i V SUMMARIES OF PAPERS IN THIS ISSUE [August, 1966Summaries of Papers in this IssueRapid Automated Determination of Biphenyl in Citrus Fruit RindA totally automated analytical procedure for determining the fungistatbiphenyl in citrus fruit rind has been developed. Small pieces of hand-peeledrind are automatically homogenised in water and steam-distilled to liberate thebiphenyl, which is trapped in cyclohexane solution ; this solution is exhaustivelyextracted'to remove interfering steam volatiles and the biphenyl remaining isread a t 246nip in a continuous flow recording spectrophotometer. Timerequired from the introduction of rind samples to read out of biphenyl presentis 9 minutes, with about 15 minutes for the first sample. The useful range isfrom 1 to about 150 p.p.m.on a whole-fruit basis with a reproducibility ofabout 3 per cent. The method has been applied most extensively to Valenciaoranges.F. A. GUNTHER and D. E. OTTDepartment of Entomology, University of California Citrus Research Center andAgricultural Experiment Station, Riverside, California.Analyst, 1966, 91, 475-481.Determination of Carbon in Steel by a Dynamic Infrared SystemA simple, automatic apparatus has been developed for the rapid deter-mination of carbon in steel. I t is based upon the continuous measurementof carbon dioxide evolved during the high temperature combustion of steeli n oxygen by using a specially designed infrared gas analyser and integrationsystem. When this is used in conjuction with a conventional resistance-tubefurnace, the speed of the determination varies from 40 to 55 seconds for mildand low alloy steels, and slightly longer for highly alloyed materials.G.WHITE and P. H. SCHOLESBritish Iron and Steel Research Association, Metallurgy Division, Hoyle Street,Sheffield 3.Analyst, 1966, 91, 482-489.The Determination of Aluminium in Iron and SteelVarious colorimetric reagents have been examined for their suitabilityin a standard method for determining aluminium in ferrous metals. Inter-fering elements are removed by a mercury-cathode separation followed bycupferron - chloroform extraction. In the method adopted, aluminium isdetermined by measuring the optical density of its complex with Alizarinred S - calcium reagent. The method has been tested with a wide rangeof steels.J.A. CORBETTPhysical Metallurgy Section, Commonwealth Scientific and Industrial ResearchOrganization, Australia.and B. D. GUERINMetallurgy Department, University of Melbournc, Australia.A n~Zy.~t, 1966, 91, 490-498vi SUMMARIES OF PAPERS I N THIS ISSUEA Chemiluminescence Method for Determining Ozoneh method for determining ozone is described which is characterised bythe direct recording and automatic determination of ozone within a widerange of concentrations. The development of this method is based on theuse of a chemiluminescent solution that is stable, and shows a linear relation-ship between the light emitted and the ozone concentration. A combinationof rhodamine B with gallic acid in ethanol is satisfactory in operation and doesnot itself emit light.The electronic instrumentation used is relativelysimple. Other methods of ozone analysis based on this principle meet withmuch difficulty, owing to the direct oxidation of the chemiluminescent com-pound. The present method, by contrast, involves the use of gallic acid asan ozone acceptor, and rhodamine B, which remains unchanged during themeasurement, as a photon emitter. Observations made with an oscillographof the light emitted by single bubbles of ozonised air passing through the chem-iluminescent solution give valuable information about the response-timeof the system.D. BERSIS and E. VASSILIOUNuclear Research Center “Democritus,” Aghia Paraskevi Attikis, Athens, Greece.[August, 1966A nalyst, 1966, 91, 499-505.The Determination of Tantalum by the Solvent Extractionof a Tantalum - Pyrogallol ComplexA colorimetric procedure for determining up to 1.2 mg of tantalum inthe presence of up to 20 mg of niobium, or up to 180 mg of tungsten, hasbeen developed.The colourless tantalum - pyrogallol complex is extractedinto ethyl acetate a t pH between 4-5 to 6.0 by means of tetrahexyl or tetra-butyl ammonium iodide and back-extracted with acidified ammonium oxalate(pH 2.0). The yellow complex obtained is measured spectrophotometricallyat 400 mp.BETSY BIRABEN SCOTTFacultad de Quimica y Farmacia, Universidad Nacional de La Plata, Argentina.Analyst, 1966, 91, 506-510.Flame-photometric Determination of Sodium and Potassiumin Manganese OresTwo procedures are described; in one, the sample is dissolved in hydro-chloric acid, interfering elements are precipitated with 8-hydroxyquinolinein ammoniacal solution, and the precipitate then extracted into chloroform.Sodium and potassium are determined in the aqueous phase by means ofa filter flame photometer.The second procedure is more suitable for routine use and involves thedissolution of the sample in hydrochloric acid, followed by the addition ofsulphuric acid and aluminium nitrate to suppress interferences, and thedirect evaluation of the sodium and potassium contents of the solution bymeans of either a prism or a filter flame photometer.Comparative resultsobtained by this alternative procedure on instruments of these two typesare given.B. G.RUSSELLThe National Institute for Metallurgy, Yale Road, Milner Park, Johannesburg.Analyst, 1966, 91, 511-519.Determination of Cyclamate in Soft Drinks byGas ChromatographyShort PaperM. L. RICHARDSON and P. E. LUTONJohn & E. Sturge Ltd., Lifford Chemical Works, Kings Norton, Birmingham 30.Analyst, 1966, 91, 520-521viii SUMMARIES OF PAPERS I N THIS ISSUEThe Determination of Cyclamate in Soft Drinks by Titrationwith Sodium NitriteShort Paper[August, 1966M. L. RICHARDSON and P. E. LUTONJohn & E. Sturge Ltd., Lifford Chemical Works, Kings Norton, Birmingham 30.Analyst, 1966, 91, 522-523.The Determination of Ethanolamine and Serine inPhospholipidsShort PaperA. J. de KONINGFishing Industry Research Institute, University of Cape Town, Rondebosch, CapeTown, South Africa.AnaZyst,1966, 91, 523-525.The Paper Chromatography of Some Purines, Pyrimidinesand ImidazolesShort PuperM.N. KHATTAK, N. T. BARKER and J. H. GREENDepartment of Nuclear and Radiation Chemistry, University of New South \Vales,Kensington, Sydney, Australia.Analyst, 1966, 91, 526-528.The Determination of Total Sulphur in Soil and Plant MaterialShort PaperI. A. CHAUDHRY and A. H. CORNFIELDChemistry Department, Imperial College of Science and Technology, London, S.W. 7.Analyst, 1966, 91, 528-530.A Simple Colorimetric Finish for the Johnson - NishitaMicro-distillation of SulphurShovl PaperG. A. DEANCommonwealth Scientific and Industrial Research Organization, Division of Soils,W.A.Regional Laboratory, Nedlands, \Vestern Australia.Ancrlyst, 1966, 91, 530-532.The Oxidation of Hydroxylamine in Sodium Hydroxide in thePresence of Copper(1r)Shovt PaperJ. H. ANDERSONLong Ashton Research Station, University of Bristol, Somerset.Analyst, 1966, 91, 532-535x SUMMARIES OF PAPERS IN THIS ISSUE [August, 1966Limits of Sensitivity of Detection of Aluminium in Amorphous andCrystalline Aluminium Oxide by X-ray DiffractometryShort PaperC. J. TOUSSAINT and G. VOSEuratom, Chemistry Department, Analytical and Mineral Chemistry Section,Ispra, Italy.Analyst, 1966, 91, 535-537.Nitrogen Factor for KidneyReport prepared by the Meat Products Sub-committeeANALYTICAL METHODS COMMITTEE14 Belgrave Square, London, S.W.1.Analyst, 1966, 91, 538-539.Nitrogen Factor for Cod FleshReport prepared by the Fish Products Sub-committeeANALYTICAL METHODS COMMITTEE14 Belgrave Square, London, S.W.1.Analyst, 1966, 91, 540-542.Methods for the Analysis ofNon-Soapy Detergent (NSD) ProductsbyG. F. LONGMAN, B.SC., F.R.I.C. & J. HILTON, B.Sc., A.R.I.C.(Unilever Research Laboratory, Port Sunlight)Society for Analytical Chemistry Monograph No. I-0-This Monograph describes in detail the methods of analysisdeveloped in Unilever’s Laboratories for the identificationand assay of components of NSD Products.-0-Available ONLY fromThe Editor, “The Analyst,” I 4 Belgrave Square, London, S.W. I(Not through Trade Agents)Price 15s.or U.S. $2.00 Post freeA remittance made out to “Society for Analytical Chemistry” should accompany every order.Members of the Society may purchase copies at the special price of 5s., post freeAEl SCIENTIFIC APPARATUSBULLETIN NO. 4ANALYSIS OF SOLIDS BY SPARKSOURCE MASS SPECTROMETRYSince the possibility of using spark source ionization for the analysis of solidswas first recognised ten years ago, the design of the double focusing massspectrometer using Mattauch geometry has been considerably improved. Andtoday one instrument-namely the AEI MS7-is capable of detecting im-purities at levels as low as 1 part in IO@. As a result the MS7-which, incident-ally, was the first commercially available double focusing instrument to bebuilt expressly for the analysis of solids-has found wide applicationsparticularly where overall coverage of all elements and comparison analysiswithout standards are valuable.Now the MS7 technique has been further improved by the careful control ofcertain parameters, and very good reproducability and accuracy can beguaranteed.This bulletin reviews the parameters affecting analytical accuracyand outlines the methods of control developed by AEI engineers.I t has been shown that the sensitivityof the majority of elements differs fromsome standard such as iron by no morethan a factor of 3 . In other words, mostrelative sensitivity factors lie between0.3 and 3 . In the case of the MS7, thedetermination of relative sensitivityfactors is considerably simplified bythe fact that the response of the instru-ment is linear over a very large rangeof concentration.Indeed Hannay &Ahearn established linearity over therange 104 to 1 using doped siliconsamples. More recently, W. A. Wolsten-holme (AEI Consultant Lab.) hasreported on the investigation of golddoped titanium samples covering a con-centration range from 0.5% to .02 ppmby weight; a range of more than lo5to 1.These samples were chosen because oftheir suitability for neutron activationand wet chemical analysis. Figure 1shows the relative ion intensity of goldplotted against the concentration de-+,ermined by chemical or neutron acti-FOOTNOTE :I Hannay and Ahenrn (1954) Anal. Chem 261956.reported in tables 1 and 2.In copper and steel relative sensitivitiesare very similar for chromium (1 -8 and1.4) and for tin ( 1 -3 and 1.1).Only forsome low BP elements is there amarked dependence on the matrix, e.g.(1.4 and 2.6) for lead.Homogeneity ofstandards and samplesCare has to be taken to use homo-geneous standards or alternatively toincrease the rate of consumption ofsample above the usual 5 to 10 milli-grams.As it happens, however, the possibilityof inadvertently using an inhomogene-ous sample has been materially reducedby the introduction of more reliablemethods of sample preparation.FIG. 1TABLE 1RELATIVE SENSITIVITIES ANDREPRODUCIBILITY OF R.F. SPARKSOURCE ANALYSIS OF COPPER:JOHNSON MATTHEY CA230kV r f.; 19.5kV accel. volts; pulse length andrepetition rate varied in the analytical plates,vation techniques.The “just detect-able” line is an individual assessmentfor ion intensity and the “densito-metric” line indicates that a micro-densitometer was used to scan thespectral lines; the two graphs havebeen displaced for the sake of clarity.Matrix effects are generally very small,as is illustrated by the relative sensi-tivities for copper and steel standards11014013016021 0250110 --lelativensitivit)3.851.41 .I1.31.43.11.8 -O htandard deviationseparatcnalyses31303225232720 -5 repealrposure131 1189.2211321 ReproducibilityThe most important improvements inreproducibility have been achieved bycareful control of certain instrumentalparameters.As a result of recent in-vestigationse it is now clear that themost important factor is one that iscomparatively simple to control ; name-ly, the ion accelerating voltage. Whenthis is always set to the same value, andother source conditions are kept asconstant as possible, reproducible ana-lytical results are obtained. The princi-pal reason is the improved constancyof the relative sensitivity factors fordifferent elements. Table 2 shows howclose agreement with known values isobtained when such relative sensitivityfactors are measured and used to cor-rect observed concentrations.Variations in PhotoplateThe standard deviations on identicalexposures on a photoplate using ahomogeneous aluminium standard in-dicate that the best standard deviationobtained for the different elements isabout loo/,.The standard deviation onisotope .ratios, i.e. where the relativeFOOTNOTE :2 Halliday, Swift and U'olstenholme; Quantita-tire Analysis by Spark Source Mass Spectro-metry, International Mass SpectrometryConference, Paris 1964.TABLE 2ACCURACY OF ANALYSIS AFTERZALIBRATION: BUREAU OF ANALYSEISTANDARDS LTD. MILD STEELRESIDUAL SERIES SPECTROGRAPHICSTANDARD SSl4Pulse length 100 microsecs ; 300 pulses/sec. :30kV r.f.; 19.5kV accel. volts.Concenrration % wtInpurityCrcoNicuZrNbMoSnPbLIST. average of MS7, CONeCtOd liven Spectmeight analyses using SSl2 graphic value0.26 0.18 0.1850.1 9 0.18 0.190.1 0 0.13 0.130.047 0.029 0.040.017 0.007 (0.005)0.072 0.025 (0.05)0.053 0.060 0.070.019 0.017 0.020.01 8 0.007 (0.0075)( ) not certified-approximatesensitivity factor is not involved andwhere a limited area of plate is used,show better figures of 3 to 4%.Thisindicates the likely variation due to theplate itself, and represents a limit thatwould remain even if a calibrationspectrum of a standard was placed onthe same photoplate as the sample tobe analysed.Reproducibility of AnalysesWhen different photoplates are used inseparate analyses the standard devia-tion increases.In repeat analyses different photoplateswill be used and also source conditionsmay vary slightly. To cover the fullrange of exposures (10' to 1) it is some-times necessary to vary the pulse lengthand repetition rate of the spark whichmight also contribute to the variations.An internal standard, although notessential, eliminates the need for sucha wide range of exposures.When the spark pulse-length and repe-tition rates for the analysis of a copperstandard shown in Table 1 were heldconstant throughout the series of suc-cessive analyses of one of the standards,the minimum standard deviation did infact decrease towards the minimumvalue previously obtained for repeatexposures (about 1 PA).Accuracy of AnalysisThe accuracy which can be attainedwith the MS7 when all parameters areproperly controlled is perfectly exem-plified by the data set forth in Table 2.In this case, of course, a standard ofreasonable homogeneity has been usedto establish relative sensitivity factorsfor the impurities in the given matrix.It will be noted that only one standardis used, and that the levels of concen-tration in the standard do not corres-pond too closely to those in the un-known.A steel matrix has been takenas an example because studies of sourceconditions indicated that it seemed themost likely (compared with the alu-minium and copper standards) to givepoor results should the source con-ditions change. It is clear, however, thatthe results once corrected for relativesensitivity factors are in excellent agree-ment with those given by chemicalanalysis.Further information on the MS7can be obtained from AssociatedElectrical Industries Ltd., ScientificApparatus Department, BartonDock Road, Urmston, Manchesteror your nearest AEI off ice.8 I01