ISSN:0003-2654    

DOI:10.1039/AN97904FX037

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

ANALAO 104 (1 242) 897-992 (1 979)ISSN 0003-2654October 1979THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYCONTENTS897 Kinetics of a- and p-Molybdosilicic Acid Formation-Victor W. Truesdale, Peter J.Smith and Christopher J. Smith91 9 Spectrophotometric Determination o f Acetaminophen and SalicylamideThrough Nitrosation and Subsequent Chelation-Saied F. Belal, M. Abdel-HadyElsayed, A. Elwalily and H. Abdine928 Spectrophotometric Determination of Aliphatic lsocyanates in the OccupationalAtmosphere. Part 1. Determination of Total Isocyanate Concentration-R. F. Walker and M. A. Pinches937 Influence of Conditioning Agents on the Determination o f Metallic Content ofSewage Sludge by Atomic-absorption Spectrophotometry with Electro-thermal Atomisation-M.J. T. Carrondo, R. Perry and J. N. Lester944 Determination o f Antimony and Other Trace Elements in Irons and Steels byAtomic-absorption Spectrophotometry w i t h Introduction of Solid Samplesinto an Induction Furnace-A. M. Aziz-Alrahman and J. B. Headridge952 Quantitative Determination o f Lead Dioxide Polymorphs by X-ray PowderDiffractometry-P. R. Skidmore and R. R. Schwarz961 Calcium lon-selective Electrode Measurements in the Presence of ComplexingLigands-A. Craggs, G. J. Moody and J. D. R. Thomas973 Interferences of a Barium lon-selective Electrode Used for the PotentiometricTitration o f Sulphate-Dilys L. Jones, 6. J. Moody, J. D. R. Thomas and MagdolnaHangosSHORT PAPERSDetection of L-Cysteine, Methionine, Thiourea, Allylthiourea and a-Naphthyl-thiourea in Sub-milligram Amounts Using Acidified Potassium ChromateSolution-M.Nasim Beg, Fasih A. Siddiqui, M. Mohtashim Beg, R. Shyam andM. Arshad977979 Implementation of a Sensitive Method for Determining Mercury in SurfaceWaters and Sediments by Cold-vapour Atomic-absorption Spectrophoto-metry-R. L. Lutze983 Spectrophotometric Determination o f Trace Amounts o f Boron in SolutionsContaining Large Amounts of Nitrate-H. J. Rosenfeld and A. R. Selmer-Olsen985 Spectrophotometric Determination of Nitrite Using 4,5-Dihydroxycoumarin-Motoshi Nakamura and Akira Murata989 Book ReviewsSummaries of Papers in this Issue-Pages iv, vi, vii, x, xiiPrinted by Heffers Printers Ltd Cambridge EnglandEntered as Second Class at New York, USA, Post OfficANALYTICAL SCIENCES MONOGRAPH No.4Electrothermal Atomization forAtomic Absorption Spectrometryby C. W. FullerAt the present time the two most successful alternatives to the flame appear to bethe electrothermal atomizer and the inductively-coupled plasma. In this book anattempt has been made to provide the author's views on the historical develop-ment, commercial design features, theory, practical considerations, analyticalparameters of the elements, and areas of application of the first of these twotechniques, electrotherma I atomization.The chapter headings are as follows: History; Theoretical Aspects of theAtomization Process; General Experimental Conditions; Analytical Conditionsfor the Determination of the Elements by Atomic Absorption Spectrometry;Applications (Oil and Oil Products; Metals; Rocks, Minerals and Soils; Waters;Plants; Food and Drugs; Biological Fluids; Biological Tissues; Air Particulates;Refractory Oxides and Related Materials; Other Analytical Applications;Theoretical).Clothbound 135pp 82" x 5" 0 85186 777 4 f6.75 ($13.50)CS Members f5.50THE CHEMICAL SOCIETYDistribution Centre, Blackhorse Road, Letchworth,Herts., SG6 1 HN, EnglandNotice to SubscribersFor information on subscription rates for The Analyst, AnalyticalAbstracts and Proceedings write to:The Marketing Department,The Chemical Society,B urli ngton House,London WIV OBN, England.Telephone 01 -734 9864.Telex: 268001.Orders and claims for missing issues should be sent to:The Chemical Society, Distribution Centre,Blackhorse Road, Letchworth, Herts., SG6 1 HN, Englan

ISSN:0003-2654    

DOI:10.1039/AN97904BX039

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

SUMMARIES OF PAPERS IN THIS ISSUE October, 1979Summaries of Papers in this IssueKinetics of a- and P-Molybdosilicic Acid FormationThe kinetics of formation of a- and P-molybdosilicic acid have been investi-gated and rate equations proposed, together with possible reaction sequencesinvolving individual molybdate species. The a-acid was found to formaccording to a sum of exponentials model [At = 8, + B,exp( Q) + 8,exp( 8,t)lrelating absorbance with time. In contrast, the P-acid formed according toa simpler exponential model [A, = 8, + e,exp(O,t)], which is consistentwith pseudo-first-order behaviour with respect to silicate concentration.For the first time in a kinetic study of heteropolymolybdate formation, theeffects of changes in molybdate speciation upon the kinetics have been con-sidered in detail.The concentrations of individual molybdate species a tpH between 0.5 and 5.0, and total molybdate concentrations up to 0.20 M asmolybdenum, have been calculated using equilibrium constants together withan existing computer program. The most interesting result of the study isthat the observed a-acid kinetics demand the presence of a third silicon speciesin addition to the reactant silicate and the a-acid product. At this stage itis not known whether the third species should be another silicate condensationproduct or a molybdosilicic acid. The general application of the results ofthis study to silicate analysis is discussed. Methods of non-linear parameterestimation have been used extensively in fitting theoretical models to observeddata.Keywords ; Molybdosilicic acids ; molybdate speciation ; kinetics ; modellingkineticsVICTOR W.TRUESDALE, PETER J. SMITH and CHRISTOPER J. SMITHInstitute of Hydrology, Maclean Building, Crowmarsh Gifford, Wallingford, Oxford-shire, OX10 8BB.Analyst, 1979, 104, 897-918.Spectrophotometric Determination of Acetaminophen andSalicylamide Through Nitrosation and Subsequent ChelationA nitrosation reaction has been adopted for the spectrophotometric deter-mination of acetaminophen and salicylamide. The selectivity of the reactionis increased through utilisation of the nitroso derivatives as chelating agentsfor cobalt(I1) and copper(I1) ions. The optimum experimental conditionsfor the application of nitrosation and nitrosation with subsequent chelationwere established.The proportions of reactants in each method and theinstability constants for the products were determined. The nitroso deriva-tives and their chelates obey Beer’s law and their absorbances were used forthe determination of acetaminophen and salicylamide in pharmaceuticalformulations. The proposed methods gave more accurate results than theofficial methods.Keywords : Spectrophotometry ; acetaminophen determination ; salicylamidedetermination ; nitroso derivatives ; chelatesSAIED F. BELAL, M. ABDEL-HADY ELSAYED, A. ELWALILY and H.ABDINEDepartment of Pharmaceutical Analytical Chemistry, Faculty of Pharmacy, Uni-versity of Alexandria, Alexandria, Egypt.Analyst, 1979, 104, 919-927vi SUMMARIES OF PAPERS I N THIS ISSUESpectrophotometric Determination of Aliphatic Isocyanates inthe Occupational Atmosphere.October, 197’9Part I.Determination of Total Isocyanate ConcentrationThe paper describes a spectrophotometric method for use in the field for thedetermination of aliphatic isocyanates and their oligomers in air. Theatmosphere being tested is drawn through a mixture of hydrochloric acidand dimethylsulphoxide a t a sampling rate of 2 lmin-l for 10 min. Any isocyanates or oligomer present are hydrolysed to the corresponding amine.l-Fluoro-2,4-dinitrobenzene is added and forms coloured derivatives with theamines. The absorbance of each derivative is measured a t approximately353 nm, and can be directly related to the amount of isocyanate. Concentra-tions down to 0.002 p.p.m.can be determined.Keywords : Aliphatic isocyanate determination ; spectrophotometry ; 1-jluoro-2,4-dinitrobenzene reagentR. F. WALKER and M. A. PINCHESOccupational Medicine and Hygiene Laboratories, Health and Safety Executive403-405 Edgware Road, Cricklewood, London, NW2 6LN.Analyst, 1979, 104, 928-936.Influence of Conditioning Agents on the Determination ofMetallic Content of Sewage Sludge by Atomic-absorptionSpectrophotometry with Electrothermal AtomisationConditioning agents are often used to aid the de-watering of sewage sludgesprior to disposal to agricultural land. These might interfere in the electro-thermal atomic-absorption spectrophotometric analysis of heavy metals.Possible interferences by inorganic and polyelectrolyte conditioners werestudied a t the higher rates of addition normally used in sewage practice.Analyses for cadmium, chromium, copper, nickel, lead and zinc were per-formed by flame and electrothermal atomisation methods in conditionedand unconditioned samples.The organic polyelectrolytes tested did notinterfere, nor did most inorganic conditioners a t rates of addition consistentwith normal sewage treatment practice. However, interferences occurredwith aluminium chlorohydrate a t normal and very high addition rates, andwith other inorganic conditioners a t very high addition levels.Keywords: Electrothermal atomic-absorption spectvophotometry; cadmium,chromium, copper, nickel, lead and zinc determination ; sewage sludgeanalysis ; interferences ; inorganic and polyelectrolyte conditionersM.J. T. CARRONDO, R. PERRY and J. N. LESTERPublic Health Engineering Laboratory, Imperial College of Science and Technology,London, SW7 2AZ.Analyst, 1979, 104, 937-943viiOctober, 1979 SUMMARIES O F PAPERS I N THIS ISSUEDetermination of Antimony and Other Trace Elements in Irons andSteels by Atomic- absorption Spectrophotometry withIntroduction of Solid Samples into an Induction FurnaceAtomic-absorption spectrophotometry with an induction furnace has beenused for the determination of 0.5-350 pg g-l of antimony in 1-20 mg samplesof irons and steels dropped into the furnace. Calibration graphs of peakabsorbance vevsus the mass of antimony have been constructed by usingstandard steels.Information is presented on the accuracy and precision ofthe method for 13 irons and steels. The limit of detection for antimony isCalibration graphs have also been obtained for indium in a nickel-basealloy and thallium, tin, selenium, tellurium, zinc and cadmium in steels inorder to establish the conditions that will be necessary to determine theseelements using a procedure similar to that employed for antimony. Factorsaffecting the volatility of trace elements with boiling points below 2300 "Care discussed.0.12 l"g g-!.Keywovds ; Antimony determination ; tvace-element detevminations ; ivon andsteel analysis ; atomic-absovption spectvophotometvy ; induction fuvnaceA. M. AZIZ-ALRAHMAN and J. B. HEADRIDGEDepartment of Chemistry, The University, Sheffield, S3 7HFAnalyst, 1979, 104, 944-951.Quantitative Determination of Lead Dioxide Polymorphsby X-ray Powder DiffractometryThe X-ray diffractometric methods used by previous workers to analysepositive plate material from lead-acid batteries for a-, p- and amorphouslead dioxides are critically appraised a In addition two alternative approaches,not previously applied to this problem, are assessed. An internal standardmethod is considered to be the best,Keywovds : X-vay diffyaction ; quantitative phase au2alysis ; lead dioxide ; leadsulplzate ; lanthanum lzydvoxideP. R. SKIDMORE and R. R. SCHWARZBerec Group Limited, Group Technical Centre, St. Ann's Road, London, N15 3TJAnalyst, 1979, 104, 952-960

ISSN:0003-2654    

DOI:10.1039/AN97904FP085

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

X SUMMARIES O F PAPERS I N THIS ISSUE October, 1979Calcium Ion- selective Electrode Measurements in thePresence of Complexing LigandsCalcium ion-selective electrodes based on calcium bis[di (4-octylphenyl) -phosphate] sensor and dioctyl phenylphosphonate solvent mediator havebeen used for monitoring free calcium-ion levels (from below lo-' to above10-3 M) in the presence of citrate, malate, malonate, oxalate, EDTA, NTA,sulphate, orthophosphate, tripolyphosphate and pyrophosphate anion ligandsystems under conditions of constant ionic strength maintained by sodiumchloride. Log /3 data fall in the range of those previously measured for thevarious equilibria by alternative methods, thus demonstrating that calciumion-selective electrodes of the type used here can be employed for free calcium-ion measurements to below the detection limits characteristic of calibrationswith serial dilution standards and without disturbing the equilibria of com-plexat ion.Equilibria existing in the tripolyphosphate and pyrophosphate systemswere discerned by application of the MINIQUAD program for computing forma-tion constants and species distribution. Predictions concerning the existenceof [CaP30,J3- and [Ca(P,O,,),]*- in tripolyphosphate and of [CaP,0,I2- and[Ca( P,O,) J6-.in pyrophosphate systems are briefly discussed.These calcium ion-selective electrodes are not affected by added phosphateexcept insofar as free calcium-ion levels are lowered by complexation.Keywovds : Calcium ion-selective electvodes ; formation constants ; coiaplexationequilibriaA.CRAGGS, G. J. MOODY and J. D. R. THOMASChemistry Department, University of Wales Institute of Science and Technology,Cardiff, CF1 3NU.Analyst, 1979, 104, 961-972.Interferences of a Barium Ion- selective Electrode Used for thePotentiometric Titration of SulphateCertain cations and anions interfere with the potentiometric titration ofsulphate with barium chloride as titrant and a barium ion-selective indicatorelectrode. Thus, cations, for example potassium, which can be sensed by thebarium ion-selective electrode, give distorted titration curves, while pH valuesbelow 1.5 lead to electrode breakdown. Cations, such as calcium, whichinteract with sulphate give low sulphate recovery, while anions that inter-act with the barium titrant lead to apparently high sulphate.Treatment of samples with cation-exchange resins in the sodium form canremove cation interferences, but acidification to pH 2 with hydrochloric acidprevents interference from anions, such as phosphate, carbonate - hydrogencarbonate and organic anions.Keywords : Baviuwz ion-selective electvode ; sulphate titvation ; intevferencesDILYS L.JONES, G. J. MOODY and J. D. R. THOMASChemistry Department, University of Wales Institute of Science and Technology,Cardiff, CF1 3NU.and MAGDOLNA HANGOSInstitute for General and Analytical Chemistry, Technical University, 1502 BudapestXI, Hungary.Analyst, 1979, 104, 973-976October, 1979 THE ANALYST xib Collaborative studiesb Original papers onnew techniques,applications,0 authentic data ofcomposition studiesleading to methodsdevelopmentb Referee reportsPublished six times annually.Official publication of theAssociation of Official AnalyticalChemists-the internationallyknown organization devotedto developing and publishingreliable thoroughly testedmethods of analysis for:b foodsh drugsb cosmeticsb colorsb beveragesb flavorsb vitaminsb preservativesb fats and oilsh feedsb fertilizersb pesticidesb disinfectantsb hazardous substancesASSOCIATION OFOFFICIALANALKICALCHEMISTSJOURNALAssociation of Official Analytical ChemistsBox 540, Benjamin Franklin StationWashington, D. C.20044U.S.A.Please enter -subscription(s) to JOURNAL OF THE AOAC for197-. (Subscriptions are on calendar basis for 6 issues: Janyry, March,May, July, September, November.Back issues are sent to subscribers.)Price: $45.00 (domestic): $50.00 (foreign). $12.00 single copy.Name(please print)AddressZip Code - CityState or Country0 Payment enclosed 0 Send invoiceSignature DateIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIxii SUMMARIES OF PAPERS I N THIS ISSUE October, 1979Detection of L-Cysteine, Methionine, Thiourea, Allylthiourea anda-Naphthylthiourea in Sub-milligram Amounts Using AcidifiedPotassium Chromate SolutionShort PaperKeywords : Thio compound detection ; acidified potassium chromate CO~OZLVreagentM. NASIM BEG, FASIH A.SIDDIQUI, M. MOHTASHIM BEG,R. SHYAM and M. ARSHADDepartment of Chemistry, Aligarh Muslim University, Aligarh-202001, India.Anaylst, 1979, 104, 977-979.Implementation of a Sensitive Method for Determining Mercury inSurface Waters and Sediments by Cold-vapour Atomic-absorptionSpectrophotometryShort PaperKeywords Mercury deternzination ; atomic-absorption spectvophotometry ;cold-vapour method comparisons ; environmental samplesR. L. LUTZEState Pollution Control Commission, Central Square Building, 323 Castlereagh Strcet,Sydney 2001, N.S.W., Australia.Analyst, 1979, 104, 979-982.Spectrophotometric Determination of Trace Amounts of Boronin Solutions Containing Large Amounts of NitrateShort PaperKeywords : Boron determination ; nitrate interfevence ; hydvazine hydvate ;cavminic acid ; spectrophotoinetryH.J. ROSENFELD and A. R. SELMER-OLSENAgricultural University of Norway, Chemical Rescarcli Laboratory, N-1432 Ais-XLH,Norway.Analyst, 1979, 104, 983-986.Spectrophotometric Determination of Nitrite Using4,5- DihydroxycoumarinShort PaperKeywords : Nitrite determination ; spectvophotometvy ; 4,5-dilzydro~~3,cou.tnavinMOTOSHI NAKAMURA and AKIRA MURATAFaculty of Engineering, Shizuoka University, 3-5- 1, Johoku, Hamamatsu-shi,Shizuoka-ken, 432, Japan.Analyst, 1979, 104, 985-988October, 1979 THE A...NALYST XlllACS PublicationsModern Classicsin AnalyticalChemistry Vol. 2Compiled by Alvin L. BeilbyA selection from the best featurearticles that appeared in issues ofAnalytical Chemistry from 1 970 to1975.Ideal assupplementaryreadingfor theadvanced student of analyticalchemistry. Includes spectroscopy,electroc hemistry, chromatography,automation and instrumentation,measurement techniques, analyticalmethods, and art conservation.Paperbound 314pp I l ” x 8 ~ ”0 841 2 0332 6 f 6.50ReagentChemicals:5th EditionACS Specifications for 320 reagentchemicals; includes 50 pages ofdefinitions, tests, and reagentsolutions. Features flame and flame-less atomic absorption methods;new polarographic and chromato-graphic procedures; and newcolorimetric test for arsenic.Clothbound 685pp 93”x 6$’10 841 2 021 0 9 f 30.00both available from:The Chemical Society,D ist ri but io n Cent re,Blackhorse Road,Letc h wo rt h,Herts. SG6 1 HNEnglandI To: May & Baker Ltd (Wall-planner Offer) I Essex House, Station Road, Upminster, Essex RM14 2JTPlease sendI enclose€NameIIII Address II !wall-planners (1980/81) at f l eachpostal order/cheque payable to May& Baker Ltd II(please print clearly) LA16001L i r i ~ i m r ~ ~ r ~ r m r - civ THE ANALYST October, 1979CLASSIFIED ADVERTISEMENTSThe Rate for Classified Advertisements i s,9.j 0 per single column centimetre.Box Numbers are charged an extra 60p.Deadline for classified copy i s 20th of themonth preceding month of issue.All space orders, copy instructions andenquiries should be addressed toThe A dvertisement Department,The Chemical Society, Burlington House,Piccadilly, London WIV o B N .Telq5hone 01-734 9864 Telex 268001PATENTSBRITISH PATENT No.1 346 389.An Anti-Adhesive Composition.Owner desires commercial exploitationon reasonable terms by license or sale.Inquiries Fitzpatricks, Chartered Patent Agents, 14-18 CadoganStreet, Glasgow, G2 SQW and Warwick House, Warwick Court,London, WClR 5DJ.BUREAU OF ANALYSEDSAMPLES LTDNewham Hall, Newby,Middlesbroug h,Cleveland, TS8 9EA(Telephone : Middlesbrough 31 721 6)announce a new series ofLOW ALLOY CAST STEELSforSPECTROSCOPIC ANALYSISSS NOS. 611 /I -61 5/1Chief Analyst - HalifaxWe are a Group whose operations are world-wide.One of the reasons we have achieved a leadingposition in the confectionery market is that webelieve in the importance of the scientific effortinvolved in developing our products and in main-taining their quality. This vacancy reflects theemphasis placed on the work.We now seek a Chief Analyst to take charge, in thefirst instance, of the Analytical Section of thelaboratories at our Halifax factories where weproduce Quality Street, Toffo, Chewing Gum andother products.The work involves the analytical con-trol of a wide range of confectionery ingredients andproducts as well as analytical research and develop-ment. An interest in modern industrial techniques isessential.The post will be of particular interest to men orwomen who now wish to widen their experience as astep towards achieving further responsibilities.You should have a degree in food science, chemistryor physics, with a background of analytical orproducts research work and be between 25 and 35years of age.We provide a wide range of benefits- life assurance,profit sharing and removal expenses, and a widerange of social, sporting and educational facilities.Please apply quoting ref. LA.673 to:Miss D.M.N. Dick, Staff Office,Rowntree Mackintosh Ltd.,York YO1 IXY.Rswntree Mackintosh

ISSN:0003-2654    

DOI:10.1039/AN97904BP091

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

OCTOBER 1979 Vol. 104 No. 1243 The Analyst Kinetics of a- and P-Molybdosilicic Acid Formation Victor W. Truesdale Peter J. Smith and Christopher J. Smith Institute of Hydrology Maclean Building Crowmarsh Giflord Wallingford Oxfordshire 0x10 81113 The kinetics of formation of tl- and /3-molybdosilicic acid have been investi-gated and rate equations proposed together with possible reaction sequences involving individual molybdate species. The a-acid was found to form according to a sum of exponentials model [At = dl + d2exp( d32) + e,ex.p( e,t)] relating absorbance with time. In contrast the /3-acid formed according to a simpler exponential model [At = dl + d2exp(e,t)] which is consistent with pseudo-first-order behaviour with respect to silicate concentration. For the first time in a kinetic study of heteropolymolybdate formation the effects of changes in molybdate speciation upon the kinetics have been con-sidered in detail.The concentrations of individual molybdate species at pH between 0.5 and 5.0 and total molybdate concentrations up to 0.20 M as molybdenum have been calculated using equilibrium constants together with an existing computer program. The most interesting result of the study is that the observed a-acid kinetics demand the presence of a third silicon species in addition to the reactant silicate and the a-acid product. At this stage it is not known whether the third species should be another silicate condensation product or a molybdosilicic acid. The general application of the results of this study to silicate analysis is discussed.Methods of non-linear parameter estimation have been used extensively in fitting theoretical models to observed data. Keywords Molybdosilicic acids ; vnolybdate speciation ; kinetics ; modelling kinetics In this paper we describe the results of a study of the kinetics of formation of a- and /3-molybdosilicic acids. Although there were several reasons for conducting this study the most immediate need was to give a firm foundation to our earlier observations that at room temperature (20 “C) the P-acid forms between about pH 1.0 and 3.8 and the oc-acid between about pH 1.8 and 4.8.l These observations formed the basis for rejecting the earlier view that the conditions for molybdosilicic acid formation could be defined by the use of a variable known as “acid to molybdate ratio.”2 This study is part of a wider programme of research whose original objective was to try to verify the existence of a variety of silicate species in natural waters; this wider programme has already included an examination of the spectrophotometric characteristics of the molybdosilicic acids3 and a study of their transformation and decomposition.4 Detailed information about these systems is also needed to provide a firmer foundation to the widely used analytical methods for silicate that rely on the formation of the molybdosilicic acids.The molybdosilicic acids are also interesting because they are members of the large family of heteropolymolybdates ; conclusions reached about the behaviour of one member of the family can often have a bearing upon that of another.This study is novel in that it attempts to overcome the major problem that must be encounted with any study of heteropolymolybdate formation that of having an adequate description of the molybdate system under given conditions of pH total molybdate con-centration etc. Hence although it seems to be agreed that several molybdate species exist in solutions of pH between 0.0 and 5.0 little agreement exists as to which and how much of each of the individual species is present under given conditions.5 Indeed at one time or another molybdate systems have been defined using different combinations of molybdate condensation products containing 2 3 4 6 7 8 10 12 16 or 24 molybdenum atoms. Moreover in the aforementioned pH range the problem is aggravated by the fact that each condensation product can occur in several protonated forms.Presented with 89 898 TRUESDALE SMITH AND SMITH KINETICS OF Analyst Vol. 104 several alternative definitions of the molybdate system we found it impossible to choose the best one objectively. This was particularly so because the concentrations of the different species were not determined directly but were inferred from mathematical models that appeared to fit observed changes in relaxation or optical spectra etc. induced by changes in pH.5 These mathematical models contain different combinations of several of the ten molybdate condensation products (plus protonated forms) mentioned above. Accordingly, in this study where at best we can only hope to lay down a strategy for future studies we have found it worthwhile to consider only one of the molybdate speciation models.6 It is our intention to consider the alternatives later.The objective of this study then was to find the relationship between the rate of formation of the molybdosilicic acids and the concentrations of molybdate-molybdenum hydrogen ions and silicate-silicon. Earlier work suggested that there would be few problems of stoicheiometry. Moreover although separate pH regimens exist for the acids the /3-acid nevertheless transforms4 spontaneously into the oc-acid. Initially we chose to examine the rate of molybdosilicic acid formation using the approach in which the effects of changes in the concentration of a given variable are studied with the other variables being kept constant.Throughout the study all silicate solutions were undersaturated with respect to silica solu-bility and kinetic effects due to dissolution of solid phases' were avoided. Experimental Apparatus and Reagents Absorbance measurements were made on a Unicam SP 1800 spectrophotometer equipped with a constant-temperature housing for the cell. An EIL Model 23A direct-reading pH meter was used which was standardised at pH 4.0 by means of Soloid buffer tablets (Burroughs Wellcome & Co.) and at pH 1.0 by using a mixture of 25.0 ml of 0.20 M potassium chloride solution and 67.0 ml of 0.20 M hydrochloric acid.* Readings of pH relating to those solutions to which sodium chloride had not been added could be used directly to obtain the hydrogen-ion concentration.However when a background concentration of chloride (as sodium chloride) was maintained the observed pH had to be corrected for salt effect on the electrodes before it was converted into a hydrogen-ion concentration (in 1.0 M sodium chloride the observed pH values were 0.08 too low). AnalaR-grade sodium chloride, ammonium paramolybdate [ (NH,),Mo,02,.4H,0] and hydrochloric acid were used. A stock 1.000 g 1-1 silicate-silicon standard was prepared by fusing 2.1393 g of Specpure silica with approximately 10 g of AnalaR anhydrous sodium carbonate at 960 "C and dissolving the cooled pellet in distilled water to make 1.000 1 of solution. Method Mixtures (approximately 50.0 ml) of appropriate amounts of molybdate sodium chloride and acid in plastic bottles were brought to the required temperature (h0.05 "C).The mixture was inoculated with an appropriate amount of a silicate standard solution which was also at the same temperature and after mixing an aliquot was quickly transferredinto the spectrophotometer cell at the same temperature. For the more rapid reactions inocula-tion was performed directly into the cell. In these instances the contents of the cell were stirred continuously by mechanical means and afterwards the contents of the cell were established by weighing. In both instances the reaction mixture was retained until itspH had been measured. Molybdate Speciation As mentioned above a facility for quantifying changes within the molybdate system is a prerequisite of any satisfactory interpretation of the kinetics of molybdosilicic acid forma-tion.As the results obtained in this area are not only crucial to our work but also pervade it it is worthwhile to describe them now. The results were obtained by applying the computer program Haltafall,g which is designed to calculate the equilibrium concentrations of the species in mixtures of several components to existing equilibrium constants5t6 for the formation of the molybdate species under conditions of pH and total molybdate encountered in molybdosilicic acid formation (Table I). The prime consideration was to identify th October 1979 a- AND P-MOLYBDOSILICIC ACID FORMATION TABLE I VALVES OF THE EQUILIBRIUM CONSTAXTS ( @ q n ) FOR FORMATION O F THE MOLYBDATE SPECIES H,,(hfoO,),l'"' -'"- FROM p PROTONS AND 4 MOLYBDATE MONOMERS USED I N THIS STUDY 899 P 4 Loglo( isw) 1 1 3.53 2 1 7.26 8 7 52.80 9 7 57.46 10 7 60.84 12 8 71.56 form of the curves that relate the concentration of individual molybdate species to pH or total moljybdate concentration as this form can give a clue to the relationship between a given rate constant and the concentration of the relevant molybdate species.When discussing the effect that varying the total molybdate concentration has upon the concentra-tion of individual members of a family of species that are characterised by a given number of niolybdate atoms for example HM00,- and H,MoO, it is sufficient to consider 1 .o 0.8 0 I-4d 2 5 0.6 C s -a % 0.4 E .- - m 2 0.2 0 0.05 0.10 0.15 0.20 Total molybdate concentration/M as molybdenum Total molybdate concentration/M as molybdenum Total molybdate concentration/M as molybdenum Fig.1 . \'ariation i n the concentration of (a) Mo, ( b ) Mo; and (c) &lo8 species caused by change i l l total iiiolybdate concentration a t the three given pHs of ( A ) 1.0 (B) 5.0 and (C) 6.0 900 TRUESDALE SMITH AND SMITH KINETICS OF Analyst Vol. 104 only one of the family as they are linear multiples of one another. Moreover as ultimately it will be necessary to compare the behaviour of different families each of which can be present at markedly different concentrations it is convenient to normalise the concentration ranges . The graphs of normalised species concentration against total molybdate concentration (0.0-0.20 M molybdenum) for the families Mo, Mo and Mo are shown in Fig.1. This shows that the form of the variation of concentration accompanying change in total molybdate concentration is virtually the same at any pH between 0.6 and 5.0. The individual values for relative concentration at 0.050 M molybdate (Table 11) illustrate this well. Moreover these individual values together with Fig. 1 show that at pH between 5.0 and 6.0 the pattern of variation is very sensitive to changes of pH. The changes in the concentrations of Mo,O,,6- and Mo,02,4- species are very interesting (Fig. 1) because at pH below 5.0 for the Mo species and below 4.0 for the Mo species the graphs of relative species concentration vemm total molybdate concentr<ation are approximately linear. Above these pH limits as in the case already discussed the pattern of variation changes markedly with small changes in pH (Table 11 Fig.1 ) ; the linear approximation is lost. EFFECT OF pH ON THE SHAPES OF GRAPHS OF THE TYPE SHOWN IN FIG 1. Changes in shape are detected as a change in the concentration of the given molybdate species at 0.050 M total molybdate-molybdenum relative to that at 0.200 M total molybdate-molybdenum. PH 1.o-f 0.6 1.6 2.0 2.6 3.0 3.6 4.0 4.6 5.0t 5.6 6.0-f Relative concentration of individual molybdate species* Mo species M’o species Mo species 0.832 0.275 0.228 0.834 0.281 0.235 0.836 0.285 0.238 0.835 0.284 0.237 0.833 0.278 0.231 0.830 0.270 0.224 0.824 0.256 0.211 0.820 0.250 0.205 0.816 0.241 0.197 0.808 0.225 0.181 0.739 0.121 0.089 0.002 0.450 0.004 f A I * Concentration of species at 0.050 M total molybdate-molybdenum.Concentration of species at 0.200 M total molybdate-molybdenum. Full graph given in Fig. 1. ____ Fig. 2 shows how the partitioning of the molybdenum in solution at a given total molybdate concentration changes with pH. Thus at pH above 6.0 the predominant form is the unprotonated ion. As the pH decreases the Mo@246- HMo7?24‘- H2M0702,4-, Mo,O,,~- and H,MoO succeed the Moo4,- a s predominant forms. It is clear therefore, that any consideration of the effect of pH changes on the apparent formation rates of the molybdosilicic acids must take into account the substantial changes that pH induces in the molybdate speciation. Mathematical Modelling As in our earlier study of the transformation of the P-acid, it has been necessary to fit experimental data to mathematical models.In essence the numerical “optimisation” approach used here is similar to that explained earlier. However here it has been necessary to estimate the parameters of the sum of exponentials model A = 8 + 02exp(8,t) + B,exp(B,t). To overcome numerical problems inherent in the use of this model particularly that of preventing the Hessian matrix becoming ill-conditioned near the optimum solution, it has been necessary to reinforce our earlier numerical approach4 with an additional sub-routinelo that makes an approximation to the Hessian matrix that involves only first-orde October 1979 a- AND p-MOLYBDOSILICIC ACID FORMATION 901 1 .o 0 5 - m 0 ,,- 0.8 0 K 0 0 + w .- +.’ L Y- 0.6 u) m C .- CI 2 0.4 0, C 8 8 0.2 MOO 2-0 1 .o 2.0 3.0 4.0 5.0 6.0 PH Fig.2. Variation in the concentration of Pr400,~- HMo0,- H,MoO, M O ~ ~ , HMo,O,,~- H,Mo,O,,~- and Mo,O,,~- species with change in pH at a total molybdate concentration of 0.025 M as molybdenum. The curve for HMo04- is indistinguishable from the abscissa. derivatives (the Jacobian). The solving of the sum of exponentials models is acknowledged as being fundamentally difficult,ll and is encountered in many disciplines e.g. hydrology, nuclear physics and chemical kinetics. However our success results from the combined facts that the data are relatively free of “noise” and only two exponents are involved. Typical examples of fitting the sum of cxponentials model are given in Fig.3. As is demonstrated later in general the individual rate constants within any given over-all reaction scheme that gives the form characteristic of the sum of two exponentials model are related to the parameters Oi i = 1 . . . 5 of the model through non-linear functions. Moreover as there are fewer rate constants than parameters the set of equations are over-determined. Therefore to obtain estimates of the rate constants it has been necessary to perform a further “optimisation” stage. Results and Discussion During the formation of over 100 @-acid solutions at a wide variety of pHs and ionic strengths the increase in absorbance with time has been found to fit the simple exponential model A = + 6,exp(O3t) . . - - (1) with 8 = -O1 63 < 0 and where A is the absorbance at time t and el 6 and e3 are con-stants.As the molybdate-molybdenum concentration of each mixture was greatly in excess of that required for the formation of the ,&acid this behaviour is consistent with pseudo-first-order kinetics with respect to silicate-silicon that is, = K[Si] d W - MSAI - d[Si] --dt dt and therefore [/I - MSA] = [@ - MSA],(l - eckt) as C[Si] = [Si] + [/3 - MSA] where [p - MSA] and [Si] are the concentrations of the fi-acid and unreacted silicate-silicon a 902 TRUESDALE SMITH AND SMITH KINETICS OF Analyst Vol. 104 time t [/3 - MSA] is the final concentration of p-acid and k is an apparent rate constant. As in this instance absorbance is proportional to concentration, A = A (1 - e-kt) giving from equation (l) O1 = -02 = A and 8 = -K.Following accepted procedure,12 the sensitivity of the model parameters to changes in initial reactant concentration was tested to ensure that the adopted model was appropriate; where a model is inappropriate marked changes in the parameters generally occur in this test. The absence of marked changes in the parameters of equation (1) accompanying a five-fold change in initial concentration of silicate (Table 111) but under set conditions of pH and ionic strength supports the appropriateness of the simple exponential model for /3-acid formation. The values of the parameters obtained for the two initial silicate-silicon concentrations do differ slightly under some of the set conditions. However these differences are far too small to affect the appropriateness of the model; they can be attributed to experimental error.In fact these experiments are difficult to perform as they necessitate the use of extreme ranges of absorbance; also it is difficult to match the pHs of the final mixtures precisely because of the high carbonate content of the standard silicate solution. During the formation of a similar number of a-acid solutions the increase in absorbance with time was found to fit the sum of exponentials model A = + e2exp(8,t) + B,exp(e,t) . . TABLE I11 EFFECT OF CHANGING THE SILICATE-SILICON CONCENTRATION ON THE VALUE OF THE PARAMETERS ei (i = 1 . . . 5) I N EQUATIONS (1) AND (2) ( a ) Tests with solutions containing 0.050 M total molybdate-molybdenum and up to 0.1 M added sodium chloride at 22.0 "C-Conditions of test f A r -l PH Silicate-silicon 1.01 1.992 1.01 1.970 1.01 0.421 1.01 0.408 1.52 1.877 1.52 1.973 1.52 0.419 1.52 0.413 ( 0.03) concentration/mg 1-1 Product ,%Acid p-Acid I Parameters of equation (1) A 1 ell ( a b s o r b a n ~ e ) ~ ~ ~ 0 " ~ per mg l-l Si B,/min-I 0.430 - 1.58 0.430 - 1.59 0.444 - 1.60 0.446 - 1.65 0.430 - 2.50 0.428 - 2.48 0.439 - 2.58 0.438 - 2.63 0.358 -2.17 0.357 -2.12 0.372 - 2.30 0.372 - 2.27 0.278 - 1.50 0.277 - 1.48 u- + P-Acids U- + P-ACidS 0.298 - 1.62 0.299 - 1.55 0.267 - 0.65 0.267 - 0.65 0.264 - 0.60 0.265 -0.58 2.56 1.989 2.56 1.920 2.52 0.409 2.52 0.403 3.50 1.975 3.50 1.908 3.58 0.402 3.58 0.406 4.18 2.108 4.17 2.111 4.13 0.426 4.13 0.437 u- Acid October 1979 903 a- AND P-MOLYBDOSILICIC ACID FORMATION TABLE III(continued) ( b ) Tests with solutions containing 0.050 M total ~ o l ~ ~ b d a t e - ~ p z u l ~ ~ O d p n u m m i d 1 .0 bi added sodium cklorsde at 22 “C-Conditions of test --h----7 PH Silicate-silicon (k0.03) concentration/mg 1-1 1 .oo 1.883 1 .oo 1.958 1.00 0.393 1 .oo 0.413 1.55 1.862 1.55 1.965 1.51 0.383 1 i 1 1.51 0.380 J I 2.53 1.842 2.52 1.883 Product P-Acid P- Acid 1 a- + /I-Acids 2.49 0.386 2.49 0.385 3.52 1.894 3.52 1.868 3.48 0.381 ]i 3.49 0.398 a-Acid Parameters of equation (1) A 7 -7 el/ ( absorbance):!:l,:”’ per mg 1-1 Si 0.448 0.453 0.468 0.469 0.429 0.432 0.447 0.449 0.318 0.319 0.327 0.330 0.287 0.287 0.296 0.293 B,/mi 11 -- 1.70 - 1.65 - 1.70 - 1.67 - 2.50 - 2.48 - 2.62 -2.51 - 1.65 - 1.64 - 1.76 - 1.79 - 0.67 - 0.68 -0.78 -0.79 4.0 Fittings of experimental data to equation ( 1 ) inappropriate (G) l e s t s with solartions contaanang 0.200 M total n?ol~ibdatr-nzod,ibde~~i~na and 1 .O ni added s o d i i m chloride at 25.0 “C a% whach the a-aczd as fovrned accovdzng to the s w n of cxpoiirntzals n w d e -Conditions of tcst A I’arameters of cquation ( 2 ) - -7 -7 Silicate-silicon concentration/ - ___ PH mg 1-1 01 * e * 04* O,/min-’ O,/min-’ 3.41 3.227 0.269 -0.136 -0.138 -6.83 -1.09 3.39 1.308 0 266 -0.130 -0.139 -6.92 -1.17 * (Absorbance):9:l,P‘” per mg 1-1 Si. ~ i t h the parameters O, O, 8 and 8 all negative.[In extreme instances where 8 is much greater than 8, or vice versa and 6 and O4 are comparable a “rapid” initial increase in product is followed by a “slow” approach to equilibrium Fig. 3 (h).] The appropriateness of the model was confirmed in a similar test to that used with the ,&acid solutions where the initial silicate-silicon concentration was varied. This test showed (Table 111) that, within the hypothesis of marked changes the parameters 0 and 8 are independent of the initial silicate concentration but that the parameters 8, 0,. and 8 are linearly dependent upon it. Also the possibility that the goodness of the fitting of the model was due to a shortage of relevant molybdate species was rejected when it was observed that the addition of more silicate to a reaction mixture in a “slow” growth phase re-established a “rapid” growth phase.Experience has also shown that at low molybdate concentrations and low ionic strengths the sum of exponentials model for oc-acid foi-iiiation reduces to the single exponential model [equation (l)]. The approximation to pseudo-first-order kinetics at these low molybdate-molybclenuiii concentrations was shown to be satisfactory (Table 111) when the apparent rate constant under given conditions was found to be largely independent of the initial silicate-silicon concentration. More details of this will lie given below 904 TRVESDALE SMITH AX D SBIITH KINETICS OF . 4 t l 4 S t Yol. 701 0 1 2 3 4 5 6 0 12 24 36 48 60 Ti me/min Fig. 3. Typical examples of fitting experimental data points (x) taken directly froni the spcctro--1bsorbances a t 390 nm photomcter recortlcr chart to the sum of exponontials model jequation (2)].4-cm path length. Choice of Silicate Standard The results reported in this paper were obtained using working standard silicate solutions that contained 100.0 mg 1-1 or less of silicate-silicon. In preliminary work the silicon in stronger solutions was found to show different kinetic characteristics to that present in more dilute solutions. Thus in one experiment where the formation of the /3- ancl r.-acids was investigated at pH 1.03 ancl 4.09 at a molybdate-molybdenum concentration o f 0.050 M but without added sodium chloride pseudo-first-order rate constants for the formation of the acids were found to increase sympat1ietic;illy with a decrease in inoculum-silicate con-centration (Table 11').(By careful choice of pH the rate constants for both acids were TABLE IV EFFECT OF r m t i T I x G THE STOCK SILJCATI. IKOCULUM SOLUTIOK ox THE KIXETICS OF FORMATIOX OF Q- AND /~-ACII)S AT BOTH LOW IOXIC STRESGTH (0.050 11 J I O L I ~ ) E N ~ ~ J I ) ANT) AT PH 4.09 ,4m 2.03 RESPECTIVELY (0.0 BI EXCESS OF SODIUM CHLORIDE) AS11 LOW TOTAL MOLYBIIATE Tcnipcraturc not recorded but probably 17.0 "C. Concentration of silicate stock solution/ m g 1-' silicon 1000 500 400 200 100 50 20 First-ortlcr apparent rate constant requation (1) ; r-k,/niinrl 1.3s 1.53 1.55 1.57 1.61 1.67 1.64 1.36 1.49 1.58 1.55 1.63 1.64 1.67 made similar and indeed the two sets of results are almost identical.) At a given pH all of the reaction mixtures had tlie same composition; the volume of inoculum used was decreased as tlic silicate concentration was increased and compensating volumes of distilled water were added prior to inoculation.The results (Table IY) show that whereas there i October 1979 a- AND /~-MOLYBDOSILICIC ACID FORMATION 905 little or no change in the formation rate constant with up to 100.0 mg 1-1 of silicate-silicon, there is a marked change at higher concentrations. Despite its affecting the formation kinetics a change in the concentration of the inoculuni was founcl not to affect the yield of molybdosilicic acid. Thus in a separate experiment two sets of P-acid mixtures each of six replicates prepared in the aforementioned manner and using either 1000 or 100.0 mg 1-1 silicate-silicon standards yielded mean absorbances (390 nm 4 cm) of 0.800 and 0.801 after subtraction of a mean blank absorbance of 0.149 obtained in an identical fashion but without added silicate.As the standard deviations of these two sets of results is 0.005 in both instances the difference between the means is not significant. Although the changes referred to above in tlie rates of formation of the molybdosilicic acids are undoubtedly caused by silicate speciation it is not yet known how the change in speciation should be apportioned between the dilution and the concomitant change in pH of the silicate standard. Effect of Molybdate Concentration The effect that changing the total rnolybdate concentration has upon 13- and %-acid forma-tion was studied at pH 1.2 and 3.5 respectively in a background of 1.0 M sodium chloride.The sodium chloride was added to minimise the effect of ionic strength clianges that occur together with changes in total molybdate concentration. Under these conditions molvbdate salts crystallise within 1-2 h of mixing if the niolybctate concentration is in excess of approxi-mately 0.040 M as molybdenum; to circumvent this problem mixtures were prepared and tested within approximately 0.5 h. Blank determinations showed no change in absorbance within the test period. Although from earlier work1 it might appear that the pH of 3.5 is too low for a-acid production as will be explained later at these higher ionic strengths the pH regimen where only the a-acid f o r m is extended to this lower pH.,&Acid fownation At pH 1.2 the kinetics of @id formation are pseudo-first-order throughout the range 0.00-0.200 SI niolybdate-molybdenum. I'alues of tlie pseudo-first-order rate constants obtained at 25.0 "C are given in Table V. The data can be seen to fit the model well by the fact that the root-mean-square error is less than 1.99 x in all instances that is, less than 0.4o/b of the change in absorbance. For each instance tliis error measures the average deviation of all the data points from the fitted curve and is calculated by taking the square root of the quotient of the total sums of squared deviations and the number of data points. molybdate M O Q ~ ~ ~ - 3 1 0 ~ 0 ~ ~ ~ - or A 1 0 0 ~ ~ - concentratinns in any simple waj-.investigation showed however that the pseudo-first-order rate constant k of Preliminary investigations showed that the apparent rate constant is not related = k[Si] d[/3 - MSA] _ _ _ _ ~ ~ - ~ ~ dt where [#I - MSA; and [Si] are the concentrations of tively at time t is given by giving to total Further 13-acid and unreacted silicate respec- . . . . (3) 1 + n Of course an analogous equation in which SIo replaces Mo8 could have been developed. However because the Slo species is predominant in the pH range where tlw ,!?-acid forms (l5g. 3) it seems likely that the 110 species is the appropriate candidate 906 TRUESDALE SMITH AND SMITH KINETICS OF A~aaEyst Vol. 104 TABLE V APPARENT RATE CONSTANTS AND ROOT MEAN SQUARE ERRORS OBTAINEII BY FITTIKG EXPERIMENTAL DATA FOK P-ACID FORMATION TO THE EXPONENTIAL MODEL OF EQUATION (1) Total molybtlate concentration/ M molybdcnuin 0.200 0.175 0.150 0.125 0,100 0.090 0.080 0.070 0.060 0.050 0.040 0.030 0.020 0.010 Xpparcnt rate constant k b (- B,)/min-l 3.613 3.66 3.45 3.51 3.33 3.20 3.17 3.21 3.21 3.08 3.02 2.97 2.82 2.98 2.73 2.75 i"! 2.00 Root mean square error x lo4 9.73 19.2 13.5 11.9 6.94 6.83 13.1 11.5 19.9 11.6 18.8 18.9 0.81 14.2 8.8 12.9 0.6 12.0 1.0 14.7 14.0 In practice the curve of the graph of iMol]/k versus l/[Mog] was found to fit a linear model well (correlation coefficient 0.9870) with a gradient and intercept of 7.06 x 10-11 and 1.144 x respectively accompanied by standard errors of 0.25 x 10F1 and 0.013 x 10-8 respectively.These parameters yield values for a and h of 8.74 x lo7 1 g-ion-1 min-1 and 6.17 x lW3 g-ion 1-* respectively. As can be seen from Fig. 4 the experimental data lie close to the derived equation. Total moiybdate concentration/M as molybdenum Fig. 4. Comparison bctween observed (A) and predicted values [line derived from equation (3)] of the apparent rate constant for p-acid formation at pH 1.80 & 0.02 1.0 hi added sodium chloride, 25.0 "C and various total molybdate concentrations October 1979 u- AND /3-MOLYBDOSILICIC ACID FORMATION 907 This form of rate law [equation (3)] suggests that an equilibrium step occurs prior to the rate-determining step.A possible reaction path is k , k-1 Si(OH) + MOO^^- I + H 2 0 k 2 I + Mo8022- -+ products slow The rate-determining step is the reaction between the Mo species and the intermediate complex I which contains one silicon and one molybdenum atom. Assuming a steady-state condition for I this gives This equation is identical in form with the experimental rate equation and by comparison, values of 8.74 x 1071g-ion-lmin-l and 6.17 g-ionl-1 are obtained for k and the ratio k-,/k, respectively. As the concentration of the intermediate I could not be measured it was not possible to calculate the magnitude of the individual rate constants k- and k,. u-Acid formation At pH 3.5 the formation of a-acid followed the sum of exponentials model [equation (2)].Values of the parameters obtained at 25.0 "C at each of several molybdate-molybdenum concentrations are gven in Table VI. The data can be seen to fit the model well by the fact that the root mean square error is less than 1.83 x in all instances. Estimates of TABLE VI Total molybdate concentra-tion/M molyb-denum 0.20 0.150 0.100 0.090 0.080 0.070 0.060 0.050 0.040 0.030 0.020 0.010 0.005 PARAMETERS AND THEIR VARIANCE ESTIMATES OBTAINED FROM FITTING EXPERIMENTAL DATA RELATING TO DIFFERENT TOTAL MOLYBDATE CONCENTRATIONS TO THE SUM OF EXPONENTIALS MODEL Root mean 8 + square Var Var 8 O3 (.-kl)/ Var e3 Var O4 Ob(.-ks)/ Var Oh 8$ -k error 8 x lo6 8 x 109 min-1 x 106 O4 x 109 min-1 x 103 O4 x lo* 0.816 1.53 0.808 1.24 0.799 42.6 0.792 26.0 0.802 1.89 0.785 1.93 0.801 1.85 0.811 1.59 0.794 2.03 0.787 1.90 0.790 6.45 0.802 1.27 0.796 -0.789 -0.799 22.2 -0.388 - 0.366 - 0.371 - 0.348 -0.342 - 0.354 -0.331 - 0.303 -0.289 -0.272 -0.193 - 0.796 -0.789 -0.185 - 0.382 7.38 5.35 93.1 61.6 36.1 52.4 30.3 197 127 170 14 400 197 000 --2 20 - 6.89 - 7.11 - 5.54 - 3.36 - 3.22 - 2.84 - 2.51 - 1.95 -1.84 - 1.46 - 0.91 0 - 1.08 - 0.417 -0.217 - 0.034 8.50 6.62 8.57 3.48 4.18 1.92 3.66 2.30 1.82 43.7 15.9 10.3 --0.012 - 0.414 - 0.41 3 - 0.417 - 0.413 - 0.443 - 0.437 - 0.445 - 0.489 -0.516 - 0.613 - 0.481 - 0.500 - -- 0.572 6.84 4.55 44.3 60.7 30.9 48.1 25.7 189 111 152 14000 193000 - -336 - 0.887 - 0.898 - 0.780 - 0.623 -0.678 - 0.604 - 0.482 - 0.581 - 0.526 - 0.447 -0.471 - 0.495 - -- 0.144 0.138 0.014 7.73 0.109 0.013 7.19 1.8.5 0.016 18.3 0.249 0.008 14.5 0.145 0.011 6.62 0.07 0.002 7.62 0.193 -0.001 F.45 0.131 0.002 6.37 0.100 -0.002 6.09 1.93 0.002 7.88 0.27 -0.004 5.50 0.153 0.006 8.80 - - -- - -0.001 0.042 10.9 Time for 9876 formation/ min 3.7 3.6 4.2 5.2 4.9 5.5 6.9 5.8 6.5 11.0 r r I ./ 7 .4 9.4 51.0 18.0 the variance for each parameter are also given in Table VI. These values were obtained from the variance - covariance matrix which in turn is calculated from the inverse of the approximation to the Hessian matrix that involves first derivatives of the error function with respect to the parameters evaluated at their optimum values4 In all instances the variances are negligible when compared with the magnitude of the parameters.At 0.010 and 0.020 M molybdenum concentrations the sum of exponentials model collapsed to the simpler one, and estimates of the parameters 8 (-82) and e3 were obtained using the model represented by equation (1). At a 0 . 0 0 5 ~ molybdenum concentration however the data reverted to being better fitted to the sum of exponentials model [equation (a)]. Nevertheless the values of these parameters appear to be compatible with those obtained at the higher concentrations (Table VI). Within the limits of experimental and numerical method errors the paramete 908 TRUESDALE SMITH ASI SMITH KISETICS OF Aiialyst lJol.101 O1 was constant for all total molybdate concentrations and the sum of the parameters el 8, and 0 was equal to zero across the range of molybdenum concentrations studied; the para-meters 8 and 8 vary antipathetically. The results (Table VI) also show that notwith-standing the presence of “rapid” and “slow” growth phases tlie over-all rate of formation of tlie %-acid (to 98yA completeness) at a given pH increases with total molybdate concentra-t ion. Several cliemical reaction schemes that yield the sum of exponentials model can be written (Table VII) ~ ,411 involve combinations of first-order reactions that link tliree silicon species that are either synthesised after or introduced during inoculation of the reaction mixture with silicate-silicon ; this leads to the species Y having in the two instances an initial concentration (yo) o f zero or non-zero respectively.The equations that describe each system are gi\wi in full as in our experience several text-b~oks’~-l~ do not always express them correctly and their derivation is not simple. Further examples of schemes that give the sums of exponentials models can be obtained by making reversible any of the uni-directional reactions in Table VII. The identification of the actual scheme is probably best accoinplislied by combining chemical experimentation with nunierical modelling. Altlrougli a unique sclieine has not been identified using this approach it is worthwlde discussing further the applicability of schemes such as (i)-(v) (Table VII) as they assist in the appreciation of the chemistry involved.A s our results (Table VI) sliow that tlie parameters 8 and O4 of equation (2) are negative, reaction scheme (i) (Table VII) which requires 8 and 8 to have opposite signs (with yo and zo set to zero) can be rejected. Also all those scliemes involving the cx-acid in an cqui-librium with one or both reactants e - g . scheme (iv) in Table VII can also be rejected on stoicheiometric grounds as they are not consistent with the observation of full formation of %-acid across a wide pH range. This is supported by equation (9) where at equilibrium (t-+m) the concentration of x-acid z is only a fraction k 3 / ( k + k5) of tlie initial silicate concentration so; the remainder is present as the intermediate Y 1 equation @)I.The rate constants of schemes (ii) antl (iii) (Table 1’11) have been obtained (Table VIII) by using the values of the parameters (02-e5) found from the earlier fitting. This was not necessarjy with scheme (v) (Table VII) because there tlie reaction rate constants (k and K 3 ) can be considered to be the parameters 8 and O5 of the sum of exponentials model. As with the original fitting (Table i71) tlie magnitude of the root mean square errors given in Table l V I I I indicate that the fittings are satisfactory as they are less than 7.09 x 10-3, Similarly the magnitude of the variance associated with each estimate o f the rate constant is small when compared with the value of the respective rate constant itself.A surprising feature of the results in Table \;I11 is that the estimates of k, var k and root mean square error for schemes (ii) antl (iii) are tlie same for any given total molybdate concentration. Subsequent to observing this behaviour we have been able to show mathematically that this must be so at least for k,. Notwithstanding this we are pleased that the numerical approach adopted here lias demonstrated this fact as this has provided an unsolicited test of the methods used. The proof of this unexpectecl property of these reaction schemes will be given elsewliere. A clear understanding of the difference between schemes (ii) and (iii) (Table VII) can be obtaintd from Fig. 5 where for both cases tlie concentrations of reactant intermediate and product are plotted against time.The data for this exercise arise from a single kinetic run antl the result:; displayed in Fig. 5 are obtained by re-introducing the experimentally derived reaction rate constants into the relevant equations that clescribe the concentration of each species at an). time c.g. equations ( 5 ) (6) and (7) (Table WIT). The graphs show that although in both instances the concentration of the interniediate Y increases to a maximum after a short reaction period tlie maximum amount of interniediate formed differs in the two instances. After the maximum amount 11 as been attained this discrepancy is compen-sated for 1,)- an equal and opposite one in tlie concentration of unreacted silicate. It can be seen that in sclienic (iii) tlie unrcactcd silicate concentration is continuously maintained ltiglicr than that encountered in scheme (ii).To facilitate further interpretation of the results from Tables VI and VIII graphs of the rate constanti; w i w s total iiiolS.bdate-iiiolSrbdenuiii concentration have been plotted (Fig. 6). From these it can be seen that for all tliree models the rate constant k, appears to vary linearly with total niolybdatc concentration. It has already been shown (Fig. 1) that th TABLE VII EXAMPLES OF REACTION SCHEMES THAT FOR THE PRODUCT CONCENTRATION YIELD THE SUM Reactants or intermediates X and Y are converted into product 2 (molybdosilicic acid) in first-order steps ki (i = 1 . . . 5 ) . The concentrations of X Y and 2 a t any time t are denoted by x y and z respectively, yo and zo. in the present work demand that y o = zo = 0 in schemes (i)-(iv) and zo = 0 in Where appropriate terms hl and At which consist of combinations of k, k and k, have been Scheme Description Equations (i) Consecutive reactions __ x = x exp (- k,t) (v) Parallel reactions x = x o exp ( - k,t) [ k y = yo exp ( - k,t) z = x o + y o i- z - x o exp (- klt) - y o exp (- k,f) S k1 910 TRUESDALE SMITH AND SMITH KINETICS OF Analyst Vol.104 variation in the concentrations of both M o ~ O ~ ~ ~ - and Mo,O,,~- species is approximately linear with total molybdate concentration. Thus both observations are consistent with the rate constant K, being first order with respect to either Mo,O,,6- or Mo~O,,~- concentra-tion. Similarly in scheme (ii) (Table VII) the rate constant k, would also appear to be first order with respect to either of these species.The remaining rate constants are not related linearly to total molybdate (Fig. 6) and consequently their behaviour is more difficult to understand with the possibility of the otheir molybdate species being involved. TABLE VIII PARAMETERS (PSEUDO-FIRST-ORDER RATE CONSTANTS) AND VARIANCE ESTIMATES OBTAINED FROM FITTING THE PARAMETERS OF THE SUM OF EXPONENTIALS MODEL o, o, e AND e5 [EQUATION (2)1 TO TWO REACTION SCHEMES ( a ) The coinpetitzve - consecutive reactton scheme [the coej5caents and exponents of equation (6) scheme (ii), Table V 1 4 -Total molybdate concentration/ M molybdenum 0.200 0.150 0.100 0.090 0.080 0.070 0.060 0.050 0.040 0.030 0.005 k , 3.79 3.01 1.92 1.80 1.59 1.38 1.14 1.03 0.817 { ::% 0.116 0.090 Var 12 x lo3 138 186 8.82 6.47 0.371 0.219 0.074 4 0.905 0.593 1.95 1.76 7.56 15.2 k2 3.09 2.52 1.44 1.42 1.25 1.13 0.812 0.813 0.642 0.288 0.448 0.028 0.037 Var k x lo3 147 209 9.7 18.6 7.96 0.453 0.296 1.35 0.927 3.56 0.393 2.80 10.7 k , 0.887 0.780 0.623 0.678 0.604 0.482 0.581 0.526 0.447 0.470 0.495 0.034 0.012 Var k x lo3 265 362 16.7 28.6 12.3 0.713 0.418 1.68 1.08 3.06 2.83 0.132 13.9 Root mean square error 6.07 7.09 3.81 4.99 3.27 0.787 0.602 0.348 1.21 0.969 1.63 0.188 4.07 x 103 (b) The coiisecutivc reaction with reversible steps scheme [the coe8cients and exponents of equation (7) scheme ( i i i ) Table V11]-Total molybdate concentration/ M molybdennm 0.200 0.150 0.100 0.090 0.080 0.070 0.060 0.050 0.040 0.030 0.005 kl 3.79 3.01 1.92 1.80 1.59 1.38 1.14 1.03 0.817 { :::XE 0.116 0.090 Var k x loa 138 186 8.82 6.47 0.371 0.219 0.0744 0.905 0.593 1.95 1.76 7.56 15.2 k2 2.37 1.87 0.973 0.886 0.776 0.733 0.398 0.398 0.291 0.070 0.099 0.020 0.032 Var k x lo3 394 543 19.8 33.7 14.4 0.915 0.340 0.128 1.41 0.253 0.907 0.630 7.34 k4 1.61 1.44 1.09 1.21 1.08 0.877 0.995 0.941 0.798 0.688 0.844 0.043 0.016 Root mean square error Var k x lo3 x lo3 678 6.07 913 7.09 35.4 3.81 57.7 4.99 24.6 3.27 1.54 0.787 0.609 0.602 0.214 0.348 2.35 1.21 0.743 0.969 2.13 1.63 3.73 0.186 0.24 4.07 An important constraint upon the three sc‘hemes is that they should accommodate the observed collapse of the sum of exponentials niodel [equation (2)] to that of the simple one [equation (l)] at low total molybdate concentrations.This can be seen by re-writing the sum of exponentials model as where el& = 8 and = 04 and then comparing these coefficients and exponents with those of equations (6) (7) and (10) (Table VII) after yo and zo are set to zero where appro-priate and the equation is yre-multiplied by eZ the absorptivity of the cc-acid. Then for the individual schemes we have the following October 1979 a- AND P-MOLYBDOSILICIC ACID FORMATIOX 91 1 (a ) x + y + z 3 4 5 6 0 1 2 3 4 5 6 1 2 Time/min Fig.5. Variation with time of the concentrations of unreacted silicate intermediate and a-molvbtlosilicic acid when the a-acid forms by (a) the competitive - consecutive reaction jschcme ( i i ) Table \'LIj and by (b) consecutive reactions with a reversible step [scheme (iii) 'Table 1'11 j The coiiimon experimental data upon which these graphs are based arise from that shown in Fig. 3 ( a ) . (a) Schenze (ii). At low total molybdate concentration (Fig. 6) both k and k tend to k,. Therefore by comparing coefficients and letting k = k = k = k we obtain 1 - _ + == k 2 -k - k - k , Therefore A = eZxO(l - e-Xt) which is analogous to equation (1). k, k tends to zero.k = 0 we obtain ( b ) Sclzeme (iii). At low total molybdate concentration (Fig. 6 ) whereas k tends to Therefore by comparing coefficients and letting k = k = k and A = lim ( k + 6) = k and A = lim ( k - 6) = k 6+0 6+0 Therefore A = E ~ X ~ ( I - eckt) which is analagous to equation (1) 912 TRUESDALE SMITH AXD SMITH KINETICS OF Analyst Vol. 104 (c) Scherne (21). At low total molybdate concentration (Fig. 6) k tends to k,. Therefore, hy comparing coefficients and letting k = k == k we obtain 462 = - xo 0 = - k,= - k $4 = -Yo 0 5 - - - k 3 = - k Therefore A = eZ(xO -+- yo) (1 - e-kt) which is analogous to equation (1). 4.0 7 3.0 E E .-I w C C 2.0 2 2 1.0 a) c 0 0.05 0.10 0.15 0.20 Total molybdate concentration/M as molybdenum 0 0.05 0.10 0.15 0.20 Total molybdate concentration/M as molybdenum 8.0 C t 7 6.0 E 5 4.0 I-1 *.’ s 2 2.0 (u + 0 0.05 0.10 0.15 0.20 Total molybdate concentration/M as molybdenum \-dues of the individual rate constants obtained by fitting the parameters of the sum of exponentials model (Table VI) to three reaction schemes (Table 1’11) (a) competitive - consecutive reactions, schcme (ii) ; ( b ) consecutive reactions with a reversible step scheme (iii) ; and (c) parallel reactions scheme (v).In each graph a given symbol represents one curve; i t has no wider significance I;ig. 6. Xote that each curve is fitted “by eye” to its points. I t is not yet known why the fitting reverts to the sum of exponentials model at 0.005 M niolybdate-molybclenum.This observation is inconsistent with the analysis of the three models given above as they will remain as the simple exponential model. However it is suspected that the reversion is caused by there being insufficient total molybdate within the system a constraint that must be encountered if the total molybdate concentration is progressiveljr reduced. In this instance it is inevitable that the sum of exponentials model would offer the better fitting as it includes two extra parameters although it would not be the correct one. We believe therefore that the observation does not detract from the aforementioned analysis October 1979 a- AND P-MOLYBDOSILICIC ACID FORMATION 913 The possibility that the intermediate Y in scheme (ii) (Table VII) is P-molybdosilicic acid has also been considered.[A similar condition for scheme (iii) was rejected because of the unlikelihood of a reversible reaction between reactants and the @-acid.] In this instance the observed absorbance A t is composed of contributions from both acids Y and 2 which have absorptivities ey and eZ respectively Thus, A = Eyy + EzZ . . which from Table VII leads to Equation (12) is equivalent to the sum of exponentials model. However the inclusion of the P-acid in scheme (ii) (Table VII) yielded negative values for k at all but two of the thirteen total molybdate concentrations studied and consequently has been eliminated. As yet further elimination of the possible reaction schemes has not proved possible. Thus, schemes (ii) and (iii) (Table VII) cannot be ranked by any statistical method as the root mean square errors for each fitting within the sequence of total molybdate concentrations (Table VIII) are identical and moreover they both have the same number of parameters.Further a statistical comparison of these with scheme (v) (Tables VI and VII) is impracticable as the numerical method has not been taken to an equivalent state in the two instances; for scheme (v) it was possible to use the exponents of the more complex model directly (Table VI) . Effect of Hydrogen-ion Concentration on Formation of the Acids The greatest problem likely to be encountered in any study of the effect of variation in hydrogen-ion concentration on molybdosilicic acid formation is that of understanding what happens in the pH region where the acids form together.Outside this region the problems are only those of measuring and interpreting changes in the magnitude of pseudo-rate constants for the formation of a single product. Within the region of co-formation it is necessary to resolve the over-all change in reaction variables e.g. absorbance into a part due to the a-acid and a part due to the P-acid. Our experience suggests that this will be very difficult and perhaps impossible if the over-all change in absorbance during formation is determined by the combined effects of the a-acid formation following the sum of expo-nentials model and the P-acid formation following the single exponential model. Fortu-nately however some idea of the extent to which both a- and P-acid systems react to hydrogen-ion concentration can be gauged from studies at low total molybdate concentrations and low ionic strengths.There the sum of exponentials model for the a-acid collapses to the simpler one [equation ( l ) ] which is more amenable to treatment and moreover the over-all change in absorbance with time in the region of co-formation also follows the single exponential model [equation (l)] . The variation with pH of the apparent rate constant of formation was studied between pH 0.9 and 4.2. Various amounts of sodium chloride were added to each solution to main-tain a constant chloride concentration in all samples. Thus whereas samples with a low pH gained chloride from hydrochloric acid used in reaction with molybdate samples with higher pH required the addition of sodium chloride.In this way the ionic strength contri-buted by chloride salts to all of the solutions was kept within 1.00 0.01. The results show that the over-all apparent rate constant attains a maximum value at approximately pH 1.8 (Fig. 7). The variation in the final absorbance of each mixture as predicted by the model [S, equation (l)] is also shown in Fig. 7. This graph of absorbance against pH is similar to that which was originally usedl to de-limit the pH regimens of a- and P-acid formation but which was obtained by direct observation of solutions that had been allowed t o react for up to 30 min 914 TRUESDALE SMITH AND SMITH KINETICS OF Analyst Vol. 104 Further treatment of the data is dependent on the assumption of mechanisms for a- and /3-acid formation.Initially one likely possibility involved the competitive formation of the acids by independent first-order reactions : a:-acid 7 \ k b si\ /?-acid Also it had already been suggested2 that the formation of the P-acid could precede that of the a-acid in consecutive first-order reactions. Scheme (i) (Table VII) describes this mechanism when the intermediate Y is the P-acid and the product 2 the a-acid. Although both schemes can accommodate the observation that the "pure" solutions of a- and P-acids are formed according to the single exponential model the consecutive scheme is rejected as it does not represent the co-formation adequately. Moreover our earlier estimates4 of the rate constants of the second step of the consecutive reaction scheme (that is the trans-formation of the P-acid) are too low to support the observed over-all rates of a-acid formation.2.0 1.6 rl I C .-E \ tu 44 C 1.2 8 + 0.8 a w F 2! a C 8 0.L 0 - 0.8 - 0.6 -d E s- 0 0, W C - 0.4 5 2 2 n - 0.2 4 - 0.0 1 .o 2.0 3.0 4.0 PH Fig. 7. Variation with pH of the over-all apparent rate constant and the individual apparent rate constants for formation of a- and fi-acids a t low ionic strength and low total molybdate concentrations where the single exponential model [equation (l)] applies. The variation with pH of the final absorbance predicted by the model (6,) is also shown in the upper graph. The results were obtained a t 17 "C with 1 . 0 M sodium chloride and 0.025 M molybdate as molyb-denum present in the reaction mixture.In the lower graphs A represents the over-all (observed) apparent rate constant B that for the /I-acid and C that for the a-acid. Thus at 17.0 "C typical values for the transformation rate constant4 are only 0.05 times the over-all observed rate constant - O3 [equation (l)] for a-acid formation. Accordingly, the over-all apparent formation rate constant k(-0,) [equation (l)] has been divided int October 1979 a- AND /3-MOLYBDOSILICIC ACID FORMATION 915 separate rate constants k and kp appropriate to formation of a- and P-acids respectively. Under these circ~mstancesl~ the sum of the individual rate constants is equal to the over-all rate constant the ratio of k and kp is equal to the ratio of the final concentrations ceca and Pm respectively and the sum of concentrations of a- and P-acids at equilibrium is equal to the total initial silicate concentration so.Thus, k = k + kp . . . . (13) so = aca + p w - - . . . . . . (15) Substituting k from equation (14) into equation (13) and rearranging gives k p = ( a00 Bw + p m ) k = ( y ) k . . . . Also substituting am from equation (15) into equation (ll) which sums the absorbances contributed by the two acids with absorptivities E and E D gives or Thus from equations (16) and (17) we obtain Similarly, . . x k k . . . . (17) . . (18) . . . . (19) Through the use of equations (18) and (19) it has been found that the graphs of the apparent rate constant vemws pH for both a- and P-acid formation have an inverted U-shape.Although not drawn in Fig. 7 because of paucity of appropriate data these curves can be expected to approach the pH axis (zero rate constant) asymptotically. The apparent rate constants for a- and P-acid formation attain maximum values at pH 3.2 and 1.7 respectively. The changes in each of the two apparent rate constants caused by change in pH (Fig. 7) are much more difficult to interpret than the corresponding changes that resulted from variation in total molybdate concentration. Of course both interpretations are crucial to a complete understanding of the system. The difficulty that precludes full interpretation of Fig. 7 arises because a change in pH means not only a change in hydrogen-ion concentra-tion but also a concomittant change in the concentration of each of the several molybdate species present.Nevertheless it is possible to gain a semi-quantitative appreciation of Fig. 7 that substantiates the earlier interpretation of the effects of varying total molybdate at constant pH as well as indicating where future work might be most usefully directed. Thus there is a noticeable correlation between the shapes and positions of the graphs relating the apparent rate constants for formation of a- and P-acids to pH (Fig. 7) and the shapes and positions of the graphs relating the concentrations of H,Mo,O,,*- and Mo802,’-ions to pH (Fig. 2). These correlations are consistent with the mechanisms that were obtained from the total molybdate studies described earlier which include the ion H,Mo702:-as an important precursor of the a-acid and the ion Mo,O,,~- as a precursor of the P-acid.Moreover it would have been surprising if the variations in the apparent rate constants fo 91 6 Analyst Vol. 104 the formation of both a- and P-acids had been totally explained by changes in concentration of just these ions. Therefore any discrepancy between the observed variation in apparent rate constant and that predicted by the mechanisms can be attributed to the omission of other ions or species. With cc-acid formation which is now believed to be first order with respect to the concentration of H2M~70244- the inclusion of these other species would have to account for the small departure from linearity of the graph of apparent rate constant against H2Mo,02,4- concentration (Fig. 8). One plausible explanation for this is that a term for hydrogen-ion concentration should appear in the denominator of the rate equation, perhaps as a result of hydrogen ions being a product of one or more elementary reaction steps.With P-acid formation this approach is unsuitable because the relationship between the apparent formation rate constant and the concentrations of the identified Mo and Mo, precursors is not simple [equation (3)] and the appropriate graph cannot be drawn. As the description of the molybdate system used here is tenuous it would seem to be provident to reconsider the alternative descriptions of the molybdate system before attempting to force analvsis of Fig. 7 further. TRUESDALE SMITH AND SMITH KINETICS OF I 3.15 0 0.2 0.4 0.6 Concentration of H2 Mo 0244- ion Fig.8. Graph of the apparent rate constant for u-acid forination against the concentration of H,Mo,~,,~- ion expressed as a fraction of the total molybdate con-centration of 0.025 M as molybdenum. The solid line would be given by a perfect correla-tion. The dotted line indicates the sequence of the experimental points. General Discussion The results presented here have reinforced the view that the pH boundaries for a- and 13-molybdosilicic acid formation are kinetically determined and they depend on the time allowed for reaction. A practical implication of this is that observed boundaries for full formation can be in error if insufficient time is allowed; in our earlier work,l for example, we limited reaction times to 30 min. A further but less obvious implication is that owing to the instability of the P - a ~ i d ~ it is worthwhile to consider the state of the molybdosilicic acid system in both the short term (reaction times used in analysis) and the long term (infinite reaction time).In the short term the over-all pH boundaries for the formation of molybdosilicic acid arise from competition between formation and decomposition reactions. At low ionic strengths and room temperature these boundaries are approximately pH 1.0 and 5.0. Increases in ionic strength decrease the competitiveness of the formation reaction at the upper boundary October 1979 a- AND p-MOLYBDOSILICIC ACID FORMATION 91 7 thereby moving it to approximately pH 4.0 (not 3.5 as printed in an earlier paper4) at an ionic strength of approximately 1 .O.Within the over-all pH domain for molybdosilicic acid formation competition between separate formation reactions produces distinct pH regimes for a- and P-acid formation as well as a third where mixtures of the acids form. At low ionic strengths these pH regimes for exclusive a- and ,&acid formation are approximately 1.0-1.8 and 3.8-5.0 respective1y.l Increases in ionic strength increasa the competitiveness of the a-acid formation reaction so that at an ionic strength of approximately 1.0 only a-acid formation occurs at a pH as low as approximately 3.5. While this increas? in competitive-ness could be due to an increase in a-acid formation rate a decrease in /I-acid formation rate an increase in the rate of transformation of the P-acid4 or a combination of all of these changes the precise contribution that each makes to the over-all change is still unknown.In the long term at room temperature the pH boundaries for a-acid formation are as wide as the over-all pH boundaries applying in the short term. Thus a greatly increased reaction time allows all of the /I-acid to transform to the a-acid. Nevertheless there must be doubt about the composition of the product occurring at the lowx pH boundary where, under these circumstances decomposition also affects the a-acid. Under these conditions there remains the possibility of a complicated steady-state system existing which includes some /?-acid formed from the products of the decomposed oc-acid. As temperature also affects the rate of each reaction operating within the molybdosilicic acid system it also can affect the position of the pH boundaries for formation.Indeed the acceleration of the /3-acid transformation reaction is the most significant change brought by temperature increase. Thus the use of temperatures of approximately 100 "C during formation of molybdosilicic acid results in the oc-acid being the sole product at pHs where in the short term and at room temperature both of the acids would have been formed.16 As explained it has not been possible to identify a unique mechanism that accounts for the observed kinetics of a-acid formation. This must await further chemical information about the system because the numerical approach cannot offer further assistance. The nature of the species Y (Table VII) is of particular interest as it indicates the presznce of either an undiscovered silicate condensation product or an undiscovered molybdosilicic acid.Neither of these suggestions is inconsistent with any of the information obtained in this study. Thus whereas the observations concerning silicate speciation in strong stock silicate solutions offers some support to the former suggestion the fact that the individual rate constants obtained from fitting the various reaction schemes for a-acid formation to experi-mental data (Table VII) are dependent on the molybdate concentration supports the latter. One way of resolving these two possibilities is to examine the product(s) of the formation reaction as the reaction proceeds. Indeed exploratory experiments in which the absorption spectrum of the tin(I1) chloride reduction product1 was examined have shown that the a-acid is not the only product.In fact the observed progressive changes in the shape of the absorption spectrum were consistent with the presence of a small proportion of P-acid; however the amount was too low to conflict with the numerical approach's strong rejection of the P-acid being the intermediate Y (Table VII). Equally however these changes in the spectrum would be consistent with the presence of much larger amounts of another molybdosilicic acid that has a lower absorptivity. Notwithstanding this as these tests were conducted at pH 3.5 (in the presence of 0.10 M molybdate-molybdenum and 1.0 M sodium chloride) the effect could be due merely to entering inadvertently the /I-acid regime, proper.Our continuing work is directed towards solving this intriguing problem. The accuracy of our model for the kinetics of formation of the molybdosilicic acids can be questioned. However there seems to be no reason to question the logic of the over-all approach we have used in which the acid - total molybdate system is modelled separately and prior to the molybdosilicic acid system particularly as the acid - total molybdate system is known to equilibrate within a fraction of a second.6 This over-all approach is much superior to the earlier in which the total molybdate concentration was used as a variable in the rate equations and in which the apparent rate constant for formation was assumed to be dependent on a separable function of the hydrogen-ion concentration; that is a change in pH did not imply a concomitant change in molybdate speciation.We believe that the approach we have used would also be of value in studies of both the kinetics and stoicheio-metry of formation of other heteropolymolybdates where undoubtedly molybdate speciation should be taken into account.1 918 TRUESDALE SMITH AND SMITH Conclusions A proper understanding of the kinetics of molybdosilicic acid formation is entirely dependent upon having an accurate model for molybdate speciation; it is not justifiable to use total molybdate concentration as a variable in the rate equations. When examining the effect of changes in hydrogen-ion concentration on the kinetics of molybdosilicic acid formation it is essential to irecognise that concomitant changes occur in the concentration of the various molybdate speizies; indeed it is likely that the major part of any change resulting from pH variation is due to molybdate speciation.Silicate speciation in stock standard solutions affects the kinetics of formation of both molybdosilicic acids. During formation of the P-acid the absorbance increases according to a simple exponential model [equation (l)] ; this implies a reaction of which the velocity is first order with respect to unreacted silicate. During formation of the a-acid the absorbance increases according to a sums of exponen-tials model [equation (S)]; this implies a reaction scheme with first-order steps involving at least two other silicon species as well as the a-acid. The mechanism for p-acid formation is consistent with a rate-determining step in which an Mo species reacts with an intermediate the product of a pre-equilibrium involving monomeric silicate and molybdate species.The mechanism for formation of the a-acid directly from monomeric silicate species is consistent with a rate-determining step in which an Mo species reacts with a silicate monomer. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Truesdale V. W. and Smith C. J. Analyst 1975 100 203. Strickland J . D. H. J . Am. Chem. SOG. 1952 74 862 868 and 872. Truesdale V. W. and Smith C. J. Analyst 1975 100 797. Truesdale V. W. Smith C. J. and Smith P. J . Analyst 1977 102 73. Sillen L. G. and Martell A. E. “Stability Constants of Metal-ion Complexes,” Supplement No. I, Aveston J. Anacker E. W. and Johnson J . S. Inorg. Chem. 1964 3 735. Krauskopk K. B. Geochim. Cosmochim. A d a 1956 10 1. Weast R. C. Selby S. M. and Hodgman C. D. Editors “Handbook of Chemistry and Physics,” Ingri N. Kakolowicz W. Sillen L. G. and Warnquist B. Talanta 1967 14 1261. Gill I?. E. and Murray W. in “The National Physical Laboratory Optimisation Software Library,” Lanczos C. “Applied Analysis,” Prentice-Hall Englewood Cliffs N. J. 1956. Bunnett J . F. in Weissberger A. Editor “Techniques of Chemistry Volume VI Part I Investi-Wiberg I<. B. in Weissberger A. Editor “Techniques of Chemistry Volume VI Part I Investi-Frost A. A. and Pearson R. G. in “Kinetics and Mechanism,” John Wiley New York 1961, Szabo Z. G. in Bamford C. H. and Tipper C F. H. Editors “Chemical Kinetics,” Volume 2, Andersson L. H. Acta Chem. Scand. 1958 12 495. Hargis L. G. Analyt. Chem. 1970 42 1494. Backwith P. M. Scheeline A. and Crouch S . R. Analyt. Chem. 1975 47 1930. Special Publication No. 25 The Chemical Society London 1971. Forty-sixth Edition Chemical Rubber Co. Cleveland Ohio 1965-66 p. D73. National Physical Laboratory Teddington 1977. gation of Rates and Mechanisms of Reactions,” John Wiley New York 1974 pp. 129-209. gation of Rates and Mechanisms of Reactions,” John Wiley New York 1974 pp. 771-776. p. 176. Elsevier Amsterdam 1969 p. 30. Received January 23rd 1979 Accepted A p r i l 24th 197

ISSN:0003-2654    

DOI:10.1039/AN9790400897

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Analyst, October, 1979, Vol. 104, PP. 919-927 919 Spectrop hotomet ric Determination of Acetaminophen and Salicylamide Through Nitrosation and Subsequent Chelation Saied F. Belal, M. Abdel-Hady Elsayed, A. Elwalily and H. Abdine Deflartment of Pharmaceutical Analytical Chemistry, Faculty of Phavmacy, Uvzivevsity o f A lexandvia, Alexandria, Egypt A nitrosation reaction has been adopted for the spectrophotometric deter- mination of acetaminophen and salicylamide. The selectivity of the reaction is increased through utilisation of the nitroso derivatives as chelating agents for cobalt(I1) and copper(I1) ions. The optimum experimental conditions for the application of nitrosation and nitrosation with subsequent chelation were established. The proportions of reactants in each method and the instability constants for the products were determined.The nitroso deriva- tives and their chelates obey Beer’s law and their absorbances were used for the determination of acetaminophen and salicylamide in pharmaceutical formulations. The proposed methods gave more accurate results than the official methods. Keywords : Spectrofilzotovtzetvy ; acetaminophen determination ; salicylavtzide determination ; nitroso derivatives ; chelates The official compendia describe an ultraviolet spectrophotometric assay for the determination of acetaminophen1 (N-acetyl-9-aminophenol) and non-aqueous titration for the determina- tion of salicylamide.2 The former is highly sensitive to diverse co-existing substances that are often present in pharmaceutical formulations, while the latter is not specific.Various colour reactions have been utilised for the determination of acetaminophen through its hydrolysis and subsequent reaction with alkaline 2-naphth01,~ diazotisation and ~oupling,~ or condensation with ani ill in,^ anisaldehyde,6 p-dimethylamin~benzaldehyde~ or $-dimethylaminocinnamaldehyde.8 Other spectrophotometric methods were based on the reaction of acetaminophen with phenol together with an oxidising ion such as hypobr~mite,~ hypochlorite in the presence of meta-arsenitelO or hexacyanoferrate(1II)ll; nitration of acetaminophen has also been utilised for its spectrophotometric determination.l21l3 How- ever, these methods lack the simplicity needed for routine analysis. The fluorimetric methods were developed either by oxidation of acetaminophen with alkaline he~acyanoferrate(III)~*J~ or hydrolysis followed by reaction with benzylamine.16 These methods are subject to interferences, the materials present in syrup formulations causing the most interference. Acetaminophen has also been determined by cerimetric titration,17,18 gravimetry,lg titra- tion of the hydrolytic product with nitrite20 or polarography.21 These methods are not sufficiently sensitive.The different spectrophotometric methods proposed for the determination of salicylamide were based on its reaction with iron(II1) potassium hexacyanoferrate(II1) ,24 potassium hexacyanoferrate(II1) plus dimethyl-fl-phenylenediamine,25,26 and 3,5-dichloro-P- benzoquinone ~ h l o r i m i n e . ~ ~ ~ ~ 8 Non-aqueous t i t r a t i ~ n , ~ , ~ ~ iodimetric titration30 and complexometric titration31 have been employed for the assay of salicylamide.Inamdar and Kadji32 utilised the nitrosation of acetminophen for its determination in dosage forms. They reported that the stability of the chromogen produced in an acidic medium was increased by the addition of alcohol. claimed that a nitro rather than nitroso derivative was produced on treatment of acetaminophen with nitrous acid. In this work, the chromogen was adapted to the determination of acetaminophen in some commercial preparations purchased locally. More important, the nitroso derivative was used to form chelates with cobalt(I1) and copper(I1) ions. The formation of metal chelates Other920 BELAL et al. : SPECTROPHOTOMETRIC DETERMINATION OF Analyst, VoZ. 104 of the reaction product indicates that the formation of a nitroso derivative of acetaminophen is more probable than the nitro derivative.Furthermore, application of this type of reaction was extended to the determination of salicylamide. Experimental Apparatus and Reagents All reagents were of analytical-reagent grade. Acetaminophen (BP 1973). Chemical Industries, Egypt. Salicylamide. Spectrophotometer. p H meter. Pye, Model 79. Supplied by the Alexandria Company for Pharmaceutical and Supplied by CID Company, Cairo, Egypt. A Prolabo photoelectric spectrophotometer with 1-cm silica or glass cells. General Procedure for Nitrosation (Colour Development and Preparation of Standard Calibration Graphs) A cetaminophen Transfer 5.0 ml of the assay solution (containing 0.5 mg of acetaminophen) into a 25-ml calibrated flask, add 2 ml of 3% sodium nitrite solution and 1 ml of 1 N hydrochloric acid.Mix and leave to stand for 5 min. Render alkaline with 3 ml of 1 N sodium hydroxide solution. Dilute to volume with water and measure the absorbance at 430nm using a blank prepared in the same way, but omitting the drug substance. Calculate the concentration of acetaminophen using the linear equation in Table I, which describes the calibration graph prepared by applying the same procedure to solutions of acetaminophen reference standards, in the concentration range 0.004- 0.036 mg ml-l. Mix and leave to stand for a further 5 min. TABLE I EXPERIMENTAL PARAMETERS FOR THE UTILISATION OF THE NITROSO DERIVATIVES AND THEIR CHELATES IN QUANTITATIVE ANALYSIS Concentration Regression equation Percentage Compound Derivative Medium A,.range/mg-% A = a + b C fits Acetaminophen . . . . o-Sitroso Alkaline 430 0.4-3.6 A = 0.0056 + 0.2411C 99.68 Cobalt(11) chelate Chloroform 400 0.4-3.6 A = 0.0044 + 0.02105C 99.97 Copper(I1) chelate Aqueous 535 10.0-80.0 A = 0.005 8 + 0.021 22C 99.97 Salicylamide . . . . p-IVjtroso Acidic 398 0.2-2.0 A = 0.0503 + 0.4515C 99.91 p-h I troso Alkaline 393 0.1-1.0 A =z 0.0147 + 0.9766C 99.74 Copper(I1) chelate Aqueous 520 4.0-40.0 A = 0.0080 + 0.0257C 99.47 Salicylamide Add 2 ml of 3% sodium nitrite solution followed by 0.1 ml of 1 N hydrochloric acid. Heat the mixture over a boiling water-bath for 5 min and then cool. Next, either transfer directly into a 25-ml calibrated flask (acidic method) or add 1 ml of 1 N sodium hydroxide solution and then transfer into a 25-ml calibrated flask (alkaline method).Dilute to volume with water and measure the absorbance at A,,,. 398 nm using a blank prepared in the same way, but omitting the drug substance. Calculate the concentration of salicylamide using the linear equations (see Table I), which describe the calibration graphs prepared by applying the same procedure to salicylamide reference standards in the concentration ranges 0.002-0.02 and 0.001-0.01 mg ml-l for the acidic and alkaline methods, respectively. Molecular ratio of reactants solution with the reagents, according to the volumes and concentrations given in Table 11. Transfer 5.0 ml, containing 0.3 mg of salicylamide, into a 25-ml beaker.In a 25-ml calibrated flask mix the standard solutions of the drugs and sodium nitrite General Procedure for the Formation of Cobalt( 11) Chelate of the Nitroso Derivative of Acetaminophen and Preparation of Standard Calibration Graph Into a 100-ml separating funnel transfer 5.0 ml of the assay solution (containing 0.5 mg of drug), and add 2 ml of 3% sodium nitrite solution and 1 ml 1 N hydrochloric acid. MixOctober, 1979 ACETAMINOPHEN AND SALICYLAMIDE BY NITROSATION AND CHELATION 921 and leave to stand for 5 min. Extract the chelate three times with a total volume of 25 ml of chloroform. Measure the absorbance of the chloroform extract at 400 nm against a blank, prepared in the same way but omitting the drug substance.Calculate the concentration of acetaminophen using the linear equation (Table I), which describes the calibration graph prepared by applying the same procedure to solutions of acetaminophen reference standards in the concentration range 0.004-0.036 mg ml-l. Add 5 ml of a 0.125 M solution of cobalt(I1) bromide. TABLE I1 REAGENTS AND THEIR AMOUNTS USED IN THE DETERMINATION OF THE MOLAR RATIO These reagents are added to 25-ml calibrated flasks along with 2 ml of 304 sodium nitrite solution and 1 N hydrochloric acid (0.3 ml for acetaminophen and 0.5 ml for salicylamide), and made up to 25 ml with water. No. of flask r A Reagents 1 2 3 4 5 6 7 8 9 V,/ml of drug solution* . . . . . . 0.5 1.0 1.5 2 2.5 3 3.5 4 4.5 V,/ml of copper acetate solutionf . . . . 4.5 4 3.5 3 2.5 2 1.5 1 0.5 Acetic acid solution (2% V/V)/ml .. . . - 0.5 1.0 1.5 2 2.5 3.0 3.5 4.0 * The molar concentration of acetaminophen is 4 x lo-" M and of salicylamide is 2 x lo-" M. t The copper acetate is solubilised in acetic acid solution ( 2 ml of glacial acetic acid diluted to 100 ml with water). The molar concentration of copper acetate for acetaminophen is 4 x lo-" M and for salicylamide is 2 x M. General Procedure for the Formation of Copper( 11) Chelate of the o-Nitroso Derivatives of Acetaminophen and Salicylamide and Preparation of Standard Calibration Graphs Into a wide test-tube transfer 10.0 ml of the assay solution containing 10 mg of acetamino- phen (or 5 mg of salicylamide). Add 2 ml of 3% sodium nitrite solution and 2 ml of 1% copper(I1) acetate solution.Mix well and add 0.3 ml of 1 N hydrochloric acid (0.5 ml with salicylamide). Cool and quantitatively transfer into a 25-ml calibrated flask. Dilute to volume with water and measure the absorbance at the corresponding A,,,. (535 nm for acetaminophen and 520 nm for salicylamide) against a blank, prepared in the same way but omitting the drug substance. Calculate the concentration of acetaminophen and salicylamide using the linear equation (Table I), which describes the calibration graphs prepared by applying the same procedure to solutions of acetaminophen and salicylamide reference standards in the concentration ranges 0.1-0.8 and 0.04-0.4 mg ml-l for acetaminophen and salicylamide, respectively. Molecular ratio of reactants solution with the reagents, according to the volumes and concentrations given in Table 11.Place the reaction mixture in a boiling water-bath for 25 min. In a 25-ml calibrated flask mix the standard solutions of the drugs and sodium nitrite Preparation of Assay Solutions for Real Analysis Tablets one tablet, into a 100-ml calibrated flask. dilute to volume with water. filtrate, after a suitable dilution, as mentioned under the general procedure. Weigh and powder 20 tablets. Transfer an accurately weighed amount, equivalent to Dissolve as completely as possible in water and Treat the Filter and discard the first portion of the filtrate. Syrups or drops general procedure. Dilute an aliquot volume with water to attain a concentration suitable for use in the922 BELAL et a,?. : SPECTROPHOTOMETRIC DETERMINATION OF A?Zab.St, VOz.104 Calculation of the concentration in any of the above formulations was made using the Re-standardisation for such equations was occasionally regression equations given in Table I. checked. Wavelengthhm Fig. 1. Absorption curves of nitroso derivatives of: A, acetaminophen (1.75 mg-yo) in acidic medium (-) and alkaline medium (- - - -) ; B, salicylamide ( 1 mg-yo) in acidic medium (. . . .) and in alkaline medium (-.-.-)- Results and Illiscussion Nitrosation Reaction Acetaminophen and salicylamide, owing to the presence of aromatic hydroxy groups, are capable of forming nitroso derivatives (yellow chromogen) by reaction with sodium nitrite in an acidic medium. To investigate the molecular ratio of the reactants, Job’s method of continuous variation34 was employed.Lines A in Figs. 2 and 3 show that the maximum absorption occurs when the ratio of the reagent to the drug is 1 : 1. This is confirmed qualitatively by carrying out a test (coupling with a diazonium salt) €or the presence of the second free position ortho to the hydroxy group; Absorption curves are shown in Fig. 1. 0.6 Q) 0 (CI * 0.4 0, 2 0.2 0 0.2 0.4 0.6 0.8 1 .o 0 0.2 0.4 0.6 0.8 1.0 VM VM -t vL VM V 7 Fig. 2. Continuous variation graph for : A, nitrous acid - acetaminophen ; Fig. 3. Continuous variation graph for: B, o-nitroso-p-acetamidophenol - copper( I I) A, nitrous acid - salicylamide; B, o-nitro- chelate. sosalicylamide - copper(I1) chelate.October, 1979 ACETAMINOPHEN AND SALICYLAMIDE BY NITROSATION AND CHELATION 923 therefore, an 0- or $-nitroso derivative is indicated.( K J , calculated by the Harvey and Manning equation,35 are 1.01 x which indicate the high stability of these nitroso derivatives : Their respective instability constants and 3.68 x 10-5, (.'C) (ba'C)b c (1 - a') .Ki = where a' = degree of dissociation, C = molar concentration and b = number of molecules of the reagent. Compound I1 is unstable in acidic medium but can be stabilised by addition of methan01.3~ It was found that alkalinisation of the medium resulted in the formation of a highly stable compound and a bathochromic shift, together with hyperchromic effect. This is probably due to the contribution of an extra pair of unshared electrons in the interaction with the aromatic nucleus (Scheme A, compounds I11 and 111').OH OH 0- NHCOCH, NHCOCH, NHCOCH, &n ax. 370 nm hrnax.430 nm (1) ( 1 1 ) ( 1 1 1 1 5.-.- NHCOCH, ( I l l ' ) hmax. 400 nrn NHCOCH3 Scheme A $-Nitrososalicylamide (VI) attains its maximum colour intensity and stability when 0.1 ml of 1 N hydrochloric acid is used (as higher or lower values resulted in a decrease in the colour intensity) and when the reaction mixture is heated for 5 min on a boiling water- OH OH 1 N=O CONH;! @ CONH;! N-0 ( V I I ' ) Scheme B924 BELAL et at. : SPECTROPHOTOMETRIC DETERMINATION OF Analyst, VoZ. 104 bath. As compound VI is stable in either acidic or alkaline medium, it is assumed that the chromogen exists in the most stable tautomeric form, the p-quinonoidal structure, Scheme K (compounds VI' and VII'), in both media.This may explain why the absorbance maxi- ma of chromogens of this type do not exhibit a bathochromic shift on alkalinisation of the medium. Nitrosation with Subsequent Chelation In order to increase the selectivity of the nitrosation reaction, our investigation was extended to test for the chelating power of the nitroso derivatives. Nitroso-9-acetamido- phenol is capable of forming a chelate with cobalt(I1) ion, as the chelating agent contains a phenolic OH group (a group with an easily replaceable proton) ortho to the N=O group; the latter offers a lone pair of electrons to the cobalt(I1) ion. This chelate (Scheme A, IV), is sparingly soluble in water but easily extractable into chloroform to give a coloured solution, which can be measured at 370 nm.This is evidence that nitrosation occurs favourably in the para position to the phenolic OH group. Meanwhile, introduction of the nitroso group in the ortho position to the phenolic OH group is accomplished by the action of nitrous acid in the presence of a copper(I1) salt. This copper(I1) salt is essential for stabilising the nitrosyl radical and for ensuring that the ovtho (but not the pava) nitroso derivative is Under such conditions the copper chelates of the ovtho derivative of salicylamidle (with A,,,. 520 nm) and acetaminophen (with A,,,, 535 nm) are formed. The stoicheiometric ratio [o-nitroso derivative to copper( 11)] determined by Job's method of continuous variation is 2 : 1 (lines B in Figs. 2 and 3). The instability constants, calculated using the Harvey and Manning equation, are 4.813 x for acetaminophen and salicylamide, respectively.The maximum intensity of the chelates were obtained by the addition of 1 N hydrochloric acid in volumes of 0.3 ml for acetaminophen and 0.5 ml for salicylamide (Table 111), and by heating the reaction mixture in a boiling water-bath for 25min. Fig. 4 shows their absorption curves. This method is less sensitive than the two other recommended methods, yet it has the advantage of measuring the coloured chelate in aqueous solution without the need for its solvent extraction. On the other hand, p-nitrososalicylamide fails to give a chelate with cobalt(I1) ion. and 2.17 x TABLE 111 EFFECT OF THE VOLUME OF 1 N HYDROCHLORIC ACID ON THE COLOUR INTENSITY OF THE COPPER(II) CHELATES OF THE DRUGS" Volume of 1.0 N HCl/ml 0.05 0.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Measured pH* of the final diluted solution 4.8 4.4 4.1 3.78 3.00 2.48 2.40 2.34 2.13 2.00 1.79 Acetaminophen chelate, Amax. 535 nm 0.300 0.470 0.780 0.960 t t t t 0.000 0.000 0.000 Salic ylamide chelate, Amax. 520 nm 0.086 0.186 0.280 0.332 0.440 0.468 0.400 0.372 0.214 0.133 0.000 * Concentration of acetaminophen 80 mg-?(, and salicylamide 20 mg-%. Precipitate formed. Methods of Assay The high stability of the nitroso derivatives of acetaminophen and salicylamide, and their chelates, stimulated further investigation to develop spectrophotometric methods for the assay of these drugs.October, 1979 ACETAMINOPHEN AND SALICYLAMIDE BY NITROSATION AND CHELATION 925 0.6 C *-‘ .- 0.4 0 2 a 0.2 \ \-- - 480 520 560 600 Wavelengthhm Fig.4. Absorption curves of copper- (11) chelates of o-nitroso derivatives of acetaminophen (-) and salicyl- amide (- - - -) Standard calibration graphs for different derivatives were prepared by making serial dilutions of acetaminophen or salicylamide and applying the conditions specified under the corresponding General Procedure. Beer’s law is valid within the concentration ranges of acetaminophen or salicylamide calculated in the final dilution, given in Table I. Using the method of least squares3’ the regression equations for the different calibration graphs were derived and utilised for the calculation of unknown concentrations in dosage forms. The percentage fits38 for different calibration graphs were calculated (Table I).The validity of the regression equations was tested by analysing laboratory-made tablets. The results obtained had good accuracy and high precision (Tables IV and V). These results, including the precision of replicates, the agreement between the three versions of the recommended spectrophotometric methods and the good recovery from synthetic mixtures, encourage the application of the proposed methods in the routine analysis of pharmaceutical preparations of acetaminophen and salicylamide. Subjecting the results of the proposed method and pharmacopoeia1 methods1p2 to statistical analysis, calculated t exceeds theoretical t ; therefore, the null hypothesis is rejected and there is a significant difference between the results of the two methods,37 the proposed method giving more accurate results (Tables IV and V).TABLE IV APPLICATION OF THE NITROSATION REACTION TO THE ASSAY OF DIFFERENT DOSAGE FORMS O F ACETAMINOPHEN The results for the laboratory-made tablets are the percentage recoveries and the results for the commercial preparations are a percentage of the labelled claim. Mean (yo) and coefficient of variation* Pharmacopoeia1 Nitroso derivative Cobalt(I1) Copper(1i) method, (alkaline method), chelation, chelation, Dosage form n = 12 n = 16 n = 16 n = 16 Laboratory-made tablet . . . . - 100.5 100.5 100.5 Paracetamol tablet . . . . .. 101.0 99.6 99.4 99.5 ‘:;3”.)5 103.5 Paracetamol syrup .. .. 104.7 103.4 Pyral syrup . . .. . . .. 107.5 K . ) 4 105.5 105.4 101.5 Pyral syrup . . .. . . .. 102.3 101.5 101.5 (0.38) (0.3) (0.3) (0.3) (0.3) (0.3) (0.2) (0.3) (1.1) (0.2) (0.2) (0.4) (0.7) (0.2) (0.3) (0.3) (0.2) * The figures in parentheses are the coefficients of variation and n is the number of experiments.926 BELAL et al.: SPECTROPHOTOMETRIC DETERMINATION OF Analyst, VoZ. 104 TABLE V APPLICATION OF THE NITROSATION REACTION TO THE ASSAY OF DIFFERENT DOSAGE FORMS OF SALICYLAMIDE The results for the laboratory-made tablets are the percentage recoveries and the results for the commercial preparations are a percentage of the labelled claim. Mean (yo) and coefficient of variation* , 1 Nitroso derivative Pharmacopoeia1 A , Copper(I1) method, Acidic method, Alkaline method, chelation, n = 12 Dosage form n = 5 n = 12 n = 12 Laboratory-made tablet . . - (0.7) Cidal fort tablet .. . . 99.6 * The figures in parentheses are the coefficients of vaxiation and n is the number of experiments. Although the official ultraviolet spectrophotometric assay for acetaminophen1 is simple, its accuracy is greatly influenced by the interferences from diluents and binders in tablets or colouring matter, sweetening agents and preservatives in syrups and drops. The official non-aqueous titration for salicylamide2 suffers from non-selectivity and low sensitivity. The recommended methods can be applied only to formulations containing either acetamino- phen or salicylamide singly in unit dose preparations. However, the application of the proposed methods for the assay of formulations containing both acetaminophen and sali- cylamide is under investigation.References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. “British Pharmacopoeia 1973,” HM Stationery Office, London, 1973, p. 340. “The Pharmacopoeia of Japan 197 1 ,” Eighth Edition, Society of Japanese Pharmacopoeia, Tokyo, Kos, J., Favmaceutski Glasn., 1966, 22, 51; AnalLyt. Abstr., 1967, 14, 3529. Agarwal, S. P., and Walash, M. I., Indian J . Pharrn., 1974, 36, 47. Vaughan, J . B., J . Pharm. Sci., 1969, 58, 469. D’Sauza, A. A., and Shenoy, K. G., Can. J . Phavm. Sci., 1968, 3, 90. Vishwasrao, D. R., Indian J . Pharm., 1973, 35, 172. Inoue, T., Tatsuzawa, M., Lee, S.-C., and Ishii, T., Eisei Kagaku, 1975, 21, 313; Analyt. Abstr., Welch, R. M., and Conney, A. H., Clin. Chem., 1965, 11, 1064.Davis, D. R., Fogg, A. G., Thorburn Burns, D., and Wragg, J. S., Analyst, 1974, 99, 12. Deodhar, R. D., Shastri, M. R., and Mehta, R. C., Indian J . Pharrn., 1973, 35, 120. Hanegraaff, C., and Chastagner, N., Annls Pharm. Fr., 1969, 27, 663. le Perdriel, F., Hanegroaff, C., Chatcugner, N., and DeMontetry, E., Annls Pharm. Fr., 1968, 26, Kaito, T., Sagara, K., Yoshida, T., and Ito, Y., Yakugaku Zasshi, 1974, 94, 633 and 639; Analyt. Kaito, T., Kasuya, K., and Inoue, T., Bunseki Kagaku, 1971, 20, 801; Analyt. Abstr., 1972, 23, Kaito, T., and Sagara, K., Yakugaku Zasshi, 1974, 94, 639; Analyt. Abstr., 1976, 30, 2E33. Kalinoswka, 2. E., and Hasztar, H., Farrnacja Pol., 1965, 21, 570; Analyt. Abstr., 1965, 13, 7092. Kalinowska, Z. E., and Hasztar, H., Farmacja Pol., 1967, 23, 447; Analyt.Abstr., 1968, 15, 4978. Poethke, W., and Kohne, H., Pharm. Zentralhalle Dtl., 1965, 104, 630; Analyt. Abstr., 1967, 14, Inamdar, M. C., Saboo, J. C., Kamdar, C. N., and Sanghavi, N. M., Indian J . Pharm., 1973, 35, 187. Brockelt, G., Pharmazie, 1965, 20, 136. Zoltai, E., Acta Pharm. Hung., 1967, 37, 74; Analyt. Abstr., 1968, 15, 3551. Ruttkowski, R., Arzneimittel-Forsch., 1954, 4, 209. Ensor, P. C., J, Ass. Publ. Analysts, 1972, 10, 56. Murai, K., Arch. Proc. Pharm., Japan, 1961, 21, 58; Analyt. Abstr., 1963, 10, 3870. Tatsuzawa, M., and Hashiba, S., Bunseki Kagaku, 1968, 17, 478; Analyt. Abstr., 1970, 18, 1218. Joy, J,, and Szekeres, L., Mikrochirn. Acta, 1975, 11, 125; Analyt. Abstr., 1976, 30, 2E30. Szekeres, L., Harmon, R. E., and Gupta, S. K., Microchem. J., 1973, 18, 101; Analyt. Abstr., 1973, 1971, p. 674. 1976, 31, 1E29. 227. Abstr., 1976, 30, 2E33. 2776. 370. 25, 4046.OCtObe?’, 1979 ACETAMINOPHEN AND SALICYLAMIDE BY NITROSATION AND CHELATION 927 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. Grabowska, I., and Weclawska, K., Farmacja Pol., 1976, 32, 363; Analyt. Abstr., 1977, 31, 5E42, Miszczuk-Lucka, B., and Taborska, H., Przem. Chena., 1955, 11, 706; Analyt. Abstv., 1957, 4, 1016. Gagewska, M., and Ciszewsk, M., Acta Pol. Pharm., 1975, 32, 607; Analyt. Absti.., 1976, 31, 1E27. Inamdar, M. C., and Kadji, N. N., Indian J . Pharm., 1969, 31, 79. Chafetz, L., Daly, R. E., Schriftman, H., and Lomner, J . J., J . Pharm. Sci., 1971, 60, 463. Rose, J., “Advanced Physico-Chemical Experiments,” Pitman, London, 1964, p. 54. Harvey, A. E., and Manning, D. L., J . A m . Chew. Soc., 1950, 72, 4488. Finar, I. L., “Organic Chemistry, The Fundamental Principles,” Volume 1, Fifth Edition, Long- Spiegel, M. R., “Theory and Problems of Probability and Statistics,” McGraw-Hill, New York, Davies, 0. L., and Goldsmith, P., “Statistical Methods in Research and Production,” Fourth Edition, mans, Green, London, 1967, p. 676. 1975, pp. 259 and 215. Oliver and Boyd, Edinburgh, 1972, p. 178. Received April 24th, 1978 Amended December 1 lth, 1978 Accepted February 14th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9790400919

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

928 Analyst, October, 1979, Vol, 104, pp. 928-936 Spectrophotometric Determination of Aliphatic lsocyanates in the Occupational Atmosphere Part 1. Determination of Total Isocyanate Concentration R. F. Walker and M. A. Pinches Occupational Medicine and Hygiene Laboratories, Health and Safety Executive, 403-405 Edgware Road, Cracklewood, London, N W2 6LN This paper describes a spectrophotometric method for use in the field for the determination of aliphatic isocyanates and their oligomers in air. The atmosphere being tested is drawn through a mixture of hydrochloric acid and dirriethyl sulphoxide a t a sampling rate of 2 1 min-l for 10 min. Any isocyanates or oligomer present are hydrolysed to the corresponding amine. l-Fluorc1-2,4-dinitrobenzene is added and forms coloured derivatives with the amines.The absorbance of each derivative is measured at approximately 353 nm, and can be directly related to the amount of isocyanate. Concentra- tions down to 0.002 p.p.m. can be determined. Keywords : Aliphatic isocyanate determination ; spectvophotornetry ; l-fluoro- 2,4-dinitrobenzene reagent In general, urethane polymers made from aromatic isocyanates tend to yellow on prolonged exposure to sunlight (possibly because ultraviolet light promotes autoxidation of these polymers to form quinone - imide based products). However, aliphatic isocyanates are used to make urethane polymers with better resistance to ultraviolet-induced discoloration. Consequently, polyurethanes requiring exceptional light stability are largely based on aliphatic isocyanates. Such two-pack polyol-cure paint systems are being used increasingly in the vehicle re-finishing industry.The following aliphatic isocyanates are among the most widely used. This is unsuitable for the production of surface coatings because of its volatility and physiological activity. By treating HMDI with water (3 : 1 molar ratio) it is possible to produce a polyisocyanate of high relative molecular mass, which is no longer volatile and which is used to make polyurethane coatings for air drying and stoving. This compound has two isocyanate groups of different reactivities (the cycloaliphatic group is about ten times less reactive than the aliphatic grouping1), enabling adducts and oligomers to be produced that have a low residual diiso- cyanate monomer content (about 0.7%).These products are used in the production of transparent elastomers, foams, moisture-curing one-pack and polyol-curing two-pack paint systems. Methylene bis(4-cyclohexylisocyanate) (MCHI) . This is used in the manufacture of colourless, light-stable foams and coatings. Trimethylhexamethylene diisocyanate (TMDI). This material consists of a mixture of the 2,2,4- and 2,4,4-isomers. It is used in the production of flexible foams, elastomers and fibres. Like other aliphatic isocyanates, TMDI is mainly used in the form of adducts, condensates or oligomers when used for surface coatings. These diisocyanates indicate the chemical types in the aliphatic isocyanate class which, when combined with polyfunctional polyethers, polyesters or other compounds containing hydroxy groups, are used in the production of polyurethanes.Until recently it was thought that the toxic hazard associated with the application of polyurethane-based surface coatings was a result of monomeric isocyanate vapour in the atmosphere. However, medical evidence2 now suggests that vapour phase inhalation is not the sole method by which isocyanates can enter the respiratory tract, and that inhalation of spray droplets is equally hazardous to health. Therefore, despite the low vapour pressures 1,6-Hexamethylene diisocyanate (HMDI). Isophorone diisocyanate (IPDI). Crown Copyright.WALKER AND PINCHES 929 of the adducts and oligomers gencrally employed in the formation of surface coatings, hazardous concentrations of these polyisocyanates may occur as droplets or as aerosols in industrial atmospheres where the surface coatings are sprayed. The only aliphatic diiso- cyanates with published threshold limit values (TLVs) are IDPI and hICHI,3 but it is wise t o assume that the atmospheric concentration of all aliphatic diisocyanates should not be allowed to exceed 0.01 p.p.m. The spectrophotometric method of von E i ~ k e n , ~ later modified by Pilz and Johann,5 has been used to determine concentrations of atmospheric HMDI. More recently Dunlap et aZ.697 have developed chromatographic methods for determining concentrations of aliphatic isocyanates.Both methods are based on the reaction of isocyanates with N-4-nitrobenzyl- N-propylamine followed by analysis with thin-layer chromatography6 or high-performance liquid chr~matography.~ The method proposed by von Eicken4 was selected as being suitable for adaptation to a general field-test procedure for the aliphatic isocyanates commonly used in industry.Basically the method consists in hydrolysing any isocyanate present to the corresponding amine, followed by reaction with l-fluoro-2,4-dinitrobenzene to form a 2,4-dinitrophenyl- amine derivative. The absorbance of the derivative is then measured at about 353 nm. Both aromatic and aliphatic isocyanates respond quantitatively to this method. A critical examination of the various steps of the von Eicken method was carried out in order to determine the optimum conditions for the collection and analysis of aliphatic isocyanates and their related oligomers. Experiment a1 Preparation of Standard Atmospheres Dynamic standard atmospheres of a range of aliphatic isocyanates were required in order t o assess the efficiency of the sampling procedure and for application in the development of the analytical method.A double-dilution atomisation apparatus* was used, which generated known concentrations of isocyanate in toluene into a constant-flow stream of dried air. The atmospheres were monitored in duplicate; one sample was analysed by the proposed spectro- photometric method, and the other by a long path length infrared gas analyser (Miran l A , Wilks - Foxboro Analytical). The infrared gas analyser was calibrated in advance by injecting known volumes of each aliphatic isocyanate into a closed-shop calibration system, which could be attached to the analyser when required. The results are summaried in Table I.TABLE I ANALYSIS OF ALIPHATIC ISOCYANATE STANDARD ATMOSPHERES Number Isocyanate of runs HMDI . . 5 8 4 IPDI . . . . 7 6 10 MCHI . . 5 9 4 TMDI . . 11 7 4 Calculated concentration* of isocyanate atmosphere a t 20 "C, p.p.m. 0.048 0.105 0.230 0.045 0.112 0.230 0.049 0.091 0.216 0.043 0.104 0.224 Measured concentration of isocyanate atmosphere, p.p.m. By infrared By spectrophotometry r----J------7 & Standard Standard Mean deviation Mean deviation 0.033 0.01 1 0.032 0.004 0.075 0.009 0.071 0.006 0.171 0.006 0.165 0.007 0.037 0.010 0.039 0.005 0.093 0.005 0.097 0.007 0.191 0.008 0.179 0.004 0.042 0.006 0.048 0.006 0.075 0.005 0.078 0.003 0.178 0.004 0.174 0.006 0.030 0.006 0.034 0.004 0.078 0.010 0.085 0.006 0.169 0.01 1 0.149 0.009 r - * Based on the purity of each isocyanate as determined by HPLC.Q By monitoring the stretching vibration of the N=C=O bond at approximately 4.45 pm and using a cell with a 20-m path length, the lower limit of detection using the infrared930 WALKER AND PINCHES : SPECTROPHOTOMISTRIC DETERMINATION OF Analyst, VoZ.104 method was 0.03 p.p.m. The results given in Table I show that good agreement was obtained between the analyses by the two methods for the range of isocyanate atmospheres examined. The standard deviation for each set of measurements is also given. Once a correlation had been established between the analyses by the infrared and spectrophotometric methods, atmospheres at the 0.01 p.p.m. level could be determined by the proposed spectrophotometric method. The latter method gave excellent straight line graphs over the concentration range 0.002-0.03 p.p.m.The coefficients a and b for the straight-line equations A = a -t bc where .A is the absorbance obtained from an isocyanate concentration c , and the correlation coefficients, y2, are given in Table I1 for this range of concentrations. The results are based on a 20-1 sample. TABLE I1 ALIPHATIC ISOCYANATE CALIBRATION LINES A = a + bc OVER THE CONCENTRATION RANGE 0.002-0.03 p .p.m. Isocyanate a b r2 HMDI . . . . 0.0025 5.461 5 0.999 2 0.9996 IPDI . . . . -0.0040 6.4109 MCHI . . . . -0.0020 6.9099 0.9995 TMDI . . . . -0.0115 9.6942 0.9995 Oligomer isocyanates with high relative molecular masses were found to be unsuitable for standard atmosphere generation.In these instances solutions of the isocyanate, diluted with dimethyl sulphoxide, were used instead. Selection of Absorbing Solution Previous work4y5 had shown that dimethyl sulphoxide (DMSO) was the only available common organic solvent with a high boiling-point that was suitable for the absorption of aliphatic isocyanates. However, bottles containing DMSO would often freeze during the winter months because of the high freezing-point (18 "C) of this solvent.1° Therefore, absorbing solutions of equal volumes of DMSO and 4.0 M hydrochloric acid were prepared in the laboratory prior to sampling. In addition, no ageing effects were noticed when using this mixture. The original method4 devised for the determination of HMDI in air made use of a mixture of DMSO and 0.1 M hydrochloric acid as the absorbing solution.Subsequent work5 showed that only partial hydrolysis of HMDI was achieved at this acid concentration and suggested 0.8 8 0.7 0.5 0.3 1 I 0.2 1 0.1 ~ H M D I oligomerj Acid concen t r a t ion/M Fig. 1. Effect of hydrochloric acid concentration on the hydro- lysis of 3 x hi HMDI and 2 x M HMDI oligomer.October, 1979 ALIPHATIC ISOCYANATES I N THE OCCUPATIONAL ATMOSPHERE. PART I 931 the use of 1.2 M hydrochloric acid. The present investigation shows that although mono- meric isocyanates are readily hydrolysed by 0.1 M hydrochloric acid at ambient temperatures, oligomeric isocyanates are completely hydrolysed only by acid concentrations in excess of 4.0 M. Fig.1 shows the effect of varying the acid concentration on the extent of hydrolysis for HMDI and its commercial oligomer after 15 min at 25 "C. Thus, 4.0 M was selected as being the optimum hydrochloric acid concentration for the hydrolysis of both aliphatic isocyanates and their oligomers. The efficiency of the original sampling procedure was examined by passing a standard HMDI atmosphere through three impingers in series at a rate of 2 1 min-l; each impinger contained 8 ml of the absorbing solution. HMDI was chosen for this experiment because it has a relatively low boiling-point compared with those of the other aliphatic isocyanates studied. Standard atmosphere concentrations were generally found to be 30% less than their theoretical value because of isocyanate loss on the glass walls of the apparatus.The results are indicated in Table 111. The three-impinger sampling system is assumed to have a collection efficiency of lO0yo; it can be seen that there is 14% isocyanate breakthrough from the first impinger when sampling a 710,~.gm-~ (0.10 p.p.m.) HMDI atmosphere at 2 1 min-I for 20 min. Most field sampling is done in atmospheres with 0.01-0.02 p.p.m. concentrations and hence a single impinger with a collection efficiency of approximately 94% at these concentrations can be used (Table 111). TABLE I11 EFFICIENCY OF SAMPLING PROCEDURE FOR THE COLLECTION OF HMDI IN 4.0 M HYDROCHLORIC ACID AND DMSO (1 + 1 V / V ) Calculated concentration* Number of atmosphere a t of runs 20 "C/pg m-3 3 25.34 2 25.34 2 75.28 2 75.28 2 165.71 3 165.71 3 320.50 2 320.50 2 483.67 3 483.67 2 710.33 3 710.33 3 710.33 * See Table I.Sanipling rate/l min-l 1.0 2.0 1.0 2.0 1 .o 2.0 1.5 2.0 1 .o 2.0 1.0 1.5 2.0 Sampling time/ in in 10 15 15 15 10 15 15 20 20 15 20 20 20 Mean concentration of isocyanate collccted/pg n1r3 Trap 1 16.1 17.9 52.1 56.3 127.5 121.0 243.5 258.3 381.1 369.7 583.2 570.4 525.5 Trap 2 Trap 3 1.5 0.0 1.2 0.0 1.3 0.0 3.2 0.0 6.4 0.0 11.8 0.6 29.3 1.4 22.6 0.8 39.4 1.1 47.7 1.3 38.6 0 . 9 49.7 1.7 83.2 2.1 Choice of Reagent l-Fluoro-2,4-dinitrobenzene was introduced as a quantitative reagent for primary and secondary amines by Sangerll in the determination of the free amino groups in proteins and peptides. Subsequently it has become one of the most important reagents in the analysis of amino acids and peptides.12 In addition to l-fluoro-2,4-dinitrobenzene, l-chloro-2,4- dinitrobenzene, I-cl~loro-3,5-dinitrobenzcne and l-fluoro-3,5-dinitrobenzene were also examined in order to find which gave both a fast reaction rate with aliphatic amines and a high molar absorptivity.l-Fluoro-2,4-dinitroberizene was found to be the most suitable reagent and was used for this work. I t should be noted that l-fluoro-2,4-dinitrobenzene has been shown13 to give positive results in tests for mutagenesis, and should therefore be con- sidered a possible carcinogen. Reaction Conditions In the original method, complete reaction was achieved by heating the reagents to approximately 70 "C. Standard solutions of the aliphatic isocyanates were made up, hydrolysed and treated with 0.1 ml of 1 yo and 1076 l-fluoro-~,~-dinitrc,benzene solutions over a range of temperatures in order to determine the optimum temperature required for complete reaction.These concentrations of l-Auoro-2,4-dinitrobenzene were considered to932 WALKER AND PINCHES : SPECTROPHOTOMETRIC DETERMINATION OF Analyst, T/d. 104 be the two extremes and both produced similar curve profiles. The results are shown in Fig. 2 for 15-min heating periods using a 1% solution. It can be seen that heating the reagents above 80 "C causes some degradation of the coloured products. Below 70 "C a longer period of heating is required for full colour development; 75 "C was chosen as the optimum temper at ure . 0.4 0.3 u C m a 0.2 Ll 6 0.1 0 p' MCH I i 60 70 80 90 Temperature/" C Fig.2. Effect of varying temperature on the reaction between l-fluoro-2,4-dinitro- benzene and aliphatic iso- cyanates. (Isocyanate con- centration was 9 x lop6 M.) 0.8 r 1 0.7 c I C 0.6 0.5 ' 0.4 6 0.3 0.2 0.1 W 2 a 0 0.2 0.4 0.6 0.8 1.0 Reagent concentration, % V/V Fig. 3. Effect of varying the concentration of l-fluoro-2,4-dinitrobenzerie on the absorb- ance a t 353 nm of the diamine derived from HMDI, with varied heating periods. (Iso- cyanate concentration was 1.8 x M.) A, 5 min; B, 15 min; and C, 30 min. A series of experiments was carried out for each aliphatic isocyanate using increasing concentrations of l-fluoro-2,4-dinitrobenzene to find the optimum reagent concentration. The results are shown in Figs. 3-7 for heating periods of 5, 15 and 30 min at 75 "C with 0.1 ml of reagent and a 1.8 x In each instance an increase in the reagent concentration eventually produced a decrease in the heating period required for full colour development of the amine derivative. Heating for periods greater than the optimum time causes a marked degradation in the absorbance of the colour derivatives.I t can be seen that by using 0.1 ml of a 0.3% l-fluoro- 2,4-dinitrobenzene solution, full colour development was obtained within 15 min at 75 "C. M concentration of isocyanate. 0.7 0.6 a, 0.5 0.4 Q 0.3 13 I) 0.2 O.' 0 L1 0.2 0.4 0.6 0.8 1.0 Reagent concentration, % V/V Fig. 4. Effect of varyin,? the concentration of 1 -fluoro-2,4-clinitrobenzene on the absorb- ance a t 352 I i m of the dialnine derived from IPDI, lvith varied heating periods.(Iso- cyanate concentration was 1.8 :* 10F5 M.) A, 5 min; B, 15 min; and C, 30 min.October, 1979 ALIPHATIC ISOCYANATES IN THE OCCUPATIONAL ATMOSPHERE. PART I 933 0.2 0.4 0.6 0.8 1.0 Reagent concentration, % V/V Fig. 5. Effect of varying the concentration of l-fluoro-2,4-dinitrobenzene on the absorb- ance at 355 nm of the diamine derived from MCHI, with varied heating periods. (Iso- cyanate concentration was 1.8 x l O - 5 ~ . ) A, 5 min; B, 15 min; and C, 30 min. This time is considered satisfactory for the purposes of the proposed field test. Although a more concentrated solution required less heating time, the blank solutions produced in this way had an absorbance that was too great for accurate measurement of the sample solutions. 1 .o 0.8 u CJ 0.6 2 a I) 0.4 0.2 0 L 0.2 0.4 0.6 0.8 1 .o Reagent concentration, % V / V Fig.6. Effect of varying the concentration of 1- fluoro-2,4-dinitrobenzene on the absorbance a t 353 nm of the diamine derived from TMDI, with varied heating periods. (Isocyanate concentration was 1.8 x M.) A, 5 min; B, 15 min; and C, 30 min. Determination of Wavelength Absorbance Maxima Aliquots containing 15 pg of the standard isocyanate solutions were dispensed into 8 ml of the absorbing solution by use of an adjustable micropipette (Pipetman, Gilson Commercial Service). The mixtures were agitated and allowed to stand for 10 rnin at room temperature in order to ensure complete hydrolysis of the isocyanates to the corresponding primary amines. Colour derivatives were developed by using the proposed field method and double- extracted with AnalaR 1,1,2-trichloroethane. The absorbances of the derivatives were measured in the range 320-380 nm against a blank solution.The wavelengths of maximum absorbance were found to be approximately 353 nm for the aliphatic isocyanate monomer derivatives and 350 nm for the HMDI oligomer derivative.934 WALKER AND PINCHES : SPECTROPHOTOMETRIC DETERMINATION OF Analyst, VoZ. 104 u 0.2 0.4 0.6 0.8 1 .o Reagent concentration, % VIV Fig. 7. Effect of varying the concentration of 1- fluoro-2,4-dinitrobenzene on the absorbance at 350 nm of the triamine derived from HMDI oligomer, with varied heating periods. (Isocyanate concentration was 1.8 x 1 0 - 5 ~ . ) A, 5niin; B, 15min; and C, 30 min. Molar absorptivities, E , have been calculated for each reaction product (Table IV).It is interesting to note that the molar absorptivity of the HMDI oligomer is considerably higher than that of HMDI itself. Thus, it is unlikely that the central isocyanate grouping of the oligomer (see Fig. 1) is sterically hindered, as is the case for certain secondary amines to this reaction, e.g., diisopropylamine and dicy~lohexylamine.~~ A commercial sample of HMDI oligomer was analysed by use of HPLCg and found to contain 0.82 If 0.05% of free HMDI. A correction factor has therefore been included in the calculation of the molar absorptivity of the oligomer (based on three functional amino groups). TABLE IV MOLAR ABSORPTIVITIES OF THE SECONDARY DIAMINES FORMED BY THE REACTION OF ALIPHATIC DIISOCYANATES WITH 1-FLUORO-2,4-DINITROBENZENE Isocyanate .. HMDI IPDI MCHI TMDI HMDI oligomer f . . . . 33729 f 873 38610 f 649 41.803 f 837 57885 & 1567 45550 3 7171 ~ m a x . l n m * * 353 352 355 353 350 Preparation of Calibration Graphs Calibration graphs for the various aliphatic i:socyanates studied were prepared by using standard isocyanate solutions and adding known aliquots to 8 ml of the absorbing solution. This was considered to be a more reliable method than standard atmosphere generation, where flow fluctuations, albeit minor, and isocyanate adsorption and desorption may occur. Some difficulty was encountered in quantitatively dissolving the necessary amounts of HMDI oligomer (about 200 mg) in DMSO. Satisfactory dissolution was achieved by weighing accurately about 200mg of the oligomer on to a glass slide, which was then placed in a 50-ml glass beaker containing 25 ml of DMSO for approximately 20 min.The solution was then transferred to a 100-ml flask and the procedure repeated, if necessary, in order to remove residual traces of the oligomer from the slide. The solution was made up to volume with DMSO. Aliphatic isocyanate monomer solutions were prepared by dispensing suitable aliquots of the isocyanates directly into 100-ml flasks and making up to 100 ml with DMSO. Occasionally the solutions became opalescent on being made up to volume, but vigorous shaking for 2 or 3 min dissipated any cloudiness. Calibration graphs, over the concentration range 0-20 pg, were prepared for all of the isocyanates studied, by using the optimum conditions determined for each stage of the procedure.A linear response was exhibited between the concentration of each isocyanate and the absorbance of its tliamine. Table V shows the absorbances,October, 1979 ALIPHATIC ISOCYANATES IN THE OCCUPATIONAL ATMOSPHERE. PART I 935 TABLE V ABSORBANCE VALUES PRODUCED BY THE PROPOSED FIELD METHOD Isocyanate ~ HMDI IPDI MCHI TMDI HMDI oligomer L - f - 7 r - - - \ r - - - - - - ~ 7- Amount of Standard Standard Standard Standard Standard isocyanatelyg Mean deviation Mean deviation Mean deviation Mean deviation Mean deviation 5 0.208 0.09 0.170 0.06 0.155 0.04 0.265 0.06 0.051 0.03 10 0.392 0.08 0.349 0.08 0.324 0.07 0.564 0.05 0.213% 0.02 15 0.595 0.05 0.531 0.09 0.485 0.07 0.829 0.04 0.255 0.09 20 0.803 0.05 0.697 0.10 0.648 0.09 1.117 0.11 0.330 0.07 measured on a Pye Unicam SP6-500 spectrophotometer at the appropriate wavelengths (see Determination of Wavelength Absorbance Maxima) with a 10-mm path length quartz cell.Five determinations were made at each concentration. Interferences N-Ethylmorpholine, NN-dimethylcyclohexylamine and other tertiary amines commonly used as catalysts in the polyurethane industry do not interfere in the proposed field method. Primary amines, if present, will interfere and the determination of aliphatic iso- cyanates in the presence of primary amines, together with data from field trials, will be the subject of Part I1 of this paper. Proposed Field Method for the Determination of Aliphatic Isocyanates in Air Apparatus 1-cm path length quartz cell.inlet tube and a flat-bottomed receiver. 2 1 min-1. ture to 100 "C. Spectrophotometer. Impingers. Sampling pump. Heating block. Mechanical stirrer. Fisons Whirlimixer. A Pye Unicam SP6-500 spectrophotometer equipped with a 4-ml, All-glass midget impingers of the Greenberg - Smith type, with a tapered jet A sampling pump capable of drawing air through the apparatus at A heating block (Techne, Model DB-3) with a range from ambient tempera- Reagents All the reagents were of analytical-reagent grade unless otherwise stated. Absorbing solution. Standard sodium hydroxide solutions, 0.1 and 0.4 M. Sodium borate bufer solution. Dilute 36 ml of concentrated hydrochloric acid (sp. gr. 1.19 at 20 "C) to 100 ml with distilled water. Mix 50 ml of the resulting solution with 50 ml of DMSO. Dissolve 30.0 g of boric acid in 700 ml of distilled water.Add slowly, while stirring, 4.0 M sodium hydroxide solution until the pH is 8.8, as indicated by a pH meter. Pipette 0.3 ml of 1 -fluoro-2,4-dini t ro benzene into 5ml of DMSO in a 10-ml flask. Prepare a fresh solution daily. Suitable amounts of the aliphatic isocyanates to be moni- tored are dissolved in DMSO and the solutions made up to 100 ml. For example, 35 pl of HMDI (Bayer AG) are dissolved in 100 ml of DMSO (see Preparation of calibration graphs). Freshly prepared solutions were used. 1 - Fluoro-2,4-dinitro benzene solution. Dilute the solution to volume with DMSO. Standard isocyanate solutions. 1 , l ,Z-Trichloroetlzane. AnalaR-grade material was used. Procedure impinger receivers.and a second to which 10 p1 of the appropriate standard isocyanate solution are added. standard is included in the test procedure as a check on the calibration graphs. In an uncontaminated atmosphere pipette 8.0 ml of the absorbing solution into the midget- Set aside one impinger so that its contents act as the blank solution The Insert the936 WALKER AND PINCHES inlet tubes and position the impingers at the sampling site in a vertical position (the position was found to be especially important when monitoring aliphatic isocyanates in paint aerosols). Alternatively, a midget impinger can be attached to the worker as close to his breathing zone as possible, usually to his lapel. Attach the pump to the impinger and draw a sample of the atmosphere being tested through the absorbing solution at a flow-rate of 2 1 min-1 for at least 10 min.Do not attach rubber or poly(viny1 chloride) tubing to the air inlet of the impinger as the tubing will adsorb isocyanate vapour or spray and this adsorption causes low results. Disconnect the impinger from the pump, remove the impinger and its contents to an uncontaminated atmosphere, and allow them to stand for 10 min to ensure complete hydrolysis of the isocyanate collected. Next lift the inlet tube clear of the liquid in the impinger receiver tube and expel the liquid from the inlet tube with a blow-ball. Neutralise the solution in the impinger with 4.0 ml of 4.0 M sodium hydroxide solution (checking the pH with indicator paper). By US(: of a pipette, add 1.0 ml of sodium borate buffer solution and mix.Then add 0.1 ml of 0.3% l-fluoro-2,4-dinitrobenzene solution and again mix. Heat the mixture for 15 min at 75 “C on a heating block (taking care not to heat it above 75 “C). Below 70 “C a longer period of heating is required for full colour development. Remove the impinger receiver tubes from the heating block and allow their contents to cool for 5 min. Next add 4.0 ml of 0.1 M sodium hydroxide solution and mix. Pipette 3.0 ml of 1,1,2-trichloroethane into each receiver tube and mix the contents on a Whirlimixer for 30 s, then transfer the liquid to a 25-ml separating funnel, allowing the two phases to separate, and filter the lower organic layer through a folded sheet of 7-cm Whatman No. 2 filter-paper into a 5-ml flask.Return the aqueous layer back into the receiver tube and repeat the extraction with 2.0 ml of 1,1,2-trichloroethane. Combine the second extract with the first, and make up the solution to 5 ml with 1,1,2-trichloroethane. Finally, shake the flask. The absorbance is measured a t the appropriate wavelength (see Table IV), using the blank solution to zero the spectro- photometer. Determine the amount of isocyanate by reference to the appropriate calibra- tion graph. The solution is stable for at least 24 h. Conclusion and and 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. A spectrophotometric method for use in the field has been developed for the determination of low concentrations of atmospheric aliphatic isocyanates. The limits of detection are significantly below the current threshold limit value. The proposed method has been satisfactorily assessed under field conditions at a number of car re-finishing industries. The equipment necessary to carry out the analysis is portable requires only an electrical power supply to operate the spectrophotometer, heating block mechanical stirrer. A determination can be completed in about 1 h. References “Isophorone Diisocyanate IPDI,” Product Information Sheet 22-ME-871-6, Veba-Chemie AG, Bunge, W., Ehrlicher, H., and Kimmerle, G., Zentbl. Arbeitsmed. Arbeitsschutz Prophylaxe, 1977, 4, Health and Safety Executive, “Threshold Limit Values for 1977,” Guidance Note EH15/77, HM von Eicken, S., Mikrochinz. Acta, 1958, 731. Pilz, W., and Johann, I . , Mikrochim. Acta, 1970, 351. Keller, J . , Dunlap, K. L., and Sandridge, R. I>., Analyt. Chern., 1974, 46, 1845. Dunlap, K. Id., Sandridge, I<. L., and Keller, J . , Analyt. Chem., 1976, 48, 497, Meddle, D. \V., and Wood, K., Cheiny I n d . , 1968, 1635. Cox, G. U., and Sugden, K., Analytica Chiw.. Acta, 1977, 91, 365. Weast, R. C., Editor, “Handbook of Chemistry and Physics,” Fifty-third Edition, CRC Press, Cleveland, Ohio, 1973, p. C-501. Sanger, F., Biochem. J . , 1945, 39, 507. Pataki, C . , “Thin Layer Chrornatography in Aminoacid and Peptide Chemistry,” W. de Gruyter, Purchase, I. F. H., Longstaff, E., Ashby, J., Styles, J . A., Anderson, D., Lefevre, P. A., and West- ?kTcIntire, F. C., Clements, W. M., and Sproull, M., Analyt. Chew.., 1953, 25, 1758. Gelsenkirchen, Germany, 1971. 1. Stationery Office, London, 1978. Berlin, 1966. wood, I;. I<., R r . J . Cancer, 1978, 37, 873. Received February 19th, 1979 Accepted A p r i l 19th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400928

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Analyst, October, 1979, Vol. 104, $9. 937-943 937 Influence of Conditioning Agents on the Determination of Metallic Content of Sewage Sludge by Atomic-absorption Spectrophotometry with Electrothermal Atomisation M. J. T. Carrondo, R. Perry and J. N. Lester Public Health Engineering Laboratory, Imperial College of Science and Technology, London, S W7 2AZ Conditioning agents are often used to aid the de-watering of sewage sludges prior to disposal to agricultural land. These might interfere in the electro- thermal atomic-absorption spectrophotometric analysis of heavy metals. Possible interferences by inorganic and polyelectrolyte conditioners were studied at the higher rates of addition normally used in sewage practice. Analyses for cadmium, chromium, copper, nickel, lead and zinc were per- formed by flame and electrothermal atomisation methods in conditioned and unconditioned samples.The organic polyelectrolytes tested did not interfere, nor did most inorganic conditioners a t rates of addition consistent with normal sewage treatment practice. However, interferences occurred with aluminium chlorohydrate a t normal and very high addition rates, and with other inorganic conditioners a t very high addition levels. Keywords: Electrothermal atomic-absorption spectrophotornetry ; cadmium, chromium, copper, nickel, lead and zinc determination ; sewage sludge analysis ; interferences ; inorganic and polyelectrolyte conditioners During the process of wastewater treatment, heavy metals contained in the incoming sewage tend to be concentrated in the sludge pr0duced.l In the UK it is estimated2 that 40% of the sludges from all inland sewage treatment works are disposed to agricultural land.Liquid digested sludge is the most favoured form for land application, but because of handling problems many sludges are de-watered prior to application. To improve the de-watering process, conditioners are generally added.3 These include iron( 11) sulphate (copperas), iron(II1) chloride, aluminium sulphate, hydrated aluminium chloride (aluminium chloro- hydrate) and lime, or mixtures of t h e ~ e . ~ ? ~ More recently, organic compounds, mainly cationic polyelectrolytes, have been ~ s e d . ~ - ~ Inorganic conditioners must be added to the sludge in large amounts to achieve their effect (see Table I), whereas the organic conditioners will generally not exceed 1% of the dry matter present in the ~ l u d g e .~ - ~ Because of the toxic6 or phytotoxic7 nature of some of the heavy metals present, their concentrations must be determined prior to disposal to land to ensure that maximum permissible levels8 are not exceeded. The methods available for the determination of heavy metals in sewage sludges require some form of pre-treatment to destroy organic matter prior to analysis and this is time consuming. Recently, a rapid electrothermal atomic-absorption spectrophotometric method, utilising homogenisation as the only pre-treatment, has been successfully developed for the determination of cadmium, copper, chromium, nickel, lead and zinc in unconditioned sewage sludge.gJ0 Various causes have been suggested for the occurrence of interferences in graphite furnace atomic-absorption spectrophotometry.It has been suggested that interferences are due to physical effects, formation of molecular compounds in the gas phase resulting in molecular absorption, formation of refractory compounds and interaction between elements and gaseous compounds.ll The work of Sturgeon et a1.12 indicated that losses of some analytes might be due to the formation of the more volatile chlorides. Occlusion of the analyte in the matrix has also been proposed as a mechanism of interference.l3Sl4 The literature on cationic and anionic interferences occurring in electrothermal atomic-absorption spectro- photometry has been reviewed.15 Most of the work on interferences has involved the determination of a particular element in a particular matrix and the results are usually reported as an enhancement or depression of the signal when one or two additions of the938 CARRONDO et al.INFLUENCE OF CONDITI0N'"G AGENTS ON THE Analyst, Vol. 104 interferent are added to the analyte. This approach may lead to erroneous conclusions when attempting to predict interferences on different complex matrices with various inter- ferents and at different interferent to analyte ratios. Under these conditions interferences may either compensate or enhance each other or synergistic or antagonistic effects may occur.lG The work reported here was undertaken to determine if some of the common conditioning agents would cause interferences in the rapid electrothermal atomic-absorption spectro- photometric analysisg~10 of cadmium, chromium, copper , nickel, lead and zinc in sewage sludge.As interferences normally increase with an increase in the interferent to analyte ratio, the conditioners were used at two rates: the maximum addition normally found in sewage treatment practice and an additional rate 50% higher than that, to cover any abnormal situations (see Table I). Analysis of the unconditioned sludge by flame atomic- absorption spectrophotometry after digestion by a nitric acid - hydrogen peroxide procedure was also undertaken. TABLE I RATES AND METHODS FOR ADDITION OF INORGANIC CONDITIONERS Range of dosing rate in normal practice Inorganic conditioner (yo m/m dry solids) Method of addition Lime , ... . . . . 10-25 Added as 10% mlm slurry + copperas . . . . . . 5-20 Always added before lime as a 25% mlm solution Lime . . . . . . . . 10-15 Added as 10% mlm slurry + iron(II1) chloride . . . . 1-5 A.dded before lime as a 40% mlrn solution Aluminium chlorohydrate . . 0.5-3.0 Dose expressed as yo mlm aluminium oxide on dry solids Rates used in these experiments (% mlm dry solids) - Rate 1 Rate 2 25 37.5 20 30 15 22.5 5 7.5 3 4.5 Experimental Apparatus Flame atomic-absorption analysis was undertaken using a Perkin-Elmer, Model 603, atomic-absorption spectrophotometer equipped with deuterium background correction. The conditions for analysis were those recommended by the instrument manufacturer, with the exception of those for chromium for which slit 3 (0.2 nm) was used to reduce spectral inter- f erences .17 Electrothermal analyses were undertaken using the same spectrophotometer in conjunc- tion with a Perkin-Elmer HGA 76 heated graphite atomiser.The conditions and working ranges for electrothermal atomic-absorption analysis are presented in Table 11. Aliquots of 20 or 50 p1 were injected into the electrothermal atomiser with an Eppendorf micropipette (Anderman & Co. Ltd., Surrey). Analysis was performed by direct comparison with standards in 1% V/V nitric acid. Nitric Acid - Hydrogen Peroxide Digestionls the volume was reduced to 10ml. peroxide were added repeatedly until the digestate turned a pale straw colour. A sample of 20 ml was digested on a thermostatic hot-plate with 30 ml of nitric acid until After cooling, 2ml each of nitric acid and hydrogen Homogenisation Approximately 250 ml of the sample, previously diluted 50-fold and acidified to 1% V/V with nitric acid, was homogenised in a 2-1 tall-form borosilicate beaker with an Ultra Turrax T45N homogeniser (Scientific Instrument Co.Ltd., London) for 5 min at 8000 rev min-l. To avoid contamination of the sample by chromium and nickel the original stainless-steelOctober, 1979 DETERMINATION OF METALLIC CONTENT OF SEWAGE SLUDGE BY AAS TABLE I1 OPERATING CONDITIONS FOR THE ATOMIC-ABSORPTION SPECTROPHOTOMETER AND ELECTROTHERMAL ATOMISER Conditions Wavelength/nm . . . . Slit width/nm . . . . . . Drying time/s . . . . .. Ashing time/s . . .. .. Atomisation time/s . . . . 20 $/mg 1-l . . . . . . 50 pl/mg 1-1 . . . .. . Drying temperature/"C . . Ashing temperature/"C . . Atomisation temperature/"C . . Working range using Working range using * Gas stop used. Metal Cd 228.8 0.7 100 30 250 40 2 100 5 0.005-0.03 0.002-0.02 Cr 357.9 0.2 100 30 1100 30 2 770 5 0.02-0.2 0.01-0.1 c u 324.8 0.7 100 30 700 30 2 770 4 0.05-0.4 0.02-0.2 Ni 232.0 0.2 100 30 800 30 2 770 5 0.1-1 .o 0.05-0.4 Pb 283.3 0.7 100 30 350 40 2 300 5 0.025-0.4 0.01-0.2 939 - Zn 307.6 0.7 100 30 450 35 2 500 5* 0.2-2.0 0.05-1.0 shaft, stator and rotor of the homogeniser were replaced with a replicate made of titanium, The suitability of this shaft for the homogenisation of samples to be analysed for cadmium, chromium, copper, nickel, lead and zinc has been reported previo~s1y.l~ Addition of Inorganic Conditioners Three commonly used inorganic conditioners were studied: lime and copperas, lime and iron(II1) chloride and aluminium chlorohydrate.The range of rates of addition of inorganic conditioners in normal practice, and also the dosing rates used in the experiment, are shown in Table I. Each conditioner was dosed at two rates, the maximum of the range used in normal practice and a rate 50% higher than this. Conditioning agents are added to sludge on the basis of its total solids content; the rate of addition is therefore expressed as a percentage by mass of the dry solids content of the sludge. Addition of Polyelectrolytes Three cationic polyelectrolytes were studied : Aquafloc 4051 (Dearborn Chemicals Ltd., Lancashire), Flocbel 170 (Float-Ore Ltd., Middlesex) and Zetag 94 (Allied Colloids Manu- facturing Co.Ltd., Yorkshire). Aquafloc 4051 and Flocbel 170 are solid polyelectrolytes and are 100~o active. Zetag 94 is a liquid polyelectrolyte containing 15% rn/m of active material. This again was added as a 0.1% m/m solution. The normal upper limit of polyelectrolyte addition in sewage treatment is 1% m/m of the dry solids content of the sludge. Each polyelectrolyte was added at two dose rates, 1.0 and 1.5% mlwz dry solids. They were applied as dilute 0.1% m/m solutions. Incorporation of Conditioners required amount of inorganic conditioner or polyelectrolyte solution. then added to make up the volume to 1250 ml and the solution was mixed by stirring. solution was then homogenised as described for unconditioned sludges. To 25 ml of mixed primary sludge in a 2-1 tall-form borosilicate beaker was added the Distilled water was The Reagents BDH Aristar-grade reagents were used for all analyses. Nitric acid, 70%, sp.gr. 1.42. Hydrogen peroxide, 100 volume. Standards were prepared by serial dilution of 1000 mg 1-1 metal stock solutions.940 CARRONDO et d. INFLUENCE OF CONDITIONING AGENTS ON THE Analyst, Vol. 104 Results and Discussion A sample of mixed primary sludge (total solids content 6.26%) was collected in a polythene container and acidified to 1% V/V nitric acid. For each of the pre-treatments (digestion and homogenisation of unconditioned and conditioned sludge) five replicates and two blanks were taken. The digested samples were analysed by flame atomic-absorption spectro- photometry and the homogenised samples by electrothermal atomic-absorption spectro- photometry. The results were treated statistically and the rnean values, within-group relative standard deviations (RSD) and the results of an analysis of variance by the F-test20 are presented.Tukey’s procedure20 was used to identify which means were different at the 0.05 probability level. The values obtained with the inorganically conditioned sludge are presented in Table 111 and there are highly significant effects for all metals, with the exception of zinc, for which no significant differences were found. The results obtained for cadmium, chromium and nickel in the presence of aluminium chlorohydrate, at both rates, are statistically different from the others; in all instances lower results were obtained. For these same elements with other conditioners, at both rates, no significant differences were found.The results obtained for copper with aluminium chlorohydrate were also significantly different from the others, but the higher rate of addition enhanced the results and sharply increased their scatter, as indicated by the RSI). Higher results were also obtained with the higher dose (rate 2) of lime and copperas (CaO + FeSO,). The results obtained in the presence of the other conditioners agree well with the values used for comparison, i.e., hydrogen peroxide digested samples analysed by flame atomic-absorption spectrophoto- metry and the homogenised samples analysed by electrothermal atomic-absorption spectro- photometry. Finally, for the determination of lead the high dosages (rate 2) of lime and copperas and lime and iron(II1) chloride yielded significantly higher results with a larger RSD.The presence of these conditioners at the maximum rate normally used in sewage practice (rate 1) or the presence of aluminium chlorohydrate did not seem to affect the results. As the presence of aluminium chlorohydrate, at both rates of addition, significantly affected the results obtained for cadmium, chromium, copper and nickel, the influence of this com- pound on the analysis of aqueous acidic standards was examined. If a similar effect for the standards as in the sludges had been observed, then the reproduction in the standards of the aluminium chlorohydrate content of the sludge would have allowed the use of a calibration method foI the analysis.However, the test showed that no effects occurred, i.e., the standaids yielded the same absorbance in the presence and absence of aluminium chlorohydrate. When polyelectrolyte conditioners were added no significant differences between the treated and untreated sludge were found for cadmium, copper, nickel and zinc (Table IV) but both rates of addition of Aquafloc 4051 caused an increase in the scatter of results for cadmium and lead. Zetag 44 appeared to affect the determination of lead (and to a lesser extent cadmium) in a similar mannner when added at rate 2. The determination of chromium and zinc with Flocbel 170 at the high dosage rate yielded results marginally lower than those obtained in some other instances.For both these elements, the results may be grouped into “high results” and “low results.” For chromium, the “high results” contain all but those obtained for Flocbel 170 dosed at rate 2 and the “low results” contain all but those for the digestion procedure. Thus, the evidence for rejecting the method for the determination of chromium and zinc, at least in the presence of polyelectrolytes equivalent to the maximum additions used in practice, is very weak. The addition of inorganic conditioners increased the amouhts of some ions, calcium, iron, aluminium, sulphate and chloride, for example, which are known to cause interferences in the assays of some or all of the six metals determined.13-15 At the maximum rates of addition of the conditioners used in sewage practice, ratios of analyte to interfering substances of 1 : 1000 or even greater will occur.I t would therefore be expected that interferences would occur in the presence of the conditioners. The results show that these conditioners did not interfere at either rate for zinc and that, apart from aluminium chlorohydrate, there were no interferences in the determination of cadmium, chromium or nickel. For copper and lead at some of the higher rates of addition, interferences did occur.October, 1979 DETERMINATION OF METALLIC CONTENT OF SEWAGE SLUDGE BY AAS TABLE I11 EFFECT OF DIFFERENT CONDITIONERS ON THE DETERMINATION OF CADMIUM, CHROMIUM, COPPER, NICKEL, LEAD AND ZINC IN SEWAGE SLUDGE BY ELECTROTHERMAL ATOMIC-ABSORPTION SPECTROPHOTOMETRY Metal Cd . . Cr . . c u .. Ni . . Pb . . Zn . . Pre-treatment * . . H202-HN03 Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) AlC1, (1) AlC1, (2) Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) AlC1, (1) AlCl, (2) Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) AlCl, (1) AlC1, (2) .. H20,-HN03 .. H202-HNO, .. H20, - HNO, Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeCl, (2) AlC1, (1) AlC1, (2) .. H202-HN03 Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeCl, (1) CaO + FeCl, (2) AlCl, (1) NC1, (2) .. H20a-HN03 Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) MC1, (1) NC1, (2) F-test level of Mean concentration?/ significance mg 1-1 KSD,S % 0.01 0 . 2 6 ~ 4.3 0 . 2 5 ~ 5.5 0 . 2 6 ~ 10.2 0 .2 4 ~ 3.8 0 . 2 4 ~ 2.7 0 . 2 4 ~ 4.9 0.20b 6.5 0.18b 5.5 0.01 3 . 4 ~ 3.2ac 3.2ac 3 . 4 ~ 3.2ac 3 . 3 ~ 2.9bc 2.6b 0.01 1 6 . 9 ~ 1 7 . 2 ~ 1 7 . 5 ~ 2 1 .'Ob 1 7 . 7 ~ ~ 1 6 . 7 ~ 12.4d 20.3bc 0.01 6 . 5 ~ 6 . 2 ~ 6 . 4 ~ 6 . 5 ~ 6 . 2 ~ 5.7ab 5.lb 3 . 1 ~ 6.3 5.9 4.7 4.6 3.5 5.1 4.0 2.4 3.8 5.0 5.6 9.7 5.9 7.6 5.6 20.7 7.8 6.7 9.5 11.0 5.3 9.4 15.6 16.7 0.01 40a 7.8 43a 5.9 46a 8.4 57b 9.4 41a 5.9 58b 9.1 42a 7.3 42a 6.7 N.S.3 35a 4.8 34a 5.1 31a 5.2 33a 5.8 33a 5.7 35a 6.1 32a 6.1 34a 7.2 941 * H202 - HNO, = hydrogen peroxide - nitric acid digestion; Homog. = pre- treatment by homogenisation; CaO lime; FeSO, = copperas; FeC1, = iron(II1) chloride; AlCl, = aluminium chlorohydrate; (1) = rate 1 (Table I ) ; (2) = rate 2 (Table I). t Means not followed by a common letter are statistically different at the 0.05 significance level.RSD = relative standard deviation. 3 N.S. = not significant at the 0.05 significance level.942 Analyst, Vol. 104 The polyelectrolytes, which are organic in nature and added in much smaller amounts, would be expected to have much less effect, if any. It can be seen from the data presented that there was no effect on the results for cadmium, copper, nickel or lead and there was CARRONDO et al. : INFLUENCE OF CONDITIONING AGENTS ON THE TABLE IV EFFECT OF DIFFERENT POLYELECTROLYTES ON THE DETERMINATION OF CADMIUM, CHROMIUM, COPPER, NICKEL, LEAD AND ZINC I N SEWAGE SLUDGE BY ELECTROTHERMAL ATOMIC-ABSORPTION SPECTROPHOTOMETRY F-test level of Mean concentrationt / Metal Pre-treatment* significance mg 1-l RSD,$ % Cd Cr c u Ni Pb Zn ... . .. .. . . .. .. .. HZOZ - HNO, Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Hz02 - HNO, HZOZ - HNO, HSO, - HNO, H20z - HNO, HZOZ - HNO, N.S.$ 0.26a 0.25a 0.26a 0.26a 0.24a 0.24a 0.25a 0.24a 0.05 N.S. N.S. N.S. 0.05 3.4a 3.2ab 3.lab 3.0b 3.lab 3.2ab 3.4a 3.3ab 16.9a 17.2a 17.4a 16.7a 17.6a 16.4a 16.Oa 16.0a 6.5a 6.2a 6.0a 6.3a 6.5a 6.4a 6.3a 6.3a 40a 43a 41a 42a 45a 46a 46a 47a 35a 34ab 31ab 30b 32ab 31ab 32ab 33ab 4.3 5.5 6.6 7.9 12.8 10.3 6.9 8.6 6.3 5.9 8.2 6.6 4.1 7.7 8.6 6.3 3.8 5.0 8.6 5.3 4.3 3.3 5.9 6.0 7.8 6.7 5.9 5.1 4.5 6.3 2.3 5.7 7.8 5.9 7.7 7.2 11.0 10.2 7.9 12.5 4.8 5.1 6.5 6.3 4.6 4.5 5.9 4.9 * Flocbel = Flocbel 170; Aquafl = Aquafloc 4051; Zetag = Zetag 94.Other footnotes as in Table 111.October, 1979 DETERMINATION OF METALLIC CONTENT OF SEWAGE SLUDGE BY AAS 943 only weak evidence for an effect on the determination of chromium and zinc. Although only three polyelectrolytes were tested it is thought that similar results would have been obtained with others. Conclusion The rapid electrothermal atomic-absorption spectrophotometric procedure described here can be used for the determination of cadmium, chromium, copper, nickel, lead and zinc in sludges that have been conditioned with polyelectrolytes.When high rates of conditioners are used the method is not suitable for these analytes. However, the method could be used at rates of addition consistent with normal sludge treatment practice for all of the con- ditioners tested except aluminium chloroh ydrat e. We acknowledge financial support for this work from the Department of the Environment and the Department’s approval to publish these results. One of us (M. J. T. Carrondo) is also grateful to the Instituto Nacional De Investiga@o Cientifica, Lisboa, Portugal, for the award of a postgraduate scholarship. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.20. References Neufeld, R. O., Gutierrez, J., and Novak, R. A., J . Wat. Pollut. Control Fed., 1977, 49, 489. Ministry of Housing and Local Government, “Taken for Granted,” Report of the Working Party on Vesilind, P. A., “Treatment and Disposal of Sewage Sludges,” Ann Arbor Science Publishers, Ann Clark, E. I., and Fisk, B., Waf. Pollut. Control, 1978, 77, 509. Gale, R. S., and Baskerville, R. C., Filtr. Sep., 1970, 7, 37. Dulka, J. J., and Risby, T. H., Analyt. Chem., 1976, 48, 640A. Page, A. L., “Fate and Effects of Trace Elements in Sewage Sludge when Applied t o Agricultural Department of the Environment, “Report of the Working Party on the Disposal of Sewage Sludge Carrondo, M. J. T., Perry, R., and Lester, J . N., Analytica Chim. Acta, 1979, 106, 309. Stoveland, S., Perry, R., and Lester, J . N., Sci. Total Environ., in the press. Baudin, G., Chaput, M., and Feve, L., Spectrochim. Acta, 1971, 26B, 425. Sturgeon, R. E., Chakrabarti, C. L., and Langford, C. H., Analyt. Chem., 1976, 48, 1792. Smeyers-Verbeke, J., Michotte, Y., Van den Winkel, P., and Massart, D. L., Analyt. Chem., 1976, Cruz, R. B., and Van Loon, J . C., Analytica Chim. Acta, 1974, 72, 231. Fuller, C. W., “Electrothermal Atomization for Atomic Absorption Spectrophotometry,” Analytical Woodis, T. C., Hunter, G. B., and Johnson, F. J., Analytica Chim. Acta, 1977, 90, 127. Kirkbright, G. F., and Sargent, M., “Atomic Absorption and Fluorescence Spectroscopy,” Academic Geyer, D., Martin, P., and Adrian, P., Korrespondenz Abwasser, 1975, 22, 369. Stoveland, S., Astruc, M., Perry, R., and Lester, J . N., Sci. Total Environ., 1978, 9, 263. Sewage Disposal, HM Stationery Office, London, 1970. Arbor, Mich., 1974. Land-A Literature Review Study,” U.S. EPA-670/2-74-005, 1974. t o Land,” HM Stationery Office, London, 1977. 48, 125. Sciences Monograph No. 4, Chemical Society, London, 1977. Press, London, 1974. Bowker, A. H., and Lieberman, G. J., ‘ N.J., 1972. ‘Engineering Statistics,” Prentice-Hall, Englewood Cliffs, Received April 3rd, 1979 Accepted May 9th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400937

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

944 14nalyst, October, 1979, Vol. 104, $9. 944-951 Determination of Antimony and Other Trace Elements in irons and Steels by Atomic-absorption Spectrophotometry with Introduction of Solid Samples into an Induction Furnace* A. M. Aziz-Alrahman and J. B. Headridge Depavtment of Chernistvy, The Univevsity, SheBeld, S3 7HF Atomic-absorption spectrophotometry with an induction furnace has been used for the determination of 0.5-350 p g g-l of antimony in 1-20-mg samples of irons and steels dropped into the furnace. Calibration graphs of peak absorbance vevsus the mass of antimony have been constructed by using standard steels. Information is presented on the accuracy and precision of the method for 13 irons and steels. The limit of detection for antimony is Calibration graphs have also been obtained for indium in a nickel-base alloy and thallium, tin, selenium, tellurium, zinc and cadmium in steels in order to establish the conditions that will be necessary to determine these elements using a procedure similar to that employed for antimony.Factors affecting the volatility of trace elements with boiling-points below 2300 “C are discussed. 0.12 rf.g g-:. Keywovds : A ntipnony detevmination ; trace-element detevnzinations ; ivon and steel analysis ; atomic-absovption spectvophotometvy ; induction furnace The presence of antimony is known to be detrimental to certain mechanical properties of low-alloy steels. As little as 10 pg g-l produces 350 “C embrittlement and temper brittle- ness.l Low concentrations of antimony also adversely affect the hot workability of stainless steels2 There is a continuing need to determine antimony in steels within the range 1-500 pg 8-l.Many methods are available for the determination of antimony above 100 pg g-l but few methods are suitable below this level and only the most sensitive can be employed to determine concentrations of antimony below 10 pg g-l. Information about some of these methods is given in Table I. TABLE I METHODS FOR THE DETERMINATION OF ANTIMONY IN STEEL Method Molecular-absorption spectrophotometry of rhodamine B complex D.c. arc-emission spectrography after pre-concentration by precipita- . . . . . . .. . . after solvent extraction . . . . tion of Sb,S, on CuS . . . . . . . . . . . . . . Hollow-cathode emission spectrometry on solid samples .. . . Flame atomic-absorption spectrophotometry . . . . .. . . pentan-2-one after solvent extraction . . . . . . . . . . tion after dissolution of the alloy . . . . . . . . . . Flame atomic-absorption spectrophotometry of Sb(II1) in 4-methyl- Flame atomic-absorption spectrophotometry after generation of SbH, Atomic-absorption spectrophotometry with carbon-furnace atomisa- Concentration range of antimony presentlpg g-l 10-500 5-500 -0.540 20-350 2-350 1-350 Reference 3 4 5 6 7 8 9 10 Methods involving solvent extraction or other pre-concentration procedures are relatively slow and the most attractive methods so far reported are the atomic-absorption methods involving hydride generation,8 atomisation in resistively heated carbon furnacesgtio and hollow-cathode emission ~pectrometry.~ Headridge and co-workers have already reported * Presented a t the Meeting on “Research and Development Topics in Analytical Chemistry,” Heriot- Watt University, Edinburgh, July, 1979.AZIZ-ALRAHMAN AND HEADRIDGE 945 on the determination of bisrnuth,ll silver12 and lead13 in steels using atomic-absorption spectrophotometry with the introduction of solid samples into an induction furnace.In this paper, results for a similar study on antimony are reported along with data on other volatile elements that might at some time have to be determined in steels containing very low concentrations of these elements. Experimental Materials of Standards, USA, and alloys from the Institutet for Metallforskning, Sweden. that no more than three pieces need to be added to the furnace core at the same time.Standard irons and steels. Samplesfor analysis. British Chemical Standards, alloys from the National Bureau These should preferably be millings or turnings of irons or steels so Apparatus and Method for Determining Absorbances for a Series of Solid Samples The apparatus and the method were identical with those described previouslyll except that the graphite core and side-arms were made from Ultra Superior Purity Grade graphite (Ultra Carbon) and baked for 6 h under vacuum at about 1900 "C before use. Information on the hollow-cathode lamps used in this work is given in Table 11. TABLE I1 DATA ON HOLLOW-CATHODE LAMPS Element Manufacturer Line employed/nm Lamp current used/mA Antimony Indium . . Thallium Tin .. Selenium Tellurium Zinc . . Cadmium .. . . Pye Unicam . . . . Pye Unicam .. . . Perkin-Elmer .. . . Activion . . . . Activion . . .. EM1 .. . . Activion .. . . Activion 231.1 303.9 276.8 286.3 196.0 214.3 213.9 228.8 10 5 20 20 6 5 8 13 In all instances, except for selenium, a slit width of 0.2 nm was used on a Perkin-Elmer 300s atomic-absorption spectrophotometer ; the slit width used for selenium was 0.7 nm. The experimental conditions for the determination of antimony are shown in Table 111. Calibration Graphs A ntimmy For the determination of antimony in irons and steels containing less than 150 pg g1 of antimony, calibration graphs of peak absorbance versus the amount of antimony are obtained by dropping increasing amounts of 0.2% carbon steel JK2C, which contains 29 pg g1 of antimony, into the graphite core under conditions capable of producing absorbances of up to 1.0 (see Table 111).For steels containing 150-350 pg g-l of antimony, a calibration graph is prepared in a similar way using mild steel BCS 328 containing 260 pg g-l of antimony. Other elements Although series of irons and steels were not analysed for these elements, the conditions necessary for obtaining a calibration graph for each element in steel were established using one standard alloy in each instance. These conditions are given in Table IV.946 AZIZ-ALRAHMAN AND HEADRIDGE : DETERMINATION OF ANTIMONY Analyst, Vol. 104 TABLE TI1 EXPERIMENTAL CONDITIONS FOR THE DETERMINATION OF ANTIMONY Concentration Mass range of Core temperature/ Scale rangelpg g-' sample/mg "C Damping* expansion <5 10-20 2 620 1 x 5 5-150 1-10 2 550-2 640 1 x l 150-350 1-6 1810 4 x l * Damping positions 1 and 4 are for time constants of 0.2 and 10 s, respectively.Determination of Antimony in Irons and Steels Irons and steels with a silicon to antimony ratio >5: 1 When a series of irons and steels is to be analysed, suitable masses are dropped into the graphite core over a period of 2-3 h and, during the same run, various masses of JK2C or BCS 328 are also added, generally at the beginning of the run, for the purpose of constructing a calibration graph. During a run the temperature of the core should not alter by more than &lo "C. When the run is completed the calibration graph is drawn and the mass of antimony in each sample is obtained from the graph.The concentrations of antimony in the samples are then calculated. Irons and steels with a silicon to antimony ratio <5 ; 1 The procedure is similar to that above except that 5 g of silicon powder (BDH Chemicals Ltd.) are added to the cold core and well tamped down with a glass rod before the core is heated at approximately 1900 "C, as read on the optical pyrometer, for a sufficient period of time to remove all antimony from the silicon (12 h in our case); samples may then be added to the core. TABLE 1IV EXPERIMENTAL CONDITIONS FOR OBTAINING CALIBRATION GRAPHS FOR CERTAIN MORE VOLATILE ELEMENTS I N STEEL Damping position 1 (time constant 0.2 s) was used in all instances except zinc, for which damping position 5 (time constant 30 s) was employed.Element Indium . . .. Thallium . . . . Tin . . . . . . Selenium . . .. Tellurium . . . . Zinc . . .. . . Cadmium . . .. Standard alloy R3387* BCS 312 BCS 323 SRM 361 SRM 361 JKlC J K2C Content/ -10 240 CLg g-' 0.37 40: 2.0 -0.03 6: Mass range Core temperature/ of sample/mg "C 9-15 2 370 1-13 2 330 2-20 2 450 3-30 2 430 1-7 2 350 1-7 1570 8-35 2 350 Scale expansion x l x 5 x l x l x 1 x l x6 * A nickel-base alloy. t See reference 14. 1 Value for SRM 1261. Results The calibration graphs for the range 0-360ng of antimony obtained with steel JKZC passed through the origin and were slightly curved at the upper end. The mass for 1% absorption was 1.4 ng, this value being obtained from a typical calibration graph for antimony at 2630 "C. The calibration graph for the range 0-1.6 pg of antimony obtained with steel BCS 328 also passed through the origin but was appreciably curved because of the damping applied to the amplifier signals.This damping was necessary in order to keep the apparent absorbance readings below 1 .O. Samples of a series of irons and steels were dropped into the furnace in order to obtain their antimony contents using the conditions outlined in Table I11 and in the section entitled Determination of Antimony in Irons and Steels. The results are shown in Table V.October, 1979 AND OTHER TRACE ELEMENTS IN IRONS AND STEELS BY AAS TABLE V RESULTS FOR THE DETERMINATION OF ANTIMONY IN IRONS AND STEELS Calibration graph prepared using JK2C (29 pg g-' of antimony), except for results marked $, when BCS 328 (260 pg g-l of antimony) was used.Alloy BCS 260/4 336 337 456 457 459 460 SRM 361 362 363 364 365 JKlC Antimony reported/ clg g-' < 10 28, 29* 22t 110 290 70 20 42 130 20§ 3407 <0.5)1 2.111; 1, 2* Antimony Number of samples foundlpg g-1 analysed 2.2 6 31 7 29 8 116 7 270: 8 62 19 19 8 Relative standard deviation, yo 15 12 7 8 18 8 16 41 7 14 135 7 8 16 8 8 0.65 11 9 337: 7 7 1.3 12 23 947 * The figures shown were obtained by Frech.9 t Result obtained by Burke.l5 $ Si was added to the core prior to addition of the samples because Si: Sb < 5 : 1. 5 Certificate value for SRM 1263; SRM 363 is the same material in the form of chips. 7 Certificate value for SRM 1264; SRM 364 is the same material in the form of chips. (1 Single result quoted on the certificate.All other figures in the second column are certificate values. With the other elements, straight-line calibration graphs passing through the origin were obtained for indium, thallium, selenium, tellurium and cadmium. The calibration graph for tin passed through the origin but was curved towards the mass axis, particularly for masses of tin in excess of 2.5 pg. The calibration graph for zinc was extensively curved because of the high degree of damping necessary to keep the apparent absorbances below 1.0. From these graphs the mass for 1% absorption was obtained in each instance. These data are shown in Table VI. TABLE VI MASSES REQUIRED FOR 1 yo ABSORPTION FOR VARIOUS ELEMENTS Element . . . . .. . . In T1 Sn Se Te Zn Cd Mass for 1% absorptionlng .. -0.7 0.14 46 12 0.18 0.043 -0.11 Discussion Steel JK2C has not been fully standardised for the amount of antimony it contains, but contents of 32 and 27 pg g1 are reported on the certificate and Frech9 has found antimony concentrations of 27 and 29pgg-l. The average value of 29 pgg-l was selected as the antimony content of this steel. Steel BCS 328 has been fully standardised and contains 260 pg g1 of antimony. The accuracy of the results for antimony shown in Table V is considered to be good, the agreement between our results and those for standardised steels being very satisfactory. The precision of the method is acceptable, although not so good as that for a similar method for lead reported previ0us1y.l~ This is probably because, in this study, the furnace was usually operated at a temperature in the range 2550-2640 "C, which is about the maximum temperature that can be achieved with the equipment already described.ll Corrosion of the graphite surfaces of the furnace, from trace amounts of reactive gases left in the argon stream and from air not completely removed from the graphite during the heating-up procedure, is more noticeable at higher temperatures and this leads to more scatter in the results.948 AZIZ-ALRAHMAN AND HEADRIDGE : DETERMINATION OF ANTIMONY A%aL'yst, Vd.104 Lundberg and lirech16 have reported that scatter due to inhomogeneity should not be the major contribution to the relative standard deviation when milligram masses of steel are analysed for antimony by adding turnings to a furnace. For the first time since using this design of induction furnace, a matrix effect has been observed for the determination of an element in steel.Steels BCS 457 and SRM 364 pro- duced consistently high results for antimony when added to the normal graphite core. Both alloys were analysed on four different occasions using five or more samples each time and the average results were 449 pg g-l for BCS 457 and 521 pg 8-l for SRM 364. A close examination of the chemical composition of all the steels, which had been added to the core, showed that BCS 457 and SRM 364 were exceptional only in that the ratios of the concentra- tions of silicon to antimony were low at 1.6: 1 and 2.6: 1 , respectively, while in all other instances, the ratios were in excess of 6: 1. I t was felt that the presence of silicon must have reduced the rate of diffusion of antimony in molten steels, when the ratio of silicon t o antimony was in excess of 5: 1, such that the rate of release of antimony into the gas phase was diminished.I t was decided to add silicon powder to the graphite core in order toensure that the ratio of silicon to antimony in the molten metal was in excess of 5: 1 and when this was done, acceptable results were obtained for BCS 457 and SRM 364 as reported in Table V. Incidentally, steels with a silicon to antimony ratio of less than 5: 1 are unusual so the need to add silicon to a core will occur very occasionally. The atomic-absorption spectrophotometer was operated without a background corrector. To check that there was negligible molecular absorption and negligible light scattering from relatively high concentrations of the elements present in steels that are volatile at high temperatures, the instrument was peaked up at 231.1 nm using the antimony hollow-cathode lamp, which was then replaced with a deuterium lamp, and samples of Specpure manganese (Johnson Matthey & Co. Ltd.), JK2C, BCS 457 and SRM 364 were added to the furnace core at 2540 "C.In no instance was there an absorbance reading in excess of 0.01 unit. Manganese is present in almost all steels and boils at 1962 "C. The limit of detection of the method is 0.12 pg g-l of antimony and was taken as twice the standard deviation for sample SRM 365. This limit of detection is lower than any other reported in Table I. Over a period of 4 years a considerable amount of data has been accumulated on the determination of trace elements in steels using atomic-absorption spectrophotometry with the introduction of solid samples into a constant-temperature induction furnace and an assessment of the real value of the method can now be given.Detailed results have already been reported for bisrnuth,ll silver12 and lead13 in irons and steels, and have just been given for antimony in irons and steels. The method can be applied to the determination of many trace elements that are sufficiently volatile at 2600 "C or lower temperatures. These elements include indium, thallium, tin, lead, antimony, bismuth, selenium, tellurium, silver, zinc and cadmium, for which calibration graphs have been obtained. In theory the method should also be suitable for the determination of trace concentrations of magnesium, calcium, aluminium, gallium, manganese and copper in steels.However, the method is so sensitive that appreciable background absorption from magnesium, aluminium, manganese and copper continuously released from the graphite core has been obtained by us even using vacuum-degassed USP grade graphite. Suitable calibration graphs for very low masses of these elements have not yet been obtained. Gallium should present no problems with respect to background absorption. The masses required for 1% absorption for various elements have already been reported in this paper (see Results). In Table VI the sensitivity for zinc is for a damped signal and the mass for 1% absorption in the absence of damping will be considerably less than 43 pg.The results for indium and cadmium may well be significantly in error. The indium content of the nickel-base alloy R3387 used to construct the calibration graph is nominally 10 pg g-l but could be appreciably lower. The cadmium content of JK2C is given as 0.03 pg 8-l on the certificate, from a single determination, but this figure seems to be too high because the mass for 1% absorption would be expected to be similar to that for zinc. The reliable masses for 1% absorption for the elements reported here along with those for bismuth,ll silver12 and lead,13 reported earlier, have been converted into concentrations for 10-mg samples and are shown in Table VII. Also shown in this table are the concentrations of these elements in solution (pg ml-l) producing 1% absorption in an air - acetylene flame,October, 1979 AND OTHER TRACE ELEMENTS IN IRONS AND STEELS BY AAS 949 using the same resonance lines for the elements as for the furnace work,17 and these con- centrations converted into pg g-l for steels assuming that 2 g of steel are dissolved in 100 ml of solution in each instance.At the furnace temperatures employed, all these elements are monatomic in the vapour phase.18 In the air - acetylene flame with a temperature of about 2300 O C , all these elements with the exception of tin should exist mainly in the monatomic form. Tin will be present both as the monatomic form and as the monoxide.19 It will be noticed that the CIA values for thallium, lead, bismuth, tellurium and silver are all between 1300 and 1800 while those for tin, antimony and selenium are much lower.Admittedly the sensitivities reported in column A were not all obtained at the temperature of the air - acetylene flame; the sensitivity increases with increasing temperature. For example, 1 pg of lead from BCS 327 gave absorbances of 0.15 and 0.27 at 2010 "C and 2150 "C, respcctively,13 and 20 ng of tellurium from SRM 361 gave an absorbance of 0.29 at 2190 "C and of 0.59 at 2350 "C. However, if all the results in column A had been obtained at the temperature of the air - acetylene flame, then the C / A values for thallium, lead, bismuth, tellurium and silver would have been in excess of 1000 and those for tin, antimony and selenium would have been lower than those reported in Table VII.The results shown -in Table VII are very interesting. TABLE VII DATA ON CONCENTRATIONS FOR 1% ABSORPTION OBTAINED BY ADDING SOLID SAMPLES TO AN INDUCTION FURNACE AND FROM NEBULISING 2% m/V SOLUTIONS OF STEELS INTO AN AIR - ACETYLENE FLAME Concentration for 1% absorption Element Thallium . . . . Tin . . .. . . Lead . . . . . . Antimony . . . . Bismuth . . . . Selenium . . . . Tellurium . . .. Silver .. .. Furnace Resonance temperature/ line/nm "C 276.8 2 330 286.3 2 450 283.3 2 000 231.1 2 630 306.8 2 070 196.0 2 430 214.3 2 350 328.1 2 270 r A * / Pg g-' 0.014 4.6 0.017 0.14 0.015 1.2 0.018 0.001 8 Btl c:/ pgm1-l CLgg-l 0.5 25 3.5 175 0.5 25 1.1 55 0.5 25 0.5 25 0.5 25 0.06 3 CIA 1786 38 1471 393 1667 21 1389 1667 * Using induction furnace. t Following dissolution of steel in acid and atomisation in an air - acetylene flame, assuming $ After converting pg ml-l into pg g-l of steel.that the sensitivities are the same in the presence and absence of iron. These CIA values of approximately 1500 in Table VII are found for the elements that quickly diffuse out of the molten globule of steel into the gas phase within our induction furnace. These are also the elements that should be readily removed from steel on vacuum melting. Indeed, Chernov and Ageev20 reported that lead and bismuth were readily removed from iron on vacuum induction melting, tin and antimony were removed more slowly and arsenic was not removed a t all. The boiling-points of the eight elements under discussion are shown in Table VIII.On boiling-point alone, tin should be the most difficult element to remove by vacuum melting or by volatilisation in our furnace but, intuitively, one would expect these elements, which show some non-metallic properties, to be held back in molten steel because of bonding between the elements and iron atoms. These elements include tin, antimony and selenium and, indeed for them, CIA values much less than 1000 were obtained. If tin were entirely monatomic in the air - acetylene flame, the sensitivity would be less than 175 pg g-l and the CIA value would be less than 38 as given in Table VII. TABLE VIII BOILING-POINTS OF VARIOUS TRACE ELEMENTS FOUND IN STEEL Element . . . . . . . . T1 Sn Pb Sb Bi Se Te Ag Boiling-point/"C . . . . . . 1457 2270 1740 1750 1560 685 990 2212950 AZIZ-ALRAHMAN AND HEADRIDGE : DETERMINATION OF ANTIMONY Analyst, VoZ.104 In this study, attempts were made to obtain a calibration graph for arsenic in steel - arsenic sublimes at 613 "C at atmospheric pressure, but without success. Even at 2570 "C there was no evidence for the presence of arsenic atoms in the gas phase, the absorbance being zero in each instance when milligram masses of steels containing arsenic were added to the graphite core through which the resonance line for arsenic at 193.7 nm was being passed. The limits of detection for the elements that readily volatilise from steel have been estimated and are shown in Table IX. For a single-beam atomic-absorption spectrophotometer such as the Perkin-Elmer 300s and an air - acetylene flame, the relative standard deviation of a determination is often approximately 1 yo when solutions are nebulised to give absorbances between 0.2 and 0.8.Under these conditions the limit of detection is frequently one fifth to one tenth of the concentration for 1% absorption. With our induction furnace the precision is poorer with a relative standard deviation of 5-10y0, and hence one would reasonably expect a limit of detection about the same as the concentration for 1% absorption for a 10-mg sample. The estimated limits of detection in Table I X are taken as the concentrations for 1% absorption (pg g-I) that have been given in Table VII (column A ) . This was carried out as follows. TABLE KX ESTIMATED AND ACTUAL LIMITS OF DETECTION FOR TRACE ELEMENTS READILY VOLATILISED FROM STEELS Element Thallium ... . Tin . . .. . . Lead . . .. . . Antimony . . . . Bismuth .. . . Selenium .. . . Tellurium . . . . Silver . . .. . . Estimated limit of detection in steel/ Pg g-l 0.014 4.6 0.017 0.14 0.015 1.2 0.018 0.001 8 Actual limit of detection in steel/ Pg g-l - < 0.02* 0.12t 0.004S - - 0.005§ * See reference 13. t Reported in this paper. $ See reference 11. 3 See reference 12. Considering the assumptions that have been made in the method for estimating the limit of detection, the agreement between actual and estimated limit of detection in the instances where actual limits have been determined, is reasonable. I t can be said with certainty that the actual limits of detection for thallium and tellurium and also zinc and cadmium will be well below 0.1 pg g-l but that the limits of detection for tin and selenium at the tempera- tures given in Table VII will be not as good.Of course, the limits of detection for both tin and selenium may well be better at temperatures in excess of 2600 "C. The determination of trace elements in metals by atomic-absorption spectrophotometry with the introduction of solid samples into electrically heated graphite furnaces has been shown to be convenient and reliable on many o c c a ~ i o n s . ~ ~ - , 1 ~ ~ ~ ~ - ~ ~ There is evidence that introducing samples into a graphite furnace being maintained at constant temperature is to be preferred.21 The furnaces may be inductively or resistively heated but the cost of the latter is cheaper, unless a spare induction generator is available.It is to be hoped that instrument manufacturers will soon provide graphite atomisers to which metal samples can be conveniently added at controlled constant temperatures. Facilities for measuring both peak height and peak area will also be desirable. We are indebted to the BSC/BISPA Chemical Analysis Committee and the British Steel Corporation for a grant to buy the Perkin-Elmer 300s atomic-absorption spectrophotometer.October, 1979 AND OTHER TRACE ELEMENTS IN IRONS AND STEELS BY AAS References 95 1 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Capus, J. M., Iron Steel, Lond., 1965, 38, 594. Lynch, D. W. P., Proc. Elect. Furn. Steel Conf. A m . Inst. M i n . Metall. Petrol. Engvs, 1961, 19, 220. “Methods of Chemical Analysis of Iron and Steel,” British Steel Corporation, Sheffield, 1974, p. 19. Balfour, B. E., Jukes, D., and Thornton, K., Appl. Spectrosc., 1966, 20, 168. Thornton, K., Analyst, 1969, 94, 958. Barnett, W. B., and Kerber, J . D., Atom. Absorption Newsl., 1974, 13, 56. Headridge, J. B., and Smith, D. R., Lab. Pract., 1971, 20, 312. Fleming, H. D., and Ide, R. G., Analytica Chirn. Acta, 1976, 83, 67. Frech, W., Talanta, 1974, 21, 565. - Barnett, W. B., and McLaughlin, E. A., Analytica Chzm. Acta, 1975, 80, 285. Andrews, D. G., and Headridge, J . B., Analyst, 1977, 102, 436. Aziz-Alrahman, A, M., and Headridge, J. B., Talanta, 1978, 25, 413. Andrews, D. G., Aziz-Alrahman, A. M., and Headridge, J . B., Analyst, 1978, 103, 909. Burke, K. E., Appl. Spectrosc., 1974, 28, 234. Burke, K. E., Analyst, 1972, 97, 19. Lundberg, E., and Frech, W., Analytica Chim. Acta, 1979, 104, 67. “Analytical Methods for Atomic Absorption Spectroscopy,” Perkin-Elmer Corp., Norwalk, Conn., “Encyclopedia of Science and Technology,” McGraw-Hill, New York, 197 1. Mavrodineanu, R., Editor, “Analytical Flame Spectroscopy,” Macmillan, London, 1970, p. 27. Chernov, B. G., and Ageev, P. Ya., Stal’, 1968, 28, 1003. Lundberg, E., and Frech, W., Analytica Chim. Acta, 1970, 104, 75. Lundberg, E., and Frech, W., Analytzca Chznz. Acta, in the press. Backman, S., and Karlsson, R., Analyst, in the press. Marks, J . Y . , Welcher, G. G., and Spellman, R. J., Appl. Spectrosc., 1977, 31, 9. 1976. Received, April 20th. 1979 Accepted May 18th. 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400944

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

952 Analyst, October, 1979, VoE. 104, p p . 952-960 Quantitative Determination of Lead Dioxide Polymorphs by X-ray Powder Diffractometry P. R. Skidmore and R. R. Schwarz" Berec Group Limited, Group Technical Centre, St. Ann's Road, London, N15 3TJ The X-ray diffractometric methods used by previous workers to analyse positive plate material from lead-acid batteries €or a-, 8- and amorphous lead dioxides are critically appraised. In addition two alternative approaches, not previously applied to this problem, are assessed. *4n internal standard method is considered to be the best. Keywords : X-ray diflraction ; quantitative phase analysis ; lead dioxide ; lead sulphate ; lanthanum hydroxide. The object of this work was to develop a reliable method for the quantitative phase analysis of lead dioxide present in the positive plates of lead-acid batteries. Because the electro- chemical properties of the two crystalline forms of lead dioxide found in positive plate material, cc-lead dioxide (orthorhombic) and 13-lead dioxide (tetragonal), differ considerably, plate composition markedly affects battery performance and durabi1ity.l-4 I t has also been found that in some samples a certain amount of the lead dioxide content determined by chemical analysis cannot be detected as crystalline material by X-ray diffraction.132 Further evidence for this amorphous form has come from microscopic examination and thermal analysis of positive plates5 In addition to the lead dioxide the plates may contain lead sulphate and an organic binder.Several previous studies attempted to correlate X-ray diffraction line intensities with the amount of each phase pre~ent.ll~s~-~ Reappraisal of this work revealed its limitations and two different approaches were considered, the internal standard methodlOJ1 and the double dilution method.12 Experimental The X-ray diffraction patterns were obtained with a Philips X-ray diffractometer.Apparatus and Reagents A broad focus tube giving copper Kcc radiation was used at 60 mA and 45 kV and a nickel /?-filter was placed in the diffracted beam. Slits used were : divergence, +' ; receiving, 0.2 mm; scatter, 8'. The samples were milled in a McCrone micronising mill using alumina grinding elements. The material was packed into an aluminium holder, backed with Selotape, and pressed from the back against a glass microscope slide to give a plane surface.The linear regression lines were calculated using a least-squares program on a Texas Instruments programmable 57 calculator. Preparation of the lead dioxide standard samples The work by Bagshaw et al.13 on the preparation of the polymorphs of lead dioxide revealed that different preparation methods do not give samples with uniform X-ray diffraction properties. They differ in their degree of orientation, crystallinity, crystallite size and purity. The practical importance of this was shown by Dodson3 who found that different preparations of lead dioxide gave different X-ray diffraction calibration graphs. Moreover, even the purest lead dioxide samples are slightly non-stoicheiometric and probably contain a certain amount of amorphous lead dioxide.Wiesener and co-workers1p2 obtained values for the total crystalline lead dioxide in excess of lOOyo for many of their samples, probably because of the high amorphous content of their standard materials. Thus, it must be concluded that, despite taking every care with the preparation of the lead dioxide poly- morphs, it is possible that two sources of systematic error will remain: different degrees of preferred orientation of the standards and of the samples to be analysed and the presence of amorphous lead dioxide in the supposedly completely crystalline standards for pure a- and * Retired.SKIDMORE AND SCHWARZ 953 pure P-lead dioxide. To minimise these sources of error, the preparations chosen as standards for our work had the intensities of their diffraction lines as nearly as possible in the same ratios as the positive plate material and had narrow diffraction lines, indicating that they were highly crystalline.The two crystalline polymorphs of lead dioxide were prepared by several of the methods given by Bagshaw et aZ.13 For cc-lead dioxide these were: (i) (ii) (iii) oxidation of lead acetate by ammonium per~ulphatel~; electroformation of a hand-pasted plate in 0.7 M sodium sulphate solution with two steel counter electrodes at 0.15 A for 2 d ; oxidation of orthorhombic lead monoxide by fusion with a sodium nitrate - sodium chlorate mixture. Method (i) gave material with broadened diffraction lines, method (ii) gave an impure sample and so the product of method (iii) was chosen as the standard for the calibration lines.After difficulty was experienced in preparing reproducible samples by method (iii), a short investi- gation was undertaken. Yellow lead monoxide (orthorhombic) was reacted with fused sodium nitrate and sodium chlorate at 330 "C for 10 min and the product was washed and dried. I t contained a mixture of Pb,O, and PbO,,,, identified by their Xaray diffraction patterns.15 After repeating the fusion, which introduced no new compounds, the product was washed with 3 M nitric acid overnight. This procedure did not remove all the PbO1.55 from the fusion product, although it did appear to remove the Pb,O,. The source of the inconsistencies previously noted in a-lead dioxide preparations seemed to be this contamination with PbO,.,,.However, boiling the fusion product with 500ml of 5 M nitric acid for 15 min gave an a-lead dioxide virtually free from PbO,.,,. No P-lead dioxide was detected in this material, but prolonged boiling (for more than 1.5 h) did lead to a small amount of @-lead dioxide contamination. Oxidation of tetragonal lead monoxide by the sodium nitrate - sodium chlorate fusion gave only PbO,.,,. This sample was partly converted into a-lead dioxide by boiling with 5~ nitric acid but its diffraction pattern was less crystalline than that from yellow lead monoxide. The suspension was then heated to 60 "C and filtered. For /3-lead dioxide the preparation methods used were : (iv) reaction of red lead (Pb,O,) with nitric acid; (v) hydrolysis of lead tetraacetate; (vi) electroformation of a hand-pasted plate in 3.8 M sulphuric acid with a platinum counter electrode at 1 A for 24 h. Method (v) gave a very finely divided material that was difficult to handle and method Method (iv) was used to prepare the standard P-lead dioxide.(vi) gave an impure sample. Lanthanum hydroxide internal standard The choice of an internal standard material was difficult because many otherwise desirable compounds exhibit diffraction lines that overlap those of a- and P-lead dioxide. In particular, corundum (A1203), which was suggested as suitable by Chungll and which has been used by the Joint Committee on Powder Diffraction Standards (J.C.P.D.S.) in their recent semi-quantitative work, cannot be used. Material marketed as lanthanum oxide (La,O,) was chosen because it was found to have a clear line at 15.8 "20 (Cu Ka radiation).However, when the pattern obtained was com- pared with the J.C.P.D.S. card file it was found that this material was in fact lanthanum hydroxide.16 A survey of several batches of so-called lanthanum oxide from different manufacturers (Hopkin and Williams, BDH Chemicals and Koch-Light) revealed that even when the bottles were previously unopened the sample would consist of a mixture of La(OH), and La203.17 Lanthanum oxide rapidly absorbs water on exposure to air giving lanthanum hydroxide, causing a marked increase in volume of the material. A previous referencels to the use of lanthanum oxide in quantitative X-ray diffraction analysis (as a "heavy absorber" to overwhelm the effects of the rest of the matrix) noted that different batches of "lanthanum oxide" showed a variable stability to hydration.I t is probable that here, too, the author was using lanthanum oxide - hydroxide mixtures. Lanthanum hydroxide precipitated from solution is gelatinous and gives a poorly resolved954 Analyst, vol. 104 X-ray diffraction pattern with broad peaks. The internal standard material was therefore prepared by stirring the lanthanum oxide - hydroxide mixture obtained commercially with water for 30 min and drying the product over pliosphorus pentoxide or in an oven at 110 "C. We have found the diffraction patterns of various batches of lanthanum hydroxide prepared in this way to be indistinguishable, but re-calibration would normally be carried out with each new batch as it is brought into use.SKIDMORE AND SCHWARZ : QUANTITATIVE DETERMINATION OF Preparation of the X-yay s9ecirnen To achieve reproducibility in any X-ray diffraction analytical method, the preparation of the specimen must aim to miniinise preferred orientation, to give maximum intensities by reducing primary extinction and microabsorption and to mix any added material (for example, an internal standard) uniformly into the sample. In this study a small ball mill was used to grind the sample, moistened with propan-2-01, and at the same time to mix in the added material. Klug and AlexanderlO point out that it is desirable to reduce the size of the crystallites to about 5 p m but that caution must be exercised concerning prolonged grinding. Apart from the general danger of introducing stress into the lattice and producing very small particles and thus causing the broadening of the diffraction lines,1° it has been suggested that over-grinding may cause a transformation of P-lead dioxide into a-lead dio~ide,19~Jg*~~ the coating of particles of cc-lead dioxide by P-lead dioxide,' or the thermal decomposition of cc-lead dioxide.6 An investigation of the optimum milling time was there- fore carried out. The effect on diffraction line intensities of milling a 1 + 1 + 1 mixture of lanthanum hydroxide, cc-lead dioxide and P-lead dioxide is shown in Fig.1. No dramatic peak-height fluctuations are seen, indicating that the mixing-in of the lanthanum hydroxide is rapidly achieved, The initial increase in intensity of one of the P-lead dioxide lines paralleled by a fall in intensity of the other indicates that the i3-lead dioxide sample shows a small amount of preferred orientation that can be removed by reducing the particle size.X La (OH), (100) Y a(110) v La (OH), (1 10) 0 1 2 4 7 10 Mi I I i ng ti me/m in Fig. 1. Effect of milling time on the X-ray diffraction line intensities on a 1 + 1 + 1 mixture of a-lead dioxide, /3-lead dioxide and lanthanum hydroxide. Measurement of difracted line intensity Peak area is a better measure of the diffracted intensity than peak height, particularly for materials such as the lead dioxides, which can show considerable peak broadening because of the small size of the crystallites. The areas were calculated from step scan data, which gave the counts accumulated in 100 s taken at every 0.05 "20.The background was estimated using the regions just before and just beyond the peak and was subtracted from the total number of counts under the peak to give the peak area.October, 1979 LEAD DIOXIDE POLYMORPHS BY X-RAY POWDER DIFFRACTOMETRY Procedures 955 Direct intensity method Wash the mixture out of the mill with propan-2-01 as required and remove the solvent. intervals of 0.05" and a step time of 100 s. the total counts and subtracting the background. calibration line. Weigh 1 g of the sample into the McCrone mill, add 3 ml of propan-2-01 and mill for 4 min. Obtain the X-ray diffraction pattern from 21 to 27.5 "26 by a step scan procedure using Calculate the area under the peaks by adding Obtain the phase composition from the Internal standard method Proceed as in the direct intensity method, obtaining the diffraction pattern between 14 and 27.5 "26.Weigh 1 g of the sample and 0.5 g of lanthanum hydroxide into the McCrone mill. Double dilution method and 27.5 "26 only. mix thoroughly in an agate pestle and mortar. 21.5 and 27.5 "26. Carry out the internal standard method, but obtain the diffraction pattern between 21.5 Add a further 1 g of lanthanum hydroxide to 1 g of the internal standard mixture and Obtain the diffraction pattern again between Calibration lines 1 g of lead dioxide and treat as the sample in the methods above. Weigh out the a- and P-lead dioxide standards in appropriate amounts to give a total of Lead sulphate cowection factor Weigh out lead sulphate (PbSO,, Hopkin and Williams, G.P.R.grade) and the ,&lead dioxide standard in appropriate amounts to give a total of 1 g and treat as the sample in the internal standard method above, obtaining the diffraction pattern between 14 and 24.5 "20. Results and Discussion Most quantitative calibration methods are derived from the well known equation given by Klug and AlexanderlO : where Iij is the intensity of the ith line of component J , x, is the weight fraction of J , ,ii* is the total matrix mass absorption coefficient and k , is a constant containing the instru- mental factors and the density of J . Rearranging equation (1) to give an expression for xJ yields xJ = KiJP*IiJ . . . . . . . . - * (2) where Ki, is the reciprocal of k,. Direct Intensity - Weight Fraction Calibrations component (p;) of the sample multiplied by their respective weight fractions : The total matrix mass absorption coefficient, p*, is the sum of the coefficients of each N - (3) The mass absorption coefficient of a compound at a particular wavelength depends solely on the atoms it contains and so there is only one mass absorption coefficient for any lead dioxide.956 Analyst, Vol, 104 Hence, as the positive plate material always consists mainly of lead dioxide, the matrix mass absorption coefficient can be assumed to be constant and the approximation can be made that SKIDMORE AND SCHWARZ : QUANTITATIVE DETERMINATION OF .. . . - (4) ~j cc Iij . . .. This approach was used by Wiesenei- et aZ.l and gave straight-line calibration graphs.We repeated the calibration with four mixtures of our standard lead dioxide (Table I) and also obtained straight-line graphs of intensity against weight fraction with correlation coefficients better than 0.97. Although the method works well for mixtures of standard lead dioxides, it is considered necessary by Wiesener et aZ.l to wash the plate material to remove lead(I1) compounds before X-ray analysis to ensure that the X-ray sample contains only lead dioxide in line with the initial assumption in relation (4). Evidence suggests, however, that some of the lead dioxide will dissolve during this procedure,21 possibly leading to a change in the a : ,8 ratio. More- over, although an external standard was used to indicate instrumental effects such as variations in incident beam intensity, there are other effects such as variations in the sample packing or its position in the goniometer for which an external standard does not compensate.The total crystalline lead dioxide is then given by Sekido and YokohQ use equation (4) with proportionality constants ka and kb. T , = kaIia -t kbIjp . . . . . . . . .. For a large set of Ii, and I j p data, these authors used a least-squares numerical method to calculate k a and kb by minimising the function where T , is the total lead dioxide in each sample determined chemically. However, when the material contains unknown and variable amounts of amorphous lead dioxide, T , > T , by an unknown and variable amount for each sample and equation (6) cannot be used. TABLE I DATA FROM a- AND P-LEAD DIOXIDE STANDARD MIXTURES of E-PbO, (xR) (1,) of P-PbO, (XP) (Id XPl% I p l I u Intensity of ( 1 10) line Intensity of (1 10) line Weight fraction of a-PbO,, counts Weight fraction of P-PbO,, counts 0.201 2.45 x 105 0.799 3.77 x 106 3.98 15.35 0.334 3.59 x 105 0.666 2.60 x lo6 1.99 7.25 0.500 7.06 x 105 0.500 2.23 x lo6 1.00 3.16 0.801 9.74 x 105 0.199 1.10 x 106 0.248 1.13 Intensity Ratio - Weight Fraction Ratio Calibration equation (2) Taking the ratio of the intensities of two components of a mixture, J and K , gives from Dodson3 and, following his paper, Kordes,6 Fedorova et aZ.' and Dugdales plotted a semi- logarithmic graph of x, against log(Ijp/Ii,).By comparing this relation with equation (7) it can be seen that the graph of x, against Ijp/Ii, would not be expected to give a straight line. Only a brief, unsatisfactory derivation of this relationship has been given,6 and it is possible that the logarithmic relationship was introduced to give the appearance of direct proportionality to the graph.We constructed a calibration graph of q/x. against I j p / I i a from the same four mixtures used for the direct intensity - weight fraction calibrations (Table I) and obtained a straight line (correlation coefficient 0.998). Because the cc-lead dioxide is in effect being used as a standard for the ,&lead dioxide, variations in matrix effects or in instrumental performance are eliminated, but no estimate of the total crystallinity of the material can be obtained.October, 1979 LEAD DIOXIDE POLYMORPHS BY X-RAY POWDER DIFFRACTOMETRY 957 Because of the limitations of these direct methods of relating intensity and concentration, two further methods were applied to this analysis, the internal standard method and the double dilution method.Internal Standard Method standard S added in a known constant amount. Internal standard analysis is based on equation (7) where component K is the internal . . . . . . (8) That is: X J -7, - = c x - . . X S 4 s where C is the proportionality constant. TABLE I1 CALIBRATION DATA FOR INTERNAL STANDARD METHOD Weight of ct-PbO, Weight of La(OH), Intensity of x-PbO, (1 10) line Intensity of La(OH), (111) line Weight of P-PbO, Intensity of P-PbO, (1 10) line Weight of La(OH), Intensity of La(OH), (111) line 0.061 3 0.2509 0.495 8 0.751 1 0.904 4 1.001 0.997 6 1.253 1.498 1.749 1.952 0.075 38 0.211 6 0.367 7 0.6142 0.726 9 0.7785 0.779 3 0.993 6 1.273 1.510 1.619 1.945 1.758 1.493 1.251 6 1.100 0.966 1 1.005 0.742 3 0.493 2 0.229 2 0.065 1 4.417 3.948 3.228 2.745 2.157 2.017 2.304 1.675 1.074 0.503 9 0.1372 Eleven standard mixtures (Table 11) were made up and an X-ray diffraction pattern of a Least-squares typical mixture is given in Fig.2, showing the lines used in the analysis. regression analysis on the data in Table I1 gave the following lines: for u- + 0.0320 weight a-PbO, - intensity a-PbO, weight La(OH), - ''''O weight P-Pbo, = 0.444 x intensity P-PbO, + o.0279 weight La(0H) intensity La(0H) , intensity La(OH), for p- Angle, "28 Fig. 2. X-ray diffraction pattern of a 1 + 1 + 1 mixture of or-lead dioxide, /$lead dioxide and lanthanum hydroxide.958 SKIDMORE AND SCHWARZ : QUANTITATIVE DETERMINATION OF Analyst, Vol.104 both with correlation coefficients of 0.997. The two small positive intercepts imply that the background was slightly overestimated, probably because it is rising under both the peaks. The situation is shown, exaggerated, in Fig. 3. The background, estimated as a straight line between the averages of the two regions b, and b,, lies above the actual curved back- ground. Thus, the internal standard method gives satisfactory calibration lines and has several advantages over the methods used by previous workers. It provides a compensation factor for both matrix and instrumental variability. Because the crystalline forms are determined separately the results can be used to estimate the non-crystalline lead dioxide by subtraction from the lead dioxide content determined chemically. Contents of cc- or ,&lead dioxide below about 2% will therefore not be detected. I Angle," 28 Fig.3. Over-estimation of the background below the peaks. Internal Standard Method in the Presence of Lead Sulphate The presence of lead sulphate in some positive plate samples causes a problem because its (111) line is superimposed on the or-lead dioxide (011) line. Rather than attempt to remove lead sulphate from the matrix by washing, it should be possible to make a correction for its presence, using the (011) line, which is free from overlap, to estimate the lead sulphate con tent . The total intensity of the compound line I i , is given by x,/x, can be obtained from another line l of component K : Thus, substituting from equation (10) into equation (9) and rearranging: The ratio of proportionality constants e/d is given by .... .. .. . . .. . . .. A set of mixtures was prepared containing lead sulphate, P-lead dioxide and lanthanum hydroxide to give a calibration line from equation (12) for the estimation of the ratio e/d. A point from lead sulyhate alone was also included. Performing a least-squares regression on the data given in Table I11 gives a straight line with a gradient of 0.432 and an intercept of 11 797 counts (correlation coefficient 0.997). The intercept arises from the overestimationOctober, 1979 LEAD DIOXIDE POLYMORPHS BY X-RAY POWDER DIFFRACTOMETRY TABLE I11 CALIBRATION DATA FOR THE LEAD SULPHATE CORRECTION 959 Intensity of PbSO, Intensity of PbSO, Weight of PbSO, Intensity of PbSO, (011) line (011) line, counts (111) line, counts Weight of La(OH), Intensity of La(OH), (111) line _ _ _ _ _ _ _ _ 17 175 45426 97 586 146 909 158213 179 050 202 166 252 183 337 301 1 055 491 23061 38 094 53 305 81 107 90 669 86 628 97 393 116553 132 323 474583 0.090 6 0.204 1 0.397 9 0.598 2 0.51 1 7 0.667 7 0.798 2 0.9950 1.208 - 0.046 1 0.1242 0.255 2 0.406 4 0.329 1 0.446 6 0.5157 0.647 5 0.876 8 __ of the background and should be neglected to avoid the systematic error of including it twice in the determination of the a-lead dioxide content.Least- squares regression on the data in Table I11 gives The data can also be used to determine the lead sulphate content of the sample.intensity PbSO, (01 1) - 0.0288 weight PbSO4 = 0.724 weight La(0H) intensity La(0H) with a correlation coefficient of 0.996. is negative, implying that the background is underestimated. In this instance the background correction constant Double Dilution Method The internal standard method cannot compensate for the effects of microabsorption that arise because diffraction from the. sample and from the standard take place within different particles. The double dilution method, which should eliminate both matrix and micro- absorption effects, has been reviewed by Clark and Preston.12 One portion of “diluent,” a compound absent from the sample matrix, is added, the intensity of a line of each com- ponent is measured, a further portion of diluent is added and the intensities of the same lines are measured again. Then, where Iijf and Iij” are the intensities when one and two portions, respectively, of diluent are added.Several of the mixtures used for the internal standard calibration were diluted with a further 1 g of lanthanum hydroxide and the diffraction patterns obtained again. Although Clark and Preston recommend that the diluent should be amorphous and should have a low mass absorption coefficient, this standardisation was successfully carried out with a crystalline high mass absorption material. The results are given in Table IV. TABLE IV CALIBRATION DATA FOR THE DOUBLE DILUTION METHOD F ( I ) = I&J’I&J’’/(IU’ - IiJ”). Weight fraction, cc F(1,) Weight fraction, /3 F(1p) 0.125 1.19 0.875 34.24 0.249 2.20 0.751 28.18 0.498 4.85 0.502 20.34 0.628 5.99 0.372 16.25 0.752 6.05 0.248 8.90 0.884 7.14 0.116 2.29960 SKIDMORE AN11 SCHWARZ The equations of the lines are calculated to be == 0.122 F(I,) - 0.0342 for cx with a correlation coefficient of 0.984 and xp = 0.0245 F(1p) + 0.0278 for ,f3 with a correlation coefficient of 0.993, where F ( I ) is the right-hand side of relation (13).The background corrections shown by the intercepts are significantly large and limit the precision of the method and its application at low levels. The two mixing-in stages that are required for this method result in greater random errors than for the internal standard method and make the analysis more time consuming. Large random errors also arise from the counting and particle statistics in the more dilute sample when the content of one phase is low.Because of these sources of error, the double dilution method is considered to be inferior to the internal standard method. Conclusion This investigation has demonstrated that internal standard X-ray diffraction provides a more reliable method for the phase analysis of lead dioxide in positive plate material from lead-acid batteries than those used by previous workers, Analysis of positive plates of lead-acid batteries for a- and P-lead dioxide are now carried out routinely by the internal standard method in this laboratory and an estimate of the amorphous lead dioxide content of the plates is obtained from the difference between the crystallographic and chemical determination.The authors thank the Directors of Rerec Group Limited for permission to publish this work and D. M. Holton for her help with the experimental work. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. References Wiesener, K., Hoffmann, W., and Rademacher, (I., Electrochim. Acta, 1973, 18, 913. Wiesener, K., and Reinhardt, P., 2. Phys. Chem., 1975, 256, 285. Dodson, V. H., J . Electrochem. Soc., 1961, 108, 406. Ness, P., Electrochim. Acla, 1967, 10, 161. Caulder, S. M., and Simon, A. C., J . Electrochem. SOC., 1974, 121, 1546. Kordes, D., Chemie-Ingr-Tech., 1966, 38, 638. l?edorova, N. N., Aguf, 1. A., Levinson, L. M., and Dasoyan, M. A., 2av. Lab., 1964, 30, 727. Dugdale, I., in Collins, D. H., Edztor, in the discussion of the paper by Acton, R. G., “Power Sources,” Sekido, S., and Yokoh, T., J . Electrochem. SOC. Japan, 1963, 31, 15. Klug, H. P., and Alexander, L. E., “X-ray Diffraction Procedures,” Second Edition, Wiley- Chung, F. H., A h . X-ray Analysis, 1974, 17, 106. Clark, N. H., and Preston, R. J . , X-ray Spectrom., 1974, 3, 21. J3agshaw, N. E., Clarke, R. L., and Halliwell, B., J . Appl. Chern., 1966, 16, 180. Angstadt, K. T., Venuto, C. J . , and Ruetschi, P., J . Electrochem. SOC., 1962, 109, 177. Joint Committee on Powder Diffraction Standards, Powder Diffraction File card numbers 23-331 Joint Committee on Powder Diffraction Standards, Powder Diffraction File card number 6-0585. Joint Committee on l’owder Difiraction Standards, Powder Diffraction File card number 5-602. Flinter, B. H., Neues Jahrb. Maneral Monatsh., 1973, 5, 216. Jenkins, R., A d v . X-ray Analysis, 1974, 17, 32. Weigelt, D., and Schrader, R., 2. Anorg. Allg. Chem., 1970, 372; 228. Reutschi, P., Sklarchuk, J., and Angstadt, 13. T., zn Collins, D. H., Editor, “Batteries,” Volume 1 , Pergamon Press, Oxford, 1966, p. 142. Interscience, New York, 1974. and 27-1200. Pergamon Press, Oxford, 1962, p. 89. Received December 22nd, 1978 Accepted April 20th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400952

出版商:RSC    

年代:1979

数据来源: RSC