ISSN:0003-2654    

DOI:10.1039/AN97601FX037

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

ANALAO 101 (I 207) 761-832 (1 976)ISSN 0003-2654October 1976THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYCONTENTS76176877778679079880380881 5820825828830ORIGINAL PAPERSEvaluation o f Extraction Techniques f o r the Determination o f Metals in AquaticSediments-Haig Agemian and A. S. Y. ChauApplication o f High-performance Liquid Chromatography t o the Separation o fCardenolides and the Assay o f Digoxin in Digitah /anata Leaf-Paul H. CobbIonic Polymerisation as a Means o f End-point Indication in Non-aqueous Therm-ometric Titrimetry. Part IX. Determination o f Dithiocarbamates andPhosphorodithioates-E. J. Greenhow and L. E. SpencerA Digital Logic Automatic Potentiometric Titrator-John T. Stock and K. D. WolterDetermination o f Gold in Blood Fractions by Atomic-absorption SpectrometryUsing Carbon Rod and Carbon Furnace Atomisation-H.Kamel, D. H. Brown,J. M. Ottaway and W. E. SmithA Kinetic Theory of Atomisation f o r Atomic-absorption Spectrometry with aGraphite Furnace. Part IV. Assessment o f Interference Effects-C. W.FullerDevelopment o f Fluxes for the Analysis o f Ceramic Materials by X-ray Fluores-cence Spectrometry-H. Bennett and G. J. OliverFI uo r i metric Deter m i nation of Tet racyc I i n es i n B i o I og i ca I M ate ria I s-H . Po ig erand Ch. SchlatterDetermination of Halogens in Marine Algae by Use o f an lon-selective Electrode-J. N. C. Whyte and J. R. EnglarMethod f o r Determination o f Glyphosate Residues in Natural Waters Basedon Polarography of the N-Nitroso Derivative-Jan 0. Bronstad and Hikon 0.FriestadAuto-standardisation in Spectrophotometry-Michael Thompson and D. J. W. JonesCO M M U N I CAT1 0 ND e t e r m i nation o f D i xa n t h og e n by H i g h - perform a n ce Li q u i d C h r o m a t o g rap h y-R. A. HastyBook ReviewsSummaries of Papers in this Issue-Pages iv, v, viii, xPrinted by Heffers Printers Ltd, Cambridge, EnglandEntered as Second Class at New York, USA, Post Offic

ISSN:0003-2654    

DOI:10.1039/AN97601BX039

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

iv SUMMARIES OF PAPE:RS I N THIS ISSUE October, 1976Summaries of Papers in this IssueEvaluation of Extraction Techniques for the Determination of Metalsin Aquatic SedimentsA brief study was made of the extractability of a large number of metals fromsediments, using different extraction techniques. Total, acid-extractableand cold-extractable metal extraction methods were compared in order toobtain the best technique applicable to environmental analysis. Extractiontechniques for use with easily extractable metals were studied in detail,including their behaviour with respect to organic carbon content and typeof lake or river sediment.HAIG AGEMIAN and A. S. Y . CHAUCanada Centre for Inland Waters, Water Quality Laboratory, 867 Lakeshore Road,P.O. Box 5050, Burlington, Ontario, Canada.Analyst, 1976, 101, 761-767.Application of High- performance Liquid Chromatography to theSeparation of Cardenolides and the Assay of Digoxin in Digitalislanata LeafHigh-performance liquid chromatography on microparticulate silica gel hasbeen used for the assay of digoxin in Digitalis Zanata leaf.The precision ofthe determination has been improved by the incorporation of an internalstandardisation procedure in the chromatographic evaluation of leaf extracts.Factors affecting the extraction of digoxin from the leaf have been investi-gated, and it is concluded that maceration of the leaf with aqueous ethanol,followed by a de-acetylation reaction, is necessary to release the full digoxincontent.Retention data for a large number of cardenolides have been determined inorder to demonstrate the specificity of the assay method, and a relationshipbetween chemical structure and retention time has been revealed.Specimen chromatograms are included in order to illustrate the applicabilityof the chromatographic method to the separation and quantitative evaluationof groups of cardenolides.PAUL H.COBBCentral Analytical Laboratories (Chemical), The Wellcome Foundation Limited,Dartford, Kent.Analyst, 1976, 101, 768-776Ionic Polymerisation as a Means of End-point Indication inNon - aqueous Thermometric TitrimetryPhosphor o dithioatesAmmonium, cadmium, bismuth, iron, lead, manganese, nickel, piperidiniumand zinc dithiocarbamates and nickel and zinc phosphorodithioates have beendetermined in amounts down to 0.002 mmol by catalytic thermometric iodi-metric titration of their solutions in diniethylformamide and acrylonitrile.The end-points in most instances correspond to the formation of polyiodides.The reaction stoicheiometries depend on the nature of the ligand as well as onthe central ion.The method has been compared with aqueous iodimetry,and the results agree to within about 1%.Solutions of dithiocarbamates and phosphorodithioates in hydrocarbonoils, used as oil additives, can be determined in concentrations down to 0.05%.High-boiling alcohols and other oil additives, including hindered phenols,diphenylamine derivatives and phenothiazine, do not interfere. The resultsshow precisions ranging from 0.44 to 2.01(:(,.E.J. GREENHOW and L. E. SPENCERDepartment of Chemistry, Chelsea College, University of London, Manresa Road,London, SW3 6LX.Analyst, 1976, 101, 777-785.Part IX. Determination of Dithiocarbamates anOctober, 1976 SUMMARIES OF PAPERS I N THIS ISSUEA Digital Logic Automatic Potentiometric TitratorA digital logic automatic potentiometric titrator is described and results aregiven for its use in various titrations.JOHN T. STOCK and K. D. WOLTERDepartment of Chemistry, University of Connecticut, Storrs, Conn. 06268, USA.Analyst, 1976, 101, 786-789.VDetermination of Gold in Blood Fractions by Atomic-absorptionSpectrometry Using Carbon Rod and Carbon Furnace AtomisationA comparison of procedures that involve the use of either carbon rod orcarbon furnace atomisation in order to determine, by atomic-absorptionspectrometry, the level of gold in whole blood, plasma and serum frompatients undergoing gold treatment for rheumatoid arthritis is described.A procedure using carbon furnace stomisation is preferred because of itssimplicity and sensitivity. The detection limits for gold, obtained by usingthe preferred procedure, in serum, plasma and whole blood are 0.002, 0.002and 0.004 5 ,ug ml-l, respectively.The relative standard deviations are 1.9%for 0.063,ugml-l in serum, 2% for 0.061 pgm1-l in plasma and 7.3% for0.030pgml-l in whole blood. The method is used to confirm that mostgold is carried in the serum fraction of blood, to determine the gold level inwhite cells and to demonstrate that the gold level in the ultra-filtrate islow.H.KAMEL, D. H. BROWN, J. M. OTTAWAY and W. E. SMITHDepartment of Pure and Applied Chemistry, University of Strathclyde, CathedralStreet, Glasgow, G1 1XL.Analyst, 1976, 101, 790-797.A Kinetic Theory of Atomisation for Atomic- absorptionSpectrometry with a Graphite FurnaceInterference effects in atomic-absorption measurements using electrothermalatomisation are well known but poorly understood. The effects of variationsin the rate of atomisation, the rate of loss of atoms and the efficiency ofatomisation on absorbance uevsus time profiles, in a graphite furnace, havebeen established. These effects are proposed as a basis for establishing thecauses of interferences that occur in atomic-absorption spectrometry using agraphite furnace atomiser.C.W. FULLERTioxide International Limited, Stockton-on-Tees, Cleveland, TS18 2NQ.Part IV. Assessment of Interference EffectsAnalyst, 1976, 101, 798-802.Development of Fluxes for the Analysis of Ceramic Materials byX-ray Fluorescence SpectrometryFrom experience of the use of both lithium metaborate and tetraborate asfluxes for a wide variety of minerals and refractories, it is postulated thatwhen the sample contains a preponderance of acidic oxides (e.g., silica),lithium metaborate has considerable advantages. The melts from meta-borate are much more fluid and decomposition can be carried out by heatingover gas burners. For samples that contain high contents of basic oxides(e.g., carbonate rocks), the more acidic tetraborate is recommended. A 1 + 4mixture of metaborate and tetraborate forms a universal flux for samplescontaining silica and/or alumina as major contents. The most suitable fluxcan be predicted from the acidity or basicity of the sample.H. BENNETT and G. J. OLIVERThe British Ceramic Research Association, Queens Road, Penkhull, Stoke-on-Trent,ST4 7LQ.Analyst, 1976, 101, 803-807

ISSN:0003-2654    

DOI:10.1039/AN97601FP077

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

viii SUMMARIES OF PAPERS I N THIS ISSUE October, 1976Fluorimetric Determination of Tetracyclines in Biological MaterialsA method for the fluorimetric determination of tetracyclines in biologicalmaterials based on solvent extraction of mixed tetracycline - calcium trichloro-acetate ion pairs from aqueous solutions is described. The extraction oftetracycline (TC) and chlorotetracycline (CTC) is almost quantitative, whereasonly very poor extraction occurs with oxytetracycline (OTC). However,saturation of the aqueous phase with sodium chloride results in completeextraction of OTC into the organic phase. This effect enables OTC to bedetermined in the presence of TC and CTC. Ethyl acetate was found to bethe most suitable extractant. The fluorescence of the organic phase ismeasured after addition of magnesium ions and a base.The excitation maxima of all three tetracycline derivatives are a t about400 nm and emission maxima are a t 500 nm.They differ only very slightlyand cannot be used for differentiation bet ween the derivatives. The followingdetection limits for CTC in biological materials were found: serum, 0.05pg ml-l; muscle tissue, 0.125 pg g-l; kidney, 0.15 pg g-l; liver, 0.3 p g g-l; andmilk powder with high fat content, 2 pgg-l. Recoveries were between 30and 45%.H. POIGER and Ch. SCHLATTERInstitute of Toxicology, Federal Institute of Technology and University of Zurich,8603 Schwerzenbach-Zurich, Switzerland.Analyst, 1976, 101, 808-814.Determination of Halogens in Marine Algae by Use of anIon- selective ElectrodeA method for determining the concentrations of chloride, iodide and ether-extractable bromide in marine algae is described. Alkali fusion and ashingof the dry algal tissue or an ether extract of the algal tissue affords aqueoussolutions of the total and organic solvent soluble halides, respectively, whichare subsequently determined by argeritimetric titration using an iodide-selective electrode.The accuracy of t:he procedure is established by theanalysis of standard compounds and the simplicity, reliability and convenienceof the method are confirmed by the analysis of eleven species of marine algae todetermine their halogen contents.J. N. C. WHYTE and J. R. ENGLARFisheries and Marine Service, Department of the Environment, VancouverLaboratory, 6640 N.W.Marine Drive, Vancouver, B.C., Canada, V6T 1x2.Analyst, 1976, 101, 815-819.Method for Determination of Glyphosate Residues in NaturalWaters Based on Polarography of the N-Nitroso DerivativeA differential pulse polarographic method for the determination of residuesof glyphosate herbicide in natural waters down to 35 pgl-1 is described.The material is concentrated by chromatography on an anion-exchange column,which at the same time serves as the only clean-up step required. A polaro-graphically active glyphosate derivative is obtained by nitrosation performeddirectly in a fraction of the eluate from the column. The precision in termsof relative standard deviation varies from 22.8% a t the detection limit to2.3% at 210 pg I-'.JAN 0.BR0NSTAD and HAKON 0. FRIESTADChemical Research Laboratory, Agricultura.1 University of Norway, 1432 AS-NLH,Norway.Analyst, 1976, 101, 820-824.The total analysis time for one sample is 3-4 hx SUMMARIES OF PAPERS I N THIS ISSUE October, 1976Auto -standardisation in SpectrophotometryPoor precision has been obtained in the spectrophotometric determinationof fluoride with a version of the alizarin fluorine blue method in whichthe total volumes of the solutions are small and relative variations in the finalvolumc were substantial. In some analytical methods, this kind of situationcan be ameliorated by means of internal standardisation, but the wideabsorption bands typical of ui traviolet - visible light spectrophotometryrender the technique impracticable for this application. However, the reagentitself can be used as an internal standard if absorbances are measured a tan isosbestic point.MICHAEL THOMPSON and D. J. W. JONESApplied Geochemistry Research Group, Department of Geology, Imperial Collegeof Science and Technology, London, SW7 2BP.Analyst, 1976, 101, 825-827.Determination of Dixanthogen by High-performanceLiquid ChromatographyCommunicationR. A. HASTYDepartment of Chemistry, University of the Witwatersrand, Johannesburg,South Africa.Analyst, 1976, 101, 828-829

ISSN:0003-2654    

DOI:10.1039/AN97601BP081

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

OCTOBER 1976 The Analyst Vol. 101 No. 1207 Evaluation of Extraction Techniques for the Determination of Metals in Aquatic Sediments Haig Agemian and A. S. Y. Chau Canada Centre for Inland Waters, Water Quality 1-aboratory, 867 Lakeshore Road, P.O. Box 5050, Burlington, Ontario, Canada A brief study was made of the extractability of a large number of metals from sediments, using different extraction techniques. Total, acid-extractable and cold-extractable metal extraction methods were compared in order to obtain the best technique applicable to environmental analysis. Extraction techniques for use with easily extractable metals were studied in detail, including their behaviour with respect t o organic carbon content and type of lake or river sediment. In order to determine metals in sediments by means of atoniic-absorption spectroscopy, it is first necessary to bring them into solution.Extraction methods have been well documented1 and involve fusion or acid dissolution; the latter type of technique has several advantages. Mineral acids can be obtained in a sufficiently pure form that their use does not introduce any appreciable impurities and acid decomposition methods, unlike fusion techniques, do not allow large amounts of salts to be introduced into the solution; a high salt content can cause instability and lead to high instrument background readings. In addition, fusion techniques are restricted to the determination of the total metal content of silicates only. On the other hand, the concentrations of acids can be varied by dilution and therefore selective dissolution of several components of sediments can be effected.In classical chemical analysis,l as well as in geochemical2 and ~ o i l ~ - ~ analysis, five mineral acids, namely hydrochloric, nitric, sulphuric, perchloric and hydrofluoric acids, have been very widely used. For the simultaneous extraction of a large number of metals, sulphuric acid is disadvantageous, owing t o the formation of some insoluble sulphates.2 Hydrofluoric acid has the one notable property of dissolving silica. Thus, it has been used in conjunction with nitric, hydrochloric or perchloric acid in the total decomposition of silicates. Nitric acid has been used separately or with either hydrochloric or perchloric acid. Such methods provide a high degree of metal extraction but do not dissolve silicates completely; they destroy organic matter, dissolve all precipitated and adsorbed metals and leach out a certain amount of the metals from the silicate lattice.Therefore, depending on the strength and type of acid mixture used, a rock mineral would be partially attacked if these methods were applied to it. Much weaker extracting agents have also been used to extract metals of a non-residual origin only. and 0.05 N ethylene- diaminetetraacetic acid697 take little account of rock types as they dissolve complexed, adsorbed and precipitated metals in sediments with minimum attack on the silicate. A mixture of 1 N hydroxylammonium chloride and 25% acetic acid8sg has been used to dissolve ferromanganese and carbonate minerals and adsorbed trace elements in sediments and is similar to the above two methods.The intention of this work was to study the relative simultaneous extraction of a large number of metals from aquatic sediments in order to obtain a rapid, simple technique for measuring non-residual metal. The non-residual metal phase includes the exchangeable metal, carbonate, organic and sulphide phases, as well as oxides and hydroxides of manganese and iron. Trace-metal partitioning schemes for the selective dissolution of the different phases of sediments, such as those reported by Presley et nZ.1° or Gupta and Chen,ll are beyond the scope of this study. Because of the time and expense required for such schemes, they can be used only for specialised studies involving a limited number of samples.761 Methods involving the use of 0.5 N hydrochloric762 AGEMIAN AND CHAU : EVALUATION OF E:XTRACTION TECHNIQUES Analyst, Vol. 101 Experiment a1 Apparatus photometer, equipped with a triple-slot air - acetylene burner. lamps were used for all determinations. digestion bomb (Parr Instrument Co., Moline, Ill., USA) .12 All determinations were made on a Perkin-Elmer, Model 503, atomic-absorption spectro- Perkin-Elmer Intensitron Digestions for total metal determination were carried out in a PTFE Parr 4745 acid- Reagents High-purity, certified reagents were used for all analyses. Hydrochloric acid, 36 N. Nitric acid, 16 N. Aqua regia (hydrochloric - nitric acid, 3 + 1). Perchloric acid, 60%. Hydropuoric acid, 48%. Acetic acid, glacial (99.8%).Hydroxylammonium chloride, crystals. Ethylenediaminetetraacetic acid (disodium salt), crystals. Standard metal solutions. Standards were prepared by serial dilution of 1 000 mg 1-1 All standard solutions were prepared containing the same reagents metal stock solutions. that were added to the samples. Procedure of less than 80 mesh (<0.177 mm). All sediment samples were dried in air at room temperature and sieved to a particle size Cold-extractable metal content A 5-g amount of sediment was shaken overnight at room temperature with 100-ml solutions of: 0.05 N ethylenediaminetetraacetic acid at pH 4.8; 1 N hydroxylammonium chloride plus 25% acetic acid; and 0.5 N hydrochloric acid. Acid-extractable metal content A l-g amount of sediment was digested as follows: with 25 ml of nitric acid, boiled to dryness twice; with 25 ml of aqua regia, boiled to dryness twice; with 25 ml of nitric - perchloric (1 + 1) acid, boiled to dryness twice.The residue was dissolved in dilute hydrochloric acid in each instance. Total metal content The <80-mesh sample was crushed to about 200 mesh and 100 mg of this powder were digested with 6 ml of hydrofluoric acid, 4 ml of nitric acid and 1 ml of perchloric acid in a PTFE bomb, as has been described by Agemian and Chau.12 Results and Discussion For geochemical and environmental purposes, the preferred analytical technique is that which gives the greatest contrast between anomalous and background samples. Two factors affect the contrast: the particle size of the sample and the form of metal ( i e ., cold-extractable or total) chosen for analysis. Firstly, Oliver13 showed that the size of sediment particles strongly influenced the extractable metal content of the samples. Hawkes and Webb14 have shown that the (8O-mesh portion of a sediment provides the greatest contrast between anomalous and background samples. Thus, the <80-mesh portion of the sample was analysed in this study. To facilitate the dissolution necessary for determining the total metal, a sub-sample from the (80-mesh portion was ground to about 200 mesh. Secondly, it has been shown that partial extraction techniques provide higher contrast7J4 than total analyses. Therefore, in the present work, the cold-extractable techniques wereOctober, 1976 FOR DETERMINATION OF METALS IN AQUATIC SEDIMENTS 763 studied in more detail.However, for completeness, it was necessary to study extraction by use of these techniques relative to the stronger leaching agents. This relationship between methods is different for each metal, depending on its geochemistry. Table I shows the degree of extraction of several metals by use of the methods under consideration. The sample used was a typical sediment analysed in this laboratory. The methods used were three of the four types of extraction techniques, namely, those which extract total (last column), acid-extractable (columns 5-7) and cold-extractable (columns 2 4 ) metal. The type of extraction not studied was that of exchangeable metal, with com- pounds such as ammonium acetate. This last method is very useful in soil studies but does not provide sufficient information for environmental and geochemical studies.The results obtained from total metal extractions reflect the rock type,7 as well as mineralisation. Cold- extractable methods are designed to indicate all but the rock type of the samples. Thus, any anomalies occurring with the latter methods are easier to interpret as natural variations due to changes in bedrock are virtually eliminated.7 The acid-leaching techniques (columns 5-7) show varying degrees of attack oil the crystal lattice and thus give an intermediate value between cold-extractable and total metal extractions. The results (Table I) reflect this postulation. Metal A1 Ca Fe :: Ba Cd co Cr cu Li Ni Pb Zn TABLE I COMPARISON OF THE EXTRACTION OF METALS FROM A LAKE ONTARIO SEDIMENT CONTAINING 2.8% OF ORGANIC CARBON AND 0.17% OF ORGANIC NITROGEN USING DIFFERENT EXTRACTION SYSTEMS 1 N NH,OH.HCl + 25% CH,COOH 0.05 N EDTA 400 970 21 400 22 700 4 800 6 800 3 800 4 800 550 620 20 80 2.2 2.2 6 6 5.5 15 25 16 0.4 1.0 16 20 55 52 97 122 All results are in mg 0.5 N HCI 4 000 23 100 12 500 6 900 620 100 2.0 8 22 33 2s 56 149 5.4 HNO, (boiling) 15 500 27 000 32 000 12 000 750 1100 31 49 44 33 4.0 32 70 218 kg-l.Aqua regia (boiling) 25 300 27 000 32 000 13 000 800 1600 31 48 40 38 4.0 25 70 206 HNO, - HCIO, HF - HNO, - HC10, (6 + 4 + 1) PTFE bomb (1 + 1) (boiling) 38 500 30 000 34 000 12 000 750 2 600 36 15 50 44 6.0 33 $0 229 43 000 30 000 42 000 16 000 4 500 2 $00 40 200 110 50 50 200 100 290 It is apparent from Table I that for the sample studied, perchloric acid does not liberate all of the metal from the silicate matrix.The amount of metal extracted by perchloric acid depends on the type of sample (both type of mineral and organic matter content). For many types of sample, this acid is suitable for total metal extraction. For example, Hossner and Ferrara15 have used digestion with perchloric acid for the total extraction of manganese from soils. For some geochemical or environmental purposes this acid is suitable as a measure of total metal content as the rock type is reflected in the results obtained by its use. The nitric acid used in this method (Table I, column 7') serves only as a safety measurel6-18 if large amounts of organic matter are present.2 The use of perchloric acid alone for the sample in Table I gave results identical with those obtained by using nitric - perchloric acid.The low boiling-p~intl~ of chromyl chloride (CrO,Cl,) of 116 "C, compared with about 200 "C for perchloric acid,2 probably results in volatilisation losses. With nitric acid or aqua regia, these losses do not occur because the boiling-points2 of nitric and hydrochloric acids are lower. Holmes et aL9 used boiling nitric acid to extract the zinc and cadmium that is bound in organic or sulphide compounds. They extracted about 85% of the zinc and cadmium from the sedi- ments. Jones20 reported extractions of 75% of the zinc and 60% of the cadmium from some bottom sediments. The fraction of the total metal extracted by any partial extraction technique will depend on the type of sample used.However, it is widely agreed that There was one unsatisfactory recovery with perchloric acid, vix., that of chromium. Aqua regia and nitric acid are weaker extracting agents than perchloric acid.764 AGEMIAN AND CHAU : EVALUATION OF EXTRACTION TECHNIQUES Anahst, VoZ. 101 leaching with nitric acid or aqua regia will give only partial extraction. Aqua regia (Table I) is a stronger oxidising and extracting agent than nitric acid as a result of the presence of free or nascent chlorine.2y21 It should be pointed out that nitric acid, aqua regia and perchloric acid have their strongest leaching effect when they are boiling; perchloric acid, especially, is a strong leaching, dehydrating and oxidising agent only when it is hot and concentrated. Sample 4 5 6 11 12 13 TABLE 11 EXTRACTION OF MAJOR METALS FROM SEDIMENTS OF VARIOUS ORGANIC CARBON CONTENTS WITH FOUR EXTRACTION SYSTEMS All results are in mg kg-l.Organic carbon, 0.15 0.43 0.90 2.25 2.47 2.81 2.95 3.15 3.35 4.10 4.85 10.75 12.30 % 7 a 99 140 140 320 900 5 50 240 160 1090 100 400 400 380 Manganese Iron 7 r-------- - b c d a C 120 120 227 990 1400 1590 11 700 220 300 655 1800 3770 10940 35000 160 160 253 1980 2580 4170 14400 380 380 718 3360 4160 8790 30900 1010 1070 1100 5780 7 740 12900 36 300 620 620 787 4750 6780 12520 34700 300 280 633 3190 3740 8970 30800 200 220 600 3780 4540 8770 27 300 1210 1170 1340 8 300 10 500 17 500 38800 120 140 417 6750 7110 9960 27500 440 440 1070 3140 3 570 7 710 24600 440 480 - 3380 3770 8120 - 440 420 533 4180 3 790 6160 22 300 7 a 9 70 660 140 790 340 400 620 1400 1 400 7 500 650 1100 1200 Aluminium b C 1400 1700 1800 6400 240 1000 1500 4800 930 4000 970 4000 1100 5 500 2 600 5400 2 TOO 7600 9 700 15 000 1200 5 100 1900 6 500 1500 4000 d 20 300 54 700 41 500 45 100 49 900 49 700 48 700 39 500 43 400 30 300 48 500 38 800 - a, 0.05 N ethylenediaminetetraacetic acid (pH 4.8); b, 1 N hydroxylammonium chloride + 257; acetic acid; c, 0.5 N hydrochloric acid; d, HF - HNO, - HC10, (total metal).As was pointed out earlier, the cold-extraction methods are the weakest as they do not attack the silicate lattice appreciably. With such methods it is usually desirable to extract the non-residual metal from the sediments. The three methods of this type studied are compared in Table 11, together with a total extraction method for manganese, iron and aluminium, for thirteen sediments of different types from a variety of locations across Canada. The extraction efficiency of the three methods is in the increasing order 0.05 N ethylenediaminetetraacetic acid, 1 N hydroxylammonium chloride plus 25% acetic acid and 0.5 N hydrochloric acid.This trend correlates with the decreasing pHs of 4.8, 1.5 and 0.3, respectively, for the above reagents. From their chemical properties, it would be expected that these methods would extract the adsorbed, precipitated and complexed metals. Table I11 gives a description of the samples used. TABLE I11 DESCRIPTION OF SAMPLES IN TABLES 11, IV AND V Sample Organic carbon, % Silicon, % 1 0.15 31.3 2 0.43 24.2 3 0.90 28.0 4 2.25 22.5 5 2.47 19.0 6 2.81 20.6 7 2.95 20.3 8 3.15 22.8 9 3.35 20.3 10 4.10 20.8 11 4.85 19.9 12 10.75 20.0 13 12.30 20.3 Sample description Orange sand Clay Sand Silt Clay Silt Silt and fine sand Silt Dark brown clay Grey clay Silt and clay Silt and clay Silt and clay Sample location* Lake Superior Rideau River Calgary Rideau River Rideau River Lake Ontario Rideau River Ottawa River Lake Huron Cardigan Bay Rideau River Rideau River Rideau River * All sample locations are in Canada.Bradshaw et a1.' stated that dilute hydrochloric acid may attack some of the less resistant silicates, such as layered silicates. Ray et aZ.22 showed that dilute hydrochloric acid will attack the lattice structures of certain clay minerals. However, in spite of the above, Billings and Ragland23 used 1 N hydrochloric acid as their tests indicated that a brief treat-October, 1976 FOR DETERMINATION OF METALS IN AQUATIC SEDIMENTS 765 ment with this solution did not leach significant amounts of metals from well crystallised common clays.Also, Bradshaw et a1.' showed that the 0.5 N hydrochloric acid attack gave no indication of the various rock types. The results given in Table I1 indicate that the cold-extraction methods extract only a very small fraction of the total aluminium from the sediments (compare methods a, b and c with method d). On the other hand, a considerable amount of the manganese is extracted. The fraction of total iron extracted by use of these methods is intermediate between manganese and aluminium.Such results are expected from the geochemistry of these elements. Chester and Hughes24 have shown that for a North Pacific deep-sea clay, SSyO of the manganese is non-residual, while only 5% of the iron is from that phase. Aluminium is the predominant metal in aluminosilicate clays and is found mainly in the crystal lattice of the sediment. Thus, both iron and aluminium are mainly found in the residual phase. About 14 -+ 10, 17 j, 10 and 32 & 6% of the total iron was extracted by methods a, b and c, respectively (Table 11). For aluminium, about 2 & 1.5, 4 j, 4 and 10 & 4% of the total was extracted, respectively, for methods a, b and c. This information shows that these methods do not affect the crystalline structure appreciably. To substantiate further the above contention, the amount of silicon extracted by the total metal and 0.5 N hydrochloric acid-extraction methods was measured.Table I11 gives the total amount of silicon in each of the samples. It was found that the mean amount of silicon extracted by use of the 0.5 N hydrochloric acid method was about 1% of the total. Table IV shows the results obtained for the same samples, with the three partial extraction methods, for the seven trace elements studied. The order of efficiency of extraction of the methods is again increasing from 0.05 N ethylenediaminetetraacetic acid to 0.5 N hydro- chloric acid except for copper, where 1 N hydroxylammonium chloride plus 25% acetic acid gives the lowest extraction. Chowdhury and Bose25 have shown that hydrochloric acid (pH 1 .O) liberates copper complexed with humic compounds isolated from soils.However, 1 N hydroxylammonium chloride plus 25% acetic acid mixture is not a strong enough com- plexing agent to compete in complex equilibria, nor acidic enough to cause dissociation of the natural complexes. The result of this effect is seen with copper (which correlates highly with organic matter26) where ethylenediaminetetraaacetic acid, which is the weakest extracting agent, shows higher values for copper than the hydroxylammonium solution. The samples TABLE IV EXTRACTION OF TRACE ELEMENTS FROM SEDIMENTS OF VARIOUS ORGANIC CARBON CONTENTS WITH THREE EXTRACTION SYSTEMS All results are in mg kg-1. Organic carbon, Sample % 1 0.15 2 0.43 3 0.90 4 2.25 9 2.47 6 2.81 7 2.95 8 3.15 9 3.35 10 4.10 11 4.85 12 10.75 13 12.30 Cadmium A a b c 0.4 0.6 0.6 0.4 0.4 0.4 0.6 0.8 0.8 0.0 0.0 0.0 2.0 2.0 2.0 2.2 2.2 2.2 0.6 0.4 0.6 1.0 0.8 1.0 1.2 1.2 1.2 1.0 0.8 1.0 0.2 0.2 0.2 0.2 0.2 0.2 2.0 1.2 2.0 Chromium * a b c 0.8 1.8 2.4 1.2 4.417 1.4 2.0 2.8 1.4 2.8 9.6 4.214 20 5.515 22 1.8 2.8 8.7 2.6 4.710 1.2 4.8 16 4.6 10 12 1.6 2.4 8.4 1.0 2.8 9.1 3.2 4.4 6.2 Copper - a b c 8 8 9 12 16 24 8 4 9 8 2 12 22 16 28 25 16 33 10 2 14 13 3 19 31 20 39 59 23 74 11 1 13 15 1 15 19 2 20 Cobalt Nickel Lead Zinc --l---4--7* a b c a b c a b c a b c 0 2 2 4 0 4 6 6 6 6 7 8 4 6 8 4 8 1 8 8 1 0 1 0 3 9 3 2 4 4 4 4 4 8 14 10 1 2 14 17 25 4 4 8 2 2 10 20 18 16 14 17 36 6 6 8 18 20 26 54 50 56 92 127 175 6 6 8 16 20 28 55 52 56 97 122 149 4 4 8 4 4 10 16 16 17 13 18 39 2 4 4 4 4 10 22 22 26 39 57 84 6 8 10 20 24 32 40 36 40 30 43 76 457 434 478 10 10 16 22 22 22 39 39 54 4 2 6 2 4 10 14 14 16 15 18 47 6 4 8 4 4 10 28 24 28 27 32 61 4 2 6 8 6 12 54 50 54 114 126 141 a, 0.05 N ethylenediaminetetraacetic acid (pH 4.8); b, 1 N hydroxylammonium chloride + 25% acetic acid; c, 0.5 N hydrochloric acid.listed in Table IV have a wide range of organic matter contents. Table V presents the results for the extraction of copper by the three methods related to results for extraction with 0.5 N hydrochloric acid. It can be seen that as the organic matter content increases, the relative extraction of copper by 1 N hydroxylammonium chloride plus 25% acetic acid is reduced. This evidence confirms the inability of this last solution to extract copper from its organic complexes.However, 0.5 N hydrochloric acid and 0.05 N ethylenediaminetetraacetic acid766 AGEMIAN AND CHAU : EVALUATION OF EXTRACTION TECHNIQUES AnaZyst, VoZ. 101 TABLE 'V EFFECT OF ORGANIC MATTER ON THE EXTRACTION OF COPPER BY THREE METHODS Data for copper from Table I11 referred to the results with the 0.5 N hydrochloric acid method. Coppe:r/mg kg-l by method a, b or c Copper/mg kg-l with 0.5 N HCl extraction I A \ Sample 1 2 3 4 6 6 7 8 9 10 11 12 13 Organic matter, yo 0.16 0.43 0.90 2.26 2.47 2.81 2.95 3.16 3.35 4.10 4.85 10.76 12.30 a, 0.05 N ethylenediaminetetraacetic acid; b, c , 0.6 N hydrochloric acid. a 0.90 0.50 0.90 0.67 0.79 0.76 0.71 0.68 0.79 0.80 0.85 1.00 0.95 b 0.90 0.67 0.44 0.17 0.57 0.48 0.14 0.16 0.51 0.32 0.08 0.07 0.10 C 1 .oo 1.00 1.00 1 .oo 1.00 1 .oo 1.00 1.00 1 .oo 1.00 1.00 1 .oo 1 .oo 1 N hydroxylammonium chloride + 26% acetic acid; liberate copper from its stable organic complexes.This property is desirable because organic matter plays a very important role in smaller lakes and rivers. The 1 N hydroxylammonium chloride plus %yo acetic acid is a special solution for the dissolution of ferromanganese minerals, the hydroxylammonium chloride dissolving the manganese oxide phase. Indeed, Chao2' used this solution for the selective dissolution of manganese oxides from ferromanganese minerals. The role of the 25% acetic acid is to dissolve the iron oxide phases. Thus, the mixture of the above reagents satisfactorily dissolved ferromanganese minerals and is suitable for other metals that do not form strong complexes with organic matter.Because of this property, it is not suitable for the simul- taneous extraction of a large number of metals. Conclusions The extracting efficiency of the methods for the ten metals studied is in the decreasing order hydrofluoric - perchloric - nitric acid mixture, boiling perchloric - nitric acid mixture, boiling aqua regia, boiling nitric acid solution, cold 0.5 N hydrochloric acid solution, cold 1 N hydroxylammonium chloride plus 25% acetic acid solution and 0.5 N ethylene- diaminetetraacetic acid solution. The two exceptions to this order are boiling perchloric - nitric acid solution if used'for chromium and cold 1 N hydroxylammonium chloride plus 25% acetic acid solution if used for copper.With the former chromium is lost by volatilisation and with the latter copper in the form of organic complexes is not extracted. Although boiling perchloric - nitric acid may give total extraction of some metals from some non-resistant samples, complete destruction of the silica matrix by hydrofluoric acid is necessary for total extraction of all metals of interest. Of the cold-extractable metal extraction techniques, 1 N hydroxylammonium chloride plus 25% acetic acid is inadequate for the simultaneous extraction of a large number of metals from a variety of sample types because it does riot extract copper that has been complexed by organic matter. Both 0.5 N hydrochloric acid and 0.05 N ethylenediaminetetraacetic acid are suitable for the simultaneous extraction of ten elements from the adsorbed, organic and precipitated phases of aquatic sediments.The 0.5 N hydrochloric acid method is preferred because, owing to some natural processes, the adsorbed metals in some sediments are bonded more strongly? rendering the 0.05 N ethylenediaminetetraacetic acid extraction method incapable of extracting them. The extraction methods that are the most informative for environmental purposes are the total metal and cold or easily extractable metal extraction techniques. The former is well defined and includes both metals from the rock matrix and the non-residual metals (it?.,October, 1976 FOR DETERMINATION OF METALS IN AQUATIC SEDIMENTS 767 those adsorbed from the aqueous medium). The latter extraction techniques show no association with the type of rock forming the sediment and give results only for the weakly held metals, which include those originating from polluted waters.The authors are grateful for the critical review and valuable suggestions of Y. K. Chau and to L. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. -.a Pielacz for her secretarial help. References Lundell, G. E. F., and Hoffman, J- I., “Outlines of Methods of Chemical Analysis,’’ John Wiley, Maxwell, J. A., “Rock and Mineral Analysis,” Interscience, New York, 1968. Black, C. A., “Methods of Soil Analysis, Part 2, Chemical and Microbiological Properties,” American Hesse, P. R., “A Textbook of Soil Chemical Analysis,” Chemical Publishing Co., New York, 1971. Jackson, M.I,., “Soil Chemical Analysis,”’ Prentice-Hall, Englewood Cliffs, N. J., 1958. Maynard, D. E., and Fletcher, W. K., J . Geochem. Explor., 1973, 2, 19. Bradshaw, P. M. D., Thomson, I., Smee, B. W., and Larsson, J. O., J . Geochem. Explor., 1974, Chester, R., and Hughes, M. J., Chem. Geol., 1967, 2, 249. Holmes, C. W., Slade, E. A., and McLerran, C. J., Envir. Sci. Technol., 1974, 8, 255. Presley, B. J., Brooks, R. R., and Kappel, H. M., J. Mar. Res., 1967, 25, 355. Gupta, S. K., and Chen, K. Y., E m i r . Lett., 1975, 10, 129. Agemian, H., and Chau, A. S. Y., Analytica Chim. Acta, 1975, 80, 61. Oliver, B. G.. Envir. Sci. Technol., 1973, 7, 135. Hawkes, H. E., and Webb, J. S., “Geochemistry in Mineral Exploration,” Harper and Row, New York, 1962. Hossner, L. R., and Ferrara, L. W., Atom. Absorption Newsl., 1967, 6, 71. “Properties and Essential Information for the Safe Handling and Use of Perchloric Acid Solution,” Chemical Safety Data Sheet SD-11, Manufacturing Chemists Association, Washington, D.C., 1965. Analytical Methods Committee, Analyst, 1959, 84, 215. Smith, G. F., Analytica Chim. A d a , 1953, 8, 397. Gorsuch, T. T., Analyst, 1962, 87, 112. Jones, A. S. G., Mar. Geol., 1973, 14, 1. Taylor, F. S. , “Inorganic and Theoretical Chemistry,’’ Heinemann, London, 193 1. Ray, S., Gault, H. R., and Dodd, C. G., Am. Miner., 1957, 42, 681. Billings, G. K., and Ragland, P. C., Chem. Geol., 1968, 3, 135. Chester, R., and Hughes, M. J.. Deep Sea Res., 1966, 13, 627. Chowdhury, A. N., and Bose, B. B., Geochem. Expl C I M Special, 1971, 11, 410. Schnitzer, M., and Skinner, S. I. M., Soil Sci., 1966, 102, 361. Chao, T. T., Soil Sci. Soc. Am. Proc., 1972, 36, 764. New York, 1938. Society of Agronomy, Madison, Wisc., 1965. 3, 209. Received March 22nd. 1976 Accepted May 24th, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100761

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

768 A.nalyst, October, 1976, Vol. 101, 99. 768-776 Application of High-performance Liquid Chroma- tography to the Separation of Cardenolides and the Assay of Digoxin in Digita/is /anata Leaf Paul H. Cobb Central Analytical Laboratories (Chemical), The Wellcomc: Foundation Limited, Dartford, Kent High-performance liquid chromatography on microparticulate silica gel has been used for the assay of digoxin in Digitalis Zanata leaf. The precision of the determination has been improved by the incorporation of an internal standardisation procedure in the chromatographic evaluation of leaf extracts. Factors affecting the extraction of digoxin from the leaf have been investi- gated, and it is concluded that maceration of the leaf with aqueous ethanol, followed by a de-acetylation reaction, is necessary t o release the full digoxin content.Retention data for a large number of cardenolides have been determined in order to demonstrate the specificity of the assay method, and a relationship between chemical structure and retention time has been revealed. Specimen chromatograms are included in order to illustrate the applicability of the chromatographic method to the separation and quantitative evaluation of groups of cardenolides. Digitalis Zanata leaf contains a mixture of cardioactive glycosides, which are classified into five series on the basis of the steroidal aglycone. The most abundant and important of these are the “C” series glycosides, among which the secondary glycoside digoxin is a particularly useful drug in the treatment of auricular fibrillation and congestive heart failure.Digoxin is excreted more quickly than most other Digitalis glycosides and is therefore less cumulative. Chromatography has long been recognised as the most valuable tool in the resolution and quantitative evaluation of cardiac glycoside mixtures. The method of Zaffaroni et aZ.1 for the separation of steroidal mixtures on filter-paper impregnated with formamide or ethylene glycol was used by Reichstein and Schindler2 for the chromatography of Digitalis glycosides. A number of author^^-^ have subsequently used paper chromatography to determine quantitatively Digitalis glycosides by elution of the resolved compounds followed by various colorimetric assay procedures. Thin-layer chrornatography on silica gel was used by Stahl and Kaltenbach7 in order to separate low loadings of cardiac glycoside mixtures, and thin- layer chromatographic separations have also been used in quantitative analysis by removal and elution of the resolved bands for colorimetric determination.8-1° Direct densitometry of developed thin-layer chromatograms is possible but, in general, the complexity of Digitalis extracts renders investigation of the individual components difficult and imprecise.Evans et al.ll simplified extracts of Digitalis purpurea by hydrolysing the glycosides to the parent aglycones, which they then measured by densitometry of thin-layer chromatograms. Extra precision was sought by incorporation of an internal standard, and a coefficient of variation of 4% was obtained for the method.A disadvantage of methods employing a hydrolytic procedure is the possibility of the formation of anhydrogenins. Stoll et aZ.12 used the differences in polarity of the cardiac glycosides to separate them on silica-gel columns, and column chromatography has been used for the enrichment of cardiac glycoside extracts prior to their evaluation by paper chr~matographyl~ or thin-layer chromatography . l4 Gas - liquid chromatography has been used to separate simple mixtures of cardiac glycosides,15~1~ but it involves prior formation of trimethylsilyl derivatives and has the added disadvantage that the elevated temperatures used can lead to thermal elimination of the C,, tertiary hydroxyl group. The advent of high-performance liquid chromatography (HPLC) has introduced a powerful technique for the resolution of complex mixtures of polar compounds.Evans17 has used ion-exchange HPLC in order to separate the components of the “A” series cardenolides, and Lotscher et aZ.18 have reported the use of chemically bonded polar phases and non-polarCOBB 769 reversed phases for the separation of a number of cardiac glycosides and aglycones. Recently Castlelg has reported the use of HPLC for the quantitative determination of low levels of digoxin, digitoxin and their metabolites, while Lindner and Frei20 have pub- lished HPLC systems for the separation of some Digitalis cardenolides on silica gel. In this paper the use of HPLC on silica gel to provide a rapid and precise method for the determination of the potential digoxin content of samples of D.Zanata leaf and the use of the method to establish the optimum conditions for the analytical-scale extraction of digoxin are described. A wide range of cardenolides have been investigated in order to establish the specificity of the method, and the retention data obtained have been correlated with chemical structure. Experimental Apparatus The HPLC apparatus used was constructed from a CE 210 coil pump fitted with an injection port (Cecil Instruments Ltd.) and a chromatographic column made from a 250-mm length of 4 mm i.d. precision-bore, seamless, stainless-steel tubing. All connections were made with low dead-volume stainless-steel couplings (Reeve Angel Scientific Ltd.) . The injections were made with a 10-pl syringe, Type 701N (Hamilton Micromesure BV), through a silicone rubber septum faced with PTFE (Hamilton Micromesure BV), and the column eluate was monitored by using a CE 212 variable-wavelength spectrophotometer fitted with a 10-pl flow cell (Cecil Instruments Ltd.).Results were displayed on an Autograph S pen recorder (Shandon Southern Instruments Ltd.). Reagents and Materials All materials were of analytical-reagent grade unless otherwise stated. Acetic acid, glacial. Chloroform. Cyclohexane, spectroscopic grade. Ethanol, absolute. Ethanol, 2074 V/V. Methanol. Tetrabromoethane. Tetrachloroethylene, spectroscopic gyade. Sodium kydroxide solution, 0.35 M. Hydrochloric acid, 0.35 M. Silica gel, average particle size 10 pm. Purified by passing through a silica-gel column. LiChrosorb SI 60 (E.Merck). Reference Compounds The majority of the reference compounds examined were obtained from Sandoz Ltd. and Boehringer-Mannheim GmbH. The mono- and bis-digitoxosides of digitoxigenin and gitoxigenin were prepared by hydrolysis of the secondary glycosides according to the methods of Kaiser et aL2I The standard digoxin sample used was a British Chemical Reference Sample, and the sulphamethoxazole was BP quality from Hoffmann - La Roche. Extraction of Digoxin from D. Zanata Leaf The hydrolysis of lanatoside C to form digoxin proceeds in two stages and involves the enzyme digilanidase, which occurs naturally in the leaf. The naturally occurring precursor of digoxin in D. lanata is lanatoside C. Desacetyl lanatoside C Lanatoside C \ e Digilanidase -* D igox in Acetyl digoxin770 COBB : HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY FOR SEPARATION Analyst, Vol.101 Partial hydrolysis of lanatoside C to digoxin occurs when the harvested D. Zanata leaf is dried, but it is necessary to allow for hydrolysis during the extraction procedure in order to release the full amount of digoxin. Any alkaline: treatment during the extraction process must be carefully controlled in order to avoid hydrolysis of the butenolide ring. Solvents such as water, methanol, ethanol and ethyl acetate have been used extensively to extract cardenolides from plant material. A variety of methods have been used to clean up the extract, including precipitation of saponins with lead acetate and extraction of fats and resins with light petroleum, diethyl ether or hexane.Tantivatana and Wright4 preferred extraction of digoxin from Digitalis leaves with 20% ethanol followed by a further extraction of the digoxin into chloroform, and this method has formed the basis of the extraction procedure used in this paper, which is as follows. Grind an amount of dried D . lanata leaf to a particle size of approximately 500pm and macerate 10.0 g with 25ml of 20% V/V ethanol in a stoppered conical flask for 24 h at 25 "C, in order to allow simultaneous enzymatic hydrolysis and extraction of the glycosides. Add a further 75ml of 20% aqueous ethanol, stopper the flask and agitate it on a wrist- action shaker for 4 h. Next pour the slurry into a 2.5 cm diameter glass column fitted with a glass-wool plug and a tap, and allow the liquor to drain off at the approximate rate of 1 ml min-1.Extract the leaf with further portions of 20% ethanol until 200 ml of liquor has been collected. Cool the liquor to 20 "C and stir in 20 ml of 0.35 M sodium hydroxide solution. Allow the solution to stand at 20 "C for 15 min to ensure de-acetylation of any acetyldigoxin present, then adjust it to pH 6-6.5 with 0.35 M hydrochloric acid. Extract the liquor with five 50-ml portions of chloroform and dry each extract by passing it through the same 10 g of anhydrous sodium sulphate, packed into a glass funnel that is fitted with a plug of glass-wool. Finally evaporate the bulked chloroform extracts to dryness. Assay of Digoxin in the Leaf Extract HPLC conditions Mobile phase. Column packing. The column was packed at a pressure of 35 MN m-2 with 10-pm silica gel as a slurry in a mixture of tetrabromoethane and tetrachloroethylene according to the method of Majors.22 Operating @essure.7 MN m-2. Flow-rate. 2 ml min-l. Injection volume. 5 p1. Detection wavelengths. 265 or 234 nm. Detector sensitivity. 2.0 or 0.1 a.u.f.s. Cyclohexane - absolute ethanol - glacial acetic acid (60 + 9 + 1). Internal standardisation The cardenolides in D. lanata leaf can be readily detected at 234 nm, which is the lowest workable wavelength with this mobile phase, and a typical chromatogram of a leaf extract, recorded at 234 nm, is shown in Fig. 1. However, the cardenolides have a very low absorbance at 265 nm, and this property has been used to develop an internal standardisation procedure with a standard that absorbs strongly a t 265 nm and is eluted from the column before digoxin. Some sulphonamides were known to have the required chromatographic proper tie^,^^ and of these sulphamethoxazole was chosen as the internal standard.On injection of a mixture of D. lanata leaf extract and sulphamethoxazole it was possible to detect sulphamethoxazole free from interference from underlying glycosides by monitoring the column eluate at 265 nm and at low sensitivity. After elution of the sulphamethoxazole the digoxin could be detected by switching the detector to a lower wavelength and a higher sensitivity. An examination of extracts of D. Zanata leaf showed that digoxin was the last significant band to be eluted from the column. It was therefore possible to select an interval between injections of an internally standardised D.Zanata leaf extract and an internally standardised reference solution of digoxin such that both internal standard bands could be detected at 265 nm before switching to 234 nm to detect both digoxin bands.October, 1976 error from inaccurate adjustment of the spectrophotometer and to save analysis time. OF CARDENOLIDES AND ASSAY OF DIGOXIN IN D. Zanata LEAF 771 By overlapping test and standard in this manner it was possible to eliminate potential 1 Pro cedw e Add approximately 5 mg of sulphamethoxazole, accurately weighed, to the dried leaf extract and dissolve the mixture in approximately 2 ml of a 1 + 1 mixture, by volume, of chloroform and methanol. Also prepare a standard solution containing accurately measured amounts of approximately 8 mg ml-l of digoxin and 2 mg ml-l of sulphamethoxazole. Inject 5 p1 of the leaf extract solution on to the HPLC column and monitor the eluate at 265 nm and over a sensitivity range of 2.0 a.u.f.s.Shortly before the internal standard band begins to elute (after approximately 6 min) inject 5 p1 of the standard digoxin solution. When the two internal standard bands have eluted (after approximately 15min) set the spectrophotometer to zero at a wavelength of 234 nm and to a sensitivity range of 0.1 a.u.f.s. and monitor the elution of the two digoxin bands. Measure the heights of the two digoxin peaks and the two internal standard peaks, and calculate the digoxin assay from the following equation : d (test) 8 (std) m 8 (test) d (std) loo% Digoxin assay = - h d (std) h s (test) s (std) m l where k d (test) is the peak height given by digoxin in the leaf extract solution; h d (std), the peak height given by digoxin in the reference solution; h (test), the peak height given by sulphamethoxazole in the leaf extract solution; h (std), the peak height given by sulpha- methoxazole in the reference solution; m (test), the mass of sulphamethoxazole in the leaf extract solution; m (std), the mass of sulphamethoxazole in the reference solution; m d (std), the mass of digoxin in the reference solution; and m 1, the mass of D.Zanata leaf taken. A typical chromatogram obtained during the assay of digoxin in D. Zanata leaf is shown in Fig. 2. I l l 1 1 I I l l l I I t 0 4 8 12 16 20 24 Tirne/m in I Fig.1. Chromatogram of a n extract of Dutch D. lanata leaf. Ti me/m in 2 Fig. 2. Typical chromatogram from the digoxin assay. Results and Discussion Calibration Graph A calibration graph of peak height against concentration was plotted over the range 0-12 mg ml-1 of digoxin by using the internally standardised procedure. Each solution examined was measured against a common standard by use of the overlapping injection technique and a linear graph was obtained.772 COBB HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY FOR SEPARATION Analyst, VOZ. IOI Effect of Maceration Time on Digoxin Yield Indian origin is shown in Table I . The effect of maceration time on the yield of digoxin for a batch of dried D. Zanata leaf of TABLE I EFFECT OF MACERATION TIME ON DIGOXIN ASSAY (INDIAN LEAF) Maceration time/h .. 0 8 16 24 36 48 64 Digoxin assay, yo . . 0.092 0.184 0.177 0.197 0.194 0.194 0.182 for enzymatic hydrolysis to occur. These results show that a maceration period of 24-36 h at 25 "C provides adequate time Hydrolysis of Acetyldigoxin A large-scale, aqueous ethanolic extract of D. Zanata was prepared and the extract wits divided into aliquots. The aliquots were treated with sodium hydroxide solutions of different strengths for different periods of time and the digoxin content of each was deter- mined by means of the HPLC method. The effect of varying the strength of the sodium hydroxide solution and the period of de-acetylation is shown in Fig. 3, and it can be seen that for any set de-acetylation period the digoxin assay reaches a peak at a certain concentra- tion of sodium hydroxide.As the concentration of sodium hydroxide increases from this point there is a reduction in the digoxin assay owing to hydrolysis of the butenolide ring. 5 0.12 .O 0.09 0.08 0 tp.- 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Concentration of sodiun.1 hydroxide added/M Fig. 3. Effect of the concentration of sodium hydroxide solution used in the de-acetylation. procedure (carried out at 20 "C). De-acetylation period: A, 15 min; B, 7.5 min; C, 30 min; and XI, 60 min. These results show that treatment with 0.25-0.40 M sodium hydroxide solution for 10-20 min at 20 "C should ensure maximum de-acetylation with the minimum hydrolysis of digoxin. TABLE I1 REPLICATE DETERMINATIONS OF THE DIGOXIN CONTENT OF D. hZah2 LEAF FROM VARIOUS SOURCES Extract 1 7- Assay 1, Assay 2, Source of leaf % YO Holland 1 .. . . 0.167 0.165 Holland 2 . . . . 0.142 0.144 Russia .. . . 0.096 0.095 W. Germany.. . . 0.118 0.121 India . . .. . . 0.224 0.220 E. Europe . . . . 0.146 0.148 Extract 2 +L--7 Assay 1, Assay 2, % % 0.167 0.165 0.142 0.144 0.094 0.094 0.128 0.125 0.228 0.229 0.149 0.144 Mean, 0.166 0. I43 0.096 0.123 0.225 0.143 %October, 1976 OF CARDENOLIDES AND ASSAY OF DIGOXIN IN D. Zanata LEAF 773 Evaluation of the Assay Samples of dried D. Zanata leaf from different sources were chosen to represent a range of digoxin contents. Two extracts of each leaf sample were undertaken with each pair of extractions being performed simultaneously. Two determinations of the digoxin content of each extract were undertaken, and the four determinations relating to a single leaf sample were carried out consecutively.The standard deviation based on duplicate chromatographic evaluations of single extracts was 0.001 5% for digoxin in dried leaf, while the standard deviation based on four determina- tions per leaf (k, including the variations between extracts) was 0.002 674. This implies The results obtained are listed in Table 11. Series A TABLE I11 RETENTION TIMES RELATIVE TO DIGOXIN OF A RANGE Aglycone CH? Compound Digitoxigenin Digitoxigenin monodigitoxoside Lanatoside A HO OF D. Zanata CARDENOLIDES E Gitaloxin B Gitoxigenin Gitoxigenin Strospeside Lanatoside B HO I CH3 W Digoxin HO Lanatoside C Diginatin Glycosidically linked residue at C,' -D -D-D -D-D-D -D-D-D-G I Ac -D-D-D -D -D-D -D-D-D -L -D-D-D-G I Ac -D -D-D -D-D-D -D-D-D I Ac I Ac -D-D-D-G -D-D-D Relative retention ti me 0.24 0.31 0.40 0.52 1.54 0.63 0.38 0.45 0.61 0.78 0.78 2.67 0.45 0.53 0.75 1.0 0.84 0.73 3.29 1.63 *D = digitoxose, G = glucose, L = digitalose, Ac = acetyl.774 COBB : HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY FOR SEPARATION Analyst, VoZ.101 that for a sample giving a mean determination of 0.150% of digoxin, a single-assay deter- mination carried out on the same occasion will 1.ie within the range 0.145-4.155% (P = 0.95). The assay method has now been in use in this laboratory for more than a year, during which time several hundred leaf samples and leaf extracts have been examined. Small decreases in column efficiency have been noted after approximately 200 assays, but these have not proved serious enough to affect the specificity of the method.Specificity of the Assay A number of the common cardenolides of D. Zanata were examined in order to test the specificity of the assay method. It was found that the retention time increased with the number of glycosidically linked digitoxose units and in the order A<E<B<C<D, in accordance with previously observed polarity trends.24 The retention times relative to digoxin are listed in Table 111. From these results it can be seen that the chroinatographic system is highly selective for digoxin in the presence of the other abundant cardenolides of D. Zanata. By using chromato- graphic columns made by the slurry packing technique, column efficiencies of 1 500 theoreti- cal plates have been obtained reproducibly, and this efficiency is sufficient to resolve com- pletely digoxin from potential interfering compounds.Assay of Other Cardenolides The chromatographic system described has been designed specifically to resolve digoxin from the other cardenolides in D. Zanata leaf. However, Figs. 6 7 show that this system can also be applied to the resolution and quantitative determination of other groups of cardeno- lides. Retention times for the more polar compounds, such as diginatin and the lanatosides, can be decreased by increasing the proportion of ethanol in the mobile phase. 0 4 8 12 16 20 24 28 32 T irn e/m in Fig. 5. Chromatogram of the secondary glycosides of D. Zanata: 1, digitoxin; 2, gitaloxin; 3, gitoxin; 4, digoxin; 5, diginatin.0 2 4 6 8 1012 Time/min Fig. 4. Chromatogram of a mixture of aglycones : 1, digitoxigenin; 2, git- oxigenin ; 3, digoxigenin. Relationship Between Chemical Structure of Cardenolides and Retention Time Clayton2s has shown that the retention behaviour of steroids in gas -liquid chromato-Octobey, 1976 OF CARDENOLIDES AND ASSAY OF DIGOXIN IN D. Zanata LEAF L 775 h, 1 3 1 I 1 1 1 1 1 1 1 1 1 4 8 12 16 20 24 Time/min Fig. 7. Chromatonram of Fig. 6. Chromatogram of “C” series cardenolides : 1, digoxigenin; 2, dig- oxigenin monodigitoxo- side ; 3, digoxigenin bis- digitoxoside; 4 digoxin. digoxin and its a- and- p-acetyl derivatives : 1, p-acetyldigoxin ; 2, a-acetyldigoxin ; 3, digoxin. graphy is governed by the following relationship (provided that no interaction occurs between substituent groups) : t R ( n , a , b , c ) = tR(n) k&bkc where tR(n,a,b,C) is the retention time of the composite steroid; t R p ) , the retention time of the steroid nucleus; and ka, kb and k, are the functional group retention ratios of substituents a, b and c, respectively. Functional group retention ratios are calculated from the equation Functional group retention ratios were calculated for 16/3-hyd.roxyl, 12/3-hydroxyl and 3/3-(/3-~-digitoxose) groups, as shown in Table IV.Theoretical relative retention times k x = tR(n,x)/tR(n)- TABLE IV CALCULATION OF FUNCTIONAL GROUP RETENTION RATIOS n, steroid nucleus; x, functional groups. Position Substituent* n,x 3fi B-DS Digitoxin Gitoxin Digoxin 128 OH Digoxigenin Digoxigenin Digoxigenin Digoxin Diginatin monodigitoxoside bisdigitoxoside 16/3 OH Gitoxigenin Gitoxigenin Gitoxigenin Gitoxin Diginatin monodigitoxoside bisdigitoxoside Il Digitoxigenin Gitoxigenin Digoxigenin Digitoxigenin Digitoxigenin Digitoxigenin Digi toxin Gitoxin monodigitoxoside bisdigitoxoside Digitoxigenin Digitoxigenin Digitoxigenin Digitoxin Digoxin monodigitoxoside bisdigitoxoside tR(n,x) 0.52 0.78 1 .oo 0.45 0.63 0.76 1.00 1.63 0.38 0.45 0.61 0.78 1.63 tR(n) 0.24 0.38 0.45 0.24 0.31 0.40 0.52 0.78 0.24 0.31 0.40 0.52 1.00 kx 2.17 2.05 2.22 1.88 1.71 1.90 1.92 2.09 1.68 1.45 1.53 1.50 1.63 Mean k , 2.16 1.90 1.64 *D = digitoxose.776 COBB TABLE V GLYCOSIDES OF 1).Zanata OBSERVED AND CALCULATED RELATIVE RETENTION TIMES OF SOME SECONDARY n, steroid nucleus; x, fun.ctiona1 groups.n,x n tR(n) hI2-o= k I 8 - o ~ k , - ~ , * calculated observed tR tB Digitoxin Digitoxigenin 0.24 - - 2.15 0.52 0.52 Gitoxin Digitoxigenin 0.24 - 1.54 2.15 0.79 0.78 Diginatin Digitoxigenin 0.24 1.90 1.54 2.15 1.51 1.63 Digoxin Digitoxigenin 0.24 1.90 - 2.15 0.98 1.00 *D = digitoxose, were then calculated for digitoxin, gitoxin, digoxin and diginatin (Table V) and good agree- ment with the observed values was obtained. These results show that there is a correlation between chemical structure and retention time on this HPLC system, and suggest that HPLC could be used in the identification of unknown cardenolides. Conclusion The internally standardised HPLC method for the assay of potential digoxin in D. Zanntu leaf is specific and more precise than existing methods.The chromatographic system des- cribed is suitable for the rapid separation, identification and quantitative determination of a great number of cardenolides, and a relationship between chemical structure and retention time has been revealed. I thank Messrs. A. C. Caws, R. E. A. Drey and G. T. Hill for helpful discussions during I am also indebted to Mr. F. W. Harpley for his comments on the the course of this work. statistical significance of the results and to Mr. C. J. Clarke for skilful technical assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. References Zaffaroni, A., Burton, R. B., and Keutmann, H. E., J . Biol. Chem., 1949, 177, 109. Reichstein, T., and Schindler, O., Helv. Chirn. Actfir, 1951, 34, 108. Jensen, K. B., Acta Pharmac. Tox., 1956, 12, 27. Tantivatana, P., and Wright, S. E., J . Pharm. Pharmac., 1958, 10, 189. Potter, H., Pharmazie, 1963, 18, 554. Wolf, L., and Karacsony, E. M., Planta filed., 1963, 11, 432. Stahl, E., and Kaltenbach, U., J . Chromat., 1961, 5, 458. Heusser, D., Dt. ApothZtg, 1965, 103, 1101. Hauser, W., Kartnig, T., and Verdino, G., Scientia Pharm., 1968, 36, 237. Potter, H., and Baerisch, H., Pharmazie, 1972, 27, 315. Evans, F. J., Flemons, P. A., Duignan, C. F., and Cowley, P. S., J . Chromat., 1974, 88, 341. Stoll, A., Angliker, E., Barfuss, F., Kussmaul, W., and Renz, J., Helv. Chim. Acta, 1951, 34, 1460 Kaiser, F., Arch. Pharm., Berl., 1966, 299, 263. Hauser, W., Kartnig, T., and Verdino, G., Scientiit Pharm., 1969, 37, 149. Jellife, R. W., and Blankenhorn, D. H., J . Chromat., 1963, 12, 268. Wilson, W. E.. Johnson, S. A., Perkins, W. H., and Ripley, J. E., Artalyt. Chem., 1967, 39, 40. Evans, F. J., J . Chromat., 1974, 88, 411. Lotscher, K. M., Brander, B., and Kern, H., Varian Aerograph Application Notes, 1975, No. 8. Castle, M. C., J . Chromat., 1975, 115, 437. Lindner, W., and Frei, R. W., J . Chromat., 1976, 117, 81. Kaiser, F., Haack, E., and Spingler, H., Justus Liebigs Annln Chem., 1957, 603, 75. Majors, R. E., Analyt. Chem., 1972, 44, 1722. Cobb, P. H., and Hill, G. T., J . Chromat., to be published. Waldi, D., in Stahl, E., Editor, “Thin-Layer Chroma.tography,” First Edition, Springer Verlag, Berlin Gottingen, and Heidelberg, 1962, p. 275. Clayton, R. B., Nature, Lond., 1961, 192, 524. Received May 7tk, 1976 Accepted June 7th, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100768

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst , October, 1976, VoL. 101, $9. 777-785 777 Ionic Polymerisation as a Means of End-point Indication in Non-aqueous Thermometric Titrimetry Part IX.” Phosphorodithioates Determination of Dithiocarbamates and E. J. Greenhow and L. E. Spencer Department of Chemistry, Chelsea College, University of London, Manresa Road, London, SlV3 6LX Ammonium, cadmium, bismuth, iron, lead, manganese, nickel, piperidinium and zinc dithiocarbamates and nickel and zinc phosphorodithioates have been determined in amounts down to 0.002 mmol by catalytic thermometric iodi- metric titration of their solutions in dimethylformamide and acrylonitrile. The end-points in most instances correspond to the formation of polyiodides. The reaction stoicheiometries depend on the nature of the ligand as well as on the central ion.The method has been compared with aqueous iodimetry, and the results agree to within about 1%. Solutions of dithiocarbamates and phosphorodithioates in hydrocarbon oils, used as oil additives, can be determined in concentrations down to 0.05%. I-Iigh-boiling alcohols and other oil additives, including hindered phenols, diphenylamine derivatives and phenothiazine, do not interfere. The results show precisions ranging from 0.44 to 2.01%. The metal dithiocarbamates (I) and phosphorodithioates (dithiophosphates) (11) are similar in structure and properties. They are compounds in which the metal is bonded to sulphur, and which can be oxidised by iodine to the corresponding organic disulphides (I11 and IV). R 1 )-Q L I R s s RO S S OR / / Mn+ 1 Ro>( ] Mnf R’O S F$ ‘s-s ‘R, R’O S - S OR n n Ii Ill IV The NN-disubstituted dithiocarbamates and the 00’-disubstituted phosphorodithioates have a range of important commercial applications.Both find use as additives for automotive engine oils, but the phosphorodithioates, particularly the zinc salts (ZDDP), which have a multifunctional action, appear to be the more generally preferred for this purp0se.l The dithiocarbamates are employed in large amounts as agricultural fungicides, and in smaller amounts as additives in rubber compositions and plastics formulations. Some water-soluble dithiocarbamates are useful analytical reagents for the determination of trace metals by colorimetry and atomic-absorption spectroscopy.2 Technical-grade dithiocarbamates can contain as impurities various metal salts, including sulphides and sulphites.The last two are sometimes added deliberately as stabili~ers.~ Talc, kaolin, kieselguhr, silica gel and clays are used as diluents in dithiocarbamate fungicide formulations, and surface-active agents are included to make the formulations “~ettable.”~ Technical-grade phosphorodithioates, intended for use as oil additives, axe supplied as con- centrated solutions in oil. They contain small amounts of the phenols and high-boiling alcohols used in their synthesis. Clearly, in the assay of these technical-quality materials, interference by the various impurities, additives and diluents must be avoided, either by use of a separation step or by means of a selective analytical reaction. In the procedure currently recommended4s5 for the determination of metal dithiocarbamates in formulations, the sample is decomposed by treatment with acid and the carbon disulphide * For Part VIII of this series, see Analyst, 1976, 101, 421.778 GREENHOW AND SPENCER : IONIC POLYMERISATION FOR END-POINT Afia&St, VOl.101 released is absorbed in methanolic potassium hydroxide and titrated iodimetrically as the potassium 0-methyldithiocarbonate. Thin-layer chromatography appears to be the analytical technique most widely used for the analysis of metal phosphorodithioate additives in Re~ently,~ gas - solid chromato- graphy has been shown to be suitable for the determination of volatile dialkyl phosphoro- dithioates, including zinc, nickel, palladium, platinum, rhodium and chromium derivatives. Like the structurally similar 0-substituted dithiocarbonates, the substituted dithiocar- bamates and phosphorodithioates can be determined by direct iodimetric titration, but the oxidation reactions are, apparently, not as quantj tative.Thus, the thiuram disulphides (111) formed in the oxidation of the dithiocarbamates undergo a further, slow, reaction with the iodine titrant. This difficulty has been overcomelo by adding a water-immiscible solvent to the aqueous sample to remove the disulphide from the aqueous phase as quickly as it forms. It is claimed11 that the iodimetric oxidation of the phosphorodithioates is reversible and does not give sharp iodimetric end-points. Direct titration with dialkylphenylmercury( 11) acetate or the use of an iodimetric method, involving oxidation with hypoiodite, is stated to yield more accurate results.ll In Part VII12 is described the determination of metal 0-alkyl dithiocarbonates by a non- aqueous iodimetric procedure, in which the titrimetric end-point is marked by the rise in temperature resulting from the polymerisation of ethyl vinyl ether.The polymerisation is initiated by a small excess of the iodine titrant when the oxidative determinative reaction is completed. An advantage of the method is that, with many of the dithiocarbonates titrated, the end-point is indicated on the formation of the triiodide ion. In contrast, the end-point in conventional aqueous iodimetry corresponds to the formation of monoiodide. Thus, the non-aqueous method offers the possibility of increased sensitivity. In this investigation, the application of the method to the direct determination of some dithiocarbamates and phosphorodithioates of commercial and analytical importance has been studied.Experimental Reagents Laboratory-reagent grade acrylonitrile and dimethylformamide were dried over molecular sieve 4A before use. Laboratory-reagent grade ethyl vinyl ether was distilled through a 20 x 1 cm column packed with glass Fenske helices, and the fraction boiling in the range 36-39 “C was retained for use. The 0.05 M solution of AnalaR-grade iodine in dimethylformamide was standardised by the method described in Part V.13 Dithiocarbamates and Phosphorodithioates Ammonium tetramethylenedithiocarbamate, sodium NN-diethyldithiocarbamate, zinc NN-dibenzyldithiocarbamate and zinc 00’-di(2-propyl)phosphorodithioate were of laboratory- reagent grade.Lead, bismuth, iron( 111) and zinc NN-dimethyldithiocarbamates, cadmium NN-diethyldithiocarbamate, zinc NN-diethyl- and NN-dibutyldithiocarbamates, man- ganese( 11) and zinc NN’-ethylenebisdithiocarbamates, piperidine and zinc cyclopenta- methylenedithiocarbamates, cufraneb (a comp1.e~ ethylenebisdithiocarbamate of copper, manganese, zinc and iron) and a 50% m/m solution of zinc NN-diamyldithiocarbamate in lubricating oil were gifts from Robinson Brothers Limited. Solutions of dialkyl phosphoro- dithioates and lead NN-dialkyldithiocarbamates in hydrocarbon oils were gifts from K-K Greef Industrial Chemicals Limited. Iron(II1) NN-diethyldithiocarbamate was prepared from the sodium salt and was recrystal- lised from 1,2-dichloroethane.Nickel 00’-diethylphosphorodithioate was prepared from phosphorus(V) sulphide, ethanol and the metal sulphate by the method of Busev and Ivanyutin,14 and was recrystallised from ethyl acetate. The zinc NN-diethyl- and NN-dibenzyldithiocarbamates were recrystallised from acetone. The bismuth, iron(II1) and lead NN-dimethyldithiocarbamates, cadmium NN-diethyldithio- carbamate, zinc NN-dibutyldithiocarbamate and manganese(I1) NN’-ethylenebisdithio- carbamate were purified by Soxhlet extraction with chloroform, and the zinc OO’-di(2-propyl)-October, 19'16 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART IX 779 phosphorodithioate was purified by Soxhlet extraction with hexane.Where possible, purities were checked by determining the melting-point of the material. Apparatus instead of the insulated titration flask. Use the automatic apparatus described in Part 111,15 but with a 14- or 25-ml Dewar beaker Procedure Thermometric Titration Prepare a solution containing 0.2-1.0 mequiv of the sample in 10-50 ml of dimethyl- formamide or acrylonitrile as required (the concentration will depend on the solubility of the sample and the stoicheiometry of the reaction with iodine), and centrifuge the solution for 30 min in order to remove insoluble impurities. Transfer 2-5 ml of the supernatant solution into the Dewar beaker and add 2 ml of ethyl vinyl ether. To the stirred solution add titrant a t the rate of 0.2mlmin-1 from the motor-driven syringe.The volume of titrant at the end-point is taken to be the volume corresponding to the point of inflection in the titration graph. When this inflection is indistinct, the end-point is taken to be the volume correspond- ing to the point where the tangent to the main heat rise leaves the graph at its lower tempera- ture end. Carry out a blank titration by using an equal volume of the same batch of solvent with the same water content as that used for the sample solution. In the preparation of calibration graphs, use the same volume of sample solvent irrespective of the amount of sample, i.e., make up the sample volume to 2 or 5 ml, as appropriate. The chart recorder of the automatic titration apparatus is conveniently operated over the 0-100-mV range with dimethylformamide as the solvent and the 0-200 or 0-500-mV range with acrylonitrile as the solvent.Aqueous Iodimetric Titration of Dithiocarbamatesl0J6 Weigh accurately into a 250-ml flask with ground-glass neck about 0.4g of the dithio- carbamate, add 50 ml of chloroform, 50 ml of distilled water and 2 drops of phenolphthalein indicator. Stopper the flask and shake it well. If the solution becomes pink, add 0.05 N acetic acid until the pink colour is discharged. Add 5-ml increments of 0.1 N aqueous iodine solution, shaking well between each addition, until the brown colour persists. Add an additional 5 ml of titrant, shake, and back-titrate with 0.1 N sodium thiosulphate solution to the colour change of starch indicator. Aqueous Iodimetric Titration of Phosphorodithioates14 Weigh accurately into a 250-ml flask with ground-glass neck about 0.4 g of the phosphoro- dithioate, add 20 ml of distilled water and 5 ml of concentrated hydrochloric acid, stopper the flask and shake it for 10 min. Add 5-ml increments of 0.1 N aqueous iodine solution and back-titrate the excess of iodine as described above for the determination of dithiocarbamates.Results and Discussion Titration curves for the dithiocarbamates and phosphorodithioates examined are shown in Figs. 1 4 and 5, respectively. The curves shown in Fig. 6 are those obtained in the analysis of commercial samples of concentrated solutions of dithiocarbamates and phosphorodithioates in hydrocarbon oils. Single, sharp end-point inflections are achieved in the titrations of all of the compounds except antimony phosphorodithioate by using appropriate amounts of the sample, which range from 0.002 to 0.05 mmol.The sample size required to give a sharp end-point depends on the stoicheiometry of the iodimetric reaction as well as on the solubility and type of compound analysed. Two end-point inflections become apparent (Figs. 3-6) as the sample concentration is increased. The first inflection, which is usually distinct, corresponds to the completion of the iodimetric oxidation t o the disulphide (I11 or IV), while the second, less well defined, inflection is believed to measure further oxidation of the disulphide. As the sample concentration is increased even further, the two oxidation processes tend to overlap and it becomes difficult to locate the first inflection [Figs.3(b), 4(h), 6(b) and 6(j)]. The change in the shape of the titration curves for zinc dibutyldithiocarbamate and iron(I1I)780 GREENHOW AND SPENCER : IONIC POLYMERISATION FOR END-POINT Analyst, 6/01. 101 0.05 M iodine reagendrnl ( 1 division = 1 rnl) Fig. 1. Catalytic thermometric t itration of dithiocarbamates in solution in dimethylformamide. Dithiocarbamate/mg (with reaction stoicheiometry in parentheses) : a, bismuth dimethyl* 5.8 (3.0); b, cadmium diethyl* 6.0 (2.0); c, cadmium diethylt 12.1 (2.0) ; d, ammonium tetramethglenef 6.1 (4.0) ; e, cufranebt 4.5; f, iron(II1) dimethyl* 3.2 (6.0); g, iron(II1) diethyl* 1.8 (6.5) ; h, lead dimethyl? 4.1 (2.9); j, manganese(I1) ethylenebis* 6.0 (1.8) ; k, mangancse(I1) ethylenebist 14.3 (2.0) ; m, piperi- dinium cyclopentamethylenet 5.0 (5.2) ; n, zinc diethyl* 3.1 (4.0); p, zinc dibutyl* 3.0 (4.5); q, zinc dibenzyl* 15.2 (3.1); r, zinc ethylenebist 2.3 (2.7).* Recrystallised; t technical grade ; 1 laboratory-reagent grade. dimethyldithiocarbamate as the sample concentration is increased is shown in Fig. 4. It can be seen that the second inflection of the titriation curves for the iron complex differs in shape from that obtained in the titrations of other dithiocarbamates. A C 0 c --I_.- 0.05 M iodine reagent/ml 11 division = 1 ml) Fig. 2. Catalytic thermometric titration of dithiocarbamates Oxcdation of the in soluGon in aciylonitrile. Dithiocarbamate/mg (with reaction stoicheiometry in parentheses) : a, ammonium tetramethylene* 1.6 (4.8) ; b, cadmium diethylt 25.4 (2.2) ; c, lead dimethyl: 7.7 (3.2) ; d, manganese(I1) ethylenebist 2.5 (2.0) ; e, piperidinium cyclopenta- methylene: 5.25 (6.2); f, zinc dimethyl: 3.2 (4.5); g, zinc diethyl: 3.3 (4.0); h, zinc diethylt 2.0 (4.2); j, zinc dibenzylt 10.0 (3.2); k, zinc ethylenebis: 3.0 (2.3) ; m, zinc cyclopentamethylene: 3.3 (5.0).* Laboratory-reagent grade ; t recrystallised ; : technical grade.October, 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART IX I 78 1 0.05 M iodine reagent/ml ( 1 division = 1 ml) Fig. 3. Thermometric titration curves for dithiocarbamates showing double inflections. Dithiocarbamateslmg : a, bismuth dimethyl* 13.5; b, zinc dimethyl? 4.7; c, zinc dimethyl? 10.1; d, zinc diethylt 10.2; e, zinc dibutylt 9.6; f, zinc dibenzyl* 31.3; g, zinc cyclopentamethylenet 9.05.*Recrystallised : t technical grade. Solvent: a, b and f, acrylonitrile; c-e and g, dimethyl- formamide. disulphide has been suggested by Shankaranarayana and Patello as being responsible for the unsatisfactory results in the direct aqueous iodimetric titration of the dithiocarbamates. The measured reaction stoicheiometries obtained in the determinations of the dithio- carbarnates are shown in the legends to Figs. 1 and 2. The values depend on the structure of the ligand as well as on the nature of the “central” ion. For example, the stoicheiometry is approximately 3 for zinc dibenzyldithiocarbamate but ranges from 3.2 to 5.0 for the corresponding dimethyl, diethyl, dibutyl and cyclopentamethylene dithiocarbamates.This range is in contrast to the more consistent values obtained in the titrations of the zinc 0-alkyl I f 1 I I I 0.05M iodine reagent/ml ( 1 division = 1 ml) Fig. 4. Effect of sample concentration on the shape of the titration curves for iron(II1) dimethyl and zinc dibutyl dithiocarbamates. Solvent : acrylonitrile. a, Sol- vent blank. Zinc dibutyldithiocarbamate: b, 3.0 mg; c, 6.0 mg; d, 9.0 mg; e, 12.0 mg. Iron(II1) dimethyldithio- carbarnate: f, 1.0 mg; g , 2.0 mg; h, 3.0 mg; j, 4.0 mg.782 GREENHOW AND SPENCER: IONIC POLYMERISATION FOR END-POINT Analyst, VoZ. 101 dithiocarbonates.12 Magee2 has pointed out that dithiocarbamate ions should form stronger complexes than do the xanthate (O-substituted dithiocarbonate) ions, because there is a greater electron drift into the sulphur atoms from the -NR2 group than from the -OR group, and this effect increases their electron donor capacity.Thus, it is likely that the stability of the metal complex increases with the increasing inductive efl:ect of the groups attached to the nitrogen atoms of the ligand. The stability of the complex might influence the kinetics as well as the mechanism of the iodimetric reaction. It was noted in Part VIP2 that concerted reaction mechanisms, as distinct from two-stage processes involving the formation of metal iodide followed by the reaction of iodine with iodide ions, could account for the unexpected reaction stoicheiometries of the zinc O-alkyl dithiocarbonates. Similar reaction mechanisms might obtain in the oxidation of the dithiocarbonates.The cadmium and bismuth dithiocarbamates have reaction stoicheiometries of approxi- mately 2 and 3, respectively. Apparently, polyiodide ions are not formed in the iodimetric oxidation, and this suggestion has been confirmed by titrating solutions of bismuth and cadmium iodides in dimethylformamide. In both instances, the titration volume was equal to the “blank” value. It was shown in Part VIP2 that the metal iodide can be titrated to polyiodide if the corre- sponding dithiocarbonate shows a reaction stoicheiometry higher than that required to yield the “normal” iodide. Manganese( 11) and zinc ethylenebisdithiocarbamates show low reaction stoicheiometries of approximately 2 and 2.7, respectively. This effect may be related to the mechanism of the oxidation process with ethvlenebisdithiocarbamates. as it has been found4 that their decomposition ‘In acidic solutioi is a more complex process than obtains di t hiocarbamates.with the dialkyl- 0.051M iodine reagent/ml ( 1 division = 1 ml) Fig. 5. Catalytic thermometric titration of phosphorodithioates. Phosphorodithioate/mg (with reaction stoicheiometry in paren- theses) : a, zinc di(2-propyl)* 4.25 (5.3); b, zinc di(2-propyl)* 4.47 (5.2) ; c, zinc dioctylt 5.03 (4.1); d, zinc di(2-ethylhexyl)? 11.5 (4.0) ; e, zinc di(2-propyl): 12.0 (4.4); f, nickel &ethyl* 7.2 (9.1). * Recrystal- lised ; t technical grade; : laboratory-reagent grade. Solvent : a and f , acrylonitrile; b - e, dimethylformamide. The reaction stoicheiometry of the lead dimethyldithiocarbamate of approximately 3 is similar to that of the corresponding dithiocarbonate.12 The iron( 111) dialkyldithiocarbamates show high reaction stoicheiometries, ranging from about 6 to 8, depending on the compound, solvent and concentration.The values are intermediate between those expected if the iodides produced by the iodimetric reaction are FeI, and FeI,. This range suggests that the end-point occurs when a mixture of mono- and polyiodides is formed. Two non-metal dithiocarbamates, ammonium tetramethylene- and piperidinium cyclo- pentamethylenedithiocarbamate, have been titrated. Both show higher reaction stoicheio-October, 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART IX 783 metries than would be expected if the (presumably unstable) ammonium and piperidinium triiodides were formed.There is no evidence of a second inflection in the titration curves for nickel(I1) diethyl- phosphorodithioate, while a second inflection in the titration curves for the zinc phosphoro- dithioates is obtained at concentrations higher than those at which the phenomenon becomes apparent with the zinc dithiocarbamates. The reaction stoicheiometry obtained in the titration of the zinc phosphorodithioates varies from 4 to 5, depending on the structure of the ligand, and is, therefore, of the same order as the stoicheiometries obtained with the zinc dithiocarbamates and dithiocarbonates. The nickel phosphorodithioate shows high reaction stoicheiometries ranging from 7.4 to 10, depending on the sample concentration, i.e., the calibration graph is curved.k-- 0.05 M iodine reagent/mI (1 division = 1 ml) Fig. 6. Catalytic thermometric titration of solutions of dithio- carbamates and phosphorodithioates in hydrocarbon oils. Dithio- carbamates/mg (with concentration, % m/m, in parentheses) : a, zinc diamyl 16.7 (50); b, zinc diamyl 33.3 (50); g, lead dipentyl 21.3 (15). PhosphorodithioatesImg (with concentration, yo m/m, in parentheses) : c, zinc O-butyl-0’-pentyl 5.4 (87) ; d, zinc di(hepty1- phenyl) 30.2 (68) ; e, zinc di(nonylpheny1) 33.2 (50) ; f, zinc di(nony1- phenyl) 63.5 (50); h, antimony dialkyl 15.0 (16); j , antimony dialkyl 30.0 (16). Solvent: acrylonitrile. Most of the impurities in technical-grade dithiocarbamates are insoluble in the acrylonitrile and dimethylformamide solvents, and can be removed by centrifuging the sample solutions.The exceptions are sodium sulphide and sodium trithiocarbonate; both are soluble in dimethyl- formamide and consume the iodine titrant, and consequently acrylonitrile is the preferred solvent when sodium salts are present. Lubricating oils and plastics formulations may contain, in addition to the technical-grade dithiocarbarnates and phosphorodithioates with their associated impurities, other additives, such as hindered phenols, tritolyl phosphate and diphenylamine derivatives, including N-phenylnaphthylamines and phen0thiazines.l However, the iodimetric titration of butan- 1-01, pentan-1-01, phenol, 2,6-di-t-butyl-4-methylphenol, diphen ylamine, N-phenyl- 1 -naphthyl- amine, phenothiazine and tritolyl phosphate has established that none of these compounds has a significant titration value.It should be noted that primary amines and alkylamines, including N-alkylanilines and phenylenediamines, are determined by the iodimetric method and, if present, would interfere in the determination of the dithiocarbamates and phosphoro- dithioat es. The thermometric procedure is generally satisfactory for the determination of 0.5- to 10-mg amounts of the compounds. Thus, concentrations as low as 0.05% in l-ml or l-g samples can be analysed. Although a method with this sensitivity may not be required in the assay of fungicide formulations, it is useful for determining additives in lubricating oils,784 GREENHOW AND SPENCER : IONIC POLYMERISATION FOR END-POINT Analyst, VoZ.101 rubber compositions and plastics formulations, where concentrations in the range 0.14% are usual. The calibration graphs for most of the pure compounds are linear, or almost linear, with sample concentrations in the range 0-3 mg ml-l. In some instances, deviations from linearity occur not only at higher concentrations, when the two oxidation reactions overlap, but also at very low concentrations. The linear range for some of the compounds is given in Table I. In Table I, results obtained by the catalytic thermometric procedure and aqueous iodimetry are compared. The method of Busev and Ivanyutin14 and a modificationla of the method of Shankaranarayana and Patello were used for the aqueous titrations of phosphorodithioates and dithiocarbamates, respectively.Purified, technical-grade and laboratory-reagent grade materials were determined by these methods. The catalytic thermometric procedure was used to determine the purities of technical and laboratory-reagent grade compounds with the aid of the calibration graphs described above. In general, the results obtained by the two methods agree to within &1.2%. It can be seen from Table I that this difference is within the range of precision of the thermometric method for most of the compounds. The coefficients of variation (0.44-2.01%) are of the same order as those obtained in the thermo- metric titration of the dithiocarbonates. TABLE I DETERMINATION OF DITHIOCARBAMATES AND PHOSPHORODITHIOATES BY CATALYTIC THERMOMETRIC AND AQUEOUS IODIMETRIC TITRIMETRY Compound* Dithiocavbamate- Cadmium diethyl (R) Cadmium diethyl (T) Zinc dimethyl (R) Zinc dimethyl (T) Zinc diethyl (R) .. Zinc diethyl (T) . . Zinc dibutyl (R) . . Zinc dibutyl (T) . . Phosphoyodzthioate- Zinc di(2-propyl) (R) Zinc di(2-propyl) (LR) Nickel diethyl (R) Nickel diethyl (U) Thermometric titrimetry Amount taken/ mgt Coefficient of variation, % Mean content, %I Liiear range/ mgt - ,. . . 9.42 . . 4.50 . . 6.88 . . 4.26 - .. - .. - . . - . . . . 5.28 . . 4.26 - .. - 1.55 0.99 0.65 0.44 - - - - 2.01 1.33 - - 99.7 102.0 100.4 99.3 - - - - 92.6 98.7 - 0-1 6 0-10 0-9 0-10 - - - - 0-6 P 7 . 5 - - Aqueous iodimetry & Amount Mean taken/ content, mg %§ 503.0 99.9 504.4 98.8 350.6 100.4 321.7 100.8 449.3 101.3 515.2 99.2 247.8 99.9 506.0 100.0 431.4 100.1 502.6 91.6 401.9 98.7 500.6 97.6 * R, recrystallised ; T, technical grade: LR, laboratory-reagent grade; U, crude.t In 5 ml of acrylonitrile + 2 ml of ethyl vinyl ether. I Calculated from three determinations. 5 Mean of two determinations. The solutions of the dithiocarbamates and phosphorodithioates in hydrocarbon oils were miscible with acrylonitrile in the proportions required for the iodimetric titrations. The titration curves for the oil solutions of zinc diamyldithiocarbamates [Figs. 6(a) and (b)] show the development of a second inflection with the increase in sample size. A similar change in the shape of the titration curve for oil solutions of zinc di(nonylpheny1)phosphorodithioate is shown in Figs. 6(e) and (f). No distinct end-point inflections were obtained in the titrations of the solutions of antimony dialkylphosphorodithioate [Figs.6(h) and (j)], and this compound cannot be determined satisfactorily by the present procedure. Calibration graphs based on the results of these titrations are linear for the zinc dialkyl and diary1 phosphorodithioates and are almost linear for the lead and zinc dialkyldithiocarbamates (Fig. 7). The calibration graphs indicate that zinc O-butyl-0’-pentyl- and zinc di(heptylpheny1)phosphorodithioates can usefully be determined in the ranges 0-10 and 0-20mg, respectively, while the zinc dipentyl- and lead dipentyldithiocarbamates can be determined in the ranges 5-15 and 1.5-6 mg, respectively. The main disadvantage of the present method is that calibration is necessary because theOCtOber, 1976 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY.PART IX 785 - 1.2 E --. W 2 1.0 0.8 0.6 3 2 0.4 z u) 0.2 L .- 8 I I I I I I I , I J Oil solution/mg 5 10 15 20 25 30 35 40 45 Fig. 7. Calibration graphs for the determination of dithiocarbamates and phosphorodithioates in solution in hydrocarbon oil. A, Zinc O-butyl-0’- pentylphosphorodithioate (87 yo). B, Zinc dipentyl- dithiocarbamate (50%). C , Lead dipentyldithio- carbamate (15%). D, Zinc di(heptylpheny1)phos- phorodithioate (67.5”,). Solvent system: 2 ml of acrylonitrile + 2 ml of ethyl vinyl ether. reaction stoicheiometries are not exact. For routine analysis, the construction of calibration graphs can be replaced by the use of the simple bracketing technique employed in automated chemical analysis.The advantages of the thermometric method over current titrimetric procedures are that it is rapid, simple and can be used for the determination of milligram amounts of the compounds dissolved or dispersed in hydrocarbon solvents. With some compounds, the high reaction stoicheiometries give the added advantage of increased sensitivity. Messrs. Robinson Brothers Limited and K-K Greef Industrial Chemicals Limited are thanked for gifts of chemicals, and Mr. R. J. Hadley is thanked for helpful advice on the assay of the dit hiocarbamat es. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. References Gilks, J. H., J . Inst. Petrol., 1964, 50, 309. Magee, R., Rev. Analyt. Chem., 1973, 1, 333. Crozier, B., MSc Dissertation, University of London, 1973. Stevenson, A., J . Sci. Fd Agric., 1964, 15, 509. Bontoyan, W. R., J . Ass. Off. Agric. Chem., 1965, 48, 562. Killer, F. C. A., and Amos, R., J . Inst. Petrol., 1966, 52, 315. Amos, R., 2. Analyt. Chem., 1968, 236, 350. Brook, A. J. W., Davies, J. E., and King, B. M. J., in Hodges, D. R., Editor, “Recent Analytical Cardwell, T. J., and McDonough, P. S., Inorg. Nucl. Chem. Lett., 1974, 10, 283. Shankaranarayana, M. L., and Patel, C. C., 2. Analyt. Chem., 1961, 179, 263. Busev, A. I., and Teternikov, L. I., Zh. Analit. Khim., 1968, 23, 1134. Greenhow, E. J., and Spencer, L. E., Analyst, 1975, 100, 747. Greenhow. E. J., and Spencer, L. E., Analyst, 1974, 99, 82. Busev, A. I., and Ivanyutin, M. I., Zh. Analit. Khim., 1958, 11, 523. Greenhow, E. J., and Spencer, L. E., Analyst, 1973, 98, 98. Hadley, R. J . , Robinson Brothers Limited, personal communication. NOTE-References 12, 13 and 15 are to Parts VII, V and I11 of this series, respectively. Developments in the Petroleum Industry,” Applied Science, Barking, Essex, 1974, p. 97. Received April 13th, 1976 Accepted June 2nd, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100777

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

786 Analyst, October, 1976, VoJ. 101, pp. 786-789 A Digital Logic Automatic Potentiometric Titrator John T. Stock and K. 0. Wolter Department of Chemistry, University of Conneiticuf, Stows, Conn. 06268, USA A digital logic automatic potentiometric titrator is described and results are given for its use in various titrations. An automatic potentiometric titrratcx with end-point anticipation and delayed termination was described in a previous papei-.l The present device performs the same functions, but uses inexpensive solid-state devices to replace all but one of the numerous relays used in the original titrator. This new version has a much higher input impedance (about 0.5 M a), operates over a wider range of signal voltages and is more compact than its predecessor. The circuits of the titrator and of its power supply are shown in Figs.1 and 2, respectively. The components are listed in the Appendix. Replacement of the single mercury cell B, is governed by its shelf-life. The output signals from the titrator are designed to suit the adjustable two-rate burette valve described earlier.2 Switch S, allows both rising voltage and falling voltage titration systems to be handled. When connected to terminals M and with switch S, in the SET position, a pH meter or suitable voltmeter can be used to calibrate the scales of the END POINT ( E e p ) control R, and of the ANTICIPATE (Eant) control R,. If S, is moved to the TEST position, the course of the titration can be followed on the meter, which is useful when setting up the apparatus for a particular titration.The approximate maximum values of Eep and E a n t are 1 500 mV and 200 mV, respectively. Switch S, is used to apply a bucking voltage when Eep has to be negative. The WAIT control, R,,, allows from 15 to 150s to elapse between an apparent end-point and actual termination of the run (STOP). Suitably marked light-emitting diodes (LEDS) indicate the status of the titration. Having suitably adjusted the controls, the titration is initiated by momentary depression of push-button switch K, (START). Titration, here assumed to occur with a fall in cell voltage ( E c e n ) , then proceeds at the normal rate and the FAST LED, P,, lights. When Ecell has fallen to just below the value Eep + Eant, the output of operational amplifier OA, swings to the positive limit imposed by zener diode 2,; the upper input of flip-flop oscillator FF, is thus driven “high.” This change turns off the transistor combination T,T,, so that the burette valve loses its fast signal and the FAST LED goes out.However, the SLOW LED P, lights as titration continues at a slower rate. This changeover is possible because the triggering of FF, also cuts off transistor T, and thus eliminates Eant. Although the output of OA, then returns to its initial, slightly negative, state, this return cannot cause FF, to re-trigger. Eventually, Ecell falls slightly below E,, and the output of OA, again goes positive. This time the result is to drive the output of NAND gate 1 to “low,” which turns off the transistor combination T,T,, thus completely de-energising the burette valve and arresting the titration.The signal also extinguishes the SLOW LED P, and causes monostable FF, to emit a brief falling pulse. This pulse, in turn, triggers the tinier, T, and the WAIT period begins. The timer output is “high” during this period, so that the WAIT LED P, lights. As a result of the interposition of inverter I,, the bottom input of NAND gate 3 is kept at a low level, so that the output of this gate is prevented from going “low.” In instances where a false end-point occurs, Ecell will creep above E,,, so thLt the output of OA, again changes sign. This change re-starts SLOW titration and re-sets the timer. When a new end-point is reached, the flow of titrant is again arrested and the timer is re-triggered. If this end-point is true, the timer output stays “high” for the pre-set period and then goes “low.” Under these conditions all three inputs of NAND gate 3 are “high,” so that its output goes “low.” This change triggers FF,, which, in turn, turns on combination T,T,.Relay L, then “pulls in” and self-latches in this position. Only STOP LED P, is illuminated. Power cannot now be restored to the burette valve until L, is unlatched by depression of START key K,. The titration can be terminated at any time by depressing STOP key K,. An analogous sequence of events occurs in a rising-voltage titration. However, the switch- ing is such that the FAST to SLOW transition occurs when Ecell has risen just above E,, - Eant. The change of sign of Eant is brought about by OA,, which is connected as a unity-gain inverter.m Fig.1. Circuit of titrator [power connections t o the operational amplifiers ( 4 3 5 V), t o the integrated-circuit gates ( 3 - 5 V) and t o the burette 2 valves (+ 6V) are not shown]. 4788 STOCK AND WOLTER: A DIGITAL LOGIC Analyst, Vol. I01 s4 , z4 -15V I I --- Fig. 2. Circuit of power-supply unit. Although the input impedance is high, it is inadequate to accept titration systems that employ a glass electrode. However, glass-electrode titrations can be performed if a suitable pH meter is interposed between the electrode system and the titrator. For example, a Leeds and Northrup Type 7401 pH meter was found to provide adequate titrator drive. A 2.2-kQ resistor was connected across the output resistor terminals, which were then connected to the input of the titrator.Typical results obtained with the apparatus described are shown in Table I. Titration times ranged from 1-3 min, averaging 2 min. TABLE I: REPLICATE AUTOMATIC TITRATIONS OF 15.00-ml PORTIONS OF TITRAND Titrand 0.2 M HCl 0.15 M NaOH 0.01 M NaOH 0.1 M NaOH 0.1 M AgNOs 0.01 M KC1 0.02 M FeSO, Titrant 0.1 M NaOH 0.1 M HCl 0.005 M HCl 0.1 M HC1 0.05 M KCl 0.05 M AgNO, 0.01 M KMnO, Indicator electrode* Platinum? Platinum t Platinum - quinhydrone Glass: Silver - silver chloride Silver - silver chloride Platinum E.m.f. Falling Rising Rislng Rising Falling Rising Rising Number of runs 13 8 7 6 8 8 6 Standard deviationlml 0.06 0.08 0.11 0.03 0.02 0.06 0.04 * Reference electrode : saturated calomel. t Wire that had been previously heated in an oxidising flame., $ pH meter interposed between electrodes and titrator.The authors thank the Perkin-Elmer Corporation for an undergraduate research award to K.D.W. References 1. 2. 3. Stock, J. T., Analyst, 1962, 87, 908. Stock, J. T., and Fill, M. A., Autalyst, 1960, 85, 609. Kekedy, L., and Makkay, F., Tulanta, 1969, 16, 1212. Received Afiril lith, 1976 Accepted May llth, 1976October, 1976 AUTOMATIC POTENTIOMETRIC TITRATOR 789 APPENDIX List of Components Unless otherwise specified, all resistors are 0.5 W. Rl Resistor, 33 kR 23, 2 4 (BALANCE) T5-7 R3 Resistor, 8.2 kR RGl Potentiometer, 1 kR (END POINT) TR, R5 R6 R,, Rl3 Resistor, 56 kR TR3 Potentiometer, 100 S2 (ANTICIPATE) TR, Resistor, 1.2 ki2 Cl, Cll R, RlO R11 Resistor, 68 ltR c2 El, Potentiometer, 1 MS2 (WAIT) c, R13, R,, Resistor, 180 0 c4 R15 Resistor, 150 R c5-9 Rl, Resistor, 680 R ClO Rl, Resistor, 470 R Pl R18, R,,, K2, Resistor, 160 R, 2 W p 2 Rl, Resistor, 68 R, 1 W p3 Resistor. 270 R.1 W P" R2 Potentiometer 10 k '1-4 Resistor, 330 R R4 Resistor, 2.7 kR k1 Three pole, double throw switch Single pole, double throw switch Single pole, double throw switch Single pole, single throw switch Double pole, double throw push- Single pole, single throw push- Diode, 1N4001 (RISE - FALL) (BUCK) ( TEST-SET) (POWER) button switch (START) button switch (STOP) Bridge rectifier, 1 ,4 Diode, 5 A Zener diode, 6 V, 1 W Zener diode, 15 V, 1 W npn transistor 2N3392 npn transistor TIP-29-7433 Integrated-circuit regulator, 5 V, 1 A Double pole, double throw relay, 6 V Transformer, 6.3 V, 0.3 A Transformer, 12.6 V centre tap, 5 A Transformer, 25 V centre tap, 0.3 A Transformer, 6.3 V, 0.6 A Capacitor, 100 pF Capacitor, 0.01 pF Capacitor, 100 p F Capacitor, 270 p F Capacitor, 1000 pF Capacitor, 0.1 pF Red LED (STOP) Green LED (FAST) Yellow LED (SLOW) Yellow LED (WAIT) Neon pilot lamp ' Fuse, 1 A Mercury cell, 1.4 V Operational amplifier, type 741 Operational amplifier, half type 747, dual integrated circuit Assembled from type 7402 quad 2-in- put NOR gate integrated circuit Type 74121 monostable flip-flop os- cillator One-quarter type 7400 quad 2-input NAND gate integrated circuit One-third type 7410 triple %input NAND gate integrated circuit Type 7404,4 hex inverter integrated circuit Type 555 timer integrated circuit pi Fl Bl 0 A1 OA,, OA, FF,, FF2 FF3 NAND 1, NAND 2 } NAND 3 11-6 T

ISSN:0003-2654    

DOI:10.1039/AN9760100786

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

790 Analyst, October, 1976, Vol. 101, pp. 790-797 Determination of Gold in Blood Fractions by Atomic-absorption Spectrometry Using Carbon Rod and Carbon Furnace Atomisation H. Kamel, D. H. Brown, J. M. Ottaway and W. E. Smith Department of Pure and Applied Chemistry, University of .St.vathclyde, Cathedral Street, Glasgow, G 1 1XL A comparison of procedures that involve the use of either carbon rod or carbon furnace atomisation in order to determine, by atomic-absorption spectrometry, the level of gold in whole blood, plasma and serum from patients undergoing gold treatment for rheumatoid arthritis is described. A procedure using carbon furnace atomisation is preferred because of its simplicity and sensitivity. The detection limits for gold, obtained by using the preferred procedure, in serum, plasma and whole blood are 0.002, 0.002 and 0.004 5 pg ml-l, respectively.The relative standard deviations are 1.9% for 0.063 pg ml-l in serum, 2% for 0.061 pg ml-l in plasma and 7.3% for O.O30pgi~11-~ in whole blood. The method is used to confirm that most gold is carried in the serum fraction of blood., to determine the gold level in white cells and to demonstrate that the gold level in the ultra-filtrate is low. Gold complexes have been used in the treatment of rheumatoid arthritis for some time and, although they can give rise to toxic reactions in some patients, they remain among the few drugs that can cause a remission of the &~ease.l-~ ‘The original discovery of their effectiveness was accidental and, as yet, 110 definitive mechanism for their action has been postulated.As part of a programme designed to elucidate their mode of action, a method was required for the determination of the amount and distribution of gold in the blood of patients under- going chrysotherapy (gold treatment). A simple and sensitive method for this determination using carbon furnace atomic-absorption spectrometry has been developed and is reported in this paper. A variety of analytical techniques have been applied to the determination of gold in blood but, with the exception of studies involving neutron-activation analysi~,~,~ most recent studies have made use of atomic-absorption spectrometry. Gold has not yet been identified in man as a prerequisite for normal life processes and at present the gold concentration in human blood before gold treatment is below detectable levels.During treatment, the level in blood serum, the blood fraction that has been most extensively studied, is between 1 and 10 pg m1-1.lF2 Although this level is within the range of most flame atomic-absorption instruments, the dilution procedures that are necessary in order to obtain good nebulisation of serum render the methods low in sensitivity.6-H Further, it is not generally appreciated that gold complexes can readily be reduced by reagents such as amino-acids, acetates, etc., to a gold(0) colloid, which adheres at least partially to the walls of vessels. Therefore, tech- niques that require pre-treatment of the samples or a prolonged analysis time are undesirable, especially as there are considerable variations in the chemical composition of the blood of patients with rheumatoid arthritis. For example, we have recorded albumin concentrations as disparate as 3.43 and 2.47g per 100ml of blood serum.Consequently, the reducing tendency of each patient’s blood might be an important variable in such methods and would have to be checked. Carbon rod atomisation techniques have been applied to the analysis of blood serum for gold in three previous st~dies.~-ll In this paper, an assessment is made of various possible procedures and a procedure involving direct injection of blood serum into a carbon furnace is selected as being the most suitable. The simplicity and increased sensitivity of this pro- cedure are used with advantage to analyse the various fractions of whole blood and provide an answer, for the first time, to the question of the distribution of gold between cells, serum proteins and the ultra-filtrate.KAMEL, BROWN, OTTAWAY AND SMITH 791 Experimental De-ionised water and reagents of the highest available purity were used throughout.Gold(1) stock solution (250 pg ml-l). Dissolve one ampoule of the drug myochrysine (May and Baker), containing 50 mg of sodium aurothiomalate, in 100 ml of de-ionised water. Also prepare a gold(II1) stock solution (250 pg ml-l) from sodium chloroaurate. The gold(II1) solution is much more easily reduced and reacts in a complex manner with proteins. It is used to standardise the above gold( I) stock solution, which is the preferred working standard in all the procedures referred to in this paper.Preparation of Calibration Solutions Dilute 2 ml of gold(1) stock solution (250 pg ml-l) to 100 ml with water. Transfer 0, 1, 2, 3, 4, 5 and 6 ml of the solution to 100-ml calibrated flasks and dilute each to the mark with water. This procedure gives calibration solutions with concentrations of 0, 0.05, 0.10, 0.15, 0.20, 0.25 and 0.30 pg ml-l. (ii) Serum standard solutions. Dilute 4 ml of gold(I) stock solution (250 pg ml-l) to 100 ml with water. Transfer, by using a microburette, 0, 0.05, 0.1, 0.15, 0.2 and 0.25 ml of this solution into 10-ml calibrated flasks and dilute the solutions to the mark with gold-free serum. (iii) 10- and 25-fold diluted serum standard solutions. Dilute 4 ml of gold(1) stock solution (250 pg ml-l) to 100 ml with water.Transfer, by use of a microburette, 0, 0.05, 0.1, 0.15, 0.2 and 0.25 ml to a series of 10-ml calibrated flasks, each containing 1 ml of gold-free serum. Dilute the solutions to the mark with water; this procedure yields 10-fold diluted serum standard solutions. To obtain 25-fold diluted serum standard solutions use an analogous procedure, adding 0, 0.15, 0.30, 0.45, 0.60 and 0.75 ml of the gold stock solution, diluted as before, to 1 ml of serum in 25-ml calibrated flasks and dilute each to the mark. (iv) Plasma and blood standard solutions. Repeat (ii) and (iii), using plasma or whole blood instead of serum in order to obtain a similar series of plasma or blood standard solutions. Solutions of whole blood are particularly unstable but, in general, all diluted standard solutions under (i)-(iv) must be prepared fresh daily.Reagents (i) Aqueous standard solutiom. Separation of Blood Fractions Prepare serum, plasma, ultra-filtrate and cell fractions I and I1 of blood by using standard biochemical methods according to the procedure shown in Fig. 1. Separate the red and white cells by lysis of the red cells, then centrifuge the product and wash the residue frequently with physiological saline solution. This residue consists mainly of white cells with a small concentration of platelets. (The amount of red cells present is determined by difference.) Apparatus Carbon filament atomisation A Shandon Southern Instruments A3470 filament atomiser, mounted in an A3400 atomic- absorption spectrometer, was used for the measurements described under Procedure A below.Signals were measured on a Honeywell Electronik 194 recorder. Samples are added to a Type 2 graphite rod manufactured from RWO graphite (Ringsdorf Werk GmbH), which has a hemispherical sample cavity that can accept sample volumes of up to 5 p1. The design of the rod ensures that maximum temperature occurs at the sampling point while the clamped ends of the rod remain relatively cool. An attached power-control unit has variable time and temperature selectors for sequentially drying, charring and atomising the samples and, once set, this sequence of operation proceeds automatically. A Pye Unicam gold hollow- cathode lamp was used as the source and argon was used as the inert gas medium during the heating stages. Samples were transferred to the cavity by means of a micropipette.Operating conditions are given under Procedure D. Carbon furnace atomisation The instrument used for the measurements described under Procedure B was a Perkin- Elmer 306 atomic-absorption spectrometer, equipped with an HGA-72 heated graphite tube atomiser and a deuterium-arc background corrector. Atomisation signals were measured on a Servoscribe strip-chart recorder. The HGA-72 has variable time and temperature792 KAMEL et al. : DETERMINATION OF GOLD IN BLOOD FRACTIONS AnaZyst, VoL. 101 Blood sample (10 ml) I Allow t o c l o t centrifuge a t 3 000 rev min-' for 1 O min 1 I su p e h a tant serum sample Red cells Cell fraction I Resibue Fibrinogen White cells Platelets 1 ]I Add heparin to stop clotting I I Whole blood sample I Centrifuge a t 3 OOO rev min-' for 1 O rnin I Residue Platelets I Supernatant u it ra f i i tratio n Plasma sample I Residue: proteins M, > 1 000 I Ultrafiltrate: salt solution compounds M, < 1 000 Fig.1. Procedure for the separation of a blood sample into its components. selectors for sequentially drying, ashing and atomising the samples and, once set, the sequence of operations proceeds automatically. Samples are atomised in a graphite tube that is 5.3 cm long and 1 cm in diameter, under an atmosphere of argon. In this atomiser, it is also possible to increase the sensitivity by stopping the flow of purge gas automatically during the atomisation stage, thus retaining the metal atoms in the beam for a longer period of time. Operating conditions are given in Procedure D.Procedures Procedure A . Determination ojgold iqc s e w n , plasma and whole blood by a n extraction method using 4-methylpentan-2-one7 and a fomic-absorption spectrometry with carbon rod atomisation To 1 ml of gold-free serum or 1 in1 of a serum sample containing between 0.5 and 5 pg of gold, placed in a 20-ml stoppered glass tube, add 2 ml of saturated potassium permanganate solution and mix. Then add 1 ml of 6 M hydrochloric acid and mix again. After leaving it to stand for 20 min, place the tube in a water-bath for 1 h at 75 "C and then boil the contents for 2 min. Cool the resulting suspension to room temperature and add 2 ml of 4-methylpentan-2-one, stopper the tube and shake it vigorously for 2 min. Next, inject 5 p1 of the 4-methylpentan-2-one layer into the recess in the carbon rod atomiser and carry out the determination as described under Procedure D .A similar procedure is adopted for plasma and blood samples, using 1 ml of sample in each instance. Prepare standards for these pro- cedures as in (ii) above and carry out the full procedure. Procedure B. Determination of gold in serum, plasma and whole blood by direct injection or by dilution and injection into the carbon furnace atomiser Inject 5 pl of a sample of serum, plasma or whole blood containing between 0.05 and 0.3 pg ml-l of gold directly into the carbon furnace (Procedure D ) . For higher gold concen- trations (up to 7.5 pg ml-l), use the dilution procedure described in the preparation of %-fold diluted serum solutions before injection into the carbon furnace,October, 1976 793 Procedure C. Standard additions method for gold in serum, plasma or whole blood using carbon furnace atomisation Place 1 ml of a sample of serum, plasma or whole blood containing between 0.05 and 0.3 pg of gold in a 10-ml calibrated flask, transfer 0.25 ml of standard gold solution (2 pg ml-l) to the flask and make up to the mark with de-ionised water.Repeat the procedure with the addition of 0.5, 0.75, 1.0 and 1.25 ml of the 2 pg ml-l standard gold solution to further 1-ml portions of the same sample. Carry out the analysis as described under Procedure B. Procedure D. Operation of the instruments given in Table I. BY AAS USING CARBON ROD AND CARBON FURNACE ATOMISATION The carbon furnace and carbon rod atomisers were operated under the optimised conditions TABLE I OPTIMISED OPERATING CONDITIONS FOR ATOMISERS Condition Carbon filament Wavelength/nm .. .. .. .. 242.8 Spectral band width/nm . . .. .. 0.7 Lamp current/mA . . .. .. .. 10 Drying temperature/"C . . .. .. Voltage used for drying/V . . .. .. 4 Ashing temperature/'C . . .. .. Atomisation temperaturel'c . . .. .. Voltage used for atomisation/V . . .. 5 Volume of the samplelpl , . .. .. 5 - Drying time/s . . .. .. .. .. 30 Voltage used for ashing/V . . .. .. 6 Ashing time/s . . .. .. .. .. 15 - - Atomisation time/s . . .. .. .. 6 Argon flow-ratell min-' . . .. .. 1.6 Carbon furnace 242.8 10 0.7 140 30 475 30 2 000 5 6 1.5 - - - Sequentially inject samples and standards into the carbon furnace or carbon filament atomiser and record the peak height of the atomic-absorption signal during the atomisation step.Interpolate sample concentrations from the calibration graph obtained from deter- minations on standards or directly from the standard addition calibration graph. Results and Discussion The results of analyses of blood samples from six patients undergoing gold therapy, carried out by using Procedures A, B and C, are given in Table 11. As the procedures use different atomisation techniques on different instruments and different sample treatments (solvent extraction, dilution and standard addition), the reasonable agreement of the results suggests that both carbon rod and carbon furnace atomisation can be used to determine the amount of gold in serum, plasma and whole blood with acceptable accuracy.The results from Procedure A show a small negative bias compared with those from Procedures B and C and the use of Procedure B is to be preferred. These points are discussed further below. Use of Procedure A By using undiluted serum samples in an analogous manner to that described by Matousek and Stevens,12 it was found that ashing was incomplete and the residue left on the rod often obstructed the light path and led to erroneous results. The residual deposit was difficult to remove satisfactorily from the fragile rod and this operation had to be carried out after every seven samples. Increasing the temperature of ashing to red heat, as suggested by Aggett,g resulted in large losses of gold in the sample, either by spluttering or by premature atomisation of the gold.With 10-fold diluted samples, problems resulting from background absorption by the smoke were considerable, as the instrument used had no background correction facility. The use of the 4-methylpentan-2-one extraction procedure (A) was therefore essential in order to obtain reliable results. To take account of interference from other ions or molecules that are extracted with the gold, a calibration graph of the amount of gold extracted from standard794 KAMEL et aZ. : DETERMINATION OF GOLD IN BLOOD FRACTIONS Analyst, VoZ. 101 TABLE I1 DETERMINATION OF GOLD I N BLOOD FRACTIONS OF PATIENTS UNDERGOING Sample Serum Plasma Whole blood Patient 1 2 3 4 5 6 1 2 3 4 6 6 1 2 3 4 5 6 GOLD THERAPY Concentration of gold measured in pg m1-1.Procedure B Procedure A (following 10-fold dilution) P P 3.60 3.65 3.61 4.13 4.09 4.11 0.53 0.51 0.54 0.58 0.57 0.57 0.60 0.65 0.64 0.63 0.62 0.63 1.10 1.14 1.12 1.19 1.19 1.20 1.01 1.03 1.05 1.06 1.06 1.07 4.12 4.10 4.09 4.56 4.53 4.54 3.65 3.62 3.69 3.79 3.77 3.78 0.61 0.60 0.62 0.65 0.66 0.68 0.70 0.71 0.73 0.75 0.76 0.75 1.19 1.20 1.16 1.23 1.24 1.23 0.98 1.00 1.10 1.05 1.06 1.05 4.21 4.18 4.20 4.21 4.20 4.21 2.17 2.20 2.24 2.44 2.46 2.47 0.30 0.31 0.30 0.34 0.35 0.35 0.39 0.38 0.38 0.40 0.41 0.42 0.69 0.70 0.67 0.73 0.71 0.72 0.60 0.59 0.68 0.61 0.62 0.65 2.52 2.53 2.60 2.69 2.68 2.69 Procedure C 4.06 0.56 0.66 1.20 1.07 4.50 3.80 0.67 0.75 1.26 1.06 4.18 2.44 0.33 0.40 0.72 0.64 2.70 gold(1) stock solutions, prepared as in (i) above, was compared with a calibration graph of the amount extracted from standard gold serum solutions prepared as in (iii), using the conditions of Procedure A.Extraction from the serum standards was uniformly low at all concentrations ; the preparation of standards by extraction of serum solutions was therefore essential. Calibration graphs were linear in the range 0.002-0.5 pg ml-l of gold. It was found that a smaller amount of oxidant or a lesser time of oxidation than those specified in Procedure A resulted in larger discrepancies. Reproducible sample application was difficult, particularly as the carbon rod appears to be permeable to 4-methylpentan-2-one. It seems possible that the negative bias on results obtained with Procedure A is caused by the incomplete extraction of gold from a strong gold - protein complex in the samples, which is not formed during the short preparation time of the serum standards.Despite this bias, the results from all three methods can be considered to be in adequate agreement. Use of Procedure B In this procedure, optimisation of the time and temperature of ashing is important in order to ensure that all of the matrix is destroyed, that the evolution of smoke and other causes of background absorption are minimised and that none of the gold is lost at this stage. Ashing at 475 "C for 30 s satisfies all three requirements. By using method (iii) under Preparation of Calibration Solutions, samples containing between 0.05 and 0.30 pg ml-l of gold in 0-, 5-, 10- and 25-fold diluted serum solutions were prepared.The results of the analysis of these solutions using Procedure B, under the operating conditions of Procedure D, are shown in Fig. 2. As, for each serum dilution, the results give a straight line graph that passes through the origin, it is clear that the ashing procedure and background corrector are working efficiently in overcoming the effect of smoke for all serum dilutions. However, the sensitivity of the signal for a constant level of gold increases with dilution of the serum, reaching a maximum at the 10- and 25-fold dilutions. The atomisation signal for 0.1 pgml-l was observed as a function of time for 0-, 5- and 10-fold serum dilutions (Fig. 3) and it was clear that peak area was increasing with dilution as well as peak height. Blank serum gave no signal at all concentrations.In concentrated serum solutions, some gold appeared to be held up and released towards the end of the atomisation stage (Fig. 3). Possibly a very stable protein complex, which was not completely destroyed under these conditions during the ashing procedure, prevented rapid atomisation of part of the gold. Another possible explanation is that during the ashing stage some goldOctober, 1976 BY AAS USING CARBON ROD AND CARBON FURNACE ATOMISATION 14 12 2 0 795 Concentration of gold / pg ml-' Fig. 2. Results of the analysis of samples containing between 0.05 and 0.3 pg ml-1 of gold in 0- ( A), 5- (a), 10- (A) and 25-fold (A) diluted serum, compared with aqueous gold stan- dards (A). The points for the 10- and 25-fold diluted sera and the aqueous gold standards are coincident, was deposited away from the centre of the tube, thus reducing the sensitivity of the signal.This effect would be reduced as the serum level was reduced. Because there was no problem of sensitivity, we used a 10-fold diluted serum solution in the concentration range 0.05-0.3 pg ml-l of gold and a 25-fold diluted serum solution between 0.3 and 7.5 pg d-l. The major reasons for preferring Procedure B to Procedure A are its relative simplicity, its more economic use of time and materials and its higher sensitivity. A comparison of the relative standard deviation and detection limits for serum gold determinations using the two procedures is given in Table III, together with the same information using Procedure B for Fig. 3.Atomic-absorption signals as a function of time for samples containing 0.1 pg ml-1 of gold at various serum dilutions. 4, No dilution; a, 5-fold dilution ; A, 10-fold dilution. Atomisation starts at A. Chart speed, 1 cm s-l. Each division is equivalent t o 1 s.796 KAMEL et aZ. : DETERMINATION OF GOLD IN BLOOD FRACTIONS Analyst, VoZ. 101 plasma and blood samples. The accuracy of Procedure B was established by use of the standard addition method, Procedure C. An example is shown in Fig. 4 and the results are tabulated for each instance in Table 11. Both plasma and serum samples gave constant results over a period of 24 h when the samples were kept at room temperature but there was some reduction in the signal for whole blood. TABLE 1I:I COMPARISON OF PROCEDURES A AND B FOR THE ANALYSIS OF VARIOUS BLOOD FRACTIONS Mean Relative Detection limit concentration Number of standard -7 Sample Procedure found/pg ml-l results deviation, % pg ml-l g Serum 1 A Serum 2 A Serum 2 B 0.361 0.061 0.063 6 6 10 2.5 } 0.005 2.5 x 10-lf 4.1 1.9 0.002 1 x 10-11 Plasma 2 B 0.061 10 2.0 0.002 1 x 10-11 Whole blood 2 B 0.031 10 7.4 0.004 6 2.2 x 10-l1 Biochemical Implications of the Results The results demonstrate a number of possible advantages that can accrue from the use of this technique in a clinical laboratory.It is clear that most of the gold in blood is bound to the serum proteins, confirming that the gold level in serum, which is usually the deter- mination reported in clinical studies, offers a reasonable approximation to the bulk gold level in blood.Two recent studies1y2 suggest that this measurement is not a useful measure of whether or not gold therapy will be successful but there is a suggestion1y2 that there could be a link between abnormally high levels of gold and toxic skin reactions; this suggestion requires further investigation. The simplicity of Procedure B would make the routine screening of a large number of patients undergoing gold therapy a relatively simple matter. Concentration of gold / pg ml-1 Fig. 4. Example of the standard additions method for the determination of gold in serum; A, sample 1, 4.06 pg ml-l; C, sample 2, 0.56 pg ml-l; B, sample 3, 0.66 pg ml-l; and D, aqueous gold. standards. It is possible that the gold concentration in a firaction of blood other than serum is a more useful indication of the effectiveness of gold therapy.The blood of patient 6 was therefore studied in more detail, taking advantage of the higher sensitivity of the carbon furnace technique in order to determine the gold level in less heavily doped samples without the need for extra pre-concentration stages. A sample of the ultra-filtrate contained no detectable gold (less than 0.004 pg ml-l) and we have confirmed that this was not due to a reaction of the gold with the sinter. We have demonstrated that free myochrysine will pass throughOctober, 1976 BY AAS USING CARBON ROD AND CARBON FURNACE ATOMISATION 797 the sinter, so that it can be concluded that there was no free drug in this sample of blood serum, although it may be that some of the gold thiomalate anion is bound unchanged to the serum proteins.Cell fraction I (Fig, 1) contained 4.1% of the total blood gold, with not less than 90% of it in the white cells. It was difficult to obtain a more exact figure because of problems with the white cell separation procedure. It is possible that the amount of gold within particular types of cell might give a better indication of the nature of the clinical response and the simplicity and sensitivity of the carbon furnace may prove to be particularly useful in this type of study. We thank Professor W. Watson Buchanan and Dr. Alistair Kennedy of the Rheumatoid Arthritis Centre, Glasgow, for the provision of samples and for their encouragement throughout, Shandon Southern Instruments Ltd. for the loan of the A3400 and A3470 instruments, and the Royal Society for the award of a research grant to one of us (J.M.O.) for the purchase of the HGA-72 atomiser. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 13. References Jessop, J. D., and Johns, R. G. S., Ann. Rheum. Dis., 1973, 32, 228. Lorber, A., Atkins, C. J., Chang, C. C., Lee, Y. B., Starrs, J., and Bovy, R. A., Ann. Rheum. Dis., Empire Rheumatism Council Final Report, Ann. Rheum. Dis., 1961, 20, 315. Brune, D., Samsahl, K., and Wester, P. O., Clin. Chim. Acta, 1966, 13, 285. Kriesius, F. E., Markkanen, A., and Pelota, P., Ann. Rheum. Dis., 1970, 29, 232. Dunckley, J. V., Clin. Chem., 1971, 17, 992. Balazs, N. D. H., Pole, D. J., and Masarei, J. R., Clin. Chim. Acta, 1972, 40, 213. Dietz, A. A., and Rubinstein, H. M., Ann. Rheum. Dis., 1973, 32, 124. Aggett, J., Analytica Chim. Ada, 1973, 63, 473. Aggett, J., and West, T. S., Analytica Chim. Acta, 1971, 55, 349. hfaesson, F. J. H. J., Posna, F. D., and Balke, J., Analyt. Chem., 1974, 46, 1445. Matousek, J. P., and Stevens, B. J., Clin. Chem., 1971, 17, 364. 1973, 32, 133. Received March 24th, 1976 Accepted May 7 t h 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100790

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

798 Analyst, October, 1976, V d . 101, $9. 798-802 A Kinetic Theory of Atomisation for Atomic-absorption Spectrometry with a Graphite Furnace Part IV.* Assessment of Interference Effects C. W. Fuller Tioxide International Limited, Stockton-on-Tees, Cleveland, T S 18 2NQ Interference effects in atomic-absorption measurements using electrothermal atomisation are well known but poorly understood. The effects of variations in the rate of atomisation, the rate of loss of atoms and the efficiency of atomisation on absorbance veYsu.s time profiles, in a graphite furnace, have been established. These effects are proposed as a basis for establishing the causes of interferences that occur in atomic-absorption spectrometry using a graphite furnace atomiser. In a recent publication,l the basis for a kinetic model to describe the atomisation process in a graphite furnace under isothermal conditions was established.In subsequent publications, this model was used to describe and quantify various analytical operating parameters.2J Atomic-absorption spectrometry using electrothermal atomisation has been found experi- mentally to be subject to a number of interference effects. While a considerable amount of qualitative information is now building up, there is still no basic understanding of the problems to explain the cause of these effects. In this final paper in the series, the use of the kinetic model of atomisation to assess inter- ference effects that occur during the atomisation period is outlined. Experimental Atomic-absorption measurements were made using a Perkin-Elmer HGA-70 graphite furnace fitted to a Perkin-Elmer, Model 103, atomic-absorption spectrometer with a Model 166 recorder.The experimental conditions have been described previously in Parts I1 and II.2 Other conditions applicable to the work described here are given in the text where appropriate. Results and Discussion Assuming that a general equation, similar to that derived for copper,l can be written to describe the concentration of metal atonis, M , in a graphite furnace at time t, then where Mo is the initial amount of element introduced into the furnace, k, is the first-order rate constant for the formation of metal atoms and k , is the first-order rate constant for the removal of metal atoms from the furnace. On integration, equation (1) gives [exp (-k,t) - exp (- kzt)] .. . . . . M = M 1 O k , - k, k By inserting a proportionality factor, p , this equation can be rewritten to give the equation for the measured absorbance at any time t, i.e. This factor p will be a function of the oscillator strength for each element (a constant) and From equation (3), it is apparent that there are three separate factors that will influence * For Part I11 of this series, see reference list, p. 802. the efficiency with which metal atoms are produced.FULLER 799 the shape of an atomic-absorption signal profile, i.e., k,, k , and p . If during an analytical determination any one of these parameters changes then there will be a change in the absor- bance versus time profile and an interference effect will be observed. The term k, represents the dependence on the rate of formation of metal atoms in the atomiser and its value may change for the reasons given below.(i) Physical: entrainment of an element in the presence of a matrix causing either (a) increased or reduced contact between the metal salt and graphite with the possibility of a change in the reduction rate of the metal salt or (b) a faster or slower rate of evaporation of the free metal. (ii) Chemical: the formation of more or less stable compounds prior to atomisation due to the presence of other species. The rates of reduction or decomposition will normally vary for different compounds. Fig. 1 shows that for variations in the value of k, (0.01-1.0 s-1) at fixed values of k , (0.1 s-l), p (1 x 108) and M, (1 x g) the peak height value increases and the time to reach the maximum absorbance value decreases with increasing values of k,.0 1'3 20 30 40 Time/s Fig. 1. Effect of variations in the value k , on an absorbance veisus time profile: k , 0.1 s-1, p = 1 x 108, Mo = 1 x 10-8 g. The term k , represents the dependence on the rate of removal of metal atoms from the atomiser and its value may change for the reasons given below. (i) A change in the flow-rate of the inert gas through the atomiser. (ii) A change in the diffusion rate of metal atoms through the walls of the graphite furnace caused by (a) the degradation of the graphite furnace or (b) the formation of a pyrolytic graphite coating on the graphite furnace. Fig. 2 shows that for variations in the value of k , (0.01-1.0 s-l) at fixed values of k, (0.1 s-l), p (1 x lo8) and Mo (1 x 10-8 g) the peak height value and the time to reach the maximum absorbance value increase with decreasing values of k,.The term p represents the dependence on the proportion of the element, introduced into the atomiser, which is converted into vaporised metal atoms compared with the vaporisation of other species, e.g. , oxides, chlorides and organometallics. It therefore reflects any changes in the atomisation efficiency. Fig. 3 shows that for variations in the value of p (0.14.0 x lo8) at fixed values of k, (0.1 s-l), K , (0.2 s-l) and M, (1 x lo-* g) the peak height value increases with increasing values of 9. The time to reach the maximum absorbance is independent of changes in the value of 9.Further information can be obtained by examining the variation of the integrated signal with variations in the values of k,, k , and 9. From equation (3) , (4) k, exp (-k,t) - exp (-K,t) t = 00 t z 0 * . (Absorbance) = fi M0- k2--1 [ k 2 k, 1 t = 00 - p M o . . . . .. Absorbance -- = k2800 FULLER : A KINETIC THEORY OF ATOMISATION Analyst, Vol. 101 0 25 50 75 Time/s Fig. 2. Effect of variations in the value of k , on an absorbance veYsus time profile: k, = 0.1 s-1, P = 1 x lo8, M, = 1 x lO-'g. 0 25 50 Tirne/s Fig. 3. Effect of variations in the value of p on an absorbance versus time profile: R , = 0.1 s-1, k , = 0.2 s-1, M, = 1 x 10-8 g. The integrated absorbance signal is therefore independent of k, but linearly proportional to p and inversely proportional to k,.Equation (5) shows clearly why signal integration techniques cannot compensate for interference effects in all instances. Compensation is possible for only one of the three causes of interference, Le., where the rate of atomisation (k,) varies. The information described above for assessing the causes of interference effects is summarised in Table I. This summary shows that the change in any one parameter is not a positive guide to the cause of an interference effect. However, by comparing the changes in all three parameters an immediate guide to the type and cause of interference can be obtained, even when two separate interference effects occur simultaneously. TABLE I EFFECTS ON PEAK ABSORBANCE, TIME TO REACH PEAK ABSORBANCE AND INTEGRATED ABSORBANCE CAUSED BY INCREASING THE VALUES OF THE THREE PARAMETERS k,, k , AND p Effect on observed parameter f A \ Increased Time to reach Integrated variable Peak absorbance peak absorbance absorbance k l Increase Decrease Constant Decrease Decrease Decrease Increase Constant Increase The assessment of interference effects using the criteria described above is illustrated by the following three examples.Eflect of pyrolytic graphite on the absorbance signal for vanadium Fig, 4 shows the absorbance signals obtained for 40ng of vanadium when atomised in a normal graphite tube and in a graphite tube coated with pyrolytic graphite. The pyrolytic graphite coating was achieved by heating the furnace at approximately 2 200 "C for 6 min while adding 100 ml min-l of argon - methane mixture (9 + 1) to the inert gas purge entering the furnace.With the pyrolytic graphite coating, the peak absorbance value is greater, occurs slightly sooner and decays more rapidly than with the normal graphite tube. All of these observations are consistent with an increase in the value of k,. However, the integrated absorbance valueOctober, 1976 FOR AAS WITH A GRAPHITE FURNACE. PART I V 801 is greater, which is not consistent with an increase in the value of k, being the sole cause of the interference. It would appear, therefore, that the increased signal obtained for vanadium is caused by a more rapid rate of atomisation and also by a more efficient production of vanadium atoms (i.e., an increase in the value of 9). These facts could be explained by a decrease in the formation of a vanadium carbide.I I 10 5 0 Time/s Fig. 4. Effect of pyrolytic graphite on the absorbance vevsus time profile for vanadium: 20-p1 aliquot of a 2.0 pg ml-l vanadium solution. Instrument setting: programme 4 (30 s) and 10 V atomisation. A, Normal graphite tube; B, graphite tube with a pyrolytic graphite coating. There is no evidence to show that the increased signal is caused by a lowering of the diffusion rate of metal atoms through the walls of the graphite tube. This effect could still be a minor factor, however, that is masked by the other more predominant effects. Efect of reduced gasjow on the absorbame signal for cadmium The simplest method of illustrating the effect caused by changes in the rate of removal of metal atoms from the furnace is to stop the flow of inert gas.Removal of metal atoms can then only occur by vapour diffusion. This effect is shown in Fig. 5 for 0.2 ng of cadmium. The result obtained is in good agreement with that predicted from Fig. 2, i.e., an increased peak absorbance, a delayed time to reach the peak absorbance and an increased value for the integrated absorbance signal. Efect of nitric acid and copper on the absorbance signal for telluriztm Fig. 6 shows the variations in absorbance signals obtained for long of tellurium by the presence of 0.1% V/V of nitric acid and 2 pg of copper. The addition of nitric acid has no significant effect on the shape of the tellurium absorbance profile and only brings about an increase in the magnitude of the signal, i.e., enhances the value of @.This result could be explained by the stabilisation of tellurium oxide through the formation of the compound 2TeO,.HNO,. This effect could reduce the proportion of tellurium lost as tellurium oxide, producing a higher conversion to metal atoms during the atomisation stage. In the presence of copper, the tellurium absorbance profile changes in two ways. Firstly, the atomic-absorption signal does not become significant until several seconds later with the result that the peak absorbance value also occurs later. Secondly, the rate of atomisation is slower. This result indicates that a more stable compound, e g . , copper telluride, is formed in the presence of copper as the atomisation process becomes significant at a higher tempera- ture (the furnace temperature increases for the initial 5-10 s before becoming relatively constant).The atomisation reaction then proceeds at a slower rate. These examples show that the causes of interference effects occurring in atomic-absorption determinations, using electrothermal atomisation, can be assessed by careful examination802 FULLER 0.3 0.2 g -f? 2 C Q 0.1 6 0 Time/s Fig. 6. Effect of reduced gas flow on the absorbance veYsus time profile for cad- mium: 10-pl aliquot of a 0.02 pg ml-' cadmium solu- tion. Instrument condi- tions: programme 4 (30 s) and 44 V atomisation. A, Normal operation; B, zero gas flow. .10 Q, C m 42 .05 0, 2 20 15 10 5 0 Time/s Fig. 6. Effect of nitric acid and copper on the absorbance zlersws time profile for tellurium: 2 0 - 4 aliquot of a 0.5 pg ml-l tellurium solution. Instrument conditions: programme 4 (30 S) and 5 i V atomisation. A, aqueous; B, 0.1% V / V HNO,; C, 0.1% V / V HNO, + 100 pg ml-I Cu. of the absorbance vwsus time profiles. The observed effects should be compared with those illustrated in Figs. 1-3 and Table I. The interpretation of these interferences is also greatly improved by studying the effects at lower temperatures than those normally used for analytical determinations. An increased resolution of the absorbance profile is obtained at the slower atomisation rates achieved. However, care should be taken that the interference effect is not altered by changing the atomisation temperature. This work is published by permission of the Directors of Tioxide International Limited. References 1. 2. 3. Fuller, C. W., Analyst, 1974, 99, 739. Fuller, C. W., Analyst, 1975, 100, 229. Fuller, C. W., Proc. Analyt. Div. Chem. SOC., 1976, 13, 273. No=-References 1, 2 and 3 are to Parts I, I1 and 111 of this series, respectively. Received June 2nd, 1976 Accepted June loth, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100798

出版商:RSC    

年代:1976

数据来源: RSC