ISSN:0003-2654    

DOI:10.1039/AN97500FX005

出版商:RSC    

年代:1975

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97500BX007

出版商:RSC    

年代:1975

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97500FP017

出版商:RSC    

年代:1975

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97500BP023

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

FEBRUARY, 1975 The Analyst Vol. 100, No. 11 87 The Examination of Meat Products with Special Reference to the Assessment of the Meat Content A Review* D. Pearson National College of Food Technology (University of Readiwg), Weybridge, Surrey Summary of Contents Introduction The basic composition of meat Assessment of the total meat content Assessment of the lean meat content Methods of proximate analysis Examination for the type of meat present Quality of the meat used Examination of products for various additions Salt Added milk Soya additives and novel proteins Added phosphates Conclusions Introduction When Stubbs and Morel devised their classical formula for calculating the proportion of meat present in fresh sausages the only sources of protein considered were derived from the meat and the bread filler.Currently, statutory regulations in the United Kingdom prescribe minimum requirements for the total and lean meat content of several products. The filler in sausages is now usually wheat rusk and farinaceous materials used in other meat products include potato starch, potato flour and cornflour from maize. Additional ingredients that are now available include non-meat sources of nitrogen, such as milk and soya. Vegetable materials such as soya can be used as a concentrate or isolate, or in a texturised form SO that they resemble meat. Many such additions can only be detected by instrumental tech- niques, for example electrophoresis, and allowing for their presence necessitates modification of the Stubbs and More formula for assessing the meat content.The meat content of many products is calculated as the amount of raw meat used and as a result of the wide natural variation in composition the Society for Analytical Chemistry (SAC) has presented compromise mean factors for the nitrogen in many forms of flesh when calculated on the fat-free basis. Some regulations, however, prescribe minima for heat processed products, but the absence of recognised protein factors for cooked meats causes difficulties in industrial control and in enforcement. There is a lack of agreement on the composition of lean meat, which is somewhat loosely defined in the official enactments. Apart from the actual amount of meat present, the quality of the meat used is obviously of considerable importance. Advances have been made in identifying the meat used in products of unknown origin.The higher the processing temperature that is used the greater is the alteration in the proteins present and the identification of the type of meat by electro- phoretic and chromatographic means becomes increasingly difficult. Added phosphates are * Based on a paper given at the SAC Centenary Celebrations, London, July 16th-I9th, 1974. Reprints of this paper will be available shortly. For details see summaries in advertisement pages. 7374 Analyst, VoZ. 100 also of interest in view of their use for textural reasons. Overall, problems arise because of the heterogeneous and complex nature of meat itself, the effect on the results of processing and the presence of the increasing number of a.dditives used for various reasons.The Basic Composition of Meat PEARSON : EXAMINATION OF MEAT PRODUCTS WITH SPECIAL Because meat is comparatively rich in protein, the nitrogen content is commonly used in assessing the proportion of flesh present in products. As the fat to protein ratio of meat can vary from 0.15: 1 for very lean meat to over 100: 1 for outside fat, calculations are made on the fat-free (FF) basis. Various authorities1-" have recommended mean factors for the nitrogen in the fat-free material (N/FF) for pork and beef (Table I). Some later figures have been published by the Meat Products Sub-committee of the SAC'S Analytical Methods Committee (AMC) by considering literature values and results specifically obtained from the analysis of selected cuts of meat as used in meat products.TABLE I HISTORY OF RECOMMENDED AVERAGE NITROGEN (PER CENT.) IN THE FAT-FREE FACTORS FOR PORK AND BEEF Stubbs Jackson and Morel and Jones:! SPASp4 SAC6 SAC6 Year . . .. 1919 1932 1940, 1952 1961 1963 Pork . . .. 4.0 3-55 3.6 3.45 - Beef .. .. 3.75 3.42 3.4 - 3.55 Considered over the years the factors have in general become steadily lower, mainly due to the fact that Stubbs and More probably examined relatively small amounts of, and there- fore partially dried up, meat in comparison with that used on some modern manufacturing plants, where pork, for instance, is used soon after the slaughter of the pig. The Committee have also recommended N/FF factors for other meat and poultry, vix., ~hicken,~ liver,8 veal,9 turkey,lO kidney,ll tongue12 and bl00d.l~ All of these mean factors are derived from a statistical study of several samples, which show a wide biological variation.In practice, it is advisable to allow statistically for a 95 per cent. probable variation from the agreed mean factor. The AMC's Sub-committee has also examined the variation in the nitrogen content of fillers and recommended compromise factors of both wheat rusk14J5 and pearl barley.16 No allowance is necessary for the negligible amount of nitrogen present in some fillers, such as cornflour and potato starch. Assessment of the Total Meat Content It was a t the end of World War I when Stubbs and Morel published their classical formula for calculating the total meat content of fresh sausages. This involved subtracting the nitrogen due to a bread-type filler from the total nitrogen, thus obtaining the nitrogen due to meat, which was then expressed as a percentage of the mean N/FF for the particular meat to obtain the defatted meat content.Then, by adding the result for the percentage of fat the total meat in the product was obtained. Since 1919 the same basic formula has been employed, but various modifications have been made, particularly because (a) sub- tractions may have to be made for other non-meat proteins, (b) wheat rusk rather than bread is used as the filler in sausages and (c) more comprehensive data on meats have been studied statistically and have modified opinion as to what constitutes reasonable mean factors for the N/FF,3-13 Interest in the subject has increased in recent years due to the re-introduction of statutory standards for sausages and some similar products.17 The Stubbs and More formula can also be used in assessing the meat content (calculated as raw meat) of other products, e.g., sausage rolls,l8 certain meat pies18 and canned r r i e a t ~ , ~ ~ and meat pastes.30 With meat pies, both the meat filling and the pastry have to be analysed separately.Then, any fat in excess of 50 per cent. of the carbohydrate in the pastry (35 per cent. in Scottish pie) is assumed to have been due to rendering from the meat filling and is added to the calculated meat content.21p22 Apart from difficulties in interpreting the results in relation to the rather complex statutory standards, there has been controversy as to the way a pie should be cut during official sampling under the Food and Drugs Act.For certain products the regulations prescribe that the total meat content shall be calculated when raw or after curing or anyFebruary, 1975 REFERENCE TO THE ASSESSMENT OF MEAT CONTENT 75 other similar processing. Unfortunately, it is difficult to decide on a “true” mean N/FF factor for cured meat as the alteration in the composition varies according to the manu- facturing process. Thus, a corned beef containing 58.5 per cent. of water, 3.60 per cent. of nitrogen, 15.0 per cent. of fat and a negligible amount of carbohydrate gives a raw meat equivalent of 116 per cent. if the raw factor of 3.55 per cent. N/FF for beef is used. The heat processing increases the protein content, so when using N/FF factors such as 4.7 or 4.8, the calculated total meat then becomes more realistic at 92 or 90 per cent., respectively.On the Continent it is more usual to prescribe a maximum Feder number, which is essentially the ratio of the water content to the protein obtained by difference. This is mainly applied to products that contain no carbohydrate (cj., Krol and Meester23). Assessment of the Lean Meat Content Although fat is valuable in products as it contributes to quality attributes such as juiciness and texture, it is, in general, less valuable than the protein-rich lean meat from both the commercial and nutritional standpoints. Some regulation^^^^^^^^^ therefore prescribe, according to the product, a minimum lean meat content of from 50 to 60 per cent.of the statutory minimum for the total meat. For such a requirement “lean meat content” means “lean meat free of visible fat,” which is best considered as lean muscle with the outside fat cut completely away from it. In practice this is a rather theoretical material as manufac- turers use different cuts, each containing varying proportions of lean and fatty tissue. In the Stubbs and Morel calculation, the total meat content is assessed from the summation of the defatted meat and the extracted fat. This defatted meat is calculated from the total meat nitrogen, a small portion of which is derived from the outside fat. The fatty tissue, in turn, consists only of about nine tenths “pure” chemical fat. Calculation of the true lean meat must also take into account the presence of intramuscular fat, which is also included in the fat extracted in the analysis.The a ~ t h o r ~ ~ - ~ ~ has presented various formulae for the calculation, of which the following is the most general form: Lean meat (LM), per cent. = where -ATT is the total nitrogen in the product, per cent.; F, the fat extracted from the product, per cent. ; N , the N/FF for the particular meat, per cent. ; FF the fat, per cent., in the outside fat; F , the fat in lean meat, per cent.; KF the nitrogen, per cent., in cereal filler, calculated with reference to the dry carbohydrate; and C the total carbohydrate in the product, per cent. The main difficulty in using the formula is in deciding on the best figures to employ for FL and FF. The amount of fat extractable from lean meat (FL) may vary from 1-18 per cent.27 However, the author has not found lean beef to contain more than 11 per cent.of fat and the maximum figures reported by the Association of Public Analystsz8 for FL were: beef, 11.6; pork, 6.5; and lamb, 10.8. The author24 tentatively suggested that the maximum allowable figure for FL should be 10 per cent., but Coomara~wamy~~ uses Dalley’s mean figure of 4.2. The outside fat of meat contains about 90-95 per cent. of fat, and the author24 suggests that FF should be limited to 90 per cent., whereas Coomara~wamy~~ uses a figure of 95 per cent. The position should be cleared up by officially prescribing recognised values for FL and FF. If the Stubbs and Morel method is used, the defatted meat is very close to the lean meat figure obtained from the above equations.Therefore, although the earlier method combines together ( a ) , the protein in the fat with that present in the defatted meat, and (b), the intra- muscular and outside fat, there appears to be a mathematical compensating effect in practice. The lean meat calculation does not affect the figure obtained for the total meat. The lean meat content can also be assessed from the potassium content.30 Methods of Proximate Analysis The Society for Analytical Chemistry31 and the British Standards Institution32-35 have published reference methods for determining the basic components in meat and meat products.76 Analyst, VoZ. 100 These procedures have, in turn, been influenced by recommendations of the International Standards Organisat ion. Because of the heterogeneous nature of meat it is most important that all analytical determinations on meat products should be applied to thoroughly comminuted samples that have been prepared by successive mincing and mixing.In the reference methods, moisture is determined by drying at 103 0C.34 The ash is weighed after ignition at 550-600 "C in the presence of magnesium acetate as an aid to ashirig; this gives a spongy ash and shortens the combustion time.32 The nitrogen content is deteimined by the macro-Kjeldahl method using a copper catalyst and involves distillation into boric acid.33 Total fat is extracted with diethyl ether, n-hexane or light petroleum after digestion of the sample with hydrochloric acid,35 while the extractable fat is obtained by extracting the dried material with n-hexane or light petrole~m.~~ Many United Kingdom meat products contain cereal, and the starch determination involves digestion of the meat in alcoholic potassium hydroxide, dissolving the filtered starch in acid and determining the glucose formed by titration.31 Reference methods are still frequently used in both public health and industrial control.There is a need, however, in the control of manufacturing processes for more rapid procedures. Methods for determining water involving the use of higher temperatures (140-200 "C) and shorter drying times have therefore been developed by several workers.36-39 As an increase in the drying time causes relatively little change in the mass loss, it is apparent that, unlike foods such as cereals, meat contains little bound moisture.Pettinati et aL40a lso found good agreement between methods using a gravity oven, the Ohaus Infrared Moisture Balance and azeotropic distillation from cumene, m-xylene and octan-2-01. Rapid methods for the determination of total fat have been extensively reviewed by Smith,41 while Mahmood-ul-Hassan and P e a r ~ o n ~ ~ concentrated their attention on techniques involving volumetric, refractive index and density methods. For the volumetric method, cleaner fat separations are obtained in van Gulik tubes using an acetic acid - perchloric acid mixture43 instead of sulphuric acid in Gerber b u t y r o m e t e r ~ . ~ ~ ~ ~ ~ By selecting suitable solvents the fat content can be rapidly assessed from changes in the density of the solution.A com- mercial instrument based on this principle has been developed called the Foss - Let Fat Analyzer, which uses a magnetic float cell for the measurement of the density of a tetra- chloroethylene extract of the fat.45 Meat is not an ideal medium for the application of nuclear magnetic resonance techniques. Samples for fat determination have to be dried before use in %on-destructive" low-resolution instruments. With the more expensive high-resolution instruments fat and water combined can be determined at one setting and water at another, in order to obtain the value for fat by difference.46 The so-called Ground Meat Analyzer is a non-destructive method for fat in- volving absorbance in both the visible and infrared regions of the spectrum47 and Ben-Gera and Norris48 have used near infrared absorption properties of meat emulsions for the simul- taneous determination of water and fat.More conventionally, the SI-MO-FAT employs the distillation extraction of moisture and the simultaneous extraction of fat in di-n-butyl ether.49 For products containing no carbohydrate the LJltra-X balance is convenient in industrial control.50 With this last instrument the material is weighed into a dish and then weighed successively after drying, after extracting the fat with carbon tetrachloride, and finally after ignition. Having obtained the water, fat and ash from these respective operations the protein is calculated by difference. The author39 has suggested that the ash can be obtained without igniting the sample by adding an empirical figure to the salt content obtained from a rapid direct titration.The nitrogen content can be rapidly assessed by colorimetric methods based on the proportionate drop in absorbance caused by the ability of polar groups of proteins to bind with dyes of opposite ~ h a r g e . ~ ~ ~ ~ ~ Such procedures have been successfully applied to milk and cereals, but problems arise with meats due to heterogeneity and difficulties in producing intimate contact of the macerated meat with the dye. Precision tends to be poor51 and with sausages the correlation with the reference Kjeldahl method is not satisfactory unless dye- binding is applied to a product made to a standard recipe.63 The ammonia in Kjeldahl digests can be determined without distillations4 and the colorimetric measurement of the indophenol blue complex formed by reaction with phenol and hypochlorite has been applied to the Technicon AutoAnaly~er.~5 With this last system it is reported that about 20 samples can be satisfactorily examined per PEARSON : EXAMINATION OF MEAT PRODUCTS WITH SPECIALFebruary, 1975 REFERENCE TO THE ASSESSMENT OF MEAT CONTENT 77 In general, rapid methods for water and fat give results that agree reasonably well with those from reference procedures.Taken in conjunction with the ash, the protein content can readily be calculated by difference in products that contain no added cereal. Many United Kingdom products, however, contain rusk and a direct determination of the all- important protein is usually required, although as yet there is no rapid method available to replace the reliable, though laborious, macro-Kjeldahl reference procedure.Examination for the Type of Meat Present The examination of products for the type of meat present has been one of the major problems in food analysis for many decades. Formerly, the main concern was the use of horseflesh, the presence of which was suggested by either a high glycogen content or from the results of Uhlenluth's precipitin test (a biological one using rabbits).57 Neither test is suitable for cooked products, but it was sometimes possible to distinguish horse fat from other fats because of its relatively high iodine value. This is by no means specific, however, as pigs which have been given a diet containing a highly unsaturated oil can produce lard with a much higher iodine value than usual.Alternatively, horse fat could be detected because it contains a relatively high proportion of linolenic acid (up to 2 per cent.) when determined as the hexabr~mide~~ or spectrophotometrically.5g Most of the more recent work in the meat identification field has involved typing of the characteristic proteins of the juice or an extract separated by electrophoresis, the protein bands usually being stained with amido Most of the published electrophoretic methods are, however, inapplicable to heated products containing denatured protein. Never- theless, by using gel electrophoresis, Mattey et aZ.65 were able to identify quantitatively meat species after heating to 120 "C for 3-6 min. Later, it was found possible to identify soya proteins in meat products that had been heated to 127 "C for 24 min by applying a method based on laser densitometry of polyacrylamide electrophoretograms of the proteins extracted in 10 M urea.66 Hibbert and L a ~ r i e ~ ~ also investigated the use of determinations of the amino-acid 3-methylhistidine, which is present in meat but which is less likely to occur in any protein additives used.Ninhydrin is used to develop the colour of the separated amino- acid and the procedure is still applicable even when the proteins are completely hydrolysed. Most methods associated with the fat are applicable to cooked products, including the measurement of the differing absorptivities at 268 nm after alkaline isomerisation.68 Fats can be more specifically identified by gas chromatography, which can be applied to fractions of the unsaponifiable matter after separation by column chromat~graphy.~~ Castledine and Davies57 employ transesterification of the fat in methanol, separate the relatively volatile methyl esters and apply n-propyl palmitate as an internal standard.Horse fat can be identified from the pronounced linoleate and linolenate peaks. Specific methods for lard have also been described.'O Overall, by the use of more sophisticated techniques than hitherto, methods for the detection of the meat used is an area in which our knowledge is continually advancing and vast improvements have been brought about in the last decade. Quality of the Meat Used Apart from restrictions on certain 0ffals~~3~~ statutory regulations do not lay down requirements as to the quality of the cuts used. One of the most important criteria of meat quality is toughness, which is largely influenced by the collagen, although its effects can be reduced by thorough cooking.The amount of hydroxyproline present is usually used to assess the collagen content (Table 11). TABLE I1 HYDROXYPROLINE CONTENT OF SOME PROTEIN SOURCES2' Source Hydroxyproline, per cent. m/m, on dry, fat-free basis Purified c ~ l l a g e n ~ ~ - ~ ~ . . .. . . * . 13.4-14.5 Purified tend0n~8.7' . . .. .. .. 11.2-13'2 Tendon associated with muscle76 .. .. 12.3-13.3 11.0-12.0 Skeletal muscle p r ~ t e i n ~ ~ - ~ l . . . . .. 0.002-0.07 Plant materiala2 . . . . .. .. - . Negligible Skin cooked to an internal temperature of 61 oC7878 Analyst, VoZ.100 Hydroxyproline can be determined by the method of Langsl as modified by Stegemann and Stalder.83 This involves acid hydrolysis, oxidation of the hydroxyproline with chlor- amine T and the colorimetric determination a.t 560 nm of the red colour produced with 4-dimethylaminobenzaldehyde. The hydroxyproline is released by cleaving the peptide bond with the acid. A practical disadvantage of the method is that satisfactory results can be obtained only with a total hydrolysis time of the order of 16 h. The method has been em- ployed by Coomara~wamy~~ for determining not only the connective tissue in products but also the amount of cooked rind, which has been used in meat products and is relatively rich in nitrogen and hydroxyproline (Table 11).The author's e q u a t i o n ~ ~ * , ~ ~ were also extended so that the assessed lean and total meat figures are corrected by allowing for excessive con- nective tissue in sausages, luncheon meat, beefburgers and hamburgers and corned beef. Coomaraswamy based his calculations on the assumed presence of 10 per cent. of con- nective tissue and 90 per cent. of muscle meat. Unlike collagen, elastin contains little hydroxyproline. It is possible, however, to extend the Stegemann - Stalder procedure to include the elastin component by determining glutaminic acid following an enzymic con- version.84 Dahls5 examined the composition of different quality meats and showed that high-quality skeletal muscle had a low hydroxyproline content but was relatively rich in tryptophan and creatine when calculated on the basis of the total protein.It was suggested therefore that the creatine content could be employed to assess the quality of the meat cuts used in products. The method cannot be used. for cooked materials, such as corned beef, where the creatine tends to be largely lost in the cooking water. With the in- creasing price of meat, however, it is likely that cheaper cuts may be used in products than hitherto and examination for hydroxyproline, creatine, etc., may become of increasing importance. Examination of Products for Various Additions PEARSON : EXAMINATION OF MEAT PRODUCTS WITH SPECIAL Methods for assessing meat quality are seldom used in routine control. Various materials are added to meat products.These include substances that affect the flavour or texture and those which essential.1~ enhance the nitrogen content. Salt Provided that the ignition temperature does not exceed about 500 "C the chloride content can be approximately determined on the ash by application of the Mohr or Volhard methods.22P86 In the official methods the Volhard procedure is applied directly to the original sample, the organic matter being oxidised by heating to 80 "C with permanganate, the excess of which is reduced with A high salt-free ash figure suggests the presence of additions such as dried milk or phosphates.86 Added Milk For a short period in the early 1950s, owing to supply difficulties, manufacturers were able to replace up to 10 per cent. of the prescribed minimum meat content with dried milk, six parts of dried milk being taken as the equivalent of ten parts of meat.Apart from dried milk, sodium caseinate is also added to some meat products. The presence of milk increases the salt-free ash, calcium and lactose contents and these figures can be used for its detection and determination.86 The lactose, amounting to 50 per cent. in skimmed milk powder, can be determined after the other common sugars have been destroyed by fermentation with yeast.*' The direct determination of reducing sugars on an aqueous extract of stored sausage without fermentation is less accurate as the extract may also include maltose derived from rusk filler. BahP has developed an enzymic method for the assessment of skimmed milk powder in sausages, which is based on the determination of free lactose by its hydrolysis with /?-galactosidase to galactose and glucose, the amount of which is measured by the hexokinase method.The final measurement is by ultraviolet absorption. The milk protein can be determined from the phosphoprotein f r a c t i ~ n . ~ ~ ~ ~ ~ Also, casein is more heat stable than muscle proteins and the blands can be detected in electrophoretograms, even in products heated to 115 0C.g1s92 Samples are defatted with acetone and treated with 2-mercaptoethanol and urea to break down the protein coagulum. The casein and soya protein are then separated by starch-gel electrophoresis. Such a method is used in routine inspections in Holland. For the detection of sodium caseinate either thin-layer chromato- graphy93 or polyacrylamide gel electrophore~is~~~ can be used. The effect on organolepticFebrZWy, 1975 REFERENCE TO THE ASSESSMENT OF MEAT CONTENT 79 quality of additions of caseinate to sausage has been evaluated by Thomas et aLg5 Many methods for milk and caseinate have been described but it is difficult to decide on the most reliable one for control purposes.The position would be clearer if laboratories tried out the comparatively new method of Bahl,86 preferably in comparison with older methods in their routine examination of products. Soya Additives and Novel Proteins During World War I1 there occurred, for a time, a compulsory addition of 7.5 per cent. of soya flour to sausages to enhance the protein content. This regulation was later withdrawn because, although low-fat soya flour was used, problems arose due to the occurrence of bitterness and oxidative rancidity.Soya flour was detected in unheated products by the ammonia formed on incubation with urea solution owing to the presence of active urease in the soya products then used.g6 Some idea of the amount present can be obtained from the determination of the manganese, sucrose, salt-free ash and insoluble non-fermentable sugar contents. More recently, meat analogues produced by texturising soya and other vegetable protein sources have been produced by spinning and extrusion for use in various food^.^^^^^ The effect on the sensory qualities of such additions has been studied by Thomas et aLS5 Detection by the analyst, however, requires the application of more sophisticated techniques than those used for soya containing the active enzyme. Thus, workers at Nottingham UniversityGe and the Lyons laboratoriesg8 have employed electrophoretic techniques for the quantitative identification of soya additives, which can, however, be applied to both raw and heated products. The proteins are extracted into a solution of urea or 2-mercaptoethanol and separated by electrophoresis, usually on gels containing polyacrylamide.The amount of additive present is obtained from the densitograms of the stained protein bands. A histochemical method for the detection of soya novel proteins has been described, which involves controlled oxidation of the carbohydrate in plant protein products and detec- tion of the resulting aldehydes.99 It is based on the presence of specific carbohydrate con- stituents in soya, which tend to be retained in products such as novel proteins.The use of these additives has perhaps taxed the ingenuity of analysts more than any other field of meat examination. It is apparent, however, that a considerable degree of success has been achieved in a period spanning less than 5 years. Added Phosphates Ortho-, meta-, pyro- and polyphosphates of sodium and potassium, either singly or as mixtures, have been incorporated into various meat products, usually either for the retention of fluid or fat.lO0 These substances are included in the statutory permitted list of miscel- laneous additiveslOl and the amount used is restricted in most EEC countries and the USA (maximum from 0-25-043 per cent. according to the product).The addition of phosphate can be suspected if the salt-free ash is high, and it can then be detected more specifically by chemical tests or by paper chromatography1O2 or thin-layer chromatography.103 Alter- natively, phosphate can be determined in the ash by titrimetric and colorimetric techniques. Two difficulties arise in attempting to deduce the exact form and amount of phosphate originally added; firstly, all forms tend to be converted to the orthophosphate during most types of analysis, and secondly, allowance has to be made for the fairly considerable amount of phosphate that is naturally present in meat. Kamm and Coffinlo4 suggest that with ham, the result should be expressed as the phosphorus to protein ratio. Conclusions The examination of meat products of unknown origin presents several difficulties in interpretation. Apart from the wide variation in the nitrogen content of raw meat there is a lack of agreement on the amount of fat that can be allowed in lean meat. The calculation of the meat content has also to take into account the possible presence of materials containing non-meat protein, such as dried milk, caseinate and derived and novel proteins.From the analytical point of view, improvements in the future may include more efficient sampling procedures and increased use of multiple and automatic analysis. To have one multi-channel instrument that gives the analyst all the main results that he requires, with digital print-out, as is used for milk, seems a matter for the distant future.Currently, what80 Analyst, Vol. 100 would enable us to operate more efficiently, particularly in factory control, would be a rela- tively simple method for determining protein, which is all-important in the examination of meat products. The general need here is reflected in the number of firms who in the last decade have marketed multiple digestion equipment. The smaller laboratories might consider replacing the Kjeldahl distillation for protein by one of the methods involving direct titration of the ammonia formed in the digest, Perhaps the SAC, British Standards Institution and International Standards Organi- sation could give a lead here by recommending ari alternative, or, in the first instance, tentative method to encourage everyone in the field to experiment with the idea.Other improvements will surely come in the more sophisticated fields of detection of the meat used, where con- siderable advances have taken place in recent years. Also, with a world shortage of protein, increasing use will be made of other sources and new methods for their deter- mination will have to be developed. More immediately, it would be useful if the official organisations could suggest nitrogen factors for cured and heat-processed meats. Although much needs to be decided, it is obvious that we have advanced considerably since the days of Stubbs and More. PEARSON : EXAMINATION OF MEAT PRODUCTS WITH SPECIAL 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. 26. 27. 28. 29.30. 31. 32. 33. 34. 36. 36. 37. 38. 39. 40. 41. References Stubbs, G., and More, A., Analyst, 1919, 44, 125. Jackson, F. W., and Jones, O., Analyst, 1932, 57, 562. Analytical Methods Committee, Analyst, 1940, 65, 267. Analytical Methods Committee, Analyst, 1962, 77, 543. Analytical Methods Committee, Analyst, 1961, 86, 557. Analytical Methods Committee, Analyst, 1963, 88, 422. Analytical Methods Committee, Analyst, 1963, 88, 583. Analytical Methods Committee, Analyst, 1964, 89, 630. Analytical Methods Committee, Analyst, 1966, 90, 256. Analytical Methods Committee, Analyst, 1966. 90, 581. Analytical Methods Committee, Analyst, 1966, 91, 538. Analytical Methods Committee, Analyst, 1967, 92, 326. Analytical Methods Committee, Analyst, 1968, 93, 478. Analytical Methods Committee, Analyst, 1961, 86, 660.Analytical Methods Committee, Analyst, 1965, 90, 579. Analytical Methods Committee, Analyst, 1968, 93, 476. “The Sausage and Other Meat Products Regulations 1967,” S.I. 1967 No. 862 as amended by “The Meat Pie and Sausage Roll Regulations 1967,” S.I. 1967 No. 860, H.M. Stationery Office, “The Canned Meat Product Regulations 1967,” S.I. 1967 No. 861 as amended by S.I. 1968 “The Fish and Meat Spreadable Products Regidations 1968,” S.I. 1968 No. 430, H.M. Stationery Dedicoat, H., J . Ass. Publ. Analysts, 1963, 1, 85. Pearson, D., “Laboratory Techniques in Food Analysis,” Buttenvorths, London, 1973, p. 190. Krol, B., and Meester, J., Conserva, 1962/3, 11, 163. Pearson, D., J . Ass. Publ. Analysts, 1968, 6, 129. Pearson, D., J .Ass. Publ. Analysts, 1970, 8, 212. Pearson, D., and Gardiner, G. R. B., J . Ass. Publ. Analysts, 1971, 9, 102. “Food Standards Committee Report on Sausages,” H.M. Stationery Office, London, 1966, p. 8. Dalley, R. A., J . Ass. Publ. Analysts, 1969, 7 , 33. Coomaraswamy, M., J . Ass. Publ. Analysts, 1972, 10, 33. Kreuzer, W., Ring, Ch., and Schroeder, K., jFleischwirtschaft, 1968, 48, 802. Hanson, N. W., Editor, “Official, Standardised and Recommended Methods of Analysis,” Second British Standard 4401 : Part 1 : 1969. British Standard 4401 : Part 2 : 1969. British Standard 4401 : Part 3 : 1970. British Standard 4401 : Part 4 : 1970. Hall, J . L., “Proceedings of the 4th Annual Reciprocal Meat Conference,” National Livestock Perrin, C. H., and Ferguson, P. A., J . Ass.Ojy. Agric. Chem., 1967, 40, 1093. Krol, B., and Meester, J.. Fleischwirtschaft, 1963, 15, 489. Pearson, D., J . Ass. Publ. Analysts, 1971, 9, 89. Pettinati, J. D., Metzger, V. G., Van Horn, I]., and Cohen, E. H., J . Ass. Off‘. Analyt. Chem., 1973, 56, 1130. Smith, P. R., “A Review of Rapid Methods for the Estimation of Total Fat,” Scientific and Technical Surveys No, 66, British Food Manufacturing Industry Research Association, Leatherhead, 1969. S.I. 1968 No. 2047, H.M. Stationery Office, London. London. No. 2046, H.M. Stationery Office, London. Office, London. Edition, Society for Analytical Chemistry, London, 1974, p. 162. and Meat Board, Chicago, Illinois, U.S.A.February, 1975 REFERENCE TO THE ASSESSMENT OF MEAT CONTENT 81 42. 43. 44. 46. 46. 47.48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 76. 76. 77. 78. 79. 80. 81. 82. 83. 84. 86. 86. 87. 88. 89. 90. 91. 92. 93. 94. 96. 96. 97. 98. 99. 100. 101. 102. 103. 104. Mahmood-ul-Hassan, and Pearson, D., J . Sci. Fd Agric., 1966, 17, 421. Pearson, D., Fd Mf., 1972, 47 (4), 45. Talbot, A., Analyst, 1949, 74, 462. Usher, C. D., Green, C. J., and Smith, C. A,, J . Fd Technol., 1973, 8, 429. Ziemba, J. V., Fd Engng, 1962, 34 (2), 76. Dempster, D. P., Fd Canada, 1974 (Jan.), 22. Ben-Gera, I., and Norris, K. H., J . Fd Sci., 1968, 33, 64. Davis, C. E., Ockerman, H. W., and Cahill, V. R., Fd Technol., Champaign, 1966, 20, 1476. Gilardi, C., Industrie Aliment., 1969, 8 (6), 79. Torten, J., and Whitaker, J .R., J . Fd Sci., 1964, 29, 168. Moss, V. G., and Kielsmeier, E. W., Fd Technol., Champaign, 1967, 21, 33A. Pearson, D., and Parvaneh, V., J . Ass. Publ. Analysts, 1971, 9, 43. Pearson, D., in Porter, J. W. E., and Rolls, B. A., Editors, “Proteins in Human Nutrition,” Lento, H. F., and Daugherty, C. E., Fd Prod. Dev., 1971, 5, 86. Gantenbein, W. M., J . Ass. 08. Analyt. Chem., 1973, 56, 31. Castledine, S. A., and Davies, D. R. A., J . Ass. Publ. Analysts, 1968, 6, 39. Dalley, R. A., Analyst, 1950, 75, 336. Dugan, L. R., and Petheram, M., J . Ass. Ojf. Agric. Chem., 1952, 35, 767. Payne, W. R., J . Ass. Ojf. Agric. Chem., 1963, 46, 1003. Coduri, R. J., and Rand, A. G., J . Ass. 08. AnaZyt. Chem., 1972, 55, 461. Durand, M., and Schneider, R., Archs Inst. Pasteur Tunis, 1962, 39, 153.Ebermann, R., and Barna, J., 2. Lebensmittelunters. u. -Forsch., 1972, 148, 341. Hoyem, T., and Thorson, B., J . Agric. Fd Chem., 1970, 18, 737. Mattey, M., Parsons, A. L., and Lawrie. R. A., J . Fd Technol., 1970, 5, 41. Parsons, A. L., and Lawrie, R. A., J . Fd Technol., 1972, 7, 455. Hibbert, I., and Lawrie, R. A., J . Fd Technol., 1972, 7, 333. Cook, H. R., J . Ass. Off. Agric. Chem., 1962, 45, 10. Cook, H. R., and Sturgeon, J . D., J . Ass. Ofl. Analyt. Chem., 1966, 49, 877. Hubbard, A. W., and Pocklington, W. D., J . Sci. Fd Agric., 1968, 19, 571. “The Offals in Meat Products Order 1953,” S.I. 1953 No. 246, H.M. Stationery Office, London. “Food Standards Committee Report on Offals in Meat Products,” H.M. Stationery Office, London, Eastoe, J.E., Biochem. J., 1955, 61, 589. Bowes, J, H., J . SOC. Leath. Trades Chem., 1959, 43, 203. Tristam, G. R., and Smith, R. H., Adv. Protein Chem., 1963, 18, 227. Baker, L. C., Lampitt, L. H., and Brown, K. P., J . Sci. Fd Agric., 1954, 5, 226. Miyada, D. J., and Tappel, A. L., Analyt. Chem., 1956, 28, 909. Ritchey, S. J., and Cover, S., J . Agric. Fd Chem., 1962, 10, 40. Aronson, J. N., and Elvehjem, C. A., Fd Res., 1956, 21, 109. Lawrie, R. A., Pomeroy, R. W., and Williams, D. R., J . Agric. Sci., 1963, 60, 195. Lang, K., Hoppe-Seyler’s 2. Physiol. Chem., 1933, 148, 219. Clarke, E. M. W., and Ellinger, G. E., J . Sci. Fd Agric., 1967, 18, 536. Stegemann, H., and Stalder, K., Clinica Chim. Acta, 1967, 18, 267. Bernt, E., and Bergmeyer, H.-U., “Methoden der Enzymatischen Analyse,” Verlag Chemie, Dahl, O., J . Agric. Fd Chem., 1963, 11, 350. Pearson, D., “The Chemical Analysis of Foods,” Sixth Edition, Churchill, London, 1970, p. 383. McVey, W. C., and McMillin, H. R., J . Ass. Off. Agric. Chem., 1940, 23, 811. Bahl, R. K., Analyst, 1971, 96, 88. Kutscher, W., Nagel, W., and Pfaff, W., 2. Lebensmittelunters. u. -Fmsch., 1961, 115, 117. Thalacker, R., Dt. LebensmittRdsch., 1963, 59 (4). 111. Freimuth, U., and Krause, W., Nahrung, 1969, 13, 447. Olsman, W. J., Houtepen, W. H. C . , and Van Leeuwen, C., 2. Lebensmittelunters. u. -Forsch., 1969 Zwetkova, Z., Fleischwirtschaft, 1968, 48, 1486. Thorson, B., Skaare, K., and Hoyem, T., Nord. VetMed., 1969, 21, 436. Thomas, M. A., Baumgartner, P. A., Board, P. W., and Gipps, P. G., J . Fd Technol., 1973, 8, 175. Church, J. A., Fd Mf., 1970, 45 (6), 54. Guy, R. C. E., Jayaram, R., and Willcox, C. J., J . Sci. Fd Agric., 1973, 24, 1551. Coomaraswamy, M., and Flint, F. O., Analyst, 1973, 98, 642.,, Iles, N. A., “Phosphates in Meat and Meat Products-A Survey, Scientific and Technical Surveys “The Miscellaneous Additives in Food Regulations 1974,” S.I. 1974 No. 1121, H.M. Stationery Doro, B., and Remoli, S., Boll. Laboratori Chim. Provinciali, 1963, 14, 23. Rossel, T., and Kiesslich, H., 2. Analyt. Chem., 1967, 229, 96. Kamm, L., and Coffin, E., J . Ass. 08. Analyt. Chem., 1968, 51, 968. Academic Press, London, 1973, pp. 397-405. 1972. Weinheim, 1962, p. 384. 141, 253. Wood, J . C., Fd Mf., 1971, 46 (3-4), 37. No. 81, Food Research Association, Leatherhead, 1973. Office, London. Received July 1st. 1974 Accepted July 29th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9750000073

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

82 A.nalyst, February, 1975, Vol. 100, $9. 82-95 Characterisation of Rubbers, Rubber Contact Traces and Tyre Prints by Fluorescence Spectroscopy J. B. F. Lloyd West Midland Forensic Science Laboratory, Gooch Street North, Birmingham 6 Fluorescent contact traces are transferred from many rubber articles, and may be characterised, after extraction, by synchronously excited fluorescence emission spectroscopy. The fluorescence is due mainly to extender and process oils, to antioxidants, and to polynuclear aromatic hydrocarbons. Variation in these components in tyre treads, caused. by variation in manu- facture and wear, enables tread rubbers, and their prints, to be efficiently distinguished from one another. Fluorescent tyre print formation is con- trolled by the nature of the surface, and by contact time and pressure.Some of the fluorescent material solvent-extracted from such prints is present initially as non-fluorescent particles of rubber. Tyre sidewall rubbers are distinguished from tread rubbers by differences in the extraneous polynuclear aromatic hydrocarbon emissions and, in the case of ,radial tyres, principally by differences in extender oil emissions. The techniques described are also applied to traces left by footwear and by erasers, to the detection of mineral oil in failed brake seals, and can be generally applied in the comparative examination of rubbers. In ultraviolet light, rubbers and rubber additives itre often visibly fluorescent. The fluores- cence intrinsic to rubber varies between samples, and in the past its measurement has been advocated as a means whereby manufacturing processes might be visibly monitored.Additives could be evaluated likewise. The pertinent literature has been reviewed by Radley and Grant1 and by Zaitseva and Provorov.2 Although carbon black and other materials used as fillers entirely mask fluorescence, additives are extractable by organic solvetns and, in the case of some antioxidants, may be determined in such extracts spectrofluori- metri~ally.~ To date, these observations have found no general application in forensic science. How- ever, it is now known that fluorescent tyre prints are sometimes formed by the tyres of parked vehicles. Thus, a complete set of fluorescent (but otherwise invisible) prints charac- teristic of a car carrying the body of a victim of murder provided crucial evidence for the prosecution.4 From the laboratory investigations, it is evident that contact traces of rubber can be characterised by fluorescence spectroscopy under circumstances where no other technique is applicable.This paper is concerned with the origin of the fluorescence, and with the factors governing the use of fluorescence spectrloscopy in the characterisation of rubbers. Experimental The spectra, which are uncorrected, are obtained from the equipment previously des- cribed.6 Cyclohexane and chloroform are of commercial spectrophotometric grades, exhibiting no significant extraneous fluorescence emission or ultraviolet absorption. Light petroleum (boiling range 40-60 "C) is purified by percolation through silica gel.Aluminium oxide and silica gel sheets are Polygram Alox N/UV2,, and Pol.ygram Sil G from Machery, Nagel and Co. A 20-W lamp emitting at 366 nm is used to detect transferred fluorescence. Fluorescent prints are made by compression on to the receiving surface of segments cut from tyres, with any object giving pressures in the region of 50 lb in-2, or by compression together of the receiving and tyre surfaces with a pneumatic bead extender (essentially a piece of reinforced hose that is fixed round the tyre and expanded with compressed air), with which contact pressures can be conveniently varied. For spectroscopic evaluation, areas (e.g. , 16 mm2) gouged from the prints with the straight end of a Chataway micro-spatula are transferred to a centrifuge tube, soaked in cycl!ohexane or chloroform (1 ml) for 6 min, shaken and centrifuged.The supernatant liquids are generally strongly fluorescent and require dilution prior to spectroscopic evaluation in order to eliminate inner filter effects. @ Crown Copyright Reserved.LLOYD 83 From prints on other surfaces, fluorescent material is collected in a piece of glass tubing (i.d., 2-5mm; length, 100 mm) loosely plugged with Soxhlet-extracted (diethyl ether) cotton- wool at a constricted end, to which a mouthpiece is connected with PTFE tubing. Loose material, for instance in some shoe scuff marks, is sucked into the tube directly. Strongly adherent material and fluorescent surfaces are loosened with a scalpel blade or, in severe cases, with a tungsten carbide edge, and sucked into the collection tube simultaneously.When much extraneous material is present, amounts collected are generally in the range 10-50 mg. At the other extreme, 10-50 pg are ample. The tube is supported (conveniently with a paper-clip) in the mouth of a centrifuge tube, the collected material moistened and, after 5 min, eluted with solvent (0-5 ml). After centrifugation and, if necessary, dilution, the extract is ready for use. Pieces of rubber (e.g., 50 pg) are soaked with occasional shaking in volumes (e.g., 1 ml) of solvent appropriate to the required concentration for a period of 30 min. The appropriate proportions of rubber to solvent vary, as the text indicates. Erasure marks are extracted in a similar manner. The amounts of paper required are of the order of 40 mm2.It is emphasised that the spectra are uncorrected, and that other instruments may give spectra that differ to various extents. However, it is not likely that any of the conclusions drawn need be modified because of this effect. Discussion and Results Fluorescence Characteristics of Tyre Prints When a tyre rests on a flexible strip of thin-layer chromatography adsorbent, either of aluminium oxide or of silica gel, a fluorescent print appears, usually well within a period of 1 h. The fluorescence, of a blue - white colour when excited by the 366-nm mercury-vapour emission, can be analysed spectrofluorimetrically, directly on the strip, but spectra of improved resolution and intensity are obtained from extracts of the adsorbent in cyclohexane.Typical spectra of an extract excited at fixed and synchronously varied6#' wavelengths are shown in Figs. 1 and 2. When the extract is excited at fixed wavelengths (Fig. 11, emission occurs mainly in the ultraviolet region as a broad envelope with a maximum shifting to longer wavelengths as the excitation wavelength is increased, and, at the sensitivities used, is subject to interference from solvent Raman emissions, indicated by broken lines in the figure. Depending on the excitation interval, the synchronous technique either eliminates or broadens Raman emissions to the extent that they contribute no well defined features to emission spectra, whereas resolution of fluorescence emission is enhanced. Thus, in Fig. 2, synchronous excitation at intervals of between 15 and 100nm reveals the presence of a complex series of emissions not apparent under fixed excitation and negligibly influenced by the Raman effect.That the complexity is due to the presence of many fluorescent compounds, rather than to a few with complex spectra, follows from the restriction of the range of emission wavelengths of each compound to the excitation Further detail is revealed by the use of quenching effects5 Some spectra of material from the same print excited at an interval of 47 nm are shown in Fig. 3. When the cyclo- hexane solution is de-aerated, oxygen quenching is suppressed and the emission at 348 nm is approximately doubled in intensity; but the most significant change is in the emission at 387 nm, which is transformed from an insignificant shoulder to the dominant emission.At 77 K, in ahydrocarbon glass, quenching effects are suppressed even further, the 387-nm emission is intensified and new emissions at 378 and 396 nm appear. From their extensive variation between different samples, it is evident that the three emissions are due to more than one compound. Relative to the cyclohexane solution, in de-aerated chloroform at room temperature, partial quenching occurs, leaving emissions at 433, 412 and 378 nm prominent compared with the remainder. The effect remains to a modified degree when the solution is aerated. Although the over-all fluorescence is now extensively quenched, the differential suppression of the component emissions has increased significantly the number of distinguishable features present.Other features appear when the excitation interval is vaned. Given an adequate level of reproducibility, the detail obtained under the various condi- tions clearly offers excellent possibilities of differentiation between tyres. Spectra of extracts84 LLOYD CHARACTERISATION OF RUBBERS, RUBBER CONTACT TRACES Analyst, VO,!. 100 taken from various points on silica-gel prints of complete tyre treads are generally in excellent agreement with one another. Exceptions arise on the surprisingly rare occasions when bitumen is present. From such experiments (37 spectra from three tyres) the coefficient of variation of emissions, expressed in terms of their intensities relative to the most prominent emission in each spectrum, is 4.03 per cent. I 1 I 1 350 400 450 500 Wavelengthhm Fig.1. Emission spectra in aerated cyclohexane of an extract from a fluorescent tyre print. The fixed excitation wavelengths are : A, 275; B, 300; C , 326; D, 360; E, 375; and F, 400 nm. Broken lines indicate regions of Raman emission from the solvent. 300 ,350 400 450 500 Wavelengthhm Fig. 2. Synchronously excited fluor- escence emission spectra of the same solution as in Fig. 1. A-E, solvent blanks excited a t intervals of 16, 23, 30, 47 and 100nm. F-J, the solution excited a t the same intervals in the same order. The dotted line is a displaced part of J . Origins of Tyre Fluorescence Many of the spectral features are attributable to mineral oil contained in tyre com- positions at a level, on average, of 15 per cent. as an extender,8 together with smaller amounts of process oils and other oils.Spectra from two extender and two process oils are shown in Fig. 4. The spectra, recorded under the same conlditions as some of the spectra in Fig. 3, differ considerably from one another and, indeed, from the comparable spectra of the print extracts. Even so, the more prominent emissions from the extracts correlate individually with those from the oil samples. Despite the considerable variation that exists, this result is given by the majority of tyre treads. (The nature of the luminescent compounds in mineral oils has been discussed previou~ly.~) That the phenomenon of fluorescent tyre-print formation has not been observed until recently is attributed to the relatively recent use of extender oils in tyre-rubber manufacture. Other emissions inconspicuous in extender and process oils often arise from tyre prints, e.g., the emissions at 407 and 432 nm in the 23-nm interval excited spectrum of Fig.2, which are typical of polynuclear aromatic hydrocarbon^.^ Extracts from some prints, particularly when partially quenched in aerated chloroform, give spectra dominated by such emissions, together with others in the region of 460 nm. Compounds commonly producing the effect are anthracene, benzo [alpyrene, benzo [klfluoranthene, perylene and anthanthrene.O The occurrence of polynuclear aromatic hydrocarbons in rubbers is already known ,10,11February, 1975 AND TYRE PRINTS BY FLUORESCENCE SPECTROSCOPY 85 Antioxidants used in rubber manufacture are generally fluorescent and sometimes con- tribute to the spectra of extracts from tyre prints.Spectra of some amino antioxidants, in aerated cyclohexane, synchronously excited at intervals of 30 and 100 nm, are shown in Fig. 6 . All the emissions are relatively featureless, as expected for single compounds, with maxima in the range 350400 nm. Apart from the peak wavelengths, relative intensities of emissions at the two excitation intervals are features that are usefully dependent on antioxidants. Only the N-phenylnaphthylamines are sufficiently emissive to be readily distinguished from the intense extender oil emissions of tyre-tread prints, and identified by means of their excitation spectra. 1 I I I 300 350 400 460 500 Wavelengthh m Fig. 3. Emission spectra, syn- chronously excited a t an interval of 47nm, of an extract from the same print used for Figs.1 and 2 : A, in aerated chloroform ; J3, in de-aerated chloroform; C, in de-aerated cyclo- hexane; and D, in light petroleum at 77 K. Fig. 4. Synchronously excited (47 nm) spectra of process and extender oils: A, Gulf mineral process oil; B, Shell mineral process oil; C, Mobil TY30 extender; D, Dutrex 729 ex- tender. The solvent is de-aerated cyclohexane. The broken line is part of a spectrum of the Dutrex extender in light petroleum a t 77 K. Con- centrations are 10 pg ml-l. Ageing Effects on Tread Prints As the presence of N-phenylnaphthylamines results in strongly fluorescent prints that vary with time, because of decomposition of the antioxidant, the age of a print can be assessed. The effect, exhibited by a tread print containing N-phenyl-2-naphthylamine, is shown in Fig.6, where the synchronously excited (30 nm) emission at 380 nm, largely due to the antioxidant, falls to 54 per cent. of the initial value over a period of 12 days. Un- fortunately, from a forensic point of view, the use of this compound is declining. In only two of 79 treads examined have sufficient amounts been present to produce characteristic spectra. Hopefully, new techniques permitting the use of other, less emissive antioxidants in tyre-print dating will be developed. When fluorescence is due to mineral oils alone, few qualitative changes occur with time, although the print becomes increasingly faint, probably because of diffusion and, on exterior86 LLOYD CHARACTERISATION OF RUBBERS, RUBBER CONTACT TRACES Artalyst, Vol.100 Wavelengttdn m Fig. 5. Synchronously excited spectra of some amino anti- oxidants in aerated cyclohexane: (a), excited a t intervals of 30 nm; and ( b ) , excited a t intervals of 100 ncn. Spectra in each pair are plotted a t the same sensitivity. A, N-phenyl-2-naphthylamine ; B, N-phenyl-l-naphthylamine ; C, N-p henpl-N'-isopropyl-p-phenyl- enediamine ; D, acetone - diphenylamine adduct. All concentrations are 1 p g ml-l. surf aces, erosion and accumulation of quenching material. A pertinent comparison between a new and a much faded 6 months' old print on silica gel is included in Fig. 6. The identity between the two persists under a variety of spectral conditions. 325 375 425 350 400 450 Wavelengthhm Fig.6. Ageing of fluorescent tyre prints: A and B, extracts in aerated cyclohexane of 12 days' old ;and new prints from a tyre containing N-phenyl-2-naphthylaminr:, synchronously excited a t an interval of 30nm; C and D, extrxts in de-aerated chloroform of 6 months' old and new prints from a Goodyear remould tyre, synchronously excited a t an interval of 57 nm; and E, extract from a print of the original interior rubber of the latter tyre, conditions as for C and D.February, I975 AND TYRE PRINTS BY FLUORESCENCE SPECTROSCOPY 87 Variation in Tread Cross-sections The distribution of fluorescent material within a tread cross-section is generally uniform. Among seven cross-sections examined, the maximum variation (see Fig. 6) occurs in a remould tyre from which the remoulding rubber has been pared to expose, to silica gel, the original rubber.Both rubbers are styrene - butadiene compositions shown to be identical by pyrolysis gas chromatography. Even when fluorescent compounds accumulate during use from the tyre’s exterior, as do polynuclear hydrocarbons (see below) , rapid equilibration occurs throughout the tread, presumably because of flexing processes. The occurrence of two types of rubber in a remould tread does not necessarily result in a marked difference in their fluorescence characteristics. In a sample of 30 tyres, analysed by pyrolysis gas chromatography, there was no apparent correlation of rubber type with fluorescence characteristics. Effects of Contact Time and Pressure on Tread Prints Fluorescence intensities vary in a complex way with contact times and pressures. Results from some experiments using a pneumatic bead extender to compress silica-gel sheets on to a tyre tread for various periods of time at constant pressure, and at various pressures for constant periods of time, are plotted in log - log coordinates in Fig.7. [Intensities plotted are of emissions at 415 nm, from extracts (excited at 385 nm) of known dilution, from constant areas of adsorbent displaying identical parts of the tread pattern.] Variation of fluorescence intensity with time over a period from 3 min, when a fluorescent print is clearly visible, to 16 h yields a curve with a gradient always less than unity, i e . , intensities increase less than linearly with time. The phenomenon is attributed to attenuation of the adsorbent’s activity as the amount 0 I I I I I I I I I 1 1 1 I i .I I I I I I I I . 2 10 100 1( Time/mi n Pressure/lb in-* Fig. 7. Effect of time (closed circles) and pressure (open circles) on the intensity of emissions from fluorescent prints formed from a Motorway remould tyre. Further details are given in the text. of material transferred increases. Repeated contact of an area of tyre tread with fresh adsorbent does not result in prints of decreased fluorescence intensity, and therefore the effect is not due to depletion of the rubber in fluorescent material. Neither is fluorescence quenching involved, as the intensity values are from solutions purposely diluted so as to avoid such effects. Despite a 100-fold variation in the amount of fluorescent material trans- ferred, there is no difference between any of the spectra after the appropriate dilutions have been made.By means of control experiments, it is evident that times of contact can be estimated from fluorescence intensities provided that the adsorptive capacity of the surface involved is not exhausted.88 LLOYD: CHARACTERISATION OF RUBBERS, RUBBER CONTACT TRACES Analyst, VOl. 100 Intensity increases with pressure considerably more rapidly than a linear dependence requires (Fig. 7), probably because at increased pressures increased areas of tread are brought into contact with the adsorbent. When prints made at low pressures, e.g., 5-10 lb in-*, are examined visually, fluorescence is seen only at points representing prominences in the tread pattern.Because fluorescent print formation is particularly sensitive to pressure, positions occupied by the most heavily laden wheels of a vehicle can be distinguished under favourable circumstances, as in a recently reported case.4 Surface Effects on Tread Prints Tyre-print fluorescence may be modified by the nature of the surface involved and by the conditions under which observations are made. Visibly, fluorescence is varied in intensity but not significantly in shade by variation of the type of surface and of the wavelength (264 or 366 nm) characteristic of the illuminating lamp. Firm, pervious surfaces, for instance, of concrete or brick, particularly when new or little used, generally favour the effect, which has been observed after contact times of as little as 6 min.Impervious surfaces, for instance, glass, sometimes display weak prints, bituminous surf aces display none, and rarely do surfaces of wood, hardboard, paper and related materials visibly display prints. In the presence of water, prints appear brightened in intensity, probably because light-scattering effects of surface irregularities are reduced. Intensities are increased in the presence of liquid nitrogen, whic-h promotes phosphorescence, but spalling of hard surfaces could cause the technique to be hazardous. Marked differences are apparent when, by the techniques described under Experimental, material collected from fluorescent prints on thest: surfaces is compared spectrophotofluori- metrically with material from prints on silica gel or aluminium oxide.(The prints are from segments of tyre compressed at about 50 lb in-2 on to the surface for a period of 4 h.) For instance, in aerated-cyclohexane extracts of fluorescent material from a concrete surface (Fig. S), the long-lived emissions characteristic of mineral oils7 (at 338, 352 and 373 nm when silica gel is used) are much reduced in intensity. Treatment of the extract with silica gel does not reverse the effect, which is increased in cyclohexane extracts of tyre rubber. In aerated chloroform a broad emission at 415420 nm is intensified, relative to other emissions, in spectra of extracts from prints on concrete paving and flooring, glass, wood and various types of brick. Again, the effect is maximal when rubber is extracted by the solvent directly (or when direct cyclohexane extracts are evaporated to dryness and the residue is dissolved in chloroform).In view of their slight visible emission, prints on glass yield very strongly emissive solutions. Microscopic examination shows that such prints contain many minute fragments of rubber. Even non-fluorescent prints, apparent only as dirt marks on materials such a s wood or paper, contain microscopic fragments of rubber. When marks of this type are extracted with cyclohexane or chloroform, typically fluorescent solutions are obtained. Fig. 9 includes an example. It is concluded from these experiments that fluorescent material transferred from a tyre to a surface will be present in three states to varying extents: as non-fluorescent rubber particles; as a fluorescent adsorbate; and as a fluorescent extrudate when the surface, for instance glass, is not significantly adsorptive.Differences between the spectra in cyclohexane are attributed to the presence in tyre rubbers of fluorescence-quenching materials, presumably sulphur compounds, that are susceptible to solvent extraction from rubber or from transferred rubber particles, but not susceptible to transfer by adsorption. Chloroform exerts strong fluorescence-quenching effects, with which effective competition by rubber-derived quenchers in the dilute solutions used is unlikely. The varying emission in this solvent is therefore assumed to be due to emissive compounds similar i n extractive susceptibility to those causing quenching. The presence of transferred material in several possible states gives rise to fluorescence spectra that vary according to the nature of the surface involved, but the effect is of slight significance and can be accommodated by appropriate control experiments.Spectra obtained from material on the types of surface on which prints are likely to be found in practice indicate that the fluorescence of the extract is due mainly to rubber particles, however visibly bright the fluorescent print may be. When extracts of prints from various tyres on a given surface Prints tend to become complete as the pressure increases. Organic solvents disperse the prints. Some examples are shown in Fig. 9.Febrzcary, 1975 AND TYRE PRINTS BY FLUORESCENCE SPECTROSCOPY 89 are examined, spectral features characteristic of the individual tyres are reproduced.Fluores- cent and ultraviolet-absorbing contaminants are the most usual surface-derived complication ; old or exposed surfaces generally carry an accumulation of such materials, with which the fluorescence of tyre prints cannot compete. 300 350 400 450 500 Wavelengthhm Fig. 8. Synchronously excited (30 nm) spectra of extracts in aerated cyclohexane of: A, blank concrete flooring; B, a fluorescent tyre print on concrete flooring; C, tyre rubber; and D, a fluor- escent print from the same tyre on silica gel. I I I 1 350 400 450 500 Wavelengthhm Fig. 9. Synchronously excited (30nm) spectra of extracts in aerated chloroform of: A, non-imprinted brick; B, tyre print on brick; C, on silica gel; D, on wood; E, on concrete flooring; F, on glass; G, rubber taken directly from the tyre tread.None of the other blank emissions is more intense than A. Each print is from the same tyre. Direct Extraction of Fluorescent Material from Rubber Fluorescent material is rapidly solvent-extracted from the highly dispersed material of tyre prints by the technique described under Experimental. As rubber-fragment size in- creases, reduced extraction rates consistent with reduced area to mass ratios are observed. Some results obtained by soaking regularly shaped fragments, 8-612pg in mass, in 1-ml amounts of cyclohexane are shown in Fig. 10. Constant fluorescence intensities (365 nm, excited at 335 nm) are established in 10-50 min, depending on the amount of rubber used. Spectra of the final extracts, diluted to equivalent concentrations, are independent of the amounts used.The final emission intensities vary linearly (the slope of the log - log graph is unity) with the amounts of rubber taken (Fig. ll), and therefore this particular emission is not subject to a significant level of quenching by rubber-derived material. [Fluorescence intensity and concentration of quencher are both derived from the same amount of rubber in any given90 LLOYD: CHARACTERISATION OF RUBBERS] RUBBER CONTACT TRACES An6!&Stl VOi?. 100 ,r.--t-t-= C 'e B w- A d = a I I L 20 40 60 80 100 Ex tract io 11 t i me/m i n Fig. 10. Variation with extraction time of fluorescence intensity (at 365 nm, excited at 335nm) of extracts in aerated cyclohexane of rubber llragments.Frag- ment concentrations: A, 8 ; B, 20; C, 41; D, 112; E, 320; and F, 612 pgml-l. solution; hence, in the single component instance, the Stern - Volmer equation may be reduced to F = aW (1 + KbW)-l1 where F is the observed fluorescence intensity, a and b are constants] W is the mass of rubber taken and K is the Stern - Volmer quenching coefficient. When quenching] i.e., K , is significant, F becomes a non-linear function of W. More complex cases are considered elsewhere.'] Similarly, as Fig. 11 shows, the same emission from solutions of extender oil varies in intensity linearly wit;h concentration up to levels in the region 10-50 pg ml-1, when quenching intrudes. Determination of the oil content of rubbers from such results is complicated by variations in fluorescence yields of different oil samples, of which the plotted results are an example, and by the presence of fluorescent additives and extraneous compounds, whose interference can be minimised at appropriate excitation intervals. On the basis of the data in Fig.11, the oil content of the sample of rubber used is in the range 9.2-10.2 per cent. I Oil or rubber cimcentration/,ug ml-' Fig. 11. Variation of fluorescence intensity (365 nm, excited at 335 nm) with amount of oil or rubber extracted in aerated cyclohexane. Full line and closed circles : Dutrex extender oil. Broken line and open circles: Mobil extender. Full line and crosses : tyre-tread rubber.February, 1975 AND TYRE PRINTS BY FLUORESCENCE SPECTROSCOPY 91 Variation Between Tyre Treads Because transfer of fluorescence from tyres is, essentially, transfer of incipient fluores- cence in the form of rubber particles, the extent to which tyres can be differentiated is assessed from the spectra of solvent extracts of tread rubbers (collected from recent cases).The sample, representing 23 manufacturers, includes 26 radial-ply tyres, 24 cross-ply, and 11 remoulds of which nine are cross-ply. The data used are from synchronously excited (30-nm interval) spectra of extracts in chloroform, plotted for each of two fragments from opposite points on the treads of the 61 tyres; and, similarly, from spectra (30- and 100-nm intervals) of extracts in cyclohexane from 47 of the tyres. In both extractions the amount of rubber used is 83 pg ml-1. From the spectra are calculated the following : intensities from cyclohexane solutions of the 365-nm peak, excited at the 30-nm interval, relative to thescattered excitation (335 nm) signal from the pure solvent, with which the instrument’s sensitivity is standardised between each plot; the ratio of maximum intensities from cyclohexane solutions at the two excitation intervals (the 30-nm maximum relative to the 100-nm maximum) ; peak heights in chloroform solutions at 374, 387 and 417 nm in excess of and relative to intensities at 400 nm, which are usually featureless; and, from the latter spectra, peak heights at 465 nm, relative to the intensities at 400 nm, with the assumption of linear background variations between inflections on either side of the peaks.Most of the features present in rubber fluores- cence are reflected in the results obtained.Further detail is revealed under other conditions, particularly from de-aerated solutions. However, the tendency of spectral sensitivity dis- tributions of uncorrected spectrophotofluorimeters to drift with time requires comparisons of the type described to be completed rapidly and restricts the numbers of samples and techniques that can be handled. In Table I are given F ratios, expressing between-tyre variance in terms of within-tyre variance, from an analysis of variance of the spectral data. The upper 0.1 per cent. points of F distributions with both degrees of freedom of 40, and both of 60, are 2-73 and 2.25, respec- tively. Hence, the various characteristics whose F ratios lie in the range 16-45 form an effective basis on which tyres in this particular sample can be distinguished, provided that adequate allowance is made for the varying levels of correlation, e.g., by the method of Smalldon and Moffat.12 From a population characterised by the parameters of Table I, and from Smalldon and Moffat’s equation for three correlated Gaussian attributes, it can be shown13 that pairs of randomly selected individuals will match in characteristics 1, 2 and 5 , within two standard errors of each characteristic, with a probability of 0.02.i.e., the dis- criminating power is 0.98. (There is no implication intended here that from the small sample used the actual population parameters can be adequately estimated; the figures quoted are criteria of specificity for the experimental techniques applied within the limits of this particular sample.) TABLE I VARIATION IN FLUORESCENCE CHARACTERISTICS BETWEEN TYRE TREADS F ratios +- Correlation coefficients t Degrees of w \ Fluorescence characteristic* freedom F (2) (3) (4) ( 5 ) (6) (1) Cyclohexane, 365 nm .. . . 46, 47 45 0.475 0.03 0.04 -0.54 -0.49 (2) Cyclohexane, ratio of maxima (3) Chloroform, 374 nm . . . . 61, 61 21 0.37 1 at 30 and 100 nm . . .. 46,47 39 1 (4) Chloroform, 387 nm . . . . 61,61 20 . -0.02 0.39 1 (6) Chloroform, 417 nm . . .. 61, 61 37 -0.59 -0.40 -0.09 1 (6) Chloroform, 465 nm . . .. 61, 61 16 -0.62 -0.29 0.18 0.91 1 * Further details are given in the text. t Significance levels are 0.1 per cent. for the coefficients 5/6, 2/5, 2/6, 1/2, 1/5 and 116; 1 per cent. for 3/4, 3/5 and 213; and 5 per cent. for 316 (numbers refer to fluorescence characteristics).The coefficients are based on data from 47 tyres. Some Origins of Between-tread Variation Of the characteristics examined the intensity of emission in cyclohexane at 365 nm, which is proportional to the oil content of the rubbers, varies most extensively. Two factors are clearly identifiable. First, a variation occurs between tyre treads of different manufacture.92 Analyst, VoE. 100 Thus, six Michelin tyres, 5-6 mm in tread-depth., yield a mean and standard deviation of 0.211 and 0.027, respectively, whereas corresponding values for ten tyres, worn to the same degree, of various other manufacturers are 0.545 and 0.183, respectively. The difference is significant at a level of 0.1 per cent.(t distribution). Secondly, the characteristic is correlated to tread-depth. In the sample of 47 tyres, the correlation coefficient is 0.43 (significance level, 1 per cent .). When between-manufacturer variation is eliminated, in a sample restricted to 15 Dunlop tyres, the correlation is 0.73 (significance level, 0.1 per cent.). Evidently, extender oil is lost more rapidly than rubber as a tyre tread wears. Calculation of the 365-nm characteristic requires the amount of rubber extracted to be known. This is possible in practice when transEer results from impact damage; but in the examination of a tyre print, which may involve the extraction of material derived from 1-10 pg of rubber distributed through about 10 mg of surface scrapings, quantitation is obvi- ously impossible.However, the substantial correlation between this characteristic and that defined simply as the ratio of maximum intensities in the 30- and 100-nm interval excited cyclohexane spectra (characteristic 2) provides at least some remedy t_o the situation. Possibly the high F ratio of the latter characteristic derives from the degree of correlation be- tween the two. The polynuclear hydrocarbon emissions in chloroform, at 417 and 465 nm (Table I), are highly correlated. That the emissionssare not due to a single compound follows from their separation by more than the excitation interval, i.e., 30nm. However, in view of the relatively low F ratio of the 465-nm emission, together with the high correlation coefficient, the emission intensity is evidentially largely irrelevant.Although polynuclear hydrocarbons occur in carbon black, their typical emissions are generally weak in extracts of new or little- used tyres. Apparently, tread rubbers accumulate these compounds during use, probably from road surfaces. Between tread depth and the 417-nm emission, the correlation coefficient taken over 47 tyres is -0-45 (significance level, 1 per cent.). Sidewall Rubbers Sidewall rubbers are particularly liable to transfer from impacting vehicles. In general, sidewall and tread rubbers are markedly differentiated by the strong polynuclear hydrocarbon emissions derived from the latter. Thus, a paired comparison of the differences between 30-nm interval excited 417-nm emissions in chloroform (characteristic 5, Table I) of tread and sidewall rubbers from 13 tyres, of mean tread depth 5.1 mm, yields a mean and standard deviation of 0.125 and 0.226, respectively.The mean is significantly different from zero (in favour of tread rubbers) at a level of 5 per cent. (t distribution), which lends further support to the view that the presence of the hydrocarbons is a consequence of contact between road and tyre surfaces. Apart from the polynuclear hydrocarbon emissions, no other marked, general difference exists between tread and sidewall rubbers of cross-ply tyres. Radial-ply sidewalls, however, are consistently distinguished by low fluorescence yields indicating reduced oil contents. In a sample of eight cross-ply tyres from various manufacturers (mean tread depth, 5.7 mm), the mean and standard deviation of emission intensities at 365nm in cyclohexane (charac- teristic 1) of sidewall rubbers relative to tread rubbers are 1.201 and 0.388, respectively, whereas seven radial-ply tyres (mean tread depth, 5-5mm) give values of 0.304 and 0-174, respectively.The difference is significant at a level of 0.1 per cent. (t distribution). A second fluorescence feature of radial-ply sidewalls, consequent upon the reduced oil fluorescence, is the more readily distinguishable contribution made by antioxidants, which are not now restricted to the N-phenylnaphthylamines. For instance, $-phenylenediamine derivatives are sufficiently emissive to be detected. It might be expected that fluorescence from remould sidewalls would depend on the composition of the underlying rubber as well as on the new surface composition.Although from the tyres examined a considerable variation exists, no consistent trends are yet apparent. Eraser Rubbers Direct extraction of eraser rubber with cyclohexane, in amounts of 0.1-20 mg ml-l, commonly yields fluorescent solutions exhibiting synchronously excited (30-nm interval) maxima (at 338 nm) Characteristic of light mineral oils. Apart from a considerable variation in intensity (e.g,, the extremes in a sample of seven erasers varied in this respect more than LLOYD : CHARACTERISATION OF RUBBERS, RUBBER CONTACT TRACESFebruary, 1975 AND TYRE PRINTS BY FLUORESCENCE SPECTROSCOPY 93 100-fold) considerable qualitative variation in the spectra occurs both in the distribution of oil-derived emissions and in the presence of emissions from fluorescent additives. Examples from three erasers are shown in Fig.12. Features characteristic of each of these erasers are preserved in cyclohexane extracts of paper on which the erasers have been used, as the figure shows. The eraser marks (erasures) are negligibly fluorescent, hence the transferred fluores- cence is in the form of finely dispersed rubber. Because the amount of paper required is fairly small, about 40mm2, and can be recovered after extraction, correlation of erasures with erasers is clearly practicable. In one example, an intense emission at 400 nm is present, which is attributed to an antioxidant. Over a period of 8 days the emission intensity, relative to the weak 338-nm emission, decays to half of the initial value.Evidently, recent use of erasers of this type can be dated. 300 350 400 450 W aveleng t h/n m Fig. 12. Synchronously excited (30 nm) emission spectra of extracts in aerated cyclohexane of various eraser rubbers (B, C and D, full lines) and corresponding eraser marks (dotted lines). The paper blank at the maximum sensitivity used is also shown (A). A weak emission a t 335 nm in spectrum D is not reproduced. Footwear Samples of rubber from shoe soles and heels yield cyclohexane extracts that vary widely in fluorescence properties. Synchronously excited (30- and 100-nm interval) spectra of the less emissive examples are characteristic of mineral oils and polynuclear aromatic hydro- carbons. In some cases, for instance, when there has been contact with a surface subject to burning, emissions from the last-mentioned group of compounds predominate.Even the relatively weakly fluorescent rubbers leave sufficient material ( e g . , 1 mg) for characterisation in scuff marks. However, fluorescent additives giving rise to strong emissions that extend into the visible region are often present. On these occasions amounts of rubber in the range94 Analyst, VoZ. 100 1-1Opg may be characterised. In Fig. 13 are shown spectra of extracts from two scuff marks together with spectra from the shoe responsible, the sole and heel of which are shown by appearance and by pyrolysis gas chromatography to be made of matching rubbers. From the spectra, however, it can be seen that different additives are present, which cause the sole and heel scuff marks to differ markedly from one another, but to correlate closely with the relevant parts of the shoe.Oil-contaminated Rubbers Transfer of trace amounts of oil to, rather than from, rubbers is sometimes a point of importance to investigations of mechanical failures. Thus, mineral oil adventitiously present in brake fluids is absorbed by and causes failure of rubber seals, in which the presence of mineral oil can be readily demonstrated. Fig. 14 shows a series of spectra, synchronously excited at an interval of 30 nm, of cyclohexane extracts of equivalent amounts of rubber seals previously exposed to various levels of oil contamination, together with the spectrum of the contaminating oil. Apart from slight variations in the emission at 385nm, there is LLOYD : CHARACTERISATION OF RUBBERS, RUBBER CONTACT TRACES r I 300 350 400 450 500 Wavelengthhm Fig.13. Synchronously excited (30 nm) emission spectra of extracts in aerated cyclohexane of material from a shoe sole and heel rubbers (A and C) and scuff marks made by the shoe (B and D). 300 350 400 450 Wave I engt h/n m Fig. 14. Synchronously excited (30nm) spectra of extracts in de-aerated cyclo- hexane of: A, solvent blank; B, unused brake seal; C, seal used in uncontaminated fluid; D, seal exposed for 30 days to fluid containing 0-5 per cent. of mineral oil; E, seal exposed for 16 h to neat oil; and F, the con- taminating oil. Rubber con- centrations are 18 p g ml-1; the oil concentration is 5.7 pgml-l. Spectrum F is a t 61.7 per cent.of the sen- sitivity of the remainder. close agreement between the spectra of the contaminating and extracted oils. From emission intensities at 338 nm it can be calculated that in the fluid containing 0.5 per cent. oil (30 days’ exposure) the amount of oil accumulated by the seal is 12.9 per cent., whereas in neat oil (16 h exposure) the accumulation is 18.1 per cent. From the increased mass of the seals it is calculated that the actual values are 11.1 and 18.5 per cent., respectively. Evidently,February, 1975 AND TYRE PRINTS BY FLUORESCENCE SPECTROSCOPY 95 the absorption process in this example negligibly discriminates between the small amounts of fluorescent components of the oil and the remainder. Miscellaneous Many other rubbers on extraction yield fluorescent materials mainly of the types already indicated.In some instances, further possibilities for characterisation are provided by the presence of fluorescent pigments, to which the previously described techniques are applicable. Occasionally, high levels of pigments and other additives may interfere in the detection of less fluorescent components. Under these circumstances, extraction by the silica-gel technique sometimes enables more strongly adsorbed interfering materials to be removed. Conclusion The factors that determine the nature of the fluorescence of a rubber are largely inde- pendent of the major components that determine the results of other techniques of character- isation applicable to very small samples. Hence, fluorescence spectroscopy yields results inde- pendent of those obtained, for instance, by pyrolysis gas chromatography, when circumstances permit application of both techniques. Unlike the chromatographic technique, however, fluorescence spectroscopy does not require a coherent sample, and can be applied in the characterisation of contact traces much diluted by extraneous material. It is now possible, therefore, to correlate rubbers to points of contact in situations where previously no chemical evidence could be obtained. I acknowledge my indebtedness to Mr. I. W. Evett for statistical calculations, and to several members of the rubber and petrochemicals industries for various samples. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. References Radley, J , A., and Grant, J., “Fluorescence Analysis in Ultraviolet Light,” Fourth Edition, Chapman and Hall, London, 1954, Chapter XVII. Zaitseva, V. D., and Provorov, V. N., in Konstantinova-Shlezinger, M. A., Editor. “Fluorimetric Analysis,” Moscow, 1961 ; translated by Kaner, N., Israel Program for Scientific Translations, Jerusalem, 1965, pp. 233-245. Parker, C. A., and Barnes, W. J., Analyst, 1957, 82, 606. Loughran, J . H., Lloyd, J. B. F., and Watson, T. R., Nature, Lond., 1974, 250, 762. Lloyd, J. B. F., Analyst, 1974, 99, 729. Lloyd, J. B. F., Nature Phys. Sci., 1971, 231, 64. Lloyd, J. B. F., J . Forens. Sci. SOG., 1971, 11, 83. Corbin, H. E.. Rubb. Age, N.Y., 1974, 106, 49. Lloyd, J . B. F., J . Forens. Sci. SOL, 1971, 11, 235. Falk, H. L., Steiner, P. E., Goldfein, S., Breslow, A., and Hykes, R., Cancer Res., 1951, 11, 318. Smith, C. G., Diss. Abstr., 1967, 28, 522B. Smalldon, K. W., and Moffat, A. C., J . Forens. Sci. SOL, 1973, 13, 291. Evett, I. W., unpublished work. Received July lst, 1974 Accepted July 24th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9750000082

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

96 Analyst, February, 1975, Vol. 100, p p . 96-98 Determination of Zinc in Soils by X-ray Fluorescence Spectrometry Involving a Modified Background-ratio Method H. Bergseth Department of Soil Science, Agricultural University of Norway, 1432 AS-NLH, Novway The zinc content of soil samples that contained different amounts of organic matter was determined by X-ray fluorescence spectrometry with a modified background-ratio method. The ratio, 1% Ka/I$17, in which I , is the back- ground intensity at 39.50 O 2 8 , was substantially independent of matrix for the range of zinc contents investigated. The results coi-related well with those obtained by atomic-absorption spectrometry. A simple ratio method for the determination of trace elements, by X-ray fluorescence spectro- metry, in soil samples with widely different organic matter contents was published by Leake and Peachey.1 For such samples, however, the 'background-ratio method2-& can be used to compensate for variations in absorption, enha-ncement and in instrumental conditions.The aim of this paper is to show the applicability of a modified background-ratio method to the determination of zinc in soil samples. Experimental Sample Preparation Samples (with organic matter contents in the range 30-98 per cent.) from the upper horizon of forest soils in Odda (Norway) were ground to less than 0-074mm and dried overnight at 105 "C. Portions of 5.00 g were then thoroughly mixed with 1.00-g portions of wax before being pressed into pellets in aluminium cups at tt pressure of 25 tons.Standards were prepared by adding known amounts of zinc as zinc sulphate solution to 5-00-g portions of ground and dried peat soil and to mixtures of peat soil and subsoil having a known content of zinc. After drying, each poirtion was mixed uniformly with 1.OOg of wax and pressed into pellets. Apparatus Spectrograph . . .. .. .. . . Philips PW 1410 Excitation . . .. .. .. . . Chromium tube, 30 kV, 30 mA Detect or .. .. .. .. . . Scintillation Collimator . . .. .. .. .. Coarse Radiation path .. .. .. .. Air Analysing crystal . . .. .. * a LiF20? Procedure ground). Pulse height discrimination was not employed. Counts were accumulated over a period of 40 s at 41.73 "28 (Zn Ka) and 39.50 "28 (back- TABLE I: INTENSITIES (counts s-I), INTENSITY RATIOS FOR FOUR DIFFERENT MATRICES WITH (a), 1000 p.p.m.OF ZINC AND (b), 1800 p.p.m. OF ZINC AND RELATIVE STANDARD DEVIATION (€I) OF THE RATIO I z ~ g&/1i1' IZn Ka IB IZn Ktr/IB Izn K a p y cr Matrix a b a b a b a b - 1 5793 10020 481 480 12-04 20.88 4.21 7.30 0.87 0.88 2 5380 9190 442 443 12.17 20.74 4.32 7.38 0.91 0.89 3 2795 4820 268 258 10.83 18.68 4-22 7.27 1-19 1-17 4 1705 2880 165 166 10.33 17.45 4.34 7.33 1.49 1.47 * * + - - 7 *BERGSETH 97 The intensity ratio, Izn Ka/IB, in which IB is the background intensity at 39.60 "28, did not have the same value for four different matrices of a given zinc content. The matrices examined contained 1000 and 1800 p.p.m. of zinc in dry matter (Table I). In order to obtain a standard curve relating the intensity ratio to the zinc content for samples covering the present range of organic matter content, an empirical exponent t such that Results for different pairs of matrices was calculated for matrices containing 1000 p.p.m.of zinc and 1800 p.p.m. of zinc. 0.02 for matrices containing 1000 and 1800 p.p.m. of zinc, respectively, were obtained. The standard curve established for Izn ga/Ii17 veisus zinc concentration is shown in Fig. 1. Mean values for t of 1.16 & 0.04 and 1.17 Zinc in dry matter, p.p.m. Fig. 1. Standard graph representing four different matrices : 0, matrix 1, organic matter 98 per cent.; a, matrix 2, organic matter 95 per cent.; A, matrix 3, organic matter 70 per cent.; and A, matrix 4, organic matter 30 per cent. The effects of variations in the thickness of the pellets were examined (Table 11). As the sample mass was decreased from 5.0 to 1-5 g , I z n K a and I B decreased. The ratios Izi K a / I B and I z n Ka/1k17 also showed small variations. It appeared, however, that the use of pellets that contained 5-00 g of sample gave satisfactory results in the determination of zinc.TABLE I1 VARIATION WITH PELLET THICKNESS OF INTENSITIES (counts s-l), INTENSITY RATIOS AND ZINC CONTENT (OBTAINED BY USING THE STANDARD CURVE) B C The organic matter contents for samples A, B and C were 58, 87 and 92 per cent., respectively. Zinc content, Sample Sample mass/g Izn ~ g l I B I Z n K a / I B IZn Ka/1:'' p.p.m. A 1.60 2036 167 12-19 5.13 1220 2.60 2088 175 11.93 4-96 1175 6.00 2139 181 11-82 4-87 1166 8.00 2110 186 11.34 4-67 1105 1.50 1114 323 3.46 1-29 248 2.50 1146 326 3-52 1-31 263 6-00 1224 357 3.43 1.26 240 8-00 1229 362 3.40 1.25 238 1-60 1273 36 1 3-63 1.34 260 2.50 1374 3 85 3-67 1-30 260 6.00 1468 409 3-66 1.28 246; 8-00 1482 422 3-51 1.26 24098 BERGSETH As this is a fixed-time method, the counting error varies from sample to sample.At very low concentrations of zinc or in a strongly absorbing matrix it may be necessary to increase the counting time in order to obtain satisfactory precision. In the determination of zinc in samples with various organic matter contents by X-ray fluorescence spectrometry results were obtained that correlated well with those obtained by atomic-absorption spectrometry (Fig. 2). During the latter procedure, measurements were made on the solutions obtained after the samples, ignited at 500 "C, had been boiled for 3 h in 50 per cent. V/V nitric acid. 10 Zinc in dry matter, p.p.m. (atomic absorption) ( x ) Fig. 2. Comparison of zinc contents deter- mined by X-ray fluorescence spectrometry and atomic-absorption spectrometry in samples with different organic matter contents : y = 1*024x+ 3, Y = 0.997. 0-20 5 20-40 7 40-60 6 60-80 2 80-100 6 Organic matter, per cent, Number of samples The author thanks the Agricultural Research Council of Norway for grants received and Mr. Johnny Kristiansen for technical assistance. References 1. 2. 3. 4. Leake, R. C., and Peachey, D., Trans. Instn Min. Metall., Sect. B, 1973, 82, B25. Andermann, G., and Kemp, J. W., Analyt. Chem., 1958, 30, 1306. Clark, N. H., and Mitchell, R. J., X-Ray Spectrom., 1973, 2, 47. Kalman, 2. H., and Heller, L., Analyt. Chem., 1962, 34, 946. Received May 16th, 1974 Accepted September 16th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9750000096

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

Analyst, February, 1975, VoZ. 100, pp. 99-104 99 A Modified Procedure for the Determination of Nicotine in Blood S. E. Falkrnan I, E. Burrows R. A. Lundgren and B. F. J. Page Research Laboratories, A B Leo, Helsingborg, Sweden Department of Applied Research, Pedigree Petfoods Limited, Melton Mowbray, Leicestershire, LEI3 1 BB Research Laboratories, A B Leo, Helsingborg, Sweden Department of Metabolism and Pharmacokinetics, Huntingdon Research Centre, Huntingdon An analytical method for the determination of nicotine has been established. It is a modification of an earlier reported method in which the over-all recovery of added nicotine has been increased from 55 to over 80 per cent. The determination is carried out by gas - liquid chromatography and the identity of the nicotine peak has been confirmed by mass spectrometry.The method has been applied to the determination of nicotine in the blood of cigarette smokers, and concentrations a t the level of 1 ng ml-1 can be measured. For studies of human blood levels of nicotine, it is necessary to use a method of analysis that is sensitive enough to give accurate results at nicotine concentrations of a few nanograms per millilitre of blood. Most methods previously de~cribedl-~ are either difficult to apply or lack the required sensitivity for the determination of nicotine in this concentration range. Schievelbein and Gr~ndke,~ Isaac and Rand5 and Burrows et aLs have developed useful methods for the determination of small amounts of nicotine in blood. A modification of the gas-liquid chromatographic method developed by Burrows et aL6 is now reported.The purpose of our modifications has been to improve the over-all sensitivity of the methodas well as to decrease the level of interfering co-extractives. Method Apparatus Steam-distillation apparatus. All glass. Varian 1400 gas - liquid chromatograph. LKB 2091 gas chromatograph - mass spectrometer combined with an LKB 2130 data system. All reagents were of analytical-reagent grade. Water was double distilled from glass Sodium hydroxide solution, 10 N. Alkaline salt solution. saturated with sodium chloride. Sulphuric acid, 1 N and 0-5 N. Dichloromethane. Benzene. Silicone MS antifoam A . Sodium chloride. 1-Nicotine. Obtained from Koch-Light Ltd., re-distilled under vacuum and stored at -10 "C under nitrogen. QuinoZine.Obtained from E. Merck A.G. Glass Equipment All glass equipment was treated overnight with Diversey Pyroneg cleaning fluid, rinsed several times with tap water and finally rinsed with ethanol before drying at 100 "C. Fitted with a flame-ionisation detector. Reagents apparatus before use. A solution of 300g of sodium hydroxide in 700ml of water,100 A d y s t , VoZ. 100 Procedure Obtain blood samples from the subjects by venepuncture in the arm. Dilute 10ml of heparinised whole blood with 10 ml of glass-distilled water, make the solution alkaline with 10 ml of alkaline salt solution and add 0.1 g of antifoam A. Steam distil the mixture at a distillation rate of 5 ml min-l, and collect the first 70 ml of distillate in a separating funnel containing 2 ml of 1 N sulphuric acid.Extract the acidified distillate with 10 ml of dichloro- methane and discard the organic phase. Make the aqueous phase alkaline by adding 2 ml of 10 N sodium hydroxide solution and extract the mixture four times with 10 ml of dichloro- methane. Extract the combined dichloromethane layers four times with 2 ml of 0.6 N sulphuric acid. Combine the aqueous layers in a 10-ml calibrated flask containing 2.5 g of sodium chloride and make the mixture alkaline with 1 ml of 10 N sodium hydroxide solution. Extract the nicotine with 100 p1 of benzene containing 5 pg ml-l of quinoline as an internal standard. Inject 6 p1 of the benzene solution on to the gas-chromatographic column using a Hamilton syringe. Gas Chromatography The gas chromatograph was fitted with a glass column (2 m x 3 mm id.) packed with 8 per cent.m/m Carbowax 20M and 2 per cent. m/m potassium hydroxide on acid-washed Chromosorb W, 80-100 mesh, treated with hexamethyldisilazane. The optimum chromatographic conditions for the determination of nicotine were as follows : column temperature, 150 "C ; nitrogen flow-rate, 27 ml min-1; hydrogen pressure, 2.2 x lo5 N m-2; air pressure, 4-1 x lo6 N m-2; and nitrogen pressure, 3.1 x 105 N m-2. The retention time was about 6 min for nicotine and about 8 min for quinoline. At a signal to noise ratio of 3: 1, the detection limit for nicotine was 0.04 ng. Gas Chromatography - Mass Spectrometry A glass column (2.7 m x 2 mm i.d.) packed with 6 per cent. m/m Carbowax 20M and 2 per cent. m/m potassium hydroxide was used.The gas-chromatographic operating condi- tions were as follows: column temperature, 170 "C; injection port temperature, 200 "C; helium flow-rate, 30 ml min-l; and inlet pressure, 3-1 x lo5 N m--2. The retention time for nicotine under these conditions was about 11 min. The mass-spectrometric operating conditions were as follows : ion-source temperature, 270 "C; inlet tube temperature, 230 "C; electron energy, 20 or 70 eV; trap current, 50 PA; accelerating voltage, 3-0 kV. After injection of a sample, a mass spectrum was scanned every third second. The spectra were stored on magnetic tape and, after the run, the total ion current was calculated by the computer and the total ion chromatogram was printed out together with mass chromatograms for three mass numbers typical for nicotine.This technique, ''mass chromatography," to detect the presence of a compound of known mass spectrum, was described by Hites and Biemann.' The mass spectrum of nicotine has four typical fragments (see, for example, McLafferty8 and SastryQ) at m/e 162 (molecular ion), 161, 133 and 84 (basic peak). In this investigation, the mass numbers 162, 133 and 84 were used in order to confirm the identity of extracted nicotine. The method required a little more substance than the standard procedure. A male volunteer smoked two cigarettes and 100 ml of blood were drawn after 5 min. The blood was divided into ten portions and analysed according to the standard procedure. The ten extracts were combined and the volume was reduced by extraction into sulphuric acid followed by re-extraction int.0 0.1 ml of benzene as described above.A 5-p1 volume was injected into the gas chromatography - mass spectro- metry system, and chromatograms were obtained. FALKMAN et al.: A MODIFIED PROCEDURE FOR THE Results and Discussion Identification of Nicotine Chromatograms of extracts of blood from a smoker and of control blood are shown in Figs. 1 and 2. In order to establish the identity of nicotine extracted from a smoker's blood, the experiments with mass spectrometry were performed. As can be seen from the total ion chromatogram and the mass chromatograms in Fig. 3, all the typical fragments are present in the presumed nicotine peak. Mass spectra were printed at the beginning and end of the peak and both were identical with an authentic mass spectrum, thus proving that the sub-February , 1975 DETERMINATION OF NICOTINE I N BLOOD 101 stance extracted from blood was nicotine that had passed through the entire analytical procedure unchanged.Fig. 1. Gas chro- Fig. 2. Gas chro- matogram of extract matogram of extract of of blood from a smoker. control blood from a 1, Nicotine; and 2, non-smoker. 2, Quino- quinoline. line; and 3, blank. As has also been observed by other analysts working on nicotine analysis (e.g., Isaac and Rand6), it was found that blood from non-smokers gave a small peak with the same retention time as nicotine. A repetition of the mass-spectrometric experiment using control blood from a non-smoker resulted in a mass chromatogram and mass spectra in which all the fragments characteristic of nicotine were present, thus indicating that nicotine may occur in small amounts also in the blood of non-smokers.This conclusion is in accordance with the discovery of Horning et aZ.10 that detectable amounts of nicotine are excreted in the urine of non-smokers who have been exposed to tobacco smoke. The blank value was deter- mined and subtracted from the result obtained for every subject involved in the investigation, and had to be assumed to be constant during the time of the experiment. The method of analysis has been applied to several determinations of nicotine in the clinical field, the results of which will be published elsewhere, and the blank value mostly corresponded to about 5 per cent. of the internal standard or 3 ng ml-l in blood.I 1 V 0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 I I I t I I I I I < TIC 0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 84 1 33 162 - Ti me/m i n ' 1, Nicotine; and 2, quinoline. Fig. 3. Total ion chromatogram (TIC) and mass fragmentograms for mass numbers 162, 133 and 84 from extract of blood from a smoker.102 Recovery Experiments solution to be injected into the gas chromatograph consists of five steps: FALKMAN et al.: A MODIFIED PROCEDURE FOR THE Analyst, VoZ. 100 The procedure for the extraction and purification of nicotine from blood to the benzene I. Steam distillation. 11. Purification of the distillate with dichloromethane. 111. Extraction four times with dichloromethane after making alkaline. IV. Re-extraction four times with sulphuric acid.V. Extraction of aqueous solution with benzene, Three series of experiments have been carried out in order to determine the recovery of nicotine and to establish the step in which loss of the alkaloid may occur. The results are expressed in Table I. The last step (V) is a single extraction of 10 ml of saline solution TABLE I RECOVERY OF ADDED NICOTINE IN DIFFERENT STEPS OF THE PROCEDURE Added nicotinelng Step Fluid (total) V Water 100 600 I I-v Water 100 600 I-v Whole blood 100 (complete procedure) 600 Number of experiments 13 13 9 6 12 10 Recovery, per cent. (95 per cent. confidence interval) 90.6 f 3.7 88.6 f 2.3 84.8 f 2-4 83.7 f 2.4 86.4 f 6.7 79.8 f 2.1 with 1 0 0 ~ 1 of benzene and, a s can be seen from the results, 90 per cent. of the nicotine appears in the organic phase.As this level is obtained as a result of the distribution co- efficient, it is theoretically the highest possible recovery of nicotine for the whole procedure. When nicotine was added to water and steps II-V were performed, the recovery was found to be only slightly lower (about 85 per cent.), which demonstrates that the loss in the extraction procedure occurs mainly in step V. When working with blood, the extraction procedure was impracticable because of considerable foaming ; this problem was solved by introducing the steam distillation (step I) as the first phase of the operation. The last results in Table I OY 100 200 300 400 500 I I 1 I 1 Nicotinehg Fig. 4. Calibration graph for deter- mination of nicotine in smokers’ blood. Quinoline is used as an internal standard.February, 1975 DETERMINATION OF NICOTINE I N BLOOD 103 give the recovery for the entire procedure and show that it was not further reduced by the steam distillation.TABLE I1 RESULTS OF RECOVERY EXPERIMENTS WITH CONTROL BLOOD TO WHICH KNOWN AMOUNTS OF NICOTINE WERE ADDED Added/ng ml-l 10 10 10 10 10 10 10 10 10 10 10 10 Mean .. .. .. 95 per cent. confidence inter- Standard deviation, per cent. Val, per cent. . . .. Found/ng ml-1 8-40 9.59 9-38 9.10 10.78 11-41 9.94 10.08 12-11 9.10 9.59 10-36 9.99 (99.9 f 7.0) 10.5 Addedlng mi-' 50 50 60 60 60 50 60 50 50 50 Found/ng ml-l 47.5 48.4 48.8 48.4 46.8 48.2 48.6 50-6 45.6 46.1 47.9 (95.8 f 2.1) 3-1 Applications With quinoline as an internal standard, a calibration graph was obtained using blood from non-smokers fortified with various amounts (0-50 ng ml-l) of nicotine (Fig.4). After correction for the individual blank value, this graph was used for routine analysis. In order to check the precision of the method, two series of recovery experiments were carried out with blood to which nicotine was added (10 and 50 ng ml-l). The results evaluated from the calibration graph (Table 11) show a standard deviation of 10 per cent. at the lower level and 3 per cent. at the higher level. t I I I I I I 5 10 15 20 25 30 Tirne/min Fig. 5. Variation of blood concentration of nicotine after smoking one cigarette. Subjects are identified by initials on graphs.104 FALKMAN, BURROWS, LUNDGREN AND PAGE This method of analysis was used to determine nicotine in smokers’ blood.In the first investigation, four subjects (all male) smoked one cigarette after a nicotine-free period of 12 h (overnight). The smoking time was 5 min, and blood samples were drawn just before (for determination of the personal blank value) and 2.5, 5 , 7.5, 10, 15 and, in some instances, 20 and 31) min after smoking commenced. The results are shown in Fig. 5, and, as can be seen from the curves, the nicotine concentration in blood increases during the first 10 min and thereafter slowly decreases. In another experiment, two volunteers, one male (SS) and one female (GE), after a nicotine-free period of 12 h started to smoke one cigarette every hour with a smoking time of 5 min. Blood samples were drawn just before starting and then every second hour just before smoking the next cigarette.As can be seen from Fig. 6, nicotine accumulates and its concentration in blood increases during the observation time. 0 1 2 3 4 5 6 7 8 Time/h Fig. 6. Increase of blood concen- tration of nicotine during smoking one cigarette per hour. Subjects are identi- fied by initials on graphs. The authors are grateful to Miss Gun Blomkvist, Application Laboratory, LKB Produkter AB, Stockholm, for valuable help with the mass spectrometric work. Mrs. Kristina Berg and Miss Ingrid Linderot are thanked for skilful technical assistance. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. References Isaac, P, F., and Rand, M. J., Eur. J . Phurmuc., 1969, 8, 269. Turner, D. M., Biochem. J., 1969, 115, 889. Turner, D. M., BY. J . Phurmac., 1971, 41, 621. Schievelbein, H., and Grundke, K., 2. Analyt. Chem., 1968, 237, 1. Isaac, P. F., and Rand, M. J., Nature, Lond., 1972, 236, 308. Burrows, I. E., Corp, P. J., Jackson, G. C., and Page, B. F. J., Analyst, 1971, 96, 81. Hites, R. A., and Biemann, K., Analyd. Chem., 1970, 42, 866. McLafferty, F. W., Analyt. Chem., 1956, 28, 306. Sastry, S. D., in Waller, G. R., Editor, “Biochemical Applications of Mass Spectrometry,” Wiley- Horning, E. C., Horning, M. G., Carroll, D. I., Stillwell, R. N., and Dzidic, I., Life Sci., 1973, Received June 12th, 1973 Amended JuZy 29th, 1974 Accepted August 7th, 1974 Interscience, New York, 1972, p. 657. 13, 1331.

ISSN:0003-2654    

DOI:10.1039/AN9750000099

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

Analyst, February, 1975, Vol. 100, pp. 105-1 10 105 A Colorimetric Method for the Determination of Deoxyri bonucleic Acid in Adipose Tissue P. B. Curtis-Prior, T. Hanley* and N. J. Temple The Wellcome Research Laboratories, Langley Court, Beckenham, Kent, BR3 3BS, and Department of Medicine, Guy's Hospital Medical School, London, SE1 9 R T A method is described for measuring the deoxyribonucleic acid (DNA) content of small samples of adipose tissue or free fat cells. Lipids and acid-soluble nucleotides are first removed by extraction with a cold diethyl ether - ethanol mixture containing 10 per cent. m/V of trichloroacetic acid. DNA is then measured by hydroIysing the nucleoprotein residue in a 5 per cent. solution of trichloroacetic acid at 90 "C for 20 min, followed by treatment with p-nitro- phenylhydrazine and measurement of the hydrazone at 560 nm.Several aspects of the method have been critically examined in order to determine the optimum conditions. The method is satisfactorily reproducible. The metabolic activity of adipose tissue has been the subject of intensive investigation during the last two decades. Probably the best basis for expressing the metabolic activity of this tissue is in terms of its DNA content and a reliable method for measuring the DNA content of adipose tissue is therefore desirable. However, the high lipid content of this tissue, roughly 80 per cent. of the wet mass,l gives rise to handling difficulties owing to emulsion or micelle formation by lipoproteins. A commonly used technique is that described by Schneider2J in which, after a preliminary extraction in cold aqueous acid, the tissue is extracted once with cold ethanol, and then several times with a hot diethyl ether - ethanol mixture.This procedure is not only very laborious but has also been suspected of causing degradation of DNA.4p5 Whereas removal of the last traces of phospholipids is of importance when DNA is measured in terms of its organic phosphorus content,6 small amounts of phospholipid do not interfere with determinations of DNA in terms of its deoxyribose content. A less laborious technique has been developed that is based on this principle. The colorimetric method used is that established by Webb and Levy,' with some minor but important modifications. Effective removal of these lipids is essential. Experimental Reagents De-ionised, glass-distilled water was used in all experiments. Diethyl ether - ethanol mixtures.For the diethyl ether-rich mixture add 25 ml of redistilled ethanol to 75 ml of diethyl ether (reagent grade) and mix thoroughly. For the ethanol-rich mixture, use 75 ml of ethanol and 25 ml of diethyl ether. Extraction mixtures containing trichZoroacetic acid. For the diethyl ether - ethanol solutions of trichloroacetic acid, dissolve 10 g of trichloroacetic acid (analytical-reagent grade) in 100 ml of organic solvent mixture. For the aqueous extraction mixture, dissolve 5 g of trichloroacetic acid in 100 ml of water. p-Nitrophenylhydrazine colour reagent. Dissolve 20 mg of P-nitrophenylhydrazine (general- purpose reagent grade) in 2 ml of ethanol with gentle warming.n-Butyl acetate. D N A standard solutions. Use general-purpose reagent grade. Prepare a stock solution of DNA by adding 50 ml of a 5 per cent. m/V aqueous solution of trichloroacetic acid to 25 mg of salmon-sperm DNA (the purified form of the free acid) in a calibrated flask fitted with an air condenser. Heat on a boiling water bath for 60 min (see Note). Allow to cool to room temperature, add water to replace the small amount lost by evaporation, mix well and store at 4 "C. Prepare, fresh * Present address : Department of Geriatric Medicine, University Hospital of South Manchester. Manchester, M20 8LR.106 CURTIS-PRIOR et aZ. : A COLORIMETRIC METHOD FOR THE Analyst, VoZ. 100 daily, standard solutions containing 20, 10 and 5 pg ml-l of DNA, by diluting the stock solution of DNA with 5 per cent.m/V aqueous trichloroacetic acid solution. By using these standard DNA solutions, the following equation relating DNA concen- tration in micrograms per millilitre (x) and absorbance at 560 nm ( y ) was derived : y = 0.02533~. NOTE- When samples of 2-deoxy-~-ribose dissolved in 5 per cent. m/ V trichloroacetic acid were incubated, as described for DNA standard solutions, for 30 or 60 min, there was no significant difference, as measured by this p-nitrophenylhydrazine method, between the results obtained for the incubated samples and a non-incubated control. Incubation of DNA for 30 rnin in aqueous 5 per cent. m/V trichloroacetic acid was the mode of preparation of standard solutions used in the original method established by Webb and Levy.’ Further, the results given under Recovery of DNA suggest that salmon-sperm DNA is not destroyed by heating at 90 “C for 20 min.It would seem that the hydrolysis separates deoxyribose from nucleoprotein but without degrading the carbohydrate moiety. Apparatus used for the spectrophotometry. A Hitachi - Perkin-Elmer, Model 139, spectrophotometer and a Model 159 recorder were Procedure Removal of Lipid We have applied the method to (A) pieces of adipose tissue weighing up to 300mg and (B) suspensions of isolated fat cells prepared by a modificati~n,~~~ of the technique of Rodbell.10 (A). To a known mass of fresh adipose tissue (50 to 300 mg) in a chilled Potter-Elvehjemll homogeniser tube, add 4 ml of the cold diethyl ether - ethanol mixture (3 + 1) containing 10 per cent.m/V of trichloroacetic acid. Homogenise for 2 rnin and transfer to a 15-ml capacity polypropylene centrifuge tube using a further 6 ml of the trichloroacetic acid solution to rinse the homogeniser plunger and tube. Pool the first homogenate and the trichloroacetic acid rinsings, centrifuge for 10 rnin at 4000 g in chilled centrifuge buckets and discard the supernatant fluid. Re-suspend the precipitate in 5 ml of cold diethyl ether - ethanol (3 + l), without trichloroacetic acid. Mix well, centrifuge as before, aspirate and discard the supernatant fluid. (B). Dispense l-ml portions of a suspension of isolated fat cells (lipid content 20 to 80 mg) into Potter - Elvehjem tubes, add 4 ml of cold ethanol - diethyl ether mixture (3 + 1) con- taining 10 per cent.m/V of trichloroacetic acid, homogenise for about 1 rnin and proceed as before but re-suspend the precipitate in ethanol - diethyl ether (3 + 1). Tissue Hydrolysis To the slightly damp protein pellet, add 3 ml of a 5 per cent. m/V aqueous solution of trichloroacetic acid and several anti-bumping granules. Use a glass rod to fragment and disperse the protein, then incubate at 90 “C in a shaking water-bath for 20 min with frequent and thorough agitation by means of the glass stirring rod. Allow to cool, then centrifuge for 20 min and retain the maximum possible volume of supernatant. Transfer exactly 2 ml of this supernatant in the case of adipose cells (or 1 ml of supernatant plus 1 ml of aqueous 5 per cent.m/V trichloroacetic acid in the case of fresh adipose tissue samples) or 2 ml of DNA standard solution or aqueous 5 per cent. m/V trichloroacetic acid as “blank,” into a glass tube of 12-ml capacity, with a standard ground-glass neck and securing spring hooks. Hydrazone Formation Add 40 pl of freshly prepared colour reagent (20 mg of p-nitrophenylhydrazine dissolved in 2 ml of ethanol) to each sample and mix. Equip each tube with a “cold finger” cooled by tap water and secured with springs. Incubate in a boiling water bath for 20 min, then allow to cool. Removal of Excess of Colour Reagent acetate. and discard the upper phase. Transfer the whole of each sample to 15-ml centrifuge tubes and add 7 ml of n-butyl Centrifuge briefly to separate the phases; aspirate Shake vigorously for 2 min.Repeat the washing procedure twice more.February, 1975 DETERMINATION OF DNA IN ADIPOSE TISSUE 107 Spectrophotometric Measurement of the Colour Developed by Alkali Transfer 1 ml of the washed lower phase into a l-ml glass centrifuge tube, spin briefly and aspirate any visible n-butyl acetate, plus the uppermost layer of the aqueous solution. Transfer 500 p1 of the remaining solution into a glass microcuvette of 2-cm light path, and add 200 pl of fresh, aqueous 4 N sodium hydroxide solution. Stopper the cuvette or cover it with a glass cover-slip, and mix by repeated inversion. Stirring is inadequate for efficient mixing, owing to the high specific gravity of the alkali relative to the tissue extract. Read the absorbance of the magenta-coloured solution at 560 nm, exactly 3 min after the addition of alkali.Refer the absorbance of unknown solutions to the calibration equation (see above) to determine their DNA concentration. Investigation of Stages of the Procedure Stage Aspects investigated Hydrolysis of DNA Removal of excess of p-nitrophenyl- Time course of colour development and Whole procedure Reproducibility. Recovery of DNA Optimum temperature, time and acid concentration Number of n-butyl acetate washes Optimum time for reading colorimetric change. Influence hydrazine recession of pH Hot Acid Hydrolysis of DNA S~hneider,~J employing the diphenylamine reaction,12 determined the optimum conditions for hot acid hydrolysis of liver nucleoproteins to be 90 "C for 15 min with 5 per cent. m/V trichloroacetic acid, but Webb and Levy7 suggested a longer incubation.We performed three experiments (all using a mass of macerated rat epididymal adipose tissue as starting material) in which one variable was changed at a time. In the first series, incubation was effected for 20 min using 5 per cent. m/V trichloroacetic acid at five temperatures between 80 and 100 "C. In the second experiment, the temperature was maintained at 90 "C and the trichloroacetic acid concentration at 5 per cent. m/V, but the time of incubation was varied from 5 to 60 min. Finally, we examined the influence of different trichloroacetic acid concentrations. The results in Table I led us to adopt 90 "C as the temperature of incubation, 20 min as the time and 5 per cent. m/V as the concentration of trichloroacetic acid for our standard procedure.TABLE I CONDITIONS FOR HOT ACID HYDROLYSIS OF RAT ADIPOSE TISSUE NUCLEOPROTEIN Influence of incubation temperature (using 5 per cent. m/V trichloroacetic acid for 20 min)- Temperaturel'C . . .. 30 85 90 95 100 Adipose mass corrected DNA concentration in sample (mean f standard error of the mean)*/pg ml-1 . . 22-0 &- 0.9 37-9 f 0-6 39-5 & 1.1 41-6 f 1.2 35.9 f 3.0 (3) ( 5 ) (9) (4) (4) Influence of the time of incubation (at 90 "C using 5 per cent. m/V trichloroacetic acid)- Timelmin .. .. .. 5 10 20 30 60 Adipose mass corrected DNA concentration in sample (mean f standard error of the mean)*/pgml-l . 17.9 f 1 28.4 f 1-7 39.5 f 1.1 42-4 f 0.9 40.5 f 1.5 (3) (5) (9) (4) (4) Influence of trichloroacetic acid concentration (at 90 "Cfor 20 main)- 40 Trichloroacetic acid concen- Adipose mass corrected DNA concentration in sample (mean f standard error of tration, per cent.m/V . . 2-5 5-0 7.5 10 20 the mean)*/pg ml-l . . 34.9 f 0.3 39-5 f 1-1 40.8 f 0.4 41.1 f 1-1 39.3 f 1.0 33.7 f 1.4 (3) (9) (4) (4) (4) (4) Figures in parentheses indicate number of observations.108 Analyst, Vol. 100 Removal of Excess of p-Nitrophenylhydrazine Colour Reagent After hydrazone formation with the deoxyribose is complete, any excess of P-nitrophenyl- hydrazine must be removed, or it will itself form a coloured complex on addition of alkali. In the original technique by Webb and Levy7 a single wash with n-butyl acetate is recom- mended, but we found this to be inadequate if consistently low blank values were to be obtained (an important factor if only small amounts of DNA are present in the sample).The absorbance of blanks treated with a single 7-ml wash of n-butyl acetate was 0.135 -+ 0.003 (mean & standard error of the mean, number of readings = 8), whereas the value for the thrice-washed sample was 0.018 & 0.0004. In the preparation of standard curves for DNA, erratic and non-linear standard calibration lines were obtained when only one wash with n-butyl acetate was used. More prolonged shaking or an increase of the single-wash volume to 10ml produced little improvement. CURTIS-PRIOR et al. : A COLORIMETRIC METHOD FOR THE Development of Colour by Alkali The absorbance of replicate samples of a standard DNA solution was measured 3 min after addition of alkali of various concentrations, and the rate-of fall in absorbance was also recorded.Volumes of 200 ml of 1, 2, 3, 4 or 8 N sodium hydroxide solution were added to replicate samples of a DNA solution. There was an almost immediate development of maximal colour that was unrelated to the alkali concentration provided that the pH was 11 or higher.’ The fall in absorbance was almost linear during the period of examination (about 20 min after the addition of alkali) at approximately 1 per cent. min-l. The rate of colour loss was not related to the concentration of alkali present. Reproducibility of the Complete Procedure Ten replicate samples of macerated rabbit perineal adipose tissue, each weighing approxi- mately 300 mg, were found to have a DNA content of 94.2 -J-- 2.3 pg g-l of fresh tissue; the coefficient of variation was 5.4 per cent. and the standard error was 5.1 per cent.In another experiment, a homogenate of rat epididymal adipose tissue provided three batches of sample weighing 100, 200 or 300 mg. The results shown in Fig. 1 indicate good reproducibility at all three levels of sample size. 0 100 200 300 Mass of adipose tissue used/mg Fig. 1. Relationship between DNA re- covered and mass of adipose tissue sample. Each point represents the mean of seven replicate observations f the standard error of the mean. Recovery of DNA In a final test of the method a “recovery” experiment was carried out on samples prepared from a large mass of adipose tissue. The DNA concentration, measured for five samples of about 600 mg, was 362.2 pg g-l.Four other samples of mean wet mass 579 mg and calculatedFebruary, 1975 DETERMINATION OF DNA IN ADIPOSE TISSUE 109 total DNA content 208-4 pg were de-fatted, and to the protein pellets were added 317 pug of purified salmon-sperm DNA. The total DNA content was then measured by continuing the analytical procedure from this stage. The expected mean total DNA content of the sample and added DNA was 525.4pg (208.4 + 317 pg), and the mean total as found colorirnetrically was 520-7 pg. The mean “recovery” was thus 98.8 per cent. (standard error of the mean, 1-6 per cent.). This result indicates a negligible loss, and acceptable precision, from the protein-pellet stage of the analysis onwards. Discussion If DNA is determined by measuring the organic phosphorus content of tissues it is necessary to remove thoroughly lipids that contain phosphorus, and this is accomplished by using the elaborate and time-consuming Schneider procedure.I t is not necessary to remove lipids when DNA is determined in terms of its deoxyribose content. Lipid extraction is, however, still necessary with adipose tissue for the different purpose of preventing the emulsion and micelle formation that occur in homogenised adipose tissue. The much simpler extraction described here seems adequate for lipid removal; no difficulty with emulsion formation was encountered after the extraction. After the stage of colour development, any excess of p-nitrophenylhydrazine must be removed, as this will itself increase absorbance at 560 nm, in the presence of alkali.Webb and Levy recommended a single wash with n-butyl acetate, but however well done, this single wash gave blank values of about 0.135 absorbance unit and this was unacceptably high when absorbances of from 0.01 to 0.4 were recorded from DNA itself. It was found that if the blank value were to be reduced to about 0.01, three successive washes with 7 ml of n-butyl acetate were required. This added to the labour and time of the analysis, but is worthwhile because of the increased sensitivity achieved. Although the n-butyl acetate wash was originally designed to remove excess of colour reagent and trichloro- acetic acid was said to be necessary to effect quantitative extraction, it is of interest that Young13 reported the use of n-butyl acetate as a solvent for the extraction of trichloroacetic acid.As one extraction with an equal volume of n-butyl acetate gives 80 per cent. removal of trichloroacetic acid, the amount of trichloroacetic acid in solution during our second and third washes with n-butyl acetate must assume negligible proportions. A systematic investi- gation of the use of trichloroacetic acid saturated n-butyl acetate might, therefore, prove of value in reducing the number of washes. It is of interest that in another modification of this technique14 excess of colour reagent is removed by reaction with acetyl acetone and the chromophore extracted into butanol. The colorimetric test used by Schneider for the determination of DNA in the final tissue extract was either the carbazole16 reaction or the diphenylamine reaction of Dische.12 The carbazole reaction is not specific for carbohydratesl6 and the unmodified diphenylamine reaction is far less sensitive than the 9-nitrophenylhydrazine reaction.Further, the protocol described has been used to measure DNA in 20-50 samples during the working day, whereas the modified diphenylamine reactions of Burton17 or Gangulil* involve a 17-h incubation period. However, the recently published method of Abraham et aZ.,lg which is an improvement of the technique of Giles and Myer’s modification20 of the diphenylamine reaction, would appear to be a very convenient method for the determination of low concentrations of standard DNA solutions although its validity for biological samples remains unproved.We acknowledge with thanks the encouragement of Professor W. J. H. Butterfield, the valuable statistical help of Mr. F. W. Harpley and Mr. D. A. Field and the technical assistance of Mr. T. Kelly. We are grateful to Dr. R. H. Nimmo-Smith for his gift of the purified salmon-sperm DNA used as standards. References 1. 2. 3. 4. 5. Boshart, C. R., Will, L., and Pirre, A., Proc. SOC. Ex$. B i d . Med., 1962, 110, 661. Schneider, W. C., J. Biol. Chem., 1945, 161, 293. Schneider, W. C., Methods Enzyrn., 1957, 3, 680. Munro, H. N., and Fleck, A., AnaZyst, 1966, 91, 78. Hutchinson, W. C., Downie, E. D., and Munro, H. N., Biochim. Biophys. Acta, 1962, 55, 561.110 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. CURTIS-PRIOR, HANLEY AND TEMPLE Schmidt, G., and Thannhauser, S. J.. J . B i d . Chem., 1945, 161, 83. Webb, J. M., and Levy, H. B., J . Biol. Chem., 1955, 213, 107. Curtis-Prior, P. B., Guy’s Hosp. Rep., 1972, 121, 167. Curtis-Prior, P. B., and Hanley, T., Acta Endocr. Copenh., 1973, 14, 409. Rodbell, M., J . B i d . Chem., 1964, 239, 375. Potter, V. R., and Elvehjem, C. A., J . B i d . Chem., 1936, 114, 495. Dische, Z., Mikrochemie, 1930, 8, 9. Young, D. A. B., Biochim. Biophys. Acta, 1964, 90, 178. Martin, R. F., Donohue, D. C., and Finch, L. R., Analyt. Biochem., 1972, 47, 562. Gurin, S., and Hood, D. B., J . Bid. Chem., 1941, 139, 775. Hutchison, W. C., and Munro, H. N., Analyst, 1961, 86, 768. Burton, K., Biochem. J., 1956, 62, 315. Ganguli, P. K., Revue Can. Biol., 1970, 29, 339. Abraham, G. N., Scaletta, C., and Vaughan, J. H., Analyt. Biochem., 1972, 49, 547. Giles, K. W., and Myers, A., Nature, Lond., 1965, 206, 93. Received APriZ llth, 1974 Accepted Sefitember 17th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9750000105

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

A Malyst, February, 1975, Vol. 100, ~~.111-123 111 The Automatic Determination of Iodide or Iodate in Solution by Catalytic Spectrophotometry, with Particular Reference to River Water Victor W. Truesdale and Peter J. Smith Institute of Hydrology, Maclean Building, Crowmarsh Gifford, Wallingford, Oxfordshire A catalytic procedure, in which a Technicon AutoAnalyzer(1) is used, for determining iodide or iodate (50 pgl-l) added to river water (during river- flow gauging operations by the dilution method) is described. The maximum error of the analytical procedure, as determined by replicate analysis of six sets of eleven samples, is k0-36 per cent. in the range 20-100 pgl-,l. The setting up of the procedure is described in detail, according to Mark s (1973) recommendations. The shape of the most appropriate calibration graph is considered.Also, the effects of changes in temperature, reaction period and spectrophotometric variables are discussed. This information could be used directly in the development of methods for determining iodine in other types of sample, e.g., protein-bound iodine in blood or urine and iodate or iodide in sea water, etc. Further, it could also be used in developing other automatic procedures that rely upon catalytic reactions that follow a pseudo- first-order behaviour. Mark,l in his definitive paper, has reviewed the development and publication of new methods in kinetic analysis. He states that many of the papers on the development of fundamental methods published recently do not fully explain the applicability and limitations of the individual methods, and therefore that it is not clear to other analytical chemists how they could modify or adapt these methods to their particular system or analytical problem.As a first step in overcoming this defect, Mark1 has listed specific requirements that are unique for the development and description of kinetics-based analytical methods. The difficulties that we encountered in setting up an automatic catalytic procedure for the determination of approximately 50 pg 1-1 of iodide- or iodate-iodine present in river water substantiate Mark’s1 viewpoint. The procedure adopted was to utilise the catalytic property of iodide in the reaction between ammonium cerium(1V) sulphate and arsenious acid. It was required for river-flow gauging exercises conducted by the dilution method, and was to be adapted to a Technicon AutoAnalyzer. An examination of available Auto- Analyzer procedures for this concentration of iodine2-10 suggested that they had been set up empirically.In particular, we were surprised to find that although several procedures for determining almost the same concentration of iodine had been proposed, there was little correlation between the conditions used in these procedures. For example, two procedures for the determination of the same iodine concentration (50 pg 1-l) ‘in sea water9J0 have been proposed; whereas in one9 a x10 electronic range expander is used on the output of the colorimeter, in the otherlo an expander is not used. No reasons have been given for these and other procedural differences and it does not seem that use can be made of the differences between the types of sample matrix (urine, blood or sea water) to account for them.As we had already found that there is no difficulty in assembling a crude procedure, we became concerned that we should not adopt, by default, a procedure unsuited to our needs. The literature dealing with catalytic spectrophotometry did little to overcome our difficulties as we were unable to locate any source that discusses the problems of establishing a new procedure. Instead, it was found that only the general aspects of the subjectll-17 or the specific details2-9 of procedures were covered. As our experience had shown, therefore, that there is a considerable gap in knowledge of this subject, we studied the problem further. In presenting our results we have adhered to Mark’s recommendations.Although the study was directed specifically at analysis for iodine, and therefore many of the results will apply only to this problem, this paper contains information of more general interest. In particular, families of standard calibration graphs are presented. Catalytic procedures for any112 TRUESDALE AND SMITH: AUTOMATIC DETERMINATION OF IODIDE OR Afialyst, VoZ. 100 chemical species, which satisfy the assumptions inherent in the development of the cali- bration graphs for use in the iodine determination, will require the use of one of these standard graphs. Experimental Apparatus Automatic analysis was carried out with a Technicon AutoAnalyzer(1) instrument.Addi- tional optical measurements were made by using a Hilger-Watts Uvichem spectrophotometer fitted with a constant-temperature housing. For kinetic studies the output of the spectro- photometer was displayed on a Smiths' Servoscribe recorder. A Dec PDP-8 computer was used for the calculations. AnalaR reagents were used whenever possible, although only laboratory-reagent grade (Hopkin & Williams) ammonium cerium( IV) sulphate was available. Distilled water was prepared from tap water by means of a cast-iron Manesty still. Results and Discussion The Response of the Procedure to Iodine The principles involved in operating a catalytic method of analysis are different from those encountered in more conventional trace-element colorimetric methods. Usually, the optimum conditions for analysis are discovered by using a stepwise procedure in which all variables except one, e.g., the concentration of one reagent or a time of reaction, are kept constant. The effect upon the response of altering the isolated variable is then studied. This approach is valid only when the isolated variable is independent of the others.As will be seen below, however, with the analysis for iodine by the catalytic method the variables are dependent and the stepwise approach inappropriate. Instead, a mathematical approach has been found to be more suitable because it deals with all. the variables simultaneously. The approach used here is to derive the equation that describes the calibration graph obtained under any set of conditions. The calibration graph is obtained by considering the factors that control the residual cerium(1V) concentration in the reaction mixtures derived from a series of iodine standards.Rodriguez and Parduel* stated that a simplified rate expression for the reaction, which probably applies over most practical analytical conditions, is of the form [I1 d[Ce] dt - ~,[AsIII]~ + k6 [As1111 + k , [Ce] [Ce][ As1111 (k,[AsIII] + kp) Rate = - - - where K, to k, are constant under conditions of constant acidity, temperature and sodium chloride concentration and k, > k8 > k,. In this work the arsenious acid concentration, [AsIII], is maintained at an appreciably higher level than the cerium(1V) concentration, [Ce], (Ostwald isolation conditionlQ) , the reaction mixture contains a significant and constant amount of sodium chloride20 and the acid concentration is constant.Given these conditions the reaction becomes pseudo-first-order10~18~21-27 with respect to cerium(1V) and iodine concentrations, as k,[Ce] can be neglected. Thus we can write where [AS"'] (ka[As] + k'"p) k, [ASIII]~ + k6 [AsfII] ' [I3 k = [I1 ka[A~III] + kp k,[AsIII] + ks - = 127 x lo6 w [I] (relative atomic mass of I is 127) and o is constant provided the arsenious acid concentration is constant and the above conditions hold. The constant w can be considered to be the change in the rate constant (per minute) resulting from an iodine concentration change of 1 pg l-l, under the given conditions. It therefore indicates the sensitivity of the reaction to iodine. The sulphuric acid and sodium chloride concentrations recommended here are considered to approach thoseFebruary, 1975 IODATE I N SOLUTION BY CATALYTIC SPECTROPHOTOMETRY 113 which maximise w.The effect upon w of changing the arsenious acid concentration is negligible, as is demonstrated below. For our purposes, it has not been necessary to seek conditions in which the reaction is more sensitive to iodine, as some inadequacy in w can be compensated for by adjustment of the reaction time used in the analytical procedure. However, it is recognised that adjustment of w would be necessary if one required appreciably more sensitivity than that already obtained. In this procedure the difference in iodine content of any two samples is detected as the difference between the residual cerium( IV) concentrations (C, and C,) of their respective reaction mixtures, after a set reaction period (t).Integration of the above rate equation gives [Ce], = [Ce], e--kt which can be used to describe C, and C, in terms of the initial cerium(1V) concentration (C,), t and the apparent velocity constants (k, and k,). Moreover, the equations thus formed can be expressed in terms of corresponding absorbances (A,, A , and Ao), provided that Beer's law is obeyed at the wavelength of light chosen. (The relevance of the wavelength is discussed below.) Thus, .. .' (1) .. .. * * (2) AA = A, - A, = A , [exp(-k,t) - exp(-kK,t)] .. .. (3) .. * ' (4) A , = A, exp(-k,f) .. .. A , = A, exp(-k2t) .. Hence the difference (AA) between the absorbances of the two reaction mixtures is = A , (exp[-(k,-k,)t] - 1) = A, [exp(-Akt) - 13 .... .. where Ak = k, - k,. Equation (4) can be used immediately when the spectrophotometer output is recorded as absorbance. However, in many instances, and in particular with the type of Technicon AutoAnalyzer used here, the output is recorded as transmission (T) and the equation has to be manipulated further before it is of direct use. Thus, AT = T , - T - 1 0 - A ~ - ~ O - A I = ~ O - A I (lO-(As-Al) - 1) = ~ O - A I (10-AA - 1) 1 - Changing from base 10 to base e, e.g., 10-A1 = e--Al'*lO, and substituting for AA from (4), AT = e+ (exp(Z[l - exp(1 - Akt)] 1 - 1) . . .. * * (6) where Z = A , In 10 = 2.303 A,. Moreover, as the difference between the apparent velocity constants (Ak) is proportional to the difference in iodine concentrations A I (= I, - I,), Ak = wA1, where w is the coefficient of proportionality already mentioned.Substituting o A I for Ak in equation (5) gives AT = e-+ (exp[Z (1 - e-*'ut)] - 1) This equation describes the calibration graph for iodine that is obtained under any given set of conditions determined by the three variables 2, o and t. With river-flow gauging studies the background concentration is small compared with the concentration of added iodine tracer. Therefore, the lowest iodine concentration (I,) can be considered to be zero. Therefore, putting I , = 0, A I = I, = I, the equation above can be rewritten as AT = e-% (exp[Z(l - e-*Iwt)] - 1) . . .. . . (6 The general curve generated by equation (6) is j-shaped; it passes through the origin and approaches the asymptote (1 - e-.).In practice, this means that the base-line (zero iodine) on the recorder output is fixed at a transmission value corresponding to an absorbance A,, and samples containing iodine produce peaks with higher transmission values. A series of samples with progressively higher concentrations of iodine would produce peaks that converge in the region near T = 1 (100 per cent. transmission). As implied above the curve is inflected114 TRUESDALE AND SMITH: AUTOMATIC DETERMINATION OF IODIDE OR Analyst, VoZ. IOU and the second differential with respect to iodine is therefore equal to zero at some point. This occurs at a constant value of (1 - e-l) away from the asymptote. It is useful to divide the calibration graphs into the following three cases, which depend on where the point of inflection is placed.0.80 0.60 0.40 0.20 0 1. An injection point between the extreme limits of the chosen concentration range Four families of inflected calibration graphs are shown in Fig. 1 (a-d). Any one curve from any family is obtained from equation (6) by fixing A, and the compound variable at, but taking various concentrations of iodine between 0 and 100 pg 1-l. The curves of any one family &re generated by repeating this process at discrete values of wt, but with the same value of A,. Of course, both ot and A, vary continuously and an infinite number of curves are possible. Other curves can be obtained by interpolation of those described. Although the choice of the value of ot is arbitrary, the range of useful values is fixed by the choice of both A , and the iodine concentration interval. This is because AT is allowed _-.-.-.-.-.- - - - - I l l to vary only betwee; 0 and 1. 0 20 40 60 80 100 Iodine concentration/pg I" 1 .oo 0.80 I- Q 0.60 $ 0 8 0.40 CT 0.20 0 I 0 20 40 60 80 100 Iodine concentration/pg I-' '*O0 1 (b) I 1 .oo 0.80 h 0.60 $ 0 8 0.40 a 0.20 0 0 20 40 60 80 100 Iodine concentration/pg I-' ot = 0.001 - / 1 1 I I I I I 1 I I 0 20 40 60 80 100 Iodine concentration/pg I-' Fig. 1. Families of inflected curves. A,: (a) 2.0; (b) 1.0; (c) 0.7; (d) 0.5. ot (1 p g - l ) : (a) 0.015, 0.017, 0.019, 0.022, 0.025, 0.031, 0.038, 0.051, 0.076, 0.152; (b) 0.010, 0.012, 0.014, 0.017, 0.021, 0.028, 0.042, 0-083; (c) 0.005, 0.006, 0.008, 0.010, 0.012, 0.016, 0.024, 0.048; (d) 0.001, 0.003, 0.005, 0.00'7, 0.014.. -. - are asymptotes, - - - - give the points of inflection. * The curve selected and obtained practically in this study. Compare it with its logarithmic counterpart in Fig. 3. 0.008,February , 1975 IODATE I N SOLUTION BY CATALYTIC SPECTROPHOTOMETRY 115 In the problem discussed here the iodine concentration interval is between 0 and 100 pg 1-l. In other procedures it might be desirable to adopt a different concentration interval while still retaining the same shape of calibration graph. If this is so, the range of wt values must be changed in inverse proportion to the change in the iodine concentration interval. Thus, for example, in order to adapt the curves in Figs. 1 and 2 to cover the range 0-50 mg l-l, i.e., 500 times the concentration interval, it is necessary to reduce the ot values supplied in these figures 500 times.1.00 0.80 ' 0.60 - i 0 Q M 0-40 - fIT 0.20 - 0 - 0 20 40 60 80 100 Iodine concentration/pg I-' Fig. 2. A family of curves with points of inflection a t the origin. Formed when A, = (In 10)-l = 0.43. wt (1 p g - l ) : 0.010, 0*020, 0.030, 0.040, 0.050, 0.070, 0.100, 0.200. The results given in Figs. 1 and 2 show that when high A, values are chosen (say, greater than 2-0) the curves traverse almost the whole transmission scale. With lower A, values a similar distribution of curves is obtained, but, of course, the maximum response is less. The inflection also becomes less obvious as A, is decreased. 2. Injection at the origin (decreasing gradient) of inflection is placed at the origin, i.e., A family of curves is formed in a similar way to that given above except that the point which is satisfied when 2 = 1, i.e., A , = (In lO)-l. A family of these curves for values of wt between 0-20 and 0.01 is shown in Fig.2. 3. Injection at the maximum anticipated concentration (increasing gradient) concentration, which in this instance is 100 pg l-l, i.e., These curves are formed when the point of inflection is placed at the maximum anticipated They can be seen in Fig. 1 where they are the lowest members of each of the four families, la-ld. Because the point of inflection is constrained in the manner described above the maximum possible response that can be obtained is (e-l - e-a). The choice of a curve, from one of the cases, will be determined by its usefulness for a particular procedure.Clearly, the inflected curves are inherently more difficult to draw116 TRUESDALE AND SMITH: AUTOMATIC DETERMINATION OF IODIDE OR Andyst, VoZ. 100 than either of the single-sweep curves. The importance of this factor, and others concerning the distribution of the gradient, will have to be considered when choosing a curve. With some of the curves discussed above almost a full-scale deflection is obtained for 100 pg 1-l. Where calibration graphs have a maximum response close to 06T, range expansion facilities can be used to increase their usefulness. Thus almost twice the sensitivity could be obtained from four such curves described in case 1. Similarly, the curve in case 2 which offers this same over-all response could be expanded to yield an almost linear calibration graph.A similar argument can be advanced for the additional settings of x 4 and x 10 expansion that are fitted to the Technicon AutoAnalyzer, although, of course, the least amount of expansion should be incorporated so as to minimise interferences from electronic and other sources. There can be no doubt, therefore, that a large number of variations of the catalytic procedure can be assembled, which explains how, as mentioned earlier, two dissimilar pro- cedures9Jo for the same concentration of iodine in sea water could have been proposed. Further, it emphasises the need for critical appraisal of “new” procedures that might be proposed in the future. Choice of Calibration Graph The calibration graph that is eventually chosen will depend on the purposes to which the analytical results are put.In our work the degree to which we wished to resolve our measure- ments of river flow-rate has predominated. An iodine concentrate (I,) is injected at a constant rate (4) into the river, which flows at an unknown rate (Q) and contains a small background iodine concentration (Ia). After the tracer has become thoroughly mixed, the river water is sampled and the diluted concentration (Ib) is determined. The flow-rate of the river is calculated from the equation: There is no unique iodine concentration that corresponds to a given flow. Instead, at each gauging the field worker arranges for the final iodine concentration to be as close as possible to 60 pg 1-1 (an environmentally acceptable level).To avoid inconsistencies in the flow measurements, therefore, it seems worthwhile to make the relative errors of all such measure- ments constant. Thus, although the values for the flow will range over many orders of magnitude, the accuracy of any one can be calculated simply as a percentage of its value. This constraint on the relative error is transmitted to the iodine concentration and thence to the change in response yielded by the analytical method. Let us consider, therefore, the limiting case in which the least discernible change in response (7) is produced by a small change in iodine concentration from I to I + E. The resolution of the analytical procedure is described by the ratio of T to E . As this can be taken as the gradient of the calibration graph in the small interval ( I , I + E ) , we can write .... .. r d(AT) E C U -- -- (7) But, as the relative error in iodine concentration is constant throughout the range considered, we can also write where a is the constant relative error. Combining equations (7) and (8) and integrating we have 7 01 .. .. .. (9) AT =-ln,8I .. .. where ,8 is a constant of integration. Equation (9) therefore describes the form of calibration graph best suited for use in the river gauging work when only resolution is considered. The logarithmic curve and the double exponential curve derived earlier cannot be coincident across the whole range of iodine concentrations. Unlike the double exponential, the logarith-February, 1975 IODATE IN SOLUTION BY CATALYTIC SPECTROPHOTOMETRY 117 mic curve does not pass through the origin; it intersects the abscissa at 1/p.Also, whereas the double exponential curve approaches an asymptote, the logarithmic curve tends to infinity at high concentrations. By defining 1//3 as 10 pg l-l, and equating the response at 100 pg 1-1 with the value of the asymptote (1 - e-.) of the double exponential curve, we have generated logarithmic curves scaled appropriately to the unexpanded range. (Although the response at 100 pg 1-1 will obviously be slightly less than the asymptote value, this dis- crepancy has been ignored for the sake of simplicity. Also, l//l has been defined in an arbitrary way and another value between 0 and 10 pg l-l would, perhaps, be equally satisfactory.) A similar exercise in which the same /3 value was used in conjunction with a response of 0*5T, for 100pgl-l, yielded logarithmic curves suitable for use when a x2 range expansion is used.Initial attempts at fitting the two sets of curves were performed visually. By this means it was possible to discard immediately most of the calibration graphs, and therefore to narrow the choice to only a few. More rigorous fitting of the curves using mathematical methods was felt to be unnecessary as the visual method had produced so few alternatives. The logarith- mic curves for six values are shown in Fig. 3. Comparison of these with the whole range of double exponential curves showed that a close agreement is obtained where A , and wt are about 1.00 and 0.0417, respectively. For a x2 range expansion close agreement was found when A , and wt are equal to 0.4343 and 0.020, respectively.The logarithmic and double exponential curves seemed to fit better in the latter than in the previous case. As these two cases seemed to offer practicable alternatives, the study was not extended to higher-range expansion settings; the choice of conditions was, therefore, between these two. Iodine concentration/pg I-' Fig. 3. A family of logarithmic curves. AT = ?' In PI, where- = 10 pg 1-l; 2 =A,ln 10. For curves with unbroken line, ATloo pg 1-1 = (1 - e-z) A, = 2-0, 1.0, 0.7, 0.5, 0.43. For broken line ATlw pg 1-1 = 0.50. 1 B It is felt that in the context of our work, the importance of interference due to background coloration of river water is great enough to cancel the advantages gained by using the closer- fitting curve.Experience has shown that this coloration introduces a problem when scale expansion is used in conjunction with a base-line (A,) of less than approximately 0.5 unit. Under these circumstances, the base-line position on the recorder chart can change sub- stantially with each new batch of samples; it can go off the chart when densely coloured samples are introduced into the reaction mixture. The adjustments that are necessary to bring the base-line back on to the chart are time consuming. In contrast, when range expansion is not used and A , is larger, say, 1.0, the changes in the position of the base-line118 TRUESDALE AND SMITH: AUTOMATIC DETERMINATION OF IODIDE OR Analyst, VoZ.100 caused by variation in background coloration are usually very small because the increase in absorbance arising from extraneous coloration is proportionately smaller. Setting Up the Procedure The procedure for use in this work was established as follows. In essence it was necessary to convert the desired A , and wt values of 1.00 and 4.17 x respectively, into instru- mental settings and reagent concentrations. A reaction period of 20 min was chosen in order that the wt value should be at the maximum practicable, giving a value for w of 2.09 x 10-3 min-1 pg-l 1. Measurements of velocity constants of the reaction, with either distilled water or standard iodide solution incorporated in the reaction mixture, were then made at several temperature settings and empirical i;, values established. The required temperature setting for the AutoAnalyzer was then selected by comparing the desired w value with those of 6.In order to complete the sequence, the concentration of the cerium(1V) reagent was adjusted so that when iodine-free water was sampled by the AutoAnalyzer the required base-line setting (A,) was obtained. The apparent velocity constants for the reaction were measured using a mixture of 5 ml of sample (46 pg 1-1 of iodide-iodine), 20 ml of arsenious acid and 10 ml of cerium(1V) reagent. The reagents were prepared in the way described below except that the cerium(1V) concen- tration was slightly higher. The combination of reagents and sample was taken from a pro- cedure described earlier.l* The i;, values were plotted against temperature ("C) and a suitable curve drawn through the points.A straight line with gradient 4 x min-1 pg-11 "C-1 and intercept 2.5 x lo4 min-l p g l l was obtained for the range investigated. The results showed that the desired w value could be obtained at approximately 59 "C with a reaction mixture of the composition mentioned above. Tests performed with the AutoAnalyzer showed, however, that at this temperature the concentration of the cerium(1V) reagent had to be extremely high in order to provide the required base-line (A,) value. As there was doubt that the Ostwald isolation condition would hold under these circum- stances, alternative settings that relied upon a lower temperature were sought. Accordingly, the effect of doubling the flow through the sample delivery tube of the AutoAnalyzer was investigated.Although this led to a larger total flow in the AutoAnalyzer system, and therefore to a shorter reaction time, the effect was small and was neglected. (If the effect had been larger, however, the original total flow could have been re-established by redu- cing the flow from the cerium(1V) reagent pump tube by the required amount.) A second graph that describes the variation of i;, with temperature, in the modified reaction mixture, was obtained by multiplying the original 6 values by the factor by which the sample contribution had increased, namely 1.75. This revised graph showed that the required w value could be obtained at approximately 35 "C. When the concentration of the cerium(1V) reagent had been adjusted so that the correct base-line (A,) was obtained at this temperature, the anticipated calibration graph was yielded. Temperature Control Catalytic methods of analysis are inherently more sensitive to changes in temperature than are more conventional equilibrium methods. As a result, greater care has usually to be taken in selecting the type of constant-temperature bath used.It would be desirable to have some idea of the sensitivity of a new procedure to temperature fluctuation before the appropriate equipment is assembled. The relevant information for this catalytic procedure can be obtained by partial differentiation of equation (6) with respect to W . This gives the rate at which the response in transmission changes with w As, in the procedure described here the rate of change of w with temperature (8) can be considered to be a constant, 7.02 x min-l p g l l "C-l, we can writeFebruary, 1975 IODATE IN SOLUTION BY CATALYTIC SPECTROPHOTOMETRY 119 The change in transmission caused by a shift in temperature of 1 "C has been calculated by substituting the desired values of o, 2, t and I into equation (10) (Table I).The results show that the sensitivity of the procedure to temperature fluctuation is at a maximum between 40 and 50 pg l-l, when 1 "C produces a change of 1.6 per cent. in transmission. The magnitude of this figure suggests that temperature control to within k0.05 "C will be necessary to keep changes in the percentage transmission at less than 0.15, which is about the degree of certainty that one can put on the pen trace of the AutoAnalyzer.TABLE I RATE OF CHANGE OF RESPONSE (PERCENTAGE TRANSMISSION) AT VARIOUS IODINE CONCENTRATIONS WITH TEMPERATURE AND PERIOD OF REACTION A,, w and t are equal to 1.00, 2-09 x 10-8 min-* pg-l 1 and 20 min, respectively. Iodine concentration/pgl-l 10 20 30 40 50 60 70 80 90 100 Temperaturecoefficient . . 0.5 1.0 1.5 1.6 1.6 1.4 1.2 1-0 0.7 0-6 Reaction period coefficient . . 0.7 1.6 2.2 2.5 2.4 2.1 1-8 1.4 1-1 0.9 Measurements of the temperature of the oil in a standard Technicon constant-temperature bath were made during this study. A quartz thermometer was positioned near to both the bottom and surface of the oil and close to the glass coil. The pattern of temperature fluctuation was similar at both positions. A rapid rise in temperature of 0.4 "C (in 3 min) was followed by a slower fall (in 9-5 min).The mean temperature (approximately 33 "C) at the two positions differed by 0.4 "C. These relatively large fluctuations seem to be due to poor circulation of the oil in the bath. The recorder pen trace for continuous aspiration of 0,50,70 and 90 pg 1-1 standard iodine solutions changed sympathetically with temperature by 0.2, 0.6, 0.4 and 0.2 per cent. transmission, respectively, conforming closely to the pre- dicted values; as will be shown later, they are sufficiently large to limit the over-all precision of the procedure. Therefore, when precision, additional to that reported here, is required, a more efficient heating bath should be used. Accuracy of Reaction Period It is not always practicable to assemble an AutoAnalyzer manifold so that the chosen reaction period is obtained.The glass coil of the heating bath is of fixed length and it may not be possible to match it with the correct total flow through the system. As a result the empirically derived calibration graph will be different from the desired graph. A second source of error in the reaction period is inevitably introducedwith theuseof a peristaltic pump. Pulsed flow in the AutoAnalyzer might in some circumstances be of sufficient magnitude to change the reaction period significantly. Presumably, this would arise if a very short reaction period were utilised. Some idea of how important these factors are to a given manifold design can be obtained mathematically, before the manifold is assembled.Thus, partial differentiation of equation (6) with respect to time gives Values of - when 2, w , t and I have the desired values are shown in Table I, in which the results given show that the response of the procedure is sensitive to errors in the reaction period that are likely to be encountered when assembling the manifold. Care must therefore be exercised if the empirically derived calibration graph is to be close to its theoretical counterpart. at Spectrophotometric Considerations Although absorbance has been used in deriving the equation for the calibration graphs, neither an appropriate cell-path length nor a wavelength at which to measure absorbance was stipulated. In fact, these factors are less crucial to the catalytic procedure than they are to more conventional colorimetric methods.Although an absorbance term (2) is involved120 TRUESDALE AND SMITH: AUTOMATIC DETERMINATION OF IODIDE OR Analyst, VoZ. 100 in equation (6), the only major constraint on it is that Beer’s law should be obeyed at the wavelength at which it is measured. Therefore, if the wavelength is changed, but 2 kept constant [and no assumption inherent in equation (6) thereby disobeyed], the response will remain unaltered. No such comparable flexibility exists in conventional colorimetry. Such a change can be effected simply. Thus, for example, if the wavelength were decreased, the concentration of the cerium(1V) reagent would have to be increased to compensate for the apparent fall in the value of 2. The extent to which the wavelength can be increased is limited, however, because the use of large concentrations of cerium(1V) species leads to consumption of an appreciable proportion of the arsenious acid.Eventually, therefore, the assumption that the arsenious acid concentration remains constant throughout the reaction cannot be sustained and the mathematical analysis breaks down. An investigation of how, relative to 410nm, the extinction coefficient of the cerium(1V) solution changes with wavelength was made, from which (Table 11) the proportion by which the cerium(1V) concentration must be changed to compensate for a given wavelength change from 410nm has been calculated. The study was not extended beyond 455nm because, then, relatively large cerium( IV) concentrations are obviously required.Preliminary investi- gations of the effect of changing the concentration of arsenious acid in the reaction mixture have suggested that the procedure could accommodate twice the arsenious acid concentration without undue loss in sensitivity. It seems likely, therefore, that premature breakdown of the Ostwald isolation condition, caused by increased cerium(1V) concentrations, could be averted by this means, and therefore, that there is no reason why the procedure could not be operated successfully at any wavelength up to 455 nm. TABLE I1 THE FACTOR BY WHICH THE CONCENTRATION OF THE CERIUM(IV) REAGENT IS CHANGED FOR A GIVEN CHANGE IN WAVELENGTH FROM 410 nm Revised wavelength/nm 390 396 400 405 410 415 420 425 430 436 440 445 450 465 Concentration factor 0.46 0.67 0.66 0.81 1.00 1.17 1.44 1.77 2.29 2.82 3.66 4.66 6-83 7-21 By similar arguments, it can be seen that changes in the path length of the spectrophoto- metric cell used need not alter the performance of the procedure, provided that compensating changes in cerium(1V) concentration are introduced.Perhaps the main reason for using a long light path, then, is that less cerium(1V) reagent is used. Once again, it is clear that catalytic procedures involving cerium(1V) species are markedly different from their more conventional counterparts. When a new procedure is first established, the required base-line (A,) might not be obtained immediately. In this event an adjustment to the cerium(1V) reagent concentration is usually made. However, the foregoing discussion shows that when a spectrophotometer is incor- porated in the manifold, the adjustments can be made equally effectively, and probably faster, by shifting the wavelength setting.The magnitude of the change can be gauged from the results given in Table 11. Arsenious Acid A change in concentration of the arsenious acid present in the reaction mixture can be considered to have two major effects on the analytical procedure. Firstly, it alters the rate of reaction in any given reaction mixture and thereby affects the amounts of reagents con- sumed. Secondly, it alters w , the change in rate constant yielded by an additional 1 pg 1-1 of iodine, and therefore can ultimately affect the sensitivity of the procedure to iodine. However, experiments have shown that, in practice, decreasing the arsenious acid concen- tration to as low as one fifth of that recommended here does not offer any advantage to the analyst unless a harsh economic constraint is applied.The rate constants in mixtures con- taining distilled water as sample were 29 x 18 x lov3 and 9 x 10-3min-1 when the arsenious acid concentration in the entire mixture was 0-114, 0457 and 0-028 M, respec- tively. With these same arsenious acid concentrations, but with distilled water containing 46 pg 1-1 of iodide as sample, the rate constants became 66 x 60 x 10-3 and 53 x 10-3 min-1, respectively. Thus, little concomitant change in the over-all magnitude of the rate constants was observed, and w was almost constant at 0.80 x 0.91 x and 0 . 9 6 ~ 10-9 min-l pg-11, respectively.Febrzcary, 1975 IODATE IN SOLUTION BY CATALYTIC SPECTROPHOTOMETRY 121 The magnitude of o was also calculated by means of the Rodriguez and Parduel8 equation mentioned above.By this method, and allowing for the fact that here the iodine concentra- tion term refers to a 5-ml sample diluted to 35 ml, w was found to be 0.65 x low3 min-l pg-11 at the arsenious acid concentration levels studied. Therefore, the two approaches appear to agree satisfactorily. Interferences Background coloration, turbidity or presence in the water of substances capable of reducing ammonium cerium( IV) sulphate interfere in the procedure. Reaction mixtures derived from coloured or turbid waters will have a lower transmission than those derived from distilled water.Further, because the shift in transmission will be more pronounced in those reaction mixtures with high transmission (that is, when a high concentration of iodine is present), the gradient of the calibration graphs for coloured waters will be lower at any iodine concen- tration than that of the corresponding graph for distilled water. Experience has shown that whereas the shift in transmission at the base-line (0 pg 1-1 of iodine) is not greater than 0-5 per cent., it can be as much as 5 per cent. at the 100 pg 1-1 of iodine level. The argument con- cerning reducing substances is similar, but the effect on the calibration graph is in the opposite direction. Because of these interferences it seems advisable to construct a calibration graph for each batch of samples taken in a flow gauging exercise, using water that has been collected at the gauging site, but which has not been treated with iodine.This procedure also has the advantage that any other, unknown, interfering substances are present at the same concentration in both sqmples and standards, and the problems otherwise encountered in natural water analysis are circumvented. Precision The precision of the procedure at each of six concentration levels was tested by analysing six sets of eleven replicate samples of river water and the results obtained are presented in Table 111. A statistical F-test between the greatest and least variances obtained from TABLE I11 PRECISION AND RESOLUTION OF THE PROCEDURE AT SIX DIFFERENT IODINE CONCENTRATIONS Standard error ( x lo2) Iodine concentration/ 20 40 50 60 80 100 P-cg 1-1 Mean peak (a) of peak mission, T ) mission, I') 24.9 5.5 53.7 4.7 65.1 6.7 72.6 5.5 82.4 5.0 86.5 3.5 height (% trans- height (yo trans- (b) of iodine concentration 3.6 3.8 7.3 7.8 15.0 18.0 ( PI3 1-1) Relative standard error, per cent.0.18 0.09 0.14 0-13 0.16 0.18 column (a), showed that there is no significant difference a t the 1 per cent. level. The mag- nitude of the standard error (standard deviation)/dll for each set of peaks was therefore similar. Further, the precision at each level is comparable with that usually expected with an AutoAnalyzer procedure. Resolution In practice, the degree to which the iodine concentration can be resolved depends upon the precision of the response, which has been measured above as a standard error.Dividing the standard error for each response by the gradient of the calibration graph at each iodine concentration yielded the standard error in iodine concentration. Division again, by the prevailing iodine concentration, gave the relative standard error, which is a measure of the resolution (higher values indicate poorer resolution). Examination of these values shows that the resolution is similar, and is sufficiently good in each instance to justify the use of the previously derived calibration graph. We have chosen the worst value, namely 0.18 per cent., to describe the resolution of the method as a whole. It follows that the actual iodine122 TRUESDALE AND SMITH: AUTOMATIC DETERMINATION OF IODIDE OR Analyst, Vol. 100 concentration can be expected to be within 50.36 per cent.(h.2 x relative standard error) of the mean for 95 per cent. confidence, using 11 samples. ModWed Method Reagents Ammonium ceriuum(1V) sulphate. Dissolve a 7.5-g amount of the yellow solid salt [2(NH,),S0,.Ce(S0,),.3~H20] in 200 ml of water. Add 52 ml of concentrated sulphuric acid (36 N) and make the volume up to 1 1 with distilled water. Filter the solution through a Whatman glass-fibre (GF/A) filter circle before use. Arseniozls acid. Dissolve 9.8 g of arsenic(II1) oxide plus 7-0 g of sodium hydroxide in a small amount of water. Dilute the solution with a further 700 ml of water. Add concentrated sulphuric acid (36 N) dropwise until the solution is neutral to phenolphthalein. Add a further 28 ml of concentrated acid and 25.0 g of sodium chloride and make the volume up to 1 1 with water.Filter the solution through a Whatman GF/C glass-fibre filter circle. Iodine standards. Dry sodium iodide for about 1 h at 110 "C and dissolve about 0.1181 g of it in 1 1 of distilled water. Dilute the stock solution, which contains 100.0 mg 1-1 of iodide-iodine, with either river water or distilled water to give working standards. 0*040 Ammonium cerium (IV) Procedure Assemble the AutoAnalyzer manifold according to the design given in Fig. 4. With only the arsenious acid reagent and distilled water entering the manifold, set 0 and 100 per cent. transmission on the recorder. Introduce the cerium(1V) reagent. If the base-line does not stabilise at 10 per cent. transmission, correct its position in one of the ways mentioned earlier.When the base-line has settled at 10 per cent. transmission, introduce standard solu- tions and samples at the rate of 50 h-l. Read off the peak height (relative to the base-line) of the standards and construct the calibration graph. By using this graph and the peak height for each sample determine the iodine concentration of each sample. Heating Double bath coil Colorimeter 4 15-mm flow cell n Sampler sulphate 0-065 Fig. 4. The AutoAnalyzer manifold. * Numbers represent tube diameters in inches. Experience has shown that all reagents must be washed from the AutoAnalyzer manifold before it is allowed to stand overnight or for longer periods. If this precaution is not taken fine particles that form in the manifold can introduce an erratic turbidity error into subse- quent analyses.Thanks are due to Mr. A. G. P. Debney for his encouragement and support, and to the Director of the Institute of Hydrology for permission to publish this paper.February, 1975 IODATE IN SOLUTION BY CATALYTIC SPECTROPHOTOMETRY 123 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. References Mark, H. B., Talanta, 1973, 20, 257. Benotti, J., and Benotti, N., Clin. Chem., 1963, 9, 408. Comoy, E., Revue Fr. Etud. Clin. Biol., 1967, XII, 189. Foissac, L., “Proceedings of the International Symposium on Automation in Analytical Chemistry,” 1964, No. 454, Reprinted by Compagnie Technicon, Dormont, France. Riley, M., and Gochman, N., “Proceedings of the International Symposium on Automation in Analytical Chemistry, ” 1964, No. 62. Reprinted by Technicon Controls Inc., Chauncey, New York. Piette, P. C., Savoie, J. C., Piette, M., and Guillosson, J. J., Annls Biol. Clin,, 1966, 24, 393. Widdowson, G. M., and Northam, B. E., Clinica Chim. Acta, 1963, 8, 636. Zak, B., and Baginski, E., Chemist Analyst, 1962, 51, 39. Revel, J., Cah. Oce’anogr., 1969, 21, 237. Truesdale, V. W., and Spencer, C . P., Mar. Chem., 1974, 2, 33. Ingle, J. D., and Crouch, S. R., Analyt. Chem., 1971, 43, 697. Guilbault, G. C., “Enzymatic Methods of Analysis,” International Series of Monographs on Analytical Chemistry, No. 34, Pergamon Press, Oxford, First Edition, 1970. Yatsimirskii, K. B., “Kinetic Methods of Analysis,” International Series of Monographs on Analytical Chemistry, Pergamon Press, Oxford, Volume 25, 1966. Mark, H. B., Pap, L. J., and Reilley, C. N., in “Advances in Analytical Chemistry and Instrumen- tation,” Volume 2, Interscience Publishers, New York and London, 1963, p. 255. Blaedel. W. J., and Hicks, G. P., in “Advances in Analytical Chemistry and Instrumentation,” Volume 3, Interscience Publishers, New York and London, 1964, p. 105. Pardue, H. L., in “Advances in Analytical Chemistry and Instrumentation,” Volume 7, Interscience Publishers, New York and London, 1968, p. 141. Mark, H. B., and Rechnitz, G. A., “Kinetics in Analytical Chemistry,” A Series of Monographs on Analytical Chemistry and its Applications, Interscience Publishers, New York and London, Volume 24, 1968. Rodriguez, P. A,, and Pardue, H. L., Analyt. Chem., 1969, 41, 1369. Moore, W. J ., “Physical Chemistry,” Second Edition, Longmans, London, 1956. DubravCiC, M., Analyst, 1955, 80, 146. O’Neal, L. W., and Simms, E. S., Amer. J . Clin. Path., 1953, 23, 493. Foss, 0. P., Hankes, L. V., and Van Slyke, D. D., Clinica Chim. Acta, 1960, 5, 301. Deman, J., Mikvochim. Ada, 1964, 1, 67. Lein, A., and Schwartz, N., Analyt. Chem., 1951, 23, 1507. Malmstadt, H. V., and Hadjiioannou, T. P., Analyt. Chem., 1963, 35, 2157. Strickland, R. D., and Maloney, C. M., Analyt. Chem., 1957, 29, 1870. DuhravCiC, M., Analyst, 1955, 80, 295. Received March 19th, 1974 Amended September 2nd, 1974 Accepted October 2nd, 1974

ISSN:0003-2654    

DOI:10.1039/AN9750000111

出版商:RSC    

年代:1975

数据来源: RSC