ISSN:0003-2654    

DOI:10.1039/AN97196FX001

出版商:RSC    

年代:1971

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97196BX003

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

xx SUMMARIES OF PAPERS I N THIS ISSUEThe Determination of 33- Dinitrosalicyl- (5-nitrofurfury1idene)-hydrazide (Nifursol) in Animal Feeding Stuffs by Electron- captureGas ChromatographyA method is described for the determination of 3,5-dinitrosalicyl-(5-nitro-furfury1idene)hydrazide (nifursol) in animal feeding stuffs. The additive isextracted from the feed with acetonitrile, and after further clean-up withcarbon disulphide the nifursol is reacted with boron trifluoride - methanolcomplex. The product, methyl 3,5-dinitrosalicylate, is determined by gaschromatography with an electron-capture detector.B. B. WHEALS and R. E. WES'I'ONDepartment of Trade and Industry, Laboratory of the Government Chemist, CornwallHouse, Stamford Street, London, S.E.1.Analyst, 1971, 96, 78-80.The Determination of Nicotine in Human Blood byGas - Liquid ChromatographyA method has been developed for the extraction and determination ofsub-microgram amounts of nicotine in blood. I t involves steam distillationof the nicotine followed by solvent partition and column-chromatographicclean-up. The final solution of nicotine in ethanol is injected directly onto the column of a gas - liquid chromatograph fitted with a flame-ionisationdetector. The method can be used to determine down to 1 ng of nicotine,and has been applied to the measurement of nicotine levels found in theblood of smokers while smoking cigarettes.I. E. BURROWS, P. J. CORP, G. C. JACKSON and B. F. J. PAGEDepartment of Chemistry, Huntingdon Research Centre, Huntingdon.Analyst, 1971, 96, 81-84.Amperometric Method for the Determination of ProphamAn amperornetric diazotisation titration for the determination of activematerial in technical propham has been investigated. Advantages in com-parison with the official CIPAC method are the smaller amount of samplerequired, the reduced time of hydrolysis and analysis, automation of titration,more accurate evaluation and better reproducibility because no extractionis necessary.The standard deviation was 0.52 compared with 1.04 obtainedwith the CIPAC method. In addition, the method has been tested successfullywith other carbamates and their formulations.GENO KYNAST and HANS HAHNSchering AG, Zentrale Analytik, Werk Wolfenbuttel, 334 Wolfenbuttel, HalchterscheStrasse 33, Germany.Analyst, 1971, 96, 85-87.An Enzymic Method for the Determination of Skimmed MilkPowder in Raw SausagesAn enzymic method for the determination of skimmed milk powder inraw sausages is described. The method is based upon the estimation of freelactose by its hydrolysis with p-galactosidase to galactose and glucose, thelatter being determined by the hexokinase method. The determination isfree from interference by reducing sugars and other substances present insausage ingredients. The method is more rapid, accurate and reliable thanother methods currently in use.R. K. BAHLJ. Sainsbury Ltd., Stamford House, Stamford Street, London, S.E. 1.Analyst, 1971, 96, 88-92

ISSN:0003-2654    

DOI:10.1039/AN97196BP011

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

26 Artalyst, January, 1971, Vol. 96, pp. 26-36 Some Observations on Oxidation - Reduction Indicators of the Benzidine, Naphthidine and Diar ylamine BY E. BISHOP AND MRS. L. G. HARTSHORN (Chemistry Department, University of Exeter, Stocker Road, Exeter, Devon) A spectrophotometric investigation has been made of representative compounds of the benzidine class of indicators in an attempt to resolve fundamental problems concerning their oxidation mechanism in sulphuric acid media. In all instances the oxidation to the coloured compound is a single-step two-electron process. There is no detectable evidence for the formation of an intermediate, either of a benzidine from arylamines or of a semiquinone from a benzidine or naphthidine. Reduction or spontaneous decay of oxidised arylamines stops a t the benzidine stage.All oxidised indicators are unstable both intrinsically and in the presence of excess of oxidant : unsubstituted benzidine and naphthidine are, additionally, photo- sensitive in the oxidised state. The spontaneous decomposition of the oxidised form regenerates the reduced form and is a disproportionation, probably of 2: 1 stoicheiometry, but a kinetic study shows that the rate-controlling step is a unimolecular precursive reaction, which may be the relaxation of the triplet state of the coloured dication diradical oxidised form. Stability decreases on sulphonation and with increasing temperature, and increases with increasing sulphuric acid concentration and with progressive substitution of the amino hydrogen atoms. FOR many reasons, such as sparing solubility, instability of oxidised forms, side and induced reactions, low exchange currents, photosensitisation and the need with certain oxidants of providing an inductor, potentiometric measurements on di- and triarylamines and alkyldiaryl- amines (hereafter collectively termed arylamines) , benzidine, naphthidine and their derivatives are attended by severe difficulties. Interpretation of such results as can be obtained is often uncertain and equivocal.In few instances is it known beyond doubt how many electrons are involved in the oxidation of the indicator, and little beyond the fact of its occurrence is known about the decomposition or further oxidation of the oxidised form of the indicator. None of the indicators is sufficiently stable in the oxidised form to pennit accurate deter- mination of its formal potential by direct potentiometric titration, and few allow approximate estimation of either formal potential or number of electrons by this method.Instead, recourse must be made to the method of conducting a potentiometric titration of, say, iron(I1) with dichromate or cerium(1V) in the presence of a substantial amount of indicator that involves passing through the end-point region with a 10 or 100-fold diluted titrant while recording potentials and visual observations of the indicator colour. Rapid titration to avoid error from decomposition of the indicator produces inaccurate potentials ; slow titration to achieve accurate potential measurements permits significant decomposition of the indicator. Conse- quently it is more usual to report “transition potentials,” i.e., the potential at which a discernible colour first appears (all indicators of these classes are colourless in the reduced form), and “transition ranges,” i.e., the potential range from the first appearance of colour to its full development. Potentials and ranges are both dependent on experimental conditions, and are valid when the indicator is used under the same conditions.But the ranges do not unequivocally define the number of electrons involved in the indicator reaction, either by the slope of the potentiometric curve and the magnitude of the range, or by the amount of oxidant consumed. Manual spectrophotometric examination of a decomposing species is obviously difficult, and only when the decomposition is slow can reasonably accurate values of wavelengths of maximum absorption (Amax.) and of molar absorptivities ( E ) be obtained.Mechanistic 0 SAC and the authors.BISHOP AND HARTSHORN 27 investigations have been few, and are empirical and largely speculative. It is much easier to rationalise the oxidation of benzidines on a basis of a one-electron mechanism than on a basis of a two-electron mechanism. The little unequivocal evidence available supports a two-electron mechanism ; the remainder favours a one-electron mechanism. Kolthoff and Sarverl postulated that arylamines were first oxidised by an irreversible two-electron bi- molecular process to the colourless diarylbenzidine, and that this benzidine was then reversibly oxidised by a two-electron unimolecular reaction to a diquinonediimine.The blue or violet oxidised form is unstable and is destroyed by an irreversible process. Further, frequent reports suggest that an intermediate product occurs in the oxidation of the benzidine, and this is described as a molecular complex of the benzidine and the diquinonediimine. For the unsubstituted parent compound, and ignoring protonation- dipheny lamine diphenylbenzidine diquinonediimine w+-e intermediate semiquinone I? destruction This theory has since been tacitly accepted as applying generally to all arylamines and benzidines and, by analogy, to naphthidines. However, the intermediate diphenylbenzidine has not been isolated and characterised, except by reduction of the purple oxidation product, nor have the intermediate semiquinone and the destruction product been examined.Further, the oxidation product formulated as a diquinonediimine could not be expected to have an intense blue colour. It has been reported, but in only one instance, that the original reduced form of the indicator is regenerated during the decomposition reaction,2 thus suggesting a disproportionation. Many questions, even one so fundamental as how many electrons are involved in the oxidations, remain to be answered, therefore, before the behaviour of these indicators can be rationalised. This brief introduction could be annotated by 200 or more references to the literature, and be extended to include a detailed argument of the possibilities, but this has been done in a recent monograph3 and need not be repeated here. Earlier work was carried out without the benefit of modern instrumentation, and it was thought that fast scanning spectrophotometry might be used to define the number of electrons involved in the oxidations and to give some information on the formation of intermediates and on the decomposition reactions.As it seemed likely that free radicals would be involved, electron spin resonance spectrometry might also yield some useful information. EXPERIMENTAL REAGENTS- Indicators-These were purified by crystallisation from water or dilute sulphuric acid, or by precipitation from concentrated sulphuric acid by dilution and cooling, under a carbon dioxide atmosphere.2 They were examined for impurities by appropriate thin-layer chromato- graphic methods.Commercial diphenylbenzidine was found to contain about mole per cent. of dichromate, otherwise the compounds appeared to be pure individual authentic species : commercial samples contained traces of oxidised species. Determinate stock solutions, usually M, were prepared in water or concentrated sulphuric acid and diluted as required. SuZPhuric acid-Aristar or AnalaR grades, free from traces of oxidants, were used concentrated, or after dilution with distilled water. Concentrations were determined by precise density measurements.28 BISHOP AND HARTSHORN : SOME OBSERVATIONS ON OXIDATION - REDUCTION [Analyst, Vol. 96 Cerium(1V) solutions-A stock 0.1 M solution of cerium(1V) in M sulphuric acid was prepared from AnalaR ammonium hexanitratocerate(1V) by the usual method: and stan- dardised against freshly prepared primary standard 0.05 M arsenic(II1) by using osmic acid as the catalyst and tris(1,lO-phenanthroline)iron(II) as the indicator.More dilute solutions were prepared by diluting the stock solution with 1.0 M sulphuric acid. Dichromate solutions-Stock 0.01667 M aqueous solutions were prepared determinately from AnalaR potassium dichromate and diluted with water as required. INSTRUMENTS- Spectral scanning was carried out over a pre-set wavelength range on a Unicam SP800B double-beam recording spectrophotometer, with the appropriate solvent in the blank beam. The repetitive scan mode of the instrument was used in preliminary kinetic exploration of decomposition rates, the time of passing the wavelength of maximum absorption being measured by stop-watch for each scan.PTFE-stoppered 10-mm Spectrosil cells were used. A Perkin-Elmer 137 UV double-beam recording spectrophotometer was also used in a similar fashion as a check. Didymium and holmium glass wavelength standards were used for calibration. Precise absorbance measurements at fixed wavelength were made with a Hilger H700 single-beam spectrophotometer, with blank and sample cells mounted in a constant-tem- perature housing fed from a thermostat tank maintained to within +0.05 "C of the required temperature. Molar absorptivities were thus determined and precise values of Amax. checked on this instrument. Electron spin resonance measurements were made by courtesy of the Physics Depart- ment on a Decca electron spin resonance spectrometer, operating in the Q band, but were makeshift as proper solution handling components were not available. METHODS DETERMINATION OF NUMBER OF ELECTRONS AND OF APPROXIMATE VALUES FOR Amax, AND E- A series of solutions in 100-ml calibrated flasks was prepared, each containing the same amount of diluted indicator solution, so that the absorbance at the wavelength of maximum absorption would be between 1.0 and 1.5 (2 x to 1 0 - 4 ~ after making up to volume), and the required amount of sulphuric acid to give the desired concentration after making up to volume.To these were then added diluted cerium(1V) or dichromate solution in amounts corresponding to 0, 06, 1.0, 1.5, 2.0, 3.0 and 4.0 equivalents with respect to the amount of indicator present.After addition of the oxidant, the solution was made up to volume with water, quickly mixed and immediately scanned through the appropriate wavelength range to include both ultraviolet and visible peaks. The series of spectra was recorded on the same chart. The spectra over the range from 200 to 850nm were carefully examined for evidence of intermediate or other products formed during the oxidation process and also with excess of oxidant. Approximate values for Amax. for reduced and oxidised forms of the indicator were noted for precise checking later, and approximate values of E were calculated from the absorbances at the wavelength of maximum absorption. A graph of absorbance against number of equivalents of oxidant added gave the number of electrons involved.Rate of decay of oxidised form-The rate of decay of the oxidised form was explored by repetitive scanning of the samples at timed intervals. The main examination was carried out on the samples treatedwith two equivalents of oxidant, but other ratios were also examined. During the decay process the spectra were again examined over the range 200 to 850nm for evidence of other products. More precise studies of decay were made at constant tem- perature and at a fixed wavelength of Amax, in the manual spectrophotometer. From the results thus obtained, rate plots for orders of reaction from zero to fourth order were made. Precise measurement of Amax. and E values-By using the Amax, values from spectral scans as a guide, and taking advantage of the small induction period usually allowed by a purified indicator, the region t 5 to 10 nm about the approximate Amax, value was quickly surveyed on the manual instrument with a solution containing two equivalents of oxidant.The molar absorptivity at the wavelength of maximum absorption was then calculated. A similar check was made on the reduced form of the indicator.January, 197 I] INDICATORS OF THE BENZIDINE, NAPHTHIDINE AND DIARYLAMINE TYPES 29 Determination of the amount of reduced form regenerated-A solution containing the stoicheiometric amount of oxidant was allowed to decay completely, after its absorbance at the wavelength of maximum absorption had been determined immediately after mixing. After decay the spectrum was again scanned and showed no more than a slight elevation over the base-line in the visible region.The decayed solution was then divided into aliquots, which were treated one after the other with successively larger increments of oxidant, the required amount of sulphuric acid was added and the solution made up to volume and im- mediately scanned. From a graph of absorbance against amount of oxidant added, the amount of the latter required fully to oxidise the regenerated reduced indicator was found. This was checked against the maximum absorbance reached which, from Beer’s law, also gave the amount of regenerated reduced form. As will be explained, the second method is the more reliable. RESULTS AND DISCUSSION The primary questions concerned the oxidation of benzidines of the form When X=Y =Z=hydrogen, the parent compound benzidine (4,4’-diamino-l ,l’-biphenyl) appears ; when X=phenyl and Y =Z= hydrogen, the putative intermediate diphenylbenzidine formed from diphenylamine appears; when X=Y =hydrogen, the compound is a nuclear- substituted benzidine, X or Y, or both, being other than hydrogen give N-substituted benzidines.Addition of a fused benzene ring on the side opposite to 2 gives the corresponding naphthidines (4,4‘-diamino-l, 1‘-binaphthyl) . As many compounds as could be readily obtained were examined, and gave a reasonable cross-section of the whole. These were X=Y =hydrogen and Z=hydrogen, benzidine ; Z=methyl, 3,3’-dimethylbenzidine (0-tolidine) ; Z=methoxyl, 3,3’-dimethoxybenzidine (0-dianisidine) ; Z=hydrogen, X= hydrogen and Y = phenyl, diphenylbenzidine ; and X =methyl and Y = phenyl, NN’-dimethyl-NN’-diphenyl- benzidinedisulphonic acid.Naphthidine (X =Y =Z= hydrogen) and 3,3‘-dimethylnaphthi- dine (X=Y =hydrogen and Z=methyl) provided examples of the naphthyl analogues. Of the presumptive precursor diarylamines, diphenylamine, diphenylamine-4-sulphonic acid, N-methyldiphenylamine-4-sulphonic acid and 2-carboxydiphenylamine (N-phenylanthranilic acid) were also examined; the last should give a benzidine with X=hydrogen, Y=phenyl and Z=carboxyl, but the location of the carboxyl groups is indeterminate. NUMBER OF ELECTRONS- In the indicator reaction, Index + ne + Indrea the number of electrons, TZ, is in all instances 2. Graphs of absorbance of Indo, against equivalents of oxidant added are linear for all compounds examined up to 2 +_ 0.02 equiva- lents per mole of Indred, whether this be a benzidine, a naphthidine, an arylamine or an alkylarylamine : the absorbance remains constant when oxidant in excess of the two equiva- lents is added.SPECTRA- Immediate& after oxidatiofl-All spectra show a cut-off in the far ultraviolet region because of high sulphate concentrations and aromatic ring currents. Usually Ind,,d shows a peak in the mid or near ultraviolet region; the wavelength is longer for the benzidine than for the arylamine, exceptions being for Z=alkoxyl, when a double peak appears, and the naphthi- dines, when there is a double hump or a complex spectrum: none shows any other absorption above the vibrational frequencies. The Ind,, spectrum shows a single peak in the visible region, except for Z=alkoxyl, when the peak is again split, and naphthidine, which shows an additional weaker higher frequency absorption.Starting with pure Indred and scanning30 BISHOP AND HARTSHORN : SOME OBSERVATIONS ON OXIDATION - REDUCTION [Arzalyst, Vol. 96 solutions containing progressively larger increments of oxidant, the ultraviolet absorbance of Ind,,a decreases proportionately, usually to zero, and the visible absorbance of Ind,, appears and increases proportionately until two equivalents of oxidant have been added; thereafter there is no further change. At the same time the absorbance of solvated cerium(II1) appears in the ultraviolet /(Amax = 252 rnm, E = 905 1 mol-1 cm-l) and increases proportion- ately, reaching a maximum when two equivalents of cerium(1V) have been added.The changes in these absorbances are precisely linear with the number of equivalents of oxidant added, terminating at 2 f 0.02 equivalents. These peaks show no further change when more than two equivalents of cerium(1V) are added, but the absorbance of the latter (Amax, = 317 nm, E = 5 900 1 mol-l cm-l) appears and increases in proportion; there is no evidence whatever of a further oxidation product of the indicator. The series of Indred - Indo, spectra shows a characteristic isosbestic point pattern (Fig. 1). The shape of the Ind,, spectrum is the same 0.8 L 2 0-4 0.2 0 I I - 0) i 0.6 - t L I t ’ 250 275 300 325 I I 1 I 350 400 450 500 550 Wavelength/ nm Fig. 1. Illustration of the isosbestic pattern on oxidation.Solutions 4 x M in N-methyl- diphenylamine-4-sulphonic acid and 2.0 M in sulphuric acid treated with a, 0; b, 0.60; c, 1-21; d, 1-72; e, 2.0; and f, 2.2 equivalents of cerium(1V): e and f are coincident as that of Ind,,a, unless the latter is an arylamine, with merely a bathochromic shift and an increase in molar absorptivity, which are nearly constant for a given type of compound. This indicates that there is no basic change in structure but merely a decrease in mean energy of the electronic transitions. Spectral measurements are collected in Table I. TABLE I SPECTRAL CHARACTERISTICS OF SOME BENZIDINE AND NAPHTHIDINE DERIVATIVES AND SOME ARYLAMINES IN 2.0 M SULPHURIC ACID AT 25 “C Indred Indo, & - Spectral Compound Amax/nm 6/1 mol-1 cm-l Amar/nm ~ / 1 mol-l cm-I shift/kcal mol-l Cox/Cred Biphenyl 252 20000 (in ethanol) Benzidine 248 20 300 426 69 500 50-8 3.3 3,3’-Dimethyl- 248 18000 438 - 49 benzidine 250* 18000* 435* 65 000 48.6 3.6 12 800 454 31 000 49 2.4 8 100 510 23 000 41.5 2.8 31 200t 560 60 000 61.9 1.6 benzidine { E 3,3’-Dimethoxy- NN’-Diphenylbenzidine 253t N-Phenyl- N-Methyldiphenyl- 3205 24 OOOj 511 44 000 62 1.7 Diphen ylamine 220: Cut-off 565 45 000 - - anthranilic acid 256: - 524 30 000 - - amine-4-sulphonic 294: 14 OOO$ acid - 400 7 000 30 {E 11 000 527 18000 43 1.6 Naphthidine 3,3’-Dimethyl- naphthidine Complex spectrum 543 35 000 - - * In M sulphuric acid.t In concentrated sulphuric acid. $ Arylamine form. 5 In benzidine form; oxidised with two equivalents of oxidant and immediately reduced with zinc dust.January, 19711 INDICATORS OF THE BENZIDINE, NAPHTHIDINE AND DIARYLAMINE TYPES 31 When Indr,d is an arylarnine, a slight change in shape occurs for the Ind,, spectrum, but when Ind,, is reduced to the presumptive benzidine and the new Indred spectrum com- pared with that of the arylamine, the change is seen to be minor.Arylamine oxidation occurs in a single step: no evidence exists for the intermediate formation of the benzidine. Lest this be due to use of a very strong oxidant [cerium(IV)], dichromate, vanadate, iron(II1) and other oxidants were tested : either no oxidation occurred, or the single-step two-electron oxidation took place. If the arylamine is oxidised with two equivalents of oxidant and then immediately reduced with zinc dust, a scan shows that the ultraviolet peak is shifted to a longer wavelength, thus indicating the benzidine form.Subjecting this solution to the oxidation and scanning process shows that the oxidised form is identical in spectral charac- teristics with the oxidation product of the original arylamine, but the maximum absorbance is now reached at one equivalent of oxidant per mole of original arylamine, which corresponds to two equivalents per mole of the presumptive benzidine produced by the zinc dust reduction. The example shown in Fig. 1 is an alkylarylamine; reduction with zinc dust and repetition of the process gives a similar set of spectra with the ultraviolet peak shifted to 320 nm and a corresponding shift in the isosbestic point. As the sulphuric acid concentration of the medium is increased, a considerable batho- chromic shift occurs in Amax,, which is caused by the solvent effect, as shown in Table 11.Belcher has observed this shift with naphthidine derivative^.^ In sulphuric acid media there is no detectable evidence in any of the spectra for the formation of an intermediate, even transiently, either of the semiquinone type from a benzidine, or of a benzidine from an arylamine. In acetic acid some evidence exists that a different reaction may occur, and that this is specific to the presence of acetic acid or acetate ion. (Kolthoff's solutions con- tained acetic acid.l) For example, a double-humped peak in the region 300 to 420 nm was observed in the oxidation of diphenylbenzidine in a mixture of concentrated sulphuric acid, glacial acetic acid and water (2 + 78 + 20 v/v after correction for the addition of aqueous acid cerate) ; peculiarities have also been noted with o-dianisidine in acetate media.2 This, however, opens up a large field of subsidiary study, which will not be pursued at present.On decompositiort-On standing, all oxidised solutions, whether containing a deficiency, the stoicheiometric amount or an excess of oxidant, faded more or less rapidly, eventually becoming colourless and occasionally depositing a precipitate. In scanning the spectrum at intervals during this decay, as the absorbance of Ind,, decreased, the absorbance of Indred appeared and increased with time, again giving an isosbestic pattern as illustrated in Fig. 2. The absorbance of Indo, fell virtually to zero, but the absorbance of Indred increased to only about one half of its original value.Apart from N-phenylanthranilic acid, no other peaks appeared and there was no evidence for the formation of a semiquinone or any other compound, except that the base-line of the whole spectrum rose by a small but detectable and fairly uniform amount, such as may be expected from the formation of an insoluble but disperse phase. There was no change in shape or in the Amax values of either In&, or In&,d peaks during the decay process, except that an arylamine decayed to the benzidine giving the same Indred spectrum as that after zinc dust reduction of the freshly oxidised arylamine. Even in solutions containing one equivalent of oxidant, which might be expected to encourage 0.8 - a 0.2 - I I I I I I I O t 275 300 325 350 400 450 500 550 t Wavelength /nm 10 Fig.2. Illustration of the isosbestic pattern on spontaneous decomposition. A solution M in naphthidine and 2.0 M in sulphuric acid treated with 2.0 equivalents of cerium(1V) 4 x and allowed to decay. Interval between scans a to e, about 12 minutes32 BISHOP AND HARTSHORN : SOME OBSERVATIONS ON OXIDATION -REDUCTION [Analyst, Vol. 96 formation of the semiquinone, no spectral shift was observed, although decay was accelerated. Evidence for autocatalysis was found in many instances. The cerium(II1) and cerium(1V) peaks, when present, complicate interpretation of the spectra, but their resolution is possible from the law of additive absorbances.Unsubstituted benzidine and naphthidine oxidised forms decay rapidly in daylight, but in darkness, or in the cell housing of the spectrophotometer, the decay is drastically retarded, thus indicating that the process is photosensitive. Belcher, Lyle and Stephen5 have reported that 3,3'-dimethoxynaphthidine is photosensitive at the wavelength of maximum absorption and they were therefore unable to determine the Amax. or E value for this compound. The curious behaviour of N-phenylanthranilic acid is noteworthy. It is directly oxidised, without any intermediate benzidine formation, to the bluish red Ind,,. The Ind,, band at 524nm immediately begins to decay and a new band at 436 nm grows and the colour changes to green. This decay is rapid with a half-life of 10 minutes, and the set of spectra scanned at 3-minute intervals shows a good isosbestic point, smeared only slightly by the over-all slower decay of both species.Re-oxidation of the decayed solutions gave an identical Ind,, peak and the same Indred - Ind,, isosbestic behaviour. When the decay of a solution treated with excess of oxidant was followed by the scanning method, the decay was slower to begin with because of re-oxidation of the regenerated Indred by the excess of oxidant, but the cerium(1V) absorbance decayed more quickly than could be accounted for by this reaction: the cerium(1V) absorbance is therefore useless in assessing the extent of reaction. ELECTRON SPIN RESONANCE SPECTRA- Without the proper components for work with solutions the electron spin resonance spectra have only a qualitative value.Solutions of diphenylbenzidine with a range of sulphuric acid concentrations and amounts of oxidant were examined. Cerium(II1) and cerium( IV) gave no signals at the operating frequencies. A half-oxidised solution showed a signal that grew gradually to a maximum, and showed a Land4 splitting g factor of almost exactly 2, indicative of a free radical with a single unpaired electron, and then decayed at a rate similar to the rate of decay of absorbance at 560 nm previously observed spectrophoto- metrically. The band was unusually broad, probably on account of solvent effects, and may arise from an averaging effect, so that it cannot be regarded as evidence of a semiquinone, particularly in view of the absence of optical evidence.As the amount of oxidant was in- creased, finally to two equivalents, this band disappeared and was replaced by a weak complex signal that can be related to, but does not with any certainty identify, a diradical with two separated electrons of parallel spin. KINETICS OF DECOMPOSITION- The spectral scanning method gave a clue to the pattern of decomposition, there being some indication of a small induction period with pure indicators and strong evidence of autocatalysis. Precise kinetic measurements were made by following the change in absorbance at the wavelength of maximum absorption of Ind,, with time, at controlled temperature. The decay rate was found to increase with rising temperature, to decrease with increasing sulphuric acid concentration (as shown in Table 11), and to depend on the nature of the TABLE I1 SOLVENT SHIFT AND EFFECT OF SULPHURIC ACID CONCENTRATION ON DECAY OF OXIDISED DIPHENYLBENZIDINE AT 25 "C Sulphuric acid, per cent.. . 18 22 36 63 Amax. Indox/nm . . . . 560 565 570 585 Half-lifelminutes . . .. 188 200 445 1300 substituents X, Y and 2. The stabilising effect of high sulphuric acid concentrations is implicit in some earlier work,3 and may be caused by the decrease in water concentration, or by increased protonation of the various basic species. The decay of the oxidation products of the unsubstituted benzidine and naphthidine is greatly accelerated by exposure to daylight as has been noted. Sulphonation of the indicator de-stabilises the oxidised form, as has several times been observed b e f ~ r e , ~ , ~ and could well result from a shift in the charge densityJanuary, 19711 INDICATORS OF THE BENZIDINE, NAPHTHIDINE AND DIARYLAMINE TYPES 33 distribution.Partially oxidised solutions decay faster than stoicheiometrically oxidised solutions, and addition of Indred to a decaying solution accelerates the process markedly, as with diphenylbenzidine, giving support to the hypothesis that the autocatalysis is caused by the growth of regenerated Ind,ed in the solution. The kinetics of the decay process do not fit any reaction order, but come closest to first order; indeed, some show an almost perfect first-order plot. It is possible to fit the decay process for diphenylbenzidine to an equation of the form- - k, [Ind,,] + k, [product] d [Indod dt in 2 M sulphuric acid ( k , = 2.3 x Rather than present artificially fitted rate constants, the speed of the decay process is represented by the half-life in Table I11 under the conditions specified.As Indred in the benzidine form minute-l and k , = 8.2 x lW3 minute-l). TABLE XI1 HALF-LIVES OF IND,,~ IN THE ABSENCE OF EXCESS OF OXIDANT, AND IN THE INITIAL ABSENCE OF IN&&, IN 2 M SULPHURIC ACID AT 26 “C; AND THE RECOVERY OF INDred AFTER COMPLETE DECAY Compound Benzidine . . .. .. .. .. 3,3’-Dimethylbenzidine . . .. .. 3,3’-Dimethoxybenzidine .. .. NW-Diphenylbenzidine . . .. .. Diphenylamine . . .. .. .. N-Phenylanthranilic acid .. .. Naphthidine .. .. .. N-Methyldiphenylamine-4-sulphonic acid]) 3,3’-Dimethylnaphthidine . . .. Half-lifelminutes .. 500* ..264 .. 190 .. 188 .. 96 .. 6 to 105 . . 170 .. 80 200 : . - 23’11 Indrd regenerated, per cent. sot 50 46: 60 50, 51, 46 1005 - 65 40 35, 40 * Measured in darkness. Decay accelerated in daylight to about 50 minutes. t Value very high. $ In M sulphuric acid. Q Initial fast decay to green form. 11 Presumptive benzidine derivative : Ind,,d oxidised with two equivalents of oxidant, im- 7 At 30 “C. Decay much faster in daylight, or under continuous illumination a t wavelength mediately reduced with zinc dust, then re-oxidised. of maximum absorption (about 2-5 minutes). is regenerated in the decay process, the latter must be of the nature of a disproportionation with a minimum molecularity of two, and therefore the rate-controlling step must be a preceding unimolecular reaction that could be accelerated by energy exchange with Indred.Substituents in the 2 position appear to protect the amino group, as Belcher, Lyle and Stephen ~uggest,~ but this is perhaps more a matter of preventing the photochemical reaction, with the exception noted,5 than of hindering further oxidation. It is notable that the half-life of oxidised diphenylamine is about half that of diphenylbenzidine. Substituents in the X or Y position stabilise the oxidised form, and maximum stability is reached when both X and Y are substituted. The instance of NNN’N‘-tetramethyl-3,3’-dimethylbenzidine (t et ramethyl-o-t olidine) is strongly relevant. RECOVERY OF INDred- Determination of the amount of Indred regenerated would define the stoicheiometry of the decomposition reaction.Re-oxidation of the regenerated Indred gives precisely the same spectrum and decay pattern as the original indicator. Measurement of the absorbance at the appropriate wavelength of the regenerated Indred and calculation from Beer’s law will give the amount of Indred, provided proper correction for other absorbances such as cerium(II1) and decomposition products is made. Measurement of the absorbance of Ind,, for a series of solutions to which successively larger amounts of cerium(1V) have been added and a plot of absorbance against equivalents of oxidant do not give reliable results, because cerium( IV) is consumed in other reactions, as noted before. However, addition of a slight excess of34 BISHOP AND HARTSHORN : SOME OBSERVATIONS ON OXIDATION - REDUCTION [A~zalyst, Vol. 96 oxidant [checked by the appearance of a small cerium(1V) peak], measurement of the absorb- ance of Indo, and application of Beer’s law give reasonably reliable results, provided the measurement is made quickly enough to avoid significant decay and yet allow sufficient time for the full development of colour. This latter time is no more than a few seconds except for benzidine, which takes up to 30 minutes.It is not possible to check the stoicheio- metry by adding the integrated number of equivalents of cerium(1V) (e.g., four for a 2 : 1 disproportionation) so that there is no residual Indred left after decay. Indrea is regenerated even in the presence of a large excess of oxidant, and this phenomenon persists until about twenty-two equivalents of oxidant have been added, so clearly oxidant is attacking a de- composition product by ring fission faster than Indred is regenerated.If the disproportionation is formulated as m Ind,, 3 D + % Indrea where D is the primary decomposition (disproportionation) product, then excess of oxidant is consumed by three processes: (a) re-oxidation of regenerated Indred to In&,, (b) direct oxidation of Ind,, to D, and (c) attack of D to give fission products, and of these reaction (c) is the major process. Results are included in Table I11 and, on balance, favour a 2 : 1 dis- proportionation (50 per cent. regeneration) , although other stoicheiometries such as 5 : 2 (40 per cent. regeneration) are not excluded, and there are notable exceptions. CONCLUSIONS In sulphuric acid media all indicators are oxidised directly to Ind,, by a single two- electron step.Reduction of oxidised arylamine stops at the benzidine stage. Oxidation of arylamines does not proceed by a benzidine tran~formation,~ and is not, therefore, an intra- molecular reaction but an intermolecular reaction. Starting from an arylamine there are five possible products depending on which carbon atoms provide the bridge, and if the arylamine is substituted in the nucleus the possible products are multiplied in number accord- ing to the rings in which the substituents finally appear: there are fifteen possible products, for instance, for N-phenylanthranilic acid, although it is probable that a single product will preponderate in a given instance.Presupposing that the product is a p,@’-benzidine and the substituent 2 appears in the benzidine nucleus, the main reaction can be formulated as- bentidine base Y NX xi&x - protonated benzidine -2H+-2e L - arylamine (Y = phenyl) where Ind,, is a triplet diradical dication. It should be emphasised that benzidine has not been isolated except after zinc dust reduction of In&, for the case when X = 2 = hydrogen and Y = phenyl. The identity in shape of the benzidine and Ind,, spectra and the isosbestic pattern suggest that no gross change occurs on oxidation, and the bathochromic shift indicatesJanuary, 19711 INDICATORS OF THE BENZIDINE, NAPHTHIDINE AND DIARYLAMINE TYPES 35 an elevation of the mean ground-state energy of the molecular orbitals, so that electronic transitions require a lower energy; this is supported by the increase in molar absorptivity.Except for the curious instance of N-phenylanthranilic acid, Indred is invariably re- generated on decomposition (as the benzidine in the case of arylamines), together with a weakly coloured sparingly soluble decomposition product, D. Regeneration of Indred is clearly the reverse of the main indicator reaction. Neglecting the possible participation of hydrogen ion, this is- Ind,, + 2e + Ind,,d The necessary electrons must come from somewhere, and as Indo, will itself consume oxidant, it is the obvious source- The spontaneous decomposition of Ind,, is therefore a disproportionation, several examples of which are known in this field of chemistry.3 However, the minimum value of $’ for a disproportionation would be 2, and the decay does not show second-order kinetics.A monomolecular rate-controlling step must therefore precede the disproportionation, and it is not unlikely that this should be the relaxation of the triplet state, catalysed by energy exchange with Indred, to the xIndo, + y D + x e $’ Indo, -+ 4 D Y Indred hd4+J=(==J:x DQDI diquinonediimine; it could also be the deprotonation of DQDI to give the free base. The final reaction could then be formulated as for which the over-all stoicheiometry would be (a + b)/c. The product D has been neither isolated nor examined, and even if it had it would be difficult to be sure that it had not been changed in the process. Discussion of the nature of D is necessarily speculative, but there is obviously a connection between the substituents X, Y and 2 and the stability of In&,.X and Y are those which matter in the decay process, and it is worth recalling that stability increases with increasing sulphuric acid concentration, so that protonation may be influential. Moreover, biphenyl is itself unreactive so that the activity must reside in the nitrogen atoms. When neither X nor Y is a hydrogen atom, then the oxidation product, whether formulated as the diradical dication or the diquinonediimine, has no lone pairs available for protonation, and it is not possible to formulate a further oxidation stage without breaking bonds or ejecting substituents. That the tetra-N-substituted derivatives are stable in the oxidised form is well known,3y6 but a quantitative definition of stability is lacking.That the oxidised form of N-methyldiphenylamine-4-sulphonic acid does decompose and regenerate a benzidine may be the result of the de-stabilising effect of the sulphonic acid group but is still difficult to understand. When either X or Y is a hydrogen atom, then protonation is possible and the diquinonedi- imine does offer electrons accessible to oxidation. A 2 : 1 disproportionation can be formu- lated as- a Indo, + b DQDI -+ D + c Indred where the primary decomposition product, D, is the diradical dication of the DQDI. This unlikely looking species would be very reactive and may polymerise by nuclear attack. A 3 : 1 disproportionation is easier to formulate but is contrary to the 50 per cent. recovery Of Indred.36 BISHOP AND HARTSHORN When X=Y =hydrogen, the 2 : 1 disproportionation would again require formation of the diradical dication DQDI, but with hydrogen in place of Y, which opens up further possibilities. A 3 : 1 disproportionation would allow the formation of a hydrazine that is capable of being oxidised to an azo compound, which could isomerise to a dimeric diquinone- diimine or could be oxidised to an azoxy compound. The formation of a linear trimeric bis-hydrazine would give a 5 : 2 stoicheiometry and a 40 per cent. recovery of Indred. There is also the possibility of a phenazine type of polymerisation. It is not profitable to enumerate further all the possibilities because the matter remains speculative until accurate information is available on the nature of D. However, the stability of Ind,, increases as 2, Y and X are successively substituted and as the sulphuric acid concentration increases, and the recovery of Indred suggests that the oxidation of Ind,, to D is a two-electron process. REFERENCES 1. 2. 3. 4. 5. 6. Kolthoff, I. M., and Sarver, L. A., J . Amer. Chem. SOC., 1930, 52, 4179. Crawford, A. B., and Bishop, E., J . R. Tech. Coll. Glasg., 1950, 5, 62. Bishop, E., “Indicators,” Pergamon Press, Oxford, 1970, Chapter 8B. Smith, G. F., “Cerate Oxidimetry,” G. F. Smith Chemical Co., Columbus, Ohio, 1942. Belcher, R., Lyle, S. J., and Stephen, W. I., J . Chem. SOC., 1958, 4454. Jordanov, N., and Daiev, Ch., Talanta, 1963, 10, 163. Received April 16th, 1970 Accepted August 19tk, 1970

ISSN:0003-2654    

DOI:10.1039/AN9719600026

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, January, 1971, Vol. 96, pp. 37-46 37 The Determination of Carbon in .Steel by Cowlometric Titration in Partially Aqueous Medium BY H. J. BONIFACE AND R. H. JENKINS (British Steel Corporation, Strip Mills Division Research Centre, Port Talbot, Glamorgan) A method that involves coulometric titration of carbon dioxide absorbed in a partially aqueous medium is described for the precise determination of carbon in steel. It does not depend on empirical standardisation and the use of high-vacuum systems and inflammable titrants is avoided. The apparatus used is inexpensive and the technique is simple. The application of appropriate instrumentation could provide a routine control method of high precision. The analytical performance of the method compares well with that of other techniques considered suitable for reference analysis; 95 per cent. con- fidence limits of &0-0007 per cent.of carbon a t the 0.05 per cent. of carbon level of 0.008 per cent. of carbon a t the 0.9 per cent. of carbon level were obtained in this laboratory. RECENT trends in methods for determining carbon in metals, particularly in steel, have been towards rapid routine techniques, which have also provided an increase in precision at low levels. Many physical techniques have been successfully used over the past 10 years and commercial equipment based on the following principles are available : infrared gas analysis,lJ thermal conductivity and electrical conductivity. Invariably, the capital cost of the equip- ment has been relatively high. Rapidity coupled with good accuracy, however, was achieved at relatively low cost by Jones, Gale, Hopkins and Powell3 in their non-aqueous titration method, and although not as robust as those with the physical instruments it has found wide application in works’ laboratories.Most of the recently developed methods depend on empirical standardisation and are unsuitable for reference purposes. The gravimetric methods, which have been largely replaced, were unreliable at low carbon levels, although Bagshawe and Pindefl improved the sensitivity of the method by using large sample weights. For many years the low-pressure volumetric method used by Wells,5 and later modified by Cook and SpeightI6 was generally accepted as a reference method as it was based on fundamental gas laws.It has lost favour because of the fragile nature of the equipment, the skill required in its operation and the need for a supply of liquid oxygen. Dunnill and Kent7 recently proposed a modification of the method that overcomes some of these objections. The need arose in this laboratory to replace the low-pressure volumetric method by a simpler technique suitable for reference analysis. In 1959, Abresch and Lemms proposed a coulometric technique for the determination of oxygen in steel. The application of Faraday’s laws and the measurement of quantities such as current and time, which can be made with high accuracy, were the attractive features. An investigation was made into the possibility of determining carbon in steel by coulometric titration. Conditions similar to those of Abresch and Lemm were used with aqueous barium perchlorate solution as the absorbent for carbon dioxide in an oxygen stream.Promising results were obtained, but difficulties experienced in achieving efficient absorption of carbon dioxide caused the method to be unreliable. However, commercial instruments that can be successfully used for coulometric titration in aqueous media have since been marketed in Germany. Jones and his co-workers3 and Braid, Hunter, Massie, Nicholson and Pearceg had demon- strated the high efficiency of dimethylformamide containing monoethanolamine in absorbing carbon dioxide. Consequently a study was made of the coulometric titration of carbon dioxide absorbed in non-aqueous and partially aqueous solvents in which various electrolytes were used. During this investigation Whymark and OttawaylO reported their work on coulometric titration in semi-aqueous media; they found that the presence of 20 to 30 per 0 SAC and the authors.38 BONIFACE AND JENKINS: DETERMINATION OF CARBON IN STEEL [Artalyst, Vol.96 cent. of water was necessary for 100 per cent. current efficiency in pyridine and dimethyl- forrnamide, but no other electrolyte was used. The titration of acids, including carbon dioxide, by generating a base in non-aqueous solution, has been described by several authors-ll 912 9 1 3 9 1 4 9 1 5 End-points were determined by either potentiometric or visual methods. Previous experience indicated that serious difficulties could arise because of the disturbing influence of the generating current on the indicating electrodes.It was therefore decided to use a photometric end-point detection method.16 The primary criteria for a successful procedure were recognised as (i) complete current efficiency and (ii) complete absorption of carbon dioxide. These considerations form the basis of the investigation. It is emphasised that this paper describes only the basic conditions that have been found suitable for the determination of carbon dioxide evolved by ignition of steel samples in oxygen. The apparatus and technique could well be developed further to provide automatic operation. APPARATUS AND CONDITIONS- EXPERIMENTAL The apparatus used is shown diagrammatically in Figs. 1 to 3. The cell for the initial experiments consisted of a 150-ml squat beaker containing a strip of 0.6-cm wide platinum foil formed into a circle lying around the inside wall at the base of the beaker; this strip acted as cathode.A Perspex lid carried a glass cylinder with sealed-in sinter disc at its lower end containing the anolyte and a graphite rod anode. A gas delivery tube also passed through the lid; its tip was diverted off-centre of the beaker section. The current source (see circuit diagram, Fig. 3) provided a 20-mA constant current, which was measured with a milliammeter. End-point detection was provided by an Evans Electro- selenium Limited (EEL) Quantitrator with Ilford filter No. 607, the cell being positioned on the magnetic stirrer stand of the titrator. Later, an all-glass cell with a side-arm for the anode chamber was constructed (Fig.l), and a silver rod was substituted for the graphite anode. The lid was made of glass and fitted to the cell body by a ground-glass joint. Gas inlet i\ -55- Fig. 1. Absorption cell (all measurements are in millimetres)19711 BY COULOMETRIC TITRATION IN PARTIALLY AQUEOUS MEDIUM = Flow meter H = Combustion tube B = Purification tube J E Manganese dioxide and magnesium C = Resistance furnace D = Soda asbestos tube E = 10-Litre reservoir L = EEL Quantitrator F = Soda asbestos and magnesium per- M = Lamp G = Atmosphere trap perchlorate tube K = Absorption cell N = Filter and photocell 0 = Spot galvanometer chlorate tube Fig. 2. Schematic diagram of apparatus 39 TITRATION PROCEDURE- With the photometric end-point detector switched on (lamp, galvanometer and stirrer), the absorbing solution was titrated to an arbitrary deflection (within the blue region) by passing a current of 20mA.Acid or carbon dioxide was then introduced into the cell as described later and, after complete absorption, the current was switched on simultaneously with starting a stop-watch. The milliammeter reading was immediately adjusted to 20 mA and the titration was continued until the galvanometer spot returned to the starting point. Current and stop-watch were stopped exactly at this point. - - 1-1 i D1 - D 4 c1: R2 , 1 output IR4 R3 Fig. 3. Circuit diagram of current source (for values of components see Appendix)40 BONIFACE AND JENKINS: DETERMINATION OF CARBON IN STEEL [Analyst, Vol. 96 ABSORBENT AND ELECTROLYTE- In preliminary tests the solvent mixture and indicator described by Jones et aL3 were used.The composition of the mixture was as follows: dimethylformamide, 150ml; mono- ethanolamine, 5ml; and thymolphthalein (0.1 per cent. solution) 2ml (for this work the thymolphthalein was dissolved in dimethylformamide instead of methanol). Various salts were chosen as possible electrolytes and were dissolved in the solution. calculated tihation time " observed titration tirne The current x 100, of each electrolyte was measured by titration of known amounts of pure benzoic acid. The most suitable electrolyte was shown by these experiments to be potassium iodide (Table I). More reproducible efficiencies were obtained by replacing the graphite anode by a silver anode, which prevented diffusion of iodine from the anode compartment into the bulk of the solution.TABLE I CURRENT EFFICIENCIES OF VARIOUS SALTS DISSOLVED I N A DIMETHYLFORMAMIDE - MONOETHANOLAMINE MIXTURE Weight in 75 ml Electrolyte of solventlg LiCl . . .. .. 0.4 NaI .. . . .. 2 KI . . .. .. 3 RbI . . . . .. 0.5 CSI . . .. .. 0.5 MgClO, .. . . 0.5 NaC10, .. .. 2.0 CaCl, . . .. .. 1 BaC10, .. .. 1 NH,Br.. .. .. 1 .. 0.5 .. 0.5 (%H6)4NBr (%H6)4N1 . ' Efficiency, per cent. 88 to 94 95 and 97 94 to 100.5 81 83 - - - 82 80 68 Remarks Deposit formed on cathode Range for about twenty results Range for about twenty results Apparently no base generated Very slow colour change Insoluble product formed on cathode Apparently no base generated - - - - - It was found that the non-aqueous solution of potassium iodide in dimethylformamide - monoethanolamine mixture formed an insoluble product on titration of carbon dioxide, which restricted the useful life of the absorbent.It was thought that the addition of water would reduce the insolubility and about 4 per cent. of water was added (this amount gave the greatest increase in the solubility of the product compatible with the least loss of end-point definition). The composition of the absorbing solution was therefore changed to the following : dimethylformamide, 78 ml; 0.1 per cent. solution of thymolphthalein in dimethylformamide, 2 ml; water, 3 ml; potassium iodide, 3 g; and monoethanolamine, 3 ml. TITRATION EFFICIENCY- A 0.4 per cent. w/v solution of benzoic acid in dimethylformamide (2.0 ml) was added to the titration cell (fitted with a silver anode) by using a burette with a PTFE stopcock.Titration was carried out at currents of 10 and 20 mA, the flow-rate of oxygen being 200 ml minute-l. A blank was performed with 2.0 ml of dimethylformamide. Weighed fused pieces of benzoic acid were also dissolved and titrated. The results obtained are shown in Table 11. TABLE I1 TITRATION OF BENZOIC ACID Current efficiency, per cent. Benzoic acid taken 10 mA 20 m i 20 mg as fused solid. . . . - 100.8, 100.9 8 mg in solution . . . . 101.0, 100.3 101.2, 100.3 101.5 99.5, 100.6 Measured volumes of carbon dioxide were passed into the absorbing solution by using a gas pipette, oxygen being the carrier gas. The solutions were titrated and results given in Table I11 show the recovery obtained.January, 19711 BY COULOMETRIC TITRATION IN PARTIALLY AQUEOUS MEDIUM TABLE I11 TITRATION OF CARBON DIOXIDE 41 Recovery, per cent.Volume of carbon r dioxidelml 10 mA 20 mh 1.756 99.5, 100.1 99.7 101.0, 100.3 101.4, 98.9 Later, confirmation of efficient absorption was obtained : two cells were connected in series and the carbon dioxide collected in the second cell was titrated. Carbon dioxide was provided by the ignition of steel samples, by using the combustion train described in the following section. Steel sample Weightlg B.C.S. 218/3 0-37 B.C.S. 218/3 0.75 B.C.S. 220/1 0.075 B.C.S. 220/1 0.156 B.C.S. 220/1 0-301 B.C.S. 220/1 0.592 TABLE IV ABSORPTION EFFICIENCY Titration timesls r A 'L 1st cell 2nd cell Times for l-mA current divided by 20 for comparison with 1st cell Current 20 mA 264 0.07 514 0.2 1 278 0.00 576 0.19 1110 1.18 2190 3.07 Percentage of carbon dioxide not absorbed by 1st cell 0.03 0.04 0.00 0.03 0.11 0-14 Results in Table IV indicate no significant loss of carbon dioxide from the first absorbent.APPLICATION TO STEEL SAMPLES- Equipment for application of the technique to steel analysis was set up, which consisted of a conventional combustion train with resistance heating (1200" C) coupled to the coulometric cell. The oxygen supply was purified by passage through a heated refractory tube and soda asbestos. Sulphur gases were removed from the stream after combustion by using a manganese dioxide column. Selected British Chemical Standard samples were analysed by the procedure (see Method).Tin was added to all samples as a combustion initiator and blank determinations with boats containing tin alone were carried out for periods of time similar to those for the samples. The effect of oxygen flow-rates of 100 to 700 ml minute-1 was examined. Results given in Table V show that the optimum rate is 200 to 400mlminute-l. TABLE V INFLUENCE OF FLOW-RATE OF OXYGEN ON THE DETERMINATION OF CARBON IN B.C.S. 218/3 (CERTIFICATE VALUE 0.17 PER CENT. OF CARBON) Oxygenflow-ratelml minute-l 100 200 300 400 500 600 700 Carbon, per cent. . . 0.1648 0.1713 0-1702 0.1717 0.1694 0.1705 0.1693 0.1714 0.1716 0.1701 0.1705 0.1699 0-1691 A total determination time of 8 to 9 minutes was considered acceptable, which was made up of 2 minutes for purging, 2 minutes for burn-off and 4 to 5 minutes for the titration.Sample weights were adjusted according to the carbon content to enable titration to be completed in the time. The capacity of the absorbent (before precipitation occurred) was found to be about 5 mg of carbon. Fresh electrolyte solution was required after this amount had been absorbed. A titration current of 20mA was normally used but 10mA was more convenient for low carbon levels.42 REAGENTS- BONIFACE AND JENKINS: DETERMINATION OF CARBON IN STEEL [Analyst, Vol. 96 METHOD Indicator solution-Dissolve 0-1 g of thymolphthalein in 100 ml of dimethylformamide. Absor-tion solution-To 780 ml of dimethylfonnarnide in a 1-litre bottle add 20 ml of the indicator solution, a solution of 30 g of analytical-reagent grade potassium iodide in 30 ml of water and finally 30 ml of ethanolamine.Mix the solution and transfer 80 to 90 ml to the absorption cell. APPARATUS- Oxygen cylinder. Flow meter-Capable of measuring flow-rates of oxygen up to 800 ml minute-1. TNbe furnace-Capable of operating at 1200" C (or 1300" C if alloy steels are to be andysed). It is convenient to use a two-tube furnace so that the purification tube (used to oxidise carbon monoxide and other carbon compounds) and the combustion tube can be heated in one furnace. These tubes are made of aluminous porcelain and are 76 cm long and 2-2 cm i.d. An atmosphere trap (through which the samples are introduced) is fitted at the end of the combustion tube, which consists of a modified 29/32 cone and socket. The cone is joined to the combustion tube with rubber tubing. The tubing attached to the socket is drawn out into a narrow tube that can be closed with a rubber cap.Sealed to the side of the socket is a side-arm through which oxygen is passed. By using this trap the sample can be pushed into the furnace while an out-flow of oxygen excludes the atmosphere. Absorption tower cfilled with soda asbestos. A 10-Zitre aspirator bottle-This acts as a reservoir of gas to prevent suck-back of the absorption solution. Absorption tower Jilled with a layer of soda asbestos and a layer of magnesium perchlorate. Absorption tube-This contains a layer of manganese dioxide on asbestos and a layer of magnesium perchlorate. The manganese dioxide is prepared as follows : shake thoroughly 25 g of ignited asbestos fibre with 408 ml of a saturated solution of potassium permanganate.L4dd 400ml of a saturated solution of manganese(I1) sulphate and shake the mixture well. Filter on a Buchner funnel under pressure, wash the fibre twice with hot water and dry the cake at 105" C. Break up the solid mass into small granules. Absorfition cell (see Fig. 1)-This contains 85 ml of absorption solution. End-point detection afiparatus-This consists of an Evans Electroselenium Limited Current source-Capable of producing 20.00 +_ 0.01 mA. Ammeter-Capable of giving a full-scale deflection of 20 mA (e.g., Sangamo - Weston Sto$-watch-Capable of being read to 0.1 s. Alpolain No. 3 refractory combustion boats. PROCEDURE- Weigh a suitable amount (Note 1) of the steel sample into a pre-ignited combustion boat (Note 2).Cover the sample with 0.2 g of de-greased tin foil or granules and then place the boat in the mouth of the combustion tube. Allow the oxygen to flow for 2 minutes to free the train from contamination. Switch on the titration current and allow it to flow until the galvanometer light spot reaches the end-point marked on the scale (Note 3), then switch off the current. Remove the cap from the atmosphere trap at the end of the combustion tube and push the combustion boat into the hot zone of the furnace with a stainless-steel rod. Replace the cap. Adjust the oxygen flow-rate to 400ml minute-l. As the sample burns (shown by cessation of gas bubbles in the absorption cell) increase the gas flow temporarily to prevent suck-back of the solution (when a gas reservoir is being used, this is necessary only for sample weights greater than 0.5 g).Two minutes after setting the end-pointJ switch on the currrent, start the stop-watch simultaneously and check that the current is exactly 20.00 mA (or 10.00 mA if required). Allow the current to flow until the galvanometer spot returns to the marked end-point. Stop the stop-watch and switch off the current at the same time. Read the stop-watch and record the titration time. Remove the combustion boat from the furnace tube. Quantitrator and galvanometer. " Sub-standard" instrument).January, 19711 BY COULOMETRIC TITRATION IN PARTIALLY AQUEOUS MEDIUM 43 Carry out a blank determination with a combustion boat and tin flux. Allow the same interval of time to elapse between the initial end-point adjustment and the end of the titration as for the sample.At low blank levels use a current of 1 or 2 mA for titration. After nine 4-minute titrations at 20 mA it is necessary to change the absorption solution. C.4LCULATION- 12.011 i x t Percentage of carbon =- x ~ 96487 W x 10 - 0.12448 x i x t - w x 104 where i is the current, mA, t is the time, s, and W is the sample weight, g. NOTES- It is convenient to take a sample weight that will give a titration time of 4 minutes (Le., a total time of 6 minutes between the initial setting of the end-point and the end of the titration). If necessary a rough titration can be carried out to determine the sample weight required. 1. However, the following is a guide to sample weights required.Carbon, per cent. Sample weightlg 0.08 and below 1.0 0.08 to 0.10 0.75 0.10 to 0.20 0.50 0.20 to 0.40 0.25 If a titration time of much less than 4 minutes is taken, the current should be switched on a little later to maintain a total time of 6 minutes. 2. Combustion boats should be pre-ignited in a muffle furnace a t 1200OC. They should then be transferred to a desiccator to cool. 3. In a preliminary titration determine the reading of the galvanometer a t which the most rapid change in optical density occurs. Mark this point on a scale and use it as a subsequent end-point. If the galvanometer light spot appears to move irregularly when the current is off, either the position of the gas inlet must be changed or the stirrer speed must be reduced. When the current is on, a stirrer speed that is too low will cause irregular movement.BLANKS Contributions to the blank values were reduced as much as possible as follows. The oxygen stream was purified by passage through a combustion tube heated at 1200" C and then through soda asbestos. Combustion boats were pre-ignited in a muffle furnace at 1200" C. Commercially available low-carbon tin granules wese used. The use of rubber tubing was kept to a minimum by the substitution of glass tubing whenever possible. Ingress of air to the titration cell was prevented by providing a restricted outlet for the oxygen gas. Diffusion of anolyte into the main cell compartment was reduced by fitting a No. 4 porosity sintered-glass disc to the anode arm. Blanks arising from electrical causes were not significant; fatigue of the photocell was checked and found to be negligible, and no leakage occurred from the current source, which was particularly stable.Typical blank values over 6 minutes' titration time were 5 pg of carbon, most of which arose from the boat and tin initiator. Because of the risk of missing the end-point when using a 20-mA current, it was necessary to titrate these low blank tests at a current of 2 mA. This was conveniently achieved by replacing the fixed current source with a dry cell (1.5 V) and an appropriate variable resistor in series. LIMIT OF DETECTION- indicate the limit of detection of the method. Twenty blank determinations were carried out and a standard deviation was calculated to Mean of twenty blank determinations (current of 1 mA) .. Standard deviation, s . . . . . . . . . . . . 7.7 s (= 1 pg of carbon) Limit of detection (3 s = 3 pg of carbon) (0.983 probability) 34.1 s 0.0003 per cent. of carbon (1-g sample)44 REACTIONS- BONIFACE AND JENKINS: DETERMINATION OF CARBON IN STEEL [Autalyst, Vol. 96 DISCUSSION AND RESULTS The following reactions are thought to occur- KI - K+ + I- . . .. .. * ' (1) K++e-K .. .. .. .. * * (2) I-- * I , + e .. .. .. ' (3) K + HOCH,CH2NH2 - KOCH,CH,NH, + 8 H, . . ' (4) (5) CO, + HOCH,CH,NH, - HOCH,CH,-NH I I I c=o .. 0- HOCH,CH,-NH,+ HOCH2CH2-NH I I c=o I + KOCH2CH2NH2 0- HOCH,CH,-NH$ + 2HOCH2CH,NH, + 2HOCH2CH,NH, In reaction (4) the potassium liberated at the cathode reacts with monoethanolamine to form a base. This is supported by the observation that when the monoethanolamine is omitted from the solution, the current efficiency drops from 100 to 64 per cent., showing that a different reaction (probably with dimethylformamide) is taking place.The sodium derivative of monoethanolamine has been used successfully as a base in non-aqueous titration by other workers.17 Reaction (5) has been established by Braid et aZ.,9 who also suggest reactions similar to (6) and (7), but reaction (6) is believed to be the more probable. As the electrolysis takes place both in the presence and absence of water it has been assumed that water does not take part in the reactions mentioned above. APPARATUS AND TITRATION CONDITIONS- The apparatus used in this investigation was based on previous experience with aqueous coulometry (with the exception of the end-point detector).The chief advantages are simplicity and low cost, e.g., the use of a constant current and measurement of time with a stop-watch. The EEL Quantitrator could easily be replaced by a lamp - photocell - galvanometer arrange- ment constructed in the laboratory. It should be possible to provide greater flexibility by using a variable titration current, the amount of electricity being measured by an integrator. This could be designed to give high initial titration rates with end-point anticipation, and the photocell output would provide the necessary signals. However, expensive high-quality instrumentation would be needed t.o avoid loss in precision. In this way an automated rapid control method of high precision could be produced.January, 19711 BY COULOMETRIC TITRATION IN PARTIALLY AQUEOUS MEDIUM 45 ANALYTICAL PERFORMANCE- Table VI gives the results of triplicate determinations of carbon in a variety of British Chemical Standard steels.Table VII gives the results of replicate determinations of carbon in further samples of B.C.S. steels to provide a measure of the reproducibility of the method. The 95 per cent. confidence limits of the method within this laboratory range from +0.0007 per cent. of carbon at 0.048 per cent. of carbon to +O-OOS per cent. of carbon at the 0-9 per cent. of carbon level. Certificate values are given for comparison. The limit of detection calculated from the standard deviation of the blank (Wilson’s method1*) is 0.0003 per cent.on a l-g sample weight. The performance of the method compares well with that of the low-pressure volumetric method, which it is designed to replace in this laboratory. TABLE VI ANALYSIS OF SOME BRITISH CHEMICAL STANDARD STEELS Sample B.C.S. No. Weightlg 431 Mild steel . . .. .. 1 317 3 per cent. Silicon steel . . 1 432 Mild steel . . .. .. 1 239/3 Mild steel . . .. . . 0.35 434 Mild steel , . .. . . 0.20 264/1 Mild steel . . .. . . 0.20 221/1 Mild steel . . .. . . 0.18 Certificate values for carbon, mean and (range), per cent. 0.019 (0-017 to 0.020) 0.028 (0-026 to 0.029) 0.093 (0.091 to 0.096) 0.30 (0.29 to 0.30,) 0.37 (0.36 to 0-38) 0-49, (0.49, to 0.50,) 0.60 (0.59 to 0.62) Carbon found by coulometry, per cent. 0.0185, 0-0184. 0.018 1 0.027 2, 0.027 4, 0.027 1 0.091 7, 0.091 3, 0.091 5 0-292, 0.291, 0.291 0.366, 0-366, 0.365 0.489, 0.492, 0.489 0.594, 0.593, 0.592 TABLE VII REPRODUCIBILITY OF ANALYSES OF SOME BRITISH CHEMICAL STANDARD STEELS B.C.S.sample and certificate values by coulometry, for carbon, per cent. Mean carbon found per cent. 265/2 Mild, 0.048 . . .. 0.047 95 333 Austenitic stainless, 0.066 0.065 2 237/1 Mild, 0.10, . . . . 0.103 4 218/3 Mild, 0.17 . . . . 0.170 8 240/2 Mild, 0.41 0.406 4 220/1 7 per cent. Tungstkn, 0.93 0.919 9 No. of deter- minations 20 20 18 18 20 20 Standard deviation, per cent. 0~000 35 0-000 42 0-000 91 0-000 68 0.002 4 0.004 1 Coefficient of variation, per cent. 0.73 0.65 0.88 0.40 0.59 0.45 CONCLUSIONS A method based on coulometric titration in a partially aqueous medium has been developed for the precise determination of carbon in steel. It is intended for use in reference analysis and involves the use of inexpensive apparatus and a simple technique.However, the application of appropriate instrumentation could provide a method of high precision for routine control purposes. The reproducibility index of the method (20) of 0.0007 per cent. at the 0.05 per cent. of carbon level is comparable with that of the low-pressure volumetric method. As it is based on the fundamental laws of Faraday, no standardisation against analysed samples or pure materials is required. The use of high-vacuum techniques, expensive instrumentation and highly inflammable titrants, characteristic of other methods, is avoided. This paper is published by permission of the Strip Mills Division of the British Steel Corporation.We thank Dr. A. O’Connor for his interest and encouragement, Mr. P. Gale and Mr. P. Hopkins (Port Talbot Works) for useful discussions, and Mr. A. Rees (Electrical Services Port Talbot Works), who constructed the current source. REFERENCES 1. 2. 3. 4. 5. 6. Tipler, G. A., Andyst. 1963, 88, 272. Scholes, P. H., Proceedings of the Nineteenth BISRA Chemists Conference, 1966, p.52. Jones, R. F., Gale, P., Hopkins, P., and Powell, L. H., Analyst, 1965, 90, 623 and 1966, 91, 399. Bagshawe, B., and Pinder, R. H., Ibid., 1958, 81, 153. Wells, J. E., J . Iron Steel Inst., 1950, 166, 113. Cook, R. H., and Speight, G. E., Analyst, 1956, 81, 144.46 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. BONIFACE AND JENKINS Dunnill, P. B., and Kent, A. E., Metallurgia, 1970, 81, 115. Abresch, K., and Lemm, H., Arch. Eisenhiittenw., 1959, 30, 1. Braid, P., Hunter, J. A., Massie, W. H. S., Nicholson, J. D., and Pearce, B. E., Analyst, 1966, Whymark, D. W., and Ottaway, J. M., Proc. SOC. Analyt. Chem., 1969, 6, 186. Johansson, G., Talanta, 1964, 11, 789. Streuli, C. A., Cincotta, J. J., Maricle, D. L., and Mead, K. K., Analyt. Chew., 1964, 36, 1371. Cotman, C., Shreiner, W., Hickey, J., and Williams, T., Tdunta, 1965, 12, 17. White, D. C., Ibid., 1966, 13, 1303. Fritz, J. S., and Gainer, F. E., Ibid., 1968, 15, 939. Snoek, 0. I., and Gouverneur, P., Analytica Chim. Acda, 1967, 39, 463. Doernberg, S., Hubacher, M., and Lysyj, J., J . Amer. Pharm. Ass. Sci. Ed., 1954, 43, 418. Wilson, A. L., Analyst, 1961, 86, 72 and 272. 91, 439. Received May 21st, 1970 Accepted August 17th, 1970 APPENDIX LIST OF COMPONENTS OF CONSTANT-CURRENT SOURCE (Fig. 3) Resistors- R, R, R* R,* Rtl = 22S-2 R, R* R* R, = 7500, = 3 3 k a = 240 k a = 1kQ = 500 (variable) = 2400 = 270S-2 Tyansfomer- 250/125 V Capacitors- CI = 100 pF, 350 V c2 = 25 pF, 25V Semi-conductors- D,, D,, D,, D, = 1B05J400 = ZB6.8 = BF259 = BF259 = 2N698 ZDl TRl TR, TR,

ISSN:0003-2654    

DOI:10.1039/AN9719600037

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Amlyst, January, 1971, Vol. 96, #$. 47-50 47 The Determination of Yttrium, Europium, Terbium, Dysprosium, Holmium, Erbium, Thulium, Ytterbium and Lutetium in Minerals by Atomic-absorption Spectrophotometry BY J. C. VAN LOON, J. H. GALBRAITH AND H. M. AARDEN (Department of Geology, University of Toronto, Toronto 5, Canada) A procedure is described for determining yttrium, europium, terbium, dysprosium, holmium, erbium, thulium, ytterbium and lutetium in zirconium and calcium rare earth silicates by atomic-absorption spectroscopy, which involves the use of a lanthanum suppressor and has potential applications to other minerals containing these interferences. Although many of the associated ingredients are found to interfere, the addition of lanthanum overcomes these problems in many instances.Sensitivities and detection limits found with the proposed procedure are given for each element. EXISTING methods for determining individual rare earths in complex materials are cumber- some and, in general, difficult to carry out. Wet-chemical, emission-spectrographic and spectrophotometric methods1 have been widely used, and recently X-ray fluorescence2 and neutron-activation method^^^^^^^^ have been applied to these determinations. Spectrographic, spectrophotometric and X-ray fluorescence methods require a complex system of standardisa- tion for accurate results. Wet-chemical methods involve complex schemes to separate the rare earths as a group, before isolating each of the individual rare earth elements. Atomic-absorption methods are generally simpler and are often relatively free from spectral interferences.Several workers have studied the atomic-absorption behaviour of the rare earths,' to l5 but, to the authors' knowledge, no method for their determination exists which is accompanied by supporting results. In the most extensive investigation, Kinnunen and Linsjo14J5 studied the atomic- absorption behaviour of all of the rare earths, with the exception of thulium, lutetium and cerium. Hollow-cathode lamps for thulium and lutetium were not available at the time of their study, and cerium determinations were unsuccessful. The same authors recommended the use of a calibration graph prepared from a rare earth matrix solution for accurate work, but no results of determinations were given. Recently, Thomas13 and Fernandez and Manningg outlined the conditions for the application of atomic-absorption spectroscopy to the determination of cerium, thulium and lutetium. In the present work a method is proposed for the atomic-absorption determination of the more sensitive rare earths, including europium.Results are given, which are substantiated by X-ray fluorescence determinations on mineral material. EXPERIMENTAL INTERFERENCES- Table I (see p. 48) summarises a study of the interference effect on the absorbance of a typical rare earth (60 p.p.m. of ytterbium), caused by the commonly associated ingredients of rare earth minerals and rocks. (Other rare earths such as 60 p.p.m. of dysprosium and 100 p.p.m. of holmium gave similar values.) The flame conditions were adjusted to give maximum absorption in each case (with a slightly reducing flame).0 SAC and the authors.48 VAN LOOK et d. : DETERMINATION OF CERTAIN RARE EARTH [Analyst, VOl. 96 TABLE I INTERFERENCE STUDY Interference Fe .. .. K .. .. Mg .. .. Na . . .. h h .. .. Ca . . .. Ti . . .. Other rare earths Al , . .. Zr .. .. HNO, .. .. HCl . . .. .. .. I .. -I .. .. .. .. .. -.{ .. Concentration 500 p.p.m. 500 p.p.m. 5 p.p.m. 50 p.p.m. 500 p.p.m. 6 p.p.m. 50 p.p.m. 500 p.p.m. 500 p.p.m. 500 p.p.m. 500 p.p.m. 200 p.p.m. 10 p.p.m. 100 p.p.m. 500 p.p.m. Absorbance change, per cent. + 20 + 78 0 + 4 + 20 f 7 + 22 + 90 + 20 + 50 - 35 4-13 to 4-21 - 7 - 22 - 60 1000 p.p.m. - 60 Corrective step required Add 1 per cent. of lanthanum Add 1 per cent. of lanthanum and use a burner-to-beam height of about 2 cm As for aluminium and also adjust flame to less reducing conditions until equal absorbances are obtained be- tween zirconium-bearing and zircon- ium-free solutions Standards and samples must have roughly comparable acidities Under the conditions given in Table I the interference arising from a 2-fold excess of silica and an %fold excess of all of the other cations is decreased to a negligible level.One per cent. of lanthanum is the smallest amount that successfully overcomes the interferences; larger amounts have no beneficial effect. N'o other element, not even strontium, was as effective as lanthanum as a suppressor. The additional adjustments required for eliminating zirconium and aluminium interferences result in 15 to 25 per cent., and at least 50 per cent., decreases in the absorbances , respectively, when compared with conditions giving maximum values.METHOD APPARATUS- A Perkin-Elmer Model 303 atomic-absorption spectrophotometer, with a 5-cm nitrous oxide burner, was used. Table I1 gives the wavelength, current and slit widths. Westinghouse hollow-cathode lamps were used for dysprosium, holmium, erbium and ytterbium, and Perkin-Elmer Intenstron lamps for yttrium, europium, thulium and lutetium. TABLE I1 INSTRUMENT PARAMETERS Element Eu Tb Dy Ho Er Th Yb Lu Y Wavelength/nm 459.4 432.6 421.2 410.4 400.8 409.4 398.8 331.2 410-2 Current/mA . . 35 30 15 30 25 30 20 30 30 Slit width/mm . . 0.3 0-3 0.3 0-3 0-3 0-3 0.3 0.1 0.3 REAGENTS- Rare earth stock solutions , 1000 fi.fi.m.-Dissolve accurately weighed, spectroscopically pure, rare earth oxides (obtained from A.D. McKay Co. Ltd.) in 25 ml of 1 + 1 nitric acid, by gently heating. Wash the solutions into l-litre calibrated flasks and dilute to the mark with water. Prepare dilute working solutions to cover a wide concentration range, by diluting the required aliquot with sufficient 10 per cent. lanthanum solution to make the final con- centration of lanthanum 1 per cent., and sufficient nitric acid to correspond with the acid content of the sample.January, 19711 METALS IN MINERALS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 49 Prepare solutions for the interference study from analytical-reagent grade chemicals. Lanthanum rtitrate solution, 10 per cent.-Dissolve a known weight of purified lanthanum oxide, La203, in 1 + 1 nitric acid and dilute with water to give a nitric acid concentration of 1 per cent.and lanthanum content of 10 per cent. Test this solution to ensure that the rare earths are absent. PROCEDURE- The rare earth content of minerals and rocks is usually small and therefore it is often necessary to concentrate the dissolved material in a small volume (5 to 10 ml). The method by which this is done will depend on the composition of the sample, which may give a sample solution containing any of the elements listed above, the interferences from which are over- come with lanthanum. Zirconium rare earth silicates-The sample (0.5 g) is treated with hydrofluoric acid to effect decomposition and to precipitate the insoluble-fluoride group.The soluble zirconium fluoride is then removed by filtration. After ignition to give the oxides, aided by the addition of a few drops of sulphuric acid, the insoluble material is fused with potassium pyrosulphate and dissolved in dilute acid. The hydroxides are precipitated with ammonia solution and treated as indicated below. A direct determination of the rare earths without removal of zirconium can be carried out: dissolution of the sample with hydrofluoric acid is followed by heating it to fumes with sulphuric acid to remove fluoride. The resulting precipitate is dissolved in dilute acid and after adding lanthanum the solution is measured by atomic-absorption spectroscopy under the conditions given in Table I. A loss of at least 50 per cent. in the absorbance makes this alternative less attractive when sensitivity is important. Calcium rare earth silicates-The sample (0.5 g) is decomposed by fusion with carbonate, followed by precipitation of the hydroxide in the dilute acid solution of the fusion mixture.The hydroxide precipitate is treated as indicated below. Other rare earth silicates-A suitable accurately weighed amount (1 to 10 g) of sample is dissolved by an established procedure.2J6J7~~~ d9 s20 One precipitation, or a combination of fluoride, hydroxide and oxalate precipitations, depending on the nature of the sample, is used to remove the rare earth group from the high salt content of the dissolution mixture. In the final step a precipitate of hydroxide is obtained and treated as indicated below. Determination of rare earths in the hydroxide preci$itate-Dissolve the precipitate in hot dilute nitric acid in a 50-ml beaker.Wash it thoroughly with hot 1 + 50 nitric acid. Add sufficient lanthanum to give a final concentration of 1 per cent., and evaporate the solution to about 2 ml. Transfer the mixture with washing into a small calibrated flask and dilute to volume with water. Run the solution against standards for each element containing a similar amount of nitric acid and 1 per cent. of lanthanum. Although these steps appear to make the procedure unnecessarily cumbersome, in practice they take a relatively short time and greatly increase the reliability. Element (as oxide) Tb02 .. DY20, * . Ho203 .. Tm203 .. Yb203 . . .. Lu203 .. Eu203 .. y2°3 .. .. .. .. .. .. .. .... .. TABLE I11 RARE EARTH DETERMINATIONS Zirconium silicate r by X-ray 0-04 0.31 0.09 0.35 0.06 0.44 0.06 0.01 3.50 Estimated standard deviation 0-004 0.006 0-004 0-007 0.006 0-008 0.006 0.00 1 0.07 by atomic absorption* 0.05 0.28 0.07 0.32 0.05 0.4 1 0.06 0.01 3.60 I Standard deviation 0.007 0.012 0.002 0.017 0.006 0.002 0-006 0.001 0.03 Calcium rare earth silicate - by by atomic X-rayt absorptiont 0.38 0.29 2.14 2-40 0.41 0.39 1.01 0.97 0.13 0.12 0.66 0.68 - 0.10 0.01 0.0 1 10.3 10.0 * Mean of three determinations. t Enough sample for only one determination.50 VAN LOON, GALBRAITH AND AARDEN [Autalyst, VOl. 96 RESULTS AND DISCUSSION A zirconium silicate and a calcium rare earth silicate were subjected to the proposed procedure, and the results obtained are given in Table 111.The results are compared with those obtained by X-ray fluorescence on similar materials. The complete atomic-absorption method for the elements determined takes 3 to 4 hours, compared with about 16 hours by X-ray fluorescence. The precision of the two techniques is about the same, although the sensitivity of the X-ray fluorescence method is greater for lutetium, terbium and dysprosium, and is comparable for the others. Serious rare earth interelemental interference occurs in the europium, holmium, ytterbium and lutetium determinations by X-ray fluorescence. SENSITIVITY AND DETECTION LIMIT- which are compared with those recorded by Slavin.21 Table IV gives the sensitivities and detection limits obtained by the proposed procedure, TABLE IV DETECTION LIMITS AND SENSITIVITIES Element Terbium .. .. .. Dysprosium . . .. Holmium .. .. .. Erbium . . .. .. Thulium . . .. .. Ytterbium . . . . .. Lutetium . . .. .. Europium . . .. .. Yttrium . . .. .. Sensitivity, p.p.m. per 1 per cent. absorption Present method Slavin r A \ 10 7.5 1.0 0.7 2.0 1.4 0.8 0.85 1.0 1.0 0.3 0.17 0-7 0.75 1.8 1.1 30 15 Detection limit, p.p.m. r A > Present method Slavin 2 2 0.5 0.4 0.3 0.3 0.4 0.1 0-3 0.15 0.1 0.04 0-3 0.2 10 3 - - Determination of lutetium is extremely difficult because it occurs in very small amounts in natural materials, and because it is relatively insensitive to atomic-absorption spectroscopy. Work is continuing on the atomic-absorption and atomic-fluorescence determination of the lighter, less sensitive, rare earths.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. REFERENCES Vickery, R. C., “Analytical Chemistry of the Rare Earths,” Pergamon Press, Oxford, London, Aarden, H. M., Ph.D. Thesis, University of Toronto, 1969. Massart, D. L., and Hoste, J., Analytica Chim. Acta, 1968, 42, 7. -- , Ibid., 1968, 42, 166. Am;s, M. D., and Willis, J. B., Spectrochim. Acta, 1966, 22, 1325. Skogerboe, R. K., and Woodruff, R. A., Analyt. Chem., 1965,34, 1977. Fernandez, F., and Manning, D. C., Atomic Absorfition Newsletter, 1968, 7, 67. Jaworowski, A. J., Weberling, R. P., and Bracco, D. J., Analytica Chim. Acta, 1967, 37, 284. Mossotti, V. G., and Fassel, V. A,, Spectrochim. Acta, 1964, 20, 1117. Joyner, T., Healy, M. L., Chakravarti, D., and Koyanagi, F., Environ. Sci. Tech., 1967, 1, 417. Thomas, P. E., Resonance Lines, 1969, 1, 6. Kinnunen, J., and Lindsjo, O., Chemist Analyst, 1967, 56, 25. Kinnunen, J., and Wennerstrand, B., Ibid., 1967, 56, 24. Cruft, E. F., Ingamells, C. O., and Muysson. J.. Geochim. Cosmochim. Ada, 1965, 29, 581. Hughson, M. R., and Sen Gupta, J. G., Amer. Miner., 1964, 48, 937. Kolthoff, I. M., and Elving, P. J., Editors, “Treatise on Analytical Chemistry,” Part 11, Volume 8, Interscience Publishers, a division of John Wiley & Sons, New York and London, 1963, p. 3. Eyring, L., Editor, “Progress in the Science and Technology of the Rare Earths,” Volume 1, Per- gamon Press, Oxford, London, New York and Paris, and The Macmillan Company, New York, 1964. Slavin, W. W., “Atomic Absorption Spectroscopy,” Interscience Publishers, a division of John Wiley & Sons, New York, London and Sydney, 1968. Received August llth, 1969 Accepted July 14th, 1970 New York and Paris, 1961. , Ibid., 1968, 42, 15. , Ibid., 1968, 42, 21. , -- , -- , , Ibid., 1967, 56, 76. --

ISSN:0003-2654    

DOI:10.1039/AN9719600047

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, January, 1971, Vol. 96, p p . 51-54 51 The Determination of Fluorine in Rocks and Minerals by a Pyrohydrolytic Method BY R. L. CLEMENTS, G. A. SERGEANT AND P. J. WEBB (Defiartment of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, S.E. 1) A rapid pyrohydrolytic method, with simple apparatus, is described for the determination of fluorine in rocks and minerals. The sample is heated with a three-component flux in a stream of moist air, and the liberated hydrogen fluoride is absorbed into an alkaline solution. The recovered fluorine is determined either colorimetrically or by means of a fluoride-specific electrode. The method is suitable for determining fluorine a t concentrations down to about 50 p.p.m. THE distillation method of Evans and Sergeant1 for the determination of fluorine in rocks and minerals has been shown to be reliable in routine work at this laboratory over a period of years.It is, however, a comparatively long procedure, and the possibilities of a rapid pyrohydrolytic method have been under consideration for some time. In particular, among recent methods of this type, that described by Newman2 for aluminosilicate analysis has now been examined in detail and, with modifications, forms the basis of our proposed method. EXPERIMENTAL Certain alterations were made to the apparatus as used by Newman, the most important of which concerned the outlet condenser and the method of introducing water into the reaction tube. The condenser was found not to be necessary (or desirable) with the reduced air flow of our method, as condensation was apt to prevent complete fluorine recovery.In place of the steam generator required by Newman’s system, water is contained in a nickel boat placed in the reaction tube and heated by a burner. This simple arrangement ensures that sufficient water vapour is present during each determination. Pure vanadium pentoxide, as used by Newman, was found to be an effective flux for most of the rock samples tested, with a few exceptions. In particular, a specimen of micaceous material (muscovite - tremolite schist) failed to yield the expected recovery of 0.40 per cent. of fluorine, and variation of the sample-to-flux ratio did not lead to significantly improved recoveries. Next, mixed fluxes were considered, and a combination of one part of bismuth trioxide with two parts of vanadium pentoxide gave improved recoveries of fluorine, but these appeared to be critically dependent on the degree of grinding of flux with the sample.The addition of sodium tungstate as a third component gave a still more active flux so that fluorine was recovered quantitatively from the micaceous material. In order to avoid spitting of the flux during pyrohydrolysis, it was necessary to prepare pre-fused flux from a mixture of the components. The dense ground flux powder was also an excellent material for further grinding with sample powder. In order to standardise the sample-to-flux ratio, further determinations of the fluorine content of the micaceous schist were made on 200-mg portions mechanically ground with various weights (100,200,400 and 800 mg) of the pre-fused flux. Complete recoveries of fluorine were obtained with all but the smallest weight of flux.To examine the effect on the pyrohydrolysis of varying the amount of water used, the operation was performed successively with 2, 4, 6 and 8 ml of water in the nickel boat. A mechanically ground mixture of the sample of micaceous schist with two parts by weight of three-component flux was used in these tests, which showed complete recoveries of fluorine except when the smallest amount of water was used. The time taken, from the time of lighting the burners, for the water to evaporate completely was 10 to 25 minutes for 2 to 8 ml, respectively. 0 SAC; Crown Copyright Reserved.52 CLEMENTS, SERGEANT AND WEBB: DETERMINATION O F FLUORINE [AIz&?"st, VOl.96 The reaction temperature in the ignition tube under working conditions was measured with a thermocouple positioned in place of the sample, and was found to be 700 to 750 "C. For the purpose of this investigation, the fluoride contents of the absorbent solutions were measured by the fluoride-specific electrode and by the lanthanum - alizarin fluorine blue colorimetric method as described by Greenhalgh and Riley.3 Good agreement was observed between the two methods over a wide range of fluoride contents. METHOD APPARATUS- The transparent fused silica tube* A (Fig. 1) is similar to that used by Newman and is 450 mm long with an internal diameter of 20 mm and a wall thickness of 2 mm.It is heated by three Bunsen burners, B, C and D, the first two of which are fitted with flame-spreader attachments. The silica boat F (50 x 15mm) contains the mixture of sample and flux. The nickel boat G (100 x lOmm, capacity approximately 8ml) is filled with water for each determination and placed near the inlet end of the tube. Compressed air, after being filtered through a cotton-wool plug, passes into the reaction tube by way of the flow meter E and the emergent air stream bubbles through the alkaline absorbent solution contained in a polythene bottle of approximately 50-ml capacity. Electrical measuring equipment-This comprised an "Ionalyzer" fluoride-activity elec- trode, Model 9449, together with the associated specific-ion meter. Spectrophotometer-A Unicam SP600 instrument was used.REAGENTS- All reagents should be of analytical-reagent grade when possible. Preparation of flux-In a nickel crucible mix 5 g of bismuth trioxide, 5 g of sodium tungstate and l o g of vanadium pentoxide. Heat the mixture to fusion over a burner, and pour the melt on to a silica plate, then break up the cooled melt and grind it to powder. Sodium hydroxide solutiout-A 0.2 M aqueous solution. Neutralising bufler solution-Dissolve 15 g of sodium acetate trihydrate and 5 ml of glacial acetic acid in water, and dilute the solution to 100 ml. Lanthanum - alizarin JEuorine blue reagent-This is prepared as described for solution A in the method of Marshall and Wood.4 Standard juoride solutiofi-Dissolve 0-1106 g of dry sodium fluoride in water, and dilute the solution to 500 ml.This solution contains 100 pg ml-l of fluorine. From it prepare, by dilution with water, a dilute standard solution containing 4pg ml-1 of fluorine. G I F A I Fig. 1. Details of apparatus * A suitable tube was made to order by Jencons Ltd., Hemel Hempstead, Herts.January, 19711 IN ROCKS AND MINERALS BY A PYROHYDROLYTIC METHOD 53 PROCEDURE- Mix 0400 g of sample powder with 1.00 g of flux in a mechanical mortar (preferably of agate), then grind the mixture for 30 minutes. Transfer the ground mixture, which is sufficient for duplicate determinations, to a stoppered specimen tube. Transfer 0.600 g of the mixture to the silica boat and place this in position in the reaction tube over burner B (Fig. 1). Fill the nickel boat to the brim with water, and place it in the reaction tube on the inlet side of burner C, as shown in the diagram.The position should be such that about 25 minutes are required for total evaporation of the water after lighting the burners. Prepare the absorbent mixture in the polythene bottle by introducing 5 ml of sodium hydroxide solution and 20 to 25 ml of water, and place the bottle so that the exit end of the reaction tube is just below the surface of the liquid. After connecting the air supply and adjusting the flow-rate to 10 litres per hour, light the two outer burners D and C, then, 2 minutes later, the central burner B. At this point the air flow-rate may need re-adjustment. Sample Felsite . . .. Diabase W-1 .. Tonalite, T-1 . . Granite, G-1 . , Biotitic green schist Porphyritic basalt Alkali-granite .. Muscovite - tremolite Granite . . .. .. .. .. .. .. .. .. .. .. TABLE I DETERMINATION OF FLUORINE IN ROCKS Fluorine found, per cent. f A Deviation Pyrohydrolytic method standard C E l e c t r o 7 d e Distillation method finish finish method from None B G None A G None E G None A G None None B A G None C G None D G None None G 0.01 1 0.01 1 0-013 0.023 0.023 0.025 0.047 0.046 0-045 0.065 0.065 0.063 0.073 0.073 0.073 0.073 0.07 1 0.127 0-127 0-126 0.181 0.181 0.178 0.395 0.395 0.395 0.01 1 0.012 0.011 0.01 1 0.022 0.023 0.022 0.023 0-048 0.046 0.047 0.046 0-065 0,063 0.065 0.064 0.072 0.074 0-074 0.074 0-074 0.072 0.130 0.120 0-129 0.126 0-181 0.180 0.183 0.180 0.400 0.400 0.400 0.395 None 1-16 1.18 1.1 1 F, B 1.16 1.18 G 1-14 1.15 Other figures - 0-025* 0-045t 0.08 0.05 0*07* * From the compilation by M.Fleisher.6 t From the publication “Standard Geochemical Sample T-1,” 1961, Geological Survey Division, A: Air flow-rate a t 5 1 hour-1. B: Air flow-rate at 20 1 hour-1. C, D, €3, F: Volume of water added to nickel boat reduced to 5 ml. Time for total evaporation G: Sample and flux mixed and ground by hand in an agate mortar for a total time of 10 minutes. Thirty minutes after lighting burner B remove the polythene bottle, turn off all the burners, and rinse the tip of the outlet tube with water into the polythene bottle. Allow the absorbent mixture to cool, add 5 ml of neutralising buffer solution, then quantitatively Ministry of Commerce and Industry, Tanganyika. of water and total period of pyrohydrolysis, in minutes, was 17(C), 10(D), 15(E) and 13(F).54 CLEMENTS, SERGEANT AND WEBB transfer the liquid to a 50-ml graduated flask and dilute to the mark.Determine the fluorine content of the solution either by using the fluoride-specific electrode, or the colorimetric method, as described below. Ion-sfiecific electrode Jinish-Prepare a set of fluoride standards in 50-ml graduated flasks by adding to each, by pipette, 5ml of sodium hydroxide solution and 5ml of neutralising buffer solution, together with aliquots of standard fluoride solution equivalent to 0.002, 0.010, 0.100 and 1.00 per cent. of fluorine in the original material. Dilute each standard to 50 ml and store in polythene bottles. With the ion-specific electrode, compare the test solution with the standards according to the manufacturer’s instruct ions.Colorimetric $finish-Transfer an aliquot of the test solution, containing not more than 20pg of fluorine, to a 25-ml graduated flask. (A 5-ml aliquot is suitable if the fluorine content of the original material does not exceed 0.1 per cent.) Add, by pipette, 10ml of lanthanum - alizarin fluorine blue reagent, dilute to 25 ml and mix. After 30 minutes measure the optical density at 625 nm of the solution contained in a l-cm cell with air as reference. Ascertain the fluorine content by reference to a calibration curve covering the range 0 to 20 pg of fluorine in 25ml of solution. Results obtained by either method should be corrected for the small blank, which is determined by completing the whole procedure without sample material. RESULTS The proposed pyrohydrolytic method was applied to a number of silicate rocks with a wide range of fluorine contents previously established by the distillation procedure; the results are shown in Table I. Also shown are the results of experiments intended to demon- strate that the method is tolerant to the effects of fairly wide deviations from the standard procedure in a number of respects, and some figures obtained by other workers. This paper is published by permission of the Government Chemist and the Director of the Institute of Geological Sciences. REFERENCES 1. 2. 3. 4. 5. Evans, W. H., and Sergeant, G. A., Analyst, 1967, 92, 690. Newman, A. C. D., Ibid., 1968, 93, 827. Greenhalgh, R., and Riley, J. P., Analytica Chim. Acta, 1961, 25, 179. Marshall, B. S., and Wood, R., Analyst, 1969, 94, 493. Fleisher, M., Geochim. Cosmochim. Acta, 1969, 33, 65. Received J d y 22nd, 1970 Accepted Awgust 21st, 1970

ISSN:0003-2654    

DOI:10.1039/AN9719600051

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, January, 1971, Vol. 96, $@. 55-58 55 A Simple Method for the Determination of Solvents Retained in Plastic Films and Laminates BY J. T. DAVIES AND J. R. BISHOP (The Metal Box Company Limited, Research and Development Department, Kendal Avenue, Westfields Road, Acton, London, W.3) A simple method is described for the determination of solvents retained in polymer films and laminates as a result of the application of inks, coatings or adhesives. A sample of the film or laminate is heated, together with a known amount of a suitable solvent as an internal standard, in a simple modified Kilner jar. After a period of heating in an oven the atmosphere in the headspace of the jar is analysed by gas chromatography with a flame- ionisation detector. Results given show the precision of the method and its application to actual samples.Replicate analyses on similar samples show a standard deviation of 5 per cent. of the mean value for each solvent determined. The method is currently used in the authors' laboratories for analysing samples containing retained solvents a t levels between 1 and 300 mg m-2. COATINGS, inks and adhesives are usually applied to polymer films as solutions in or mixtures with organic solvents. After their application the film is passed through a drying oven, but small amounts of solvent may diffuse into the film and be retained by it. It is important to know the amounts retained, as these can relate to the odour of the film when used to package foodstuffs, etc. For the past 10 years, these solvents have been determined in our laboratories by gas chromatography after a carrier distillati0n.l Application of this method requires considerable expertise and recoveries are low when it is used with certain films that have come into common use since its inception.Several alternative methods have been described recently. Martin2 heated a relatively large area of film, in a converted vacuum oven and passed a sample of the vapour through a gas-chromatographic detector. Susukida and Okuma3 heated a small sample with a carrier liquid in a sealed tube and then analysed the carrier liquid. Allavena and Grassi4 heated a sample of film under an unmeasured reduced pressure and have described a way of introducing an internal standard. These methods are not sufficiently sensitive or simple for our requirements.Wilks and Gilbert5 heated a sample of film under a measured reduced pressure and analysed the atmos- phere in the headspace of the vessel by gas chromatography. The method described below can be regarded as a logical development of the method of Wilks and Gilbert. The apparatus, operations and calculations have been changed and simplified, and the use of an internal standard has been introduced to improve the precision. An experiment described demonstrates that it is possible to show whether the recovery of each solvent from each film or laminate is complete. EXPERIMENTAL APPARATUS AND REAGENTS- Kilner jars (1 Zb)-These jars have metal lids. A brass &inch pipe union is sealed with Araldite into a hole drilled through each lid; this firmly holds a rubber septum.Jars and lids can be used many times; after use they can be cleaned by being placed separately in an oven at 100" C for half 'an hour. Gas chyomatogra@h with a pame-ionisation detector-This can be fitted with any column that will satisfactorily separate the various solvents from each other and from the internal 0 SAC and the authors.56 [Analyst, Vol. 96 standard. We have found that a suitable liquid phase for our purpose is silicone oil modified by the addition of a small amount of UCON HB2000 (polyalkylene glycol). The proportions used were 9 parts of silicone oil, 2 parts of UCON HB2000 and 100 parts of Chromosorb W. Internu2 standard-This can be any solvent that is known to be absent from the sample material. We found butyl acetate or butyl propionate to be suitable for most of our analyses.DAVIES AND BISHOP: A SIMPLE METHOD FOR THE DETERMINATION PROCEDURE- About 250 cm2 of material are crumpled loosely, placed in a clean Kilner jar and the jar is closed. Internal standard (3.00 yl) is injected into the jar through the rubber septum by using a 10-y1 syringe. The jar is then placed in an oven at 100" C for the minimum specified period (see below), at the end of which it is removed from the oven and about 2 ml of the atmosphere within are immediately extracted with an all-glass unlubricated' syringe and injected into the gas chromatograph. The ratios of the peak heights of the solvents (a, b, c, etc.) to that of the internal standard are calculated. The same amount of internal standard as used with the samples, plus 2.00 yl of each of the solvents a, b, c, etc., is introduced, for calibration purposes, into each of two more Kilner jars.These jars are heated together with those containing samples. The content of retained solvent, e g . , a, is given in milligrams per square metre by the expression- Peak height of solvent a in sample jar Peak height of internal standard in sample jar 20,000 x specific gravity of solvent a Peak height of solvent a in calibration jar Area of sample in square centimetres Peak height of internal standard in calibration jar The minimum heating period for each material is determined as follows. The film or laminate sample and internal standard are placed in a Kilner jar as described above and the jar is placed in an oven at 100" C; it is removed from the oven at 10 or 15-minute intervals and 2-ml samples are withdrawn and chromatographed. The jar is replaced in the oven immediately after each 2-ml sample has been taken.The ratios of the peak heights of solvents a, b, c, etc., to that of the internal standard are calculated. The minimum heating time is that after which continued heating produces no further change in the ratios. The method as described above is applicable for retained solvent levels between 1 and 300 mg m-2. The linearity of the amplified response of the flame-ionisation detector must be shown to be adequate throughout the range of attenuation settings used. The following to actual samples. PRECISION IN THE RESULTS results show the precision of the method and illustrate its application PREPARATION O F CALIBRATION STANDARDS- Seven jars were prepared, each containing 2.00 pl of toluene and 3.00 p1 of butyl acetate (our usual internal standard).Each jar was heated in an oven for a minimum of 46 minutes. Peak height ratio, toluene/internal standard 2.12, 2.12, 2.08. 2.06, 2.09, 2.10, 2.11 Mean 2.10 Standard deviation 0.02 DETERMINATION OF MINIMUM HEATING TIME- This experiment was carried out to ascertain the minimum heating time for the deter- mination of toluene in a laminate of polythene and polypropylene double-coated with saran [poly(vinylidene chloride)]. Time in ovenlminutes . . .. 15 25 35 45 55 80 90 160 185 210 4.2 6.1 7.0 8.1 8.3 8.7 8.8 8-9 8.8 8.8 toluene internal standard Peak height ratio, . From these figures it is deduced that a suitable minimum heating time would be 90 minutes.January, 19711 OF SOLVENTS RETAINED IN PLASTIC FILMS AND LAMINATES 57 REPEATED DETERMINATIONS ON ACTUAL SAMPLES- The following results were obtained for the determination of toluene and ethyl acetate on adjacent pieces of laminate (polythene adhesive-laminated to polypropylene double-coated with saran).The solvents originate from the adhesive. Mean s.d. Toluene/mgm-2 . . 127 135 119 127 140 122 130 119 127 7 Ethylacetate/mgm-2.. 136 145 124 141 136 133 138 128 135 6 Results as follows were obtained for the determination of toluene on adjacent pieces of printed film (polypropylene with a single saran coating). The toluene originated from the printing ink. Toluene/ mg m-2 . . . .11-9, 11-4, 12.3, 11.9, 12-5, 11-9, 12.3 Mean 12.0 mg m--2 Standard deviation 0.34 mg m--2 RECOVERY EXPERIMENTS- It is not possible to carry out direct recovery experiments because films or laminates cannot at present be prepared with known amounts of solvent evenly distributed throughout the material. However, one can say that after the minimum heating period the solvents and internal standard are partitioned between the polymer and vapour phases in proportions that remain unchanged during further heating; this indicates that one can conduct the following experiment with each solvent and each film to estimate the percentage recovery obtained in an actual determination. Two sets of calibration standards in Kilner jars are prepared, one normal and the other containing a sample of film; both sets of jars are then heated for the minimum heating time for that solvent and film.The peak height ratios of solvent to standard are measured. Any differences in ratio between the two sets of jars are caused by a preferential absorption of one compound by the film. A comparison of these ratios is a measure of the recovery of that solvent during an analysis of that film. It might be suggested that calibration jars should always contain a sample of the film under analysis, but it is seldom that such a sample known to be solvent free is available. In the example given below two sets of calibration jars were prepared, of which only one set contained in each jar 250 cm2 of polythene laminated to polypropylene double-coated with saran- Peak height ratio, toluene/internal standard Ca1ib;ation standards without Calibration standardHwith laminate laminate 1.82 1.65 1.82 1.66 1.77 1.60 1-86 1.73 Mean .. 1.82 Mean .. 1-66 These results show that for the determination of toluene in this particular laminate the recovery is greater than 90 per cent. Two solvents that have different physical properties from those of toluene are butanol and 2-ethoxyethanol. Experiments have shown that 100 per cent. of butanol is recovered from polythene and 95 per cent. from polypropylene double-coated with saran and that 85 per cent. of 2-ethoxyethanol is recovered from polypropylene double-coated with saran. In recovery experiments of this sort it is theoretically possible for recoveries to be over 100 per cent., which could occur when more of the internal standard than of the solvent under test is absorbed into the film.Such recovery figures show by exactly how much one is over- estimating this solvent during an actual analysis. DISCUSSION The method described has been devised to determine the amount of solvent retained in a film or laminate. It is currently used in the range 1 to 300 mg m-2, but the range could easily be extended at either end. Although it is still impossible to carry out true recovery experiments, an estimate of recovery can be made by the described procedure. The apparatus has been deliberately kept simple, and 100°C was chosen as a suitable temperature to which58 DAVIES AND BISHOP [Analyst, Vol. 96 the adapted Kilner jars could be repeatedly heated. We have shown that at 100°C the mini- mum heating time is reasonably short, being not more than 2 hours and often much less.This method has been used to determine many different solvents in several different substrates. The solvents include ethanol, ethyl acetate, ethyl methyl ketone, 2-ethoxyethanol, propan-1-01 and toluene. The substrates include polythene, polypropylene and cellophane, which occur individually, coated with saran or combined in laminates. Fig. 1 (a and b) shows typical chromatograms obtained from a film and a calibration jar. 4 0 5 4 0 Timelminutes I. Injection mark 4. Butanol 2. Ethanol 5. Toluene 3. Ethyl methyl ketone 6. Butyl acetate (internal standard) Fig. 1. (a) Chromatogram of solvents from a sample of printed polythene - polypropylene laminate.Chromatograph, Pye 104 with flame-ionisation detector; column, glass, 6 feet x 5 inch, packed with 9 per cent. silicone oil and 3 per cent. UCON HB2000 on Chromosorb W; oven, 80’ C; attenuation x 20,000; and chart speed, 1 cm minute-l. (b) Calibration jar prepared for sample shown in (a) Although it is possible, by using a large number of Kilner jars, to keep the gas chromato- graph running continuously and thus analyse, say, between six and twenty samples per hour, there is still a delay between receipt of a sample and the completion of the analysis. Most of this delay is caused by the heating time required. It may well be possible, by using a vessel more robust than a Kilner jar, to heat the sample to a temperature just short of polymer decomposition and so reduce the heating period substantially. In two of the examples given reference was made to polythene laminated to polypropylene double-coated with saran. Tkis laminate was chosen for these examples because the authors have previously found it to be difficult to analyse. REFERENCES 1. Bishop, J. R., “Proceedings of Conference on Odour in Packaging,’’ The Institute of Packaging, London, 1960, p. 184. 2. Martin, C. N., Chem. & Ind., 1969, 454. 3. Susukida, W., and Okuma, J., “Gas Chromatographic Measuring of Residual Solvents in the Printed Films,” Dai Nippon Printing Company, Tokyo, 1968. 4. Allavena, S., and Grassi, P., Boll. Laboratori Chim. Prov., 1969, 20, 8. 5. Wilks, R. A,, and Gilbert, S. G., Muter. Res. Stand., 1968, 8, 29. Received June 5th, 1970 Accepted Augzcst 19tit, 1970

ISSN:0003-2654    

DOI:10.1039/AN9719600055

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Artalyst, January, 1971, Vol. 96, @. 59-61 59 The Microgasometric Determination of Some Inorganic and Organic Nitrates by Reduction with Iodide Ion and Elemental Iodine BY S. S. M. HASSAN (Research Microanalytical Laboratories, Chemistry Defiartment, Faculty of Science. Ain Shams University, Cairo, U.A.R.) Simple micro methods are described for the determination of nitrate based on its reduction with iodide or iodine in the presence of halogen acids. Nitric oxide gas is liberated and iodine(1) is the oxidation product from both iodide and iodine. METHODS for the determination of the nitrate group have been reviewed in a previous work describing the microgasometric determination of nitrate by reduction with iron(I1) , titan- ium(III), mercury and hydr0quinone.l However, reduction with iron(I1) and mercury1 is not specific for nitrate salts as nitramine and nitrate esters are also reduced.The mercury method has the further limitation that any aromatic compounds present are preferentially nitrated, thus giving low results for nitrate ion. Although reduction of nitrates with titan- ium(1II)l is specific, the preparation and storage of the reagent are tedious. The present work shows that iodide and iodine can be used instead of these reductants. The nitramine group and nitrate esters are not reduced under these conditions and aromatic substances do not interfere. EXPERIMENTAL APPARATUS- As previously described.l s2 PROCEDURE- Introduce 3 to 5 mg of the nitrate sample into the reaction vessel, and add 20 to 30 mg of potassium iodide.Displace the air in the apparatus with a stream of carbon dioxide gas at the rate of 100 bubbles minute-l for 5 minutes, or until no air bubbles are collected in the nitrometer. Add 3 to 5 ml of 40 per cent. hydrobromic acid or 35 per cent. hydrochloric acid by using the funnel. Gently heat the reaction mixture for 5 minutes until no further gas bubbles are evolved. Sweep the gaseous products with carbon dioxide through the trap, containing 20 per cent. tin(I1) chloride solution in concentrated hydrochloric acid, and collect the nitric oxide over freshly prepared 50 per cent. potassium hydroxide solution in the nitro- meter. Carry out a blank experiment. Reduction with iodine is carried out in the same way, but in this event only hydrobromic acid can be used as reaction medium. CALCULATION- (V - V ) (P - $) x 273 x 14.01 (273 + t) x W x 224 x 760 Nitrate-nitrogen, per cent.= where V is the volume of nitric oxide, v is the blank, both in ml, P is the atmospheric pressure, mm Hg, $ is the water vapour pressure, mm Hg, at temperature t, which is the average room temperature and also that of the potassium hydroxide solution contained in the nitrometer, and W is the weight of sample, mg. As the reaction time is short, no correction for the solubility of nitric oxide is required when fresh potassium hydroxide solution is used in the nitrometer. 0 SAC and the author.60 HASSAN : MICROGASOMETRIC DETERMINATION OF INORGANIC AND ORGANIC [Analyst, Vol. 96 RESULTS AND DISCUSSION NATURE OF THE REACTION- Various mole ratios of potassium nitrate and potassium iodide were tried, and the degree of reduction was calculated from the volume of nitric oxide gas produced.Quantitative reduction occurred with 1-5 mole of iodide per mole of nitrate showing that the reaction is- 2N0,- + 31- + 8H+ + 6C1- = 2NO + 4H20 + 3Ic1,-. This was confirmed by carrying out reduction of potassium nitrate with various mole ratios of iodine in hydrochloric acid. Three moles of iodine were the minimum amount required for reduction of 2 mole of nitrate according to the equation- ZN0,- + 31, + 8H+ + 12C1- = 2N0 + 4H20 + 61C12-. The nitric oxide was identified by the formation of brown fumes when it was exposed to air and by its complete absorption in acidic iron(I1) sulphate solution. The aqueous solution of reduction products decolorises indigo carmine and litmus solutions , thus confirming the presence of iodine mon~chloride.~ The mean recovery of the nitric oxide on simple sweeping and collection in the nitrometer was 91.0 per cent.These low results are attributed to volatilisation of the iodine with dissolution of the vapour in the potassium hydroxide to form hypoiodite which, in turn, reacts with the nitric oxide. Attempts were made to absorb the iodine in traps containing alkaline sodium hydrogen sulphite solution, carbon tetrachloride, starch solution and tin(I1) chloride solution. The results obtained (Table I) show that 20 per cent. tin(I1) chloride solution is the most suitable. TABLE I OF NITROGEN) BY REDUCTION WITH POTASSIUM IODIDE AND HYDROCHLORIC ACID MICROGASOMETRIC DETERMINATION OF POTASSIUM NITRATE (CONTAINING 13-85 PER CENT.AFTER IODINE ABSORPTION Average nitrate-nitrogen found, Average recovery, Iodine absorption solution per cent. per cent. - 12.6 91.0 Sodium hydrogen sulphite, 35 per cent. . . 13.0 94.1 Starch solution, 20 per cent. .. .. 13.4 97.0 Carbon tetrachloride.. . . .. .. 13.6 98.2 Tin(I1) chloride, 20 per cent. . . .. 13.8 99.9 DETERMINATION OF NITRATE- Several nitrate samples were analysed by reduction with potassium iodide in the presence of hydrochloric and hydrobromic acids (Table 11). In either medium, quantitative liberation of nitric oxide gas occurred with acid concentrations in the range 6 to 11 N. The maximum deviation was kO.1 per cent. absolute. With iodine as reductant (Table 111) similar results were obtained in hydrobromic acid, but unsatisfactory results were obtained in hydrochloric acid medium.TABLE I1 MICROGASOMETRIC DETERMINATION OF SOME NITRATE SAMPLES BY REDUCTION WITH POTASSIUM IODIDE IN HYDROCHLORIC AND HYDROBROMIC ACID MEDIA Reduction Nitrate-nitrogen in hydrochloric acid. calculated, Nitrate-nitrogen Sample per cent. found, per cent. Potassium nitrate . . . . 13.85 13.9 13.8 Bariumnitrate .. . . .. 10.71 10.7 10.6 Ureanitrate . . . . .. 11.38 11.4 11.3 Guanidine nitrate . . .. 11-47 11.4 11.4 Nitronnitrate . . . . . . 3.73 3.6 3.7 Reduction in hydrobromic acid, Nitrate-nitrogen found, per cent. 13-8 13.8 10.8 10-7 11.4 11.3 11.6 11.4 3.8 3.7January, 19711 NITRATES BY REDUCTION WITH IODIDE ION AND ELEMENTAL IODINE 61 TABLE I11 MICROGASOMETRIC DETERMINATION OF SOME NITRATE SAMPLES BY REDUCTION WITH IODINE IN HYDROCHLORIC AND HYDROBROMIC ACID MEDIA Sample Potassium nitrate Barium nitrate .. Urea nitrate . . Guanidine nitrate Nitron nitrate . . .. .. .. .. .. .. .. .. .. .. Nitrate-nitrogen calculated, per cent. 13-85 10.71 11.38 11.47 3.73 Reduction in hydrochloric acid. Nitrate-nitrogen found, per cent. 12.8 12.6 9.8 9.9 10.7 10.5 10.3 10.4 3.0 3.2 Reduction in hydrobromic acid. Nitrate-nitrogen found, per cent. 13.8 13.7 10.7 10-7 11.2 11.3 11.5 11.4 3-7 3-8 In general, the nitrate was reduced more easily if hydrobromic acid was used instead of hydrochloric acid. Sulphuric acid could not be used because of its reduction to hydrogen sulphide and sulphur. Attempts to use potassium bromide in presence of hydrochloric or hydrobromic acid proved to be unsuitable for quantitative reduction of nitrate as only 70 per cent. of the theoretical nitrate-nitrogen was recovered as nitric oxide gas. REFERENCES 1. 2. 3. Awad, W. I., and Hassan, S. S. M., Talanta, 1969, 16, 1393. Hassan, S. S. M., Mikrockim. Ada, in the press. Mellor, J., “A Comprehensive Treatise on Inorganic and Theoretical Chemistry,” Longmans, Received December 301h, 1969 Accepted July 28th, 1970 Green & Co. Ltd., London, 1946, Volume 11, p. 118.

ISSN:0003-2654    

DOI:10.1039/AN9719600059

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

62 Analyst, January, 1971, Vol. 96, pp. 62-66 A Field Method for the Determination of Iron Oxide Fume in Air BY D. W. MEDDLE AND R. WOOD (Department of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, S.E.l) A field method is described for the determination of iron oxide fume in industrial atmospheres at concentrations up to 20 mg m-3. After collection on a filter-paper the iron oxide is dissolved in a hot solution of hydroxylam- monium chloride in hydrochloric acid. The iron solution is quantitatively transferred to a calibrated flask and the red complex of iron(I1) - batho- phenanthrolinedisulphonate is formed, the intensity of which is determined either spectrophotometrically or by visual comparison with a set of permanent colour standards.The procedure is simple to carry out and the time required for a complete analysis is about 25 minutes. IRON oxide fume occurs in many industrial processes that involve the melting of iron or steel, e.g., steel manufacture and the welding of iron or steel products. Although generally considered to be only a nuisance,l as it reduces visibility and causes damage by soiling, iron oxide occurs widely in industry and has been assigned2 a threshold limit value of 10 mg m-3. Consequently, and also to make possible the identification of an unknown fume as iron oxide, it appeared that there was a requirement for a simple, rapid field test for the determination of this material in air at its threshold limit value. A review of the literature revealed that no such test had previously been developed.Reynolds and Monkman3 have described a method for the determination of iron in dustfall samples, but their procedure involves wet ashing and solvent-extraction steps that would greatly limit its use as a field test. EXPERIMENTAL COLORIMETRIC DETERMINATION OF IRON- It was envisaged that any field method devised would involve the collection of the iron-containing fume from the test atmosphere on a suitable filter, dissolution of the iron from the filter and subsequent colorimetric determination with an optional visual or spectro- photometric finish. For use in the colorimetric procedure the widely used reagent for iron, 4,7-diphenyl-1 ,lo-phenanthr~line~ (bathophenanthroline), was chosen. Its main advantages are its specificity for iron(II), adequate sensitivity, rapid formation of the coloured iron bathophenanthroline complex and the excellent visual colour change from colourless to red.By using standard iron solutions, an initial study was made of a bathophenanthroline procedure described previ~usly,~ which was satisfactory when used by a skilled operator, but a disadvantage in its use as a field test was the need to extract the iron complexwith chloroform. Consequently, attention was turned to the water-soluble disodium salt of bathophenanthrolinedisulphonic acid,5 the use of which obviated the need for the extraction of the coloured complex with an organic solvent. It had been shown5 that this reagent has a similar sensitivity to bathophenanthroline for the determination of iron(I1) and was not more subject to interference from other chemical species.We therefore decided to determine the various optimum reaction parameters required for the use of this reagent in the proposed field test. Concentration of disodium bathophenanthrolinedisulphonate reagent-The combining ratio of the iron(I1) - bathophenanthrolinedisulphonate ion had previously been found experi- mentally6 to be 1 : 3.17. On this basis, the presence of 2 ml of a 0.1 per cent. w/v aqueous solution of the reagent in a final reaction solution volume of 25 ml should have been adequate for determining up to 66 pg of iron(I1). In practice, the maximum amount of iron(I1) capable of being determined by this system was only 60 pg, equivalent to 86 pg of iron oxide (Fe,O,), but this was considered adequate for the purposes of the proposed field test.0 SAC; Crown Copyright Reserved.MEDDLE AND WOOD 63 Optimum PH coditions and colour stability-Although it had previously been shown5 that the red complex of iron(I1) and the bathophenanthrolinedisulphonate ion can be formed over a pH range from 2.6 to 9, it was found in the present work that full and rapid colour development (within 5 minutes) commenced at pH 3.7, which was attained by the addition of sodium acetate solution to the acidic solution of the iron oxide fume sample (see Removal of fume samples from filter-papers). Tests showed that the presence of between 6 and 10 ml of a 16.7 per cent. w/v aqueous solution of sodium acetate trihydrate in the final reaction solution of 25ml stabilised the pH at between 3-7 and 4-2 and yielded constant absorb- ances for a known amount of iron.To obviate the effect of possible accidental addition in the field of excess of the acid used for dissolution, it was decided to use 10 ml of the buffering acetate solution. The colours produced by this sytem were found to be stable for at least 1 hour. To reduce the number of solutions required it was found possible, as suggested previously,6 to combine the aqueous hydroxylammonium chloride solution (about 10 per cent. w/v) necessary to maintain the iron as iron(I1) in the reaction solution with the sodium acetate solution. Reagent blank-The iron contamination in several commercial, analytical-grade hydro- chloric acids was found to be minimal. Although the quoted maximum limits of iron impurity in the various analytical-reagent grades of hydroxylammonium chloride and sodium acetate trihydrate would have produced reagent blanks above those allowable in the proposed method, tests showed that all of these reagents were free from iron contamination.The maximum, reagent blank for the complete procedure was always less than 1 pg of iron which, on the basis of a proposed 2.5-litre sample (see Sample size and visual colour standards), was equivalent to less than 0.6 mg m-3 of iron oxide in air, a value sufficiently low to be of no consequence in the proposed field test. In practice, a reagent blank of up to the equivalent of 1 mg m-3 of iron oxide in air would be allowable in a visual determination of iron oxide, and would entail a 20 per cent.over-estimation of the iron oxide fume in an atmosphere of half the present threshold limit value. Sertsitivity of the method-By using the parameters established above and a standard iron solution, a calibration graph, which was linear, was constructed over the range 0 to 50 pg of iron (0 to 70 pg of iron oxide). The optical densities of the various solutions were read in a 20-mm cell against a reagent blank at 538 nm, the readings for 5 and 70 pg of iron oxide being 0.11 and 1.53, respectively. Sample size and visual colour startdards--It was necessary to assess the most convenient sample size to be taken, particularly when using visual colour standards. A set of standards representing an acceptable reagent blank and 0.5, 1 and 2 times the present threshold limit value for iron oxide in air (5, 10 and 20 mg m-3) was required.A suitable maximum standard was considered to be 50 pg of iron oxide. Consequently, as a sampling rate of at least 500 ml minute-1 was considered necessary, this would involve sampling 2.5 litres of an atmosphere. By using a standard iron solution a set of standards was prepared to satisfy the above requirements, the respective colours of the solutions being easily differentiated visually when viewed through about 10mm of liquid, and this was selected for use. However, to avoid the use of liquid colour standards in the field, a set of standard discs was prepared, with the co-operation of Tintometer Ltd. , representing the intensity of colours produced by collecting TABLE I DETERMINATION OF 10 pg OF IRON (ABOUT THE EQUIVALENT OF AN ATMOSPHERE CONTAINING HALF THE THRESHOLD LIMIT VALUE OF IRON OXIDE SAMPLED BY THE PROPOSED METHOD) MAXIMUM LEVELS TESTED OF VARIOUS CHEMICAL SPECIES SHOWING NO INTERFERENCE WITH THE Chemical species Manganese (Mn2+) .. Phosphate . . Vanadate (VOs-) . . Fluoride (F-) . . .. Zinc (Zna+) . . .. Tin (Sna+) . . .. Nickel (Niz+) . . .. Titanium (Ti4+) . . .. Tungstate (WOq2-) . . Chromium (Crs+) . . .. .. .. .. .. .. .. .. .. .. Weight of interfering specieslpg 1000 1000 1000 1000 1000 200 50 50 20 2064 MEDDLE AND WOOD: A FIELD METHOD FOR THE [Analyst, Vol. 96 26litre samples of 1 (reagent blank), 5, 10 and 20 mg r r 3 atmospheres of iron oxide fume when viewed through a solution thickness of 13.5 mm (the internal diameter of glass tubes used for colour comparisons).With these standards it was possible to determine the iron oxide fume content of an atmosphere to the nearest k2.5 mg m-3 between 0 and 10 mg ~ n - ~ , and at least to the nearest 5 mg 111-~ between 10 and 20 mg m-3. Inter ferences-The effects on the proposed colorimetric method of several chemical species that might occur together with iron oxide in industrial atmospheres were examined by adding known amounts of these species to a solution containing 10 pg of iron (approximately equivalent to a 2.5-litre sample of an atmosphere containing half the present threshold limit value of iron oxide) prior to determining the latter. Under the test conditions the species listed in Table I did not interfere, at least up to the level indicated.These results concurred with a previous study,5 which included, among others, several of the species examined in the present work. The ratios (w/w) of the interfering species to iron in Table I were already well in excess of those at which the interfering species, on the basis of their respective threshold limit values,2 became more of a hazard than the iron oxide. SAMPLING AND COLLECTION OF IRON OXIDE FUME- Choice of Jilter-paper-On the basis of published information,' filters such as the Millipore, Type AA, were selected as the most suitable for the collection of iron oxide particulates; they had also previously been showns to have virtually a 100 per cent. collection efficiency for all particulates down to a size of 8 nm when sampling is carried out a t a face velocity of less than 0.4m s-l.In the proposed test with an effective collecting area of 255 mm2 and a sampling rate of 500 ml minute-l, the face velocity across the filter is only 0-033 m s-1. Removal of fume samples from Jilter-papers-Iron oxide fume atmospheres were prepared by atomising aqueous iron(II1) citrate solutions in the apparatus to be described el~ewhere.~ The use of a 10 per cent. w/v iron(II1) citrate solution gave an atmosphere containing about 3.0 mg m-3 of iron oxide fume but, as was also foundg in the preparation of a zinc oxide fume, this concentration was neither reproducible from one run to another nor constant during any one run. A series of iron oxide fume samples (sources) was collected on Millipore filters and the response of each with non-destructive X-ray fluorescence spectrometry was noted.These sources were then used in investigations to find a suitable method of dissolving the collected fume in the field, the amounts of iron remaining on the filter-papers after trials with various methods being also assessed by X-ray fluorescence spectrometry. Several dissolution agents were tried as follows: 50 per cent. v/v hydrochloric acid alone and with each of tin(I1) chloride, iodine, hydrogen peroxide, manganese dioxide and hydroxyl- ammonium chloride. The last, consisting of 2 ml of a 1 + 1 v/v mixture of 50 per cent. v/v hydrochloric acid and 20 per cent. w/v aqueous hydroxylammonium chloride solution, showed promise as over 90 per cent. of the iron oxide on a source was dissolved after a 15-minute extraction on a hot-plate.As it was difficult to reproduce the exact temperature with a hot-plate, a steam-bath was adopted as a standard source of heating; by using this extraction of thirteen iron oxide fume samples (ranging from 5 to 45 pg) with the above dissolution agent gave an average recovery of 96-3 2.9 per cent., which appeared to be independent of extraction times between 5 and 15 minutes; a 10-minute extraction period TABLE I1 SOLUBILITY IN SUCCESSIVE EXTRACTIONS WITH HYDROCHLORIC ACID - HYDROXYLAMMOKIUN CHLORIDE SOLUTION* OF IRON OXIDE FUME SAMPLES TAKEN IN THE VICINITY OF INDUSTRIAL WELDING Iron found/pg Total iron in \ 1st extraction, I A Sample 1st extraction 2nd extraction 3rd extraction per cent. 1 25.0 1.0 0.0 96.2 2 30.0 0-8 0.0 97.4 3 35.0 0.7 <0*1 97.8 4 63.8 1.7 0.0 97.4 5 180.5 4.3 0.0 97.7 * One millilitre of each of 50 per cent.v/v hydrochloric acid and 20 per cent. w/v aqueous solution of hydroxylammonium chloride.January, 197 11 65 was chosen. The best extraction results were obtained when the two components of the dissolution agent were combined immediately before use. It was noted with this procedure that the number of micrograms of iron recovered from the various samples of collected fume plotted against the respective X-ray fluorescence responses, in counts s-l, gave a straight line. The above method of dissolution proved satisfactory for iron oxide fume formed in the burner of the generat~r,~ but it was found that iron oxide fused at 1000” C for 8 hours was more difficult to dissolve as only about 75 per cent.of it was soluble under the proposed conditions. A heating time of 40 minutes was required for 95 per cent. dissolution. However, it was considered that this would be an extreme case and that iron oxide fume formed in industry, in either welding or casting operations, would not be so highly fused. To check this view, fume samples were collected in the vicinity of industrial welding operations and the extent of their dissolution was examined in the laboratory; each sample was successively treated three times by the proposed dissolution method and the iron extracted in each step was determined separately. Table I1 lists the results obtained, which show that simply one extraction would be adequate to dissolve iron oxide fume samples collected in the field.APPARATUS- DETERMINATION OF IRON OXIDE FUME IN AIR FIELD METHOD FOR THE DETERMINATION OF IRON OXIDE FUME IN AIR Filter-paper holder-To hold filter-papers 25 mm in diameter. Filter-pa$ers-Millipore, Type AA, 25 mm in diameter. Sampling PumP-A pump capable of drawing air through the filter-paper in the holder at a constant rate of 0.5 1 minute-l. (This flow-rate can be achieved by using a suitable critical orifice in conjunction with a pump capable of producing an orifice pressure differential of at least 200mm of mercury. Alternatively, sampling can be carried out with a pump, by monitoring and controlling the flow-rate with a flow-meter and control valve, respectively.) Spectrophotometer or calorimeter-An instrument capable of measuring the optical densities of solutions at 538nm.Colour standards-A comparator disc, containing coloured glass standards for this test, for use with the Lovibond “1000” comparator is obtainable from Tintometer Limited, Salisbury. Four standards are provided representing 1,5,10 and 20 mg m--3 of iron oxide in air. Glass tubes for colour comparison-These were of 13.5 mm i d . and 10-ml capacity (Tinto- meter Limited supply pairs of tubes suitable for use in conjunction with the Lovibond “lOOO’J comparator). NOTE- REAGENTS- All reagents should be of analytical-reagent grade when possible and all solutions prepared with distilled or de-ionised water. All reagents should be stored in iron-free glass or polythene bottles and kept stoppered except when in use.Hydrochloric acid (1 + 1 v/v)-Dilute concentrated hydrochloric acid with an equal volume of water. Hydroxylammonium chloride solution, 20 per cent. w/v, aqueous. Bufler solution-Dissolve 33.3 g of hydroxylammonium chloride and 60 g of sodium acetate trihydrate in 360 ml of water. Disodium 4,7-di9 henyl- l,lO-~henanthrolinedisul~ honate solution-Dissolve 100 mg of the reagent in 100ml of water. Standard iron solution-Dissolve 0.28 g of precipitated iron powder (or electrolytic iron wire) in 20 ml of hydrochloric acid (1 + 1 v/v) and dilute to 1 litre with water (solution A). To 50 ml of solution A, add 20 ml of hydrochloric acid (1 + 1 v/v) and dilute to 1 litre with water (solution B). To 100 ml of solution B, add 4 ml of hydrochloric acid (1 + 1 v/v) and dilute to 200 ml with water (solution C).Solution C contains 7-0 pg ml-l of iron, i.e., the equivalent of 10.0 pg ml-l of iron oxide (Fe,O,). All glassware should be rendered free from iron by washing it with concentrated hydrochloric acid. NOTE- When using visual colour standards for determining the iron oxide content it is important that the reagent blank should be low. If it has a coloration more intense than that of the 1 mg m-s standard, discard the solutions and prepare again with fresh reagents.66 MEDDLE AND WOOD PROCEDURE- Place a filter-paper in the filter holder, attach the assembly to the pump and draw a sample of air through the paper a t a rate of 0.5lminute-l for 5 minutes. Disconnect the holder from the pump, remove the filter-paper and place it in a small beaker (25-ml) of diameter not less than 25 mm.Add 1 ml of hydrochloric acid (1 + 1 v/v) and 1 ml of the hydroxyl- ammonium chloride solution, cover the beaker with a watch-glass and place it on a steam-bath. After 10 minutes remove the beaker from the bath, transfer the acidic solution to a 25-ml calibrated flask and wash the filter-paper thoroughly with 10ml of water. Add 2ml of disodium 4,7-diphenyl-l,l0-phenanthrolinedisulphonate solution followed by 10 ml of the buffer solution. Dilute to 25 ml with water, mix well, allow to stand for 5 minutes for maximum colour development and determine the iron as iron oxide either visually or spectrophoto- metrically as described below. VISUAL DETERMINATION OF IRON OXIDE- Fill one colour comparison tube with the reaction solution and another with water.Insert both tubes into the comparator and, by using the comparator disc, ascertain the nearest colour match between sample and standards while viewing through the sample. SPECTROPHOTOMETRIC DETERMINATION OF IRON OXIDE- Measure the optical density of the reaction solution in a 20-mm glass cell at 538 nm against a reference solution prepared from all of the reagents used. Determine the amount of iron oxide in the solution by reference to the calibration graph. Calculate the concentration of iron oxide in the sample of air, in milligrams per cubic metre, by dividing the total iron oxide in micrograms by 2.5. Preparation of calibration graph-To a series of 25-ml calibrated flasks add 0, 1, 2, 3, 5 and 6 ml of the standard iron solution C.To each flask add 1 ml of hydrochloric acid (1 + 1 v/v), 1 ml of the hydroxylammonium chloride solution, 2 ml of disodium 4,7-diphenyl- 1,lO-phenanthrolinedisulphonate solution and 10 ml of buffer solution. Dilute each to 25 ml with water, mix well, allow to stand for 5 minutes and then measure the optical densities of the solutions in 20-mm cells at 538nm, with the solution to which no iron was added as reference. Construct a graph of micrograms of iron oxide (Fe,O,) against optical density. APPLICATION AND SCOPE OF THE METHOD- Although specifically designed for the determination of iron oxide fume in the field, the above method is versatile and can be adapted, with minor modifications, to the field or labora- tory determination of other iron-containing species, including metallic iron, in the form of fume or dust, or both.Laboratory tests showed that the dusts of particle size less than 45pm (respirable range) of both iron oxides (Fe203 and Fe,O,) can be determined. The method can also be used t o determine the concentration of water-soluble iron salts (as iron) in air, a threshold limit value (1 mg m-,) for these having been first established2 in 1969. Provided a sample of exactly 17-5 litres is taken at a rate between 0.5 and 2.5 1 minute-l, then the same set of visual colour standards, i.e., the comparator disc, in this case representing 0.1, 0.5, 1.0 and 2.0 mg m-3 of soluble iron salts (as iron), respectively, can be used. This work was carried out on behalf of the Department of Employment and Productivity Committee on Tests for Toxic Substances in Air. We thank the Government Chemist for permission to publish this paper, and H.M. Factory Inspectorate for arranging the field tests. 1. 2. 3. 4. 6. 6. 7. 8. 9. REFERENCES Community Air Quality Guide for Iron Oxide, Amer. Ind. Hyg. Ass. J., 1968, 29, 4. “Threshold Limit Values 1969,” Technical Data Note 2/69, Department of Employment and Reynolds, R. G., and Monkman, J. L., Amer. Ind. Hyg. Ass. J . , 1962, 23, 416. Smith, G. F., McCurdy, W. H., jun., and Diehl, H., Analyst, 1952, 77, 418. Blair, D., and Diehl, H., Tulunta, 1961, 7, 163. Johnson, W. C., Editor, “Organic Reagents for Metals,” Volume 11, Hopkin and Williams Ltd., Chadwell Heath, Essex, 1964, p. 65. Farrah, G. H., J . Air Pollut. Control Ass., 1967, 17, 738. Megaw, W. J., and Wiffen, R. D., Int. J . Air Wat. Pollut., 1963, 7 , 501. Marshall, B. S., Telford, I., and Wood, R., in preparation. Productivity, London, 1969. Received July 6th, 1970 Accepted July 20th, 1970

ISSN:0003-2654    

DOI:10.1039/AN9719600062

出版商:RSC    

年代:1971

数据来源: RSC