ISSN:0003-2654    

DOI:10.1039/AN97499FX017

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Volume 99, No. 1178, Pages 241 -31 2 May, 1974THE ANALYSTTHE JOURNAL OF THE SOCIETY FOR AffALY71CAL CHEMISTRYCONTENTSPageR EVIEW PAPERThe Determination o f Some 1.4-Benzodiazepines and their Metabolitesin Body Fluid-. M. Clifford and W. Franklin Smyth . . I . . .ORIGINAL PAPERSA Method f o r the Direct Titrimetric Determination o f Iron(ll1) in SilicateDetermination o f Rubidium, Caesium, Barium and Eight Rare EarthElements in Ultramafic Rocks by Neutron-activation Analysis-A. 0. Brunfelt, 1. Roelandts and E. Steinnes . . . ( I I . - 6 * .The Determlnation o f Iridium and Ruthenium in Rhodium Sponge bySolvent Extraction Followed by Atomic-absorption Spectrophoto-The Use of Ascorbic Acid to Eliminate Interference from Iron in theAluminon Method for Determining Aluminium in Plant and SoilRocks-J.M. Murphy, J. 1. Read and G. A. Sergeant . . . . ..metry-M. A. Ashy and J. B. Headridge , , .. . . . . . .Extracts-T. C. Z. Jayman and S. Sivasubramaniam . . .. ..241273277285296A Method of Collecting and Concentrating Headspace Volatiles f o rGas-chromatographic Analysis-R. E. Hurst . . .. . .A Refractometric Method f o r the Approximate Measurement o f theAlcoholic Strength of Wines a t Room Temperature-J. R, Cooke . .Book Reviews . . w . * I ’. . . . . .. .. * .306310Summaries of Papers In this Issue * . . . .. . . vl, xii. xivPrinted for the Society for Analytical Chemistry by Hdfers Printers Ltd., Cambridge, EnglandAllcommunicationato beaddrer8edtothe Managing Editor,ThsAnalyst. @/lOSauile Row, London, WlX1 AFEnquiries about advertisemts should be addressed to J. Arthur Cook, 9 Uoyd Square, London, WC1X 9BAEntered as Second Class at New York. USA, Post Offic

ISSN:0003-2654    

DOI:10.1039/AN97499BX019

出版商:RSC    

年代:1974

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97499FP049

出版商:RSC    

年代:1974

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97499BP055

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

MAY, 1974 THE ANALYST Vol. 99, No. 1178 The Determination of Some 1,4=Benzodiazepines and their Metabolites in Body Fluids A Review* BY J. 81. CLIFFORDf. AND W. FRANKLIN SMYTH (RA F Institute of Aviation Medicine, Farnborough, Hampshire) (Chemistry Department, Chelsea College, Manvesa Road, LoNdon, S . W.3) SUMMARY OF CONTENTS Introduction Structure and nomenclature Physical properties Chemical properties Solvent extraction Chromatography Paper chromatography Thin-layer chromatography Gas - liquid chromatography Liquid chromatography Ultraviolet and visible spectrophotometry Spectrophotofluorimetry Polarography Mass spectrometry Nuclear magnetic resonance spectroscopy Infrared spectroscopy Comparison of analytical methods Miscellaneous methods Protein binding Analysis of the metabolites of some 1,4-benzodiazepines Chlordiazepoxide Diazepam Oxazepam Nitrazepam Lorazepam Xledazepam Flurazepam Addendum INTRODUCTION THE potential of the 1,4-benzodiazepines as useful drugs in medicine was first exploited by Randall in his studies on ch1ordiazepoxide.l Since then a great many of these compounds have been synthesised and tested for pharmacological activity. In Table I the pharmacology of seven important 1,4-benzodiazepines including observed side effects and toxicological data is briefly summarised.Comprehensive reviews by Zbinden and Randall2 and Schallek, Schlosser and Randall3 have compared the pharmacological actions of the therapeutically active members of this group. Because of the present importance in clinical practice and also the misuse of these drugs, it is fclt that a review of the rapidly increasing literature on the analytical chemistry of the 1,4-benzodiazepines and their metabolites is needed, especially with reference to their determination in body fluids. * Reprints of this paper will be available shortly.For details see summaries in advertisement pages. t Present address: G. D. Searle and Co. Ltd., Lane End Road, High Wycombe, Bucks., HP12 4HL. @ SAC and the authors. 241242 CLIFFORD AKD SMYTH : DETERMINATIOK OF 1 ,hENZODIAZEPINES [A%abSt, VOl. 99 TABLE I SUMMARY 01; THE PHARMACOLOGY OF SOME 1,4-BENZODIAZEPINES Benzodi- azepine Chlordi- azepoxide Diazepam Oxazepam Nitrazepam Lorazepam Medazepam Flurazepani Major use in man and usual dose Other uses Monkey (4 in man taming* 4 ,.1ranquilliser; Anaesthetic sedative premedicatiori ; (5-10) in alcohol Tranquilliser ; Anticonvulsant ; 4 anxiolytic muscle relaxant ; (2-10) mood elevation; withdrawal hypnotic (20-30 mg) ; minor anaesthetic Tranquilliser Anxiolytic 7 (10-30) Hypnotic Anticonvulsant 1 (5-1 0) Tranquilliser Sedative 4 Sedative : Musclc 2 anxiolytic relasant 4 Hypnotic - (2-4) ( 15-40) (1 5-30) Muscle relax ant * 3 2 Anticon- vulsant* 7 4 LD,ot/ mg kg--l 720 orally, rats 780 orally, mice 5000 orally, rats 825 orally, rats 5000 orally, rats 1020 or a 11 y , rats 870 orally, mice Side effects Drowsiness ; unsteadiness ; dependence Fatigue ; drowsiness ; unsteadiness ; dependence Drowsiness Drowsiness ; light- headedness Drowsiness Drowsiness * Experiments in animals: rank order of effectiveness. 7 LD,, is the close required to kill 50 per cent.of the population of the animals concerned. STRUCTURE AND NOMENCLATURE The individual compounds are listed in Table I1 together with their systematic and proprietary names. Two important 1,4-benzodiazepines that have slight variations on this general structure are medazepam (Nobrium) and chlordiazepoxide (Librium) : medazepam is methyl substituted on the nitrogen atom in the l-position, has methylene groups in the 2- and 3-positions, no substitution on the 5-phenyl ring and a chlorine atom in the 7-position; its systematic name is 7-chloro-2,3-dihydro-l-methyl-5-phenyl-l~-l,4-benzodiazepine. Chlordiazepoxide has the most diverse structure, which is shown below (11). R' The 1,4-benzodiazepines have the general structure shown below (I).I II PHYSICAL PROPERTIES With the exceptions of chlordiazepoxide hydrochloride and flurazepam dihydrochloride, which exhibit solubility in water, the others, which are free bases, are insoluble in this medium and are practically insoluble in n-heptane and n-hexane but soluble in methanol, ethanol and chloroform. References to the physical properties of these 1,4-benzodiazepines are given in the Merck ~nanual,~ and by Beyer and Sad6e5 and MacDonald, Michaelis and Senkowski.6May, 19741 AND THEIR METABOLITES I N BODY FLUIDS 243 TABLE I1 SOME THERAPEUTICAL1,'IT IMPORTANT 1,4-BENZODIAZEPINES Proprietary Compound R' R" R"' li"" Systcniatic name name 3H-1,4-benzodiazepin-Z-one \la1 i u m 3 H-l,4-benzodiazepin-2-one Serenid-D benzodiazepin-2-one hfogadon 3-hydroxy-2H- 1,4-benzodiazepin-2-0ne Ativan (2-fluorophcnyl)-l,3-dihydro-2H-1,4- benzodiazepin-2-one Diazepam .. CH, H H C1 7-Chloro-1,3-dihydro-l-methyl-5-phenyl- Oxazepam . . H OH H C1 'i-Chloro-l,3-dihydro-3-hydroxy-5-phenyl- Nitrazepam . . H H H KO2 1,3-Dihydro-'i-nitro-5-phenyl-2H-1,4- Lorazepam . . H OH C1 C1 7-Chloro-5-(o-chlorophenyl)-l,3-dihydro- Flurazepam . . - (CH,),. €1 F C1 7-Chloro-1-(2-diethylaminoethyl)-5- Dalmane* N(C,H,), * United States Pharmacopoeia. The pK, values, as determined by ultraviolet spe~tropl~otometry,~ are listed for seven of the compounds in Table 111. Barrett, Smyth and Davidson' have considered the ultra- violet spectra to be due to the superimposition of the spectra of two benzene rings, one mono-substituted and the other tri-substituted, within the molecule.On this basis, sites of protonation (principally at the nitrogen atom in the 4-position) and deprotonation (for those 1,4-benzodiazepines which possess an N-H bond in the 1-position) have been predicted and the differences in the observed pK, values explained. The 1,4-benzodiazepines that possess a second acid-dissociation constant therefore exist as neutral molecules at a pH greater than pK, and pH less than pK,. Chlordiazepoxide, diazepam and medazepam will exist as neutral molecules at pH values greater than their sole acid-dissociation constants. It is as neutral molecules that these drugs are generally extracted from body fluids into non-aqueous solvents. TABLE I11 pk', VALUES OF SOME 1 ,4-BENZODIAZEPINES7 Compound PK, PK2 Chlordiazepoxide .. .. 4.6 - Diazepam . . . . . . 3.3 - Nitrazepam . . . . . . 3.2 10-8 Medazepam . . . . . . 4.4 Flurazepama . . . . Oxazepam . . . . . . 1.7 11.6 - Lorazepam . . . . . . 1.3 11.5 . . 1.4 - CHEMICAL PROPERTIES Stock solutions of the 1,4-benzodiazepines, with the exception of medazepam, in methanol and absolute ethanol are stable for several weeks. Aqueous solutions are more likely to decompose, particularly under acidic and alkaline conditions. Acid-catalysed hydrolysis produces the benzophenone compound (111) ; compound (IV) is formed by fission of the amide linkage in alkaline media. The benzophenone compound (111) is particularly R' 1 R' 'I GR"'244 CLIFFORD AND SMYTH : DETERMINATION OF 1 ,PBENZODIAZEPINES [AnaZyst, VOl.99 useful in analytical procedures as its use results in greater sensitivity of ultraviolet spectro- photometric and gas - liquid chromatographic determinations than if the 1,4-benzodiazepine itself was used. The 1,4-benzodiazepines have been reduced chemically to yield products identical with those obtained by polarographic redu~tion,~-ll, e.g., the N-oxide and 4,5-azomethine groups in chlordiazepoxide were reduced separately by catalytic hydrogenation over Kaney nickel followed by the same process over platinum. The end product, the 4,5-dihydro compound, was identical with that obtained with preparative electrolytic procedures at a potential corresponding to the plateau following the first two polarographic reduction waves of the compound.Complete polarographic reduction of nitrazepam simulated the metabolism of this compound in mammals in which the 7-amino compound is one of the metabolites naturally produced. Oxidative procedures did not yield compounds that are recognised as metabolites. Other chemical reactions such as nitration, halogenation and skeletal rearrangements are fully discussed in the reviews of Childress and Gluckman,12 Archer and Sternbach13 and Sternbach.14 SOLVENT EXTRACTION The organic solvents of choice in extraction methods for the 1 ,.l-benzodiazepines and their metabolites are, in order of preference, diethyl ether, ethyl acetate and chloroform. n-Heptane has been used as a selective solvent in a few instances but not benzene as its intense ultraviolet spectrum interferes with subsequent spectroscopic determinations in that region of the spectrum.The various solvent-extraction procedures have been collated in Table IV. The influence of pH on the extraction with chloroform of various 1,4-benzodiazepines (standard amount 1 mg in 10 ml of aqueous phase) was studied by Tompsett15 and his results in eleven out of twelve extractions were in agreement with the pH region in which the readily extractable neutral molecule exists, as discussed in an earlier section. The influence of pH on these solvent extractions is shown for the chlordiazepoxide lactam metabolite (V), which could not be extracted from acidic solutions (pH 1) owing to the existence of a protonated N-oxide group.' This lactam also could not be extracted at pH 13 at which pH it existed as an anion, in contrast to the parent compound which does not possess an ionisable N-H group.V Clifford16 and Groves and Smyth8 have studied the extraction of nanogram amounts of nitrazepam and flurazepam, respectively, into a variety of solvents and found dichloro- ethane to be a particularly good solvent for nitrazepam and ethyl acetate for flurazepam. Optimum pH values for extraction were again found to be between pK, and pK,. Proelss and Lohmannl7 have also used knowledge of the pK, values of various sedatives and tranquillisers in order to facilitate their extraction from biological media prior to gas - liquid chromatography: extraction into diethyl ether at pH 8 gave optimum recoveries for the 1,4-benzodiazepines (81 to 93 per cent.). Table IV illustrates that most in vivo extractions have been carried out at a pH greater than pK, and less than pK, and the most commonly used solvent was diethyl ether.May, 19741 AND THEIR METABOLITES I N BODY FLUIDS TABLE IV SOLVENT-EXTRACTION PROCEDURES APPLIED TO BIOLOGICAL MEDIA Refer- Method of Biolog.ica1 Drug ence Chlordiazepoxide 18 Chlordiazepoxide ; 19 Chlordiazepoxide ; 20 diazepam nitrazepam (as benzophenones) Chlordiazepoxide 21 (as benzophenone) Diazepam 22 Diazepam 23 Diazepam ; 24 medazepam Diazepam 25 Diazepam ; 26 medazepam ; oxazepam Oxazepam 27 Nitrazepam 28 Nitrazepam 29 Nitrazepam 30 Nitrazepam 31 (as benzophenone) Lorazepam 32 Flurazepam 33 Benzodiazepines 34 Benzodiazepincs 17 1,C-Benzo- 35 (not nitrazepam) diazepines " determination sample Spectrofluori- Plasma metry Polarography Blood; Spectrophoto- Urine ; metry serum urine GLC Plasma GLC Blood GLC Serum GLC Blood Polar ograph y Plasm a GLC Blood ; urine TLC Plasma Spectrophoto- Plasma ; metry - TLC urine Spectrophoto- Urine metry Colorimetry Urine GLC Plasma GLC Serum Spectrophoto- Plasma ; metry and urine spectro- fluorimetry GLC Blood GLC Serum TLC Not stated Buffer used and pH 0.2 M phosphate buffer, pH 7.2 Alkaline with N%c2~ HCI (acidic reaction) (ii) 10 N HC1 + NaHCO, (neutral) (iii) 10 N NaOH (alkaline) 0.1 N NaOH, pH 9-10 1 M KH,PO,, pH 7.0 2 M acetate buffer, pH 7.0 Borate buffer, pH 9.0 1 N NaOH Phosphate buffer, pH 7-2 Buffer, pH 11-12 Magnesium oxide, pH 10 Buffer, pH 9-10 20 per cent.N+CO,, pH 9-11 Boric acid - KC1 - sodium carbonate, pH 9.0 Phosphate buffer, PH 7 Phosphate buffer, PH 9 1 M KH,PO,, pH 7 Borax buffer, pH 8 l/lOth volume of con- centrated HC1 CHROMATOGRAPHY 245 Solvent Ether : hydrolyse in glycine - HC1 buffer at pH 4.8 to form a lactam; re-extract into ether Ether : back-extract into 1 N HCl Chlordiazepoxide benzo- phenone from (i), (ii) and (iii) n-heptane + CHCl,.Nitrazepam benzophenone from (ii) and (iii) CHCl, n-Heptane - 3-methyl- bu tan- 1-01 (98.5 + 1.5) Ether : extract into 6 N HC1, hydrolyse t o benzophenone and extract with ether Chloroform Ether Light petroleum Ether (re-extract into 2.0 N HC1; neutralise, extract into ether) (b.p. 40 to 60 "C) Ether Dichloromethane - ethyl acetate (2 + 1) Ethyl acetate Ethyl acetate Diethyl ether Ether : back-extract into 12 N H,SO, and hydrolyse, neutralise and re-extract into ether Ether : back-extract into HC1, hydrolyse, make alkaline and re-extract into ether Ether Ether Chloroform It is generally necessary to separate the parent drug from its metabolites and all of them from blood plasma, urine or tissue before measuring their levels in biological specimens.Apart from a simple clean-up procedure, the use of chromatography obviates recourse to lengthy separations; however, the drawback of certain types of chromatography is a lack of sensitivity, which precludes their use in the determination of drug levels following therapeutic dosage, but as both paper and thin-layer chromatography are able to separate closely related compounds these techniques lend themselves to metabolite analysis.3G The low volatility of246 CLIFFORD AND SMYTH : DETERMINATION OF 1,4-BENZODIAZEPINES [Analyst, VOl.99 some drugs and their metabolites has imposed limitations on the use of gas - liquid chromato- graphy. However, the sensitivity of detectors, especially electron-capture detection, is greater than the non-radiochemical methods of detection used in paper and thin-layer chromatography. PAPER CHROMATOGRAPHY- Fox37 reviewed the r81e of paper chromatography in the isolatiob and identification of drugs. R, values for chlordiazepoxide, diazepam and nitrazepam have been determined by using the procedure of Curry and The RF value x 100 of diazepam when the latter is run by the reversed-phase method of Street39 was found to be 26, but values for the other 1,4-benzodiazepines were not stated.Stevens and Jenkins40 used a paper loaded with silica gel and a mobile phase containing chloroform - ethanol (49 + 1) to achieve the chromatographic separation of diazepam and nitrazepam as they found that the method of Curry and Powell38 gave poor resolution of these drugs (Table V). and are included in Table V. TABLE V PAPER CHROMATOGRAPHY OF CERTAIN 1,4-BENZODIAZEPINES RF x 100 Method of Method of Drug Curry and Stevens and Jenkins40 Chlordiazepoxide . . .. 82 2 Diazepam . . . . .. 89 84 Nitrazepam , . .. 92 59 Oxazepam . . .. .. hledazepam .. .. 43 83 - - Identification of the 1,4-benzodiazepine drugs*0 was assisted by spraying the paper loaded with silica gel with an acidified solution of potassium iodoplatinate before examination under ultraviolet light at 245 and 350 nm.Although 5 to 20 pg of the drug were loaded on to the paper at its origin it was stated that 200 ng of diazepam could be detected by examination under ultraviolet light after exposure to hydrogen chloride fumes. Koechlin and D’Arcontel8 used paper chromatography to separate chlordiazepoxide from its metabolites. The behaviour of chlordiazepoxide when subjected to paper chromatography has been reported by Macek, Vecerkova and Stani~lova.~~ The solvent used was a mixture of light petroleum, ethanol and ammonia solution. A mixture of 60 + 2 + 38 gave an R, x 100 value of 85 and a mixture of 45 + 2 + 53 an RF x 100 value of 70. Paper loaded with cation-exchange resin has been investigated as an alternative to direct extraction of drugs in a screening programme for drug abuse and data were presented for chlordiazepoxide and one of its metabolite^.^^ Walkenstein, Wiser, Gudmundsen, Kimmel and Corradino used Whatman No.1 paper for the separation of 14C-oxazepam from its metabolites and with the developing solvent butanol- ethanol - water (17 + 3 + 20) found an R, x 100 value of 83 for o x a ~ e p a m . ~ ~ THIN-LAYER CHROMATOGRAPHY- Thin-layer chromatography of some 1,4-benzodiazepines has been discussed in a review paper by de Z e e ~ w . ~ ~ The applications of this technique are essentially three-fold : firstly, concerning the detection of impurities in pharmaceuticals, Beckstead and Smith in 196844 reviewed previous studies on the thin-layer chromatography of the 1,4-benzodiazepines and then discussed solvent systems for screening chlordiazepoxide, diazepam, oxazepam and nitrazepam for impurities ; and secondly, this technique is used for the toxicological analysis of hypnotics and psychotherapeutic substances in urine.Hermans and Kamp45 prepared the benzophenone derivatives of chlordiazepoxide, diazepam, oxazepam, nitrazepam and some of their more common metabolites from ether extracts of urine and ran their benzophenones on thin-layer chromatographic plates, using benzene for development as other solvents gave unsatisfactory separation (Table VI).May, 19741 AND THEIR METABOLITES I N BODY FLUIDS 247 TABLE VI THIN-LAYER CHROMATOGRAPHY ( R F X 100 VALUES) OF 1,4-BENZODIAZEPINES AS THEIR BENZOPHENONESg5 Solvent r 1 Ethyl acetate - n-hexane - 12 per cent.ammonia Chloroform - ether solution Benzophenone Benzene (3 + 1) (60 + 27 + 25) 2-Amino-5-chlorobenzophenone .. 39 92 97 2-Amino-5-nitrobenzophenone . . . . 20 87 93 2,5-Diaminobenzophenone . . . . 39 92 97 2-Methylamino-5-chlorobenzophenone 73 95 98 Lafargue, Meunier and L e m ~ n t e y ~ ~ combined thin-layer chromatography of benzo- phenone derivatives with thin-layer and gas - liquid chromatography and spectrofluorimetry of benzodiazepines in urgent toxicological analysis of these drugs. The mono-amino- benzophenone of nitrazepam has also been separated from the parent compound on a silica- gel thin-layer chromatographic plate, by using chloroform - acetone (9 + 1) as the developing solvent4'; the RF x 100 value was 90 and when using benzene was 30.The benzophenone was identified by diazotisation and then coupling the product with N-l-naphthylethylene- diamine dihydrochloride : nitrazepam did not undergo this reaction. Kamm and Baier48 detected chlordiazepoxide and diazepam in blood 24 hours after a single therapeutic dose by subjecting their benzophenones to thin-layer chromatography. Chlordiazepoxide was measured by colorimetry and diazepam by fluorescence. The third and most important use of thin-layer chromatography is the separation of the parent 1,4-benzodiazepines from their metabolites. Two-dimensional chromatography gave better resolution and has been used to study diazepam rnetab~lism~~ and that of flurazepam50 on isotopically labelled parent drugs, which facilitated more precise measurements than evaluations based on colorimetric reactions.Thin-layer chromatographic systems and methods of detection used for the 1,4-benzo- diazepines under discussion are presented in Table VII . TABLE VII THIN-LAYER CHROMATOGRAPHIC SYSTEMS AND METHODS OF DETECTION USED 1,4-Benzo- Reference diazepjne 51 Chlor- 52, 53 Chlor- diazepoxide diazepoxide 54 Chlor- diazepoxide 19 Chlor- diazepoxide 55 Chlor- 56 Chlor- diazepoxide diazepoxide 57 Chlor- 45 Chlor- diazepoxide diazepoxide 15 Chlor- diazepoxide FOR 1,4-BENZODIAZEPINES Thin-layer chromatographic R F system x 100 Method of detection* Benzene - ethanol - 25 per cent. Acetic acid - butanol - di-n-butyl ether (10 + 40 + 80) Ammonia solution - benzene - dioxan Acetone - methanol - ammonia 64 70 10 78 ammonia (50 + 10 + 5) (5 + 60 + 35) solution (50 + 50 + 1) Chloroform - ethanol - benzene - 62 ammonia solution (80 + 40 + 17.5 + 5) Benzene - acetone (4 + 1) 41 47 53 75 0 0 10 12 75 Chloroform - methanol (90 + 10) Acetone - 25 per cent.ammonia Chloroform - methanol (50 + 50) Benzene - acetone (95 + 5) Chloroform - ether (75 + 25) n-Heptane - chloroform - ethanol (99 + 1) Cyclohexane - triethylamine (9 + 1) Ethyl acetate - n-hexane - 12 per cent. ammonia (90 + 27 + 25) (20 + 20 + 1) Dragendorff reagent; KMnO, Iodoplatinate spray Iodoplatinate spray (i) Marquis reagent (ii) p-dimethylamino- benzaldehyde in ethanol - H,SO, Iodoplatinate reagent Dragendorff reagent KMnO, Dragendorfi reagent Iodoplatinate reagent HgINO, 2 per cent. pdimethylamino- benzaldehyde in 4~ HCI Ultraviolet light248 CLIFFORD AND SMYTH : DETERMINATION OF 1,4-BENZODIAZEPINES [Analyst, VOl.99 Reference 44 58 53 59 36 60 61 40 62 TABLE VII (continued) 1,4-Benzo- Thin-layer chromatographic RF diazepine system x 100 Method of detection* Chlor- Toluene - nitromethane - methanol 11 Dragendorff reagent Chloroform - methanol (10 + 1) 37 Cerium(1V) sulphate Chloroform - toluene - methanol 13 Dragendorff reagent diazepoxide (11 + 8 + 1) hydrochloride Benzene - nitromethane (30 + 1) (10 + 9 + 1'1 0 Chlorine - o-toluidine reagent Chlor- diazepoxide Chlor- Chlor- diazepoxide diazepoxide diazepoxide Chlor- Chlor- diazepoxide Chlor- diazepoxide Chlor- diazepoxide Chlor- diazepoxide 63, modified Diazepam 19 Diazepam in 53 49 Diazepam 64 Diazepam 65 Diazepam 44 Diazepam Mithanol - aceione (12 + 88) 48 96 per cent ethanol 48 Propan-2-01- di-n-propyl ether 29 Methanol - methyl acetate - cyclo- 57 Methanol - ammonia (sp.gr. 0.88) 67 19 35 3 5 Propan-2-01- chloroform - 25 per cent. 95 ammonia (45 + 45 + 10) Chloroform - methanol (95 + 5) 96 Chloroform - methanol (95 + 5) 15 Benzene 0 Methanol 56 Ethyl acetate - methanol - diethyl- 63 1,2-Dichloroethane - methanol - 48 hexane (18 + 49 + 33) Chloroform - acetone (9 + 1) Chloroform - acetone - pyridine Chloroform - ether (75 + 25) Chloroform - acetone (90 + 10) (100 + 1.5) (75 + 25 + 2.6) amine (70 + 30 + 1) ammonia (sp. gr. 0.88) Toluene - acetone - ammonia (sp. gr. 0.88) (50 + 50 + 1) Ethyl acetate - methanol - water - ammonia (sp. gr. 0.88) (90 1- 10 + 1) 55 68 (85 + 10 + 3 + 1) (99 + 1) Methanol - ammonia (sp.gr. 0.88) Chloroform - methanol - ammonia (sp. gr. 0.88) (90 + 10 +1) Chloroform - toluene - ethanol 82 75 20 (40 + 60 + 6) Ethyl acetate - methanol - ammonia (sp. gr. 0.88) (170 + 20 + 10) Methanol - ammonia (sp. gr. 0.88) (100 + 1.5) Chloroform - ethanol - benzene - ammonia (sp. gr. 0-88) (80 + 40 + 175 + 5) Chloroform - n-heptanol- ethanol n-Heptane - chloroform - acetic acid Chloroform - acetone (90 + 10) Chloroform - n-heptane - ethanol Chloroform - n-heptane - acetic acid - ethanol (5 + 5 + 1 + 0.3) Chloroform - acetone (90 + 10) Toluene - nitromethane - methanol Benzene - nitromethane (30 + 1) Chloroform - methanol (10 + 1) Chloroform - toluene - methanol (10 + 10 + 1) (5 + 5 + 1) (10 + 10 + 1) (11 + 8 + 1) (10 + 9 + 1) 81 75 88 60 50 84 54 63 62 26 0 83 37 Ultraviolet light a t 254 nm Acidified iodoplatinate spray Not stated Not stated Dragendorff reagent ; Ehrlich reagent 1 per cent.iodine - iodo- platinate 1. Ultraviolet a t 234 and 2. Ultraviolet after HC1 3. Exposure to ammonia Bleach, phenol, starch - iodide Dragendorff reagent 350 nm fumes vapour (2 t% m1-Y Dragendorff reagent 3H-diazepam Ultraviolet light HgINO, Dragendorff reagent Chlorine - o-toluidine reagent Cerium(1V) sulphate Dragendorff reagentMay, 19741 AND THEIR METABOLITES I N BODY FLUIDS 249 TABLE VII-continued Reference 45 59 36 66 60 61 40 62 1,CBenzo- diazepine Diazepam Diazepam Diazepam Diazepam Diazepam Diazepam Diazepam Diazepam 63, modified Nitrazepam in 53 28 47 45 20 44 36 Nitrazepam Nitrazepam Nitrazepam Nitrazepam Nitrazepam Nitrazepam Thin-layer chromatographic system Chloroform - diethyl ether (75 + 25) Ethyl acetate - hexanc - 12 per cent.ammonia (90 + 27 + 25) Chloroform - acetone (9 + 1) Chloroform - acetone - pyridine Chloroform - diethyl ether (75 + 25) Chloroform - acetone (90 + 10) Propan-2-01 - chloroform - 25 per cent ammonia (45 + 45 + 10) Chloroform - methanol (95 + 5) Chloroform - ethanol (29 + 1) Chloroform - acetone (9 + 1) Benzene Methanol Ethyl acetate - propan-1-01 - diethylamine (70 + 30 + 1) 1,2-Dichloroethane - methanol - ammonia (sp. gr. 0-88) Toluene - acetone - ammonia (sp. gr. 0.88) (50 + 50 + 1) Ethyl acetate - methanol - water - ammonia (sp. gr. 0.88) (85 + 10 + 3 + 1) (99 + 1) (75 + 25 + 2.6) (90 + 10 + 1) Methanol - ammonia (sp.gr. 0.88) Chloroform - methanol - ammonia (sp. gr. 0.88) (90 + 10 + 1) Chloroform - toluene - ethanol (40 + 60 + 60) Ethyl acetate - methanol - ammonia (sp. gr. 0.88) (170 + 20 + 10) Methanol - ammonia (sp. gr. 0.88) (100 + 1.5) Ethyl acetate - propan-1-01 - diethylamine (70 + 30 + 1) 1,2-Dichloroethane - methanol - 25 per cent. ammonia (90 + 10 + Toluene - acetone - 25 per cent. ammonia (50 + 50 + 1) Chloroform - acetone (9 + 1) Benzene Methanol Benzene - methanol (99 + 1) RF x 100 Method of detection* 58 2 per cent. p-dimethyl- 80 aminobenzaldehyde in 4 N HCI 76 Not stated 64 Methanol - acetone - triethanolamine Chloroform - diethyl ether (75 + 25) Ethyl acetate - n-hexane - 12 per cent. n-Heptane - chloroform - ethanol 54 Ultraviolet light Toluene - nitromethane - methanol 23 Dragendorff reagent Benzene - nitromethane (30 + 1) Chloroform - methanol (10 + 1) 39 Cerium(1V) sulphate Chloroform - toluene - methanol 10 Dragendorff reagent Chloroform - diethyl ether (75 + 25) 18 Not recorded Chloroform - acetone (90 + 10) Propan-2-01 - chloroform - 25 per cent.ammonia (45 + 45 + 10) Chloroform - methanol (95 + 5 ) 0 (100 + 100 + 0.03) 18 2 per cent. p-dimethyl- 43 aminobenzaldehyde in 4~ HCl ammonia (90 + 27 + 25) (20 + 20 + 1) (11 + 8 + 1) 0 Chlorine - o-toluidine reagent (10 + 9 + 1) 29 94 84 44 Not stated 54 97 96 80 75 5 59 90 Ehrlich reagent (20 pg) 82 Dragendorff reagent (0.5 pg) ; 80 92 Iodoplatiiiate reagent (1 pg per ml urine) 84 92 67 1. Ultraviolet at 234 and 2. Ultraviolet after HC1 3. Exposure to ammonia 350 nm (5 pg) fumes (0.2 pg) 71 Bleach, phenol, starch - iodide (2 pg ml-1) 74 Dragendorff reagent 91 Chlorination - o-toluidine Ehrlich reagent (0.1 pg) 1) 66 52 45 Dragendorff reagent 5 85 5250 CLIFFORD AND SMYTH DETERMINATION OF l,&BENZODIAZEPINES [ArtdySt, VOl.99 1,4-Benzo- Reference diazepine 66 Nitrazepam 67 Nitrazepam 60 Nitrazepam 40 Nitrazepam 45 44 59 67 60 40 62 68 69 70 TABLE VI I-con tinued Thin-layer chromatographic system Toluene - acetone - ammonia (sp. gr. 0.88) (50 + 50 + 1) Ethyl acetate - propan-l-ol- diethylamine (70 + 30 + 1) Methanol - 1,2-dichloroethane ammonia (sp. gr. 0.88) (10 + 90 + 1) Chloroform - ethanol (29 + 1) Chloroform - acetone (9 + 1) Benzene Methanol Ethyl acetate - propan-l-ol- Ethylene chloride - methane - diethylamine (70 + 30 + 1) ammonia (sp.gr. 0.88) (90 + 10 + 1) Toluene - acetone - ammonia (SP. gr. 0.88) (50 + 50 + 1) Chloroform - toluene - ethanol (40 + 60 + 2) RF 52 Bratton - Marshall reaction x 100 Method of detection* 70 54 32 Not recorded 45 Dragendorff reagent ( 5 pg) 0 Bratton - Marshall reaction 59 KI - o-toluidine (0.1 pg) 91 (0.05 CLg). 66 52 8 Oxazepam Chloroform - diethyl ether (75 + 25) 10 20 Ethyl acetate - n-hexane - 12 per cent. ammonia (90 + 27 + 25) Oxazepam Toluene - nitromethane - methanol Benzene - nitromethane (30 + 1) Chloroform - methanol (10 + 1) Chloroform - toluene - methanol Chloroform - acetone (9 + 1) Chloroform - acetone - pyridine Chloroform - ethanol (29 + 1) Chloroform - acetone (9 + 1) Benzene (11 + 8 + 1) (10 + 9 + 1) Oxazepam (75 + 25 + 2.6) Oxazepam Oxazepam 14 0 8 14 24 39 0 25 0 Methanol 54 Ethyl acetate - propan-l-ol- 80 diethylamine (70 + 30 + 1) (sp.gr. 0.88) (90 + 10 + 1) (sp. gr. 0.88) (50 + 50 + 1) Chloroethane - methanol - ammonia Toluene - acetone - ammonia 52 40 Oxazepam Chloroform - toluene - ethanol 0 (40 + 60 + 2) Oxazepam Ethyl acetate - methanol - ammonia 54 (sp. gr. 0.88) (170 + 20 + 10) Lorazepam Chloroform - ethanol - acetone 35 (8 + 1 + 1) Ethyl acetate - ethanol - ammonia 63 (sp. gr. 0.88) (5 + 5 + 1) Lorazepam Benzene 46 (as 2-amino- 2’,5-dichloro- benzophenone) Medazepam Chloroform - acetone (9 + 1) 60 1. Ultraviolet a t 234 and 2. Ultraviolet after HC1 3. Exposure to ammonia 2 per cent. P-dimethyl- aminobenzaldehyde in 4 N HCL 350 nm fumes Dragendorff reagent Chlorine - o-toluidine Cerium(1V) sulphate Dragendorff reagent Not recorded Not recorded Dragendorff reagent (10 pg) Bratton - Marshall reaction KI - o-toluidine (1 pg) (0.1 CLg) 1.Ultraviolet light a t 254 and 2. Iodoplatinate reagent Bleach, phenol, starch - 1. Ultraviolet light a t 2. Diazotisation and 350 nm iodide 254 nm Bratton - Marshall procedure Bratton - Marshall reaction Fluorescence at 366 nm, 0.1 pg sensitivityMay, 19741 AND THEIR METABOLITES IN BODY FLUIDS 251 TABLE VII (continued) 1,4-Benzo- Thin-layer chromatographic RF Reference diazepine system x 100 Method of detection* 40 Medazepam Chloroform - toluene - ethanol (40 + 60 + 2) 50 Flurazepam Ethyl acetate - ethanol - ammonia 33 Flurazepam (2) as benzo- phenone (iz) as acrida- none (sp.gr. 0-88) (90 + 10 + 0.3) Benzene - methanol - glacial acetic acid (90 + 10 + 10) Chloroform - acetone (85 + 15) 88 1. Ultraviolet at 254 and 2. Iodoplatinate reagent 350 nm 50 Dragendorff reagent 30 Ultraviolet light 15-28 Ultraviolet light System separates acridanones from the benzophenones from which they are formed * Values in parentheses are sensitivities. GAS - LIQUID CHROMATOGRAPHY- Chlordiazepoxide-De Silva, Schwartz, Stefanovic, Kaplan and D'Arconte22 reported that the method used for the gas - liquid chromatography of the major metabolite of diazepam after hydrolysis to its benzophenone (2-amino-5-chlorobenzophenone) could also be used to determine chlordiazepoxide as its hydrolysis product, which was identical with the benzo- phenone of N-desmethyldiazepam.Although intact chlordiazepoxide hydrochloride was found to have a retention time greater than five relative to glutethimide (Korzun, Brody, Keegan, Luders and Rehm57), McMartin and Street71 reported that chlordiazepoxide hydrochloride exhibited two peaks on 2 per cent. SE-30 at 245 O C , the first peak emerging at 3.6 minutes being considered a decomposition product. Foster and F r i n g ~ ~ ~ determined a retention time of 12 minutes for this compound on 3 per cent. SE-30 at a column temperature of 205 "C with a helium flow-rate of 25 ml min-1. Proelss and Lohmann17 determined the retention time of chlordiazepoxide hydrochloride on three different liquid phases and found a linear response between the peak height ratio and concentration over the range 0 to 25mg per 100ml.Although this ratio was small, the determination of chlordiazepoxide in a patient that received 100 mg in 24 hours gave a serum concentration of 0.1 1 mg per 100 ml. The two latter reports again identified chlordiazepoxide as a single peak. Van der Kleijn, Beelen and Frederick26 did not consider gas - liquid chromatography the method of choice for the assay of chlordiazepoxide as Sadhe and van der Kleijn72 had shown that chlordiazepoxide in addition to the N-4-oxides of diazepam were thermally labile. However, Zingales21 determined plasma concentrations of chlordiazepoxide in eleven subjects, who were receiving between 5 and 150 mg daily, by gas chromatography (oven at 275 "C) with electron-capture detection of the intact 1,4-benzodiazepine. The response of the detector to 5-ng amounts of chlordiazepoxide allowed the determination of plasma concentrations of the drug after a single 5-mg dose.Gardner-Thorpe, Parsonage, Smethurst and Toothill73 claimed a sensitivity of 5 ng with flame-ionisation detection of chlordiazepoxide after gas - liquid chromatography on a 1 per cent. DCMS liquid phase; column temperatures were 230 and 250 "C and again no thermolysis was reported. Diazepam-De Silva et aZ.22 described a method for the determination of diazepam and its N-demethylated metabolite in blood and plasma. After preliminary solvent extraction into diethyl ether, the diazepam was back-extracted into 6 N hydrochloric acid and hydro- lysed to its benzophenone derivative, 2-methylamino-5-chlorobenzophenone. It was found that the benzophenone derivative was five times more sensitive to tritium electron-capture detection than intact diazepam itself.However, benzophenone formation has not been found sufficiently selective for the determination of the 1,4-benzodiazepines because of interference from metabolites, especially if metabolic changes occur in the 3-position. In addition, different 1,4-benzodiazepines can give rise to the same benzophenone, e.g., 2-amino- 5-chlorobenzophenone can be formed by the acid hydrolysis of both the desmethyl metabolite of diazepam and also chlordiazepoxide. 2-Methylamino- and 2-amino-5-chlorobenzophenone were separated on a 2 per cent. Carbowax 20M column at 190 "C; similar operating conditions252 CLIFFORD AND SMYTH : DETERMINATION OF 1,4-BENZODIAZEPINES [Analyst, vol.99 were used by Cano, Vignoli and Viala.74 The separation of intact diazepam was improved by the introduction of high-temperature liquid phases, OV-1 and OV-17, both of which are methylsiloxane polymers. De Silva and P u g l i ~ i ~ ~ considered that the gas - liquid chromato- graphic analysis of intact benzodiazepines proposed by Marcucci, Fanelli and Mussini,34 who used flame-ionisation detection, was unsuitable for the determination of blood levels of the lJ4-benzodiazepines following therapeutic dosage. Improved resolution of these lJ4-benzo- diazepines occurred when the more polar liquid phase OV-17 was used; nickel-63 electron- capture detection provided highly sensitive (1 ng ml-l) determination^.^^ Blood levels of diazepam as low as 10 ng ml-l were measured.Lafargue, Pont and M e ~ n i e r ~ ~ reported the relative retention times of four therapeutically active 1,4-benzodiazepinesJ some of their metabolites and the benzophenone products of acid hydrolysis. Gas - liquid chromatography was performed on a 3 per cent. OV-17 column at 250 "C (210 "C for the benzophenones), and flame-ionisation detection was used (Table VIII). TABLE VIII SEPARATION OF SOME lJ4-BENZODIAZEPINES AND BENZOPHENONES ON 3 PER CENT. OV-17 WITH FLAME-IONISATION DETECTION75 Nitrogen flow-rate 25 ml min-l, column temperature 250 "C (210 "C for benzophenones) Retention time, minutes, relative to diazepam Compound (10-87 minutes) Oxazepam .. .. * . 0.70 Desmethyldiazepam . . .. 1.42 Chlordiazepoxide .. .. 1.44 Nitrazepam .. .. . . - 3.52 7-Aminonitrazepam . . .. 3.78 Diazepam . . .. . . .. 1.00 Retention time, minutes, relative to diazepam Compound (10.87 minutes) 7-Acetamidonitrazepam . . 11.40 2-Amino-5-chlorobenzophenone 0.33* 2-Methylamino-5-chloro- benzophenone .. .. 0*37* 2-Amino-5-nitrobenzophenone 1*43* Oxazepam . . .. .. 1.00* * Retention time relative to oxazepam (32.74 minutes). Foster and F r i n g ~ ~ ~ described a gas - liquid chromatographic method for the determina- tion of serum diazepam. A sensitivity of 0.5 mg per 10 ml was claimed by using a 3 per cent. SE-30 column and flame-ionisation detection. Proelss and Lohmann17 determined serum diazepam levels on three-column liquid phases; a sensitivity of 2.5 mg per 100 ml was reported with flame-ionisation detection (Table IX).TABLE IX CONDITIONS FOR GAS - LIQUID CHROMATOGRAPHY OF DIAZEPAM ON THREE LIQUID PHASES17 Column liquid phase . . . . . . 3 per cent. OV-17 3.5 per cent. XE-60 5 per cent. QF-1 Polarity of liquid phase . . . . Semi-polar Semi-polar Polar Retention ratio relative to pheno- Retention time of phenothiazinel Oven temperature/"C . . . . . . 235 235 210 (programmed to 240) thiazine . . .. . . . . 3.54 2-76 3.86 minutes . . .. .. .. 4.9 4.26 6-6 Forgione, Martelli , Marcucci , Fanelli, Mussini and J0m11-i~~ reported efficient separation of diazepam and oxazepam on both 3 per cent. OV-17 and 3 per cent. OV-1 (Table X). Van der Kleijn et al.26 reported that the sensitivity of gas-liquid chromatography to tranquillisers, which included diazepam and oxazepam, was in the range 20 to 50ng by using flame-ionisation detection with a 10 or 100-fold increase in sensitivity occurring with tritium or nickel-63 electron-capture detection, depending on the number of electron- capturing moieties in the molecule.Gardner-Thorpe et ~ 1 . ' ~ separated diazepam, oxazepam and chlordiazepoxide on 1 per cent. cyclohexane dimethanol succinate in dichloromethane and with flame-ionisation detection a sensitivity of 1 ng was claimed (5 ng for chlordiaz- epoxide), although it was not stated whether or not this amount could be detected in blood samples.May, 19741 AND THEIR METABOLITES I N BODY FLUIDS 253 Gas-chromatographic determinations of blood levels of diazepam have provided blood or plasma concentrations that are more acceptable parameters in defining pharmacological effect than dosage alone.The pharmacokinetic profile of diazepam and its major metabolites was studied by de Silva et. nZ.22 by using gas chromatography of their respective benzo- phenones. Maximum diazepam blood levels of 0.18 to 0.22 pg ml-l were reported after a single 10-mg oral dose: the clearance of diazepam from the blood was found to be biphasic, a fast phase with a half-life of 2 to 3 hours was followed by a slow phase with a half-life of 27 to 28 hours. Can0 et. aZ.74 found blood levels of diazepam in the region of 0.5 pg ml-l after an oral dose of 30 mg and 24 hours later 30 per cent. of this maximum level was still detectable. Van der Kleijn, van Rossum, Muskens and R i j n t j e ~ , ~ ~ drew attention to the erratic course of diazepam blood concentrations and suggested that a triphasic concentration - time relationship occurred: only the biological half-life of the third phase could be estimated in man and this was found to be in the range 24 to 42 hours.TABLE X RELATIVE RETENTION TIMES OF THREE BENZODIAZEPINES ON A NON-POLAR OV-1 AND A SEMI-POLAR OV-17 LIQUID PHASE, AND THEIR SENSITIVITY TO FLAME-IONISATION DETECTION76 Relative retention time Drug 3 per cent. OV-1 3 per cent. OV-17 Sensitivitylpg Oxazepam . . .. 1 Diazepam . , .. 1-56 Nitrazepam . . .. 2.83 1 1.39 3.27 0.25 0.20 0.50 Oxazepam-Sad6e and van der Kleij n72 investigated thermolysis of selected 1,4-benzo- diazepines during gas chromatography. Oxazepam is more polar than diazepam and desmethyldiazepam, yet it has been found to have a shorter retention time on OV-17 (de Silva and P ~ g l i s i ~ ~ ) ; this effect was considered due to rearrangements of the diazepine ring, which occurred at the higher temperatures used for gas - liquid chromatography of intact benzodiazepines.Comparison of the mass spectrum of this reaction product of oxazepam after gas chromatography with that of quinazolinecarbaldehyde after direct-insertion mass spectrometry and the finding that an authentic sample of quinazolinecarbaldehyde gave a retention time identical with that of oxazepam under the same gas - liquid chromatographic conditions suggested that a thermolytic rearrangement, with the loss of water, had occurred. Fig. 1 illustrates the proposed scheme. Oxazepam Quinazolinecarbaldehyde Thermolytic rearrangement of 0 ~ a z e p a .m ~ ~ Fig. 1. Forgione et ~ 1 . ~ ~ also reported that the mass spectra of eluted gas-chromatographic peaks show that diazepam, desmethyldiazepam, methyloxazepam and nitrazepam were un- modified under the analytical conditions, but that oxazepam was rearranged to form 6- chloro-4-phenylquinazoline-2-carbaldehyde with the loss of a water molecule as a result of the heating. Recently, oxazepam has been determined in serum and urine as its hydrolysis product, 2-amino-5-chlorobenzophenone78 ; the thermal rearrangement product did not interfere and sensitivity down to 0.8 ng ml-l was reported. Gas - liquid chromatography of the 2-amino-5-chlorobenzophenone with electron-capture detection has been used to study the clearance of oxazepam from human plasma.A 2 per cent. XE-60 column was used at 240 0C.79254 CLIFFORD AND SMYTH : DETERMINATION OF 1,4-BENZODIAZEPINES [Analyst, VOl. 99 NitrazePam-Marcucci et aZ.34 described the gas - liquid chromatographic separation of intact nitrazepam from four other benzodiazepines on a 3 per cent. OV-1 column at a temperature of 245 "C, using flame-ionisation detection. Nitrazepam had the longest retention time (9 minutes) and a sensitivity range between 0-1 and 2.0 pg was reported for these compounds. Van der Kleijn et aZ.26 separated nitrazepam from other benzodiazepines on a 3 per cent. OV-17 column at an oven temperature of 250 "C by using flame-ionisation detectors, with the detector block at 280 "C. However, Gardner-Thorpe et aZ.73 could not detect nitrazepam after gas - liquid chromatography on any of the columns that they used. Beharrell, Hailey and McLaurin31 described the development of a sensitive gas-chromato- graphic method for the determination of nitrazepam in plasma, using clonazepam as the internal standard.The two benzodiazepines were extracted carefully into diethyl ether before hydrolysis to their respective benzophenones. Electron-capture detection was used to determine the eluted benzophenones. The two major metabolites of nitrazepam, the 7-amino and 7-acetamido derivatives, yield 2,5-diaminobenzophenone as the major product on hydrolysis and as this product is only weakly electron capturing the method is considered specific for nitrazepam. The metabolite, 1,3-dihydr0-3-hydroxy-7-nit ro-5-p henyl-2H- 1,4- benzodiazepin-2-one was reported as being normally not detectable in plasma. The sensi- tivity limit for the 2-amino-5-nitrobenzophenone was stated to be 0.1 ng per millilitre of plasma; gas - liquid chromatography of the intact nitrazepam was limited to 1 pg ml-l concentration and at lower concentrations the peak shape deteriorated and variable recoveries for intact nitrazepam and clonazepam were reported.Masudas0 described a somewhat similar determination of nitrazepam in biological materials by using gas - liquid chromatography. Acid hydrolysis of nitrazepam to the 2-amino-5-nitrobenzophenone was again performed but with flame-ionisation detection ; Masuda found a sensitivity of only 5 to 1Opg. The data obtained also included three important variants in this method: the effect of pH on the extraction of nitrazepam, the effect of hydrochloric acid concentration on the hydrolysis of nitrazepam and the optimum time required for this hydrolysis.Lorazepam-Knowles, Comer and R u e l i u ~ ~ ~ determined both free and conjugated lorazepam in human serum : lorazepam was determined as 2-amino-2',5-dichlorobenzophenone on 3 per cent. OV-17 at an oven temperature of 280 "C with the nickel-63 electron-capture detector at 320 "C. In human subjects the maximum serum level of lorazepam occurred between 1 and 6 hours after oral administration and both free and conjugated forms were present in the serum 24 hours later. The sensitivity of this method was reported to be 0-01 pg per millilitre of specimen.Marcucci, Mussini, Arioldi, Guitani and Garattinisl measured brain levels of lorazepam in mice. They found that during gas - liquid chromatography, intact lorazepam lost a water molecule and was rearranged to form 6-chloro-4-(2'-chlorophenylquinazoline)-2-carbaldehyde. This reaction product was checked by mass spectrometry in the manner developed for the study on the dehydration mechanism of oxa~epam.'~ MedazePam-De Silva and PuglisiZ4 determined medazepam in blood and urine by using a nickel-63 electron-capture detector. The retention times on a 3 per cent. OV-17 column at an oven temperature of 230 "C and the melting-points of medazepam and some of its metabolites are shown in Table XI. TABLE XI GAS - LIQUID CHROMATOGRAPHY OF MEDAZEPAM AND ITS METABOLITES ON Operating conditions: carrier gas, argon - methane (90 + 10) ; column temperature 230 "C; injection port at 270 "C; and detector at 310 "C Retention time/ 3 PER CENT.OV-17 Compound Melting-point/"C minutes Medazepam . . .. .. 95-97 4.2 N-Desmethylmedazepam . . 170-1 7 1 6.0 Oxazepam . . . . .. 205-206 6.0 Diazepam . . .. .. .. 131-135 8.7 N-Desmethyldiazepam . . .. 2 1 6-2 1 7 12-6 3-Hydroxydiazepam . . .. 118-120 21.8May, 19741 AND THEIR METABOLITES I N BODY FLUIDS 255 Oxazepam with a melting-point of 205 to 206 "C occupies an anomalous position with a retention time of only 6 minutes, which is due to thermolysis to the carbaldehyde derivative. The 3-hydroxydiazepam has a longer retention time than its melting-point would suggest and it also gave a poor response to the electron-capture detector; the formation of its tri- methylsilyl derivative resulted in a shortened retention time to 12.8 minutes and a sharper peak on the gas chromatogram.The oxazepam trimethylsilyl derivative retained the same retention time but showed some trailing of the peak, which was thought to be due to the increased polarity of the -NH group in the 1,4-benzodiazepine ring. The sensitivity limits of medazepam and N-desmethyldiazepam were 0.05 pg per milli- litre of blood, and those of diazepam and N-desmethyldiazepam were 0.02 pg per millilitre of blood. A maximum blood level of 1 pg ml-l occurred after the ingestion of a 50-mg tablet of medazepam and fell to 50 per cent. of this level in 3 hours. LIQUID CHROMATOGRAPHY- Scott and Bommer82 used liquid - solid chromatography to separate diazepam, oxazepam and ten other benzodiazepines.A 100 cm x 1 mm i.d. stainless-steel column dry-packed with Durapak OPN particles of 36 to 75pm diameter effected satisfactory separation and the eluate was scanned by an ultraviolet detector, which gave a sensitivity at the microgram level. Diazepam and oxazepam have also been separated and determined on a column containing Durapak OPN-Poracil C with tetrahydrofuran - diisopropyl ether (3 + 17) as the mobile phase. ULTRAVIOLET AND VISIBLE SPECTROPHOTOMETRY The 1,4-benzodiazepines and their metabolites have extinction coefficients of the order of lo4 1 mol-l cm-l, which limits the usefulness of the technique to the microgram per millilitre level with conventional spectrophotometers.In Table XII, the wavelengths of maximum absorption (Ama.) and the corresponding extinction coefficients (e) for those forms of the benzodiazepines which absorb in the pH range 0 to 14 are listed. The sensitivity of this method was 10 pg.s3 TABLE XI1 COMPARISON OF DIFFERENT FORMS OF 1,4-BENZODIAZEPINES (ELECTRONIC SPECTRA)' H,A+ HA A- 7- * Compound hmax.* .( x 104) hma,. €(x 104) Chlordiazepoxide Diazepam . . Oxazepam . . Nitrazepam . . Lorazepam . . Medazepam . . Flurazepam* . . .. 245 2.6 310 0.6 .. 241 2.8 286 1.3 . . 239 3.6 288 1-6 .. 217s 2.0 282 2-6 .. 240 0.8 292 3.2 . . 255 3-2 290 s 1.4 .. 242 2.2 285 1.1 250 s 2-4 - 260 2.5 - 310 0.4 - 23 1 3.3 253 s 1.7 - 23 1 4.2 236 255 s 1.8 260 s 280 0.8 310 s 0.4 217 2.6 228 260 1.7 260 s 313 s 1.2 370 215 s 2.8 232 23 1 3-6 278 s 259 s 1.2 233 2-9 252 s 2.5 231 2.7 250 s 1.4 325 s 1.2 - - - - - - 2.9 1.9 2-6 1-5 1.4 2.8 0.8 * s =: shoulder.Tables I11 and XI1 neglect to state that the diethylamino function in flurazepam will be positively Hence flurazepam has a second pK, value that is not measurable by spectro- In Table XII, the two columns refer to H,A2+ and H,A+ in the case of Lafargue et aL"3 reported spectral studies on the benzophenones of chlordiazepoxide, diazepam and nitrazepam in 2 N hydrochloric acid, 0.2 N sodium hydroxide solution and charged u p to pH 11. photometry as in Table 111. flurazepam.256 CLIFFORD AND SMYTH DETERMINATION OF 1,4-BENZODIAZEPINES [Analyst, VOl. 99 methanol. The benzophenones of both chlordiazepoxide (2-amino-5-chlorobenzophenone) and diazepam (2-methylamino-5-chlorobenzophenone) absorbed at 213, 236 and 237 nm, respectively, in the three solvents and their extinction coefficients were approximately 20 per cent.higher than those of the parent compounds. 2-Amino-5-nitrobenzophenone showed principal absorption bands at 367, 365 and 357 nm with extinction coefficients of the order of 50 per cent. higher than nitrazepam. These figures indicate an increased sensitivity of the spectrophotometric method after extraction of the benzophenone from a sample that had been subjected to acid hydrolysis. The ultraviolet spectra of these benzophenones have also been reported by A. Heyndrickx and A. de Leenher (private communication). 2-Amino- 5-chlorobenzophenone, the product of the acid hydrolysis of chlordiazepoxide, gave A,.at 214 and 255 nm; acid hydrolysis of oxazepam formed the same benzophenone as that obtained with chlordiazepoxide. 2-Methylamino-5-chlorobenzophenone, the acid hydrolysis product of diazepam, exhibited A,,,. at 214 and 248 nm and 2-amino-5-nitrobenzophenone, the acid hydrolysis product of nitrazepam, showed A,. at 235 and 370 nm. All these spectra were recorded in 4 N hydrochloric acid. Beyer and Sadbe66 have synthesised four possible metabolites of nitrazepam that could occur in humans and characterised them by their ultraviolet spectra, which were stated to be sufficiently different to permit differentiation of these metabolites in a mixture of them. The absorption maxima of the neutral form (HA) occur at 350 and 242 nm for the 7-amino metabolite, at 327 and 246 nm for the 7-acetamido metabolite and at 305 and 260 nm for the 3-hydroxynitrazepam metabolite.De Silva, Koechlin and Bader64 have used ultraviolet spectrophotometry to determine diazepam and its N-desmethyl metabolite simultaneously in urine after extraction of the neutral forms at pH 7.1 into diethyl ether. De Silva and Str0j1-1~~~ reported the spectro- photometric determination of flurazepam and three of its urinary metabolites as their benzo- phenones. In methanol they all showed a major band with A, between 236 and 234 nm, and a minor band with A,. at 392 nm for 2-amino-5-chloro-2’-fluorobenzophenone, at 405 nm for 5-chloro-2-(2-diethylaminoethylamino)-2’-fluorobenzophenone hydrochloride, at 41 2 nm for 5-chloro-2’-fluoro-2-(2-hydroxyethyl) aminobenzophenone and at 408 nm for N-[4-chloro-2- (2-fluorobenzoyl)]glycine. The structural formulae of flurazepam metabolites are shown in the metabolites section on p.269. SPECTROPHOTOFLUORIMETRY Fluorimetry is by far the most sensitive optical method used in the determination of the 1,4-benzodiazepines and their metabolites, with a sensitivity comparable with gas - liquid chromatography and polarography under optimum conditions. Koechlin and D’Arcontel8 have described a sensitive procedure for the determination of both chlordiazepoxide and its lactam metabolite (V) in plasma, based on the quantitative conversion of chlordiazepoxide into the lactam under controlled hydrolysis conditions followed by rearrangement of the N-oxide grouping in alkaline solution under the influence of light to a fluorescent derivative.The procedure is specific for chlordiazepoxide, with a sensitivity limit of 0-25 pg per millilitre of plasma; it is also reported that the procedure can be modified so as to permit an equally selective determination of the lactam metabolite. For this procedure, the excitation wave- length is 380nm and the fluorescent emission occurs at 480nm. The fluorescence of such an irradiated solution disappears instantaneously on acidification and reappears immediately on restoring alkaline conditions. Schwartz and Postma84 developed a fluorimetric method for the determination of an N-demethylated metabolite of chlordiazepoxide and slightly modified the procedure of Koechlin and D’ Arcontels so as to make possible the fluorimetric determination of each compound when all three are present in a single blood sample.Braun, Caille and Mockles5 have reported that alcoholic solutions of chlordiazepoxide, diazepam, nitrazepam and oxazepam develop a very intense fluorescence in 1 N sulphuric, phosphoric and perchloric acids. The fluorescence is peculiar to each compound, the limit of sensitivity being 5 ng ml-I for oxazepam to 100 ng ml-1 for the other three 1,4-benzodiazepines. Walkenstein et ~ 1 . ~ measured 0.5 pg of oxazepam per millilitre of urine or serum by fluorimetry. De Silva and Strojny33 determined fluorazepam and its major metabolites in blood andMay, 19741 AND THEIR METABOLITES IN BODY FLUIDS 257 urine by using the intense fluorescence of the 9-acridanone derivatives (VI), which are pro- duced by cyclisation of the appropriate benzophenones in dimethylformamide with potassium carbonate as catalyst.A sensitivity of 5 to 10 ng of the acridanone derivative per millilitre of solvent is claimed. R 0 VI Reider86 has recently described a fluorimetric method for the determination of nitrazepam in human plasma. Nitrazepam and its two main metabolites were extracted into dichloro- methane-ethyl acetate (2 + 1) after adjusting the pH of the deproteinised plasma to 10 with magnesium oxide. The metabolites were then extracted into 0.15 N hydrochloric acid and the parent drug into 3 N hydrochloric acid. In both instances the acidity of the acidic phase was then increased to give a concentration of 6 N hydrochloric acid.The re-extracted nitrazepam was then reduced to the 7-amino compound prior to heating with a methanolic solution of o-phthalaldehyde. Firstly, the aminobenzophenone was formed, then the amino group reacted with o-phthalaldehyde to form a fluorescent species. With the metabolites the hydrolysis step produced the 2,5-diaminobenzophenone. A sensitivity in the sub-nanogram range permitted the determination of a half-life of 27 hours for nitrazepam in the plasma after a single 10-mg oral dose.87 POLAROGRAPHY Although d.c. polarographic reduction waves of the 1,4-benzodiazepines have been extensively studied mechanistically and for the assay of formulations, there are fewer recorded applications of sensitive polarographic measurements for the determination of drugs in biological fluids after either t h e r a p e ~ t i c ~ ~ or toxicological d o ~ a g e .~ ~ ~ ~ ~ Direct-current polarographic methods have yielded linear limiting current - concentration relationships in concentration ranges that are suitable only in forensic investigations for determinations in body materials after overdosage. Oelschlager, Volke and Kureksg showed diazepam to have a linear relationship down to 10-5 M and claimed a sensitivity of 11.4 pg per 10 ml of buffer at pH 3.7. Cimbura and Guptalg used a d.c. polarographic procedure for the routine toxicological analysis for diazepam and chlordiazepoxide and reported a sensitivity of 1-0 pgml-l in a 10-ml blood sample for both benzodiazepines. They also carried out polarographic measurements on the benzophenones after the samples had been subjected to solvent extraction followed by acid hydrolysis.Caille, Braun and Mocklego found that spectrophotofluorimetry was superior to polarography in the determination of oxazepam, whereas with diazepam and chlordiazepoxide both methods were comparable in precision and selectivity; however, these analyses were directed to pharmaceutical preparations. Fazzari and Rigglemangl detected 12.5 pg ml-l of oxazepam in acetate buffer. Oelschlager, Volke, Lim and Spangg2 reported that oxazepam was reduced at the dropping- mercury electrode in a single wave, the wave height being dependent on the pH of the supporting electrolyte; in acetate buffer solutions they found a linear dependence of wave height on the concentration of the depolariser and the polarographic determination of oxazepam in a pharmaceutical formulation was stated to be as accurate as spectrophotometric determination.Recently, Volke, Oelschlager and Limg3 observed an anomalous dependence of the limiting current of oxazepam on potential when working in buffers of pH greater than 7.0 and attributed this effect to adsorption of the reducible species on the dropping-mercury electrode up to the potential of its reduction. Jacobsen and Jacobseng4 detected 13 pg ml-1 of chlordiazepoxide contained in a mixture of equal amounts of serum and 0.1 M sulphuric acid. Evidence was presented showing that both chlordiazepoxide and its reduction products are adsorbed at the electrode surface, the258 CLIFFORD AND SMYTH DETERMINATION OF 1,4-BENZODIAZEPINES [AndySt, VOl.99 former more strongly than the serum proteins, thus making the determination possible without prior separation. Halvorsen and Jacobseng5 have utilised the preferential adsorbability of nitrazepam over serum proteins in order to determine this 1,4-benzodiazepine in the range 0.5 to 8-0 pg of nitrazepam per millilitre of serum. Fidelus, Zietek, Mikolajek and Groch owskass reported the polarographic determination of diazepam in cadaver blood without prior solvent extraction, whereas Jacobsen, Jacobsen and Rojahng6 have extracted diazepam from a mixture of 5 ml of serum and 30 ml of benzene and determined it in the range 0.05 to 1.0 pg per rnillilitre of serum. Berry25 has reported that after therapeutic dosage of diazepam, plasma levels as low as 0.02 pg ml-1 can be detected by cathode-ray polarography, following solvent extraction into light petroleum from human plasma that was made alkaline with 1 N sodium hydroxide solution.Clifford and Smythg7 have studied the polarographic behaviour of six 1,4-benzo- d jazepines by d.c., cathode-ray and pulse-polarographic techniques and suggested the opti- mum pH ranges for analytical purposes. These data and the appropriate reducible moieties are presented in Table XIII. The limiting current - concentration relationship for nitrazepam uas constructed by using cathode-ray and differential pulse-polarographic methods at pH 5.15 and linearity was found over the range lo-* to lo-’ M for the first reduction wave, corresponding to -NO,+ -NHOH, which is equivalent to quantitative determination down to 20 ng ml-l with a detection limit of 5 to 10 ng r n F .Oxazepam gave a linear relationship TABLE XI11 OPTIMUM PH RANGES FOR POLAROGRAPHIC MEASUREMENTS ON THE 1,4-BENZODIAZEPINES* AND THE CORRESPONDING REDUCIBLE MOIETIES Compound Chlordiazepoxide . . Diazepam . . .. Oxazepam and lorazepam . Nitrazepam . . .. Medazepam , . .. Flurazepama .. pH range 3-7 (all three waves) 3-1 1 3-6 3-1 1 (first wave) 3-6 (second wave) 3-7 3-1 1 Reducible moiety \ / “+OH+, C=N // / H+ \ / / / H+ \ C=N , -N=C \ / \/ C=N , OH,+ / H + / \ -NO, -NHO+H, \ / / H+ \ / / H+ \ / / H+ C=N C=N C=N * Those pH ranges in which the 1,Cbenzodiazepines give well defined waves suitable for analytical measurements on pharmaceutical formulations or in body fluids and excluding pH<3 and pH> 11, when unwanted decomposition occurs.May, 19741 AND THEIR METABOLITES IN BODY FLUIDS 259 in the range to loA6 M at pH 3.0, which has been used for the assay of oxazepam and lorazepam in pharmaceutical preparation~.~8 A summary of the polarographic reduction waves of 1,4-benzodiazepines is presented in Table XIV.TABLE XIV POLAROGRAPHIC REDUCTION WAVES FOR SOME 1,4-BENZODIAZEPINES* Compound Chlordiazepoxide Diazepam . . Oxazepam . . Nitrazepam . . Medazepam . . Flurazepam . . - .. . . .. .. .. .. E+/V versus S.C.E. 0.38, 0.72, 1.20 0.27, 0.64, 1.15t 0.36, 0.67 0-22, 0*57t 0.68 1,ost 1.09 0.677 1-06f 0.70 0.60t 0.61t 1.02t 0.70 0.96 1.08 1-19 0.12, 0.71t 0.72, 1.267 0.70, 1-10! 1.00, 1-50:: 0-38, 1.027 1.00 0-75 Supporting electrolyte Britton - Robinson buffer, pH 2.7 0.1 N Hydrochloric acid - methanol (4 + 1) 0.05 N Ammonium chloride - methanol (4 + 1) 1 N Hydrochloric acid 0.1 M Sulphuric acid Britton - Robinson buffer (pH 3-0) - Britton - Robinson buffer (pH 9.0) - 0.1 N Hydrochloric acid - methanol (4 + 1) 0.05 N Ammonium chloride - methanol (4 + 1) 1 N Hydrochloric acid 0.1 M Sulphuric acid 0.1 M Sulphuric acid 0-2 M Acetate buffer - 10 per cent.dichloromethane Britton - Robinson buffer (pH 2.3) - dimethylformamide (4 + 1) Britton - Robinson buffer (pH 5.1) - dimethylformamide (4 + 1) Britton - Robinson buffer (pH 7.2) - dimethylformamide (4 + 1) Britton - Robinson buffer (pH 9.2) - dimethylformamide (4 + 1) 0.1 N Hydrochloric acid - methanol (4 + 1) 0.05 N Ammonium chloride - methanol (4 + 1) Britton - Robinson buffer, pH 4.4 Britton - Robinson buffer, pH 8.4 Phosphate buffer, pH 6.9 Britton - Robinson buffer (pH 6.0) - dimethylformamide (9 + 1) Britton - Robinson buffer, pH 4 dimethylformamide (95 + 5) dimethylformamide (95 + 5) (6 + 14 V / V ) Refer- ence 11 10 10 19 94 89 89 10 10 19 99 96 91 92 92 92 92 10 10 11 11 95 100 8 * Measured a t concentrations lo-' to 5x t Ei measured against AglAgC1 reference electrode.! These potentials refer to limiting current potentials. M. MASS SPECTROMETRY Mass spectrometry has been used primarily for the auxiliary detection of 1,4-benzo- diazepines and their metabolites following thin-layer and gas chromatography. High- resolution mass spectrometry has also been used solely for characterisation purposes.The mass spectra of chlordiazepoxide and diazepam obtained by using a CEC 21-110 mass spectrometer with an ionising voltage of 70 eV at a temperature of 190 "C have been included in their analytical profile.6 Thin-layer chromatography for the isolation and purification of metabolites of 3H- diazepam101 and 14C-chlordiazepoxide102 administered to the rat was followed by high- resolution mass spectrometry for characterisation. Schwartz and co-workers extended this technique in order to identify one metabolite of unlabelled flurazepam in human urine and five metabolites in dogs' urine.50 A similar technique was used to determine the urinary metabolites of lorazepam in the human, miniature pig, rat, cat and dog.68 The identification of eluted benzodiazepine peaks has been established by using mass spectrometry in combination with gas - liquid ~hromatography'~ : these workers also investi- gated the thermolysis of oxazepam (see above) and found that the structures of diazepam, N-desmethyldiazepam, N-methyloxazepam and nitrazepam were not modified by the experimental conditions.The thermolysis of 1,4-benzodiazepines during gas chromato- graphy - mass spectrometry had been reported by Sadbe and van der Kleijn72 following the260 CLIFFORD AND SMYTH DETERMINATION OF 1,4-BENZODIAZfiPINES [A??dySt, VOl. 99 studies by Sad6elo3 on the fragmentation pattern of the 1,4-benzodiazepin-2-ones under electron impact. In this work, diazepam, nitrazepam and oxazepam were found to undergo several rearrangements, which were considered analogous to reported chemical reactions in smaller ring systems.Degradation by loss of H, HCN, CO and HCO occurred under electron impact and the resulting ring systems observed were quinazolines, indolones and indazoles : by further rearrangement, condensed ring systems were generated. All spectra showed characteristic peaks at values of m/e of 205, 193, 177, 165 and 151. Milne, Fales and Axenrodlo* reported the identification of chlordiazepoxide and diazepam by chemical ionisation mass spectrometry with isobutane after a simple extraction of serum or gastric contents with chloroform. The solvent was evaporated off and the extract inserted on the direct probe of the mass spectrometer. This technique dispenses with the use of gas - liquid chromatographic separations, for in chemical ionisation mass spectrometry the collision product is an MHf ion of m/e (M + 1) and the molecular ion was very stable com- pared with the odd-numbered electron molecular ion derived from the molecule by electron impact.NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY Various workers have studied the confirmations of the diazepine ring by nuclear magnetic resonance spectro~copy.6~~05-~~~ Nuhn and Bleylo5 favoured the chair form whereas Sad6elo8 suggested an intermediate state between the pseudo-boat form of cycloheptadiene and the boat form of cycloheptatriene. Nuclear magnetic resonance spectroscopy had to be used in order positively to locate the position of the hydroxyl group on the 5-C-phenolic metabolites of diazepam after its administration to the rat.lo1 INFRARED SPECTROSCOPY Data on absorption bands of some 1,4-benzodiazepines have been collected by a number of ~ o r k e r s .6 ~ ~ ~ J ~ ~ The infrared absorption of oxazepam and nitrazepam has been described in more detail.llOJ11 Pribilla47 has used infrared spectroscopy to identify the benzophenone as the inter- mediate in the Bratton - Marshall reaction. Diazepam has been determined in the concentra- tion range 50 to 250 pg per 5 ml of urine by infrared techniques112 and Masudaso used infrared spectroscopy to examine the structure of the nitrazepam benzophenone prior to his gas- liquid chromatographic studies. COMPARISON OF ANALYTICAL METHODS Thin-layer and gas - liquid chromatography are the most widely applied chromatographic techniques for the detection and determination of 1,4-benzodiazepines in body fluids.The methods are usually based on the formation of the benzophenones prior to chromatography and therefore cannot differentiate those benzodiazepines which give the same benzophenone on acid hydrolysis, e.g., chlordiazepoxide, oxazepam and desmethyldiazepam all hydrolyse to 2-amino-5-chlorobenzophenone. Appropriately chosen thin-layer systems give adequate separation of the 1,4-benzodiazepines and their benzophenones and are best applied to the detection of 1,4-benzodiazepines in toxicological investigations. By using isotopically labelled parent drugs this method is also of use in studying metabolism. Sensitive determinations of the 1,4-benzodiazepines can be accomplished by gas - liquid chromatography, particularly if the benzophenone and electron-capture detection (e.g., nickel-63) are used.Columns containing a low loading (about 3 per cent.) of OV-1 or OV-17 are commonly used and sensitivities ranging from 1 to 50 ng per millilitre of body fluid are quoted in the literature. Oxazepam and chlordiazepoxide are thermally labile and intact nitrazepam and flurazepam do not lend themselves to gas - liquid chromatographic assay. Liquid - solid and liquid - liquid chromatography have so far been used for the resolution of mixtures of benzodiazepines at the microgram level and are therefore most applicable to formulation and toxicological analyses. The ultraviolet spectral bands of the 1,4-benzodiazepines have extinction coefficients of the order of lo4 to 3 x lo4 1 mol-l cm-l, the benzophenones of chlordiazepoxide, oxazepam and diazepam having values 20 per cent.higher than the parent compounds. The benzo- phenone of nitrazepam is 50 per cent. higher, resulting in analytical methods with a sensi-May, 19741 AND THEIR METABOLITES I N BODY FLUIDS 261 tivity of the order of micrograms per millilitre. The design of selective methods based on the ultraviolet absorption of the 1,4-benzodiazepines and their benzophenones is only possible with nitrazepam, flurazepam and their metabolites when adequate separation in the Am*. values occurs. The classic Bratton - Marshall reaction, which is only applicable to 1,4- benzodiazepines with hydrogen substituted in the 1-position, has been used for the assay of chlordiazepoxide, oxazepam and nitrazepam with a sensitivity of 2 to 20 pg per millilitre of body fluid.Spectrofluorimetric methods have the required sensitivity to follow blood levels of 1,4-benzodiazepines after therapeutic dosage but usually involve lengthy multi-step con- versions into fluorescent derivatives, e.g., in the assays of chlordiazepoxide, nitrazepam and flurazepam. Polarographic methods (those involving differential pulse, cathode-ray and a. c. polaro- graphy) under optimum conditions possess a sensitivity in the range 10 to 100 ng per milli- litre of body fluid, depending on the number of electrons transferred in the reduction process and the amount of interfering material present in the sample which is co-extracted from the body fluid.Preferential adsorbability of 1,4-benzodiazepines on the dropping-mercury electrode is an additional factor and allows the direct determination of chlordiazepoxide and nitrazepam “spiked” in serum after dilution with supporting electrolyte at the microgram per millilitre level. The selectivity of the polarographic method can be exploited with chlordiazepoxide, nitrazepam and their metabolites. Gas - liquid chromatography is therefore recommended for the assay of intact diazepam and medazepam, and of oxazepam, lorazepam and nitrazepam if the benzophenones are used. Spectrofluorimetry is the method of choice for chlordiazepoxide and polarography for intact nitrazepam and flurazepam. MISCELLANEOUS METHODS Although gravimetric and titrimetric113 as well as p~tentiometricll~ and colorimetric methods have been used for the assay of 1,4-benzodiazepines in pharmaceutical formulations, only the colorimetric methods are of use in the measurement of levels of 1,4-benzodiazepines in body fluids.The classic Bratton - Marshall procedure115 has been applied to the deter- mination of chlordiazepoxide,lJ16 which was subjected to hydrolytic cleavage to an aromatic amine after its extraction into ether. The amine was diazotised and coupled with N- 1-naphthylethylenediamine to form an azo dye, which was determined colorimetrically at 550nm. The method has also been applied to the determination of oxazepam and nitrazepam.20,2*~30,47 Sawada and Shinohara30 reported a working range of 2 to 20 yg ml-1. However, 1-N-substituted benzodiazepines, e.g., diazepam and flurazepam, did not give this reaction.PROTEIN BINDING The amount of centrally acting drug that was available for transport across the blood brain barrier has been found to be related to the fraction that existed as free drug in the plasma unbound to the plasma proteins. The degree to which a drug is bound to protein in biological media has with other physico-chemical properties of the drug to be considered in extraction techniques. Van der KleijnS9 has used preparative ultra-centrifugation for the determination of the fraction of chlordiazepoxide and diazepam bound to plasma proteins in in vivo experiments, after administration of these drugs intravenously to dogs. The figures quoted indicated that 89 to 93 per cent, of the chlordiazepoxide and 89 to 94 per cent.of the diazepam were protein bound; however, these binding parameters could not be related to partition coefficients or R, values by thin-layer chromatography. ANALYSIS OF THE METABOLITES OF SOME 1 ,4-BENZODIAZEPINES The separation and identification of these products of biotransformation is important as many of them are active biologically although they may differ in potency and action from the parent compound.262 CLIFFORD AND SMYTH DETERMINATION OF 1,4-BENZODIAZEPINES [A?Zdyd, VOl. 99 CHLORDIAZEPOXIDE- A selective spectrofluorimetric methodls for the determination of chlordiazepoxide based on the hydrolysis of this compound to its lactam derivative could be modified to measure endogenous levels of this lactam, which occurred as a natural metabolite of chlordiazepoxide : this metabolite is now termed “demoxepam.” The metabolism of chlordiazepoxide has also been studied by using the 14C-labelled compound (on the 2-position)l17: the problem en- countered due to an isomerisation reaction and its remedy by sublimation under vacuum have been fully described in a recent review.lls The analytical techniques used in such metabolic st~dies~838~9~~~ were also collected in this review and are summarised in Table XV.TABLE XV ANALYSIS OF SOME 1,4-BENZODIAZEPINE METABOLITES (TO 1967) 1,4-Benzodiazepine metabolites Chlordiazepoxide : Chlordiazepoxide : (i) lactam lactam (2;) open lactam (iii) three others Chlordiazepoxide : desmethyl metabolite Diazepam : 2-amino and 2-methyl- smino-4-h ydroxy metabolites: after hydrolysis to benzo- phenones Diazepam and N-des- methyldiazepam Diazepam : 3H-label on phenyl ring 3-Hydroxydiazepam N-Desmethyldiazepam Oxazepam Diazepam Oxazeparn Diazepam N-Desmeth yldiazepam levels Refer- Extraction solvent ence and pH of media 18 Ether, pH 7.2 117 Plasma extracted with n-heptane a t pH 7.4 to remove chlordiazepoxide Extracted with ethyl acetate a t pH 7.0 and butanol a t P H 3 84 Diethyl ether, pH 7.0 to 7.22 Extracted with 0.1 N NaOH to remove lactam 119 22 Diethyl ether extract Hydrolysis to benzo- of blood phenone, re- extracted into ether plasma 40 Ether extraction of 49 Ethyl acetate extraction of urine after enzyme hydrolysis 120 Chloroform, pH 7 to 7.5 64 Blood extracted with diethyl ether Chromatography Determination step Spectrofluorimetry Paper chromatography Liquid scintillation for metabolites counting extracted into butanol graphy with n-heptane - CHCl, - ethanol Thin-layer chromato- (5 + 10 + 1) Spectrofluorimetry Column and thin- Ultraviolet and infra- layer chromato- red spectrophoto- graphy metric comparison with authentic compounds Gas - liquid chromato- Electron-capture graphy (2% carbo- detection wax 20M a t 190 “C) Two-dimensional TLC Liquid scintillation n-Heptane - CHCl, - counting ethanol (10 + 10 + 1) n-Heptane - CHC1, - Comparison of RF values against authenticcom- acetic acid - ethanol ( 5 + 5 + 1 + 0-3) pounds.Reverse or propan-2-01 - isotope dilution ammonia (sp. gr. analysis for study 0.88) (20 + 1) on oxazepam Thin-layer chromato- Colorimetry (Bratton - graphy Marshall procedure), CHCl, - acetone - infrared spectro- ethanol (8 + 1 + 1) scopy Two-dimensional thin- Ultraviolet spectro- layer chromato- photometry graphy measured both graphy of benzo- metabolite.phenones (2 ”/, Carbowax 20M) gave a selective Gas - liquid chromato- diazepam and 3H electron capture determinationMay, 19741 1,4-Benzodiazepine metabolites Diazepam 3 Phenolic metabolites (3H diazepam) in rat Oxazepam Oxazepam glucuronide Nitrazepam : 7-amino metabolite; 7-acetamido metabolite Nitrazepam : 7-amino metabolite AND THEIR METABOLITES IN BODY FLUIDS TABLE XV-continued 263 Refer- Extraction solvent ence and pH of media Chromatography Determination step 101 Ethyl acetate extrac- Thin-layer chromato- Mass spectrometry tion of gut contents graphy Nuclear magnetic after enzymic hydro- n-Heptane - CHCl, - resonance spectro- lysis a t pH 7.0 scopy to establish ethanol ( 5 + 5 + 2) ammonia (sp.gr. hydroxyl group Propan-2-01 - position of phenolic 0.88) (20 + 1) graphy 43 1,S-Dichloroethane Paper chromato- Autoradiography Butanol - ethanol - Butanol - pyridirie - water (17 -f 3 + 20) water (6 + 4 + 3) 28 Dichloromethane, Thin-layer chromato- Spectrophotometry Ethyl acetate - pH 10 P P h Y propanol- diethyl- amine Toluene - acetone - ammonia (sp. gr. 0.88) (50 + 50 + 1) Colorimetry (Bratton - Marshall procedure) Thin-layer chromato- graphy 45 Ether, pH 6.8 Ethyl acetate - n-hexane - 120/, ammonia (90 + 27 + 26) The metabolism of chlordiazepoxide in the rat was studied in more detail by solvent extraction, thin-layer chromatography and mass spectrometry after the 14C-labelled com- pound (on the2-position) had been given to two rats whose urine was collected during 3 days.102 In order to avoid photochemical degradation of chlordiazepoxide and its metabolites, the separation procedures were undertaken in semi-darkness or under a red light.Solvent extraction was performed at pH 1 with ethyl acetate: the organic phase was washed with 0.1 N hydrochloric acid and re-extracted with ethyl acetate after adjusting the pH to 7.0. The aqueous phases were then brought to a pH of 5.5 and incubated at 37 "C for 3 hours with glusulase (P-glucuronidase and sulphatase) so as to hydrolyse the metabolites conjugated t o glucuronic acid: after this enzyme hydrolysis, the extractions with ethyl acetate at pH 1 and pH 7 were repeated.The labelled metabolites were then purified and separated by thin-layer chromato- graphy on silica gel containing a fluorescent indicator by using a solvent mixture of n- heptanc - chloroform - ethanol (2 + 2 + 1) for development. The eluates from this initial treatment were subjected to further purification by running them in ethyl acetate - ethanol (80 + 20) or (90 + 10) prior to mass-spectral analysis. As mass-spectral analysis failed specifically to locate the hydroxyl group on the phenyl ring in the metabolites, two-dimensional thin-layer chromatography against authentic reference compounds established its location in the para position. The structure of one of these phenolic metabolites is shown (VII) and it is a product of the metabolism of both chlordiazepoxide and diazepam in the rat.264 CLIFFORD AND SMYTH : DETERMINATION OF lJ4-BENZOD1AZEPINES [Analyst, Vol.99 After an oral dose of 26 mg kg-1 of 14C-labelled chlordiazepoxide had been given to a dog, 1-1 per cent. of the total radioactivity was found in oxazepam.lZ1 This metabolite was extracted from urine, both before and after hydrolysis with P-glucuronidase, with 1,2- dichloroethane as solvent, before separation by paper and thin-layer chromatography. The metabolite was located by autoradiography and liquid scintillation counting after elution of the radioactive spots. A 200-mg dose of 2-14C-labelled demoxepam (V), a biologically active metabolite of chlordiazepoxide, was given orally to man in order to study its metabolism as investigations in the dog had found that its metabolism proceeded via oxidative, reductive and hydrolytic pathways.122 In rnan,l23 samples of pooled urine were extracted with ethyl acetate at pH 7, which removed the parent drug and its lipophilic metabolites, e.g., demoxepam, in agreement with the findings of Barrett, Smyth and Hart.124 The unextracted labelled compounds were hydrolysed with glusulase so as to release the conjugated metabolites prior to their extraction at pH 7 into ethyl acetate.A final extraction at pH 2-0 removed the metabolites that formed salts at pH 7.0, e.g., the “opened lactam” metabolite (VIII). The behaviour of the metabolites against authentic compounds on two-dimensional thin-layer chromatography by using at least two different solvent systems from a choice of chloroform - ethanol - acetone n-heptane - chloroform - ethanol - am- monia solution (sp, gr.0.88) (5 + 5 + 2 + 0.1) or n-heptane - ethyl acetate - ethanol (40 + 80 + 5 ) and (40 + 80 + 10) was required to establish their identity. It was found that most of the demoxepam (the major metabolite) and all of the opened lactam detected was unconjugated but that oxazepam was mainly conjugated. It was considered that in man potential chlordiazepoxide metabolites were N-desoxydemoxepam (IX) and phenolic compounds. (80 + 5 + 5 ) , Vlll I X Schwartz et have also determined the biological half-life of demoxepam (V) by using As the 9-hydroxy and 5-(4’-hydroxyphenyl) metabolites of a fluorimetric method of assay.demoxepam produced no fluorescence, the method was considered specific. DIAZEPAM- The analytical chemistry of diazepam metabolites has been reviewed up to 1967.118~1~6 Diazepam has been shown to undergo biotransformation by N-demethylation (IX), C-3- hydroxylation (X) , 5-phenylhydroxylation (XI) and conjugation.126 The analytical methods reported in these review papers are summarised in Table XV. X XIMay, 19741 AND THEIR METABOLITES I N BODY FLUIDS 2 65 The metabolism of diazepam has also been studied in vitro in dog-liver preparations.127 Diazepam was synthesised and randomly labelled with tritium in the 5-phenyl ring. The determination of 3H-diazepam and its metabolites was effected by extraction with ethyl acetate of labelled compounds from the liver incubation mixture prior to two-dimensional thin-layer chromatography with internal and external reference compounds : the quantitative deter- mination of metabolite levels was carried out by liquid scintillation counting. Gas - liquid chromatography of diazepam and its metabolites with nickel-63 electron- capture detection in order to detect the metabolites after their separation on a 3 per cent.column liquid phase has shown differences in species between rat and mouse in the biotrans- formation of diazepam.128 The gas - liquid chromatographic separation of diazepam from its metabolite^^^,^^ has been described in the chromatography section above. More recently Zingale~l~~ determined diazepam metabolite levels in man during chronic medication. Diazepam and its metabolites were extracted from plasma, lysed red blood cells and urine (all adjusted to pH 9.5 with sodium hydroxide) with a toluene - heptane (80 + 20) solvent mixture.Under the conditions of these determinations, diazepam, desmethyldiazepam (IX) , 3-hydroxydiazepam (X) and oxazepam were partitioned between toluene (containing 2 per cent. of 3-methylbutan-1-01) - n-heptane (80 + 20) and various buffer solutions containing 1 pg ml-I of diazepam and desmethyldiazepam ; oxazepam and 3-hydroxydiazepam were extracted at a concentration of 0-1 pug ml-l (Table XVI). The low distribution coefficients of desmethyldiazepam and oxazepam when extracted from pH 2 buffers are in agreement with later pK, determinations, as is the extraction efficiency of oxazepam from the pH 14 buffer s o l ~ t i o n .~ ~ ~ ~ ~ TABLE XVI APPROXIMATE PERCENTAGES OF DIAZEPAM AND ITS METABOLITES EXTRACTED INTO TOLUENE (CONTAINING 2 PER CENT. OF 3-METHYLBUTAN-1-OL) - n-HEPTANE (80 + 20) AT VARIOUS BUFFER PHS pH of buffer solution Compound 2 5.5 9.5 14.6 Desmethyldiazepam . . . . 20 70 100 88 Diazepam . . .. . . 65 90 100 100 Oxazepam . . .. . . 40 80 100 18 3-Hydroxydiazepam . . .. 85 97 100 100 In order to detect these four compounds at sub-nanogram levels, the toluene - n-heptane phase was re-extracted with 6 N hydrochloric acid. Gas - liquid chromatographic separation of these compounds was performed on 2 per cent. OV-17 on Chromosorb W HP, 80 to 100 mesh, with an oven temperature of 235 "C and an electron-capture detector at 330 "C : under these conditions no thermolysis of oxazepam was reported, although with respect to its melting-point (206 "C) oxazepam had an anomalous retention time of 4 minutes compared with that of diazepam (5.5 minutes, m.p.131 to 135 "C). Barrett et aZ.124 also reported a differential solvent-extraction method for the separation of diazepam from its N-desmethyl metabolite (Table XVII). Their determination step was polarographic. TABLE XVII PERCENTAGE OF DIAZEPAM AND DESMETHYLDIAZEPAM (CONCENTRATION 10-6 M) EXTRACTED FROM AQUEOUS PHASE AT pH 14" INTO CHLOROFORM AND LIGHT PETROLEUM Solvent I Compound Chloroform Light petroleum Diazepam . . . . .. 95-100 95-100 Desmethyldiazepam . . 95 0-5 * The pH a t which diazepam exists as a readily extractable neutral molecule and desmethyldiazepam as an anion.266 CLIFFORD AND SMYTH : DETERMINATION OF 1,4-BENZODIAZEPINES [A?Za@St, VOl.99 OXAZEPAM- A previous study,43 in which 2-14C-labelled oxazepam was used, had shown that in the dog and pig the major metabolite of oxazepam was oxazepam glucuronide (XII), whereas in the rat seven metabolites were found. Recently, it has been reported that although in man and the miniature pig 95 per cent. of a dose of non-labelled oxazepam is excreted as the oxazepam glucuronide conjugate, numerous other minor metabolites exist and one of these (XIII), which is hydroxylated on the phenyl ring, is the major metabolite in the rat. In this study,130 oxazepam and its metabolites were extracted into diethyl ether from urine adjusted to pH 7 both before and after p-glucuronidase hydrolysis ; two-dimensional thin-layer chromatography was used to separate the compounds extracted and the developing solvents were chloroform - 95 per cent.ethanol - acetone (8 $- 1 + 1) and then ethyl acetate - 95 per cent. ethanol - ammonia solution (sp. gr. 0.88) (5 + 1 + 1). The spots on the chromatograms were viewed under long-wave ultraviolet light and made visible with 2,4-dinitrophenylhydrazine and by the Bratton - Marshall reaction. Identification of the metabolites was effected by comparing the mass spectra of authentic substances with those of the compounds eluted from the chromatograms. The technique used is presented in Table XV. Minor metabolites found in man included 6-chloro-4-phenyl-2 (1H)-quinazoline, Z-amino- 5-chlorobenzophenone, 2-benzoyl-4-chloro-2,2-dihydroxyacetamide and 7-chloro- 1,3-dihydro- 3-hydroxy-5(~-hydroxyphenyl)-2H-1,4-benzodiazepin-2-one and its conjugate.NITRAZEPAM- Reider28 has extracted nitrazepam, and its 7-amino (XIV) and 7-acetamido (XV) derivatives, from deproteinised plasma at pH 10 into an organic solvent [dichloromethane - ethyl acetate (2 + l)] prior to determining it spectrophotometrically, by using the Bratton - Marshall reagent, which reacts with the primary aromatic amino group formed after reduction of nitrazepam and hydrolysis of the 7-acetamido compound. A differential re-extraction into different concentrations of hydrochloric acid in order to separate nitrazepam from its metabolites86 has been described in the spectrofluorimetry section (see above).As both metabolites have been found to be inactive, further differentiation is less important than with other 1,4-benzodiazepines although thin-layer chromatographic methods have been reported48160 and are shown in Table XV. H H XIV xvMay, 19741 AND THEIR METABOLITES I N BODY FLUIDS 267 Beyer and Sad4e66 synthesised the 7-amino, 7-acetamido, 7-amino-3-hydroxy and 7-nitro- 3-hydroxy metabolites of nitrazepam and examined their behaviour in three solvents on thin-layer chromatography. These model metabolites were extracted into ethyl acetate from solutions at pH 8 to 10. The 7-amino and 7-acetamido metabolites and the 7-amino- 3-hydroxynitrazepam (not found in human blood) could be re-extracted from ethyl acetate with 0-1 s hydrochloric acid, and nitrazepam and the 7-nitro-3-hydroxynitrazepam meta- bolite could then be extracted with 10 per cent.hydrochloric acid. A table showing the ultraviolet spectral absorption maxima and minima of these compounds in methanol was also presented. Sawada and Shinohara131 used this method to extract nitrazepam and its metabolites from the urine of men and rabbits who had received 10 and 50mg kg-1 of nitrazepam, respectively . I t has been reported124 that in Britton- Robertson buffer solution (pH 4) both the 7-amino and 7-acetamido metabolites of nitrazepam (10P M concentration) were equally well extracted into dichloroethane. Although the latter compound exists as an un-ionised species whereas the 7-amino metabolite is monoprotonated at this pH, it was considered that ion-pair formation occurred with the amino compound, which enhances its extractability and makes differentiation of these metabolites difficult. A polarographic differentiation of nitrazepam from its two major metabolites was described in this report.The method is based on the ratio of wave heights in Britton - Robertson buffer (pH 4) after solvent extrac- tion of nitrazepam and its two major metabolites at pH 7 at which pH all of them exist as neutral molecules. As the first wave corresponds solely to the 4-electron reduction of -NO, NO++NHOH -NHOH%NH~ -0-27V >C=N-,XN-NH' 2e \ -0.84V --RIH.COCH3 / metabolite - 0 4 3 V )C=N%>CH NH CY Concentration of total metabolites after polarography Mixtuine pH7 and extraction of subsequent all three a t t parent drug - ------- a Concentration of r '-0*27V -0.83V Fig.2. Scheme of polarographic differentiation of nitrazepam from its two major metabolites. pH range of predominantly PIG non-ionised form Nitrazepam . . . . 3.2 and 10.8 4 to 10 -NHCOCH, metabolite 3.2 and 12.4 4 to 11.5 -NH, metabolite . . 2.5, 4.6 and 13.1 5.5 to 12.5 In appropriate pH range all three can be extracted into a suitable solvent (ethyl acetate). Separate the organic layer, evaporate off the solvent and polarograph at pH 4268 CLIFFORD AND SMYTH : DETERMINATION O F 1,4-BENZODIAZEPINES [ A nabst, Vol. 99 to -NHOH, which occurs solely in nitrazepam, and as a second wave was due to the 2-electron reduction of -NHOH to -NH, in nitrazepam and the sum of the 2-electron azomethine reductions occurring in all three compounds, the difference in wave height between the two waves could provide a measure of total metabolites; such a scheme is illustrated in Fig.2. LORAZEPAM- The metabolism of lorazepam was studied in man, the dog, cat, pig and rat.68 In the first four species more than 99 per cent. of the drug is excreted in the urine as the lorazepam glucuronide, as lorazepam, like oxazepam, has the structural requirement (hydroxyl group) for glucuronide formation. The urine was extracted at pH 7.0 with ether both before and after /3-glucuronidase enzyme hydrolysis. Ether extracted all but one metabolite, which occurred in rat urine and required the use of ethyl acetate for its complete extraction. Separation and purifica- tion of the metabolites was achieved by thin-layer chromatography on fluorescent silica gel plates developed with two solvent systems: chloroform - ethanol - acetone (8 + 1 + 1) and ethyl acetate - ethanol - ammonia solution (sp.gr. 0.88) (5 + 5 + 1). The Bratton - Marshall reaction procedure was used to detect lorazepam and its metabolites, including the 3-hydroxy and hydroxyphenyl biotransformation products. The structure of the quinazoline meta- bolite (XVI) was established by comparison of its infrared and mass spectra, and its behaviour on thin-layer chromatography, when compared with authentic reference com- pounds. MEDAZEPAM- Studies on the biotransformation products of medazepam were first reported by Reider and Rentsch13, : 5-14C-labelled compounds were extracted from plasma, adjusted to pH 9.0, into diethyl ether and from urine at pH 7.0 into ethyl acetate.Purification and separation of the metabolites were achieved with thin-layer chromatography by using the following developing systems : n-heptane - chloroform - ethanol (10 + 10 + 1) ; n-heptane - chloroform - acetic acid - ethanol (5 + 5 + 1 + 0.3); propan-2-01 - ammonia solution (sp. gr. 0.88) (20 + 1) ; n-heptane - ethyl acetate - ethanol - ammonia solution (sp. gr. 0%3) (5 + 5 + 1 + 0.3); and chloroform - acetone (9 + 1). Localisation of the radioactive areas was by autoradiography and identification by two- dimensional thin-layer chromatography against authentic reference substances. Liquid scintillation counting after elution was used for the determination of medazepam and its metabolite levels. In a later study,133 similar solvent-extraction procedures for 5-14C-labelled medazepam and its metabolites and a previously described thin-layer chromatographic system for their separation were used : identification was effected by two-dimensional thin-layer chromato- graphy against internal reference compounds suspected of being metabolites.In man, dog and rat N-desmethylmedazepam (XVII), diazepam and desmethyldiazepam (IX) were described as tentative blood metabolites, and metabolites identified in a 0 to 7-day pooled urine sample after glusulase hydrolysis were desmethyldiazepam, N-desmethyl-l,2-dehydro- medazepam (XVIII) (an intermediate metabolite SO far found only in the rat), oxazepam and 2-amino-3-hydroxy-5-chlorobenzophenone. However, these metabolites accounted for only about 14 per cent.of urinary carbon-14. De Silva and Puglisi2* used gas - liquid chro- matography with nickel-63 electron capture in order to measure medazepam and its meta- bolites diazepam, desmethylmedazepam and desmethyldiazepam in the blood of subjects receiving medazepam: oxazepam was the major metabolite in the urine. xv I XVI IMay, 19741 AND THEIR METABOLITES I N BODY FLUIDS 269 XVI I I XIX FLURAZEPAM- Schwartz, Vane and Postma50 studied the urinary metabolites of flurazepam in two human subjects who had been given a 60-mg dose (about four times the normal hypnotic dose) and found one metabolite, 7-chloro- 1 - (2-hydroxyet hyl) -5- (2-fluorophenyl) - 1,3-dihydro- 2H-1,4-benzodiazepin-2-one (XIX). The 24-hour urine samples were adjusted to pH 9 with 1 N sodium hydroxide solution and extracted twice with equal volumes of ether: in this ether fraction flurazepam and neutral and basic non-polar metabolites were found.The pH of the urine was then adjusted to 7-0 and extracted with ethyl acetate so as to remove the neutral more polar metabolites and the extraction was repeated after enzymic hydrolysis in order to release the conjugated metabolites. The 2-hydroxyethyl metabolite (XIX) was extracted by all three solvent systems and had an R, x 100 value of 71 on thin-layer chromatography when developed with ethyl acetate - ethanol - ammonia solution (sp gr. 0.88) (90 + 10 + 0.3); flurazepam had an RF x 100 value of 50. The structure of this metabolite was elucidated by high-resolution mass-spectral analysis and two-dimensional thin-layer chromatography against authentic reference com- pounds.In this study, a single dog was given 40 mg kg-l of flurazepam daily for 6 months and by a similar analytical technique the 2-ethylaminoethyl (XX) 2-aminoethyl (XXI) and a 2H-3-hydroxy metabolite were identified in addition to a phenolic derivative of 7-chloro-5- (2-fluorophenyl) -1 ,3-dihydro-2H-l,4-benzodiazepine. and dogs 5-14C-flurazepam xx XXI In a further studv of the metabolism of flurazeDam in man hydrochloride was usGd.134 Both plasma and urini were fractionated b y extraction into ether at pH 9.0 and into ethyl acetate after adjusting the pH of the biological samples to pH 7.0 and then to pH 3.0. The urine samples were also hydrolysed with P-glucuronidase. Separation and determination of the drug and metabolites were carried out by noting the amount of chromatographed radioactivity that migrated on two-dimensional thin-layer chromatography in a manner similar to that for an authentic synthesised reference compound.De Silva and S t r ~ j n y ~ ~ described a spectrophotofluorimetric method for the determina- tion of flurazepam and its metabolites in blood and urine. This somewhat lengthy procedure involved their solvent extraction into diethyl ether from the biological sample after adjusting the pH to 9. Re-extraction into 4 N hydrochloric acid was followed by hydrolysis of the parent drug and its metabolites to their respective benzophenones, which were extracted into diethyl ether after making the hydrolysate alkaline.If the original sample was urine the residue after evaporation of this second ether extraction was subjected to thin-layer chromato- graphy in order to separate the benzophenones, which were then determined by ultraviolet - visible absorption spectrophotometry. The sensitivity of this procedure was stated to be270 CLIFFORD AND SMYTH DETERMINATION OF 1,4-BENZODIAZEPINES [Analyst, VOl, 99 1 to 2 pug per millilitre of sample. The benzophenones formed from flurazepam and its metabolites in blood-plasma samples were allowed to react with dimethylformamide at 100 “C for 2 hours in the presence of potassium carbonate as a catalyst so as to form the acridanones (VI) prior to thin-layer chromatography and determination by spectrophoto- fluorimetry. This method has greater sensitivity than absorption spectrophotometry (5 to 10 ng per millilitre of sample) and avoids the interference of high blank values experienced in spectrophotometry of extracts of biological samples. Nishikawa, Mineura and Takahira135 have recently studied the metabolism of flurazepam in the rat after oral dosage of l*C-flurazepam labelled in the 5-C-position.Bile and urine were adjusted to pH 9.0 with sodium hydroxide and then extracted with ethyl acetate; the flurazepam metabolite (XX) was found in the organic phase. The aqueous layer was then placed on an Amberlite XL4D-2 column and eluted with methanol. The eluate was then subjected to further chromatography on a DEAE-Sephadex column, which was eluted first with phosphate buffer at a pH of 7-2 to give a fraction containing the metabolite XXI and then with 0.5 M sodium chloride solution, which yielded the conjugates of metabolites XX and XXI.Characterisation of the metabolites was achieved by causing them to react with 3,5-dinitrobenzoyl chloride or by benzophenone formation and comparison of their R, values on thin-layer chromatography against those of authentic substances. A further urinary metabolite was the N-1 acetic acid and one of the major metabolites, which possessed an intact 1,4-benzodiazepine ring and an N-1 aminoethyl group, was not fully identified. Barrett, Groves and Smyths have studied the acid - base equilibria that exist in aqueous solutions of some flurazepam metabolites and suggested differential solvent extraction of those compounds which are hydrogen substituted in the l-position (and thus ionisable in alkaline media at pH 13 to 14) from those which have more bulky non-ionisable substituents in this position.In addition, substituents in the l-position that contain an amino group can be differentiated from others, e.g., R’ = -CH,CH,OH (XIX), by a further solvent extrac- tion. Clifford, Smyth and Smythl36 have been successful in applying these findings to the analysis of plasma samples obtained from beagles and rhesus monkeys following intravenous administ ration. ADDENDUM I t is pertinent to consider the technique of drug determination by radioinimunoassay, which has been reviewed by Butler.l3‘ Peskar and Spector138 have recently described the determination of diazepam in blood by using the antisera to two different analogues of diazepam.It was stated that this radioimmunoassay was capable of detecting 1 ngml-l of diazepam and its major metabolite N-desmethyldiazepam. The 2-0x0 group was considered to be an important antigenic determinant, and the 3-hydroxy group (e.g., in oxazepam) did not bind well to either antibody. Blood level determinations compared well with those measured in a recent gas - liquid chromatographic it remains to be seen if this powerful new technique can be applied to the selective analysis of other benzodiazepines. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES Randall, L. O., Dis. New. Syst., 1961, 22, Suppl. Part 11, 4. Zbinden, G., and Randall, L. O., in Garattini, S., and Shore, P. A., Editors, *‘Advances in Pharma- cology,” Volume V, Academic Press, New York and London, 1967, p.213. Schallek, W., Schlosser, W., and Randall, L. O., in Garattini, S., Goldin, A., Hawking, F., and Kopin, I. J ., Editors, “Advances in Pharmacology and Chemotherapy,” Volume X, Academic Press, New York and London, 1972, p. 119. Stecker, P. J., Editor, “Merck Manual,” Eighth Edition, Merck & Co. Inc., Rathway, New Jersey, 1968. Beyer, I<.-H., and Sadde, W., Dt. ApothZtg, 1969, 109, 312. MacDonald, A., Michaelis, A. F., and Senkowski, B. Z., in Florey, K., Editor, “Analytical Profiles of Drug Substances,” Volume I, Academic Press, New York and London, 1972, pp. 15-51 and 79-99. Barrett, J., Smyth, W. F., and Davidson, I. E., J . Pharm. Pharmac., 1973, 25, 387. Groves, J., and Smyth, W. F., t o be published.Oelschlager, H., Arch. Pharm., Bed., 1963, 296, 396. Senkowski, B. Z., Levin, M. S., Urbigkit, J. R., and Wollish, E. G., AnaZyt. Chem., 1961, 36, 1991. Oelschlager, H., Volke, J., and Lim, G. T., Arch. Pharm., Bed., 1969, 302, 241. Childress, S. J., and Gluckman, M. I., J . Pharm. Sci., 1964, 53, 577.May, 19741 AND THEIR METABOLITES I N BODY FLUIDS 271 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. Archer, G. A., and Sternbach, L. H., Chem. Rev., 1968, 68, 747. Sternbach, L. H., Angew. Chem., I n t . Edn, 1971, 10, 34. Tompsett, S. L., Analyst, 1968, 93, 740.Clifford, J . M., M.Sc. dissertation, University of London, 1972. Proelss, H. F., and Lohmann, H. J., Clin. Chem., 1971, 17, 222. Koechlin, B. A. and D’Arconte, L., Analyt. Biochem., 1963, 5, 195. Cimbura, G., and Gupta, R. C., J . Forens Sci., 1965, 10, 282. Tompsett, S. L., J . Clin. Path., 1968, 21, 366. Zingales, I. A., J . Chromat., 1971, 61, 237. de Silva, J. A. F., Schwartz, M. A., Stefanovic, V., Kaplan, J., and D’Arconte, L., Analyt. Chem., Foster, L. B., and Frings, C. S., Clin. Chem., 1970, 16, 177. de Silva, J. A. F., and Puglisi, C. V., Analyt. Chem., 1970, 42, 1725. Berry, D. J., Clinica Chim. Acta, 1971, 32, 235. van der Kleijn, E., Beelen, G. C., and Frederick, M. A., Ibid., 1971, 34, 345. Kamm, G., and Kelm, R., Arzneimittel-Forsch, 1969, 19, 1659.Reidcr, J., Ibid., 1965, 15, 1134. Beyer, I<.-H., and Sadbe, W., Ibid., 1969, 19, 1929. Sawada, H., and Shinohara K., J . Hyg. Chcm., 1970, 16, 318. Beharrell, G. P., Hailey, D. M., and McLaurin, M. K., J . Chromat., 1972 70, 45. Iinowles, J. A., Comer, W. H., and Ruelius, H. W., Arzneimittel-Forsch., 1971, 21, 1055. de Silva, J . A. F., and Strojny, N., J . Pharm. Sci., 1971, 60, 1303. Marcucci F., Fanelli, R., and Mussini, E., J . Chromat., 1968, 37, 318. Gardner-Thorpe, C., Parsonage, M. J., and Toothill, C., Clinica Chim. Acta, 1971, 35, 39. de Zeeuw, R. A., Prog. Chem. Toxicol., 1969, 4, 59. Fox, A. S., “Paper Chromatography,” in Clarke, E. G. C., Editor, “The Isolation and Identification Curry, A. S., and Powell, H., Nature, Lond., 1954, 173, 1143.Street, H. V., Acta Pharmac. Tox., 1962, 19, 312. Stevens, H. M., and Jenkins, R. W., J . Forens. Sci. SOC., 1971, 11, 183, Macek, K., Vecerkova, J., and Stanislova, J . , Pharmazie, 1966, 20, 605. Kaistha, K. K., and Jaffe, J . H., J . Chromat., 1971, 60, 83. Walkenstein, S. S., Wiser, R., Gudmundsen, C. H., Kimmel, H. B., and Corradino, R. A., J . Beckstead, H. D., and Smith, S. J., Arzneiinittel-Forsch., 1968, 18, 529. Hermans, R. B., and Kamp, P. E., Pharm. Weekbl. Xed., 1967, 102, 1103. Lafargue, P., Mcunier, J., and Lemontey, Y., J . Chromat., 1971, 62, 423. Pribilla, O., Arzneimittel-Forsch., 1965, 15, 1148. Kamm, G., and Baier, R., Ibid., 1969, 19, 213. Schwartz, M. A., Koechlin, €3. A., Postma, E., Palmer, S., and Krol, G., J . Pharmac.Exp. Ther., Schwartz, M. A., Vane, F. M., and Postma, E., J . Med. Chem., 1968, 11, 770. Paulus, W., Hoch, W., and Keymer, K., Arzneimittel-Forsch., 1963, 13, 609. Cochin, J., and Daly, J. W., .I. Pharmac. Exp. They., 1963, 139, 154. Curry, A. S., “Thin Layer Chromatography,” in Clarke, E. G. C., Editor, “The Isolation and Thomas, J. J., and Dryon, L., J . Phavm. Belg., 1961, 19, 481. Tomoda, M., Kyoritsu Yakka Daigaku Kenky Neppo, 1965, 10, 18. Norifalaise, A,, .I. Chromat., 1965, 20, 61. Korzun, B. P., Brody, S. M., Keegan, P. G., Luders, R. C., and Rehm, C. R., J . Lab. Clin. Med., Roder, E., Mutschler, E., and Rochelmeyer, H., Z. analyt. Chem., 1969, 244, 45. van tler Kleijn, E., Archs Int. Pharnzacodyn. ThLr., 1969, 179, 225. Sawada, H., and Shinohara, K., Arch.Tox., 1970, 27, 71. Mul6, S. J., J . Chromat., 1971, 55, 255. Hetland, L. B., Knowlton, D. A., and Cori, D., Clinica Chim. Acta, 1972, 36, 473. Sunshine, I., Amer. J . Clin. Path., 1963, 40, 576. de Silva, J . A. F., Koechlin, B. A,, and Bader, G., J . Pharm. Sci., 1966, 55, 692. Yamamoto, J., Suzuki, M., and Okuo, S., Yakinzaigaku, 1965, 25, 23. Beyer, K.-H., and Sad&, W., Arch. Pharm., Rerl., 1969, 302, 152. Lafargue, P., Pont, P., and Meunier, J., Annls Pharm. Fr., 1970, 28, 343. Schillings, R. T., Shrader, S. R., and Ruelius, H. W., Arzlzeimittel-Forsch., 1971, 21, 1059. Schutz, C., and Schutz, H., Arch. Tox., 1973, 30, 183. Besserer, K., Henzler, S., Kohler, E., and Mallach, H. J., Arzneimittel-Forsch., 1971, 21, 2003. McMartin C., and Street, 13.V., J . Chromat., 1966. 22, 274. Sadbe, W., and van der Kleijn, E., J . Pharm. Sci., 1971, 60, 135. Gardner-Thorpe, C., Parsonage, M. J., Smethurst P. F., and Toothill, C., Clinica Chim. Acta, Cano, J. P., Vignoli, L., and Viala, A., Annls Pharm. Fr., 1967, 25, 821. Lafargue, P., Pont, P., and Meunier, J., Ibid., 1970, 28, 477. Forgione, A., Martelli, P., Marcucci, F., Fanelli, R., Mussini, E., and Jommi, G. C., J . Chromat., 1964, 36, 2099. of Drugs,” The Pharmaceutical Press, London, 1969, p. 31. Pharm. Sci., 1964, 53, 1181. 1965, 149, 423. Tdentification of Drugs,” Pharmaceutical Press, London, 1969, p. 43. 1966, 68, 333. 1972, 36, 223. 1971, 59, 163.CLIFFORD AND SMYTH van der Kleijn, E., van Rossum, J. M., Muskens, E. T. J . M., and Rijntjes, N. V. M., Acta Pharmac.Vessman, J., Freij, G., and Stromberg, S., Acta Pharm. Suecica, 1972, 9, 447. Knowles, J . A., and Ruelius, H. W., Arzneimittel-Forsch., 1972, 22, 687. Masuda, Y., Jap. J . Legal. Med., 1971, 25, 445. Marcucci, F., Mussini, E., Arioldi, L., Guitani, A., and Garattini, S., J . Pharm. Pharmac., 1972, Scott, C. G., and Bommer, P., J . Chromat. Sci., 1970, 8, 446. Weber, D. J., J . Pharm. Sci., 1972, 61, 1797. Schwartz, M. A., and Postma, E., Ibid., 1966, 55, 1358. Braun, J., Caille, G., and Mockle, J. A., Can. J . Pharm. Sci., 1968, 3, 65. Reider, J., Arzneimittel-Forsch., 1973, 23, 207. -, Ibid., 1973, 23, 212. Fidelus, J., Zietek, M., Mikolajek, A., and Grochowska, Z., Mikrochim. Acta, 1972, 84. Oelschlager, H., Volke, J., and Kurek, E., Arch. Pharm., B e d , 1964, 297, 431. Caille, G., Braun, J., and Mockle, J . A., Can. J . Pharm. Sci., 1970, 5, 78. Fazzari, F. R., and Riggleman, G. H., J . Pharm. Sci., 1969, 58, 1530. Oelschlager, H., Volke, J., Lim, G. T., and Spang, R., Arch. Pharm., Berl., 1969, 302, 946. Volke, J., Oelschlagcr, H., and Lim, G. T., J . Electroanalyt. Chem., 1970, 25, 307. Jacobsen, E., and Jacobsen, T. V., Analytica Chim. Acta, 1971, 55, 293. Halvorsen, S., and Jacobsen, E., Ibid., 1972, 59, 127. Jacobsen, E., Jacobsen, T. V., and Rojahn, T., Ibid., 1973, 64, 473. Clifford, J . M., and Smyth, W. F., 2. analyt. Chem., 1972, 264, 149. Goldsmith, J., Grant, J., Jenkins, A., and Smyth, W. F., Analytica Chim. Acta, 1973, 66, 427. Jacobsen, E., and Jacobsen, T. V., Ibid., 1972, 60, 472. Oelschlager, H., and Oehr, H. P., Pharm. Acta Helv., 1970, 45, 708. Schwartz, M. A., Bommer, P., and Vane, F. M., Archs Biochem. Biophys., 1967, 121, 508. Schwartz, M. A., Vane, F. M., and Postma, E., Biochem. Pharmac., 1968, 17, 965. SadCe, W., J. Med. Chem., 1970, 13, 475. Milne, G. W. A., Fales, H. M., and Axenrod, T., Analyt. Chem., 1971, 43, 1815. Nuhn, P., and Bley, W., Pharmazie, 1967, 22, 532. Linscheid, P., and Lehn, J. M., Bull. Soc. Chim. Fr., 1967, No. 3, 992. Bley, W., Nuhn, P., and Beundorf, G., Arch. Pharm., B e d , 1968, 301, 444. SadCe, W., Ibid., 1969, 302, 769. Chapman, D. I., and Moss, M. S., in Clarke, E. C. G., Editor, “The Isolation and Identification Drugs of Today, 1965, 1, 109. Ibid., 1965, 1, 137. Bellomonte, G., Boll. SOC. Ital. Biol. Sper., 1967, 43, 460. Grecu, I., and Barbu, S., Annls Pharm. Fr., 1968, 26, 405. Beyer, K. H., and SadCe, W., Arch. Pharm., Berl., 1967, 300, 667. Bratton, A. C., and Marshall, E. K., J . Bid. Chem., 1939, 128, 537. Frings, C. S., and Cohen, P. S., Amer. J . Clin. Path., 1971, 56, 216. Koechlin, B. A., Schwartz, M. A., Krol, G., and Oberhansli, W., J . Pharm. Exp. Ther., 1965, 148, Hirtz, J . L., “Analytical Metabolic Chemistry of Drugs,” Medicinal Research Series, Volume 4, Jommi, G.. Manitto, P., and Silanos, M. A., Archs Biochem. Biophys., 1964, 108, 334. Ruelius, H. W., Lee, M. L., and Alkin, H. E., Ibid., 1965, 111, 376. Kimmel, H. B., and Walkenstein, S. S., J . Pharm. Sci., 1967, 56, 538. Schwartz, M. A., Postma, E., and Krolis, S. J., Ibid., 1971, 60, 439. Schwartz, M. A., and Postma, E., Ibid., 1972, 61, 123. Barrett, J., Smyth, W. F., and Hart, J . P., J . Pharm. Pharmac., 1974, 26, 9. Schwartz, M. A., Postma, E., and Gaut, Z., J . Pharm. Sci., 1971, 60, 1000. Garattini, S., in Cerletti, A., and Bove, F. J., Editors, “The Present Status of Psychotropic Drugs,” Schwartz, M. A., and Postma, E., Biochem. Pharmac., 1968, 17, 2443. Marcucci, F., Mussini, E., Fanelli, R., and Garattini, S., Ibid., 1970, 19, 1847. Zingales, I. A., J . Chromat., 1973, 75, 55. Sisenwine, S. F., Tio, C. O., Shrader, S. R., and Ruelius, H. W., Arzneimittel-Forsch., 1972, 22, 682. Sawada, H., and Shinohara, K., Arch. Tox., 1971, 28, 214. Reider, J., and Rentsch, G., Arzneimittel-Forsch., 1968, 18, 1545. Schwartz, M. A., and Carbone, J. J., Biochem. Pharmac., 1970, 19, 343. Schwartz, M. A., and Postma, E., J . Pharm. Sci., 1970, 59, 1800. Nishikawa, T., Mineura, K., and Takahira, H., Yakugaki Zasshi, 1973, 73, 226. Clifford, J. M., Smyth, M. R., and Smyth, W. F., J . Pharm. Pharmac., submitted for publication. Butler, V. P., Metabolism, 1973, 22, 1145. Peskar, B., and Spector, S., J . Pharmac. Exp. Ther., 1973, 186, 167. Kaplan, S. A., Jack, M. L., Alexander, K., and Weinfeld, R. E., J . Pharm. Sci., 1973, 62, 1789. Tox., 1971, 29, Suppl. 111, 109. 24, 63. of Drugs,” The Pharmaceutical Press, London, 1969, pp. 103-122 and 688-793. 399. Marcel Dekker Inc., New York, 1971, pp. 81-89. Excerpta Medica Foundation, Amsterdam, 1969, pp. 84-89. Received September 21st, 1973 Accepted December 7th, 1973 272 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 106. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139.

ISSN:0003-2654    

DOI:10.1039/AN9749900241

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, May, 1974, Vol. 99, $9. 273-276 273 A Method for the Direct Titrimetric Determination of Iron(II1) in Silicate Rocks BY J. M. MURPHY, J. I. READ AND G. A. SERGEANT (Department of Trade and IBdustry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SE1 9NQ) The rock sample is decomposed with hydrofluoric acid in the manner of a conventional iron (11) determination, and the iron(II1) in the resulting solution is determined by titration with a solution of ferrocene (dicyclo- pentadienyliron) in 2-methoxyethanol. A modified procedure is necessary for rocks of very high magnesium content, and it is indicated that errors may arise in the conventional iron(I1) determination when applied to rocks of this type. THE iron(II1) content of silicate rocks is normally calculated by difference from the results obtained for total iron and iron(II), and the results of collaborative studies of rock analysis show that, among the major constituents, the result for iron(I.11) is one of the least reliable, mainly on account of uncertainties in the iron(I1) determination. For the total iron deter- mination various well known methods are available, while the iron(I1) is commonly determined by some variant of the Pratt method.1 In this method, the sample is first decomposed by heating in a covered platinum crucible with a mixture of dilute hydrofluoric and sulphuric acids. The contents of the crucible are then quickly added to an excess of boric acid solution and the iron(I1) is titrated with a standard solution of an oxidant such as dichromate.In the absence of a convenient and generally accepted method, the direct determination of iron(II1) in the solution thus prepared is not normally attempted but this could, in principle, provide a more accurate value for iron(II1) than that derived by difference, especially when the ratio of iron(II1) to iron(I1) is low. The titrimetric determination of iron(II1) in solution by titration with ferrocene (di- cyclopentadienyliron) has been proposed by Wolf, Franz and Hennig2 and is based on the reaction Fe(C,H,), + Fe3+ + Fe2+ + Fe(C,H,),+ In the presence of thiocyanate ions the end-point is indicated by the disappearance of the red colour of iron(II1) thiocyanate and its replacement by the blue colour of the ferriceniurn ion. As there is a high degree of specificity for iron(II1) and the reaction takes place in dilute mineral acid solution, there appeared to be a possible application to iron(II1) determination in solutions prepared as for the conventional iron(I1) procedure. The stoicheiometry of the titration was first examined and the reaction was found, under these conditions, to be quantitative.However, the presence of a suitable wetting agent appeared to be essential; the wetting agent presumably acts by ensuring the effective dispersion of the ferrocene in a finely divided and reactive form. A number of preparations were tried in this connection but some of them caused interference by interaction with the iron(II1) thiocyanate complex or by causing cloudiness in the solution. The wetting agent finally selected (see under Reagents) is satisfactory in both of these respects and is of the non-ionogenic type with a poly(ethy1ene glycol) chain.EXPERIMENTAL REAGENTS- All reagents should be of analytical-reagent grade when available. Sulphuric acid, 50 per cent. V/V-Add 1 volume of concentrated sulphuric acid (sp. gr. @ SAC; Crown Copyright Reserved.274 MCRPHY et nl. : A METHOD FOR THE DIRECT TITRIMETRIC [Analyst, Vol. 99 1.84) cautiously, with stirring, to 1 volume of water and finally dilute the mixture, when cool, back to 2 volumes. Hydrojzioric acid, 40 per cent. m/m. Boric acid solution, saturated. Boric acid. Ammonium thiocyanate solution, 5 per cent. m/V. Lissapol NDB solution-A 5 per cent. V/V solution of Lissapol NDB detergent liquid (distributed by Hopkin & Williams Ltd., by arrangement with Imperial Chemical Industries Ltd.) in 95 per cent.V/V ethanol. Ferrocene solution-Dissolve 0.5830 g of ferrocene (dicyclopentadienyliron) in 2-methoxy- ethanol and dilute the solution to 500 ml with the same solvent. This solution is stable if 5 tored in a glass-stoppered bottle. Potassium permanganate solution, 0- 1 per cent. m/V. Standard iron(l1l) solution-Dissolve 0.491 1 g of ammonium iron(I1) sulphate in about 25 ml of water containing 10 ml of the sulphuric acid (50 per cent. V / V ) and add a small excess of bromine water. Evaporate the solution until fumes of sulphuric acid appear, then cool it and cautiously add about 100 ml of water. Heat the mixture until a clear solution is obtained, then transfer it to a 500-ml calibrated flask, cool to room temperature and dilute the solution to the mark: 1 ml of this solution contains the equivalent of 0.2 mg of iron(II1) oxide.Prepare an iron(II1) spike solution by diluting one volume of standard iron(II1) solution with four volumes of water. In order to check the iron content of the ammonium iron(I1) sulphate, dissolve 1 g in water, oxidise the solution with bromine water and precipitate the iron as iron(II1) hydroxide with ammonia from the hot solution. Filter off the precipitate, ignite it at 1000 "C and weigh the resulting iron(II1) oxide (Fe,O,). PROCEDURE- To a clear, plastic 250-ml beaker add 100ml of saturated boric acid solution and to a similar 100-ml beaker add 50 ml of the same solution. To 0.5000 g of rock sample powder (ground to pass a 100-mesh sieve) in a platinum crucible of 60 to 80-ml capacity with a close-fitting lid, add a few millilitres of water and 10ml of 50 per cent.V/V sulphuric acid. Add more water approximately to half-fill the crucible and heat to boiling over a small flame protected from draughts. After a few seconds' boiling add 10 ml of hydrofluoric acid without interrupting the heating, replace the lid immediately and note the time at which boiling recommences. After 10 minutes' boiling iemove the flame and quickly rinse condensate from the lid into the crucible with boric acid solution from the 100-ml beaker. Then add further boric acid solution so as almost to fill the crucible and, with the aid of tongs, tip the contents of the crucible without delay into the 100 ml of boric acid solution in the prepared 250-ml beaker.Rinse any material left in the crucible into the beaker with the remainder of the boric acid solution in the 100-ml beaker and then transfer the whole contents into a 200-ml calibrated flask, rinsing with boiled, air-free water. Cool the flask to room temperature under running water and dilute to the mark with air-free water. Transfer a 20-ml aliquot of the solution into a 150-ml glass beaker and dilute it with water to about 75 ml. Stir the solution with a magnetic stirrer and, after adding 10 ml of ammonium thiocyanate solution and 2 ml of Lissapol NDB solution, titrate it against ferrocene solution from a burette graduated in 0-05-ml divisions. The end-point is indicated, in good white light, by the appearance of a blue colour without any trace of red iron(II1) thiocyanate colour.Add 1 ml of the iron(II1) spike solution and again titrate to the same end-point, this time noting the total volume of titrant ( X ml) added. Carry out the above titration procedure on 1 ml of iron(II1) spike solution to which has been added 1 ml of dilute sulphuric acid before dilution to 75 ml, and note the volume of titrant (s ml) required. In order to standardise the ferrocene solution transfer 10 ml of standard iron(II1) solution into a 150-ml beaker by means of a pipette and add permanganate solution dropwise until there is a small visible excess, then dilute the mixture to 75 ml and procezcl as above, including the addition of 1 ml of iron(II1) spike solution.Note the total volume of titrant (S ml) required. Set these items aside.May, 19741 CALCULATION- DETERMINATION OF IRON(III) IN SILICATE ROCKS 275 Iron(II1) oxide in sample, per cent. = 4 (X - s) s - s RESULTS AND DISCUSSION With the exception of PCC-1 and DTS-1, the rocks listed in Table I were analysed by the procedure given above. This standard procedure gave very low results (0.21 and 0-04 per cent. for PCC-1 and DTS-1, respectively) when applied to these very magnesium-rich rocks. Cloudy solutions were invariably obtained, but this did not seriously obscure the end-point of the titration and it was suspected that the white insoluble material, although essentially magnesium fluoride, might contain coprecipitated iron(II1) fluoride. This white solid could be brought into solution by adding a further 3 g of solid boric acid to the solution in the 200-ml flask and heating on the water-bath for about 45 minutes.After cooling and diluting to the mark, 20-ml aliquots were taken for titration in the usual way and a marked increase in the iron(II1) value was found (Table I). As this result could have been caused, at least in part, by atmospheric oxidation of iron(I1) in the solution, determinations were carried out on 100-ml aliquots of the same solution of iron(I1) by titration with dichromate solution. TABLE I RESULTS OBTAINED ON SOME STANDARD ROCKS BY USING THE PROPOSED METHOD r- Rock Granite G-1 . . . . Diabase (dolerite) W-1 Granite G-2 . . . . Granodiorite GSP-1 . . Andesite AGV- 1 . . Peridotite PCC-1 .. Dunite DTS-1 . . . . Basalt BCR-1 . . . . Operator 1 0-86 1.33 1.0s 1-63 4.39 2.06: 0.38: 3.3G Iron(II1) oxide, per cent. A 7 Operator Operator 2 3 Other methods 0.84 0-86 0*87* 1.47 1.38 1-40* 1.63 1-67 1.60t 2.06; 2.06: 3-49? 0.41: 0.371 0.S5t 1.06 1-06 1-01? 4.42 4.43 4-39? 3.45 3.47 3.45t * Flanagan.3 1 Abbey.4 The results of these determinations for PCC-1 and DTS-1 (5.39 and 7.21 per cent. of iron (11) oxide, respectively) do not support this atmospheric oxidation explanation. The results can be compared with those (5.24 and 7.23) in the most recent report by Flanagan3 on values for international reference samples and are considerably higher than those appearing in the earlier Flanagan report5 (4.94 and 6.79) and also higher than the figures proposed by Abbey4 (5.14 and 6-98) after a critical study of this earlier report.The tendency to obtain maximum values for both iron(II1) and iron(I1) after dissolving the magnesium fluoride precipitate indicates that both forms of iron are, to some extent, segregated in the precipitate. So far as the present authors are aware this is a possible source of error in the conventional iron(I1) determination that has not previously been reported and which should be borne in mind when rocks of very high magnesium content are being analysed. The blue colour of the ferricenium ion, when sufficiently intense, tends to obscure the end-point of the titration, which, for this reason, is most successful with relatively dilute solutions of iron(II1). In practice, under the conditions prescribed for the procedure, the most satisfactory working range of the method is up to about 5 per cent.of iron(II1) oxide in the sample. This, however, is not a serious limitation as most silicate rocks fall within this range. The proposed method will not avoid certain errors that affect the accuracy of the con- ventional iron(I1) determination. These errors are discussed in some detail by Maxwell6 and commonly include those due to the presence of carbonaceous matter and sulphides in the sample. However, the mineral pyrite, which is the most frequently encountered sulphide Variant of method, used for magnesium-rich rocks (see text).276 MURPHY, READ AND SERGEANT in silicate rocks, is not normally attacked by the acid mixture used in this method and so will not affect the iron(II1) titration. It would, if not allowed for, be recorded as iron(II1) if that value was calculated by difference from the total iron and iron(I1) results. The use of a spiking procedure in the titration serves to correct for any iron(II1) in the titration reagents. It also allows the first tentative end-point to be over-run and the final definitive end-point to be judged with greater precision. This paper is published by permission of the Government Chemist and the Director of the Institute of Geological Sciences. REFERENCES 1. 2. 3. 4. 5. 6. Pratt. J. H., Amer. J . Sci., 1894, 48, 149. Wolf, L., Franz, H., and Hennig, H. Z., 2. Chemie Lpz., 1961, 1, 220. Flanagan, F. J., Geochim. Cosmochim. Acta, 1973, 37, 1189. Abbey, S., Can. Spectrosc., 1970, 15, 10. Flanagan, F. J., Geochim. Cosmochim. Acta, 1969, 33, 81. Maxwell, J. A., “Rock and Mineral Analysis,” Interscience Publishers, New York, 1968, pp. 203 Received November 29th, 1973 Accepted January 7th, 1974 to 210.

ISSN:0003-2654    

DOI:10.1039/AN9749900273

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, May, 1974, Vol. 99, $$. 277-284 277 Determination of Rubidium, Caesium, Barium and Eight Rare Earth Elements in Ultramafic Rocks by Neutron-activation Analysis* BY A. 0. BRUNFELT, I. ROELANDTST (iMineralogica1-Geological Museum, University of Oslo, Sarsgt. 1, Oslo 5, Norway) AND E. STEINNES (Institutt for Atomenergi, Isotope Laboratories, Kjeller, Norway) A method for the simultaneous determination of rubidium, caesium, barium and eight rare earth elements in rocks by neutron-activation analysis is described. The method is suitable for use with samples with particularly low abundances of the elements concerned. After fusion with alkali and determination of rubidium and caesium in the supernatant liquid, the hydroxide fraction is carried through an anion-exchange step in chloride media and extracted with tri-n-butyl phosphate in order to remove inter- fering nuclides.Subsequently, barium, strontium and the rare earths are separated into three groups by mixed-solvent anion exchange. The method is shown to be suitable for application to ultramafic rocks. Results for rubidium, caesium, lutetium, ytterbium, terbium, europium, samarium, neodymium, cerium and lanthanum in dunite DTS-1 are pre- sented. Granite (G-2) was also analysed in order to assess the accuracy of the method. As mafic and ultramafic rocks are presumably derived from the upper mantle of the Earth' they provide the material upon which geochemical inferences of data for the elemental com- position and of homogeneity of the mantle are based. Even though these rocks, especially the ultramafic rocks, are deficient in elements such as the alkali, alkaline earth and rare earth elements, the abundances of elements give information concerning the geochemical processes that affect these rocks.Ultramafic rocks with their high content of ferromagnesium minerals show low concen- trations of aluminium, calcium and the alkali elements. Moreover, a number of trace elements are often present in concentrations that are 2 to 4 orders of magnitude lower than in most common silicate rocks. For these reasons, methods that are otherwise of wide use in rock analysis may not be applicable to this class of rocks. In terms of neutron-activation analysis, instrumental methods with germanium (lithium) detectors are of limited use in this case, and even methods involving radiochemical separations will often be inapplicable to ultramafic rocks, although they may be useful for other types of rocks.In this paper, a radiochemical method based on neutron-activation analysis and german- ium (lithium) y-ray spectrometry for the simultaneous determination of rubidium, caesium, barium and eight rare earth elements, which has been specially designed for application to ultramafic rocks, is described. Dunite (DTS-1) was chosen as the test rock and basalt, BCR-1, was used as the comparative standard for the activation-analysis procedure. For assessment of accuracy, granite (G-2) was also analysed. These three rocks are currently being distributed by the U.S. Geological Survey. The method described below for the simultaneous determination of these elements is based on decomposition with alkali followed by leaching with water.The alkali elements are determined in the supernatant liquor, while the hydroxide fraction, which contains most of the strontium, barium and rare earth elements, is dissolved in hydrochloric acid and passed * Presented at the 3rd Symposium on Recent Developments in Neutron Activation Analysis, Churchill t P:esent address : Laboratory of Geology, Petrology and Geochemistry, University of Likge, 7 Place du @ SAC and the authors. College, Cambridge, 2nd to 4th July, 1973. XX Aout, 4000 Likge, Belgium.278 [Analyst, Vol. 99 through an anion-exchange column in order to separate a number of interfering radionuclides, the most important being cobalt-60, iron-59, antimony-122 and antimony-124.The effluent is subsequently extracted with tri-n-butyl phosphate in order to remove the rather large amounts of scandium-46 that are not eliminated by the preceding steps. After evaporating the resulting aqueous phase to dryness, the residue is dissolved in a nitric acid - methanol mixture, and a chromatographic anion-exchange separation is performed so as to fractionate three groups of elements : lutetium and ytterbium ; strontium, terbium, barium, europium, samarium and gadolinium ; and neodymium, cerium and lanthanum. Desai, Krishnamoorthy Iyer and Sankar Dasl and Brunfelt and Steinne~~9~ used similar anion-exchange separation methods from mixed solvents as described in this work for the determination of some of the rare earth elements by neutron-activation analysis with sodium iodide (thallium) y-ray spectrometry.The conditions chosen for these methods as well as for that described here were based on the anion-exchange studies of rare earth elements made by Faris and Warton,* Korkisch and J a n a ~ e r , ~ and Stewart, Bloomquist and Faris.6 In respect of the present work, it should be noted that Higuchi, Tomura and Hamaguchi’ used cation exchange with EDTA in order to separate the rare earth elements into one light fraction (lanthanum - terbium) and one heavy fraction (holmium - lutetium) in radiochemical neutron-activation analysis of rocks combined with germanium (lithium) y-ray spectrometry. The procedure for automatic radiochemical separations in activation analysis developed by Samsahls also includes the division pf the rare earth elements into two main groups by ex- tract ion with h ydrox ydi (2-e t hylh ex yl) phosphoric acid.EXPERIMENTAL REAGENTS- Anion-exchange resin for removal of interfering elements-Before use, pre-condition the resin Dowex 1-X8 (100 to 200 mesh; chloride form) with 9 M hydrochloric acid. Anion-exchange resin for fractionation of barium, strontium and rare earth elements-Convert the resin Dowex 1-X8 (100 to 200 mesh; chloride form) into the nitrate form by treatment with 7 M nitric acid. Before use, pre-condition the resin with the mixture nitric acid - methanol BRUNFELT et al. : DETERMINATION OF Rb, cs, Ba AND RARE EARTHS (10 + 90 V / V ) . Tri-n- butyl phosp hate-Analytical-reagent grade. Carrier solutions-Prepare stock solutions of mixed carriers from the elements corres- ponding to about 1 mg for rubidium, caesium, terbium and thulium and about 0.1 mg for strontium, barium, lanthanum, cerium, neodymium, samarium, europium, gadolinium, ytterbium and lutetium.Add 0.1 mg of scandium and cobalt as holdback carriers. APPARATUS- Anion-exchange columns-Load columns of 1 cm i.d. with the pre-conditioned Dowex 1-X8 resin to give a resin bed of 14 cm for barium, strontium and rare earth fractionation, and 8.5 cm for removal of interfering elements. Counting equipment-The detector used for this study was a Canberra co-axial germanium (lithium) detector with a resolution of 2.5 keV (full width at half maximum; FWHM) and a peak to Compton ratio of 23: 1 for the 1331 keV peak of cobalt-60; the relative efficiency was 5.7 per cent.[as compared with that of a 3 x 3-inch sodium iodide (thallium) crystal at 1331 keV]. The associated electronics consisted of a Hewlett-Packard 200 MHz ADC inter- faced to a Nord-1 digital computer (16 K, 16 bit). A precision pulser was applied for the determination of counting time as well as for dead-time correction (as described by Andersg). IRRADIATION- Accurately weighed rock samples (about 200 mg) were wrapped in 30 x 30-mm sheets of aluminium foil. Irradiations were carried out in the JEEP-I1 reactor (Kjeller, Norway) at a thermal neutron flux of about 1.7 x 1013 neutrons cm-2 s-l. RADIOCHEMICAL PROCEDURE- By pipette transfer exactly 1 ml of the mixed carrier solution into a nickel crucible and evaporate it carefully to dryness under an infrared lamp. Unwrap the aluminium foil and pour the irradiated sample quantitatively into the crucible containing the carriers plus 2 g of sodium hydroxide pellets and 0.5 g of sodium peroxide.Heat the crucible with its cover over an electrothermal burner for 5 After irradiation, allow the samples to decay for 5 days.May, 19741 I N ULTRAMAFIC ROCKS BY NEUTRON-ACTIVATION ANALYSIS 279 minutes and wait 30 s before removing the cover. Then transfer the crucible carefully into a 250-ml beaker containing 20 ml of water, digest the resulting fusion cake on a water-bath, allow the mixture to cool at room temperature and, after it has stood for at least 2 hours, centrifuge the hydroxide precipitate. Wash the precipitate twice with water and combine the supernatant liquids for rubidium and caesium determination.For the ultramafic rocks with low contents of rubidium and caesium, precipitate these elements by adding 15 ml of a 1 M sodium hydroxide solution containing 1 per cent. of sodium tetraphenylboron. Dissolve the hydroxide precipitate in 10 ml of 9 M hydrochloric acid and heat the solution for a few minutes on a water-bath so as to obtain a clear solution. After cooling it, transfer this solution to the pre-equilibrated Dowex 1-X8 anion-exchange column and collect the effluent in a 50-ml polyethylene screw-capped bottle. Wash the centrifuge tube with 5 ml of 9 M hydrochloric acid and transfer the washings to the column after the sorption solution has reached the resin bed.Repeat this procedure with three further 5-ml portions of 9 M hydrochloric acid. Transfer the collected effluents into a 250-ml separating funnel, extract them with 25 ml of tri-n-butyl phosphate and discard the organic phase. Transfer the aqueous phase (containing strontium, barium and rare earth elements) into a 100-ml beaker, evaporate it to dryness on a water-bath and dissolve the residue in 5 ml of 15.6 M nitric acid. After complete dissolution, dilute the solution with 45 ml of methanol and pass this solution through the prepared Dowex column at a flow-rate of approximately 0.8 ml min-l in order to remove the residual activity of scandium-46. After the sorption step, elute the column with 200ml of 7-8 M nitric acid-methanol (15 + 85) at the flow-rate of 0-6 ml min-l and collect this fraction in a 250-ml polyethylene screw-capped bottle for the determination of lutetium and ytterbium.Continue eluting with a 200-ml portion of 7-8 M nitric acid - methanol (45 + 55) at the flow-rate of 0.5 ml min-1. This fraction is analysed for strontium, terbium, thulium, barium, europium, samarium, and gadolinium. Finally, elute neodymium, cerium and lanthanum with 100 ml of water at the flow-rate of 0.7 ml min-l and again collect the effluent in a screw-capped polyethylene bottle. The radiochemical group separation scheme is summarised in Fig. 1. Irradiated sample ( 200 mg 1 Fuse with NaOH + Na202 in presence of carriers. Dissolve melt in H20 I I Supernatant Precipitate with tetraphenylboron I Rb, Cs I Residue Dissolve in 9 M HCI Solution Pass through an 85-mm column of Dowex 1-X8,100 to 200 mesh, diameter 10 mm Effluent solution Extract with tributyl phosphate I I I Aqueous phase Evaporate and dissolve residue in 5 ml of HNO3 + 45 ml of CHBOH I Solution Pass through a 140-mm column of Dowex 1-X8, 100 to 200 mesh, diameter 10 mrn and discard sorption eluate.Elute column with: A, 200 ml of 7-8 M HNOB-CH~OH (15+85 V / V ) -+ Lu, Yb, Sr; B, 200 ml of 7.8 M HNO3 -CH30H (45 -+ 55 V / V ) + Sr, Tb, Ba, Eu, Sm, Gd; C, 100 ml of H20 + Nd, Ce, La Fig. 1. Radiochemical group separation scheme I Organic phase Discard280 BRUNFELT et al. : DETERMINATION OF Rb, Cs, Ba AND RARE EARTHS [Analyst, Vol. 99 ACTIVITY MEASUREMENTS- The activities of the separated fractions were assayed by germanium (lithium) y-ray spectrometry and the peak area calculations were performed according to Cove1110 or Ster- linski.11 Fig.2 shows the elution curves obtained in an initial experiment carried out to establish the working conditions to be used in this part of the procedure. In Fig. 3 the y-spectra of the eluted fractions for BCR-1 are shown. The relevant nuclear data are given in Table I. 80 w 60- aJ n 4-- HN03- 7.8 M HNOS-CH3OH [x M HNOB-CH~OH GO (1 0 i- 90 v / v ) NdtCei-La (1 5 t 85 v/v) (45 -t 55 v/v) loo- r- CH,OH - ii I i I 1 11 sc DETERMINATION OF CHEMICAL YIELD- The chemical yield was determined by a reactivation technique as described by Brunfelt and Steinnes,14 that of rubidium and caesium being found to be about 90 per cent, In experimental work to be described elsewhere,l5 the chemical yield, after carrying out the anion-exchange separation and extraction with tri-n-butyl phosphate following the procedure described above, has been determined on more than forty different rock samples.From these experiments, no significant fractionation of the rare earth elements during these steps was found. The chemical recovery varied from 75 to 85 per cent., while the variation of the individual rare earth element carriers in the same rock was less than 3 per cent. in each instance. The recovery, including the mixed-solvent separation, of barium and strontium was found to be better than 90 per cent. This high yield may be explained in part by the presence of carbonate impurity in the sodium hydroxide used ; otherwise, a somewhat higher loss would be expected in the fusion steps owing to partial leaching of barium and strontium into the supernatant liquor.Counts per channel A 0 0 0 0 4 -L 0 0 0 -..0 0 - 0 A -L -. 0 0 0 0 0 o--L 0 0 0 0 0 0 0 -L 0 -. 0 1 I I I \\I I I I \ \ I I I I I Xo La 328.8 lL0 La 432.5 1282 BRUNFELT et al. : DETERMINATION OF Rb, Cs, Ba AND RARE EARTHS [Analyst, Vol. 99 TABLE I DATA FOR NUCLEAR REACTIONS INVOLVED IN THE ANALYSIS Radionuclide Energy* Element measured used/keV Hali-lifet Rb 86Rb 1078.8 18.66 days c s 134cs 605.0; 797-0 2.07 years Sr E5Sr 5 14.0 64 days Ba 131Ba 496.3 11.5 days La 140La 487.0; 1596.6 40.27 hours Ce 141Ce 145.5 32.5 days Nd la7Nd 531.0 11.06 days Sm 153Sm 103.2 47.1 hours Eu 152E~ 344.2; 1408.1 12.2 years Gd 153Gd 103.2 236 days Tb l60Tb 298.6; 879.3 73.0 clays Yb 160Yb 177.0 30.6 days Yb 175Yb 396.1 101 hours Lu 177Lu 208.3 6.75 days * Results taken from Filby, Davis, Shah, Wainscott, Haller and Cassatt.12 t Results taken from Adams and Dams.13 RESULTS AND DISCUSSION The data reported on the ultramafic rock, DTS-1, so far are scarce.Table I1 shows the experimental results in duplicate obtained by the present method calculated relative to BCR-1. As no recommended values for rare earths in BCR-1 had been issued, results from the recent work by Brunfelt and Steinnesl41ls were taken as assigned values. Most of these results were obtained during a systematic investigation of RCR-1 as a standard for analysis of lunar samples. The values obtained for G-2 are also included in Table I1 and are in most instances in good agreement with those from the compilation of Flanagan17 as well as with other reported neutron-activation data,2,3,7,18,19-21 In Table 111, the mean values obtained for DTS-1 are compared with other available neutron-activation results.For rubidium, our value of 0.057 p.p.m. is slightly higher than the upper limit (0.043 p.p.m.) reported by Rey, Wakita and TABLE I1 RESULTS FOR ELEMENTS DETERMINED IN DUNITE DTS-1 AND GRANITE G-2 (p.p.m.) Elements Rb cs Sr Ba La Ce Nd Sm Eu Gd Tb Yb Lu r Concentration used for BCR-1 standard16 0.97 50 312 580 23.7 53 30.5 6-6 1-94 7.5 0.96 3.21 0.64 DTS-1 A \ Experimental 0.060, 0.055 0.057 0.0052, 0.0053 0.0053 0.70 0.70 0-036, 0.027 0.032 0.091, 0.075 0.083 0.025, 0.044 0.035 0.0064, 0.0058 0.0061 0.0013, 0.0019 0.0016 0*0007, 0.0007 0.0007 0.018, 0.010 0.014 0.0025, 0.0024 0.0025 results Mean value - - - - - - G-2 - Experimental results 194, 181 1.44, 1.37 484, 482 1502, 1578 99, 81 141, 157 61,54 7.27, 7.35 1-33, 1.36 5.41, 4.97 0.43, 0.44 0.69, 0.76 0.108, 0.115 Mean value 188 483 1540 90 149 57 1.41 7.31 1.35 5-19 0.44 0.73 0.112 Schmitt21; the value given by Gangadharam and Reddy22 is, however, much higher (0.4p.p.m.).For caesium our value of 0.0053 p.p.m. is in good agreement with the value of 0.006 p.p.m. obtained by neutron-activation analysis recently reported by Muller23 and 0-008 p.p.m. by Heier and Br~nfelt,~4 although other results reported are significantly higher, as shown in Table 111. For the rare earth elements, the results obtained in this work are in good agreement with those reported by Rey et but those reported by Higuchi, Tomura, Onuma and Hama-May, 19741 I N ULTRAMAFIC ROCKS BY NEUTRON-ACTIVATION ANALYSIS 283 guchi26 are systematically lower by a factor of about 2.The values obtained by Garmann, Brunfelt, Finstad and HeierZ7 are in good agreement for lanthanum, cerium and samarium, while those for europium, ytterbium and lutetium are systematically higher by a factor of about 2. 1.0 0.1= 0.01 0~001 TABLE I11 - - - - 0 0 if----- A A -@-.. / /. ?+= .'A, / A // o A *--\ \ a / 0 0 A '\// r 0 A - 0 - - Pr Eu Dy Er Yb 0.098 0.092 0.089 0.085 0.105 \ I \ / / I La Ce Nd Sm Gd Tb Ho Trn Lu COMPARISON OF REPORTED NEUTRON-ACTIVATION RESULTS IN DUNITE DTS-1 (p.p.m.) Rb cs Ba La Ce Nd Sm Eu Gd Tb Yb Lu Previous results 0.0322, 0*0062', 0-00824, 0*0225 <0*04321, 0.425, < l 2 5 .. O*02G2l, 0*03626, 0*0327 . . 0.07621, 0-03926, 0.0827 . . <0*02421 . . . . . . 0*005021, 0.002826, 0*00627 0.001 121, 0.000726, 0 ~ 0 0 3 ~ ~ <0.01021 . . . . . . <0.000321, 0*000326 . . 0*01821, 0.00572s, 0*0427 0.003121, 0.001426, 0*00527 Present results . . . . . . 0.057 . . . . . . 0.0053 .. . . . . 0.032 . . . . . . 0.083 . . .. .. 0.035 . . . . .. 0.006 1 .. . . .. 0.0016 . . . . . . 0.0007 . . . . . . 0.014 . . .. * . 0.0025 0.70 - .. .. . . The distribution pattern of rare earth elements relative to chondrites of DTS-1 is plotted in Fig. 4, the chondritic values used being taken from Haskin, Wildeman and Haskin.28 As previously mentioned,*l it appears that the logarithmic graph of normalised abundances exhibits smooth curves, and a marked inflection point at terbium or dysprosium is clearly seen.It should be noted that our results approximate more closely to the average results previously published than those obtained by Higuchi et aLZ6 and are less scattered about a smooth curve than those found by Rey et ~ ~ 1 . ~ 1 Although not attempted in this work, the method presented could easily be extended t o the simultaneous determination of the additional elements cobalt, copper, iron, gallium, tungsten, cadmium, arsenic and antimony by fractional elution of the elements removed on the first anion-exchange column, following the procedure given by Johansen and S t e i n n e ~ . ~ ~ Fig. 4. Chondrite-normalised pattern for the dunite DTS-1: a, this work; 0, Higuchi et aZ.26; A, Rey, Wakita and Schmitt21; and 0, Garmann, Brunfelt, Finstad and Heier27284 BRUNFELT, ROELANDTS AND STEINNES It should also be useful for other categories of geological samples that contain the elements concerned in low concentrations, such as ores and special minerals.The separation steps preceding the mixed-solvent separation would in most instances remove most of the interfering nuclides. The method may also be used for most common silicate rocks, as exemplified by the analysis of granite G-2, in most instances of which, however, satisfactory methods involving less work are available. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. REFERENCES Desai, H.B., Krishnamoorthy Iyer, R., and Sankar Das, M., Talanta, 1964, 11, 1249. Brunfelt, A. O., and Steinnes, E., Chem. Geol., 1967, 2, 199. -- , Analyst, 1969, 94, 979. Faris, J. P., and Warton, J. W., Analyt. Cheun., 1962, 34, 1077. Korkisch, J., and Janauer, G. E., Talanta, 1962, 9, 957. Stewart, D. C., Bloomquist, C. A. A., and Faris, J. P., Rep. Congr. Atom. Energy Commn U.S., Higuchi, H., Tomura, K., and Hamaguchi, H., J . Radioanalyt. Chem., 1970, 5, 207. Samsahl, K., Sci. Total Environ., 1972, 1, 65. Anders, 0. U., Nucl. Instrum. Meth., 1969, 68, 205. Covell, D. F., Analyt. Chem., 1959, 31, 1785. Sterlinski, S., Ibid., 1968, 41, 1995. Filby, R. H., Davis, A. I., Shah, K. R., Wainscott, G. G., Haller, W. A., and Cassatt, W. A., Rep. Adams, F., and Dams, R., J . Radioanalyt. Chem., 1969, 3, 99. Brunfelt, A. O., and Steinnes, E., Ibid., 1973, 13, 1 1 . Roelandts, I., Thesis, University of Liege, in preparation. Brunfelt, A. O., and Steinnes, E., Talanta, 1971, 18, 1197. Flanagan, F. J., Geochim. Cosmochim. Acta, 1969, 33, 81. Green, T. H., Brunfelt, A. O., and Heier, K. S., Ibid., 1972, 36, 241. Gordon, G. E., Randle, K., Goles, G. G., Corlies, J. B., Beeson, M. H., and Oxley, S. S., Ibid., 1968, Morrison, G. H., Gerard, J. I., Travesi, A., Currie, R. L., Peterson, S. F., and Potter, N. M., Analyt. Rey, P., Wakita, H., and Schmitt, R. A., Analytica Chim. Acta, 1970, 51, 163. Gangadharam, E. V., and Reddy, G. R., Radiochinz. Acta, 1969, 11, 90. Miiller, O., Earth Planet. Sci. Lett., 1970, 8, 283. Heier, K. S., and Brunfelt, A. O., Ibid., 1970, 9, 416. Tomura, K., Higuchi, H., Takashi, H., Onuma, N., and Hamaguchi, H., Analytica Cka’m. Acta, Higuchi, H., Tomura, K., Onuma, N., and Hamaguchi, H., quoted in reference 21. Garmann, L. B., Brunfelt, A. O., Finstad, K. G., and Heier, K. S., to be published. Haskin, L. A., Wildeman, T. R., and Haskin, M. A., J . Radioanalyt. Chem., 1968, 1, 337. Johansen, O., and Steinnes, E., Talanta, 1970, 17, 407. ANL-6999, 1965. WSUNRC-97(2), Washington State University, Nuclear Radiation Center, 1970. 32, 369. Chem., 1969, 41, 1633. 1968, 42, 623. Received August 6th, 1973 Accepted November 30th, 1973

ISSN:0003-2654    

DOI:10.1039/AN9749900277

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, May, 1974, Vol. 99, $9. 285-295 285 The Determination of Iridium and Ruthenium in Rhodium Sponge by Solvent Extraction Followed by Atomic-absorption Spectrophotometry* BY M. A. ASHY AND J. B. HEADRIDGE (Department of Chemistry, The University, Sheffield, S3 7HF) A method is described for the determination of trace amounts of iridium and ruthenium in rhodium sponge. The sponge is dissolved by treatment with concentrated hydrochloric acid and sodium chlorate in a sealed glass tube a t 250 "C. Chloro-complexes of rhodium and other noble metals are produced. From such a solution, 2 M in hydrochloric acid, iridium and ruthenium are extracted into chloroform that is 1 per cent. rn/V in methyl- triphenylphosphonium chloride ; rhodium is not extracted. The organic phase is evaporated to dryness and the residue dissolved in acetonitrile containing lithium perchlorate, which is the interference suppressant for the subsequent determination of iridium and ruthenium by atomic-absorption spectrophoto- metry. Osmium can be determined in rhodium sponge in a similar manner if precautions are taken so as to prevent loss of osmium when the tube is opened.The results are in good agreement with those obtained by emission spectrography and the limits of detection for iridium and ruthenium in the sponge are 7 and 4.5 pgg-l, respectively. A COMPLEX series of chemical operations is required for the preparation of pure rhodium sponge from ore samples. In the final stages of the refining process rhodium sponge can be produced that contains trace amounts of other noble metals.The efficiency of the process can be ascertained by determining the iridium and ruthenium contents of the sponge. When this work was started in 1972, the detection limits for the spectrographic determination of platinum, ruthenium, osmium and iridium in rhodium sponge were 5, 30, 30 and 40 pg 8-1, respectively. It was felt that the detection limits for iridium, osmium and ruthenium could be improved by solvent extraction of these trace elements from an aqueous solution of the rhodium metal into an organic solvent, and subsequent spraying of the organic solvent into the flame of an atomic-absorption spectrophotometer. Headridge and co-workers have found that the detection limits for the determination of aluminium,l antimony,2 bismuth3 and tin4 in steels following solvent-extraction procedures are improved by factors of thirty to sixty compared with methods involving direct atomic- absorption spectrophotometry in aqueous solutions.Slavin5 states that the concentrations of iridium, osmium and ruthenium that produce 1 per cent. absorption are 8, 1 and 0.3 pg m1-1, respectively, using the atomic lines at 208.9 nm, 290-9 nm and 349.9 nm, respectively. Iridium and ruthenium were determined with the air - acetylene flame but osmium was determined with the nitrous oxide - acetylene flame. A search of the chemical literature has revealed that 10 to 150 pg of iridium can be separated from 1 to 2 mg of rhodium by means of a single extraction.6 The extraction was carried out by adding 3 ml of a 2 per cent.solution of tetraphenylphosphonium bromide in water to 15 to 20 ml of iridium(1V) in approximately 0.1 M hydrochloric acid and shaking the mixture for 3 minutes with 10 to 25ml of chloroform. An extractable ion-association complex is formed with the doubly charged hexachloroiridate(1V) ion, but not with the triply charged hexachlororhodate(II1) ion. F0k7 has also shown that iridium(1V) can be extracted quantitatively from 0.1 M hydro- chloric acid with an equal volume of chloroform containing 1 per cent. m/V of tetraphenyl- arsonium chloride. The hexachlororhodate(II1) anion is not extracted. He also states that palladium(I1) and (IV), osmium(IV), ruthenium(1V) and platinum(1V) also form extractable compounds with the tetraphenylarsonium cation. * Presented a t a meeting of the Society for Analytical Chemistry and Analytical Division of the Chemical Society, Sheffield, July 12th and 13th, 1973.0 SAC and the authors.286 ASHY AND HEADRIDGE: DETERMINATION OF Ir AND Ru IN Rh SPONGE [Analyst, Vol. 99 If rhodium sponge containing trace amounts of osmium, ruthenium, iridium, palladium and platinum could be brought into solution in hydrochloric acid such that the oxidation states of the chloro-species were rhodium(III), ruthenium( IV), osmium(IV), iridium( IV), palladium(I1) or (IV) and platinum(IV), then it seemed likely that trace amounts of iridium, and possibly also ruthenium, osmium, palladium and platinum, could be readily separated from rhodium by solvent extraction. Dissolution of the noble metals can be achieved by reaction at 250 "C in a sealed tube with chlorine produced from sodium chlorate and concentrated hydrochloric acid.* This method of dissolution was used as chloro-complexes of the noble metals are produced.A consideration of the inorganic chemistry of the noble metals indicated that treatment with sodium chlorate and concentrated hydrochloric acid at 250 "C should produce hexachloro- rhodate( 111) , hexachloroiridate( IV) , hexachloroplatinate( IV) , probably mainly hexachloro- ruthenate(1V) and hexachloropalladate(IV), and probably osmium(VII1) tetroxide together with the dioxotetrachloroosmate(V1) anion, [OSO,C~,]~-.~~~O After diluting the solution to make it 2 M in hydrochloric acid (before the solvent extraction), the osmium will probably be present as a mixture of osmium(V1) and osmium(IV), particularly if the solution is allowed to stand at some stage.Both osmium tetroxide and the dioxotetrachloroosmate(V1) ion will oxidise the chloride ion in hydrochloric acid to chlorine, the former much more rapidly than the latter. In fact, when standard solutions of rhodium, iridium, ruthenium and osmium in 6 M hydrochloric acid were prepared by the above treatment with concentrated hydrochloric acid and sodium chlorate, and these solutions diluted to make the hydrochloric acid concentration 2 M, it was found that the ultraviolet and visible absorption spectra of the solutions indicated that they contained predominantly the hexachloro-complexes of rhodium(III), iridium(IV), ruthe- nium(1V) and osmium(1V). In this paper, a method is described for the dissolution of rhodium metal with concen- trated hydrochloric acid and sodium chlorate in a sealed glass tube at 250 "C, and the removal of iridium(1V) and most of the ruthenium(1V) from a solution of the rhodium sponge in 2 M hydrochloric acid by extraction into chloroform containing methyltriphenylphosphonium chloride.The chloroform layer is evaporated to dryness, the residue taken up in acetonitrile, lithium perchlorate added as an interference suppressant and the solution made up to 5 ml with acetonitrile. The concentrations of iridium and ruthenium in this solution, and hence in rhodium sponge, are determined by use of atomic-absorption spectrophotometry. The method should also be suitable for the determination of trace amounts of osmium, if present, in rhodium sponge provided that care is taken to prevent any loss of trace amounts of volatile osmium tetroxide when the tube containing the treated rhodium sponge is opened.EXPERIMENTAL APPARATUS- Pyrex glass tubes, 9 mm in internal diameter, 1.5 mm in wall thickness and sealed at one end, were prepared. These were drawn out near to the open end to form a constriction in the tube with a minimum internal diameter of 2 to 3mm. The tubes could be sealed by heating these constrictions with an oxy-propane torch. When 100 mg of noble-metal sponge were to be dissolved, the sealed tube was 25 cm long, but for samples of 500 mg, longer tubes (50 cm) were used. Before heating was started, the tube was enclosed in a steel casing with a steel screw-top.The steel casing contained 20 g of calcium carbonate powder in order to neutralise acid, which would escape from the sealed tube in the unlikely event of the tube exploding. A small escape hole for gases was also drilled through the side of the casing. Atomic-absorption spectrophotometric measurements were made with Unicam SP90 Series 1 and Unicam SP1900 instruments fitted with air - acetylene or nitrous oxide - acetylene burners. REAGENTS- Hydrochloric acid, sp. gy. 1-18-Analytical-reagent grade. Sodium chlorate-General-purpose reagent. Lithium @erchlorate-General-purpose reagent, dried at 100 "C. Chloroform-Analytical-reagent grade.May, 19741 BY SOLVENT EXTRACTION AND ATOMIC-ABSORPTION SPECTROPHOTOMETRY 287 Acetonitvile-General-purpose reagent, redistilled at 82 "C before use.~ethyltriphenylphosphonium chloride-This reagent was prepared from methyltriphenyl- phosphonium bromide (Koch-Light Laboratories Ltd.) by passing an aqueous solution of the salt through a column of De-Acidite FF resin in the chloride form. The effluent was evaporated at 100 "C nearly to dryness and the solid was then dried in an oven at 90 "C for 6 hours. Rhodium, iridium, ruthenium and osmium sponges-These were of Specpure quality and were supplied by Johnson Matthey Chemicals Ltd. Potassium hexachloropnlladate ( I V ) and potassium hexaclzloroPlatiizate(rV)-These salts were prepared from palladium( 11) chloride and platinum (IV) chloride, respectively. Standard iridium solution A (1000 pg ml-l)-Weigh accurately 100 mg of iridium sponge into a small glass tube, transfer it quantitatively into a Pyrex tube (25 cm long when sealed) and add 5 ml of concentrated hydrochloric acid (sp. gr.1-18>. Immerse the tube in liquid nitrogen or cardice until the contents just start to solidify and then add 0.2 g of sodium chlorate crystals. Remove the tube from the coolant, seal it and allow it to attain room temperature. Place it in a steel casing fitted with a screw-cap and heat in an oven at 250 "C for 6 hours so as to dissolve all of the sponge. Allow the tube to cool to room temperature and cool it again in liquid nitrogen or cardice until the contents just start to solidify. Open the tube by scratching the glass near to the sealed end and giving it a sharp tap. (As the tube breaks, chlorine, under pressure in the tube, may escape.) Transfer the contents into a 100-ml calibrated flask containing 45 ml of concentrated hydrochloric acid, rinse the tube several times with distilled water and make the volume up to the mark so that the concen- tration of hydrochloric acid in the final solution is approximately 6 M.Standard iridium solution B (100 pg ml-l)-Dilute 10 ml of solution A to 100 ml with 2 M hydrochloric acid immediately before use. Standard ruthenium solutions A (1000 pg ml-l) and B (100 pg ml-1)These solutions are prepared in a similar way to the iridium solutions. Standard rhodium solution (5000 pg ml-l)-This is prepared in a similar way to the above solutions, except that 500 mg of rhodium sponge, 10 ml of concentrated hydrochloric acid and 0.6 g of sodium chlorate are used in a 50 cm long sealed tube.Standard osmium solution (1000 pg mZ-l)-This is prepared in a similar way to the above solutions, but before opening the tube, the contents are solidified completely by immersing the tube in liquid nitrogen. The opened tube is then placed upright in a 100-ml beaker and held clear of the sides of the beaker by means of tongs fitted with glass end-pieces. From a beaker immersed in cardice, 45 ml of cooled, concentrated hydrochloric acid are then added to the tube so as to fill it completely, the excess of acid being added to the 100-ml beaker. A long test-tube is inverted over the glass tube containing the osmium such that the open end of the tube is immersed in the acid in the beaker. This outer tube is used in order to direct into the beaker any solution that may spray from the inner tube as the solid chlorine melts and then vaporises.The beaker and tubes are allowed to stand until they attain room temperature; as the liquid in the tube expands, it overflows into the beaker. Rinse the outer test-tube with distilled water and collect the rinsings in the beaker. Then transfer approximately 5 ml of liquid from the tube into the beaker with a narrow diameter pipette and rinse the outside of the tube with distilled water, collecting the rinsings in the beaker. Add the contents of the tube to the beaker and rinse the tube with distilled water. Finally, transfer the contents of the beaker into a 100-ml calibrated flask and dilute to the mark. By use of this method no loss of volatile osmium tetroxide can occur.EXTENTS OF EXTRACTION OF THE NOBLE METALS- Rhodium-Two extractions with 10-ml volumes of chloroform that was 1 per cent. m/V in methyltriphenylphosphonium chloride were carried out on 40 ml of 2 M hydrochloric acid containing 1000 pg of rhodium, prepared from standard rhodium solution. The aqueous solution was diluted to 50 ml. The amount of rhodium in the aqueous phase was determined by atomic-absorption spectrophotometry. Palladium and platinum-A similar procedure to that for rhodium was applied to solutions of palladium(1V) and platinum(1V) in 2 M hydrochloric acid.288 ASHY AND HEADRIDGE : DETERMINATION OF Ir AND Ru IN Rh SPONGE [Analyst, Vol. 99 Iridiwm-Two extractions were carried out with 60-ml volumes of chloroform that was 1 per cent. m/V in methyltriphenylphosphonium chloride on 240 ml of 2 M hydrochloric acid containing 3000 pg of iridium, prepared from standard iridium solution A.The aqueous phase was evaporated to dryness and the residue taken up in 5 ml of 2 M hydrochloric acid. The concentration of iridium was determined in this solution by atomic-absorption spectro- photometry. Ruthenium-A similar procedure was applied to a solution of ruthenium in 2 M hydro- chloric acid, prepared from standard ruthenium solution A. Osmium-A similar extraction procedure was applied to a solution of osmium in 2 M hydrochloric acid prepared from standard osmium solution. As solutions in 2 M hydrochloric acid prepared from the standard osmium solution absorb strongly at 372 nm, solution absorp- tion spectrophotometry was used in order to determine the concentration of osmium remaining in the aqueous phase.The extents of extraction of the above six elements into chloroform are shown in Table I. TABLE I EXTENTS OF EXTRACTION OF NOBLE METALS INTO CHLOROFORM Element Amount extracted, per cent. Rhodium . . .. . . 0 Iridium . . .. .. .. 100 Ruthenium . . .. .. 90 Osmium.. .. .. .. 100 Palladium . . .. .. 43 Platinum . . . . .. 93 TENTATIVE METHOD FOR THE DETERMINATION OF IRIDIUM AND RUTHENIUM I N RHODIUM SPONGE- A tentative method for these determinations could now be devised and was as follows. To 0.5 g of rhodium sponge contained in a Pyrex tube, add 10 ml of concentrated hydrochloric acid. Cool the tube and its contents in liquid nitrogen and add 0.6 g of sodium chlorate, then seal the tube and heat it for a minimum of 6 hours at 250 "C.Open the tube and transfer the contents quantitatively into a 250-ml separating funnel, then add 10 ml of concentrated hydrochloric acid and dilute the solution to 120 ml. Add 30 ml of chloroform that is 1 per cent. m/V in methyltriphenylphosphonium chloride to the mixture in the funnel and shake it vigorously for 2 minutes. Allow the layers to separate and transfer the organic phase into a 100-ml beaker, then add 30ml of chloroform solution and extract again. Combine the chloroform phases and evaporate them to dryness on a steam-bath, as chloroform has unsatisfactory burning characteristics in flames. Take up the residue in 2 ml of acetonitrile and transfer the solution into a 5-ml calibrated flask, washing the beaker with acetonitrile and finally diluting the solution to the mark with the same solvent.Determine the flame absorbances for iridium and ruthenium in this solution by nebulising the solution into the appropriate flame of an atomic-absorption spectrophotometer. By use of suitable calibration graphs, calculate the concentration of these elements in the rhodium sponge. ATOMIC-ABSORPTION SPECTROPHOTOMETRIC DETERMINATION OF IRIDIUM, RUTHENIUM AND Iridium-By using the Unicam SP90 spectrophotometer, the 263.9 nm atomic line from an iridium hollow-cathode lamp and an air - acetylene flame, a calibration graph of absorbance Venus concentration was constructed for iridium (0 to 2000 pg ml-l) in 2 M hydrochloric acid by appropriate treatment of standard iridium solution A. The instrument conditions for this graph are given in Table 11.In order to obtain the calibration graph for iridium in acetonitrile, the following method was used. To six separating funnels add 0, 1, 2, 3, 4 and 5 ml of standard iridium solution B and make up the volumes of the solutions to approximately 40 ml by adding 2 M hydrochloric acid. In each instance extract the solution twice by vigorous shaking with 10-ml volumes of chloroform that is 1 per cent. m/V in methyltriphenylphosphonium chloride. Transfer the combined chloroform phases into 100-ml beakers and evaporate to dryness on the steam- bath. Take up the residue in each beaker with 2 ml of acetonitrile and transfer the solutions OSMIUM-May, 19741 BY SOLVENT EXTRACTION AND ATOMIC-ABSORPTION SPECTROPHOTOMETRY 289 into 5-ml calibrated flasks marked 0, 20, 40, 60, 80 and 100 pg ml-1 of iridium, Dilute to the marks with the acetonitrile solvent and obtain the absorbance for each solution nebulised into the flame under the conditions given in Table 11.Calibration graphs for iridium in solution in 2 M hydrochloric acid (0 to 250 pg ml-1) and in acetonitrile (0 to 80 pg ml-l) were also obtained by using the Unicam SP1900 spectro- photometer under the conditions given in Table 111. Ruthenium and osmium-By using the Unicam SP90, calibration graphs were also obtained for ruthenium in solution in 2 M hydrochloric acid (0 to 400 pg ml-l) and in aceto- nitrile (0 to 100 pg ml-l), and for osmium in 2 M hydrochloric acid (0 to 2000 pg ml-l) and in acetonitrile (0 to 200 pg ml-1). The solvent-extraction procedures were identical with those used for iridium. The conditions for the atomic-absorption spectrophotometric deter- minations are shown in Table 11.TABLE I1 INSTRUMENT CONDITIONS FOR THE ATOMIC-ABSORPTION SPECTROPHOTOMETRIC DETERMINATION OF IRIDIUM, RUTHENIUM AND OSMIUM WHEN USING THE UNICAM SP90 INSTRUMENT Instrument parameter Air a t 30 p.s.i. . . .. . . C2H2 a t 7 p.s.i. (aqueous solution) N,O a t 30 p.s.i. . . .. . . C2H2 at P.s-i. (CH3CN) . . C2H2 a t 15 p.s.i. (CH3CN) . . CaH, a t 15 p.s.i. (aqueous solution) .. .. .. .. . . .. . . . . .. Wavelength of line used/nm Slit width/mm . . . . Lamp current/mA . . . . . . .. Distance of centre of light Air - C2H2 .. ..path above burnerlmm. {N20 - C2H2 .. .. Iridium 5 1.2 1.0 - 263.9 0.03 15 9 Ruthenium 5 1.4 1.2 5 2.8 2-6 349.8 0.03 15 5 5 Osmium - 5 3 2.8 290-9 15 10 0.03 - Calibration graphs for ruthenium (0 to 100 pg ml-l) and osmium (0 to 400 pg ml-1) in 2 M hydrochloric acid, and for ruthenium (0 to 40 pg ml-1) and osmium (0 to 80 pg ml-1) in acetonitrile, were also obtained by use of the Unicam SP1900 instrument under the conditions given in Table 111. TABLE I11 INSTRUMENT CONDITIONS FOR THE ATOMIC-ABSORPTION SPECTROPHOTOMETRIC DETERMINATION OF IRIDIUM, RUTHENIUM AND OSMIUM WHEN USING THE UNICAM SP1900 INSTRUMENT Instrument parameter Air a t 30 p.s.i. . . . . .. C2H, a t 10 p.s.i. (CH,CN) N20 a t 30 p.s.i. . . .. . . C2H2 a t 10 p.s.i. (CH,CN) C2H2 a t 10 p.s.i. (aqueous solution) C,H2 a t 10 p.s.i.(aqueous solution) .. . . .. .. .. .. .. .. .. .. .. .. .. ,. .. .. .. .. .. . . .. .. . . .. .. .. Wavelength of line used/nm Slit width/mm . . Sensitivity . . Integration timels Lamp current/mA . . Distance of centre of light Aqueous solution .. path above burner/mm { CH,CN . . .. .. Iridium 5 1.2 0.7 - 263.9 0.03 490 4 15 9 9 Ruthenium 5 1.2 5 2-8 349.8 0.03 580 4 15 8 10 - - Osmium - 6 3 2.8 290.9 0.06 518 4 15 10 10 The flow-rates shown in Tables I1 and I11 were those found to produce the maximum absorbances for solutions of the noble metals. The nitrous oxide - acetylene flames were lit and extinguished via air - acetylene. The concentrations of iridium, ruthenium and osmium in 2 M hydrochloric acid and in acetonitrile that produce 1 per cent.absorption using both the Unicam SP90 and Unicam SP1900 atomic-absorption spectrophotometers are shown in Table IV.290 ASHY AND HEADRIDGE: DETERMINATION OF I r AND Ru IN Rh SPONGE [Analyst, Vol. 99 TABLE IV CONCENTRATIONS OF IRIDIUM, RUTHENIUM AND OSMIUM PRODUCING 1 PER CENT. ABSORPTION Sensitivity in aqueous Sensitivity in aceto- Element Flame type solution/pg ml-l nitrilelpg ml-l & r- SP90 SP1900 SP90 SP1900 Iridium .. . . Air- C,H, 17 16 2.2 1.6 Ruthenium . . . . Air-C,H, 14 0.6 20 - NZO - CZH, 29 2.5 10-7 0.8 Osmium .. . . N,O-C,H, 65 5 12.5 1.5 For 0.5-g samples of rhodium sponge, the figures in the last column of Table IV corres- pond to 16, 8 and 15 pg g-l of iridium, ruthenium and osmium, respectively. The limits of detection would be lower than these values and a method based on the tentative method for the determination of iridium, ruthenium and osmium in rhodium sponge will be more sensitive than the spectrographic method (see above).However, before such a method could be applied to rhodium sponges, it was necessary to investigate the possible interfering effects of other extractable noble metals on the absorbances of solutions of iridium, ruthenium and osmium. INTERFERING EFFECTS OF RUTHENIUM, OSMIUM, PLATINUM AND PALLADIUM ON IRIDIUM- Twenty solutions, each containing 200 pg of iridium in addition to 200 to 1000 pg of ruthenium, osmium, palladium or platinum, were extracted in a similar way to that described for the iridium in acetonitrile calibration graph. The flame absorbances (air - acetylene flame) of the solutions of iridium containing different amounts of other co-extracted noble metals in acetonitrile were measured on the Unicam SP90 instrument and compared with the flame absorbance of acetonitrile containing iridium alone. All of these added elements affected the iridium absorbance, as shown in Fig. 1.0.1 1 0.1 0 L - ; 0.09 01 2 0.08 - 0 - 8 007 c 7J I L 2 0.06 a 0 05 Os, Pd, Pt or R u with LiCIO, added 0.04 Amount of interfering element added/pg Fig. 1. Effects of extracted osmium, palladium, platinum and ruthenium on the absorbance for extracted iridium. The amounts of elements are those originally present in the aqueous solutions before extraction INTERFERING EFFECTS OF IRIDIUM, OSMIUM, PALLADIUM AND PLATINUM ON RUTHENIUM- Twenty solutions, each containing 300 pg of ruthenium in addition to 250 to 1250 pg of .. .. .. .. . .. . ~ * . . .. I . - . . iridium, osmium, palladium or platinum, were extracted in a similar way to that describedMay, 19741 BY SOLVENT EXTRACTION AND ATOMIC-ABSORPTION SPECTROPHOTOMETRY 291 for the iridium in acetonitrile calibration graph. The flame absorbances (nitrous oxide - acetylene flame) of the solutions of ruthenium in acetonitrile containing different amounts of other co-extracted noble metals were measured on the Unicam SP90 instrument and compared with the flame absorbance of acetonitrile containing ruthenium alone. All of these added elements affected the ruthenium absorbance, as shown in Fig. 2. =I oeo6 Ir, Os, Pd or Pt with LiCIO, added + - - -1- - -I-- - -1- - + - - [I v- 0.05 cn a 0 c3 0.04 z cc 0) 2 0.03 5 8 2 0.02 0.01 0 200 400 600 800 1000 1200 1400 L I Pt Pd Ir O.O1 0 1 200 400 600 800 1000 1200 1400 Amount of interfering element added/pg Fig.2. Effects of extracted iridium, osmium, palladium and platinum on the absorbance for extracted ruthenium. The amounts of elements are those originally present in the aqueous solutions before extraction INTERFERING EFFECTS OF RUTHENIUM, IRIDIUM, PALLADIUM AND PLATINUM ON OSMIUM- Twenty solutions, each containing 250pg of osmium in addition to 250 to 1250pg of ruthenium, iridium, palladium or platinum, were extracted in a similar way to that described for the iridium in acetonitrile calibration graph. The flame absorbances (nitrous oxide - acetylene flame) of the solutions of osmium in acetonitrile containing different amounts of other co-extracted noble metals were measured on the Unicam SP90 instrument and com- pared with the flame absorbance of acetonitrile containing osmium alone.Iridium did not interfere but the other added elements affected the osmium absorbance, as shown in Fig. 3. ELIMINATION OF INTERFERENCES IN THE DETERMINATION OF IRIDIUM, RUTHENIUM AND OSMIUM- Pannetier and Toff olill successfully used lithium sulphate as an interference suppressant when trace amounts of platinum, iridium, rhodium and palladium were determined in matrices of hexachlororhodic, hexachloroiridic and hexachloroplatinic acids. Lithium perchlorate is readily soluble in non-aqueous solvents and when it was added to solutions of the noble metals in acetonitrile the interferences were eliminated.Thus, 1 ml of a 5 per cent. m/V solution of lithium perchlorate in acetonitrile was added to the solutions of the noble metals in acetonitrile before the volumes were adjusted to 5 ml. The concentration of lithium ions in the solutions to be nebulised was, therefore, approximately 6OOpgml-l. As shown in Figs. 1, 2 and 3, the addition of lithium perchlorate not only removed the interferences due to the other noble metals, but also enhanced the flame absorbances for ruthenium and osmium and allowed these two elements to be determined with considerably greater sensitivity. No enhancement in flame absorbance occurred for iridium when lithium perchlorate was present in the solutions.292 ASHY AND HEADRIDGE: DETERMINATION OF Ir AND Ru IN Rh SPONGE [Analyst, Vol.99 0'036 I - + - - 1 - - -1- - + - _ I - - - -- Ir, Pd, P t or Ru with LICIO, added c; 0.028 0.020 - 0.004 - Pd lr P t Ru I I I I I I I 0 250 500 750 1000 1250 1500 1' i0 Amount of interfering element added/pg Fig. 3. Effects of extracted iridium, palladium, platinum and ruthenium on the absorbance for extracted osmium. The amounts of elements are those originally present in the aqueous solutions before extraction FINAL METHOD FOR THE DETERMINATION OF IRIDIUM AND RUTHENIUM IN RHODIUM SPONGE To 0.5 g of rhodium sponge, containing up to 0.1 per cent. each of iridium and ruthenium, in a Pyrex tube (50 cm long when sealed), add 10 ml of concentrated hydrochloric acid. Immerse the tube in liquid nitrogen or cardice until the contents just start to solidify and then add 0.6 g of sodium chlorate crystals.Remove the tube from the liquid nitrogen, seal it and allow it to attain room temperature. Place it in a steel casing fitted with a screw-cap and heat it in an oven at 250 "C for at least 6 hours so as to dissolve all of the sponge. Allow the tube to cool to room temperature and cool it again in liquid nitrogen or cardice until the contents just start to solidify. Open the tube by scratching the glass near to the sealed end and giving it a sharp tap. (As the tube breaks, chlorine, under pressure in the tube, may escape.) Transfer the contents of the tube quantitatively into a 250-ml separating funnel, then add 10 ml of concentrated hydrochloric acid and dilute the solution to 120 ml.Add 30 ml of chloroform that is 1 per cent. m/V in methyltriphenylphosphonium chloride to the mixture in the funnel and shake it vigorously for 2 minutes. Allow the layers to separate and transfer the organic phase into a 100-ml beaker, then add 30 ml of chloroform solution and extract again. Combine the chloroform phases and evaporate them to dryness on a steam-bath. Take up the residue in 2 ml of acetonitrile and transfer the solution into a 5-ml calibrated flask. Rinse the beaker with 1 ml of acetonitrile and add the rinsings to the flask, then add 1 ml of a 5 per cent. m/V solution of lithium perchlorate in acetonitrile, ignoring a small precipitate of methyltriphenylphosphonium perchlorate that forms, and dilute the solution to the mark with the same solvent.Determine the flame absorbances for iridium and ruthenium in this solution by using air - acetylene and nitrous oxide - acetylene flames, respectively, and the instrument con- ditions outlined in Table 111. If the rhodium sponge contains not less than 70 pg g-1 of iridium and 200 pg g-l of ruthenium, the instrument conditions outlined in Table I1 can be used. Read off the concentrations of iridium and ruthenium from calibration graphs prepared from solutions, the flame absorbances of which were determined at the same time as those of the solutions prepared from the rhodium sponge samples. PREPARATION OF THE CALIBRATION GRAPHS- Iridium-The method is identical with that described under Iridium in the section entitled Atomic-absorption spectrophotometric determination of iridium, ruthenium andMay, 19741 BY SOLVENT EXTRACTION AND ATOMIC-ABSORPTION SPECTROPHOTOMETRY 293 osmium, except that a l-ml volume of a 5 per cent.m/V solution of lithium perchlorate in acetonitrile is added to each 5-ml calibrated flask before diluting to the mark. Ruthenium-The method is identical with that used for iridium except that standard ruthenium solution B is used. ANALYSIS OF SYNTHETIC MIXTURES OF RHODIUM SPONGE WITH TRACE AMOUNTS OF IRIDIUM, Seven synthetic mixtures were prepared by adding fixed volumes of standard solutions of iridium and ruthenium to 0.5-g amounts of Specpure rhodium sponge in seven Pyrex tubes. Dissolution was achieved in a similar way to that described in the final method and the iridium and ruthenium contents of these solutions were determined as outlined in the final method. Four of these solutions also contained known trace amounts of osmium, which were later determined by atomic-absorption spectrophotometry in a manner similar to that for iridium, but this osmium was added after the dissolution procedure and before the solvent extraction (see Discussion).ANALYSIS OF RHODIUM SPONGE SAMPLES- by means of the final method, as was a l-g sample of rhodium trichloride. RUTHENIUM AND OSMIUM ADDED- Eleven samples of rhodium sponge were analysed for iridium and ruthenium contents RESULTS The sensitivities of the atomic-absorption spectrophotometric determinations of iridium, ruthenium and osmium in acetonitrile containing lithium perchlorate are given in Table V.TABLE V CONCENTRATIONS OF IRIDIUM, RUTHENIUM AND OSMIUM PRODUCING 1 PER CENT. ABSORPTION I N ACETONITRILE CONTAINING LITHIUM PERCHLORATE Sensitivity in acetonitrile containing LiClO,/pg ml-1 Element Flame type SG90 SP1400 Iridium . . .. Air-C2H, 2.2 1.6 Ruthenium . . . . Air-C,H, 15 N2O - C,H2 3 0.4 - Osmium . . .. NZO-CZH, 7 1 Results for the analysis of synthetic mixtures of rhodium containing trace amounts of iridium, ruthenium and osmium are shown in Table VI. These results are corrected for the trace amount of iridium present in the Specpure rhodium sponge; no ruthenium or osmium could be detected in this sponge. The results for mixtures 1, 2, 3 and 4 were obtained with the Unicam SP90 instrument. The Unicam SP1900 instrument was used in the analysis of mixtures 5, 6 and 7.TABLE VI RESULTS FOR THE ANALYSIS OF SYNTHETIC MIXTURES OF RHODIUM CONTAINING TRACE AMOUNTS OF IRIDIUM, RUTHENIUM AND OSMIUM Mixture 1 2 3 4 5 6 7 Iridium, per cent. & Added Found 0.060 0.058 0.040 0.037 0.020 0.017 0.100 0.099 0.020 0.022 0.080 0.082 0.160 0.155 Ruthenium, per cent. +--7 & Added Found Added Found 0.040 0.041 0.120 0.115 0.100 0.100 0.080 0.075 0.020 0.019 0.160 0.150 0.060 0.061 0.040 0.042 0.080 0.075 - - 0.060 0.065 0.020 0.020 - - Osmium, per cent. - - Results for the determination of iridium and ruthenium in rhodium sponges by the described method are given in Table VII, together with results for the spectrographic deter- mination of these elements in the sponges. The atomic-absorption results are the averages of two determinations on each sample.294 ASHY AND HEADRIDGE: DETERMINATION OF Ir AND Ru IN Rh SPONGE [Analyst, Vol.99 TABLE VII COMPARISON OF RESULTS FOR THE DETERMINATION OF IRIDIUM AND RUTHENIUM I N RHODIUM SPONGE BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY AND EMISSION SPECTROGRAPHY Iridium content, per cent. f A > Spectrographic Atomic-absorption Sample analysis spectroyhotometr y A 0.1 0.098 B 0.001 N.D. C 0.01 0.012 D 0.015 0.014 E 0.007 0.008 F 0-03 0-027 G 0.006 0.010 H 0.03 0.029 0.03 0.027 0.007 0.008 J K Specpure sponge - 0.005 RhCl,.xH,O - 0.042 N.D. = Not detected. Ruthenium content, per cent. A f \ Spectrographic Atomic-absorption analysis spectrophotometry 0.07 0-0770 N.D. N.D. 0.020 0.0190 0.0060 0.005 0.005 0.0070 0.007 0.0075 0-007 0~0080 0.015 0.0150 0.001 N.D.0.005 0.0055 - N.D. - 0.02 18 DISCUSSION The results obtained for the determination of trace amounts of iridium in actual and simulated rhodium sponge samples are considered to be satisfactory. Only with iridium for sample G is there some discrepancy between the spectrographic and atomic-absorption spectrophotometric results. The limits of detection for the determination of iridium and ruthenium in rhodium sponge were 7 and 4-5 pg gl, respectively, where the limit of detection is defined as that concentration of the element which gives a signal equal to twice the standard deviation of a series of ten determinations near to the blank level. The spectrographic results were obtained in the laboratories of Johnson Matthey Chemicals Ltd.by using solid samples in a d.c. arc, intensities being recorded photographically on an Ebert 3-m instrument. Rhodium sponge seldom contains osmium in detectable amounts but if such a deter- mination has to be made, it is advisable to freeze completely the contents of the Pyrex tube by immersing it in liquid nitrogen before it is broken and to treat the contents in the manner specified in the method for preparing a standard solution of osmium. If a known trace amount of osmium is added to 0.5 g of rhodium sponge and the sponge dissolved according to the procedure of the final method, in which the contents of the tube are cooled until they just start to solidify, there is only a 97 per cent. recovery of osmium, presumably because a small amount of volatile osmium tetroxide escapes with the excess of chlorine from the tube when it is broken.Rhodium sponge is particularly difficult to dissolve, but the use of sealed glass tubes in the dissolution procedure was very helpful. If simple precautions are taken, these tubes are not dangerous and are easy to handle. No explosions occurred when tubes containing 1 g of sodium chlorate and 10 ml of concentrated hydrochloric acid were heated to a maximum temperature of 260 "C. The use of a PTFE-lined steel pressure vessel for the dissolution of rhodium, iridium and ruthenium with sodium chlorate and concentrated hydrochloric acid has been investigated. Unfortunately, neither rhodium nor iridium could be dissolved completely at a temperature of 220 "C on heating for 14 hours. When the temperature was increased, the PTFE started to soften and gases escaped from the pressure vessel.Ruthenium was found to dissolve at 220 "C on treatment in the pressure vessel for 24 hours but iridium still resisted complete dissolution under similar conditions. It is well known that noble-metal hollow cathodes have complicated emission spectra, and the lack of sensitivity for ruthenium and osmium in atomic-absorption spectrophotometric determinations when the Unicam SP90 instrument was used is due to the inexpensive mono- chromator of the instrument, which could not resolve the resonance lines from the emission spectra. When iridium was determined, the less sensitive resonance line at 263.9 nm was used as it was less noisy than the primary resonance line at 208.9 nm.Precipitates slowly formed during the extraction procedure when more than 600pg ofMay, 19741 BY SOLVENT EXTRACTION AND ATOMIC-ABSORPTION SPECTROPHOTOMETRY 295 iridium, 800 pg of ruthenium and 1200 pg of osmium were extracted into 10 ml of chloroform that was 1 per cent. m/V in methyltriphenylphosphonium chloride. Therefore, an immediate transfer of the chloroform layer was necessary when the amounts of noble metals exceeded these limits. The precipitates are very soluble in acetonitrile. The mutual interferences have been eliminated by the addition of lithium perchlorate. Not only were the interferences overcome, but also, for ruthenium and osmium, an enhance- ment in the absorption signals occurred. The causes of mutual interferences among the noble metals are not understood and very little is known about the mode of action of the reagents that suppress them.As can be seen from Tables IV and V, the concentrations of iridium, ruthenium and osmium in acetonitrile containing lithium perchlorate that produce 1 per cent. absorption with the Unicam SP1900 instrument are 1.6, 0.4 and 1 pg ml-l, respectively, compared with the best values of 16, 0.6 and 5 pg ml-l for aqueous solutions. If, for a direct method without solvent extraction, a typical concentration of rhodium in aqueous solution is taken as being 1 per cent. m/V, then the concentrations of iridium, ruthenium and osmium in rhodium sponge corresponding to 1 per cent. absorption are 1600,60 and 500 pg g-l, respec- tively. With the solvent-extraction procedure, the concentrations of iridium, ruthenium and osmium in rhodium sponge corresponding to 1 per cent. absorption are 16, 4 and 10 pg g l , which is a marked improvement. No scale expansion was used on the Unicam SP1900 instrument in these determinations, but if scale expansion is used it is anticipated that the limits of detection for iridium and ruthenium in rhodium sponge by the described method will be 4 and 1 pgg-l, respectively, or even better.12 Very recent results for the limits of detection for iridium and ruthenium in rhodium sponge using a d.c. arc on an Ebert 3-m spectrograph are 10 and 1 pg g l , respectively. The atomic-absorption spectrophotometric method gives similar sensitivity. We are indebted to the University of Riyadh, Saudi Arabia, for a maintenance grant for M. A. Ashy and to Johnson Matthey Chemicals Ltd. for the loan of Specpure noble metals and rhodium sponge samples. We also thank Johnson Matthey Chemicals Ltd. for allowing us to use their Unicam SP1900 atomic-absorption spectrophotometer and for providing us with information on the detection limits for noble metals in rhodium sponge when carrying out emission spectrography. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Headridge, J. B., and Sowerbutts, A., Analyst, 1973, 98, 57. Headridge, J . B., and Smith, D. R., Lab. Pract., 1971, 20, 312. Headridge, J. B., and Richardson, J., Analyst, 1970, 95, 930. Headridge, J. B., and Sowerbutts. A., Ibid., 1972, 97, 442. Slavin, W., “Atomic Absorption Spectroscopy,” Interscience Publishers, New York, 1968, pp. 114, Neeb, R., 2. analyt. Chem., 1957, 17, 154. Beamish, F. E., “The Analytical Chemistry of the Noble Metals,” Pergamon Press, Oxford, 1966, Cotton, F. A., and Wilkinson, G., “Advanced Inorganic Chemistry,” Third Edition, Interscience Griffith, W. P., “The Chemistry of the Rarer Platinum Metals,’’ Interscience Publishers, New Pannetier, G., and Toffoli, P., Bull. Soc. Chim. Fr., 1971, 3775. Thomerson, D. R., Scan, 1973, 1, 12. 138 and 154. Fok, J . S.-K., D ~ s s . Abstr., 1965, 25, 3815. p. 22. Publishers, New York, 1972, p. 990. York, 1967. Received November 7th, 1973 Accepted December 28th, 1973

ISSN:0003-2654    

DOI:10.1039/AN9749900285

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

296 Analyst, May, 1974, Vol. 99, pp. 296-301 The Use of Ascorbic Acid to Eliminate Interference from Iron in the Aluminon Method for Determining Aluminium in Plant and Soil Extracts BY T. C. 2. JAYMAN AND S. SIVASUBRAMANIAM (Tea Research Institute of Ceylon, Agricultural Chemistry Department, St. Coombs, Talawakele, Sri Lanka) The aluminon method of Hsu for the determination of aluminium in soil has been examined for interference from the presence of iron and copper. Only iron(II1) interfered. The use of thioglycollic acid not only eliminated this interference but bleached the colour of the aluminon reagent and caused a reduction in sensitivity. Ascorbic acid, however, eliminated it without bleaching the colour of the reagent, the addition of 1 ml of a 0.50 per cent. solution of ascorbic acid suppressing the interference from up to 500p.g of iron(II1). When 200 pg of copper were added to aluminium standards no interference occurred.ALUMINIUM in plant and soil extracts has been commonly determined with aluminon. Cheneryl used thioglycollic acid to complex the iron and many workers have subsequently used his method or modifications of it. Middleton2 used 8-hydroxyquinoline to extract both iron and aluminium in the same extract. The aluminon method of Cheneryl has not always yielded reproducible values for aluminium. Pellowe and Hardy3 have commented on the heating time adopted and observed that the ambient temperature at which the final solutions containing the developed lake have stood is very important if reproducible results are to be achieved.Furthermore, in most of the aluminon methods so far published, stabilisers have been incorporated. However, Corey and Jackson4 have used the aluminon method without the use of a protective colloid. Page and Bingham5 have eliminated interfering ions by passing the test solution through an ion-exchange column and accomplished colour formation at pH 4.2 without the use of a stabiliser. In most aluminon methods, emphasis has been placed on the interference of iron. In his method, however, Hsu6 has not considered this aspect. If the ratio of aluminium to iron is large, which is usually the case with aluminium accumulators such as the tea plant, then the values for aluminium determined without the use of an inhibitor for iron will be reasonably accurate.For the analysis of the foliage of plants grown in nutrient solutions excluding aluminium but incorporating a high concentration of iron, however, values for aluminium determined without the use of an iron inhibitor will be incorrect. In order to modify Hsu’s method so as to make it more readily applicable, and prompted by the fact that erratic results were obtained for tea-leaf samples by the method of Chenery,l the authors undertook the present investigation. EXPERIMENTAL REAGENTS- Aluminon - acetate bufer-Dilute 120 ml of glacial acetic acid to about 900 ml with distilled water, add 24g of sodium hydroxide, mix and dissolve in the resulting solution 0.35 g of aluminon; dilute to 1 litre with distilled water and mix. The pH of this solution should be 4.2.When lord of this reagent are added to a test solution containing 2.0 to 4-0 ml of 1 N hydrochloric acid, dilution with water to 50 ml gives a final pH of 4.0 to 3.7. Ascorbic acid solution, 0.50 per cent.-Dissolve 0-50 g of ascorbic acid in a 100-ml cali- brated flask in, and make the contents up to the mark with, distilled water. Standard aluminium solution-Dissolve 1 g of pure aluminium wire in 25 ml of 6 N hydro- chloric acid and dilute to 1 litre with distilled water. This solution contains 1000 pg ml-l of aluminium. Dilute 10 ml of the above solution to 1000 ml with distilled water so as to obtain a working standard containing 10 pg ml-l of aluminium. @ SAC and the authors.JAYMAN AND SIVASUBRAMANIAM METHODS CALIBRATION OF STANDARD GRAPH AS OUTLINED BY HSU*- 297 Transfer, by pipette, into 50-ml calibrated flasks aliquots containing from 0 to 60pg of aluminium. Add 2.0 ml of 1 N hydrochloric acid to each flask and place the flasks on a steam-bath for 30 minutes.Cool and dilute the solutions to about 35 ml with distilled water, add exactly 10 ml of the aluminon reagent, make the contents up to the mark with distilled water, mix and allow the solutions to stand for at least 2 hours. Read the colours produced against the reagent blank set at zero in 1-cm cells at 530 nm. Any suitable colori- meter can be used. The graph obtained is linear. CALIBRATION OF STANDARD GRAPH AND DETERMINATION OF ALUMINIUM IN UNKNOWN SAMPLES By pipette, transfer into 50-ml calibrated flasks aliquots containing from 0 to 50pg of aluminium.Add sufficient distilled water to make the volumes up to approximately 10 ml. Then add, in order, 1 ml of 0.50 per cent. ascorbic acid solution and 2.0 ml of 1 N hydro- chloric acid (this volume may be varied so that the final pH with 10 ml of aluminon reagent added should be between 4.2 and 3.8, the slight shift in the range from 4-0 to 3.7 causing a negligible error) and place the flasks on a steam-bath for 30 minutes. Cool and dilute the solutions to about 35 ml with distilled water, add 10 ml of aluminon reagent, make the volumes up to the mark with distilled water, stopper the flasks and mix. After 2 hours or longer, read the colours obtained against the reagent blank in 1 or 2-cm cells at 530nm. ASHING OF PLANT SAMPLES- Weigh 0.20 g of finely ground sample into glass ashing tubes held in place with a stainless- steel tray and leave them overnight in a muffle furnace at 450 "C.Cool, add to the tubes a few drops of distilled water in order to moisten the ash, then add by pipette 2 ml of the digestion mixture nitric acid - hydrochloric acid - water (25 + 25 + 50 V / V ) and evaporate the mixture to dryness on a hot-plate. Add exactly 10 ml of 0.05 N hydrochloric acid, warm the tubes so as to dissolve the contents, stopper them and shake them to mix well. Leave the tubes aside for 1 hour so as to allow silica to settle. Use suitable aliquots for the deter- mination of aluminium. BY THE MODIFIED PROCEDURE- ALUMINIUM EXTRACTABLE FROM SOIL WITH AMMONIUM CHLORIDE SOLUTION- Weigh into a 4-02 Kali shaking bottle 10 g of air-dried soil that has been sieved to pass a 2-mm mesh, add 100 ml of 1 N ammonium chloride solution, stopper the bottle and shake it overnight in a reciprocating shaking machine.Filter the mixture into a clean, dry receiver by using a Whatman No. 542 filter-paper. Determine the aluminium on a suitable aliquot by the modified procedure. INTERFERENCE FROM IRON- Standards calibrated as outlined by Hsu6 obeyed Beers' law over the range 0 to 60 pg of aluminium per 50 ml. When iron was added, however, there was a definite enhancement of colour that resulted in a positive error. The results obtained are shown in Table I. RESULTS AND DISCUSSION TABLE I INTERFERENCE FROM IRON WHEN ADDED TO SAMPLES BEING ANALYSED FOR ALUMINIUM Aluminium present/ p g per 50 ml 20.0 20.0 20.0 20.0 20.0 40.0 40.0 40.0 40.0 40.0 Iron added/ pg per 50 ml 0 5 10 20 40 0 5 10 20 40 Aluminium found/ pg per 50 ml 20.0 21.0 22.5 23.5 27-6 40.0 41-0 43-0 45.0 48.0298 JAYMAN AND SIVASUBRAMANIAM: USE OF ASCORBIC ACID TO ELIMINATE [Analyst, Vol.99 THIOGLYCOLLIC ACID AS AN INHIBITOR FOR IRON INTERFERENCE- The effect of different volumes of 1 per cent. V/Y thioglycollic acid solution, with and without added iron, on the recovery of 40 pg of aluminium contained in the solution in the 50-ml calibrated flasks was investigated. The results are shown in Table 11. TABLE I1 EFFECTIVENESS OF THIOGLYCOLLIC ACID AS AN INHIBITOR FOR IRON INTERFERENCE WHEN ADDED TO SAMPLES BEING ANALYSED FOR ALUMINIUM Aluminium present/ pg per 50 ml 40.0 40.0 40.0 40.0 40-0 40.0 40.0 40-0 Iron added/ pg per 50 ml 0 100 0 0 0 100 100 100 Volume of thioglycollic acid/ml - 1.0 2.0 4-0 1.0 2.0 4.0 Aluminium found/ pg per 50 ml 40.0 61.5 25.0 19.0 10.0 25-0 19.0 10.0 From the results in Table 11, it is clear that the addition of 1 ml of thioglycollic acid solution eliminates the interference from 100 pg of iron, but that the thioglycollic acid also suppresses full colour development, resulting in a low recovery of aluminium.In order to ascertain whether various amounts of thioglycollic acid would affect the colour of the aluminon reagent itself, an experiment was carried out in which 0, 1, 2, 3, 4, 5 and 6 ml of 1 per cent. V/V thioglycollic acid solution were added to the solutions in the respective 50-ml flasks and the colours developed as outlined by HSU.~ The colours were read against distilled water at 530 nm in l-cm cells.I i Volume of thioglycollic acid solution /ml Fig. 1. Bleaching action of thioglycollic Final pH of all acid on the aluminon reagent. solutions 4.0 to 3.9 The results are illustrated in Fig. 1, from which it is clear that thioglycollic acid bleached the aluminon reagent and that the addition of only 1 ml of the thioglycollic acid solution caused more than 50 per cent. of the colour to be bleached. The gradients of the calibration graphs for aluminium standards with and without the addition of thioglycollic acid were determined by carrying out the following experiment. To three sets of aluminium standards containing from 0 to 50 pg of aluminium per 50 ml, 0, 1 and 3 ml of 1 per cent.V/V thio- glycollic acid were added. All of the graphs were linear, but their slopes were 0.60, 0.30 and 0.22, respectively. Another set of aluminium standards containing from 0 to 50pg of alu- minium per 50 ml, and a ten-fold increase in iron at each level of aluminium, was calibrated with the addition of 1 ml of the thioglycollic acid solution. The graph was again linear with a slope of 0.32 and fitted that which contained 1 ml of thioglycollic acid solution, but without iron. Fig. 2 illustrates the results of the above experiment.May, 19741 INTERFERENCE FROM IRON I N THE ALUMINON METHOD 299 OTHER CHELATING AGENTS- From Fig. 2, it is clear that the thioglycollic acid caused a reduction in the sensitivity of the method by almost half. Two other chelating agents, i.e., 5 per cent.m/V tartaric acid solution and 1 per cent. m/V EDTA solution, were therefore tried, but in both instances colour development was markedly suppressed. 301-- 25 - 20 - CJ, U .- !! 1 5 - aJ + 5 . l o - 5 - Aluminium/pg per 50 ml Fig. 2. Loss of sensitivity owing to the use of thioglycollic acid : A, without iron, ascorbic acid or thioglycollic acid, gradient 0.60; B, with iron (A1 to Fe ratio 1 : 10) and ascorbic acid, gradient 0.60; C, without iron but with 1 ml of 1 per cent. V/V thioglycollic acid solution, gradient 0.30; D, with iron (A1 t o Fe ratio 1 : 10) and 1 ml of 1 per cent. V / V thio- glycollic acid solution, gradient 0.32; and E, without iron but with 3 ml of 1 per cent. V / V thjoglycollic acid solution, gradient 0.22 REDUCING AGENTS- As both iron(II1) and aluminium ions behaved in a similar way, it was decided to investi- gate the effect of reducing the iron(II1) to the iron(I1) state, and the two reducing agents ascorbic acid and hydroxylammonium chloride were therefore tried.When aluminium standards containing iron(II1) were calibrated with either of these reducing agents, the graphs300 JAYMAN AND SIVASUBRAMANIAM : USE OF ASCORBIC ACID TO ELIMINATE [Analyst, Vol. 99 were linear and fitted that which contained no iron(II1). As it was more convenient to use ascorbic acid, all further investigations were carried out with it. ASCORBIC ACID AS A REDUCING AGENT- It is apparent from the above experiment that iron present only in the iron(II1) state interfered and it was necessary to determine the most suitable concentration of ascorbic acid to use for the elimination of this interference.In a series of experiments carried out with this objective, it was found that 1 ml of a 0.50 per cent. m/V solution of ascorbic acid was sufficient to eliminate the interference from up to 500 pg of iron(II1) in the presence of 50 pg of aluminium per 50ml. A set of aluminium standards containing a ten-fold increase in iron(II1) at each level of aluminium was calibrated with the addition of 1 ml of 0.50 per cent. ascorbic acid. The graph was linear and fitted the line for aluminium standards without the addition of either iron(II1) or ascorbic acid (see Fig. 2). It was now clear that ascorbic acid not only eliminated the interference from iron(III), but also left the colour of the aluminon reagent unimpaired and therefore the sensitivity of the method unaffected.INTERFERENCE FROM COPPER- Pellowe and Hardy3 mentioned that copper in the presence of aluminon and thioglycollic acid imparts a mauve coloration, which interferes, and that if copper is present in large amount it should be either removed or suitably compensated for in the standards. In order to determine whether copper interfered in the aluminon method with ascorbic acid, 200 pg of copper were introduced into the standards. This amount of copper caused no interference. Even with tea leaves, on which fungicides with copper formulations are often sprayed, the copper contents in the aliquots used for aluminium determinations rarely exceed 200 pg. COMPARISON OF RESULTS FOR ALUMINIUM DETERMINED IN TEA-PLANT FOLIAGE BY THE Leaf samples from an experimental plot were sampled, dried and ashed as outlined above and suitable aliquots were taken for aluminium determination by the method of Chenery.1 The results obtained are presented in Table 111.TABLE I11 RESULTS OF THE DETERMINATION OF ALUMINIUM BY THE METHOD OF CHENERY METHOD OF CHENERY AND THE MODIFIED PROCEDURE- Aliquot size/ Aluminium ml added/pg 0.40 - 1.00 - 2.00 - 0.40 10.0 1.00 20.0 2.00 30.0 Aluminium found/ pg 2.75 4.75 6-00 10.20 16.25 20-50 Aluminium in plant Recovery, sample, p.p.m. per cent. 344.0 - 238-0 - 150.00 - - 80.00 - 65.6 - 56.9 It will be noted from the results in Table I11 that the amount of aluminium found is dependent on the aliquot size and that increasing the aliquot size decreased the amount of aluminium recovered, probably as a result of interference by silicates.TABLE IV DETERMINATION OF ALUMINIUM AND RECOVERY TESTS ON TEA-LEAF SAMPLES BY THE MODIFIED METHOD Aliquot size/ml 1.00 2.00 2.50 3.00 1.00 1.00 1.00 1.00 Aluminium Aluminium in plant Aluminium Aluminium Recovery, found/& sample, p.p.m. addedlpg recovered/pg per cent. 9.00 450.0 - - - 18.00 450.0 - - - 22.50 450.0 - - 27-00 450.0 - - - - - 10.0 19.0 100 - - 20.0 29.0 100 - - 30.0 39.0 100 - - 40.0 49-0 100May, 19741 INTERFERENCE FROM IRON I N THE ALUMINON METHOD 301 The same plant-leaf solution was analysed by the modified procedure and the sample size was varied in order to ascertain whether it would have any bearing on the amount of aluminium recovered.The results obtained are shown in Table IV. It can be seen from the results in Table IV that the aliquot size had no influence on the amount of aluminium found and that the recovery of added aluminium by this method was excellent. For plant samples, therefore, this method appears to be very reliable. ALUMINIUM EXTRACTABLE FROM SOIL WITH AMMONIUM CHLORIDE SOLUTION- A soil from an experimental plot was extracted with ammonium chloride solution as detailed earlier and the aluminium determined in the extract. Aluminium was also added to separate aliquots and to one series from 50 to 200 pg of iron(II1) were added. The results of the determination are presented in Table V. TABLE V RECOVERY OF ADDED ALUMINIUM IN EXTRACTS OF SOIL WITH AMMONIUM CHLORIDE SOLUTION I N THE PRESENCE OF ADDED IRON BY THE MODIFIED PROCEDURE Sample aliquot/ml 0.50 0.50 0-50 0.50 0.50 0.50 0.50 0.50 Aluminium added/pg - 10.0 20.0 30.0 10.0 20.0 30.0 Iron(II1) Aluminium added/pg found/ pg - 17.20 - 17-20 - 27.50 - 37.20 - 47.50 50.0 27.50 100.0 38.00 200.0 47.50 Aluminium Recovery, recoveredlpg per cent.- - 10.30 103 20 00 100 30.30 101 10.30 103 20.80 104 30.30 101 Mean . . 102 It can be seen from the results in Table V that the modified method is precise and accurate even for determining aluminium extractable from soil with ammonium chloride solution and, furthermore, only one calibration graph needs to be constructed for both alu- minium in plant samples and that extractable from soil with ammonium chloride solution. The presence of ammonium chloride does not interfere in the method. Hsu6 in his method examined the effect of phosphates and silicates, the time of heating and temperature of the solution, and the pH range for maximum colour development. His work on these aspects was so comprehensive that no further examination of the parameters has been carried out, other than of the effect of phosphates, and our investigations confirmed Hsu’s findings. His method, as modified in order to eliminate the interference from iron by the use of ascorbic acid, has been found to be extremely reliable and simple to use. The colours are stable for at least 24 hours, hence large numbers of samples can conveniently be determined. The authors gratefully acknowledge the skilful technical assistance given by Mr. P. Nalliah during this work. They thank the Director of the Tea Research Institute for permission to publish this paper. REFERENCES 1. 2. 3. 4. 5. 6. Chenery, E. M., Analyst, 1948, 73, 501. Middleton, K. R., Ibid., 1964, 89, 421. Pellowe, E. F., and Hardy, F. R. F., Ibid., 1954, 79, 225. Corey, R. B., and Jackson, M. L., Analyt. Chem., 2953, 25, 624. Page, A. L., and Bingham, F. T., Proc. Soil Sci. SOC. Amer., 1962, 26, 351. Hsu, Pa Ho, Soil Sci., 1963, 96, 230. Received October 15th, 1973 Accepted November Ist, 1973

ISSN:0003-2654    

DOI:10.1039/AN9749900296

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

302 Analyst, May, 1974, Vol. 99, pp. 302-305 A Method of Collecting and Concentrating Headspace Volatiles for Gas- chromatographic Analysis By R. E. HURST (Fisheries Research Board of Canada, Vartcouver Laboratory, 6640 N . W. Marine Drive, Vancouve, 8, B.C., Canada) Headspace analysis by gas chroinatography has been widely used in the field of food science in order to characterise flavours and odours, and as an index of quality. One of the major difficulties encountered in this work lies in obtaining a sufficient concentration of volatiles to permit detection and identification. A method of collecting and enriching headspace volatiles and of introducing the total concentrate on to a gas chromatograph is described. GAS-CHROMATOGRAPHIC analysis of the headspace over various foods has been used for such purposes as the characterisation of flavours and odours, and as an index of quality.Direct sampling of the volatiles with a gas-tight syringe is the most elementary procedure used in headspace analysis and for certain applications is very sati~factory.l-~ Generally, however, this method does not provide a sufficient concentration of volatiles for adequate detection by gas chromatography, and various devices and techniques have been developed to overcome this Hughes4 used a sweeping gas to carry volatiles from cooked herring into a U-trap cooled with liquid oxygen. When the collection was complete the condensed volatiles in the U-trap were expanded into a glass syringe of large capacity (50 or 100 ml) as the gas warmed to ambient temperature. Samples were then removed by syringe through a septum in the system and injected in a convenient way on to a gas chromatograph.The method described here is a modification and extension of this procedure, which permits volatiles at ambient temperature to be collected and the total volume to be injected on to the column, METHOD GAS CHROMATOGRAPHY- Analysis was performed on a Microtek MT220 gas chromatograph with flame-ionisation detection. The injection port septum nut of the instrument was modified by adding a Teflon- lined tube on the top to act as an additional seal for the injection needle (Fig. 1). Further, a Whitey toggle valve was fitted to the carrier-gas line so that the flow could be shut off during injection. A 1.8-m long column was used, constructed of heavy-walled glass tubing, 8 mm 0.d.x 2 mm i.d., with 6 mm 0.d. heavy-walled tubing on both ends to accommodate the &-inch Swagelok fittings at the inlet and detector port of the instrument. The column was packed with Porapak Q, 80 to 100 mesh. Temperature programming was used in the range of 50 to 180 "C at a rate of 2 "C min-l and the carrier gas was nitrogen with a flow-rate of 25 ml min-1. The instrument was attenuated at 1 x 16 so as to give maximum sensitivity with an accept- able signal to noise ratio. PROCEDURE- A detailed description of the apparatus is given in Fig. 1. The sample flask with two side-arm tubes containing septa is connected by Vacutainer 22G needles ( l i inches) (Becton, Dickinson and Company, Rutherford, New Jersey) and silicone rubber tubing to a 50-ml glass syringe and to a collection tube consisting of a heavy-walled test-tube (9 mm i.d., 13 mm 0.d.and 70 mm long with a volume of approximately 3 ml) stoppered with a 4-ml Vacutainer stopper. The sample flask is charged with sample which, in the two examples given, was glycerol containing a mixture of alcohols of low relative molecular mass at concentrations of 40 p.p.m. and liquor from canned salmon. The syringe plunger is withdrawn so as to give a 50-rnl capacity and a glass stopper placed securely in the neck of the sample flask so that the system is sealed and the syringe is held at the 50-ml mark. Liquid nitrogen @ SAC and the author.HURST 303 - contained in a small vacuum flask is brought into contact with the collecting tube so that approximately the bottom two thirds of the tube is immersed, which ensures that the upper area of the tube does not become too cold and thus cause freezing-out at the inlet needle.Stirrer Vacutainer stopper 'Collection tube -Vacuum flask T -Syringe 50 40 - - 30 2c - 1c - B n Aluminium cap Septum (1 3 mm) Hole Vacutainer needle PTFE Modified septum nut Septum (1 Column 6mm 0.d. 8mm 0.d. 2 mm i.d. 3 mm) Fig. 1. Schematic diagram of t h e collection device As the headspace gas from the sample flask condenses in the collection tube, the syringe plunger moves down to compensate for the reduced volume of gas in the system. When the syringe plunger has moved sufficiently to replace 50 ml of condensed headspace, the liquid nitrogen is removed from the collection tube, and as the temperature rises the syringe plunger moves upwards in order to accommodate the expanding gases.When the expanded volume has reached 45 ml, the liquid nitrogen is again brought into contact with the collection tube and the procedure is repeated so that a further 45 ml of headspace is condensed. It should be noted here that the principal condensation was of 50 ml; however, expansion to a volume of only 45 ml was permitted, thus ensuring that a sufficiently low temperature is maintained within the collection flask so that the condensed volatiles essentially remain during expansion. The temperature at this stage is approximately -50 "C. After an appropriate number of condensations have been made, which in this work was four, the needle is withdrawn from the septum of the collection tube while the latter is still immersed in liquid nitrogen.The tube is then removed from the liquid nitrogen, immediately inserted into the aluminium jacket and the cap quickly screwed down tightly (Fig. 2A). The rubber stopper in the collection tube is now held securely in place so that no volatiles escape as the pressure within the tube increases with warming. The jacketed tube is placed in a water-bath at 50 "C ready for injection on to the gas chromatograph. Carrier-gas flow to the gas-chromatographic column is shut off by means of the toggle valve, which will cause a negative response on the recorder. The long end of a Vacutainer 22G needle (14 inches) is inserted into the modified injection nut and through the septum.Next, the cap on the aluminium jacket containing the collection tube is brought down on to304 [Analyst, Vol. 99 the protruding short end of the Vacutainer needle in such a way as to enter the hole in the cap and pierce the cap septum and tube stopper (Fig. 2B). This procedure discharges the volatiles under pressure on to the column and an immediate pressure spike will ensue on the recorder. When the pen drops back to the negative response position, the carrier gas is turned on and the needle is withdrawn from the septum of the instrument, leaving the upper end of the needle to remain in the collection tube. The instrument is then set to temperature programme, HURST: A METHOD OF COLLECTING AND CONCENTRATING HEADSPACE DISCUSSION Gas chromatograms of alcohols of low relative molecular mass are shown in Fig.3. Headspace was collected above glycerol containing 40 p.p.m. each of methanol, ethanol, propan-1-01 and butan-1-01. Sampling by the standard method using a 10-ml Hamilton 1010 gas-tight syringe is compared with a single condensation and with four condensations in order clearly to demonstrate the enrichment by this method. Identification of volatiles associated with canned salmon is currently being undertaken and the method described above was developed for this purpose. Gas chromatograms of headspace volatiles collected from the liquor in canned salmon are shown in Fig. 4. The enrichment of volatiles from successive condensations can be seen by comparing trace 1 from a single 50-ml condensation with trace 2, which represents four condensations totalling 185 ml. Single 50-ml condensations have been successfully used in a routine manner to demonstrate, by gas-chromatographic profiles, differences related to the quality of product.Care should be taken in the procedure to ensure that the needle entering the collection tube just passes through the Vacutainer stopper so that plugging will not occur from freezing during condensation. Once the collection tube has been secured in the aluminium jacket with the cap, no problem has been encountered with the escape of gas prior to injection, although no attempt has been made to store the collected volatiles in this manner. It was necessary to modify the septum nut of the instrument in order to prevent the escape of volatiles during the pressure injection, which is approximately 16 atm in the collection tube at ambient temperature.This amount of pressure has presented no hazardous problems. The apparatus has been thoroughly tested by using a thin-walled glass collection tube instead of the heavy-walled tube so as to encourage breakage under pressure. In every instance in which the aluminium cap was not in place the rubber stopper released the pressure in much the same manner as the pressure valve in some domestic pressure cookers. Once the tube is secured in the jacket any breakage would be contained; however, the use of a thick-walled collection tube makes this possibility remote. Choice of a column packing material centred on the porous polyaromatic polymer type of beads because of their compatibility with water.The Porapak series, P, Q, S and T (Waters Associates), were tested and the best results were obtained on 80 to 100-mesh Porapak Q. A I B Fig. 3. Gas chromatograms of alcohols: 1, methanol; 2, ethanol; 3, propan-1-01; and 4, butan-1-01, collected from the headspace above glycerol containing 40 p.p.m. of each alcohol, A, 10-ml sampling with gas-tight syringe; B, one 50-ml conden- sation ; and C, four condensationsFig. 2. A, Exploded view showing the assembly of the collection tube, aluminium jacket and cap; and B, the jacketed collection tube in position during injection [To face p. 304May, 19741 VOLATILES FOR GAS-CHROMATOGRAPHIC ANALYSIS 305 Fig. 4. Gas chromatograms of headspace collected from canned salmon liquor.Trace 1, one collection of 50 ml; and trace 2, enrichment from four collections (total 185 ml) of headspace Recently, Krumperman* has drawn attention to possible erroneous peaks associated with Porapak Q when the latter is used to trap volatiles. This material was therefore checked in order to determine if an analogous problem would be encountered when it was used as column packing. The column was conditioned for 4 days at a temperature of 250 "C and a carrier-gas flow-rate of 30 ml min-l. The collection tube, sample flask and syringe were carefully cleaned, and the condensation procedure was then carried out without a sample in the sample flask. No peaks were observed after the pressure spike. This system has been used for the past 5 months in our laboratory and it has not presented any maintenance or operational problems. 1. 2. 3. 4. 5. 6. 7. 8, REFERENCES Buttery, R. G., and Teranishi, R., Analyt. Chem., 1961, 33, 1439. Guadagni, D. G., Bomben, J. L., and Hudson, J. S., J . Sci. Fd Agric., 1971,22, 110. Lorenz, K., and Maga, J., J . Fd Sci., 1971, 36, 936. Hughes, R. B., J . Sci. Fd Agric., 1964, 15, 290. Mendelsohn, J. M., Steinberg, M. A., and Merritt, C., J . Fd Sci., 1966, 31, 389. Jennings, W. G., and Nursten, H. E., Analyt. Chem., 1967, 39, 521. Jennings, W. G., Wohleb, R., and Lewis, J. M., J . Fd Sci., 1972, 37, 69. Krumperman, P. H., J . Agric. Fd Chem., 1972, 20, 909. Received July 23rd, 1973 Amended December loth, 1973 Accepted January 14th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900302

出版商:RSC    

年代:1974

数据来源: RSC