ISSN:0003-2654    

DOI:10.1039/AN96287FX005

出版商:RSC    

年代:1962

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96287BX007

出版商:RSC    

年代:1962

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96287FP027

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

February, 19621 THE ANALYST xviiThe rate f o r classified advertisements is 5s. a line (or spaceequivalent of a h e ) , with. an extra charge of Is. f o r theuse of a Box Number. Semi-displayed classifiedadvertisements are 60s. per single-column inch.Analyticalchemist(Honours Degree, Grad.R.1.C. orA.R.I.C. status)We have a vacancy for an analyticalchemist in our York laboratories. Thework, which is interesting and varied,will include the chemical and physicalexamination of many types of food-stuff by both traditional and moderntechniques.Full consideration of all relevant detailswill be given when discussing startingsalary.Our pension fund, sickness benefitsand other similar facilities are in keep-ing with the best standards of BritishIndustry today.Applications giving full details of quali-fications and experience should beaddressed to Mr.M. Kershaw, StaffOfficer (M), Rowntree & CompanyLimited, The Cocoa Works, York.CITY OF SALFORD HEALTH DEPARTMENTAPPOINTMENT O F PUBLIC ANALYSTApplications are invited for the appointment of PublicAnalyst at a salary in accordance with Scale 3 of theJoint Negotiating Committee for Public Analysts (L1710rising by annual increments of LTO and L5.5 to a maximumof jJ955 per annum).Particulars of appointment and forms of applicationarc available from Dr. J. L. Burn, Medical Officer ofHealth 143 Regent Road Salford, 5, by whom completedapplicition forms should i e received by March 16th, 1962.NALYST male or female about H.N.C. standard forAsmall rese)arch group handling a wide variety of chemicalproblems. This is not a routine job, and the successfulcandidate will be encouraged to work with minimum super-vision and to obtain a further qualification. Please writeto the Personnel Officer, Hilger and Watts Ltd., 98 St.Pancras Way, Camden Koad, N.W.I.-THE SOUTH STAFFORDSHIRE WATERWORKSCOMPAKYAssistant Chemist (Graduate or equivalent qualifications)required for chemical, bacteriological and biologicalcxamination of water and associated problems. Housingavailable.Superannuation. Salary according to qualifi-cations and experience.Applications to Enginccr-in-Chief, 3U Slircpcote Strcet,B~rniingham, 12.IIELANEY GALLAY LTD.require the followingA QUALIFIED METALLURGISTfor their Cricklewood Laboratories.The Companyspecialises in Heat Exchange and Metal Joining and thesebring many diverse problems which require invcstigationand con,rrol. Ideally the successful applicant will be inthe 22-25 age group, will be a British National, and willhave a degree or membership of the Royal Institute ofChemists.Apply Personnel Manager, Delaney Gallay Ltd., VulcanWorks, ISdgware Road, Cricklewood, N.W.2. GLA 2201.LOUGHBOROUGH COLLEGE OF TECHNOLOGY,LEICESTERSHIRE DEPARTMENT O F APPLIEDCHEMISTRYApplications are invited from suitably qualified persons(male or female) for appointments as Research Assistants,liesearch Scholars or Resew-ch Fellows. Research iscurrently being carried out in the Department in mostbranche(5 of Chemistry, but special opportunities exist atpresent for fundamental work in the fields of Analysis(under ).he direction of the Reader in Analytical Chem-istry), Radiochemistry, Photochemistry, Electrochemistryand Polymer Chemistry.Salary within the range ,C;350-L1400 depending on qualifications, age and experience.Applicarion forms may be obtained from the AcademicIZcgistrar, quoting Ref. 34/BB.ACANCY in Consulting Food Laboratories in North-vWesl London for Chief Analyst preferably Branch E ,F.R.I.C. Apply Box No. 4049, The Analyst, 17 GreshamStreet, London, E.C.Z.ALLEN & HANBURYSLIMITEDWARE, H ERTFOR DSH I R Erequire a young graduate for aninteresting post in their AnalyticalF!esearch Department.The workwill involve analytical investigationinto a wide range of products andwill be of a non-routine nature.Applicants should be under 28years of age, have a B.Sc., B.Pharm.osr A.R.I.C. and preferably someexperience of pharmaceutical orfine chemical analysis. Pleaseapply in writing to the PersonnelManager, giving full details andquoting reference No. A.R.3ssi i THE ;\NALYST [February, 1962TABLES OF WAVENUMBERSFOR THE CALIBRATION OFINFRA-RED SPECTROMETERSInternational Union of Pure and Applied Chemistry.Price 40s.This collection of infra-red wavenumbers for use in the calibration of infra-red spec-trometers over the range 4000-600 cm-l has been prepared by the Commission onMolecular Spectroscopy of the International Union of Pure and Applied Chemistry.The Commission includes authoritative workers in the field in many countries whohave made a critical examination before drawing up the recommendation.CARBON-14 COMPOUNDSJ.R. Catch Price 30s.This book is a guide to the literature and use of carbon-14 compounds. It is intendedto direct the reader's attention to the many and varied publications on all aspects ofwork with carbon-14 compounds, and to help the newcomer in particular to benefitby the past experience of others; but it is hoped that even the experienced andknowledgeable reader will find it useful on occasion.RADIOACTIVATION ANALYSISPrice 30s.Radioactivation Analysis is a relatively new technique. During the last few years ithas been perfected. All sciences, but especially biology, biochemistry, geology andchemistry have much to gain by its use, which enhances the accuracy of analyticalmethods, and, in many cases, with its help much smaller amounts can be determinedthan had been possible hitherto by using conventional methods.QUANTITATIVEINORGANIC ANALYSISA Laboratory Manual for Students and Practising Chemists.R. Gelcher and A. J . Nutten 2nd Edn. Price 35s.This text has been designed as a practical teaching course in quantitative inorganicanalysis suitable for Universities and Technical Colleges, although much of its con-tents will be of use to the practising analytical chemist.-BUTTERWORTHS4-5, Bell Yard, London, W.C.

ISSN:0003-2654    

DOI:10.1039/AN96287BP041

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

FEBRUARY, 1962 THE ANALYST Vol. 87, No. 1031 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETING AN Ordinary Meeting of the Society was held at 3 p.m. on Thursday, February 8th, 1962, at the Wellcome Building, Euston Road, London, N.W.l. The Chair was taken by the President, Dr. A. J. Amos, O.B.E., BSc., F.R.I.C. The subject of the meeting was “New Analytical Reagents in Colorimetric Analysis,” and the following papers were presented and discussed: “The Search for New Reagents for Absorptiometry : Some Theoretical Considerations,’’ by Professor H. M. N. H. Irving, M.A., D.Phil., D.Sc., F.R.I.C., L.R.A.M. ; “A Critical Examination of Some Chloranilates as Colori- metric Reagents,” by J. T. Yardley, B.Sc., F.R.I.C., and F. A. Fryer, B.Sc.; “The Search for New Reagents for Absorptiometry : Some Practical Considerations,” by T.S. West, B.Sc., Ph.D., A.R.I.C. ; “Alizarin Complexan,” by &I. A. Leonard, BSc., Ph.D. ; “Newer Colori- metric Reagents for Iron,” by H. J. Cluley, RISc., Ph.D., F.R.I.C., and E. J. Newman, BSc., A.R.I.C. NEW MEMBERS ORDINARY MEMBERS John Leonard Bull; Joan Connor; Ann Patricia Dorrat, B.Sc. (Dunelm.) ; Youssef Argalious Gawargious, BSc. (Cairo), MSc. (Birm.) ; Lysbeth Chaston Graham, B.Sc. (Dunelm.) ; Ian Douglas Morton, MSc. (New Zealand), Ph.D. (Cantab.), F.R.I.C. ; Ian Malcolm Muten; Keith Eric Neuper, Dip.Pharm., Dip.Clin.Chem. (S. Africa), M.P.S. ; William M. Plank, BSc. (Fordham) ; Jack David Pryce, L.M.S.S.A., M.B., B.S., M.D. (Lond.) ; Giovanni Scotti; Archie Taylor, A.R.I.C. ; Alan Townshend, B.Sc.(Birm.) ; Babu Chandulal Vasa, B.Sc. (Bombay), BSc. (Manc.), A.M.C.T., A.R.I.C.; Hugh Westaway; Kenneth Whittaker, A.R.I.C.; Dennis Ewart Winsor. JUNIOR MEMBERS Susan Ruth Allen; Anthony Richard Dale; Sheila Mary Kirkwood; Iain Lovat Marr, B.Sc. (Aberdeen) ; Eoghan F. O’SulIivan, B.Sc. (N.U.I.) ; Peter Brandon Richards; John William Start up. SCOTTISH SECTION A JOINT Meeting of the Section with the Glasgow Sections of the Chemical Society and the Society of Chemical Industry, and the Glasgow and West of Scotland Section of the Royal Institute of Chemistry was held at 7.15 p.m. on Friday, December 8th, 1961, in the Royal College of Science and Technology, George Street, Glasgow, C.l. The Chair was taken by Professor R. A. Raphael, Ph.D., D.Sc., A.R.C.S., F.R.S.E., F.R.I.C.The following paper was presented and discussed: “The Structure of Natural Products by Direct X-ray Analysis,’’ by Professor J. Monteath Robertson, M.A., Ph.D., D.Sc., F.Inst.P., F.R.I.C., F.R.S.E., F.R.S. WESTERN SECTION A JOINT Meeting of the Section with the Cardiff and District Section of the Royal Institute of Chemistry and the South Wales Section of the Society of Chemical Industry was held at 7 p.m. on Friday, December 8th, 1961, at University College, Cardiff. The Chair was taken by Dr. G. V. James, M.B.E., M.Sc., F.R.I.C. A talk on “Analytical Research” was given by J. Haslam, D.Sc., F.R.I.C. MIDLANDS SECTION AN Ordinary Meeting of the Section was held at 7 p.m. on Wednesday, December 13th, 1961, in the Hills Lecture Theatre, The University, Edgbaston, Birmingham, 15.The Chair was taken by the Chairman of the Section, Dr. H. C. Smith, M.Sc., F.R.I.C. 8182 PROCEEDINGS [Vol. 87 The following paper was presented and discussed: “Fluorescent Indicators for the Determination of Metals,” by W. I. Stephen, B.Sc., Ph.D., A.R.I.C. MICROCHEMISTRY GROUP THE thirty-second London Discussion Meeting of the Group was held at 6.30 p.m. on Wednes- day, December 13th, 1961, a t “The Feathers,” Tudor Street, London, E.C.4. The Chair was taken by the Chairman of the Group, Mr. C. Whalley, BSc., F.R.I.C. This was a Review Meeting, at which many subjects related to those already covered in the Discussion Meetings were raised. BIOLOGICAL METHODS GROUP THE seventeenth Annual General Meeting of the Group was held at 6.30 p.m. on Thursday, December 14th, 1961, in the Restaurant Room of “The Feathers,” Tudor Street, London, E.C.4.The Chair was taken by the Vice-chairman of the Group, Mr. W. A. Broom, B.Sc., F.R.I.C. The following Officers and Committee Members were elected for the forthcoming year :-Chairman-Mr. J. S. Simpson. Hon. Secretary and Treaszcrer-Mr. K. L. Smith, Standards Department, Boots Pure Drug Co. Ltd., Notting- ham. Members of Committee-Dr. J. M. Bond, Mr. L. C. Dutton, Miss A. M. Jones, Dr. M. W. Parkes, Dr. G. F. Somers, and Mr. G. Sykes. Mr. D. M. Freeland and Dr. J. H. Hamence were re-appointed Hon. Auditors. Immediately following the Annual General Meeting a Discussion Meeting on “Assess- ment of Anti-atherosclerotics” was opened by G.S. Boyd, Ph.D., A.R.I.C. Vice-Chairman-Mr. W. A. Broom. Obituary GEORGE FREDERICK HALL GEORGE FREDERICK HALL died on October 8th at the age of 63. He was born on April lSth, 1898, and was educated at the Nottingham High School. On leaving school at the end of 1913, he entered the analytical laboratory of Boots Pure Drug Co., where, apart from war service during the 1914-1918 War, he spent the whole of his working life. He retired in 1955 after nearly 45 years of service with the firm. His early training was under the late Major S. R. Trotman, a former City Analyst for Nottingham. He was commissioned with a Yorkshire regiment and spent the years of his active service on the Western front. An.act of gallantry during this service brought a recommendation for the V.C., but as the incident occurred behind the lines the award could not be given.Instead Hall received an award of M.B.E. (Military). He left the army with the rank of Captain and obtained a Government Grant to enable him to complete his education. He took this up in September, 1919, at Nottingham University College, and studied there for the London B.Sc. degree, which he obtained with honours at the end of 1922. In January, 1923, he rejoined Boots Pure Drug Co. in his former position and was put in charge of a section of the laboratory. When the main laboratory moved to the Beeston Works in 1932, Hall remained in charge of the Nottingham laboratory until enemy action rendered it untenable and the two sections were arnalgamated at Beeston, where he remained as a principal analyst until his retirement. Hall’s chief characteristics as an analyst were the thoroughness with which he undertook any investigation that came his way and the care he gave to the training of his junior staff. He represented the Society for Analytical Chemistry, and, after its formation, the Midlands Section of the Society, on the Chemical Panel of the Regional Advisory Council for the Organisation of Further Education in the East Midlands, until May 1959. He also played a considerable part in the formation of the East Midlands section of the Institute of Chemistry, which he represented on the Council of the Institute from 1943 to 1946. His knowledge and experience of drug analysis were recognised by his appointment to various committees concerned with the revision of the British Pharmacopoeia and British Pharmaceutical Codex, on which he served until his retirement. In his private life he was at one time a keen gardener and touring motorist. He leaves a widow but had no family. His War service began in 1916 when he joined the Officers Training Corps. A. D. POWELL.

ISSN:0003-2654    

DOI:10.1039/AN9628700081

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

82 PROCEEDINGS [Vol. 87 Obituary GEORGE FREDERICK HALL GEORGE FREDERICK HALL died on October 8th at the age of 63. He was born on April lSth, 1898, and was educated at the Nottingham High School. On leaving school at the end of 1913, he entered the analytical laboratory of Boots Pure Drug Co., where, apart from war service during the 1914-1918 War, he spent the whole of his working life. He retired in 1955 after nearly 45 years of service with the firm. His early training was under the late Major S. R. Trotman, a former City Analyst for Nottingham. He was commissioned with a Yorkshire regiment and spent the years of his active service on the Western front. An.act of gallantry during this service brought a recommendation for the V.C., but as the incident occurred behind the lines the award could not be given.Instead Hall received an award of M.B.E. (Military). He left the army with the rank of Captain and obtained a Government Grant to enable him to complete his education. He took this up in September, 1919, at Nottingham University College, and studied there for the London B.Sc. degree, which he obtained with honours at the end of 1922. In January, 1923, he rejoined Boots Pure Drug Co. in his former position and was put in charge of a section of the laboratory. When the main laboratory moved to the Beeston Works in 1932, Hall remained in charge of the Nottingham laboratory until enemy action rendered it untenable and the two sections were arnalgamated at Beeston, where he remained as a principal analyst until his retirement. Hall’s chief characteristics as an analyst were the thoroughness with which he undertook any investigation that came his way and the care he gave to the training of his junior staff.He represented the Society for Analytical Chemistry, and, after its formation, the Midlands Section of the Society, on the Chemical Panel of the Regional Advisory Council for the Organisation of Further Education in the East Midlands, until May 1959. He also played a considerable part in the formation of the East Midlands section of the Institute of Chemistry, which he represented on the Council of the Institute from 1943 to 1946. His knowledge and experience of drug analysis were recognised by his appointment to various committees concerned with the revision of the British Pharmacopoeia and British Pharmaceutical Codex, on which he served until his retirement. In his private life he was at one time a keen gardener and touring motorist. He leaves a widow but had no family. His War service began in 1916 when he joined the Officers Training Corps. A. D. POWELL.

ISSN:0003-2654    

DOI:10.1039/AN9628700082

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

February, 19621 PARKER AND REES: FLUORESCENCE SPECTROMETRY. X REVIEW 83 Fluorescence Spectrometry A Review* BY C. A. PARKER AND W. T. REES (Admiralty Materials Laboratory, Holton Heath, Poob, Dorset) SUMMARY OF CONTENTS Introduction The nature of fluorescence and phosphorescence Quenching processes and inner-filter effects Quenching Inner-filter effects with right-angle illumination Inner-filter effects with frontal illumination Abnormal fluorescence General considerations Light sources Monochromators Detector system Design of sample compartment Choice of sample container Some specific instruments Correction of spectra Standard fluorescent substances Sensitivity Apparatus Applications to organic analysis Fluorescence of inorganic substances Applications involving inorganic - organic compounds Conclusion ALTHOUGH the techniques for measuring fluorescence emission spectra have been in use by physical chemists for many years, the method has until recently found comparatively little use in analytical chemistry. One of the main reasons is that no commercial instruments have been available, and indeed it is only during the last 10 to 15 years, with the development of high-sensitivity photodetectors, that photo-electric recording of fluorescence spectra has been able to compete in terms of sensitivity with the less convenient photographic method.On the other hand, the associated technique of fluorimetry has long been known to the analytical chemist, although this also has benefited from the use of more sensitive detection systems. Fluorimetry has not always been regarded as reliable by some workers, partly because it has sometimes been applied without sufficient attention to the principles of the method and partly through lack of suitable equipment.Since spectrofluorimetry is an extension and refinement of fluorimetry, most of the basic principles are common to both. Every fluorimeter or spectrofluorimeter, no matter how simple or complicated, contains three basic items (a) a source of light with which to irradiate the sample, (b) a sample holder and (c) a detector to observe or measure the resulting fluorescence light emitted by the sample. The complexity of the instrument depends on the degree to which the frequency and intensity of the exciting light can be controlled, the degree to which specific frequencies of fluorescence light can be selected and the sensitivity and precision with which the selected fluorescence can be measured.In filter fluorimeters, selection of frequency is made by inserting filters in the beams of exciting and fluorescence light. If a monochromator is used to select the frequency of either beam, the instrument becomes a spectrofluorimeter. I t can then be used to measure the spectral distribution of the fluorescence light-the “fluorescence emission spectrum”--or the variation of fluorescence intensity with frequency of exciting light-t he “fluorescence excitation spectrum.” Spectrofluorimetry bears a relationship to fluorimetry similar to that borne by absorption spectrophotometry to filter absorptiometry. It has the advantages over absorption spectrophotometry that it is * Reprints of this paper will be available shortly. For details, please see p.169.84 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol. 87 potentially far more sensitive and provides two spectra as criteria for identification instead of one. It has the disadvantage that, in order for it to fluoresce, a substance must absorb light, but not every absorbing substance exhibits fluorescence. It is also more difficult to measure an “absolute fluorescence intensity,” analogous to molecular extinction coefficient , and measurements are usually made by reference to some arbitrarily chosen standard substance. A series of excellent reviews on fluorescence analysis has been published by White,l and these include a wide literature coverage. More recently a general account of spectrofluori- metry has been published by Bartholomew.2 It is not therefore proposed to give here a com- plete bibliography of spectrofluorimetry, but rather to outline the theoretical principles on which the technique is based, to discuss some of the more useful types of apparatus and to give references to some typical and some unusual applications. It will be as well at the outset to clear up any confusion that may have arisen over the name of the technique itself.As noted by Pringsheim (reference No. 3, p. 17), any photo- meter can be used for measuring the intensity of fluorescence, but photometers for this purpose have been advertised as either “fluorometers” or “fluorimeters” ; the former mainly in the U.S.A.and the latter in this country. Since the term “fluorometer” has also been used to describe an instrument for measuring fluorescence lifetime (reference No. 3, p. 114), it might perhaps be argued that the term fluorimeter should be used for equipment to measure fluorescence intensity. However, Pringsheim has suggested the term “fluoro-photometer” for the latter, and an apparatus incorporating a dispersing unit would on this basis be called a spectrofluorophotometer. On the other hand, Bowman, Caulfield and Udenfriend? in their pioneering work with an instrument incorporating a dispersing unit in the beams of both the exciting and fluorescence light, called their instrument a “spectrophotofluorometer.” To be completely analogous to absorption spectrometry, we would suggest “fluorescence spectrometry,” but, for the sake of brevity, prefer “spectrofluorimetry.” Whatever term is met in the literature, it seems that the prefix “spectro-” implies that at least one spectro- meter is incorporated in the equipment.THE NATURE OF FLUORESCEKCE AND PHOSPHORESCENCE When a beam of light passes through a material its energy will reappear in a variety of forms. Part of the light will be absorbed, part will be reflected, part will be transmitted and a part will be scattered in various ways. Absorption takes place in discrete units or quanta, the size of which are equal to the product hv, where v is the frequency of the light and 62 is Planck’s constant. Quanta of visible or ultraviolet light carry energy sufficient to raise a molecule to an electronically excited state, from which the energy may be converted into rotational, vibrational or kinetic energy (k, heat) or into chemical energy (i.e., the molecule undergoes photochemical change to give products having a greater chemical potential), or part of the energy may be re-emitted as quanta of lower energy, Le., as fluorescence or phos- phorescence.The latter are in principle clearly distinguishable from the other forms of emission (e.g., Tyndall, Rayleigh and Raman scattering) by the fact that, in order to fluoresce or phosphoresce, the substance must first absorb some of the exciting light. The various scattering processes are of interest in spectrofluorimetry mainly because they can interfere with measurement of the fluorescence of extremely dilute solutions: Tyndall and Rayleigh scattering produces light having the same frequency as that of the incident light, although different frequencies will be scattered to different degrees. Raman scattering appears in a region of lower frequency and will be discussed in more detail later.The mechanism by which a substance normally fluoresces was recently described5 and will only be briefly outlined here. Depending on the frequency of the light absorbed, a molecule is raised to a particular vibrational level of one of the upper electronically excited states (see Fig. 1, which shows in conventional form the possible transitions between the various energy levels of a photo-excited molecule). After excitation, almost all complex molecules rapidly drop back to the lowest vibrational level of the first excited state, and it is from here, therefore, that the return to the ground state with emission of fluorescence takes place.If all the molecules originally excited by absorption of light return to the ground state by emitting fluorescence, then one quantum of fluorescence light is emitted for every quantum of exciting light absorbed and the fluorescence efficiency of the solution is unity. A proportion of the excited molecules may return to the ground state by other processes, for example, by conversion to the triplet state (Kt in Fig. l), by collision with solvent moleculesFebruary, 19621 PARKER AND REES : FLUORESCENCE SPECTROMETRY. A REVIEW 85 (KO in Fig. 1) or with molecules of another solute (a quenching agent-see below) or some may undergo photochemical change. The fluorescence efficiency is then less than unity and may be zero.The rate of emission of fluorescence is by definition equal to the rate of light absorption (quanta per unit time) multiplied by the quantum efficiency of fluorescence, ie.- where F = total fluorescence intensity in quanta per unit time, I , = intensity of exciting light in quanta per unit time, c = concentration of fluorescing solute, d = optical depth of solution, E = molecular extinction coefficient of solute and = quantum efficiency of fluorescence. . . - - (1) F = [I0 (1 - 10-wCd) J [+] . . .. Ideally, for very dilute solutions in which only a small fraction of light is absorbed- Thus, for any given solute and instrumental geometry, if the exciting frequency and intensity are kept constant, the detector response is directly proportional to the concentration.Further, since with modern photomultipliers and amplifiers very low fluorescence intensities can be measured, it is possible, by making the exciting intensity large, to detect low concentrations of fluorescent substances. This is the fundamental difference from absorption spectrometry, in which the detection limit is set by the minimum detectable difference in intensity between the incident and transmitted light and extremely precise measurements of light intensity must be made to attain high sensitivity. .. * * (2) F = [Io (2*3ecd)] [+] . . .. . . Photochemical reaction - - 2nd Excited - state, S2, (sing I e t) 1 s t Excited -state, SI, (singlet) Ground -state, So.(singlet) Fig. 1. Relationship between transitions giving rise to absorption, fluorescence, phosphorescence and quenching pro- cesses (vibrational levels of electronic states are not shown) The mechanism of fluorescence already discussed leads immediately to two rules of spectrofluorimetry of great practical value to the analytical chemi~t.~ First, the shape of the fluorescence emission spectrum of a solution containing only one fluorescent solute will be independent of the frequency of the light used to excite it, as emission always takes place from the same level, no matter to what level the molecule was originally excited. Secondly, according to equation (2) , the shape of the excitation spectrum, i.e., the plot of the intensity of the fluorescence band as a function of the frequency of the exciting light, will be directly86 PARKER AND REES FLUORESCENCE SPECTROMETRY.A REVIEW [Vol. 87 proportional to the molecular extinction coefficient of the solute, E , provided that (a) the quantum intensity of the exciting light is kept constant as the frequency is varied, (b) the solu- tion is sufficiently dilute for its absorption to be small at all frequencies, (c) only one fluorescent solute is present and (d) the fluorescence efficiency is independent of the exciting frequency (this is true of a great many substances). If a solution does not obey these two rules, the presence of more than one component should be suspected, although there are some substances not obeying them, e.g., substances present in more than one form in solution.A third rule of considerable theoretical and some practical value concerns the shape and position of the fluorescence emission spectrum in relation to the absorption spectrum. The emission spectrum normally appears on the low-frequency side of the absorption spectrum. Its shape bears approximate mirror-image relationship to that of the low-frequency absorption band, which it often partly overlaps. If the first and second absorption bands (corresponding to transitions to the first and second excited singlet states) themselves overlap, this mirror- image relationship will naturally not be apparent. However, if the fluorescence emission band appears far removed from the absorption band, then either the solute exhibits some form of abnormality (e.g., dissociation in the excited state-see below) or a trace of fluorescent impurity must be present.Although the phenomenon of phosphorescence does not strictly come within the field indicated by the title of this review, it is closely related to fluorescence, and its analytical possibilities are being inve~tigated.~s~s~ It will be as well, therefore, to indicate briefly the difference between phosphorescence and fluorescence; more detailed accounts of the origin of phosphorescence are given el~ewhere.~,~ The essential difference is in the duration of the emission after the exciting light is shut off. Fluorescence decays almost instantaneously, i.e., usually within about second, whereas phosphorescence may have a lifetime from 10-4 second up to tens of seconds.The difference is due to the nature of the excited state from which the transition giving rise to phosphorescence originates. Referring to Fig. 1, excited molecules arriving in the lowest vibrational level of S, may pass over to the lowest triplet level, T, from which they might be expected to return to the ground state, So, by emis- sion of a quantum of light (although of lower frequency than that of the fluorescence, because the energy difference between T and So is less than that between S, and So). However, radiative transitions between states of different multiplicity, eg., between triplets and singlets, are theoretically-in spectroscopic terms-"forbidden" and should not therefore be expected to occur. In practice, it is found that they do occur, but only with extremely low probability, the probability frequently being less than one millionth of that for normal (e.g., singlet - singlet) transitions, and hence the lifetime of the molecules in the lowest triplet state, in the absence of quenching processes, is correspondingly increased.Because of their long lifetime, the triplet molecules are much more susceptible to quenching processes, and phosphorescence in solution is usually completely quenched by exceedingly minute traces of impurity, especially oxygen; for this reason it has only rarely been o b ~ e r v e d . ~ J ~ , ~ ~ By dissolving the substance in a rigid medium and cooling to low temperature, quenching can be greatly reduced; under these conditions phosphorescence can be observed from many substances.' QUENCHING PROCESSES AND INNER-FILTER EFFECTS Both quenching processes and inner-filter effects are of considerable practical importance to the analytical chemist using spectrofluorimetry.Lack of appreciation of the elementary principles governing these two effects, or confusion between the two, has in the past led to the use of unsatisfactory equipment or procedures and to the incorrect interpretation of fluorescence measurements, and this is one of the main reasons why the technique of fluorimetry has sometimes been regarded with suspicion by many analysts. We shall therefore define what we mean by quenching. By quenching we mean all those processes that result in the true fluorescence efficiency being reduced to below unity.Except for energy transfer,12 which is significant under rather special conditions and usually a t concentrations greater than 10-3 M, they have been discussed above and are indicated in Fig. 1. They are those processes that divert the light energy absorbed by a potentially fluorescent molecule into channels other than the emission process responsible for fluorescence. Inner-filter effects, on the other hand, have no influence on the primary process of emission, but simply reduce the intensity of the observed fluorescence by absorption of the exciting light or of the fluorescence light within the material being tested.February, 19621 PARKER AND RE,ES : FLUORESCENCE SPECTROMETRY. A REVIEW 87 QUENCHING- shown in Fig. 1 and is given by the expression- The fluorescence efficiency can be derived in terms of the rates of the various processes k f where kf is the rate constant for the fluorescence process itself (and is equal to the reciprocal k , is the sum of the rate constants of all internal and solvent quenching processes, kt is the rate constant for triplet formation and kQ is the rate constant for the bimolecular quenching occurring on collision with a ' = k f + ko + kt + kQ [QI of the fluorescence lifetime), molecule of added quenching agent of concentration Q.When Q = 0, then- kf $' = kf + k , + kt ' Hence the ratio of the fluorescence intensity without quencher to that with quencher is given by the Stern - Volmer quenching equation- & = '' [Q1 = 1 + K [Q] ' + kr + k , + kt The maximum value for k , will be observed when every encounter between excited molecule and quenching agent is effective in causing quenching.The encounter frequency for molecules in solution (in litres per second per mole) is given by the expression13- k . = SRT/3000q where q is the viscosity of the solvent. For cyclohexane, water and ethanol at 20" C, k has the values 0-65, 0.64 and 0.38 x lolo, respectively. Thus, if we know the fluorescence lifetime (it can be calculated from the absorption spectrum14), the maximum possible degree of quenching for any given concentration of quenching agent can be calculated by assuming unit fluorescence efficiency in the absence of quencher ( i e . , k , = kt = 0). Since the fluores- cence lifetime (ie., l / k f ) is typically about lobs second, the minimum concentration of quenching agent required to produce 50 per cent.quenching in aqueous solution is about Further, this applies to a substance for which $, is unity; if 4, were less than unity, a correspondingly greater concentration of quenching agent would be required. In practice, therefore, quenching can usually be avoided by suitable dilution of the solution, the exception being quenching by oxygen, which is often appreciable and occasionally large in an air- saturated solution; for example, air-saturated ethanol is about 1 0 - 3 ~ in oxygen. If the fluorescence has an exceptionally long lifetime, as occasionally happens, then quenching in an air-saturated solution can be considerable (pyreneX5 is a case in point). Formation of a complex between the fluorescent solute and a second solute may also cause a reduction in the intensity of the observed fluorescence.Although this is not strictly quenching in the sense defined above (since the complex is a new species with its own fluores- cence characteristics), the absorption spectrum of the complex may differ only slightly from that of the molecule itself, and it may therefore appear that collisional quenching is taking place (the borate - benzoin compound in oxygenated solution may be an example of this effect16). INNER-FILTER EFFECTS WITH RIGHT-ANGLE ILLUMINATION- The two kinds of inner-filter effect are best described by reference to Figs. 2 (a), 2 (b) and 2 (c), which represent the two most frequently used arrangements of sample container in relation to the beam of exciting light (I,) and the direction in which the fluorescence is viewed (F) ; they are generally known as the right-angle and frontal methods of illumination.In the first arrangement, Figs. 2 ( a ) and 2 ( b ) , the exciting light passes through the sample in a direction at right-angles to that along which the fluorescence is viewed, and it is usually arranged that the photomultiplier "sees" only the illuminated liquid and not the illuminated cell faces. The advantage of this arrangement is that interference by stray light arising from reflection at the cuvette faces or fluorescence of the cuvette itself is minimised. Since the exciting light has to pass through a depth of liquid AD (= d ) before reaching the region viewed by the photomultiplier, its effective intensity will be reduced by a factor where M.88 PARKER AND REES FLUORESCENCE SPECTROMETRY. A REVIEW [Vol.87 D is the total optical density per centimetre of the solution at the exciting frequency. (If the fluorescence is viewed over a comparatively wide area BC a somewhat more complicated correction factor must be used.le) The right-angle arrangement is clearly suitable only for weakly absorbing solutions, which generally means dilute solutions of the fluorescing sub- stance, although large concentrations of a transparent solute can be tolerated. If Dd is less than 0.02 (4.6 per cent. absorption) no correction is generally necessary, For higher values of Dd the correction factor can be applied, but for strongly absorbing solutions the fluorescence becomes concentrated near the front face of the cuvette and practically none will be observed by the phot~multiplier.~~ The use of too concentrated a solution will have a pronounced effect on the shape of the fluorescence excitation spectrum.An example of this is shown in Fig. 3 (curves A, B and C) by reference to solutions of quinine bisulphate. As the concentration is increased from 1 to 10 pg per ml, the height of the main excitation peak (corresponding to maximum absorp- tion) is decreased relative to the low-frequency peak. As the concentration is further increased to 100 pg per ml, the absorption of exciting light at the frequency of the main peak becomes so great that the observed excitation spectrum (curve C) passes through a minimum at this frequency and the remainder of the spectrum is also distorted.Fig. 2. Methods of illumination and viewing: (a) and ( b ) , right-angle method; (c), frontal method; ( d ) straight-through method Distortion of the spectrum can equally well be produced by the presence of a second solute showing strong absorption bands. The effects of various concentrations of benzene on the main excitation band of a dilute solution of anthracene are shown by the curves in Fig. 4. The benzene absorption is superimposed on the anthracene excitation curve, as indicated by the fact that the minima in the distorted anthracene spectra correspond to maxima in the benzene excitation spectrum, h., to the absorption bands of benzene. The second type of inner-filter effect is that arising from absorption of the fluorescence light by the solution; it may be absorption by an excessive concentration of the fluorescent solute itself (self-absorption) or it ma.y be .absorption by other solutes.For right-angle illumination, the magnitude of the effect can be approximately calculated if the path length of the fluorescence beam through the liquid is known and re-emission of fluorescence by absorption of the original fluorescence is neglected. Self-absorption mainly affects the high-frequency side of the fluorescence emission band5 because the latter is always on the low-frequency side of the absorption band, which it partly overlaps. The presence of a second solute absorbing in the region where the first fluoresces will naturally produce distortion ofFebruary, 1962j PARKER AND REES: FLUORESCENCE SPECTROMETRY. ,4 REVIEW 89 the emission spectrum.An example of this is given in Fig. 5, which shows the absorption bands of neodymium chloride superimposed on the emission spectrum of quinine bisulphate when both were present in the same solution. Admittedly, this particular example is some- what hypothetical, as the measurement of quinine bisulphate in such a concentrated solution of neodymium chloride would not normally be attempted because of the strong quenching of the quinine fluorescence by the neodymium chloride. Neodymium was chosen to demon- strate the general principle because it shows a series of sharp absorption bands. A substance showing a broad absorption band could produce an equally serious, although less spectacular, distortion.I I I 4.4 4.0 3.5 3.0 2.5 Frequency of exciting 1ight.p-l Fig. 3. Effect of excessive absorption of exciting light on excitation spectrum of quinine bisulphate in 0.1 N sulphuric acid at 20° C (measurements made in cell 1 cm square, with right-angle illumination): curve A, 1 pg per ml; curve B, 1Opg per ml (sensitivity decreased); curve C, 100 pg per ml (sensitivity further decreased); curve D, 10 pg per ml, with frontal illumination and optical depth 15 mm. For excitation the band width was 0.04 p-l at 3.0 p-l; for fluorescence i t was 0.2 p-l at 2-17 p-l 4.4 4.2 4.0 3-8 Frequency of exciting light,f Fig. 4. Effect of a second absorbing solute on excitation spectrum of a solution of anthracene in ethanol (0.1 pg per ml) ; measurements made in cell 1 cm square, with right-angle illumination.Concentrations of benzene were: curve A, nil; curve B, 0.05 per cent.; curves C and F, 0.125 per cent.; curve D, 0.26 per cent.; curve E, 0.5 per cent. For curves A, B, C, D and E, the fluorescence monochromator was set at 2.5 p-l, with half-band width 0.16 p-l, i.e. anthracene fluorescence: for curve F, it was set at 3-6 p-l, with half-band width 0.1 p-l, i.e., benzene fluorescence INNER-FILTER EFFECTS WITH FRONTAL ILLUMINATION- With the method of frontal illumination shown in Fig. 2 (c) the inner-filter effects are in some respects less serious; indeed, the main advantage of this arrangement is that it can (with suitable precautions) be used to measure the emission spectrum of a solution containing a high concentration of the fluorescent substance.However, great care must be exercised in interpreting the results, as serious distortion of both excitation and emission spectrum can also occur with this arrangement. With dilute solutions of a fluorescent substance, through which most of the light is transmitted, the observed fluorescence intensity is propor- tional to concentration and extinction coefficient-see equation (2) , p. 85-and the true excitation spectrum is recorded. If, however, the concentration is increased until the fraction of light absorbed becomes significant , the fluorescence intensity is no longer proportional to concentration and extinction coefficient-see equation (1)-and the excitation spectrum90 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol.87 is distorted. The greatest distortion occurs where the absorption is greatest, i.e., at the maxima, which become decreased in height relative to the minima. In the limit, with complete absorption of the exciting light at all frequencies, the observed fluorescence intensity will be independent of frequency (assuming 4 to be constant) and the measured excitation spectrum will consist of a horizontal straight line and will clearly be of no value as a criterion for identification. This effect has already become appreciable for the solution containing 10 pg of quinine bisulphate per ml (Fig. 3, curve D) , which gave a main excitation maximum lower than that obtained by right-angle illumination. Frequency of fluorescence, y-' Effect of second absorbing solute on emission spectrum when right-angle illumination is used: curve A, 80 pg per ml of quinine birulphate in 0.1 N sulphuric acid (uncorrected); curve B, as for curve A, but solution 0-12 M in neodymium chloride and sensitivity increased to compensate for quench- ing ; curve C, sensitivity of E.M. I. 9558 photo- multiplier with quartz monochromator. Optical depth for fluorescence was 2.6 cm, and excitation was a t 2.73 p-l; measurements made a t 20°C. The band width of the fluorescence mono- chromator was 0.01 p-l at 2.0 p-1 Fig. 5. If a second absorbing solute is present, it will not interfere so long as the total fraction of light absorbed is small; for complete absorption, however, each solute will absorb a fraction of the exciting light proportional to its own value of D, and, with high concentrations of the second solute (and no energy transfer), the fluorescence of the first will be almost extinguished (not quenched). A simple example of this is observed in the measurement of an excitation spectrum in a solvent such as carbon tetrachloride, which shows a cut-off in the quartz ultraviolet region; beyond the cut-off, the observed excitation spectrum will fall to zero.Frontal illumination of concentrated solutions has mainly been used for measuring emission spectra, but even here some remarkable distortions of the spectra can be observed. An example is given in Fig. 6, which shows the effect of adding naphthacene to a solution of N-phenyl-2-naphthylamine. Exciting light of frequency 2.73 p-l was comparatively weakly absorbed and penetrated the solution; as a result, much of the observed fluorescence had to pass back again through a thick layer of solution and was thus partly absorbed by the naphthacene, which therefore superimposed its absorption spectrum on the emission spectrum of the phenylnaphthylamine (Fig.6, curve B). Exciting light of frequency 3.19 p-lFebruary, 19621 PARKER AND REES : FLUORESCENCE SPECTROMETRY. A REVIEW 91 was more strongly absorbed and did not penetrate so far; hence the fluorescence had to travel back through a smaller depth of solution and was less strongly absorbed by the naphthacene (Fig. 6, curve C). Similarly, it is possible to record different fluorescence emission spectra from concentrated solutions of substances showing a large absorption - fluorescence overlap, the amount of self-absorption depending on the depth of penetration of the exciting light and hence upon its frequency.Thus, when excited by frontal illumination with light of frequency 2.0 p-1, a concentrated solution of fluorescein shows a normal emission spectrum because the exciting light is all absorbed in a very thin layer of solution (Fig. 7, curve A). If, however, a freauencv of 2.7 ,u-1 is used for excitation. the emission sDectrum shows the effect of con- siderible <elf-absofption, because exciting into the solution. light of this freiuency penetrates much further Frequency of fluorescence, p-' Fig. 6. Effect of second absorbing solute on emission spectrum when frontal illumin- ation is used: curve A, 100 pg per mlof N-phenyl-2-naphthyl- amine in benzene (excitation with 2.73- or 3.19-p-1 light); curve B, as for curve A, but 40pg per ml of naphthacene present (excitation with 2.73-p-1 light) : curve C, as for curve B, but with more strongly absorbed exciting light (3.19 p-1).Spectral-correction curve as in Fig. 5. Symbols (1) and (2) indicate, respect- ively, absorption and fluores- cence of naphthacene. The band width of the fluorescence monochromator was 0.1 p-1 at 2.0 p-l The arguments for and against the Frequency of fluorescence, p-' Fig. 7. Effect of self- absorption on fluorescence emission spectrum of 4 x M fluorescein a t pH 11 when frontal illumination is used: curve A, excitation with 2.04-p-1 (490 mp) light; curve B, excitation with 2-73-p-l (366 mp) light. Sensitivity and intensity of exciting light were the same foreboth curves method of frontal illumination will be discussed further in thve section dealing wzh the design of a spectrofluorimeter.So far as measurements on liquids for analytical purposes are concerned, the only worth-while advantage seems to be the ability to measure concentrated solutions, but in view of the difficulties likely to arise from inner-filter effects, particularly with a solution of unknown composition, we prefer whenever practicable to dilute the solution sufficiently to reduce the absorption to manageable proportions and to use right-angle illumination. Each component then produces its own fluorescence spectra undistorted by absorption.92 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol.87 ABNORMAL FLUORESCENCE- Under this heading we group a variety of effects, all of which involve a rapid change of molecular species between light absorption and emissi0n.l' Since the absorbing molecule undergoes a reversible photochemical change, the intensity of its characteristic fluorescence emission is reduced, and the effects to be discussed here are therefore special instances of fluorescence quenching by photochemical reaction. They differ from other quenching pro- cesses in that, as the normal emission spectrum is quenched, it is often replaced by a new emission from the excited photochemical product. Recognition of these effects is of importance to the analytical chemist because they are common, particularly among ionisable fluorescent solutes, and can lead to the observation of more than one fluorescence emission spectrum from the same absorbing species, depending on the composition of the solution.The effect is well illustrated by aqueous solutions of 2-naphthol,ls a weak acid having a pK value of 9.6. A strongly alkaline solution shows the absorption spectrum of the naphtholate ion and the blue fluorescence of the excited naphtholate ion. At pH 7, the sub- stance is present almost entirely as undissociated naphthol molecules and exhibits the corre- sponding absorption spectrum. However, the fluorescence spectrum shows two bands, the expected violet band due to the excited undissociated molecule and the blue band of the excited naphtholate ion, which is not present in the ground state at this pH. The blue band persists even down to pH 3 and is only suppressed in strongly acid solutions.Clearly, the excited naphthol molecule is a far stronger acid than the normal molecule, and, when the latter is excited in the pH range 3 to 7, some of the excited molecules have time to dissociate before fluorescing- although there is not time for complete establishment of equilibrium in the excited state before fluorescence occurs.19 Since the ionisation requires the co-operation of a proton acceptor, its rate will be influenced by the addition of other bases, e.g., acetate ion, to the solution,20 and the concentration of buffer as well as the pH of the solution will affect the relative intensities of the two fluorescence bands. This must obviously be taken into account when devising methods and interpreting analytical results.Excited-state ionisation is also shown by 2-naphthylamine and its naphthylammonium ion.'* Both are more strongly acidic in the excited state and dissociate according to the equations- ROH* + H2O ---+ R-O-* + H,O+ RNH2* + OH- -+ RNH-* + H,O RNH3+* + H20-+RNH,* + H30+ In strongly alkaline solution (PH 14), the absorption spectrum shows that the naphthylamine is present only as the neutral molecule, but the fluorescence spectrum shows not only the violet band due to the excited neutral molecule, but also a green component due to the excited naphthylamide ion, RNH-*, formed by dissociation of the excited neutral molecules. These naphthylamide ions are unstable in aqueous solutions at all pH values. The dissociation of excited naphthylammonium ions can be observed by taking measurements at lower pH values.The pK value of 2-naphthylamine in its normal state is 4.07, and at pH 2 practically no free amine molecules are present in the solution. Nevertheless, in the fluorescence spectrum only the violet band of the free amine is observed, and clearly the naphthylammonium ion after excitation splits off a proton according to the above equation. In order to observe the fluorescence spectrum of the naphthylammonium ion itself it is necessary to use strongly acid solutions; the band appears close to the absorption band of the ion in the ultraviolet region. In contrast to the hydroxy and amino compounds, the aromatic carbonyl compounds and carboxylic acids show a decrease in acidity on excitation.Thus, 2-naphthoic acid21 is an extremely weak base having a pK value of -6.9 corresponding to the equilibrium- //OH+* O* +H+ + R - C # R - C 'OH 'OH Nevertheless, the corresponding half-way change in the fluorescence spectrum takes place at only moderate acidities, corresponding to a pK value of 0 for the excited molecule. The fact that a phenolic group increases its acidity on excitation, whereas a carboxylic groupFebruary, 19621 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW 93 increases its basicity can lead to interesting effects when both are present in the same molecule. In methanolic solution, for example, salicylic acid absorbs in the ultraviolet region (3.3 p-1) as the neutral molecule, but emits in the visible region (2-3 p-l). There is thus a Stokes shift of unusual magnitude, and this is attributed to an intramolecular proton transfer in the excited state, the blue fluorescence being attributed to the excited zwitter ion.OH OH / C / /OH C C /OH C ____, (Rapid) (x \OH+* OH OH 0-H 0- / C / Fluorescence (Rapid) //\/ No ’y‘OH+* ( x \ O H + - 0- \\ 4A 0- As might be expected, esters of salicylic acid show the same large Stokes shift, but the ethers a normal one.22 A molecule in its excited state can undergo changes other than dissociation. For example, at concentrations up to lo4 M, solutions of pyrene in benzene show the well defined ultraviolet fluorescence emission spectrum of the pyrene molecule itself. At high concen- trations a component in the blue region appears; this is due to a dimer formed only in the excited state by combination of an excited pyrefie molecule with a second unexcited molecule.= NQ dimerisation takes place in the ground state, as the absorption spectrum of the concen- trated solution obeys Beer’s law.APPARATUS The number of commercially available spectrofluorimeters is still rather small, and most of them incorporate two wide-aperture monochromators and are therefore expensive. If, however, an absorption spectrophotometer is available, it is possible to make use of the monochromator and photomultiplier of this instrument and to do useful work with a com- paratively moderate outlay for additional equipment. Even when it is required to use two monochromators it may sometimes be desirable for the user to set up his own instrument, as commercial instruments have limited versatility with regard to the types of test that can be carried out.t Detector t ’ Monochromator for fluorescence Monochromator Source for exciting light Fig. 8. Schematic diagram of a general-purpose spectrofluorimeter GENERAL CONSIDERATIONS- It consists of a source of “white” light (visible and ultraviolet), a monochromator to select the required A diagram of a typical general-purpose spectrofluorimeter is shown in Fig. 8.[Vol. 87 frequency for excitation, a sample container and a second monochromator fitted with a photomultiplier to analyse the fluorescence light. With such an apparatus, one can (a) choose any frequency of exciting light and measure the spectrum of the fluorescence emitted by the sample or (b) set the fluorescence monochromator on the frequency of the fluorescence band of the substance and observe how the intensity of this fluorescence varies with the frequency of the exciting light used.The principle is simple, but in practice it is difficult to achieve high sensitivity with this system because so much light is wasted. The light from the source is emitted in all directions, and only a limited proportion finds its way into the first mono- chromator. Of this light, only a narrow band of frequencies is selected for passing to the sample. It is usually arranged that less than 1 per cent. of this light is absorbed, the remainder passing on and being of no further use. The resulting fluorescence light is again emitted in all directions, and, again, of necessity only a comparatively small proportion is collected by the second monochromator This light is finally dispersed again, and a narrow band of frequencies is collected, so that the final intensity falling on the detector is a small fraction of the light originally emitted by the source.There are several ways in which the limitations imposed by these light losses can be minimised. Obviously it is desirable to use the most powerful lamp, the most sensitive detector and the largest monochromators available. Even with very large and expensive monochromators, the sensitivity will still be rather low if it is attemped to take measurements with very narrow slits on both monochromators simultaneously. Fortunately, it is not usually necessary to do this, and a considerable increase in sensitivity can be attained by using wide slits on one or other monochromator.For example, ParkerM has recorded well resolved excitation spectra of anthracene a t a concentration of less than one part in 100 million; he used monochromators of moderate size, but with wide slit settings on the fluorescence monoc hromat or. Instead of using a monochromator with wide slits, it is frequently possible to dispense with one monochromator altogether and to select either the exciting or the fluorescence light by means of filters. When a mercury lamp and filters are used, a much larger intensity can be obtained than is possible even with a large monochromator. For example, with a 125-watt mercury lamp and filters, the dose rates obtainable were about 100 times greater than those obtained when the same lamp was used with a small monochromator.To achieve dose rates similar to the former, it was necessary to use a 1-kW lamp and a large monochromator having a 6-inch prism.ls As discussed below, the limitations imposed by the intensity of the exciting light are most severe at frequencies greater than 3-3 p-l (300 mp.). This region is important because many simple organic compounds do not absorb at lower frequencies, and, to obtain adequate sensitivity, the lower discrimination associated with the use of filters may have to be accepted. Such single-monochromator experiments have been used by Weber and Teale,25s26s27 by White, Hoffman and Magee2* and by many others. LIGHT SOURCES- The determination of fluorescence excitation spectra requires a lamp giving a high- intensity continuous spectrum throughout the visible and ultraviolet region , preferably with a constant quantum output at all frequencies to minimise correction factors.Unfortunately, the sources at present available fall far short of this ideal. The high-power tungsten-filament lamp can be used for excitation in the visible region, but its intensity falls off rapidly towards the violet and in the near-ultraviolet region is too low for most purposes. The hydrogen lamp provides a reasonably constant energy distribution in the ultraviolet region, but again its intensity is too low for many purposes. For most work on excitation spectra, a compact- source xenon-arc lamp available commercially in sizes from 250 watts to 2-5 kW is used.Although desirable from the point of view of high intensity, the larger sizes of lamp, together with their associated control equipment, are expensive, and also the dissipation of the heat is an inconvenience. The source is small and of high intrinsic brightness, so that a comparatively large proportion of the light can be focused on the entrance slit of the excitation monochromator. It has two disadvantages; the intensity falls off in the higher frequencies of the ultraviolet region, and the visible spectrum consists of a continuum on which are superimposed a large number of lines (many lamps also contain traces of mercury, which also produces a weak line spectrum). When working at low resolution these lines are not troublesome, but at high resolution they can be.94 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW Lamp powers of: 250 to 500 watts are most generally used.February, 19621 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW 95 When measuring fluorescence emission spectra it is often not necessary to have the frequency discrimination of exciting light given by a continuous source ; much higher intensi- ties, with greater spectral purity, can be obtained by isolating one of the principal lines from a mercury lamp.16 924 Again, to obtain high intrinsic brightness, a compact-source high- pressure lamp is desirable, although this gives broader lines and a more intense continuous background. The relative and absolute intensities of the lines vary somewhat from lamp to lamp.Typical dose rates obtainable in the sample cuvette when a medium-sized mono- chromator is used are of the order 0.1 to 0.5 micro-einstein per minute at 2.73 p-l (366 mp), depending on the slit When suitable high-transmission filters are available, e g . , for lines at 366, 405, 436, 546 and 678 mp, higher intensities can be obtained by this means, but generally at the expense of spectral p ~ r i t y . ~ ~ , ~ ~ The intensity of the spectrum of high- pressure mercury lamps below 3.3 p-l (300 mp) is rather low, partly owing to the reduced transmission of the thick silica envelope, which decreases further during the life of the lamp. Rather better intensities can be obtained in this region by means of the linear type of high- pressure mercury tube. When a 500-watt tube of this type was used in conjunction with a large quartz prism monochr~mator~~ the dose rates attained were as shown in Table I.TABLE I DOSE RATES WITH 500-watt HIGH-PRESSURE MERCURY TUBE Frequency, Wavelength, Dose rate, CL-l m P micro-einstein per minute 2.29 436 0.03 6 2-47 405 0.023 2.73 366 0.064 2.99 334 0.013 3-19 313 0.015 0.014 0.005 3.3 1 3.37 3.55 282 3.73 268 0-007 3-94 254 (band) 0.014 Z} Additional excitation frequencies can be obtained if a mercury - cadmium lamp is used. Unfortunately, compact-source lamps of this type are usually manufactured in large sizes (2.5 kW) for studio lighting and, with their control gear, are expensive, They do, however, provide a line source of high intensity, giving useful cadmium lines at 644, 509, 480, 468, 361, 347 to 350, 340 and 326 mp as well as the usual mercury spectrum.They have been used for photochemical work, and dose rates of lines isolated from them by means of a mono- chromator have been measuredw; typical relative intensities are shown in Table 11. TABLE I1 QUANTUM INTENSITIES OF LINES FROM MERCURY - CADMIUM LAMP ,-=-, Relative I-------h-, Relative Wavelength, intensity, Wavelength, intensity, Element ml-L quanta Element mP quanta Cd 276 1.2 Cd 347 to 350 19 Cd - Hg 284 to 285 1.7 Cd 361 25 Cd - Hg 288 to 289 2.8 365 to 366 19 436 12 Hg 297 6.5 Hg 302 5.1 313 7-8 Cd 480 56 319 3.5 Cd 609 78 Hg Cd 325 to 326 6.5 546 22 3 644 106 334 2.8 340 12 Hg Cd Line Line Hg 405to408 7.5 308 3.0 ?i 466to 468 34 Hg 577 to579 23 3 Cd - Hg The quantum dose rate arriving at the sample cuvette is one of the two factors deter- mining the over-all sensitivity of a spectrofluorimeter, the other being the sensitivity of the fluorescence monochromator - detector combination.The analytical chemist is naturally mainly interested in the over-all sensitivity of his instrument, but, in comparing it with those of other workers, it would be useful to compare dose rates of exciting light, so that, if it is suspected that the over-all performance is inferior, it can be decided where the fault lies.96 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol. 87 Unfortunately, few publications have so far quoted such dose rates (see, however, papers by Parker and ~ t h e r s l ~ , ~ , ~ ~ ~ ~ ~ ) , although they can readily be determined by means of the ferrioxalate a~tinometer.~~ 330 Although it is generally desirable to obtain the greatest possible excitation intensity consistent with adequate spectral purity, solutions are occasionally photosensitive, and the exciting intensity may have to be decreased or the time for measurement limited.The most usual example of interference arising from photosensitivity is that in which a low concentration of a fluorescent substance has to be measured in the presence of a large concen- tration of a substance that decomposes to give a fluorescent product. Measures to minimise such interference have been described.16 9% MONOCHROMATORS- For a given band width and dispersing element (prism or grating) , the amount of light passed by the monochromator is proportional to d2/f, where d is the diameter of the prism or grating, and also of the collimating mirrors or lenses, and f is the focal length of the latter.The choice between prisms and gratings is controversial. Large quartz prisms are expensive, and, moreover, the dispersion with a grating is larger than that with a prism, except at the higher frequencies. On the other hand, a grating produces a series of overlapping spectra of different orders, and this causes some inconvenience. For example, a grating instrument used in the first order and set at 500 mp will also pass light of wavelength 250 mp from the second order. Thus, if used as a fluorescence monochromator, it will show a scattered-light peak at 500 mp when 250-mp light is used for excitation. However, such interference only applies to wavelengths longer than 400 mp when working in the first order, because excitation wavelengths shorter than 200 mp are not used.When working at wavelengths longer than 400 mp, interference from spectral orders higher than the first can be overcome by inserting a glass filter in the beam of fluorescence light. When working at high sensitivity, such a filter must be completely non-fluorescent , or interference will be observed from excitation of filter fluorescence by exciting light scattered by the sample. The grating instrument, with a linear drive on the grating, gives a recorded spectrum with a linear wavelength scale, and this is regarded by many workers as an advantage. How- ever, excitation and emission spectra are best plotted on a linear frequency scaleJ6 and direct drive to a quartz prism gives a scale that in many parts of the spectrum is more nearly linear than that of a grating.The ideal is to have a cam drive to give a precisely linear scale whichever type of instrument is used, but this means additional expense, and, for many workers who wish to set up their own instrument, it is not possible. In view of the greater dispersion of quartz at the higher frequencies, it has been suggested that the ideal instrument should consist of a prism monochromator to isolate the exciting light and a grating monochromator to analyse the fluorescence, the first monochromator in general being required to work at higher frequencies than the second. However, the excitation monochromator frequently has to work at the lower frequencies also, and the best arrangement is probably to use two grating monochromators (particularly if mechanical arrangements can be incorporated to record with a linear frequency scale) and to accept the inconvenience introduced by the necessity for inserting a filter in some spectral regions.Nevertheless, most of our work has been done with two prism instruments, and reasonably high sensitivity has been attained. DETECTOR SYSTEMS- In view of what has already been said about light losses in the spectrofluorimeter, it will be realised that even with the largest monochromators and light source, the amount of light reaching the detector is still small, and it is necessary to use a detector system of the highest sensitivity if it is desired to measure the spectra of very dilute solutions. The most important part of the system is the photomultiplier tube itself, which serves to convert the weak beam of light into photo-electrons and to amplify this electron current by factors up to 108.The output from the photomultiplier is measured by means of a sensitive galvano- meter or, for the highest sensitivity, is further amplified and then measured on a suitable meter or passed to a pen recorder or oscilloscope. The choice of amplifier is less critical than that of the photomultiplier, as the sensitivity is ultimately limited by the minimum detectable signal-to-noise ratio, and, if an inferior photomultiplier is used, it will give a large current even in complete darkness owing to the thermal ejection of electrons.Although the directFebruary, 19621 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW 97 current component of this output can be “backed off” in the amplifier, the statistical fluc- tuations remain and will appear on the meter or recorder pen. It is true that these fluctuations can be considerably reduced by cooling the tube in dry ice or liquid nitrogen, but this involves considerable complication of the apparatus; it is better to use a more expensive photomultiplier and so avoid these complications. I t is not proposed here to give a list of recommended tubes. and since then great improvements have been and are still being made, so that it is best to apply to the leading manufacturers for the latest information. There are now available 11- and 13-stage tubes, fitted with quartz windows, of extremely high sensitivity and low dark current, and, although such a tube will cost L50 or more, this investment will be well worth-while.For example, a 13-stage tube with a caesium - antimony photocathode has a minimum sensitivity of 2000 amps per lumen with a dark current of 0.1 p A and is sensitive to frequencies of 1-7 p-l and higher. If it is desired to make measurements at lower frequencies, h., in the orange and red region, an 11-stage tube with a “tri-alkali” photocathode can be obtained; this has a somewhat lower sensitivity of 200 amps per lumen with a dark current of 0.004 PA. Full details for operating these tubes are provided by the manufacturers. They require a stabilised d.c. source, variable up to 2000 volts, which can be obtained commercially.Variation in the supply voltage provides a ready means of altering the sensitivity over a wide range of values.15 There is wide scope in the choice of amplifier. A high-resistance d.c. voltmeter such as is used to measure glass-electrode potentials can be set up to measure the voltage drop across a 1- to 10-megohm resistor through which the output from the photomultiplier passes. Some workers have used a.c. amplifiers. For example, B ~ w e n ~ ~ excited the fluorescence by means of a mercury lamp operated from the 50-cycle mains supply, so that the exciting light and hence also the fluorescence was modulated at 100 cycles. The output from the photomultiplier was then detected by means of a 100-cycle amplifier. Instruments described by Parker and Barnesfs and Parker24y32 chop the exciting light a t 800 cycles and include an 800-cycle tuned amplifier.Theoretically, the use of chopped exciting light gives a lower minimum-detectable signal, as half of the light available is not used, but, to offset this, a.c. amplification has the convenience that a small amount of light leakage into the instrument is of no consequence ; further, short-lived fluorescence can be distinguished from long-lived phosphorescence without a phosphoroscope, which involves the use of two choppers. Thus, when chopping the exciting beam, the a.c. amplifier measures only the fluorescence, whereas, if the chopper is ?laced in the fluorescence beam, the sum of the fluorescence and phosphores cence is recorded.The sensitivity of the detector system can be increased by integrating the signal over a longer period of Itime, i.e., by increasing the time constant of the amplifier or recorder. (This decreases the noise level appearing at the recorder pen, but, to obtain the same resolu- tion, the speed of recording must be decreased.) For this reason it is undesirable to rely only on an oscilloscope recorder, which records the complete spectrum in perhaps 5 seconds, when, by using a pen recorder and a recording time of 5 minutes, greatly decreased noise level and increased sensitivity could be achieved. There is naturally a limit to the recording time, which is set by considerations of convenience, long-term stability of the electronics, photo- sensitivity of the sample, etc.DESIGN OF SAMPLE COMPARTMENT- The geometrical arrangement of the beams of exciting and fluorescence light in relation to the sample is the most controversial point in the design of a spectrofluorimeter. This is partly because of the different uses to which the instrument is put by different workers, partly because of the different sensitivities of the instruments, which determine the relative importance of stray fluorescence from the sample container, etc., and partly because of the different assessments of the relative advantages and disadvantages of the various systems. There are in general two types of sample to be considered; dilute solutions or gases, for which light absorption is small at all relevant frequencies, and concentrated solutions or opaque solids, for which light absorption is large at some or all frequencies of interest.The choice of arrangement for opaque solids and solutions is straightforward; some form of frontal illumination, such as that shown in Fig. 2 ( c ) , must be used. Examples of this type of sample. are solid phosphors, zones on a chromatographic column and materials adsorbed on filter- paper. The choice for weakly absorbing solutions is also clear; here, the sensitivity with a good spectrofluorimeter is nearly always limited not by the instrumental sensitivity itself, The types were reviewed some years ago by98 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol. 87 but by the over-all “blank” due to scattered light, fluorescence from the container walls, impurities in the solvent, etc.Since scattered light and fluorescence from the cuvette are less with right-angle illumination (where the photomultiplier does not “see” the illuminated cuvette walls) than with almost all forms of frontal illumination (where it does), the former is clearly the best choice. It has, however, been argued2 that the over-all instrumental sensitivity is theoretically much greater for frontal illumination if the fl uorescence is viewed in almost the same direction as that oE the exciting light, because the whole length of illumi- nated liquid is observed instead of a comparatively small part of it as with right-angle illumination-see Fig. 2 (a). This argument is only valid if the instrumental sensitivity is so low that the “stray-light” factor is negligible.The choice of method for moderately absorbing solutions will be decided by the nature of the investigation and the degree of inner-filter effects to be expected; the latter have been discussed fully in a previous section. Although it is occasionally required to measure con- centrated solutions in analytical work, it is usually possible to dilute the solution until the inner-filter effects are either negligible or sufficiently small to be corrected for when right- angle illumination is used. Under these conditions each fluorescent species present contributes its own spectra independently of the others. Frequency of fluorescence, P-’ Fig. 9. Fluorescence emission spectra of fused quartz and synthetic silica: curve A, optical-grade fused quartz (sensitivity x 1); curve B, as for curve A, but sensitivity x 20; curve C, synthetic silica (sensitivity x 540); curve D, as for curve C, but sensitivity ).: 60,000.Excitation with 4.03-p-1 (248 mp) light; spectral-correction curve as for Fig. 10 One other occasionally used arrangement is that shown in Fig. 2 (d), which has the merit of simple optics as compared with the frontal method. It has the disadvantages that, with strongly absorbing solutions self-absorption is a t its greatest, and with weakly absorbing solutions the photomultiplier has to withstand the full power of the beam of exciting light when the frequency settings of the two monochromators coincide. The ideal system for general use is one in which cell compartments can be readily inter- changed, so that right-angle or frontal illumination can be used or a Dewar vessel or other arrangement for low-temperature measurements can be substituted. There is therefore a good argument for having the excitation monochromator, lamp and monitor, etc., as one unit and the fluorescence monochromator with its photomultiplier as a second unit, so that the relative dispositions of the two units can be readily altered to suit the various requirements.CHOICE OF SAMPLE CONTAINER- Glass cuvettes are transparent down to about 3.1 p-1 (323 mp), depending on the wall thickness, and, if some loss of exciting intensity is acceptable, they can be used for determining fluorescence emission spectra with exciting light having a frequency as high as that of the 3.19-p-l (313 mp) mercury line. However, at frequencies less than 2.8 p-l, the absorptionFebruary, 19621 PARKER AND REES: FLUORESCENCE SPECTROMETRY.A REVIEW 99 of the cell wall will introduce distortion of the excitation spectrum, and below about 3.3 p-l practically all light is cut off. Special glasses transparent down to 4-0 p-l can be obtained, but for general-purpose spectrofluorimetry some form of quartz or silica cell is desirable. As it is frequently necessary to work with very dilute solutions, fluorescence of the cell walls can be troublesome even with right-angle illumination, where it is arranged that the photo- multiplier does not view the illuminated cell faces directly. Fused natural quartz, even of “optical” quality, nearly always contains traces of impurity (mainly aluminium) that render it fluorescent when excited with higher frequencies in the region of 4.0 p-l.Synthetic silica is of much greater purity and, after suitable heat treatment, is much less fluorescent. I t should be emphasised, however, that even this pure silica is fluorescent if it has been subjected to incorrect heat treatment, and, when purchasing silica cells, it is desirable to specify material of low fluorescence. The wide variation in the fluorescence of various types of silica can be seen from the spectra in Fig. 9. The fluorescence from the best sample of synthetic material (curve D) was some 100,000 times lower than that from a sample of optical-quality fused quartz (curve A). To minimise interference from any residual fluorescence in the cuvette, it is desirable when working at very high sensitivity to use right-angle illumination and to design the cell compartment so that the beam of exciting light after passing through the cell is dissipated as rapidly as possible and is not reflected back to illuminate the parts of the cuvette wall visible to the photomultiplier. SOME SPECIFIC INSTRUMENTS- There are in general two ways in which the spectra can be recorded, analogous in many respects to single- and double-beam recording in absorption spectrophotometry.The easiest way is simply to measure directly the amplified output from the photomultiplier or to feed this output directly to a recorder, which will then record the “apparent” emission or excitation spectrum as one or other of the monochromators is scanned. For many purposes this system is satisfactory, provided that short-term fluctuations of the light source, of the E.H.T.supply and of the amplifier gain are kept to 2 per cent. or less, depending on the precision required. Long-term fluctuations are much less important, as measurements are almost always made by comparison with a standard fluorescent solution. The most serious errors are usually those arising from fluctuations in the intensity of the discharge tube (mercury or xenon), particularly when operated from ax. mains; such fluctuation can be overcome by various kinds of double-beam operation. One of the easiest to set up32 involves a beam-splitter consisting of a clear silica plate situated in the beam of exciting light in front of the sample at an angle of 45”, so that it deflects a small proportion of the light on to a liquid fluorescent screen viewed by a monitoring photomultiplier.The outputs from the fluorescence photo- multiplier and the monitoring photomultiplier are fed to a ratio recorder, and compensation for fluctuations in the light intensity is thereby automatically made. This method also has the further advantage that, if the fluorescent screen is suitably chosen so that its fluorescence output is proportional to the quantum intensity of the exciting beam, irrespective of frequency, the instrument will record a “true” excitation spectrum.32 This is a great advantage because, owing to the rapid variation in intensity of a xenon arc with frequency, the uncorrected spectrum is a highly distorted version of the true one. Slavin, Mooney and Palumbo3’ have described a spectrofluorimeter incorporating a somewhat similar principle, but in which the balancing beam falls on a thermopile, so that the spectra are recorded in terms of energy rather than quantum units.Their instrument also incorporates servo-operated slits on the fluorescence monochromator, so that, as the emission spectrum is scanned, automatic correction is made for the spectral sensitivity of the photomultiplier - monochromator combination, and a “true” rather than an “apparent” emission spectrum is recorded. Although this instrument records spectra in energy rather than quantum units, the same principle could be used for the latter. Lipsett3* has constructed a versatile automatic-recording spectrofluorimeter involving two grating monochromators.The instrument includes a cam-adjusted amplifier that automatically corrects the experi- mental emission spectra as they are recorded. Fluctuations in the intensity of the exciting light are compensated for by the use of a monitoring photomultiplier. The instrument does not include facilities for the direct recording of true excitation spectra, but it could clearly be modified to do this. A useful feature is the kinematic trolley on which various stands may be placed so that specimens may be measured in a variety of conditions.[Vol. 87 Although the beam-splitter system described by Parker32 takes care of the most im- portant source of error-that arising from the lamp fluctuations-it does not compensate for fluctuations in the photomultiplier or amplifier ; on the contrary, it introduces additional sources of fluctuation, as a second photomultiplier is used.One British firm39 has under development a spectrofluorimeter in which a fraction of the exciting light, controlled by a calibrated attenuator, is made to alternate at the detector with the fluorescence light. Measurement is made by adjusting the attenuator until the beam outputs are equal, as indicated by a null-point meter. The fluorescence intensity is thus measured as a fraction of the exciting intensity, and the system is said to overcome most of the problems of stabi- lisation. The instrument has a xenon-arc source, two quartz-prism monochromators and frontal illumination of the sample. Two American commercial spectrofl~orimeters,~~ sgl both said to be very sensitive, have large grating monochromators, xenon arc, right-angle illumination, single-beam operation and facilities for either oscilloscope or pen recording.One of these instruments is also available in modified form as a spectropho~phorimeter.~~ In a German instrument,42 with two monochromators and single-beam operation, use is made of a somewhat unconventional arrangement in which the fluorescence is viewed through the polished bottom of the cuvette. The sensitivity claimed is that 1 pg per ml of quinine sulphate in 0.1 N sulphuric acid produces full-scale deflection for excitation at 355 mp and fluorescence at 460 mp when the band-widths of both monochromators are 10mp. It is not intended here to discuss purely filter fluorimeters nor fluorescence attachments to absorption spectrophotometers in which only the photometer unit of the latter is used, i.e., attachments that convert the instrument into a filter fluorimeter.There is, however, at least one absorption spectrophotometer available in this country with an attachment for measuring fluorescence emission spectra.& It incorporates a high-pressure mercury arc and filter for the exciting light, and the sample holder can be used for the frontal illumination of opaque materials or concentrated solutions or for right-angle illumination for measurements on dilute solutions. The sensitivity is claimed to be such that it will determine 1 p.p.m. of quinine sulphate in dilute sulphuric acid with a photometric accuracy of 0.2 per cent.when a band width of 2.4 mp at 460 mp is used. Various other instruments have been described in the literature apart from those referred t o earlier. As long ago as 1947 Burdett and Jonesu made an attachment for the Beckman spectrophotometer for measuring fluorescence spectra, and in 1948 StudeF described an instrument for the same purpose ; he used a photomultiplier and constant-deviation spectro- meter. Since then, many adaptations of commercial absorption spectrophotometers have been described for measuring either emission or excitation spectra. For example, modifica- tions have been described for the Beckman manual spectrophotometer,46 to 53 the Beckman recording ~pectrophotometer~~~~ and the Cary instrument .56 The use of a Unicam spectro- photometer for emission measurements has been described by Dammers-de-Klerk and Spruit ,57 and a Perkin - Elmer instrument has been modified for the same purpose.68 Bartholomew and his co-workers2 359 used two prism monochromators, with right-angle illumination of the sample and detection by photon counting.Olsonso described an ingenious microspectro- fluorimeter with which the fluorescence spectra of microscopic objects could be continuously scanned. Three new designs of recording spectrofluorimeter and their comparison with a commercial instrument have been reported by Goldzieher and Givner.61 The first was an adaptation of a Beckman recording spectrophotometer, the second incorporated Perkin - Elmer components, and the third was fitted with Bausch and Lomb grating monochromators.A Russian paper62 gave details of a spectrometric installation of high power for investigating luminescence ; a large grating monochromator was used. Some other instruments have also been described.63 to 68 100 PARKER AND REES : FLUORESCENCE SPECTROMETRY. A REVIEW CORRECTION OF SPECTRA- When a simple, single-beam spectrofluorimeter is used at constant slit width, the excita- tion and emission spectra obtained are a function not only of the absolute fluorescence characteristics of the substance being measured, but also of the characteristics of the particular instrument. Such “apparent” spectra are frequently grossly distorted versions of the true spectra, and, before they can be compared with results obtained with other instruments, they must be corrected to give the true spectra.For purely routine analytical work, the uncorrected spectra are often adequate, but, when results are to be published, it is obviouslyFebruary, 19621 PARKER AND REES : FLUORESCENCE SPECTROMETRY. A REVIEW 101 desirable to correct the spectra or to provide correction data so that the information is of maximum value to other workers. Methods of correcting spectra have been described in detail,5 and only a brief outline of the principles will be given here. The apparent excitation spectrum is a plot of I& against the frequency of exciting light -see equation (2)-and to obtain the true excitation spectrum the intensity of the exciting light as a function of frequency must be determined. Because the fluorescence efficiency of many substances is independent of the frequency of the exciting light, the true excitation spectrum is proportional to the molecular extinction coefficient of the compound and therefore reproduces its absorption spectrum.The measurement of true excitation spectra thus pro- vides a method of measuring absorption spectra of fluorescent compounds at concentrations far lower than would be needed to measure the absorption spectrum directly. It also makes it possible to detect the absorption spectrum of one fluorescent compound from a mixture of absorbing solutes. For this reason, the use of corrected excitation spectra, even in routine qualitative work, is an advantage, as they can be compared directly with known absorption spectra. VC’ith complicated excitation spectra, the point-by-point correction can be tedious, and it is a great advantage to have an instrument that directly records the true spectrum.32 The curve obtained when, with constant exciting-light intensity, the fluorescence mono- chromator is scanned at constant slit width is the apparent emission spectrum.The true emission spectrum is a plot of fluorescence intensity (measured in quanta per unit frequency interval) against frequency. To deterrnine this, the apparent curve has to be corrected for the changes with frequency of (a) the photomultiplier sensitivity, (b) the band width of the monochromator and (c) the changing transmission of the monochromator. Full details of the methods of correction when a calibrated tungsten lamp is used have been de~cribed.~ At frequencies greater than about 2.5 p-l a tungsten lamp is not particularly suitable. For approximate calibration in the ultraviolet region, the fluorescent screen monitor32 can be used to measure the relative quantum output from the excitation monochromator when the latter is set up to illuminate a magnesium oxide screen viewed by the fluorescence mono- chromator.The intensities recorded by the fluorescence photomultiplier for various fre- quency settings of the excitation monochromator can then be used to derive the sensitivity curve of the fluorescence monochromator - photomultiplier combination. It can be arranged to vary the setting of the monochromator slits or the amplifier gain in such a way that the instrument automatically applies its own correction, and a trueemission spectrum is recorded directly.Slavin, Mooney and P a l ~ m b o ~ ~ have described such an instrument, although they chose as units energy per unit wavelength interval, which is theoretically less significant than quanta per unit frequency interval. Lipsett38 has also devised a self-correcting instrument. Choice of units has been discussed previ~usly.~ For the fluorescence emission spectrum, there are at least three good reasons for using quanta per unit frequency interval rather than energy per unit wavelength interval (although correction data in the latter units have been reported by White, Ho and Weimer for one commercial i n s t r ~ m e n t ~ ~ ) . First, with the latter system, the fluorescence spectrum obtained is a distorted version not bearing the correct relationship to the corresponding absorption band.Secondly, if wavelength rather than frequency is used for the horizontal scale, the bands are unnaturally bunched together in the short-wavelength region, an undue amount of space being given to the long-wavelength region, for which less information is available. Thirdly, if energy units rather than quanta are used, the integrated area under the curve is not proportional to the fluorescence efficiency, since the quanta of lower frequency carry proportionately less energy. For the horizontal (frequency) scale the reciprocal micron has been recommended as a convenient unit (1.0 to 5 . 0 ~ - l corresponds to 1000 to 200mp), although most workers in the past have used the reciprocal centimetre.The vertical scale is plotted in relative units, but results can be made quantitative by quoting the fluorescence efficiency or by recording the emission spectrum of a standard substance of known efficiency a t a stated concentration or optical density per centimetre on the same scale.5 The absolute fluorescence efficiency of the new substance can be derived by comparing the areas under the two curves and the optical densities of the two solutions at the frequency of the exciting light ~ s e d . ~ , l ~ STANDARD FLUORESCENT SUBSTANCES- The determination of the sensitivity curve, i.e., the correction curve, of the spectro- meter - photomultiplier combination is subject to many errors, some of which have been102 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol.87 discussed by Lippert and co-worker~.~~ The correction curve could be more conveniently determined and readily re-checked if there were available a series of standard fluorescent substances whose corrected fluorescence emission spectra had been determined by several independent observers, so that the true spectra could be internationally agreed upon for standardisation purposes. Ideally, a series of standard substances should be chosen, having broad and structureless fluorescence bands overlapping one another, so that the complete series covered the whole of the visible and quartz ultraviolet regions. Other requirements of an ideal standard substance are that it should be solid and readily and cheaply prepared in a pure state of precisely known composition.It should be stable to air and light and soluble in aqueous solution. It should have a high fluorescence efficiency, with little overlap of absorption and fluorescence spectra. If, in addition to their spectra, the absolute fluorescence efficiencies of the standard substances were also determined and agreed upon, their value would be greatly increased. It is likely to be a long time before the ideal system of standards is achieved, particularly for the ultraviolet region, but some progress has already been made. For example, several groups of workers have published and compared emission spectra of solutions of quinine bisulphate and anthra~ene,~~ s70 p 7 1 972 and Lippert and his co-workers58 have determined the emission spectra of five substances covering the spectrum from about 1.1 to 3.0 p-1 and have proposed their use as standards for calibrating fluorescence spectrometers.Absolute fluorescence yields have from time to time been directly determined for a variety of substances during the past 30 to 40 years3 Some recent publications of particular value are those by F O r ~ t e r , ~ ~ who has reviewed the methods of measurement, by Weber and Teale,74 by M e l h ~ i s h , ~ ~ , ~ ~ by Gilmore, Gibson and M c C l ~ r e ~ ~ and by Forster and Livingston.78 For purely routine determinations within one laboratory, the fluorescent solution used for comparison need not be one whose spectral emission curve is accurately known. All that is required is a stable solution with which to set the fluorimeter, and some workers have even used fluorescent glasses as standards.SENSITIVITY- The absolute fluorescence sensitivity of a substance, Le., the intensity of fluorescence (of all frequencies) emitted for a given frequency and quantum intensity of exciting light is proportional to y k . Since measurements are usually made by selecting a narrow band of frequencies at the fluorescence maximum, the observed intensity will also depend on the effective half-band width of the fluorescence spectrum (H), and the observed sensitivity will be proportional to +E/H. As a more convenient unit than E , the optical density per centimetre for a concentration of 1 pg per ml (D) has been recommended.6 This value varies with the frequency of the exciting light and for most substances is maximal at the peak of the principal absorption maximum, i.e., +Dmax./H.Values at other frequencies can be calculated from this maximum value by reference to the absorption spectrum or, more strictly, the excitation spectrum. If data on the relative fluorescence sensitivity of substances were expressed in this form, they would be independent of the instrument used to obtain them and hence of maximum value to other workers. The analytical chemist using spectrofluorimetry is also interested in another kind of sensitivity, viz., the sensitivity of his own instrument as compared with those of other workers, This may be defined as the minimum detectable signal-to-noise ratio and can be expressed in terms of a minimum detectable concentration of some standard substance with a chosen frequency for excitation under carefully specified conditions.16 I t depends on the charac- teristics of the photomultiplier - amplifier combination, the intensity of the light source, the light-gathering power of the monochromators and the frequencies and band widths at which they are used.However, the factor limiting sensitivity is frequently not the electrical signal-to-noise ratio, but the over-all fluorescence blank, which may include contributions from the fluorescence of the cuvette or impurities in the reagents, scattered exciting light or even the Raman spectrum of the solvent. Measurement of the latter is a useful test of the performance of a spectrofluorimeter.24 APPLICATIONS TO ORGANIC ANALYSIS Most aliphatic compounds that absorb strongly in the ultraviolet region contain groups such as nitro, carboxyl or iodo, which quench fluorescence.Aliphatic ketones have been claimed to fluoresce, although there is some doubt about the interpretation of the spectraFebruary, 19621 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW 103 observed ; biacetyl produces a green luminescence of comparatively long lifetime.3 Some highly conjugated aliphatic compounds fluoresce, e.g., vitamin A and carotene.3 However, fluorimetric methods for determining the latter have not been adopted so widely as have those for determining the heterocyclic vitamins thiamine and ribofla~in.l~t7~ The most fruitful field in which spectrofluorimetric analysis has been applied is that of aromatic and heterocyclic compounds, both simple and complex. Benzene itself has a comparatively low fluorescence efficiency in solution at room temperat~re,~ and this, coupled with the fact that it requires rather high frequencies for excitation and shows rather weak absorption in the accessible spectral region, probably accounts for the fact that fluorimetric methods for its determination have not been popular; its fluorescence spectrum is compared with those of some other simple aromatic compounds in Fig.10. Frequency of fluorescence, p-' Fluorescence emission spectra of simple aromatic compounds in aerated ethanol: curve A, 0.32 pg of anthracene per ml (sensitivity x 1); curve B, 12 pg of naphthalene per ml (sensitivity x 15); curve C, 100 pg of phenol per ml (sensitivity x 5); curve D, 170 pg of benzene per ml (sen- sitivity x 50) ; curve E, sensitivity of quartz monochromator with E.M.I.6256 photomultiplier. (Spectra are uncorrected; correction curve as indicated.) All solutions were measured with the same intensity of exciting light (4.03 p-1)- and constant slit width on the monochromator corresponding to half-band widths of 0.0 1 p-l a t 3.6 p-I and 0.02 p-I a t 2.5 p-I. The optical density per cm a t 4.03 p-1 was 0.2 for all solutions Fig. 10. Condensed-ring aromatic hydrocarbons have long been known for their complex fluores- cence emission spectra, which, until the advent of the high-sensitivity spectrofluorimeter, were measured spectrographically. As recently as 1961, NeScoviC and SoSkiCso have reported the detection of 10-9 g of 3,4-benzopyrene directly from chromatographic paper by spectro- graphy, and Davies and Wilmshurstsl have detected g per ml of the same compound by using this technique.Bentley and Burgans2 have also reported a spectrographic method ior determining 3,4-benzopyrene in tobacco and tobacco-smoke condensate after chromatographic separation. Several workers have reported methods for determining various condensed-ring aromatic hydrocarbons by means of photo-electric measurement of emission spectra.@ @ For their determination as traces in complex mixtures, the use of a concentration procedure and then, perhaps, chromatography is necessary, but it is in just such circumstances that the advantages of spectrofluorimetry are most apparent. Its high sensitivity permits identification of sub-microgram amounts of materials separated, for example, on a paper chromatogram.It has the further advantage over absorption spectrophotometry that it104 PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol. 87 produces two spectra instead of one, and the excitation spectrum of an impure zone is less distorted by background than is the absorption spectrum. The method has also been used for the direct analysis of mixtures of aromatic hydrocarbons without preliminary separation. Thommes and Leiningers3 determined anthracene, phenanthrene and fluorene in a mixture by measuring the emission spectrum with excitation at two frequencies ; with 365-mp light, only anthracene fluoresces and can be directly determined. With excitation at 265 mp, all three compounds fluoresce, but, by measuring the fluorescence intensity at 316 and 350 mp, their individual concentrations can be determined. For compounds such as anthracerie and phenant hrene, when an excitation frequency can be chosen to which one component is completely transparent, very small concentrations of the second component can be determined in the presence of high concentrations of the first.For example, unpublished work by C. A. Parker and C. G. Hatchard has shown that, by excitation of a 1 per cent. solution of phenanthrene in ethanol with the 2.73-p-1 mercury line, less than 2 p.p.m. of anthracene in phenanthrene can be detected. At the other extreme, there are complex systems of fluorescent substances for which interpretation of the emission spectra obtained with a variety of excitation frequencies becomes difficult.we be^-86 has recently shown how the constancy of the spectral distribution of the fluorescence from a given substance excited with light of different frequencies permits the development of a simple quantitative criterion for determining the number of independent fluorescent components in a mixture. Williams87 has compiled some useful results on the fluorescence characteristics of simple aromatic compounds. He concludes that monosubstituted benzenes, in general, may fluoresce if they contain an o-$-directing group, but not if the substituent is m-directing. Thus, at a suitable pH, phenol and aniline and their alkylated derivatives fluoresce, whereas nitro-compounds, carboxylic acids and the cations of amines do not.On the other hand, a group such as CH,CO-NH-, although o-$-directing, gives non-fluorescent compounds. He found that, although benzoic acid is non-fluorescent, its hydroxy derivatives are. Williams also records results for polycyclic and heterocyclic phenols and coumarin derivatives. The greatest number of organic applications has so far been in the biochemical and related fields, mainly owing to the pioneering work of Bowman, Udenfriend and their col- leagues. In 1955 they described an experimental spectrofluorimeter,4 a practical adaptation of which was developed commercially. They have subsequently listed fluorescence data for more than fifty compounds of biological interest and have discussed the application of pH variation as an additional criterion for identification.8s They concluded that the structural requirements for fluorescence were the presence of either (a) an aromatic nucleus substituted by at least one electron-donating group or ( b ) a conjugated unsaturated system capable of a high degree of resonance.Members of the same team have also investigated more than fifty compounds of pharmacological interestsg and have studied the partition characteristics of a representative group of compounds to develop double-extraction procedures for separating drugs from tissue homogenates and body fluids. Other papers have dealt with specific methods, for example, the determination of 5-hydroxytryptamine in biological tissues,gO of tryptamine in tissues after solvent extraction, condensation with formaldehyde and de- hydr~genation,~l of tyrosine in plasma and tissues after reaction with l-nitroso-2-naphtho192 and of other substances.It is alreadv clear that spectrofluorimetry is fast becoming a standard analytical method in this field. An example of a less direct application of spectrofluorimetry to biochemical problems is given by the work of Leaback and Walker,93 who used the determination of methylumbelliferone in a method for assaying the enzyme N-acetyl- ,%glucosaminidase. Spectrofluorimetry has proved to be of considerable value in the investigation of protein structure. The residues of the three aromatic amino acids phenylalanine, tyrosine and tryptophan are the only components of the polypeptide fabric of non-conjugated proteins that are capable of fluorescen~e.~~ The fluorescence characteristics of the three acids them- selves in aqueous solution have been described by Teale and Weber.26 Each has a characteris- tic fluorescence band in the ultraviolet region, the quantum efficiencies being, respectively, 0.05, 0.2 and 0.2.TealeP4 systematically investigated the fluorescence of many proteins and found that tyrosine fluorescence was observed only in those proteins not containing tryptophan, and then only in small yield. It was also observed that tyrosine fluorescence was quenched by substances containing the charged carboxyl group -COO-, and comparison of the tyrosine fluorescence yields in proteins with the relative abundance of carboxylate groups suggestedFebruary, 19621 PARKER AND REES : FLUORESCENCE SPECTROMETRY. A REVIEW 105 that it was indeed this group that was responsible for the quenching in the proteins themselves.It was also concluded that the tryptophan yield in a native protein is a function of the specific constellation of groups around each aromatic residue and characterises each protein. As a result, the effects of changes in hydrogen bonding and mobility of groups and of chemical modification of groups on the protein surface can be studied fluorimetrically. Fluorescence polarisation spectra have not been dealt with in detail in this review, as they are not, at this stage of development, likely to be used by the average analytical chemist. Mention should, however, be made of Weber’s recent work on tyrosine, tryptophan and related compounds95 and on proteins.96 The absolute value of the polarisation of fluorescence in solution can decrease as a result of molecular rotations or by intermolecular transfer of the excited state.In the absence of these effects, i.e., in dilute viscous solutions, the polarisation reaches a maximum absolute value, the principal polari~ation.~5 A plot of the principal polarisation against wavelength or frequency of exciting light-the principal polarisation spectrum-reflects to a first approximation the relative orientations of the transition moments associated with the absorption and emission of light. A study of the spectrum thus makes it possible to distinguish between different overlapping electronic transitions. In concen- trated solutions energy transfer between molecules can be detected, and in dilute solutions of complex molecules energy transfer between different parts of the same molecule can be observed .96 The application of spectrofluorimetry in the field of more conventional organic analysis is still comparatively limited, although pamphlets from two instrument manufacturers list large numbers of organic compounds detectable at low concentration by this mean^.^^,^^ Hercules and Rogersg7 have determined 1- and 2-naphthols in mixtures by measurement of fluorescence at two wavelengths and have studied the absorption and fluorescence sFectra of 1- and 2-naphthols and seven naphthalene di01s.~~ Kokoski, Kokoski and Slamag9 des- cribed the fluorescence of many vegetable drugs after treatment with various reagents, and Hornsteinloo has determined the ultraviolet fluorescence spectra of organic pesticides.Among other organic compounds investigated fluorinietrically are glycerol (determined after conver- sion to quinolinelol), naphthionic acid,lo2 coumarins,lo3 0- and m-hydroxybenzoic acids,l04 many 2,5-diarylo~azoles~~~ and some salicyloyl hydrazones.lo6 Apart from its application to the determination of specific substances, spectrofluorimetry can be used in an empirical way for detecting and determining complex mixtures of natural or artificial origin, provided that their composition remains reasonably constant from sample to sample. For example, Parker and Barneslo7 have determined the fluorescence emission and excitation spectra of a variety of lubricating oils, all of which were found to have similar characteristics.They made this the basis of a method for determinining oil mist in air at concentrations down to less than 1 pg per litre. Mihul, Ruscior and Pop108 have investigated the spectra of illuminating oils and made recommendations for identifying products of diverse origin. The fluorescence spectra of petroleum oils and their fractions in liquid form and on chromatographic columns have been investigated in detail by Kats and Sidorov.109 An unusual application of spectrofluorimetry has been described by Armstrong and Grant,llO who investigated the products of radiolysis of aqueous solutions of calcium benzoate and devised a highly sensitive chemical dosimeter for ionising radiation. By exciting the irradiated solution with light of wavelength 290 mp and measuring the fluorescence at 400 mp, the salicylic acid produced by the radiolysis can be determined with negligible interference from the other two isomers.Biphenyl is also a radiation product, but does not interefere. The fluorescence intensity is a linear function of the concentration of salicylic acid, and the latter is itself proportional to the dose of X- or gamma-radiation over the range 5 to 100 rads. The application of a similar principle to the determination of radiation in the visible and quartz ultraviolet regions would be attractive. The most sensitive chemical actinometer at present available is based on the photolysis of potassium ferrioxalate and determination of the ferrous iron so formed by absorption ~pectrophotometry.~~ y 3 0 If a substance could be found that photolysed reliably to give a product determinable by spectrofluorimetry, an actinometer of even greater sensitivity would be possible.From the above examples of applications to organic analysis, it is clear that there is a wide scope for the technique in this field, and the number of papers dealing with organic analysis will undoubtedly increase greatly during the next few years. Apart from papers dealing specifically with fluorescence spectra, there are many dealing with analysis by filter fluorimetry, which could often be better done by spectrofluorimetry. The reviews by106 PARKER AND REES FLUORESCENCE SPECTROMETRY. A REVIEW [Vol. 87 White1 give many additional references to analytical methods, including the determination of steroids, riboflavin, tocopherols, beiizaldehyde and salicylaldehyde derivatives, deoxy- ribonucleic acid, 5-hydroxyindoles, alkaloids , reserpine and derivatives, oestrogens and flavines and flavones. In fact, almost any substance absorbing in the quartz ultraviolet region is worth testing for fluorescence when analytical methods for its determination are being considered.FLUORESCENCE OF INORGANIC SUBSTANCES Although few inorganic compounds are appreciably fluorescent in solution, a wide variety of solid polycrystalline phosphors can be prepared in which the spectral distribution of the luminescence and its lifetime are determined primarily by the presence of small amounts of impurities or “activators.” Well known examples of polycrystalline base materials used in phosphors are zinc, cadmium, calcium and strontium sulphides, potassium chloride, zinc selenide , calcium and magnesium tungstates, beryllium, zinc and cadmium silicates and many others.Impurity activators include copper, silver, manganese, antimony, thallium, lead, rare earths, bismuth and uranium. Accounts of the preparation and properties of inorganic polycrystalline phosphors are given in t e x t - b ~ o k s , ~ * ~ ~ ~ and it is not intended to discuss them in detail because, except for uranium, few analytical applications of the luminescence have been made, possibly because the luminescence is usually critically dependent on the precise conditions under which the phosphors are prepared. Two applications depending on the quenching of the luminescence will be mentioned.Starik, Starik and KostYrevll2 have deter- mined chromium by measuring the quenching of the luminescence of a uranium-activated sodium fluoride bead after it had been dipped into a chromium solution and fused. Bradley and Sutton113 have detected traces of copper, silver, mercury and platinum by their effects on the luminescence of a silver-activated cadmium sulphide phosphor. The determination of uranium by observing the fluorescence produced when traces of the element are fused with sodium fluoride was first developed by Hernegger and Karlikll4 and has been discussed by Rodden115 in a paper on the determination of naturally occurring radioactive elements; the minimum amount detectable is said to be g. Other references to the method are given by White.l A more recent paper by Vozzella, Powell, Gale and Kelly116 describes its determination in zirconium and hafnium after extraction, as uranyl nitrate, into ethyl acetate.It should be mentioned that many inorganic glasses and single crystals, as distinct from polycrystalline phosphors, emit luminescence when suitably excited, and again this lumines- cence is dependent on the impurities pre~ent.~ Thus, chromium has been determined in ruby by means of its luminescence in the infrared region.l17 Unpublished work by C. A. Parker has shown that the luminescence of fused quartz and synthetic silica can give a sensitive indication of the presence of impurities (see Fig. 9), although the interpretation of the spectra is complicated by the fact that the luminescence depends also on the method of heat-treatment of the sample.The observation of luminescence from inorganic compounds in solution has been restricted to the ions of only a few elements, notably uranium and the rare earth^.^ The photo- luminescence of uranium does not appear in all compounds, but only those containing the uranyl ion, UO,+. However, the nature of the fluorescence in solution depends on the anion present and the solvent used, and it is concluded that the spectra originate from a variety of complexes. Sill and Petersonll* have developed a sensitive visual method for detecting uranium in phosphoric acid solution with use of excitation by the 254-mp mercury line. They investigated the effects of quenching agents and found that, in sulphuric acid solution, the only other inorganic ions giving fluorescence were those of stannous and j3-stannic tin and some rare earths.They avoided interference from these elements by appropriate chemical pre-treatment, but, no doubt by the use of a spectrofluorimeter, the spectrum of the uranyl compound could be distinguished from those of the other ions, and the method could probably be adapted for determining the latter also. Pant and KhandelwaPlg have found the fluores- cence emission spectrum of uranyl acetate solutions to show structure at room temperature and have also made measurements at the temperature of liquid air. The same workers have also investigated the spectrum of uranyl perchlorate at room temperature.120 Other recent investigations of uranium solutions have been made by Bistl21 and NovZk.l22 The fluorescence emission spectra of solutions of inorganic salts of the rare earths have been discussed by Pring~heim.~ Because the transitions involve electrons in the innerFebruary, 19621 PARKER AND REES FLUORESCENCE SPECTROMETRY.A REVIEW 107 4f shell, which are well shielded from external influences, the spectra are frequently very sharp, particularly in crystalline solids (e.g., with europium, samarium, gadolinium, terbium and dysprosium). Generally, in solution, the lines forming a group are less sharply separated, although the lines of gadolinium sulphate are relatively narrow. When the salts are not completely dissociated, e.g., at high concentration in aqueous solution or in solutions of com- plex organic compounds in organic solvents, the spectra show considerable variation in their structure.Pringsheim3 discusses the fluorescence in solution of samarium, europium, terbium, gadolinium, erbium, praseodymium, neodymium, cerium and lanthanum, some of which require short-wavelength ultraviolet light for the efficient excitation of fluorescence. Fassel and Heide1123 have devised a spectrofluorimetric method for determining terbium in hydro- chloric acid solution, a hydrogen arc being used for excitation; their calibration graph extended down to about 4 pg of terbium per ml. Varsanyi and Dieke124 have described an interesting experimental arrangement for investigating the excitation of the various line emissions of a rare earth (in a crystal) as a function of the excitation frequency.As the excitation mono- chromator was scanned, a second motor drive moved the spectrograph-plate holder vertically, so that the resulting exposure showed the fluorescence spectrum as a function of the exciting frequency. (The rare earths are an exception to the rule that the fluorescence emission spectrum is independent of the exciting frequency.) Few direct fluorimetric methods for other elements have been described. Sill and Peterson125 have used the fluorescence from thallous ion in sodium chloride solution for its detection. They used irradiation with short-wavelength ultraviolet light and claimed a sensitivity of one part in 50 million. These workerslls have also observed fluorescence from solutions containing /3-stannic and stannous tin, and Gudymenko, Belyi and Skachko126 have measured the fluorescence produced by the latter in sulphuric acid solution.APPLICATIONS INVOLVING INORGANIC - ORGANIC COMPOUNDS During the past 20 years, a large number of methods has been described for deter- mining inorganic substances by reaction with an organic compound to give a fluorescent product. Most of the investigations dealt with systems giving visible fluorescence, in which observations have been made visually or with a filter fluorimeter. In recent years, however, spectrofluorimetric investigations of such systems have been described. Collat and Rogers127 have devised a composite method for determining aluminium and gallium by extracting their 8-hydroxyquinolinates into chloroform and measuring the fluorescence intensities produced by excitation with the 366- and 436-mp mercury lines.The emission spectra of the two quinolinates are similar, but their absorption spectra differ, and hence their respective excitation efficiencies are different at the two wavelengths. Ohnesorge and Rogers6’ have investigated the fluorescence emission spectra of some metal chelate compounds of 8-hydroxyquinoline and have observed the effects of solvent, acid, alkali and ultraviolet radiation. White, Hoffman and Magee28 have compared the absorption, excitation and emission spectra of some metal chelates used in determining aluminium, beryllium, thorium, zirconium, boron and lithium. The same team has investigated the fluorescent metal chelates of a variety of o,o’-dihydroxyazo-compounds128 and has also devised methods for determining tin with flavonoll29 and magnesium with bis (salicylidene) ethylenediamine.l30 Bartholomew2 has discussed the analytical significance of intramolecular energy transfer within organic complexes of rare earths.Thus, Weissman131 showed that solutions of inorganic salts of europium are weakly fluorescent, but solutions of europium benzoylacetonate and europium salicylaldehyde emit strongly the characteristic europium fluorescence when excited by frequencies absorbed by the organic part of the molecule. Similar effects were observed by Kuznetsova and Se~chenkol~~ with various organic complexes of europium, samarium and thorium. Energy transfer in the reverse direction appears to occur in gadolinium sali- cylate, in which the blue-violet fluorescence of free salicylic acid is greatly enhanced owing to the strong absorption by the gadolinium at the excitation frequency.133 With the advent of modern high-sensitivity filter fluorimeters and spectrofluorimeters, the over-all sensitivity of a method is frequently limited not by the instrumental sensitivity, but by the magnitude of the “blank” reading, as discussed by Parker and Barnes16 and Parker.24,134 The “blank” may include fluorescence arising from contamination of the reagents by traces of the inorganic substance being determined, contributions from various instrumental defects already discussed and Raman emission from the solvent.Further, when organic reagents are used to determine traces of inorganic substances, a major contribution to the108 PARKER AND REES : FLUORESCENCE SPECTROMETRY. A REVIEW [Vol.87 “blank” is frequently made by traces of organic impurity in the reagent or by photochemical decomposition products of the reagent. When designing an ultra-sensitive fluorimetric method, it is of the utmost importance to determine the relative magnitudes of the various contributions to the “blank” so that steps can be taken to eliminate them, and this can only be done by investigating the emission spectra of the “blank” and “test” solutions with a spectrofluorimeter. Once the “blank” has been satisfactorily decreased, the application of the method may well be relegated to a carefully designed filter fluorimeter. An example of the use of a spectrofluorimeter in the development stages of a method was given by Parker and Barnes,l6tU who described the fluorimetric determination of boron down to 0-002 pg.Apart from boron in the reagents, the main contribution to the “blank” value came from photochemical decomposition of the reagent and Raman emission from the solvent. These effects were investigated spectrofluorimetrically, and, by suitable choice of conditions, the “blank” value was decreased; the method was ultimately applied with a filter fluorimeter. Similar investigations were made by Parker and Harvey31 in developing a fluorimetric method for determining traces of selenium. I t is not proposed here to give a complete bibliography of fluorimetric reactions that have been or could be used for analytical purposes. Many references are given in the reviews by White,l and an idea of the scope of the method can be gained from a brief r&um& of some of the reagents that have been used and the elements that have been determined.Probably more fluorimetric methods have been described for aluminium than for any other element. The reagents used include 8-hydroxyquinoline, 8-hydroxyquinaldine, 3-hydroxy-2-naphthoic acid, Pontachrome blue black R, salicylaldehyde condensation products and various hydroxy- azo dyes. Various reagents have also been recommended for gallium, including the first two listed above for aluminium, dihalo-8-hydroxyqi~inolines, morin and various dyes. Beryllium can also be determined with 8-hydroxyquinaldine, morin and hydroxyanthraquinone derivatives. Other elements have received somewhat less attention, but White1 gives refer- ences to methods for zirconium, thorium, germanium, boron, silicon, lithium, copper, zinc, tungsten, tin, thallium, tellurium, fluorine, vanadium, ruthenium and others.Probably the best method of determining traces of boron is by reaction with b e n ~ o i n ~ , ~ ~ ~ ; the same reagent can also be used for determining zinc.136 Some workers in Australia have published a series of papers137 to 141 on the determination of stannous tin with various aromatic nitro- compounds; the latter are reduced to give fluorescent products. Veening and Brandt142 have recently developed a method for determining ruthenium in the presence of platinum metals, and Kosta and D ~ b a r l ~ ~ report a method for determining niobium as 8-hydroxyquinolinate. Hanker and his co-workers have devised rather unusual methods for determining cyanide and sulphide.The first depends on the release of 8-hydroxyquinoline-5-sulphonate from its palladium complex by either CN- or S2- and its determination by formation of the fluorescent magnesium complex144; a sensitivity of 0.02 pg of cyanide per ml is claimed. In the second method, hydrocyanic acid was separated and was converted to cyanogen chloride by chlor- amine-T. The cyanogen chloride was then made to cleave the pyridine ring of nicotinamide to give a product strongly fluorescent in alkaline s01iition.l~~ A method for determining sulphate, fluoride and phosphate ions has been reported by Nazarenko and S h u ~ t o v a l ~ ~ ; this involves release of a fluorescent dye from a thorium chelate.Various fluorimetric methods have been proposed for determining molecular oxygen. For example, Konstantinova-Shlezinger and Krasnova147 found that alkaline solutions of adrenalin develop a bright yellow-green fluorescence in the presence of small traces of oxygen. The same workers have also reported a method for ozone148; this depends on the oxidation of dihydroacridine to acridine. Molecular oxygen can also be determined by its quenching action on fluorescence or phosphorescence. Chleck, Brinkerhoff, Hadley and Ziegler149 used a liquid scintillator to measure the oxygen content of a gas in the range 0.1 to 100 per cent. Kautsky and Muller report that the effect of oxygen on the luminescence of trypaflavin adsorbed on silica gel permits the determination of mole of oxygen.150 CONCLUSION From the analytical point of view, the present state of development of spectrofluorimetry may be compared with that of absorption spectrophotometry 20 to 25 years ago, and the subsequent advances in the latter technique will give some idea of the prospects for spectro- fluorimetry. Commercial instruments are now becoming wide-spread, and we may expectFebruary, 19621 PARKER AND REES : FLUORESCENCE SPECTROMETRY.A REVIEW 109 a rapid increase in the publication of spectral data. Probably one of the most urgent require- ments, therefore, is the production of instruments that will record directly the true emission and excitation spectra, so that analytical chemists can readily record and publish data in a form that will be of the most value to themselves and other workers.The utility of spectro- fluorimeters would also be greatly increased by the development of convenient sources giving a higher output in the spectral region below 300mp and preferably extending to 180mp. Excitation need not be restricted to the visible and ultraviolet regions. The analytical use of beta-ray e~citationll.~ 9149 has already been reported, and the analytical possibilities of excitation by cathode rays and X-rays and possibly even gamma-rays will probably also be investigated. The largest proportion of work so far described deals with measurements in solution at room temperature. As convenient apparatus is developed, more attention will undoubtedly be given to measurements at low temperature, where a wider range of substances fluoresces, the spectra are more characteristic and phosphorescence is common, so that an additional pair of spectra is provided for identification. The analytical chemist to-day has not only to report on the main chemical composition of materials, but is increasingly required to provide information about more subtle differences between the behaviours of samples, differences that may depend, for example, on the presence of exceedingly minute amounts of a trace impurity, traces that can sometimes best be deter- mined by fluorescence measurements.Even more subtle differences may be shown up by measurements of fluorescence polarisation, which, apart from the possibility of providing an additional criterion for identification, can, as Weber has shown, give an insight into the interactions between different parts of a complex molecule.Such information can now reasonably be expected to come within the general field of interest of the analytical chemist. 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. 30. 31. 32. 33. 34. 35. 36. REFERENCES White, C. E., Anal. Chem., 1949, 21, 104; 1950, 22, 69; 1952,24, 85; 1954, 26, 129; 1956,28, 621; Bartholomew, R. J., Rev. Pure and Appl. Chem., 1958, 8, 265. Pringsheim, P., “Fluorescence and Phosphorescence,” Interscience Publishers Ltd., London, 1949. Bowman, R. L., Caulfield, P. A., and Udenfriend, S., Science, 1955, 122, 32, Parker, C. A., and Rees, W. T., Analyst, 1960, 85, 587. Freed, S., and Salmre, W., Science, 1958, 128, 1341. Chen, M.-W., “Bibliography of Phosphorescent Molecules,” The American Instrument Co.Inc., Parker, C. A., and Hatchard, C. G., Analyst, in the press. Boudin, S., J . Chim. Phys., 1930, 27, 285. Kautsky, H., Hirsch, A., and Flesch, W., Ber., 1935, 68, 152. Parker, C. A., and Hatchard, C. G., Trans. Faraday SOC., 1961, 57, 1894. Forster, Th., Ann. Phys., 1948, 6, 55. Bowen, E. J., Trans. Faraday SOC., 1954, 50, 97. Bowen, E. J., and Wokes, F., “Fluorescence of Solutions,” Longmans, Green & Co. Ltd., London, Parker, C. A., and Hatchard, C. G., Nature, 1961, 190, 165. Parker, C. A., and Barnes, W. J., Analyst, 1957, 82, 606. Bowen, E. J., Photoelectric Spectrometry Group Bulletin, 1961, No. 13, 331. Forster, Th., 2.Elektrochem., 1950, 54, 531. Weller, A., 2. Elektrochem., 1952,56, 662; Z. Phys. Chem., 1955,3/4, 238. -, 2. Elektrochem., 1954, 58, 849. Forster, Th., “Photochemistry in the Liquid and Solid States,” John Wiley & Sons Inc., New York, 1960, p. 10. Weller, A., 2. Elektrochem., 1956, 60, 1144. Forster, Th., and Kasper, K., Ibid., 1955, 59, 976. Parker, C. A., Analyst, 1959, 84, 446. Weber, G., and Teale, F. W. J., Trans. Faraday SOC., 1958, 54, 640. Teale, F. W. J., and Weber, G., Biochem. J . , 1957, 65, 476. Teale, F. W. J., Ibid., 1960, 76, 381. White, C. E., Hoffman, D. E., and Magee, J . S., Spectrochim. Ada, 1957, 9, 105. Parker, C. A., Proc. Roy. SOC. A , 1953, 220, 104. Hatchard, C. G., and Parker, C. A., Ibid., 1956, 235, 518. Parker, C. A., and Harvey, L.G., Analyst, 1961, 86, 54. Parker, C. A., Nature, 1958, 182, 1002. Parker, C. A., and Rees, W. T., Nature, 1959, 184, 1223; J . Chim. Phys., 1959, 761. Parker, C. A., and Barnes, W. J., Analyst, 1960, 85, 828. Turk, W. E., Photoelectric Spectrometry Group Bulletin, 1952, No. 5, 100. Bowen, E. J., Ibid., 1953, No. 6, 124. 1958, 30, 729; 1960, 32, 47R. Silver Spring, Maryland, U.S.A. 1953.110 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. 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. PARKER AND REES: FLUORESCENCE SPECTROMETRY. A REVIEW [Vol. 87 Slavin, W., Mooney, R. W., and Palumbo, D.T., J . Opt. Soc. Amer,, 1961, 51, 93. Lipsett, F. R., Ibid., 1959, 49, 673. Pamphlets of Unicam Instruments Ltd., Cambridge. Pamphlets of the American Instrument Co. Inc., Silver Spring, Maryland, U.S.A. Pamphlets of the Farrand Optical Instrument Co. Inc., New York. Pamphlets of the Carl Zeiss Optical Company, Jena, Germany. Pamphlets of Optica U.K. Ltd., Gateshead on Tyne. Burdett, R. A., and Jones, L. C., J . Opt. SOC. Amer., 1947, 37, 554. Studer, F. J., Ibid., 1948, 38, 467. Priestley, W., jun., paper presented a t the Conference on Analytical Spectroscopy, Pittsburgh, Derr, A. J., J . Opt. Soc. Amer., 1951, 41, 872. Lauer, J. L., and Rosenbaum, E. J., Ibid., 1951, 41, 450. Huke, F. B., Heidel, R. H., and Fassel, R. H., Ibid., 1953, 43, 400. McAnally, J .S., Anal. Chem., 1954, 26, 1526. Gornall, A. G., and Kalant, H., Ibid., 1955, 27, 474. Nadeau, G., and Joly, L. P., Ibid., 1957, 29, 583. McCarter, J. A., Ibid., 1958, 30, 158. Chaudet, J. H., and Kaye, W. I., Ibid., 1961, 33, 113. Gemmill, C. L., Ibid., 1956, 28, 1061. Sill, C. W., Ibid., 1961, 33, 1579. Dammers-de-Klerk, A., and Spruit, F. J., J . Opt. SOC. Amer., 1956, 46, 556. Lippert, E., Nagele, W., Seibold-Blankenstein, I., Staiger, U., and Voss, W., 2. anal. Chem., Bartholomew, R. J., Dalgliesh, C. E., and Wootton, I. D. P., Biochem. J., 1957, 65, 2 7 ~ . Olson, R. A., Rev. Sci. Instrum., 1960, 31, 844. Goldzieher, J . W., and Givner, M. L., Spectrochim. Acta, 1960, 16, 620. Neporent, B. S., and Klochkov, V. P., Izvest. Akad. Nauk S.S.S.R., 1956, 20, 601. De Francesco, F., Ann.Chim., 1958, 48, 390. King, G. W., Photoelectric Spectrometry Group Bulletin, 1956, No. 9, 227. Moss, D. W., Clin. Chim. Acta, 1960, 5, 283. Jokl, J., Chem. Listy, 1958, 52, 1370. Ohnesorge, W., and Rogers, L. B., S$ectrochim. Acta, 1959, 14, 27 and 41. Olson, J. M., and Amesz, J., Biochim. Biophys. Acta, 1960, 37, 14. White, C. E., Ho, M., and Weimer, E. Q., Anal. Chem., 1960, 32, 438. Kortum, G., and Hess, W., Z . phys. Chem., N.F., 1959, 19, 142. Melhuish, W. H., J . Phys. Chem., 1960, 64, 762. Nebbia, G., Boll. Sci. Fac. Chim. Ind., Bologna, 1960, 18, ‘35. Forster, Th., “Fluoreszenz Organische Verbindungen,” Vandenhoeck and Ruprecht, Gottingen, Weber, G., and Teale, F. W. J., Trans. Faraday SOC., 1957, 53, 646. Melhuish, W. H., N.Z. J .Sci. Tech., B., 1955, 37, 142. -, J. Phys. Chem., 1961, 65, 229. Gilmore, E. H., Gibson, G. E., and McClure, D. S., J . Chem. Phys., 1952, 20, 829. Forster, L. S., and Livingston, R., Ibid., 1952, 20, 1315. The Association of Vitamin Chemists Inc., “Methods of Vitamin Assay,” Second Edition, Inter- NeBcoviC, B. A., and SoSkiC, V., Nature, 1961, 190, 1199. Davies, W., and Wilmshurst, J. R., Brit. J . Cancer, 1960, 14, 295. Bentley, H. R., and Burgan, J. G., Analyst, 1958, 83, 442. Thommes, G. A., and Leininger, E., Talanta, 1961, 7, 181. Muel, B., Hubert-Habart, M., and Buu-HOT, N. P., J . Chim. Phys., 1957, 54, 483. Hirshberg, Y., Anal. Chem., 1956, 28, 1954. Weber, G., Nature, 1961, 190, 27. Williams, R. T., J . Roy. Inst. Chem., 1959, 83, 611; Photoelectric Spectrometry Group Bulletin, Duggan, D. E., Bowman, R. L., Brodie, B. B., and Udenfriend, S., Arch. Biochem. Biophys., Udenfriend, S . , Duggan, D. E., Vasta, B. M., and Brodie, B. B., J . Pharmacol. Exp. Therapy, Udenfriend, S., Weissbach, H., and Clark, C. T., J . Biol. Chem., 1955, 215, 337. Hess, S. M., and Udenfriend, S., J . Pharmacol. Exp. Therapy, 1959, 127, 175. Waalkes, T. P., and Udenfriend, S., J . Lab. Clin. Med., 1957, 50, 733. Leaback, D. H., and Walker, P. G., Biochem. J., 1961, 78, 151. Teale, F. W. J., Photoelectric Spectrometry Group Bulletin, 1961, No. 13, 346. Weber, G., Biochem. J . , 1960, 75, 335. -, Ibid., 1960, 75, 345. Hercules, D. M., and Rogers, L. B., Anal. Chem., 1958, 30, 96. -- , Spectrochim. Acta, 1959, 14, 393. KokAski, C. J., Kokoski, R. J., and Slama, F. J., J . Amer. Pharm. Ass., Sci. Ed., 1958, 47, 715. Hornstein, I., J . Agric. Food Chem., 1958, 6, 32. Eisenbrand, J., and Raisch, M., 2. anal. Chem., 1960, 177, 1. Eisenbrand, J., and Meyer, H., Ibid., 1960, 174, 414. February, 1950; abstracted in Anal. Chem., 1950,22, 509. 1959, 170, 1. 1951, 143. science Publishers Ltd., London, 1951. 1961, No. 13, 339. 1957, 68, 1. 1957, 120, 26.February, 19621 PARKER AND REES FLUORESCENCE SPECTROMETRY. A REVIEW 111 103. 104. 105. 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. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. Wheelock, C. E., J . Amer. Chem. SOG., 1959, 81, 1348. Thommes, G. A., and Leininger, E., Anal. Chem., 1958, 30, 1361. Ott, D. G., Hayes, F. N., Hansbury, E., and Kerr, V. N., J . Amer. Chem. SOC., 1957, 79, 5448. Chen, P. S., Anal. Chem., 1959, 31, 296. Parker, C. A., and Barnes, W. J., Analyst, 1960, 85, 3. Mihul, C., Ruscior, C., and Pop, V., An. Stiint. Univ. “ A l . I. Cuza,” IaSi, 1956, 2, 199. Kats, M. L., and Sidorov, N. K., Uchen. Zap. Saratov. Univ., 1954, 40, 3. Armstrong, W. A., and Grant, D. W., Nature, 1958, 182, 747. Leverenz, H. W., “Luminescence of Solids,” Chapman & Hall Ltd., London, 1950. Starik, I. E., Stank, E. F., and Kostyrev, G. V., Trudy Radiev. Inst. Akad. Nauk S.S.S.R., 1956, Bradley, A., and Sutton, N. V., Anal. Chem., 1959, 31, 1554. Hernegger, F., and Karlik, B., Sitzungber. Akad. Wiss. Wien, Math. naturw. Klasse IIA, 1935, Rodden, C. J., Anal. Chem., 1949, 21, 327. Vozzella, P. A., Powell, A. S., Gale, R. H., and Kelly, J. E., Ibid., 1960, 32, 1430. Feofilov, P. P., and Kuznetsova, L. A., Inzhener-Fiz. Zhur., 1958, 1, 46. Sill, C. W., and Peterson, H. E., Anal. Chem., 1947, 19, 646. Pant, D. D., and Khandelwal, D. P., Curr. Sci., 1958, 27, 242. -- , J. Sci. Ind. Res., India, B, 1959, 18, 126. Bist,’H. D., Ibid., 1959, 18, 387. Noviik, M., Jaderncl Energie, 1957, 3, 44. Fassel, V. A., and Heidel, R. H., Anal. Chem., 1954, 26, 1134. Varsanyi, F., and Dieke, G. H., J . Chem. Phys., 1959, 31, 1066. Sill, C. W., and Peterson, H. E., Anal. Chem., 1949, 21, 1266. Gudjimenko, K. F., Belyi, M. U., and Skachko, M. A., Zavod. Lab., 1958, 24, 1066. Collat, J. W., and Rogers, L. B., Anal. Chem., 1955, 27, 961. Freeman, D. C., and White, C. E., J . Amer. Chem. Soc., 1956, 28, 2678. Coyle, C. F., and White, C. E., Anal. Chem., 1957, 29, 1486. White, C. E., and Cuttitta, F., Ibid., 1959, 31, 2083. Weissman, S. I., J . Chem. Phys., 1942, 10, 214. Kuznetsova, V. V., and Sevchenko, A. N., Izvest. Akad. Nauk. S.S.S.R., Ser. Fiz., 1959, 23, 2. Tomaschek, R., Reichsber. Physik, 1944, 1, 139. Parker, C. A., Photoelectric Spectrometry Group Bulletin, 1961, No. 13, 334. Elliott, G., and Radley, J. A., Analyst, 1961, 86, 62. White, C. E., and Neustadt, M. H., Ind. Eng. Chem., Anal. Ed., 1943, 15, 599. Anderson, J. R. A., and Garnett, J. L., Anal. Chim. Acta, 1953, 8, 393. Anderson, J. R. A,, and Lowy, S. L., Ibid., 1956, 15, 246. Garnett, J. L., and Lock, L. C., Ibid., 1957, 17, 574. Anderson, J. R. A., Crawford, B., and Garnett, J. L., Ibid., 1958, 19, 1. Anderson, J. R. A., Garnett, J. L., and Lock, L. C., Ibid., 1958, 19, 256; 1960, 22, 1. Veening, H., and Brandt, W. W., Anal. Chem., 1960, 32, 1426. Kosta, L., and Dubar, M., Talanta, 1961, 8, 265. Hanker, J. S., Gelberg, A., and Witten, B., Anal. Chem., 1958, 30, 93. Hanker, J. S., Gamson, R. M., and Klapper, H., Ibid., 1957, 29, 879. Nazarenko, V. A., and Shustova, M. B., Zavod. Lab., 1958, 24, 1344. Konstantinova-Schlezinger, M. A., and Krasnova, V. S., Ibid., 1945, 11, 567. Chleck, D. J., Brinkerhoff, J., Hadley, W., and Ziegler, C. A., Rev. Sci. Instrum., 1959, 30, 37. Kautsky, H., and Miiller, G. O., 2. Naturforsch., 1947, 2A, 167. 7, 111. 144, 217. , , Ibid., 1939, 8, 957. -- Received August 14th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9628700083

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

112 GORSUCH: LOSSES OF TRACE ELEMENTS [Vol. 87 Losses of Trace Elements during Oxidation of Organic Materials The Formation of Volatile Chlorides during Dry Ashing in Presence of Inorganic Chlorides BY 1’. T. GORSUCH ( U . K , Atomic Energy Authority, The Radiochemical Centre, Amersham, Bucks.) An investigation has been caxried out into the possible volatilisation losses occurring when microgram amounts of antimony, chromium, ferric iron, lead or zinc are heated in the presence of certain inorganic chlorides. Where the data are available, the free-energy changes for the postulated reactions are presented ; these are in agreement with the experimental findings. DURING the dry ashing of organic material before the determination of trace elements, part or all of a trace element may be lost by volatilisation.References are to be found in the literature that attribute some of these losses to interaction between the trace element and an inorganic chloride, particularly sodium chloride, leading to the formation of a volatile chloride.1$2 No direct experimental demonstration of such losses appears to have been reported, and the evidence adduced has mainly been the inability to recover trace amounts of materials added. Investigations of this kind are not able to distinguish between losses arising from volatilisation and those caused by retention of the trace element on the vessel or to allow for inaccuracies resulting from the treatment necessary to determine the element. The recent ready availability of radioactive tracers has made it possible to carry out a simple and unambiguous investigation of the reality of such volatilisation losses.Con- sideration was given to several common elements having chlorides volatile at 600” C and for each of which a gamma-emitting isotope was available ; physical, chemical and radiochemical data for the elements selected are shown in Table I. TABLE I PHYSICAL, CHEMICAL AND RADIOCHEMICAL DATA FOR ELEMENTS TESTED Chemical Volatile Boiling- or Tracer Element form used form melting-point, * isotope Half -life Antimony . . . . Antimonite SbCl, 223 (b.p.) lZ4Sb 60 days Chromium . . . . Chromate CrO,Cl, 116 (b.p.) 5lCr 27.5 days Iron . . .. . . Nitrate FeC1, 315 (b.p.) 59Fe 45 days Lead .. .. . . Nitrate PbC1, 501 (m.p.) 212Pb 10-6 hours Zinc . . . . . . Nitrate ZnC1, 262 (m.p.) 65zn 245 days * Values taken from “Handbook of Chemistry and Physics,” Fortieth Edition.3 “C The lead-212 tracer was prepared as already described4; the other tracers were normal Radiochemical Centre production materials.METHOD TRACER SOLUTIONS- Separate solutions of antimony, chromium, iron, lead and zinc were prepared, each containing 2 p g per ml of the element in the chemical form shown in Table I and each labelled with the corresponding active isotope. CHLORIDE SOLUTIONS- were prepared, each containing 200 mg of the anhydrous chloride per ml. Separate solutions of sodium, ammonium, magnesium, calcium and barium chloridesFebruary, 19621 DURING OXIDATION OF ORGANIC MATERIALS I13 PROCEDURE- Place in a small silica or porcelain crucible 1 ml of one of the tracer solutions and 0.05 ml of one of the chloride solutions.Evaporate the contents of the crucible to dryness by heating under an infrared lamp, allow the crucible to cool, and place it on the crystal of a gamma- scintillation counter, ensuring that the exact position of the crucible is readily reproducible. Obtain a measure of the activity present in the crucible by observing the rate of counting, in counts per minute, and applying a correction for the background count of the equipment. Transfer the crucible to a muffle furnace, and heat it for a number of hours at the selected temperature. Remove the crucible from the furnace, allow to cool, and return it to the gamma- scintillation counter, ensuring that its position on the crystal is exactly the same as before.Re-determine the activity by again observing the count rate. Correct the second count for background and for any decay of the nuclide since the time of the first count. The percentage recovery of the trace element is then given by the expression- Corrected counts per minute after heating x 100 Counts per minute before heating RESULTS Table I1 shows the results obtained when each of the elements listed in Table I was heated in the presence of sodium chloride or ammonium chloride and also results for zinc heated in the presence of magnesium, calcium and barium chlorides; in each experiment, the crucible was heated at 600" C for 16 hours. TABLE I1 RECOVERIES OF TRACE ELEMENTS AFTER HEATING IN PRESENCE Element Antimony .. Chromium . . Iron111 .. Lead . . .. Zinc . . .. OF INORGANIC CHLORIDES Compound added Recovery, % Sodium chloride . . .. . . . . lol,looD1oo,loo Ammonium chloride . . .. .. 11, 6, 19, 8 Sodium chloride . . .. .. .. Ammonium chloride . . .. .. Sodium chloride . . . . . . . . Ammoniumchloride . . . . . . Sodium chloride . . .. .. .. Ammonium chloride . . .. .. Sodium chloride . . .. a . .. Ammonium chloride . . .. .. . . Magnesium chloride .. .. .. Calcium chloride . . .. .. .. 4 ..( i Barium chloride . . .. .. .. 97, 99, 95 99, 100, 97, 97 98, 100 93, 96, 97, 96 95,95 69, 75, 64, 76 102,100 7, 7, 9, 6 53, 52, 52 99,101,100,100 4D 4D 4, TABLE I11 FREE-ENERGY CHANGES IN REACTIONS OF OXIDES Reaction Fe20, + 6NaCl = 2FeC1, + 3Na20 . . . . . . Sb20s + 6NaC1 = 2SbC1, + 3Na20 . . . . . . PbO + 2NaC1= PbCI, + Na20 .. .. .. ZnO + 2NaCl= ZnC1, + Na20 . . .. .. Sb20s + 6HC1= 2SbC1, + 3H20 . . .. .. Fe,O, + 6HC1= 2FeC1, + 3H2O . . . . .. PbO + 2HC1= PbCl, + H20 . . .. .. ZnO + 2HC1= ZnC1, + H20 .. .. .. Free-energy change (AG?) at 600" C, kilocalories +261 +313 + 73 + 79 - 24 + 26 - 24 - 11 DISCUSSION OF RESULTS A major advantage of the procedure used in this investigation is that the penetrating nature of the gamma radiation permits a measure to be taken of the amount of radioactive material inside the crucible without in any way disturbing the contents. The counts before114 GORSUCH: LOSSES OF TRACE ELEMENTS [Vol. 87 and after the heat treatment are therefore directly related. This avoids the confusion that might arise from incomplete extraction of the trace element or from any inaccuracy in pre- paring the solution for the determination.Further, any loss occurring through volatilisation of the trace element cannot be made up by contamination from apparatus or reagents, as all such contamination is inactive and will not be measured. Determinations of this kind demonstrate losses in a clear and unequivocal way, and, by carrying out all measurements in situ, the losses measured are confined to those arising from volatilisation. Radio- active decay is essentially a statistical process, and the standard deviation for a given count is equal to the square root of the number of counts recorded. In most experiments sufficient counts were recorded to give a standard deviation on each result of 0.5 per cent. or less.In the experiments involving lead-212, the count rates were appreciably lower and the standard deviations were approximately 1 per cent. As well as the statistical variations the possibility of geometrical errors must be con- sidered. If in an experiment the material within the crucible moved during the heating and cooling, or if the position of the crucible on the crystal was not exactly the same before and after heating, then the geometrical relationship between the source and the detector would be altered and the efficiency of counting altered with it. Finally, the duration of each experiment involving lead-212 was long compared with the half-life of the nuclide, and large corrections for decay had to be made before the initial and final counts could be compared.The position was aggravated by the need to delay count- ing to ensure that the daughter nuclide, bismuth-212, was in equilibrium with its parent. After taking account of all the points mentioned above, it is considered that the results reported should be accurate to within about +5 per cent. An estimate of the reproducibility of the results can be made if it is assumed that all recoveries greater than 90 per cent. are complete. This is reasonable when the nature of the volatile species is considered and account taken of the magnitude of the -losses experienced in the presence of ammonium chloride. The coefficient of variation for the results reported is then found to be 2.9 per cent. The results in Table I1 for the experiments with sodium chloride show that no loss occurred for any of the elements investigated.This is in direct opposition to statements by Jackson,l Monier-Williams2 and others, in which the risk of losing iron during heating with sodium chloride is specifically stressed. As these findings are at variance with accepted analytical opinion and as sodium chloride is of such wide-spread occurrence in biological materials, further corroboration was sought. The feasibility of a chemical reaction is governed by the free-energy change associated with it, and free-energy data are available for many oxides and chlorides over a wide range of temperatures. From Osborn's data,5 supplemented when necessary by those of Kelloggc and Ellingham,' free-energy changes were calculated for the reactions postulated in Table I11 between various oxides and sodium chloride or hydrochloric acid ; thermodynamic data for the formation of chromyl chloride were not available.In the calculations, it was assumed that the metal compounds used would react as the oxides and that the reaction between the oxides and hydrochloric acid would represent the situation existing in the presence of ammonium chloride, with which a high degree of dissociation occurs at elevated temperatures. It can be seen from these calculations that none of the reactions involving sodium chloride is feasible, which agrees with the experimental findings. Indeed, the high value for the free energy of formation of sodium chloride (AG$98 = -183,820 kilocalories, for 2 NaC1) indicates that few, if any, metal oxides are likely to react with sodium chloride to form the corresponding metal chlorides at temperatures used in the dry ashing of organic com- pounds.However, it can also be seen that the reactions of hydrochloric acid with the oxides of antimony, lead and zinc can lead to the formation of the chlorides, but that the reaction with ferric oxide cannot. The comparatively small loss of lead in the presence of ammonium chloride can probably be attributed to the low vapour pressure of lead chloride at 600" C; Smiths gives this as 2.8 mm of mercury. Table I1 includes three sets of results for the recovery of zinc after heating in the presence of magnesium chloride, calcium chloride or barium chloride. The thermodynamic data for calculating the free-energy changes to be expected when zinc oxide reacts with these three chlorides are a ~ a i l a b l e , * ~ ~ ~ ~ but the results obtained from such calculations are not fully in The accuracy of the results so obtained is somewhat difficult to define precisely.This, again, is in complete accord with experiment.February, 19621 DURING OXIDATION OF ORGANIC MATERIALS 115 agreement with the experimental results. The free-energy changes at 600” C for the reaction- ZnO + MC1, = ZnC1, + MO where M can be magnesium, calcium or barium, are: magnesium chloride, AGO = -10 kilo- calories ; calcium chloride, AGO .= +SO kilocalories; barium chloride, AGO = +52 kilocalories. On this basis, the reaction of zinc oxide with calcium chloride should not lead to loss of zinc; in fact, the loss is almost complete. In the free-energy calculations, however, it is assumed that the anhydrous chlorides are involved, whereas the experimental procedure leads to the production of hydrated chlorides, to which the data will not necessarily apply. I thank Dr. J. Down and Mr. D. A. Lambie for assistance in preparing the manuscript. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. Jackson, S. H., Ind. Eng. Chem., Anal. Ed., 1938, 10, 302. Monier-Williams, G. W., “Trace Elements in Food,” Chapman & Hall Ltd., London, 1949, p. 257. Hodgman, C. D., Weast, R. C., and Selby, S. M., Editors, “Handbook of Chemistry and Physics,” Gorsuch, T. T., Analyst, 1960, 85, 225. Osborn, C. J., Trans. Amer. Inst. Min. Met. Eng., 1950, 188, 600. Kellogg, H. H., Ibid., 1950, 188, 862. Ellingham, H. J. T., J . SOC. Chem. Ind., 1944, 63, 125. Smith, A. N., Trans. Brit. Ceram. SOC., 1949, 48, 85. Fortieth Edition, Chemical Rubber Publishing Company, Cleveland, Ohio, 1958. Received September 1 Sth, 1961

ISSN:0003-2654    

DOI:10.1039/AN9628700112

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

[Vol. 87 116 MAEHLY: A MICRO TECHNIQUE FOR IDENTIFYING A Micro Technique for Identifying Barbiturates in Forensic Chemistry BY A. C. MAEHLY (Government Laboratory for Forensic Chemistry, Stockholm 60, Sweden) A simple apparatus is described by means of which sublimation in vacuo on to potassium bromide discs can be carried out with as little as 10 to 20 pg of barbiturate. The infrared spectrum and the melting-point can then be determined on the same sublimated sample. AN efficient method recently described for determining barbiturates in tissues, blood and urine1 consists in extraction of the barbiturate into chloroform and subsequent paper chromato- graphy of the chloroform solution. Quantitative results are obtained by ultraviolet spectro- photometry of the material eluted from the spots on the chromatogram.Identification by infrared spectrophotometry and by melting-point determinations presents no great problem when only one derivative of barbituric acid is present in reasonably large amount. Lists of infrared spectra of many barbiturates have been published by Levi and Hubley,2 Manning and O’Brien3 and Cleverley4; the spectra presented in these papers are listed in Table I. Alha and Tamminen5 tried a number of micro techniques for obtaining the infrared spectra of a variety of toxicologically important compounds; they used a Perkin - Elmer instrument incorporating a microscope. We had independently found that evaporation to dryness of concentrated chloroform solutions of barbiturates on a potassium bromide disc, and especially direct sublimation on to such a disc, gave reasonably satisfactory results.However, at the temperatures necessary for sublimation, some decomposition of the barbiturates was often obierved. made subsequent determinations of melting-points difficult or impossible, This caused changes in the infrared TABLE I SUMMARY OF PUBLISHED INFRARED SPECTRA OF BARBITURATES spectra and Barbiturate 5-Allyl-5-phenylbarbituric acid 5-Isopentyl-5-ethylbarbituric acid 5-Allyl-5-isopropylbarbituric acid 5,5-Diethylbarbitunc acid (barbitone) &Ethyl-$( 1-methylprop yl) barbituric (amobarbitone) (aprobarbitone) acid (butabarbitone) Spectriim shown in reference No.* Barbiturate 5-Ethyl-5- (1-methylbutyl) barbituric 5-Ethyl-5-phenylbarbituric acid 5-Ethyl-5-isopropylbarbituric acid 5-Methyl-5-phenylbarbituric acid acid (pentobarbitone) 2J 3l (phenobarbitone) (probarbitone) (Rutonal) l 2 3, 4 5-n-Butyl-&ethylbarbituric acid 5-Allyl-5-isobutyl) barbituric acid 5- (Cyclohexen- 1-yl) -5-ethylbarbituric acid 5-Allyl-5- (cyclopenten-kyl) barbituric acid 5,5-Diallylbarbiturk acid 5- (Cyclohexen-2-yl) - lI5-dimethylbarbituric 5-Ethyl-1-methyl-5-phenylbarbituric (butethal; butobarbitone) 2, 4 (sandoptal) 5-Allyl-5-( 1-methylbuty1)barbituric acid 5-Allyl-5-( 1-methylpropyl) barbituric 5-Allyl-5- ( 1 -methylbutyl) thiobarbituric 5-Ethyl-5-( 1-methylbutyl) thiobarbituric 5-Ethyl-5- ( 1-methyl- 1-butenyl) barbituric (cyclobarbitone) 3, 4 (secobarbitone) 2, 3, 4 (cyclopal) 4 acid (talbutal) 3 (dial ; allobarbitone) 2, 3 acid (thiamylal) 4 acid (hexobarbitone) 3 acid (thiopentone) 3 acid (mephobarbitone) 2, 3 acid (vinbarbitone) 3 * See reference list, p.120. Spectrum shown in reference No. * 2l 3J 2 2, 4 4 If several barbiturates are present in the same extract, and in amounts less than 100 pg, the experimental difficulties are considerably increased. We obtained satisfactory spectra by using Mason’s micro-pellet technique,6 the main disadvantage of which is the loss occurring when the sample is re-extracted from the pellet; such re-extraction is often necessary for obtaining material for melting-point determinations. Further, the compounds apparentlyFebruary, 19621 BARBITURATES I N FORENSIC CHEMISTRY 117 undergo certain non-reproducible changes during the pressing of the pellets, as observed by Jensen,' who recommended that the pellet be warmed to 100" C after it had been pressed in order to obtain reproducible spectra.A similar technique was suggested by Cle~erley.~ To vacuum and outlet tubes Fig. 1. Apparatus for sublimation on mlcro scale. The water inlet and outlet tubes are con- nected to the cooling block by thick-walled rubber tubing. Standard glass joints (B34 and B45) are used t s .- 5 2 U Q 2 I I 1 - 1 I I 1 1 I 1 I I 1 1 - 1 1 4000 3000 2000 1800 1600 1400 1200 1000 800 Wavelength, cm-' Fig. 2. Infrared spectrum of dial extracted from liver tissue (Case No. 1476/60) and sublimed in vacuo on to a potassium bromide disc. Spectrum recorded on a Hilger H800 instrument (with microscope) fitted with a sodium chloride prism: slit width, 0.16 mm a t 4000 cm-1; gain, 5 ; damping, 3; recording time, 22.5 minutes APPARATUS AND TECHNIQUE In our institute, a simple apparatus for sublimation in vacuo on the micro scale was constructed and has served its purpose well.This apparatus is shown in Fig. 1 and consists of three glass parts (the main body, the lower portion and inlet and outlet tubes for cooling water) and a cylindrical brass cooling block. A solution of the barbiturate in chloroform (generally the eluate from a paper-chromatographic spot) is transferred to the small finger of the lower part of the apparatus, which is then placed on a bath of hot water. A fine118 MAEHLY: A MICRO TECHNIQUE FOR IDENTIFYING [Vol. 87 Wavelength, cm-' Fig. 3. Infrared spectra of dial in the region 1100 to 1500 cm-l: curve A, same sample as in Fig.2; curve B, reference sample. Recording conditions as for Fig. 2, but spectra recorded for 5 minutes stream of air is directed on to the surface of the solution to prevent bumping, and, after evaporation of the solution to dryness, the lower part of the apparatus is allowed to cool. A cover-glass is then placed on the flat surface above the finger, the apparatus is assembled, the cooling water is turned on, a vacuum pump (an oil pump giving 1 mm of mercury) is connected and started, and sublimation of one fraction is carried out by warming the finger in a paraffin bath heated by a small gas burner. The apparatus is opened, the cover-glass replaced by a potassium bromide disc, the apparatus re-assembled, and the main fraction sublimed on to the disc.Heating for a few minutes at 100" to 120" C generally suffices for TABLE I1 RESULTS FOR BARBITURATES IN VARIOUS FORENSIC CASES Amount of barbiturate*- Melting-point of- -7 f A -l found per 100 g of isolated for Infrared pure Case No. Sample sample, sublimation, spectrum sample, barbiturate, mixture, mg Pg "C "C "C Death in tra@c accident- 1476/60 Liver; 100 g -0.3 each of -300 of D Fig. 3, curve A 169 to 172 171 to 173 167 to 170 A and D { -300 of A Fig. 4, curve A 137 to 140 139 to 141 137 to 139 Suicide; overdose of baybiturates- 0.5 of P, 2.0 of A 2486/60 Liver; 30 g 2.5 of V and -500 of A Characteristic 138 to 142 139 to 141 138 to 140 Suicide; overdose of barbiturates plus glutethiwzide- 3582/60 Liver; 50 g 15 of C -250 of C characteristic 170 to 173 167 to 170 171 to 174 4588/60 Liver; 30 g 3.1 of V and -200 of A Characteristic 139 to 142 139 to 141 138 to 140 1.8 of A 5621/60 Liver; 30 g 2.2 of V and -40 of V Fig.4, curve C 164 to 166 161 to 164 Impure 1.6 of A fraction 673/61 Kidney; 42 g 0.3 of V and -120 of V Characteristic 160 to 163 161 to 164 159 t o 162 Death from secondary effects of barbiturate intoxication- 0.6 of P Efiileptic; drowned while under the influence of bavbiturate- 7311/61 Liver; 50g 9.1 of P -250 of P Characteristic 175 to 176 174 to 176 175 to 177 * A = aprobarbitone; C = cyclobarbitone; D = dial; P = phenobarbitone; V = vinbarbitone.February, 19621 BARBITURATES IN FORENSIC CHEMISTRY 119 completing sublimation. Amounts of barbiturates from 10 to 100 pg have been collected on one O.5-inch potassium bromide disc, and the optimal amount is from 20 to 50 pg.The disc is then placed on the stage of the infrared microscope; we use a Hilger H800 spectrophotometer fitted with a Hilger reflecting microscope. The zero line is adjusted by attenuation of the reference beam (a match placed diagonally over the beam gives good results), and the signal is set to a maximum at an absorption peak by trial and error, the disc being moved in the horizontal plane of the microscope stage. Infrared spectra of the barbiturates differ, especially in the region 1100 to 1500 cm-l (6.7 to 7.1 p). and comparisons are best made in this region. Fig. 2 shows the complete spectrum (650 to 4000cm-l) of dial (5,5-diallylbarbituric acid) extracted from liver tissue ; Figs.3 and 4 show the spectra of three different barbiturates in the region 1100 to 1500 cm-l, reference spectra also being included. (The reference samples were sublimed in the same way as the barbiturates isolated from tissue.) n I I I I I I500 I300 I 100 Wavelength, cm-' Fig. 4. Sample and reference spectra of (a) apro- barbitone and (b) vinbarbitone. Curve A, aprobarbitone from liver tissue (Case No. 1476/60); curve B, reference spectrum; curve C , vinbarbitone from liver tissue (Case No. 5621/60); curve D, reference spectrum. Recording conditions as for Fig. 3 After the spectrum has been recorded, the potassium bromide disc is placed upside down on a clean glass slide, and the melting-point of the sample is determined with an electrically heated microscope stage (Kofler block).The material on the cover-glass is used for determination of a mixed melting-point. In this way, spectra and melting-points can be obtained from as little as 10 to 20pg of barbiturate with minimal loss. APPLICATION OF THE TECHNIQUE Case No. 1476/60-B.K., a 61-year-old male, was run over by an automobile; death was caused by fractures of the skull and right leg. Analysis for ethanol (both by titration and enzymatically) gave the values: blood, 0.15 per cent.; kidney, 0.17 per cent; urine,120 MAEHLY [Vol. 87 0.25 per cent. Analysis of 100 g of liver and 30 ml of blood for acidic and neutral drugs by Bonnichsen, Maehly and Frank’s method1 indicated the presence of two barbiturates (dial and aprobarbitone) and caffeine; spectrophotometricl~8 determination gave the results shown below.Drug . . .. .. .. .. . . Dial Aprobarbitone Caffeine Amount found per 100 ml of blood, mg . . -0.2 -0.2 -0.9 Amount found per 100 g of liver, mg . . -0.3 -0.3 -0.8 The two barbiturates were separated and purified by paper chromatography,l and known amounts were eluted from the chromatograms and subjected to sublimation i.p2 vacuo as described above; the results are included in Table 11. On the basis of these analyses, the medical examiner stated that the victim was run over while under the influence of alcohol and barbiturates, and this statement was used in court. The results obtained when these techniques were applied to six other cases are also summarised in Table 11. Column 3 lists the concentrations of the individual barbiturates as computed from spectrophotometric measurements, and column 4 the amounts of each barbiturate isolated and transferred to the sublimation apparatus. 1. 2. 3. 4. 5. 6. 7. 8. REFERENCES Bonnichsen, R., Maehly, A. C., and Frank A., J . Forensic Sci., 1961, 6, 411. Levi, L., and Hubley, C. E., Anal. Chem., 1956, 28, 1591. Manning, J. J., and O’Brien, K. P., Bull. on Narcotics, 1958, 10, 25. Cleverley, B., Analyst, 1960, 85, 582. Alha, A. R., and Tamminen, V., Ann. Med. Exp. Fenn., 1959, 37, 167. Mason, W. B., Paper presented at the Pittsburg Conference on Analytical Chemistry and Applied Spectroscopy, March, 1958, and distributed by the Perkin - Elmer Corporation, Nonvalk, Connecticut. Jensen, J. B., Dansk Tidskr. Farm., 1958, 32, 205. Bonnichsen, R., Maehly, A. C., and Nordlander, S., J . Chromatography, 1960, 3, 190. Received July 20th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9628700116

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

February, 19621 DUCKITT AND GOODE 121 A Method for High-precision Assay of Uranium Metal BY J. A. DUCKITT AND G. C. GOODE (U.K. Atomic Energy Authority, Atomic Weapons Research Establishment, Aldermaston, Berks.) A method is described for the rapid determination of uranium in natural and enriched uranium. Reduction of uraniumV1 to uraniumN is avoided by solution of the metal in orthophosphoric acid, and precipitation of uranous phosphate is precluded by careful control of acidity. A weighed excess of potassium dichromate is added, and the excess is titrated potentiometrically with standard ferrous solution added from a weight-burette. The procedure has a standard deviation of 0.02 per cent., and it was necessary to use uranium metal of known purity as a primary standard. The effects of some interfering elements have been studied.AS in the analyses of other metals, determination of the uranium content of “pure” uranium by difference after the determination of impurities has been considered the most reliable method. For routine work, however, in which many samples have to be analysed, such a procedure is unsuitable and a rapid high-precision assay is required. Many volumetric procedures have been described for the precise determination of uranium, usually involving the oxidimetric titration of uraniumIv after reduction of uraniumV1 by a suitable reducing agent. Reduction of uraniumv1 to uraniumIv is some- times accompanied by formation of uraniumIII, but this can be readily oxidised to uraniumIV by aeration. Ceric sulphate or potassium dichromate has been used as oxidant, and titra- tions are carried out visually, with a suitable indicator, or potentiometrically.Some reviews of such procedures have appeared.l Y 2 Previous experience of the assay of uranium metal with lead or zinc amalgam as reducing agent has shown that a standard deviation of 0.1 per cent. can readily be attained when 1-g samples are used. Increased precision can be obtained by suitable refinements in titration technique, e.g., by the use of weight titration, but the considerable care needed when using these reducing agents prompted a search for an alternative procedure. Ideally, such a method would involve no transference of solutions during the determination, and the reduction step would be eliminated by dissolving the sample to the quadrivalent state by means of a suitable non-oxidising solvent.In one such pr~cedure,~ a mixture of hydrochloric, sulphuric and fluorosilicic acids was used as solvent. This paper describes investigations into the use of orthophosphoric acid for this purpose and the evaluation of a rapid high-precision procedure for assaying uranium and its alloys. EXPERIMENTAL SOLUTION OF THE METAL- The solution of uranium and its alloys has been discussed by Larsen.* The pure metal is reported to dissolve at a moderate rate in 85 per cent. w/w orthophosphoric acid to give a clear solution of uraniumIV, but use of this solvent in the analysis of uranium is generally avoided because of the precipitation that results on dilution with water and the undesirable interference of phosphates in many determinations.It has been shown that 1-g samples of the metal rapidly dissolve in the hot acid, giving a clear green solution, and that precipitation can be avoided if a 5 per cent. v/v solution of orthophosphoric acid in 3 M sulphuric acid is used as diluent. TITRATION PROCEDURE- The procedure chosen for titrating the uraniumIv solution involved addition of an excess of potassium dichromate and titration of the excess with a ferrous solution. To achieve the maximum possible precision, additions of oxidant and titrant were made by weight and titrations were carried out potentiometrically. Small polythene weight-burettes capable of dispensing drops of titrant weighing 0.01 to 0.015 g were found to be convenient for the titration.Although the weight-titration technique was somewhat more time-con- suming than a conventional volumetric procedure, this was more than off set by the elimination of the reduction step.122 DUCKITT AND GOODE: A METHOD FOR [Vol. 57 EFFECTS OF IMPURITIES- A number of elements could give rise to interference in this procedure by solution to valency states capable of consuming titrant and also by possible catalytic oxidation of uraniumxv. Some investigations of these interferences have been made with a view to their removal by oxidation of the elements to higher-valency states or to their stoicheiometric correction. The stability of solutions of uraniumIv phosphate towards oxidation by air suggested that simple aeration of solutions might be an effective means of overcoming some of these interferences. Iron-Experiments were carried out in which known amounts of pure iron granules were added to uranium metal of known purity, and the mixture was dissolved in orthophos- phoric acid and assayed.Results were calculated by assuming stoicheiometric correction for the amount of iron present and showed such a method to be inaccurate and not reproducible. Results of further determinations on samples containing added iron, but in which solutions were aerated for 15 minutes before potassium dichromate was added, are shown below, the uranium content being calculated without correction for the iron present. Sample No. . . .. 1 2 3 4 5 6 Iron added, yo . . . . Nil 0.1 0.1 0.1 1.0 5.0 Uranium found, % . . 99-99* 99.97 99.99 99.98 101.60 105.9 *Mean of ten determinations.Aeration for 45 minutes of another sample containing 1 per cent. of added iron gave a result for uranium of 100.31 per cent., but the increased time of aeration was without effect on a sample to which no addition had been made. Iron, which is one of the major impurities in uranium, is rarely present to an extent greater than 500 p.p.m., and aeration for 15 minutes would satisfy normal requirements. Chromium-Preliminary experiments with pure chromium metal showed it to dissolve readily in orthophosphoric acid, but difficulties were experienced in the detection of end-points when solutions were titrated. Aeration of the solutions removed these difficulties ; results of assays in the presence of added chromium were- Sample No.. . .. .. 1 2 3 4 Uranium found, yo . . I . 99.99" 100~01 99.97 99.99 Chromium added, yo . . . . Nil 0.1 1.0 5.0 * Mean of ten determinations. These results show that the procedure is entirely satisfactory in the presence of up to 5 per cent. of chromium, and the applicability of the method to chromium - uranium alloys was confirmed by analysis of an alloy containing 0.12 per cent. of chromium. The mean of the results found for uranium by two analysts was 99.91 per cent., and their individual results were- Uranium found by analyst A, yo . . . . 99.91 99-92 Uranium found by analyst B, Yo . . . . ' 99-92 99.88 TABLE I RESULTS FOR URANIUM IN PRESENCE OF MOLYBDENUM AND VANADIUM Sample No. Molybdenum as impurity Impurity added, p.p.m. Nil 10 10 100 100 250 500 1000 7 Uranium found, 99-97" 99.94 99-96 100~00 99-98 99.99 100.12 100.16 % Vanadium as impurity r Impurity added, p.p.m.Nil 10 10 100 100 250 500 1000 * Mean of five determinations. Uranium found, % 99.97" 99.93 99.96 99.95 99.95 99.96 100.04 100.15 Molybdenum and vanadium-The effects of these elements were again determined by addi- Vanadium metal dissolved readily in orthophosphoric acid, but the use tion experiments.February, 19621 HIGH-PRECISION ASSAY OF URANIUM METAL 123 of pure molybdenum was precluded by its low solubility, and a uranium alloy containing 10 per cent. of molybdenum, which dissolved readily, was used. Results of assays in the presence of known amounts of these elements are shown in Table I; a 15-minute period of aeration was used. These results show that the effect of up to 250 p.p.m.of either element can be overcome by aeration, but, for larger concentrations, interference is significant. The possibility of stoicheiometric correction for these elements in the absence of aeration was investigated by assaying several alloys of known composition. Again, results were not reproducible and did not correspond to any simple change in valency. Other elements-The effects of some other elements were investigated, the aeration procedure being used, and Table I1 shows results of assays in the presence of copper, nickel, zinc and cobalt. TABLE I1 RESULTS FOR URANIUM IN PRESENCE OF COPPER, ZINC, NICKEL AND COBALT Sample No. Impurity added Uranium found, yo 1 Nil . . . . .. .. . . .. .. .. .. 99.97* 2 1000 p.p.m.of copper . . . . . . . . .. . . .. 99.95 3 1000 p.p.m. of copper plus 500 p.p.m. of nickel . . . . .. 99.98 4 1000 p.p.m. of copperplus 500 p.p.m. each of nickel, cobalt and zinc 99.99 * Mean of five determinations. METHOD APPARATUS- Titrations were carried out with a Cambridge electro-titration unit fitted with platinum and saturated-calomel electrodes. The polythene weight-burette used was capable of dis- pensing drops weighing 0.01 to 0.015 g; it was conveniently made by drawing out the neck of a 2-02 polythene bottle into a fine jet. REAGENTS- All reagents should be of recognised analytical grade. Orthophosphoric acid, 88 per cent. w/w. Sulphuric acid, approximately 3 M. Nitric acid, diluted (1 + 1). Sulphuric acid - orthophosphoric acid wash solution-Add 50 ml of 88 per cent.ortho- Potassium dichromate-Grind the crystalline solid to a powder, heat to 150" C, and allow Ammonium ferrous sulphate, approximately 0.05 N-DiSSOlVe approximately 20 g of phosphoric acid to 450ml of 3~ sulphuric acid. to cool in a desiccator. (NH4),SO,.FeSO4.6H,O in water, add 20 ml of 3 M sulphuric acid, and dilute to 1 litre. PROCEDURE- Before analysis, remove grease from the samples of metal by treatment with a suitable solvent. Remove any surface oxide by warming with nitric acid diluted (1 + 1) until a bright surface is produced, and then wash thoroughly with water to remove all traces of acid. Finally, rinse with acetone, and dry with compressed air, avoiding the use of heat. Weigh 1 0-005 g of the prepared sample to within 0.00002 g, transfer to a Pyrex-glass 100-ml beaker, add approximately 10 ml of orthophosphoric acid, and cover the beaker with a watch- glass.Heat on a hot-plate until the metal begins to dissolve, and then remove to a cooler part of the plate; solution is usually complete in a few minutes, but occasionally small amounts of insoluble material remain, necessitating more prolonged heating. Allow to cool, but, while still warm, rinse the watch-glass and sides of the beaker with the acid wash solution from a wash-bottle. Pass a stream of air through the solution for about 15 minutes, remove the delivery tube, and rinse it thoroughly with a further portion of wash solution. Prepare a solution containing 0.435 & 0.001 g, weighed to within 0.00002 g, of potassium dichromate, and add it to the uranium soIution; then add 20 ml of 3 M sulphuric acid, and stir until solution is complete.Titrate the solution potentiometrically against 0.05 N am- monium ferrous sulphate added from the weight-burette until 1 drop of titrant gives full-scale deflection with the instrument set at maximum sensitivity. Determine the weight of titrant124 DUCKITT AND GOODE [Vol. 87 used to the nearest milligram. Standardise the ammonium ferrous sulphate solution against weighed amounts of potassium dichromate, adding 20 ml of 3 M sulphuric acid and 10 ml of orthophosphoric acid before titration. RESULTS The precision of the method was ascertained from two independent series of determina- tions on a bulk sample of uranium metal shown, by determination of impurities, to contain 99.95 per cent.of uranium. The results are shown in Table 111, from which it can be seen that the procedure is free from operator bias and that the precision is of the order expected. However, when the mean result is compared with the value obtained by difference after the determination of impurities, a bias of 0.04 per cent. is evident. TABLE I11 RESULTS FOR A SAMPLE OF NATURAL URANIUM Mean uranium Standard Analyst Uranium content found, content, deviation, % % % 0.02 0.02 99.97, 99.99, 100*01, 99.97, 100.01 99.96, 99.96, 100*00, 99.99, 99-96 { 100.04, 100-00. 99.97, 100-00, 99.99 } A { 99.97, 99.99, 100*00, 99.97, 99.97 } 99.99 99*99 The significance of this bias was investigated by means of further assays on a standard uranium metal obtained from the New Brunswick laboratory of the United States Atomic Energy Commission.This material was a certified reference standard containing 99.99 per cent, of uranium, and results for its assay with potassium dichromate from the same batch as used in previous determinations are shown in Table IV. These results show a positive bias of 0-03 per cent., which correlates well with that of the previous determinations. TABLE IV RESULTS FOR A SAMPLE OF STANDARD URANIUM Mean uranium Standard Analyst Uranium content found, content, deviation, % % % 0.01 100*01, 100.03, 100.00, 100.03, 100.04, 100.03 B 100.01, 100-03, 100.01, 100.03 100.02 A { 100*01, 100.01, 100.02, 100*01, 100.01, 100.01 } 1oo.02 CONCLUSIONS The procedure developed for assaying uranium metal has been shown to possess the required precision and is simple and rapid to carry out. It has been clearly shown that, for assays in which high precision is required, careful choice of primary standard is necessary ; even materials of analytical-reagent grade, which meet normal requirements, can be unsuitable. It is therefore recommended that uranium metal of known purity be used as primary standard in this procedure and that the potassium dichromate be standardised from this uranium. Many ions interfere seriously in the titration of uranium, and corrections for their presence are unreliable if high-precision assay is required. Aeration of solutions before titration has been shown to cause no significant oxidation of uranium, but eliminates inter- ference from iron , chromium, vanadium and molybdenum at levels considerably higher than those normally encountered in the metal. We thank Mrs. P. C. Terry and Mr. G. A. Wood for carrying out some of the determinations. REFERENCES 1. I. 9 - , Anal. Chem., 1959, 31, 1940. 3. 4. Rodden, C. J., U.S. Atomic Energy Commission Report TID-7555, Oak Ridge, Tennessee, 1957, Long, J. L., U.S. Atomic Energy Commission Report RFP-139, Denver, Colorado, 1959. Larsen, R. P., Anal. Chem., 1959, 31, 645. p. 24. Received September 13th, 1961

ISSN:0003-2654    

DOI:10.1039/AN9628700121

出版商:RSC    

年代:1962

数据来源: RSC