摘要:

ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 Ronald Belcher Memorial Lectureship In recognition of the late Professor 2. Ronald Belcher’s interest in education, the Council of the Analytical Division has instituted a Belcher Memorial Lecture to be given annually on an analytical topic by a graduate student. The award will be considered by the Honours Committee, acting on behalf of the Council of the AD. Students will be considered to be eligible if they are registered for an academic session in the year of the award. The decision concerning the award will be made in December. The aim of the award is to commemorate Professor Belcher as a teacher, by encouraging students to make a positive contribution to, and take an active part in, the profession of analytical chemistry.The paper, to be written by the student (see rule 2 below), should be in the form required for submission to The Analyst. 3. The award shall be made annually in December and shall be based on an over-all assessment of the originality of the work described in the paper and the significance of its contribu- tion to analytical chemistry. The winner of the award will be expected to present his or her work at the Research and Development Topics Meeting following the award. Rules 1. Candidates must currently be regis- tered postgraduate students of a 4. The award will take the form of a London W1V OBN. The closing date is British University. The merits of a particular candidate may be brought to the notice of the Honours Committee by any super- visor of postgraduate students regis- tered with a British University who desires to recommend the candidate, by letter addressed to the President of the Division.The letter shall be accompanied by a paper (or manus- cript ready for submission for publi- cation) of which the student is a principal co-author . presentation scroll plus a sum of f300. Thursday, September 30th, 1993. 111 conference. of the Analytical Division. ... The sum is to assist the candidate to attend a national or international conference. It will be given to the candidate, up to two years after the granting of the award, on presen- tation of satisfactory evidence of the candidate’s intention to attend such a 5 An award shall not be made if it is considered by the Honours Com- mittee that none of the papers sub- mitted reaches the required standard. 6. The decision of the Council of the Analytical Division shall be final. 7. Any alteration to these Rules shall be subject to the approval of the Council Submissions should be sent to the Presi- dent, Analytical Division, The Royal Society of Chemistry, Burlington House,

ISSN:0144-557X    

DOI:10.1039/AP993300P001

出版商:RSC    

年代:1993

数据来源: RSC

ISSN:0144-557X    

DOI:10.1039/AP99330FX021

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

June 1993 ANPRDI 30(6) 241-288 (1 993) Analytical Proceedings Proceedings of the Analytical Division of The Royal Society of Chemistry CONTENTS 241 Retirement of Professor Stuart Bark, Emeritus Professor of Analytical Chemistry, University of Salford 241 Reports of Meetings 242 VAM Viewpoint 'Analytical Chemistry and the 21st Century' 259 262 266 268 272 243 European Analytical News 'Euroanalysis VIII: European Conference on Analytical Chemistry' 'Bimolecular Photoabsorption Spectroscopy' by D. L. Andrews 'Polarographic Behaviour of N-Oxides of Oxazolidines and Their Analytical 255 257 243 IUPAC Draft Recommendations 244 SUMMARIES OF PAPERS 244 SAC 92 244 245 Application' by J. Konigstein and B. Steiner 'Lasers and Ion Trap Mass Spectrometry' by C.S. Creaser and K. E. O'Neill 'Measurement Uncertainty in Chemical Analysis' by Alex Williams 'Smuggling and Separation' by Terry Gough 'International Guide for Laboratory Accreditation' by David Holcombe 'Selective Particle Counting by Means of Affinity Ligands Linked to Microscopically 246 248 251 252 253 Visible Labels' by Derek Craston, John Francis, Harmesh Aojula, David Clarke and Robert Jepras 'European Community Measurement and Testing Programme' by Ronald F. Walker 'Proficiency Testing: How Laboratories Measure up to the Competition' by Ronald F. Walker 'Heavy Metals in Deep-sea Holothurians' by Heather M. Moore, David Roberts, Michael Harriott and Duncan Thorburn Burns 'Quantitative Analysis of the Pyrolysis Products of Perfluorocarbon Fluids by Gas Chromatography and Spectroscopic Techniques' by T.K. P. O'Mahony, A. P. Cox and David J. Roberts 'Determination of Vitamins KI, K2 (MK-4) and K3 in Animal Feeds by High- performance Liquid Chromatography' by Stephen White 'Analysis of an (Isodecyl End-capped) Propylenediol Adipate Polyester Using Coupled High-performance Liquid Chromatography-Fourier Transform Infrared Spectrometry' by Alan M. Robertson, Dell Farnan, David Littlejohn, Michael Brown, Christopher J. Dowle and Elizabeth Goodwin 'Precursors to and Evolution of Elemental Organic Tube Combustion Analysis Over the Last Two Hundred Years' by D. Thorburn Burns 276 Analytical Viewpoint 'Preliminary Operations: A Pending Goal of Today's Analytical Chemistry' by M. Valcarcel, M. D. Luque de Castro and M. T. Tena 280 Equipment News 285 Publications Received 286 Conferences and Meetings 287 Courses 288 Analytical Division Diary iii Ronald Belcher Memorial Lectureship (Rules) I Ill Typeset and printed by Black Bear Press Limited, Cambridge, England ... 111 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30

ISSN:0144-557X    

DOI:10.1039/AP99330BX023

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 24 1 Reports of Meetings North East Region The twenty-seventh Annual General Meeting of the Region was held at 10.30 a.m. on Wednesday, February 17th, 1993, at BP Chemicals, Hull. The Chair was taken by the Chairman of the Region, Dr. J. D. Green. The following office bearers were elected for the forthcoming year: Chairman-Dr. J. Marshall. Vice-Chair- man-Professor A. F. Fell. Honorary Secretary - Dr . G. M. Greenway, School of Chemistry, University of Hull, Hull HU6 7RX. Honorary Treasurer-Mr. M. Daniel. Honorary Assistant Secretary - Dr. J. R. Dean. Members of Committee- Professor K. D. Bartle, Professor M. Cooke, Dr. J. D. Green (ex oficio), Mr. A. Honeybone and Dr. R. Reeve. Mr. F. C. Shenton and Mr. J. Whitehead were re-appointed as Honorary Auditors. North West Region The sixty-eighth Annual General Meeting of the Region was held on Thursday, March 18th, 1993, at Millipore (UK) Ltd., Chester. The Chair was taken by the Chairman of the Region, Dr. I. D. Wilson. The following office bearers were elected for the forthcoming year: Chair- man-Dr. I. D. Wilson. Vice-Chair- man - Dr. D . Griffin. Honorary Secre- tary-Dr. G. Davison, 34 Beechfields, Doctors Lane, Eccleston, Chorley, Lan- cashire PR7 5RE. Honorary Treasurer- Mr . B . Taylor. Honorary Assistant Secre- tary-Mr. J. W. Ogleby. Members of Committee-Mr. E. R. Adlard, Mr. A. S. Clarkson, Mr. D. S. Corrigan, Dr. J. R. Ebdon, Dr. P. R. Fielden (co-opted) and Mr. A. Handley. Mr. K. Watson and Mr. J. Williams were re-appointed as Honor- ary Auditors.

ISSN:0144-557X    

DOI:10.1039/AP993300241b

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

242 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 Analytical Chemistry and the 2 l s t Century The consensus view of scientists from around the world indicates that there is a perceived need to do something more to improve the quality and comparability of chemical measurements. For example, the last few decades have seen the devel- opment of enormously powerful tech- niques for the analysis of complex samples in increasing detail. This, in turn, has fuelled increased customer expectation with regard to both the technical specifi- cation of measurements and their value for money. Data quality is under increas- ing scrutiny and the provision of rapid results is often also required. These trends are certain to continue and present a serious challenge to the analytical chemist and society.In addition, there are clear signs that some data are not reliable and results produced in different laboratories often do not agree. However, we increasingly live in a world where activities in one place can have a serious impact on people thou- sands of miles away. Trade, transport, health and the environment are just some areas where interdependence is increas- ing. Thus, it is essential that analytical laboratories across the world have appro- priate means of comparing and mutually agreeing their data. Support structures in the form of national institutes, professional and learned bodies and international organi- zations have been under constant deve- lopment for over 150 years and the pace of development is rapidly increasing. In Europe, for example, EURACHEM was established in 1989 to provide a focus for analytical quality improvement in the EC and E R A countries and FECS has for many years provided a forum for national chemical societies to develop pan-Euro- pean co-operation. Similar developments are taking place in other parts of the world and existing international organizations, such as BIPM, IUPAC, AOAC and ILAC, are expanding their activities.These initiatives are, however, not enough on their own to meet the chal- lenges of the 21st century. In addition, we need to develop further the systems which help analysts make reliable measure- ments and develop an international infrastructure which promotes and facili- tates reliable measurements. As part of this wider initiative, EURA- CHEM organized a successful two-day workshop on Comparability and Tracea- bility in Belgium last November (see Anal.Proc., 1993, 30, 67). More recently a broader-based international one-day workshop on the development of an inter- ternational chemical measurement system was held in association with the Pittcon Conference in Atlanta in March, 1993. The workshop commenced with a number of keynote lectures by some of the leading players in analytical chemistry, which set the scene for a constructive and lively discussion session in which 45 senior delegates from over 20 countries had the opportunity to formu- late ideas and proposals on the way forward for analytical chemistry in the 21st century. The keynote presentations covered a number of important broad-based topics including describing the progress to date in developing an international chemical measurement system (Dr. Bernard King, LGC), the role of the BIPM in high level metrology (Dr.Harry Hertz, NIST), the use of international standards as an aid to harmonizing good practice (Dr. Michael Parkany, ISO), the needs of industry (Dr. Paul La Fleur, Eastman Kodak Co.) and views from Russia on the way forward (Dr. Fridman). The discussion sessions extended the initiatives proposed at the previous EURACHEM workshop in Bel- gium and, most importantly, agreed that a new international activity be established to provide a mechanism for co-operation between key organizations and current initiatives with a view to improving the comparability of chemical measurements made in different laboratories in different countries. One key to this mechanism will be to establish traceability from the work- ing level to internationally recognized reference materials and methods and, where feasible, to the chemical SI unit- the mole.Harmonization of analytical quality practice is also an essential requirement. The new activity will also provide a mechanism for improving co- operation between key organizations and institutions with a view to improving communication, promoting harmoniza- tion of good working practices, identify- ing gaps and minimizing overlap. In view of the need to take prompt action on these matters, a 15 member working group (chaired by Dr. Bernard King) was set up at the close of the workshop and a constructive planning meeting held next day.The meeting reviewed the list of possible tasks pro- posed at the main Workshop and agreed that for 1993 the following tasks should be given priority. (a) The compilation of a dictionary of certified reference materials (CRMs) in the pipeline. This is of im- portance, not only to users of CRMs, who need to be aware of potentially-available materials, but also producers who would not wish to utilize valuable resources in duplicating a CRM already being pro- duced. The initiative will be channelled through REMCO, the IS0 committee dealing with reference material issues. ( b ) The preparation of a guide on the quality requirements for the production of refer- ence materials. There are over 100 recog- nized CRM producers around the world and many more commercial organizations who produce chemical standards in one form or another.However, at present, there is little evidence of compliance with internationally agreed quality system standards ( e . g . , IS0 9000 and IS0 Guide 25) for the production of such materials. The proposed guide will be progressed jointly by ISO-REMCO and EURA- CHEM and will provide detailed infor- mation on the requirements of reference material producers to meet the quality requirements necessary to produce refer- ence materials fit for purpose. (c) Prep- aration of a directory of chemical metrology activities. Many of the current initiatives are conducted in isolation or with an absence of knowledge of what other groups are doing. A directory which identifies such groups and their activities is a vital first component towards estab- lishing a truly integrated chemical measurement infrastructure.The work will be led by the National Institute of Standards and Technology (NIST). (d) Defining the criteria for establishing tra- ceability to the mole. At present, many chemical measurements lack traceability because of insufficient information on the total uncertainty of the measurement process, a lack of high quality reference materials, in turn due to a lack of primary methods, and no recognized system for comparability of traceable chemical measurements. The proposed criteria should provide valuable information to any group or organization concerned with improving the traceability and hence the comparability of their chemical measure- ment data. Overall, the workshop and the delib- erations of the new working group pro- duced many valuable proposals on the way forward for analytical chemistry into the 21st century. A further international workshop will be held at Pittcon '94 in Chicago next February. For further information on EURA- CHEM or the new working group please contact: The EURACHEM Secretariat, P.O. Box 46, Teddington, Middlesex TWll ONH.

ISSN:0144-557X    

DOI:10.1039/AP9933000242

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 243 IUPAC Draft Recommendations Instrumentation for the Spectral Disper- sion and Isolation of Optical Radiation IUPAC has made available for public review a draft document dealing with nomenclature, symbols and their usage in the field of instrumentation for the disper- sion and isolation of optical spectra in the wavelength region from 50 nm to 1 mm, as applied in analytical atomic and molecu- lar emission, absorption and fluorescence spectroscopy. Eleven chapters cover vari- ous aspects of dispersive and non-disper- sive spectral apparatus, including spectral filters and interferometers. Definitions are given for spectral instruments with and without detection and measuring facilities. The properties of optical com- ponents of dispersive and non-dispersive spectral instruments are defined in detail, with emphasis on such fundamental figures of merit as spectral purity, resolu- tion, resolving power, conductance of optical systems, characteristic wave- lengths and polarization. Terms closely related to the optimum use of spectral instruments, e.g., optimal slit width and height, theoretical and practical effective spectral linewidth, line to background radiant power ratio, are given. Terms for various forms of mountings for spectral apparatus are included in the vocabulary. IUPAC would welcome comments on these recommendations, before publica- tion in Pure Appl. Chern. Copies of the text may be obtained on request from Dr. A. D. McNaught, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF. Comments should be submit- ted by November 30, 1993.

ISSN:0144-557X    

DOI:10.1039/AP993300243b

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

244 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 SAC 92 The following are summaries of fourteen of the papers presented at the tenth SAC Conference held on September 20-26, 1992, in the University of Reading. The conference incorporated the third Spectroscopy Across the Spectrum Conference and the 150th anniversary celebrations of the Laboratory of the Government Chemist. Bimolecular Photoabsorption Spectroscopy D. L. Andrews School of Chemical Sciences, University of East Anglia, Norwich NR4 7TJ The development in recent years of new analytical methods based on non-linear optics has already been rewarded with a remarkable range of applications. In principle, any chemical sample can display optical non-linearity , and non-linear response to laser light in particular underlies the operation of a wide range of modern techniques such as CARS (coherent anti-Stokes Raman scattering), laser mass spectrometry and surface harmonic generation. 1-3 Most types of optical non- linearity owe their origin to the essentially simultaneous interaction of sample molecules, atoms or ions with two or more laser photons: this is precisely why highly intense light is required. However, it has recently become evident that under such high intensity conditions, individual photons can simul- taneously interact with two or more sample molecules in a type of role-reversal of conventional optical non-linearity . Bimolecular photoabsorption generally entails the co-opera- tive excitation of pairs of molecules in close proximity to each other, usually at nearest neighbour distances, through a mechanism similar to Forster energy transfer.In the simplest situation a single photon provides the energy for the excitation of two chemically different molecules A and B A + B + hv + A* + B* Given that neither the spectrum of the pure component A nor that of pure B displays absorption at the frequency v, then the above mechanism necessarily generates spectral features that can be unambiguously identified as originating from species A in interaction with B.4 The essential energetics can be understood from Fig. 1. Two different excitation schemes can be entertained for two- photon bimolecular absorption,’ as shown in Fig. 2. In Fig. 2(a), both photons are absorbed from a single high- intensity laser beam operating at a frequency that is precisely the average of the absorption frequencies of A and B.In Fig. 2(b) two photons of different frequencies are absorbed (for example, one from a pump and the other from a dye laser) and two chemically equivalent molecules are simultaneously excited. Both can be regarded as instances of mean-frequency absorption. The scheme shown in Fig. 2(a) is of more analytical potential as it identifies the proximity of two components: the main virtue of the scheme shown in Fig. 2(b) is that source tunability offers a means of exploiting one of several resonance enhancement mechanisms. Bimolecular photoabsorption affords a unique in situ method for detecting the physical proximity of two chemically different species. As such it offers considerable scope for the study of any inhomogeneous fluid in which two chemically different constituents are both present only in certain areas, as, for example, at micellar boundaries or other types of liquid- liquid interface.Another obvious area of application lies in the microanalysis of surfaces and solid interfaces. In this respect bimolecular excitation is a great deal more surface-specific and also offers considerably more scope for chemical identification than harmonic generation. The method is also applicable to the analysis of samples in which a given analyte comprises either I * I A B Fig. 1 are necessarily chemically different species Energetics of bimolecular single-photon absorption. A and B t lh lh I 6 A2 Fig. 2 Energetics of bimolecular two-photon absorption: (a) single- frequency excitation involving two chemically different species; ( b ) two-frequency excitation involving two molecules of the same speciesANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 245 one or both of the partner species: here it is possible to improve detection limits for species that are normally considered optically transparent. Experimental studies have already demonstrated the occur- rence of bimolecular photoabsorption in a wide range of media.Most studies have been performed in the gas phase or in crystals, but the effect has also been exhibited in solutions and inert gas matrices (extensive citations are given in refs. 4 and 5). A familiar example is the dimol absorption of liquid oxygen, which, through bimolecular absorption at 634 nm, is responsible for the blue colour of the liquid. Detection is typically based either on absorption or fluorescence measure- ment: the latter benefits from a much better signal-to-noise ratio, and generally requires monitoring of the fluorescence from only one of the analytes. In certain instances, where the product A* or B* spontaneously undergoes chemical reaction or dissociation, chemical or mass spectrometric detection methods can be employed: a simple example is provided by the mixture of H2C0 and D2C0.6 As further studies continue to increase the familiarity of bimolecular excitation, it is to be hoped that experimental developments will enable the full analytical potential of the technique to be realized, so that it can begin to take its rightful place alongside other methods of laser analytical spectroscopy. References Struve, W.S . , Fundamentals of Molecular Spectroscopy, Wiley, New York, 1989. Levenson, M. D . , and Kano, S. S . , Introduction to Nonlinear Laser Spectroscopy, Academic Press, Boston, 1982. Shen, Y. R., The Principles of Nonlinear Optics, Wiley, New York, 1984. Andrews, D . L., andBittner, A. M., Chem. Phys., 1992,165,l. Andrews, D . L., and Hopkins, K. P., Adv. Chem. Phys., 1990, 77, 39. Andrews, D. L., and Harlow, M. J . , J . Chem. Phys., 1983, 78, 1088. Andrews, D . L., and McCoustra, M. R. S., Spectrosc. Int., 1992, 4, 28. Polarographic Behaviour of N-Oxides of Oxazolidines and Their Analytical Application Eo 50 .- a, 1 Y a a" 30 10 J. Konigstein and B. Steiner Institute of Chemistry, Slovak Academy of Sciences, CS-842 38 Bratislava, CSFR - - - Electroanalytical determinations of N-oxides, which complete the classical redox potentiometric and chromatographic deter- minations, have been reviewed by Devinsky. ' Rapid, repro- ducible and precise determination of the N-oxides of substi- tuted 3-alkyl-5-chloromethyloxazolidines has been achieved by exploitin the readily accessible technique of d.c.polar- ography? which has also been used for controlling the synthesis of N-oxides by oxidation of the original oxazolidine derivatives with Hz02.3 This technique enables the mixture of reaction products to be determined, including adducts of N- oxides and oxazolidines, which do not interfere. It employs the polarographic currents controlled by diffusion of the active compound to the electrode.The diffusion current for a two-electron reduction limits the sensitivity of the techni ue to concentration levels of approxi- mately 5 x lo-' mol 1- . As an increase in sensitivity and the ability to determine a larger number of N-oxides are precondi- tions for the wider application of polarographic methods, it was decided to investigate the use of currents based on adsorption and catalytic effects. These are determined by adsorption of the active compound on the surface of the mercury drop electrode or by its influence on the reduction of the hydrogen overvoltage at this electrode. The polarographic behaviour of 23 derivatives of N-oxides of oxazolidines, and for comparison that of N, N-dimethyl- dodecylamine N-oxide, and their adducts with hydrogen peroxide, was investigated.The different substituents, as indicated in the general formula below, are in positions 2, 3 (alkyl; n = number of methylene units) and 5 (R = H, Cl). 9 H2C - CHCH2R The changes of substituent were always performed at one position only, while the substituents at the other positions remained unchanged. The strong electron accepting substitu- ent C1 in position 5 increases the strength of the oxazolidine ring. Consequently, the N-oxides are less easily protonated, polarographic currents are controlled by diffusion, there are fewer changes in the current and the half-wave potential, Eg, shifts with the pH. The temperature coefficient does not exceed 3% in each interval. The correlation coefficient ( r ) for the dependence of iD on h' ( h is the height of mercury above the mercury drop electrode) is 0.99 [Fig.1, curve 1; Fig. 2, current iz(D,]. Stabilization of the oxazolidine ring is due to intra- molecular interactions between the chlorine atom and the strongly polar N-oxide group in the half chair conformation, which agrees with the findings of Pihlaja and A l a j ~ k i , ~ based on NMR data. \ \. \ '. 3 \. \ \ 3.0 5.0 7.0 9.0 11.0 PH Fig. 1 Dependence of polarographic currents of oxazolidine N-oxides on pH. 1, Diffusion current of N-oxides; 2, H202; and 3, mixed (catalytic and adsorption) currents of N-oxides; 2 x , twice decreased246 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 When R = H, the 2-substituted 3-alkyl-5-methyloxazolidine N-oxide becomes less stabilized and the catalytic and adsorp- tion effects are more important and the mixed currents play a larger role [Fig.1, curve 3; Fig. 2, current iI((K+A)]. Changes in Eh are observed and the current i corresponds to protonation. An increase in the length of the alkyl chain in position 3 (n increases from 0 to 18) increases the adsorption properties of the N-oxide. Derivatives of 2,2-disubstituted 3-dodecyl-5- methyloxazolidine N-oxides (Fig. 1, curve 3) are typical representatives of this group. The 3-octyl-5-methyl derivative also belongs to this group. Substituents in position 2 do not have such a strong influence when they are strongly electronegative, e.g., dinitrophenyl, p- nitrophenyl, p-bromophenyl or trichloromethyl substituents, and they have an opposite effect on the oxazolidine ring to that of a chloro substituent in position 5.This effect itself influences the character of the polarographic current i3(D+A) or i4(K+A) (see Fig. 2). When Ed and the current are not dependent on pH, the curve for H202 limited by diffusion [Fig. 1, curve 2, il(D)] can be compared with the curve for the acyclic N,N-dimethyl- dodecylamine N-oxide [Fig. 2, i3(D+A)], which is strongly protonated. In the pH range 2-10 the value of changes by about 500 mV, whereas in the pH range 2-8 the current does not change at all. The current decreases with increasing temperature; this behaviour is typical of an adsorption current. Similarly, there is a slight increase in the current with an increase in the height of the mercury reservoir (non-linearly with the square root), viz., an 8% increase in the range 25-81 cm.On the other hand, the catalytic current of 3-dodecyl-5- E versus SCE - Fig. 2 Typical polarogram of polarographic currents [i( ,I. D, diffusion; K, catalytic; and A, adsorption. 1, H202 and addition products; 2, 5-chloromethyloxazolidine N-oxide; 3, dodecylamine N- oxide; 4, N , N-dimethyl-5-methyloxazolidine N-oxide; and 5 , support- ing electrolyte: Britton-Robinson buffer; pH 8, from -0.8 V versus SCE; sensitivity, 1.6 x A; temperature, 20 "C methyloxazolidine N-oxide grows rapidly with increasing temperature (the temperature coefficient in the range 20-70 "C is 5.6% per gradient), but decreases slightly with an increase in the height of the mercury above the electrode (about 9% in the range 25-81 cm).Catalytic and adsorption currents are easier to observe at lower concentrations of polarographically active compounds and when the lifetime of the mercury drop is short. The values of Ed for H202 and the addition products are: il(,-,) = -1.0 V; for the N-oxides: i 2 ( ~ ) = -1.15 V, i3(D+A) = -1.35 V, i4(K+A) = -1.5 V; decomposition of the supporting electrolyte occurs at - 1.7 V. The precise polarographic data for individual N-oxides are given elsewhere .2,3,5 Statistical evaluation of the concentration dependencies, employing the limited experimental data available, showed that both groups of compounds are readily determined in the concentration range from 5 x 10-~ to 1 x moll-' using the diffusion-limited current of the addition products ( Y = 0.999; a = 1.82 k 0.2 mm) (a = intercept) and that of the N- oxides (Y = 0.99; a = 2.0 & 0.6 mm).6 By using the mixed polarographic current the sensitivity can be increased (concentration range from 1 x lov6 to 1 x mol l-'), but the statistical parameters deteriorate ( e .g . , Y = 0.9). There is a strong dependence of i on the concentration of the N-~xide,~ but the accuracy of the determinations decreases, thus limiting the sensitivity. Polarographic waves are better developed and determined when the height of the mercury reservoir above the mercury drop electrode is large and the lifetime of the mercury drop is short. The influence of pH and temperature indicates that the analytical application of mixed currents of N-oxides requires strict control of the conditions, particularly pH and temperature.It might be possible to optimize the determination so that several N-oxides and their addition products with H202 could be determined in their mixtures, although, in general, the sensitivity would be rather limited. In addition, it might also be possible to differentiate N-oxides with similar properties. The significance of the polarographic technique described here is that it is suitable for controlling the purity and stability of N-oxides and for monitoring their synthesis. References 1 2 Devinsky, F . , Acta Fac. Pharm. Univ. Comeniae, 1988, 42, 191. Konigstein, J., and Steiner, B., Proceedings of the 21st Sym- posium o t Tensid$s and Detergents in JetYichovice, House of Technics CSSTS, Usti nad Labem, 1987, p. 78. 3 Konigstein, J., and Steiner, B., Proceedings of the 13th Conference of Organic Chemists, Smolenice, STU Bratislava, 1986, p.118. 4 Pihlaja, K., and Alajoki, K., Finn. Chem. Lett., 1982, 1. 5 Konigstein, J., and Steiner, B., Chem. Zvesti, 1992, 46, 28. 6 Miller, J.C., and Miller, J.N., Statistics for Analytical Chemistry, Ellis Horwood, Chichester, 1984, p. 82. 7 Bifiovec, V., and Mrazek, J., Programmable Calculator for Scientific and Technical Calculations, Texas Instruments, Praha, 1978, p. 39 Lasers and Ion Trap Mass Spectrometry C. S. Creaser and K. E. O'Neill* Department of Chemistry and Physics, Nottingham Trent University, Clifton Lane, Nottingham NG I I 8NS The combination of lasers and mass spectrometry provides a spectrometric technique .I3* Quadrupole ion trap mass spec- powerful method for extending the range of the mass trometers are particularly well suited to laser techniques because of the long irradiation times, which can be used for ion photodissociation (PD) studies, and the compatibility of these Leicestershire LEll 2RH.devices with pulsed ionisation techniques such as laser * Present address: Fisons plc, Bakewell Road, Loughborough,ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 247 desorption (LD). Various laser interfaces and applications have been reported for ion trap mass spectrometry using either direct laser introduction or fibre optic^.^" Fibre optic inter- faces are easy to install, require little modification of the ion trap and allow both PD and LD analytical experiments to be performed. Fig. 1 shows two fibre optic interface configurations im le- mented on a research ion trap mass spectrometer (ITMS).In one configuration a 1 mm fibre optic is threaded through a vacuum seal mounted in the upper flange of the vacuum housing and enters the trap through a hole in the ring electrode, whilst the other uses a flexible 0.6 mm fibre fitted into a probe, which can be introduced into the spectrometer via the solids probe lock without breaking the vacuum. In both instances, the laser light is coupled to the fibre by using a three- axis delivery system outside the vacuum housing. 1: Photodissociation Tandem mass spectrometry (MS-MS) has been used widely for structural studies and for the quantitative determination of analytes in the presence of a complex matrix. The dissociation method predominantly used for MS-MS is collisionally acti- vated dissociation (CAD),8 but it is also possible to use photodissociation for ions with suitable absorption characteris- t i c ~ .~ PD has the advantage over CAD that, in principle, the energy deposition may be controlled, since it is a function of the photon wavelength. This is illustrated by the PD/MS-MS product ion spectra of the isolated molecular ion of butylben- zene, mlz 134, acquired using photons of 488 nm (2.6 eV) and 350 nm (3.7 eV). The branching ratio of the mlz 91 (C7H7+) and m/z 92 (C7H8+.) ion intensities in the product ion spectrum reflects the energy deposition. An enhanced ratio is observed at the lower wavelength since the high energy dissociation process leading to the mlz 91 fragment is preferred at this wavelength, whereas this ion is absent when the ion is irradiated at 488 nm.It is difficult to control the energy in this way by using CAD activation. CAD and PD/MS-MS spectra may provide complimentary transfer structural information for selected ions because differences in these activation processes allow access to alternative fragmen- tation pathways. This is demonstrated by the MS-MS product ion spectra of the isolated molecular ion of naringenin, m/z 272, [2-(4-hydroxyphenyl)-5,7-dihydroxychroman-4-one] acquired by CAD and PD at 350 nm (Fig. 2). The product ions formed from the two activation methods are quite different. The CAD spectrum shows ions at rnlz 254, [M - CO]+' and mlz 166, [M - CHC6H40H]+', both of which are absent from the PD spectrum, which shows simple cleavage products at m/z 179 [M - C6H40H]+, m/z 153 [(H0)&6H2CO]+ and m/z A further advantage of using PD with an ion trap spec- trometer is the high dissociation efficiencies (10-100%) which may be obtained compared with those for linear or beam spectrometers (<1%).The sensitivity of PD with the ion trap has been shown to be compatible with chromatographic introduction of samples at the low nanogram level; an analysis which is not possible with the low efficiencies of a beam experiment .7 137 [(HO)ZC~H~CO]+. Laser Desorption The fibre optic interface can also be used for the analysis of non-volatile, thermally labile or ionic compounds by laser desorption. Here the compatibility of the ion trap with pulsed ionization techniques is important .2 Those samples which are ionic, for example the dye rhodamine 6G, need only be desorbed to yield gas-phase ions and the positive ion spectrum [Fig.3(a)] shows the cationic part of the dye at mlz 443. The spectrum was obtained simply by placing approximately 1 pl of a solution of the dye in glycerol on the tip of the probe- mounted fibre optic, which was inserted via the probe lock. The sample was subjected to laser pulses (532 nm) from a Nd- YAG laser. At this wavelength no fragments are formed but tandem mass spectrometry on the cation (mlz 443), following laser desorption, yields characteristic product ions. Fig. 3(b) shows the LD/MS-MS analysis of rhodamine 6G using CAD activation. Stainless I probe shaft steel Vacuum port p 1" to g Swagelok reducing union Fig.1 Fibre optic interfaces for ion trap laser desorption and photodissociation248 272 ?H 0 (4 100 100 150 200 250 Q) m Q) 2 100 a c - 100% = 575 300 50 100 150 200 250 300 m/z Fig. 2 MS-MS product ion spectra of the m/z 272 ion of 4‘,5,7- trihydroxyflavanone using: ( a ) , PD (350 nm); and (b), CAD Non-ionic analytes subjected to LD must undergo the dual functions of desorption and ionization. Cationized products such as [M + Na]+ or [M + K]+ are commonly formed by LD in the presence of alkali metals. However, some molecules can be ionized directly by a multi-photon ionization (MPI) mechanism, if there is sufficient energy in the laser beam. For example, the laser desorptionhonization spectrum of naringin, the glucosylrhamnose derivative of naringenin, using a single 355 nm pulse from a Nd-YAG laser at a power of 360 pJ, shows the molecular ion (mlz 580) and a fragment ion at mlz 271 resulting from the loss of the glycoside group.The simplicity of interfacing continuous wave and pulsed lasers with the ion trap using fibre optics therefore provides a convenient system for a range of laser based analytical applications. We wish to acknowledge financial support from ICI Agro- chemicals and the British Mass Spectrometry Society. 100 a, C m -a c 3 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 (a) 443 100.00% = 204 825 I I I I I - % 250 300 350 400 450 500 250 300 350 400 450 500 m/z Fig. 3 Laser desorption ion trap mass spectra of rhodamine 6G: ( a ) , LD-MS; and (b), LD/MS-MS of m/z 443 References Lasers and Mass Spectrometry, ed.Lubman, D. M., Oxford University Press, New York, 1990. March, R. E., and Hughes, R. J . , Quadrupole Storage Mass Spectrometry, J. Wiley, New York, 1989. Louris, J. N., Amy, J. W., Ridley, T. Y., and Cooks, R. G., In?. J. Mass Spectrom. Ion Proc., 1989, 88, 97. Glish, G. L., Goeringer, D. E., Asano, K. G., and McLuckey, S. A., In?. J. Mass Spectrom. Ion Proc., 1989, 94, 15. Heller, D. N., Lys, I., and Cotter, R. J., Anal. Chem. , 1989, 61, 1083. Creaser, C. S . , McCoustra, M. R. S . , and O’Neill, K. E., Proceedings of the XIX Meeting, British Mass Spectrometry Society, 1992, p. 110. Creaser, C. S . , McCoustra, M. R. S . , and O’Neill, K. E., Org. Mass Spectrom. , 1991 , 26, 335. Louris, J. N., Syka, J. E. P., Kelley, P.E., Stafford, G. C., Jr., and Todd, J. F. J . , Anal. Chem. , 1987, 59, 1677. Louris, J. N., Brodbelt, J. S . , and Cooks, R. G., In?. J. Mass Spectrom. Ion Proc., 1987, 75,345. Measurement Uncertainty in Chemical Analysis Alex Williams Laboratory of the Government Chemist, Queens Road, Teddington, Middlesex TWI I OLY One of the first problems faced by analysts is whether they have the methodology to provide a result of the required ‘accuracy’. The word accuracy is in inverted commas to denote the colloquial use rather than that given in the standard definition of metrological terms. However, after carrying out an analysis, it is very unusual for an analyst to give any indication of the ‘accuracy’ of the result. This means that the user of the result is unable to make any judgement on the confidence to be placed in it, nor is it possible to compare in a rational way the results of an independent analysis of the same sample.It is now becoming recognized that a statement of a result is not complete without including in it information about the ‘accuracy’ or the ‘uncertainty’ of the result. Indeed many customers now insist that the ‘uncertainty’ be given and this is a requirement of accreditation standards. Before going into the concept of uncertainty and itsANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 249 evaluation, it is essential to be clear about the meaning of the terminology used, since terms such as ‘accuracy’, ‘error’, ‘trueness’ and ‘uncertainty’ have colloquial meanings that are not necessarily the same as those of the formal definition.Definitions The following definitions are based on those given by the International Organization for Standardization (ISO) but have been simplified for the purpose of this paper. Error is the result of a measurement minus the true value; accuracy is the closeness of agreement between the result of a measurement and the true value; uncertainty is an estimate of the range of values within which, at a certain level of confidence, the true value is asserted to lie. These terms will be used without inverted commas to indicate that their meaning is as given in these definitions. From these definitions it is clear that there is a significant difference between the meanings of error and uncertainty. Firstly, error is the difference of two values whereas uncer- tainty is a range.Secondly, error requires a knowledge of the true value, or at least a nominal true value, and hence only has meaning when this value is known, e.g., comparison of the result of a measurement of a standard with the value quoted for the standard. Thus, the error cannot be evaluated for measurement of a sample when the value is not known and therefore is not very useful for stating the ‘accuracy’ of a result of an analysis of a real sample. Uncertainty, however, is an estimate of the range of values which is expected to include the true value, and obviously can be used when the true value is not known. In order to understand how the uncertainty can be evaluated it is necessary to appreciate these differences. Evaluation of Uncertainty Detailed evaluation of uncertainty is carried out as common practice, by national standards laboratories on the realization of base and derived units.The techniques they use, sometimes called the ‘genealogical’ approach, rely on assessing possible causes of error then, by means of subsidiary experiments and/ or theoretical analysis, determining the correction for each cause and building up an uncertainty budget by evaluating the uncertainty on the correction, even if the best estimate of the correction is zero. This can be a tedious and time-consuming operation, although often the components of the uncertainty budget associated with just a few corrections dominate. Fortunately this ‘genealogical’ approach can be considerably simplified for more routine measurements by calibration of the measurement system with ‘traceable’ measurement standards, since the calibration can reduce considerably the number of uncertainty components that have to be evaluated.Traceability For routine measurements the correct use of ‘traceable’ standards is essential to ensure reliable results and to enable the uncertainty to be evaluated. Traceability means that results are traceable to national or international measurement standards through an unbroken chain of comparisons with the uncertainty being stated at each stage. An essential element of this traceability chain is that the uncertainty is given at each stage; thus the uncertainty on the working standard used for calibration is known whichever level of the chain is used to provide the working standard.The Measurement System It is essential that the measurement system, shown schemat- ically in Fig. 1, is in what is known as a state of ‘statistical control’, i.e., repeated measurements over a period of time of standard samples processed right through the system from point A are consistent with the measured variance (relative Fig. 1 Schematic diagram of measurement system standard deviation sR) of the system. For the sake of simplicity sampling has not been included in the measurement system but in many instances it will need to be. This measured variance of the measurement system identi- fies one component of the uncertainty budget, since the over- all uncertainty must be at least as large as the uncertainty on repeated measurements of a standard sample.The system can be calibrated in a number of ways, each requiring a different number of steps in evaluating the uncertainty. Calibration can be carried out by: ( I ) measure- ment of a certified reference material (CRM), processed through the complete measurement system (traceability provided by producer of CRM); (2) by addition of known amount of analyte to unknown sample (traceability via calibrated weights and compounds of known composition); (3) addition of a tracer (spiking), e.g., isotopically or radiolabelled compound (traceability from known amount of tracer); ( 4 ) measurement of a ‘pure’ standard of the analyte, used to calibrate just the ‘comparator’ with the calibration standard introduced at point B (traceability via calibrated weights and compounds of known composition).The second component of the uncertainty budget is the uncertainty on the value of the calibration standard itself, which it is assumed can be stated as a relative standard deviation sCs. This should certainly be so if the calibration standard is a CRM, in other instances the uncertainty on the calibration standard may have to be evaluated using the ‘genealogical’ approach. The third component of the uncertainty budget arises from the reproducibility of the calibration measurements, relative standard deviation SCM. If the same calibration procedure has been used in calibrating the measurement system when determining the long-term reproducibility then it will be also necessary to consider whether or not sCM has been included in SR.The fourth component of the uncertainty budget, and the one most difficult to evaluate, arises from the suitability of the method of calibration for calibrating the whole system. Unfortunately, this component has many sub-components that have to be evaluated using the genealogical approach. If the calibration was carried out using a CRM based on a real sample it would only be necessary to consider the effect of differences in the level of contaminants and any difference in the level of the analyte in the CRM and in the sample. If the composition of the CRM was different from that of the sample then the effect of this would have to be evaluated. Also, if there was a significant difference between the calibration factor using methods ( I ) , (2) or (3) and method ( 4 ) this would have to be investigated and the associated uncertainty evaluated.Calibration by method (4) alone is clearly unsatisfactory, since it is necessary to apply the genealogical approach to all of the measurement system between points A and B, a formidable task. Thus, using the genealogical approach it is possible to assess the contributions to the uncertainty budget arising from any lack of suitability of the method used for calibration to calibrate the complete system. If this is assessed as sG (again as relative standard deviation) then the over-all uncertainty is given by:250 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 where k is a factor agreed with the user of the result and in most instances a value of two is used. Example A detailed evaluation of uncertainty has been carried out by the Laboratory of the Government Chemist (LGC) on the reference material used to calibrate breathalysers.The refer- ence material is a solution of ethanol in water and the ethanol concentration is measured by quantitative oxidation of the ethanol with potassium dichromate, using a CRM for the preparation of the dichromate solutions. The procedure is shown schematically in Fig. 2; dichromate in slight excess is added and this excess is determined by titration with ammo- nium iron(::) sulfate which is itself calibrated against the standard potassium dichromate. First the uncertainties arising from the reproducibility of the calibration procedures are evaluated; these are shown in Fig. 3, the largest arising from the reproducibility of the titration. dichromate NlST reference Ethanol oxidation I Titrate excess of 1 Standardize ammonium sulfate dichromate ri Result Fig.2 Procedure for ethanol determination 0.14 I 0.06 0.04 0.02 n Fig. 3 Physical uncertainty components (%) for ethanol determination Then considering those arising from the suitability of the calibration procedure to calibrate the complete system, there is a component arising from the conversion of mass to volume since the ethanol reference material value is required in mg ml-', and there are three further components arising from the 'chemistry' of the method used: ( I ) the purity of the dichromate standard; (2) the extent of oxidation which may not be complete or which may proceed further with the oxidation of the product of the initial oxidation; and (3) the uncertainty arising from the determination of the end-point of the titration.These are shown in Fig. 4; combining these components gives an over-all uncertainty of 0.5%, using a value of k = 2. 0.20 Fig. 4 Chemical uncertainty components (%) for ethanol determination Proficiency Testing A very good way of checking the derived uncertainty is by participation in proficiency testing, providing the organizers of the scheme give the uncertainty on the assigned value. Participation is particularly valuable for this purpose when the proficiency scheme utilizes real or natural samples. Summary It is now accepted that the statement of an analytical result should include a statement on the uncertainty of the quoted value, evaluated from: reproducibility of measurement system, sR; uncertainty on calibration standard, SCS; reproducibility of calibration measurement sCM; uncertainty derived from 'genealogical' approach, SG; all of the above giving an over-all uncertainty using expression (1).Although calibration of the measurement system using traceable standards considerably simplifies the assessment of the uncertainty, it still relies on the skill of the analyst in assessing the suitability of the calibrated measurement system for carrying out the analysis required, but after all this is what the analyst is paid to do! The author acknowledges the contribution of Dr. S. Ellison and Mr. T. Farrant in calculating the uncertainties on the data which were kindly supplied by Mr. S. Evans of the Forensic and Customs Division, LGC.ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 25 1 Smuggling and Separation Terry Gough Forensic and Customs Division, 1 aboratory of the Government Middlesex TW7 7 OLY Many different types of materials are smuggled across interna- tional boundaries, usually with a view to financial gain.By far the most publicity is given to the smuggling of drugs of abuse. The process of separation, at least for drugs from natural sources, begins at the point of harvesting. Separation occurs again at the point of seizure of the drugs by the customs or police officer. Internally concealed drugs (stuffed or swal- lowed) may need the assistance of a surgeon to effect a separation from the smuggler. Once the suspect drug reaches the laboratory, separation science plays a vital role.Identifica- tion of drugs of abuse is most frequently based on chromato- graphic retention data in conjunction with infrared spec- troscopy or mass spectrometry. A laboratory should be able to provide much more than confirmation of the identity of drugs if it is to support enforcement activities fully. The enforcement officer and scientist should work together as a partnership. The services which a laboratory should provide include analysis, advice, training, intelligence data, and research and development. Laboratory staff should also be competent expert witnesses. The particular services required will depend very much on the nature of the investigation. Identification of substances is usually straightforward and will determine which (if any) controlled drugs are present.There are many instances where at least some of a consignment is found not to be a drug substance. Purity is relevant for powdered drugs and is often an indication of the position of the suspect in the distribution chain. Although any charges will be brought on the basis of the drug itself, the identification of impurities is valuable in assigning origins of consignments. Heroin and cannabis have characteristics which are indicative of particular geographical origins. These characteristics are partly the result of climate, but also of processing differences in different countries. The Laboratory has an extensive database of chromatographic and other data on major drugs of abuse of known origin. It is thus possible to give opinions on the likely origin of seized drugs.Drugs are processed in batches and the Laboratory can compare seizures made at different times or in different locations to identify any similarities which may indicate the same batch. Items other than drugs are also examined and the Labora- tory can often provide vital evidence of association on, for example, paraphernalia, vehicles, premises, wrapping material and documents. Examination of paraphernalia, vehicles and premises may reveal traces of drugs, which could relate the suspect to colleagues or locations. Such evidence may be of value in relation to the seizure of assets. The materials in which Chemist, Queens Road, Teddington, drugs are packaged may also give valuable information. Immediate packaging is often characteristic of a particular source.Other wrappings and particularly textiles, adhesive tapes and polythene bags often have unique compositions or other characteristics which may enable a link to be established between different seizures and suspects. Documents examin- ation can also provide vital evidence. For example, the Laboratory examines air tickets and passports for unauthor- ized alterations, relates typewritten documents to particular machines, reads typewriter ribbons and compares handwriting with that of known authorship. Scientific support is not restricted to laboratory based operations and staff are’ on-call 24 h per day. In large or complex cases, the scientist working at the point of seizure can undertake preliminary tests, identify items which may be of evidential value, and advise on the handling of exhibits which may pose a safety hazard or require special treatment.The advice of the scientist at an early stage in an investigation often enables the enforcement officer to pursue particular avenues of enquiry more rapidly. It also enables the chemist or documents examiner to proceed with his work much more quickly. It minimizes the security problems of transporting bulk quantities of drugs to the Laboratory, and enables representative samples to be taken by the scientist. The Laboratory supports intelligence on drug trafficking, for example, by studying the ‘trade marks’ or other features on drugs or their wrappings which are incorporated at the manufacturing or distribution stage.Sometimes trade marks are of immediate evidential value: for example, there have been several cases in the UK where there was a common trade mark on cloth bags containing cannabis resin. Bags with identical marks containing resin of the same chemical compo- sition were seized at various points within the UK over a period of several weeks. This indicated a connection between the suspects undertaking the smuggling. Further scientific investi- gation, in which an identical cloth bag, but containing some resin shown to be of different composition, was examined, led to the identification of other parts of the distribution chain. Adulterants and impurities can also be used as markers to correlate different seizures, and can be used to follow drugs from a given illicit operation in the source country to destinations in the user country.Much of the scientific examination of drugs-related exhibits involves chromatography. Chromatography thus makes a vital contribution to ensuring justice in drug smuggling (and many other) cases.252 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 International Guide for Laboratory Accreditation David Holcombe Laboratory of the Government Chemist, Queens Road, Teddington, Middlesex TWl 1 OLY It is increasingly accepted that accreditation, i. e . , formal recognition that a testing laboratory is competent to carry out specific tests or types of tests, makes a valuable contribution to the quality of results produced by a laboratory and is one of the key ingredients for facilitating the mutual acceptance of data.Accreditation is usually assessed against the requirements of a particular standard, by an independent, expert body. Labora- tories involved in chemical testing, seeking accreditation, have a choice of three main accreditation standards. Firstly, they may work to IS0 9000 (BS 5750 in the UK, EN 29000 in Europe). This standard, defining quality system requirements, is widely applicable to organizations in produc- tion, design and servicing. For chemical testing its main application is in laboratories providing a quality control facility to the rest of the organization. Secondly, laboratories carrying out chemical testing in support of toxicology studies, for food or drug development, may have a legal requirement to comply with the requirements of Good Laboratory Practice (GLP) based on guidelines published by the Organization for Econ- omic Co-operation and Development (OECD) in 1982 and complying with EC Directives.Lastly, and the most appropri- ate option for laboratories providing a testing or calibration service to others, is to seek accreditation against standards conforming to the guidelines set out in ISO/IEC Guide 25. The most important such standard in Europe is EN 45001 with related national standards, such as NAMAS M10 in the UK. In order to be applicable to a wide range of sectors of testing or calibration EN 45001 is written in very general terms, and it is sometimes difficult to apply EN 45001 to some types of measurement without additional technical guidance ; guidance which interprets the general requirements of the standard without introducing new requirements.Almost all of the 18 EC and EFTA countries have or are developing accreditation schemes based on EN 45001. The accreditation bodies meet to discuss common points of interest in WELAC, the Western European Laboratory Accreditation Co-operation. Useful achievements include the signing of a multilateral agreement to recognize each other’s schemes as equivalent. Five members have signed so far with a further six having applied. WELAC activities cover accreditation in all areas of measurement. In Europe, the specific interests of chemical testing are represented by EURACHEM, which is a forum of senior decision makers representing government, industry, academics and professionals from the EC and EFTA countries.Both WELAC and EURACHEM recognized the urgent need to produce technical guidance to help chemical testing labora- tories interpret EN 45001. They agreed to collaborate to produce suitable guidance notes and formed a joint working group early in 1991. EURACHEM and WELAC have since identified other areas within chemical testing which can best be addressed by mutual co-operation and are working together on further projects. Production of the Guide The first draft of the Guide was based on an existing UK document. As it developed, additional material on computers, sampling, measurement uncertainty and validation was added. A full list of subject areas covered is given in Table 1. Inevitably in such a technical exercise, with many contributors, there were often widely differing views on what guidance should be given.Throughout, the group tried to produce advice that would be Table 1 Subject areas within the Guide Introduction Scope of accreditation Staff Environment Equipment Reagents Methods and procedures Calibration Reference materials Computers Audit and review Sampling, sample handling, Quality control Measurement uncertainty Validation Bibliography sample preparation acceptable to the majority, written in simple English that would be unambiguous in translation. In the latter stages of drafting, the working group collabor- ated with an IS0 group producing a similar guide on I S 0 Guide 25. Because the EURACHEMNELAC Guide was recognized as being closer to completion, and so as not to duplicate effort, work on the I S 0 Guide has been suspended for the moment.Specific Details and Problems Definitions were taken from the International Standards Organization (ISO) wherever possible. Unfortunately, many were found to be dependent on the IS0 source used. The acknowledged definitive source ‘VIM’, the International Metrology Vocabulary, dates back to 1984 and is currently under revision. A general rule was made to use the most recent source. Other sources of definitions used were IUPAC for sampling and IBM for computing. Guidance for defining a laboratory’s scope of accreditation proved to be a contentious issue. Some favour the documen- tation of clear rules; others were keen that the matter should be left entirely open. This polarization probably arises from a number of factors.Accreditation may or may not be tied into legal requirements. EN 45001, GLP and I S 0 9000 may be the responsibility of one organization or several, depending on the country. Consequently, some wish to see these three standards amalgamated into a single standard. The strength of particular lobbies, such as industry, may influence the way the schemes are run. These issues are yet to be resolved. The reagents section is a good example of an area fairly specific to chemical testing but not covered directly in EN 45001. EN 45001 makes no specific reference to ‘reagents’ and yet it is easy to appreciate the importance of reagents to a chemical testing laboratory’s quality system. Reagents are very broadly part of the equipment used in performing chemical tests; the requirements for equipment must also be applied to, and interpreted for, reagents.Thus, the Guide provides practical advice on requirements such as choice, grade, use, methods of preparation, storage, labelling and disposal. Up to now, little guidance has been available in Europe on the validation of the use of computers for chemical testing, but it is an area of increasing importance. The Guide adopts a pragmatic approach, suggesting that within certain guidelines it is possible to assume that a computer is working properly if it produces expected answers when fed with known data. Hence, the validation of computers can be achieved as part of the whole measurement process using standards, reference mater- ials or quality control samples. EN 45001 makes only the barest mention of the important subject of audit and review.The Guide complements WELAC’s recently published document with specific guidanceANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 253 on the areas which should be examined during audit of a chemical testing laboratory. The sampling section gives brief general guidance on best practice in sampling, sample handling and sample preparation, and stresses the importance of appropriate sampling in the over-all context of analytical problem solving. It provides suitable pointers for more detailed advice, particularly in specific sectors of measurement, and stresses the importance of correct nomenclature in the sampling process. In writing the section on measurement uncertainty the aim was to produce statistically correct advice which was mean- ingful to the non-expert and yet sufficiently detailed to be useful.Production of this section was beset by problems due to inconsistent IS0 definitions and variations in what experts considered to be best practice. The philosophy in this area is still subject to debate and future developments may require this section to be updated. The section on validation gives advice on an important subject for which there is little other advice available, particularly for laboratories working in isolation. The import- ant elements of method validation are discussed with indica- tions of where more detail can be found. The emphasis these days is not to accredit the use of standard methods in preference to non-standard methods, but rather to accredit the use of properly validated methods in preference to those which are poorly validated. Publication and Use Complete endorsement of he Guide by EURACHEM and WELAC is expected soon after which it will probably be published as a working draft, available through EURACHEM, WELAC and national accreditation bodies.The latter may well publish it with a national supplement listing particularly national requirements not covered in the main text. It is expected to be in circulation in this form for about two years, during which time users will be invited to comment on its content. At the end of that time, it will be revised taking into account comments and developments of the accreditation standards themselves. It is hoped that the final text can be formally adopted and published by the European Committee for Standardization (CEN) or ISO.Conclusion The joint working group has made rapid progress to produce a guide aimed at meeting the needs of chemical testing labora- tories interpreting EN 45001 or IS0 Guide 25. It will also assist laboratories working to GLP or I S 0 9000. The problems encountered during the drafting were typical of such an exercise but none proved insurmountable. Publication is expected soon and it is hoped that the guidance will prove useful. Selective Particle Counting by Means of Affinity Ligands Linked to Microscopically Visible Labels Derek Craston" and John Francis Laboratory of the Government Chemist, Queens Road, Teddington, Middlesex 7 W 7 7 OLY Harmesh Aojula, David Clarke and Robert Jeprast Centre for Applied Microbiology and Research, Porton, Salisbury, Wiltshire Microscopically sized particles supporting affinity ligands (antibodies) have found analytical use as reagents in selective separation,' cell labelling,ll2 and homogeneous immuno- assay^.^^^ For labelling purposes, particle-based systems have the particular advantage that selective detection is still possible when only small numbers (a few tens) of antigen-antibody conjugates are f ~ r m e d .~ If this property c8n be exploited to its extreme limit, it should prove possible to detect small molecules which possess as few as two antigenic sites. Liposomes, which are spherical particles consisting of an aqueous lumen enclosed by a lipid bilayer membrane ,6 are particularly attractive options for antibody labelling appli- cations, in that (a) they are easy to produce in relatively homogeneous populations, ( b ) they can be detected as single entities, or as small clusters, by the encapsulation of an enzyme or a fluorescent dye,7 and (c) phospholipids which contain chemical functionalities suitable for attachment onto proteins can be conveniently incorporated as membrane components.This paper describes the application of liposomes containing a fluorophore as antibody labels. The system is demonstrated by the selective flow cytometric' detection of Legionella pneumo- phila, while the possible extension of this work to allow the detection of single molecules is discussed. * To whom correspondence should be addressed. -f Present address: SmithKline Beecham Pharmaceuticals, Brock- ham Park, Bletchworth, Surrey.Experimental The liposomes used in the flow cytometry work contained 15 mmol 1-l of carboxyfluorescein (CF) and 1 mmol 1-l of ethylenediaminetetraacetic acid (EDTA) dissolved in 10 mmol 1-l tris(hydroxymethy1)methylamine (Tris) buffer (pH 7.4). These were prepared by a freeze-thaw sonication procedure,6 followed by passage ten times under pressures of up to 700 psif through two polycarbonate membranes (Nucleo- pore; pore diameter approximately 0.4 pm) in an extruder (Lipex, Vancouver, Canada). Non-encapsulated fluorophore was removed by gel filtration on a Sephadex G-25M column (Pharmacia; PD-10). The proportions of the various forms of phospholipids constituting the liposome membranes were egg lecithin : cholesterol : DL-a-phosphatidyl-L-serine : 1,2-distear- oyl-sn-glycero-3-phosphorylethanolamine (PE), 13 + 4 + 6 + 1 by mass.3-(2-Pyridoxyldithio)propanoic acid, N-hydroxysuc- cinimide ester (SPDP) was covalently attached to the PE6 before liposome preparation. Formaldehyde fixed (1% d v ) suspensions of L. Pneumo- phila sero group 1 (w74/81) and rabbit antiserum to L. Pneumophila lipopolysaccharide were kindly provided by Dr. R. Fitzgeorge, Centre for Applied Microbiology and Research (CAMR). Escherichia coli K12 was also fixed with buffered formaldehyde. Antiserum was purified on immobilized protein A at CAMR. Fragments of these antibodies were prepared by pepsin digestion [ImmunoPure F(ab')z Preparation Kit; t 1 psi = 6.89476 x lo3 Pa.254 r ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 Pierce] , with subsequent conversion to Fab' achieved by reduction with dithiothreitol.The Fab' fragments were cross- linked with the liposomes via SPDP-derivatized PE by mixing the two overnight under a nitrogen atmosphere.6 The molar ratio of Fab' fragments : liposomes was approximately 30000: 1. Macroscopic fluorescence measurements on liposome popu- lations were performed on a Baird Nova-3 spectrofluorimeter, with excitation at 494 nm and emission at 515 nm. The flow cytometer used was an Argus 100 (Skatron Ltd.), with a mercury arc lamp excitation beam of wavelength 470495 nm, and fluorescence detection at 520-550 nm. Results and Discussion Optimization of Liposome Populations Spectrofluorimetric investigations were performed on popula- tions of liposomes prepared from egg lecithin and cholesterol in equimolar proportions and constant concentrations, in the presence of a range of concentrations of CF.Fig. l(a) shows a plot of the relative fluorescence intensity against the concen- tration of entrapped dye, which, as previously reported2 gives a maximum fluorescence at 15 mmol 1-' CF. Concentrations of CF greater than 15 mmol 1-' lead to a self-quenching,2 with weak fluorescence signals observed when the concentration of the entrapped dye is greater than 50 mmol 1-l. Liposome stability was determined by monitoring the rate of CF leakage for a PO ulation prepared from a solution containing 50 mmol l-'CF. The initial fluorescence of the 1 0 - ~ I O - ~ lo-' 1 10 102 Carboxyfluorescein concentration/mmol I-' 15 I I I (b' 0 //a I 50 100 Tim e/d 150 Fig. 1 Fluorescence studies of liposomes containing carboxyfluor- escein (CF). ( a ) Relative fluorescence versus concentration of entrapped CF; (b) % leakage (as defined in text) versus time; A , 100% egg lecithin; B, 50% moYmol egg lecithin : cholesterol.Unless otherwise stated the liposomal membranes consisted of a 1 + 1 (by mass) mixture of egg lecithin and cholesterol. The liposomes were prepared by the hydration of thin films of the lipid mixture with CF solutions, followed by five cycles of freeze-thawing (using liquid nitrogen), and finally by transmembrane extrusion (Nuclepore; pore diameter 100 nm) liposome suspension was low , owing to self-quenching. How- ever, the fluorescence increased steadily over a period of 4 months as the CF leaching from the liposomes became diluted and thus unquenched.The proportion of fluorophore lost to the external medium was calculated as the ratio of the rise in fluorescence of the liposome solution to that which occurred when the membranes were deliberately disrupted (by the addition of Triton X-100). Fig. l(b) shows that liposomes composed of egg lecithin leach less than 5% of their entrapped fluorophore per month when stored refrigerated under air and in buffer having the same ionic strength as the liposomally encapsulated solution. Stability is further improved by the inclusion of cholesterol in the liposome membrane. The data in Fig. l(b) were obtained with an external medium of pH 7.4; when stability studies were performed at pH 6.4, the leakage rates were considerably higher.This was attributable to the higher solubility in lipids of CF with undissociated carboxylic acid groups at pH 6.4, whilst lipid bilayer properties might also be pH-dependent . Selective Detection of L. pneumophila Light scattering signals [Fig. 2(a)] are observed when L. pneumophilu cells flow through the beam of a flow cytometer;' however, under normal circumstances these signals do not correlate with any rise in the detected fluorescence above background. Thus, since other bacteria also scatter light [as illustrated in Fig. 2(a) with E. coli], it is not possible to identify the presence of L. pneumophila definitively in complex solutions containing many bacterial forms by flow cytometr unless the cells are first selectively labelled with fluorophore.l Intensity Fig. 2 Flow cytometric data for a sample containing liposomes with surface-bound Fab' fragments of anti-L. pneumophilu antibodies, L. pneumophilu, and E. Coli: ( a ) forward angle light scattering signal; (b) forward angle light scattering coincident with fluorescence. The x-axis gives the relative intensity of the light scattering signal, with the y-axis indicating the numbers of particles producing that response. Lipo- somes were manufactured in CF (15 mmol 1-'). The gain voltage on the forward angle light scattering photomultiplier detector was 550 V. Specimens were introduced into the laminar stream at a rate of 0.5 p1 min-'ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 255 Liposomes containing CF are weaker light scatterers than bacteria, but they do produce fluorescent signals which are discernible above background in a flow cytometer.When a population of these lipsomes carrying Fab’ fragments specific for L. pneumophila was mixed with a solution containing L. pneumophila, conjugates were formed which could be selec- tively measured by configuring the flow cytometer to detect only particles which produced simultaneous response at both the fluorescent and light scattering detectors [Fig. 2(b)]. Thus, liposome conjugation provides a possible route to the selective measurement of bacterial cells, in an approach which is analogous to that of conventional fluorescein isothiocyanate (FITC) labelling,’ but has the potential advantage that it may be applied where there are very few antigenic sites present on the surface of the bacteria.6 In order to measure L.pneumo- phila quantitatively by liposome labelling it is essential that all the bacteria in the sample are present as conjugates, and this inevitably requires that the liposomes are present in excess. In our experience, the required ratio varies from a few tens to a few hundreds between different batches of liposomes, and therefore some care is required in selecting reagent strengths appropriate for a Legionella assay. Nevertheless, we have quantitatively measured L. pneumophila at concentrations as low as lo3 ml-’ in specimens which contains a 10000-fold excess of E. coli. Detection of Single Molecules We feel that it should be possible to detect and count individual molecules by flow cytometry , using microscopically visible fluorescent labels coupled to antibodies.The principle of such a measurement is that the antigenic sites on the molecule act as a ‘glue’ for the formation of small clusters of labels, which can be distinguished by the magnitude of their fluorescence signal. A necessary prerequisite for this approach to be successful is that the population of the labels are monodisperse with respect to the level of their fluorescence under illumination. Unfortu- nately, however, the liposome populations described above did not fulfil this criterion, as indicated by Fig. 3, which shows that a wide spectrum of fluorescence intensities was observed in flow cytometric investigations of liposome samples.The reason for this broad range of fluorescence signals is unclear. However, since we have found by electron and fluorescence microscopy and by photon correlation spectroscopy that liposome populations can be prepared with a relatively monodisperse size distribution, it seems likely that the CF is unevenly distributed between liposomes. Intensity Fig. 3 Flow cytometric data for a sample of liposomes which contained 15 mmol 1-1 CF. The x-axis gives the relative intensity of the fluorescence signal, with the y-axis indicating the numbers of particles producing that response. The gain voltage on the fluorescence photomultiplier detector was 650 V. Specimen flow rates and sheath fluid were as in Fig. 2 Conclusions Liposomes containing high concentrations of fluorescent dyes, after they are coupled with an appropriate antibody, can serve as alternative labels for the selective measurement of bacteria by flow cytometry.They are relatively stable with respect to long-term storage, and their high fluorescence makes them suitable for measuring cells which possess relatively few antigenic sites. The application of such labels to the detection of molecules might yet be possible. However, this will require some improvements to the existing preparatory methods to provide greater uniformity in the population. References 1 Ugelstad, J., Adv. Urg. Coat. Sci. Technol. Ser., 1991, 13,507. 2 Truneh, A., Machy, P., and Horan, P. K., J. Immunol. Methods, 1987, 100, 59. 3 Monroe, D., J. Liposome Res., 1989, 1, 339. 4 Clarke, D. J., and Aojula, H.A., Br. Pat. 5 Truneh, A., and Machy, P., Cytometry, 1987, 8, 562. 6 New, R. C. C., Liposomes: A Practical Approach, Oxford University Press, Oxford, 1990. 7 Childers, N. K., Michalek, S. M., Eldridge, J. H., Francis, R. D., Berry, A. K., and McGhee, J. R., J. Irnmunol. Methods, 1989, 119, 135. 8 Loken, M. R., and Stall, A. M., J. Immunol. Methods, 1982,50, R85. European Community Measurement and Testing Programme Ronald F. Walker Laboratory of the Government Chemist, Queens Road, Teddington, Middlesex TW17 OLY The scientific activities carried out by the European Commun- ity (EC) go back to the mid-1950s when nuclear energy research was first carried out under the EurAtom treaty. By the early 1980s, a much broader range of activities, collectively known as the EC R & D Framework Programme, had been laid down which attempts to provide even coverage of those scientific areas judged to be important to the Community as a whole.For example, energy, the environment, industry, food and agriculture and raw materials. Each Framework pro- gramme runs for five years and, in 1990, the European Council adopted the 3rd Framework Programme which runs until 1994. The main objectives of the 3rd Framework Programme are essentially to: (a) improve industrial competitiveness (the emphasis being on pre-competitive research rather than fundamental blue skies research or near to the market development); ( b ) direct the attitude of the industrial sector towards further pan-European initiatives; (c) ensure scientists and engineers are firmly aware of the Community dimension and spirit during their training; ( d ) increase economic and social cohesion within the Community whilst maintaining the quality of R & D projects; ( e ) take into account the protection of the environment and quality of life; (f> face the challenges relating to the completion of the Single Market particularly with respect to Directives and standards.In practical terms, EC R & D largely boils down to international collaborative projects jointly funded by the256 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 European Commission and by the participants - industry, academia, government departments and other organizations with a vested interest in R & D across the Community. The current programme has an over-all budget of 5.7 billion ECU, or about 24 billion, and contains 15 specific research and technical development activities (Table 1).Perhaps the most important activity for analytical chemists is the measurement and testing programme which is run by the Community Bureau of Reference (BCR). The Community Bureau of Reference The principal objective of the BCR programme is to improve the reliability of chemical analysis and physical metrology in order to achieve harmonization of measurement results across the Community. Much of the BCR work programme is therefore geared towards solving measurement problems which could lead to trade disputes or hinder the operation of the Single Market. Based on this premise, the type of projects which tend to be looked on favourably by the BCR include those where: the accuracy of measurements is not sufficient for the industrial requirements related to manufacture and quality control; measurement discrepancies between laboratories are greater than commercial specifications allow; measurement inaccur- acies may induce heavy losses in the trade of goods of high value or in large volumes (food, raw materials); inaccuracies related to pollution or health care may lead to wrong conclusions with considerable damage to industry, the public or the environment.In some instances, the strategic importance of measurements for European industry is also taken into account, for example, measurements in the field of microelectronics. In the context of these criteria, the BCR has for the past five years or so concentrated essentially on the following priority areas: agriculture and food analysis; environmental analysis; applied metrology measurements; testing and measurements of indus- trial products; biomedical analysis.There are essentially two ways in which the BCR programme is funded. Firstly, concerted action, which is used where the aim of the project is to co-ordinate the work of several laboratories in order to improve a measurement or test, for example, improvements in the accuracy of calibrations between national metrology laboratories. In this instance, the BCR will pay up to 100% of the costs of meetings, the scientific and administrative co-ordination of the project and the processing and publication of the data produced. The second, and most common, means of funding is known as shared-cost action which is used to fund collaborative studies between Community laboratories studying improvements in methodology or when certifying a reference material.In this instance, projects are funded on a 50 : 50 basis between BCR and the participants, although for certain non-industrial Table 1 EC R & D Framework Programme 1 Information technologies 2 Communications technologies 3 Development of telematics systems 4 Industrial and materials technologies 5 Measurement and testing 6 Environment 7 Marine science and technology 8 Biotechnology 9 Agricultural and agro-industrial research 10 Biomedical and health research 11 Life sciences and technologies for developing countries 12 Non-nuclear energies 13 Nuclear fission safety 14 Controlled thermonuclear fusion 15 Human capital and mobility participants, for example universities, it is possible to secure 100% funding.The usual approach adopted by the BCR is to invite laboratories from as many Member States as possible, with perhaps an upper limit of 20-30, to carry out a number of collaborative studies such that any measurement discrepancies are reduced to what is considered to be an acceptable level of uncertainty. As regards chemical measurements, the labora- tories will then often participate in a certification exercise whereby they rigorously characterize a candidate reference material. In this way, participating laboratories gain consider- able expertise which is then disseminated to other European laboratories who use the certified reference materials (CRMs) produced.It should be noted that the certification exercise should always be undertaken by a number of laboratories using different independent methods of analysis. In this way, any measurement bias is reduced to a minimum. Of course, this approach to certification may occasionally lead to the elimina- tion of the data from one or more of the methods being used if their inaccuracies are shown to be unacceptable and, at the limit, data from only one method used by relatively few laboratories may be all that is suitable for certification. The New (1992-94) BCR Programme Since its formation in 1973, the BCR programme has carried out over 600 measurement projects and, as a result, has produced over 400 reference materials.Recently, the BCR has had a major call for proposals in advance of carrying out its 1992-94 work programme. Essen- tially, the new programme breaks down into four main areas: Area 1, Support to Directives and Regulations (12 MECU); Area 2, Support to Standards (11.5 MECU); Area 3, Common Means of Calibration (12 MECU); Area 4, New Methods of Measurement (12 MECU). The distribution of the budget is shown in parentheses (1 MECU = 1 million ECU) and includes the cost of personnel and management for running the programme. In other words, a total of 47 MECU or less than 1% of the total Framework Programme budget. Area 1 will cover the work required to improve those methods of measurement and testing which are necessary for the implementation of Directives. This is vital for the Single Market because although European Directives provide the legal basis for the harmonization of national regulations, their implementation sometimes requires measurements and ana- lyses which are beyond the capabilities of many laboratories at present.Typical topics of interest will include the analysis of food products (e.g., pesticides in cereals, vegetables and fruit; nitrate in baby food, cheese and milk powder; and heavy metals in olive oil), agricultural products (e.g., drug residues, vitamins and toxic impurities in animal feeds; and cow’s milk in goat’s cheese), environmental analyses [e.g., pollution in the sea; leachable heavy metals from building and demolition waste; and polycyclic aromatic hydrocarbons (PAHs) , poly- chlorinated biphenyls (PCBs) , and dioxins in contaminated soils], air in the workplace ( e . g . , aldehydes, isocyanates, nitrosamines, asbestos and heavy metals), noise measurements (e.g., the evaluation of hearing protectors and the transmission of noise in buildings), biomedical analysis (e.g., toxic sub- stances in blood and urine) and quality assurance (e.g., supporting collaboration between the national quality assur- ance systems). Area 2 will focus on the development of standardized methods for the testing of industrial products as specified by European standards. Topics so far identified include the development of test methods for determining the fire behav- iour of upholstered furniture, constructive products, hearing protectors against impulse noise and cold spots in microwave ovens. Area 3 has been the area where the BCR has concentratedANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 257 most of its effort over the years, essentially producing CRMs and improving calibrations between the National Physical Laboratories. In the new programme this area will produce CRMs mainly for Area 1 and possibly Area 2, for example, certified nitrate content of baby food. Finally, Area 4 will concentrate on developing instrumental measurement methods either as a result of work carried out under Areas 1-3, or for the on-line or in situ determination of physical, chemical or biological parameters where either no suitable direct methods currently exist or where there are severe operational conditions. Summary In conclusion, European scientific collaboration in areas such as the Framework Programme is becoming increasingly important. For example, depending on the type of projects undertaken, it is often possible to realize substantial benefits such as: ( a ) sharing the cost and risks of R & D by making use of complementary skills and common facilities, thus enabling participation in projects where the scale of investment would otherwise be beyond an individual organization; ( b ) gaining a commercial advantage from tapping into the scientific exper- tise of European laboratories/organizations; ( c ) achieving a more significant role in the development of international scientific standards; (d) establishing business contacts with overseas counterparts in order to be well placed to take advantage of the Single Market. It should be noted, however, that EC programmes are often oversubscribed and the preparation of a suitable proposal can be extremely time consuming and therefore expensive. But, that said, the benefits of participation are often great and in most instances easily outweigh the costs of participation.

ISSN:0144-557X    

DOI:10.1039/AP9933000244

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 257 Proficiency Testing: How Laboratories Measure up to the Corn petit ion Ronald F. Walker laboratory of the Government Chemist, Queens Road, Teddington, Middlesex TWI I Ql Y As mankind moves towards the twenty-first century, the importance placed on obtaining reliable chemical measure- ments is constantly increasing. It is now clearly recognized that chemical analysis represents a vital contribution towards securing our commercial and industrial prosperity, towards ensuring the integrity of forensic evidence in justly enforcing the law and in monitoring both our own health and that of our planet. In such a climate, it is clearly not sufficient to be able to achieve the correct answer, it is necessary for a laboratory to demonstrate the accuracy or comparability of its data by some form of external assessment of the quality of its results.Such a demonstration of a laboratory’s competence is known as external quality assessment. There are two main ways by which a laboratory can do this, which are essentially complementary in nature. The first is by physical inspection of the laboratory to ensure that its in-house quality assurance procedures comply with well established, recognized standards; in other words, third-party assessment or accreditation. The second, known as proficiency testing, is by the assessment of its performance in interlabora- tory comparisons using centrally distributed samples. Essentially a proficiency testing (PT) scheme checks the competence of a group of participating laboratories by a statistical evaluation of the data they obtain on analysing distributed materials.Each laboratory is then provided with a numerical indicator of its performance, together with infor- mation on the performance of the group as a whole, enabling its proficiency relative to the group to be compared and evaluated. Participation in a proficiency testing scheme also reinforces an interest in quality control and provides the basis for any corrective action in those laboratories whose data do not meet the required level of acceptability. Over the years proficiency testing schemes have been introduced under a number of different circumstances, some are ‘open’ to any laboratory, usually for an annual fee, others are ‘closed’ or by invitation only.Regardless of the circum- stances, there are basically two main types of scheme. Firstly, there are those set up to measure the competence of a group of laboratories to undertake a very specific analysis, e.g., lead in blood or the number of asbestos fibres on membrane filters. Secondly, there are those where there is a need to judge the competence of a laboratory across a certain field or type of analysis. Because of the wide range of possible analyte/matrix combinations it is not practicable to apply comprehensive testing. Instead, a representative cross-section of analyses is chosen [e.g., trace metal analysis by atomic absorption spectrometry or the detection of drugs of abuse by high- performance liquid chromatography (HPLC)] . Each of these two main types of proficiency testing schemes can be further sub-divided into three categories.(a) Where the sample to be tested is circulated successively from one laboratory to the next. In this instance the sample may be returned to a central laboratory sometimes before being passed on to the next testing laboratory in order to determine whether any changes to the sample have taken place. ( b ) Where randomly selected sub-samples from a bulk homogeneous supply of material are distributed simultaneously to participat- ing laboratories, by far the most common type of proficiency testing scheme. (c) Where samples of a product or a material are divided into several parts with each participating laboratory testing one part of each sample. This is frequently referred to as ‘split sample’ testing.A number of PT schemes are currently operating in the UK, over 20 main schemes were identified in a survey carried out in 1990 (Table 1). For example, in the area of chemical measurements there are important schemes concerned with Table 1 Typical UK proficiency testing schemes Name of scheme Nature of tests Food Analysis Performance Assessment Scheme (FAPAS) Workplace Analysis Scheme for Proficiency (WASP) AQUACHECK UK External Quality Assessment Schemes (UKEQAS) Regular Interlaboratory Counting Proficiency Test Scheme for Alcohol Exchange (RICE) (PROTAS) Proximates, trace contaminants in food Metals, organic vapours in air Most aspects of water quality Most aspects of clinical and microbiological testing Asbestos fibre counting Alcoholic content of beverages258 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 food analysis (FAPAS), hazardous substances in the work- place atmosphere (WASP), water quality (AQUACHECK), clinical chemistry (UKEQAS) , asbestos fibre counting (RICE) and the alcoholic strength of beverages (PROTAS).These schemes have developed independently over the past 20 years; the UKEQAS scheme started in 1969, the PROTAS scheme as recently as 1991. Aims of Proficiency Testing Although the over-all aim of proficiency testing is to encourage good performance, within the scope of this general aim a successful scheme must provide certain types of information for both the participants and the organizers. Firstly, it must enable a laboratory to compare its performance at a particular time against an external standard of performance.It must also enable a laboratory to compare its performance at a particular time with its performance in the past. It must enable a laboratory to compare its performance with that of other laboratories at a particular time. It must enable the organizers to identify the participants whose performance is unsatisfac- tory, and finally it must enable the organizers to see whether there is a general improvement in performance with time. Organization of Proficiency Testing Schemes In general proficiency testing schemes should be organized in a sequence of clearly defined stages (Fig. 1). The organizing body should lay down a protocol for the conduct of the tests and the interpretation of the data which is then circulated to the participants so that they understand exactly how the scheme is run and how their results are assessed.Materials should be chosen such that they are, as far as possible, representative of the type of material that is normally analysed in terms of the matrix and the concentration range of the analyte. Materials also need to be tested for homogeneity before distribution since the effective interpretation of all the test data for the scheme is clearly based on this assumption; each separate batch of test material must therefore be checked for this, an exercise which is generally carried out by a single expert laboratory. In a homogeneity study it is important to use a method which achieves very good precision. The accuracy of the technique is less important. It is not necessary to know whether the true Identify requirements of the proficiency test I 1 Prepare and characterize bulk material I Distribute samples to test laboratories Test laboratories analyse samples and report results I r I I Evaluation of results and performance assessment 1 I i I Test laboratories notified of their performance score I Next round commences Fig. 1 Typical framework for a proficiency test value of the determinand is 3 or 6 ppm as long as the technique used produces reproducible results.Of course, non-homogen- eity is always possible even after the material has been homogenized and distributed. Therefore, if only a sub-sample of the material supplied by the organizers is to be analysed, it is clearly sensible to rehomogenize it by mixing before use.There is no experimentally established optimum frequency for the distribution of samples. The consensus view is that the minimum frequency should be about four rounds per year. Tests that are less frequent than this would probably be ineffective in reinforcing the perceived need for maintaining quality standards or for following up marginally poor per- formance. A frequency of one round per month for any particular type of analysis is the maximum that is likely to be effective. Postal circulation of samples and results would usually impose an absolute minimum of two weeks for a round to be completed and it is possible that over-frequent rounds might have the effect of discouraging some laboratories from conducting their own routine quality control. Once the samples have been analysed and the results reported, the first stage in producing a score from each individual result x is to obtain an estimate of its bias (x - X), where X is the true concentration or amount of analyte.Clearly, the quality of any proficiency test depends on using as reliable a value for X as possible and there are essentially three methods available to obtain a working estimate of the true value. (a) The addition of a known amount or concentration of analyte to a base material containing none. This method is completely satisfactory in many cases, especially when it is the total amount of the analyte rather than the concentration that is subject to testing. (b) The use of a consensus value produced by a group of expert or referee laboratories using best possible methods.This is probably the closest approach to obtaining true values for the test materials, but it may well be expensive to do so. (c) The use of a consensus value, produced in each round of the proficiency test, and based on the results obtained by the participants. The consensus is usually estimated as the mean of the test results after any outliers have been rejected. This approach is clearly the cheapest way of obtaining an estimate of the true value, but problems might arise if there is not a real consensus among the participants or the consensus is biased because of the general use of faulty methodology, neither of which is rare, for example, in the determination of trace constituents in natural matrices. The choice between a consensus either from expert labora- tories or from the participants depends on whether the aim of the proficiency test is to encourage the production of true results or merely to obtain agreement among the participants.In spite of the extra cost, it clearly makes sense to obtain the best estimate of the true value, in other words using method (b) if possible, because it is only by this approach that the scheme can hope to form part of the wider chemical measurement system within which comparable, traceable measurements are essential. For example, one can imagine a scheme using method (c) as a closed system within which there is an over-all systematic bias in the results for which, over time, the participants have learnt to compensate in order to achieve good performance indices, and as a result build in the same bias for all their routine samples.Clearly, it would be impossible for a laboratory within the scheme to compare its results with a laboratory outside the scheme and hope to achieve comparability. If, on the other hand, the organizers of the scheme are attempting to obtain the best estimate of the true value, then it follows that compara- bility of measurement with outside laboratories should follow. Having obtained a working estimate of the true value of the analyte being measured, most PT schemes compare the estimate of the bias with a standard error. One fairly common way to do this is by means of a z score:ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 259 where o is a standard deviation which is chosen either as an estimate of the actual variation encountered in a particular round of the scheme or a target value representing the maximum allowed variation consistent with achieving valid data.The main assumption in using the z score is that z will be approximately normally distributed with a mean of zero and a unit standard deviation, and on this basis analytical results can be described as ‘well behaved’. A classification based on z scores can be made: z < 2 Satisfactory ( 2 ) 2 < z < 3 Questionable (3) 2 > 2 Unsatisfactory (4) An alternative scoring, known as Q scoring, is also sometimes used: Q = ( X - X ) / X but is based not on the standardized value, but on the relative bias. This type of score tends to be used when the participants in a scheme may have wide standards of performance and there is no basis for a common value of o.At the end of each round, participants need to be informed of the outcome as soon as possible after the closing date for the reporting of results in order to respond to any problems. Usually the results are reported in the form of a computer printout which includes detailed information on the partici- pants’ performance and ranking, the number of outliers, the over-all distribution of results, and so on. In the early rounds of most schemes there is usually an over- all significant improvement in performance, although for a number of laboratories there can be a lack of consistency; achieving a good performance in one round, but not being able to sustain it over a long period, suggesting that they do not have an adequate quality system in place.Current Developments In the current climate of widespread interest in the quality of data it seems likely that more and more proficiency testing schemes will be introduced, particularly since the accreditation authorities would like to include schemes as part of the requirements for achieving accreditation. However, while proficiency testing schemes have been useful in helping to improve the quality of analytical data, they could also act as a barrier to the mutual recognition of such data unless there is some degree of uniformity in their procedures. For example, one of the defects of existing schemes as a whole is the diversity of performance indices which makes it difficult both to appreciate the meaning of an index in an unfamiliar scheme and to compare the relative performance or effectiveness of different schemes. Chemac, a UK committee concerned with improving the validity of analytical data, has therefore established a Proficiency Testing Working Group to consider this issue.A major achievement of the Group has been to produce a draft harmonized protocol containing general guidelines on the organization of schemes. Following consider- ation by a joint ISO/IUPAC/AOAC committee, it is hoped that the protocol will be published as an IS0 document or Guide in the not too distant future. Although the harmonized protocol is an extremely useful document, it describes only the operation of PT schemes and does not consider wider issues such as the relationship of PT to other aspects of AQA or the economic aspects of PT schemes, i.e., the relationship of costs to resultant benefits.The UK National Measurement System Policy Unit (NMSPU) has therefore commissioned the Laboratory of the Government Chemist (LGC) to undertake a more detailed and objective study of both the theoretical basis and practical aspects of PT schemes, not only from the viewpoint of the organizers, but also from those of the participants and users. Amongst other things, this study will consider: ( a ) efficacy of schemes; what features.of a given scheme help a laboratory to identify and rectify causes of poor analytical performance (e. g. , the effect of sample distribution frequency and the methods used for performance scoring); (b) structure and operation of PT schemes (e.g., the role and function of the steering committee and the type of feedback to participants); ( c ) statistical procedures; a comprehensive evaluation of the fundamental requirements of a scoring system will be undertaken, including the possible use of robust statistics, the need for a minimum number of participants and the treatment of outliers; ( d ) a detailed protocol will be produced which will include a distillation of the above studies and which will be validated and actively promoted through both national and international bodies such as Chemac, Eurachem and ISO. In summary, then, it can be seen that proficiency testing makes an important contribution to the quality of analytical data and should therefore be implemented whenever it is appropriate and technically feasible to do so.Participating in a PT scheme and achieving good performance scores gives a laboratory confidence that it is well on the way to measuring up to the competition. Heavy Metals in Deep-sea Holothurians Heather M. Moore and David Roberts School of Biology and Biochemistry, The Queen's University of Belfast, Belfast BT9 5AG Michael Harriott and Duncan Thorburn Burns School of Chemistry, The Queen's University of Belfast, Belfast BT9 5AG The production, sinking and decomposition of organic par- ticles are important factors controlling the distribution of trace elements within the ocean. During sinking these particles can absorb and scavenge trace elements from the dissolved phase and thereby remove them to depth.',* Particles ultimately reach the sea floor where the adsorbed and incorporated elements may be released or transformed by chemical and biological proce~ses.~ Sedimentation processes provide a major pathway by which metals are transported to oceanic sediment^.^ The flux of organic matter to the sea floor is no longer viewed as a continuous slow detrital rain over all the ocean floor, since certain areas of ocean floor have been shown to receive large periodic inputs of organic In addition to the accumulation of heavy elements by the transport from surface layers to depths, there is a possible secondary source due to increasing pressure to use deep ocean sites for dumping sewage sludges, which are known to contain elevated levels of heavy It is therefore important to establish baseline data on metal levels in deep-sea organisms prior to the large-scale introduction of anthropogenic waste.The study site was selected because it is an area known to have260 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 a large and rapid detrital flux which may accentuate metal transfer to deep-sea sediments. Holothurians (sea cucumbers) were selected for this study as they comprise the dominant group of large organisms on the deep-sea floor which ingest sediment. Experimental Holothurians were collected either by trawling or dredging during the following cruises: RRS Discovery 185 (1989); 194 (1990); and RV Challenger 79 (1991). Samples of different species of holothurian were taken from a selection of abyssal (area with true deep-sea fauna) and bathyal (slope area, with a transitional fauna comprising species from both shallow and deep water) sites in the north east Atlantic (Fig.1). In addition, samples of a shallow-water holothurian were collected from the Irish coast off Donegal for comparison. On board ship small samples (2 X 2 cm) of body-wall tissue were dissected from both the dorsal and ventral surface of each animal, frozen and stored at -70 or -20 "C in plastic bags. On return to the laboratory, tissue was dried at 80°C for 2 d to constant mass and then ground to a fine powder. Acid digestions were carried out on known amounts (0.02-0.5 g) of powdered tissue in a PTFE 'bomb' with 2 ml of a HN03-HC1 (1 + 1) acid mixture (AnalaR grades) for each digestion." After the bulk of the reaction had subsided the bombs were sealed and placed in an oven at 80 "C for 1 h.After cooling, the digests were diluted with doubly distilled water to 20 ml. Digests were initially screened using inductively coupled plasma atomic emission spectrometry (ICP-AES) (Perkin- Elmer Plasma 40 instrument). A number of samples were analysed by both ICP-AES and by atomic absorption spec- trometry (AAS) for comparison. Results The initial multi-element scan of digests of tissue from deep-sea and shallow-water holothurians using ICP-AES revealed the presence of seven metals (Al, Co, Cu, Fe, Mn, Zn and V) in the tissue digests which were not present in control samples of sea-water. Subsequent work concentrated on the determi- nation of the seven metals in the body-wall samples from RRS Discovery 1851194 cruises.Analyses of 26 split samples by ICP-AES and AAS revealed no significant differences using the paired t-test at p = 0.05 (Table 1). Owing to more ready availability of equipment, AAS was used for subsequent analyses of the samples from the Challenger 79 cruise. Interference problems were encountered whilst attempting to measure Co and V in the RRS Discovery samples and determination of these metals was not attempted 60"N ,""" 20" w L 10'W 0" w --. 60" N 0 5 50" N 40" N 30"N -1 3" -' ) t 3 0 " N 00 1 Fig. 1 Map of sampling sites: coastal Donegal (1); RRS Discovery 185 (1989) (2); 194 (1990) (3); RV Challenger 79 (1991) abyssal site (4); and slope site (5) Table 1 Concentrations of Mn, A1 and Cu in holothurian body-wall samples collected during Discovery cruises 185 (1989) and 194 (1990) determined by ICP-AES and AAS Mean (PPm) Paired t-test Significance Metal ICP-AES AAS t-value at p C0.05 Mn 25.6 29.1 0.285 None A1 446.5 415.9 0.279 None c u 73.3 108.2 0.791 None for the Challenger 79 samples; however, cadmium concen- trations were determined in this later series.Dorsal and ventral body-wall sections were analysed from the Challenger 79 samples. No significant differences were observed by analysis of variance using the UNISTAT program between these two regions (Table 2); therefore, mean body- wall values were determined in further analysis. Inter-species Differences The levels of Cd, Mn, Cu, Zn, A1 and Fe in the body walls of five species (Oneirophanta mutabilis, Psychropotes longicauda, Pseudosticopus villosus, Paroriza spp. , Benthogone rosea) of deep-sea holothurian are shown graphically in Fig.2. The lowest levels noted were for Cd (range 1.82-5.07 yg 8-l dry mass) and highest levels for Fe (range 75.38-243.04 yg 8-l dry mass). Significant inter-species differences in body-wall metal concentrations were revealed for Cd, Mn, Cu, Zn and A1 by analysis of variance and for Fe, Cd and A1 when the slope (bathyal) species B. rosea was excluded to remove major 'site' differences. When compared with the four abyssal species, B. rosea showed lower levels of Mn, Cu, Zn, A1 and Fe, but higher concentrations of Cd. The elevated levels of Cd may reflect the greater proximity to anthropogenic metal sources of the slope- dwelling species. The apparent absence of Cu in B.rosea is an anomaly requiring further investigation. Site and Season Differences Oneirophanta mutabilis was the most common species found in the various hauls. The abundance of this species provided sufficient material to allow more detailed investigations, the data from which are summarized in Tables 3 and 4. Significant differences were noted in body-wall concen- trations of Zn, Cu, A1 and Mn in 0. mutabilis samples obtained at different times of the year (Table 3) and for Cd, A1 and Mn from different sampling sites at the same times of the year (Table 4). Discussion Deep-sea organisms can acquire metals from a variety of sources including rivers, the atmosphere and hydrothermal input^.^ Apart from incidental waste disposal on major trade routes there has to date been no large-scale dumping of wastes Table 2 Concentrations of six metals in dorsal and ventral body-wall sections of five species of deep-sea holothurians from Challenger 79 samples Metal Zn Fe Cd c u A1 Mn Dorsal Ventral 34.9 30.7 200.3 193.04 3.54 3.2 116.4 103.2 150.6 123.0 16.5 22.6 Significance at p <0.05 None None None None None NoneANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 261 250 200 - 150 0.Q1 - - 4 100 50 0 Oneirophanta P. villosus Ps ychropotes Paroriza B. rosea ( n = 23) ( n = 6 ) ( n = 3 ) ( n = 3 ) ( n = 8 ) Species Fig. 2 Metal levels in the body walls of five species of deep-sea holothurian collected during Challenger cruise 79 (1991) Table 3 Comparison of mean concentrations of selected metals in the body wall of Oneirophanta mutabilis showing possible inter-seasonal variation Mean concentration (ppm) P Discovery 185 Challenger 79 (Anova) Aug./Sept . 1989 MayIJune 1991 significance Metal ( n = 20) ( n = 23) at p <0.05 Zn 31.8 k 2.17 24 k 2.28 <0.05 Fe 333.6 k 43.9 249.0 f 59.8 None c u 118.5 k 14.1 208.5 f 17.3 <0.001 A1 304.5 k 28.6 149.2 k 24.4 <0.001 Mn 19.9 k 1.73 9.01 f 0.93 <0.0001 Table 4 Mean concentrations of selected metals in the body wall of Oneirophanta mutabilis showing inter-station variation Mean concentration (ppm) Station 52701#6 Metal ( n = 7) Zn 21.8 k 2.55 Fe 401.8 k 166 Cd 4.32 f 0.29 c u 208.3 f 29.2 A1 247.5 k 33.7 Mn 12.45 k 1.58 Station 52701# 17 ( n = 16) 25.3 k 3.35 172.6 f 16.8 2.95 f 0.19 208.5 k 22.7 90.1 k 13.5 6.94 k 0.44 P Anova significance at p <0.05 None None <0.001 None <0.001 <0.0001 containing metals in the deep sea.The Porcupine Abyssal Plain has already been used to dispose of low-level radioactive waste and there are proposals to use the same site for disposal of sewage sludge.**” The data obtained in the present work provide a baseline for future studies using holothurians or some other bioindicator organisms which ingest recently sedimented material on the deep-sea floor. Data obtained for deep-sea holothurians compared with those reported by Bryan” for shallow-water echinoderms show that (Table 5 ) mean concentrations of Zn, Fe, A1 and Mn are lower, and Cd and Cu levels are higher in the deep-sea holothurians compared with shallow-water echinoderms. Cad- mium and Cu are two pollutants frequently associated with anthropogenic pollution and their presence in parts of the environment is increasing.The high levels of these elements may reinforce the idea of immobilization of toxic metals as Table 5 Mean concentrations of metals (pg 8-l dry mass) for deep-sea holothurians (DSH) and shallow-water echinoderms (SWE) Metal DSH SWE Zn 31.97 100 Fe 197.5 250 Cd 3.34 2 c u 116.8 10 A1 135.7 160 Mn 18.6 40 inclusions in body tissues when species respond to stress by sequestering potentially toxic metals in forms that are probably biologically inactive. l3 Previous X-ray microanalysis on deep-sea holothurian tentacular tissue showed dense ovoid bodies in the tissue as the localization site of a number of elements.14 Future work is aimed at detailed analysis of inclusions in body-wall tissue of scavenging bioindicating species.In addition, analysis of deep- sea sediment and holothurian gut contents is planned to provide information on the biological processes having an impact at the sediment-water interface on the fate of trace metals in the deep sea. This work was funded by EC contract No. MAST-CT90-0037 in collaboration with the Institute of Oceanographic Sciences (NERC-UK) DEEPSEAS programme. Grateful thanks is expressed to the masters and crews of RRS Discovery and RV Challenger (NERC) for the provision of support facilities which enabled the samples to be collected. We also acknow- ledge the assistance received from Queen’s University Questor Centre and the Analytical Services and Environmental project unit.References 1 Hart, B. T., Hydrobiologia, 1982, 91, 299. 2 Levinton, J. S . , in Lecture Notes on Coastal and Estuarine Studies I , eds. Lopez, G. R., Taghon, G., and Levinton, J., Springer-Verlag, Berlin, 1990. 3 Heavy Metals in the Aquatic Environment, ed. Henkel, P. A., Pergamon Press, Oxford, 1975. 4 Fowler, S. W., andKnauer, G. A., Prog. Oceanogr., 1986,16, 147. 5 Billett, D. S. M., Lampitt, R. S, Rice, A. L., and Mantoura, R. F. C., Nature (London), 1983, 302, 520.262 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 6 Lampitt, R. S . , Deep Sea Res., 1985, 32, 885. 7 Rice, A. L., Billett, D. S. M., Fry, J., John, A. W. G., Lampitt, R. S., Mantoura, R. F. C . , and Morris, R. J., Proc. R . SOC. Edinburgh, Sect. B , 1986, 88, 265. 8 Davies, G., Mar. Pollut.Bull., 1988, 19, 2. 9 van Dover, C. L., Grassle, J. F., Fry, B . , Garritt, R. H., and Starczak, V. R., Nature (London), 1992, 360, 153. 10 Donaghey, C. A., Ph.D. Thesis, Queen’s University Belfast, 1989. 11 Davies, G., Mar. Pollut. Bull., 1987, 18, 59. 12 Bryan, G. W., Mar. Ecol. 1976, 5 , 1239. 13 Roesijadi, G., Young, J. S., Crecelius, E. A., and Thomas, L. E., Biol. SOC. Wash. Bull., 1985, 6, 311. 14 Hayes, G., and Roberts, D., unpublished work. Quantitative Analysis of the Pyrolysis Products of Perfluorocarbon Fluids by Gas Chromatography and Spectroscopic Techniques T. Karl P. O'Mahony, A. Peter Cox and David J. Roberts School of Chemistry, University of Bristol, Cantock's Close, Bristol BS8 ITS Chromatography and spectroscopic techniques have been combined to characterize the breakdown products of perflu- orocarbon fluids, 'Flutecs'. These fluids (Figs.1 and 2) are chemically inert, non-flammable and have excellent dielectric properties. They have numerous medical and industrial applications including electronics testing, vapour-phase solder- ing, fluid tagging, electrical switching, eye surgery, use as flame retardants, oxygen carriers and CFC replacements. They are designated 'non-toxic' fluids. However, under severe con- ditions they decompose, yielding smaller perfluorinated mol- ecules which can be extremely toxic. One such molecule is PPI Perfluorohexane C6F14 PP30 Perfluoropropane C3F8 (gas) perfluoroisobutylene (PFIB), Fig. 3, probably the most toxic of all perfluorinated molecules with a threshold limit value (ceiling limit) of 10 ppb.Perfluoroisobutylene is a colourless, odourless gas, 10 times more toxic than phosgene. PPlC Perfluoromethylcyclopentane CSFT2 PP2 Perf I uoromet h y lcyclo hexane F2 PP5 Perfluorodecalin CloFls PP3 Perf I uorodi met h y lcyclo hexane CSFl6 F2 F PP9 Perfluoromethyldecalin CllFz0 PP 1 1 Perf I u or0 p hena n t h ren e c14F24 PPlO Perfluorofluorene C13F22 PP50 Perfiuoropentane C5F12 Fig. 2 Ring Flutecs PP80 Perfluorooctane c8F18 -c Fig. 1 Linear Flutecs Fig. 3 Perfluoroisobutylene (PFIB)ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 263 Manometer avacuum I I , 7 I J Baratron gauge NaCl windows - 0.2 cm3 Flutec sample Fig. 4 Vacuum and pyrolysis apparatus t ar Ts 0 Q VJ 2 0 0 w 4ir PFIB 10 20 Retention time/min Experimental The fluids were thermally decomposed at 900 "C in the absence of air by leaking de-gassed perfluorocarbon vapour at reduced pressure ( ~ 0 .1 mbar)* through a hot quartz tube. The rate of flow was controlled via the vapour pressure of the liquid perfluorocarbon using appropriate temperature baths. A static vacuum vessel was used to collect a total product vapour sample. The vacuum and pyrolysis apparatus is displayed in Fig. 4. The pyrolysis products were analysed by gas chroma- tography and Fourier transform infrared spectroscopy (FTIR) . For the gas chromatography a packed column (5% Krytox 143AC on 60-80 mesh Graphpac GB) was used with electron- capture (GC-ECD) and mass spectrometer (GC-MS) detec- tors. After several trial s t ~ d i e s , ~ Krytox, a poly(perfluor0- alkyl) ether was found to be the best stationary phase for the separation of perfluorocarbons and is represented structurally below CF3 I F-[ --CF--CF20-] .--CF*CF3 For the infrared analyses, a Perkin-Elmer 1750 spectrometer (resolution 2 cm-') was used. A standard 10 cm gas cell was employed with known sample pressures (10-100 mbar). Characteristic peak - 1751 cm-' 4000 3000 2000 1000 Wavenum berkm- ' Fig. 7 FTIR spectrum for PFIB Fig.5 GC-ECD trace for PFIB 74 PPm / Concentration Fig. 6 GC-ECD calibration graph for PFIB. Limit of detection, 1.5 PPb -j; 80 c .- C 3 L- e 2 60 4- .- I 40 0 Q f 20 : 0 2 4 6 8 10 12 14 PressurelTorr 400 Fig. 8 R I R calibration graph for PFIB: r = 0.9991; slope = 6.65 k 0.24 Torr-'; intercept = 0.006 k 0.004 (1 Torr = 133.322 Pa) * 1 bar = lo5 Pa.264 -0.05 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 - d & & A L L Results A literature search revealed the most likely breakdown products.Where possible, samples of standard gases were obtained and characteristic peaks were assigned from chro- matograms and spectra. Calibration graphs were produced for 300 000 v) C 3 w .- 2 225000 2 e *-' .- I W 150000 Q v) 2 s 75000 L w 0 0 0 I I I I 1 I I I 4 8 12 16 20 24 28 32 36 Retention time/min Fig. 9 PP1 decomposition GC-MS trace C4F7+ 181 Parent 40 60 80 100 120 140 160 180 200 rnlz Fig. 10 Mass spectrum for PFIB each standard using gas pressure versus peak area. A PFIB GC-ECD chromatogram and corresponding calibration graph are shown in Figs. 5 and 6, respectively. In addition, Figs.7 and 8 display an FTIR spectrum and resulting calibration plot for PFIB , respectively. Chromatograms and infrared spectra of the pyrolysis products were compared with those of the standards in order to quantify the individual components. A PP1 decomposition GC-MS chromatogram is displayed in Fig. 9, with the mass fragmentation pattern for PFIB in Fig. 10. The FTIR spectrum for the above decomposition is shown in Fig. 11. An attempt was made to identify any remaining peaks from the literature. Both GC-ECD and FTIR were employed for the quantitative analyses. GC-MS was useful for qualitative analysis of larger breakdown products (C5-C6). Tetrafluoro- propyne, CF&%CF, was identified and analysed in detail via its microwave ~pectrum,~ Fig. 12. Discussion A total of 95% of the volatile decomposition products are quantitatively accounted for, Table 1.The remainder, apart from 1% undecomposed Flutec vapour, are given qualitatively 0.95 ; 0.75 w .- C 3 2 e .- 2 0.55 - W 0.35 3 $ a q 0.15 II"'" \I 4000 2942.15 1942.35 1519.21 1096.07 672.93 Wavenum berlcm-' Fig. 11 PP1 decomposition FTIR spectrum I J = 11-10 12-1 1 13-1 2 30.0 32.0 34.0 FrequencyIG Hz Fig. 12 Microwave spectrum of tetrafluoropropyne 36.0ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 265 Table 1 Quantitative results (YO) PP1 PPlC COF2 0.1 0.1 C2F4 15 22 CZF6 50 35 C3F6 11 1 5 2 C3F8 C4Fs Cycle C4F8 cis-2-ene - 10 C4F8 trans-2-ene 1 8 C4Fs is0 4 0.6 CF4 8 13 - - PP2 0.3 15 13 52 1 1 9 2 0.7 0.6 PP3 14 14 51 1 2 1 6 3 <o. 1 0.06 PP5 0.1 13 23 40 0.8 3 1 13 2 0.5 PP9 0.6 10 20 43 1 3 0.4 14 - 0.7 PPlO 0.02 17 14 58 0.7 1 0.2 4 0.4 - PPll 2.0 13 19 44 2 2 3 6 5 1 PP30 0.2 7 25 47 5 11 - - - 2 PP50 0.4 5 28 44 10 3 - - 1 4 PP80 0.02 15 13 46 6 3 7 3 2 - Table 2 Qualitative results PP1 PPlC PP2 co v v v COZ v v v v SiF4 v v CZ8 l-ene - - C4F6 butyne v v v C4F6 diene - v v C3F4 - v v C5F8 v v v C5Flo CyClO v v v C5FI0 3m ene* d v v v v - v - v FTIR - - GC-MS v - C5F12 C6FlO C6F12 - * 3m ene = perfluoro-3-methylbut-l-ene.PP5 v v v v v v v v v v v v v PP9 v v v v v v v v v v v v v PPlO v v v v v v v v v v v v - PPll v v v v v v v v v v v v - PP30 PP50 v v v v v v - v PP80 v v v v - - v v v v v v v ~~~~~~~~ ~ ~ Table 3 PFIB analysis (% of volatile products; mg g-' of flutec) Flutec PFIB (YO) PFIB/mg g-' PP1 3.83 26 PPlC 0.65 5 PP2 0.68 4 PP3 0.10 1 PP5 0.55 3 PP9 0.74 4 PPlO 0.39 2 PP11 1.29 7 PP30 1.81 11 PP50 4.18 28 PP80 2.02 13 in Table 2.A total of 23 compounds were identified, of which 10 were measured quantitatively. The major pyrolysis products were CF4, C2F4, C2F6 and C3F6. Considerable amounts (up to 4%) of PFIB were found. Pyrolysis of the straight-chain perfluorocarbons generated more PFIB than that of other fluids examined, see Table 3. Initial breakdown of the perfluorocarbon occurs at the carbon-carbon bond, resulting in the generation of CF2, CF3, C2F5 and C3F, radicals. These react further to form new decomposition products. The main product is c2F6 (35-50%). This suggests that CF3 is one of the major radicals formed during pyrolysis. CF3 + CF3 + C2F6 (1) The other main product is C2F4, which is formed by the re- combination of CF2 radicals.CF2 + CF2 + C2F4 ( 2 ) C3F6 may be formed from the atom abstraction reaction involving the CF3 and C3F7 radicals. CF3 CF~(CFZ)&F~ -+ CF3CF2CF2 * CF3CF=CF2 + CF4 (3) 'PP80. The mechanism for the formation of PFIB is believed to proceed via C3F6. Here, a CF2 radical inserts preferentially at the CF3 end of the C3F6 molecule CF3, /F CF3/ c=c \F (4) CF3CFeCF2 + CF2 * v Certainly where large amounts of PFIB are found, C3F6 is also produced in significant amounts. For example, linear Flutecs produce more C3F6 and PFIB than the ring fluids. The main feature for the ring structures is the formation of other C4Fs alkenes, instead of PFIB and the generation of C5 and c6 compounds. Evidently the breakdown of the cycle stucture has a significantly different mechanism from that of the straight- chain Flutecs.Scavenger Experiments Experiments have been carried out to alter the pyrolysis products formed, firstly, with the hope of establishing the above mechanisms and secondly to suppress the generation of PFIB. This involved decomposition of the fluid in the presence of a train of dry air, nitrogen and nitrogen-oxygen mixtures. Nitrogen suppressed the formation of PFIB by 50% with dry air reducing this amount by a further 25%. Oxygen-rich mixtures greater than 80% prevent the production of PFIB, with COF2 as the major pyrolysis product with CF4, CZF4 and C2F6- The authors thank Rhhe-Poulenc (ISC division), Bristol, for financial support via a SERC CASE studentship, Dr.John Davey for his interest in this work and Mr. P. Webb and Mrs. H. Hart for their help with the use of the GC-MS instrument at266 ANALYTICAL PROCEEDINGS , JUNE 1993, VOL 30 Avonmouth. We also thank Dr. G. Nickless and K. Ball for help with the chromatography. 2 3 References 1 Flutec PP Perfluorocarbon Liquids, Technical Bulletin, Rhdne- 4 Poulenc (ISC Division), Avonmouth, Bristol. Turbini, L. Z., and Zado, F. M., Electron. Packag. Prod., 1980, 20, 49. O'Mahony, T. K. P., Cox, A. P., and Roberts, D. J . , unpublished work. Cox, A. P., Ellis, M. C . , Summers, T. D., and Sheridan, J . , J. Chern. SOC., Faraday Trans., 1992, 88, 1079. Determination of Vitamins K I , K2 (MK-4) and K3 in Animal Feeds by High-performance Liquid Chromatography Stephen White Laboratory of the Government Chemist, Queen's Road, Teddington, Middlesex TWI 1 OLY During the process of blood coagulation, inactive prothrombin is activated to thrombin, which converts soluble fibrinogen into insoluble fibrin, thus forming a clot.This latter conversion is dependent on the presence of a coagulation factor known as vitamin K. In 1939, two compounds were identified with antihaemor- rhagic activity: phylloquinone (vitamin K1) and the menaqui- nones (vitamin K2). 0 ij Phylloquinone (vitamin K1) 0 0 Menaquinone (vitamin K2) ( n = 3-1 3) Phylloquinone occurs naturally as the trans isomer in green plants, while synthetic preparations contain both the cis and trans isomers. However only the trans isomer is responsible for antihaemorrhagic activity. Vitamin K2 refers to a series known as the menaquinones, which are synthesized by bacteria.The menaquinones have side-chains consisting of 4-13 isoprene units ( n = 3-13) attached to the methylnaphthaquinone nucleus. The parent compound of vitamin K is 2-methyl-l,4-naphtha- quinone, which is known as menadione or vitamin K3. Menadione can be prepared synthetically and since it is cheaper than phylloquinone it is often added to animal feeds in this form, or as a water-soluble derivative such as menadione sodium hydrogensulfite [menadione sodium bisulfite (MSB)] . 0 0 Menadione (vitamin K3) Menadione sodium hydrogensulfite (MSB) Existing Methodology In the past the determination of vitamins K1, K2 and K3 has required at least two separate chromatographic systems.Determination of K3, described by Manz and Maurer' and by Laffi et al., * is achieved by normal-phase high-performance liquid chromatography (HPLC) with ultraviolet (UV) detec- tion. Vitamin K1 and several menaquinones can be determined by either normal-phase HPLC with UV detection (Haroon et al. 374) , or by reversed-phase HPLC with fluorescence detection. Detection by fluorescence is made possible by ost- column reduction using either chemical (Haroon et al. ) or electrochemical methods (Guillaumont et al. 6 ) . The use of fluorescence measurement as a detection method has the advantage of lower detection limits. However, these systems tend to be more difficult to set up and to stabilize and give less reproducible chromatography, than those using UV detection, which require no derivatization or reduction of the analytes.For the above reasons, and the fact that K1 and K3 could be determined on separate normal-phase systems, an attempt was made to adapt a single normal-phase system to resolve both (and possibly K2). P Method Development Initial findings suggested that K1 and K3 could be resolved on a 5 pm Si column using 10% CC14 in dichloromethane (DCM) as the eluent, at a flow rate of 1 ml min-l. Owing to the hazards associated with CC14 an alternative solvent was desirable. An eluent of hexane-DCM proved a suitable substitute. By modification of eluent composition, and hence the polarity, and column temperature, it became possible to resolve K1, menatetrenone (MK-4) (vitamin K2, when n = 3) and K3, see Table 1.The optimum resolution was observed under the following conditions: column, LiChrosorb Si 60 5 pm (250 x 4 mm i.d.); column temperature, 30 "C; eluent, hexane-DCM (1 + 1); flow rate, 1 ml min-l; detection wavelength, 264 nm. The method developed utilizes a simple chloroform extrac- tion, as described by Laffi et aZ.,2 with base reduction of any menadione salts (e.g., MSB) to free menadione. After Table 1 Chromatographic separation results Vitamin Retention timelmin Phylloquinone (Kl) 3.5 Menadione (K3) 10.0 Menatetrenone (K2; MK-4) 4.0ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 2 1 i 5ml of 25% 109 ammonia celite-Na2SOo l o g sample 6-i (3 + 10, 4 Shake Shake Shake 3min 3min 20min I Neutralize Inject solution chromatograph into *---- Centrifuge 10min 2500 rev mi n ' Fig.1 Schematic diagram of sample preparation A 1 0 5 10 0 5 10 Ti m e/m i n Fig. 2 Chromatograms obtained for ( a ) unspiked animal feed and (b) animal feed spiked with 10 ppm of K1 (peak 1); K2 (MK-4) (peak 2); and K3 (peak 3). Injection volume, 50 pl; other conditions as given in text 267 neutralization, centrifugation and filtration, 50 yl of the chloroform extract are injected-into the chromatograph (Fig. Quantification is achieved by measurement of peak areas and comparison with those obtained for standard solutions. Since MSB has a different vitamin K activity to menadione, a factor of 1.92 should be applied to calculated K3 concen- trations to convert to MSB concentration. Typical chromato- grams are shown in Fig. 2. Recoveries for vitamins K1, K2 (MK-4), K3 and MSB from animal feed and pet foods were as shown in Table 2._- 1). Table 2 Recoveries of vitamins (%) from animal feed and pet food Spike (10 ppm) Animal feed Dog food Cat food K1 87.6 87.6 77.3 K2 (MK-4) 99.0 97.0 89.0 K3 105.2 104.1 96.9 MSB 85.5 95.5 79.8 MSB * 90.3 83.6 86.5 * 20 ppm. Conclusion A system has been developed for the simultaneous determi- nation of vitamins K1, K2 (MK-4), K3 and MSB in animal feeds and pet foods down to 100 ng g-l, which is a considerable improvement over previous methods using UV detection. The method is simple to use and gives high recoveries for all four analytes. Work is to continue, in the hope of developing a clean-up/ concentration step which should improve detection limits further. Also, repeatability for the four forms of vitamin K is to be determined, and a possible comparison between this and the statutory m e t h ~ d .~ The author thanks the Ministry of Agriculture, Fisheries and Food (Chemical Safety of Food Branch B) for funding this work. References 1 Manz, U., and Maurer, R., Int. J. Vitam. Nutr. Res., 1982, 52, 248. 2 Laffi, R., Marchetti, S . , and Marchetti, M., J . Assoc. Off. Anal. Chem., 1988, 71, 826. 3 Haroon, Y., Shearer, M. J., and Barkhan, P., J. Chromatogr., 1980, 200, 293. 4 Haroon, Y., Shearer, M. J., and Barkhan, P., J . Chromatogr. 1981, 206, 333. 5 Haroon, Y., Bacon, D. S., and Sadowski, J. A., J . Chromatogr., 1987, 384, 383. 6 Guillaumont, M., Leclercq, M., Gosselet, H., Makala, K., and Vignal, B., J. Micronutr. Anal., 1988, 4, 285.7 Agriculture, The Feeding Stuffs (Sampling and Analysis) Regula- tions 1982, Statutory Instruments, No. 1144, HM Stationery Office, London, 1982.268 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 Analysis of an (Isodecyl End-capped) Propylenediol Adipate Polyester Using Coupled High-performance Liquid Chromatography-Fourier Transform Infrared Spectrometry Alan M. Robertson, Dell Farnan and David Littlejohn Department of Pure and Applied Chemistry, Thomas Graham Building, University of Strathclyde, Cathedral Street, Glasgow GI IXL Michael Brown Unicam Ltd., York Street, Cambridge CBI 2PX Christopher J. Dowle and Elizabeth Goodwin ICI Wilton Research Centre, P.O. Box 90, Middlesbrough TS6 8JE Plasticizers are normally added to polymers to satisfy per- formance requirements associated with product applications.Polymeric plasticizers are a group of high molecular weight esters based on aliphatic and aromatic diacids and glycols. Their large molecular size renders them less vulnerable to plasticizer extraction and migration, which makes them attractive alternatives to poly(viny1 chloride) in food-wrap cling film applications. The polyadipate plasticizer analysed in this work was prepared at ICI by treating adipic acid (hexane-l,6-dioic acid), propane-l,2-diol and 2-methylnonan-1-01 with an organotin catalyst at 210°C. The water of reaction was thermally removed and the excess of 2-methylnonan-1-01 was used to drive the reaction to completion. When the acid value of the mixture was below 2 mg of KOH per gram and the expected volume of water had been removed, the excess of 2- methylnonan-1-01 was purged by applying a vacuum and sparging with nitrogen.As the 2-methylnonan-1-01 was removed, the viscosity of the product increased and this was continued until the target viscosity and hence required average molecular weight was achieved. The reaction is summarized below and the desired value of n is 9: ( n + l)H02C(CH2)4C02H + nHO(CH2)30H + 2C10H210H = (2n + 2)H20 + ClOH2102C(CH2)4C02[ (CH2)302C(CH2)4C02],ClOH21 (1) High-performance liquid chromatography (HPLC) was used by ICI Wilton to analyse the polymer product. The method involved isocratic elution with a methanol-water (9 + 1) eluent, pumped at a flow rate of 1 ml min-' through a Chromegabond 5 pm pentafluorophenol HPLC column (250 X 4.6 mm) at 45 "C.Elution of separated solutes was detected by ultraviolet (UV) absorption at 210 nm. The HPLC-UV trace obtained for this method of analysis displayed two sets of absorbance peaks believed to be associated with two distinct homologous series (see Fig. 1). The second set of peaks was thought to be associated with an isodecyl end-capped polyester species but the identity of the first set was unknown. As detection by UV absorption provided insufficient structural information, it was decided to investigate the potential of on- line Fourier transform infrared (FTIR) spectrometry for indentification of the components in the chromatogram. HPLC-FTIR Spectrometry The coupling of HPLC with FTIR spectrometry has been the subject of much research for more than a Two main approaches to the interfacing of the techniques have been considered, viz., the use of micro flow cells3-' or solvent For most HPLC separations, removal of the solvent prior to FTIR analysis is preferred.Solvent elimination procedures in HPLC-FTIR have been most successful in normal-phase applications, owing to the high volatility of most organic solvents and their compatibility with conventional alkali halide sampling media."' Conversely, the low volatility of aqueous based reversed-phase HPLC solvents has caused problems and led to the development of a wide variety of methods for elimination of water. Kalasinsky et aL8 added 2,2-dimethoxypropane to the HPLC column effluent to produce acetone and methanol, which were easily evaporated after deposition on to KC1 powder.A less expensive solvent elimination system was described by Fujimoto et aL9 A stainless-steel wire mesh replaced the metal halide substrate and the solvent was evaporated using a heated gas flow; FTIR analysis was then performed in the transmission mode. More recently, solvent elimination methods have been described which utilize the flow of an inert gas to promote evaporation of the mobile phase. The three most successful systems were described by Gagel and Biemann,l0,l1 Somsen et al., l2 and Lange et al. l3 The interface proposed by Gagel and Biemannlo>l1 used a nitrogen gas nebulizer to deposit solutes on to the surface of a rotating reflective disc. Reflectance absorbance spectra of the deposited solutes were obtained by rotating the disc substrate in the sample compartment of an FTIR spectrometer.The interface has recently become com- mercially available and the analysis of a number of environ- mentally important samples has been demonstrated. l4 Somsen et al.12 described a spray-jet assembly which used a heated nitrogen gas flow to aid evaporation of the HPLC solvent. The solutes were deposited on a zinc selenide substrate and analysed by FTIR microscopy. Lange et al.13 evaporated the solvent by passing warm helium gas through the outer of two concentric fused-silica tubes while the effluent from the HPLC system was passed through the inner tube. The solutes were deposited on to a rotatable zinc selenide substrate and FTIR measurements were made in the transmission mode.In general, these devices are less efficient when the HPLC solvent mixture contains even small concentrations of water. However, Robertson and co-~orkers~' have developed a system which can cope with relatively high aqueous solvent flow rates. The monodisperse aerosol generation interface for combination of HPLC with FTIR (MAGIC-HPLC-FTIR) has been used with aqueous solvent flow rates of up to 0.3 ml min-' without effluent heating, although the solute deposition efficiency was only about 10%. Jansen16 described an interface system consisting of a thermospray, moving belt and optical reflectance accessory. Various polymers were analysed and it was reported that the system could be used with 30% water in methanol at a flow rate of 0.5 ml min-l. We have also developed a thermospray interface system" similar in concept to that outlined by Jansen.The thermospray was used to evaporate the mobile phase and deposit solutes on to a moving belt located in the diffuse reflectance accessory of an FTIR spectrometer. The interface is compatible with most normal- or reversed-phase solvent systems and has coped effectively with 100% aqueous effluents at flow rates of up to 1 ml min-l. The operational character- istics of the thermospray have been described18 and com-ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 269 pounds analysed by the HPLC-FTIR combination include sugars, fatty acids, amino acids and plastic additives. In this paper, the application of HPLC-FTIR to the analysis of a complex industrial sample, a polyadipate plasticizer, is described.Experimental Apparatus The HPLC system was a Unicam PU 4100 liquid chromato- graph (Cambridge) with a Rheodyne 7125 syringe loading injector (Berkeley, CA, USA) and a 20pl sample loop. The analytical column was a Hichrom stainless-steel (250 X 4.6 mm i.d.) Chromegabond 5 pm pentafluorophenol (PEP) column (Berkshire) . Various compositions of methanol-water mobile phases were pumped through the column at a flow rate of 0.5 ml min-l. A Unicam PU 4110 UV-visible detector (Cam- bridge) was used during the HPLC method development prior to FTIR spectrometric detection. A zmooth stainless-steel ribbon (width 13 mm, thickness 0.02 mm), was used as the substrate for thermospray de osition. The belt was moved at a constant speed of 1 cm min-'to transport the solutes into the FTIR spectrometer.Effluent from the HPLC system was transferred to the thermo~pray'~~'~ through stainless-steel tubing (250 pm i.d., 1.5 mm 0.d.) coupled to a 30 cm length of narrow bore stainless-steel tubing (125 pm i.d., 1.5 mm 0.d.). The narrow bore tubing was inserted into a 10 cm length of copper tubing (1.5 mm i.d., 4 mm 0.d.). Philips thermocoax heating wire (Cambridge) was brazed around the outside of this copper tubing. The narrow bore tubing protruded 1.5 cm from the heating assembly. A Farnell 'L' series 120 W (variable) power supply (Wetherby) was used to provide power to the thermo- spray. A Fluke Type 52 digital thermometer (Watford) and 'K' type thermocouple were used to monitor the temperature at the tip of the thermospray.Infrared spectrometric data were obtained using a Unicam PU 9800 FTIR spectrometer (Cambridge) equipped with a Spectra-Tech Collector diffuse reflectance accessory (Stam- ford, CT, USA). The FTIR instrument was equipped with a deuteriated triglycine sulfate (DTGS) detector. The IR data, in the form of single-scan interferograms, were collected at 16 cm-I resolution and stored using standard Unicam FTIR computer software (Cambridge). Interferograms were trans- formed to IR transmittance spectra using the fast Fourier transform algorithm, post-run. Background reflectance spectra were obtained from part of the clean substrate surface immediately before the solute deposit of interest. Specially developed Unicam computer software was used to process these data and construct FTIR functional group chromato- grams (FGCs).These were obtained by calculating the integrated IR absorbance across various wavenumber windows (corresponding to particular functional groups), as a function of time. Reagents Standard diisodecyl adipate and polyester plasticizer samples were provided by ICI Wilton Research Centre (Cleveland). HPLC-grade methanol and tetrahydrofuran were obtained from Rathburn Chemicals (Walkerburn), and water was distilled prior to analysis. Preliminary Work The HPLC-UV method of analysis for the polyadipate plasticizer, outlined by ICI Wilton, was modified in order to make it more compatible with the operational characteristics of the thermospray interface. The flow rate of the mobile phase was lowered from 1.0 to 0.5 ml min-' to reduce the optimum thermospray operating temperature from 165 to 135 "C and hence reduce the risk of any thermal degradation of the plasticizer sample during thermospray deposition.The mobile- phase composition was then altered from 90 + 10 to 92 + 8 methanol-water to complete the HPLC separation within a reasonably short period (about 25 min). The slight reduction in the aqueous content of the mobile phase probably also contributed to the use of the lower thermospray temperature. Procedure A 1.2 g specimen of the polyadipate plasticizer sample was dissolved in 2 ml of tetrahydrofuran, 0.5 ml of water was added and the solution was made up to 10 ml with methanol. A 20 pl volume of this solution was injected into the HPLC system and the solutes were detected by UV absorption at 210 nm, prior to thermospray deposition on to the moving substrate for detection by FTIR spectrometry.Results and Discussion Fig. 1 shows an overlay of HPLC traces for the analysis of the polyadipate plasticizer sample using ( a ) UV absorption at 210 nm and ( b ) FTIR spectrometric detection (wavenumber window = 1790-1680 cm- I). Two distinct chromatographic regions are evident in the UV chromatogram. For the purpose t 0 5 10 15 20 25 0 5 10 15 20 25 Ti rne/mi n Fig. 1 Overlay of HPLC traces for (a) UV detection at 210 nm and (b) FTIR detection (V window = 1790-1680 cm-I) of a polyadipate plasticizer. Conditions: column, PFP; eluent, CH~OH-HZO (92 + 8); flow rate, 0.5 ml min-l; sample concentration, 120 mg ml-'; thermospray temperature, 135 "C; belt speed, 1 cm min-' 0.250 g 0.200 s c m ? 0.150 a (0 0.100 rr 0.050 4000 3000 2000 1500 1000 Fig.2 FTIR absorbance spectrum corresponding to the compound with a retention time of 3 min in the chromatogram shown in Fig. l(b). Conditions as in Fig. 1 Compressed wavenum bedcm-'270 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 of this discussion these regions have been denoted region A (retention times <7 min) and B (retention times >7 min). Comparison of the UV and F-TIR chromatograms in Fig. 1 indicates that the peaks in region A of the IR chromatogram were smaller then expected. This might suggest that the species contained in region A were more volatile then those in region B and were less efficiently deposited on to the moving belt by the thermospray.Similar observations were made when other wavenumber windows (e.g., 3000-2800 cm-' and 1500-1400 cm-l) were used to produce FTIR functional group chromatograms. Fig. 2 shows the IR absorbance spectrum of the largest peak in region A, which occurred at a retention time of about 3 min. The main diagnostic features of this spectrum and the spectra for peaks at 2 and 3.5 rnin are as follows. (a) A weak absorption band is apparent at 3550-3400 cm-l, which was assigned to the 0-H stretching vibration. [This absorption band would not be present for a polyester species containing two isodecyl end- caps, as in eqn. (l).] ( b ) A medium intensity absorption band exists at 300G2850 cm-', which was assigned to the aliphatic hydrocarbon stretching vibration. ( c ) The intense absorption band at 1790-1680 cm-' was assigned to the carbonyl stretching vibration.( d ) There are two relative1 weak absorption bands at 1480-1430 and 1390-1340 cm-[ which were assigned to the hydrocarbon deformation. (e) The series of absorption bands at 1300-1000 cm-' could be due to either the C-0 stretching vibration or the 0-H deformation. The general features of this spectrum are consistent with those of an aliphatic ester. The relative intensity of the carbonyl to hydrocarbon absorption bands and the apparent presence of the 0-H stretching vibration suggests that the solute eluted at 3 rnin and those at 2 and 3.5 rnin correspond to polyester molecules that may only have one isodecyl end-cap. As the IR spectra of the species in region A were very similar, it is probable that they represent a homologous series of polyesters without two isodecyl end-caps. Fig.3 shows the IR absorbance spectrum of the first absorbance peak in region B, which occurred at about 7.5 min. The main diagnostic features of the IR spectrum are similar to those of the compound that eluted at 3 min. However, the 4000 3000 2000 1500 1000 Fig. 3 FTIR absorbance spectrum corresponding to the compound with a retention time of 7.5 min in the chromatogram shown in Fig. l(b). Conditions as in Fig. 1 Compressed wavenumber/cm-' Table 1 Ratio of CH to CO peaks in the FTIR spectra obtained for the chromatogram in Fig. l(b) Retention time/ min 3.0 3.5 9.0 10.5 12.0 14.0 16.5 Region A 2.0 Region B 7.5 CH: CO peak absorbance ratio 0.7 0.4 0.4 1 .o 0.9 0.7 0.3 0.3 0.3 relative size of the carbonyl to hydrocarbon absorption bands and the absence of the 0-H stretch feature imply that the polyester species formed was probably doubly end-capped. The IR spectra obtained for the first six components in region B were similar, but showed a gradual decrease in the relative magnitude of the hydrocarbon stretching and bending absorp- tion bands compared with the CO band, as indicated in Table 1.The spectrum of the component at 16.5 rnin is given in Fig. 4. The trends observed suggest that region B is a homologous series of isodecyl double end-capped polyesters where n, the number of adipate ester groupings, increases with retention time. An IR spectrum of diisodecyl adipate is given in Fig. 5. A standard diffuse reflectance FTIR method was used to obtain the spectrum of the adipate on a substrate of KBr powder.The main features of the spectrum are similar to those in Fig. 3 and support the suggestion that the components in the first part of region B have isodecyl end-caps and a small number of adipate ester groups in the polymer (i.e. , low It value). The proposed explanation is consistent with the expected separation characteristics of the PFP column. Large molecules will exhibit greater retention times, so the isodecyl double end- capped homologous series should be eluted after the single end-capped series. Also, within a series, it is expected that the more polar molecules (i.e., larger n values) will have longer retention times. The presence of single and double isodecyl end-capped homologous series has also been supported by liquid chroma- tography-nuclear magnetic resonance (LC-NMR) measure- ments obtained by analysis at ICI Wilton in collaboration with Bruker Analytische Messtechnik and co-workers at the Institut fur Organische Chemie, Tubingen, Germany.l9 The LC-NMR measurements at 400 MHz required developments in high- resolution probe-heads and the possibility to perform efficient multiple solvent suppression. This study has provided a further example of the usefulness of FTIR detection in the HPLC analysis of complex mixtures. From the structural details obtained from the IR spectra of the 4000 3000 2000 1500 1000 Compressed wavenumbedcm-' Fig. 4 FTIR absorbance spectrum corresponding to the compound with a retention time of 16.5 rnin in the chromatogram shown in Fig. l(6). Conditions as in Fig. 1 r I 4000 3000 2000 1500 1000 Compressed wavenumber/cm-' Fig. 5 FTIR absorbance spectrum of diisodecyl adipate. Spectrum obtained using KBr powdered substrate and diffuse reflectance accessoryANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 271 separated components, it was possible to deduce information 7 8 that could not be derived from UV detection alone. There was some evidence that the initial compounds eluted from the column were not deposited as efficiently or did not absorb as strongly in the IR region as the latter species. This has not been 9 10 11 12 fully explained and merits further investigation. One of us (A.M.R.) acknowledges support provided by ICI Wilton Research Centre, Middlesbrough. 13 14 15 16 17 18 19 References Hellgeth, J. W., and Taylor, L. T., J. Chromatogr. Sci., 1986, 24, 519. Griffiths, P. R., and de Haseth, J. A., in Fourier Transform Infrared Spectroscopy, eds. Elving, P. J., and Winefordner, J. D.. Wiley, New York, 1986. Fujimoto, C., Vematsu, G., and Jinno, K., Chromatographia, 1985, 178, 159. Hellgeth, J. W., and Taylor, L. T., Anal. Chem., 1987,59,295. Kuehl, El., and Griffiths, P. R., J. Chromatogr. Sci., 1979, 17, 471. Johnson, C. C., andTaylor, L. T., Anal. Chem., 1984,56,2642. Wood, D. J., Spectrosc. Int., 1990, 2, 36. Kalasinsky, V. F., Whitehead, K. G., Kenton, R. C., Smith, J. A. S., and Kalasinsky, K. S., J. Chromatogr. Sci., 1987, 25, 273. Fujimoto, C., Oosuka, T., and Jinno, K., Anal. Chim. Acta, 1985, 178, 159. Gagel, J. J., and Biemann, K., Anal. Chem., 1987, 59, 1266. Gagel, J. J., and Biemann, K., Mikrochim, Acta, 1988,2,185. Somsen, G. W., van de Nesse, R. J., Gooijer, C., Brinkman, U. A.Th., and Velthorst, N. H., J . Chromatogr., 1991, 552, 635. Lange, A. J., Griffiths, P. R., and Fraser, D. J. J., Anal. Chem., 1991, 63, 782. LC-Transform, Commercial Literature, Lab Connections Inc., Marlborough, MA, USA, 1991. Robertson, R. M., de Haseth, J. A., and Browner, R. F., Appl. Spectrosc., 1990, 44, 8. Jansen, J. A. J., Fresenius’ J. Anal. Chem., 1990, 337, 398. Robertson, A. M., Wylie, L., Littlejohn, D., Watling, R. J., and Dowle, C. J., Anal. Proc., 1991, 28, 8 . Robertson, A. M., Littlejohn, D., Brown, M., and Dowle, C. J., J. Chromatogr., 1991, 588, 15. Spraul, M., Hofmann, M., Glauner, H., Gans, J., and Albert, K., NMR Instrumentation, 1992, 12.

ISSN:0144-557X    

DOI:10.1039/AP9933000257

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

272 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 Precursors to and Evolution of Elemental Organic Tube Combustion - Analysis Over the Last Two Hundred Years D. Thorburn Burns Department of Analytical Chemistry, The Queen’s University of Belfast, The tube combustion procedures for organic elemental analysis go back to the pioneering studies of Lavoisier using vari- ous oils. The problems of early workers with the determination of hydrogen, caused by difficulties with the absorption of water, were solved elegantly by Prout for the analysis of carbohydrates and organic acids using a known volume of oxygen for the combustion. The princi- ples of modern automatic CHN analysers can be seen to derive from the work of earlier analysts such as Rigg, who used solid oxidants, coupled with the micro- practices developed by Pregl in the post- Leibig era.Introduction Since the time of Robert Boyle it has been stated that ‘an element is the ultimate product of analysis’. Early chemists’ methods of analysis were not completely specific and it was a long time before the concepts of the Aristotelian elements, earth, air, fire and water, were refined into the chemical elements as now under- stood.’ Fire was at one stage both an element and a resolving agent or agent for analysis: Boyle’s ‘Sceptical Chemist’ is an essay on the validity of fire assay,2 Lavoi- sier solved the problem of the nature of combustion, reaction with oxygen, and considered that organic compounds were com osed of carbon, hydrogen and oxy- gen! Later, Berthollet dicovered the presence of nitrogen in organic com- p o u n d ~ , ~ the determination of which was surveyed by Stephen’ and Thorburn Burns6 at the Kjeldahl Centenary Meet- ing in 1983.Antoine Laurent Lavoisier (1743-1794) Lavoisier, born in Paris in 1743, became an able scientist and public administrator; however, the latter was the cause of his death by guillotine in 1794 after trial by the Revolutionary Tribunal.’ He is justly remembered for his discovery of the role of oxygen in chemical reactions and for his reform of chemistry, including the elimination of the concept of phlogiston. Lavoisier, as noted by Szabadvary’” and by Belcher,’ was the first to attempt to determine the composition of organic compounds. First, he examined various oils by combustion in a somewhat compli- cated apparatus (Fig.l),37’0>1’a the description and operation of which is readily available in Kerr’s translation of ‘Elements of Chemistry’.’* The bulk of the water produced was condensed and collected (at 16 in Fig. l), the remainder being removed by an absorption tube (19- 20) containing deliquescent salts. Carbon dioxide was absorbed in a series of bubblers (22-25) containing caustic alka- line solution, a total of at least nine, the later ones filled with lime water (to check for complete absorption). The air was collected in a gazometer (connected at 30) and small portions withdrawn and assayed for residual oxygen by absorption with potassium sulphide. The original plates were drawn by Madame Lavoisier who, as can be seen from her portraits, was elegant as well as useful. (Marie Anne Paulze married Lavoisier in 1771, aged 13.) The original plate of the combustion apparatus contains an error: the tubes in the carbon dioxide absorbers were con- nected the wrong way round; Ken cor- rected this mistake. l2 Lavoisier wisely did not attempt to combust volatile com- pounds such as alcohols in this a aratus because of the risk of explosionR‘>” but used much simpler apparatus and obtained 29% C for spirit of wine.Lavoi- sier, in later experiments, made use of solid oxidants for the analysis of sugar.’“ Although his results were only approxi- mate his work is important in showing the correct approach to the problem of or- ganic elemental analysis. Richard Rigg (1799-1861) Rigg was one of the earliest British organic elemental analysts.Very little is Belfast BT9 5AGf Northern Ireland known about him; Boase gives but a brief biography13 based on Lonsdale’s ‘Worth- ies of Cumberland’.14 The Royal Society, to which he was elected in 1839, has some archival material mainly concerned with his publications. His Royal Society nomi- nation certificate was signed by Faraday , Daniell, Graham and Phillips, amongst others. Rigg published 16 papers in the period 1836-1846, dealing with chemical changes in the fermentation and germina- tion of seeds, growth of plants and organic analysis. His method of analysing or anic compounds appeared in Phil. in 1838 and in the Arcana of Science,17 although it had been shown to the Royal Society about 2 years earlier. The appara- tus consisted of two small glass tubes connected by a caoutchouc (rubber) collar (C) supported on a wire frame (Fig.2). The compound to be analysed was mixed with black copper oxide [cop- per(r1) oxide] packed into tube A and covered with 1 inch or more of the same copper oxide. The end of the tube was filled with dry amianthus (fibrous asbes- tos) or cotton wool. This section of the tube was heated to drive off the moisture. The tube was cooled, weighed and con- nected up to the capillary tube B, the bent end of which was placed in a mercury trough under the graduated collection tube. The pure copper oxide section was brought to red heat with a spirit lamp, then the whole tube was heated to white heat and rotated in the flame. The com- bustion was carried out slowly in order to T R A I T E E L E M E N T A I R E D E C H I X I E Fig.1 Lavoisier’s apparatus for the combustion analysis of oils (1784)ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 273 x \ Fig. 2 Rigg’s elemental analysis apparatus (1838) avoid the formation of carbon monoxide. The tube was next allowed to cool. The loss in mass was equal to the mass of N2 + C02. The analysing tube was then reheated to drive off water, cooled and re- weighed. The nitrogen and carbon diox- ide were collected in the graduated tube. The carbon dioxide was removed by absorption with potassium hydroxide solution. A correction was made for the gaseous products left in the capillary tube and in the analysing tube (2/10 of its interstices). The ultimate analysis was calculated from the weights of water, carbon dioxide and nitrogen, and the volumes of C 0 2 and N2.Rigg’s technique and results were praised by J. B. Readel’ who, it should be noted, also signed his Royal Society Nomination Certificate. The results were, in fact, far from reliable and led Rigg, by now an FRS, to the erroneous conclusion that carbon was not an element but a compound body made by plants.” An editorial footnote to Riggs’s paper15 on the preparation of black copper oxide drew attention to Prout’s contribution to organic elemental analysis. William Prout (1785-1850) Prout was born on a farm in Horton, Gloucestershire, but spent most of his life in London, where from 1812 he practised as a physician until his death in 1850.2@22 He studied medicine in Edinburgh (1808- 1811) where he came under the influence of the distinguished chemist T.C. Hope. He made significant contributions to physiological chemistry including the dis- covery that the stomach secretes free hydrochloric acid, and is remembered for his unitary matter theory, i.e., that the chemical elements possess atomic weights which are integral multiples of the atomic weight of hydrogen,23 now known as Prout’s Law or Prout’s Hyp~thesis.~~-~’ This law, published anonymo~sly,~~ gave a great impetus to the determination of accurate atomic weights. Prout’s contri- butions to organic analysis have been overlooked. Szabadvary’ dismisses them as ‘difficult to understand’ and comments that the results were unexpectedly precise. Prout’s first contribution to the analysis of organic substances,28 published in Thornson’s Annals of Philosophy, con- cerned the use of Wollaston’s chemical equivalents in calculating formulae and in drying samples prior to analysis, for which purpose he designed and used a vacuum desiccator wherein the sample is dried at 212°F and exposed to concentrated sul- phuric acid (Fig.3). He produced excel- lent results for the analysis of urea29 and reported that black copper oxide was a better oxidant than potassium chlorate when compounds contained nitrogen. He then gave a detailed description of his well-engineered apparatus3’ in which the combustion tube was vertical and a circu- lar burner was used to heat the tube evenly (Fig. 4). He determined water either by differences (which he thought was best) or by condensation coupled with a calcium chloride drying tube.Early elemental analysts found considerable difficulties in drying substances and hand- ling hygroscopic powdered substances. Prout solved the problem in a most elegant manner, by noting that if a compound containing hydrogen, carbon and oxygen is burnt in oxygen, the volume of oxygen remains unchanged if the hydrogen to oxygen proportion is the same as in water; the determination of carbon dioxide then gave a complete analysis.31 Corrections could be made if the volume of the combustion product was greater or less than the original Fig. 3 Prout’s vacuum dessicator (1815) volume of oxygen. His second elemental analysis apparatus used a horizontal com- bustion tube with vertical mercury man- ometer tubes, which also acted as pumps to drive the oxygen back or forth over the sample (Fig.5). The results were amazingly accurate (Table 1). Little is known of Prout’s personal life in London but he was a successful physician who specialized in digestive and urinary com- plaints, as indicated by much of his chemical publications. Unfortunately, deafness made him avoid scientific con- tacts after 1830. He was extremely religious and was invited to write one of the eight Bridgewater Treatises32 which had the general title ‘On the Power, Wisdom and Godness of God, as Mani- fested in the Creation’. He dealt with his personal interests in ‘Chemistry, Metero- logy and the Function of Digestion’.33334 Tube combustion methods were improved by Leibig and Dumas, then refined and transformed to the micro- scale by Pregl .8,9 Fritz Pregl (1869-1930) Pregl was born in Laibach, Austria (now Ljubljana, Yugoslavia) in 1869.35,36 Edu- cated at the local Gymnasium, he entered the University of Graz to read Medicine, where he gained an intense interest in physiology which led, as with Prout, earlier, to the need to develop appropri- ate analytical methodology for his prob- lems.Whilst working on bile acids and protein chemistry he found the then analytical methods too complicated, lengthy and inexact, and in addition requiring large samples which were diffi- cult to obtain. He thus focused on organic microanalysis, which gradually claimed his full attention. He was appointed Professor of Medical Chemistry at Inns- bruck in 1910. His first task was to obtain a sufficiently sensitive microbalance.This was made for him by W. H. Kuhlmann of the P. Bunge Works, based on an earlier balance made for F. Emich. Pregl scaled down the methods of Leibig (Fig. 6) and Dumas and, after a deal of effort and close attention to detail, developed reli- able methods which spread world-wide. He extended the work, developing micro- methods for sulphur, halogen, carboxyl and other functional groupings. The results of these studies were published in 1917: ‘Die Quantitative Organische M i k r o a n a l y ~ e ’ . ~ ~ , ~ ~ The impact was such that in 1923 he was awarded the Nobel Prize for Chemistry, the first to be awarded for accomplishment in Analyti- cal Chemistry. In 1913 he had been recalled to Graz, where he remained active in research till his death in 1930. He is commemorated in the Fritz Pregl prize for Microchemistry, awarded by the Aus- trian Academy of Science and by the Fritz Pregl medal of the Austrian Society for Analytical Chemistry.274 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 L ~ ~ Y z v .~ /rorBfl&m& ,&/ah’- hrB&vm &u%& & Ji, 2&T#68b f i r . J/&L iuaa Fig. 4 Prout’s elemental analysis apparatus (1820) Post-Pregl to Full Automation of Analysis Following the establishment of the Pregl methods modifications became necessary as the range of organic and organometal- lic compounds requiring analysis increased. The developments and modifi- cation may be summarized under four main heading^.^^,^' 1. Combustion tube fillings (oxidation fillings, nitrogen oxide absorbents, etc. ) .2. Combustion procedures (partial auto- mation and rapid procedures). 3. Apparatus design. 4. Determination of carbon dioxide and water (chemical, manometric and chromatographic procedures). Fig. 5 (1 827) Prout’s elemental analysis apparatus The next major development following Simon’s work4* in the early 1960s was Perkin-Elmers’ production of the Model 240 Elemental Analyser for C, H and N. Its semi-automatic operation is well docu- mented.40.42 The sample (2-3 mg) is com- busted in oxygen at about 950°C. After oxidation interfering components are removed and the products H20, C02 and oxides of nitrogen are swept to a reduc- tion stage by means of helium. The final combustion products H20, COz and N2- He are homogenized. The detection system measures thermal conductivity before and after the absorption of H20 and CO,; finally only N2 and He remain, which are related to pure helium to give a signal for N2.Modifications to combus- tion tube packings, developed to deal with fluorinated compounds43 and for use when large numbers of halo en and be analysed, are now recommended. Operational modifications such as to the ladle45 and the digitization of output data46 have been described. sulphur containing compounds4 ‘F have toANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 1 St n Fig. 6 Pregl’s micro elemental analysis train (1917) Full automation following sample en- capsulation is achieved in the Carlo Erba l102.40,42 It uses the same basic Leibig chemistry4’ but the final products of combustion/reduction are separated by gas chromatography and measured by a katharometer.Provided that flow rates are carefully controlled and widely differ- ent sample composition standards are used excellent results can be attained.48 1 2 3 4 5 6 7 8 9 10 11 12 References Read, J . , Through Alchemy to Chemistry, Bell, London, 1961. Boyle, R., The Sceptical Chemist. . . , J. Cadwell for J. Crooke, London, 1661. Lavoisier, A. L., Trait6 Elementaire de Chemie.. .. Cuchet, Paris, 1781. Berthollet, C. L., J. Phys., 1786, 28, 272. Stephen, W. I., ‘Determination of Nitrogen in Organic Compounds in the Years before Kjeldahl’s Method’, Anal. Proc., 1984, 21, 215. Thorburn Burns, D., ‘Kjeldahl, the Man, the Method and the Carlsberg Laboratory’, Anal. Proc., 1984,21,210. Guerlac, H ., ‘Antoine-Laurent Lavoi- sier’, in Dictionary of Scientific Bio- graphy, ed. Gillispie, C. C.. Scribner, New York, 1973, vol. 8, p. 66. Szabadvary, F., History of Analytical Chemistry, Pergamon Press, Oxford, 1966; 8a p. 285; 8b p. 289. Belcher, R., ‘The Elements of Organic Analysis’, Proc. Anal. Div. Chem. Soc., 1976, 13, 153. Lavoisier. A. L., ‘MCmoire sur la Com- binaison du Principe Oxygine avec L’Espirit-du-vin, L’Huile & Differens Corps Combustibles’, Mem. Acad. Sci., Paris, 1784, 593. Lavoisier, A. L., Oeuvres de Lavoisier, Imp. Imperiale, Paris, 1964-93, vols. 1- 6. l l a vol. 1, p. 346; l l b vol. 2. p. 586; Lavoisier, A. L. (trans. Kerr, R.), Elements of Chemistry . . ., Wm. Creech, Edinburgh, 1790 (5th edn., 1802). Ilc vol. 2, p. 773. 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Boase, F., Modern English Biography .. ., Netherton and Worth, Truro, Lonsdale, H., Worthies of Cumberland, G. Routledge & Sons, London, 1867- 1875, vols. 1-6. Rigg, R., ‘On a Method of Analysing Organic Compounds’, Phil. Mag., 1838, 12, 31. Rigg, R., ‘Further Observations on the Ultimate Analysis of Organic Com- pounds’, Phil. Mag., 1838, 12, 232. Rigg, R., ‘Method of Analysing Organic Compounds’, in Arcana of Science and Art, J. Limbard, London, 1838. Reade, J . B., ‘On the Chemical Compo- sition of Vegetable Membrane and Fibre: With a Reply to the Objections, Professor Hemslow and Professor Lind- ley’, Phil. Mag., 1837, 11, 421. Rigg, R., Experimental Researches; Chemical and Agricultural Shewing Car- bon to be a Compound Body, Made by Plants and Decomposed by Putrefac- tion, Smith.Elder & Co, London, 1844. Brock, W. H., ‘William Prout’. in Dic- tionary of ScientiJc Biography. ed. Gil- lispie, C. C., Scribner, New York, 1973, vol. 174. Glasstone, S., ‘William Prout (1785- 1850), J. Chem. Educ., 1947, 24, 478. Copeman, W.S., ‘William Prout M.D. F.R.S., Physician and Chemist’, Notes Rec. Roy. SOC., 1969, 247. 273. (Prout, W.), ‘On the Relation between the Specific Gravities of Bodies in their Gaseous State and the Weights of their Atoms’, Ann. Phil., 1815, 5 , 321. Siegfried, R., ‘The Chemical Basis for Prout’s Hypothesis’, J. Chem. Educ., 1956, 33, 263. Benfey, 0. T., ‘Prout’s Hypothesis’, J. Chem. Educ., 1952, 29, 78. Kendall, J . , ‘The Adventures of an Hypothesis’, Proc.Roy. SOC. Edin- burgh.. 1949-50, 53A, 1. Brock, W. H., ‘Studies in the History of Prout’s Hypotheses Parts I and 11’. Ann. Sci., 1969, 25, 49 and 127. 1892-1921, VO~S. 1-6. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 275 Dr. Prout, ‘Some Observations on the Analysis of Organic Substances’, Ann. Phil., 1815, 6, 269. Prout, W., ‘Observations on the Nature of Some of the Proximate Principles of Urine . . .’, Ann. Phil., 1818, 11, 352. Prout, W., ‘Description of an Appara- tus for the Analysis of Organised Sub- stances’, Ann. Phil., 1820, 15, 413. Prout, W., ‘On the Ultimate Compo- sition of Simple Alimentary Substances; with some Remarks on the Analysis of Organised Bodies in General’, Phil. Trans., 1827, 355. Brock, W., ‘Prout’s Chemical Bridge- water Treatise’, J.Chem. Educ., 1963, 40, 652. Prout, W., Chemistry, Meteorology and the Function of Digestion Considered with Reference to Natural Theology, W. Pickering, London, 1834; (2nd edn., 1834, 3rd edn., 1845, 4th edn., 1855). Brock, W. H., ‘William Prout and Barometry’, Notes Rec. Roy. Soc., 1969, 24, 281. Lieb, H., ‘Fritz Pregl’, Berichte, 1931, 64b, 113. Lieb, H., ‘Fritz Pregl’, Mikrochemie, 1931, 3, 105. Pregl, F., Die Quantitative Organische Mikroanalyse, J. Springer, Berlin, 1917. Pregl, F., (transl. Fylman, E.), Quanti- tative Organic Microanalysis, J . & A. Churchill, London, 1924. Ingram, G., Methods of Organic Ele- mental Microanalysis, Chapman and Hall, London, 1962. Bance, S . , Handbook of Practical Organic Analysis, Ellis Horwood, Chi- Chester, 1980. Simon, W., Sommer, P. F., and Lyssy, G. H., ‘Complete Automation of the Microdetermination of Carbon and Hydrogen in Organic Compounds’, Microchem. J., 1962, 6, 239. Belcher, R., Instrumental Organic Ele- mental Analysis, Academic Press, Lon- don, 1977. Macdonald, A. M. G., and Turton, G. G., ‘The Automated Analysis of Highly Fluorinated Organic Materials. A Note’, Microchem. J . , 1968, 13, 1. Gustin, G., and Tefft, M. L., ‘Improved Recovery of Rapid Micro Carbon and Hydrogen Method by Modified Com- bustion-Absorption Techniques’, Mi- crochem. J.. 1966, 10, 236. Thorburn Burns, D., McKnight, H. B., and Swindall, W. J., ‘Improved Com- bustion Ladle for an Elemental Ana- lyser’, Lab. Pract., 1978, 27, 650. Thorburn Burns, D., McKnight, H. B., Quigg, K. K., and Swindall, W. J . , ‘Automated Data Handling System for the Perkin-Elmer 240 Elemental Ana- lyser’, Analyst, 1980, 105, 544. Pella, E., and Colombo, B., ‘Study of Carbon, Hydrogen and Nitrogen Deter- mination by Combustion Gas Chroma- tography’, Mikrochimica Acta, 1973, 697. Swindall, W. J., and Thorburn Burns, D., ‘Improvements to the CHN Perfor- mance of a Carlo-Erba 1106 Elemental Analyser by Blank Evaluation, Drift Correction Using a Pair of Dissimilar Standards and Modifications to the Gas Flow system’, 2. Analyt. Chem., 1988, 31, 730.

ISSN:0144-557X    

DOI:10.1039/AP9933000272

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

276 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 Analytical Viewpoint The following is a member of a continuing series of articles providing either a personal view of part of one discipline in analytical chemistry (its present state, where it may be leading, etc.), or a philosophical look at a topic of relevance to chemists in general or analytical chemists in particular. These contributions need not have been the subject of papers at Analytical Division Meetings. Persons wishing t o provide an article for publication in this series are invited to contact the editor of Analytical Proceedings, who will be pleased to receive manuscripts or t o discuss outline ideas with prospective authors. Preliminary Operations: A Pending Goal of Today's Analytical Chemistry M. Valcarcel, M.D. Luque de Castro and M. T. Tena Department of Analytical Chemistry, Faculty of Sciences, University of Cordoba, E- 14004 Cordoba, Spain Introduction Analytical chemistry has developed dramatically over the past few decades. The incorporation of an R&D component has allowed the principles, techniques, methodologies and appli- cations associated with its discipline to respond expeditiously to the new social, economic, scientific and technological demands, as well as those which have arisen from advances in other scientific-technical areas. The primary objective of today's analytical chemistry is gaining access to as much accurate, reliable chemical information of the highest possible quality by expending increasingly less material, time and human resources, taking the lowest hazards and incurring the smallest expenses.Accomplishment of this multiple oal is currently facilitated by breakthroughs in automation! mini- aturization, hyphenated instrumental techniques, computers and chemometrics, among others. However, on the verge of the XXI century, one must admit that the historical development of analytical chemistry has not been too harmonious. The analytical process, viz., the set of operations performed to obtain the desired results from a raw (uncollected, untreated, unmeasured) sample, is executed in three stages: (a), preliminary operations (sampling, sample Data acquisition Fig. 1 ment of the three basic stages Analytical process: relative significance and stage of develop- preservation and treatment, subjection to separation tech- niques); (b), measurement and transducing of the analytical signal, which entails using an instrument; and (c), data acquisition and processing (see Fig.1). The last two have reached a stage of development that could hardly be envisaged not long ago. The quality of the currently available instrumen- tation is beyond doubt, as is the affordability and processing power of (micro)computers and chemometric software. While growth in this field is still possible, genuinely innovative ideas do not emerge so easily, so further, worthy developments are few and far between. On the other hand, advances in preliminary operations have run at a much slower pace than at the other two stages of the analytical process despite their, doubtless, decisive significance in obtaining quality analytical information in an expeditious, economic and human and environmentally safe way.However good the instrumentation and chemometric software used may be, the result can never be of adequate quality if preliminary operations are not per- formed appropriately. Basic Features of Preliminary Operations Preliminary operations share some general features (Fig. 2), which can be summed up as follows. (a) They are rather variable. In fact, each analytical process n pz\ demanding of errors -- Diff control icu It Fig. 2 Features of the first stage of the analytical process SourceANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 277 has some peculiarities that call for different approaches to preliminary operations on the basis of: the objectives of the analytical application in question; the type of analysis involved (qualitative, quantitative, structural, surface, speciation, etc.); the state of aggregation, nature and availability/affordability of the sample to be analysed; the type, concentration and number of analytes to be determined; and the analytical technique (instrument) required or available for the purpose.(6) They entail intensive human participation, which is in clear contrast to the other two stages of the analytical process. Because of their technical complexity, preliminary operations are rather difficult to automate. Also, because of their variability, instrument manufacturers only undertake the automatiodenhancement of preliminary operations in those instances where a wide market (e.g., clinical samples) is anticipated.Some ‘automatic analysers’ only reduce human participation in the simplest of these operation^.^ (c) They are time-consuming or even tedious. In fact, they typically take up 70-95% of the time devoted to the over-all analytical process- this is particularly evident when the fast instruments and computers currently available are used. (d) They are the source of major bias or accidental errors, which have a decisive influence on the quality of the results obtained. Preliminary operations are much more liable to introduce considerable errors than are the other two stages of the analytical process. Inappropriate sampling, incomplete dissolution or disaggregation, inefficient removal of inter- ferents, etc., may give rise to much larger errors than do the other two stages.Such errors arise from the very technical complexity of preliminary operations as well as the increased involvement of the operator, which reflects in the so-called ‘human factor’. Therefore, preliminary operations are as significant as, or even more than, the other two stages of the analytical process. (e) They defy systematic control, which is in contrast to the ease with which an instrument can be calibrated. Controlling and matching every preliminary operation is rather a complex, tedious task which can be addressed in three ways, namely: (l), by relying on the quality assurance principles set in the applicable quality manual and performing external or internal audits; (2), by using certified reference materials similar to the constituent materials of the samples concerned; (3), by using a combination of the previous two.1 Raw sample I (f) They are a major source of potential hazards to both laboratory personnel and the environment. Some operations involve using acids, solvents, a pressurized gas, evaporators, digesters, etc., which may be harmful to the operator immediately (e.g., accidents) or in the medium-to-long term (e.g., exposure to toxic vapours). The disposal of toxic laboratory waste poses special problems as preliminary operations typically use large volumes of reagents. Common Steps of Preliminary Operations Each analytical problem calls for a specific experimental set-up to be used in the preliminary operations in order to fill the gap between the raw sample and the analytical instrument used for measurements. The operations to be performed are those essential to assuring the maximum possible representativeness, integrity, sensitivity to the analytes, selectivity against inter- ferents and precision.In some instances (e.g., in vivo sensors), preliminary operations are minimal. Fig. 3 illustrates the more common analytical operations performed in order to suit the raw sample to the instrument used to measure the analytical signal (second stage of the analytical process). A detailed description is beyond the scope of this paper. However, a few brief comments are in order. Thus: (a), not all of them are required by every analytical problem; ( b ) , the stated order may vary between applications; (c), in many instances several operations are performed in a single step (e.g., simultaneous separation and analytical reaction; dissolution and decomposition of organic matter; sampling and preservation); (d), operations not shown in Fig.3 are occasionally included in the process; (e), solid raw samples call for a larger number of preliminary operations than do liquid and gas samples; (f), analytical separation techniques play a prominent role in this context as they indirectly enhance sensitivity and selectivity through precon- centration and interference removal, respectively; (g), measur- ing a mass or volume entails using an instrument to acquire analytical information that is essential to obtaining accurate final results, so it is not correct to assign this type of measuring operation exclusively to the second stage of the analytical process; (h), the variability and complexity of preliminary operations and the difficulties involved in their automation and globalization are self-evident.Mass-volume Instrument Fig. 3 Typical operations of the first stage of the analytical process278 ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 Automatic modules for < multi-sample treatment Robotic Process 0 analysers 0 I Non-chromatographic continuous separation I techniques Fig. 4 Recent developments improving on the fe n U / / intro2ction Direct introduction processes :atures of preliminary operations Recent Advances in Preliminary Operations The large number of hindrances inherent in preliminary operations, particularly those of a medium-to-high technical complexity, has fostered intense R&D activity aimed at minimizing them.It should be noted that most of the more recent developments, whether commercially available or otherwise, rely on current trends in science and technology including automation, miniaturization and the use of new materials and techniques. Fig. 4 shows some recently de- veloped approaches, most of which are available in commercial form, which, to the authors' minds, are anticipations of foreseeable future developments. Robotic stations334 are the most comprehensive approach to the automation of laboratory processes. Their greatest asset lies in the fact that they can implement some operations forbidden to most analysers (e.g., weighing, dissolution/ disaggregation, solid-liquid extraction, evaporation, solvent changeover, centrifugation, preparation of chromatographic vials) in a completely unattended fashion.As with computers, the current high price of robotic systems is bound to drop as their performance is boosted and their scope of application is further expanded. There is no doubt that robotic stations have a promising future , particularly in relation to analytical processes involving a large number of complicated preliminary operations. Non-chromatographic continuous separation techniques (e.g., sorption, extraction, dialysis, gas diffusion)' have so far solved a wide range of analytical problems associated with sample treatment and also, occasionally, collection. Their virtues (utmost simplicity and ease of automation) are not offset by their virtually exclusive applicability to liquid samples: few systems of this type can handle gas or solid samples.The development of physical and (bio)chemical sensors responding in a rapid continuous, reversible and precise manner to variations in the concentration of speciesG8 is another highly promising prospect in this context as it should allow preliminary operations to be dramatically simplified. Notwithstanding the thousands of papers published on this topic, only a few sensors have proved to be fit for tackling real problems and are commercially available (e.g., solid-state gas sensors). Much R&D effort is needed to develop sensors that can be used routinely. One major hindrance in this respect is the obvious difficulty involved in processing solid samples.Process analysers (on-line, in-line and non-invasive) are of great interest not only for industrial, but also for environmental and biotechnological applications, among others.' They mini- mize the need for preliminary operations which can be executed in a highly automated fashion. Because such operations must obviously be fairly simple, they occasionally do not guarantee the obtainment of quality results, which is also heavily dependent on accurate calibration. A number of treatment (heating, digestion, stirring, evap- oration, extraction) modules were developed in the last decade to handle several (4-24) samples simultaneously and automati- cally. While they do not encompass all possible preliminary operations, they do offer substantial savings in terms of labour and time, so they are being increasingly used in routine analyses.Attempts at simplifying preliminary operations on solid samples have so far been aimed at avoiding dissolution, which can be accomplished by introducing samples directly into the analytical system used (e.g., an atomic or indirectly, as slurries. Supercritical fluid extraction allows the rapid, efficient isolation of the analytes in a solid matrix located in an extractor. The use of unusual sources of energy including ultra- sound,12 microwave^'^ and laser radiation in automatic models allows samples to be more rapidly and efficiently attacked (dissolved, disaggregated) than do traditional systems used for this purpose. The adaptation of former industrial and preparative chemistry innovations to analytical applications has been facilitated by the inception of technically capable instruments.Such is the case with analytical lyophilizers for both certified reference materials and biological samples, l 4 and with super- critical fluid extractors''-"6 for direct processing of solid samples. Final Remarks Preliminary operations are dealt with much less frequently in both primary (papers) , secondary (monographs) and tertiary analytical literature (text-books) than are the other two stages of the analytical process (instrumental measurement and data acquisition and processing). Also, the literature on this topic, with few exception^,'^"^ is widely dispersed and unsystematic. However, handbooks dealing with analytical methodologies in detail do contain sizeable sections devoted to preliminary operations, which is proof of their undeniable significance. The reasons for this relative disregard of preliminary operations are obvious.R&D endeavours aimed at developing new or improving existing analytical techniques and chemo- metric procedures usually meet with more brilliant results. In addition, universal instruments (e. g., chromatographs and spectrometers) are much more profitable than specific appara- tus for each of the wide variety of preliminary operations with which analysts may be confronted.ANALYTICAL PROCEEDINGS, JUNE 1993, VOL 30 279 The influence of preliminary operations on the quality of analytical results and laboratory work is unquestionable. With this paper, the authors wish to express their own concern with this aspect of analytical chemistry, which is bound to become a priority topic in the near future.Extensive R&D endeavours must be made by both researchers and instrument manufac- turers in order to develop capable systems that facilitate accomplishment of the challenges faced by today’s analytical chemistry (i. e., the obtainment of more information of a higher quality by expending less material, time, labour and economic resources and taking smaller risks) which will redound to further social and economic progress. Current developments in this respect (e.g., sensors or supercritical fluid extractors) must be redoubled in order that automation, miniaturization and simplification reach all possible preliminary operations involved in routine analyses.References 1 Valcarcel, M., Fresenius’ J . Anal. Chem., 1992, 343, 814. 2 Valcarcel, M., and Luque de Castro, M. D., Automatic Methods of Analysis, Elsevier, Amsterdam, 1988. 3 Hurst, W. J . , and Mortimer, J. W., Laboratory Robotics, VCH, New York, 1987. 4 Hawk, G. L., and Strimaitis. J . , Advances in Laboratory Automation Robotics, Zymark Co, Hopkinton, USA, 1984, vol. I , 1985, vol. 11, 1986, vol. 111, 1988, vol. IV, 1989, vol. V and 1990. vol. VI. 5 6 7 8 9 10 11 12 13 14 1s 16 17 18 Valcarcel, M., and Luque de Castro, M. D., Non-chromato- graphic Continuous Separation Techniques, Royal Society of Chemistry, Cambridge, 1991. Valcarcel, M., and Luque de Castro, M. D., Lab. Robot. Autom., 1991, 3, 199. Valcarcel, M., and Luque de Castro, M.D., Analyst, 1993,118, 593. Janata, J . , Principles of Chemical Sensors, Plenum Press, New York, 1988. Nichols, G . D., On-line Process Analyzers, Wiley, New York, 1988. Scheeline, T., and Coleman, M., Anal. Chem., 1987, 59, 118SA. Miller-Ihli, N. J . , Anal. Chern., 1992, 64, 964A. Linares, P., Lazaro, F., Luque de Castro, M. D., and Valcarcel, M., J . Autom. Chem., 1988, 20, 88. Kingston, H. M., and Jassie, L. B., Introduction to Microwave Sample Preparation, American Chemical Society, Washington, DC, 1980. Izquierdo, A . , and Luque de Castro, M. D., J . Autom. Chem., 1990, 12, 267. Wenclawiak, B . , Analysis with Supercritical Fluids: Extraction and Chromatography, Springer-Verlag, Heidelberg, 1992. Tena, T . , Luque de Castro, M. D., and Valcarcel, M., La Extraccidn con Fluidos Supercriticos en el Proceso Analitico, Editorial RevertC, Barcelona, Spain, 1993.Smyth, M. R., Chemical Analysis of Complex Matrices, Ellis Honvood, Chichester, 1992. Anderson, R., Sample Pretreatment and Separation, Wiley , Chichester, 1987. A joint symposium of the North West Regions of the Analytical and Industrial Divisions and the Manchester Sections of the Royal Society of Chemistry and the Society of Chemical Industry. TOTAL QUALITY MANAGEMENT IN THE CHEMICAL INDUSTRY STRATEGIES FOR SUCCESS University of Salford September 22-24,1993 This symposium will cover the integration of TQM with other business strategies, including responsibl re, integrated pollution control, investing in people and human resource development. Sessions will also be included on excellence in other industries, TOM in R & D, current research in TQM and self-assessment schemes.The main two day symposium will be preceded by a half-day introductory session for those relatively new t o TQM. This session will cover quality standards, BS 5750, IS0 9000 and statistical process control. Programme The programme is not yet complete but will include the following: An Introduction t o TOM, J. Gilbert (Akzo Chemicals Ltd.); Statistical Process Control, G. R. Turner (FMC Process Additives Division); How t o Succeed with IS0 9000, D. Halsted (Zeneca Specialties); Tracking the Integration of Total Quality Management: Methods and Results, M. Wilcox, P. Gallagher, B. G. World Class Quality-Strategic Imperative, L. W. Allen (Anchor Chemical Group plc); Planning for Quality Leadership, J. M. Cullen (Rover Group Ltd.); The Corporate, Divisional and European Perspective, K. Miller (British Nuclear Fuels plc); Total Quality Management: Research into the Critical Factors, L. Porter and S. Black (Bradford University Self Appraisal: Practical Experience in Zeneca Specialties, G. J. Sharp (Zeneca Specialties); Investors in People, G. Waterhouse (Bradford and District Training and Enterprise Council); Human Resource Development, J. Gilbert (Akzo Chemicals Ltd.). Manchester M31 IRF. Dale and R. Boaden (Quality Management Centre, Manchester School of Management, UMIST); Management Centre); For further information contact Mrs. C. L. Sharp, Conference Secretary, 41 Exeter Road, Davyhulme,

ISSN:0144-557X    

DOI:10.1039/AP9933000276

出版商:RSC    

年代:1993

数据来源: RSC