ISSN:0003-2654    

DOI:10.1039/AN97196FX029

出版商:RSC    

年代:1971

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97196BX031

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

August, 1971 SUMMARIES OF PAPERS I N THIS I S S U ESummaries of Papers in this IssueObservations on the Theory of Action of Visual IndicatorsEarlier treatments of visual indicator theory have made the tacitassumption that the molar absorptivities of the two forms of the indicatorare equal, so equating the conditional indicator constant with the transitionpoint and hence causing some confusion in certain instances. A more exacttreatment of oxidation - reduction and ion-combination indicator parameters,conditional constant, transition point, transition range and transition interval,including the effects of molar absorptivities, indicator concentration andstoicheiometry is outlined.E. BISHOPChemistry Department, University of Exeter, Stocker Road, Exeter, EX4 4QD.Analyst, 1971, 96, 537-549.An Automatic Capillary ViscometerPart 11.Automatic Apparatus for Viscometric TitrationsThe basic principles of viscometric titrations are considered and examplesof applications to avariety of analytical and other determinations are suggested.An addition to the automatic capillary viscometer described in Part Iprovides for the introduction of successive measured volume increments ofa solvent or solution, admixture with the solution already present in a sus-pended-level viscometer and the electronic measurement of the flow timeincluding printing-out the results.By using the new apparatus i t is possible expeditiously to determinethe B-coefficients of the Jones - Dole equation for the viscosity of electrolytesand the intrinsic viscosity coefficients for solutions of polymers, and to performa wide variety of acid - base titrations.Practical examples of these deter-minations are given in illustration.R. B. SIMPSON, J. S. SMITH and H. M. N. H. IRVINGDepartment of Inorganic and Structural Chemistry, The University, Leeds 2.Analyst, 1971, 96, 550-561.An Inert Dilution Method for the X-ray Fluorescence Analysisof Niobate - Tantalate Mineral ConcentratesAn X-ray fluorescence method, in which iron(II1) oxide is used as inertdiluent, is described for the determination of niobium, tantalum, tin andtitanium, in concentrates of polymineral composition. No time-consumingfusion procedures or tedious calculations are involved. The simplicity andrapidity of the method merit its use in routine analysis.Good agreementwith chemical analysis is obtained.Y. C. WONG and S. SEEVARATNAMGeoIogical Survey of Malaysia, Perak, Malaysia.Analyst, 1971, 96, 562-564.Determination of Ammonium in Soil Extracts by anAutomated Indophenol MethodAn automated method is described for the determination of ammoniumin 2 N potassium chloride soil extracts. The determination is based on theindophenol blue method following a dialysing step. As little as 0-03 mg 1-1of ammonium-nitrogen can be determined. The recoveries have been investi-gated for a number of different soil types and satisfactory results wereobtained. Determinations can be carried out a t the rate of thirty samplesper hour.Agricultural College of Norway, Vollebekk, Norway.A.R. SELMER-OLSENAnalyst, 1971, 96, 565-568,vi THE ANALYST [August, 1971These are used as oil-soluble standards in the spectrographicanalysis of traces of metals in oils and fats, in petroleum derivativesand in lubricating agents.The petroleum industry first felt the need for the evaluationof the metallic content both of the Crude oil and of the products obtained insubsequent stages of purification.Later the need extended to the motor industry for the study oflubricants and of the behaviour of materials,The analysis of metals in non-aqueous media is carried out withspectrographs and atomic absorption spe ctrophotometers using samples of knowncontent as controls. Therefore it was necessaryto study and develop organometallic com pounds and organic saltsof metals, having a known metal content.The stability is obtained by the use of solubilisingagents such as 2-Ethylhexanoic acid, 6-Methyl-2,4-heptandione, 2-Ethyl-hexylamine,and bis-(2-Ethylhexyl)dithiocarbamic acid-bis-(2-ethylhexyl)ammonium salt, with Xylene.Thus, clear at-@ ,stable solutions in an oil base are obtained, with concentrationsup to 500 ppm of metal.It is also possible to prepare .solutions containing more than one metal, bearing in mind that mixturesof metals are more soluble than the individual constituents.The Carlo Erba RS metallorganic standards are available in 5 g. vials as follows:Aluminium cyclohexanebutyrate RSAluminium 2-ethylhexanoate RSBarium cyclohexanebutyrate RSCadmium cyclohexanebutyrate RSCalcium 2-ethylhexanoate RSChromium (Ill)-tris-(2’-hydroxyacetophenone) RSCobalt cyclohexanebutyrate RSCopper (11)-bis-benzoylacetonate RSCupric cyclohexanebutyrate RSDibutiltin bis(2-ethylhexanoate) RSFerric cyclohexanebutyrate RSIron (Ill)-tris-benzoylacetonate RSLead cyclohexanebutyrate RSLithium cyclohexanebutyrate RSMagnesium cyclohexanebutyrate RSManganous cyclohexanebutyrate RSMenthyl borate RSNickel cyclohexanebutyrate RSOctaphenylcyclotetrasiloxane RSSilver cyclohexanebutyrate RSSilver 2-ethylhexanoate RSSodium cyclohexanebutyrate RSStrontium cyclo hexanebutyrate RSVanadyl-bis-benzoylacetonate RSZinc cyclohexanebutyrate RS( CABLO ERBA’] DlVlSlONE CHIMICA INDUSTRIALE I VIA C. IMBONATI 24 i 20159 MILAN

ISSN:0003-2654    

DOI:10.1039/AN97196FP125

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

viii SUMMARIES OF PAPERS I N THIS ISSUEA Field Method for the Determination of Zinc Oxide Fume in AirA method is described for the determination of zinc oxide fume inindustrial atmospheres at concentrations up to 20 mg m-3 of zinc oxide. Thefume is collected on a filter and dissolved in acid, and the zinc is determinedspectrophotometrically or visually with 4-(2’-thiazolylazo)resorcinol reagent.The apparatus used is simple and the time required for a determination isabout 20 minutes. A dynamic method for the generation of atmospheresof zinc oxide is also described.[August, 1971B. S. MARSHALL, I. TELFORD and R. WOODDepartment of Trade and Industry, Laboratory of the Government Chemist,Cornwall House, Stamford Street, London, S.E. 1.Analyst, 1971, 96, 569-578.Polarographic Determination of Uranium in Monazite SandsA practical method is described for the extraction of uranium frommonazite sands and its subsequent polarographic determination.The sample is decomposed by fusion with potassium hydrogen di fluoridefollowed by fusion with potassium pyrosulphate.Uranium is extracted fromnitric acid solution with tributyl phosphate in 2,2,4-trimethylpentane, withaluminium nitrate as a salting-out agent, and back-extracted from the organicphase with water.The final polarographic determination is carried out in 2 M acetic acid -2 M ammonium acetate - 0.1 M ascorbic acid solution as supporting electrolyte.Neither a maximum suppressor nor removal of oxygen is needed.The interference by lead and some factors influencing the extractionof uranium are studied.The results are reproducible and agree with those obtained by other,more laborious, techniques. The proposed procedure is suitable for thedetermination of uranium in monazites and monazite sand concentratescontaining not less than 0.005 per cent. of uranium oxide, and is superiorin speed, reliability and convenience to other methods previously reported.R.W. MARTRES and J. J. BURASTEROAdministracibn Nacional de Combustibles, Alcohol y Portland, Centro de Investi-gaciones Tecnol6gicas, Pando, Uruguay.Analyst, 1971, 96, 579-583.A Modified Spectrophotometric Method for the Determination ofAmmonia (and Amino-acids) in Natural Waters, withParticular Reference to Sea WaterStrickland and Parsons have described a method for determining ammonia(and amino-acids) in sea water.This method, in which ammonia is firstoxidised to nitrite with sodium hypochlorite, was found to be very timeconsuming and often imprecise. The results of a critical examination of theunderlying chemistry have led to a modified procedure in which the timerequired for analysis has been reduced from 3.5 hours to 17 minutes. Themodified procedure gave a coefficient of variation of 1.6 per cent. from theanalysis of thirteen replicate aliquots of a sample of sea water containingapproximately 20 pg 1-1 of ammonium-nitrogen, and hence is suitable fordetermining ammonia concentrations in the order of 0 to 200 pg 1-1. Theresponse for 1 pg 1-1 of ammonium-nitrogen, in either sea water or de-ionisedwater, was equivalent to an optical density change of 1.6 x per cmof solution.The procedure is also suitable for the quantitative determinationof ammonia (and amino-acids) in other natural waters.VICTOR W. TRUESDALEEast African Marine Fisheries Research Organization, P.O. Box 668, Zanzibar,East Africa.Analyst, 1971, 96, 584-590x SUMMARIES OF PAPERS I N THIS ISSUEPhotometric Determination of Small Amounts of Oxygen inWater with 3,3’- DimethylnaphthidineThe oxidation products of manganese obtained in Winkler’s classicalmethod for the determination of oxygen can be determined by a photometricmethod based on their reaction with 3,3’-dimethylnaphthidine in an acidicmedium. This reagent is used in preference to o-tolidine because of its highermolecular weight, which results in greater sensitivity, and in the intensered - violet colour of the oxidised form of 3,3’-dimethylnaphthidine.[August 1971HUBERT FADRUS and JOSEF MALY-Central Sewage Treatment Plant, Brno-Modi-ice, Czechoslovakia.Analyst, 1971, 96, 591-597.Rapid Determination of Solvent Occluded in SomePharmaceutical ChemicalsAn infrared spectrophotometric method has been developed for thedetermination of occluded solvent in orciprenaline siilphate and warfarinsodium.In lesser detail ampicillin sodium, colchicine, the calcium and sodiumsalts of novobiocin and streptomycin sulphate have also been examined.THOMAS R. LOWTHER and W. D. WILLIAMSDepartment of Pharmaceutical Chemistry, University of Strathclyde, Glasgow.Analyst, 1971, 96, 598-600.The Gas - chromatographic Determination of Arsanilic AcidMethods are described for the determination of arsanilic acid and carb-arsone in animal feeding stuffs. These additives are extracted from the feedwith an aqueous solvent. Carbarsone is converted by hydrolysis with sodiumhydroxide into arsanilic acid, which is then reduced with Raney alloy to aniline.The aniline, separated from the reaction mixture by steam distillation andsolvent extraction, is determined by gas chromatography with flame-ionisationdetection.R. E. WESTON, B. B. WHEALS and M. J. KENSETTDepartment of Trade and Industry, Laboratory of the Government Chemist,Cornwall House, Stamford Street, London, S.E. 1.Analyst, 1971, 96, 601-603.and Carbarsone in Animal Feeding Stuff

ISSN:0003-2654    

DOI:10.1039/AN97196BP129

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

AUGUST, 1971 THE ANALYST Vol. 96, No. I145 Observations on the Theory of Action of Visual Indicators BY E. BISHOP (Chemistry Department, University of Exeter, Stocker Road, Exeter EX4 4QD) Earlier treatments of visual indicator theory have made the tacit assumption that the molar absorptivities of the two forms of the indicator are equal, so equating the conditional indicator constant with the transition point and hence causing some confusion in certain instances. A more exact treatment of oxidation - reduction and ion-combination indicator parameters, conditional constant, transition point, transition range and transition interval, including the effects of molar absorptivities, indicator concentration and stoicheiometry, is outlined. VISUAL indicators, with few exceptions, share the nature of the titrimetric reaction to which they are applied. Basically, there are only two types of reaction: oxidation - reduction, in which electrons constitute the common factor, and ion-combination, in which main and indi- cator reactions share a common ion.To conduct visual indicator titrations intelligently, knowledge is required of the equivalence point, pB, or potential, and the quality, Q,l of the titration, and the conditional constant, transition point, transition range and transition interval of the indicator. Earlier accounts of indicator theory2y3 have neglected the possibility that the molar absorptivities of the two forms of the indicator may be different, and so have equated indicator constant and transition point. The approximation is often adequate in practice, but has led to confusion when the disparity in molar absorptivities is large, for example, in the case of tris(l,10-phenanthroline)iron(II)/(III),4 where the ratio of molar absorptivities is 20: 1, and while the formal potential in 1.0 M sulphuric acid is 1.06 V, the transition point is 1.14 V.The influence of indicator concentration for unsymmetrical stoicheio- metries is also demonstrated, and some further points on choice of indicators are made. Un- symmetrical stoicheiometries are rare among oxidation - reduction indicators, but not un- common among ion-combination indicators ; the unsymmetrical case is, however, generally treated first and the solution then simplified to the more common symmetrical case, in accordance with current physicochemical practice.Derivations of equations are given in the Appendix. OXIDATION - REDUCTION INDICATORS The indicator reaction can be represented as a Indo, + nind e + b Indred . . .. .. (1) and the few that are properly reversible show Nernstian potentials- The two forms, Ind,, and Indred, display different signals of colour, fluorescence, etc. When both forms are present at unit concentration or, if a = b, at equal concentration, then the logarithmic term becomes zero and Eind = EAind, whefe Eiind is the conditional potential, the indicator constant for this class of indicator. The conditions implied are those of the titration to which the indicator is to be applied, and the measurement of ELlnd made under those conditions, and the value of Eiind applies to those precise conditions alone and will change if the conditions are changed.TWO-COLOUR INDICATORS- Transition point-For a two-colour indicator the transition point is the potential at which the two colours have equal intensity. If the oxidised and reduced forms of the indicator have molar absorptivities of cox and €red, respectively, then from Beer’s law the two colours are of equal intensity when €ox [Indoxltrans = €red [Indred] trans .. .. * (3) 537538 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS [Analyst, VOl. 96 .. .. .. . [Indoxltrans €red [Indred]trans COX - The transition point potential will therefore be, when a = b, Etrans = Eiind + - 2.3 RT €red = Eiind +- * - (5) 2.3 RT [Indoxltrans nind F loglo [Indred] trans nind F loglo 6; Hence the indicator constant and transition potential are equal only when eOX = €red.When a # b, and reaction (1) is unsymmetrical because of a change in molecularity, then the system becomes dependent on concentration and also on the values of the stoicheio- metric coefficients. To minimise the complexity of the expressions, let the total concentration of the indicator at the appropriate point in the titration (here the transition point) be ex- pressed in terms of the form in which the indicator is added (here the reduced form, Cindred). This concentration is calculated from the amount of indicator added and the total volume of the solution at the appropriate point in the titration. Thus, if Vind ml of Indred of concen- tration Cind mol 1-1 are used, the sample volume is V d ml and the titrant volume is Zttrans ml, then (total volume) = uind + Vd + Vtrans and Cindred = Cind Vind/(tOtal Volume).The transi- tion point potential is If the indicator is added in the oxidised form to give a total concentration of Cindox, substitution of this value and the coefficient, b, in the numerator of the finallogarithmic term inequation (6) gives the transition potential The transition range of a two-colour indicator is subjective in that the human eye does not perceive all tints equally and the visual response varies among individuals. Properly, therefore, the transition range should be determined experimentally by the individual, but the traditional convention that when colour 1 has ten times the intensity of colour 2 the indicator appears to be completely of colour 1 can be applied to assess the transition range.Thus, when cox [Indo,] = 10 €red [Indred] the indicator appears to the eye to be completely oxidised, and when 10 cox [Index] = €red [Indred] the indicator appears to be completely reduced. With this substitution made in equation (2) for the case where a = b, the indicator changes colour over the range of potentials When a # b, the transition range, from Elow to Ehigh, is dependent on the indicator concentration, the stoicheiometry and further on the molar absorptivities- (9) Rigorously, Cindox is different at Elow and Ehigh because of the additional volume of titrant consumed in traversing from one point to the other. The difference is small in practice, and it is sufficient in a subjective appraisal of this kind to use the equivalence point volume of titrant in calculating Cindox when the indicator is a valid choice for the titration.If the indicator is added in the reduced form, substitution of (a Cindred) for (b Cjndox) in the third term of Elow and of Ehigh gives the corresponding values. An equation of the form of equation (8) cannot therefore be written, because the third terms in equation (9) differ fromAugust, 19711 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS 539 the third terms in equation (6) or (7). The transition range is not symmetrically disposed around the transition potential. The transition interval, which defines the sharpness of the colour change, is the difference between the upper and lower potentials of the transition range. For symmetrical indicator reactions (a = b) this is independent of concentration and molar absorptivities, but depends primarily on nind, and to a very small degree on temperature.By expressing the interval as AE V. .. .. .. .. .. 2.3 RT A E = 2 x - nindF or, roughly, 120/nind mV. For unsymmetrical reactions, a # b, the interval is a function of the stoicheiometric coefficients and the molar absorptivities as well as of nind- and the transition interval is not symmetrically disposed around the transition potential. ONE-COLOUR INDICATORS- When one form absorbs entirely outside the visible range, it appears colourless to the eye. Conventionally the transition point is taken as the point when half the colour has developed, and therefore for symmetrical indicator reactions (a = b) the indicator constant, EO'ind, and the transition potential, Etrans, are identical, For unsymmetrical reactions, the transition potential is independent of indicator concentration and of which form is coloured, but dependent on the stoicheiometry- It is more difficult to define the transition range of a one-colour indicator.The full colour end, that is, the point at which the colour appears fully developed when the colourless form of the indicator is used or at which the first perceptible fading occurs when the coloured form is used, may be subjectively assigned to the condition when the colour is lO/ll-ths developed, that is, lO/ll-ths of the colourless form has been consumed. This point again depends on stoicheiometry and indicator concentration but is independent of molar absorp- tivity.When Indo, is the coloured form, the upper limit of the transition range is defined as Cindred + a log,, a - + b log,, 101 - (13) b Eupper = E h n d + 2'3 -- RT [(a - b) log,, 1.1 nind F When Indred is the coloured form, it is now the lower limit of the transition range that is defined as 1 a b Elower = &nd + 2'3 - RT [(a - b) logloT Cindox + b log,, - - a log,, 10 . - nind F (14) The other limit of the transition range is the point at which the first perceptible colour appears, and this depends on the molar absorptivity of the coloured form, the concentration of the indicator in the solution, the depth of solution (Zcm) through which the colour is viewed, the minimum absorbance ( A ) that the individual is able to detect a t the particular wavelength, and the stoicheiometry.If Indo, is the coloured form, and the indicator is added as the colourless reduced form to give a concentration at the end-point of Cindred, then the first perceptible colour will appear at the lower end of the indicator transition range If Indred is the coloured form, and the indicator is added as the colourless oxidised form to give a resultant concentration a t the end-point of Cindox, then the first perceptible colour540 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS [Analyst, Vol. 96 will appear at the upper end of the transition range When Indo, is coloured, the transition range is from Elower defined by equation (15) to Eupper defined by equation (13) and the transition interval is [ (a EOX Cindred - b A)b .lob ] . . . . . . (17) (a I eOX cin&ed/l*l) (b - a) When Indred is coloured the transition range is from Elower defined by equation (14) to Eupper defined by equation (16) and the transition interval is 2.3RT nind F loglo AE = - (b A)& and when a = b = 1 the transition interval is, in both cases, 2.3 RT I EOX Cind - A 2.3 RT A E = - loglo A +- .. .. . . (19) nind F nind F ION-COMBINATION INDICATORS For a titrimetric reaction of the type B + A + BA (neglecting charges), with a condi- tional formation constant K = [BA]/( [B] [A]) , the indicators participate in similar equilibria of two kinds, depending on whether Ind replaces B or A. If Ind replaces B, it is an indicator base or anionic indicator having an equilibrium and conditional formation constant as in equation (20), where A can be hydroxyl ion, Replacement of A gives an indicator acid or cationic (metallochromic) indicator having an equilibrium and conditional formation constant as in equation (21), where B can be hydrogen ion , Practically all ion-combination indicators are weak acids or bases or salts thereof, and when used for purposes other than hydrogen-ion indication , the true equilibria involve hydrogen ion@).However, by buffering, the hydrogen-ion concentration can be held constant and its effect absorbed into the conditional constant, Kind. Ringbom has made extensive use of such conditional constant^.^ For this class of indicator the indicator constant is the Kind in equations (20) or (21) and is conditional.TWO-COLOUR INDICATORS- indicator occurs when When cationic indicators are used for the illustration, the transition point of a two-colour EBbInde [Bb In&] = EInd [Inn] .. .. . . (22) For symmetrical indicators (a = b) substitution from equation (22) into equation (21) gives .. .. .. .. . . (23) p-J=-*- EInd 1 EBInd Kind and the transition point is therefore, in ion exponent form When a # b, the transition point becomes dependent on indicator concentration and theAugust, 19711 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS 541 stoicheiometry. In addition to equations (21) and (22), the total concentration, Clnd, of the indicator at the end-point is required and is given by equation (25)- Cind = [Ind] + a [Bb Ind,] .. .. .. . . (25) The transition point can then be derived, and is 1 PBtrans = [log10 Kind + (a - 1) log10 Cind + a lo&o EBbInda - log10 EInd + ( - a) log10 (a EInd + EBblnda)] * * .. . . (26) The transition range can be appraised on the basis of the loll and 1/10 convention. Ehd[Ind] = lo fBbInds [Bb Ind,] .. .. . . (27) The indicator appears to be completely in the free form when and to be completely in the complex form when 10 EInd [Ind] = EBbInda [Bb In&] .. .. . . (28) For a symmetrical indicator reaction (a = b) the transition range is given by the limits pB = log10 Kind + log10 =d 5 1 EInd = PBtrans & 1 - .. .. .. . . (29) When a # b the transition range is from - log10 EInd + (1 - a) log10 (a EInd + 10 EBbInda) + a log10 101 - (30) to 1 b pB,@,x = - [loglo Kind + (a - 1) loglo Cind + a log10 EBbInda - log10 EInd + (1 - a) log10 (10 a EInd + EBbInda) - log,, 101 .. (31) Although the first four terms are common to equations (26), (30) and (31), the difference in the fifth term prevents writing the transition range in terms of the transition point a s in equation (29). The transition interval is given by the difference between pBheind of equation (30) and pBcomplex of equation (31) and is This interval is not symmetrically spread around the transition point. transition range simplifies to 2 pB units and is symmetrically disposed point. Analogous expressions are readily derived for anionic indicators. ONE-COLOUR INDICATORS- 111 .. . . (32) When a = b = 1 the around the transition For one-colour indicators the indicator constant is Kind and when a = b the transition point is necessarily PBtrans = logloKind.When a # b the colour is half developed when [BbInda] = cind1(2a) and [Ind] = Cind/2, no matter which is the coloured form. Substitution of these values into 1 [BI trane [B b I nda] gives 1 - (Kind Cind(a - 1) a): [Bltrans 2(8 - 1)542 BISHOP OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS [ A nUlySt, VOl. 96 Taking logarithms gives the transition point- To illustrate the transition range, consider the case of a cationic indicator for which the coloured form is the complex. The point at which the colour appears to be completely devel- oped, or at which the first perceptible fading occurs, when the total indicator concentration at the end-point is Cind, will be defined by a [Bb Ind,] = 10 [Ind] .. . . . . . . (33) from which, with the aid of equations (21) and (25), the lower ion exponent of the transition range can be derived as The upper limit at which the first perceptible colour appears depends on the absorbance, A , detectable by the eye of the individual at the particular wavelength and the depth of solution viewed, I cm, and is For symmetrical indicator reactions, a = b, the limits simplify to PRlower = log10 Kind $- 1 . . .. .. . . (36) $upper = log10 Kind + log10 ( I Cind EBbInda - A ) - log10 A * . . (37) Analogous expressions can be derived for the other three cases: cationic indicator - coloured indicator, anionic indicator - coloured complex and anionic indicator - coloured indicator.For one-colour indicators, all the limits and ranges are concentration dependent except for the case of one limit for symmetrical indicator reactions as exemplified by equation (36). USE OF ANIONIC INDICATORS FOR CATION INDICATION- By using the simplest possible case as an illustration, that is, an indicator for which a = b and a titrimetric reaction B + A + BA, and by making the substitution of in equation (20), and the transition the transformed indicator constant becomes point is then, in contrast with equation (24), This expression now involves both the formation constant, KBA, and the concentration, [BA], of the product of the titrimetric reaction, so that the transition point and the transition range will shift with a change in concentration of the reactants in the titration.The convention of using logloKBA/Kind as the indicator constant is strongly to be discouraged because it conceals the significance of the term log,, [BA]. Furthermore, although Kind may have a small temperature coefficient, KBA may have a large one; this is particularly relevant to the use of anionic indicators in acid - base titrations. TITRATION CONDITIONS AND CHOICE OF INDICATOR- The quality of a titration has been quantitatively assessed by means of Q functions.19617 For a volume, Vd ml, of sample of Concentration cd mol titrated with a reagent of con- centration ct moll-1 and requiring vtml of titrant to reach the equivalence point, let theAugust, 19711 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS 543 smallest increment it is required to distinguish in locating the end-point be Avml.The precision of the titration is Av/vt and the percentage precision is 100 Av/vt. The quality, Q, of the titration is then the change in ion exponent or potential on adding this increment, Av, symmetrically disposed around the equivalence point. Q = AE = I E a t v = ~ t - 0 . 5 ~ u - E a t v = v t + 0 . ~ ~ v I Q = ApB = I pB at v = zt - 0.6 AU - pB at v = v t + 0.5 AZ I . . (40) For a satisfactory visual indicator titration, the indicator should change, that is, the end-point should occur, within the required increment, Av, of the equivalence point, vt. If Q is less than the transition interval of the indicator, then the required precision is not directly accessible, and some other method, such as photometric titration, must be used, or the titration conditions adjusted to increase Q.Furthermore, the indicator must be so chosen that its transition point lies within the required interval around the equivalence point, pBeq pt or Eeq pt, of the titration. - PBtrans = pBeq pt & (Q - transition interval)/2 = Eeq pt & (Q - transition interval)/2 .. . . (41) where, for a reaction m B + n A + B, An, and for titration of a reductant Ox, -j- n, e + Red, of conditional potential oxidant Ox, + n, e s Red,, of conditional potential ElOl with an (43) It is unlikely that, in a real titration, the volume, v, added just before the last split drop will be exactly vt - 0.5 Av. Were it so, the indicator would change colour completely on adding the final increment.If the volume added before the final split drop exceeds vt - 1.6 Av, a perceptible colour change will occur and the volume will be within &0.5 Av of vt and so within the required precision. The main titration is, of course, interrupted while the indicator is being titrated, and consumption of titrant by the indicator will constitute an error. For vindml of indicator of concentration Cind, the error will be negligible if ct vt > Cind vlnd, and this can be assured if the indicator is chosen so that This is not often satisfied : either the indicator has too low a molar absorptivity, or too much of it is used. The pernicious habit of stating indicator concentrations in per cent. w/v is to blame for concealing this. If “two drops” of a 0.1 per cent.solution of an indicator of molar mass 200 are used, C;nd = 0.005 M and vind = 0.1 ml, and the error is up to 0.2 per cent. on a 25-ml titration with a 0.01 M reagent. The author thanks Mr. P. L. Bailey and Dr. J. M. Ottaway for checking the arithmetic and for helpful criticism of the text. DERIVATION OF EQUATIONS- Appendix Equation (6)-From equation (1) the mass balance equation is Substitution of [Indred] from equation (6a) into equation (3) gives544 BISHOP : OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS [Andyst, Vol. 96 whence €red Cindred €OX + a E r e d .. . . . . .. b [Indoxl trans = Substitution of [Ind,,] from equation (6a) into equation (3) gives whence a -€ox Cindred €red -k ~ c o x b a [Indred] trans = ... . . . .. .. =( €red Cindred €ox +-.€red b a >: which simplifies to . . Substitution of equation (6g) into equation (2) then gives equation (6). .. .. .. Equation (7)-Following the same steps as for equation (6) the mass balance equation from equation (1) is Substitution of [Indred] from equation (7a) into equation (3) gives whence b a €red - Cindox €OX -k €red 2 .. b [Indoxl trans = Substitution of [Index] from equation (7a) into equation (3) gives whence COX Cindox €red -k 5 COX .. a [Indred] trans = .. . . .. .. .. . [Indox]&s = ( Zndox>'. ( d + a b €OX) a .. €OX + - €red Cox Cindox [Indre d l trans which simplifies to €reda b Cindox a( a COX + b €red .. .. .. .. .. .. .. .. . . .. .. .. .. .. ..August, 19711 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS 545 Substitution of equation (7g) into equation (2) then gives equation (7).Epation (9)-Elow, with indicator added in the oxidised form. equation is (7a), and the absorbance relationship is The mass balance 10 [Indo,] COX = [Indred] €red - . .. .. Substitution of [Indred] from equation (7a) into equation (ga) gives . . b a 10 [Index] COX = €red- (Cindox - [Index]) - + whence b €red Clndox * .. b [Indoxllow = (10 Cox + €red) .. .. Substitution of [Indo,] from equation (7a) into equation (9a) gives whence . . .. .. 10 €OX Clndox (€red + 10 5 COX) a [ Indred] low = . . . . . . .. .. .. .. Substitution of equation (9g) into equation (2) gives Elow. For Ehigh, with indicator added in the oxidised form, the mass balance equation is (7a) [Indo,] COX = 10 [Indred] €red - ... .. . . (9h) and the absorbance relationship is Substitution of the value of [Indred] obtained from equation (7a) into equation (9h) gives whence b 10 €red clndox €ox + 1oi.ered .. . . . . .. b [Indoxlhigh = Substitution of [Ind,,] obtained from equation (7a) into equation (9h) gives whence COX Clndox 10 €red -k jj COX . . a [Indredlhigh = . . . .546 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS [A?Za@Si?, VOl. 96 which simplifies to Substitution of equation (9n) into equation (2) gives Ehigh. A similar derivation based upon the mass balance equation (6a) for the indicator added in the reduced fonn and the absorbance equations (9a) and (9h) will give the results noted in the text following equation (9).Equation (1 1) is obtained by subtracting Elow from Ehigh in equations (9) and simplifying. Equation (12)-If the coloured form is Ind,, and the reduced form is added to give a then at the transition point, from equation (1) total finishing concentration of Cindred mol Substitution in equation (2) gives which simplifies to equation (12). If the coloured form is Indred and the indicator is added in the oxidised form to give a finishing concentration of Cindox mol l-l, then at the transition point, from equation (l), b a [Indred] = 0.5 - Cindox; [Indo,] = 0.5 Cindox Substitution in equation (2) gives which also simplifies to equation (12). Equation (13)-1nd0, is the coloured form and the indicator is added in the reduced form to give a finishing concentration as in the mass balance equation (6a).The optical condition is [Ind,,] = 10- [Indred] . . .. .. . . (13a) a b From equations (13a) and (6a) a Cindred . . b Cindred - 2 [Index] = -~ ) b 1.1 . (13b) a b Cindred [Innred] = Cindred - - 10- [Indred] == - b a 11 * * . . (13c) = c-)'" - b, l o b , . .. . . (13d) which, when substituted into equation (2), gives equation (13).August, 19711 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS 547 Equation (14)-Indred is the coloured form, and the indicator is added in the oxidised form to give a finishing concentration as in the mass balance equation (7a). The optical condition is now . . (Ma) [Indred] = 10 - [ Ind,,] .. .. .. b a From the simultaneous equations (7a) and (14a), . . (14b) a b Cindox [Indo,] = Cindox - 6 10; [Index] = - .... 11 From equations (14b) and (14c), which, when substituted into equation (2), gives equation (14). Equation (15)-Indo, is the coloured form, the indicator is added in the reduced form to give a finishing concentration as shown in the mass balance equation (sa), and from Beer's law the first perceptible colour of Indo, appears when . . .. . . .. . . (15a) A z [Indo,] = where A is the minimum absorbance detected by the eye when viewing through I cm of solution. From equations (15a) and @a), .. . . . . . . (15b) b A [Indred] = Cindred - 2 * - 1 COX Re-arrangement of equation (15c) and substitution into equation (2) gives equation (15). Equation (16)-hdred is the coloured form, the indicator is added in the oxidised form as in the mass balance equation (7a), and from Beer's law the first perceptible colour of Indr,d appears when .... . . .. . . (16a) [Indred] - 1 €red A From equations (7a) and (16a), a A [Indo,] = Cindox -- . . .. . . . . (16b) b G Re-arrangement of equation (16c) and substitution into equation (2) gives equation (16). (13) and equation (14) from equation (16), respectively, and simplifying. are obtained by solving the optical equation (22) and the mass balance equation (25). Equations (17) and (18)-These are obtained by subtracting equation (15) from equation Equation (26)-Concentrations of free indicator, [Ind], and of the complex, [BbInd,!,548 BISHOP : OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS [A?ZU&St, V O l . 96 . . (26a) .. EInd (Cind - a [BbInda]) = EBbInda [BbInda] whence .. (26b) Making the converse substitution, whence EInd [Indl Z= EBbInda (Clnd - [Ind])/a . . . . . . (26c) .. * . .. . . (26d) Cind EBbInda a EInd + EBbInda [Ind] = From the formation constant in equation (Zl), [IndIaKind .. .. . . (26e) 1 [Bl ( [BbInda] ) . ’ Substitution of the transition point concentrations from equations (26b) and (26d) gives Taking logarithms to base 10 of both sides of equation (26f) gives PBtrans in equation (26). Equatio.iz (30)-The mass balance equation is equation (25) and the optical condition is equation (27). Substitution for the concentration of the complex gives EInd [Ind] = 10 EBbInda (Cind - [Ind])/a .. .. . . (30a) whence * . .. .. . . (30b) 10 EBbInda Cind [Ind] = a EInd + 10 EBbInda The converse substitution gives EInd (Cind - a [BbInda]) = 10 EBbInda [BbInda] .- .. . . (30c) whence Substitution into equation (26e) gives and conversion into logarithms to base 10 gives equation (30). Equation (31)-The mass balance equation is (25) and the optical condition is now equation (28). Substitution of the value of [BbInda] from equation (25) into equation (28) gives 10 EInd [Ind] = EBbInda (Cind - [Ind])/a .. .. . . (31a) .. The converse substitution gives 10 fInd (Cind ..August, 19711 BISHOP: OBSERVATIONS ON THE THEORY OF ACTION OF VISUAL INDICATORS 549 Substitution into equation (26e) gives Conversion into decadic logarithms gives equation (31). Equution (34)-From equations (33) and (25), .. .. . . (34a,b) Cind 10 Clnd [Ind] = - ; [BbInd,] = - 11 11 a Substitution into equation (26e) gives Conversion into decadic logarithms gives equation (34). Equation (35)-The first perceptible colour of the complex appears when .. .. .. . . (35a) A [BbInd,] =- 1 EBbInd, Substitution into equation (25) gives Cind = b d 1 -k Ga .. .. .. . . (35b) a A whence Substitution into equation (26e) then gives Conversion into logarithms to base 10 gives equation (35). 1. 2. 3. 4. 5. 6. 7. REFERENCES Bishop, E., Proceedings of the SAC Conference, Nottingham 1965, Society for Analytical Chemistry, Kolthoff, I. M., and Stenger, V. A., “Volumetric Analysis,” Volume I, Interscience Publishers Inc., Bishop, E., in Wilson, C. L., and Wilson, D. W., Editors, “Comprehensive Analytical Chemistry,” Walden, G. M., Hammett, L. P., and Chapman, R. P., J . Amer. Chem. Soc., 1931,53, 3908. Ringbom, A., “Complexation in Analytical Chemistry,” Interscience Publishers Inc., New York, Bishop, E., “Indicators,” Pergamon Press, Oxford, 1971, Chapters 2 and 8b. -, “Coulometric Analysis,” in Wilson, C. L., and Wilson, D. W., Editors, “Comprehensive Received December 29tk 1970 Accepted March 16th, 1971 London, 1965, p. 291. New York, 1942. Volume IB, Elsevier, Amsterdam and London, 1960. 1963. Analytical Chemistry,” Volume IIB, Elsevier, Amsterdam and London, in the press.

ISSN:0003-2654    

DOI:10.1039/AN9719600537

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

5.50 Analyst, August, 1971, Vol. 96, pp. 550-561 An Automatic Capillary Viscometer Part 11.';' Automatic Apparatus for Viscometric Titrations BY R. B. SIMPSON, J. S. SMITH7 AND H. M. N. H. IRVING: (Department of Inorgalzic and Stvuctuval Chemistry. The Uiziversity, Leeds 2) The basic principIes of viscometric titrations are considered and examples of applications to a variety of analytical and other determinations are suggested. An addition to the automatic capillary viscometer described in Part I provides for the introduction of successive measured volume increments of a solvent or solution, admixture with the solution already present in a sus- pended-level viscometer and the electronic measurement of the flow time including printing-out the results. By using the new apparatus it is possible expeditiously to determine the B-coefficients of the Jones - Dole equation for the viscosity of electrolytes and the intrinsic viscosity coefficients for solutions of polymers, and to perform a wide variety of acid - base titrations.Practical examples of these deter- minations are given in illustration. TITRIMETRIC procedures for carrying out the reaction between one material, S (the sample), and another material, T (the titrant), can form the basis of quantitative measurements provided that the end-point can be detected by suitable means and various other obvious conditions are fulfilled. S + T +- products (P) A titration graph is constructed with some linear function of the concentration of T along the abscissa and some function of the concentration of S (or of P) along the ordinate.The end-point of the titration (which may or may not coincide with the stoicheiometric end-point) is then indicated by the value or values of [TI at which there is an extremum or inflection. Although measurements, for example, of optical absorption, optical rotation, conduc- tivity, electrode potentials, diffusion currents, radioactivity and thermal and magnetic effects are among the many physical properties that have been made use of, there would seem to be no previous reference to the measurement of viscosity as the basis of a titration pro- cedure. This is scarcely surprising if the process is envisaged as one in which samples of the titration mixture are removed after successive additions of titrant for measurements of their viscosity in one or other of the conventional instruments.Practical considerations, such as loss of material, problems of cleaning the viscometer between successive samples and possible changes in calibration, problems of temperature control and the excessive time required, have clearly discouraged analysts from considering viscometric titration as a useful experiment a1 technique . In Part I1 we described an instrument by use of which the process of determining the viscosity of a liquid can be carried out automatically as often as required and with high precision, the flow time being measured electronically to the nearest millisecond. Provision for printing out the results has since been added. We have now extended the usefulness of this equipment by combining it with an auto- matic burette (Radiometer ABU12b) that is capable of adding aliquots of the titrant, T, in precisely pre-determined volumes, to the sample solution, S, contained in the reservoir of a suspended-level viscometer maintained at a constant temperature.By using a combination of logic units controlling a source of compressed gas through a series of valves, it is possible to program a sequence of operations in which the sample solution receives an aliquot portion * For particulars of Part I of this series, see reference list, p. 561. t Present address : Department of Chemical Sciences,The Hatfield Polytechnic, Hatfield, Hertfortlshire. t Requests for reprints should be sent to Professor Irving. 0 SAC and the authors.SIMPSON, SMITH AND IRVING 55 1 of titrant, the whole is then well mixed, its viscosity is measured as previously describedl and finally the flow time is printed out.The whole sample is then returned to the reservoir and a further addition of titrant is made, the cycles being repeated automatically to give a series of flow times which, when plotted against the volume of titrant, constitute the visco- metric titration curve. Changes in flow time can then be used to reveal changes in the viscosity of the mixture and thus changes in its composition. Where sudden changes in the slope of the graph (break-points) coincide with changes in the stoicheiometry of the reaction the results can readily be used for the quantitative determination of the sample. We shall show later that in certain acid - alkali determinations the changes in viscosity may actually be more informative than changes in, for example, pH.Under strictly reproducible conditions break-points can be used for quantitative measurements even when the process is non- stoicheiometric. POSSIBLE APPLICATIONS OF THE METHOD- Viscometric titrations could be used, in principle, for any systems that remain homogeneous throughout. They need not be restricted to purely aqueous solutions and they may well be applicable to systems for which conventional methods of following changes in the concentra- tion of a sample (or products) fail. The essential new feature is that there must be a sufficient difference between the effects of the sample S, the titrant T and admixtures with the reaction products P on the viscosity of the solvent, so that in the course of a titration, changes in the relative concentrations of these species can be indicated by changes in flow time.For example, the sudden change in hydrogen-ion concentration at the end-point of an acid - alkali titration is reflected in the sudden change of slope in a graph of flow time against volume of titrant. Quite apart from the possibility of its use to discover viscometric titrations of analytical interest, the present apparatus is particularly suitable for studying the viscosity of mixtures ; we give below examples of its use for deriving the constants A and B of the Jones-Dole equation (2) for a solution of an electrolyte- and to obtain the constants k and [77] for the equation- which is extensively used in the characterisation of polymers. We would certainly anticipate differences between the effect of a base and its conjugate acid on the viscosity of the solvent, although the magnitude of this difference will depend on the nature of the species involved.Indeed, any reaction in which the products are of a significantly different size, or charge, or both, and which could interact with the solvent to give aggregates of different viscosity, should be suitable for viscometric titration. The interactions between cations and EDTA and between borates and cis-diols are obvious fields for study and many complexation and adduction reactions that proceed only in non-aqueous solvents are open to investigation. We have found measurable differences between the viscosities of certain cis and trans isomers of co-ordination compounds and this opens up the possibility of investigating the kinetics of isomerisation when other methods fail.The kinetics of other reactions, such as polymerisations, condensations, depolymerisations, hydrolyses and substitutions can be conveniently studied in the automatic apparatus now described. While many preliminary studies will be necessary to explore the field and to develop new procedures it is not premature to recognise that when sufficiently large changes in viscosity are found to occur they could (with appropriate apparatus) be used to monitor a reaction and to control plant. DESCRIPTION OF THE AUTOMATIC VISCOMETER- The fill - empty cycle of the visconieter, as defined and described previously,l remains unmodified in the present extended sequence, which may be described as a fill - empty - add - mix cycle.The start and stop-level detector circuits are essentially the same as before, the main differences being the addition of a line from the output of the start-level timer unit, and the introduction of the operations-mode circuit. The output from the decade counter used, as described in Part I, to pre-determine the number of measurements on the single sample, is now used both to inhibit the fill process .. .. ' - (2) .. .. .' (3) qrel. = 1 + A d c + Bc qrel. = 1 + L.13~ + kc2 . . . .552 SIMPSON, SMITH AND IRVING : AN AUTOMATIC CAPILLARY VISCOMETER [AlZUlySt, VOl. 96 Timecounter circuit Start and stop-level j 1 Operations 1 1 i detector and timer circuit -mode circuit Fig.1 (a). Start-level and stop-level detector circuits, operations-mode circuit and time-counter circuit Add-counter circuit Variable mark - space ratio oscillator I;ig. 1 (b). Variable mark - space ratio oscillator and add-counter circuit (via an extra memory unit) and to energise a variable mark - space ratio oscillator. The output of this oscillator is used to cause increments of the titrant to be added to the viscometer from the automatic burette, the number of increments being counted on another decade counter (the add-counter circuit). When a pre-determined number of increments has been added this counter output goes to a 1 level (as previously defined1) and the oscillator is de-energised. This same counter output is also used to close the solenoid valves in a sequence that causes mixing by a brief passage of inert gas through the viscometer.After mixing has been completed, the output of the memory unit inhibiting the fill - empty cycle is returned to the 0 level and the complete cycle is recommenced. The start and stop-level detector circuits are shown in Fig. 1 (a), the oscillator and add-counter circuit in Fig. 1 (b) and the mix circuit in Fig. 1 (c). We now describe in detail the two new aspects of the process, namely the addition and the mixing of liquid.August, 19711 PART 11. AUTOMATIC APPARATUS FOR VISCOMETRIC TITRATIONS 553 output 50-------.----c7>1 7 7 I 06 3 0 4 0 1 G Fig. 1 ( c ) . Mix circuit ADDITION OF LIQUID TO THE VISCOMETER- Provision is made for two possible modes of addition by introducing between the basic start-level output and the time-counter circuit a simple circuit [Fig.1 (a)], which we term the operations-mode circuit. With the counter connected to Y the addition of an increment of titrant takes place after the flow time of the existing mixture has been determined. How- ever, with the switch in position X, the addition of titrant to the reservoir commences a short time after the measurement of flow time begins and while the previous mixture is still flowing down the capillary. This addition of fresh titrant cannot affect the composition of the liquid of which the viscosity is being measured, but there can be a substantial saving of over-all time as the addition can be made to take place during the period required to measure the flow time.Logic level Fig. 2. Oscillator output wave form Let us suppose that an initial charge of liquid, sample S, has been placed in the viscometer, and that the required number of flow time measurements has been taken (in the manner described in detail in Part Il). The number of flow time measurements required is pre-set on the time-counter switch. The output of this counter goes to 1 and is remembered by both first and second memory units [Fig. 1 (a)]. The output of the second memory unit returns to the stop-level circuit to inhibit the fill process and the output of the first memory unit goes to the variable mark - space ratio oscillator via terminal 2 [Fig. 1 (b)]. The output of this oscillator goes between 0 and 1 in square-wave form as shown in Fig. 2.The length of pulse, e, (Fig. 2) and the separation of the pulses, d, (Fig. 2) depend on the values of the capacitors and resistors used with the two timer units (Fig. 3, D and E). The output of the oscillator is used in the present equipment to drive a Radiometer Autoburette ABUlZb, each pulse causing one increment of titrant, T, to be delivered. This instrument can be supplied with a range of burette assemblies, but in each case one increment is 0.04 per cent. of the total burette volume. The delivery of an increment is triggered by energising the ABUl2b internal relay with approximately 70 V d.c. and the output of the oscillator [out- put 11, Fig. 1 (b)] is used to effect this operation via a low power, thyristor trigger, output unit (YL6023/01), as shown in Fig. 4.(In a modified version of the apparatus constructed by one of us (J.S.S.) in London, the output from the low power amplifier 21A60 is used to drive a 600-IR miniature relay. The input socket 2 on the back of ABU12 is taken to the contacts of this relay so that the ABU12 delivers one increment each time they close.)554 SIMPSON, SMITH AND IRVING: AN AUTOMATIC CAPILLARY VISCOMETER [Analyst, VOl. 96 When it is required to add large amounts of titrant, the switch at the input of the oscillator is opened and the autoburette is operated manually. The Radiometer Autoburette has a gearbox that provides a choice of delivery rates. As it is necessary for the increment to be completely delivered before the start of the next command pulse, the values of the para- meters d and e must depend on the delivery rate chosen.Further, it is a characteristic of this instrument that the command pulse must be held for 60 per cent. of the time taken to deliver one increment. Both these factors make it imperative to choose carefully the values of the capacitors and resistors associated with the two timer units in the oscillator. It has been found that 10-pF fixed capacitors, used in conjunction with resistor chains connected to a pair of ganged rotary switches, provide the most convenient way of matching the oscillator to the requirements of the autoburette at its various delivery rates (see Fig. 3). C1 Fig. 3. Adjustable oscillator settings (d and e) Switch (percentage of total burette Space Pulse compatible with autoburette delivery rate Autoburette delivery rate position volume minute-') (d)/s (;y 1 20 2.7 2 10 2-7 2.0 3 5 5.4 3.0 4 2.5 6.4 5.7 The number of pulses generated by the oscillator and hence the number of increments added by the autoburette is counted by what we have called the add-counter circuit.This comprises two decade counters in series so that the total number of increments added is the product of the numbers on each counter; this produces a wider range of additions than with one counter alone. In early experiments with this sytem it was found that one of the four flip-flops in the first decade was not re-setting during the time required for its output to be transferred to the second decade. This weakness was overcome by introducing the time delay, F, of approximately 1 second's duration.When the pre-set number of increments have been delivered the output of the second counter goes to 1 and immediately re-sets the first memory unit of the time-counter circuit via terminal 3, thus de-energising the oscillator and ending the process of addition. The output is also taken to that part of the circuit controlling the mixing process [Fig. 1 (c)J, which we now describe. fa- l l (add) relav n 13 I Fig. 4. Output stage to autoburetteAugust , 19711 MIXING OF LIQUID IN THE VISCOMETER- PART 11. AUTOMATIC APPARATUS FOR VISCOMETRIC TITRATIONS 555 Auto \ (fill) I I 24V 1 1 1 (mix) D12 Fig. 5. Output stage to solenoid valves The mix circuit [Fig. 1 (c)j is so designed that, 5 s after the add process has been com- pleted, output I11 goes to the 1 level (provided that the measurement of flow time has been completed and the subsequent stop-level time delay, B, has elapsed).This output is held at the 1 level for a period of from 1 to 10 s, as controlled by a potentiometer at the time unit, H, and is then cancelled by breaking the add-counter memory circuit via terminal 6. Output 111 is taken to the output stage and valve circuit (Figs. 5 and S), thus causing the mix process. It is also taken to the mix memory circuit, which remembers the 1 level at output I11 after it has fallen back to 0 level. After a fixed period of 20 s in this case, the output of timer unit J goes to 1 level and cancels the time count second memory unit via terminal 4, thus allowing the fill-empty cycle to recommence.The mix memory circuit is cancelled im- mediately after the next filling has been completed. Fig. 6. Solenoid valve assembly. TTalves A and H are normally open and valves C and D are normally closed OUTPUT STAGE AND VALVES- The solenoid valve output stage is shown in Fig. 5 and the associated assembly of solenoid valves in Fig. 6. It will be noted that this is a modification of the circuit described in Part I of this paper and that it is now necessary to use four valves in order to carry out the filling and mixing operations. The valves are again those manufactured by Skinner Precision Industries. Valves A and B are of the normally open type (specifically V51 DA 2125556 SIMPSON, SMITH AND IRVING: AN AUTOMATIC CAPILLARY VISCOMETER [AnaZyst, Vol.96 24/DC), whereas valves C and D are normally closed (B2 DA 1400 24/DC). These valves receive power, as before, from a separate supply in order to avoid large voltage swings on the logic supply. Outputs I and I11 are taken to the output stage via a pair of switches, which may be used to operate the valves manually. With these switches in the "auto" position, the valves will operate as directed by the process control unit in the manner described above, but when these switches are at the manual setting the valves can be operated by placing the "centre-off" switch, which is connected to the 24-V supply, in the appropriate position. One difficulty that occurred in initial trials was the gradual build-up of pressure behind the solenoid valve C between successive flow time measurements.Sometimes this led to the formation of gas bubbles and in any case to less reproducible conditions for precise measure- ment of flow times. The problem was solved by introducing a short delay in energising valve B on the command to fill. This was effected without additional logic units by making use of the mix timer unit, H, as shown in Figs. 1 (c) and 5. EXPERIMENTAL In all applications up to the present time the viscometer used in conjunction with the automatic control system has been of the suspended-level type with a capillary bore of 0.6 mm. All measurements of flow time were made at a constant temperature of 25" j= 0.005 "C. In preliminary trials with pure water as the test material, thirty consecutive measurements of flow time over a period of approximately 2 hours produced a standard deviation of only 2 ms in a flow time of approximately 2.5 minutes.Standard solutions of hydrochloric acid, sodium hydroxide, ammonia and sulphuric acid were prepared; except where otherwise stated, all analytical chemicals used in this work were of reagent grade and were used without further purification. The results of some typical applications will now be described to illustrate the potentialities of the apparatus. TABLE I DETERMINATION OF THE JONES - DOLE EQUATION FOR HYDROCHLORIC ACID AT 25 "c Additions of titrant (N HCl)/ ml 0 0.020 0-050 0.100 0.150 0.200 0.250 0.300 0.400 0.500 0-750 1-000 1.250 1.500 1.750 2.000 2-250 2.500 Flow time 156.270 156,302 156.310 156.328 156.346 156.360 156.377 156.391 156.432 156-459 156.543 156-626 156.704 156.777 156.846 156-915 156.987 157.060 ( t ) Is Relative density (&el.) 1.000 00 1.000 01 1.000 04 1.000 08 1.000 12 1.000 16 1.000 20 1.000 24 1.000 32 1.000 39 1-00058 1.000 77 1.000 95 1.001 16 1.001 31 1.001 48 1.001 64 1.001 81 Concentration mole 1-' 0.000 00 0.000 89 0.002 22 0.004 42 0-006 62 0.008 8 1 0.01099 0.01316 0.01747 0-02 1 74 0.032 26 0.042 55 0.052 63 0.062 50 0.072 16 0-081 63 0.090 91 0~10000 (4 / vrel 1~00000 1.00022 1.000 30 1.000 46 1-000 61 1.000 74 1.000 89 1.001 02 1.001 36 1.001 61 1-002 33 1.003 06 1.003 74 1-004 41 1.005 0 1 1.005 62 1.00624 1-006 88 DETERMINATION OF THE A AND B COEFFICIENTS OF THE JONES-DOLE EQUATION (2) FOR Pure water (22.50 ml) was placed in the viscometer reservoir initially and increments of M hydrochloric acid were added.The final acid concentration of 0.1 M was achieved by a total addition of 2.50 ml of acid. Flow time readings for each solution are shown in Table I. Relative densities, &,., were calculated from the equation- HYDROCHLORIC ACID AT 25.0" & 0.005 "C- .. .. - * (4) drel. = 1 + 0.0181 c .. based on data by Huckel and Schaaf2 and valid for the range c = O.OOO8 to 0.15, where c isAugust, 19711 PART 11. AUTOMATIC APPARATUS FOR VISCOMETRIC TITRATIONS 557 the concentration in mole 1-l; these are shown in the third column of Table I. From the calculated concentrations (column 4) the relative viscosities (column 5) were determined by using equation (5)- - * (5) tsolution taolrent x drel. .. .. .. %el. = ~ where t is the flow time in seconds.The coefficients A and B of the Jones - Dole equation (2) were then calculated from equations (6) and (7)- whence when c is in the range 0.00089 to 0-10000 mole 1-1. The standard deviation of yrel. (qrel. = &0.004 per cent.) lies well within the estimated experimental error for such results. The present results are also in reasonably good agreement with those given by Hiickel and Schaaf" for hydrochloric acid at 25 "C. yrel. = 1 + 0.00252/;+ 0.06015~ .. .. . . . . (8). Their results are summarised in equation (9)- where CHARACTERISATION OF POLYMERS BY SOLUTION VISCOSITY- The usefulness of solution viscosity as a technique for measuring the molecular weights of polymers is well e~tablished,~ and the new apparatus provides for complete polymer characterisation by a rapid and accurate process.The results of measurements on solutions of polymers are generally fitted to an equation of the form- where c is the concentration of the polymer in grams per 100 ml; [r)] and k are constants for the sample of polymer under investigation. Both [r)] and k depend upon the average molecular weight of the sample and, if the form of this dependence is known for a particular polymer, the values of [r)] and k obtained by experiment can be used to find the molecular weight of the sample. The subject chosen to demonstrate this characterisation of polymers was a sample of poly(viny1 alcohol), supplied by B.D.H., with a molecular weight of approximately 14 000. An initial charge of 22.5 ml of pure water was placed in the viscometer and a total of 2.00 ml of a stock solution of poly(viny1 alcohol) was added from the autoburette in 0.250-ml aliquots.Values of flow time (t s) and concentration (c g per 100 ml) are shown for each solution in Table 11. TABLE I1 DETERMINATION OF THE INTRINSIC VISCOSITY OF POLY(VINYL ALCOHOL) (PVA) I N WATER AT 25 "c Additions of titrant, PVA stock solution*/ml 0 0.250 0.600 0.750 1.000 1-250 1.500 1.750 2.000 Flow time (t) Is 156*150(to) 159.517 162.783 165.73 1 169- 151 172.196 175.194 178.219 180.972 Concentration (c),g per 100 ml 0 0,0552 0.1092 0.1621 0.2138 0.2644 0.3140 0.3626 0.4102 * Concentration of stock solution was 5.0244 g per t This point was omitted from the final analysis. (t-to) /toe 0 0.3904 0-3890 0-3790t 0.3894 0.3887 0.3884 0-3898 0-3876 100 ml.558 SIMPSON, SMITH AND IRVING: AN AUTOMATIC CAPILLARY VISCOMETER [Analyst, VOl.96 Now, .. .. * (3) qm1. - 1 = [TIC + kC2 .. . . (%el* - ') = C + kc or .. .. .. . . (10) -- (t - to) - [y] + kc . . toc Values of (t - t,)/t,c are also tabulated in Table 11. Equation (10) was subjected to a least squares analysis to find [TI, the intrinsic viscosity and k, whence and z.e., where Difficulty was experienced in achieving complete mixing with these viscous solutions without introducing persistent air bubbles that would invalidate measurements of flow time. Clearly, in its present form the viscometer itself is not suited to this type of measurement and a viscometer should be chosen that is specially designed for work with solutions of high viscosity.TITRATION OF BASES WITH A STRONG ACID- The previous two applications, the production of Jones - Dole graphs and the charac- terisation of polymers, merely demonstrate the automation of well established experimental techniques, but the technique of being able to make and follow changes of species in a solution by a continuous monitoring of solution viscosity is, we believe, a novel use of such measure- ments. Viscometric titrations show many similarities with conductimetric titrations both in the shapes of the resultant graphs and in the fundamental theory. As each ion has its own conductivity, so is it possible to assign to it an ionic viscosity derived from the B-coefficients of the Jones - Dole equation for different electrolyte^.^ For example, B values for salt pairs with the same anion but different cations have con- stant differences and therefore additivity is adduced to the separate ions.Just as the iso- electronic K+ and C1- ions are assigned the same ionic radius so they are considered to contribute equally to the total viscosity of a solution of potassium chloride. Reference 4 lists the B-coefficients for such a solution up to 0.1 M at 25 "C as -0.014. It is therefore apparent that the assigned ionic viscosity for both K+ and C1- is -0.007. In a similar manner, by using this principle of additivity, the ionic viscosities of the other ions can be determined. Table I11 lists some of these values that are particularly relevant to the present work. TABLE I11 IONIC VISCOSITIES IN WATER AT 25 "C Concentrations normally <O- 1 M Ion .. . . . . K+ c1- H+ Na+ OH- Viscosity (B-coefficient) - 0.007 - 0.007 + 0.067 + 0.086 + 0.1 19 It can be predicted from the values given in Table I11 that the viscosity changes during a titration of sodium hydroxide solution with hydrochloric acid a t the temperature and concentration stated will consist of a relatively sharp decrease in viscosity (and flow time) up to the end-point as hydroxide ions (large B-coefficient) are replaced by chloride ions (smaller B-coefficient), followed by a more steady increase beyond the end-point as hydro- chloric acid is added to the solution of sodium chloride that results a t the end-point, soAugust, 19711 PART 11. AUTOMATIC APPARATUS FOR VISCOMETRIC TITRATIONS 559 increasing the total concentration of ions.Such a titration was performed, a single flow time being recorded after each addition of acid. The results are expressed graphically as the titration “curve” in Fig. 7. Similar acid-to-base titrations were carried out by using hydro- chloric acid and a variety of bases including ammonia solution, pyridine and ethylenediamine. In each instance graphs similar in general appearance to that in Fig. 7 were obtained, all with clearly discernible break-points coincident with the stoicheiometric end-points. In the case of ethylenediamine two such break-points were evident. Also, as expected, no significant changes were observed when the titrations were repeated in solutions of high, constant ionic strength (e.g., in a molar solution of sodium nitrate). 0 1.0 2.0 3.0 N Hydrochloric acid/ml Fig.7. Viscometric titration of 0.1 N sodium hydroxide solution (13 ml) with N hydrochloric acid TITRATIONS OF ACIDS WITH A STRONG BASE- A large number of acids have been titrated viscometrically to date with sodium hydroxide solution as the base and titrant. Fig. 8 shows two typical graphs recorded for the strong acid, hydrochloric acid, and for phenol (pK, = 10.00 at 25 “C). The difference in acid strength appears to have little or no effect on the significance or validity of the end-points. Polybasic acids have also been successfully titrated with the capillary-flow titrimeter. These include sulphuric, oxalic, tartaric, citric and orthophosphoric acids. With the last, dis- tinct breaks were produced for all three end-points, including that for the removal of a proton from HP04% (pK, = 12-30), which is not readily detectable by other analytical techniques.164 162 160 158 1.26 1.16 0 1 .o 2.0 3.0 N Sodium hydroxide solution/ml Fig. 8. Viscometric titration of 0.1 N hydrochloric acid (12 ml) and 0.1 N phenol (13 ml) with N sodium hydroxide solution : 0, hydrochloric acid; and 0, phenol560 SIMPSON, SMITH AND IRVING : AN AUTOMATIC CAPILLARY VISCOMETER [A'lZdySt, VOl. 96 156.9 - 3 156.7 - 156.5 0.57 1.28 1.92 2.56 0 1 .o 2.0 3.0 0.1 N Sodium hydroxide solution/ml Fig. 9. Viscometric titration of 0.01 N EDTA solution (25 ml) with 0.1 N sodium hydroxide solution However, an excellent example of the potential of viscometric titrimetry is the neutralisa- tion of ethylenediaminetetraacetic acid (EDTA ; H,Y) with sodium hydroxide.This titration was carried out by adding aliquots of 0.1 N sodium hydroxide solution t o a solution of 0.01 N EDTA initially contained in the viscometer reservoir. Fig. 9 depicts the viscometric titration graph for the pure acid (EDTA), break-points attributable to all four end-points being clearly visible. In order to show clearly the association between these break-points and changes of species in solution, the distribution of the species H,Y, H,Y-, H,Y2-, HY3- and Y4- in solution during the course of the titration was calculated and is shown together with the calculated pH titration curve in Fig. 10. In this case it is evident that the new technique of titration by viscosity monitoring provides much more information than does pH measurement about the basic changes occurring during the titration and the relationship of these changes to the stoicheiometry of the neutralisations.Although only simple acid - base neutralisations have been discussed in this report on applications of our capillary-flow titrimeter, it is probable that the instrument and technique will lend themselves to the study of innumerable analyti- cally important systems. Further, the electronic logic circuitry has been designed so that additional or alternative circuits to extend the versatility of the automatic operation can be incorporated as the need arises and with the least possible disturbance of existing circuitry. L P .- I u a Fig. 10. Graphs of percentage [HjY] for EDTA against a (the equivalents of alkali added per mole of EDTA).The graphs relate directly to the viscometric titration shown in Fig. 9. Also shown is the pH titration curve for the neutralisation of EDTA. 0, H,Y: 0, H,Y-; A, H,Y2-; A, HY3-; m, Y4-; and ---, pHAugust, 19711 PART 11. AUTOMATIC APPARATUS FOR VISCOMETRIC TITRATIONS 561 One of us (J.S.S.) thanks the Royal Society and the Chemical Society for grants towards the cost of his equipment and another (R.B.S.) is grateful to the Science Research Council for a Research Studentship. REFERENCES 1. 2. 3. 4. Smith, J. S., Irving, H. M. N. H., and Simpson, R. B., Analyst, 1970, 95, 743. Hiickel, E., and Schaaf, H., 2. Phys. Chem., 1969, 21, 326. Billmeyer, F. W., jun,, “Textbook of Polymer Chemistry,” Interscience Publishers Inc., New Stokes, R. H., and Mills, R., “Viscosity of Electrolytes and Related Properties,’’ Pergamon Press, NOTE-Reference 1 constitutes Part I of this series. Received December 1 lth, 1970 Accepted March 16th, 1971 York, 1967, pp. 80-87. Oxford, 1966. Appendix LIST OF COMPONENTS Resistors- R,, R,, R,, R7 = 100-kQ a-W %d, R6 = 270-kR, &-W R6 = 1.6-kR, 6-W Capacitors- Diodes- GI c, = 10-pF low-leakage polycarbonate type Norbit 2 comfionents- Quantity 2 9 1 4 16 1 or 1 Symbol lettering Description PS Pulse shaper PS90 TU* Timer unit TU60 LP Low power amplifier LPA6O PA Medium power amplifier PA60 - Twin 4 input Norbits 2NOR60 YL6023/01 Low power thyristor trigger output unit 24 V Miniature relay (see text) * Timer units- Letter Time delay/s A 1 to 11 Variable B 1 to 11 Variable C 6 Fixed D 7 Oscillator E (see Fig. 3) F 1 Fixed G 6 Fixed H 1 to 11 Variable J 20 Fixed K 1 t o 3 Variable Terminal inputs- Letter Description R Manual re-set input S Switch-on re-set input For a more detailed explanation of components and design, including decade counters, used in the circuitry of this paper consult “A design guide for Norbit 2” and associated literature obtainable from Industrial Electronic Controls (Mullard) Ltd.

ISSN:0003-2654    

DOI:10.1039/AN9719600550

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

562 Analyst, August, 1971, Vol. 96, $9. 562-564 An Inert Dilution Method for the X=ray Fluorescence Analysis of Niobate - Tantalate Mineral Concentrates BY Y. C. WONG AND S. SEEVARATNAM (Geological Survey of Malaysia, Perak, Malaysia) An X-ray fluorescence method, in which iron(II1) oxide is used as inert diluent, is described for the determination of niobium, tantalum, tin and titanium in concentrates of polymineral composition. No time-consuming fusion procedures or tedious calculations are involved. The simplicity and rapidity of the method merit its use in routine analysis. Good agreement with chemical analysis is obtained. THE mineral concentrates submitted to the Geological Survey of Malaysia for the deter- mination of niobium and tantalum are usually high grade columbite or tantalite - columbite containing various proportions of ilmenite, rutile, cassiterite, monazite, xenotime, wolframite and zircon; the latter concentrate is particularly difficult and tedious to analyse by conven- tional chemical methods.Rose and Brown1 have described an X-ray fluorescence method in which matrix effects are minimised by fusing the sample with a mixture of lithium borate and lanthanum oxide. The present investigation was undertaken to find a more rapid and cheaper method of determining niobium, tantalum, tin and titanium in concentrates of poly- mineral composition. Such a method would be of direct benefit to the mining industry in instances when quick and accurate results are required, as with beneficiation processes and the evaluation of mineral dumps.According to Sherman2 and Vera M&e,3 if a sample can be mixed with a definite proportion of an arbitrary diluent, it is possible to derive simple inert dilution equations containing measurable quantities only. With this technique, X-ray fluorescence measure- ments are made directly on a portion of the original sample, and on another portion after it has been mixed with a definite proportion of an inert diluent. By solving the intensity equations of these samples, an expression is obtained which is independent of the absorp- tion coefficient of the original sample. This inert dilution equation may be written as X = K.D(X).R(I), where the concentration, X , is dependent only on the intensity relation, R(1). D ( X ) is the dilution factor. The constant K for each analytical line, diluent, tube voltage and reference sample is determined from the inert dilution equation by using a standard sample.This value of K can be used with any sample provided the latter is prepared and measured under conditions similar to those used with the standard. The same two sample portions, the original and the diluted, are used for the determination of niobium, tantalum, tin and titanium. Equipment variables are almost completely eliminated by using a reference sample. The inexpensive calcined iron(II1) oxide (supplied by British Drug Houses Ltd.) is used as inert diluent. Iron(II1) oxide was chosen because it has been found to mix homo- geneously with heavy mineral concentrates that pass a 300-mesh sieve.* Among the major elements normally encountered in the tantalite - columbite concen- trates, yttrium and zirconium can interfere in the determination of niobium.To ascertain how these interferences could be overcome, five spiked samples containing various proportions of xenotime and zircon were prepared. The concentration of yttrium oxide (Y203) and zirconium dioxide (21-0,) in the spiked samples ranged from 5 to 30 per cent. Five concentrate samples were also analysed by the inert dilution method to determine niobium and tantalum pentoxides, tin dioxide and titanium dioxide, and the results obtained were compared with those determined chemically. 0 SAC and the authors.WONG AND SEEVARATNAM 563 EXPERIMENTAL SAMPLE PREPARATION- All samples were ground to less than 300 mesh in a Siebtechnik vibratory disc-type mill.The mixing of sample and diluent was also carried out in this mill. The diluted sample was prepared by mixing 0.5000 g of sample and 16000 g of iron(II1) oxide in the mill for 5 minutes. It was then made into a boric acid backed and edged button under a pressure of 10 tons inch-2. The original sample was prepared by pressing 2 g of the powdered concentrate into a button under the same pressure as for the diluted sample. INSTRUMENTATION- A fully stabilised all-vacuum Philips X-ray spectrograph, Model PW1540, with a pulse- height analyser attachment and a tungsten-target X-ray tube was used in this investigation. Measurements of X-ray intensities were made in air with the aid of a lithium fluoride crystal (2d = 0.4208 nm).Other relevant instrumental conditions are summarised in Table I. TABLE I OPERATING CONDITIONS FOR X-RAY ANALYSIS Pulse-height analyser L l W i n d o ; l / V Detector Voltage/ Current/ Element Line kV mA 30 10 0.650 1.225 Scintillation 32 20 0.850 1.700 Scintillation Tin Ku 40 20 0.638 0.925 Scintillation Titanium Ku 32 22 0.650 1.300 Flow-proportional Niobium KBI Tantalum La, For niobium, tantalum and tin the net counts of each element were obtained by sub- tracting the average value of the background counts from that taken at the analytical line. There was no need to correct for background in the determination of titanium. Table I1 lists the line and background angles (20) for the various elements. TABLE I1 SETTINGS FOR LINE AND BACKGROUND ANGLES Element Line/' Background/' Niobium .. . . 18.97 18-00; 19.90 (low Zr), 19-50 (high Zr) Tantalum . . . . 44-40 43-70 Tin . . . . . . 13.97 13.00; 15-00 ANALYSIS TIME- The analysis time for one sample analysed in duplicate was about 2 hours, which included preparation of samples, measurement of X-ray fluorescence intensities and calculations. Depending on the concentration of the element under analysis, the counting times on the line and background of the concentrate sample and diluted sample vary from 10 to I00 s. CALCULATION- The inert dilution equation for calculating the percentage of metal oxide is K.D ( X ) .Ic.Id Metal oxide, per cent. = I r (Ic - I d ) where K and D(X) are constants, and Ic, I d and Ir refer to the net counts of the concentrate, the diluted sample and reference sample, respectively.RESULTS AND DISCUSSION Analysis of the spiked samples showed that it was necessary to correct for background when the niobium K& line was used because of the proximity of the zirconium KP1 line. It can be seen in Table I11 that when background correction was not made the positive error became particularly large at low concentrations of niobium pentoxide. If the sample does not contain yttrium the use of the niobium Ka line is preferred because it gives better sensitivity. In the determination of tantalum, the second-order reflection of niobium Ka is664 WONG AND SEEVARATNAM almost completely removed by pulse-height analysis while the scattered tungsten Lor, is well resolved by the lithium fluoride crystal with the aid of fine collimation.Results presented in Table I11 show that niobium and tantalum can be determined accurately in the presence of yttrium and zirconium. TABLE I11 EFFECT OF BACKGROUND ON THE ANALYSIS OF NIOBIUM AND TANTALUM PENTOXIDES Nb,O,, per cent. r A 1 Sample Calculated Found Found* s1 53.7 53-4 54.5 s 2 46.1 46-6 48.8 s 3 30.7 30.4 33.6 s 4 15.4 15.1 18.4 s5 7.70 7.75 10.7 T+O,, per cent. Calculated Found Found* 10.9 11-1 11.3 9.30 9.48 9.80 6.20 6.40 7.15 3-10 3.18 4.16 1.55 1.46 2.62 I A I * These results have not been corrected for background. A comparison of results obtained by X-ray and chemical analysis is presented in Table IV. Results obtained by the inert dilution method are within +5 per cent. of those determined chemically. TABLE IV COMPARISON OF RESULTS OBTAINED BY X-RAY AND CHEMICAL ANALYSIS Nb,O,, per cent. Sample x-al A 59.4 58-4 B 58.5 58.3 C 67.6 56.8 D 21-8 22.3 E 19.9 20.5 TqO,, per cent.x-al 14.4 15.1 10.8 10.2 16.9 17.7 38.7 39.2 45.3 45-8 SnO,, per cent. x-al 1.50 1.57 1.79 1.88 2-24 2.29 6-14 5.28 8.00 7.97 TiO,, per cent. X E c a l 2-24 2-22 3.87 3.72 1.49 1.43 1-62 1.66 3.98 4-24 PRECISION- Ten samples of B were prepared and analysed four times. The calculated coefficients of variation corresponded to 0.30 per cent. for niobium pentoxide, 0.28 per cent. for tantalum pentoxide, 0.12 per cent. for tin dioxide and 0.08 per cent. for titanium dioxide. CONCLUSION The results presented here demonstrate that the powder X-ray fluorescence method, with iron(II1) oxide as inert diluent, can be used to provide rapid and accurate determinations of niobium, tantalum, tin and titanium in concentrates of different mineralogical composition. The simplicity of this method merits its use in routine analytical work. This paper is published with the permission of the Director, Geological Survey of Malaysia. We thank Dr. T. R. Sweatman for valuable discussion, Mr. E. H. Yin and Mr. A. G. Darling for their comments and suggestions in the presentation of this paper, Mr. Y. L. Tong (Senior Chemist) and his assistants, Mr. M. H. Yap and Mr. F. C. Thong, for carrying out the chemical analyses, Mr. S. H. Ow for his able experimental assistance and Mrs. Y. C. Wong for her encouragement. REFERENCES 1. 2. 3. 4. Rose, H. J., jun., and Brown, R., Adv. X-ray Analysis, 1963, 7, 598. Sherman, J., Ibid., 1957, 1, 231. Vera Mbge, R., Analyt. Chem., 1969, 41, 42. Sweatman, T. R., Wong, Y . C., and Toong, K. S., Trans. Instn Min. Metall., 1967, 76, B149. Received October 19t12, 1970 Accepted December 2nd, 1970

ISSN:0003-2654    

DOI:10.1039/AN9719600562

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, August, 1971, Vol. 96, pp. 565-568 565 Determination of Ammonium in Soil Extracts by an Automated Indophenol Method BY A. R. SELMER-OLSEN (Agricultural College of Norway, Vollebekk, Norway) An automated method is described for the determination of ammonium in 2 N potassium chloride soil extracts. The determination is based on the indophenol blue method following a dialysing step. As little as 0.03 mg 1-1 of ammonium-nitrogen can be determined. The recoveries have been investi- gated for a number of different soil types and satisfactory results were obtained. Determinations can be carried out at the rate of thirty samples per hour. INORGANIC nitrogen in soils may exist as ammonium, nitrate or nitrite. An automated method for determining nitrite and nitrate has previously been rep0rted.l Ammonium determinations, however, are mostly carried out by distillation with magnesium oxide, followed by titration or spectrophotometric determination.2 Hanawalt and SteckeP and Keay and Menage4y5 have automated the distillation procedure.A disadvantage of their technique, however, is that a special distillation apparatus is required. This paper describes an automated indophenol method, in which 2 N potassium chloride extracts of soils are analysed to determine exchangeable ammonium2 or ammonia formed by anaerobic incubation, which involves the use of a dialyser unit. METHOD APPARATUS- A Technicon AutoAnalyzer with a manifold construction, as shown in Fig. 1, was used. I 1.60 Wash v Dialyser 2.50 Sample n v Waste 2.00 Air 2.00 Air 2.50 Sodium hydroxide 1.60 Sodium phenate - 7 v 7 " \ f - - v Single 1.20 Sodium hypochlorite mixing coil Single 1 mixing coi I - 7 U J 2.50 F Waste + v Delay Colorimeter Recorder Flow diagram of apparatus used in automatic method for deter- coil Fig.1. mining ammonium in soil extracts 0 SAC and the author.566 REAGENTS- SELMER-OLSEN : DETERMINATION OF AMMONIUM IN SOIL [Analyst, Vol. 96 Potassium chloride, 2 N. Sodium hydroxide - sodium tartrate solution-Dissolve 1 g of sodium hydroxide and 5 g of potassium sodium tartrate in water and make the volume up to 1 litre. Alkaline Phenol solution-Dissolve 62.4g of phenol in 100 ml of 27 per cent. sodium hydroxide solution, cool, add 20 ml of acetone, and make the volume up to 1 litre with water. Sodium hypochlorite solution-Dilute 50 ml of commercial sodium hypochlorite solution (containing about 12 per cent. of available chlorine) with water to 1 litre.PROCEDURE- Shake 5-g samples of soil for 1 hour with 25 to 100 ml of 2 N potassium chloride, depending on the ammonium content, and filter the extracts through Schleicher and Schull No. 589 white ribbon filters. In determining ammonium after anaerobic incubation, the solution was diluted with an equal volume of 4 N potassium chloride so as to obtain a solution 2 N with respect to potassium chloride. The filtered extracts seemed to remain stable for several hours at room temperature. For longer storage, however, they should be placed in a refrigerator. Because the soil extracts can be coloured or contain particles the samples are dialysed in the AutoAnalyzer into a sodium hydroxide solution containing tartrate (to prevent precipita- tion of hydroxides).The ammonium present produces a blue colour with the alkaline phenol and hypochlorite. If the laboratory temperature is not constant it is advisable to place the delay coil in a heating bath at 22 oC.6 EFFECT OF pH- Interference will occur if the concentration of acids in the extracts is too high, because the amounts of alkaline reagents used in the system will not then neutralise the acids. On the other hand, if the extracts are alkaline there will be some loss of ammonia. Soils with pH values in the range 3.5 to 7.6 were extracted with 2 N potassium chloride. Known amounts of ammonium were added to aliquots of the extracts, and the ammonium contents were determined. Table I shows recoveries of 97 to 102 per cent.TABLE I RECOVERY OF AMMONIUM ADDED TO EXTRACTS FROM SOILS WITH DIFFERENT pH VALUES Ammonium-nitrogenlmg 1-1 soil Peat . . .. .. .. .. Sand (large organic matter content) Coarse sand . . .. .. .. Sand .. .. .. .. .. Sandy clay .. .. .. .. Clay . . .. .. .. .. Sandy peat . . .. .. .. Coarse sand . . .. .. .. Sand . . .. .. .. .. Coarse sand . . .. .. .. Sand .. .. .. .. .. P H 3-6 4.1 4-3 4.5 5.2 6-3 5.4 6.4 6.6 6-8 7.0 7.4 7.4 7.6 7.6 , Found in extracts without addition 0.71 0.71 0.46 0.2 1 2.15 0.67 0.54 0.53 1.69 1.83 0.53 0.48 0.42 0.18 0-52 Found in extracts with addition of 3.43 mg 1-1 of ammonium-nitrogen 4-10 4.05 3-80 3.60 6.60 4.10 4.00 3.93 5.20 5-30 3.95 3-93 3.90 3.62 3.90 I Recovered 3.39 3-34 3.34 3.39 3-45 3.43 3.46 3.40 3.51 3.47 3.42 3.45 3.48 3.44 3.38 INTERFERENCES- Known amounts of ammonium were quantitatively recovered from 2 N potassium chloride solutions containing at least 2 g 1-1 of the following salts: sodium sulphate, sodium sulphite, potassium nitrate, potassium nitrite, dipotassium hydrogen orthophosphate, potas- sium bromate, copper sulphate, aluminium sulphate, zinc sulphate, iron(I1) sulphate,August, 19711 EXTRACTS BY AN AUTOMATED INDOPHENOL METHOD 567 iron(II1) chloride, calcium sulphate, magnesium chloride, urea and guanidine.Up to 0.5 g 1-1 of asparagine, betaine, glucosamine, glutamine, galactosamine and diphenylamine did not cause noticeable interference; 10 per cent. of the nitrogen contained in hydroxylammonium chloride was determined as ammonium, probably because of the presence of ammonium in this chemical. More than 1 g 1-1 of manganese sulphate interfered seriously because precipi- tation occurred.If present in amounts exceeding 0.1 g l-l, methionine decreased the recovery of ammonium. RECOVERY OF ADDED AMMONIUM- Known amounts of ammonium chloride were added to 10-g samples of different soils prior to extraction with 50 ml of 2 N potassium chloride. The ammonium contents found are given in Table 11, and the results indicate that the recovery of added ammonium was satisfactory for all of the soils investigated. TABLE I1 RECOVERY OF AMMONIUM ADDED TO DIFFERENT SOILS Ammonium- nitrogen added per 100 g of soil/mg .. 0 4 8 20 Ammonium-nitrogenlmg f A 7 Sample Found Clay .. . . 2.15 3.50 8.00 2-80 Sand .. . . 4.10 3.50 Peat . . . . 8-20 12.1 20.0 22.0 - Found Recovered 6.30 7-40 6.90 8.00 7.20 12.1 12-1 16.2 23.8 25.9 4-15 3.90 4-10 4-10 3.90 3.70 3.90 4-10 3.8 3.9 7 Found 10.2 11.6 16.2 11.2 12.2 11.5 16.0 20.3 - - Recovered 8.05 8.10 8.20 8.40 8.10 8.00 7.80 8.20 - 7 Found 20.2 23.3 28.4 22-4 24-0 22.6 27.2 31.6 38.2 40.2 - Recovered 18-05 19.8 20.4 19.6 19.9 19.1 19.0 19-5 18.2 18.2 SENSITIVITY AND PRECISION OF THE METHOD- With the manifold shown in Fig. 1, ammonium in 2 N potassium chloride extracts could be determined within the range 0.1 to 20mg1-1 of ammonium-nitrogen. If a small peak follows a very large peak some contamination may occur, in which event it would be advisable TABLE I11 AMOUNTS OF AMMONIUM FOUND IN 2 N POTASSIUM CHLORIDE EXTRACTS BY DISTILLATION AND BY AUTOANALYZER Ammonium-nitrogen found in 100 g of soil/mg I A > Soil By distillation By AutoAnalyzer Clay .... .. 5.1 4.82 7.7 7.55 5.2 4.83 5.4 5.40 3.9 3.57 4.4 4.17 7.2 6.58 6.4 5-98 7.0 7.60 5.4 5.23 Silty clay . . .. .. 5.9 5.88 Silt . . .. .. .. 2.4 2-40 Sand .. . . . . 1.9 1-97 Mud from sewage . . . . 29 25.0 57 56.3 Sandy clay .. .. 4.6 4.73568 SELMER-OLSEN to carry out a wash between samples. If 5-g soil samples are extracted with 25 ml of 2 N potassium chloride, then 0.05 mg of ammonium-nitrogen per 100 g of soil can be determined. The sensitivity can be increased by using the range expander at x 2, x 4 or x 10 expansion. With x 10 expansion, the determination can be carried out in solutions containing 0.03 mg 1-1 of ammonium-nitrogen.On the basis of forty-one samples, the standard deviation of the method was found to be &O.OS mg 1-1 of ammonium-nitrogen within the range 0.4 to 15 mg 1-1 of nitrogen. On the basis of thirty-eight samples, with the range expander at x10 expansion, the standard deviation was found to be &0.012 mg 1-1 of ammonium-nitrogen within the range 0.03 to 0.3 mg 1-1 of nitrogen. The standard deviation given above is of the same order of magnitude as the errors in the reading of results from the calibration graph. COMPARISON OF METHODS A series of soil extracts was prepared, and the ammonium contents were determined directly by the automated method and by steam distillation with magnesium oxide, according to Bremner.2 The results shown in Table I11 indicate that the distillation method gave slightly higher results than those given by the AutoAnalyzer method, which is probably caused by interference from labile organonitrogen compounds destroyed by the steam distilla- tion with magnesium oxide and determined as ammonium. REFERENCES 1. 2. 3. 4. 5. -*- , Ibid., 1970, 95, 379. 6. Henriksen, A., and Selmer-Olsen, A. R., Analyst, 1970, 95, 514. Bremner, J. M., “Inorganic Forms of Nitrogen in Methods of Soil Analysis,’’ Part 2, Agronomy Hanawalt, R. B., and Steckel, J. E., “Automation in Analytical Chemistry, Technicon Symposium, Keay, J.. and Menage, P. M. A., Analyst, 1969, 94, 895. Davidson, J., Mathieson, J., and Boyne, A. W., Ibid., 1970, 95, 181. No. 9, American Society of Agronomy, 1965. New York,” 1966, Volume 1, p. 133. Received November 16th, 1970 Accepted April 5th, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600565

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, August, 1971, Vol. 96, pp. 569-578 569 A Field Method for the Determination of Zinc Oxide Fume in Air BY B. S. MARSHALL, I. TELFORD AND R. WOOD (Department of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, S.E. 1) A method is described for the determination of zinc oxide fume in industrial atmospheres a t concentrations up to 20 mg m4 of zinc oxide. The fume is collected on a filter and dissolved in acid, and the zinc is determined spectrophotometrically or visually with 4- (2’-thiazolylazo)resorcinol reagent. The apparatus used is simple and the time required for a determination is about 20 minutes. A dynamic method for the generation of atmospheres of zinc oxide is also described. THE main environmental health hazard from zinc, a metal widely and diversely used in indus- try, is from the inhalation of fume of freshly formed zinc oxide.It is well known that over- exposure to this can cause metal-fume fever.l Two of the main industrial processes that produce zinc oxide fume are the casting of zinc-containing alloys and the welding of galvanised steel sheet. A relatively low threshold limit value for zinc oxide fume of 5 mg per cubic metre of air is recommended at present.2 No doubt the total zinc content of an industrial atmosphere could be determined by the use of a suitable physical technique such as atomic-absorption spectrophotometry. However, in view of the toxicity and widespread occurrence, often in the smaller industrial establish- ments, of zinc oxide fume, the need appeared for a field test that required no sophisticated equipment yet could be used to determine the total zinc content of an air sample quickly and reliably. It was envisaged that any such test developed would involve the collection of the zinc-containing fume and dust on a suitable filter, rapid dissolution of the sample in acid and finally the colorimetric determination of the total zinc.It was recognised that the results obtained from such tests might include some zinc not originally present in the air as the oxide, but any over-estimation in this way of the zinc oxide content would err on the side of safety. PREPARATION AND CALIBRATION OF ATMOSPHERES OF ZINC OXIDE FUME Atmospheres of zinc oxide fume were required for use in the development of the field test.A search of the literature revealed no simple way of generating atmospheres of zinc oxide fume and in view of this the assistance of the Chemical Defence Establishment, Porton, was sought. They suggested a system whereby suitable atmospheres were produced by the pyrolysis in a bunsen burner flame of a zinc acetate aerosol issuing from a Collinson atomiser.3 A generator incorporating these basic principles was constructed. After several modifications had been made to the original generator to increase the yield of fume produced, the model shown in Fig. 1 was finally adopted as it provided the most consistent results. PREPARATION- Air at 15 p.s.i. (1.034 x lo5 N m-2) was passed through a Collinson atomiser, A (Fig. l), containing initially 100 ml of an aqueous solution of zinc acetate, B.The resulting aerosol was fed into the burner chamber, C, where it was used as the air supply to the Meker burner, D, which had a ceramic grid 24mm in diameter w5th 37 holes, each 2 mm in diameter. The burner had no gas adjustment and consumed about 4 1 minute-l of town gas, the un- wanted aerosol escaping through a vent for excess air. The burner chamber was fashioned from a tin can and the chimney, E, which was 0.72 m in height and 0-15 m in diameter, from tin plate. 0 SAC; Crown Copyright Reserved.570 MARSHALL, TELFORD AND WOOD: A FIELD METHOD FOR [Amdyst, Vol. 96 Air E ___) Excess air Fig. 1. Metal oxide aerosol generator: A, Collinson atomiser; B, zinc acetate solution ; C, burner chamber ; D, burner ; E, chimney; and F, filter-paper holder In the early experiments with the fume generator, the chimney was sited directly on top of the burner chamber.With this arrangement it was necessary to sample at a con- siderable height above the chimney to avoid distortion of the filters by heat; also, it was impossible to obtain reproducible results from duplicate samples taken simultaneously at the same site. The arrangement shown in Fig. 1, with a 12.5-mm gap between burner chamber and chimney, obviated these disadvantages and it was possible to sample with the filter-paper holder, F, just inside the top of the chimney. When compressed air, containing only 4 to 5mgm-3 of water vapour, was used to atomise the zinc acetate solution, a progressive increase in the concentration of the fume generated was noted.This was attributed to a gradual evaporation of water from the solution by the flow of air. The use of compressed air that had been passed through a water saturator prior to the atomiser minimised this effect. CALIBRATION- Our normal practice in developing a field test for any particular contaminant is to generate a constant known atmosphere of this contaminant. Despite the modifications described above, the generator was not capable of producing constant atmospheres of zinc oxide fume. There was invariably a tendency for a progressive increase in the concentration of successive atmospheres generated from any one zinc acetate solution, although the use of 5 and 15 per cent. w/v solutions in the above generator initially gave aerosol atmospheres of 4.4 and 9.2 mg m-3 of zinc oxide fume, respectively.In view of this inability to produce a constant atmosphere, the required calibration was carried out on collected samples of zinc oxide fume, which were analysed by different techniques. A series of these samples, which were adjudged to be in the range 0 to 100 pg, was first examined by a non-destructive method of analysis, X-ray fluorescence spectrometry, and the relative responses were recorded. EachAugust, 19711 THE DETERMINATION OF ZINC OXIDE FUME I N AIR 57 1 sample was then dissolved in acid and the zinc content determined by the proposed spectro- photometric method and an atomic-absorption method for which calibration graphs had been prepared previously by using a standard zinc solution.A graph of the X-ray fluorescence response against the average zinc content found by use of the two standard methods gave a straight line. These results were further analysed by means of a suitable computer pro- gramme to obtain the best straight line (see Fig. 2), which was subsequently used as the zinc calibration curve in the various stages of the development of the field test. SAMPLING AND COLLECTING TOTAL ZINC CHOICE OF FILTER-PAPER- Any sampling technique used in this work had to be capable of quantitatively collecting all particles with a diameter down to 10 nm, this being the lower limit of zinc oxide fume particle size.4 As it had been shown previously5 that the Millipore Filter, Type AA, is virtu- ally 100 per cent. efficient for trapping particles with a diameter down to at least 8 nm, when collected at a face velocity of 0.4 m s-l, this filter was considered suitable for the work in question.I t has also been independently recommended for the collection of metal fume.6 140 1 X-ray fluorescence response, counts s-’ ( x 10) Fig. 2. Relationship between concentration and X-ray fluorescence response for the zinc content of collected zinc oxide fume samples REMOVAL OF FUME SAMPLES FROM FILTERS- The linearity of the graph in Fig. 2 suggested that a collected sample of zinc oxide fume could be dissolved easily and quantitatively from a Millipore filter with dilute acid. This was confirmed by X-ray fluorescence analysis when only trace amounts of zinc (less than 1 pg) were detected on a series of filters, from each of which the zinc oxide fume (in the range 0 to 130 pg) collected on it had been dissolved by soaking it in 2 ml of 5 M hydrochloric acid for a few minutes.The removal of mixed metal fumes from filters is described later (see Interferences). COLORIMETRIC DETERMINATION OF ZINC A comprehensive survey of the literature showed that zinc is one of the most difficult metals to determine specifically except by a physical technique. A number of reagents that give colorimetric reactions were evaluated, but most were found to be subject to interference by one or more of the metals that could occur with zinc in an industrial atmosphere, e.g., lead, copper, tin and iron. Although zincon’ and compounds of the thiazo1ylazophenoP type were considered worthy of further examination as reagents, the relatively rapid fading of the zinc - zincon coloured complex detracted from the usefulness of this reagent in a field method. Of the several thiazolylazophenols examined by Kawases for the spectrophotometric determination of zinc, 4-(2’-thiazolylazo)resorcinol (TAR) was selected for the present work572 MARSHALL, TELFORD AND WOOD: A FIELD METHOD FOR [Alzdyd, VOl.96 because of its almost unique facility for forming a water-soluble zinc chelate and its com- mercial availability in the United Kingdom. Kawase8 had determined most of the optimum reaction parameters but it was decided, in the light of our requirements for a field test, to re-investigate the formation of the red zinc - TAR complex when using a 0.002 M hydrochloric acid solution containing 12.5 pg ml-l of zinc.CONCENTRATION OF TAR REAGENT- The original use8 of the reagent whereby the final reaction mixture contained 1 pg ml-l of TAR was found to give an adequate response for a spectrophotometric determination of zinc. OPTIMUM pH AND CHOICE OF BUFFER- By using a buffer solution of borax with sodium acetate, to which dilute hydrochloric acid was added to vary the pH, the previous worker8 found that the absorbance of the zinc-TAR complex was constant over the pH range from 7.4 to at least 8-4 and selected 7.5 as the optimum value. It'e substantially confirmed this finding but preferred to use a solution having a higher buffer capacity and also capable of giving a pH nearer the middle of the constant range. On theoretical grounds, and also to accommodate the acid in the eluate from the column used to separate zinc from iron (see section on Separation of iron), a buffer solution of triethanolamine and sodium hydroxide was devised such that the presence of 10 ml of this in the final reaction mixture gave a solution of pH 8.The buffering capacity of this final reaction solution can be gauged from the fact that the addition of 2.5 ml of N sodium hydroxide solution increased the pH to only 9.1 and the addition of the same volume of N hydrochloric acid gave a pH of 7.6. In neither case was the optical density resulting from a known amount of zinc altered. In contrast, similar experiments with the borax- sodium acetate buffer revealed a much reduced buffering capacity, for example, the effect of adding 2 ml of N acid to this buffer was to reduce the pH from 9.1 to 5.8.Later, to minimise the number of reagents required, it was found possible to make up the TAR reagent in the triethanolamine - sodium hydroxide buffer solution. This mixture was stable for at least 14 days at temperatures up to 30 "C. FORMATION AND STABILITY OF ZINC -TAR COMPLEX- The formation of the zinc-TAR complex appeared to be rapid and complete within the time lapse (about 1 minute) before a spectrophotometric reading could be taken. The colour was then stable for at least 30 minutes. The slight observed drop (approximately 3 per cent.) in optical density over the next hour was due to an equivalent increase in that of the reagent blank against which all samples were measured; this was not investigated further. At ambient temperatures, 15" to 30 "C, the colour developed was independent of temperature.PURITY OF TAR REAGENT- I t was intended to check the reliability of solid TAR reagent from various sources, but with material available from only one manufacturer in the United Kingdom, this was not possible. However, materials with identical batch numbers but received more than 1 year apart, and also a sample of the original material received (which had been re-crystallised from aqueous methanol after extraction of any impurities by using a sodium hydroxide solution - diethyl ether partition) , were compared with respect to their spectrophotometric responses with standard amounts of zinc. Identical optical densities were obtained. Further confirmation of the purity of the reagent was obtained by thin-layer chromatography on a cellulose plate by using a 2-ethoxyethanol- methanol - water mixture (4 + 1 + 2 v/v) as the mobile phase. Each TAR sample gave a single spot of RF 0.85.SPECTROPHOTOMETRIC DETERMINATION OF ZINC- It was established that 530nm was the optimum wavelength for the measurement of the zinc - TAR complex with respect to the reagent blank. Also, a plot of the optical density against the weight of zinc was linear over the range 0 to 70 pg, 70 pg of zinc (equivalent to 87.14 pg of zinc oxide) giving an optical density of 0.75 in a 10-mm cell. To ensure that repro- ducible samples of the test atmosphere would be taken, a sampling rate of 1 1 minute-l for a period of 5 minutes was chosen. Thus, the above calibration curve enabled atmospheresAugust, 19711 THE DETERMINATION OF ZINC OXIDE FUME I N AIR 573 of zinc oxide fume to be determined at concentrations up to 17.4 mg m-3, which is considerably more than twice the present threshold limit value.The choice of the spectrophotometric procedure finally adopted for use was governed largely by one requirement, namely the necessity to accommodate in the procedure the aqueous acetone - hydrochloric acid eluate from the ion-exchange column used to remove iron contamination prior to the determination of the zinc (see Separation of iron). This eluate, 20 ml in volume, contained 12 ml of acetone and 8 ml of 1.25 M hydrochloric acid. Thus, for samples containing no iron (the presence or absence of iron in an atmosphere can normally be ascertained a t the site of testing) the collected fume was dissolved in an equivalent amount of acid (2 ml of 5 M) and 12 ml of acetone were added.A uniform procedure was established thereafter whether or not the samples contained iron. With this procedure 10 ml of TAR reagent were added, the mixture was diluted to 50 ml with water and the resulting colour measured spectrophotometrically. VISUAL DETERMINATION OF ZINC- A visual method for the determination of zinc was devised, based on colour standards, by using the TAR reagent at a reduced concentration. Tests showed that the use of the reagent a t three twentieths of the concentration of that used in the spectrophotometric method gave the best colour differentiation between standards representing 0, 12.5, 25 and 50pg of zinc oxide.The colour differentiation was further improved by screening the yellow background colour of the TAR reagent by the addition of a Pontamine sky blue (C.I. 24410) dye solution. Later, it was found possible to add the dye to the dilute TAR reagent when the latter was being prepared. In this visual technique full colour development again occurred within 1 minute and the colour was stable for at least 20 minutes. However, the dilute TAR reagent was stable for only 24 hours. With the co-operation of Tintometer Ltd., a set of standard discs was prepared repre- senting the intensity of colours produced by collecting 5-litre samples of 0,2.5,5 and 10 mg m-3 atmospheres of zinc oxide. Colour matching was carried out by comparing a 50-mm depth of the sample solution with the standard discs.Although the visual method is not as precise as the spectrophotometric method, it permits the determination of the zinc oxide content of an atmosphere to at least 2.5 mg m-3 (k, half of the present threshold limit value) when a 5-litre sample is taken over 5 minutes. INTERFERENCES It has already been mentioned that the two main industrial processes that produce zinc oxide fume in the atmosphere are the smelting of alloys containing zinc and the welding of galvanised steel plate. Obviously, fumes of other metals may occur in concentrations sufficient to interfere in the determination of zinc with the TAR reagent. Consequently, an investigation was carried out to establish the interference that might be caused by each of a selection of metals present in the ratio of twice the threshold limit value of interfering metal to one half of the threshold limit value of zinc.(This has been a normal criterion upon which to base interferences in other field tests developed in the past. If as much as twice the threshold limit value of an interfering substance occurs in an atmosphere it is a hazard TABLE I INTERFERENCES O F METALS PRESENT I N A 5-LITRE SAMPLE AT TWICE THEIR THRESHOLD LIMIT VALUES IN THE DETERMINATION OF 1Opg OF ZINC* BY THE PROPOSED METHOD Amount of metal Indicated zinc/ Interference, Metal presentlpg Pg per cent. Antimony . , .. .. Arsenic . . .. .. Cadmium .. .. .. Calcium . . .. .. Copper . . .. .. Iron .. .. .. Lead . . .. .. Tin .. .. .. 5 10.0 5 10.0 0.97 10.2 35.8t 10.1 1 10-4 16.5 10.5 20 10.1 7;t 0 0 2 t I t 4 t 65 t 5t I t * Equivalent to 5 litres of an atmosphere of zinc oxide a t half its threshold limit value, Derived from the threshold limit value of the metal oxide.574 MARSHALL, TELFORD AND WOOD: A FIELD METHOD FOR [Analyst, Vol. 96 in its own right.) Solutions, each containing 10 pg of zinc (equivalent to 5 litres of an atmos- phere of zinc oxide at one half of its threshold limit value) and an amount of each interfering species (equivalent to 5 litres of an atmosphere of the interfering species at twice its threshold limit value), were analysed by the proposed method for zinc.The results are shown in Table I. Considering the aforementioned criterion, these interference levels, apart from iron, were considered to be sufficiently low to be ignored in the proposed field methods for the deter- mination of zinc, so that a further series of interference tests with 1 to 1 w/w ratio of zinc to interfering metal was carried out, this time leaving out iron.It was found that the percentage interference in the proposed test did not increase proportionately with the rise in concen- tration of interfering metal. Apart from copper, all interferences were less than 13 per cent. up to the 50-pg level. The interferences of copper were 60 and 55 per cent. a t the 10 and 40-pg levels, respectively. To test the effect of copper interference under conditions akin to those of field testing, mixed atmospheres of copper and zinc oxide fumes that were expected to correspond to a zinc- to-copper w/w ratio in the range from 10 to 1 to 120 to 1, were generated and sampled.(At ratios of less than 40 to 1 copper would be the greater hazard.) The zinc and copper content of each sample was first determined non-destructively by X-ray fluorescence , the sample was removed, apparently quantitatively, from the filter by using 2 ml of 5 M hydrochloric acid and the zinc content was then determined by the spectrophotometric version of the proposed field test. The results in Table I1 indicate that the presence of copper within the stated ranges in the collected mixed fume did not interfere either with the dissolution of the zinc fraction or in its determination by the proposed field method. TABLE I1 COMPARISON OF THE RESULTS OBTAINED BY ANALYSIS OF ZINC OXIDE FUME SAMPLES X-RAY FLUORESCENCE SPECTROMETRY CONTAINING COPPER FUME BY THE PROPOSED FIELD TEST AND BY Sample 1 2 3 4 5 6 7 8 Zinclpg, found by P X-ray fluorescence Field test spectrometry 187 195 194 185 177 183 285 294 361 357 39 1 396 582 575 695 684 Copper/ Clg found by X-ray fluorescence spectrometry 13.3 12.5 13.6 2-6 11.7 26.7 4.8 11.7 Ratio of zinc to copper 14.7 14.8 13.5 30-5 14.8 58.5 113 120 SEPARATION OF IRON- It was apparent in the treatment of samples taken when iron and zinc co-exist, e.g., in certain welding processes, that the iron would have to be separated from the zinc prior to the determination of the latter.The use of ion-exchange columns was investigated. Anionic resins suffered from the disadvantage that the iron had to be removed from the column before the zinc.However, on the basis of previous work9 it appeared that good separation of the two metals could be obtained by using a strong cationic resin in the acidic form and acetone - 0.5 M hydrochloric acid solution (3 + 2 v/v) as eluting agent, iron being retained on the column. Various parameters of the column procedure were examined in the light of our requirements. The use of prepared coarse Zeo-Karb 225 resin (1 g of 52 to 100 mesh) in a glass column (Fig. 3) was finally selected as it facilitates a rapid and efficient column procedure. This column was found to have an iron retaining capacity of 100 pg, Le., greater than the amount of iron present in a 5-litre sample of an atmosphere containing twice the threshold limit value of iron oxide.The fume samples requiring column treatment were dissolved in 1 ml of 50 per cent. v/v nitric acid and then diluted with aqueous acetone. This solution had no adverse effect on the functioning of the column. The use of nitric acid at this stage of the procedure was necessary to avoid the premature elution of the column under non- standard conditions, as would have occurred had hydrochloric acid been used.August, 19711 THE DETERMINATION OF ZINC OXIDE FUME IN AIR 575 The efficiency of the column was assessed by use of collected mixed fume samples of iron and zinc oxides that were expected to contain a zinc-to-iron ratio over the range from 1 to 1 to 1 to 8, w/w. Each sample was dissolved from its filter-paper as described above. The zinc content of an aliquot was determined by the atomic-absorption technique.Another aliquot was passed through the column and its zinc content determined by the spectrophotometric version of the field test. Iron was not detected in the column eluate. No attempt was made to determine accurately the iron content in the sample extract before its passage through the column because X-ray fluorescence examination of the filter-paper after its treatment with nitric acid revealed the presence of undissolved iron that comprised between 40 and 50 per cent. of the amount expected to have been collected. The amounts of zinc remaining on the various papers were minimal (less than 1 pg) and apparently independent of the total zinc collected, and were not considered to invalidate the proposed procedure for the dissolution of the mixed fume sample.The results in Table I11 indicate that the column procedure effected a separation of zinc from iron that allowed the determination of zinc by the proposed field test to proceed without interference. TAB~E I11 EFFICIENCY OF COLUMN PROCEDURE FOR THE SEPARATION OF ZINC FROM SAMPLES OF MIXED FUMES OF ZINC OXIDE AND IRON OXIDE Sample 1 2 3 4 5 6 7 8 Estimated* iron contentlpg 9 18 13 30 25 52 26 51 Zinclpg found by Atomic absorption Proposed field test before column after column procedure procedure 8.5 9.1 18.0 20.3 6.6 5.9 15.0 14.6 6.3 5.5 13.0 15.2 3.2 4.7 6.3 6.5 I A 5 * Estimated on the basis of the output of the metal fume generator when atomising an iron solution of known concentration. CONTAMINATION OF APPARATUS- A number of spurious results were obtained, which were traced to the incomplete removal of the zinc - TAR complex from the various pieces of glassware used, i.e., spectrophotometer cells, visual comparison tubes and the 50-ml flasks used for the development of the complex.Such contamination can be avoided by rinsing all such glassware with concentrated hydro- chloric acid. DETERMINATION OF ZINC OXIDE FUME IN AIR APPARATUS- Filter-paper holder-A holder suitable for Millipore AA filters (0.8 pm), 25 mm in diameter. Sampling PumP-A pump capable of drawing air through the filter-paper in the holder at a steady rate of about 1 1 minute-1. [This can be achieved either by using a critical orifice (such as Gelman Catalogue No. 7041) in conjunction with a pump capable of producing a pressure differential of at least 300mm of mercury, or by the use of a pump such as the Dymax IIA (Chas.Austen Ltd.) with a variable eccentric that can be adjusted to give a specific flow-rate.] lon-exchange column holder-A glass tube of shape and dimensions as shown in Fig. 3. TGbes for c o l o ~ ~ comparison-Flat-bottomed glass tubes, 10 mm in internal diameter and of a height that permits viewing vertically through a liquid depth of 50 mm. (Tintometer Ltd., Salisbury, supply pairs of tubes suitable for use in conjunction with the Lovibond “1000” comparator.)576 MARSHALL, TELFORD AND WOOD: A FIELD METHOD FOR [Analyst, VOl. 96 Fig. 3. Ion-ex- change column holder (dimensions in mm): A, glass tube 7 mm i.d. REAGENTS- should be made up with distilled or de-ionised water.at 20 "C) to 100 ml with water. All reagents should be of recognised analytical quality when possible and all solutions Hydrochloric acid, 5 M-Dilute 43.5 ml of concentrated hydrochloric acid (sp.gr. 1.19 Triammonium citrate solution-A 10 per cent. w/v aqueous solution, Nitric acid (1 + 1 v/v)-Dilute concentrated nitric acid with an equal volume of water. Acetone solution, aqueous-Dilute 6 volumes of acetone with 4 volumes of water. Acetone - hydrochloric acid solution-Add 10 ml of 5 M hydrochloric acid to 60 ml of acetone and dilute the mixture to 100ml with water. Preparation of ion-exchange resin-Place about 50 g of Zeo-Karb 225 ion-exchange resin (52 to 100 mesh, 8 per cent. divinylbenzene, sodium form) in a large glass column about 500 mm in length and 30 mm in internal diameter with a sintered-glass disc (porosity No.1) fitted in the lower part and terminating in a tap. Back-wash the column with distilled water to remove the fines, then calculate the bed volume of the resin (height x cross-sectional area of the wet resin). Drain off the water and wash with four bed volumes of the triammonium citrate solution, followed by a similar volume of 5 M hydrochloric acid. Finally, wash the resin with water until it is free from chloride, as indicated by testing the eluate with silver nitrate solution. Remove the excess water with a filter-pump and allow the resin to dry in air. Store in a screw-topped bottle. Preparation of ion-exchange column-Place a small plug of cotton-wool at the narrow end of the ion-exchange column holder and add a slurry of 1 g of the prepared resin in 20 ml of the acetone - hydrochloric acid solution. Allow nearly all of the liquid to pass through the resin and add a further 20 ml of the same solution to equilibrate the resin.Stopper the lower end of the column holder with a piece of plugged rubber tubing just as the meniscus enters the resin. Transport the prepared columns in the vertical position. 4-(2'-Tlziaxolylaxo)resorcinol (TAR) solution-A 0.1 per cent. w/v solution in methanol. TAR reagent (sohtion A)-Weigh 15 g of triethanolamine into a standard 100-ml flask, add 60 ml of N sodium hydroxide, 1.5 ml of TAR solution and, 1.5 ml of 0.168 per cent. w/v aqueous Pontamine sky blue (C.I. 24410) solution and dilute the mixture to 100ml withAugust, 19711 THE DETERMINATION OF ZINC OXIDE FUME IN AIR 577 water.Store at temperatures below 30 "C and renew after 24 hours. (Solution A is for use with the visual method for determining zinc.) TAR reagent (solution B)-Weigh 15 g of triethanolamine into a 100-ml flask, add 60 ml of N sodium hydroxide solution and 10 ml of TAR solution and dilute the mixture to 100 ml with water. Store below 30 "C and renew after 14 days. (Solution B is for use with the spectro- photometric method for determining zinc.) Standard zinc oxide solution-Dissolve 125.0 mg of zinc oxide in 4 ml of 5 M hydrochloric acid and dilute to 1 litre with water. Dilute 25 ml of this solution to 250 ml to give a solution containing the equivalent of 12.5 pg ml-l of zinc oxide.All standard glassware must be washed with concentrated hydrochloric acid and rinsed well with water to remove traces of metal ions and any TAR complex adhering from previous determinations. PROCEDURE- Place a filter in the filter holder, attach the assembly to the pump and draw a 5-litre sample of the atmosphere through the paper at a constant rate of about 1 1 minute-l. Dis- connect the holder from the pump, remove the filter and place it in a small beaker (diameter 25 to 30 mm). Samples containing no iron-Add 2 ml of 5 M hydrochloric acid and, after 5 minutes, transfer the acidic solution with a few millilitres of water to a 50-ml standard flask. Add 12 ml of acetone to the flask and determine the zinc either visually or spectrophotometrically as described below.SampZes containing iron-Add 1 ml of nitric acid (1 + 1 v/v) and, after 5 minutes, add 8ml of the aqueous acetone solution. Transfer the liquid on to the ion-exchange column, wash the beaker with 2 ml of aqueous acetone solution and add the washings to the column. Remove the stopper from the bottom of the column and allow the solution to pass through the resin until just before the meniscus reaches the top of the resin. Discard this eluate, add 20 ml of the acetone - hydrochloric acid solution and collect all the subsequent eluate from the column in a 50-ml standard flask. Determine the zinc either visually or spectro- photometrically as described below. VISUAL DETERMINATION OF ZINC OXIDE- Add 10 ml of TAR reagent solution A to the 50-ml flask, dilute to 50 ml with water and mix well.Fill a colour comparison tube to a depth of 50 mm with the solution and compare the colour in turn with each of the zinc oxide colour standards prepared at the same time and contained in similar tubes, viewing down the depths of the liquids against a white (paper) back- ground. Alternatively, a comparator disc containing coloured glass standards for this test is available from Tintometer Ltd., Salisbury, and should be used with the Lovibond "1000" comparator. Preparation of zinc oxide colour standards-To four 50-ml standard flasks add 0, 1, 2 and 4 ml of the standard zinc oxide solution (12.5 pg ml-l), respectively. Then to each add 20 ml of the acetone - hydrochloric acid solution and 10 ml of the TAR reagent solution A, and dilute to volume with water.For a 5-litre air sample these standards represent, respectively, 0, 2.5, 5 and 10 mg of zinc oxide per cubic metre of air. SPECTROPHOTOMETRIC DETERMINATION OF ZINC OXIDE- Add 10 ml of TAR reagent solution B to the 50-ml flask, dilute to volume with water and measure the optical density of the solution at 530nm in a 10-mm glass cell against a reference solution prepared at the same time from all of the reagents used. Determine the amount of equivalent zinc oxide in the solution by reference to the calibration graph. The concentration of zinc oxide present in the sample of air taken is given by X / 5 mg m-3, where X is the total amount of zinc oxide found in micrograms. Preparation of calibration graph-To a series of 50-ml standard flasks add 0, 1, 2, 3, 4, 5 and 6 ml of the standard zinc oxide solution.Then, to each flask add 20 ml of the acetone - hydrochloric acid solution and 10 ml of the TAR reagent solution B, and dilute each mixture to 50 ml with water. Measure the optical densities of the solutions at 530 nm in a 10-mm cell with the solution containing no zinc oxide as reference. Plot a graph of the weight of zinc oxide in micrograms against the optical density.578 MARSHALL, TELFORD AND WOOD APPLICATION AND SCOPE OF THE METHOD The procedures described above have been satisfactorily assessed with samples taken under field conditions at industrial establishments. Mixed atmospheric fumes of zinc oxide and copper were sampled in the vicinity of brass alloy casting. The mixed zinc oxide - iron fumes generated during the welding of galvanised steel sheet were also sampled.In the experiments, for the purpose of check testing and to obviate any possible errors caused by non-reproducibility of sampling, it was decided not to use the method of duplicate sampling whereby one sample would be examined by the proposed field test and the other by an independent method. Instead, the collected fume samples were returned to the laboratory where they were first examined non-destructively by X-ray fluorescence spectrometry and then by the appropriate field method. The apparatus required for the test is portable and requires only an external electrical power supply to operate the pump. A determination can be completed in 20 minutes. Permanent glass standards are available that allow the determination of zinc oxide fume in air over the range 2.5 to 10 mg m-3 from a sample taken at a rate of 1 1 minute-l for 5 minutes. More precise determinations are possible with the use of a spectrophotometer and a previously prepared calibration graph. Here, the sample volume need not be restricted to 5 litres taken at a rate of 1 1 minute-1, as when using visual permanent standards, and sample volumes ranging from 2.5 to 50 litres have been taken at sampling rates ranging from 0.5 to 5 1 minute-l, the zinc oxide contents being successfully determined. The use of a spectrophotometer and a calibration graph therefore allows for greater flexibility in sampling parameters. This work was carried out on behalf of the Department of Employment Committee on Tests for Toxic Substances in Air. We thank the Government Chemist for permission to publish this paper, H.M. Factory Inspectorate for arranging the field tests, the Chemical Defence Establishment, Porton, for advice on the generation of zinc oxide fume atmospheres and Mr. D. M. Groffman for his technical assistance. 1. 2. 3. 4. 6. 6. 7. 8. 9. REFERENCES Browning, E., “Toxicology of Industrial Metals,” Second Edition, Butterworths, London, 1969, “Threshold Limit Values 1969,” Technical Data Note 2/69, Department of Employment and Green, H. L., and Lane, W. R., “Particulate Clouds: Dusts, Smokes and Mists,” E. & F. N. Spon “Characteristics of Particles and Particle Dispersoids,” SRI J., 1961, 5; 95. Megaw, W. J., and Wiffen, R. D., Int. J . Air Wat. Pollut., 1963, 7, 501. Farrah, G. H., J . Air Pollut. Control Ass., 1967, 17, 738. Rush, R. M., and Yoe, J. H., Analyt. Chem., 1954, 26, 1345. Kawase, A., Talanta, 1965, 12, 195. Fritz, J. S., and Rettig, T. A., Analyt. Chem., 1962, 34, 1562. p. 352. Productivity, London, 1969. Ltd., London, 1957, p. 39. Received March loth, 1971 Accepted March 31st, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600569

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, August, 1971, Vol. 96, p p . 579-583 579 Polarographic Determination of Uranium in Monazite Sands BY R. W. MARTRES AND J. J. BURASTERO (Administracidn Nacional de Combustibles, Alcohol y Portland, Centro de Investigaciones Tecnoldgicas, Pando, Uruguay) A practical method is described for the extraction of uranium from monazite sands and its subsequent polarographic determination. The sample is decomposed by fusion with potassium hydrogen difluoride followed by fusion with potassium pyrosulphate. Uranium is extracted from nitric acid solution with tributyl phosphate in 2,2,4-trimethylpentane, with aluminium nitrate as a salting-out agent, and back-extracted from the organic phase with water. The final polarographic determination is carried out in 2 M acetic acid - 2 M ammonium acetate - 0.1 M ascorbic acid solution as supporting electrolyte. Neither a maximum suppressor nor removal of oxygen is needed.The interference by lead and some factors influencing the extraction of uranium are studied. The results are reproducible and agree with those obtained by other, more laborious, techniques. The proposed procedure is suitable for the determination of uranium in monazites and monazite sand concentrates containing not less than 0.005 per cent. of uranium oxide, and is superior in speed, reliability and convenience to other methods previously reported. MANY methods involving the application of classical and instrumental analytical techniques have been developed in recent years for the determination of uranium. However, only a few titrimetric, gravimetric, fluorimetric, spectrophotometric or polaro- graphic procedures that are specifically applicable to the determination of uranium in monazite have been described.Many of them proved to be too time consuming, insufficiently sensitive or lacking in accuracy for this purpose. The polarographic method has found considerable application in the determination of uranium, and its sensitivity is adequate for normal uranium contents in monazite sands. Moreover, rigorous preliminary purification is not required, provided a suitable supporting electrolyte is used. However, the polarographic procedures described for the specific determination of uranium in monazite1s2s3s4 are not reliable and are too time consuming because of the laborious purification steps needed.Further, they do not give good waves, and one of them in particulafl gives very poorly defined waves. By contrast, the proposed method is rapid and accurate. The preliminary separation of uranium is accomplished easily with a single solvent extraction after rapid decomposition of the sample. In the chosen supporting electrolyte the wave is well formed and easy to measure, as shown in Fig. 1. APPARATUS- EXPERIMENTAL A Sargent Polarograph, Model XV, was used to record all polarograms. The electrolysis cell was an H-type polarographic cell, Sargent S-29405, that permits the use of a sample as small as 1 to 2 ml. However, in routine determinations, a simpler polaro- graphic cell with a mercury pool can be used. The conventional dropping-mercury electrode was used as the cathode.The reference electrode was the S.C.E., which filled one chamber of the cell. 0 SAC and the authors.580 MARTRES AND BURASTERO : POLAROGRAPHIC [Analyst, Vol. 96 Applied voltagelv versus S.C.E. Fig. 1. Typical polarogram MONAZITE SAN- Samples I, I1 and I11 were separated from black sand concentrates obtained from the Uruguayan deposits of Atlhtida, Aguas Dulces and San Luis, respectively. Sample IV was separated from a monazite sand from Cleveland County, North Carolina, U.S.A.* REAGENTS - Standard uranium solution-A solution containing about 1 mg ml-l of uranium was pre- pared by dissolving uranyl nitrate in distilled water, diluting to an appropriate volume and mixing. The solution was standardised by reducing the uranium(V1) to uranium(1V) in a Jones reductor and then titrating with a standard potassium dichromate solution.Commercial grade 2,2,4-trimethylpentane (isooctane) t and commercial grade tributyl phosphate$ were used without further purification. PROCEDURE- Place 0.5 g of finely ground monazite (200 mesh) in a platinum crucible and mix it with 2 g of potassium hydrogen difluoride. Heat the mixture gently over a small flame. Gradually increase the flame to bring the contents to the melting-point. Cool for a few seconds, add 0.5 g of potassium pyrosulphate and continue heating until a clear melt is obtained. Cool, add 3 ml of 65 per cent. nitric acid, and evaporate to dryness. Repeat the treatment twice, taking care to ensure that the residue is well mixed with the acid after each addition.Dissolve the residue in M nitric acid and filter the solution into a separating funnel. Wash the crucible and filter until a final total volume of about 30 ml is obtained. Add 15 g of aluminium nitrate. Extract the uranium by shaking the solution for 2 minutes with 20ml of a 10 per cent. 2,2,4-trimethylpentane solution of tributyl phosphate. Discard the aqueous phase. Back-extract the uranium from the organic phase by shaking it with three successive 20-ml portions of water. Add 3 ml of 72 per cent. perchloric acid to the combined aqueous extracts and evaporate the resulting solution to dryness. Repeat the treatment with perchloric acid and again evaporate the solution to dryness. Dissolve the residue in 5 ml of 4 M acetic acid - 4 M ammonium acetate solution.Transfer this solution to a 10-ml calibrated flask containing 176 mg of ascorbic acid. After dissolution of the ascorbic acid, dilute to volume with water and mix thoroughly. Transfer a suitable volume of this solution to the polarographic cell and record the polarogram between -0.2 and -0.7 V veysus S.C.E. Measure the height of the step and determine the uranium con- centration by reference to a calibration graph. DECOMPOSITION OF MONAZITE- Many methods for the decomposition of monazite sands5,6’7,8 have been considered with a view to selecting the most suitable according to the nature of the determination required. * Supplied by Ward’s Natural Science Establishment Inc., Rochester, N.Y., U.S.A. t Supplied by Phillips Petroleum Corporation, U.S.A.$ Obtained from Matheson Coleman & Bell, U.S.A. RESULTS AND DISCUSSIONAugust, 19711 DETERMINATION OF URANIUM I N MONAZITE SANDS 581 The classical treatment with sulphuric acid was excluded because sulphate interferes when it is present in large amounts in the uranium extraction, making its removal necessary. Fusion with sodium peroxide is very effective but produces a bulky precipitate that is difficult to filter off, while a mixed flux of sodium fluoride and potassium pyrosulphate sometimes failed to bring about complete decomposition. On the other hand, fusion with potassium hydrogen difluoride followed by fusion with potassium pyrosulphate permits complete and rapid decomposition, and the residue is easily filtered off. After the fusion, repeated evaporations with 65 per cent.nitric acid ensure complete removal of fluoride. Moreover, in the presence of the sulphate introduced during the decomposition step, most of the lead remains in the residue when the melt is leached with M nitric acid. This moderate amount of sulphate does not interfere in the uranium extraction. EXTRACTION OF URANIUM- Among the various methods reported in the literature for the separation of uranium, solvent extraction with tributyl phosphate appears to be a simple and rapid way of isolating the element prior to the polarographic determination. If aluminium nitrate is used as a salting-out agent, virtually complete removal of uranium is achieved with a single extraction. The interferences caused by the phosphate present in the sample and the fluoride used for decomposition of the sample are eliminated by using aluminium nitrate at high concen- tration as salting-out agent.s In the method of attack used sulphate will be present.Preliminary experiments indicated that aluminium nitrate was also effective in preventing interference by the sulphate, which may be present in amounts up to 1 g. The results given below were obtained by extracting an aqueous phase, consisting of 30 ml of M nitric acid containing 15 g of aluminium nitrate, 2.56 mg of U,O, and various amounts of sulphate, with 20 ml of a 10 per cent. v/v solution of tributyl phosphate in 2,2,4-trimethylpentane. The uranium was back-extracted and determined polarographically- Sulphate added/g . . .. .. 0 0.35 0.71 1.06 1.42 Uranium found, step heightlmm .. 82 82 83 81 77 The presence of nitric acid is advantageous as it shifts the equilibrium to the left, thereby favouring the formation of uncomplexed uranyl ion.10 2,2,4-Trimethylpentane was used as inert diluent because the solution of tributyl phosphate in 2,2,4-trimethylpentane separates cleanly and rapidly from the aqueous phase, BACK-EXTRACTION OF URANIUM- Many aqueous solutions of different substances have been investigated as extraction media in the back-extraction of uranium from the organic phase into the aqueous phase, selection being made according to the efficiency of back-extraction and the method used subsequently for the uranium determination. In all respects, water is the most suitable back-extraction agent from the standpoint of the subsequent uranium determination.However, it has not been found very effective with tributyl phosphate containing nitric acid, the complete back-extraction being tedious and time consuming. The nitric acid concentration of the organic phase has a marked influence on the efficiency of back-extraction. In general, the lower the nitric acid concentration of the organic phase, the more effective is the aqueous back-extraction. In the procedure proposed in the present work, we use a M nitric acid solution to dissolve the uranium residue prior to the extraction with tributyl phosphate . Under these conditions, complete back-extraction from the organic phase was achieved with three successive extrac- tions with equal volumes. INTERFERENCE BY LEAD- Lead interferes seriously in the supporting electrolyte used, as it gives a wave that coalesces additively with the uranium wave, and its extraction under the conditions used was, therefore, studied.The results given in Table I indicate that the amount of lead U022+ + HSO, + UO2SO, + H+582 MARTRES AND BURASTERO : POLAROGRAPHIC [Analyst, Vol. 96 extracted decreases as the nitric acid concentration in the aqueous phase increases and as the tributyl phosphate concentration in the organic phase decreases. Further, in the presence of sulphate the amount of lead extracted also decreases. Therefore, sulphate introduced into the analysis during the fusion is beneficial. TABLE I EXTRACTION OF LEAD Aqueous phase: 15 ml of solution containing 10 mg of lead Organic phase: 10 ml of a solution of tributyl phosphate in 2,2,4-trimethylpentane Salting-out agent : 7.5 g of aluminium nitrate Aqueous phase concentration/M Sulphate added/g r A \ Nitric acid 0.1 - 1 0.1 - 1 1 0.2 - - Tributyl phosphate in organic phase, Lead extracted, per cent.per cent. 20 1.66 20 0.70 10 0.80 10 0.60 10 0.30 In the procedure proposed the uranium is extracted from a M nitric acid solutioncon- taining 0.5 g ml-l of aluminium nitrate with a 10 per cent. solution of tributyl phosphate in 2,2,4-trimethylpentane. Molar nitric acid is selected as it prevents the formation of uranyl sulphate complex, decreases the amount of lead co-extracted and prevents the forma- tion of precipitates (of titanium, thorium, etc.) by hydrolysis, which produce emulsions. On the other hand, this level of acidity is sufficiently low to permit rapid and complete back-ex trac tion with water.Under the conditions of the proposed method and for the amounts of uranium and lead usually present in monazites, lead is not extracted in interfering amounts. However, with ratios of lead to uranium higher than about 5 : 1 , trace amounts of lead may cause a small error. In this event, separation of lead is accomplished by extraction with dithizone as des- cribed below. Add 1 ml of acetic acid and 100 mg of hydroxylammonium chloride, which prevents the precipitation of uranium, to the combined aqueous extracts, neutralise with ammonia solution (to a yellow colour with methyl red) , and shake with a solution of dithizone (20 mg 1-1) in chloroform. If metals are precipitated by the ammonia solution, add 200 mg of citric acid and then neutralise the solution (with bromothymol blue as indicator) to give a higher pH than that obtained when methyl red is used.If citrate is used, it must be completely removed from the final solution to avoid changes in the diffusion current and in the wave form. It can be removed by repeated alternate evaporations with nitric and perchloric acids. POLAROGRAPHY OF URANIUM : CHOICE OF SUPPORTING ELECTROLYTE- The polarography of the uranyl ion has been studied by numerous workers, who used different supporting electrolytes. Whichever supporting electrolyte was used, difficulties were experienced in the polaro- graphic determination of uranium in monazite sands because of interfering elements that had to be removed.Many ions interfere in this determination, usually because they are reduced in the same potential region as the uranyl ion itself. Consequently, in the analysis of uranium ores, various methods have been used to remove interfering elements prior to the polarographic determination, namely, electrolysis at a mercury pool, chromatography, ion exchange and solvent extraction. After separation of the uranium from the bulk of the impurities, the selected supporting electrolyte must enable a well defined wave to be obtained with no interference by any remaining impurities accompanying the uranium. DeSesa, Hume, Glamm and DeFordll studied the polarographic characteristics of various metal ions in 2 M ammonium acetate - 2 M acetic acid solution containing 0.01 per cent.of gelatin. The uranyl ion showed a well defined and reversible wave with E, - 0-45 V versus S.C.E. , corresponding to reduction to the +5 oxidation state, so that the use of this solution for analytical purposes has been suggested. Although the sensitivity in this supporting electrolyte is only moderate, it is adequate for determining normal uranium contents in monazite sands.August, 19711 DETERMINATION OF URANIUM I N MONAZITE SANDS 583 Among the elements present in monazite sands, only lead (E, - 0.50 V vemm S.C.E.) can interfere with the uranium wave, as discussed above. Iron(II1) extracted together with uranium may interfere with the base-line of the uranium wave if present in relatively large amounts. This interference is eliminated by reducing it to the iron(I1) state, which is discharged at much more negative potentials than those of the uranium wave. This reduction is frequently accomplished by treating the solution with hydroxylamnionium chloride and warming it for a few minutes.However, the reduction is more easily carried out by addition of ascorbic acid.12 Ascorbic acid has also been proposed as supporting electrolyte for the polarographic determination of uranium in the presence of many other cations. l3 The polarographic properties of ascorbic acid as supporting electrolyte are based on its strong reducing power and complex- forming capacity. Further, at pH values above 3.5 it is not necessary to remove dissolved oxygen by bubbling an inert gas through the solution because ascorbic acid rapidly reduces oxygen.Thus, the supporting electrolyte finally selected consisted of a solution 2 M in ammonium acetate, 2 M in acetic acid and 0.1 M in ascorbic acid. In this medium the wave form is excellent, with E, - 0.455 V versus S.C.E. and a well developed base-line. It is not necessary to de-aerate the solution, or use a maximum suppressor. to 1.3 x 1 0 - 3 ~ uranium (VI) concentration. APPLICATION OF PROCEDURE- The given procedure is suitable for the rapid and accurate determination of uranium in monazites and monazite sand concentrates containing not less than 0.005 per cent. of uranium oxide (U 308). The results obtained by the method proposed agree with those found by the peroxide - spectrophotometric method5J4 and the thiocyanate - spectrophotometric method,15 as shown in Table 11.TABLE I1 COMPARATIVE URANIUM DETERMINATIONS IN MONAZITE SAMPLES The calibration graph was almost linear over the range 2.3 x U,O,, per cent., by Relative Sample Other methods Proposed method value deviation per cent. I 0.26 0.27, 0.24 0.25 0.015 6.0 I1 0.19 0.19, 0.19 0.18 0.014 7.8 I11 0.14 0.14, 0.13 0.14 0.010 7.1 IV 0.36 0.35, 0.37 0.37 0.015 4.1 r A \ Mean Standard standard deviation, 0.24, 0.26 0.16, 0.18 0.13, 0.15 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 0.38 REFERENCES Burstall, F. H., and Wells, R. A., Analyst. 1951, 76, 396. Habashi, F., Metallurgia, 1962, 65, 255. Schoeller, W. R., and Powell, A. R., “Analysis of Minerals and Ores of the Rarer Elements,” Smales, A. A., P ~ o c . Roy. SOC. Edinb., 1948, 63B, Part 11, 125. Grimaldi, F. S., May, I., Fletcher, M. H., and Titcomb, J., “Collected Papers on Methods of Powell, R. A., and Kinser, C. A., Analyt. Chem., 1958, 30, 1139. Sill, C. W., Ibid., 1961, 33, 1684. Vickery, R. C., “Analytical Chemistry of the Rare Earths,” Pergamon Press, New York, 1961, p. 6. Burastero, J. J.. and Martres, R. W., “Octavo Congreso Latinoamericano de Quimica,” Buenos Collopy, T. J., and Stock, D. A., US. Atomic Energy Commission Re+ort, NLCO-801, 1960. DeSesa, M. A., Hume, D. N., Glamm, A. C., jun., and DeFord, D. D., Analyt. Chem., 1953, 25, 983. Mashall, J., and Kendler, J., Analytica Chtim. Acta, 1964, 31, 490. Susic, M. V., Gal, I., and Cuker, E., Ibid., 1954, 11, 586. Rodden, C. J., Editor, “Analytical Chemistry of the Manhattan Project,” McGraw-Hill, New Clinch, J., and Guy, M. J., Analyst, 1957, 82, 800. Third Edition, Griffin, London, 1955, p. 110. Analysis for Uranium and Thorium,” US. Geological Survey Bulletin, 1954, No. 1006. Aires, 1962, p. 148. York, 1950, p. 147.

ISSN:0003-2654    

DOI:10.1039/AN9719600579

出版商:RSC    

年代:1971

数据来源: RSC