventional) flames. In both flames, the background is minimal in the slightly fuel-richAugust, 19681 EMISSION AND ATOMIC-FLUORESCENCE SPECTROSCOPY 100- 90 80 70 60 50 40 30 20 10 Acetylene flow-rate , litres per minute + - - - - - - - - - Acetylene flow-rate , iitres per minute - 525 Fig. 3. Effect of acetylene flow-rate on: (a), CN emission a t 388.3 nm; (b), OH emission at 306.4 nm. Curve A, unseparated flame alone; curve B, unseparated flame, water nebulised; curve C , separated flame alone; and curve D, separated flame, water nebulised flame, while the beryllium emission reaches a maximum in the same flame. A flame burning at this mixture strength was used for all other investigations. The observed variation in the beryllium emission can be explained by the need for a fuel-rich, reducing flame for the efficient maintenance of the free beryllium atom population, and the decrease in its emission in very rich flames by a slight drop in flame temperature under these conditions.Fig. 4(a) also shows that the background-emission intensity and noise level obtained in the slightly fuel-rich, separated flame is lower than in the conventional flame at 234436 nm, whereas the beryllium emission is slightly more intense in the separated flame than in the unseparated flame despite the slightly lower temperature of its interconal zone. This may be attributed to the more efficient maintenance of a population of atoms by the protected reducing atmosphere in this zone, and to its extended size.Acetylene flow-rate, litres per minute ___j Acetylene flow-rate, litres per minute - Fig. 4. Effect of acetylene flow-rate : (a) on beryllium thermal emission and background emission a t 234-86 nm; curve A, flame background, unseparated; curve B, flame background, unseparated, water nebulised; curve C, flame background, separated ; curve D, flame background, separated, water nebulised ; curve E, beryllium emission for 1 p.p.m. in separated flame; and curve F, beryllium emission for 1 p.p.m. inunseparated flame ; ( b ) , on atomic-fluorescence emission for 1 p.p.m. beryllium solution a t 234.86 nm526 HINGLE et al. : DETERMINATION OF BERYLLIUM BY THERMAL- [Autalyst, Vol. 93 SENSITIVITY AND PRECISION- Linear calibration graphs were obtained for the beryllium emission a t 234.86nm in the slightly fuel-rich flame over the range 0.3 to 10 p.p.m., with an optimised slit width of 200 pm (0.66-nm band width). The limit of detection, which is defined as that concentration producing a signal-to-noise ratio of 1 : I, was 0.1 p.p.m.The limit of practical measurement with the experimental arrangement used was considered to be about 0.25 p.p.m. In a conventional fuel-rich flame (unseparated), the high noise levels result in a detection limit of about 2 p.p.m. of beryllium. Repetitive introduction (10 times) of a 3 p.p.m. beryllium solution into the separated flame under the optimal conditions used to obtain calibration values indicated a relative standard deviation in the emission intensity of 1.3 per cent.EFFECT OF DIVERSE IONS- The effect of 100 and 1000-fold excesses by weight of a range of cations and anions on the emission intensity produced by 3 p.p.m. of beryllium under optimised conditions by using a 200-pm slit width was investigated. The following elements did not interfere (k, they produced a variation in emission intensity of not greater than twice the standard deviation) when present in 1000-fold excess: Bi, Cd, Cu, Fe(III), Hg(II), K, Mg, Mn(II), Na, Pb, Se(IV), Te(IV), U(VI), Zr and PO:-. No interference was observed when beryllium was aspirated in the presence of 5 per cent. sulphuric acid or 10 per cent. nitric, hydrochloric and hydrofluoric acid solutions. On the other hand, 1000-fold excesses of the following ions produced interference, frequently by obstruction of the burner slit by deposition of solid, but the presence of 100-fold excesses caused no interference: Al, As(V), Ca, Cr(VI), Mo(VI), Nb(V), Ti(IV), W(V1) and borate. A 100-fold excess of the following ions produced the error in the beryllium emission given in parentheses: Co(I1) (+20 per cent.), Ni(I1) (+95 per cent.), Si(1V) (+22 per cent.), Sn(I1) (+27 per cent.) and V(V) (+SO per cent.).It is difficult t o assign the cause of these interferences unequivocally, although it seems that the effect of tin(II), nickel and cobalt(I1) is possibly due to spectral interference from their nearby atomic lines. Substantial emission is in fact observed from the introduction of tin(II), nickel and cobalt(I1) solutions in the absence of beryllium. An indication of the potential of the measurement of the beryllium thermal emission, under the recommended conditions, was obtained by determination of the beryllium content of a beryllium - copper alloy.Analysis of the alloy sample by this method gave a result corresponding to 1-86 0.03 per cent. of beryllium. Independent analysis of the sample by atomic-absorption spectroscopy in a conventional 5-cm nitrous oxide - acetylene flame gave a result of 1.75 0.03 per cent. of beryllium. ATOMIC-FLUORESCENCE MEASUREMENTS- The experimental arrangement used for the fluorescence measurements is shown in Fig. 1. The beryllium microwave-discharge tube was placed immediately behind the modu- lating chopper, and the radiation was focused into the interconal zone of the flame through the entrance window and side-arm of the separator.The flame and separator were positioned so that the exit window was only about 2 mm from the monochromator entrance slit. The discharge tube was operated at 50 watts with moderate air-cooling of the exterior of the cavity. The modulated flame emission was observed in the presence of dilute beryllium solutions. As for the thermal emission, the only intense fluorescence emission observed was that from the resonance line at 234.88 nm. Thus the photomultiplier receives both the unmodulated beryllium thermal emission and the modulated fluorescence emission at 234.86 nm, but the tuned amplification system amplifies only the fluorescence emission efficiently. The effect of the acetylene flow-rate on the beryllium fluorescence emission in the separated flame is shown in Fig.4(b). The maximum emission is again found to coincide with the use of a slightly fuel-rich flame. Linear calibration graphs were obtained for the fluorescence emission in the range 0.03 to 1 p.p.m., with an optimised slit width of 300 pm (0.99 nm band width). Higher concen- trations were also examined, and linear calibration graphs were obtained between 1 and 20 p.p.m. The limit of detection, which is defined as that concentration producing a signal- to-noise ratio of 1:1, was 0.01 p.p.m. Repetitive introduction of a 0.5 p.p.m. beryllium solution into the flame under the optimal conditions gave a relative standard deviation in fluorescence of 3 per cent.August, 19681 EMISSION AND ATOMIC-FLUORESCENCE SPECTROSCOPY 527 EFFECT OF DIVERSE IONS- The effects on the intensity in atomic-fluorescence spectroscopy produced by 0.5 p.p.m.of beryllium of 100 and 1000-fold excess by weight of those ions examined for interference in thermal emission were investigated. As expected, ions that produced no interference in the thermal emission also did not affect the atomic-fluorescence signal. In addition, no interference was observed for cobalt(II), nickel and tin(I1). Interference was again observed from solutions containing very high concentrations (1000-fold excess) of the ions aluminium, calcium and molybdenum(VI), whereas the presence of a 100-fold excess of these ions caused no interference. DISCUSSION The slightly fuel-rich, separated nitrous oxide - acetylene flame provides an “atom reservoir” of high temperature and reducing properties.The reducing atmosphere serves to maintain a high population of free ground-state atoms of beryllium, which otherwise have a pronounced tendency to form a refractory (BeO) species in more conventional flames. The high flame temperature provides sufficient energy for the excitation of the beryllium resonance line at 234.86 nm, which requires 5-28 eV. High sensitivity is thus obtainable in the deter- mination of beryllium by thermal-emission spectroscopy in this relatively low background flame, and the protection afforded to the interconal zone in the separated flame results in a reducing atmosphere of extended physical dimensions. Consequently, even higher sensi- tivity should result from the use of a monochromator of greater light-gathering power than that used in this study.The selectivity obtainable in the determination by thermal spectro- scopy is high, as the hot flame suppresses the tendency for compound formation between beryllium and other compounds introduced into the flame. Because of the maintenance of a high population of beryllium atoms in the flame used, the sensitivity of the atomic-fluorescence spectroscopic determination of beryllium is also high. When an intense beryllium source is used, such as the beryllium microwave-excited, electrodeless discharge tube used in this study, the sensitivity obtainable in atomic fluores- cence, with the same instrumentation, is about one order of magnitude greater than that obtained in thermal emission. The selectivity of the determination of beryllium by atomic fluorescence, with a modulated source, is also an improvement on that of the determination by thermal emission. There is no significant overlap of d.c. atomic emission from other elements on the a.c. beryllium fluorescence and they are not excited by the beryllium source. We are grateful to Murex Ltd. (Rainham, Essex), for the provision of equipment, and for granting study leave to one of us (D. N. H.). REFERENCES 1. Gilbert, P. T., jun., “Symposium on Spectroscopy,” A.S.T.M. Special Technical Publications No. 269, 1960; Bulletin 753, 1959, Bulletin 753A, 1961, Beckman Instrument Inc., Fullerton, California. -- “Oxycyanogen Flame Photometry,” mimeographed publication, Beckman Instrument Inc., Fuilerton, California, 1958. Ramakrishna, T. V., Robinson, J. W., and West, P. W., Analytica Chim. A d a , 1967, 39, 81. Koirtyohann, S. R., and Pickett, E. E., Spectrochirn. Acta, 1968, 23B, 235. Kirkbright, G. F., Semb, A., and West, T. S., Talanta, 1968, 15, 441. 2. 3. 4. 5. Received Januavy 25th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300522

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

528 Analyst, A4ugust, 1968, Vol. 93, @. 528-531 Spectrophotometric Determination of Iron(II1) Ions with 2,4=Dihydroxybutyrophenone BY MISS M. H. GANDHI AND MAHENDRA N. DESAI (Chemistry Department, University School of Sciences, Gujarat University, A hmedabad 9) The colour reaction between iron( 111) and 2,4-dihydroxybutyrophenone has been studied to determine the optimum conditions for the analytical use of this reagent. The strongest colour developed at pH 2-9 to 3-0, and the complex exhibits an absorption maximum a t 470 mp. The composition of the complex is 1 to 1, as indicated by mole-ratio studies. The absorbance of the solution at 470mp obeys Beer's Law in the iron(II1) concentration range 1.0 to 56.0 p.p.m. and the sensitivity is 0.074 pg per ml of the iron(II1) solutions.The reagent is fairly selective for iron. ~,~-DIHYDR~XYBUTYR~PHENONE has not so far been investigated as an analytical reagent. This paper deals with a study of the reaction between iron(II1) ions and 2,4-dihydroxy- butyrophenone, and the use of the latter as a specific reagent for the spectrophotometric determination of iron, EXPERIMENTAL APPARATUS- A Unicam SP500 spectrophotometer was used, with matched l-cm silica cells, for absorbance measurements. A Beckman pH meter, Model H2, was used for all pH deter- minations. REAGENTS- 2,4-Dihydroxybutyrophenone was synthesised by condensing resorcinol (1 gram mole) with butyric acid (1 gram mole) in the presence of freshly distilled phosphorus oxychloride (2 gram moles).l The reaction was allowed to proceed at room temperature for 24 hours, and the reaction mixture then poured on to a mixture of 4OOg of crushed ice and 200ml of 50 per cent.hydrochloric acid. A yellow mass separated which, on recrystallisation from chloroform, gave colourless plates of 2,4-dihydroxybutyrophenone (m.p. 67" CJ2 yield 85 per cent .) . A solution of 2,4-dihydroxybutyrophenone in 40 per cent. ethanol was used in all of the experiments. PROPERTIES OF THE IRON(II1) - 2,4-DIHYDROXYBUTYROPHENONE CHELATE- The complex has a reddish violet colour between pH 2.0 and 3.0, which is stable, and a reddish brown colour above pH 3.0. The optimum pH range is 2.9 to 3.0, and the complex exhibits a maximum absorbance at 470 mp. The absorption spectra of the reagent, iron(II1) nitrate and the iron(II1) - 2,4-dihydroxybutyrophenone complex are given in Fig.1. The colour of the complex is formed instantaneously and is stable for 3 hours. The sequence of adding the reactants has no effect on the absorbance of the complex and Beer's law is obeyed between 1.0 and 56.0 p.p.m. of iron. 0 SAC and the authors.GANDHI AND DESAI 529 A .-*-a. I 'a / *-a a I 'a -a'* B c 5 A $ 0 400 440 480 520 560 Wavelength, m p Fig. 1. Absorption spectra of 2,4-di- hydroxybutyrophenone, iron(II1) nitrate and iron(II1) - 2,4-dihydroxybutyrophenone com- plex: graph A, 2 m l of 0.01 M iron(1II) nitrate + 20 ml of 0.01 M 2,4-dihydroxybutyrophenone; points x for graph B, 2 ml of 0.01 M iron(II1) nitrate; and points 0 for graph C, 20 ml of 0.01 M 2,4-dihydroxybutyrophenone. Final vol- ume 50 ml; pH 2-95 w c 0 13 L s n Q Moles of reagent;moles of iron Composition of the iron(II1) - 2,4-dihydroxybutyrophenone complex by the mole-ratio method a t 470 mp.p = 0.1 NaC10, (2 ml of 0.01 M iron (111) nitrate + varying amounts of 2,4-dihydroxybutyro phenone; final volume 60 ml) ; pH 2-95 Fig. 2. MOLE-RATIO STUDIES- A study of the composition of the complex by the mole-ratio method3 (Fig. 2), the method of continuous v a r i a t i ~ n ~ , ~ , ~ (Fig. 3), the slope-ratio method7 (Fig. 4) and the Bent and French method* (Fig. 5) indicates that the complex has a 1 to 1 composition. aJ c m 2 0 -0 Q In 0; Iron (Ill), ml 2 4 6 8 10 Reagent, ml 10 8 6 4 2 Fig. 3. Composition of the iron (111) - 2,4-dihydroxybutyrophenone com- plex at pH 2.95 by Job's method of continuous variation a t 470 mp. x ml of 0-01 M iron(II1) nitrate + (12 - x ) ml of 2,4-dihydroxybutyrophenone (graph A, 0.02 M reagent; graph B, 0.01 M reagent; and graph C, 0.005 M reagent) + 5 ml of M NaC10,; final volume 50 ml 0 K n L s 2 0,ool I 1 I I 1 1.0 1.2 1.4 1.6 1.8 2.0 Varying component, mI Fig.4. Composition of the iron(II1) - 2,4-dihydroxybutyrophenone complex a t pH 2.95 by the slope-ratio method a t 470mp. p = 0-1 NaC10, (graph A, 10ml of 0.01. M reagent + varying amounts of 0-002 M iron (111) nitrate; and graph B, 10 ml of 0.01 M iron (111) nitrate + varying amounts of 0.002 M reagent; final volume 50 ml)530 GANDHI AND DESAI : SPECTROPHOTOMETRIC DETERMINATION OF [ A %dy.St, VOl. 93 0 t 0 L n 2 13 m M --I 2-8 2.6 2.4 4.0 4.2 4.4 4.6 4.8 3.0 Log concei:tration Fig.5. Composition of the iron(II1) - 2,4-dihydroxybutyrophenone complex at pH 2-95 by the logarithmic method of Bent and French at 470 m p . p = 0.1 NaC10, (graph A, 5 ml of 0.01 M reagent + varying amounts of 0-002 M iron(II1) nitrate; and graph B, 5 ml of 0.01 M iron(II1) nitrate + varying amounts of 0.002 M reagent; final volume 50 ml) RECOMMENDED PROCEDURE FOR PREPARATION OF CALIBRATION CURVES- Transfer 0.25, 0.50, 1.0, 2.0, 3-0, 4.0 and 5.0 ml of a standard iron(II1) solution (0.01 M) into 50-ml calibrated flasks, add 7.5 ml of 0.1 M solution of the reagent in 40 per cent. ethanol, mix well and adjust the pH to 2.95. Make up to volume with distilled water and mix thoroughly. The series contains 2.8, 5.6, 11.2, 22.4, 33.6, 44.8 and 56.0 p.p.m.of iron, respectively, when diluted to 50ml. Measure the absorbance at 470mp after half an hour, with an iron(II1) solution of corresponding concentration as blank. Plot the absorbance as a function of iron concentration. SENSITIVITY AND OPTIMUM CONCENTRATION RANGE- iron(II1) ion in 107 parts of the solution. in the absorbance range 0.2 to 0.7. 7.4 to 26-1 p.p.m. REPRODUCIBILITY- Six aliquots of 2 ml, corresponding to 11.2 p.p.m. of iron at the final dilution, were analysed by the recommended procedure. The standard deviation was & 0.0025 unit of absorbance, i.e., 1 per cent. EFFECT OF DIVERSE IONS- The effect of 37 ions was investigated. At an iron concentration of 11.2 p.p.m., 1000 p.p.m. concentrations of AP+, Ba2+, Be2+, Cd2+, Cs+, Ca2+, Fe2+, Mg2+, Mn2+, Ni2+, K+, Sr2+, Th4+, U022+, Zn2+, Zfl+, NH,+, C1-, ClO,-, NO,- and 750 p.p.m.of Br- and Co2+, 500 p.p.m. of CP+, Pb2+ and Ag+, 150 p,p.m. of I-, 60 p.p.m. of Pd2+, 25 p.p.rn. of V4+, 15 p.p.m. of Cu2+ and 5 p.p.m. of molybdate and tellurate could be tolerated; PO:-, F-, citrates, oxalates and vanadates interfered at all levels. APPLICATIONS- were analysed for iron according to the recommended procedure. in Table I. The sensitivity of the reaction is 0.074 bg per ml of Fe3+ at 470 mp, i.e., 0.74 part of The spectrophotometric errors is at a minimum This corresponds to an optimum concentration of Eight samples were prepared containing known concentrations of diverse ions. These The results are shownAugust , 19681 IRON(II1) IONS WITH 2,4-DIHYDROXYBUTYROPHENONE 531 An indigenous iron ore, with the following percentage composition: Fe,O, 92.8, A120, 1.54, SiO, 3.07, CaO 0.3, Ti02 0.01, MgO 0.28, P 0.10 and S 0.03, was also analysed.A sample of the ore (0.4752 g) was brought into solution by the standard method and diluted to 1 litre. Two 2-ml aliquots of the solution were then analysed according to the proposed procedure. The absorbance values were 0.270 and 0.275, which corresponded to 12.22 and 12.45 p.p.m. of iron, respectively. The results are in good agreement with the calculated amount of iron, which is 12.24 p.p.m. TABLE I ANALYSIS OF SYNTHETIC SOLUTIONS Sample, p.p.m. Absorbance Fe3+ Fea+ 11.2 + Sra+ 1000 + Mn2+ 1000 + Ca2+ 1000 . . . . . . 0.250 Fe3+ 11.2 + Sra+ 1000 + Pb2+ 250 + Ca2+ 1000 . . . . 0.250 Fe3+ 11-2 + Ca2+ 1000 + Znz+ 1000 + K+ 1000 + AP+ 1000 0.245 Fe3+ 11.2 + Cu2+ 10 + Nia+ 250 .. .. . . . . . . 0.249 Fe3+ 11.2 + Cr3+ 250 + Nia+ 250 . . . . . . . . . . 0.252 Fe3+ 11.2 + Cr3+ 250 + Nia+ 250 + Na+ 1000 . . .. . . 0.252 Fe3+ 11.2 + Na+1000 + K+ 1000 + Baa+ 1000 . . 0.249 Fe3+ 11.2 + CUB+ 6 + CP+ 100 + zn2+ 100 + caz+’ioo +‘Sra+ 100 . . + AP+ 100 + Cd2+ 100 . . . . . . . . 0.24 found, p.p.m. 11.32 11.32 11 -09 11.27 11-41 11.41 11.27 10.86 DISCUSSION The proposed method is not as sensitive as some methods for the spectrophotometric determination of iron, but its optimum range of 7-4 to 26.1 p.p.m. is a distinct advantage and should be useful for the determination of large amounts of iron.1° It is reasonable to expect that the concentration-range limit could be extended significantly by the use of longer cells.This is especially promising, as the reagent does not absorb light in the visible region. The tolerance of diverse ions shown suggests that 2,4-dihydroxybutyrophenone can be used as a specific reagent for iron(II1) ions at a relatively high concentration level in the presence of large amounts of other ions. According to Sandel1,ll one great advantage of o-phenanthroline over some other reagents for iron is that it can be used in slightly acidic solutions, so that metal hydroxides are not precipitated. As the optimum pH for the 2,4-dihydroxybutyrophenone reagent is 2.9 to 3.0, this requirement is met. Moreover, the absorbance values of the complex formed are almost constant in the spectral range 460 to 480 mp, thus giving rise to a 20-mp wide spectrophotometric absorption plateau, which enables this reagent to be adapted for colour measurements with filter photometers. We thank Professor B. K. Vaidya for his kind interest, and the Gujarat University for the award of a research scholarship to one of us (M.H.G.). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Shah, R. C., and Bose, J . L., Indian Patent 57,68!,, October 29th, 1958. Heilbron, I. M., “Dictionary of Organic Compounds, Volume 2, Eyre and Spottiswoode, London, Yoe, J. H., and Jones, A. L., Ind. Engng Chem. Analyt. Edn, 1944, 16, 111. Job, P., Annls Chim. Phys., 1928, [10]9, 113. -, C. R. Hebd. Shanc. Acad. Sci., 1925, 180, 928. Harvey, A. E., and Manning, D. L., J. Amer. Chem. Soc., 1950, 72, 4488. Bent, H. E., and French, C. L., Ibid., 1941, 63, 568. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience Watts, H. L., Analyt. Chem., 1964, 36, 365. Sandell, E. B., op. czt., p. 108. 1965, p. 1061. -, Ibid., 1936, [11]6, 97. Publishers, New York and London, 1959, pp. 83 and 97. First received June 26th, 1967 Amended Febvuary 20th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300528

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

532 Analyst, August, 1968, Vol. 93, $9. 532-534 The Determination of Iron(I1) Sulphide in Soil in the Presence of Iron(II1) Oxide BY G. PRUDEN AND C. BLOOMFIELD (Rothamsted Experimental Station, Harpenden, Herts.) Hydrogen sulphide is partially oxidised by iron(II1) when soils containing sulphide and iron(II1) oxide are acidified. Satisfactory recoveries of hydrogen sulphide are obtained by using a solution of tin(I1) chloride in hydrochloric acid to decompose the sulphide. IN studying the reduction of sulphate in waterlogged soil it was necessary to determine the amount of iron(I1) sulphide that was produced as one of the reaction products. Attempts to determine this as the iron(I1) iron liberated when the soil was treated with a solution of a mercury(I1) salt were unsuccessful. The difficulties of this method are aggravated by the need to correct for soluble iron compounds, also formed during anaerobic incubation, and the determination of this quantity is uncertain, particularly under the conditions imposed by the requirements of the main experiment.The only alternative seemed to be to determine iron(I1) sulphide as hydrogen sulphide liberated on acidification. As iron(II1) oxide would also be present in the reaction mixture, we determined the extent to which this would interfere, according to the reaction 2Fe3+ + S2- = 2Fe2+ + S, with calcium sulphide as the source of sulphide. The apparatus used is illustrated in Fig. 1. Fig. 1. Details of apparatus 0 SAC and the authors.PRUDEN AND BLOOMFIELD 533 METHOD REAGENTS- Ammoniacal zinc sulphate solution-Dissolve 50 g of zinc sulphate, ZnSO,.?H,O, in 250ml of water.Add 250ml of concentrated ammonia solution, allow to stand overnight in a stoppered container and filter. Potassium iodate solution, 0.1 N. Potassium iodide solution, 1 per cent. w/v, apueous-Freshly prepared. Hydrochloric acid (1 + 1 v / v ) . PROCEDURE- About 0.1 g of calcium sulphide and various amounts of hydrated iron(1I-I) oxide were weighed into the conical flask (Fig. 1) and the apparatus flushed with nitrogen. About 25 ml of dilute hydrochloric acid were added from the dropping funnel, and the liberated hydrogen sulphide swept out with a gentle stream of nitrogen. When the initial vigorous effervescence had ceased, the contents of the reaction flask were heated to boiling.The hydrogen sulphide evolved was absorbed in about 5 ml of the ammoniacal zinc sulphate solution diluted with 25 to 30 ml of water. When all of the hydrogen sulphide was expelled, 2 to 3 ml of potassium iodide were added to the receiver, followed by starch solution and by dilute hydrochloric acid to dissolve the precipitate of zinc sulphide ; the hydrogen sulphide thus liberated was titrated immediately with 0.1 N potassium iodate to the first appearance of a permanent blue colour. TABLE I THE EFFECT OF IRON(III) OXIDE ON THE RECOVERY OF HYDROGEN SULPHIDE Sulphide taken, mg 30.8 28-3 29-7 28.8 28-5 28-8 29.7 28.7 28.8 Iron(II1) oxide added, mg 20 60 140 200 400 900 50* 1 oo* 500* Sulphide recovered, mg 30.6 26-9 22.1 17.5 14.4 13.9 24.9 21.7 15-7 Recovery, per cent.99.4 95.1 74.4 60-8 50.5 48-3 88-2 76-9 55.6 * Added as subsoil. The iron(II1) oxide used to obtain the results given in Table I was a synthetic goethite that dissolved readily in dilute acid, and it might therefore have caused a bigger proportional error than the less readily soluble iron(II1) oxide of a soil. However, when subsoil from the Rothamsted farm was substituted for the laboratory preparation, the loss of sulphide was not appreciably smaller. The error caused by iron(II1) oxide is thus considerable; the failure of Chaudhry and Cornfield1 to recover more than 60 per cent. of sulphide added to air-dried soil can presumably be ascribed to this effect. Because of the lack of any method for determining the iron(I1)-to-iron(II1) ratio in a reduced soil it is impossible to generalise, but the few determinations we have made of the sulphide contents of field samples suggest that the largest iron( 111) oxide-to-sulphide ratios in Table I are probably those most relevant to actual soils.As the concentration of iron(II1) oxide decreases during the anaerobic incubation of a soil, the extent of the error would vary throughout our experiments, and the results obtained by this method would be quite meaningless. Attempts to limit the interference by adding metallic zinc or tin to the sulphide before acidification failed, but by adding tin(I1) chloride to the acid used to decompose the sulphide (about 2 g per 25 ml), the loss of sulphide was largely overcome (Table 11). When water was added to the sulphide - iron(II1) oxide mixture, satisfactory recoveries were obtained when the concentration of the acid used to decompose the sample was increased to give a final value of about 5 N.534 PRUDEN AND BLOOMFIELD TABLE I1 THE RECOVERY OF HYDROGEN SULPHIDE IN THE PRESENCE OF IRON(III) OXIDE WITH TIN(II) CHLORIDE - HYDROCHLORIC ACID Sulphide taken, Iron(II1) oxide taken, Sulphide recovered, Recovery, mg mg m g per cent.30.6 45 30.4 99-5 29.2 100 28.9 99.0 29.2 200 29.2 100.0 29.4 900 28.6 97.5 28.7 60* 28.4 99.0 The effect of the tin(I1) chloride reagent on other sulphur compounds present, or likely to be present, in the incubated reaction mixture was tested. Thiosulphate was reduced quantitatively to sulphide, but the effect on sulphate and elemental sulphur was negligible. Pyrite reacts slowly, thus causing a positive error. It is unlikely that pyrite is formed in our experiments, but, as it is often the chief sulphur compound in sulphidic soils, it could cause serious error when dealing with a natural soil. In testing sulphate, sulphur and pyrite, the hydrogen sulphide was absorbed in sodium hydroxide solution and titrated with 0*001 N mercury(I1) acetate, with dithizone as indicator.2 The tin(I1) chloride reagent has been used for several years in our work on sulphate reduction. The iodate method described above is unsuitable for the amounts with which we are concerned; instead, hydrogen sulphide is absorbed in 25 ml of 0.05 N hypochlorite solution (0.4 N with respect to sodium hydroxide), and the excess of hypochlorite determined iodimetrically . REFERENCES * Added as subsoil. 1. 2. Chaudhry, I. A., and Cornfield, A. H., PI. Soil, 1966, 25, 474. Archer, E. E., Analyst, 1956, 81, 181. Received November 23rd, 1967

ISSN:0003-2654    

DOI:10.1039/AN9689300532

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

AutaZyst, August, 1968, Vol. 93, &5. 535-539 535 A Titrimetric Method for Determining Total Sulphur in Mineral Soils BY D. S. JENKINSON (Pedology Department, Rothamsted Experimental Station, Harpenden, Hevts.) A rapid method for the determination of total sulphur in mineral soils is described. The various sulphur compounds in soil are first oxidised to sulphate by potassium dichromate in phosphoric acid, and the sulphate thus formed is then reduced to sulphur dioxide by heating with activated charcoal. The evolved sulphur dioxide is trapped in hydrogen peroxide, barium per- chlorate added to the resulting sulphuric acid to precipitate barium sulphate, and the excess of barium determined by titration against standard potassium sulphate, with sulphonazo I11 as indicator. The proposed method gave acceptable results when tested on pure compounds and on soils of known sulphur content.No interferences are to be expected from normal soil con- stituents. WHEN soils containing charcoal are strongly heated with phosphoric acid, sulphur dioxide is evolved from any sulphates that may be present. A new method, based on this observation, was developed for determining total sulphur in soil that compares favourably, both in speed and simplicity, with other current chemical techniques for this determinati0n.l 12 s3 Inorganic sulphur is often present in soil as sulphate, held by anion exchange, but can be found, particularly in soils from arid regions, as calcium and magnesium sulphates ; it occurs as sulphides in certain soils developed under waterlogged conditions.Some of the organic sulphur in soil is combined in sulphur-containing a-amino-acids, and some as ester sulphate, but the chemical nature of much of the soil organic sulphur is obscure; for a review see Freney.* In all chemical methods for determining total sulphur in soil, the different sulphur compounds are first quantitatively converted into a single form; the sulphur in this is then isolated from the soil residue for determination. Analytical methods for determining total sulphur in soil have been briefly discussed by Bloomfield2 and by Bardsley and Lancaster.1 The proposed method is in three steps : organic and inorganic soil sulphur are oxidised to sulphate by refluxing with potassium dichromate and dehydrated phosphoric acid (oxidation step); the sulphate thus formed is reduced to sulphur dioxide by heating with activated charcoal in phosphoric acid (reduction step); the evolved sulphur dioxide is oxidised to sul- phuric acid by hydrogen peroxide, barium perchlorate added to precipitate barium sulphate, and the excess of barium determined by titration against standard potassium sulphate solution , with sulphonazo I11 as indicator.5 Back-titration is used rather than direct titration because of the erratic colour changes that sometimes occurred at the beginning of direct titrations, caused, most probably, by the absence of barium sulphate precipitation nuclei. Preliminary work showed that it was essential to pre-treat the phosphoric acid by heating to 370" C with activated charcoal if low blank values were to be obtained.During this pre- treatment the orthophosphoric acid is dehydrated, giving a mixture of ortho, pyro and meta- phosphoric acids. The charcoal and the dehydrated acid are then both used as reagents in the determination. Use of the dehydrated acid has an incidental advantage in that excess of dichromate remaining after oxidation of the organic matter during the first step is decomposed at the temperature of reflux (about 305" C). If, for example, 88 per cent. orthophosphoric acid is used in the oxidation step, the reflux temperature is lower (about 170" C), excess of dichromate remains in the reaction mixture and then reacts violently when heated with charcoal during the reduction step. 0 SAC and the author.536 JENKINSON A TITRIMETRIC METHOD FOR [AaaZyst, Vol.93 METHOD APPARATUS- This is shown in Fig. 1. The gas stream carrying sulphur dioxide from the 150-ml flask, A, is sucked through traps, B and C , each containing a sintered-glass thimble (porosity No. 2, diameter 14mm) to stop acid spray, before going through a water condenser and into the absorption trap, D. The absorption trap also contains a sintered-glass thimble of the same size and porosity as those in traps B and C. The sintered-glass thimble in the absorption trap, D, must fit snugly into the dimple in the absorption flask, almost touching the bottom. Construction of the absorption flask is simplified if, as in Fig. 1, a screw-thread adaptor with 1 Sulphur dioxide Scale 3 0 I5 cm I " ' I I I I 7-- Jf - B19 C 4 BI vacuum line needle valve A = 150-ml reduction flask 6 and C = Traps containing a sintered- glass thimble (porosity No. 2, diameter 14 mm) D = Absorption trap S = Stopper Fig.1. Details of the apparatus a PTFE washer (Quickfit and Quartz Ltd.) is used to connect the tube carrying the sinter to the B19 cone. Connections are as short as possible and made with poly(viny1 chloride) tubing. The incoming air is passed through a U-tube containing self-indicating soda limeAugust, 19681 DETERMINING TOTAL SULPHUR I N MINERAL SOILS 537 to remove oxides of sulphur. To facilitate control of the gas flow, a length of capillary tubing (1 cm long, 0-25-mm i.d.) is inserted between the U-tube containing soda lime and flask A. Except for the 50-ml absorption flask (D, Fig. l), all the glassware is available commercially. REAGENTS- Phosphoric acid-Heat 750 ml of 88 per cent.analytical-reagent grade orthophosphoric acid with 15g of activated charcoal (Merck's Pro Analysi grade) in a distillation flask connected to a water condenser. Collect 200ml of the distillate, then allow the suspension to cool to about 150" C, and filter under suction through a No. 3 glass sinter, discarding the first few millilitres of filtrate. Traces of suspended charcoal in the filtered phosphoric acid can be disregarded. Any crystals that appear on standing can be re-dissolved by keeping the acid overnight in an oven at 80" C. A burette, 3 cm in diameter with a tap and "jet" of 6-mm id., is convenient for dispensing the viscous acid. Active charcoad-Wash the charcoal remaining on the sinter with three 250-ml portions of boiling distilled water, making sure that the charcoal is well dispersed in each wash, and dry overnight at 105°C.Barium perclzlorate-A 0.03 N solution of barium perchlorate trihydrate (general-purpose reagent grade) in water. Standard sulphate solution-Dissolve 0.271 7 g of analytical-reagent grade potassium sulphate, dried overnight at 105" C, in water and make up to 1 litre; 1 ml of this solution contains 0.05 mg of sulphur. Iutdicator-A 0.1 per cent. solution of sulphonazo I11 in water. Potassium dichromate-Analytical-reagent grade. Hydrogen peroxide, 100 voZ-Microanalytical-reagent grade. A cetone-Analytical-reagent grade. PROCEDURE- Weigh 1 g (less if the sulphur content exceeds 500 p.p.m.) of air-dried soil (ground to pass a 30-mesh sieve or finer) into the dry reaction flask.Moisten with 1 ml of water, add 2 g of potassium dichromate and add 25 ml of phosphoric acid. Connect the flask to a water- cooled reflux condenser, bring the solution to the boil, swirl the flask, and heat it under steady reflux for 10 minutes. Allow to cool, add 0.5 g of activated charcoal and 4 ml of water and connect the flask to the apparatus as shown in Fig. 1. Place 0.2 ml of hydrogen peroxide in the absorption trap, D. Adjust the stream of air passing through the apparatus to give a steady rapid flow of bubbles in the absorption trap. Heat the reduction flask on a bunsen burner: if foaming occurs when the mixture begins to boil, momentarily remove the flame. After the mixture has come to the boil, heat for 5 minutes.The rate of heating should be such that steam reaches the top of the water condenser between 2 i and 34 minutes after coming to the boil. Two minutes after the burner is turned off, remove stopper, S, then disconnect the absorption trap and wash the sinter inside and out with 15ml of acetone, adding the washings to the absorption flask. Add 3 drops of indicator and run in 1 ml of barium perchlorate by using a Van Slyke pipette. If the solution remains purple add a further 1 ml of perchlorate. Titrate against the standard potassium sulphate solution, with a 10-ml burette. Carry out a blank determination, omitting the soil, with the same volume of barium perchlorate solution. Sulphur, mg = (ml of potassium sulphate solution required by blank - ml of potassium Between determinations, empty the water collected in traps B and C.If the reduction mixture is allowed to cool in the flask it solidifies: to avoid this pour it out while still liquid and then wash the flask with hot water. RESULTS AND DISCUSSION The colour change is from blue to purple. sulphate solution required by sample) x 0.05. Table I gives the sulphur contents of some organic and inorganic compounds as found by the proposed method, together with the calculated values. For analysis, 2 g of potassium dichromate and 25 ml of phosphoric acid were added to 1 ml of an aqueous solution of the compound containing about 0.4mg of sulphur. The rest of the determination was as described above. In Table I and subsequent tables results are means of triplicate determinations unless otherwise indicated.Results were quantitative, within +2 per cent.538 JENKINSON : A TITRIMETRIC METHOD FOR [Arcalyst, VOl. 93 TABLE I SULPHUR CONTENTS OF SOME ORGANIC AND INORGANIC COMPOUNDS Compound* Sulphamic acid? . . .. .. S-Benzylthiuronium chloride? . . Sulphanilic acid'; . . .. .. Methionine . . .. .. .. Cystine . . .. .. .. Copper(I1) sulphate pentahydratel Potassium sulphate: . . .. .. Sulphur calculated, per cent. 33-0 16.8 18-5 21.6 26.7 12-8 18-4 Sulphur by proposed method, per cent. 32.9 16.7 18.7 21-3 26.7 13-0 18.1 * All dried over anhydrous magnesium perchlorate. t Microanalytical-reagent grade. 1 Analytical-reagent grade. TABLE I1 SULPHUR CONTENTS OF SOME ANALYSED Sulphur nominal, Material* per cent.Basicslag? . . .. . . .. 0.16 Anthracite? . . .. . . . . 2-33 Iron pyrites . . .. .. .. 44.6 Shale . . .. .. . . .. ' 5.07 Portland cement? . . .. .. 1.21 MATERIALS Sulphur by proposed method, per cent. 0.197 2.25 354 (44.0): 1.23 5.24 * All results (except pyrites) on oven-dried (24 hours at 80" C) basis. t Obtained from the Bureau of Analysed Samples Ltd. : Result in parenthesis is after hydrogen peroxide treatment. Table I1 gives the sulphur contents of five analysed materials. The nominal sulphur contents of the basic slag, anthracite and Portland cement were those given by the Bureau of Analysed Samples Ltd. ; those given for pyrites and shale were obtained by the gravimetric procedure described by Vogel.g Acceptable agreement was obtained, except with pyrites. Unchanged pyrites crystals were found in the reaction mixture at the end of the dichromate oxidation stage.The more coarsely ground the pyrites, the less was oxidised. Hydrogen peroxide in the cold will oxidise pyrites.' More sulphur was recovered (see Table 11) from pyrites if, instead of moistening with 1 ml of water before oxidation, 1 ml of 100-vol hydrogen peroxide was added, the mixture left to stand overnight, and the analysis then continued as above. This peroxide treatment is probably only necessary with soils containing much coarsely crystalline pyrites. TABLE I11 SULPHUR CONTENTS OF SOME ROTHAMSTED SOILS soil No.+ 1 2 3 4 6 6 sampling depth. inches 0 to 6 0 to 6 0 to 9 9 to 18 18 to 27 0 to 9 organic carbon, pH percent. '7.8 2.43 8.1 0.97 7.8 2.89 7.9 0.8 1 7.7 0.66 4-6 1.98 Carbonate carbon, per cent.0.14 0.16 0.19 0.04 0 0 Nitrogen, per cent. 0.246 0-107 0.261 0.098 0.076 0.168 clay' per cent. 18 18 24 50 62 20 * All results on oven-dried (24 hours at 80' C) basis. Sulphur by proposed method, p.p.m. 611 200 419 167 108 326 Sulphur by Bloomfield's method, p.p.m. 492 187 41 1 149 116 330 Table I11 gives the sulphur contents of a range of Rothamsted soils as determined by Bloomfield's method2 and by the proposed method. Agreement was close. Table IV gives the sulphur contents of some Australian soils as found by the proposed method and that of Steinbergs, Iismaa, Freney and Barrow ; with these, discrepancies between the two methods were slightly greater than with the Rothamsted soils, but neither method gave values consistently higher than those by the other.August, 19681 DETERMINING TOTAL SULPHUR I N MINERAL SOILS 539 INTERFERENCES- The addition of 1000 p.p.m.of Fe 3+ , Pb2+, AP+, CUB+, Co2+, Mn2+, Sn2+, Sr 2+, Zn2+,Mg2+, Ba2+, NH,+, K+, Na+, Se0,2-, F-, C1-, Br-, B,0,2-, NO,- or NO2- tosoil 2 (Table 111) did not significantly alter the value found by the proposed method for the sulphur content of this soil (200 p.p.m.). Of the cations and anions tested, iodide alone interfered; in the presence of 1000 p.p.m. of iodide (as potassium iodide) the sulphur content of soil 2 was found to be 171 p.p.m. PRECISION- minations. method is precise enough for most soil work. Iodine concentrations as large as this are most unlikely in soils. Table IV gives the sulphur contents of six soils; each value is the mean of four deter- The standard errors of these means range from +2 to +_9 p.p.m., so that the TABLE IV SULPHUR CONTENTS OF SOME AUSTRALIAN SOILS soil No.* 1 2 3 4 6 6 Sulphur by Sampling Organic proposed depth, carbon, Nitrogen, method, Type inches pH per cent.per cent. p.p.m. Red - brown earth 0 to 4 6.9 0.66 0.043 68 f 3.2t Red-brown earth 0 to 4 6.8 2.23 0.136 168 f- 3.1 Red - brown earth 0 to 4 6.9 2-70 0.162 209 f 2.1 Podsol 0 to 2 5.6 3-89 0-306 326 f 1.9 Alpine humus Oto 4 4.9 7.60 0.409 442 f 3.9 Kraznozem 0 to 2 6.1 7.71 0.676 964 f 9.4 * All results on oven-dried (24 hours at 80” C) basis. 1 Standard error of mean (four replicates). CONCLUSIONS Sulphur by S teinbergs Iismaa, Freney and Barrow’s method,* p.p.m.67 167 199 343 473 928 The proposed method is rapid and may be of value when the number of determinations to be carried out does not warrant more complex apparatus. It is convenient to perform the oxidation step in batches of six and to duplicate the apparatus in Fig. 1, so that two reductions can be done at the same time. Six determinations then take 19 hours. The sensitivity is adequate for the range of sulphur contents common in mineral soils. If needed, the sensitivity could probably be increased by abandoning the titrimetric finish, collecting the sulphur dioxide in sodium tetrachloromercurate instead of hydrogen peroxide, and deter- mining the sulphur dioxide colorimetrically by the rosaniline method.* A disadvantage is that the hot phosphoric acid attacks the glass reaction flask, A : flask life is limited to twenty to fifty determinations.BudESinskjr and Krumlov6 have recently shown that dimethyl sulphonazo I11 is a slightly better indicator than sulphonazo I11 for the titration of sulphate with barium perchlorate.$ This reagent could probably be used as an alternative to sulphonazo I11 in the proposed method. I thank Dr. C. H. Williams for supplying the Australian soils, together with the results in Table IV characterising them, and also British Drug Houses Ltd. and Dr. B. BudZ:Sinsk$ for samples of sulphonazo 111. REFERENCES 1. Bardsley, C. E., and Lancaster, J. D., in Black, C . A., Evans, D. D., White, J. L., Ensminger, L. E., and Clark, F. E., Editors, “Methods of Soil Analysis,” American Society of Agronomy Inc., Madison, 1965, Part 2, p. 1102. 2. Bloomfield, C., Analyst, 1962, 87, 686. 3. Steinbergs, A., Iismaa, O., Freney, J. R., and Barrow, N. J., Analytica Chim. Acta, 1962,27, 168. 4. Freney, J. R., in McLaren, A. D., and Peterson, G. H., Editors, “Soil Biochemistry,” Marcel 5. BudEhskq, B., Analyt. Chem., 1966, 37, 1159. 6. Vogel, A. I., “A Textbook of Quantitative Inorganic Analysis, Theory and Practice,’’ Second 7. Gad, M. H., and Le Riche, H. H., Geochim. Cosmochim. Acta, 1966, 30, 841. 8. West, P. W., and Gaeke, G. C., Analyt. Chem., 1966, 28, 1816. 9. BudE!Sinskf, B., and Krumlovb, L., Analytica Chim. Acta, 1967, 39, 376. Dekker Inc., New York. 1967, p. 229. Edition, Longmans, Green & Co., London, New York and Toronto, 1951, p. 406. Received January lst, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300535

出版商:RSC    

年代:1968

数据来源: RSC