ISSN:0144-557X    

DOI:10.1039/AP98219FX001

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

2 REFERENCE MATERIALS Anal. Proc. Reference Materials The following are summaries of six of the papers presented at a Meeting of the Analytical Division held on February 4th, 1981, at the Scientific Societies’ Lecture Theatre, Savile Row, London, W.1. Some Recent Work on Reference Materials at the National Physical Laboratory J. D. Cox Division of Chemical Standards, National Physical Laboratory, Teddington, Middlesex, T W11 OL W In this account of recent work on Certified Reference Materials (CRMs) at the NPL, the charac- terisation of chemical substances and materials by accurate measurement, both chemical and physical, is emphasised. Work on CRMs at Teddington began over 25 years ago as a result of a research programme on accurate measurement of the thermal properties of organic compounds.A prerequisite of that measurement programme was a stock of highly purified substances, which we prepared in-house and certified for purity.l It became our practice to offer for sale samples drawn from the same purified batches as were used for property measurement. The second phase of CRM activities began 12 years ago in response to a proposal from the Ministry of Agri- culture, Fisheries and Food that we should utilise our expertise in purification, for the preparation of pesticide chemicals as reference substances of certified high purity. The emphasis was initially on the analysis of commercial pesticides and pesticide formulations but interest was soon shown by environmental chemists. Pure samples of contaminants and metabolites were later included in the programme.By 1973 organic reference materials were being issued to a wide variety of customers, who made us aware of other unfilled needs for CRMs of various types. A decision was then made to launch a research programme into the provision of CRMs required by industry, with the Metrology and Standards Requirements Board of the Department of Industry as the major source of funds. This third phase of CRM activities has run from 1973 to the present but it will shortly give way to a fourth phase, when the resources available will be severely reduced as part of the: Government’s economy measures. From April 1981, most of the effort available will be used in sustaining NPL’s present range of CRMs, so that expansion of the range will be much slower than in the recent past.Some examples of groups of CRMs which have recently been developed by research at the NPL and are now, or will shortly be, available for distribution outside NPL, are now given. The tariff is given in a catalogue2 available free from NPL. In all of these examples CRMs are offered along with an NPL certificate, which thereby provides the user with traceability to national standards of measurement. In the field of chemical analysis the range of pesticide CRMs has been extended to 90 substances, and CRMs are offered for other areas of analysis where pure specimens are needed for reference purposes. For example, 16 metal - organic substances, which are oil-soluble, are offered for use in determining the metal content of lube oils by atomic-absorption or emission spectroscopy; these CRMs are certified for their metal contents.In the field of powder technology, the NPL offers eight particulate materials that are certified for specific surface area, by nitrogen adsorption measurements. Four of theseJanuary, 1982 REFERENCE MATERIALS 3 materials were certified by a collaborative campaign organised by the Society of Chemical I n d ~ s t r y , ~ and a set of four grades of a-alumina was characterised at NPL alone. Our general objective in certification is to make the measurements traceable, which in the context of specific surface area implies direct measurement in terms of square metres and grams. It can be argued that the existing certification in terms of nitrogen adsorption, involving a particular physical model of the adsorption process and an assumed value for the cross- sectional area of the nitrogen molecule, is not as directly traceable to national standards of measurement as would be desirable.Therefore, research into improved traceability is under way. In the case of one material, Sterling FT graphitised carbon black, the particles are sufficiently round, crack-free and homogeneous in size for a group of particles to be photographed in an electron microscope and subsequently counted in size-categories. The geometrically calculated mean surface area was found to agree with that derived from nitrogen adsorption measurements. It should be mentioned that because individual particles in a batch of powder are likely to differ from one another in their various physical properties, it is important to draw a representative sub-sample from a CRM, which in turn must constitute a representative sample of the whole CRM batch.For this reason NPL employs a rotating riffle machine to sub-divide a CRM batch into representative portions for sale. A recent development in the NPL programme concerns the provision of RMs for the calibration and checking of apparatus for measurement of low vapour pressures. It is becoming necessary for the physicochemical properties of new commercial chemicals to be measured as part of an assessment of their environmental impact. Vapour pressure is one of the parameters to be measured, which is not easy in the pressure range to lo2 Pa. Two solid CRMs certified for vapour pressure, naphthalene4 and hexamethylbenzene,5 are available, and two liquid substances, dibutyl phthalate and butyl 2,4-dichlorophenoxyacetate, have recently been evaluated for use at ambient temperatures (where the vapour pressure of the liquids is Pa).The method employed for measurements on the liquids was a novel one,6 and it has yet to be established whether the results show any method dependency. The final development area considered concerns CRMs certified for thermal properties ; here the nature and quality of the certification depend strongly on the likely end-use of the CRM. For example, for check measurements of the performance of melting-point apparatus, manual or automatic, a set of CRMs is offered, covering the temperature range 51-285 "C; these are certified for meniscus temperature and liquefaction temperature, for a stated rate of heating.For calibration of calorimeters, including differential-scanning calorimeters, we offer CRMs certified for enthalpy of fusion7 and final melting-point temperature at equi- librium, measured to 0.01 "C. For more precise thermometric measurements (to 0.001 "C) we offer triple-point cells containing water, benzene and phen0xybenzene.s A very recent development, not yet completed, concerns triple-point cells for accurate realisation of a temperature near 36.4 "C. Such cells are expected to have an important role in temperature measurement in enzymology and elsewhere in the biomedical field. The examples given above show that the NPL has been responsive to new, and often specialised, needs for calibrants and check substances.Suggestions regarding other unfilled needs for CRMs are always welcomed. 1. 2. 3. 4. 5. 6. 7. 8. References Cox, J. D., Proc. SOC. Anal. Chem., 1966, 3, 61. "Certified Reference Materials and Transfer Standards. A Catalogue of Items Available from the National Physical Laboratory, United Kingdom, and from the Community Bureau of Reference of the Commission of the. European Communities," Department of Industry, National Physical Laboratory, Teddington, 1979. Everett, D. H., Parfitt, G. D., Sing, K. S. W., and Wilson, R., J . APPZ. Chem. Biotechnol., 1974, 24, 199. Ambrose, D., Lawrenson, I. J., and Sprake, C. H. S., J . Chem. Thermodyn., 1975, 7 , 1173. Ambrose, D., Lawrenson, I. J., and Sprake, C. H. S . , J . Chem. Thermodyn., 1976, 8, 503.Hales, J. L., Cogman, R. C., and Frith, W. J., J . Chem. Thermodyn., 1981, in the press. Andon, R. J. L., and Connett, J. E., Thermochim. Acta, 1980, 42, 241. Cox, J. D., and Vaughan, M. F., MetroZogia, 1980, 16, 105.4 REFERENCE MATERIALS Anal. Proc. The Activities of the Community Bureau of Reference H. Marchandise Commissio.Pz of the Eurofiean Communities, Community Bureau of Reference, Rue de la Loi 200, B-1049 Brussels, Belgium The Community Bureau of Reference (BCR), which is a department of the General Directorate for Research, Science and Education of the Commission of the European Communities, is responsible for a programme on Applied Metrology and Reference Materials. The reason for this activity is that the Community has specific needs for measurements and in particular for measurements which should be made everywhere with the same level of accuracy.The need for the compatibility of the measurements is obvious from the point of view of trade. Whereas the uniformity of the basic units is ensured at the level of the metrology laboratories through Bureau International des Poids et Mesures (BIPM), there are many other measurements which present difficulties. Disagreements in the results of measure- ments could be the origin of trade barriers and it is precisely the objective of the Community to avoid trade barriers and to abolish them where they exist. The work on metrology started only at the end of 1979. It consists essentially in inter- comparisons of various kinds of measurements and in improvement of transfer standards.It has already proved very successful in establishing useful collaborations. The work on reference materials (RMs) started earlier (1973). The first reference materials were certified in 1977; at present 40 RMs are available and this number will increase to 70 by the end of 1981. To illustrate the importance for the Community, one could consider the case of a directive on pollution control that specifies maximum levels of carcinogenic compounds and of toxic elements (lead, cadmium, mercury). When the directive is adopted those who have to implement it would like to be sure that it is enforced in the same manner everywhere, that the measure- ments are made with equal care and that they are comparable. Otherwise, when industry is involved there could be distortions of competition.For use in environmental analyses we have prepared pure polycyclic aromatic hydrocarbons (carcinogenic) for the calibration of chromatography, and we are on the way to certifying toxic elements in plants, milk powder, blood, meat, liver, kidney, fly ash and sludges. We also hope to prepare reference materials for water analysis. We have in progress a project on reference gas mixtures, and we have RMs for flash point determination that can be useful in the context of the regulations on dangerous chemicals. On industrial products we have prepared or are preparing RMs for fertilisers, crystal glass, coke, etc. Of course, we welcome any request from industry that would point out specific needs to us. In this context it is perhaps of interest to mention that there is in our programme a group of highly skilled laboratories for analyses of non-metallic elements in metals, and as a result of the work done with their collaboration we have a series of metals (zirconium, titanium, aluminium, lead, copper, molybdenum and nickel) which are certified for oxygen content and sometimes for nitrogen content.We have also developed a programme on the properties of powders (particle size distribution, surface area, pore size). Finally, we have started working on reference materials for biomedical analyses. This field is in great need of reference materials on a large scale. All the work on analyses and measurements is carried out by laboratories of the member countries and, for some projects, with contributions from the laboratories of the Joint Research Centre of the Community.For the sake of brevity, details of the selection and preparation of the materials, the study of stability and homogeneity are omitted. The measurements are made by the best laboratories we can find in the Community. When the results are available, they are discussed in detail. Some of them are eventually rejected when reasons for inaccuracy have been identified as a result of the discussion of the methods. When the work is terminated, the draft report is submitted to the Certification Group, composed of experts from the member countries who have responsibilities in their respective countries for metrology or for reference materials. When all steps of the work have been found to be satisfactory the material is finally certified.Samples are available to anyone who requests them and are delivered with a certificate and a complete report. The objective is in principle the same as for metrology.January, 1982 REFERENCE MATERIALS 5 The final batch of reference materials is only one result of our work. The other important consequence is the benefit of the collaboration for all participants and the result of it is better compatibility and general improvement of the quality of the measurements. Lists of BCR Reference Materials certified before 1980 and in preparation are available from the author on request. NBS: Current Work and Future Plans in Reference Materials Stanley D. Rasberry Ofice of Standard Reference iwaterials, National Bureau of Standards, Washington, D.C. 20234, USA Standard Reference Materials (SRMs) have been produced, certified and issued by the NBS since 1906.These materials, with specific chemical or physical properties certified by the NBS, have found wide acceptance especially by industrial users needing to maintain or increase high levels of productivity. Today, 84 of the 100 largest manufacturers in the USA buy and use SRMs. Together with 10000 other users, they buy over 40000 SRM units per year at a total price of about $3.0 x lo6. The SRMs have a vital role in promoting industrial product- ivity by helping companies maintain : reliability and uniformity of materials in meeting specifications ; quality control in raw materials and produced goods or services ; and inter- changeability of materials and sub-components. In addition to the industrial customers, major SRM users also include federal and state governments, universities and non-profit research establishments, particularly for SRMs used in the areas of health, environmental protection, metrology and forensic science.Inter- national distribution of SRMs accounts for about 20% of the total. The last 14 years has seen the NBS-SRM programme greatly expand the nature and scope of its activities. In 1966, the NBS offered 559 different SRMs for sale, which were intended to serve the measurement needs of rather limited but important segments of the US economy. At that time, the majority of SRMs consisted of compositional standards (primarily certified for bulk constituents or purity), designed for use in quality control systems during mining and manufacturing of raw and processed industrial materials. The SRM inventory primarily consisted of metal bearing ores, metals and alloys, cement, glass and inorganic chemicals.By 1980, the SRM programme was considerably strengthened through the participation of 23 NBS technical divisions. This enabled the NBS to increase the number of SRMs to about 1000. Now included are SRMs in many new areas such as computer technology, fire safety, forensic science and radiopharmaceuticals, with particular recent emphasis on the two important areas of environmental and clinical measurements. The expansion of the SRM programme was accompanied by a corresponding expansion of the customer base as indicated by SRM sales.In the period 1969-80, SRM sales more than doubled to $3.0 x lo6. The number of units sold each year also increased steadily over this period, from 29600 in 1971 to approximately 40800 in 1980. Several con- siderations have helped shape these plans. From our perspective the four most important considerations or, as we call them, “external trends,’’ affecting the future of SRM production are outlined below. Table I summarises the types of SRMs now certified by the NBS. The scope of our future plans for SRM production is indicated in Table 11. 1. Increased Importance of Measurement Compatibility in a Technological World During the next decade, international SRM activities are expected to increase dramatically, including the joint development of multi-national certified reference materials and increased requirements for the use of reference materials to meet international standards.Great care must be taken to ensure that international standards, which require the use of reference materials, foster international trade rather than serve as a trade barrier. 2. Increased Cost of Raw/Processed Materials and Energy Recent increases in the cost of raw and processed materials, as well as the cost of energy, have resulted in the development of more stringent procurement specifications for materials6 REFERENCE MATERIALS TABLE I SRM TYPES CERTIFIED BY NBS Minerals Refractories Carbides Glasses Cements Trace elements Nuclear materials X-ray diffraction Isotopics Ion activity Mechanical and metrology Superconducting Freezing-points Melting-points Calorimetric Vapour pressure :a1 Radioactivity Anal.PYOC. Steels Steelmaking alloys Cast irons Cast steels Non-ferrous alloys Gases in metals High-purity metals Electron probe microanalytic Primary chemicals Clinicals Biologicals Botanicals Environmentals Industrial hygiene Metallo-organic compounds Fertilisers Ores Thermal conductivity Thermal expansion Thermal resistance Thermocouple materials Magnetic Optical Gas transmission Permittivity Reference fuels Resistivity Rubber materials Computer tapes Sizing standards Colour Photographic Surface flammability Smoke density sold in national and international commerce. SRMs are being increasingly used to provide a basis for the arbitration of disputes between producers and users of materials, as well as in the traditional role as part of production quality control systems.3. Increased Development of New Materials in High-performance Applications The development and use of materials for high-performance applications has increased considerably in recent years. This trend is leading to the development of a wide variety of new high-performance materials serving many diverse uses. Examples range from high- temperature, corrosion-resistant alloys used in aircraft turbine blades to plastic foams used in building construction. Many of these high-performance materials are used in critical applica- tions where their failure in service could result in serious safety hazards. This has led to the development of rigidly enforced trade specifications and to the need for reliable production quality control, which in turn has resulted in increased demand for SRMs.4. Increased Development and Implementation of Government Imposed Health and Safety Regulations The trend to greater federal, state and local government regulation of the environment, health care system and transportation systems is also resulting in intense demand for new SRMs. Effective development and enforcement of regulations depend on the availability of TABLE I1 PROJECTION OF DEMAND FOR SRMs FROM NBS IN 1982-86 SRM category Metals . . .. .. Chemicals/rubber/plas tic Nuclear . . .. .. Non-metalslglass . . .. Radioactivity . . .. Engineering . . .. Environmental gases . . Environmental liquidslsolids Health . . .. .. Science/metrology . . SRM Current renewals? inventory* 1982-86 ..316 53 .. 74 15 . . 105 23 .. 30 10 .. 156 71 .. 146 146 .. 54 145 .. 35 21 .. 42 33 . . 105 20 Current SRMs needed discontinued NewSRMs t o b e 1982-86 1982-86 80 108 30 10 24 32 20 0 20 38 46 58 25 10 30 6 25 6 10 48 Inventory needed 1986 288 94 97 50 138 134 69 59 61 67 Total . . .. .. .. 1063 537 310 316 1057 * As of April 30th, 1980. 7 Includes a number of multiple renewals in the Radioactivity, Engineering, Environmental gases and Health categories.January, I982 REFERENCE MATERIALS 7 accurate, reliable and compatible measurement systems. The NBS response to this trend has been the establishment of a number of programmes aimed in part at providing a reliable measurement base for regulatory agencies. These include the Environmental Measurements Programme, Nondestructive Evaluation Programme, Resource Recovery Programme, Recycled Oil Programme and the Nuclear Safeguards Programme.A programme for Measure- ments and Standards for Nuclear Waste Management has also been recently proposed. All of these programmes will require the development of new SRMs to serve as measurement tecli- nology transfer mechanisms. Geological Reference Materials with Particular Emphasis on Multi-element Trace Analysis Alan R. Date Institute of Geological Sciences, 64-78 Gray’s I n n Road, London, WClX 8NG Geochemical analysis, in common with many other branches of analytical chemistry, suffers from an absence of true standards, i.e., samples with accurately fixed element concentrations. The significance of this failing was not recognised for many years, probably as a result of the traditional lack of liaison between geologist and chemist and the latter’s faith in the absolute methods of analysis used in classical schemes.Following the pioneering work of Larsen in 1938,l which demonstrated the poor agreement achieved by two to four “better” analysts for six amphibole (rock-forming mineral) samples, a collaborative programme of analysis was initiated in the late 1940s by the Geology Depart- ment of the Massachusetts Institute of Technology, the Geophysical Laboratory and the United States Geological Survey. This project, intended initially to demonstrate that spectrographic analysis could match the precision and accuracy attained by wet chemical methods, involved major element determinations for the first two “standard” rocks, the granite G-1 and the diabase (dolerite) W-1. The results submitted by 34 chemists from 25 laboratories in 10 countries were published in 1951,2 and proved to have much more far reaching influence.The wide variation shown for most major elements was very disturbing, the data for Na20 and K20 being particularly instructive. Soda was found to vary from 2.61 to 3.95% in G-1 and from 1.61 to 2.62% in W-1, and potash was found to vary from 3.85 to 6.88% in G-1 and from 0.41 to 1.30% in W-1. Major element data are normally quoted as oxides to provide summation to 100% for the commonly reported components, and because one scheme of silicate rock classification depends on the relative proportions of normative minerals (a theoretical combination determined from chemical data), in which soda and potash play a crucial role. The ranges shown for G-1 encompassed almost all the world’s known granite variation, although several distinct groups may be identified under the microscope. Soda and potash may be particularly difficult major element determinations, but the position with trace elements is expected, by definition, to be far worse. The term trace in this context includes any element normally reported in the parts per million range or below.The period following publication of the first report on G-1 and W-1 may be considered in two parts. Although a few trace element results were included in this first report in 1951, during the succeeding 15-year period the number of determinations proliferated, and several additional reports were produced.In the process, a few papers drew attention to the possibility of between-bottle variation for some trace elements in G-l,3-5 and in 1967 Kleeman6 concluded that both rocks as originally prepared (G-1, -80 mesh, 180 pm; W-1, -100 mesh, 150 pm) were too coarse to serve as reference standards. In parallel with the work on G-1 and W-1, other reference materials became available, notably the syenite Sy-1 (in conjunction with a sulphide ore, Su-1) from Canada,’ and a series of silicate rocks from France.* By 1962 the supply of G-1 was almost exhausted and a suitable replacement was prepared. By the end of 1966, Flanagan and Gwyn could record more than 50 geological reference materials from 12 sources, although many were of very limited application. Their paper,g published in Geoclzimica et Cosmochirnica Acta in 1967, was followed by an Editorial Notice announcing The granite G-2 and five other silicate rocks were issued in 1964.Anal.Proc. future limits on the publication of reference material papers. In a very broad sense it may be said that the first 15-year period was characterised by unexpurgated data compilations, with statistical analysis taking account of all reported results. The initial impetus provided by the demands of igneous petrology, which resulted in a preponderance of silicate-based reference materials, was continued into the second 15-year period, with the issue of further reference materials from the USA, Canada and France, and new reference materials from South Africa, Japan, Scandinavia and the Eastern Bloc countries.Reference to the range of silicate rocks currently available may be obtained from a recent review paper by Abbey.lo It is interesting to note at this point that in the first compilation of datall for the six USGS rocks issued in 1964, the variation for K20 in G-2 was found to be worse than that for G-1 almost 20 years earlier. As a result of the difficulty of dealing with such data sets, the tendency during the second 15-year period has been towards the “select laboratories” concept as a means of limiting the initial spread of results subjected to statistical analysis. The increasing importance of geochemical prospecting during this period led to the intro- duction of six geochemical exploration reference materials from the United States Geo- logical Survey12 and four soils from the Canadian Certified Reference Materials Project .l3 Stream sediment and lake sediment reference materials are promised (CCRMP) .The range of sample types widened further to include reference materials of a more specialised nature, such as the USGS manganese n0du1es.l~ The number of ore standard reference materials increased dramatically, with contributions from several sources including the Institute of Geological Sciences.15 There is still very poor coverage for marine sediments and for sedi- mentary rocks in general, although the field is represented to a certain extent by standard reference materials from the National Bureau of Standards in the United States and the Bureau of Analysed Samples in the UK.The lead taken by Geochimica et Cosmochimica Acta in 1967 in limiting the number of reference material papers accepted for publication resulted in the formation of Geostandards Newsletter in 1977,16 and this journal should be consulted in any search for geological reference materials. 8 REFERENCE MATERIALS Multi-element Trace Analysis The first compilation of data for the USGS geochemical exploration reference materials (GXRs)12 serves as an illustration of the difficulty in standardising geological materials for trace elements. Although atomic-absorption spectrometry would be expected to provide reliable data for copper, the range of values reported for GXR-4 by this technique is 5- 8800 pg g-l. This also illustrates the danger in defining a method of analysis by the technique of measurement only; such anomalies owe more to sample attack.Although the user of such reference materials must be aware of the problems associated with data compila- tions, he may usually rely on the work of others in selecting a series of suitable reference materials for a particular project. In the first compilation of data for the six USGS rocks issued in 1964,11 results for the ten most frequently determined trace elements (Ba, Co, Cr, Cu, Ni, Pb, Rb, Sr, Zn, Zr) were obtained by six techniques, including spectrophotometry. Although atomic-absorption spectrometry and isotope dilution mass spectrometry may be developed for simultaneous determination of several elements, only atomic-emission spectrometry, X-ray fluorescence spectrometry and neutron-activation analysis can be considered true multi-element methods, offering simultaneous determination of more than ten elements.In a review paper in 1977, RubeSkal’ used this approach to identify a trend away from traditional methods of analysis (including atomic-emission spectrometry) towards X-ray fluorescence spectrometry, neutron- activation analysis and isotope dilution mass spectrometry. The growth of inductively coupled plasma techniques is expected to reverse this trend, and the advent of laser ablation allied with plasma emission spectrometry or plasma source mass spectrometry will create demand for a wider range of reliable solid reference materials for geochemical analysis. Reference Material Selection The author’s recent interest in geological reference materials is related to the multi-element analysis of Scottish stream sediment samples submitted under the Institute’s RegionalJanuary, 1982 REFERENCE MATERIALS 9 Geochemical Reconnaissance Programme.The samples cover a composition range similar to silicate rocks, with extremes for iron and manganese. Up to 30 elements are determined by d.c. arc direct-reading emission spectrometry. International geological reference materials are too valuable to be used for calibration purposes, but are used to monitor long-term precision and accuracy. In selecting a series of reference materials to cover a suitable range of concentration for all 30 elements, compilations of data such as the series published by Abbey have been consulted.The “1979” edition of Abbeylo may be used to illustrate potential problems in selection. Abbey lists reported values for major, minor and trace elements in all reference materials that may be applied to silicate analysis. The data he considers reliable (using various criteria, including the “select laboratories” concept) are shown unqualified. Other reported values are shown with question marks. The ratio of the first group to the second is the degree of acceptability. The selection of elements with acceptability in excess of 50% includes eight of the ten most frequently reported trace elements. Of these, zinc and chromium have similar determination ranges by direct reader, and matching reported ranges of concentration. A selection of the more readily available international reference materials provides good cover for zinc, while many have chromium below our detection limit (10 pg g--.l) or above the required top standard (2000 pg g-l).Reference materials with intermediate concentrations of chromium, the Canadian MRG-1 (450 pg g-l) and the French BR (38Opgg-l) are of great use. The second group of elements, with acceptability ratios below SO%, includes tin with a reported range of concentration from 1.4 to 1900 pgg-1. Twenty-five reference materials, from a total of thirty, have concentrations below our detection limit (10 pg g-l), and only two of the remaining five have reliable values at 11 and 1900 pg g-1. A similar exercise for rare earth data, and for environmentally important elements normally reported in the ng g-l range, shows that the recent upsurge in interest in these elements has not been matched by an increase in the number of reliable geological reference materials.Synthetic Reference Materials In the first report on G-1 and W-1 it was suggested that analytical error could be separately identified from sampling error by the use of a synthetic rock standard. The author is unaware that such a standard was ever prepared. The nearest approximation to such a scheme is the series of glasses prepared by Corning Glass for private use by the US Geological Survey, and reported in 1976.lS Glass reference materials are available from the US National Bureau of Standards and from ANRT in France (VS-N).IS Each contains a very wide range of trace elements in nominally identical concentrations, and in the case of VS-N the concentrations are unusually high.In an attempt to avoid calibration standards prepared by solid dilution, the author developed a method for the preparation of synthetic silicates in which the element concentra- tions could be changed at will. This method, a development of the technique used to manufacture starting materials for experimental petrology, was described in detail in The AnaZyst in 197tX2O It involves the preparation of two solutions, the first a solution of tetra- ethyl orthosilicate in ethanol, to render it miscible with the second solution, which is an aqueous phase containing the remaining major, minor and trace elements. The two solutions are mixed, ammonia is added and a “flash hydrolysis” occurs, producing a solid “gel” with the trace elements entrained in a silicate framework.The gel is dried, ignited a t a tempera- ture sufficiently high to cause slight sintering, which removes its hygroscopic properties, and ground lightly to reduce it to a convenient powder form ((300 mesh, 53 pm). The range of major element reference materials currently used in this work is shown in Table I. Only two major element reference materials from an earlier series have been independently analysed.20 The application of the current series as calibration standards in d.c. arc direct-reading emission spectrometry is illustrated in Fig. 1, a calibration graph showing magnesium channel count against concentration. There is good agreement between synthetic reference materials and international standard rocks.The trace element reference materials have been prepared in a limited number of matrices. The series currently used is shown in Table 11. With the exception of molybdenum, the They have very limited use in geochemical analysis.10 REFERENCE MATERIALS TABLE I MAJOR ELEMENT “GEL” SILICATE REFERENCE MATERIALS Major element data in yo. Anal. PYOC. Reference material Element 50, .. .. M203 . . .. TiO, .. . . Fe,O, . . .. JIgO . . .. CaO .. .. li,O . . .. Sa,O . . .. M n . . .. DR ES FR 70.0 40.0 30.0 18.0 14.4 5.72 2.0 1.0 0.5 2.0 1.0 5.0 2.0 20.0 45.0 2.0 20.0 10.0 2.0 - 0.2 2.0 2.0 2.0 - 1.0 1.0 GR 60.0 1.83 1 .o 20.0 1 .o 1 .o 10.0 2.0 2.0 HS IR 65.0 65.0 16.0 14.0 4.0 2.0 2.0 5.0 0.5 5.0 10.0 5.0 0.5 2.0 2.0 2.0 JR 45.0 29.7 0.5 0.5 1.0 20.0 0.5 2.0 0.5 KR 70.0 9.07 0.1 1 .o 0.5 0.5 1 .o 2.0 10.0 LR MR’ 50.0 70.0 32.8 3.38 0.2 0.2 10.0 0.5 2.0 10.0 2.0 1 .o 1.0 5.0 2.0 2.0 5.0 - trace elements appear to be present at the intended concentrations.Several trace element reference materials have been analysed by independent methods. Data for the current involatile element series (with high levels of zinc) are shown in Table 111. The author is confident that the variation shown is analytical rather than compositional in nature. The use of these reference materials as calibration standards is illustrated in Fig. 2, a graph showing lithium channel count against concentration. Here too there is good agreement between synthetic and natural reference materials. This method of reference material preparation is limited in several respects: (a) some elements, e.g., Be and B, are unstable under the silicate hydrolysis procedure, and have to be added by liquid - solid dilution; ( b ) some elements, e.g., Sn and V, are dissolved with the aid of standardised NaOH solution, which assumes the presence of sodium in the matrix; and (c) the technique is currently limited to small amounts ((5OOg).For silicate analysis, however, it allows one to prepare standard reference materials to order. TABLE I1 TRACE ELEMENT “GEL” SILICATE REFERENCE MATERIALS Trace element data in pg g-l. Element Li .. .. JSe . . .. I3 . . .. v . . .. Cr .. .. 3tn . . .. co . . .. Xi .. .. c u . . .. Zn .. .. Ga . . .. Ge . . .. Rb . . . - Sr . . .. Y .. .. Zr .... No* . . .. Sn . . .. Ha . . .. La . . .. Pb .. .. Bi .. .. +g .. .. f DRlO DR09 DR08 2 50 5 5 2 20 - 50 25 - 50 - 200 - 20 - 50 - 50 - - - - - - - 20 10 2 2 loo - 2: 5 20 - 50 100 - 20 - - 500 - 33 12 5 2 20 - 50 200 - 20 - 20 200 50 20 - 50 - - - - - - - DR07 100 - - 50 100 500 50 100 200 - - - 200 50 1000 - - 500 50 500 - Reference material DR06 DR05 10 200 10 - 100 - - 100 - 200 - 1000 - 100 - 200 100 - 50 500 10 10 A - - 100 I - 500 - 100 - 2 000 81 10 - 100 - - 1000 - 100 100 2000 100 - - DR04 20 20 200 - - 200 100 20 20 20@ - - 176 20 200 - 200 200 DR03 500 - - 250 500 2 000 200 500 1000 - - - 1000 200 5 000 - - 2 000 200 5 000 DR02 50 50 500 - - 500 200 50 50 500 - - 44 1 50 500 -- - 500 500 -7 DRO 1 1000 - - 500 1000 5 000 500 1000 2 000 - - - 2 000 500 10 000 - - 5 000 500 10000 - * Determined by solution absorptiometry (found to be in error).January, 1982 REFERENCE MATERIALS TABLE I11 ANALYSIS OF DR MATRIX TRACE ELEMENT REFERENCE MATERIALS BY THREE TECHNIQUES Techniques used : a, atomic-absorption spectrometry (P.T. S. Sandon, Geochemical Division, IGS) ; b, atomic-emission spectrography (B. A. R. Tait, Geochemical Division, IGS) ; and c, instrumental neutron-activation analysis (J. Herrington, AWRE, Aldermaston). 500 f 8 2 200 - C lu 11 - - DR09 DR07 DROB DR03 DROl (-'-, r-----. -7 & - Expected, Found, Expected, Found, Expected, Found, Expected, Found, Expected, Found, Element p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. co .. .. 20 158 50 48a 100 1 0 s 200 200% 500 5408 2 i c 5SC ll0C 220c 510C 15b 44b lOOb 220b 433a Cr .... 50 54c 100 l0OC 200 180C 500 460C I000 9ooc .53oa 1000 1070a 2000 2050a 5000 5200" >In . . .. 200 -b 200a 500 190b 600b 95Ob 2150b 19oc 520c 980C 2 oooc 4 7OOC 540a 1000 1070a Ni .. .. 50 508 100 100a 200 2108 500 v .. .. 25 23b 50 45b 100 80b 250 200b 500 590b < 5OC 61C 120c 290C 590C Zn .. .. 100 1ooa 200 210a 500 5ooa 1000 1050a 2000 2000a 21oc 29oc 560C 1 oooc 190oc The author is grateful to Mrs. Dawn Hutchison and Miss E. Waine for critically reading the manuscript. "This paper is published with the permission of the Director, Iistitute Geological Sciences (NERC) . 999 I -1 100 50 20 .* 'I .* 0 * 0 $ * * * +i I I I I I 0.5 1.0 2.0 5.0 10 20 5p fl 999 7 8 ** I I I of 2.0 5.0 10 20 50 100 200 Magnesium concentration, % MgO Lithium concentration/vg g-' Fig. 1. Use of major element reference materials as calibration standards. 0, syn- Fig. 2. Use of trace element reference materials thetic reference (mean of three) ; and as calibration standards. Symbols as in Fig. 1. *, international standards. Data from Abbey.lo 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Larsen, E. S., Jr., A m . J . Sci., 1938, 35, 94. Fairbairn, H. W., Schlecht, W. G., Stevens, R. E., Dennen, W. H., Ahrens, L. H., and Chayes, F., Stevens, R. E., Niles, W. W., Chados, A. A., Philvy, R. H., Leininger, R. K., Ahrens, L. H., Fleischer, Butler, J. R., and Thompson, A. J., Geochim. Cosmochim. Acta, 1962, 26, 516. Butler, J. R., and Thompson, A. J., Geochim. Cosmochim. Acta, 1962, 26, 1349. Kleeman, A. W., J . Geol. SOC. Aust., 1967, 14, 43. Canadian Association for Applied Spectroscopy, Report of Nonmetallic Standards Committee, Roubault, M., de la Roche, H., and Govindaraju, K., Sci. Tewe, Nancy, 1966, 11, 105. Flanagan, F. J., and Gwyn, M. E., Geochim. Cosmochim. Acta, 1967, 31, 1211. Abbey, S.. Geostand. Newsl., 1980, 4, 163. Flanagan, F. J., Geochim. Cosmochim. Acta. 1969, 33, 81. Allcott. G. H., and Lakin, A. W., "Statistical Summary of GeochEmical Data Furnished by 85 US Geological Survey, Open U.S. Geol. Surv. Bull., No. 980, 1951. M., and Flanagan, F. J., U.S. Geol. Surv. Bull., No. 1113, 1960. A+@ Spectrosc., 1961, 15, 159. Laboratories for Six Geochemical Exploration Reference Samples, File Report. Denver. Colo., 1974.12 REFERENCE MATERIALS Anal. Proc. Bowman, W. S., Faye, G. H., Sutarno, R., McKeague, J. A., and Kodama, H.. Geostand. Newsl., Flanagan, F. J., and Gottfried, D., U.S. Geol. SUYV. Prof. Pap., No. 1155, 1980. Lister, B., Trans. Inst. Min. Metall., 1977, B86, 133. Geostand. Newsl., 1977, 1, No. 1. RubeSka, I., Geostand. Newsl., 1977, 1, 15. Myers, A. T., Havens, R. G., Connor, J. J., Conklin, N. M., and Rose, H. J., Jr., U.S. Geol. Surv. de la Roche, H., and Govindaraju, K., Analusis. 1973, 2, 59. Date, A. R., Analyst, 1978, 103, 84. 13. 14. 15. 16. 17. 18. 19. 20. 1979, 3, 109. Prof. Pap., No. 1013, 1976.

ISSN:0144-557X    

DOI:10.1039/AP9821900002

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

57 ANALYTICAL DIVISION DIARY The Analyst January, 1982 NOW AVAILABLE SEPARATELY The Analyst is an international journal of high repute containing original papers on the theory and practice of all aspects of analytical chemistry. It is of interest to workers in a wide range of fields including pharmaceuticals and drugs, environmental analysis, air and water pollution, food analysis, forensic analysis, clinical analysis, metals and metallurgy, pesticides, agrochemicals, animal feedstuffs, water analysis. All modern techniques are covered with a high proportion of papers on atomic absorption and related spectroscopic techniques, chromatography and electrochemical methods. The Analyst provides complete coverage of all major developments and includes papers with a strong practical bias as well as papers of a theoretical nature. For the first time since I954 The Analyst is available separately and does not have to be subscribed to in conjunction with Analytical Abstracts. 12 issues per annum (plus index) 1982 Subscription: UK 585.00, US $201.00, Rest of World 590.00 Orders should be sent to : The Royal Society of Chemistry, Distribution Centre, Blackhorse Road, Letchworth, Herts. SG6 1HN

ISSN:0144-557X    

DOI:10.1039/AP98219BX003

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

12 REFERENCE MATERIALS Anal. Proc. Standards for the Analyst C. A. Watson Hopkin and Williams Ltd., P.O. Box 1, Romford, Essex, RIM1 1HA Analytical chemistry is essentially a comparative science and is always dependent on the availability of suitable standards. Such standards may be fundamental, e.g., balance weights, physical, e.g., photometric, or chemical, e.g., elemental. Fortunately, some procedures are stable and frequent recourse to standards is not always necessary. This is particularly true of fundamental standards, but many procedures are dependent on the availability of chemical or physical standards, some examples of which are discussed below. Standards for Elemental Analysis The first materials that were prepared commercially, specifically as a range of standards, were for elemental analysis.This exercise was undertaken during the 1950s following recom- mendations of the Microchemical Methods Group of the Society for Analytical Chemistq7.l The objective was to make available a range of substances that were suitable for checking procedures for organic elemental analysis, which was developing rapidly at that time with increasing use of the oxygen-flask combustion method, more sophisticated combustion furnaces and, eventually, automatic analysers. Criteria for such a range were straightforward and can be summarised as follows. (a) Pwity There was no point in including materials that would not have an assay of at least 99.5% by the various methods for which the standard would be employed. As microchemical techniques have improved this figure may no longer be adequate, but no requests have been received to indicate that more stringent assay limits are needed, although the range was updated following a further report from the Microchemical Methods Group. (b) Stability It was therefore essential that, as each pack is likely to be in use for some years, its composition should remain constant.(c) Range of elements The original range of about 35 items included compounds containing 8-94y0 of carbon and 1 .5-Sy0 of hydrogen. Other elements were less comprehensively covered, but for the common elements such as nitrogen, oxygen and chlorine the coverage was still extensive. The initial range of materials commercially available was twice modified, following the recommendations of the Microchemical Methods G r o ~ p , ~ , ~ but still consisted of about 35 com- pounds.Unfortunately, the decreasing amounts of standard required for each check analysis has meant that consumption of these materials has fallen markedly in recent years, leading to a lack of commercial viability for many of the items. As a result, the range of commercially available materials had fallen to 22 items by 1976 and has since dropped to 17, i.e., less than half of the original range. Elemental standards are used in small amounts, often in a “control chart” situation. Volumetric Standards During the period from 1960 to 1975 the Standards Sub-committee of the Analytical Methods Committee studied a number of substances for their potential use as volumetricJanuary, 1982 REFERENCE MATERIALS 13 standards.This work involved not only purification of many of the substances, but also the development of suitable analytical procedures, and is described in a number of reports.*-8 Certain of the substances that were studied by the Sub-committee, together with a few other compounds that could be produced in a high state of purity, and for which assay methods of the required accuracy and precision were available, were made available on a commercial basis in 1960 by Hopkin and Williams. The range known as PVS consisted of potassium hydrogen carbonate, later replaced by sodium carbonate, benzoic acid, sulphamic acid, potassium iodate and potassium dichromate. The preparation of these substances, which were required to have an assay of 100.00 0.02% and to have no impurities detectable by normal methods of analysis that could affect the assay, presented considerable challenges, both in manufacture and analysis.Success was achieved only with careful selection of starting materials, operation under dust-free conditions and constant monitoring of each stage of the process. Many of the processes, such as rapid filtration on fine sintered-glass filters, were not easily scaled up, limiting batch sizes to only 1-2 kg. This resulted in a cost that was of the order of 100 times that of an equivalent AnalaR grade for a substance with an assay tolerance typically five times better, i.e., 100.00 & 0.02y0 rather than 100.0 &- 0.1%. This high cost proved to be commer- cially unacceptable, as analysts were not prepared to pay the premium price and the range has now, unfortunately, been withdrawn.Volumetric Solutions From the commercial point of view, this is the one area where preparation of standard sub- stances can be regarded as successful. Standard volumetric solutions and buffer solutions have long been available and, except for stability problems in a few instances, present little difficulty in manufacture or analysis. Recently these ranges have been extended by the addition of solutions of metals and anions, usually at a level of 1000 p.p.m. The main problem in producing such a range, once a suitable matrix has been selected, is ensuring that the solutions are adequately stable in the type of pack in which they are to be sold. In order to carry out stability trials and to monitor normal production, assay methods had to be developed.These ranged from simple EDTA titrations, in instances where interfering elements could be shown to be absent in significant amounts by spectroscopic means, to differential spectroscopic methods, which could achieve the required precision, i.e., about &O.lyo. It is worth noting that, although in many instances these solutions can be prepared by weighing a pure substance and exact volumetric dilution, this route leads to unacceptable costs and in any event assay methods are still required for stability testing. Other developments in this area include concentrated standard solutions and the investiga- -tion of multi-element solutions. These developments have been necessitated by the tremend- ous advances in ICP - OES, which is essentially a multi-element technique applied to solutions.There are still, however, considerable problems in this area, not least of which is the choice of element mixtures and their indiyidual concentrations, which arise owing to the very different ratios of elements which are found in different areas of analysis. Incompatibility of various ion mixtures in some matrices is also a problem, as is the development of test methods of the required precision. The apparently simple solution to these difficulties is to prepare individual solutions for each user’s needs by using pure substances and volumetric dilution; however, although this approach has been tried by at least one supplier, costs are very high and com- mercial viability is extremely doubtful.P ho top hysical Standards Many laboratories rely on the output of ultraviolet - visible spectrophotometers or spectro- fluorimeters, but do not have the facilities for the regular preparation of standards to ensure accurate and reproducible instrument operation. Two possibilities exist for overcoming this problem: the use of standard solutions and the use of solid or sealed standards. The former approach is straightforward from the supplier’s point of view, but is of limited use to the user because storage conditions are often critical, many solutions are unstable to the normal labora- tory atmosphere and many of the test substances are toxic or harmful in solution. The second14 REFERENCE MATERIALS Anal. PYOC. approach presents a number of manufacturing difficulties, but it has been possible to make available a series of solid solutions of fluorescent compounds that are suitable for standardisa- tion and checking for wavelength accuracy and resolution of spectrofluorimeters.These are available in two series, the contents of which are listed together with their general. properties in Table I. Set A is intended for the laboratory employing many different methods at a wide range of wavelengths and set B is more suitable for the laboratory engaged in quantitative analysis over a more restricted range. TABLE I STANDARDS FOR SPECTROFLUORIMETRY Substance Excitationlnm Type Anthracene and naphthalene . . Ovalene .. .. .. .. Perylene . . .. .. .. 7,8-Benzoquinoline . . .. Coronene .. .. .... Rhodamine . . .. .. Triphenylene . . .. .. Tetraphenylbutadiene . . .. p-Terphenyl . . .. .. Compound 610 . . .. .. 296 350 396 320 340 485 310 360 300 400 Mixed Sharp Mixed Sharp Sharp Broad Sharp Broad Broad Broad Two series of photometric standards in sealed 10-mm quartz cuvettes have been made avail- able for checking the performance of spectrophotometers. One of these consists of five solu- tions of potassium dichromate and a blank, which are intended primarily for checking photo- metric accuracy. This may be achieved either by standardisation using a high-accuracy spectrophotometer, or by checking the peak - trough ratios. The solutions are prepared at pH 2.8 in perchloric acid, so the data of Burke et aL9 can be used to calculate the appropriate ratios, some of which are given in Table 11.The spectra of a typical set of standards is illustrated in Fig. 1. TABLE I1 CALCULATED ABSORBANCE RATIOS FOR POTASSIUM DICHROMATE SOLUTIONS (pH 2.9 IN PERCHLORIC ACID) r----h Ratio v A 850 - A,,, & A, A313 Cuvette No. 2 2.215 2.962 2.544 3 2.215 2.970 2.551 4 2.215 2.978 2.558 5 2.215 2.986 2.565 6 2.215 3.002 2.578 The second set of standards consists of one potassium dichromate solution, which can be used as above, a lithium carbonate solution to check for stray light below 225 mm, a cuvette containing benzene vapour to check for instrument resolution (both optical and electronic in the case of a scanning instrument) and a samarium perchlorate solution, which has 28 peaks at known wavelengths between 230 and 560 mm, to be used for wavelength calibration.Photometric standards prepared in cuvettes have advantages over glass filters in that the geometry is the same as that of a normal sample and it is possible to prepare standards with sharp peaks for wavelength and resolution calibration. It should be mentioned, however, that although these standards are now available commercially, their preparation is not easy andJanuary, 1982 REFERENCE MATERIALS 15 considerable effort has gone into finding suitable filling and sealing procedures that ensure long-term stability. Conclusion Reagent suppliers have been active in the field of standards and reference materials for many years, but this activity has been of limited effectiveness for two reasons. Firstly, it has proved very difficult to ascertain exactly the users’ requirements, and secondly, the preparation of such substances is time consuming and therefore expensive.The continuing supply of standards and reference substances will depend upon establishing effective communication between the potential user and the supplier, and acceptance by users that standards and reference materials are costly to prepare. 1. 2. 3. 4. 5. 6. 7. 8. 9. 200 300 4 Wavelengthhm 0 Fig. 1. Spectra of potassium dichromate solu- Experimental ratios for instrument used to tions. record spectra : Cuvette A,,, A, No. A313 A313 A313 2 2.236 2.868 2.461 3 2.216 2.853 2.412 4 2.216 2.849 2.378 5 2.217 2.922 2.483 6 2.215 2.964 2.523 References Microchemistry Group, Analyst, 1953, 78, 258. Microchemistry Group, Analyst, 1962, 87, 304. Microchemical Methods Group, Analyst, 1972, 97, 740.Analytical Methods Committee, Analyst, 1965, 90, 251. Analytical Methods Committee, Analyst, 1967, 92, 587. Analytical Methods Committee, Analyst, 1975, 100, 675. Analytical Methods Committee, Analyst, 1977, 102, 955. Analytical Methods Committee, Analyst. 1978, 103, 93. Burke, R. W., Deardorff, E. R., and Menis, O., “Liquid Absorbance Standards,” NBS Special Publication No. 378, National Bureau of Standards, Washington, D.C., 1973.January, 1982 REFERENCE MATERIALS 15 considerable effort has gone into finding suitable filling and sealing procedures that ensure long-term stability. Conclusion Reagent suppliers have been active in the field of standards and reference materials for many years, but this activity has been of limited effectiveness for two reasons.Firstly, it has proved very difficult to ascertain exactly the users’ requirements, and secondly, the preparation of such substances is time consuming and therefore expensive. The continuing supply of standards and reference substances will depend upon establishing effective communication between the potential user and the supplier, and acceptance by users that standards and reference materials are costly to prepare. 1. 2. 3. 4. 5. 6. 7. 8. 9. 200 300 4 Wavelengthhm 0 Fig. 1. Spectra of potassium dichromate solu- Experimental ratios for instrument used to tions. record spectra : Cuvette A,,, A, No. A313 A313 A313 2 2.236 2.868 2.461 3 2.216 2.853 2.412 4 2.216 2.849 2.378 5 2.217 2.922 2.483 6 2.215 2.964 2.523 References Microchemistry Group, Analyst, 1953, 78, 258.Microchemistry Group, Analyst, 1962, 87, 304. Microchemical Methods Group, Analyst, 1972, 97, 740. Analytical Methods Committee, Analyst, 1965, 90, 251. Analytical Methods Committee, Analyst, 1967, 92, 587. Analytical Methods Committee, Analyst, 1975, 100, 675. Analytical Methods Committee, Analyst, 1977, 102, 955. Analytical Methods Committee, Analyst. 1978, 103, 93. Burke, R. W., Deardorff, E. R., and Menis, O., “Liquid Absorbance Standards,” NBS Special Publication No. 378, National Bureau of Standards, Washington, D.C., 1973.January, 1982 REFERENCE MATERIALS 15 considerable effort has gone into finding suitable filling and sealing procedures that ensure long-term stability. Conclusion Reagent suppliers have been active in the field of standards and reference materials for many years, but this activity has been of limited effectiveness for two reasons.Firstly, it has proved very difficult to ascertain exactly the users’ requirements, and secondly, the preparation of such substances is time consuming and therefore expensive. The continuing supply of standards and reference substances will depend upon establishing effective communication between the potential user and the supplier, and acceptance by users that standards and reference materials are costly to prepare. 1. 2. 3. 4. 5. 6. 7. 8. 9. 200 300 4 Wavelengthhm 0 Fig. 1. Spectra of potassium dichromate solu- Experimental ratios for instrument used to tions. record spectra : Cuvette A,,, A, No.A313 A313 A313 2 2.236 2.868 2.461 3 2.216 2.853 2.412 4 2.216 2.849 2.378 5 2.217 2.922 2.483 6 2.215 2.964 2.523 References Microchemistry Group, Analyst, 1953, 78, 258. Microchemistry Group, Analyst, 1962, 87, 304. Microchemical Methods Group, Analyst, 1972, 97, 740. Analytical Methods Committee, Analyst, 1965, 90, 251. Analytical Methods Committee, Analyst, 1967, 92, 587. Analytical Methods Committee, Analyst, 1975, 100, 675. Analytical Methods Committee, Analyst, 1977, 102, 955. Analytical Methods Committee, Analyst. 1978, 103, 93. Burke, R. W., Deardorff, E. R., and Menis, O., “Liquid Absorbance Standards,” NBS Special Publication No. 378, National Bureau of Standards, Washington, D.C., 1973.18 REFERENCE MATERIALS Anal. Proc. Reference Materials in Clinical Chemistry S.S. Brown Division of Clinical Chemistry, MRC Clinical Research Centre, Harrow, Middlesex, HA 1 3 U J and M. Hjelm Department of Chemical Pathology, Institute of Child Health and Hospital for Sick Children, Great 0rm.ond Street, London, W C l N 3JH The demands of clinical medicine for reliable, quantitative chemical assays of body fluids, tissues and excreta have increased dramatically during the past few decades, in parallel with our understanding of biochemical events and disease processes at the cellular and molecular levels. This paper sets out to trace one particular development in laboratory medicine, that of reference materials in clinical chemistry, from both the semantic and pragmatic points of view. Evolution of the Concepts of Reference, Calibration and Control Materials The recognition of the need for reference materials for the control of accuracy in clinical chemistry arose from experience of the 1950s and 1960s in the interpretation of quality assur- ance surveys that had been conducted in Europe and North America.In particular, the College of American Pathologists’ Standards Committee, in 1966, reviewed the poor relative accuracy and lack of precision in measurements of serum cholesterol. It was noted that some commercial preparations of cholesterol used for calibrating assay systems were of doubtful purity, and it was expected, reasonably, that if pure crystalline cholesterol could be made available, the performance of clinical laboratories for this assay would improve.1 As a direct result, the US National Bureau of Standards (NBS) embarked upon a programme of develop- ment of “Standard Reference Materials” (SRMs) for use in clinical laboratories; cholesterol SRM was the first such product. The College of American Pathologists and the American Association of Clinical Chemists continued to debate the whole question of “standardisation.” In a seminal paper, Radin2 considered the general status of standards in clinical chemistry and the possible adoption of proposed definitions of “standard” and “reference” materials.He described the situation as “particularly deplorable,” and noted that some manufacturers were marketing serum controls or standards with claims that could be misleading. Radin drew attention to the pure sub- stance SRMs that were then available from the NBS for the clinical laboratory: these included standard buffers for the measurement of blood pH, cyanmethaemoglobin for the measurement of total haemoglobin in blood and bilirubin, cholesterol and various biochemicals including amino acids and carbohydrates.Radin recognised a special need for standard preparations for the assay of plasma total protein and also pointed out requirements for standardised units (especially in enzyme measurements) and standardised nomenclature. In a far seeing way, he also recognised the essential link between reference materials and reference methods, and offered definitions of “definitive methods” of analysis, “reference methods” of analysis, and “acceptable (or standard) methods” of analysis, which were influential in establishing the formal definitions put forward later by the International Federation of Clinical Chemistry (IFCC).3 Radin did not actually offer a definition of a “reference material,” but Hanson,* on behalf of the College of American Pathologists’ Standards Committee, suggested that a reference material or reference control was “a control serum or similar material commonly used in control pro- grammes or which may of necessity be occasionally employed as calibration materials.” Hanson noted that the chemical composition and the physical characteristics of such materials were intended to simulate the patients’ specimens to be analysed, and that these very qualities precluded their use or definition as standards.Hanson properly drew attention to the relatively incomplete characterisation and poor stability of clinical chemistry control materials, in comparison with the reference materials issued by national standards organisations such as NBS.Nevertheless, there is no reason, both in principle and in practice, as described later, why serum preparations should not fulfilJanuary, 1982 REFERENCE MATERIALS 19 the general definition of reference materials that was put forward by COX,^ i.e., “substances or artifacts which are suitable for calibration, checking or testing measuring instruments and measuring procedures generally; reference materials are credited with certain property values and hence act to transfer a scale of measurement from one place to another.’’ In a state-of-the-art review, Mears and Youngs expounded the principles of rational measure- ment embodied in SI, and the link with a material’s purity.They emphasised that the purity of the standard that is used for calibrating a chemical system should not be the factor that limits the accuracy of the results for a given assay. In other words, standard solutions should be known with greater certainty than is required in the analysis of an actual sample, so that errors in their concentrations have a negligible effect on the total analytical error. Mears and Young listed the NBS SRMs that were then available and that were of potential use in the clinical laboratory. Most of those uses would not nowadays be regarded as of importance; thus few clinical laboratories require benzoic acid to calibrate melting-point apparatus, nor sodium oxalate for standardising the analysis of serum calcium by the Clark - Collip method.Nevertheless, Mears and Young’s paper is an important landmark in the evolution of the modern concept of a reference material. A Discussion Meeting on Reference Materials and Methods in Clinical Chemistry was held in conjunction with the First European Congress of Clinical Chemistry in 1974. The Report of that meeting3 drew attention to the urgent need for continuing and strengthening international collaboration in these areas if further advances were to be made in assuring the accuracy, as well as the precision, of clinical laboratory performance. It is gratifying to note that sessions on “Reference Technology” have been included in several other major international congresses7 and that clinical chemists have participated in the deliberations of the International Standard- ization Organization (ISO) Committee on Reference Materials (ISO-REMCO) which was set up in 1975 : ISO-REMCO recognises that a wide variety of international organisations have strong interests in reference materials.In general, however, the metrologist and technologist users of reference materials have not been aware of the ramifications of biomedical standardisation, nor of the efforts of the World Health Organization (WHO) and the international societies such as IFCC and ICSH (International Committee for Standardization in Haematology), which have been directed towards identifying and solving the manifold problems of establishing a basis of accuracy for clinical laboratory assays.Thus IFCC established an Office for Reference Materials and Methods in 1974, the objectives of which are as follows: (1) to gather and ex- change information on reference materials and reference methods in clinical chemistry, here called reference technology, at the national and international levels ; (2) to offer technical assistance and co-ordination of studies on reference technology ; (3) to promote international consensus on reference technology and written standards; and (4) to evaluate and promote international reference materials, and assist in their characterisation and distribution. A centralised information system on national and international activities concerning refer- ence methodology and the production of reference materials and written standards is currently being set up, and the establishment of IFCC Working Groups on Reference Methodology is in progress.Availability of Biomedical Reference Materials The summary of NBS Standard Reference Materials suitable for clinical measurements given by Meinkes included an extended list of organic SRMs-cholesterol, urea, uric acid, creatinine, D-glucose and bilirubin-together with the inorganic SRMs, calcium carbonate, potassium chloride and pH buffers; also neutral glass filters and orchard leaves (for the standardisation of trace metal analyses). The latest NBS catalogue of SRMs for the clinical laboratory, as out- lined by Ra~berry,~ includes cortisol, VMA, Tris and lead nitrate, together with reference material absorbance standards, a spectrophotometer cell and clinical thermometers.The availability of a human serum SRM (characterised by isotope dilution - mass spectrometrylO for several electrolytes and for cholesterol) was announced in November 1980 and a reference serum for antiepileptic drug assay is in course of preparation. All of the reference materials so far mentioned are linked to assays which are commonly performed in clinical chemistry laboratories. Certain laboratories, however, have specialised interests which range across haematology, coagulation, immunology, pharmacology, nuclear medicine and medical physics. For the purpose of standardisation in these laboratories, it is20 REFERENCE MATERIALS Anal. Proc. important to recognise the contributions which have been made by WHO and by the Inter- national Atomic Energy Agency (IAEA) .So far as biological reference materials are concerned, the network of WHO Collaborating Laboratories has for many years maintained a large catalogue,ll the chief purpose of which is to “provide a means of ensuring uniformity throughout the world in the designation of potency of preparations administered to man that are used in the prophylaxis, therapy or diagnosis of human disease.” The substances in question cover a wide range of pharmaceuticals, vaccines and diagnostic reagents, with special emphasis on products which cannot be characterised adequately by chemical or physical methods. The clinical laboratory’s requirements for these products remains strong, in that intrinsically complex substances such as proteins and hormones need to be analysed in many different centres worldwide.Radioimmunoassay is commonly used for these kinds of analyses and proper calibration of the necessary counting equipment is essential. The IAEA’s Analytical Quality Control Services issue annually a listing of Certified Reference Materials,12 which is an extension of previous activities of IAEA, the International Directories of Isotopes and of Certified Radioactive Materials. Experience in operating the Analytical Quality Control Services of IAEA has shown that the most important cause of error in the measurement of radioactivity is incorrect calibration of the measuring device. The most straightforward way to eliminate this systematic error is to calibrate the measuring apparatus with a source of known activity.The Directories facilitate the selection of the proper calibration source or solution with a certified radioactive content which is best adapted to the particular calibration problem. Terminology The term “reference material” has been promulgated in clinical chemistry for many years, with a tacit adoption of the usage of COX,^ referred to above. Cox’s definition, however, and those of ISO-REMCO* present real semantic difficulties because none of them clearly distin- guishes between the operations of accurately calibrating a measuring instrument or system and controlling the over-all accuracy of the whole analytical process, including the sample preparation steps. Hence the Reference Method for the determination of total calcium in serum13 is calibrated by means of SRM calcium carbonate but the over-all accuracy of the whole measurement process is monitored by incorporating into it aqueous and matrix reference materials.Dybkaer14 intuitively recognised the distinction when he described the term refer- ence material as “insufficient and ambigious.” As an alternative, he advocated “comparison material,” as being a “comprehensive term supplanting the misused word ‘standard.’ A given comparison material is named by at least three words describing : (1) composition type, e.g., high-purity, simple, complex ; (2) purpose, e.g., calibration, control, comparison; and (3) state, e.g., solid, liquid, gas and solution, suspension or material. Examples are pure calibration gas (with confidence interval for, e.g., the substance concentration of a specified component) and simple control liquid suspension (with confidence interval for, e.g., the substance concentration of a specified component) and simple control liquid suspension (with confidence interval for, e.g., the number concentration of erythrocytes).One of the following terms may precede these elements of the name : International, National, Local.’’ For these reasons, the IUPAC Analytical Division’s Commission on Analytical Nomenclature in Clinical Chemistry and the Clinical Chemistry Division’s Commission on Automation is considering discouraging the use of the term reference material, so as to make clear the distinc- tion between materials used for calibration on the one hand and the control of accuracy or precision on the other.A discussion paper on this topic, concerned with various aspects of nomenclature for automated and mechanised analysis, will soon be published. * Reference material (RM) : a material or substance one or more properties of which are sufficiently well established to be used for the calibration of an apparatus or for the verification of a measurement method. Generally, any reasonably small part of an RM sample should exhibit the property value(s) established for the RM as a whole, within the stated uncertainty limits. NOTE: Such material is intended to transfer the value of a measured quantity (physical, chemical, tech- nological) between one place and another. It may be in the form of a pure or mixed gas, liquid or solid, or even a simple manufactured object.One or more of the properties of a given batch of reference materials and their adequate stability will have been established before the batch is issued for use. Certified Reference Material (CRM): an RM accompanied by, or traceable to, a certificate stating the property value(s) concerned, issued by an organisation that is generally accepted as technically competent.January, 1982 REFERENCE MATERIALS Production of Calibration and Control Materials in Clinical Chemistry 21 The general way of producing such a material is little different from that required for com- parison materials for other branches of metrology and includes careful consideration of (a) the user’s particular specification of the material; (b) the pre-analytical preparation steps ; (c) the analytical method and statistical procedure used to quantify the component(s) to be analysed; and (d) the system to be established for storage and distribution.Hjelm15 has reviewed the general implications of these four aspects, but the following points deserve emphasis here. For calibration materials, the analyte should be highly purified and as well defined as the state of the art allows. Such materials are used to establish the relationship between the instru- mental response and the concentration of the analyte.16 For control materials, the analyte commonly needs to be incorporated into a biological matrix such as serum, urine or a tissue fraction. The composition and the properties of the matrix have to match as closely as possible the biological specimens to be analysed, if a matrix control material is to fulfil its functions.Such materials serve to verify the performance of the entire analytical procedure.l6 One component of the over-all model for the production of matrix control materials, the pre-analytical handling of the material, is of special importance. This is because improper handling of the material during this phase of its production may make it unsuitable as a bio- medical control material, even though procedures for subsequent characterisation, storage and distribution may be quite satisfactory. Thus, for a biomedical control material based on blood plasma, relevant factors include: (a) proper choice of human or animal blood; (b) proper conditions for specimen collection, including sex and age of the donor, fasting or non-fasting state at time of collection, type of container for the product; (c) conditions for further process- ing, e.g., sterility, preservatives, temperature ; and ( d ) handling before analysis. The reason why it is essential to take such factors into account is that biomedical matrix materials such as plasma are effectively complex interactive biological systems which at present cannot be adequately described in physical and chemical terms.Hence detailed specifications for the pre-analytical handling of the matrix must ensure that any crucial characteristics of the final material are maintained and that these particular characteristics are consistent for consecutive batches of the same type of material.A matrix, as will be proposed by the IUPAC bodies mentioned above, can be defined as “the assemblage of the components other than the analyte in a specimen or a sample.” Despite the simplicity of the definition, the complexity of blood plasma can be illustrated by the fact that there are more than 1000 compounds that can be characterised in such a matrix, including inorganic, low and high relative molecular mass organic compounds, enzymes, hormones and vitamins. Many of the compounds might react and interact if there were improper handling of the matrix due to redox processes, degradation, ligand formation, polymerisation or other structural changes, etc. Such changes might render a matrix control material inappropriate for a defined use, even if there were an accurately assayed content of the certified analyte.This would be the case, for example, if systematic error were introduced into a routine method because the matrices of the control material and the biological specimens differed. The kinds of difficulties associated with matrix materials which have been prepared in different ways can be illustrated from a recent investigation,17 in which either spiked lyophil- ised or liquid human sera were used as calibration materials for the assay of the hormone cortisol in plasma. All specimens were assayed using a radioimmunological method and with either of the two types of calibration materials in order to establish the calibration graph. In addition, all specimens and calibration materials were assayed by isotope dilution - mass spectrometry (ID - MS).In this way the difference between the value obtained by the radio- immunological method and the accurate value obtained by ID - MS could be calculated for each of the specimens assayed. It was shown that the spiked lyophilised human serum used as calibration material was unsuitable, despite the fact that the assigned value of the analyte, cortisol, was correct as determined by ID - MS. Results obtained by using liquid human sera (which had been stored deep frozen) as calibration materizl for the radioimmunological pro- cedure were acceptable. The findings of this investigation indicate that considerable syste- matic research is called for in order to develop (a) suitable procedures to handle biomedical matrix materials for calibration and control purposes, either in lyophilised or liquid form, and (b) adequate methods for characterising the critical matrix properties of such materials.22 DERIVATIVE SPECTROSCOPY Anal. Proc. References Brown, S. S., Ann. Clin. Biochem., 1973, 10, 146. Radin, N., Clin. Chem., 1967, 13, 55. Rinsler, M. G., and Mitchell, F. L., 2. Klin. Chem. Klin. Biochem., 1974, 12, 558. Hanson, D. J., Am. J . Clin. Pathol., 1970, 54, 451. Cox, J . D., Chem. Ind. (London), 1975, 420. Mears, T. W., and Young, D. S., Am. J . Clin. Pathol., 1968, 50, 411. Brown, S. S.. in Carroll, D. M., Burns, D. T., Brown, D. A., and MacDaeid, D. A., Editors, “Euro- Meinke, W. W., Anal. Chem., 1971, 43, 28A. Rasberry, S., Anal. Proc., 1982, 9, 5 . Lawson, A. M., Lim, C. K., Richmond, W., Samson, D. M., Setchell, K. D. R., and Thomas, A. C. S., in Lawson, A. M., Lim, C. K., and Richmond, W., Editors, “Current Developments in the Clinical Applications of HPLC, GC and MS,” Academic Press, London, 1980, pp. 135-153. World Health Organization, “Biological Substances : Lists of International Biological Standards, International Biological Reference Preparations, and International Biological Reference Reagents,” WHO, Geneva, 1979. International Atomic Energy Agency, “Certified Reference Materials, Reference Materials and Samples for Intercomparisons (LAB/243 Circ.) ,” IAEA, Vienna, 1980. Brown, S. S., Healy, M. J. R., and Kearns, M., J . Clin. Chem. Clin. Biochem., in the press. Dybkaer, R., in “Proceedings : International Conference on Standardization of Diagnostic Materials, June 5-8, 1973,” US Department of Health, Education, and Welfare, Atlanta, 1974, pp. 11-26. Hjelm, M., in Voller, A., Editor, “Immunoassays for the 198Os,” MTP Press, Lancaster, 1981, p. 185. Stamm, D., J . Clin. Chem. Clin. Biochem., 1979, 17, 283. Lantto, 0.. Bjorkhem, I., Blomstrand, R. F., and Kallner, A., Clin. Chem., 1980, 26, 1899. analysis I11 : Reviews on Analytical Chemistry,’’ Applied Science, London, 1979, pp. 75-92. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

ISSN:0144-557X    

DOI:10.1039/AP9821900012

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

22 DERIVATIVE SPECTROSCOPY Anal. Proc. Derivative Spectroscopy and its Applications in Analysis The following are summaries of the seven papers presented at a Joint Meeting of the Scottish, North East and North West Regions and Special Techniques and Joint Pharmaceutical Analysis Groups held at Heriot-Watt University, Edinburgh, on March 19th, 1980. Derivative Spectroscopy : Theoretical Aspects. Thomas C. O'Haver Plenary Lecture Department of Chemistry, University of Maryland, College Park, Md. 20742, USA The derivative technique is becoming increasingly popular in analytical spectrophotometry 'as a resolution enhancement technique, to facilitate the detection and location of the wave- lengths of poorly resolved components of a complex spectrum: and as a background correction technique, to reduce the effect of spectral background interferences in quantitative analytical spectrophotometry.2 A significant disadvantage to the derivative technique, however, is that the signal to noise ratio (SNR) becomes worse at progressively higher derivative orders3 The purpose of this paper is to investigate on a theoretical basis the effect of differentiation on SNR so as to provide the experimenter with a rational approach to the optimisation of SNR in derivative spectroscopy.Consider an experimental spectrum that has been measured by a discrete sampling technique and consists of a series of amplitudes (or absorbances, in absorption spectrometry) taken at discrete, equally spaced wavelength increments a,, a2, a3, a4, . . ., where a, represents the amplitude at wavelength 1, and so on.Consider that this series consists only of noise, i.e., that there is no signal. Let us assume that the noise is independent of the amplitude of the signal, has a white spectral distribution and a Gaussian amplitude probability distribution and that each element in the series is statistically independent of its neighbours. The noise may therefore be expressed as the standard deviation of all the elements in the series, which we will give the symbol oo, the subscript referring to the "zeroth"-order derivative. It can be shown that each element in the nth derivative calculated by the method of successiveJanuary, 1982 DERIVATIVE SPECTROSCOPY 23 differences is a linear combination of n + 1 adjacent elements from the original (zeroth order) series and the weighting coefficients are given by the binomial coefficients of order n : 92 n! & = (-1P * (n - ai+m m=O where d; is the ith element of the nth derivative.The standard deviation of the nth-order derivative can be calculated by the usual rules for error propagation and is simply the weighted quadratic sum of the standard deviation of its terms: The signal depends on the shape of the spectrum. The derivatives of Gaussian and Lorentzian bands have been evaluated analytically.4 We define the “signal” as the difference between the largest (most positive) and the smallest (most negative) value of the derivative. Relative SNRs defined in this way for the first four derivatives of Gaussian and Lorentzian bands are listed in Table I, where W is the number of points in the peak full-width at half-maximum (FWHM).Note that the SNR decreases rapidly with increasing derivative order when W is large. Typically, W must be at least 5-10 to define the width and height of the peak adequately and it is often much larger. If W = 10 the relative SNRs of the first four derivatives of a Gaussian band are 0.2, 0.032, 0.008 and 0.0017. Clearly, if W is large, the SNR of the higher derivatives will be very poor even if the SNR of the original spectrum is good. TABLE I RELATIVE SIGNAL TO NOISE RATIOS OF UNSMOOTHED DERIVATIVES Band shape* Derivative (-A 7 order Gaussian Lorentzian 0 1 1 1 2.021 w 1.84lW 3 8.101 v-3 16.71 W8 4 17.81 W 641 W4 2 3.261 W2 4.11 ws * W = number of points in the FWHM of band. How, then, is it possible to use the derivative technique effectively? The answer is that all practical differentiation techniques include some degree of Zow-pass jltering or smoothing to control the increase in noise that is the inevitable result of differentiation of a noisy signal.Smoothing is also sometimes used in normal (zeroth order) spectroscopy, but its value there is largely cosmetic; the degree of SNR improvement that can be achieved by smoothing is limited, typically 2- to &fold. However, it turns out that smoothing is much more important in derivative spectroscopy; that is, the obtainable degree of SNR improvement is much greater than in normal spectroscopy. The reason for this is that the increase in noise caused by differentiation is largely due to an increase in high-frequency noise, the process of differentiation being effectively a kind of high-pass filtering. Thus, the extra noise is especially easy to reduce by a low-pass smoothing process.The process of smoothing involves a convolution of the data series with a smoothing function consisting of a set of weighting coefficients. Each point in the smoothed array is a linear combination of a group of adjacent points in the original series each multiplied by a weighting coefficient. The different types of smooths differ only in the way the coefficients are calculated. The simplest type of smooth is the equally weighted sliding average, in which the weighting coefficients are equal. Each point in the smoothed series is the simple average of N adjacent points in the original series, where N is the smooth width.The value of the smoothing co- efficients is therefore simply 1/N. The effect of smoothing a peak-type signal is both to reduce noise, which is desirable, and to distort the signal, which is undesirable but unavoidable. The distortion is seen as an attenuation in peak height and a slight increase in peak width. The extent of this distortion24 DERIVATIVE SPECTROSCOPY Anal. PYOC. depends on the ratio of the width of the smooth to the FWHM of the peak, which is called the smoothing It was seen in Table I that the SNR decreases as the derivative order increases. Therefore, it is logical that the amount of smoothing would have to be increased as the derivative order is increased in order to control this SNR degradation.This must be done, however, not by increasing the smoothing ratio but rather by increasing the number of times that the smoothing is passed throagh the data series. Fig. 1 shows the trade-off between attenuation and SNR for the second derivative of a Gaussian band smoothed by one, two and three passes of a sliding average smooth. This plot shows that three passes of the smooth give a better trade-off between SNR improvement and attenuation than one or two passes. In general, it can be shown that for the nth derivative, n + 1 passes are required and the standard deviation of the final series is . . (2) where n is the derivative order and N is the number of points in the smooth. It can be seen from the denominator of this term that the noise reduction with increasing N will be greatest for the higher derivatives.Similarly, the standard deviations of the zeroth, first and second derivatives with, respectively, one, two and three passes of an N-point quadratic-cubic smooth6 are found to be 1.7aO/.t/N, ~ c T ~ / N ~ . ~ and 2 2 0 , / N ~ . ~ . For a given value of N , the quadratic-cubic is less effective at reducing noise but also causes less peak attenuation than the simple sliding average. We have found that, as for the sliding average smooth, the quadratic-cubic smooth also requires n + 1 passes for the nth derivative. Using the above analysis to calculate the noise, the effect of smoothing on the SNRs for any given band shape can be easily evaluated by determining the p2ak-height attenuation caused by smoothing noiseless model bands.The trade-off between attenuation and SNR as a function of smoothing ratio is illustrated in Fig. 2 for the second derivative of a Gaussian band smoothed by three passes of a sliding-average smooth. The optimum SNR occurs at a smoothing ratio of about 1.5, at which point the peak-height attenuation factor is about 0.1. 3 2 .^ o 1 4- 2 $ 0.5 0 0 .- c, 0.2 C G) z 0.1 1.0 0.8 0.6 0.4 0.2 Attenuation factor Fig. 1. The trade-off between signal to noise ratio and peak- height attenuation for the second derivative of a Gaussian band smoothed by one, two and three passes of a sliding average smooth, showing that three passes is always better than one or two. 1 .o 0.9 .O 0.8 2 0.7 .? 0.6 0.5 0.4 6 0.3 c. Q) ix 0.2 0.1 0 1 2 3 Smoothing ratio Fig.2. Attenuation factor and relative signal to noise ratio as a function of smoothing ratio for the second derivative of a Gaus- sian band smoothed by three passes of a sliding average smooth.January, 1982 DERIVATIVE SPECTROSCOPY 25 Although the peak-height attenuation caused by smoothing will not cause a systematic error in a relative analytical method calibrated by appropriate standards, excessive attenuation is undesirable for two reasons. Firstly, in many real applications the measured band will be accompanied by a background signal that is less strongly attenuated by smoothing than the measured band itself, thereby reducing the signal to background ratio. Secondly, smoothing increases the width of the peak and reduces the effective spectral resolution. Inasmuch as the usual reasons for using the derivative technique in the first place are to reduce the effect of spectral background and/or to improve the effective spectral resolution, smoothing acts in the opposite manner and, if overdone, can cancel the advantages offered by the derivative method.Thus it will always be advantageous to consider the trade-off between SNR im- provement and peak-height attenuation as smoothing ratio is increased. This can be done conveniently by plotting SNR versus attenuation, with smoothing ratio being the “hidden parameter.” Such plots for the zeroth, first and second derivatives of Gaussian and Lorentzian bands with eight-point FWHM smoothed with n + 1 passes of sliding average and quadratic- cubic smooths are shown in Figs. 3,4 and 5.From an inspection of these figures we can make the following observations : 1. The achievable relative SNR improvement (relative to the SNR without smoothing) is much greater for the derivatives than for the normal (zeroth derivative) spectrum. Had the peak width W been larger than 8, the difference would have been even more dramatic. 3 2 .- E l 2 .g 0.5 S 0 Y - 0.2 iTj 0.1 S m 0.05 1.0 0.8 0.6 0.4 Attenuation factor .- 0 2 +-’ 2 al 1.0 0.5 .- 0 +-’ - P .$ 0.2 Q) P al .- 2 0.1 - LT 0.05 Fig. 3. The trade-off between signal to noise ratio and attenuation factor for the zeroth, first and second derivatives of a Gaussian band with an eight-point FWHM smoothed by one, two and three passes (respec- tively) of a sliding average smooth. 1.0 0.8 0.6 0.4 Attenuation factor Fig.4. As Fig. 3, except for a quadratic- cubic smooth. 1.0 0.8 0.6 0.4- Attenuation factor Fig. 5. As Fig. 4, except for a Lorent- zian band, quadratic- cubic smooth. 2. With sufficient smoothing, the SNR of the derivatives can exceed that of the unsmoothed normal spectrum (relative SNR = l.O), although this occurs at progressively greater attenuation as the derivative order increases. Ultimately, the SNR of the derivatives can exceed that of the normal spectrum smoothed to optimum SNR, but this occurs only at rather drastic attenuations. 3. The effect of differentiation on SNR depends on the peak shape, type of smooth and especially on the smoothing ratio. However, for the most commonly used smoothing ratios in the range 0.2-0.5, the previously stated generalisation’ that the SNR decreases by a factor of 2 for each stage of differentiation is supported approximately.26 DERIVATIVE SPECTROSCOPY Anal.Proc. 4. The quadratic-cubic and sliding average smooths achieve about the same ultimate SNRs, but the quadratic-cubic gives a better trade-off between SNR and attenuation by about a factor of 2. 5. Gaussian and Lorentzian derivatives exhibit the same general behaviour, but Lorentzian derivatives suffer significantly more attenuation when smoothed to a given SNR improve- ment. (As expected, bands with mixed shapes exhibit intermediate behaviour.) These conclusions can be demonstrated and supported by differentiation and smoothing of realistic computer-generated noisy model bands. Fig. 6A shows an isolated single Gaussian band with a 50-point half-width and an r.m.s.SNR of 100. The noise is white. The second derivative, without smoothing, is shown in Fig. 6B. The SNR of this derivative is predicted (Table I) to be 100 (3.26/502) = 0.13, so it is no surprise that a signal cannot be detected here. Figs. 6C, D and E show the result of progressively larger amounts of sliding average smoothing, Fig. 6. Graphic examples of the effect of differentiation and smoothing on the signal to noise A : Original (zeroth derivative) B: Second deriva- C: Second derivative after two passes of an 11-point sliding average D: After three passes of an 11-point sliding E: After three passes of 25-point sliding average smooth, ratio and peak height of- the second derivative of a Gaussian band.band; Gaussian; 50 point FWHM; signal to noise ratio = 100; white noise. tive, without smoothing. smooth, greater vertical scale expansion than B. average smooth, same scale as C. same scale as C. showing both the effect of the number of passes and the smoothing width. The trade-off between SNR improvement and attenuation is evident. In Fig. 6E the smoothing ratio is 0.5. The SNR can be calculated by multiplying the SNR without smoothing (0.13, from Table I) by Nn+Oa5 and multiplying the result by the attenuation factor (0.63, from Fig. 2) : (0.13)(3125)(0.63) = 2.6 x lo2 Thus the derivative in Fig. 6E has been smoothed to a SNR better than that of the original (unsmoothed) spectrum. Fig. 7 demonstrates that with even more smoothing (smoothing ratio = 1.0 for sliding average smooth) the SNR of the smoothed derivatives equals that of the optimally smoothed original spectrum.However, the figure also illustrates why this is not usually a useful goal; the width of the central maximum of the second derivative has been broadened by smoothing to such an extent that it is now no narrower than the smoothed original peak, i.e., the resolu- tion advantage of the second derivative has been neutralised by smoothing. The same be- haviour has been observed up to the sixth derivative and is probably general for all derivative orders. The advantage of the quadratic-cubic smooth over the sliding average, which is its better trade-off between distortion and SNR improvement, is especially important when the balance between those two factors is critical, as in the example in Fig.8. The spectrum (Fig. 8A) consists of two overlapping Gaussian components of equal width separated by slightly less than their FWHM. The right-hand component has half the intensity of the left-hand com- ponent. In this example, the second derivative mode is used in resolution-enhancement service in order to detect and locate the position of the smaller (right-hand) component. That this is feasible under ideal conditions is demonstrated by Fig. SB, which is the theoretical shape of the second derivative without noise and without smoothing. The two minima correspond closely to the positions of the two components in the original spectrum. However, the separation of these two components is close to the The r.m.s. SNR is 100 (white noise).January, 1982 DERIVATIVE SPECTROSCOPY 27 Fig.7. Zeroth, first and second derivatives of a noisy Gaussian band smoothed to nearly maximum signal to noise ratio, illustrating that smoothing to such a large extent avoids the SNR degradation of differentiation but also cancels the resolution advantages of the higher derivatives. Sliding average, smoothing ratio = 1.0. theoretical minimum for resolution of Gaussians in the second derivative.* In such a case, the broadening effect of smoothing is especially troublesome. Figs. SC-J show second derivatives of the original noisy spectrum smoothed with sliding average (Figs. SC-J) and quadratic-cubic (Figs. 8H- J) smooths with different smoothing ratios. Theeffect of broadening and loss of effective spectral solution is evident, particularly for the sliding average smooth.The superiority of the quadratic-cubic smooth is clear, as it gives about a factor of two lower distortion and better SNR than the sliding average. On the other hand, the quadratic- cubic smooth requires more computation time than the sliding average, both because the required smoothing ratios are about twice as large and because the computations themselves are more complex, requiring N multiplications and N additions, whereas the sliding average requires only N additions and one multiplication. In general, the selection of the optimum smoothing ratio depends on the purpose for which the derivative technique is used. In using the even derivatives for resolution enhancement purposes, relatively small smoothing ratios (of the order of 0.2 for the sliding average or Normal Ideal shape (SNR = 100 unsmoothed) noise or smoothing of 2nd derivative without Normalised second derivatives C f Sliding beverage I smooths Three-passes smoothing ratio given Quadratic-cubic smooths -+ Fig.8. Effect of smoothing ratio and smooth type on the trade-off between signal to noise ratio and effective resolution. A : The original spectrum, consisting of two overlapping Gaussian bands of equal width and a 2: 1 intensity ratio separated by 0.8 of their width; white noise, SNR = 100, unsmoothed. B: The theoretical shape of the second derivative, without noise and without smoothing. The two minima correspond to the two components in the original band. C-G: Second derivatives of the original noisy spectrum, smoothed with three passes of a sliding average smooth, smoothing ratios 0.1, 0.15, 0.2, 0.3 and 0.5.H-J: Same, but with quadratic-cubic smooths, smoothing ratios 0.4, 0.5 and 0.6. The quadratic-cubic smooth is seen to offer a better trade-off between SNR and distortion. All derivatives are normalised to the same height.28 DERIVATIVE SPECTROSCOPY Anal. Proc. 0.5 for the quadratic-cubic) will assure that very little loss in effective resolution will result. In this case a significant loss in SNR will have to be tolerated. In quantitative analytical applications, in which the derivative technique is used to remove or reduce a broadly curved background, significantly larger smooth ratios may be profitably employed, typically 0.5 for the sliding average and 1.0 for the quadratic-cubic.Larger smoothing ratios will yield little SNR improvement at the expense of considerable further peak attenuation and broadening. I t is rarely advisable to smooth to the optimum SNR. In quantitative work the aim in selecting the smoothing ratio is to optimise the trade-off between random measurement errors caused by the noise and the measurement errors, random and systematic, caused by the presence and variability of the background. If the shape and variability of the back- ground can be predicted, then in principle it should be possible to arrive at a truly optimum smoothing ratio that minimises the sum of noise errors and background errors. The width W of the measured spectral peak, here expressed as the number of points in the FWHM, is determined by the spectral scan rate and, in a digital system, by the data sampling rate.It can be shown that if the smoothing ratio is held constant, the SNR is proportional to 4 W for all derivative orders. Therefore, the values of SNR given in Figs. 3, 4 and 5 for an eight-point wide peak can be converted into any other peak width W by multiplying by dW/S. The usual way of changing W is to change the spectral scan rate. If the scan rate is changed at a constant data sampling rate, W changes inversely with scan rate and thus SNR is inversely proportional to the square root of scan rate. If the data rate is changed in proportion to the scan rate, W remains constant but the averaging time varies inversely with the data rate and noise varies with the square root of the averaging time.Therefore, in this case also the SNR is inversely proportional to the square root of the scan rate. The same holds for real-time differentiators; as the scan rate is changed and the system response time is varied to keep the smoothing ratio constant, the band width varies directly with scan rate, and the noise varies inversely with the band width. This result is significant because it is contrary to the expectation based on signal amplitude alone; if the scan rate is increased, the derivative output signal increases but, if the smoothing ratio is kept constant, the SNR decreases. On the other hand, because one generally uses smoothing ratios below the maxi- inum SNR, increasing the scan rate at a constant differentiation response time will increase the SNR because the smoothing ratio is increased. References 1.2. 3. 4. 5. 6. 7 . 8. Butler, W. L., Methods Enzymol., 1979, 56, 501 and references cited t‘nerein. Talsky, G., Mayring, L., and Kreuzer, H., Angew. Chem., Int. E d . Engl., 1978, 17, 735. Cahill, J. E., Am. Lab., 1979, 11, 79. Morrey, J . R., Anal. Chem., 1968, 40, 905. Enke, C. G., and Nieman, T. A., Anal. Chem., 1976, 48, 705% Savitzky, A,, and Golay, M., Anal. Chem., 1964, 36, 1627. O’Haver, T. C., and Green, G. L., Anal. Chem., 1976, 48, 312. Pavlath, A. E., and Millard, M. M., AppZ. Spectrosc., 1979, 33, 502. Higher Derivative Methods in Ultraviolet - Visible and Infrared Spectrophotometry Anthony F. Fell and Geoffrey Smith Department of Pharmacy, Heriot- Watt University, Edinbwgh, EH1 2HJ About 60 years ago Lord Rutherford suggested the first derivative for more sensitive mass spectrometric detection of gas excitation potentia1s.l In 1953, the concept of second and higher derivative detection was patented by two British industrial chemist^,^-^ who first demonstrated the usefulness of higher derivatives in analytical spectroscopy, as discussed later by Martin.5*6 Morrey’s classic essay on the properties of computer-generated derivative functions’ and Butler’s computer work on fourth and higher derivative spectra*-1° coincided with the adventJanuary, 1982 DERIVATIVE SPECTROSCOPY 29 of low-noise operational amplifiers.These led to the high-quality electronic differentiators used in early work on derivative luminescence spectroscopyll 9 1 2 and on derivative infrared spectroscopy.13 More recently, there have been several studies in the biomedical area based on second or higher derivatives of ultraviolet - visible absorption spectra.14-18 The renaissance of interest in derivative spectroscopy is such that the method has been patented again, to give an elegant device for low-noise derivatives up to eighth order.lg Theoretical studies, however, indicate that for resolution enhancement in the ultraviolet - visible range, the most useful derivatives will be second or fourth order.20 These can be generated by careful combination of commercial units for accurate quantitative analysis.21 9 2 2 Recent developments in microcomputer tech- n01ogy~~ permit good quality higher derivatives and enable the derivative method to be simplified, In this paper, the rationale for developing analytical methods based on higher derivative spectroscopy is discussed with reference to applications in the biomedical sciences.Principal Features of Derivative Functions The basic properties of derivative functions have been discussed elsewhere in the context of spectroscopy.' 920924 The derivative technique is, however, perfectly general in its application, and can be used equally for chromatographic or densitometric data, e.g., gas - liquid chromatography (GLC), high-performance liquid chromatography (HPLC) or densito- metric scans of thin-layer plates or electrophoret~grams.~~ >26 Although the first derivative has been more widely known for the detection and characterisation of spectroscopic peaks,27 it turns out that the second and higher even derivatives of absorbance or signal intensity (d2A/dA2, d4A/dA4, .. .) are potentially more useful where Gaussian or Lorentzian functions approximate the data profile. By contrast, in cases where titration curves, reflectance data or thermal analysis curves are recorded, the odd derivatives are used to transform the disperse inflection to a peak for interpretation. The present discussion is concerned with the even derivatives of Gaussian or Lorentzian curves, which are seen as bipolar functions of alter- nating sign, flanked by satellites equal in number to the derivative order. The centroid coincides with the original peak, whose band width decreases progressively with derivative order.This is the key to the potential enhancement of resolution of overlapping bands. Significantly, the derivative process discriminates against broad bands, emphasising sharper features to an extent that increases with increasing derivative order, because for Gaussian or Lorentzian bands the amplitude, Dn, of the nth derivative is inversely related to the original band width, W , raised to the nth degree: Thus, for two coincident bands of equal intensity, the derivative amplitude of the sharper band (X) is greater than that of the broader band (Y) by a factor that increases with in- creasing derivative order (Fig. 1) : This feature, first reported in 19708928 and recently discussed by Cahi11,29 may account for the increase in effective detection sensitivity often observed in second and fourth derivative ~pectra.8.l~ 930 scattering and higher order polynomial interference in spectroscopy, or drifting base lines in chromatograms, are progressively suppressed as a logical consequence of the derivative technique.20 Derivative transformation does not intrinsically increase the information content of spectroscopic or chromatographic data (in fact some information is lost, e.g., constant factors).Rather does the derivative method permit more satisfactory discrimination against inter- ference, and emphasise subtle features of the data by presenting them in a new and visually more accessible way. Greater resolution of overlapping bands, reduction of systematic errors caused by matrix interferences and a sensitive qualitative profile for compound characterisa- tion can be conferred by careful application of the higher derivative method.Other types of background disturbance, such as RayleighAnal. Proc. 30 DERIVATIVE SPECTROSCOPY Strategy for Method Development In essence, the development of a derivative spectroscopic assay is little different from a regular spectroscopic assay. The principal requirements remain assay linearity, accuracy, reproducibility, analyte independence of matrix and satisfactory comparison with an accepted referee method. Any difference arises from the greater number of instrumental variables to be controlled, depending on the derivative module employed. The electronic analogue RC device computes the derivative with respect to time, as the spectrum is scanned at constant speed, S(dh/dt) : dtn - dh” [:In The “true” wavelength derivative is linearly related to the time derivative recorded, the magnitude of which is directly affected by scan speed and spectral band width.For qualitative work, the second or fourth derivatives offer a convenient method for enhancing sharp spectral features (Figs. 2 and 3), although for broad peaks the highest practicable derivative order is usuaIIy second. The lowest derivative order required for the analytical objective should be selected. When using analogue electronic or microcomputer derivative modules, the spectral displace- ment in the direction of scanning is a function of differentiator time constant, scan speed and peak half-width. Clearly, for comparison purposes, the same instrumental conditions must be used for qualitative work. dnA dnA -- First, however, the order of derivative should be considered.Fig. 1. Effect of derivative cu 6 0 3 ‘0 0 1 .O 0.8 Q 0.4 0 220 250 250 300 hln m orde; (zeroth, second and fourth) Fig. 2. Zero-order and second-deriva- on the relative amplitudes of two tive spectra of diphenhydramine hydro- coincident (a) Gaussian or (b) chloride (DPH). 750 p g ml-l in water, Lorentzian bands, X and Y, of equal of matrix M diluted 1 + 3 with water, intensity and band-width ratio 1 : 3. and of diphenhydramine in diluted Curve S represents the combined matrix (DPH + M). The derivative peak. amplitude used for assay is coded D285. In background correction, if a particular matrix interference overlapping the analyte peak approximates a quadratic, the second derivative reduces it to a constant, while the nth is required for an rcth degree polynomial, P: P = a, + aJ + a,h2 + .. . + a,hn dnP dhn -- - n!an Thus, the polynomial of degree n is reduced to a constant in the nth derivative and eliminatedJanuary, 1982 DERIVATIVE SPECTROSCOPY 31 in the (n + 1)th derivative. In fact, the constant leads to displacement of the whole deriva- tive spectrum, a feature readily compensated for by measuring graphically between the derivative peak itself and an adjacent satellite peak. Having established the minimum derivative order required to eliminate a matrix interference, the appropriate peak amplitude measure must be selected, there being several measurement options as discussed by O'Haver and Green.12 There may, of course, be several peaks associated with a particular analyte.Normally a series of standards is set up with the matrix at constant concentration, and the linearity of each peak amplitude measure is examined. An interaction study is performed, where the analyte level is kept constant while the matrix concentration is varied from 0 to 120% (Fig. 4). The best amplitude measure is selected on the basis that it is least affected by the matrix variation, and that it gives the best calibration statistics. The basis of quantita- tion in derivative spectroscopy remains the assumption that the Beer - Lambert law is obeyed over the range of interest: d"A d"e - - -.cb at h [dhn dh" where A is absorbance, E is molar absorptivity (1 mol-l cm-l), c is concentration (moll-l).and b is path length (cm). - + Q x *s O U - 1 .o 0.8 0.4 0 DPH 220 250 300 Alnm E E 100 2 8 50 0 50 100 mg per 100 ml 0 10 20 30 % VIV Fig. 4. (a) Calibration and (b) interaction graphs for diphenhydramine hydrochloride (DPH) in matrix diluted 1 + 3 with water. The concentration of matrix in the calibration graph was 25% V / V . The concentration of DPH in the interaction graph was - - 600 pg ml-l. Fig. 3. Zero-order and fourth-derivative spectra of diphenhydramine hydrochloride (DPH) , 750 pg ml-l in water, and of matrix diluted 1 + 3 with water (M). As an example, the derivative ultraviolet assay of diphenhydramine hydrochloride (DPH) in a multi-component, coloured sucrose vehicle typifies the approach developed in the authors' laboratory. The fine structure of the aromatic nucleus is readily apparent in the second and fourth derivative spectra of the pure aqueous solution (Figs.2 and 3). Dilution of the dosage form with water (1 + 3) yields the matrix at 25% concentration; addition of varying amounts of DPH to the formulation matrix and measurement of each second derivative amplitude indicated that all were linear with concentration up to 1250 pg ml-l (Fig. 4). An interaction experiment with the matrix varying from 0 to 30% V/V in the diluted sample showed that for a constant DPH level of 600 pg ml-l the best amplitude measure was that at 265 nm (to the long wavelength satellite). Using a Perkin-Elmer Model 200 spectrophotometer, the scan speed (120 nm min-l), slit width (2 nm), derivative module gain or time constant (No.5 ; Hitachi 0507 module), the32 DERIVATIVE SPECTROSCOPY Anal. Proc. absorbance scale and the test concentration range were adjusted for optimum signal to noise ratio. Comparison of results for a batch by the derivative method with the official USP XIX procedure (which involves time-consuming extraction and back-extraction steps) gave com- parable recoveries (94.3 and 93.9y0, respectively) and no significant t-test difference (n = 12). The relative standard deviation was 1.4y0, compared with 2.2% (n = 12) by the USP XIX assay.31 As in all quantitative derivative assays, the method must be externally standardised using (aqueous) standards in bracketting sequence to minimise the effect of drift. For this particular formulation, the method is rapid and accurate.Clearly, if the formulation were changed, the interaction study would have to be re-examined. Although the present discussion has concerned the “analyte-rich” situation typical of pharmaceutical and similar production systems, the derivative spectroscopic method is potentially applicable at the trace amount level as illustrated by some recent toxicological studies on paraquat in cases of poisoning,30 where the detection limit was reduced 10-fold by exploiting the second derivative approach. There are many instances in biochemical and environmental analysis where the derivative technique may, conceivably, be found useful. Future Projections Rapid progress in microcomputer technology is leading to high-quality, low-noise derivative modules,23 which may exploit the Savitzky - G01ay~~ or the Fourier transform method.33 The microcomputer permits data manipulation to give automatic peak amplitude measure- ment between defined spectral features.The presentation of “satellite-free” higher derivatives is also a feasible proposition. Automatic multi-component mixture analysis by the matrix inversion method in derivative domain is now possible by microcomputer. This approach relies on an appropriate archive of derivative spectra, which may also be used for rapid con- firmation of identity, as recently demonstrated in a commercial linear diode array spectro- photometer.34 Qualitative characterisation in the derivative domain will be useful in infrared spectroscopy, where the second derivative method has already found application in plastics and fibre and where theoretical studies indicate that fourth and higher derivatives should be useful for resolution enhancement .20 Use of the second and higher derivatives in HPLC and GLC will permit improved resolution of overlapping peaks21 725 s3’ and present interesting possibilities for densitometric scanning of electrophoretograms and TLC plates.26 938 s30 The development of multi-channel spectro- photometers coupled with the widespread availability of instruments with derivative capability, and their rapid acceptance by the analytical community, will lead to further interesting applica- tions of this refined and versatile technique.Thanks are expressed to Perkin-Elmer Ltd., Beaconsfield, Buckinghamshire, for the generous loan of equipment.The authors also acknowledge the skilled assistance of I. E. Aitchison (Department of Computer Science, Heriot-Watt University) and Mrs. Ann E. Mair (Western General Hospital, Edinburgh) and stimulating discussions with Dr. B. P. Chadburn, G. L. Collier, Dr. A. E. Martin, Dr. J. N. Miller, Professor T. C. O’Haver and A. C. M. Panting during the course of this work. 1. 2 . 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References Dymond, E. G., Proc. Camb. Phil. SOC., 1923-25, 22, 405. Singleton, F., and Collier, G. L., “Improvements in or Relating to Spectroscopes,” BY. Pat., 760729, Collier, G. L., and Singleton, F., J. Appl. Chem., 1956, 6, 495. Collier, G. L., and Panting, A. C. N., Sfiectvochim. Acta, 1959, 14, 104.Martin, A. E., Nature (London), 1957, 180, 231. Martin, A. E., Spectrochim. Acta, 1959, 14, 97. Morrey, J . R., Anal. Chem., 1968, 40, 905. Butler, W. L., and Hopkins, D. W., Photochem. Photobiol., 1970, 12, 439. Butler, W. L., and Hopkins, D. W., Photochem. Photobiol., 1970, 12, 451. Butler, W. L., Methods Enzymol., 1979, 56 (Part G), 501. Green, G. L., and O’Haver, T. C., Anal. Chem., 1974, 46, 2191. O’Haver, T. C., and Green, G. L., Anal. Chem., 1976, 48, 312. Keighley, J. H., and Rhodes, P., Proc. Inst. Elect. Rad. Eng., 1971, 22, 397. Ichikawa, T., and Terada, H., Biochim. Biophys. Acta, 1977, 494, 267. Talsky, G., Mayring, L., and Kreuzer, H., Angew. Chem., Int. Ed. Engl., 1978, 17, 785. 1956.Januavy , 1982 DERIVATIVE SPECTROSCOPY 33 16. 17.18. 19. 20. 21. 22. 23. 21. 25. 263. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Fell, A. F., Proc. Anal. Div. Chem. Soc., 1978, 15, 260. Balestrieri, C., Colonna, G., Giovane, h., Irace, G., and Servillo, L., Eur. J . Biochem., 1978, 90, 433. Jones, K. G., and Sweeney, G. D., Clin. Chem. (Winston-Salem), 1979, 25, 71. Talsky, G., “Gerat zum Auswerten von Messkurven,” Ger. Offen., 2806826, 1979. Fell, ,4. F., U V Spectrom. Group Bull., 1980, 8, 5. Fell, A. F., UV Spectrom. Group Bull., 1979, 7, 5 . Ishii, H., and Koh, H., Nippon Kagaku Kaishi, 1980, 203. O’Haver, T. C., and Smith, A., Am. Lab., 1981, 13, 43 O’I-Iaver, T. C., Anal. Chem., 1979, 51, 91.4. Fell, A. F., Anal. Proc., 1980, 17, 512. Okamura, K., Clin. Chzm. Acta, 1979, 96, 273. Shibata, S., Angew. Chem., I n t .Ed. Engl., 1976, 15, 673. Gulyaev, R. A., and Litvin, F. F., Biofizzka, 1970, 15, 670. Cahill, J . E., A m . Lab., 1979, 11(11), 79. Fell, A. F., Jarvie, D. R., and Stewart, &I. J., Clin. Ch9m. (Winstotz-Sa!em), 1981, 27, 286. Fell, A. F., to be published. Savitzky, A., and Golay, M. J . E., Anal. Chem., 1964, 36, 1627. Horlick, G., Anal. Chem., 1972, 44, 943. James, G. E., “UV/Vis Application Note,” No. 1, Hewlett-Packard, Palo Alto, Calif., 19SO. Pump, W., and Woltjes, D., Kunststoffe, 1979, 69, 317. Maddams, W. F., and Tooke, P. B., Macvomol. Scz., in the press. Zelt, I>. T., Owen, J . A,, and Marks, G. S., J . Chromatogr., 1980, 189, 203. Machold, W., Meister, A., and ,4dler, K , Photosynthetica, 1971, 5, 160. Cottrell, C. T., “Derivative and Log Spectrophotometry,” SP8 Series Accessories Applications, €’ye Unicam, Cambridge, 1980.Numerical Methods for Generating Derivative Spectra Peter Gans Department of Inorganic and Structural Chemistry, The University, Leeds, L S 2 9 J T We begin by assuming that the spectrum is available in the form of s pairs of numbers xi, yi (i = 1, 2, . . ., s). The simplest method (I) for generating the derivative spectrum is to fit a polynomial of degree h (order k , k = h + 1) to k data points at a time. Secondly (11), a polynomial of degree h can be fitted by the method of least squares to m data points at a time, when m > k . Lastly (111), a spline function of degree h can be fitted to all s data points by the method of least squares. I t is convenient, but not essential, that the data be evenly spaced along the x axis, i.e., xi+l - xi = constant = Ax.The convenience lies in the simplification of the algebra intro- duced into methods I and 11. For the former we obtain the expressions dy - Yi+l - Yi-1 - __--__ 2Ax k:= 3: - ax d2y - Y&L- 2Yi + Yi-1 _- d X 2 - (Ax) (3) for the simplest cases, k = 2 and 3. to equation (1) and, as can be seen, it is no more difficult to compute. extended and applied to spectra by Butler and H0pkins.l For method I1 we consider the simple case, k = 3 and m = 5. result from the algebraic solution of the least-squares equations. For the first derivative equation (2) is to be preferred The method has been Equations (4) and (5) dy ~ 1 (- 2Yi-2 - Yi-1 + Yi+l + 2YiS2) dx lOAx gJ ~.1 (2Yi-2 t Yi-1 - 2Yi 3- Yi+l + dx2 7Ax2 (4) It is clear that these equations contain the terms in equations (2) and (3) and additional terms arising out of the extra points used, which should give a better approximation. The method34 DERIVATIVE SPECTROSCOPY Anal. PYOC. was developed for chemists by Savitzky and Golay,2y3 and has been widely used. The Savitzky - Golay tables require that m be odd, but this is not essential; if m is even the derivatives can be calculated at the points mid-way between the data points by a similar technique. To compare the various methods, let us determine the derivatives of the functiony = x2/2 + x at x = 2, given values of y that are subject to various levels of noise as shown in Table I. The derivatives are shown in Table 11.TABLE I SIMULATED DATA X .. 0 1 2 3 4 Noise . . . . 0.2 0 0.2 -0.2 0.2 Y * . . . 0 1.5 4 7.5 12 y + noise . . 0.2 1.5 4.2 7.3 12.2 Data 1 y + noise/2 . . 0.1 1.5 4.1 7.4 12.1 Data 2 y + noise/4 . . 0.05 1.5 4.05 7.45 12.05 Data 3 y + noise/8 . . 0.025 1.5 4.025 7.475 12.025 Data 4 The significance of these results is as follows. As the noise tends to zero, the derivatives obtained by the Savitzky - Golay technique converge towards the noise-free values of 3 and 1 more rapidly than the derivatives obtained by other methods. For that reason method I1 has been the method of choice, even though it imposes the experimental restriction of equally spaced data points. Methods I and I1 share the characteristic that a low-degree polynomial (e.g., a parabola for k = 3) is fitted to a subset of the data consisting of k and m points, respectively.An advantage of this is that the methods can be applied to real-time differentiation of spectra: some manufacturers offer method I1 in microprocessor-based attachments for spectrophoto- meters. The disadvantage of method I1 is the lack of long-range continuity in the calculated derivative. Also, the choice of m must be subjective. TABLE I1 DERIVATIVES CALCULATED NUMERICALLY Data 1 2 3 4 g, equation (1) 3.1 or 2.7 3.3 or 2.6 3.4 or 2.55 3.45 or 2.525 $, equation (2) 2.9 2.95 2.98 2.987 5 2, equation (4) 2.98 2.99 2.995 2.997 5 - ::, equation (3) 0.4 0.7 0.85 0.925 2 dx2' equation (5) 1.085 7 1.042.9 1.021 4 1.010 7 Continuity is built into method 111, the spline function method. A spline function of degree 12 consists of a number of pieces each of which is a polynomial of degree h and the pieces are joined end-to-end at places on the x axis that are termed knots (or nodes).The function and its derivatives up to that of order h - 1 are continuous throughout the spectrum; the derivative of order h is constant within each polynomial piece, but changes discontinuously at the knots. The spline function is constructed as a linear combination of basis functions, which are defined completely by the knots, using the principle of least squares, and the derivatives are obtained directly from the spline function. All s data points are used so that the spline function and its derivatives (which are also spline functions) have the defined property of continuity throughout the spectrum.We have explored various schemes for knot placement with splines of degree 3, 4 and 5 and have applied the method to the analysis of Raman spectra. The signal to noise ratioJanMary , 1982 DERIVATIVE SPECTROSCOPY 35 (SNR) tends to increase as the number of knots is decreased, so that a scheme that places knots according to need is preferred to a scheme of regularly placed knots. We have attempted to make the choice of number of knots and placement scheme automatic, determined by objective criteria. The results obtained on equally spaced data indicate that the spline function method is more objective and generates derivative spectra with a higher SNR than the Savitzky - Golay method. The enhanced resolution of the second derivatives has been successfully utilised to aid the qualitative analyses of the Raman spectra of cyano - silver complexes dissolved in liquid amm~nia.~ Initial attempts to understand these spectra using the Savitzky - Golay method were vitiated by the relatively poor SNR.References The main reason for this is clearly the use of all the data simultaneously. 1. 2. 3. 4. Butler, W. L., and Hopkins, D. W., Photochem. Photobiol., 1970, 12, 439 and 451. Savitzky, A., and Golay, M. J. E., Anal. Chem., 1964, 36, 1627. Steinier, J., Termonia, Y . , and Deltour, J., Anal. Chem., 1972, 44, 1906. Gans, P., Gill, J . B., Griffin, M., and Cahill, P. C., J . Chem. SOC. Dalton Trans., 1981, 968. Square-wave Wavelength Modulation System for Use in Atomic Spectrometry” J. Sneddon, L. Bezur, R.G. Michel and J. M. Ottaway Department of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow, G1 1XL In atomic spectroscopy, wavelength modulation can be used to measure derivative spectra but a more important application is for background correction when an atomic line is super- imposed on a continuum background signal. Thus, the technique has been used for background correction in atomic-emission spectrometry using electrothermal atomisers,l ,2 flame^^-^ and plasmas,’ ,* in continuum-source atomic-absorption spectrometryg-12 and to a lesser extent in continuum-source atomic-fluorescence spectrometry.13 ,14 In many of these instances, the application of wavelength modulation has provided sig- nificant advantages, In carbon furnace atomic-emission spectrometry, detection limits were improved substantially,l in some instances by several orders of magnitude, using a quartz refractor plate mounted on a mechanical oscillator and placed near the entrance slit of a 0.75-m monochromator.In one of the early applications, Snelleman et aL3 demonstrated that the spectral interference of CaOH emission bands and/or a continuum can be minimised in the measurement of barium in a dinitrogen oxide - acetylene flame, Improvements in the detection limits of alkali and alkaline earth elements in the presence of many matrix ions were also reported and both effects were achieved by a quartz plate, vibrating at 145 Hz, and mounted behind the entrance slit of the monochromator. Epstein and O’Haver5 used a plane-parallel quartz refractor plate mounted on the shaft of a limited rotation torque motor to achieve correction for spectral interferences due to line, band and continuum radiation in flame emission spectrometry.Another interesting application in flame emission was the reduction in the interference of CH band emission on the determination of al~minium.~ One of the more important recent applications of wavelength modulation has been to background correction in continuum-source atomic-absorption spectrometry.ll ?12 A quartz refractor plate was mounted at the entrance slit of an kchelle polychromator and used to achieve simultaneous multi-element atomic-absorption analysis with automatic background correction in all channels. In the literature referred to above, wavelength modulation has invariably been achieved by employing an oscillating refractor plate. The refractor plate is usually positioned at or near the entrance slit of the monochromator and is made to oscillate about the vertical axis using commercially available scanning motors.The result of this is that the wavelength is scanned sinusoidally. Such a waveform is also highly suited to the production of derivative atomic spectra. O’Haver et aZ.15 have described computer modelling studies designed to national Confercnce on Analytical Chemistry (SAC 80) a t the university of Lancaster in June, 1980. * The system described in this paper was also demonstrated a t a Workshop Session at the Fifth Inter-36 DERIVATIVE SPECTROSCOPY Anal. Proc. determine the effects of the nature of the modulation waveform and amplitude, the shape of the modulated line and the signal processing technique used.Sinusoidal and square-wave modulation of Gaussian, triangular and Lorentzian line profiles were treated and the results indicated that the shape of the line profile has little effect, but that square-wave modulation offers a signal to noise ratio advantage approaching 2 in the 2F mode, compared with sinusoidal modulation, for a system limited by background carried photon shot noise. Koirtyohann et a1.l6 have also demonstrated experimentally that square-wave modulation gives an improve- ment over sinusoidal modulation. The improvement appears to be greater for the second compared with the first harmonic mode and increases if larger plate amplitudes are required. With scanning motors, square-wave oscillation is more difficult to achieve than sinusoidal oscillation at frequencies suitable for atomic spectrometry.Commercial equipment is avail- able,16 however, which incorporates sufficient control over the position of the refractor plate to achieve square-wave modulation at a frequency of 17 Hz. In recent reviews of wavelength modulation applied to atomic spectroscopy, O'Haverl' y18 has indicated that the oscillating refractor plate whether operated in sinusoidal or square waveform is the only current method of generating wavelength modulation. We have recently describedlg a new mechanical arrangement that incorporates a rotating quartz mechanical chopper and that results in the wavelength being repetitively scanned with a square waveform.The construction of this system and applications to atomic spectroscopy investigated to date are reviewed in this paper. When the standard refractor plate goes through an oscillation, the angle of incidence of the dispersed radiation from the monochromator at the plate changes continuously, and the spectral intensity distribution at the exit slit is modulated over a discrete wavelength interval in a sinusoidal waveform. The displacement, d, at any time is given by d = ta(n - 1)/n, where t (mm) is the thickness of the plate, cc is the angle of incidence of the light beam at the refractor plate in radians and n is the refractive index of the quartz plate. As t is constant for the quartz refractor plate, d depends only on a.The rotating quartz chopper operates on the same principle, except that in order to vary d, a is maintained constant and t is varied. The chopper was constructed of four quartz quadrants of varying thickness, mounted at an incident angle of 45" to the light beam at either the entrance or exit slits of the monochromators used. Two opposite quadrants were made of quartz of equal thickness (2.5 mm), and the other two were made of greater (4.0 mm) and lesser (1.0 mm) thickness. When the two identical quadrants were in the optical beam, light at the analyte atomic wavelength, AA, was allowed to pass through the exit slit. The other quadrants, when present in the optical beam, gave displacements to (AA + Ah) and (AA - Ah), respectively, and allowed the back- ground signal to be measured on each side of the atomic line.The modulation interval is therefore 2AX and its magnitude depends on the dispersion of the monochromator used. A detailed description and the incorporation of this system in two different spectrometers have been published elsewhere.lg~Z0 To date the rotating mechanical chopper has been used to achieve automatic background correction in flame atomic-fluorescence spectrometry,lg 921 flame emission spectrometryz1 and carbon furnace atomic-emission spectrometry.20922 Using a 300-W xenon arc as a continuum light source, correction for background signals caused by scatter and flame emission has been demonstrated in atomic-fluorescence measurements of a wide range of elements in an air - acetylene flame. Correction for flame emission background signals superimposed on atomic emission has also been demonstrated with the same instrument with the light source switched off.A new system has also been incorporated into a high-resolution echelle spectrometer used for the measurement of sensitive carbon furnace atomic-emission signals.20 In this instance, platform atomisation in an HGA-72 atomiser was used and the improvement in detection limits was achieved by a combination of factors derived from the use of the platform, high-resolution spectrometer and wavelength modulation system. The separate contributions of each factor have yet to be identified. However, the modulation system clearly offers the potential for automatic correction of background from molecular emission and matrix scatter, as well as continuum radiation from the graphite tube.The square-wave wavelength modulation system has been shown to give excellent back- ground correction in all three techniques mentioned above, when used for the determination of trace elements in clinical materials.lg~21~22 Detection limits of 20-30 pg ml-l for chromium and manganese have been used to develop simple direct methods for the determination ofJanuary, 1982 DERIVATIVE SPECTROSCOPI’ 37 these elements in blood and/or urine using carbon furnace atomic emission.22 The mechanical arrangement is no more complex than the traditional rotating glass choppers used for in- tensity modulation in spectrometric measurement. Application to continuum-source atomic- absorption measurements has yet to be demonstrated, but should be no more complicated than the above systems.It may also present a simpler means of background correction than the oscillating refractor plate for emission systems exhibiting complex spectra, such as the inductively coupled plasma. 1. 3. 1. 6. S . 9. 10. 11. 12. 13. 14. 15. 16. 17. 1s. 19. 30. 31. “ 2 . 7 -. a . - 1 References Epstein, &I. S., Rains, T. C., and O’Haver, T. C., Appl. Spectrosc., 1976, 30, 324. Hutton, R. C., Ottaway, J . M., Epstein, M. S., and Rains, T. C., Analyst, 1977, 102, 658. Snelleman, W., Rains, T. C . , Yee, K. W., Cook, H. D., and Menis, O., Anal. Chem., 1970, 42, 394. Rains, T. C., and Xenis, O., Anal. Lett., 1974, 1, 715. Epstein, M. S., and O’Haver, T. C., Spectrochim. Acta, 1975, 30B, 135. Syder, R.J., and Hieftje, G. M., Anal. Chprn., 1976, 48, 535. Iiawaguchi, H., Okada, M., Ito, T., and Mituike, A., AmaZ. Chirn. Acta, 1977, 95, 143. Rose, O., Mincey, D. W., Yacynych, A. M., Heineman, W. R., and Caruso, J . A, AnalJfst, 1976,101, 753. Zander, A. T., O’Haver, T. C . , and Keliher, P. N., Aazal. Chern., 1976, 48, 1166. Snelleman, W., Spectrochzm. Acta, 1968, 23B, 403. Harnley, J . M., and O’Haver, T. C., Anal. Chem., 1977, 49, 2187. Harnley, J . M., O’Haver, T. C., Golden, R., and Wolf, W. R., Anal. Chem., 1979, 51, 2007. Fowler, IV. K., Knapp, D. O., and Winefordner, J. D., Anal. Chem., 1974, 46, 601. Lipari, F., and Plankey, F. W., Anal. Chem., 1978, 50, 386 O’Haver, T. C., Epstein, M. S., and Zander, A. T., Anal. Chem., 1977, 49, 458. Koirtyohann, S. K., Class, E.D., Yates, D. A,, Hinderbnrger, E. J., and Lichte, F. E., Anal. Chet?z., O’Haver, T. C., Anal. Chem., 1979, 51, 91A. O’Haver, T. C., “Contemporary Topics in Analytical and Clinical Chemistry,” Plenum, New York, Illichel, R. G., Sneddon, J., Hunter, J . I<., Ottaway, J . M., and Fell, G. S., Analyst, 1981, 106, 288. Ottaway, J. M., Bezur, L., and Marshall, J., Awlyst, 1980, 105, 1130. Ottaway, J . M., Hall, M. L., Michel, R. G., Sneddon, J , , and Fell, G. S., in “Trace Element Analvtica Chemistry in Medicine and Biology,” Walter de Gruyter, Berlin, New York, 1980, p. 255. Ottaway, J. M., Bezur, L., Fakhrul-Aldeen, R., Frech, W., and Marshall, J., in “Trace Element Analytical Chemistry in Medicine and Biology,” Walter de Gruyter, Berlin, Yew York, 1980, p.575. 1977, 49, 1121. 1979, p. 1. Derivative Fluorescence Spectroscopy J. N. Miller and T. A. Ahmad and A. F. Fell Defircrtment of Chemistry, Lozdghborough University of Technology, Loughbovough, Leicestershire, LE11 3T U Ihpnvtment of Pharmacy, Heriot- Watt University, Edinburgh, E H 1 2HJ Fluorescence spectroscopy finds its main applications in two fields : (i) biochemical studies, including clinical, pharmacological and pharmaceutical analysis, and (ii) environmental analyses. These fields have in common a number of experimental problems; in particular, they both require extremely sensitive analytical methods (because of the small samples and/or low analyte concentrations) and they also demand highly selective techniques (because of the extreme complexity of the samples).The sensitivity of fluorescence spectroscopy is well established : picogram concentrations of strongly fluorescent solutes can be determined using modern instruments. I n principle, the selectivity of fluorimetry should also be superior to that of absorption spectroscopy, as each solute can be characterised by two wavelengths, those of excitation and fluorescence. In practice, this advantage is often more apparent than real, as the large band widths of fluorescence signals often yield strongly overlapping spectra. Additional selectivity can be obtained by combining fluorimetry with separation techniques or with biochemically specific methods, or by the use of modified forms of spectroscopic analysis. Amongst the latter, derivative methods have recently received much a t t e n t i ~ n .l - ~ A further approach, not possible in absorption spectrometry, is that of synchronous scanning fluorimetry* 9 5 ; combination of the derivative and synchronous methods was suggested by John and SoutarS6 Derivative fluorescence spectroscopy, both alone and in combination with38 DERIVATIVE SPECTROSCOPY Anal. Proc. synchronous scanning, has thus far been largely applied to environmental analysis. This paper describes some applications in the field of analytical biochemistry. Experimental Derivative luminescence spectra (second derivative in this study, unless otherwise stated) can be obtained from conventional excitation , fluorescence and synchronous spectra by electronic (ie. , analogue) differentiation or by numerical (digital) methods. In the work described here the former approach was exemplified by the use of a Fluoricord spectrofluori- meter (Baird-Atomic, Braintree, Essex), in conjunction with a Hitachi derivative module (Model 200-0507, obtained from Perkin-Elmer, Beaconsfield , Buckinghamshire) ; this fluori- meter generated spectra that were uncorrected for instrumental characteristics.Derivative spectra computed numerically were obtained using a Perkin-Elmer MPF 44B spectrofluori- meter equipped with a DCSU-2 microprocessor unit ; this unit was sometimes used to generate corrected spectra in addition to derivative spectra. All spectra were obtained at room temperature using samples of the highest available purity; water was distilled at least twice from a silica still.Results and Discussion The major benefits of derivative techniques are expected to lie in the increased resolution of overlapping spectral bands and in the enhancement of relatively minor spectral features. Both of these advantages are well demonstrated in Fig. 1, which shows the fluorescence excitation spectrum of fluorescein, and the second derivative of the same spectrum. A further advantage is illustrated in Fig. 2, where it is shown that a fluorescence peak superimposed on a sloping base line (in this instance fluorescein in a blood serum sample with high background fluorescence) can be readily separated from the background signal. However, it is important to note that the effectiveness of derivative spectroscopy is a function of the band width of the zero-order spectrum : spectra with large band widths will, other circumstances being equal, yield less intense derivative signals than spectra with narrow band widths.This point is 250 350 450 5"3 I l n m Fig. 1. Fluorescence excitation spectrum of fluorescein (A! = 510 nm) and its second derivative obtained using the DCSU-2 unit. 3 Vnrn Fig. 2; Zero-order and second-derivative (DSCU-2) spectra of fluorescein in a diluted serum sample. In the zero-order spectrum the fluorescein fluorescence peak (A! = 515 nm) overlaps the serum background (A! = 465 nm), but the second-deriva- tive spectrum allows the fluorescein to be deter- mined without background interference. Excitation wavelength : 400 nm.January, 1982 DERIVATIVE SPECTROSCOPY 39 well illustrated in Fig.3, which illustrates a broad fluorescence band with a sharp Raman scattering band superimposed on i t ; thc second derivative spectrum shows the Raman band clearly, but the fluorescence band is scarcely detectable. In these circumstances the synchronous scanning approach may be extremely valuable. This method involves scanning both the fluorimeter monochromators simultaneously with a fixed wavelength difference between them. The result is the production of a spectrum with a much narrower band width.5 Synchronous spectra may therefore be capable of conferring increased selectivity on their own, but in other instances the additional resolution provided by derivative spectroscopy is necessary. These principles have been applied to the solution of a well known problem in analytical biochemistry, the detection of tyrosine fluorescence in the presence of tryptophan.These two amino acids have very similar excitation spectra, and their fluorescence spectra overlap strongly. In addition, as the molar absorptivity of tryptophan at 280 nm is about four times greater than that of tyrosine, a solution of tryptophan is approximately 2.25 times as fluorescent as an equimolar solution of tyrosine excited at the same wavelength [the quantum yields are (4f)trp = 0.12 and (4f)tyr = 0.211. When the two amino acids are incorporated in the same protein molecule, the identification of tyrosine fluorescence is made even more difficult by two further factors: there is a blue shift in the fluorescence maximum of the tryptophan residues (i.e., increasing the spectral overlap) and the tyrosine fluorescence intensity is reduced by energy transfer and quenching effects.An extra problem is the small Stokes shift of tyrosine (Aex- m 280 nm; Afl. m 303 nm), which may introduce interference by Rayleigh and Raman scattered light signals.' Thus, although it is easy to detect tryptophan fluorescence in the presence of tyrosine (e.g., by using excitation wavelengths longer than 295 nm, when 400 500 600 )i/n rn Fig. 3. Derivative spectro- scopy emphasising spectral features with narrow band widths. In the zero-order spectrum a fluorescamine- labelled amino acid ( A t = 480nm) is overlapped by a Raman peak (A = 420 nm); in the second-derivative (DSCU-2) spectrum, only the Raman signal is clearly defined.Excita- tion wavelength = 370 nm. I 300 400 300 400 hdnm Lr.Jnm Fig. 4. Use of synchronous scanning to produce well defined derivative spectra ; (a) fluorescence emission spectrum of tryptophan (-20 pg ml-l) and its second derivative (DSCU-2, excitation wave- length 287 nm) ; (b) synchronous spectrum. of the same solution (AA = 58 nm) and its second derivative. The double-headed arrows show that the synchronous spec- trum has less than half the band width of the conventional spectrum.40 DERIVATIVE SPECTROSCOPY Anal. Proc. tyrosine does not absorb) the detection of tyrosine fluorescence in the presence of tryptophan may be exceedingly difficult. The problem is not only important in its own right (e.g., in studies of protein photochemistry and structure), it also exemplifies many of the problems encountered in numerous other fluorimetric assays in biochemistry.A similar problem, that of resolving tyrosine and phenylalanine fluorescence, has been satisfactorily resolved using derivative spectroscopy alone.8 This approach is not available when tryptophan is under study, because of the large band width (about 65 nm at half-height) of tryptophan fluorescence : even intense zero-order fluorescence spectra yield only weak (second) derivative spectra. The synchronous spectrum of tryptophan, obtained at a mono- chromator wavelength difference (Ah) of 58 nm, has a band width of only about 25 nm, and hence yields excellent second-derivative spectra (Fig. 4). In an earlier papers it was shown that tyrosine and tryptophan might be resolved by synchronous spectroscopy alone, Ah values of less than 15 nm giving spectra characteristic of tyrosine, and Ah values of greater than 60 nm giving spectra characteristic of tryptophan.An example of the quantitative applica- tion of this procedure, both with and without the use of derivative methods, is given in Fig. 5 . An alternative and less cumbersome approach is to use a synchronous scanning interval, which reveals both tyrosine and tryptophan components; these components can then be resolved using second-derivative spectroscopy (Fig. 6). This separation can be achieved 0 2 4 6 8 1 0 TyrIpM Fig. 5. Determination of tyrosine in the presence of tryptophan using (a) syn- chronous and (b) second- derivative synchronous fluori- metry (AA = 10 nm).All solutions were made up so that the total Tyr + Trp concentration was 10 p ~ . I n (b) the 2 D ~ line uses the long-wavelength satellite of the derivative spectrum to measure intensities, and the 2 D ~ line the short-wavelength satellite . 230 330 21 0 310 hexjnm 111111- 260 360 270 370 hernjnm Fig. 6. Resolution of tyrosine and tryptophan fluorescence using derivative synchronous spectroscopy with AA values of (a) 30nm and (b) 60nm. For details, see text. either at moderate Ah values (about 30 nm) or at much larger AX values (70 nm). In the latter instance use is made of a small but well defined peak that appears in the excitation spectrum of tyrosine at 234 nm. These methods have also been used with success in resolving three- component mixtures (phenylalanine, tyrosine and tryptophan) and in studies of proteins.January, 1982 DERIVATIVE SPECTROSCOPY 41 For example, in agreement with earlier results obtained using more complex technique^,^^"-' the tyrosine contribution to the fluorescence of lysozyme was found to be very small, whereas the tyrosine contribution to human serum albumin fluorescence was substantial.A number of other mixtures of biologically active materials have been studied and these results will be reported separately.ll Conclusion The advantages of derivative spectroscopy are just as applicable in fluorimetry as in In fluorimetry the additional device of synchronous scanning is Derivative synchronous spectra will Using second-derivative spectra, A number of fields in which derivative Replacement of the chromatographic step in many quantitative analyses that currently utilise HPLC or TLC to provide selectivity.Clearly there will be instances where closely similar materials cannot be resolved spectroscopically,l2 but equally there will be many cases where existing procedures can be improved upon. (iz) Combinations of derivative spectroscopy and phosphorimetry, especially at room temperature.13 Room-temperature phosphorescence (RTP) is of considerable value in, for example, environmental analyses, but RTP spectra are less sharply featured than those obtained at 77 K. Derivative methods will yield more characteristic spectra of added value in qualitative and quantitative analysis. These molecules or fluorophores exhibit changes in their fluorescence wavelengths and intensities as their environments change, on binding to a macromolecule, for e~arnp1e.l~ In such instances derivative methods could distinguish between probe molecules in different environments, and this approach would be of great importance in the study of drug - protein binding interactions, and in the development of fluorescence immunoassays.Studies of this type are proceeding in the authors’ laboratory. (iv) No great technical difficulty attaches to the determination of fourth-derivative (and higher order) spectra. These will merit further study, although the complexity of the derivative spectra themselves and signal to noise ratio problems may offset the potential advantages of still greater discrimination and resolution 0btainab1e.l~ absorption spectroscopy.of help in studying solutes with broad band spectra. thus be of special value in resolving difficult mixtures. quantitative analysis should present few problems. luminescence spectroscopy will be valuable can be identified. (i) These include : (iii) The use of derivative fluorimetry in combination with fluorescence probes. We thank the Medical Research Council for provision of funds to buy the recording spectro- fluorimeter, and Dr. C. S. Lim for generous assistance. References 1. O’Haver, T. C., in Wehry, E. L., Editor, “Modern Fluorescence Spectroscopy,” Plenum, New York, 1976. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 18. Green, G. L., and O’Haver, T. C., Anal, Chem., 1974, 46, 2191. O’Haver, T. C., Clin. Chem., 1979, 25, 1548. Lloyd, J . B.F., Nature Phys. Sci., 1971, 231, 64. Lloyd, J . B. F., J . Forensic Sci. Soc., 1971, 11, 83. John, P., and Soutar, I., Anal. Chem., 1976, 48, 520. Miller, J . N., Prog. Biophys. Mol. Biol., 1974, 28, 41. Terhaar, D. A., and Di Cesare, J . L., Perkin-Elmer Fluorescence Report, 1979, NO. F2-71. Miller, J . N., Proc. Anal. Div. Chem. Soc., 1979, 16, 203. Miller, J . N., and King, L. A., Biochim. Biophys. Acta, 1975, 393, 433. Miller, J. N., and Fell, A. F., in preparation. Fell, A. F., and Miller, J . N., unpublished work. Vo-Dinh, T., and Gammage, R. B., Anal. Chim. Acta, 1979, 147, 261. Edelman, G. M., and McClure, W. O., Acc. Chem. Res., 1968, 1, 65. Cahill, J . E., 30th Pittsburgh Conference, Cleveland, Ohio, 1979, Abstract No. 174.42 DERIVATIVE SPECTROSCOPY Anal.PYOC. Derivative Spectroscopy in the Laboratory: Advantages and Trading Rules Brian P. Chadburn Perkin-Elmer Ltd., Beaconsfield, Buckinghamshive The derivative technique is useful for extracting both qualitative and quantitative information from spectral curves composed of unresolved bands. However, its application involves the choice of two new parameters, and an inappropriate choice can artificially limit the advantages of the technique and lead to an excessive reduction in the signal to noise ratio. In this study, the advantages of derivative spectroscopy are considered and the trading rules involved in its application in ultraviolet - visible spectroscopy are discussed. Derivative spectroscopy is a simple yet powerful technique for enhancing the fine structure of spectral curves.l It involves plotting the first, second or higher order derivative of a spectrum with respect to wavelength rather than the spectrum itself.Usually, the derivative is obtained by an electronic RC or microcomputer device in series with the spectrophotometer and plotted as the spectrum is scanned. This process acts as a filter that favours the short- range changes in spectrum. The result is an enhancement of fine structure, which is accom- panied by a decrease in the signal to noise ratio (S/N). The technique is especially useful in areas such as the visible and ultraviolet spectra of liquid solutions, which usually have high signal to noise ratios but are lacking in structure. In this present paper the advantages of the derivative technique are identified and the trading rules between S/N, derivative order and the wavelength range over which the derivative is averaged are discussed. Advantages When the even derivative spectrum is plotted, two general advantages result.The first is an effective enhancement of resolution, which can be extremely helpful when a solution contains two or more components whose spectra overlap. This effective enhancement of resolution will often result in a separate feature for one or more of the components, which can then be used for quantitative measurements. The second general advantage is a dis- crimination in favour of the sharper features of a spectrum.2 Through this discrimination effect the signal from that component of a solution whose spectrum contains the sharpest structure will be enhanced relative to the signals from the other components whose structure is broader.This effect is useful for selectively eliminating interference by broad-band con- s tituents . Trading Rules The following equation can be derived for S/N for nth derivative spectrum : where the subscripts 0 and n indicate the zero-order and nth derivative spectrum, W is the full band width at half-maximum (FWHM) and AA is the wavelength range over which the derivative is averaged, and is assumed to be small relative to W. The optimum values of n and AA will vary with each application, and this must be determined by trial and error. Ah should be set as high as the required discrimination and resolution permit in order to obtain good S/N. While this will distort the derivative shape, the linearity of the derivative signal with concentration is not affected.Instrumentation The advent of microcomputer ultraviolet - visible spectrophotometers has realised numerous benefits in terms of a faster set-up procedure and more accurate results. Using the numerical keyboard entry, any value for the ordinate minimum and maximum readings can be con- veniently used for the recorder scale. Derivative spectroscopy on a modern microcomputer instrument is now a standard feature, with the inherent flexibility and convenience of aJanuary, 1982 DERIVATIVE SPECTROSCOPY 43 numerical keyboard entry for many of the parameters. The derivative technique is useful in those situations where resolution enhancement can greatly increase the fingerprinting use of ultraviolet spectra and sometimes will separate components in a mixture.This discrimina- tion effect separates spectra with sharp structure from broad interferences for both identifica- tion and quantitation. Although UV spectra have been considered explicitly in this study, the conclusions apply to any other composite curves such as infrared spectra and chromato- grams. Both the advantages and the noise increase with derivative order; the second derivative is almost always more useful than the first derivative. In addition to greater resolution and discrimination, the characteristic second-derivative minimum at the peak of an individual sub-band is easier to identify than the first-derivative shape. The applicability of the higher derivatives is often limited by noise and fine structure in the background.The derivative must be generated by averaging over a finite wavelength range. This wavelength range should generally be about half the width (FWMH) of the sub-bands when resolution is desired, and about equal to the FWMH for the best S/N in quantitative work, where a broader interference is to be suppressed. The distortion of the derivative by these large values is more than com- pensated for by the ability to trade the large resultant S/N for greater matrix suppression. Current developments in analytical instrumentation point to the increased use of the derivative method in spectroscopy and chromatography. References 1. 2. O’Haver, T. C., and Green, G. L., Int. Lab., 1975, 5(3), 11. Cahill, J .E., Int. Lab., 1980, 10(1), 64. Extending the Applications of Derivative Spectrophotometry Christopher T. Cottrell Pye Unicam Ltd., Cambridge The absorbance peaks encountered in ultraviolet - visible spectrophotometry are by nature normally broad, owing to the overlap of the many molecular transitions taking place. For this reason, the precise determination of A,,,. of a broad peak or resolution of multi-component mixtures where peak overlap is present can be difficult. The systematic error encountered with background or sample turbidity can cause similar problems, unless care is taken prior to sample measurement. Many techniques have been proposed for overcoming or reducing these analytical problems, e.g., mathematical correction, variation of measurement wavelength and adaptation of instrumentation.Unfortunately, they all suffer from a variety of compromises, such as time, cost and possibility of introducing errors in the interpretation of the data. One of the more elegant alternatives is the application of derivative spectrophotometry, which was first described by Hammond and Price1 in 1953 and was soon followed by the work of Morrison2 and French et aZ.3 Theoretical aspects of the technique were described by Giese and French4 and derivative orders of second and higher were discussed by Martin.5 Further studies by O’Haver and Green6 concerned the performance of the technique when applied to various band overlap situations, complemented by a discussion of the possible systematic and random errors associated with various methods of measurement .’ The types of presentation obtained by the application of first to fourth derivatives on a variety of peak configurations have been extensively reviewed.8-11 Various instrumental methods have been developed for the generation of derivative spectra.6s10J2 The approach adopted in the design of the Pye Unicam derivative accessory module was that of electronic differentiation of the spectrophotometer analogue output.It was found that this principle allowed the full advantages of the derivative technique to be applied, particularly the ability of using a number of accessory modules in series to generate derivatives up to fourth order, as well as allowing the presentation of derivative spectra at a fixed wavelength for applications such as densitometry and kinetics, etc.Because the derivative signal is, in fact, obtained with respect to time rather than wave- length, the wavelength scanning speed significantly affects both the amplitude of the derivative44 DERIVATIVE SPECTROSCOPY Anal. Proc. spectrum and also the effect of background noise. Thus, increasing the scanning speed will present a more rapidly changing absorption signal to the derivative circuit, with a significant improvement in the apparent signal to noise ratio. Applications and Advantages of Derivative Spectrophotometry Derivative spectrophotometry offers a convenient solution to a number of well defined analytical problems, such as resolution of multi-component systems, removal of sample turbidity, matrix background and enhancement of spectral details.The following briefly describes the major advantages of the application of derivative techniques in ultraviolet - visible spectrophotometry. Applications of derivative techniques in other areas where significant improvements in both analytical performance and speed of analysis can be expected are discussed. Derivative spectrophotometry offers an extremely valuable means of resolving fine detail barely visible in the normal, zero-order, spectrum. For liquid samples, second and fourth derivative orders are the most convenient owing to their ease of interpretation. For resolution of gaseous sample spectra, however, the second derivative is likely to be the most usefu1.13-15 The first derivative for the precise determination of sample maximum absorption wavelength can be very useful, especially when dealing with broad absorbance peaks, where accurate and precise estimations of the latter can prove to be something of a cornpromise.ls-l8 One of the classic analytical problems for any worker in the field of ultraviolet - visible spectrophotometry is the resolution of a number of components in a mixture. For applications of this type, the use of second and fourth derivatives offers considerable advantages.Where small peaks or shoulders appear on the side of a sloping major component background, increasing derivative orders will progressively flatten the major peaks to almost a straight line and resolve the minor components more sharply. Increasing derivative orders will also increase the apparent band sharpening and will, therefore, provide an increase in analytical sensitivity.Most publications concerning derivative ultraviolet - visible spectrophotometry have featured the application of the technique to multi-component resolution and this, in itself, aptly demonstrates the importance of this approach. Derivative techniques have been successfully applied to biological16 and biochemical problems, particularly amino acidslO 3 1 9 s20 and proteins,20-22 and to inorganic1* and pharmaceuticalll systems. Thus, when scanning in the derivative mode, the gradually increasing turbidity will not cause any marked change in the spectrum and will therefore be substantially eliminated, allowing the absorbing component to be readily resolved. This effect of the derivative on sample turbidity suggests a wide variety of applications particularly in water analysis,lO the analysis of pharma- ceutical preparationsll and the examination and measurement of components in biological matrices.23 - 26 Turbidity generally produces an absorbance that increases as wavelength decreases. Extended Applications of the Derivative Technique The derivative technique can also enhance spectral detail generated by methods other than straightforward ultraviolet - visible spectrophotometry. Colour measurement, liquid chroma- tography and densitometry are areas where derivative spectrophotometry can, if applied with care, provide improvements in both qualitative and quantitative information, and significantly reduce analysis time. The technique of colour measurement using an integrating sphere is now an accepted means of obtaining diffuse or total reflectance using tristimulus values and chromaticity co-ordinates, biased for the accepted illuminants and observer angles.In certain situations, the application of the derivative method can provide a useful means of accentuating small differences in detail. The enhancement of analytical detail from HPLC records using the first-derivative approach has been reported using a rapid-scanning diode array spectrometer% to identify small amounts of benz(a)anthracene co-eluting with chrysene, in a reversed-phase separation of atmospheric particulate^.^^ The application of second and fourth derivatives and the implications of the method for quantitation of HPLC data have recently been proposed and demon~trated.~~,~~ The effect of applying the second derivative to a zero-order liquid chromatogram has been examined in our laboratories, using a Pye Unicam LC-XPD pump, SPS-150 spectrophotometer and the liquid chromatography accessory.8 A 2O-pl sample of a mixture of polynuclearJnnzinry, 1982 WATER RESEARCH CENTRE TECHNICAL REPORTS 45 aromatics in an acetonitrile - water eluent was separated using a Partisil 10-ODs-2 column with detection at 245 nm.The application of the second derivative produced an apparent 10-fold improvement in sensitivity. The application of derivatives may prove to be an extremely convenient means of obtaining improved resolution if it has been found impossible to improve conditions further by chromatographic means. The application of derivative spectrophotometry as a means for expanding analytical detail in densitometry has been described by Machold et aL31 and 0kamu1-a.~~ Machold et al.separated various chlorophyll protein complexes by polyacrylamide gel electrophoresis and enhanced the spectral scan detail of fixed pigment-complex zones using the second derivative. Okamura used the second derivative of absorbance versus distance to resolve the ill components in the globulin fraction of human serum proteins, previously separated by cellulose acetate membrane electrophoresis. Lsing a cell compartment densitometer, described by Crane,33 the derivative technique has been applied to a variety of samples, including an autoradiograph obtained by the electrophoretic separation of a complex mixture of radioactively labelled (33S) proteins.The second derivative significantly improved the apparent resolution by reducing the effect of the background and sharpening the absorbance bands such that quantitation could be readily attempted.8 The possibility of applying derivative densitometry to normal acrylamide gels, and more interestingly unstained gels, measured in the ultraviolet region, is being investigated further. The added advantage of being able to scan the zone of interest at a fixed point in the gel with second or fourth derivative recording suggests a powerful new analytical tool for the haematologist. References 1. 3. 4. 5. 6. S. 9. 10. 11. 12. 13. 14. 15. 16. 17. I S . 19, 20. 21. 22. 23. 24. 25. 26. 2 7 . 2s 29. 30. 31. 32. 33. > -. I . Hammond, V. J . , and Price, It'. C. J., J . Opt. Soc. Am., 1953, 43, 924. Morrison, J . D., J . Chem. Phys., 1953, 21, 1767. French, C. S., Church, A. B., and Epplev, R. IV., Cavnepie Inst. Wash. Yeavb., 1954, 53, 182. Giese, A. T., and French, C. S., Apil: Spectrosc., 1955, 9, 78. Martin, A. E., Spectrochim. Acta, 1969, 14, 97. O'Haver, T. C., and Green, G. L., Int. Lab., 1975, 6(5). 11. O'Haver, T. C., and Green, G. L., Anal. Chem., 1976, 48, 312. Cottrell, C. T., "Derivative and Log Spectrophotometry," SP8 Series Accessories Applications, Pye Unicam, Cambridge, 1980. O'Haver, T. C., Clzn. Chem., 1979, 25, 1548. Talsky, G., Mayring, H., and Kreuzer, H., Angew. Chem., Int. Ed. Engl., IS Fell, 4. F., Proc. Anal. Div. Chem. Soc., 1978, 15, 260. 178, 17, 783. Gunders, E., and Kaplan, B., J . Opt. SOC. Am., 1965, 55, 1094. Grum, F., Paine, P., and Zoeller, L., Appl. Opt., 1972, 11, 93. IVilliams, D. T., and Hager, R. N., Appl. Opt., 1970, 9, 1597. Hager, R. N., Anal. Chem., 1973, 45, 1131.4. Shiga, T., Shiga, K., and Kuroda, M., Anal. Biochem., 1971, 44, 291. IVahbi, A. M., and Ebel, S., Anal. Chim. Acta, 1974, 70, 57. Shibata, S., Angew. Chem., I n t . Ed. Engl., 1976, 15. 673. Fell A. F.. I . Pharm. Pharmacol.. 1979. 31. 23P. , ~~ ~ Fell, A. F.: "CrV Spectrom ~ r o u p E u i . , 1979; 71 5. Natsushima, >I., Inoue, I., and Shibata, K., Anal. Biochem., 1975, 65, 362. Brandts, J . F., and Kaplan, L. J , , Biochemistry, 1973, 12, 2011. Schtnitt, .I., J . Clin. Chem. Clin. Biochemistry, 1977, 15, 303. Tones, K. G., and Sweenev, G. D., Clin. Chem., 1979, 25, 71. gotten, D., Instvum. Are&, 1975, 25, 14. French, C. S., and Harper, G. E., Carnegze Inst. Wash. Yearb., 1957, 56, 281. Crane, R. T., "Colour Measurement," SP8 Series -4ccessories Applications, Pye Unicam, Cambridge, Milano. >I. I.. and Grushka. E.. 1. Chromatoev.. 1977. 133. 352. 1978. Fox, li. and Staley, S. w., x n a l . Chem.,s1976, 481 992: Fell, A. F., Anal. Proc., 1980, 17, 512. Machold, O., Meister, -4., and Adler, K., Photosynthetica. 1971, 5, 160. Okamura, K., Clin. Chim. Acta, 1979, 96, 273. Crane, R. T., "Densitometry," SP8 Series Accessories Applications, Pye Unicam, Cambridge, 1978. Water Research Centre Technical Reports The following recently published reports are AIedmenham Laboratory (P.O. Box 16, Henley available from the LVater Research Centre, Road, Medmenham, Marlow, Buckinghamshire,46 EQUIPMENT NEWS SL7 2HD) or Stevenage Laboratory (Elder Way, Stevenage, Hertfordshire, SGl 1TH) : TR 15 I The Calculation of Equilibrium TR 158 TR TR Trace Metal Speciation and Solu- bility in Aqueous Systems by a Computer Method, with Particular Reference to Lead. 52 Calculation of Lead Solubility in 53 Use of Chelsting Ion-exchange TR TR Water. 59 60 Artal. PYOC, Resins for the Determination of Trace Metals in Drinking Waters. A Survey of Polycyclic Aromatic Hydrocarbon Levels in British Waters. Organic Micropollutants in Drinking Water. A Review of Photo-oxidation for the Determination of Total Organic Carbon in Water.

ISSN:0144-557X    

DOI:10.1039/AP9821900022

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

46 EQUIPMENT NEWS Artal. PYOC, Equipment News Further details of all items reported below are available from the companies concerned. For rapid information please complete the Reader Enquiry Service card (circle the appropriate number mentioned), rather than approaching companies direct. Conductivity Meter The CDM83 measures conductivity from 1300 pS cm-l to 1300 mS cm-l full scale, with auto- matic switching to the optimal range and frequency. Temperature correction to a refer- ence temperature and full adjustment of the cell constant are provided. In addition to conductivity, the 20-character alphanumeric display may simultaneously show sample or reference temperature from - 10.0 "C to + 105.0 OC, or temperature coefficient or cell constant. Radiometer A/S. Circle 510. GC - MS Column for Drug Analysis The HP59858 GC - LC - MS fused silica system permits the separation of acidic and basic drugs on a single column.Hewlett Packard Ltd. Circle 511. Electrophoresis Power Supply The Model 500/200 has a range of 0-500 V and 0-200 mA; either parameter can be set to a constant level. Overload protection is provided. Bio-Rad Laboratories Ltd. Circle 5 12. Photoacoustic Cell The Model 410.10 has been designed for use in Fourier transform infrared and laser-based systems. The cell accepts samples up to 6 mm in diameter by 2 mm thick and entrance windows are interchangeable. A power-supply and pre-amplifier with 1OOx gain and 20 kHz band width are provided. EDT Research. Circle 513. Cryo-microtome The Cryo-Cut I1 incorporates the Spencer rotary microtome, fast-feed control, automatic de-frost cycle and quick-freeze facility. Reichert- Jung UK.Circle 514. Filters for Liquid Chromatography The SSI pump inlet type is rated at lOpm, surface area 7.7 cm2 and total internal volume 1150 p1. The high pressure line type is rated at 0.5 pm, 2.5 cm2 and 580 pl. The element is corrosion resistant. Scientific Glass Engineering (UK) Ltd. Circle 51 5. Laser Ion Mass Analyser This analyser measures the concentrations at the p.p.m. level of elements in the relative46 EQUIPMENT NEWS Artal. PYOC, Equipment News Further details of all items reported below are available from the companies concerned. For rapid information please complete the Reader Enquiry Service card (circle the appropriate number mentioned), rather than approaching companies direct.Conductivity Meter The CDM83 measures conductivity from 1300 pS cm-l to 1300 mS cm-l full scale, with auto- matic switching to the optimal range and frequency. Temperature correction to a refer- ence temperature and full adjustment of the cell constant are provided. In addition to conductivity, the 20-character alphanumeric display may simultaneously show sample or reference temperature from - 10.0 "C to + 105.0 OC, or temperature coefficient or cell constant. Radiometer A/S. Circle 510. GC - MS Column for Drug Analysis The HP59858 GC - LC - MS fused silica system permits the separation of acidic and basic drugs on a single column. Hewlett Packard Ltd. Circle 511. Electrophoresis Power Supply The Model 500/200 has a range of 0-500 V and 0-200 mA; either parameter can be set to a constant level.Overload protection is provided. Bio-Rad Laboratories Ltd. Circle 5 12. Photoacoustic Cell The Model 410.10 has been designed for use in Fourier transform infrared and laser-based systems. The cell accepts samples up to 6 mm in diameter by 2 mm thick and entrance windows are interchangeable. A power-supply and pre-amplifier with 1OOx gain and 20 kHz band width are provided. EDT Research. Circle 513. Cryo-microtome The Cryo-Cut I1 incorporates the Spencer rotary microtome, fast-feed control, automatic de-frost cycle and quick-freeze facility. Reichert- Jung UK. Circle 514. Filters for Liquid Chromatography The SSI pump inlet type is rated at lOpm, surface area 7.7 cm2 and total internal volume 1150 p1.The high pressure line type is rated at 0.5 pm, 2.5 cm2 and 580 pl. The element is corrosion resistant. Scientific Glass Engineering (UK) Ltd. Circle 51 5. Laser Ion Mass Analyser This analyser measures the concentrations at the p.p.m. level of elements in the relative48 ANALYTICAL PROCEEDINGS January, 1982 Analytical Sciences Monographs No. 4 Electrothermal Atomisation for Atomic Absorption Spectrometry by C. W. Fuller Since the introduction of atomic absorption spectrometry as an analytical technique, by Walsh, in 1953, the use of alternative atomization sources to the flame has been explored. At the present time the two most successful alternatives appear to be the electrothermal atomiser and the inductively- coupled plasma. In this book an attempt has been made to provide the author's views on the historical development, commercial design features, theory, practical considerations, analytical parameters of the elements, and areas of application of the first of these two techniques, electrothermal atomisation. Hardcover 135pp 0 85186 777 4 fl8.00 ($38.00) RSC Members fl3.50 No. 5 Dithizone by H.M. N. H. Irving The author of this monograph, who has been closely associated with the development of analytical techniques using this reagent for man! years, and who has made extensive investigations into the properties of its complexes, has gathered together a body of historical and technical data that will be of interest to many practising analytical chemists. Hardcover 112pp 0 85186 787 1 f12.50 ($27.00) RSC Members f9.50 No.6 lsoenzyme Analysis Edited by D. W. Moss This monograph attempts to draw together the most important experimental techniques which have resulted from the modern recognition that enzymes frequently exist in multiple molecular forms. This monograph also indicates the advantages and limitations in isoenzyme studies of these modern experiments. Brief Contents: Multiple Forms of Enzymes; Separation of Multiple Forms of Enzymes; Selective Inactivation of Multiple Forms of Enzymes; lmmunochemistry of Multiple Forms of Enzymes; Catalytic Differences between Multiple Forms of Enzymes, Methods of Obtaining Structural Information, Selection of Methods of Analysis. Hardcover 171pp 0 85186 800 2 f12.00 ($26.00) RSC Members f9.00 No. 7 Analysis of Airborne Pollutants in Working Atmospheres The Welding and Surface Coatings Industries by J.Moreton and N. A. R. Falla This Monograph covers the following: Part I The Welding Industry: Airborne Pollutants in Welding; Sampling of Welding Workshop Atmospheres; Analysis of Welding Fumes and Pollutant Gases. Part II The Surface Coatings Industry: Origin of Airborne Pollutants in the Surface Coatings Industry; Collection and Analysis of Gaseous Atmospheric Pollutants in the Surface Coatings Industry; Collection and Analysis of Particulate Atmospheric Pollutants in the Surface Coatings Industry; Future Trends Relating to Sampling and Analysis in the Welding and Surface Coatings Industries. Hardcover 192pp 0 85186 860 6 f15.00($32.00) RSC Members f12.00 No. 8 The Sampling of Bulk Materials by R.Smith and G. V. James The literature of analytical chemistry exhaustively covers the many techniques now available to the analyst. feature common to a l l analyses, is in contrast only sparsely documented. Comparatively few original papers on this subject have been published in the last fifty years; there are very few reviews available, and perhaps as a result sampling is badly neglected in most instructional courses in analytical chemistry. This Monograph will go some way towards filling a gap in the literature and should stimulate interest in the development of sampling as a field of study. Brief Contents Introduction; Glossary of Terms; Establishment of a Ssmpling Scheme; Sampling Theories; Apparatus for Sam2ling; Sampling Methods; Appendices 1-4.Hardcover 200pp 0 851 86 81 0 X f16.50 ($35.00) RSC Members f10.75 Orders: RSC Members should send their orders to: The Membership Officer, The Royal Society of Chemistry, 30 Russell Square, London WC1B 5DT All other orders should be Sent to: The Royal Society of Chemistry, Distribution Centre, Blackhorse Road, Letchworth. Herts. SG6 1 HN Sampling, the one The Royal Society of ChemistryJanuary, 1982 ANALYTICAL PROCEEDINGS 49 RSC Publications S PEClALlST PER 10 DICAL R EPO RTS Nuclear Magnetic Resonance Vol. 10 Senior Reporter: G. A. Webb A review of the literature published between June 1979 and May 1980. Brief Contents : Theoretical and Physical Aspects of Nuclear Shielding; Applications of Nuclear Shielding; Theoretical Aspects of spin-spin couplings; Applications of spin-spin couplings; Nuclear Spin Relaxation Fluids; Solid State N.M.R.; Multiple Resonance; Natural Macromolecules; Synthetic Macromolecules; Conformational Analysis; N.M.R.of Paramagnetic Molecules; N.M.R. of Liquid Crystals and Micellar Solutions; "A subject index at the end, many tables and a clear table of contents at the beginning of each chapter complete this volume, which ought to be present in all university chemistry libraries." European Spectroscopy News, reviewing Vol. 8 Hardcover 372pp 0 851 86 332 9 Price f 51 .OO ($1 09.00) Photochemistry Vol. 11 Senior Reporter: D. Bryce-Smith This volume is the eleventh in the series of reviews in the field of phDtochemistry and covers the literature published between July 1978 and June 1979.Brief Contents : Part I : Physical Aspects of Photochemistry; Part I I : Photochemistry of Inorganic and Organometallic Compounds; Part II I; Organic Aspects of Photochemistry; Part IV: Polymer Photochemistry; Part V: Photochemical Aspects of Solar Energy Conversion; Part VI: Chemical Aspects of Photobiology. "All Chemistry libraries which have at least a pretence of completeness should have a copy of this volume and a standing order for future volumes in the series. The quality of the printing, reproduction of figures, and the paper itself is excellent, which is refreshing in these days of shortcuts and photoreplication. The senior reporter and his colleagues are to be congratulated on a demanding assignment in which a very high standard of performance has been set and consistently achieved".- Journal of Medicinal Chemistry, reviewing Vol. 8 Hardcover 704pp 0 85186 095 8 Price f70.00 ($149.00) Annual Reports On the Progress of Chemistry Vol.77 Annual Reports provides for the general reader critical coverage of the significant advances in the major areas of chemistry on an annual basis. From 1980, Section A became devoted solely to inorganic chemistry and therefore exhibited enhanced coverage of this area compared with previous years. Section C, a new Annual Reports volume on physical chemistry has been introduced which consists of authoritative critical accounts of progress in major areas by acknowledged experts. Section B covers organic chemistry. Vol. 77 1980 (to be published in Autumn 1981 ) Section A Inorganic Chemistry f 36.50 ($90.50) Section B Organic Chemistry f40.50 ($1 00.00) Section C Physical Chemistry f36.50 ($90.50) Orders should be sent to: Distribution Centre, Chemistry The Royal Society of The Royal Society of Chemistry, Blackhorse Road, Letchworth, Herts.SG6 1 HNAnal. PYOC. 50 ANALYTICAL CHEMISTRY TRUST FUND atomic mass range 1-500, e-g., hydrogen, in micron sized areas of bulk samples (e.g., welds), within a time of 1OOpS. A 20ns energy pulse from a laser is focused on to a 5pm spot on the sample. About 10pm3 of sample is vapourised, the ions extracted, accelerated and directed into a time of flight mass analyser. The ions are detected by an electron multiplier and the output is displayed on a storage oscilloscope and held in a transient recorder.Cambridge Consultants Ltd. Circle 516. PID Controller The Lauda R25 is suitable for use with platinum resistance thermometers and is compatible with Lauda thermostats with relay units R2 and R3 to give accuracies as high as f0.005 "C. It may be used as a proportional or on/off con- troller. Roth Scientific Co. Ltd. Circle 517. Rotating Disc Electrode The E628 can be equipped with a variety of interchangeable electrode tips for analysis and work on the kinetics of electrode reactions. Electronic control and stabilising circuits are incorporated. Two speeds are available, one for measurement and one for regeneration by polishing. can be. chosen over a wide range. Any desired number of revolutions Roth Scientific Co. Ltd. Circle 518. Software for Microprocessor -controlled Titrator A new software package extends the capabilities of the Titroprocessor. Additional parameters (set by the operator) include the quality of the end-point, so that significant inflections can be selected, an end-point window, for investigation of part of a curve, the identification of equi- valence points and a volume read-out for end- point titration. There is also the ability to work with one of two different burettes and two electrode chains. Roth Scientific Co. Ltd. Circle 519. Flow Cell for Spectrophotometry A sipper system for the DU-8 ultraviolet - visible spectrophotometer has been introduced. It consists of a temperature-controlled 8O-pl flow cell, a peristaltic pump and plug-in soft- ware (Compuset). The system provides fast and accurate measurement of concentration or rate results by comparison with a stored standard curve. The temperature (from 10 to 50 "C) is controlled with a precision of 0.05 "C. Beckman-RIIC Ltd. Circle 520.

ISSN:0144-557X    

DOI:10.1039/AP9821900046

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

50 ANALYTICAL CHEMISTRY TRUST FUND Anal. PYOC. Analytical Chemistry Trust Fund SAC Fellowships The Trustees invite proposals for research projects likely to make major contributions to the advancement of analytical chemistry in the UK. Projects with original ideas for research are particularly requested. Proposals will be considered from applicants with a first-class research background in non-academic establish- ments, such as industrial organisations, govern- ment laboratories and Research Associations, as well as from those in academic institutions. Applications must be made by prospective Fellows only. The value of a Fellowship is related to the Lecturer scale for non-clinical academic staff. SAC Studentships The Trustees invite proposals from super- visors for research projects likely to make important contributions to the advancement of analytical chemistry in the UK and which are suitable for well qualified postgraduate students.Projects will be assessed particularly on the originality of the research proposed. Applications may be submitted by research supervisors, who must be members of the Analytical Division of the Royal Society 'of Chemistry of at least 2 years' standing. Proposals for projects to start in the Autumn term of 1982 will be considered early in that year when a tentative award may be made, subject to the Trustees being satisfied by the Summer of 1982 that a student acceptable to them is available. The value of a Studentship is between kl640 and i2770 per annum mini- mum, according to circumstances, plus fees up to UK rates. Application Regulations for the Fellowships and the Studentships can be obtained from the Secre- tary, Analytical Division, Royal Society of Chemistry, Burlington House, London, W1V OBN. The closing date for applications is January 29th, 1982.

ISSN:0144-557X    

DOI:10.1039/AP9821900050

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

January, 1982 OBITUARY 51 Publications Received Flow Injection Analysis. Jaromir RfiiiCka and Elo Harald Hansen. Chemical Analysis Series, Volume 62. Pp. xiv + 207. John Wiley. 1981. Price k20. ISBN 0 471 08192 2. This is the first comprehensive treatise on this new and rapidly expanding technique. It pro- vides in-depth answers to the most frequently asked questions on both theoretical and practical aspects of this new analytical technique. It covers the construction and components of apparatus, techniques and methods and experi- mental approaches including such practical flow injection analysis techniques as stopped-flow, merging zones, titrations, solvent extraction and diffusion, a thorough description of experi- mental techniques, including five exercises to aid the beginner, a review of methods covering the assay of more than 60 different species and the use of microcomputers in connection with flow injection analysis with several basic pro- grams attached as appendixes.The book also treats in detail the parameters governing dis- persion and describes how these parameters can be manipulated to optimise a given analytical procedure. Photometric and Fluorometric Methods of Analysis, Nonmetals. Foster Dee Snell. Pp. xiv + 818. John Wiley. 1981. Price,656. ISBN 0 471 81023 1. As with the previous volume 011 metals, each of the 40 chapters is a short monograph on the determination on one element or radical with citation of literature from more than 25 years in 3000 references. The book assumes that the user has a basic background knowledge of laboratory procedures.Annual Reports on NMR Spectroscopy. Volume 11A. Edited by G. A. Webb. Pp. x + 282. Aca- demic Press. 1981. Price ,635; $84. ISBN 0 12 505311 8; ISSN 0066 4103.52 PUBLICATIONS RECEIVED A n d PYOC. Owing to the expansion in the applications of nitrogen NMR, Volume 11 is divided into two parts. This part, Volume 11A, covers areas of molecular science that are .critically dependent upon NMR investigations. The reviews are : “NMR of Amino Acids, Peptides and Proteins (1977-1979),” by H. W. E. Rattle; “Carbon- Carbon Coupling Constants : Data,” by V. Wray and P. E. Hansen; “Calcium and Magnesium NMR in Chemistry and Biology,” by S. Forsen and B. Lindman; and “Carbon-13 NMR of Group VIII Metal Complexes,” by P. S. Pregosin.Volume 11B will cover nitrogen NMR. Trace Element Analytical Chemistry in Medicine and Biology. Proceedings of the First International Workshop, Neuherberg, Federal Republic of Germany, April 1980. Edited by Peter Bratter and Peter Schramel. Pp. xviii + 851. Walter de Gruyter. 1980. Price DM180. The workshop, of which this book is the pro- ceedings, was organised by Gesellschaft fur Strahlen- und Umweltforschung mbH and the Working Group on “Trace Elements in the Life- Sciences” of the Arbeitsgemeinschaft der Grobforschungseinrichtungen in cooperation with the International Atomic Energy Agency and the European Communities. The main general theme was essential trace elements, and aspects of research in medicine and biology were discussed. The plenary lectures, papers, posters and discussions are all included in the book.ISBN 3 11 008357 4. The Analysis of Explosives. By Jehuda Yinon and Shmuel Zitrin. Perga- mon Series in Analytical Chemistry, Volume 3. Pp. xii + 310. Pergamon Press. 1981. Price $60; fj25 (hardback): $22.50; fj9.35 (softback). ISBN 0 08 023846 7 (hardback) ; 0 08 023845 9 (softback). Presenting an overview of all the various meth- ods and techniques, the book describes the principles of different analytical methods, how these methods are used for the analysis of explosives and reviews the major analytical work that has been carried out in this field. Chemical methods, chromatographic techniques, polarography, thermal analysis, ultraviolet and visible spectroscopy, infrared spectroscopy, magnetic resonance methods and mass spectro- metry are covered.Also included are recent advances in the field, such as novel mass spectrometry methods and HPLC. Separate chapters have been devoted to survey the various methods for the detection of hidden explosives and to describe the special problems and the methodology of post-explosion residue analysis. A list of explosives and related compounds has been included. Catalogue. English Translations of German Standards 1981. Edited by DIN, Deutsches Institut fur Normung e.V. Pp. 244. Beuth Verlag GmbH. 1981. Available from the National Standards Bodies. Chemical Methods of Rock Analysis. Third Edition. P. G. Jeffery and D. Hutchison. Pergainon Series in Analytical Chemistry, Volume 4. Pp. ’ xvi + 379. Pergamon Press. 1981. Price L25; $60.In this Third Edition the authors have been more selective in the material presented, cutting out details of older methods that have failed to keep their place in the laboratory in favour of methods based on newer ideas and techniques. The chapter on statistical methods has been omitted, as most analysts are familiar with these techniques. There are chapters on Composition of Rock Material, Sample De- composition, Classical Scheme for the Analysis of Silicate Rocks and Rapid Analysis of Silicate Rocks, followed by 44 chapters covering all elements to be found’in rocks. ISBN 0 08 023806 8. Mathematics for Physical Chemistry. Robert G. Mortimer. Pp. x + 405. Collier- Macmillan . 198 1, Price k6.95 (sof tback) . ISBN 0 02 384000 5. The first nine chapters of this survey of mathe- matics needed for physical chemistry courses at the undergraduate level are constructed around a sequence of mathematical topics : Numbers, Mathematics and Science ; Mathematical Vari- ables and Operations ; Mathematical Functions and Differential Calculus ; Integral Calculus ; Mathematical Series ; Calculus with Several Independent Varia.bles ; Differential Equations and the Motions of Objects; Operators, Matrices and Group Theory; Algebraic Equations.Chapter 10 is a discussion of mathematical topics needed in the analysis of experimental data and chapter 11 is a brief introduction to computer programming in BASIC. Particle Size Measurement. Third Edition. Terence Allen. Powder Technology Series. Pp. xxii + 678. Chapman and Hall. 1981.Price fj24.50. Many techniques that have been developed for ISBN 0 412 15410 2.January, 1982 MEETINGS measuring particle size are covered in this Third Edition. The approach is informed and critical, which will help in the selection of equipment and assessment of the value of such measurements. The high degree of technological change within the field since the Second Edition demands that this edition be updated and enlarged, and a statement on the German approach to mathe- matical handling of size data has been presented. The chapter on centrifugal methods includes a full statement of the Coulter principle. The chapters on surface area and pore size deter- mination have also been expanded and a new chapter on On-line Particle Size Analysis has been added.The lists of equipment manu- facturers and suppliers in the appendix have been updated. 53 Chemical Technicians’ Ready Reference Handbook. Second Edition. Gershon J. Shugar, Ronald A. Shugar, Lawrence Bauman and Rose Shugar Bauman. Pp. xxiv + 867. McGraw-Hill. 1981. Price L27.95. ISBN 0 07 057176 7. This handbook is designed to provide each step to be followed when performing normal laboratory procedures. It details what equip- ment is needed, what each piece of equipment looks like (with photographs and drawings), how the pieces are used, the sequential steps to be taken in performing the determinations or acquiring the experimental data, the pre- cautions that must be observed and the raw data, observations and calculations that may be needed to utilise the experimental data. Amino Acid Analysis. Edited by J. M. Rattenbury. Pp. 380. Ellis Horwood. 1981. Price L27.50. ISBN 0 85312 194 X; 0 470 27141 8. This monograph developed from a symposium on the role of amino acid analysis in clinical chemistry at the University of Edinburgh. The work has been divided into four parts. Part 1 covers established and novel techniques, such as fluorime tric detection, high-performance liquid chromatography and collaborative trials ; Part 2 relates these techniques to physiological processes ; Part 3 is concerned with the investi- gation of systemic diseases ; and Part 4 surveys the role of amino acids in congenital disorders.

ISSN:0144-557X    

DOI:10.1039/AP982190051b

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

January, 1982 MEETINGS 53 Workshop Course: Elements of Analytical Instrument Electronics March 23-27, 1982, Guildford This “appreciation” course for novices, with practical work, will include aspects of chromato- graphic detectors and microprocessors. For further details contact Dr. E. Reid, Wolfson Bioanalytical Unit, Robens Institute, University of Surrey, Guildford, GU2 5XH.

ISSN:0144-557X    

DOI:10.1039/AP982190053b

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

54 ANALYTICAL DIVISION DIARY Anal. PYOC. Ana lyt ica I Division Diary JANUARY Wednesday, 20th, 4 p.m. : Belfast Northern Ireland Region, jointly with the Northern Ireland Branch of the Insti- tute of Water Pollution Control. Analytical Quality Control in Water An a I ys i s. Speaker: M. J. Gardner. Room LG25, Chemistry Department, Queen's University, Belfast. Contact: Mr. W. J. Swindall, Depart- ment of Chemistry, David Keir Building, Queen's University, Belfast, BT9 5AG. (Tel. 0232-661 11 1, Ex. 4428). Thursday, 21st, 12 noon: London Annual General Meeting. Joint Pharmaceutical Analysis Group: Aspects of the European Pharma- copoeia. Speaker: H. S. Grainger. Pharmaceutical Society of Great Britain, 1 Lambeth High Street, London, SEI. Contact; Miss I. Ladden, British Pharma- copoeia Commission, Room 1706, Market Towers, 1 Nine Elms Lane, London, SW8 5NQ.(Tel. 01-720- 9844, Ex. 31 05). Thursday, 21 st, 6.30 p.m. : Glasgow Atomic Spectroscopy Group: Annual General Meeting, followed by a joint meeting with the Scottish Region. Has Plasma Emission Made Atomic Absorption Redundant? Atomic absorption spectroscopy has become a routine technique in most laboratories for rapid and precise determination of a wide range of elements in solution. The recent emergence of plasma emission spectroscopy into the routine laboratory has forced many users and potential users of both techniques to question the advantages of each. With increasing sophistication, not to mention cost, involved in instrumental purchase, failure to assess all parameters can be disastrous in today's economic climate.This in- formal discussion, led by a panel of eminent speakers, offers the opportunity for everyone to air their opinions, voice their grievances, and contribute to a lively and interesting discussion. Discussion to be led by E. J. Newman and A. M. Music Room, Staff Centre, University of Strat hc I yde, G lasg ow. Registration is requested but there is no registration fee. Contact: Mr. A. F. Fell, Department of Pharmacy, Heriot-Watt University, 79 Grassmarket, Edinburgh, EHI 2HJ. (Tel. 031 -225-8432, Ex. 225). Ure. FEBRUARY Friday, 5th, 7 p.m. : Moreton North West Region. Quality Control as Applied to the Pharmaceutical Industry. Speaker: S. Williams. E. R. Squibb B Sons Laboratories, Reeds Lane, Moreton, Wirral. Contact: Mr.T. Hanley, 5 Old Hall Court, Ashton, Chester. (Tel. 0829- 51 609). Wednesday, loth, 2 p.m.: Birmingham Analytical Division. Pattern Recognition : Algorithms and Applications. The increasing use of automated instruments with their large data outputs, and the many complex problems that now have to be con- sidered by chemists, have lead to great interest in the application of automated methods for the reduction and inter- pretation of large data sets. Various approaches have been investigated to achieve reliable computer-assisted inter- pretations of chemical data, involving statistics, information theory, etc. Pattern recognition met hods seem particular I y useful for the classification of objects (e.g., substances, materials) into discrete classes on the basis of measured features.A set of characteristic features (e.g., a spectrum) of an object is considered as an abstract pattern containing informa- tion about a property of the object (e.g., molecular structure or biological activity) that is not itself directly measureable. Pattern recognition methods attempt to establish relationships between the pattern and the property, without involve- ment of chemical prejudices. The [continued on p. 55January, 1982 ANALYTICAL DIVISION DIARY 55 Analytical Division Diary, continued February, continued meeting is intended to introduce the general concepts of pattern recognition and to indicate, without excessive stress on mathematical techniques, how these concepts can be applied in situations of general interest to analytical chemists."Basic Philosophy of Pattern Recognition in Chemistry," by Olav H. J. Christie. "Applications of Pattern Recognition in the Life Sciences," by Professor D. L. Massart. "Pattern Recognition in Practice: Rapid Identi- fication of Micro-organisms by Pyrolysis Gas Chromatography," by H. J. H. MacFie. "Data Enhancement and Pattern Recognition Applied to Thermal Wave Imaging," by Professor M. S. Beck. Haworth Lecture Theatre 101, The Uni- versity, B irm i ng ham. Registration is necessary. Cost f4 to members of RSC, f 6 to non-members and no charge to students. Contact: Miss P. E. Hutchinson, Analytical Division, Royal Society of Chemistry, Burlington House, London, WIV OBN. (Tel. 01 -734-9971 ). Wednesday, loth, 10 a.m.: London Hectroanalytical Group, jointly with the Electrochemistry Group of the Faraday Division.Electrochemical Measurement of Dissolved Oxygen. "Fundamental Aspects and Design of Electro- chemical Oxygen Sensors," by M. L. Hitch- man. "Dissolved Oxygen Measurement in Natural and Waste Waters," by A. E. Hey. "Dissolved Oxygen Measurement in Power Station Waters," by J. Gyllenspetz. "Measurement of Dissolved Oxygen in Liquid Metals," by B. C. H. Steele. "ln-situ Oxygen Measurement in Deep Seas," by J. M. Saunders. "Design and Operation of an Oxygen Sensor for Incorporation in a Geophysical Probe," by J. V. Dobson. "Gas Sensors for Clinical Medicine," by Pro- fessor J. Albery. "Dissolved Gas Measurements in Medicine," by D. Parker. Imperial College, London, SW7. Registration is necessary.Cost f 20 (including lunch) ; f 1 2 for students accompanied by supervisor. Contact: Mr. A. E. Bottom, Kent Industrial Measurements Ltd., EIL Analytical I nstru ments, H an wort h Lane, C h ertsey, Surrey, KT16 9LF. (Tel. 093-28- 62671 ). Friday, 12th, 5 p.m.: Exeter sula Section of RSC. Western Region, jointly with the Penin- Analytical Techniques in Forensic Science. Speaker: D. G. Sanger. Newman Building, The University, Stocker Road, Exeter. Contact: Mr. F. W. Sweeting, Wessex Water Authority, P.O. Box 95, The Ambury, Bath, BAI 2YP. (Tel. 0225- 31 3500). Tuesday and Wednesday, 16th and 17th, Edinburgh Scottish Region and Chromatography and Advances in Chromatography: In- dustrial and Petrochemical Applica- tions. Tuesday, 16th: 2 p.m.- Plenary Lecture: "Sample Pre-concentration and Detection Techniques in Environmental Analysis by HPLC," by Professor R.W. Frei. "Monitoring of the Factory Environment by Diffusive Samplers and GLC," by R. H. Brown. Wednesday, 17th: 9.1 5 a.m.- Plenary Lecture: "Strategy for Method Develop- ment in Petrochemical Analysis by GLC," by H. Poppe. "Algorithms for Oil Fingerprinting by GC - MS," by N. J. Haskins. "Role of CGC - MS in Hydrocarbon Base Line Studies," by S. J. W. Grigson. "Standardisation of H PLC Columns and Reten- tion Index Systems: Relevance to Official Methods," by R. M. Smith. Plenary Lecture: "The Application of Chromato- graphy to Industrial Pollution Problems," by D. Simpson. "Chromatographic Combinations for Drug Evalu- ation in Industry," by B. Scales. "Automated Chromatography Systems for Industrial Control," by K.E. Lieper. The meeting will include two sessions of parallel Discussion Seminars. Each Seminar will be presented by an acknowledged expert in the field, and will deal with a contemporary topic in depth. Participants are encouraged to submit problems beforehand, so that each problem can be discussed in workshop style in the relevant Seminar. The Seminars are as follows: "Automated Sample Handling," by R. H. Brown. [continued on p. 56 Electrophoresis Group.56 ANALYTICAL DIVISION DIARY Anal. Proc. Analytical Division Diary, continued Micro-organisms Using Pyrolysis - MS," by f ebruary, continued "Post-column Reaction Detectors," by Pro- fessor R. w. Frei. "Environmental Base Line Monitoring Tech- niques," by S.J. W. Grigson. "Capillary GC," by N. G. Haskins. "Automation in Process Control," by K. E. "Pre-scale HPLC," by H. Poppe. "Chromatographic Techniques in Drug Develop- ment," by B. Scales. "A Consultant's Approach to Chromatography," by D. Simpson. "On-column Derivative Techniques (With Special Reference to Metal Ions)," by R. M. Smith. Department of Chemistry, The University, King's Buildings Campus, West Mains Road, Edinburgh. Cost f25 to RSC members; f35 to non-members; f 12 to students. Contact; Dr. D. Simpson, Analysis For Industry, Factories 2/3, Bosworth House, High Street, Thorpe-le-Soken, Essex, C016 OEA. (Tel. 0255- 861 71 4). Lie per. Registration is necessary. Wednesday, 17th, 6.30 p.m. : London South East Region and Microchemical Plasma Emission Spectrometry.The speaker will show why plasma emission spectrometry is gaining in popularity for the determination of metals. Discussion to be introduced by E. J. Newman. Savoy Tavern, Savoy Street, London, w.c.2. Contact: Dr. A. H. Andrews, Beecham Pharmaceuticals, Clarendon Road, Worthing, Sussex, BN14 8QH. (Tel. Methods Group. 0903-39900, EX. 428). Thursday, 18th, 10.30 a.m. : Bristol Western Region and Automatic Methods Automated Combination Techniques with Mass Spectrometry. "The Application of Automatic Techniques for Laboratory and Process Control," by T. Long. "An Automated GC - MS Assay for Salbutamol in Plasma," by R. Tanner. "On-line Process Control of Reactors in Chemical Plants Using MS," by J. H. Scrivens. "Prospects for Automated Identification of Group.C. Gutteridge. by D. Games. Wills in the morning. 30. "Automated Methods in Combined LC - MS," There will be a Laboratory Tour of W. 0. & H. 0. Number restricted to W. D. & H. 0. Wills, Hartcliffe, Bristol. Registration is necessary. Cost f 7 to members of RSC; f 12 to non-members; f 4 to students; f3.50 extra for lunch. Contact: Dr. C. J. Jackson, Health and Safety Executive, Occupational Hygiene Laboratory, 403 Edgware Road, London, NW2 6LN. (Tel. 01- 450-891 1, EX. 227). 0 Tuesday, 23rd, 4.15 p.m.: Lough- borough Midlands Region, jointly with Lough- borough University Students Chemical Society. Fluorescence in Forensic Analysis. Fluorescence effects have been applied in forensic science for many years, e.g., for the visualisation of contact traces from scenes of crime, in the characterisa- tion of materials and in qualitative "spot" tests.Examples of such applications will be presented. However, very much more information and evidential signifi- cance can now be derived by the use of fluorescence spectrometry both in the earlier applications and in others. Examples of the techniques used and the results obtained will be described. Speaker: J. 8. F. Lloyd. Lecture Theatre JOOl, Edward Herbert Building, University of Technology, Loughborough. Contact: Mr. H. E. Brookes, 35 Dunster Road, West Bridgford, Nottingham, NG2 6JE. (Tel. 0602-231769). Wednesday, 24th. 10.15 a.m.: Billing- ham North East Region and Automatic Some Aspects of the Use of Micro- processors in Automatic Analysis. Methods Group. "Interfacing Microprocessors to Analytical Systems," by B.J. Millard. [continued on p. 57January, 1982 ANALYTICAL DIVISION DIARY 57 Analytical Division Diary, continued February, continued "A Versatile Interface for Automatic Analysis," by R. Broadridge. "The Use of Microprocessors in Continuous Flow Analysis," by G. W. Moody. "An Automated System for Monitoring Non- volatile Alkalinity in Boiler Water," by H. M. Webber. "A Microprocessor Controlled Automatic Analyser for the Determination of Quinizarin in Hydrocarbon Oils," by A. Honeybone. "Automatic Analysis System Linked to Micro- reactors," by G. B. Fish. "Minicomputers in Atomic Spectroscopy," by P. Goddard. The meeting will include a demonstration of microprocessor controlled analytical equip- ment. Registration is necessary; cost f 5 in- cluding buffet lunch. Contact; Mr. C. L. Denton, 20 Bedford Road, Nunthorpe, Middlesbrough, Cleveland, TS7 OBZ. (Tel. 0642- 31 5721 ). Thursday, 25th, 4.1 5 p.m. : Aberdeen Scottish Region, jointly with the Aberdeen and North of Scotland Section of the RSC and Aberdeen University Student Chemical Society. Quantitative Analysis of Environ- mental Samples by Gas Chromato- graphy - Mass Spectrometry. The term "quantitative" often needs to be re-defined when gas chromatography - mass spectrometry with a data system is applied to the separation and measure- ment of organic contaminants and pollutants in complex biological extracts. Examples of the use of fused silica GC - MS - DS will be given to support this argument. Speaker: D. E. Wells. Chemistry Department, University of Aberdeen, Old Aberdeen. Contact: Mr. A. F. Fell, Heriot-Watt Uni- versity, Department of Pharmacy, 79 Grassmarket, Edinburgh, EHI 2HJ. (Tel. 031 -225-8432, Ex. 225).

ISSN:0144-557X    

DOI:10.1039/AP9821900054

出版商:RSC    

年代:1982

数据来源: RSC