摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYEDITORAL ADVISORY BOARD"Chairman: H. J. Cluley (Wembley)'L. S. Bark (Salford)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)L. R. P. Butler (South Africa)E. A. M. F. Dahmen (The Netherlands)A. C. Docherty (Billingham)D. Dyrssen (Sweden)J. Hoste (Belgium)H. M. N. H. Irving (Leeds)H. Kaiser (Germany)M. T. Kelley (U.S.A.)W. Kemula (Poland)"W. T. Elwell (Birmingham)"J. A. Hunter (Edinburgh)"G. F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)"J. M. Ottaway (Glasgow)"G. E. Penketh (Billingham)"T. B. Pierce (Harwell)E. Pungor (Hungary)D. I. Rees (London)"R. Sawyer (London)P. H. Scholes (Sheffield)"W. H.C. Shaw (Greenford)S. Siggia (U.S.A.)A. A. Smales, O.B.E. (Harwell)A. Walsh (Australia)T. S. West (Aberdeen)A. L. Wilson (Medmenham)P. Zuman (U.S.A.)"A. Townshend (Birmingham)"Members of the Board serving on The Analyst Publications CommitteeREGIONAL ADVISORY EDITORSDr. J . Aggett, Department of Chemistry, University of Auckland, Private Bag, Auclcland, NEWProfessor G. Ghersini, Laboratori CISE, Casella Postale 3986, 201 00 Milano, ITALY.Professor L. Gierst, Universit6 Libre de Bruxelles, Facult6 des Sciences, Avenue F.-D. Roosevelt 50,Professor R . Herrmann, Abteilung fur Med. Physik., 63 Giessen, Schlangenzahl 29, GERMANY.Professor W. E. A. McBryde, Dean of Faculty of Science, University of Waterloo,Waterloo, Ontario,Dr.W . Wayne Meinke, KMS Fusion Inc., 3941 Research Park Drive, P.O. Box 1567, Ann Arbor,Dr. I. Rubeska, Geological Survey of Czechoslovakia, Kostelni 26, Praha 7, CZECHOSLOVAKIA.Dr. J. RGiikka, Chemistry Department A, Technical University of Denmark, 2800 Lyngby, DENMARK.Professor K. Saito, Department of Chemistry, Tohoku University, Sendai, JAPAN.Dr. A. Strasheim, National Physical Research Laboratory, P.O. Box 395, Pretoria, SOUTH AFRICA.ZEALAN D.Bruxelies, BELGIUM.CANADA.Mich. 481 06, U.S.A.Published by The Chemical SocietyEditorial: The Director of Publications, The Chemical Society, Burlington House,London, W1 V OBN. Telephone 01 -734 9864. Telex No. 268001.Advertisements: J. Arthur Cook, 9 Lloyd Square, London, WC1 X 9BA. Telephone 01 -837 631 5.Subscriptions (non-members): The Chemical Society Publications Sales Office, Blackhorse Road,Letchworth, Herts., SG6 1 HN.Volume 101 No 1203@ The Chemical Society 1976June 197

ISSN:0003-2654    

DOI:10.1039/AN97601FX021

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

ANALAO 101 (1203) 409-496 (1976)ISSN 0003-2654J u n e 1976THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYCONTENTS409421433439445455458463469478485491ORIGINAL PAPERSApplication of Microprocessors t o Chemical Analysis : Sequential Titration ofAdipic and Boric Acids-0. Betteridge, E. L. Dagless, P. David, 0. R. Deans, G. E.Penketh and P. ShawcrossIonic Polymerisation as a Means of End-point Indication in Non-aqueous Thermo-metric Titrimetry. Part VIII. Solvent Effects i n the Determination ofPolyfunctional Carboxylic Acids and Phenols-E. J. Greenhow and A. A. ShafiSolid-state lon-selective Electrodes Based on Thin lon-selective Layers Depositedon Ionic Conductors-R. E. Van de LeestIn Vitro Studieson the Dissolution Rate o f Industrial Retarded Urea Feedingstuffsby Use of a Selective Electrode. Application of the Potentiometric UreaEnzyme Electrode i n Measurements of Dissolution Rate-lldik6 Fritz, G6zaNagy, Lajos Fodor and Ern0 PungorSuppression of Interfering Ions i n the Analysis of Plants t o Determine FluorideUsing the Fluoride Ion Selective Electrode-Brian Vickery and Margaret L.VickeryDetermination of Iodine i n Plant Material by a Neutron-activation Method-0.Johansen and E. SteinnesMicro- and Submicro-scale lodimetric Determination o f Antimony(ll1) in Anti-bilharzial Compounds by Use of Amplification Reactions-Y. A. Gawargious,L. S. Boulos and Amir BesadaAssay Methods for Use i n Stability Studies of 2-Amino-6-methyl-5-oxo-4-n-propyl-4,5-dihydro[l,2,4]triazolo[l,5-a]pyrimidine-G. E. Kitson, H. E. Hudsonand N. A. DickinsonA Polarographic and Ultraviolet Spectral Investigation of the PharmacologicallyActive 1 -Benzhydryl-4-(6-methyl-2-pyridylmethyleneimino)piperazine-M. R. Smyth, W. Franklin Smyth, R. F. Palmer and J. M. CliffordThe Determination of Warfarin i n Animal Tissues by Electron-capture Gas -Liquid Chromatography-E. M. Odam and M. G. TownsendAutomated Colorimetric Determination of Phosphorus in Silicate Rocks i n thePresence of Silicon-D. Whitehead and S. A. MalikBook ReviewsSummaries o f Papers in this lssue-Pages iv, v, viii, ixPrinted by Heffers Printers Ltd, Cambridge, EnglandEntered as Second Class at New York, USA, Post Offic

ISSN:0003-2654    

DOI:10.1039/AN97601BX023

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

SUMMARIES OF PAPERS I N THIS ISSUE June, 1976Summaries of Papers in this IssueApplication of Microprocessors to Chemical Analysis :Sequential Titration of Adipic and Boric AcidsAn automatic titrator has been designed so that titrations and processingof the results are controlled by a microprocessor. The titrant is added step-wise, in equal volumes, to the sample and the potential is measured aftereach addition. The instrument has been used for the successive titration,against sodium hydroxide, of a mixture of adipic and boric acids. Themethod for the determination of the equivalence volumes is based on thedifferential plot of ApH/AV against V . The effect on the results of varyingthe rate of addition of titrant, the rate of stirring and the orientation ofthe electrodes and delivery tube have been examined.The accuracy andprecision obtained compare favourably with those of the standard potentio-metric titration procedure, in which mannitol is added for the titration ofboric acid and the titrant is added until a pre-selected potential is reached.D. BETTERIDGEDepartment of Chemistry, University College of Swansea, SingletonS A 2 8PP.Park, Swansea,E. L. DAGLESSDepartment of Electrical Engineering, University Collcge of Swansea, Singleton Park,Swansea, SA2 8PP.P. DAVIDDcpartmcnt of Chemistry, University College of Swansea, Singleton Park, Swansea,SA2 8PP.D. R. DEANS, G. E. PENKETH and P. SHAWCROSSPetrochemicals Division, Imperial Chemical Industries Limited, Wilton, Cleveland,TS6 8JE.Analyst, 1976, 101, 409-420.Ionic Polymerisation as a Means of End-point Indication inNon-aqueous Thermometric TitrimetrgPolyfunctional Carboxylic Acids and PhenolsA systematic study has been made of the effect of the solvent for the sampleon the end-point sharpness and the reaction stoicheiometry in the catalyticthermometric titration of polyfunctional carboxylic acids and polyhydricphenols.The mechanisms proposed for the reactions involved in the neutralisationand indicator processes are based on the assumption that the titrants canexist as mixtures of ion pairs, solvent-separated ion pairs and free ions in thesample solution.It is suggested that the ratio of free ions to ion pairs is thefactor that controls the reaction stoicheiometry, and the cation solvatingpower of the solvent is more important than the dielectric constant of thesolution in influencing this ratio.Solvent and titrant combinations are recommended for the determinationof the total acidities of a range of benzene carboxylic acids, polyhydric phenolsand hydroxybenzoic acids by the catalytic thermometric procedure.E.J. GREENHOW and A. A. SHAFIDepartment of Chemistry, Chelsea College, University of London, Manresa Road,London, SW3 6LX.Analyst, 1976, 101, 421-432.Part VIII. Solvent Effects in the Determination ofAcrylonitrile was used as the end-point indicatorJune, 1976 SUMMARIES OF PAPERS I N THIS ISSUESolid- state Ion- selective Electrodes Based on Thin Ion- selectiveLayers Deposited on Ionic ConductorsA new type of ion-selective electrode has been developed.The electrodeconsists of a supporting material that is a good ionic conductor, covered witha thin ion-selective layer. The electrode response is fast and reproducible.Electrodes for measuring chloride, bromide, iodide and hydrogen ortho-phosphate ions are described.R. E. VAN de LEESTPhilips Research Laboratories, Eindhoven, The Netherlands.Analyst, 1976, 101, 433-438.VIn Vifro Studies on the Dissolution Rate of Industrial Retarded Ureathe Potentiometric Urea Enzyme Electrode in Measurementsof Dissolution RateAn enzyme electrode based on the neutral carrier type ammonium ionselective electrode was prepared in order to study the rate of dissolutionof urea. A procedure was developed that was found to be applicable formonitoring the urea dissolution process in different industrial feedingstuffadditives containing urea.ILDIKO FRITZInstitute for Agricultural Chemical Technology, Technical University, Budapest,Hungary.GEZA NAGYEGYT Pharmacochemical Works, Budapest, Hungary.LAJOS FODORInstitute for Agricultural Chemical Technology, Technical University, Budapest,Hungary.and ERN6 PUNGORInstitute for General and Analytical Chemistry, Technical University, Budapest,Hungary.Feedingstuffs by Use of a Selective Electrode. Application ofAnalyst, 1976, 101, 439-444.Suppression of Interfering Ions in the Analysis of Plants to DetermineFluoride Using the Fluoride Ion Selective ElectrodePlant ashes may contain sufficient aluminium and/or iron to interfere seriouslyin the determination of fluoride ions when using the fluoride ion selectiveelectrode.In the presence of these metals the known additions methodgave erroneous results, as did that involving the attempted formation ofcomplexes with ethylenediaminetetraacetic acid, disodium salt, or 1,2-cyclo-hexylenedinitrilotetraacetic acid. Good recoveries of fluoride ion wereobtained in the presence of aluminium, iron, magnesium or silicate, usingsodium citrate as the complexing agent. The application of the citratecomplex method to ashes of commercial tea, high in aluminium and iron,gave recoveries of fluoride ion of greater than 90%.BRIAN VICKERY and MARGARET L. VICKERYKenyatta University College, P.O. Box 43844, Nairobi, Kenya.Analyst, 1976, 101, 445-454

ISSN:0003-2654    

DOI:10.1039/AN97601FP041

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

...Vlll SUMMARIES OF PAPERS I N THIS ISSUE June, 197Determination of Iodine in Plant Material by a Neutron-activationMethodA method for the determination of trace amounts of iodine in dry plantmaterial is described. The method is based on the radiochemical separationof iodine-128 using alkaline fusion of the sample in the presence of an iodinecarrier, followed by isolation of iodine by a solvent extraction and a precipi-tation step. The precision of the method is about 5% a t the 0.05 p.p.m. leveland 10% at the 0.005 p.p.m. level of iodine. Results for two internationalreference plant materials are reported.0. JOHANSEN and E. STEINNESInstitutt for Atomenergi, Isotope Laboratories, Kjeller, Norway.Analyst, 1976, 101, 455-457.Micro - and Submicro- scale Iodimetric Determination ofAntimony( 111) in Antibilharzial Compounds by Use ofAmplification ReactionsSimple, rapid and highly sensitive iodimetric amplification methods aredescribed for the micro- and submicro-determination of antimony( 111) inantibilharzial compounds.The methods are based on oxidation of thesample with an excess of periodate in acetate buffered medium, maskingof the unreacted periodate with molybdate, and finally, iodimetric titrationof the iodate that is released. Specific reactions have been found to occurfor the four different compounds analysed owing to the presence, in thesame molecule, of one or two antimony(II1) and a-diol functions, both ofwhich are oxidised quantitatively and simultaneously with periodate.Thus3-, 9-, 15- and 18-fold amplification is achieved with triphenylstibine, tartaremetic, stibophen and diantimony piperazine tartrate, respectively. Thetotal mean recovery is 99.49 per cent.Y. A. GAWARGIOUS, L. S. BOULOS and AMIR BESADAMicroanalytical Research Laboratory, National Research Centre, Dokki, Cairo,Analyst, 1976, 101, 458-462.Egypt.Assay Methods for Use in Stability Studies of 2-Amino-6-methyl-5- 0x0 -4- n-propyl- 4,5- dihydro [ 1,2,4] triazolo [ 1,5- alpyrimidineTwo assay methods were developed for use in the study of the stability ofthe drug 2-amino-6-methyl-5-oxo-4-n-propyl-4,5-dihydro[ 1,2,4] triazolo-[1,5-a]pyrimidine. A difference in the results obtained by use of the twoprocedures was noted with solutions containing either lactose or 1 ,Cdioxan.The reasons for this difference have been investigated.G. E.KITSON, H. E. HUDSONPharmaceuticals Division, Imperial Chemical Industries Ltd., Hurdsfield IndustrialEstate, Macclesfield, Cheshire, SKlO 2NA.and N. A. DICKINSONDepartment of Pharmacy, University of Manchester, Oxford Road, Manchester,M13 9PL.Analyst, 1976, 101, 463-468June, 1976 SUMMARIES OF PAPERS IN THIS ISSUEA Polarographic and Ultraviolet Spectral Investigation of thePharmacologically Active 1 - Benzhydryl-4- (6-methyl-2 -pyridyl-methy1eneimino)piperazineThis paper presents a detailed study of the ultraviolet spectral and polaro-graphic behaviour of a pharmacologically active benzhydrylpiperazinederivative. I t explains the contribution made by various parts of themolecule to the ultraviolet spectrum and predicts the different forms of themolecule existing in aqueous media, The molecule was found to reduce ina 4e process when the nitrogen atom a t position 4 became protonated and in a2e process otherwise.This information, along with distribution studies inaqueous and biological media, made i t possible to develop a polarographicmethod for measuring the concentrations of this molecule in plasma andurine following therapeutic administrations to test animals.M. R. SMYTH, W. FRANKLIN SMYTHChemistry Department, Chelsea College, Univei sity of London, Manresa Road,London, SW3 6LX.R. F. PALMER and J. M. CLIFFORDG. D. Searle and Co. Ltd., High Wycombe, Buckinghamshire.Analyst, 1976, 101, 469-477.ixThe Determination of Warfarin in Animal Tissues byElectron-capture Gas - Liquid ChromatographyA method is described for the quantitative analysis of warfarin residues inanimal tissues.An extract is made with acetone, purified by column chro-matography and solvent partitioning and the final ether solution is methylatedwith diazomethane. The methylation products have been identified, andelectron-capture gas - liquid chromatography has been used to determine4-methylwarfarin, with decachlorobiphenyl as an internal standard. Therecoveries have been determined by radiochemical techniques and shown tobe suitable for the routine determination of warfarin at the 0.1 p g g-l levelin tissue.E. M. ODAM and M. G. TOWNSENDPest Infestation Control Laboratory, Ministry of Agriculture, Fisheries and Food,Tolworth, Surrey, KT6 7NF.Analyst, 1976, 101, 478-484.Automated Colorimetric Determination of Phosphorus in SilicateRocks in the Presence of SiliconAn automated colorimetric method has been developed for the determinationof phosphorus as the blue molybdophosphate complex in the presence ofsilicon following fusion of rock samples with sodium hydroxide or a lithiumtetraborate - lithium carbonate mixture.Operating conditions and reagentconcentrations have been chosen so as t o achieve complete reaction and highsensitivity. Values obtained for some French reference standards, using theproposed method, compare well with recommended values and those usingthe molybdovanadate method. Coefficients of variation of less than 2.5%and 1 yo were obtained for phosphorus concentration levels [expressed asphosphorus(V) oxide] of 0.15 and 0.5y0, respectively.D. WHITEHEAD and S. A. MALIKDepartment of Geology, University of Reading, Whiteknights, Reading, RG6 2AB.Analyst, 1976, 101, 485-490

ISSN:0003-2654    

DOI:10.1039/AN97601BP045

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

JUNE 1976 The Analyst Vol. 101 No. 1203 Application of Microprocessors to Chemical Analysis: Sequential Titration of Adipic and Boric Acids D. Betteridge E. L. Dagless P. David D. R. Deans, G. E. Penketh and P. Shawcross Department of Chemistry, University College of Satansea, Singleton Park, Swansea, SA2 8 P P Department of Electrical Engineering, University College of Swansea, Singleton Park, Swansea, SA 2 8PP Department of Chemistry, University College of Swansea, Singleton Park, Swansea, SA 2 8PP Petvochemicals Division, Impevial Chemical Iytdustries Limited, Wilton, Cleveland, TS6 8 J E An automatic titrator has been designed so that titrations and processing of the results are controlled by a microprocessor. The titrant is added step- wise, in equal volumes, to the sample and the potential is measured after each addition.The instrument has been used for the successive titration, against sodium hydroxide, of a mixture of adipic and boric acids. The method for the determination of the equivalence volumes is based on the differential plot of ApH/AV against V . The effect on the results of varying the rate of addition of titrant, the rate of stirring and the orientation of the electrodes and delivery tube have been examined. The accuracy and precision obtained compare favourably with those of the standard potentio- metric titration procedure, in which mannitol is added for the titration of boric acid and the titrant is added until a pre-selected potential is reached. By using modern electronic components, in particular the microprocessor, it is possible to build a computer1-* that is both physically small and inexpensive.The microprocessor is the heart of such a computer and consists of a single integrated circuit on a silicon chip, and costs about E20. A microcomputer can be assembled using two or three integrated circuit boards, which are then connected inside a desk-top module. It can be programmed to perform control operations and arithmetical processing. By putting the program in a permanent store (Read Only Memory), the computer becomes dedicated to performing a single programmed task and, as such, can be connected to become part of the equipment it is ~ontrolling.~ For example, it can be used to automate the sequential titration of adipic and boric acids as required for the control of the manufacture of nylon.During the process, cyclohexane is oxidised to cyclohexanol and cyclohexanone, with organic acids as side products ; adipic acid is representative of these side products. The concentration of the organic acids and boric acid must be controlled carefully and maintained within limits of h 1 . 5 and &0.5%, respectively. At present, the concentrations of these acids are determined by titration to pre-selected end-points using an automatic titrator. The boric acid is determined after the manual addition of mannitol to improve the end-point and, effectively, two titrations have to be performed. The rate of addition of titrant is variable, being fast at the beginning of the titration and slowing down as the pre-set potential is approached.The equivalence volumes are either read from the burette or calculated manually using the pH graph obtained from a chart recorder. A schematic diagram of the apparatus used is shown in Fig. 1. The system was developed so that the microprocessor controlled the titration and its pro- cessing power was used to determine the equivalence points from the full titration curve. This meant that the addition of mannitol and titration to pre-set potentials were no longer necessary. Consequently, the pH meter could be replaced by a lower quality voltmeter. 409 Boric acid is used as a catalyst.410 BETTERIDGE et aZ. : APPLICATION OF MICROPROCESSORS TO THE Andyst, VoZ. I01 1 Burette Voltmeter Recorder Variable control d - Fig. 1. Apparatus for titration to pre-set end-points.11 1L Titrant was added at a fixed rate, controlled by the computer, and the two equivalence volumes were calculated and printed on a teletype. The arrangement of the apparatus is shown in Fig. 2. 1 Microcomputer - Teletype Determination of End-points In a conventional titration, the system is arranged so that the addition of one drop of titrant brings about an easily observable change in colour or potential. I t is a well tried system, but it demands careful control of conditions or accurate measurement of potential. Weak acids, such as boric acid, and mixtures of acids, such as result from the oxidation of cyclohexane, are difficult to determine by direct titration, because it is not easy to ensure that the equivalence point and end-point are coincident.6 These difficulties may be overcome by using the whole titration curve for the determination of the equivalence points, but calculations by desk calculators are tedious and graphical interpolations are often inaccurate.The micro- processor permits all of the points on the titration curve to be used for computation of the equivalence points, performs the calculations as the titration is proceeding and displays the results at its conclusion. Accordingly, several procedures for determining equivalence points were examined. Results obtained from titrations of known mixtures of adipic acid and boric acid against standard sodium hydroxide solution were used for a preliminary investigation to determine the best method of calculating the equivalence points. Gran Plots For a weak acid, a plot of V x 10-PH against V , the volume of titrant, before the equivalence point, and (Vo + V ) x lopH against V after the equivalence point should result in' straight lines that intercept the V-axis at the equivalence volume (Vo is the original volume of acid).I t was found that the functions were linear only as the equivalence point was approached, and although the intercepts lay within the region of the equivalence volume, the results were not consistent enough for the required accuracy (see Fig. 3). Other related functions*-11 were also examined and similar results were obtained. Functions, developed by Gran,' were applied to the titration curves. Differential Plots The differential plot of ApH/AV against V proved to be the most satisfactory method for the determination of the equivalence points.This plot measures the rate of change of pH with volume and attains a maximum where there is a point of inflection in the titration curve. For a symmetrical acid - base titration, theoretically the equivalence point coincides with the point of inflection, but for asymmetrical titrations, the inflection point just precedes theJune, 1976 SEQUENTIAL TITRATION OF ADIPIC AND BORIC ACIDS 41 1 equivalence point.G For all practical purposes, however, it can be assumed that the equival- ence point, point of inflection and end-point are coincident. Titrations were initially performed by adding the titrant continuously and, as a result, the solutions never reached equilibrium. This effect was reflected in the differential plots as noise, which was reduced by adding the titrant in equal increments and recording the pH after each addition. This procedure has the added advantage that the volume is known accur- ately, whereas with continuous addition the volume added is dependent on the flow-rate and the timing loops of the microcomputer, which are independent variables. As the computation of a differential curve amplifies noise, a smoothing routine was included in the program of the microprocessor.Three smoothing techniques were examined: a running average of three or five, median smoothing and Hanning smoothing,12 of which the running average of three proved to be the most satisfactory. The effect of smoothing the differential curve containing background noise is shown in Fig.4. The differences required for the plot of ApH/AV against I' can be calculated directly from the voltage readings from the pH meter, conversion into the corresponding pH values being unnecessary. Provided that the titrant is added in equal increments, it is also unnecessary to divide the change in voltage reading by the change in volume, as the latter will be constant. The difference values are obtained by subtracting successive readings. The differential method gave the most consistent results. The automatic titrator, controlled by the microcomputer, was therefore designed using this method for the determination of the equivalence volumes. I -Volume of titrant/cm3 Volume of titrant/cm3 Fig. 4. Smoothing of the dif- ferential curve, containing back- ground noise: A, the unsmoothed curve : B, the smoothed differential Fig.3. Gran plots for adipic acid before and Theoretical equi- after the equivalence point. valence volume 4.64 cm3. of the smoothed titration curve. Experimental Hardware Analogue to digital converter and associated amplifier An Analogic 12-bit analogue to digital converter (ADC), Type MP2412, having a maximum range of -10 to +lo V, was used to convert the voltage output from the pH meter into a 2's complement binary number.13 The computer can then manipulate the numbers, which represent the pH readings, as required. The output, 0.5 V pH-1, was amplified in order that the whole range of the ADC would be utilised over the pH range of the titration. A variable gain amplifier was built such that the gain could be controlled by the microprocessor.Depend- ing on the logic level of two inputs to the amplifier, the gain was x 8, x 4, x 2 or x 1. Relay switch titrant. which resulted in the addition of one unit (0.1 cm3) of titrant. A reed relay switch, operated by the microprocessor, was used to control the addition of Closure of the switch short-circuited one of the output sockets of the Autoburette,412 Digital to analogue converter (DAC) them back into an analogue voltage. chart recorder. of -10 to +lo V, was used. BETTERIDGE et al.: APPLICATION OF MICROPROCESSORS TO THE AnnZyst, VoZ. I01 The data obtained by the ADC and stored bithe microcomputer are fed to a DAC to convert The voltage is used to drive an oscilloscope display or a An Analog Devices, 10-bit DAC, type DAC-102-3, having a maximum range Microcomputer The organisation of the microcomputer is shown in Fig.5, together with the associated hardware units. The console is provided as a program development aid and would not be present in the completed equipment. A CYBA-O* microcomputer system was used, contain- ing an Intel 8008 microprocessor. If the computer were dedicated to this particular application, the components, costing about &loo, would be a microprocessor, a read-only memory of 1K words, a read - write memory of 2K words and digital input - output. I/O Relay DAC Memory Fig. 5. The organisation of the microcomputer and asso- I/O, input/output ; ADC, analogue to digital ciated hardware. converter; Amp, amplifier; DAC, digital to analogue converter. Software The basic control and processing sequence for the automatic titrator is shown in Fig.6. Data collection Before any addition of titrant is made, the ADC reading is stored iiz line 0 of the memory. After the first addition, it is stored in line 1. Thus the value of L, the location in the store of the pH reading, gives the number of 0.1-cm3 additions of titrant which have been made. The microprocessor is programmed to close the relay switch, thereby adding an increment of titrant, to wait a pre-set time interval before triggering the ADC, and then to store the pH reading. When the number of aliquots of titrant required to complete the titration have been added, the addition of titrant is stopped., Variable time interval The necessary timing involved in the execution of the program is achieved by using loop routines.Such a routine is used to vary the time interval between the addition of titrant and triggering of the ADC, depending on the number of times the microprocessor is instructed to go through the loop. This enabled the effect of varying the rate of addition of titrant to be investigated. Diference and smoothing routines The difference routine subtracts the contents of the previous location of the memory from the present location and stores the result. The smoothing routine adds the contents of the present location to that of the two previous locations, and divides the result by four. This scaling is performed in order to prevent overflow errors in the computer and has no effect on the final equivalence volume.(In the microcomputer, division by two, and multiples of two, is simpler than division by three.13) Both the difference and smoothing routines are applied to the titration curve as the data are being collected. * CYBA-0 is the prototype of a microcomputer system developed by the Department of Electrical and Electronic Engineering, University College of Swansea, for Datalab, Mitcham, Surrey. The result is stored in the memory.J w e , 1976 SEQUENTIAL TITRATION OF ADIPIC AND BORIC ACIDS Clear rncmory Add increment of titrant fl ~~ [ Compute smoothed and difference values I Display current data I I Locate two maxima Check if they are within required limits 1 413 Fig. 6. Flow diagram of t h e basic control and pro- cessing sequence for t h e automatic titrator.The differences of the smoothed titration curve are also calculated and stored. Finally, these differences are smoothed and stored. Data display In order to monitor the titration, a display sequence is included in the data collection phase of the program. This enables the titration curve, its differential and the smoothed differential414 BETTERIDGE et al. : APPLICATION OF MICROPROCESSORS TO THE Analyst, VoZ. 101 curve to be displayed on an oscilloscope screen, while the titration is proceeding. Slight modification of the routine enabled the data to be output to a chart recorder. Peak detection When the titration is completed, the microprocessor locates the maxima in the differential curve, and in the smoothed differential curves, corresponding to the adipic and boric acid equivalence points.The maximum value of a set of data is first located by scanning the memory according to the flow diagram in Fig. 7. The second maximum is identified by scan- ning the data a second time, omitting the region that contains the first located maximum. The routine ignores the first 10 points of data, as the high values at the beginning of the differential curve are comparable with the boric acid maximum (see Fig. 4), and confusion could otherwise arise. 1 Select data to be scanned Set L = 10 t I Increment L Subtract contents of line L of memory from B. Load memory in L to register B I icya c i v c : \/ s Store value of L in register D I Store contents of B and D in memory Fig. 7. Flow diagram of peak detection routine.The maximum value is stored i n register B and its location in register D. Interpolation The locations of the maxima, and hence the equivalence volumes, are determined more precisely by performing the difference on the differential curve, i.e., calculating the rate of change of slope of the titration curve. The latter will change from a positive to a negativeJune, 1976 SEQUENTIAL TITRATION OF ADIPIC AND BORIC ACIDS 415 value as the titration curve passes through a point of inflection, and will be zero a t the actual equivalence point. Knowing the last positive value and the first negative value, it is a straight- forward arithmetical problem to determine the volume at which the double difference is zero, i.e., to determine the equivalence volume. outpzct The equivalence volunies of the adipic and boric acids are printed on a teletype, and the titration curve, its differential and the smoothed differential curves are output to a chart recorder.In this work an available teletype was used, but the programs could be arranged to suit any other convenient output devices. A ssenzbliag the program The program was written in the assembly code of the p r o c e ~ s o r ~ ~ ~ * J ~ and the machine code generated by an assembler program which runs on a 1904s computer; output is on paper-tape, which is then loaded into the memory of the microcomputer. An example of the assembler output is shown in Fig. 8. When the program works satisfactorily, it can be stored in the Read Only Memory (ROM), which then becomes part of the microcomputer system.ADDRESS 00 041 00 042 00 044 00 045 00 050 00 052 00 053 00 G56 00 060 00 061 00 063 OBJECT CODE 060 006 001 125 106 302 000 006 000 125 106 300 000 006 002 125 006 000 125 SOURCE STATEMENT COMMENT INL LA1 1 Trigger Relay OUT 10 for addition of CAL CLOCK Titrant. LA1 0 OUT 10 CAL TIME LA1 2 OUT 10 LA1 0 OUT 10 Wait time interval. Trigger ADC. Fig. 8. :In example of the assembler output. Titrator Assembly The automatic titrator comprised a Radiometer titrator, Type TTT2, with a dual end-point titrator module, Type PHASP1, for titration to two ddferent end-points, and a Radiometer Autoburette, Type ABU 13, fitted with a 25-cm3 burette was used. Reagents Adipic acid, ap~mximateZy 0.06 M. acid, minimum assay 99.99yo, in water. Boric acid, approximately 0.22 M.mum assay 99.5y0, in water. Sodium hydroxide solution, 0.500 M. MannitoZ. Prepared by dissolving Nylon 6 : 6 salt-grade adipic Prepared by dissolving AnalaR-grade boric acid, mini- AnalaR-grade material, minimum assay 99.0%. Procedure Titration to $re-set end-points For titrations to pre-set end-points, the automatic titrator has to measure the pH of the solution accurately. The instrument was therefore standardised, using buffer solutions, before any titrations were performed. The end-points were set at pH 6.6 for the adipic acid and 8.6 for the boric acid determination. A mixture comprising 20 cm3 of the standard adipic acid solution and 20 cm3 of the boric acid solution was placed in the titration vessel and titrated against the sodium hydroxide, contained in the Autoburette, until the first end-point416 BEITERIDGE et aZ.: APPLICATION OF MICROPROCESSORS TO THE AnaZyst, VoZ. 101 was reached. The equivalence volume of the adipic acid was recorded. About 12 g of man- nitol were then added to the solution in the titration vessel and the addition of titrant was continued until the second end-point was reached, when the total titre was recorded. The boric acid equivalence volume was obtained by difference. Automatic titrator using the microcomputer As the differential plot is used as the basis for the determination of the equivalence volumes, accurate measurement of pH is unnecessary. The apparatus was arranged as in Fig. 9. vessel / I I t- Microprocessor Oscilloscope :ifiLEEF' output Fig. 9. Arrangement of apparatus for titrations controlled For meaning of abbreviations see by the microcomputer.legend to Fig. 6 . In fact, the titrator had to be adjusted so that the amplified output from the titrator for the whole titration lay within the limits -10 to +lo V, the range of the ADC. By setting the amplifier for a gain of x 4, virtually the whole range was utilised. Samples comprising 20 cm3 of adipic acid solution and 20 cm3 of boric acid solution were again titrated against sodium hydroxide. The total titre required for the titration was about 14 cm3; a parameter in the program was set so that the titration stopped after 150 additions of 0.1 cm3. The titration was started by depressing a switch on the console of the microcomputer, and it automatically stopped after the 150 additions of titrant.The equivalence volumes of the adipic and boric acids were then printed on a teletype and the titration curve and its differential (smoothed if necessary) were output to a chart recorder (see Fig. 10). t I a t Boric ~ c- Volume of titrant/cm3 Fig. 10. Titration curve and its differential, both unsmoothed, for the titration of a mixture of adipic and boric acids against sodium hydroxide.J w e , 1976 SEQUENTIAL TITRATION OF ADIPIC AND BORIC ACIDS 41 7 For the purpose of statistical analysis of the results, the equivalence volumes are quoted to three decimal places. The true significance is evident from the statistical parameters. Similarly, €or comparison purposes only, a ‘ I theoretical” equivalence volume is quoted to three decimal places. This “theoretical” equivalence volume is derived from the masses of adipic and boric acids that were used to prepare the standard solutions, assuming lOOyo purity.The Autoburette digital volume readout measures to two decimal places only. Investigation of Experimental Variables discussed below. The effects of changing various experimental parameters were examined and the results are Orientation of titrant delivery tube It was observed that when titrant was ejected directly on to the electrodes, the pH meter tended to overshoot and swing slowly back. The differential curves under these conditions contained background noise. By directing the delivery hole away from the electrodes, the titrant could be thoroughly dispersed in the solution before reaching the electrodes, thereby eliminating the noise (see Fig.11). The direction of stirring and the position of the electrodes are also important in this respect and would affect the values of the equivalence volumes. A 6 Fig. 11. Effect of orientation of titrant delivery tube. A, Unsmoothed differential curve contained background noise (cf., Fig. 4) ; B, unsmoothed differential curve had little noise (cf., Fig. 10). S, stirrer: E, electrodes; TD, titrant delivery tube. With this arrangement, the smoothing routines were not required and were not used to obtain the results reported below. Stirring speed The effect of varying the speed of stirring was examined using a magnetic stirrer. With titrant being added at a slow rate, and a time interval of 5.5 s between an addition of titrant and triggering the ADC, six titrations were performed for each of three stirring speeds, “fast,” “medium” and “slow.” The mean equivalence volumes and their standard deviations are shown in Table I.The mean equivalence volumes and standard deviations were statistically examined16 using the t-test and F-test, respectively, with 95% confidence limits being taken as the level of significance. Statistical analysis showed that the slower the stirring the greater TABLE I EFFECT ON THE RESULTS OF VARYING THE RATE OF STIRRING Sample Equivalence volume Adipic acid Theoretical/cm3 hlean/cm3 Standard deviation* /cm3 Mean/cm3 Standard deviation*/cms Boric acid Theoretical/cm3 * Based on 6 results. Rate of stirring r 4.637 4.637 4.625 4.618 0.004 4 0.005 4 h Fast Medium 1 Slow 4.637 4.618 0.030 8 8.777 8.777 8.777 8.766 8.769 8.812 0.029 6 0.030 2 0.123 741 8 BETTERIDGE et al.: APPLICATION OF MICROPROCESSORS TO THE ANalyst, VoZ. 101 was the standard deviation, although the mean equivalence volumes were not significantly different. With slow stirring, the differential curves contained background noise. When adequate stirring was provided and the titrant delivery tube was directed away from the electrodes, the differential curves did not contain background noise, and hence the equival- ence volumes could be determined from the unsmoothed curves. Range of the ADC Under the normal conditions of the titration, roughly half the range of the ADC was used for the adipic acid portion of the titration curve.The voltage output from the pH meter was amplified so that virtually the whole range, -10 to +10 V, was utilised. This was used to decide whether the resolution of the ADC affected the accuracy and repeatability of the results significantly. Table I1 shows the mean equivalence volumes for adipic acid and the standard deviations. Ten titrations were performed for each scale. TABLE I1 EFFECT ON THE RESULTS OF INCREASING THE RANGE OF THE ADC Sample Adipic acid Equivalence volume Theoretical/cm3 Mean/cm3 Standard deviation* /cm3 Range Half scale Full scale A I -7 4.583 4.583 4.573 4.572 0.005 4 0.007 5 * Based on 10 results. There is no significant difference between the equivalence volumes or the standard deviations ; thus the discrimination of the ADC is not a limiting factor.T i m e interval between an addition of titrant and digitising voltage The time interval between an addition of titrant and triggering the ADC was varied by changing a parameter in the program of the microprocessor. At the fastest rate, time interval 0.35 s, addition of titrant was almost continuous and a titration was completed in approxim- ately 1 min. Six titrations were performed for each rate of addition. Table I11 shows the mean equivalence volumes for adipic acid and boric acid for each rate, the corresponding standard deviations, and the percentage errors from the “theoretical” equivalence volume. At the slowest rate, time interval 5.50 s, titrations took about 15 min. TABLE I11 EFFECT OF VARYING THE RATE OF ADDITION OF TITRANT Run A 7 I___._ 7- 1 2 3 4 5 6 7 8 Time interval/s r A \ Sample Equivalencevolume 0.35 0.45 0.70 1.20 1.80 2.10 3.00 5.50 Adipic Theoretical/cm3 4.639 acid Mean/cm3 4.837 4.772 4.733 4.667 4.651 4.627 4.614 4.619 Standard deviation*/cm3 0.048 3 0.015 1 0.015 2 0.011 4 0.017 3 0.010 3 0.021 1 0.010 7 Error, yo +4.3 +2.8 +2.15 +0.64 +0.21 -0.21 -0.64 -0.64 Boric Theoretical/cm3 8.774 acid Mean/cm3 8.758 8.725 8.729 8.787 8.834 8.764 8.768 8.764 Standard deviation*/cm3 0.083 2 0.055 2 0.038 5 0.052 2 0.070 9 0.030 4 0.026 6 0.020 5 Error, % -0.18 -0.56 -0.51 -0.08 $0.68 -0.11 -0.07 -0.11 * Based on 6 results.Different trends are observed for the mean equivalence volumes for adipic and boric acids. The former decrease exponentially as the time interval between the addition of titrant and the triggering of the ADC increases.They are significantly different from each other until the time interval is between 2.1 and 3.0 s and, apart from runs 5 and 6, they are all significantlyJune, 1976 SEQUENTIAL TITRATION OF ADIPIC AND BORIC ACIDS 419 different from the “theoretical” equivalence volume. The mean equivalence volumes for boric acid show no significant difference from each other (apart from a slight anomaly with run 5), nor are they significantly different from the “theoretical” value (apart from a slight anomaly again with run 5 ) . Even though the percentage errors of the mean equivalence volumes for adipic acid are comparable with those for boric acid, only the former are signifi- cantly different from the “theoretical” value. This is a consequence of the lower standard deviations of the mean equivalence volumes for adipic acid, and the greater volumes involved in the boric acid determination.For all practical purposes, however, the results are within the required limits. Comparison of the Automatic and Manual Methods Ten titrations were performed using the automatic system and ten by the manual method, titrating to pre-selected potentials. Titration by the second method included the manual addition of mannitol and took approximately 7 min. Using the automatic system, the length of the titration ranged from about 1 min upwards. Table IV shows the mean equivalence volumes and standard deviations of the two methods. A slow rate of addition of titrant (approximately 5 s between the addition and recording the voltage) was employed, although the same accuracy would have been obtained using a faster rate, as sh.own by the results in Table 111.TABLE I V COMPARISON OF THE AUTOMATIC AND MANUAL METHODS Sample Adipic acid Boric acid * Based on 10 results. Equivalence volume Theoretical/cm3 Meanlcm3 Standard deviation* /cm3 Error, yo Theoretical/cm3 Mean/cm3 Standard deviation*/cm3 Error, yo 7- Automatic 4.878 4.863 0.005 5 -0.31 8.738 8.761 0.018 8 +0.29 Method 7 -A- Manual 4.878 4.84 0.006 7 - 0.67 8.738 8.80 0.007 0 +0.73 The mean equivalence volumes for adipic acid are significantly different from each other and from the “theoretical” value. The corresponding standard deviations are not significantly different. The mean equivalence volumes for boric acid are also significantly different from each other and from the “theoretical” value. The standard deviations are also significantly different.For the adipic acid determination, lower standard deviations and percentage errors are obtained when using the automatic system. In the boric acid determination, although a lower percentage error is obtained when using the automatic system, a better standard deviation is obtained when using the manual system. Conclusion The results given by both systems are within the limits required for the analysis of the (i) The addition of mannitol for the boric acid determination is unnecessary. (ii) Titrations can be performed in less than half the time required for titration to pre-set end-points, with the same accuracy. The titration could be further speeded up, without any loss in precision, by adding titrant at a fast rate until a peak has been sensed, slowing the addition until the equivalence point has passed, then speeding up addition until the second peak is detected. (iii) Although only the equivalence volumes were printed on the teletype, the program would need little alteration in order to print the concentrations of the component acids.(iv) There is no need to calibrate the pH meter or to pre-set the equivalence potentials. (v) Provided that the voltage from the electrodes could be suitably amplified, it should be organic acid - boric acid mixture. The automatic system, however, has obvious advantages :420 BETTERIDGE, DAGLESS, DAVID, DEANS, PENKETH AND SHAWCROSS possible to connect them directly to the ADC, making the pH meter redundant with a con- siderable saving in cost.(vi) As the differential method does not require any prior knowledge of the end-points of the titration, the instrument should be capable of dealing with other logarithmic titrations. By modifying the program, it should also be capable of dealing with linear titrations, where extra- polation to determine the end-points can be difficult. The instrument is thus one of general applicability, the cost is much the same as that of the conventional system, it gives equally repeatable results and is far more convenient to operate. One of us (P.D.) thanks the Trustees of the Analytical Chemistry Trust Fund for the award of a maintenance grant. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. References Dessy, R. E., and Titus, J . A., Aqzalyt. Chem., 1974, 46, 294A. “An Introduction to Microprocessors,” Workshop Manual, Department of Electrical Engineering, Dagless, E. L., “Proceedings of Computer Technology Conference,” Institution of Electrical Dessy, R. E., Vauren, P. J., and Titus, J. A., Analyt. Chem., 1974, 46, 917A. Dessy, R. E., Vauren, P. J., and Titus, J. A., Analyt. Chem., 1974, 46, 1066A. Laitinen, H. A., “Chemical Analysis,” McGraw-Hill Book Co., New York, 1960, pp. 39-65. Gran, G., Analyst, 1952, 77, 661. Johansson, A., Analyst, 1970, 95, 536. Midgeley, D., and McCallum, C., Talanta, 1974, 21, 723. Ivaska, A., Talanta, 1974, 21, 1167. Ivaska, A., Talanta, 1974, 21, 377. Bendat, J. S., and Piersol, A. G., “Random Data: Analysis and Measurement Procedures,” Wiley- Lewin, D., “Theory and Design of Digital Computers,” Nelso? London, 1972, p. 15. Husson, S. S., “Microprogramming : Principles and Practices, Prentice-Hall, Englewood Cliffs, N. J ., Nicond, J. S.. Euromicvo Newsl., 1975, 1(3), 22. Davies, 0. L., and Goldsmith, P. L., “Statistical Methods in Research and Production,” Oliver and Received December 8th, 1976 Accepted January 22nd, 1976 University College of Swansea. Engineers, London, 1974, p. 64. Interscience, New York and London, 1971, p. 318. 1970. Boyd, Edinburgh and London, 1972.

ISSN:0003-2654    

DOI:10.1039/AN9760100409

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst, June, 1976, Vol. 101, $9. 421-432 421 Ionic Polymerisation as a Means of End-point Indication in Non-aqueous Thermometric Titrimetry Part VIII.” Polyfunctional Carboxylic Acids and Phenols? Solvent Effects in the Determination of E. J. Greenhow and A. A. Shafi Department of Chemistry, Chelsea College, University of London, Manresa Road, London, S W3 6LX A systematic study has been made of the effect of the solvent for the sample on the end-point sharpness and the reaction stoicheiometry in the catalytic thermometric titration of polyfunctional carboxylic acids and polyhydric phenols. The mechanisms proposed for the reactions involved in the neutralisation and indicator processes are based on the assumption that the titrants can exist as mixtures of ion pairs, solvent-separated ion pairs and free ions in the sample solution.It is suggested that the ratio of free ions to ion pairs is the factor that controls the reaction stoicheiometry, and the cation solvating power of the solvent is more important than the dielectric constant of the solution in influencing this ratio. Solvent and titrant combinations are recommended for the determination of the total acidities of a range of benzene carboxylic acids, polyhydric phenols and hydroxybenzoic acids by the catalytic thermometric procedure. Acrylonitrile was used as the end-point indicator. In an earlier evaluation of the catalytic thermometric method,l it was shown that, when acrylonitrile was the thermal indicator, resorcinol could be determined as either a mono- functional or a difunctional acid by using the appropriate titrant and solvent system.The higher stoicheiometry was obtained when the sample was dissolved in acrylonitrile alone and the titrant was potassium hydroxide in propan-2-01, while one acidic function was determined with tetrabutylammonium hydroxide in toluene - methanol solution as the titrant and a mixture of acrylonitrile and dimethylformamide as the solvent. Both acidic functions can be determined by using acetone, cyclohexanone2 or a mixture of acetone and acrylonitrile as the indicator/solvent and potassium hydroxide as the titrant. When the heat evolved from anionic polymerisation is used to locate the end-point, the shape of the titration curve in some determinations has been shown to be influenced by the solvent composition3 and the solubility of the ample.^ In an “ideal” titration, the deter- minative and indicator reactions occur in immediate succession.Ideal and non-ideal titration curves are shown in Fig. 1 ; Fig. 1 (b) is a typical curve obtained in the titration of solutions of the very weak acid sulphaguanidine in mixtures of dipolar aprotic solvents and acrylo- The rounded end-point inflection is probably caused by the overlap of the deter- minative and indicator reactions. Curve (c) was obtained in a determination of sulphanilamide, and is similar in shape to the ideal titration curve (a), but the end-point corresponds to a sub-stoicheiometric reaction. It has been suggested3 that the effective acidity of the sul- phanilamide is reduced during the course of the titration by cyanoethylation of the amide group.Curve (d) is typical of the curves obtained in the titration of some slightly soluble cat echo la mine^^ and shows that the anionic polymerisation is initiated almost immediately on addition of the titrant, but is then inhibited as the titrant/catalyst is consumed by the sample as the latter dissolves in the (now heated) solution, to which is being added increasing amounts of the solvent in the titrant, and a second inflection occurs at the true end-point. In the present investigation, a systematic study was made of the effect of the composition of the solvent for the sample on the measured reaction stoicheiometry when the sample was titrated with tetrabutylammonium hydroxide in a mixture of toluene and methanol or with potassium hydroxide in propan-2-01, The aim of the investigation was to establish the * For Part VII of this series, see Analyst, 1975, 100, 747.t Based in part on a paper presented a t a meeting of the Analytical Division of The Chemical Society, Loughborough, July 7th and 8th, 1975.422 GREENHOW AND SHAFI: IONIC POLYMERISATION FOR END-POINT Artalyst, Vol. 101 I I 0 $ 1 + Q 6 I- - 1) I I I /+ I I I I <T ,I I Titrandm1 Fig. 1. Catalytic thermometric titration curves. a, Ideal titration curve ; b, polymerisation retardation without complete inhibition ; c, limited inhibition of polymerisation ; and d, delayed inhibition of polymerisation. T , Expected stoicheiometric titre. optimum conditions for the determination of the total acidity of some representative organic acids, including acids in which the active functions are hydrogen bonded intramolecularly. Experimental Reagents Acetone, benzene, toluene, dioxan, pyridine and propan-2-01 were of analytical-reagent grade ; acrylonitrile and the other solvents were of laboratory-reagent grade.All reagents were dried over molecular sieve 4A before use. Potassium hydroxide in propan-2-01 and laboratory-reagent grade tetrabutylammonium hydroxide, 0.1 M in toluene - methanol, were standardised against benzoic acid (analytical- reagent grade) in dimethylformamide by the thermometric method. Polyfunctional Organic Acids Pyrocatechol, hydroquinone, resorcinol, quinhydrone, salicylic acid, phloroglucinol and pyrogallol were of analytical-reagent grade and the other sample compounds were of laboratory-reagent grade.Apparatus The automatic apparatus described in Part 1115 was used, but with unsilvered Dewar beakers (capacities 14 and 30 ml) instead of the titration flasks insulated with polystyrene foam, so that precipitations and colour changes could be observed. Procedure Prepare a solution of the sample compound in the most effective solvent of the solvent mixture to be used, pipette an aliquot of the solution into the titration beaker and add the acrylonitrile and any other solvents. Introduce the titrant at a constant rate (usually 0.2 ml min-1) and record the temperature and titrant volume on a millivolt chart recorder (100-rnV scale) at a chart speed of either 600 mrn h-l or 30 mrn min-l, depending on the concentration of the titrant.Results and Discussion In Fig. 2 are shown some of the titration curves obtained when acrylonitrile and mixtures of acrylonitrile and aprotic solvents are treated with 0.1 and 0.5 M potassium hydroxide and 0.1 M tetrabutylammonium hydroxide solution. When the potassium hydroxide titrants are used, it is necessary to add a significant volume of the titrants to acrylonitrile, mixtures of acrylonitrile with benzene and a 3 + 1 V/V mixture of acrylonitrile with pyridine beforeJune, 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART VIII 423 the end-point is indicated. With mixtures that contain higher proportions of pyridine, e.g., 2 + 2 and 1 + 3 V/V, the temperature increase occurs on addition of the first drops of titrant. When dimethylformamide or dimethyl sulphoxide is included in these solvent mixtures, even in proportions as low as 1 + 8 V/V, not only is the end-point inflection much sharper but also the blank titration value is greatly reduced [cf., Fig.2(a) and (b), and 2(c) and (d)]. b C P Titrant/mI (1 division=l ml) Fig. 2. Effect of solvent composition on the shape of catalytic thermometric titration curves. Solvent*/ml and titrantt/M: a, A 4 and K 0.1; b, A 4 + D 1 and K 0.1; c, A 2 + B 2 and K 0.1; d, A 2 + B 2 + D 1 and K 0.1; el A 4 and K 0.5; f , A 3 + B 1 and K 0.5; g, A 2 + B 2 and K 0.5; h, A 1 + B 3 and K 0.5; j, A 3 + P 1 and IC 0.1; k, A 3 + P 1 and K0.5; m, A 2 + P 2 a n d K0.5; n, A 3 + P 1 + D 0 . 5 a n d K O . l ; o , A 4 a n d Q O . l ; p , A 2 + B 2 a n d Q 0.1; q, A 3 + P 1 and Q 0.1.* A, acrylonitrile; D, dimethyl- formamide; B, benzene; P, pyridine. t K, potassium hydroxide; Q, tetrabutylammonium hydroxide. When 0.1 M tetrabutylammonium hydroxide is used as the titrant, the end-point inflections are consistently sharp in the titrations of all of the solvent mixtures examined, and the blank titration values approximate to zero [Fig. 2(0)-(q)]. A comparison of Fig. 2(a) with 2(e), 2(c) with 2(g) and 2(j) with 2(k) shows that the effect of changing from the 0.1 to the 0.5 M potassium hydroxide titrant is approximately to halve the titration volume, and not to reduce it to one fifth, as might have been expected. Apparently the blank titration values are influenced not only by the concentration of the potassium hydroxide but also by the volume of the propan-2-01 introduced into the titration mixture.The effect of adding climethylformamide or dimethyl sulphoxide to the solvent mixtures can be explained if one assumes that the position and sharpness of the end-point inflections are influenced by the rate of the ionic polymerisation of the acrylonitrile. The rate of ionic polymerisations has been shown to depend on ( a ) , the dielectric constant of the monomer solution and ( b ) , the degree of ionisation of the polymerisation catalyst. Factors (a) and (b) are related in that, in general, the higher the dielectric constant of the solvent the more readily the catalyst ionises. Tetrabutylammonium hydroxide has a bulky cation and will tend to ionise completely to yield hydroxyl ions; these ions will be unsolvated unless the solvent contains anion-solvating components.In contrast, potassium hydroxide would be expected to exist as free ions, ion pairs and solvent-separated ion pairs in non-aqueous solvent systems, and the ratio of free ions to ion-pair species would depend on the solvent system. The propan-2-01 solution of potassium hydroxide contains a significant proportion of propan-2-oxide ions as well as hydroxyl ions,* and both anions are involved in the polymerisation process.424 GREENHOW AND SHAFI : IONIC POLYMERISATION FOR END-POINT AnaZyst, VoZ. 101 It has been established' that free ions are much more effective than ion pairs as catalysts in anionic polymerisation, which explains why a sharp and immediate temperature increase occurs when tetrabutylammonium hydroxide is used as the titrant for the acrylonitrile - solvent mixtures, irrespective of the nature of the solvent.The sharp increases in temperature achieved with the potassium hydroxide titrant when dimethylformamide, dimethyl sulphoxide or pyridine (in excess) are present in the solvent mixture can be explained in a similar manner. Thus, dimethylformamide and dimethyl sulphoxide have high dielectric constants and, in addition, are known to solvate the potassium ion and to liberate unsolvated anions,* and it is these anions that cause the immediate polymerisation and accompanying temperature increase. Compared with dimethylformamide and dimethyl sulphoxide, pyridine has a low dielectric constant (E = 12.3) and its addition to acrylonitrile (E = 33) would lower the dielectric constant of the mixture; its effect on the end-point inflection when it is used as a diluent for the acrylonitrile must, therefore, be due to some other characteristic property of pyridine, probably its electron-pair donicity (D = 33.1),9 which is considerably higher than that of acetonitrile (D = 14.1) and is likely to be higher than that of acrylonitrile.Solvents of high donicity, such as pyridine, should solvate potassium ions and thus cause the release of unsolvated hydroxyl and propan-2-oxide ions, although pyridine is, apparently, not as effective as dimethylformamide (D = 26.6) as an ionising solvent. When resorcinol in solution in mixtures of acrylonitrile with dimethylformamide, and some other dipolar aprotic solvents, is determined by the catalytic thermometric method, using the 0.5 XI potassium hydroxide titrant, the measured reaction stoicheiometry is dependent on the composition of the sample solvent (Fig.3). I t can be seen that resorcinol is determined ! I O L - d - - L - - L - - - - 4.0 3.5 3.0 2.5 2.0 1.5 1.0 Acrylonitri le/mI Solventh I Fig. 3. Solvent effects in the titration of 0.1 mmol of resorcinol (0.5 M potassium hydroxide reagent). Solvents : a, acetone ; b, pyridine ; c, sulpholane; d, hexamethylphos- photriamide ; e, 1-methyl-2-pyrrolidone ; f, NN-dimethyl- acetamide ; g, dimethylformamide ; h, dimethyl sulphoxide. as a dibasic acid if the solvent for the sample contains only a small amount of the dipolar aprotic solvents, e.g., about 10% V/V, and, for all of the solvents except sulpholane, as a mono- basic acid when the content of this solvent is high, e.g., about 75% V/V.An explanation for this effect is that the higher the proportion of the dipolar aprotic solvent, the greater the degree of solvation of the potassium ion, the higher the concentration of the unsolvated hydroxyl and propan-2-oxide ions and the more readily the ionic polymerisation occurs. Thus, when the dipolar aprotic solvent is present in excess, the polymerisation is initiated almost immediately after the stronger of the two phenolic hydroxyl groups has been neutralised, while when the solvent mixture contains only a small amount of the dipolar aprotic componentJ m e , 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY.PART VIII 425 the second phenolic hydroxyl group is completely neutralised before the potassium hydroxide, essentially in the ion-pair form, initiates the polymerisation. This mechanism implies that the ion-pair form of the titrant combines readily with the weakly acidic second hydroxyl group of the resorcinol. A possible reaction route is one that involves a four-centre inter- mediate : HE+ - AS- K+OH-+HA 3 : : -+ K+A-+H,O HO- - K+ The results in Fig. 3 indicate that, of the dipolar aprotic solvents examined, dimethyl sulphoxide is the most effective for solvating the potassium ion, dimethylformamide, dirnethyl- acetamide and 1-methyl-2-pyrrolidone are similar in their solvating powers and sulpholane is the least effective. The dielectric constant of sulpholane is high (E = 42.0) but its donicity (D = 14.8) is low compared with that of dimethylformamide and dimethyl sulphoxide (D = 29.8).Included in Fig. 3 are the results obtained with acetone and pyridine as the solvents. These results show that both pyridine and acetone are inferior to the dipolar aprotic solvents in terms of their ability to release free anions for the anionic polymerisation, because the reaction stoicheiometries approximate to 2, irrespective of the acetone or pyridine contents of the mixtures with acrylonitrile. A comparison of titration curves obtained with acetone alone and mixtures of acetone and acrylonitrile as the indicator solvents shows clearly that with the mixtures acrylonitrile is an effective thermometric indicator. Stoicheiometry versus solvent composition graphs for salicylic acid, pyrocatechol, hydro- quinone, 2,2’-dihydroxybiphenyl and 2,2’-dihydroxydiphenylmethane, shown in Fig.4, are similar to those obtained with resorcinol as the sample. The solvent was a mixture of acrylo- nitrile and dimethyl sulphoxide. All of the reaction stoicheiometries obtained with tetra- butylammonium hydroxide as the titrant approximated to 1 , as shown by the stoicheiometry versus solvent composition graph for this titrant ; only one line is shown because all five almost coincide. 0 I I ’ . - - L . _ _ L 2.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Acrylonitri I e/ml 2.0 0.5 0.5 0.5 0.5 0.5 0.5 0.5 L- I I I - - J Dimethyl sulphoxide/ml Fig. 4. Solvent effects in the titration of (a) salicylic acid, (b) pyrocatechol, (c) 2,2’-dihydroxydiphenylmethane, (d) hydro- quinone and (e) 2,2’-dihydroxybiphenyl (0.5 M potassium hydroxide reagent); f , compounds a-e titrated with 0.1 hi tetrabutylammonium hydroxide.0.05-mmol samples. The acid functions of all the compounds, except those of hydroquinone, can undergo intramolecular hydrogen bonding. This bonding develops its maximum strength when the 0 - H * - * 0 bonds are collinear, which is possible without requiring the straining of bonds426 GREENHOW AND SHAFI : IONIC POLYMERISATION FOR END-POINT Analyst, VoZ. 101 only with the 2,2’-dihydroxybiphenyl and 2,2’-dihydroxydiphenylmethane. The hydrogen bonding causes one of the two acidic hydrogen atoms to be more acidic than the other,lO the difference in acidities being dependent on the strength of the hydrogen bond.The relative values of the measured reaction stoicheiometries of the five compounds shown in Fig. 4 are generally in accord with the acid strengths expected if intramolecular hydrogen bonding occurs. Thus, one of the acidic functions in 2,2‘-dihydroxybiphenyl and 2,2’-di- hydroxydiphenylmethane is easily titrated but the second is more difficult to titrate than are the second, weaker, acidic functions in pyrocatechol and salicylic acid (Figs. 8 and 10 and ref. 11). The low acidity of the second acidic function of hydroquinone cannot be ex- plained in terms of intramolecular hydrogen bonding. The reaction stoicheiometries obtained when the tetrabutylammonium hydroxide is used are as expected if one accepts that this titrant is highly dissociated in all of the solvent mixtures.In determinations with potassium hydroxide reagent as the titrant, the solvent system at the end-point will contain a significant amount of propan-2-01 if a significant amount of titrant is consumed, e.g., the final solution will contain 10% V/V of propan-2-01 when both acidic functions in a 0.025 M solution of resorcinol in the solvent - acrylonitrile mixture are titrated with the 0.5 M reagent. The effect of propan-2-01 on the sharpness and position of the end-point is, therefore, of some importance. The effect on “blank” titrations and on titrations of benzoic acid and resorcinol, when propan-2-01 is included in the solvent mixtures, is shown in Fig. 5. When propan-2-01 is present in a ratio of about 1 : 4 with solvents that have significant blank titration values, e.g., acrylonitrile, mixtures of acrylonitrile with benzene and 3 + 1 mixtures of acrylonitrile with pyridine, the effect is to reduce the blank titration value to zero and to sharpen the end-point inflection (Figs.2 and 5). As the proportion of propan-2-01 is increased, the inflection marking the end-point becomes less steep but the titration value remains at about zero [Fig. 5(a) and (b)]. t 0.1 M potassium hydroxide reagent/ml (1 division=l ml) Fig. 5. Effect of propan-2-01 on the shape of catalytic thermometric titration curves (0.1 M potassium hydroxide reagent). Solvents*/ml: a, A 3 + Pr 1 ; b, A 1 + P r 3 ; c, A 2 + Pr 1 + B 1; d, A 1 + Pr 1 + B 2 ; e, A 2 + P r 2 + P 1 ; f , A 2 + Pr 1 + P 2; g, A 3 + Pr 2 + D 1; h, A 1 + Pr 3 + D 1.Benzoic acid titrations (0.1 mequiv): j , A 5; k, A 4 + D 1; 1, A 4 + Pr 1; m, A 1 + D 4 ; n, A 1 + Pr 1 + D 3; 0, A 1 + Pr 2 + D 2. Resorcinol titrations (0.1 mequiv): p, A 1 + D 4 ; q, A 5; r, A 4 + Pr 1; s, A 1 + Pr 1 + D 3; t, A 1 + Pr 4. * A, acrylonitrile; Pr, propan-2-01 ; B, benzene ; P, pyridine ; D, dimethylformamide. It can be seen in Fig. 5(1), (t) and (r) that when propan-2-01, mixed with acrylonitrile, is used as the solvent for the sample in the catalytic thermometric titrations of benzoic acid and resorcinol, the temperature increase at the end-point is less steep in most instances than when the sample solution contains dipolar aprotic solvents [Fig. 5(k), (m) and (p)]. The propan-2-01 in the sample solutions also influences the temperature changes in the deter-June, 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY.PART VIII 427 minative stage of the titration. Thus, there is an increase in temperature at this stage if propan-2-01 is the solvent for the sample, while the reverse effect obtains when dimethyl- formamide, or another dipolar solvent, is used instead. With appropriate mixtures of propan- 2-01 and dimethylformamide as the solvent for the sample, it is possible to obtain titration curves in which there is very little temperature change during the course of the determinative reaction [Fig. 5(n) and (s)]. The presence of propan-2-01 does not appear to influence the position of the end-point in the titration of benzoic acid; in the titration of resorcinol, however, the indicated end-point corresponds to a high reaction stoicheiometry (approaching 2) when propan-2-01 is the solvent for the sample.This result suggests that propan-2-01 does not solvate the potassium ion, if one accepts cation solvation as the explanation for the effect of the sample solvents on reaction stoicheiometry. The effect of propan-2-01 on the temperature changes in the determinative stage of the titration can be attributed to hydrogen-bonding reactions that involve the hydroxyl group of this solvent. The knowledge gained in this systematic study of solvent effects has been used to devise solvent - titrant combinations for use in the determination of the acidic functions in some representative polycarboxylic acids, polyh ydric phenols and hydroxybenzene carboxylic acids.Typical titration curves are shown in Figs, 6-10. The reported dissociation constants (pKa) for the ten benzene carboxylic acids titrated range from 1.40 to 6.96. Thus, in aqueous solution, even the weakest carboxylic acid group is far more acidic than phenol, which has a PI<, value of 9.9. As phenol can be determined without difficulty by non-aqueous catalytic thermometric titrimetry, it might be expected that all of the carboxylic acid functions in the ten acids examined should be titratable by this method. However, the total acidity of only five of the acids, namely benzoic acid, the three benzene dicarboxylic acids and benzene-l,3,5-tricarboxylic acid, could be determined when 0.1 M tetrabutylammonium hydroxide was used as the titrant, even though no pre- cipitation occurred during the titrations.All of the acids that were titrated incompletely when the end-point was indicated contain adjacent carboxylic groups. The possibility arises, therefore, of the low reaction stoicheiometries being related in some way to the partial neutralisation of adjacent groups. Phthalic acid is exceptional in being the only acid with adjacent groups to be completely titrated by the quaternary hydroxide. The end-point inflections in all of the titrations were acceptably sharp and unambiguous (Fig. 6). 0.1 rJ tetrabutyiarnrnoniurn hydroxide reagent/rnl (1 division = 1 rnl) Fig. 6. Titration curves for benzene carboxylic acids (0.1 M tetrabutylamnionium hydroxide reagent). Acids/mmol and sol- vents*/ml (with reaction stoicheioinetry in parentheses) : a, benzoic 0.1 and A 4 + D 1 (1.0); b, phthalic 0.05 and A 4 + D 0.5 (2.0); c, isophthalic 0.05 and A 4 + D 0.5 (2.0) ; d, terephthalic 0.05 and A 4 + D 1 (2.0); e, 1,2,3-tricarboxylic 0.05 and A 4 + D 1 (2.3); f, 1,2,4-tricarboxylic 0.05 and A 4 + D 1 (2.4); g, 1,3,5-tricarboxylic 0.05 and A 4 + D 1 (3.0); h, 1,2,4,5-tetracarboxylic 0.05 and A 4 + D 1 (3.0); j, pentacarboxylic 0.05 and A 4 + D 1 (4.0); k, hexa- carboxylic 0.05 and A 2.5 -+ D 3 + M 1 (4.2).* A, acrylonitrile; D, dimethylformamide; M, methanol.428 GREENHOW AND SHAFI : IONIC POLYMERISATION FOR END-POINT Analyst, VoZ. 101 The acidic functions in nine of the ten acids can be titrated with 0.5 M potassium hydroxide reagent if an appropriate solvent system is used (Fig.7). Mellitic acid, the hexacarboxylic acid, is insoluble in acrylonitrile and the dipolar aprotic solvents, and it was necessary to incorporate a high proportion of methanol in the solvent mixture in order to achieve a clear solution. The titration curve [Fig. 7(k)] resembles those obtained when propan-2-01 was a major solvent constituent, ie., there is a marked temperature increase during the neutralis- ation step. The low measured reaction stoicheiometry, 5.6, may be due to the partial methylation of the acid. In the titrations of terephthalic acid, benzene-l,3,5-tricarboxylic acid and benzene-1,2,4,5-tetracarboxylic acid [Fig. 7(d), (g) and (h)], smaller inflections that occur before the final temperature increase correspond to the titration of more acidic func- tions.The fourth group in the 1,2,4,5-tetracarboxylic acid is particularly difficult to titrate in non-aqueous solution, and the first inflection in the titration curve [Fig. 7(h)], corresponding to the neutralisation of three acidic groups, is much sharper than the final inflection, which resembles that of Fig. l(b), and may be due to retardation of polymerisation. 0.5 M potassium hydroxide reagent/ml (1 division=l ml) Fig. 7 . Titration curves for benzene carboxylic acids (0.5 M potassium hydroxide reagent). Acids* and reaction stoicheio- metryt: a, benzoic 1.0; b, phthalic 2.0; c, isophthalic 2.0; d, terephthalic 2.0(1) ; e, 1,2,3-tricarboxylic 2.8; f, 1,2,4-tricarboxylic 3.0; g, 1,3,5-tricarboxylic 3.0( 1) ; h, 1,2,4,5-tetracarboxylic 4.0(3) ; j, pentacarboxylic 5.0; k, hexacarboxylic 5.6.The end-points are denoted by arrows. * 0.1-mmol samples, except for h (0.05 mmol). t Values in parentheses are reaction stoicheiometries indicated by the inflections preceding the final end-point. Solvents: a-e, g and j, 4 ml of acrylonitrile + 1 ml of dimethylformamide; f, 4 ml of acrylonitrile + 1 ml of l-methyl-2-pyrrolidone; k, 5 ml of acrylo- nitrile + 0.7 ml of dimethyl sulphoxide + 1 ml of methanol; h, 8 ml of acrylonitrile + 1.7 ml of sulpholane -t- 0.25 ml of dimethyl- formamide. When polyhydric phenols are titrated with potassium hydroxide reagent, the titration curves (Fig. 8) show a variation in the sharpness of the end-point inflection, which depends on the location of the hydroxyl groups.Resorcinol can be titrated without difficulty if a titration solvent containing a high proportion of acrylonitrile is used, and the titration curve [Fig. 8(b)] shows a single inflection corresponding to the neutralisation of both hydroxyl groups. If air is not excluded from the apparatus, colour changes occur during the titration of the polyhydric phenols. Streuli12 has attributed such colour changes (observed in non- aqueous potentiometric titrimetry) to oxidation processes involving the ionised phenol molecules. This oxidation process was found to affect the sharpness of the end-point in- flection in some of the catalytic thermometric titrations, particularly in that of pyrogallol, and the end-point inflection was improved by de-oxygenating the solvents and passing nitrogen through the titration vessel.The lack of sharpness in the final inflections of the titration curves for pyrocatechol and 2,2’-dihydroxybiphenyl [Fig. 8( a) and (h)] can be partly attributed to the intramolecular hydrogen bonding of the hydroxyl groups ; the sharper intermediate inflections correspond to the neutralisation of the stronger acidic functions. The titrationJune, 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART VIII 429 curve for 2,2’-dihydroxydiphenylmethanel1 is similar to that of 2,2’-dihydroxybiphenyl. Intra- molecular hydrogen bonding is less in the tetrahalogenated pyrocatechols and 3,5,6,3’,5‘,6’- hexachloro-2,2‘-dihydroxydiphenylmethane [Fig. 8(f), (g) and (j)] than in the unsubstituted phenols, and both acidic functions can be titrated without difficulty.!---------- -I Potassium hydroxide reagent/ml ( 1 divisionxl ml) Fig. 8. Titration curves for polyhydric phenols (0.6 and 1.0 M potassium hydroxide reagent). Compounds/mmol and solvents* /ml (with reaction stoicheiometry in parentheses?) : a, pyrocatechol 0.2 and A 16 + S 2 (2.0;l); b, resorcinol 0.1 and A 3.5 + S 0.5 (2.0) ; c , hydroquinone 0.2 and A 20 + S 2 (2.0;l) ; d, phloroglucinol 0.05 and A 4 + D 0.5 (3.0) ; e, pyrogallol 0.05 and A 10 + D 1 (3.0); f, tetrachloropyrocatechol 0.1 and A 4 + D 1 (2.0); g, tetrabromopyrocatechol 0.1 and A 4 + S 1 ( 2 . 0 ) ; h, 2,2’-dihydroxybiphenyl 0.1 and A 12 + S 1.5 (2.0; 1.2); j, 3,5,0,3’, 5‘, 6’-hexachloro-2,2’-dihydroxydiphenylmethane 0.05 and A 3 + D 1 ( 2 . 0 ) ; k, 2,2-bis(4-hydroxyphenyl)propane 0.1 and A 4 + D 1 (2.0).The end-points are denoted by arrows. a and c, 1.0 M reagent. * A, acrylonitrile; S, dimethyl sul- phoxide; D, dimethylformamide. t The second values in parentheses are the stoicheiometries indicated by the inflections preceding the final end-points. Although intramolecular hydrogen bonding is not possible in hydroquinone, the titration curve resembles that of pyrocatechol, i.e., the final end-point is indistinct and the intermediate inflection corresponds to the titration of one hydroxyl group. A possible explanation for this behaviour is that the second hydroxyl group is resonance-stabilised by the anion formed a t the half-neutralisation stage. In the quinonoid canonical form : H O G 0 the negative charge adjacent to the hydroxyl group would oppose ionisation of the hydroxyl hydrogen.The weak acidity of the second hydroxyl group in pyrocatechol could be explained in a similar manner. In this instance, the stabilising canonical form would be: Go OH The quinonoid structures are not possible with resorcinol, which has a more acidic second hydroxyl group than the other dihydric phenols. Both acidic groups in hydroquinone, but only one in pyrocatechol, can be determined when acetone is used as the solvent and thermo- metric indicator.13 An acetone - hydroquinone complex, which enhances the acidity of the430 GREENHOW AND SHAFI : IONIC POLYMERISATION FOR END-POINT Analyst, VoZ. 101 latter, may have been formed. Neither intramolecular hydrogen bonding nor resonance stabilisation is likely in 2,2-bis(4-hydroxyphenyl)propane, and the single sharp end-point inflection corresponds to the titration of both hydroxyl groups.Precipitates form during the titrations of the hydrogen-bonded phenols with the potassium hydroxide reagent, and the slower rate of reaction of the resulting heterogeneous mixture may be partly responsible for the overlap of the determinative and indicator reactions that leads to the indistinct final inflections in some of the determinations. The sharpness of the final inflections can be improved by using more dilute sample solutions and more concentrated titrants, e.g., 1.0 instead of 0.5 M potassium hydroxide solution, although a decrease in precision is a limiting factor in these variations. Mono-, di- and trihydroxybenzoic acids were titrated with tetrabutylammonium hydroxide (Fig.9) and potassium hydroxide (Fig. 10) reagents. The end-points obtained when the former titrant is used correspond, in all but one of the titrations, to incomplete neutralisation of the total acidity, even though the solvent mixture contains a high proportion of acrylo- nitrile and there is no precipitate at the end-point. The results are similar, therefore, to those obtained in the titration of the benzene carboxylic acids with tetrabutylammonium hydroxide. The 2-hydroxy- and 2,6-dihydroxybenzoic acids are determined as monofunctional acids, and the 2,4-, 3,4- and 3,5-dihydroxy- and 2,3,4- and 3,4,5-trihydroxybenzoic acids are determined as difunctional acids. Both acidic functions in the 4-hydroxy acid are titrated and the 3-hydroxy acid gives a measured reaction stoicheiometry of 1.7.It is significant that intramolecular hydrogen bonding is unlikely to occur in these last two acids but can occur in the other hydroxy acids, with the notable exception of 3,5-dihydroxybenzoic acid. I I I I I I 0.1 M tetrabutylammonium hydroxide reagent/ml Titration curves for hydroxybenzoic acids (0.1 hi tetrabutylammonium hydroxide reagent). Benzoic acids and reaction stoicheiometry : a, 2-hydroxy 1.0; b, 3-hydroxy 1.7; c, 4-hydroxy 2.0; d, 2,4-dihydroxy 1.9; e, 2,6-dihydroxy 1.0; f, 3,4-dihydroxy 2.0; g, 3,5-dihydroxy 1.9; h, 2,3,4-trihydroxy 2.0; j, 3,4,5-trihydroxy 2.0. Solvent: 4 ml of acrylonitrile + 0.5 ml of dimethylformamide. Samples: 0.05 mmol. (1 division = 1 ml) Fig.9. All of the acidic functions in these hydroxybenzene carboxylic acids are determined by titration with the 0.5 M potassium hydroxide reagent when a suitable solvent for the sample, containing a high proportion of acrylonitrile, is used. As in the determinations of the benzene carboxylic acids, some of the final end-point inflections are indistinct and are preceded by sharper inflections corresponding to the neutralisation of stronger acidic functions. These intermediate inflections are noticeable in the titrations of 2,6- and 3,4-dihydroxy- and 2,3,4-tri- hydroxybenzoic acids [Fig. lO(e), (f) and (h)]. Unlike the quaternary ammonium salts, the potassium salts of the hydroxybenzoic acids are precipitated from solution during the titra- tions and the lower reactivity of the precipitates could be responsible for the lack of sharpness in some of the end-point inflections. Colour changes similar to those seen in the titration of the polyhydric phenols occur during the titration of di- and trihydroxy acids if air is not excluded.The end-point inflection for the 3,4,5-trihydroxybenzoic acid, in particular, was improved by maintaining a nitrogen atmosphere in the apparatus. Although the total acidities of most of the polyfunctional acids have been determined byJzcne, 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART VIII 431 using 4 + 1 or 8 + 1 mixtures of acrylonitrile with dimethylformamide or dimethyl sulph- oxide as the solvents, in some titrations with 0.5 and 1.0 M potassium hydroxide reagents other solvent mixtures were found to lead to sharper end-point inflections.Thus, l-methyl- 2-pyrrolidone and mixtures of dimethylformarnide with dimethyl sulphoxide and sulpholane have been used to obtain improved titration curves for 1,2,4-benzenetricarboxylic acid, 1,2,4,5-benzenetetracarboxylic acid, 2,6-dihydroxybenzene carboxylic acid and 2,3,4-tn- hydroxybenzene carboxylic acid. In the determinations of the two hydroxybenzoic acids, the addition of a small amount of water to the solvent mixture (1 + 75) further improved the end-point sharpness [Fig. lO(e) and (h)]. Potassium hydroxide reagent/ml (1 division = 1 ml) Fig. 10. Titration curves for hydroxybenzoic acids (0.5 M potassium hydroxide reagent). Benzoic acids* and solventst /ml (with reaction stoicheiometry in parentheses:) : a , 2-hydroxy and A 4 + S 0.5 (2.0;l) ; b, 3-hydroxy and A 4 + D 1 (2.0) ; c, 4-hydroxy and A 4 + D 1 (2.0; 1) ; d, 2-4-dihydroxy and A 4 + D 1 (3.0); e, 2,6-dihydroxy and A 11 + S 4 + D 1 + H,O 0.2 (3.0); f , 3,4-dihydroxy and A 10 + D 1 (3.0;2); g, 3,5-dihydroxy and A 4 + D 1 (3.0); h, 2,3,4-trihydroxy and A 11 + S 4 + D 1 + H,O 0.2 (3.9;3;1); j, 3.4,ti-trihydroxy and A 5 + D 1 (4.0).The end-points are denoted by arrows. * a-c and e-g, 0.1 mmol; d, h and j, 0.05 mmol. t A, acrylonitrile; S, dimethyl sul- phoxide; D, dimethylformamide. $ The second and third values in parentheses are the stoicheiometries indicated by the inflections pre- ceding the final end-points. The shapes of the titration curves in which sharp and less sharp inflections occur give some indication of the relative strengths of the stronger and weaker acidic functions in the analytes.A clearer indication of the difference in acidity between acidic functions in the same compound can usually be obtained from a comparison of the reaction stoicheiometries measured by using the tetrabutylammonium and potassium hydroxide titrants, e g . , by comparing the results shown in Figs. 6 and 7 and in Figs. 9 and 10. The titrimetric procedures most widely investigated for the determination of polyhydric phenols and hydroxybenzoic acids are those employing potentiometric end-point detection. Both hydroxyl groups in pyrocatechol, hydroquinone and resorcinol, and the hydroxyl and carboxyl groups in salicylic acid, have been determined by using an antimony indicator electrode in conjunction with sodium aminoethoxide as the titrant and lJ2-diaminoethane as the solvent.14-16 The results obtained when the more conventional glass indicator electrode was used are not entirely in agreement.Thus, Deal and Wyldl’ were able to titrate both acidic functions of pyrocatechol and hydroquinone, in dimethylformamide solution, with tetrabutylammonium hydroxide, but Cundiff and MarkunuslS reported that only one acidic group in hydroquinone and resorcinol could be determined by this procedure. Later, Streuli and Mironls and Streuli,12 using pyridine as the solvent for the sample, confirmed the latter findings and showed also that only two of the three hydroxyl groups in phloroglucinol could be determined. Streuli suggested that the results implied that “two ionisations in a single molecule’’ are difficult to achieve in a solvent with a low dielectric constant, such as pyridine.Our titration results with solutions of resorcinol in solvent mixtures containing pyridine and432 GREENHOW AND SHAFI the dipolar aprotic solvents of high dielectric constant (Fig. 3) differ from Streuli’s results in that the solvent mixture with the lower dielectric constant leads to the higher reaction stoicheiometry. In their investigation of 4-methyl-2-pentanone as a solvent for the sample, Bruss and Wyld20 showed that both acidic functions in salicyclic acid were determined by titration with potassium hydroxide in propan-2-01, while only the carboxylic acid was titrated with tetrabutylammonium hydroxide.Their results are more in accord with our own than is Streuli’s hypothesis. However, it is difficult to compare solvent effects in potentio- metric titrimetry with those in catalytic thermometric titrimetry because the mechanisms of the processes that occur at the end-points are different. Although attempts have been rnadel7y2I to titrate the strongly hydrogen-bonded compounds 2,Y-dihydroxybiphenyl and 2,2’-dihydroxydiphenylmethane and some of their derivatives by the potentiometric method, none appears to have been successful in determining the second, weaker, hydroxyl group. In general, it may be claimed that catalytic thermometric titrimetry is faster and less affected by precipitates than is potentiometric titrimetry. In addition, it is able to determine some weakly acidic functions that have not previously been determined by the latter procedure. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. References Greenhow, E. J., and Hargitt, R., Proc. SOC. Analyt. Chem., 1973, 10, 276. Greenhow, E. J., Chemy Ind., 1974, 456. Greenhow, E. J., and Spencer, L. E., Analyt. Chem., 1975, 47, 1384. Greenhow, E. J., and Spencer, L. E., Analyst, 1973, 98, 485. Greenhow, E. J., and Spencer, L. E., Analyst, 1973, 98, 98. Harlow, G. A., Noble, C. M., and Wyld, G. E. A., Analyt. Chem., 1956, 28, 784. Bywater, S., in Jenkins, A. D., Editor, “Progress in Polymer Science,” Volume 4, Pergamon Prcss, Owensby, D. A., Parker, A. J., and Diggles, J . W., J . Ant. Chem. SOC., 1974, 96, 2682. Gutmann, V., Chemy Britain, 1971, 7, 102. Sprengling, G. R., J . A m . Chem. Soc., 1954, 76, 1190. Greenhow, E. J., Hargitt, R., and Shafi, A. A., Angew. Makromol. Chem., 1975, 48, 55. Streuli, C. A., Analyt. Chem., 1960, 32, 407. Vaughan, G. A., and Swithenbank, J. J., Analyst, 1965, 90, 894. Moss, M. L., Elliott, J. H., and Hall, R. T., Analyt. Chem., 1948, 20, 784. Katz, M., and Glenn, R. A., Analyt. Chem., 1952, 24, 1157. Greenhow, E. J., and Smith, J. W., Analyst, 1959, 84, 457. Deal, V. Z., and Wyld, G. E. A., Analyt. Chem., 1955, 27, 47. Cundiff, R. H., and Markunus, P. C., Analyt. Chem., 1956, 28, 792. Streuli, C. A., and Miron, R. R., Analyt. Chem., 1958, 30, 1978. Bruss, D. B., and Wyld, G. E. A., Analyt. Chem., 1957, 29, 232. Harlow, G. A., and Bruss, D. B., Analyt. Chern., 1958, 30, 1833. Oxford, 1975, p. 35. Received December 22nd, 1975 Accepted February 6th. 1976.

ISSN:0003-2654    

DOI:10.1039/AN9760100421

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Amlyst, J i m e , 1976, Vol. 101, pp. 433-438 433 Solid-state lon-selective Electrodes Based on Thin lon-selective Layers Deposited on Ionic Conductors* R. E. Van de Leest Philips Research Laboratories, Eindhoven, The Netherlands A new type of ion-selective electrode has been developed. The electrode consists of a supporting material that is a good ionic conductor, covered with a thin ion-selective layer. The electrode response is fast and reproducible. Electrodes for measuring chloride, bromide, iodide and hydrogen ortho- phosphate ions are described. For compounds that are good ionic conductors1 one expects a rapid ion-exchange process at the solid - solution interface and the rapid establishment of equilibrium in the solid, owing to fast ion diffusion. These materials should therefore be useful for the production of repro- ducible electrodes with a fast r e s p ~ n s e .~ ~ ~ Unfortunately, many ionic conducting compounds are not stable in aqueous solution and therefore cannot be used. This problem can be overcome by covering these compounds with a thin layer of an active material. The thin layer has to meet the following require- ments: it must be stable in aqueous solution; it must entirely cover the supporting material so that there is no contribution by the supporting material to the measured potential; the thickness and composition of the layer must be fixed, if the electrodes are to be reproducible; and the layer should be very thin. These requirements can be met when the layer is made by chemical or electrochemical reaction at the surface of the supporting material.In this way electrodes with a repro- ducible and fast response to chloride, bromide and iodide ions have been made by covering different supporting materials with a thin layer of silver chloride, silver bromide or silver iodide. Thus, materials of low conductivity can be used as the active layer. Experimental Reagents Analytical-reagent grade chemicals were used in all the experiments. De-ionised water was distilled in an all-silica apparatus. The following compounds were used as supporting materials : Ag,S (Merck Suprapur) ; Ag,SBr (reference 4) ; Ag,SI (reference 4) ; and Ag,,I,,P,O, (reference 5). These materials were prepared in our laboratory and characterised by use of methods such as differential thermal analysis, X-ray diffraction and electrical conductivity measurements.Pellets of 5-mm diameter were made from these materials and standard pre-treatment of the pellets included de-greasing with alcohol and treatment with dilute ammonia solution. Apparatus Measurements of e.m.f. were made with a Philips PW 9414 digital ion-activity meter. A Wenking potentiostat with an SMP 69 motor control was used to establish current - X-ray diffraction patterns were produced with a Philips PW 1050 diffractometer, using Electrical conductivities were measured with a Hewlett-Packard 4800 A vector impedance potential characteristics, which were recorded on a Philips PM 8120 X - Y recorder. Cu Kcc radiation. meter. Preparation of Thin Layers of AgX Chemical method The compounds to be used as supporting materials are allowed to react with gaseous chlorine, bromine or iodine, which results in the formation of a thin layer of silver chloride, * Presented a t the International Reference and Ion-selective Electrode Conference held a t the University of Newcastle upon Tyne, January 7-9th, 1976.434 VAN DE LEEST : SOLID-STATE ION-SELECTIVE ELECTRODES BASED Analyst, VoZ.101 silver bromide or silver iodide on the surface of the supporting material. The reaction is carried out at an elevated temperature, depending on the supporting material and the nature of the layer to be formed. The apparatus used is shown in Fig. 1. By using this method, thin layers of silver chloride, silver bromide and silver iodide have been made for each of Ag,S, Ag,SBr, Ag,SI and Ag,,I,,P,O, as supporting material.However, it is not necessary to make a silver iodide layer on Ag,SI because the latter compound is selective for iodide ions and is also less soluble in aqueous solution. Similarly, it is not necessary to make a silver bromide layer on Ag,SBr. Furnace Sample 1- Glass tube I'\ Furnace Fig. 1. Preparation of thin-layer (AgX) electrodes. Electrochemical method Thin layers of AgX can be made on the surface of good silver(1) ion conducting compounds by electrical generation of silver(1) ions in a solution that contains the appropriate counter ion. The silver(1) ions react with the counter ions to yield an insoluble compound at the surface of the supporting material, By integrating the generation current it is possible to calculate the thickness of the layer. Fig.2 shows, as an example, the curve for the formation of silver bromide at the surface of silver sulphide. In the presence of bromide ions (curve B) the anodic current observed at 0.7 V corresponds to the generation of silver(1) ions and the subsequent formation of the silver bromide layer. Silver-contacted pellets were used in these experiments. The feasibility of the method can be estimated from current - potential curves. 0 0.2 0.4 0.6 0.8 Potential versus S.C.E./V Fig. 2. Current - potential curve for silver sulphide. Solution: A, 10-1 M sodium nitrate; B, 10-1 M sodium nitrate + M bromide ion. Scan speed, 0.1 V min-l. Electrodes : working electrode, silver sulphide ; auxiliary electrode, platinum; reference electrode, S.C.E.June, 1976 ON THIN ION-SELECTIVE LAYERS DEPOSITED ON IONIC CONLIUCTORS 435 Thin layers of silver bromide and silver iodide have been made electrochemically for each of Ag,S, Ag,SBr, Ag,SI and Agl,Il,P,07 as supporting material, and a thin layer of silver chloride was also made for Agl,I15P,07.Preparation of Thin Layers of A&P04 of silver orthophosphate, P,074- ions being released into the solution. Agl,Il5P,O7 reacts with a concentrated solution of orthophosphate to yield a thin layer Results and Discussion Not all of the electrodes have been extensively tested. Only a few selected examples will be described. Ion-selective Electrode for Chloride Ions Pressed pellets (diameter 5 mm) of Ag,S, Ag,SBr (specific ionic conductivity 3 x loA3 R-l cm-1) and Ag,SI (specific ionic conductivity 8 x 0-l cm-l) are covered by a thin layer of silver chloride when heated in an atmosphere of chlorine.The reaction for silver sulphide can be represented by the equation Ag,S + C1, = 2AgC1+ S From a thermodynamic viewpoint this method is a good one for the preparation of well defined electrodes because, according to the phase rule, no degrees of freedom for this system are left at a given temperature. Initially, the system has three degrees of freedom but by heating at a fixed partial pressure of chlorine three parameters, i.e., the temperature, the total pressure and the chemical potential of chloride in the solid phase, which is proportional to the partial pressure of chlorine, are also fixed. Therefore the composition of the layer is fixed and, as the ion-selective layer on an electrode, should show reproducible behaviour.As no peaks due to Ag,S or to Ag,SI are observed, it is assumed that these materials are entirely covered with a layer of silver chloride. From electron microscope photographs the average size of the silver chloride crystallites is found to be about 1 pm. The layer of silver chloride is dissolved in ammonia solution and the amount of silver dissolved is determined by means of atomic-absorption spectroscopy. The layer thickness, as determined by this method, was found to vary between 4.5 and 5.5 pm. Therefore it can be concluded that silver sulphide is covered by several layers of crystalline silver chloride, which is in agreement with the information derived from the X-ray diffraction data.Measurements of e.m.f. made with different electrodes showed excellent reproducibility. The slope of the e.m.f. veysus concentration graph was 54 mV per decade (see Table I). X-ray diffraction patterns (Fig. 3) prove the silver chloride layer to be crystalline. The layer thickness is determined by chemical analysis. TABLE I E.M.F. RESPONSE OF THIN AgX LAYERS TO X- X = Cl, Br, I. E.m.f./mV veYsus S.C.E. [Cl-1, [Br-] or f L \ [I-I/M Ag2S - AgCl Ag2S - AgBr Ag2S - AgI 1 x 10-6 - + 130 + 46 1.1 x 10-6 + 226 + 107 - 85 1.1 x 10-4 + 194 + 58 - 143 1.1 x 10-3 + 141 fl - 203 6.1 x 10-3 + 113 - 39 - 248 8.1 x 10-3 + 102 - 45 - 255 The response of the thin-layer electrode to a change in chloride ion activity is faster than the response of the classical electrode, which consists of a pressed mixture of silver sulphide and silver chloride.The two types of electrode were placed in the same flow-through cell and the concentration of chloride ion in solution changed in two steps between 1 x low2 and 2 x lo-* M. The response of both electrodes was simultaneously recorded (Fig. 4). A436 VAN DE LEEST : SOLID-STATE ION-SELECTIVE ELECTRODES BASED Analyst, VoL. 101 1. I I I I I 1 10 "26 40 30 20 Fig. 3. X-ray diffraction diagrams obtained with Cu Ka radiation for: (a), Ag,S - AgCl; and (b), Ag,SI - AgCl. marked difference was observed between the response of the two electrodes, especially when changing from high to low concentration. This difference can be explained by the fact that silver chloride crystals prepared by a solid-state method are usually superior to those made by wet chemical methods, usually having a smaller solubility product, and as the active layer of an ion-selective electrode they should perform better at low concentrations.Thin layers of silver chloride on Ag,S and on Ag,SI give the same results in the measure- ments of e.m.f., and no difference in speed of response is observed. The specific ionic con- ductivity of Ag,SI (8 x lo-, S2-l cm-l) is higher than that of Ag,S L2-l cm-I). Hence the ionic conductivity of the electrode is not the only factor that determines the speed of electrode response.June, 1976 ON THIN ION-SELECTIVE LAYERS DEPOSITED ON IONIC CONDUCTORS 437 > E h w 150 100 50 Fig. 4. Response of two different ion-selective electrodes to chloride ion: -, thin-layer electrode; and - - - -, classical electrode. Ion-selective Electrode for Bromide and Iodide Ions When Ag,S or Ag,SBr is heated in an atmosphere of bromine a thin layer of silver bromide is formed at the surface of these compounds.Likewise, thin silver iodide layers can be made at the surface of Ag,S and Ag,SI. The resulting electrodes, made under thermo- dynamically well defined conditions, show reproducible behaviour as ion-selective electrodes for bromide or iodide ions. X-ray diffraction diagrams show that a homogeneous layer of silver bromide or silver iodide is formed, which completely covers the supporting material. The electrode response is fast and the e.m.f. values obtained with the bromide and iodide ion-selective electrodes are given in Table I.Ion-selective Electrode for Hydrogen Orthophosphate Ions selective electrode for hydrogen orthophosphate ions. The reaction can be represented by the equation A thin layer of silver orthophosphate at the surface of Ag,,I,,P,O, behaves as an ion- X-ray diffraction data show the layer to consist of silver orthophosphate and silver iodide. 3Agl,I15P,0, + 4PO4,- = 4Ag,P04 + 45AgI + 3P20,4- > E \ u! x Concentration of hydrogen orthophosphate/N Fig. 5. Ag.,,I,,P,O, - Ag,PO, as ion-selective electrode for hydrogen orthophosphate. Five electrodes were used. Solution, 100 ml of 10-1 M sodium nitrate (pH = 10). Magnetic stirring. Room temperature.438 VAN DE LEEST As &hydrogen orthophosphate ions give no response, this electrode is sensitive to pH, and when measuring phosphate ions the pH of the solution should therefore be kept constant.Hydrogen orthophosphate ions are measured at pH 10 because at this pH they are the only ionic phosphate species in solution. At concentrations between and N the graph of electrode response versus concentration is a straight line with a slope of 30 mV per decade. In Fig. 5 the responses of five different electrodes are compared and a reasonable repro- ducibility is shown. As the solubility product of silver orthophosphate is rather high (Ksp = 10-15-s),s many interferences can be expected. Conclusions The new type of ion-selective electrode, which consists of a thin ion-selective layer on a supporting material that is a good ionic conductor, has interesting properties. The electrodes are reproducible because they are made under thermodynamically defined conditions. The electrode response is fast. Some interferences can be reduced because the supporting material is entirely shielded from the solution and voltammetric determinations are possible because of the low internal resistance of the thin-layer electrode. The author thanks Mr. Bastings for chemical analysis, Mr. Langereis for X-ray diffraction experiments and Miss Stienstra for electron microscope experiments. References 1. 2. 3. 4. 5. 6. Owens, B. B.. in van Gool, W., Editor, “Fast Ion Transport in Solids,” North Holland Publishing Koebel, M., Ibl, N., and Frei, A. M., Electrochim. Acta, 1974, 19, 287. Beekmans, N. M., and Heyne, L., paper presented at 24th Meeting of I.S.E., Eindhoven, 1973. Keuter, B., and Hardel, I<., 2. Anorg. Allg. Chem., 1965, 340, 158. Takahashi, T., Ikeda, S., and Yamamoto, O., J . Electrochem. Soc., 1972, 119, 477. “Stability Constants of Metal-ion Complexes,” Special Publication No. 17, The Chemical Society, Received December lst, 1975 Accepted January 201h, 1975 Company, Amsterdam, 1973, p. 593. London, 1965.

ISSN:0003-2654    

DOI:10.1039/AN9760100433

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst, Jane, 1976, Vol. 101, pp. 439-444 439 In Vitro Studies on the Dissolution Rate of Industrial Retarded Urea Feedingstuffs by Use of a Selective Electrode. Application of the Potentio- metric Urea Enzyme Electrode in Measurements of Dissolution Rate lldiko Fritz Geza Nagy Lajos Fodor and Ern0 Pungor Institute for Agricultural Chemical Technology, Technical University, Budapest, Hungary EG Y T Pharmacochemical Works, Budapest, Hungary Institute for Agricultural Chemical Technology, Technical Uiziversity, Budapest, Hungary Institute for General and Analytical Chemistry, Technical University, Budapest, Hungary An enzyme electrode based on the neutral carrier type ammonium ion selective electrode was prepared in order to study the rate of dissolution of urea. A procedure was developed that was found to be applicable for monitoring the urea dissolution process in different industrial feedingstuff additives containing urea. It is well known that the proteins produced by the micro-organisms living in the intestines of ruminants can partly or even completely provide the protein demand of the organism.Non-protein nitrogen is converted into protein in the biosphere of the rumen. By taking advantage of this process a significant amount of the proteinic food supply can be saved. In order to increase this saving an appropriate amount of non-protein nitrogen is administered, usually together with the feedingstuff, to domestic ruminant animals, i.e., cattle and sheep, as a source of nitrogen for protein synthesis. Promising results have been obtained by using urea as the source of nitrogen.However, the urea, or the ammonia produced in its hydrolysis, is absorbed from the intestines and reaches a high level of concentration in the blood, thus poisoning the animal itself. The higher the concentration of urea in the intestines the greater the possibility of poisoning. Therefore it is advantageous to administer the urea in the form of a special preparation having a relatively low dissolution rate and thus the rate of dissolution of urea from different industrial feedingstuff preparations is an important property of the pr0duct.l The dissolution properties of different pharmaceutical products are very oft en examined in quality-control measurements.2 Different, more or less standardised, apparatus and procedures, both continuous and periodic, are generally used in these studies.In most instances different spectral detection techniques, combined with periodic sampling, are applied for the measurement of dissolution rates. However, electroanalytical detection with selective sensors offers several advantage~,~ as demonstrated by using a continuous voltam- metric method and developing appropriate apparatu~.~ The concept of enzyme electrodes, introduced by Updike and Hicks5 and Guilbault and Mont alvo,6 has broadened the applicability of electrochemical detection.',8 Guilbault and several of his co-workers developed different enzyme electrodes for the determination of urea6 and certain improvements in the electrodes have been reported.+ll These improvements were achieved by using ammonium ion selective electrodes of greater selectivity and by using reaction layers of improved enzyme stability.9-14 This paper describes the development of an appropriate procedure and apparatus for the determination of the rate of dissolution of urea from different industrial feedingstuff additives containing urea.The work was done in order to provide the analytical background for technological studies that had been carried out to improve the quality of special feeding- stuff additives of retarded dissolution rate. By using the techniques described in this paper comparisons can be made between the dissolution properties of different products.440 FRITZ et al. : In vitro STUDIES ON THE DISSOLUTION Experimental Analyst, VoZ. 101 The urea enzyme electrode was prepared in our laboratory, basically according to the procedure given by Guilbault and Nagy.12 Thus, the sensing membrane of the ammonium ion selective electrode was made of silicone rubber film containing the antibiotic nonactin as neutral carrier compound.15 The membranes, in the form of discs of diameter 5 mm and thickness about 0.2 mm, were placed over the ends of glass tubes having an outer diameter of 6mm.Ammonium chloride solution M) was used as inner reference electrolyte and a silver - silver chloride electrode was used as inner reference electrode. The neutral carrier type ammonium ion selective electrode was converted into a urea electrode by covering the active surface with a thin film of reaction layer containing the enzyme urease.The reaction layer was made of a thin-film disc of urease entrapped in polyacrylamide gel, or of urease solution (200 mg ml-l). In both instances the electrode surface was covered by a dialysis membrane cap. The thin discs of urease - polyacrylamide gel were cut from a gel bar prepared by light-initiated polymerisation of an enzyme monomer solution held in a glass tube of inner diameter 5 mm. In order to prepare a very thin reaction- layer disc the end of the gel bar was pushed out of the tube by the required amount by use of a plunger and the discs were cut by means of a wet razor blade. The preparation and polymerisation of the monomer solution, using 0.58 g of NN'-methyl- enebisacrylamide, 5 g of acrylamide, 3 mg of potassium peroxodisulphate (K,S20,) and 3 mg of riboflavine in 25 ml of 10-1 M tris buffer solution, 1.0 ml of the polymer solution being added to 175 mg of urease, is described by Guilbault and Montalvo.6*l0 The electrodes were kept at 4 "C in 10-1 M tris buffer solution, pH 7, when not in use.The apparatus with which the measurements were made is shown schematically in Fig. 1. T Fig. 1. Schematic design of the measuring apparatus used: T, thermostatic unit; J, water- jacket; E, enzyme electrode; C,, measuring half-cell; s, stirrer; B, salt bridge; R, reference electrode (S.C.E.) ; C,, reference half-cell ; M, millivoltmeter; K, impedance transformer amplifier; D, recorder. The measuring half-cell, C,, was a 100-ml beaker surrounded by a water-jacket, with an electric stirrer, operating at a constant frequency (Radiometer, Type 22).The standard solutions of urea for the calibration and the solvent mixture for the urea dissolution studies were poured into this beaker, the temperature of the water circulating in the jacket being maintained at 25 "C by means of a thermostat. The enzyme electrode, E, and one end of an agar-agar salt bridge prepared with 10-1 M lithium chloride solution were also placed in the measuring half-cell. The enzyme electrode potential was measured against that of the satura- ted calomel electrode with a millivoltmeter (Radiometer, Type pH M 26) and was recorded as a function of time with a Radelkis recorder, Type OH-814/1. Between the measuring cell and the recorder a Keithley, Type 604, electrometer was connected as impedance transformer.Measurements were made with solutions containing tris buffer solution, pH = 7, at a concentration of 10-1 M. The chemicals used were all analytical-reagent grade and the stock urea solution was prepared freshly every day, standard solutions being made by serial dilution. The declared activity of the enzyme urease (lyophilised preparation, obtainable from Merck, Darmstadt) was 5 U m g l .June, 1976 RATE OF INDUSTRIAL RETARDED UREA FEEDINGSTUFFS 441 Results and Discussion The most important electrochemical properties (sensitivity, selectivity, response time, stability, etc.) of the silicone rubber based, neutral carrier type ammonium ion selective electrode, used in our experiments as base sensor, have been discussed in detail elsewhere.12 The function of the urea-selective enzyme electrode depends on the following enzyme- catalysed reaction, which takes place in the reaction layer of the electrode.0 II Urease NHZ-C-NH, + 2H,O,2NH,+ + C0s2- . . .. - * (1) E.C. 3.5.1.5 When the enzyme electrode is dipped into a solution that contains urea, the urea diffuses to the electrode and reaction 1 proceeds, giving rise to an increase in the ammonium ion activity in the reaction layer at the sensing membrane of the ammonium ion selective electrode. In this way the electrode potential changes, and when the diffusion process and the reaction have achieved a steady state, a steady-state electrode potential can be recorded. The steady state is only to be expected if the concentration relationships in the bulk of the sample solution can be considered constant.However, during a lengthy measurement the auto-hydrolysis of the urea in the sample solution and the “reactor effect” of the enzyme electrode can seriously decrease the con- centration of urea and increase the ammonium ion activity in the sample solution. In order t o study this effect the urea enzyme electrode was dipped into 50 ml of M urea solution and stirring was applied, the electrode potential being recorded over a period of 2-3 h. In a relatively short time a virtually steady-state electrode potential was reached and subse- quently only a small change (1-1.5 mV h-l), noticeable only after 1-2 h, was observed. The electrode was placed in a fresh portion of the sample solution and the original reading was obtained for the virtually steady-state electrode potential, indicating that the subsequent small change in the electrode potential was caused by the slow decomposition of the urea in the sample.As a result of the experiments it was found that, for measurement of the dissolution of urea from feedingstuffs over a period of 2 h (if the complete dissolution were to require a measurement of that length), the hydrolysis of the urea does not have a significant effect on the enzyme-electrode potential. The urea enzyme electrodes were calibrated in each instance before and after the measure- ments. Two calibration graphs of the steady-state electrode potential versus the logarithm 125 100 > 75 E 1. .S 50 + Is + 25 a, U g o 8 - Lu -2 5 -751 ’ I 1 1 I 1 2 3 4 5 - Log (concentration of urea/M) Fig.2. Calibration graphs of a urea enzyme electrode prepared with enzyme - polyacrylamide gel reaction layer. A, freshly prepared electrode; B, after 15 d.442 FRITZ et al. : In vitro STUDIES ON THE DISSOLUTION Analyst, VoE. 101 of the urea concentration are shown in Fig. 2. The shape of the calibration graphs can be explained easily on the basis of the theoretical considerations and the equation given by Guilbault and Nagy.12 The electrode responded to urea in the concentration range 10-1- 10-5 M, a linear response generally being obtained between concentrations of and 10-4 M. The concentration of the solution in the measuring half-cell should not exceed the upper measuring limit of the electrode. For studies on the dissolution of urea from feedingstuffs, the amount of sample must be chosen so that the concentration of urea in the sample solution is under 10-1 M, preferably about M, when the dissolution of urea is complete.Of course, it is necessary to know the percentage of urea in the sample feedingstuff for this condition to obtain. One of the most important parameters controlling the applicability of an enzyme electrode is its stability. The variation in the enzyme-electrode response with time is a complicated function of the following three factors: alteration in the base sensor function; change in the factors controlling the mass transport processes in the diffusion layer of the electrode; and change in the specific enzyme activity of the reaction layer. In our work the separate examination of these factors was not undertaken.Instead, the electrodes were calibrated daily and changes in the function of the enzyme electrode were observed. The difference between the steady-state enzyme-electrode potential measured for and 10-3 M solutions in the instance of an electrode having a reaction layer consisting of urease entrapped in a polyacrylamide gel was plotted against the time elapsed after the preparation of the electrode (Fig. 3). From the results shown in Fig. 3 it is obvious that the electrode had a lifetime of more than 20 d. However, the slope of the calibration graph in the concentration range examined increased for 12 d and then gradually decreased. The increase probably occurred as a result of the change of the parameters controlling the transport processes.The decrease must be due to the enzyme deactivation process. Electrodes having a reaction layer made of a film of urease solution were found to have a shorter lifetime of about 2-6 d. The reaction layer of electrodes of either structure can easily be renewed. The activity of the urease gel bar from which the thin discs that were used to form the reaction layer were cut remained constant for 2 months when the bar was stored at 4 "C. 0 3 6 9 12 15 18 21 Tirne/d I I I I I I I Fig. 4. Effect of the age of the electrode on the measured dissolution value ( T s , ~ , ) for 10 1 0 3 6 9 12 15 18 21 Tirne/d urea powder. Fig. 3. Dependence of the slope of the urea calibration graph on the age of the electrode in the range 10-8-10-~ M (polyacrylamide reaction layer).In order to study the response time of the electrodes the electrode potential was recorded after dipping the electrode, previously kept in a buffer solution free of urea, into a 1 0 - 2 ~ solution of urea. The time taken for 90% of the total change in the electrode potential to be completed was considered to be the electrode response time. The response time of electrodes prepared with an enzyme - gel reaction layer was found to be usually in the range 60-80 s. However, for thick reaction layers the response times were in the range 120-240 s, and for liquid-film reaction layers response times as short its 5-10s were measured.June, 19 76 RATE OF INDUSTRIAL RETARDED UREA FEEDINGSTUFFS 443 It can be concluded that rapid changes in the concentration of urea in solution cannot be followed by using urea enzyme electrodes as detectors.However, as the concentration change during the dissolution process was expected to be much slower than the electrode response the electrode appeared to be suitable for dissolution studies. A change was observed, however, in the response time of the electrodes during their use and the effect of this change on the measurement of dissolution rates was examined. The dissolution curve of pure urea powder was recorded periodically by using the same electrode. An amount of urea powder was added to 100 ml of buffer solution such that after complete dissolution the concentration of urea was 1 0 - 2 ~ and response time, as defined above, was plotted as a function of the time elapsed after the preparation of the electrode (Fig.4). From Fig. 4 it can be seen that the dissolution curve, which in the instance of urea powder is obtained by monitoring a relatively fast process, is not affected for 15 d by the change in the electrode behaviour. However, when the response time of the electrode increases to the extent that the electrode response is slower than the dissolution process the curve recorded by using the enzyme electrode as detector cannot be considered to be a dissolution curve. 100 75 > E 50 k .- w & 25 $ 0 + Q L + - w" -25 -50 -75 I I I I I 1 I I 1 I 20 40 60 80 100 Time/min Fig. 5. Electrode potential veY.sz4.s time of dissolution graph recorded for a special white starch - urea complex adduct. The dissolution of urea from the sample feedingstuffs examined was a much slower process Thus, by using a freshly made electrode for only 2 weeks than that from urea powder. Tirndmin Fig.6. Urea dissolution graphs: A, fatty acid - urea adduct; and B, white starch - urea adduct.444 FRITZ, NAGY, FODOR AND PUNGOR the dissolution curves could not be affected by the increase in response time caused by the ageing of the electrode. Measurements of dissolution rates were obtained as follows. One hundred millilitres of 10-1 M tris buffer solution, pH 7, were poured into the measuring half-cell and when the electrode potential reached a constant value an appropriate amount of the sample was added to it. Continuous stirring was applied and the electrode potential versus time curve was recorded. The curve recorded for a white starch-urea adduct is shown in Fig.5. By using the appropriate calibration graph the electrode potential veysus time curves were converted into concentration of dissolved urea veysus time curves (Fig. 6) or, to make the comparison easier, the dissolution curves were given in the form of percentage of urea dissolved versus time curves (Fig. 7). 0 20 40 60 Ti me/min 80 100 Fig. 7. Urea dissolution graphs given in the form of percentage of urea dissolved versus time dependence. A, urea powder; B. starch - urea adduct I (original form) ; C, starch - urea adduct I ground (powder form) ; D, starch - urea adduct 11; and E, starch - urea adduct 111. The times required for the dissolution of 50 and 90% of the urea in the sample, T5, and T,,, were used in order to compare the dissolution rates of different products.The retardation of the dissolution rate was characterised with retardation factors R,, and R,,, defined as T,,(product) /T,,(urea powder) = R,, and Tgs( product) /Tg,(urea powder) = R,. The procedure and the technique described provide a convenient method of in vitro dissolution studies. However, it must be said that the best method of comparison of feeding- stuff additives containing urea would be to measure the concentration of urea in vivo both in the blood and in the rumen of the animal while it is fed by the product. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References Lusztig, E. K., Hagony, P. L., Kollar, L., and Barks, J., Olaj, Szappan, Kozmetika, 1974, 23, 10. Hersey, J. A., Mfg Chem. Aerosol News, 1969, 40, 32. Mason, W. D., Needham, T. E., and Price, J. C . , J . Pharm. Sci., 1971, 60, 1756. FehCr, Zs., Nagy, G., T6th, K., and Pungor, E., Analyst, 1974, 99, 699. Updike, S., and Hicks, G. P., Nature, Lond., 1967, 214, 986. Guilbault, G. G., and Montalvo, J. G., J . Am. Chem. Soc., 1969, 91, 2164. Nagy, G., and Pungor, E., Hung. Sci. Instrum., 1975, 32, 1. Rechnitz, G. A., Chem. Engng News, 1975, 53 (4), 29. Guilbault, G. G., and Montalvo, J. G., Analyt. Lett., 1969, 2, 283. Montalvo, J. G., and Guilbault, G. G., Analyt. Chem., 1969, 41, 1897. Guilbault, G. G., and Shu, F. K., Analyt. Chem., 1972, 44, 2161. Guilbault, G. G., and Nagy, G., Analyt. Chem., 1973, 45, 417. Guilbault, G. G,, Nagy, G., and Kuan, S. S., Analytica Chim. Acta, 1973, 67, 195. Guilbault, G. G., and Tarp, M., Analytica Chim. Acta, 1974, 73, 355. Scholer, R. P., and Simon, W., Chimia, 1970, 24, 372. Received October 20th, 1975 Accepted December 15th, 1975

ISSN:0003-2654    

DOI:10.1039/AN9760100439

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst, June, 1976, Vol. 101, pp. 445-454 445 Suppression of Interfering Ions in the Analysis of Plants to Determine Fluoride Using the Fluoride Ion Selective Elect rode Brian Vickery and Margaret L. Vickery Kenyatta University College, P.O. Box 43844, Nairobi, Kenya Plant ashes may contain sufficient aluminium and/or iron to interfere seriously in the determination of fluoride ions when using the fluoride ion selective electrode. In the presence of these metals the known additions method gave erroneous results, as did that involving the attempted formation of complexes with ethylenediaminetetraacetic acid, disodium salt, or 1, Z-cyclo- hexylenedinitrilotetraacetic acid. Good recoveries of fluoride ion were obtained in the presence of aluminium, iron, magnesium or silicate, using sodium citrate as the complexing agent.The application of the citrate complex method to ashes of commercial tea, high in aluminium and iron, gave recoveries of fluoride ion of greater than 90%. Plants normally contain less than 10 pg g-l of fluoride,l but a number are known to accumulate fluoride in their tissues. Some of the latter group biosynthesise organofluorine compounds.2 Natural fluoride accumulators occur in the genera Dichapetal~m,~ the^,^ Gastrolobi~m,~ CamelZin,6 O~ylobium,~ Acacias and Pnlico~rea.~ Plants may also absorb fluoride into their leaves as a result of atmospheric pollution, sometimes with adverse consequences.1° Until recently, the analysis of plant material to determine fluoride was usually carried out either by the Willard - Winter distillation method,ll or by diffusion of liberated hydro- fluoric acid followed by its quantitative determination as the alizarin fluoro-complexonate.12 With the introduction of the fluoride ion selective electrode, the analysis of plants to determine fluoride has been simplified, and a number of methods involving the use of this electrode have been published.13 These methods are not generally suitable for plants that contain substantial amounts of aluminium, iron and silica, which form complexes with fluoride in solution.Cannon14 quotes the results of analyses of more than 80 plant species, representing all types of vegetation, showing that the average aluminium content of the plant ashes was 8 610 pg g-l, while the average iron content was 6 740 pg g-l.Louw and Richards15 developed a method for the determination of fluoride in sugar cane, using the fluoride ion selective electrode. They expressed the belief that their method could be applied to other types of plant tissues, particularly those which contained high concen- trations of aluminium, iron and silica. Louw and Richards did not, however, quote the amounts of these elements present in their samples, and in sugar cane these amounts are known to vary over a wide range, viz., 0.004-0.3% for aluminium, 0.00~0.3~0 for iron and 1-3% for si1ic0n.l~ A variety of methods have been used to minimise the interference from aluminium, iron and silica when using the fluoride ion selective electrode, and these methods can be divided broadly into two classes, the known additions, or spike, method and the method involving addition of a compound that forms complexes preferentially with interfering elements, thus releasing fluoride ions.We have investigated the application of these methods to the determination of fluoride in plant ashes. Experimental The fluoride ion concentrations were determined by use of a Corning fluoride ion selective electrode and a saturated calomel reference electrode connected to a Corning - EEL, Model 12, pH meter. The sample solutions were stirred slowly with a magnetic stirrer and temperature changes were kept to a minimum by placing insulation, consisting of three asbestos mats, between the stirrer motor and the beaker containing the solution under analysis. There was a considerable variation in the time taken to reach equilibrium, those solutions most446 VICKERY AND VICKERY: SUPPRESSION OF INTERFERING Analyst, VoZ.101 dilute in fluoride taking the longest. Equilibrium was considered to be established when there was no change in e.m.f. over 2 min. Standard graphs were prepared from known fluoride ion concentrations introduced into the appropriate buffer solution. All reagents were of analytical-reagent grade. The concentrations of aluminium and iron in commercial tea ashes were determined according to the methods given in the "Methods of Analysis of the Association of Official Agricultural Chemists. "16 In order to determine their fluoride ion concentration 1-g samples of tea were treated with 10 ml of sodium hydroxide solution (67 g 1-1) and heated in nickel crucibles at 110 "C for 24 h.The crucibles were next transferred to a muffle furnace, heated at 500 "C for 2 h and then at 800 "C until the contents had fused. When cool, the fused mixture was dissolved in a mixture of glacial acetic acid (2.5 ml) and water (10 ml). The resulting solution was transferred into a calibrated flask, 5 ml of sodium citrate solution (2.5 M) were added and the solution was made up to 50 ml with water. The pH was adjusted to 8.5 with sodium hydroxide solution (67 g 1-l), which was added dropwise. A standard graph was prepared from known concentrations of sodium fluoride in 5 ml of sodium citrate solution (2.5 M), 10 ml of sodium hydroxide solution (67 g 1-l) and 2.5 ml of glacial acetic acid, adjusted to pH 8.5 and diluted to 50 ml with water.The Known Additions Method The basic work on the known additions or spike method as applied to fluoride ion analysis in solution was done by Baumann.17 In order to prevent interference by metals that form complexes with fluoride she worked with solutions that were greater than 1 M in ortho- phosphoric acid and less than 1 0 - 3 ~ with respect to fluoride. In solutions of such high acidity most of the fluoride was complexed as hydrofluoric acid, and this acid was not sensed by the electrode. Baumann showed that under these conditions a consistent proportion of the fluoride ion was complexed as hydrofluoric acid irrespective of the total fluoride ion concentration. She also demonstrated that results obtained by using the known additions method, giving a slope of 59 mV per decade, were comparable with those obtained directly from calibration graphs, which might not have slopes of 59 mV per decade.In work designed to investigate the equilibria obtaining in aqueous hydrofluoric acid solution, Srinivasan and Rechnitz18 showed inter alia that when the concentration of hydrogen ions was 1.5 x 10-1 M and the fluoride ion concentration varied from 4 x to 10-1 M a variable proportion of fluoride ion was uncomplexed. Table I has been obtained from data given in Table I of their paper. The difference between Baumann's results and those of Srinivasan and Rechnitz can be attributed to the fact that under Baumann's conditions of high hydrogen ion concentration and low fluoride ion concentration, hydrofluoric acid was the only complex formed in a substantial amount.In Srinivasan and Rechnitz's work, with lower hydrogen ion and higher fluoride ion concentrations, HF,- was formed to an extent which varied with the total fluoride ion concentration. The equilibria present in aqueous hydrofluoric acid are shown below [equations (2) and (3)]. TABLE I TOTAL AND FREE FLUORIDE ION CONCENTRATIONS IN AQUEOUS SOLUTIONS OF (IONIC STRENGTH = 1.0 M) HYDROFLUORIC ACID AT CONSTANT HYDROGEN ION CONCENTRATIONS Calculated from Table I of Srinivasan and Rechnitz.18 Total F- concentration/M 4 x 10-3 8 x 10-3 2 x 10-2 4 x 10-2 6 x 8 x 10-2 1 x 10-1 Free F- concentration/M 3.7 x 10-6 6.0 x 2.1 x 10-4 4.7 x 10-4 8.5 x 10-4 1.4 x 10-3 2.2 x 10-3 H+ concentration/M 0.15 0.15 0.15 0.15 0.15 0.15 0.15 Free F-, % 0.92 0.75 1.05 1.18 1.42 1.75 2.20 A theoretical justification for Baumann's method can be obtained by using equation ( l ) , 1 9 which gives the results shown in Table 11.June, 1976 IONS I N THE DETERMINATION O F FLUORIDE I N PLANTS 447 N N 0 0 ..* . (1) Z&[L]i + - X(T1 - iT&[L]i = 0 .. where N is the maximum co-ordination number of the metal (6 in the aluminium - fluorine system), Pi the ith over-all stability constant, [L] the concentration of free ligand, T I the total concentration of ligand and Tm the total metal ion concentration. TABLE I1 FREE FLUORIDE ION CONCENTRATION AT 1 M TOTAL HYDROGEN ION CONCENTRATION FOR VARIOUS TOTAL FLUORIDE ION CONCENTRATIONS Total F- Free F- concentration/nz concentrationliv Free F-, yo 10-6 10-4 10-3 7.2 x 10-9 7.2 x 7.2 x lo-* 7.2 x 7.2 x 10-7 7.2 x 10-2 Srinivasan and Rechnitz18 found the following equilibria and equilibrium constants in aqueous hydrofluoric acid : H+ (as.) + F- (as.) = HF (aq.); K , = 794 .. .. ' * (2) HF (aq.) + F- (as.) = HF,- (aq.); K , = 6 . . The ratio K,: K , = 132. Thus, under conditions where two (or more) successive equilibria are present, a constant proportion of the ligand is present at equilibrium only under certain conditions, viz., a very high concentration of the cation and a very low concentration of the ligand. For the complexes of aluminium (Al3+) with fluoride (F-), stepwise (K) and over-all (13) stability constants are as follows20: AP+ AlF2f (aq.) + F- (aq.) = A1F2+ (as.); K , = 105-02, /3, = (aq.) + F- (aq.) = A1F2+ (as.); K, = 106J3, /I, = l O 6 - l 3 * - (4) * * (6) .. (5) AlF,+ (aq.) + F- (aq.) = AlF, (aq.); K , = 103-s5, 13, = lo1, AlF, (aq.) + F- (aq.) = AlF, (as.); K , = 102-74, p4 = 1017.74 AlF,- (as.) + F- (aq.) = A1FS2- (as.); K , = 101-63, P5 = 1019.37 A1F5,- (aq.) + F- (as.) = (aq.) ; K6 = 10°.47, p 6 = 1019-s4 . . (7) . . (8) . . (9) Note that K,: K , = 13. In order to appreciate the significance of this information in terms of the complexing of fluoride by aluminium in fluoride ion determinations using the known additions method, we have calculated the theoretical concentration of free fluoride ion in solutions containing aluminium and fluoride, at concentrations that are usual in plant analysis,14 by use of equation (1). If we assume that the aluminium ion concentration is lo-, M, and that the unspiked solution has a total fluoride ion concentration of 10-4 M, solution of equation (1) gives a free ligand (F-) concentration of 8.2 x 10-SM and 99.92% of the fluoride is com- plexed.If we were now to add a spike of negligible volume, which increased the total [F-] to M and 99.774 of the fluoride is complexed. These results are summarised in Table 111. It is clear that under these conditions the proportion of fluoride ion complexed before and after spiking is not the same and that the spike method will give erroneous results. Because the total fluoride concentration has been increased 10-fold, the more concentrated solution should have a potential 59 mV less than the weaker. When complex formation has occurred, as is shown in Table 111, instead of the ratios of free fluoride ion concentration after spiking to before spiking being 10: 1, the theoretical value, they will be 39: 1 and 32 : 1.Results calculated for the fluoride ion concentration in the original solutions will be much lower M, solution of the equation now gives [L] = 3.2 x448 VICKERY AND VICKERY: SUPPRESSION OF INTERFERING Analyst, Vol. 101 TABLE I11 PERCENTAGE OF FLUORIDE IONS COMPLEXED WITH VARIOUS ALUMINIUM AND FLUORIDE ION CONCENTRATIONS AS CALCULATED FROM EQUATION (1) in a neutral solution with an ionic strength of 0.53 M. Total AP+ Total F- Free F- F- The stability constants used in the calculations were those listed in reference 20, which were determined concentration/M concentration/M concentration/M complexed, % 10-3 10-4 8.2 x 10-8 99.92 10-3 10-3 3.2 x 99.68 10-2 10-3 8.2 x 10-8 99.99 10-2 10-2 2.7 x 99.97 than they should be.It can be demonstrated21 that the effect of increasing the fluoride ion concentration while maintaining a constant aluminium ion concentration is to increase the proportion of complexes with co-ordination numbers greater than 1. This effect is shown in Table IV. TABLE IV DISTRIBUTION OF ALUMINIUM FLUORIDE COMPLEXES AS PERCENTAGES OF TOTAL COMPLEXED SPECIES Calculated from Rossotti and Rossotti.21 Complexed species Total A13+ Total F- A concentration/M concentration/M A{FZ+ AlF,+ A1F3 AIF,- A1F62- AlG63- 10-3 10-4 99.1 0.85 0 0 0 0 10-3 10-3 75.4 24.0 0.55 0.001 5 0 0 10-2 10-3 99.1 0.85 0 0 0 0 10-2 10-2 78.7 20.9 0.4 0 0 0 Unfortunately, these theoretical results cannot be verified in solutions containing aluminium and fluoride ions as the only complexing species as the aluminium ion is appreciably acidic, showing the following equilibrium in water22 : [A1(H20)s]3+ = [A1(H2O),(OH)l2+ + H+; K = 1.12 x lo-, .. . . (10) If the pH of the aluminium ion solution is adjusted to 5, the lowest pH at which fluoride ion determinations are normally carried out with the fluoride ion electrode, the aluminium begins to precipitate as hydroxyaluminium complexes, disturbing the aluminium - fluorine equilibria. Precipitation of these complexes can be prevented by using an acetate, or other, buffer, but such a buffer will also affect the aluminium - fluorine equilibria because the buffer must be complexing to some extent with aluminium in order to render it soluble in water at this pH.A similar argument applies to iron(II1) ions, which also form a complex with fluoride. If the iron(II1) and aluminium ions are not completely complexed by the buffer, the known additions method will give results that are too low. Thus, when a known fluoride ion concen- tration (6.4 x M) was determined with the fluoride ion selective electrode, by use of the known additions method, in solutions containing aluminium or iron(II1) ions and buffered with sodium chloride (0.5 M) -sodium acetate (0.5 M) at pH 5.3, values of 5.8 x 1 0 - 2 ~ (with 0.02 M A13+) and 5.5 x Complexing fluoride with aluminium ions or other species can have another effect, which renders the known additions method valueless.Fluoride ion concentrations can only be measured with the fluoride ion selective electrode down to about M; below this con- centration the potential tends t o become constant and the method would give erroneously high results.23 It is shown in Table V, where the concentration of fluoride ion was determined at pH 5 in the presence of 0.2 M aluminium nitrate solution, that erroneously high results are obtained even when the concentration of fluoride is as high as 1.6 x M. In order to keep the aluminium in solution, and to provide a constant ionic background, the solutions were buffered with sodium chloride (0.5 M) - sodium acetate (0.5 M) solution. M (with 0.02 M Fe3+) for fluoride were obtained.June, 1976 IONS I N THE DETERMINATION OF FLUORIDE I N PLANTS 449 Carlson and K e e n e ~ ~ ~ state that “the method of standard additions may be applicable to samples which contain concentrations of complexing agents in excess of those of the ion sought.The basis of this method is that the addition of a known amount of the ion does not change the ratio of complexed to non-complexed ion in the sample solution. Thus, the ratio of the activities before and after addition of the standard would be the same as it would be if no complexes were present. The log nature of the response of the electrode causes the change in potential with the addition of standard to be constant no matter what the nature of the complex species.” M ~ w b r a y ~ ~ states that the metal must be present in a higher con- centration than the ligand for the spike method to be effective.This contention is not found to be true in practice when the complexing ion is aluminium, as is evident from Table V. TABLE V DETERMINATION OF FLUORIDE ION CONCENTRATIONS IN SOLUTIONS 0.02 M IN ALUMINIUM NITRATE AND BUFFERED WITH 0.5 M SODIUM CHLORIDE - 0.5 M SODIUM ACETATE AT pH 5 Determined from a standard graph (sg) and by the known additions method (kam). Actual F- F- F- concentrationlM concentration/M (sg) * concentration/M (kam) 1 x 10-3 1.6 x 10-4 7.5 x 10-1 4 x 10-3 1.6 x 10-4 9.4 x 10-1 1 x 10-2 1.6 x 10-4 1.0 1.6 x 1.8 x 10-4 1.0 * The standard graph was prepared by determining the potential of known concentrations of fluoride ion in a solution buffered with 0.5 M sodium chloride - 0.6 M sodium acetate. It is also claimed that the known additions method overcomes the effect of having an unknown ionic strength in solution.Baker13 states that the standard additions technique, by its very nature, cancels out the effects caused by differing pH values and ionic strengths encountered between standards and actual samples. “The calibration curve is, as it were, prepared in the sample solution.” This is not borne out in practice, as varying ionic strengths may affect the position of the slope and the theoretical Nernst relationship between the concentration of fluoride ion and the measured e.m.f. It can be seen from Table VI that both the experimental slopes and the measured e.m.f. vary with the nature of the ions in solution, even when these ions do not form complexes with fluoride.Such effects are unpre- dictable and lead to varying results when the fluoride ion concentration is determined by the known additions method (Table VI). TABLE VI EXPERIMENTAL SLOPES AND E.M.F. RECORDED BY THE FLUORIDE ION ELECTRODE 0.5 M SODIUM ACETATE AND 0.5 M SODIUM CHLORIDE - 0.5 M SODIUM ACETATE, AND THE FLUORIDE ION CONCENTRATIONS DETERMINED BY THE KNOWN ADDITIONS METHOD (kam) FOR CONCENTRATIONS OF FLUORIDE ION I N SOLUTION I N WATER, 0.5 M SODIUM CHLORIDE, Actual F- concentration 1 .o 4.0 10.0 16.0 64.0 Slope x 1 0 3 1 ~ E. m. f . /mV F- concentration x lo3 (kam)/M A r A \ r I H,O NaCl NaAc NaC1-NaAc H,O NaCl NaAc NaC1-NaAc 110 115 109 116 0.8 1.1 1.1 1.0 74 80 75 81 4.0 4.0 4.0 3.8 49 58 51 57 11.0 9.5 10.0 9.5 37 46 39 48 15.0 15.9 15.0 14.0 3 11 5 13 58.0 63.0 63.0 55.0 60 58 58 57 The Formation of Complexes In order to suppress the complexation of fluoride by a number of metal ions during water analysis, Frant and Ross26 originally made use of a very small concentration of citrate in an acetate buffer at pH 5.5.Subsequently ethylenediaminetetraacetic acid, disodium salt450 VICKERY AND VICKERY: SUPPRESSION OF INTERFERING Analyst, VoZ. 101 (EDTA) was substituted for the citrate. Hanvoodz7 later showed that 1,2-~yclohexylene- dinitrilotetraacetic acid (CDTA) was superior to EDTA in water analysis. However, a 5 g 1-1 solution of CDTA, when added 1 + 1 to the water to be analysed, gave a 97% recovery of fluoride in the presence of 2 pug g1 of aluminium, 91% with 5 pg 8-l of aluminium and only 84% with 1Opgg-1 of aluminium.In the analysis of plants the portion of sample solution taken for fluoride ion analysis may contain up to several hundred pg g-1 of aluminium ions. (Harwood cites the stability constants of EDTA and CDTA incorrectly as 16.1 and 17.6 respectively; these values are in fact for log K.) The stability constants of the complexes of aluminium and iron with EDTA and CDTA are strongly pH dependent, an important factor that seems to have been overlooked by previous workers. The co-ordinating power of these ligands depends on the extent to which the carboxyl groups are ionised ; thus co-ordinating power decreases with decreasing pH. This relationship is shown in Table VII. Moreover, the reaction between aluminium and EDTA is not instantaneousz8 and at higher pH values aluminium forms very inert hydroxo complexes.The limiting pH above which the latter complexes will be formed can be calculated from equation (1 1) .z9 Hence, complex formation with aluminium by EDTA may only be complete at pH 3.3 with 10-2 M Al3+ or pH 4 with M A13+, and at these pH values EDTA itself is not a very effective ligand (Table VII). I n fact, fluoride can be used as a masking agent for aluminium against EDTA if a 5-6-fold excess of fluoride over aluminium is added, as in boiling solutions fluoride displaces EDTA from its aluminium complex.30 TABLE VII EDTA, ALUMINIUM ION - CDTA, IRON(III) ION - EDTA AND IRON(III) ION - CDTA [H+] = (10-8 [A13+])1/3 .. .. .. . . (11) LOGARITHMS OF THE EFFECTIVE STABILITY CONSTANT, Keff, FOR THE ALUMINIUM ION - COMPLEXES AT AN IONIC STRENGTH OF 0.1 M Calculated from Tables I1 and I11 of Schwartzenbach and Flaschka.28 PH r 7 Complex 1 2 3 4 6 6 7 8 9 1 0 1 1 1 2 A13+-EDTA -1.9 2.6 5.5 7.7 9.6 11.4 12.9 13.8 14.8 15.7 16.0 16.1 AP+-CDTA -2.1 1.8 5.1 7.7 9.8 11.6 12.6 13.9 14.9 15.9 16.9 17.8 Fe3+- EDTA 7.1 11.6 14.5 16.7 18.6 20.4 21.8 22.8 23.8 24.6 25.0 25.1 Fe3+-CDTA 8.9 12.8 16.1 18.7 20.8 32.6 23.7 24.9 25.9 26.9 27.9 28.8 As can be seen from Table VIII, the recovery of fluoride when determined in the presence of a solution 0.02 M in aluminium nitrate and 0.04 M in EDTA at pH 8.5 was very low.At pH values less than 8.5 the free fluoride ion concentration was too small to be detected if the concentration of added fluoride was less than 6 x M. Clearly, fluoride forms com- plexes more efficiently with the A13+ ion than does EDTA.As neither EDTA nor CDTA is suitable as a masking agent we investigated the use of sodium citrate at various pH values. Oliver and Clayton,31 working with fluoride minerals and with the electrolyte used in the Hall process for aluminium production, showed that with 500 pg g-1 of aluminium in solution there was a linear relationship between log [F-] and potential over a wide range, when a buffer that was 0.5 M in potassium acetate and 1 M TABLE VIII DETERMINATION OF FLUORIDE ION CONCENTRATIONS IN SOLUTIONS 0.04 M IN EDTA AND 0.02 M IN ALUMINIUM NITRATE AT pH 8.5 Actual F- Experimental* F- concentration/M concentration/M 2 x 10-3 4.5 x 10-4 4 x 10-3 8.0 x 10-4 6 x 9.0 x 10-4 8 x 10-3 1.1 x 10-3 1 x 10-2 1.5 x 10-3 * Determined from a standard graph prepared from solutions 0.04 M in EDTA.June, 1976 IONS I N THE DETERMINATION OF FLUORIDE I N PLANTS 461 in sodium citrate, adjusted to pH 8, was used.Buck and R e i ~ m a n n , ~ ~ after ashing plant material and fusing the ash with sodium hydroxide, determined the fluoride ions in a solution that was 0.71 M with respect to citrate at pH 5.7. These last workers found that aluminium and magnesium interfered and that the relative error with magnesium decreased with in- creasing citrate concentration. The relative error with aluminium decreased with increasing pH, and was only 0.5% at pH 8.3. These workers advocated the use of CDTA to eliminate interference from magnesium. TABLE IX DETERMINATION OF FLUORIDE ION CONCENTRATIONS AT pH 8.5 IN THE PRESENCE OF 0 .0 2 ~ ALUMINIUM NITRATE OR 0 . 0 2 ~ IRON(III) NITRATE, AND EITHER 0 . 2 5 ~ SODIUM CITRATE OR 0.5 M SODIUM CHLORIDE - 0.5 M SODIUM ACETATE - 0.25 M SODIUM CITRATE F- (~13+) concentration/M Actual F- f A > concentrationl~ Na3 citrate* Buffert I x 10-3 1.1 x 10-3 9.7 x 10-4 4 x 10-3 4.2 x 10-3 2.1 x 10-3 1 x 10-2 1.0 x 10-2 5.0 x 10-3 1.6 x 1.6 x 7.9 x 10-3 6.4 x 10-2 6.4 x 4.5 x 10-2 F- (Fe3.+) concentration/M f A \ Na, citrate* Buffer t 1.2 x 10-3 1.2 x 10-3 4.0 x 10-3 4.0 x 10-3 1.0 x 10-2 1.1 x 10-2 1.6 x 1.7 x 6.3 x 7.2 x 10-2 * Values determined from a standard graph prepared from solutions 0.25 M in sodium citrate. 7 Values determined from a standard graph prepared from solutions containing 0.5 M sodium chloride - 0.5 M sodium acetate - 0.25 M sodium citrate.Our results show that in the presence of aluminium or iron in the concentrations likely to be met in the ashes of plants that accumulate these metals, a buffer of 0.25 M sodium citrate solution at pH 8.5 gives good recoveries of fluoride and is preferable to a 0.5 M sodium chloride - 0.5 M sodium acetate - 0.25 M sodium citrate buffer at this pH (Table IX). Deter- minations made by using a standard graph prepared in a solution that was 0.25 M with respect to sodium citrate gave slightly more accurate results than those obtained by the known additions method (Table X). At pH 6.5 citrate is not a good complexing agent for aluminium, the recovery of fluoride at this pH being too low when determined from the standard graph and too high when the known additions method is applied (Table XI).TABLE X DETERMINATION OF FLUORIDE ION CONCENTRATIONS BY THE KNOWN ADDITIONS METHOD AT pH 8.5 IN THE PRESENCE OF 0.02 M ALUMINIUM NITRATE OR 0.02 M IRON(III) NITRATE AND EITHER 0.25 M SODIUM CITRATE OR 0.5 M SODIUM CHLORIDE - 0.5 M SODIUM ACETATE - 0.25 M SODIUM CITRATE F- (~13+) concentration/M Actual F- r A 1 concentration/M Na, citrate Buffer 1.0 x 10-3 1.1 x 10-3 1.3 x 10-3 4.0 x 10-3 4.0 x 10-3 2.7 x 10-3 1.0 x 10-2 1.0 x 10-2 6.1 x 10-3 1.6 x 1.5 x 8.8 x 6.4 x 5.0 x 4.0 x F- (Fe3+) concentration/ni r A I Na, citrate Buffer 1.2 x 10-3 1.3 x 10-3 4.0 x 10-3 3.9 x 10-3 9.5 x 10-3 9.2 x 10-3 1.6 x 1.5 x 5.8 x 5.8 x The recovery of fluoride from solutions containing magnesium or silicate ions in concen- trations comparable with those found in the ashes of plants that accumulate magnesium or silica is good when a 0.25 M sodium citrate buffer solution is used at pH 8.5 and the deter- mination is made from the standard graph.The known additions method gives less accurate results (Table XII). Although silicon is known to form a complex with fluoride at low pH values there is little interference from silicate at pH 8.5 (Table XIII) and the addition of citrate is not necessary, providing aluminium, iron and magnesium are absent.452 VICKERY AND VICKERY: SUPPRESSION OF INTERFERING Analyst, VoZ. 101 TABLE XI DETERMINATION OF FLUORIDE ION CONCENTRATIONS BY THE STANDARD GRAPH (sg) AND KNOWN ADDITIONS (kam) METHODS IN THE PRESENCE OF 0.02 M ALUMINIUM NITRATE AND 0.25 M SODIUM CITRATE AT pH 6.5 Actual F- Experimental F- Experimental F- concentration/M concentration/M (sg) * concentration/M (kam) 1 x 10-3 7.4 x 10-4 1.4 x 10-3 4 x 10-3 3.4 x 10-3 6.6 x 1 x 10-2 7.4 x 10-3 1.4 x 1.6 x 1.2 x 10-2 2.5 x 6.4 x 3.8 x 2.5 x 10-l * Determined from a standard graph prepared from solutions 0.25 M in sodium citrate.Thus, the method that gives the best recoveries of fluoride in the presence of aluminium, iron and magnesium is that in which 0.25 M sodium citrate solution is used to complex the metal ions at pH 8.5 and determination of fluoride ion concentration is made from the standard graph. TABLE XI1 DETERMINATION OF FLUORIDE ION CONCENTRATIONS BY THE STANDARD GRAPH (sg) AND KNOWN ADDITIONS (kam) METHODS AT pH 8.5 IN THE PRESENCE OF 0.25 M SODIUM CITRATE AND EITHER 0.016 M MAGNESIUM NITRATE OR 0.004 M SODIUM SILICATE F- (Mg2+) concentration/x? Actual F- I A > concentration/nx sg * kam 1 x 10-3 1.0 x 10-3 1.1 x 10-3 4 x 10-3 4.0 x 10-3 4.2 x 10-3 1 x 10-2 1.0 x 10-2 1.0 x 10-2 1.6 x 1.6 x 1.5 x 6.4 x 6.4 x 6.3 x F- (S&0,2-) concentration/M r 1 sg* kam 1.1 x 10-3 1.1 x 10-3 4.2 x 10-3 4.3 x 10-3 1.0 x 10-2 1.1 x 10-2 1.6 x 1.7 x 6.4 x 6.6 x 10-2 * Determined from a standard graph prepared from solutions 0.25 M in sodium citrate.No work seems to have been done on metal ion - citrate complexes at pH 8.5, so it can only be assumed that the efficacy of citrate as a complexing agent at this pH is related to its increased ionisation at higher pH values. As is mentioned above, a similar effect is found with EDTA.That the metal ions become strongly bound to citrate ions seems certain from the above results, but only in the instance of iron(II1) - citrate complexes has a very large equilibrium constant been found.33 TABLE XI11 DETERMINATION OF FLUORIDE ION CONCENTRATIONS BY THE STANDARD GRAPH (sg) AND KNOWN ADDITIONS (kam) METHODS IN THE PRESENCE OF 0.004 M SODIUM SILICATE AT pH 5 Actual F- Experimental F- Experimental F- concentration/M concentration/M (sg)* concentration/M (kam) 1 x 10-3 4 x 10-3 1 x 10-2 1.6 x 6.4 x 1.0' x 10-3 4.0 x 10-3 9.8 x 10-3 1.6 x 6.4 x 1.5 x 10-3 4.5 x 10-3 1.0 x 10-2 1.6 x 6.3 x * Determined from a standard graph of sodium fluoride in solution in water.June, I976 IONS I N THE DETERMINATION OF FLUORIDE I N PLANTS 453 A summary of known equilibrium constants for iron(II1) - it rate^^,^^ and magnesium - ZFe3+ + 2Cit3- = [Fe2(H2Cit),12- + 2H+; log K = 21.2 (in 0.1 M NaC10, solution) Fe3+ + Cit3- = FeCit; log K = 11.85 (in 1 M NaCIO, solution) Mg2+ + Cit3- = MgCit-; log K = 3.29 (in 0.9 M NaNO, solution) No work has been done on Si - &rate3- or A13+ - &rate3- complexes.itr rate,^ complexes is as follows : Application of the Method to Plant Ashes The tea plant is known to accumulate up to 17 000 pg g-l of aluminium.35 The ash from a tea plant sample will therefore contain sufficient of this element to interfere considerably with the determination of fluoride ions, unless the aluminium is removed by complexation with citrate ions. By use of the fluoride ion selective electrode and a 0.25 M sodium citrate buffer solution we determined the total fluoride ion concentration in l-g samples of commercial tea containing an average of 2 050 pg g-l of aluminium and 2 800 pg g-l of iron.TABLE XIV RECOVERY OF SODIUM FLUORIDE ADDED TO THE ASH OF COMMERCIAL TEA SAMPLES DETERMINED WITH A FLUORIDE ION SELECTIVE ELECTRODE IN THE PRESENCE OF 0.25 M SODIUM CITRATE AT pH 8.5 F- added to ash from 1 g of tea in 50 mi of solution 0 2 5 8 32 x 104/m01 F- determined x 104/mol I Sample Sample Sample A B C 1.55 2.90 1.50 3.47 4.76 3.30 6.76 7.40 6.45 9.90 10.25 9.50 34.22 32.00 33.15 Recovery of added fluoride, yo Average - recovery Sample Sample Sample of added A B C fluoride, % 96 93 90 93 f 3 104 90 99 98 f 8 104 92 100 99 f 7 102 91 99 97 f 6 - - - - Known amounts of sodium fluoride were added to the ash from l-g samples of tea and the total fluoride ion concentration was again determined.The results given in Table XIV show that the recovery of fluoride was better than 90% when the determination was made from a standard graph prepared from samples dissolved in a solution containing sodium citrate at a concentration of 0.25 M. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. References Bollard, E. G., and Butler, G. W., A . Rev. PI. Physiol., 1966, 17, 89. Vickery, B., and Vickery, M. L., Vet. Bull., Weybridge, 1973, 43, 537. Vickery, B., and Vickery, M. L., Phytochem., 1972, 11, 1905. Zimmerman, P. W., Hitchcock, A. E., and Givertsman, J., Contr. Boyce Thomson Inst. PI.Res., McEwan, T., Nature, Lond., 1964, 201, 827. Venkateswarlu, P., Armstrong, W. D., and Singer, L., PI. Physiol., 1965, 40, 255. Aplin, T. E. H., J . Agric. West Aust., 1971, 12, 154. Oelrichs, P. B., and McEwan, T., Q. J . Agric. Sci., 1962, 19, 1. de Oliveira, M. M., Experientia, 1963, 19, 586. Lovelace, J., Miller, G. W., and Welkie, G. W., Atgnos. Envir., 1968, 2, 187. Willard, H. H., and Winter, 0. B., Ind. Engng Chem. Analyt. Edn, 1933, 5, 7. Hall, R. J., Analyst, 1968, 93, 461. Baker, R. L., Analyt. Chem., 1972, 44, 1326. Cannon, H. L., Science, N.Y., 1960, 132, 591. Louw, C. W., and Richards, J. F., Analyst, 1972, 97, 334. “Methods of Analysis of the Association of Official Agricultural Chemists,” Tenth Edition, Association Baumann, E. W., Analytica Chim.Ada, 1968, 42, 127. Srinivssan, K., and Rechnitz, G. A., Analyt. Chem., 1968, 40, 509. Beck, M. T., “Chemistry of Complex Equilibria,” Van Nostrand Reinhold Company, London, Sillkn, L. G., and Martell, A. E., Compilers, “Stability Constants of Metal - Ion Complexes,” Special 1957, 19, 49. of Official Agricultural Chemists, Washington, D.C., 1965, p. 34. 1970, p. 78. Publication No. 17, The Chemical Society, London, 1964, p. 264.454 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. VICKERY AND VICKERY Rossotti, F. R., and Rossotti, H., “The Determination of Stability Constants,’’ McGraw-Hill Book Cotton, F. A., and Wilkinson, G., “Advanced Inorganic Chemistry,’’ Second Edition, Wiley Eastern Edmond, C. R., Bull. Aust. Dev. Lab., 1969, 7, 1. Carlson, R. M., and Keeney, D. R., “Instrumental Methods of Analysis of Soils and Plant Tissue,” Mowbray, J. H., “Corning Applications Note,” No. 3, Corning Glass Works, Medfield, Mass., 1969. Frant, M. S., and Ross, J . W., Analyt. Chem., 1968, 7, 1169. Harwood, J. E., Wat. Res., 1969, 3, 273. Schwartzenbach, G., and Flaschka, H., “Complexometric Titrations,” Second English Edition, Schwartzenbach, G., and Flaschka, H., “Complexometric Titrations,” Second English Edition, Schwartzenbach, G., and Flaschka, H., “Complexometric Titraticns,” Second English Edition, Oliver, R. T., and Clayton, A. G., Analytica Chim. Acta, 1970, 51, 409. Buck, M., and Reismann, G., Fluoride, 1971, 4, 5. SillCn, L. G., and Martell, A. E., Compilers, “Stability Constants of Metal - Ion Complexes, Supple- SillCn, L. G., and Martell, A. E., Compilers, “Stability Constants of Metal - Ion Complexes,” Special Eden, T., “Tea,” Second Edition, Longmans, Green and Co., London, 1965, p. 11. Co, Inc., New York, 1961, p. 40. Private Ltd., New Delhi, 1972, p. 437. Soil Science Society of America, Madison, Wisc., 1971, p. 39. Methuen, London, 1969, p. 185. Methuen, London, 1969, p. 184. Methuen, London, 1969, p. 188. ment No. 1,” Special Publication No. 25, The Chemical Society, London, 1971, p. 412. Publication No 17, The Chemical Society, London, 1964, p. 479. Received August lst, 1975 Amended November llth, 1975 Accepted December 18th, 1975

ISSN:0003-2654    

DOI:10.1039/AN9760100445

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst, June, 1976, Vol. 101, pp. 455-457 455 Determination of Iodine in Plant Material by a Neutron-activation Method 0. J o h a n s e n a n d E. Steinnes Institutt for Atomenergi, Isotope Laboratories, Kjeller, Norway A method for the determination of trace amounts of iodine in dry plant material is described. The method is based on the radiochemical separation of iodine-128 using alkaline fusion of the sample in the presence of an iodine carrier, followed by isolation of iodine by a solvent extraction and a precipi- tation step. The precision of the method is about 5% at the 0.05 p.p.m. level and 10% at the 0.005 p.p.m. level of iodine. Results for two international reference plant materials are reported. Neutron-activation analysis is probably the most sensitive analytical technique available at present for the determination of iodine.When very low levels are encountered, it is necessary in most instances to employ a radiochemical separation technique to separate the iodine-128 activity from other activities induced in the sample by neutron activation. Distillation methods have frequently been used for this purpose, and have been applied to plant materials.lS2 Such methods are not easy to carry out, and may not bring about complete radiochemical exchange between the carrier and the induced activity if a quantitative recovery of the iodine is not obtained in the decomposition - distillation step. Alkaline fusion is an alternative to wet digestion with acid for the dissolution of the sample. Ohno3 published a method in which the fresh sample was dried in an oven at 110 “C, ground to a powder and subsequently fused with a mixture of sodium hydroxide and sodium peroxide in a nickel crucible in a muffle furnace at 450 “C.The fusion “cake” was then dissolved in water and a portion was taken for activation analysis. Contamination could be a major problem with that method unless great care was taken to avoid it. Loss of iodine might also be possible. The difficulties outlined above should not be serious if the alkaline fusion is carried out after the irradiation, in the presence of iodine carrier. This approach had previously been applied in the authors’ laboratory for the determination of iodine in and it was thought that a similar method might be used in the analysis of plant material.Experimental Reagents water to give a solution containing approximately 50 mg ml-l of iodine. to yield a concentration of approximately 0.5 pg ml-l of iodine. Irradiation Samples of 0.2-0.5 g of dry plant material were weighed into polyethylene tubes, which were then heat-sealed. Two samples and a standard were irradiated for 15 min in the pneumatic-tube facility of the JEEP-I1 reactor (Kjeller, Norway) at a thermal neutron flux of about 1.5 x 1013 neutrons cm-2 s-l. Iodine carrier solution. Iodine standard solution. Dissolve an accurately weighed amount of potassium iodide in Dilute the iodine carrier solution with 1 M ammonia solution Radiochemical Procedure Transfer the irradiated sample into a nickel crucible containing 1 ml of iodine carrier solution, which has been evaporated to dryness in the presence of excess of alkali, and about 3 g of sodium hydroxide pellets.Moisten the contents with a few drops of water and heat the crucible over an electrothermal burner, with continuous shaking. After the water has been evaporated and the sample has apparently dissolved, cover the crucible and fuse the contents for 5 min. After cooling, dissolve the fused material in about 30ml of water and transfer the resulting solution into a separating funnel containing 25 ml of carbon tetra- chloride. Add about 200mg of solid sodium pyrosulphite (Na,S,O,) to the aqueous phase456 JOHANSEN AND STEINNES : DETERMINATION OF IODINE IN AnaZyst, VoZ. 101 and then add a sufficient volume of 6 M sulphuric acid to make the solution slightly acidic.Add about 100 mg of solid sodium nitrite and 6-7 drops of concentrated nitric acid and extract the elemental iodine into the organic phase. Wash the organic phase once with water and then back-extract the iodine into 20 ml of 0.1 M potassium hydroxide solution. Add 2 ml of 1 M silver nitrate solution, acidify by dropwise addition of concentrated nitric acid and heat to boiling. Filter the solution through a blue-ribbon filter-paper (Schleicher and Schull) and wash the precipitate with water and acetone. Release the precipitate from the filter and transfer it into a glass counting vial. Withdraw a suitable volume of the standard solution with a micropipette and transfer it into a glass counting vial. After the activity measurements, dry the silver iodide precipitate for 10 min at 110 "C and determine the chemical yield by weighing.Activity Measurements sodium iodide (thallium) detector connected to a multi-channel analyser. The gamma-activity from iodine-128 of half-life 25.0 min was measured with a 3 x 3 in Results and Discussion The precision of the method was tested on two standard reference materials, namely Bowen's kale5 and Wheat Flour V-5 from the International Atomic Energy Agency, Vienna. The results obtained are given in Table I and results from duplicate analyses of five grain samples with very low iodine contents are given in Table 11. The precision appears to be of the order of 5% at the 0.05 p.p.m. level and 10% at the 0.005 p.p.m. concentration level of iodine. TABLE I DETERMINATION OF IODINE IN TWO INTERNATIONAL REFERENCE MATERIALS Iodine found, p.p.m.Aliquot No. 1 2 3 4 5 6 7 Mean . . . . .. .. Relative standard deviation, % r Bowen's kale5 0.079 0.075 0.075 0.072 0.073 0.072 0.069 0.074 4.3 I. A.E . A. Wheat Flour V-5 0.002 7 0.003 2 0.002 5 0.003 0 0.002 6 0.002 4 0.002 3 0.002 7 12.4 The accuracy of the method is not easily judged from comparison with results from other laboratories. In an initial report on Bowen's kale,5 an iodine content of 0.080 0.023 p.p.m., in good agreement with the present work, was reported as a mean of results from three laboratories using catalytic techniques. In more recent compilations on kale,6v7 results showing a considerably larger spread were reported. Concerning the Wheat Flour standard, TABLE I1 RESULTS FROM DUPLICATE ANALYSES OF WHEAT AND BARLEY SAMPLES Iodine found, p.p.m.Sample 1 2 Wheat I . . . . . . 0.002 3 0.002 9 Wheat I1 . . . . . . 0.007 3 0.006 2 Wheat I11 . . . . . . 0.003 2 0.003 5 Barley I . . . . . . 0.010 5 0.010 8 Barley I1 . . . . . . 0.006 6 0.004 5June, 1976 PLANT MATERIAL BY A NEUTRON-ACTIVATION METHOD 457 no figures have been published, but preliminary information from the International Atomic Energy Agency,s based mainly on activation analysis results, indicates a value of about 0.003 p.p.m. The method reported in this paper has been found to be very useful in connection with a project involving the determination of halogens in grain samples grown in different parts of Nonvay,S the results of which will be reported elsewhere. The method is straightforward except for the initial phase of the decomposition step, when care must be exercised to avoid too violent a reaction in the crucible, resulting in loss of material.The chemical yield is typically 60-70%. The use of an alkaline rather than an acidic dissolution process may result in a smaller risk of incomplete equilibration between carrier and induced activity affecting the results. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Bowen, H . J. M., Biochem. J., 1959, 73, 381. Tensho, K., and Ko-Ling, Yeh, Radio-Isotopes (Tokyo), 1970, 19, 574. Ohno, S., Analyst. 1971, 96, 423. L%g, J., and Steinnes, E., “Isotopes and Radiation in Soil - Plant Relationships Including Forestry,” Bowen, H. J. M., Analyst, 1967, 92, 124. Bowen, H. J. M., i n Lenihan, J. M. A., and Thomsen, S. J., Editors, “Advances in Activation Bowen, H. J. M., J . Radioanalyt. Chenz., 1974, 19, 215. P a n , R. M., personal communication. L%g, J., and Steinnes, E., unpublished work. International Atomic Energy Agency, Vienna, 1972, p. 383. Analysis,’’ Academic Press, London and New York, 1969, p. 105. Received October 29tk, 1975 Accepted December 17fh, 1975

ISSN:0003-2654    

DOI:10.1039/AN9760100455

出版商:RSC    

年代:1976

数据来源: RSC