摘要:

DECEMBER, I963 THE ANALYST Vol. 88, No. 1053 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETINGS AN Ordinary Meeting of the Society was held at 3 p.m. on Thursday, November 7th, 1963, in the Wellcome Building, Euston Road, London, N.W.l. The subject of the meeting was “Gas Analysis.” At the afternoon session the Chair was taken by the President, Dr. D. C. Garratt, Hon. M.P.S., F.R.I.C., and the following papers were presented and discussed : “Some Recent Developments in Gas Chromatography and Infrared Gas Analysis,” by D. W. Hill, M.Sc., Ph.D., F.Inst.P., A.M.I.E.E., A.R.I.C. ; “Some Problems Associated with the Analysis of Town Gas by Means of Gas Chromatography,” by G. R. Boreham, B.Sc., A.R.I.C.; “Some Methods for Moisture Determination in Gases,” by J.H. Scawin; “The Use of the Electrolytic Hygrometer for the Determination of Moisture in Gases,” by J. E. Still, B.Sc., F.R.I.C. The Chair at the evening session was taken by Dr. S. G. Burgess, B.Sc., F.R.I.C., F.Inst.Pet., and the following papers were presented and discussed: “Sonic Gas Analysers-Uses and Limitations,’’ by A. E. Martin, Ph.D., DSc. ; “Some Electrochemical Methods of Gas Analysis,” by J. H. Glover, B.Sc., F.R.I.C.; “The Use of a Slow Injection Syringe for Calibration in the Volumes per Million Range,” by H. F. Downing, BSc.; “Apparatus for Handling Gas Samples and for the Preparation of Gas Mixtures of Known Composition,” by J. E. Still, B.Sc., F.R.I.C. ; “Dispersive and Non- dispersive Infrared Analysers for In-line Measurements,” by A. W.Hough-Grassby, B.Sc., Ph.D. AN Ordinary Meeting of the Society was held at 7 p.m. on Wednesday, December 4th, 1963, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. D. C. Garratt, Hon.M.P.S., F.R.I.C. The subject of the meeting was “Differential Thermal Analysis” and the following papers were presented and discussed : “Differential Thermal Analysis and its Application to Soil and Plant Materials,” by R. C. Mackenzie, DSc., Ph.D., F.R.I.C., F.G.S., F.R.S.E. (presented by B. D. Mitchell, B.Sc., A.R.I.C., and followed by a short film dealing with the method of differential thermal analysis as it is performed at the Macaulay Institute); “The Use of Differential Thermal Analysis in the Mineralogical Investigations of Building Materials,” by H.G. Midgley, M.Sc., Ph.D., F.G.S.; “The Application of Differential Thermal Analysis to Polymeric Materials,” by D. E. Eaves, B.Sc., Ph.D. NEW MEMBERS ORDINARY MEMBERS Ijere Uguru Agwu, A.I.S.T., M.R.S.H.; David Valentine Atterton, M.A., Ph.D.(Cantab.), A.I.M. ; Mohssen Bavendi, Pharm.D. (Iran) ; Harry Leslie Blamires; Frank Arthur Chappell, B.Sc.(Lond.), A.1nst.P. ; Philip John Cooper, B.Sc.(Lond.), A.R.I.C. ; Terence Michael Duley; Charles Edward Dyer, M.P.S. ; Fred Howcroft Harrison; John Michael Hibbs, B.Sc.(Lond.) ; Robert Paul Hirsch; Alexander Hood, B.Sc.(Glas.), A.Inst.P., F.R.I.C. ; Meryl Ramsey Jackson, B.Sc.(Leeds) ; Michael Douglas Lack; Roy Basil Walter Lowndes; Lin Ma, BSc. (West China; State University of New York), Ph.D.(Leeds) ; Donald Frank Charles Morris, B.Sc., M.A., D.Phil.(Oxon.), F.R.I.C.; Eivor Eva Josefina Naddermier; Paolo Papoff; Gerald Russell, F.R.I.C., A.R.P.S. ; John Edward Saunders, A.R.I.C. ; Derek Bernard Schaverien, B.Sc.(Lond.), A.R.I.C.; James Scott, A.R.I.C.; Clemente Tarantola; Raymond Herbert Trust, L.R.I.C. ; Robert Laurence Weston. 903904 PROCEEDINGS [Analyst, Vol. 88 JUNIOR MEMBERS John Robert Ellis; John Edward Ellithorne, Dip.Tech.(Birm.) ; John Barry Pickup ; Iqbal Waheed Siddiqi, B.Sc.(Karachi) ; Robert Simpson; Brian Sims, B.Sc.(Hull) ; Roger Edward Weetman; Paul Young, Dip.Tech., A.C.T.(Rirni.). NORTH OF ENGLAND SECTION AND PHYSICAL METHODS GROUP A JOINT Meeting of the North of England Section and the Physical Methods Group with the Modern Methods of Analysis Group of the Sheffield Metallurgical Association was held at 7 p.m. on Tuesday, October 22nd, 1963, in the Conference Room of the British Iron and Steel Research Association, Hoyle Street, Sheffield, 3.The Chair was taken by the Chairman of the Physical Methods Group, Dr. W. Cule Davies, F.R.I.C. The following papers were presented and discussed : “The Determination of Oxygen in Metals by Fast-neutron Activation Analysis,” by A. L. Gray, B.Sc. ; “High-pressure Plasmas as Emission Sources, ” by S. Greenfield, L.I.M., I. L. Jones, D .L.C., A.R.I.C. , and C. T. Berry, A.R.T.C.S., A.R.I.C. (see summaries below). THE DETERMINATION OF OXYGEN IN METALS BY FAST-NEUTRON ACTIVATION ANALYSIS MR. A. L. GRAY described the technique of activation analysis and gave a short account of recent developments that made the technique practicable for routine plant analytical work.He discussed the convenience of the method for the determination of oxygen in metals by use of fast neutrons and described the development of a prototype automatic equipment, the “Analox.” The speaker described the assessment of the prototype instrument and discussed the correlation of the results obtained on a wide variety of samples with those by other methods, and outlined potentialities of the instrument. The paper concluded with a brief description of the fully developed version of the instrument and the considerations involved in its use in a steel works. HIGH-PRESSURE PLASMAS AS EMISSION SOURCES MR. I. L. JONES described high-pressure thermal plasmas and compared them with low-pressure plasmas.He discussed the production of both direct-current arc and radio-frequency induced plasmas. He then described the application of plasmas as emission sources for spectrographs and flame spectrophotometers. The speaker discussed the sensitivity and variance and the absence of interference from non-volatile compounds (as compared with flame sources) of plasma sources on flame spectrophotometers. He mentioned future applications of these sources, including the possibility of the use of direct calibration methods (obviating the need for internal standards) in emission analysis and the introduction of powdered samples into the plasma. NORTH OF ENGLAND SECTION AN Ordinary Meeting of the Section was held at 2.15 p.m.on Saturday) November 23rd, 1963, at the Old Nag’s Head Hotel, Lloyd Street, Manchester. The Chair was taken by the Chairman of the Section, Mr. C. J. House, B.Sc., A.R.C.S., F.R.I.C. The following paper was presented and discussed: “The Chemistry of Wines and Spirits,’’ by E, C. Barton-Wright, D.Sc., F.R.I.C., M.I.Bio1. SCOTTISH SECTION AN Ordinary Meeting of the Section was held at 7.15 p.m. on Friday, November Sth, 1963, at the Royal Society of Edinburgh, George Street, Edinburgh. The Chair was taken by the Chairman of the Section, Dr. R. A. Chalmers. The following paper was presented and discussed : “Atoms and the Analyst-Some Confessions of an Unqualified Chemist,” by J. M. A. Lenihan, M.Sc., Ph.D., A.M.I.E.E., F.1nst.P.December, 19631 PROCEEDINGS 905 A JOINT Meeting of the Scottish Section with the Glasgow Section of the Society of Chemical Industry was held at 6 p.m.on Friday, November 22nd, 1963, at the Royal College of Science and Technology, Glasgow. The Chair was taken by the Chairman of the Scottish Section, Dr. R. A. Chalmers. The following papers were presented and discussed : “The Presentation of Scientific Papers,” by J. D. Nisbet, M.A., B.Ed., Ph.D., F.B.P.S. ; “Editing a Scientific Journal,’’ by J. B. Attrill, M.A., F.R.I.C. MIDLANDS SECTION AN Ordinary Meeting of the Section was held at 7 p.m. on Thursday, November 14th, 1963, at the Lanchester College of Technology, Cox Street, Coventry. The Chair was taken by the Chairman of the Section, Mr. W. H. Stephenson, F.P.S., D.B.A., F.R.I.C. The following paper was presented and discussed: “Some Observations on the Use of Ion Exchange in Analytical Chemistry,” by J. E. Salmon, B.Sc., Ph.D., F.R.I.C. PHYSICAL METHODS GROUP A SPECIAL General Meeting of the Group was held at 6.30 p.m. on Tuesday, September loth, 1963, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the Chairman of the Group, Dr. W. Cule Davies, F.R.I.C. Mr. S. G. E. Stevens was elected Honorary Auditor in place of Dr. D. C. Garratt, who had ceased to be eligible on being elected to the Presidency of the Society. BIOLOGICAL METHODS GROUP AN Ordinary Meeting of the Group was held at 7 p.m. on Wednesday, October 23rd, 1963, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the Chairman of the Group, Mr. W. A. Broom, B.Sc., F.R.I.C. The subject of the meeting was “Assay of Virus Vaccines” and the following papers were presented and discussed: “The Control Testing of Vaccines,” by D. I. Magrath, B.A., Ph.D.; “Manufacture and Control of Poliomyelitis Vaccine,” by A. J. Beale, M.D., Dip.Bact.

ISSN:0003-2654    

DOI:10.1039/AN9638800903

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

906 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS [Analyst, Vol. 88 Thermogravimetric Analysis A Review* BY A. W. COATS AND J. P. REDFERN (Chemistry Department, Battersea College of Technology, London, S. W . 11) SUMMARY OF CONTENTS Introduction Techniques Kinetic studies Apparatus Applications- (2) Analytical chemistry (ii) Inorganic chemistry (iii) Organic chemistry Complementary procedures Conclusion THERMAL methods of investigation, generally referred to as thermo- or thermal analysis or thenno-analytical techniques, have found wide application in recent years1 These may be defined as experimental methods for characterising a system (element, compound or mixture) by measuring changes in physico-chemical properties at elevated temperatures as a function of increasing temperature.2 The two chief methods are (a) differential thermal analysis,3 in which changes in “heat content” are measured as a function of increasing tem- perature and ( b ) thermogravimetric analysis, in which changes in weight are measured as a function of increasing temperature.Other methods that come within this definition involve the use of changes in gas volume or pressure; changes in solid volume; changes in electrical resistance ; changes in ultraviolet, visible or infrared transmission or reflectance.4 The two techniques, (a) and ( b ) , provide information relating to certain physical and chemical phenomena, which are listed below- Physical Phenomena Chemical Phenomena Crystalline transition Second-order transition Fusion Vaporisation Sublimation Absorption Adsorption Desorption Chemisorption Desolvation (especially dehydration) Decomposition Oxidative degradation Solid-state reactions Solid - gas reactions (e.g., oxidation or reduction) The basic instrumental requirements for thermogravimetric analysis are a precision balance and a furnace that is programmed for a linear rise of temperature with time.Thermo- gravimetry can provide information on all the phenomena listed above, except crystalline transitions, fusions and those solid-state reactions that occur without change in weight. Although information can be obtained by carrying out the weighing operations manually, nowadays automatic continuous recording of the weight and temperature are usual; the continuous record of weight and temperature ensures that no features of the weight loss - temperature curve are overlooked.The results from a thermogravimetric run may be presented by- (;) weight (corrected weight-see below on “Corrections”,. p. 908), mmus temperature (or time) curve, referred to as the thermogravimetric curve, see Fig. 1 (a), or (ii) Rate of loss of weight versus temperature curve, referred to as the differential thermogravimetric curve, see Fig. 1 (b). In (i), the weight axis may be scaled in one of several ways, e.g., (a) as a true weight scale, ( b ) as a percentage of the total weight, (c) as a percentage of the total weight loss or as a fraction * Reprints of this paper will be available shortly. For details, please see p. 984December, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 907 of the total weight lost, (d) in terms of molecular-weight units, or (e) expressed in terms of a (where a = fraction decomposed).(When the molecular weight of the compound is known, method (d) affords a convenient method of plotting the results, since it is easy to extract data relating to lost fragments, for example, in the study of an inorganic complex-it also forms an easy method of comparison of a family of such compounds.) The use of method (e) should be limited to a single-stage process, i.e., a special instance of (c). The following features of the thermogravimetric curve may be identified- (i) A horizontal portion or plateau, which is indicative of constant weight. (ii) A curved portion; the steepness of the curve is indicative of the rate of weight loss, and this will obviously pass through a maximum, giving an inflection with dw - as a maximum.The shape of the curve is dependent on the variables discussed dt in the section on “Techniques” (vide infra). dw dt (iii) An inflection (at which - is a minimum, but not zero) may imply the formation of an intermediate compound. I t may, however, be due to disturbances in the heating rate or in thermocouple response. It is necessary to ensure a regular rise of temperat~re,~ and therefore it is desirable to have an independent temperature - time record to ensure the reliability of the results.6 The portion of the differential thermogravimetric curve lying on the line - = 0, see Fig. 1 ( b ) , is equivalent to the horizontal portion of the thermogravimetric curve.The peak of the differential thermogravimetric curve corresponds to the curved portion of the thermo- dw dt ~ ~- .- -.-.-. ~~ Temperature, “C - Fig. 1 (a). Thermogravimetric curve : A ( l ) , (2) and (3) are plateaux in the decomposition curve of the material. B is a point of reflexion (at which f is a minimum) Fig. 1 ( b ) . Differential thermogravimetric curve : A ( l ) , ( 2 ) and (3) correspond to the plateaux on the thermogravimetric curve (at which 2 is zero). Trough B corresponds to the point of inflexion on the thermogravimetric curve (at which $ is a minimum) gravimetric curve, whereas the peak maximum of the former is identical with the point of maximum slope of the latter. A trough on the differential thermogravimetric curve corre- sponds to an inflection at which - is a minimum on the thermogravimetric curve.The height of the trough above the line - = 0 affords some measure of the stability of the inter- mediate and the extent to which the two consecutive reactions (corresponding to the peaks on either side of the trough) overlap. The differential thermogravimetric curve offers certain advantages over the thermogravimetric curve in the matter of presentation. Features in the thermogravimetric curve that are not readily discerned are more clearly seen in the differential thermogravimetric curve, e.g., any change in the rate of weight loss may be seen immediately aw dt dw dt908 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS [Artahst, Vol. 88 as a trough, indicating two consecutive reactions, or as a shoulder to the peak, indicating two almost overlapping reactions or as a tail to a peak, which is probably an indication of strong adsorption of the volatile product on the new solid phase.’ Differential thermogravimetric curves often show considerable similarity to differential thermal analysis curves or permit a ready comparison to be made.g~g~lo The use of thermogravimetric results for evaluating thermal stability has focused atten- tion on finding a workable definition for a suitable standard for describing decomposition.The temperature at which a reaction begins in any particular thermobalance run is dependent on many variables, of which the rate of heating is perhaps the most important. This tempera- ture is neither a true decomposition temperature, below which the reaction rate is zero, nor is it a transition temperature-l1 In fact, there is often little correlation between results from isothermal runs and non-isothermal runs.5 Newkirk12 argues that knowledge of this tem- perature is useful and the term “procedural decomposition temperature” used by Doyle,l3 in his polymer studies, stresses the dependarice of this temperature on the powerfully in- fluential procedural details.Pellon14 in his work on the stability of phosphorus-containing polymers uses a temperature T,, (a temperature at which the cumulative weight change reaches 10 per cent.) as a means of defining thermal stability. A comparison of decom- position temperatures has often led to controversy owing to a lack of appreciation that the value is a function of method, apparatus and procedure.A clear statement of conditions used should be stated when quoting decomposition temperatures. Guiochonll comments : “Thermogravimetric measurements are easy to carry out, but somewhat difficult to account for.” A survey of the history of thermogravimetric analysis up to 1940 has recently been published by Duval,15 in which he traces the design and construction of the first thermobalance back to Nernst and Riesenfeldla; Wachel’ has written an article on Chevenard. The second edition of Duval’s book, which is now available,18 is divided into a discussion of (a) “The Thermobalance” and ( b ) “The Thermolysed Substances.” Review articles have been published on recording balances and instrun~entationl~ ,20 and on thermogravimetry.21 to 27 Several symposia have been held, notably in the United state^.^^^^^ The inception of the first Thermoanalysis Institute held at the Fairleigh-Dickinson University, New Jersey,30 occurred in 1962.A survey of literature, published about four times a year, commenced in May, 1962.l TECHNIQUES The first arises from the dynamic nature of the method, and the second is concerned with the factors influencing the shape of the thermogravimetric curve, and these must be taken into account to obtain meaningful and reproducible results, so that thermogravimetric data on different compounds can confidently be compared. This section is concerned with two aspects of thermogravimetry. CORRECTIONS- If a crucible, that is known not to change weight, is heated when empty, there is, generally, an apparent change in weight with increasing temperature (see Fig.2), e.g., a I I I I I 0 loo 300 5w Temperature “C. Fig. 2. Typical correction curveDecember, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 909 crucible weighing 2 g would be expected to have an apparent change in weight at 1000" C of, say, 0.1 per cent. of the crucible weight; this would amount to an apparent change in weight of 2 per cent. if the sample weighed 0.1 g. This apparent weight change is caused by the interplay of a complex combination of several factors, such as air buoyancy, convection effects within the furnace, crucible geometry, radiation effects, the atmosphere in the furnace and the fact that the crucible support is subject to a temperature gradient within the furnace.It is necessary to use the apparent weight change of the empty crucible as a correction curve to arrive at the actual weight change occurring in a sample. This is considered to be more satisfactory than the use of a tare crucible, as has been proposed,31 in a separate furnace, in view of the possible non-uniformity of the hot zones in the two furnaces. Since several factors contribute to this correction curve it is found that the corrections to be applied differ with different rates of heating. Various ~ ~ r k e r ~ ~ ~ have discussed the factors affecting the correction curves, and DuvaP and Newkirk12 have suggested the use of vents at the top of the furnace to reduce the size of the correction. This has been ~riticised~~ as amounting to critical damping of the hot-zone volume.Empirical equations have been derived35v36*37 for the correction curves, but because of the complexity of factors involved it seems simpler to use a correction curve determined practically under conditions identical to those of the actual experiment. FACTORS AFFECTING THE RESULTS OBTAINED- There are several factors that may, to a greater or less extent, influence the shape of a corrected thermogravimetric curve for a particular compound. Heating rate-For a single-stage endothermic reaction, where Ti is the procedural de- composition temperature and Tf is the temperature at which the reaction is completed, it has been shown5~38~39~40 that- where the subscripts f and s refer to a fast and slow rate of heating, respectively, @) (% > (Tf),, (ii) (Ti), > (Ti), and (iii) (Tf - Ti)f > (Tf - Ti),.Newkirk12 has shown that, at any given temperature, the extent of decomposition is greater at a slow rate of heating than for a similar sample heated at a faster rate. If the reaction involved is exothermic, the sample temperature will rise above the furnace tempera- ture; it has been shown (in unpublished work by us) that the difference between the furnace temperature and the sample temperature is greatest for the faster rate of heating when a reaction is occurring. When successive reactions are involved, the rate of heating may well determine whether or not these reactions will be separated. The appearance of a point of inflection in the thermogravimetric curve at a fast heating rate may resolve into a plateau at a slower heating rate.23 The importance of heating rates was stressed by early workers41 942 and has been much studied.43t04' Sample-The sample weight can affect the thermogravimetric curve in three ways- (i) The extent to which endo- or exothermic reactions of the sample will c a y the sample temperature to deviate from a linear rise; in general, all other factors being equal, we have found that the greater the weight of sample, the greater will be the deviation. (ii) The degree of diffusion of the product gas through the void space around the solid particles ; under static conditions, the environmental atmosphere (i.e., the atmosphere imme- diately surrounding the reacting particles) will be somewhat governed by the bulk of the material in the crucible.(2;;) The possible unevenness of the temperature throughout the sample, particularly if it has a low thermal conductivity. Thus the use of as small a weight of sample as possible, within the limits of the sensitivity of the balance, is to be preferred. The state of sub-division of the sample is important. The use of large crystals may result in spitting, as described by C. J. Keattch in a private communication. The sample may foam and bubble.48 Most thermogravimetric studies have been carried out on powders, and the effect of particle size or surface area has been widely studied.41 to 52 The smaller the9 10 COATS AND REDFERN : THEKMOGRAVIMETRIC ANALYSIS [Afidyst, VOZ. 88 particle size the greater the extent to which equilibrium is reached, and at any given tempera- ture the extent of decomposition is greater.Wendlandt53 has discussed the effect of precipitation technique on the thermogravimetric curve of ammonium molybdophosphate. Recent work on magnesium hydroxide from different sources provides a good illustration of this factor.54 Crucible-The geometry of the crucible used will profoundly affect the shape of the thermo- gravimetric curve. Crucible design has been st~died,~~y~6 and the use of a crucible fitted with a piston (where reactions take place under a pressure of 1 atmosphere of the liberated gas) or an on open tray (at the partial pressure of the gas in the atmosphere) have been compared. This latter arrangement has been used by other worker^.^^.^^ The design of a crucible to overcome possible condensation on the support rod or decrepitation before decomposition has also received a t t e n t i ~ n . ~ ~ ) ~ ~ In certain instances the material of which the crucible is constructed affects the decomposition A t mosp heye- (i) When the atmosphere does not take part in the reaction-The use of an inert atmos- phere in thermogravimetric analysis has been The r81e of the gas is to remove gaseous products from the vicinity of the sample, to ensure that the environmental atmos- phere is kept as constant as possible throughout the experiment and to prevent reaction occurring between the sample and air as normally e m p l ~ y e d , ~ ~ , ~ ~ or between the volatile components and the a t m ~ s p h e r e .~ , ~ ~ Other workers64 to 68 have used vacuum conditions to achieve the same results.The dependence of the shape of thermogravimetric curves on the pressure of the gas in the reaction chamber has been studied.69 (ii) When the atmosphere is involved in the reaction-Some work has been carried out with atmospheres either of humidified air or high pressure steam12y70y71 ; a dynamic air atmos- phere has been used to study the roasting of copper ~ u l p h i d e ~ ~ and the oxidation of reduced iron catalyst.72 A comparison of some six atmospheres in metal- gas reactions has been made by Markowitz and B ~ r y t a . ~ ~ , ~ ~ Reducing atmospheres, e.g., hydrogen, have been used.75976~77 Other atmospheres, in which reactions occur between the gas and the sample, that have been studied include hydrogen ~ u l p h i d e ~ ~ and carbon dioxide.79 The pre-history of the sample is also important.KINETIC STUDIES Thermogravimetric data can be used to evaluate kinetic parameters of reactions involving weight loss (or gain) of the following four types- AS -+ Bs + Cg80981 As + Bg -+ Cs73 As + Bs --+ cs + D, As or A1 -+ AgS2 The advantages of determining kinetic parameters by thermogravimetric methods rather than by conventional isothermal studies are (i) considerably less data are required than in the isothermal method, (ii) the kinetics can be probed over an entire temperature range in a continuous manner and (iii) when a sample undergoes considerable reaction in being raised to the required temperature, the results obtained by an isothermal method are often question- able. To these reasons may be added the advantage of using one single sample in the study.It is important to know the temperature accurately and to ensure that endo- or exo- thermic effects do not cause the rate of heating to depart from its constant ~alue.7~ The use of small samples, within the limits of sensitivity of the balance, is therefore necessary. It should be remembered that kinetic parameters derived from thermogravimetric experiments are dependent on the procedural details, e.g., crucible shape and material, particle size, pre- history of sample and heating rate.47,54 The earliest attempt to use thermogravimetric curves for kinetic data appears to have been made by Van Krevelen, Van Heerden and Huntjensqa They derived an approximate equation, from which it was possible by graphical methods to determine the order of reaction (they differentiated between orders of 0, 1 and 2).They studied the pyrolysis of coal and showed that the primary decomposition was first order. On this assumption they developed graphical methods for determining the activation energy and frequency factor, knowing the rate of heating, temperature of maximum rate of decomposition and the half-value width of the differential thermogravimetric curve.December, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 911 Kaesche-Krischer and H e i n r i ~ h ~ ~ used a similar approach in their study of the pyrolysis of poly(viny1 alcohol) in a vacuum, and Schnitzer, Turner and c o - w o r k e r ~ ~ ~ 8 ~ studied the organic matter of Podzol soils. Turner, Hoffman and Chen54 modified Van Krevelen’s approach to accommodate a two-thirds order in their work on the thermal decomposition of magnesium hydroxide. Freeman and Carrollso derived the equation- .. . . - - (1) Alog (dw/dt) - E /2-3RA T-l --x+--.- AlogW, - Alog W r where E = activation energy, R = general gas constant, T = absolute temperature, x = order of reaction and Wr = (We - W ) , where W = weight loss at time t, and W , = weight loss a t completion of the reaction. This equation was derived by assuming a rate expression- . . * * (2) = kX” . . . . . . dX dt -~ where X = amount of A at time t and the rate constant, k , is given by the simple Arrhenius expression- .. * - (3) k = Ze-EIRT . . .. .. or veysus ~ Alog W r AlogW, AT-l Alog(dw/dt) where 2 = frequency factor.Either by plotting a graph of by other suitable rearrangement of equation ( l ) 6 4 3 8 1 9 s 6 it is possible to derive values for both E and x. This derivation has been used to study several dehydration reaction~.7@~*0~S1~87 t o 91 Bar- rer and Brattg2 have used an adaptation of equation (1) in their study of non-stoicheio- metric hydrates. The decomposition reactions of calcium oxalate,80 ,81 thorium, lanthanum and uranium ~ x a l a t e s , ~ ~ calcium carbonate,80 ,81 chromium a r ~ e n a t e ~ ~ and 12-tungstomanganic acidg4 have all been studied. Bear and Wendlandtg5 have studied the effect of added salts on the decomposition of trisethylenediamine and trispropylenediamine chromium111 chlorides and thiocyanates. Similarly, the pyrolysis of several polymers has been st~died~~y86,~~J’7 However, several disadvantages appear to exist in the use of this derivation- (i) in most practical instances the derivation seems to apply to an extremely limited portion of the decomposition curve ; (ii) there is considerable difficulty in obtaining a reliable value for the order of reaction; (iii) none of the other methods of applying the derivation seems to overcome the diffi- culties; and ( i v ) in certain reactions the method does not seem to yield an answer at a1187-this may be due to the reaction itself and not inherent in the derivation.In his method,12 Newkirk assumes that all pyrolytic decompositions are first order, and his method involves calculating the first-order rate constants for a large number of tem- peratures, and then constructing the Arrhenius graph.In his study of Mylar (a plastic marketed by Du Pont Company Ltd.) he showed that the slope of the graph was independent of heating rate, but, as the rate of heating was increased, the graphs were displaced towards lower values of 1/T, compared with the results obtained from isothermal studies. Smithgs has used a similar approach to derive E , ( E , denotes that E , the activation energy, may be some- what dependent on the experimental procedure) for various polymers. Other workersM have determined apparent activation energies by plotting the reciprocal of the final decomposition temperature against the logarithm of the heating rate. Doyles2 integrated a combination of expressions equivalent to equations (2) and (3), using an approximation. He studied the zero-order volatilisation of liquid octamethylcyclotetrasiloxane and the first-order pyrolytic volatilisation of polytetrafluoroethylene. He confines himself to zero- or first-order reactions, but his approach cannot be used when the order is unknown.Horowitz and Metzgerg9 have recently developed another approach based on integration of a combination of expressions equivalent to equations (2) and (3) by an exact integral, w 1 using the substitution (T - T,) = 0 where T , is the temperature a t which - = - Wo e’ (W = weight of sample at time t, and W , is the initial weight of the sample). The theory His technique is essentially one of curve fitting.912 COATS AND REDFERN THERMOGRAVIMETRIC ANALYSIS [Analyst, Vol.88 These w* shows that a plot of In (In -) versus 8 should be a straight line of slope - W RTs2’ workers have studied four polymers and the dehydration of calcium oxalate monohydrate. In each reaction they assume a particular order of reaction. For an unknown order they suggest the use of the position of maximum rate since this is governed by the order. It is, however, not always easy to determine the position of maximum rate with accuracy. Coats and Redfernloo used a different approximation for integrating a combination of equations (2) and (3). The graph of 1 - (1 - a)1--n 1 versus - log (1 - n)T2 T is a straight line of slope - E/2-3R for the correct value of n (a = fraction decomposed and rn = order of reaction). The activation energy can have a real meaning in solid-state kinetics, corresponding to the rate-determining step or steps, which might be the diffusion of the gaseous product out of the solid, or the transport of a particular ion, or the breakage of bonds.The order of reaction cannot presumably have the meaning attributed to it, as in a gas reaction, and may best be considered as a mathematical factor in the derived equations. However, geometric models of solid systems that lead to orders of reaction of & and 8 can be set uplol; orders of 0 and 1 can also be justified. The first is a plea for the use of standard symbols in the description of kinetic processes (see, for example, Garnerlo2). The second is that it seems unlikely that the simple rate expression - = k(1 - a)n, from which all the derivations quoted are ulti- mately derived, will be applicable to all solid-state decomposition reactions.The use of computational methods for analysing thermogravimetric data is probably in wider use than the number of papers on this subject would suggest.lo0JW The advantages that these methods possess in assisting in the evaluation of kinetic parameters are obvious; their more widespread introduction will undoubtedly lead to a better coverage of some of the necessary studies on the effect of the variables, listed under “Techniques,” on kinetic parameters. APPARATUS Two points remain. da d t The comprehensive review of Gordon and CampbelllS covered the whole field of automatic and recording balances up to 1959. Many of these are directly applicable or readily adaptable for use in thermogravimetry.Lewin20 discusses six makes of commercial thermobalances on the market in 1962 and lists another four firms manufacturing thermobalances. DuvaP puts the number of thermobalance models on the market at fifty-two. Other firms104~105~106 known to us are manufacturing thermobalances. The essential components of a modern thermobalance are : (i) balance, (ii) furnace, and (iii) recorder. It is desirable to have a reaction chamber to permit work to be carried out under a wide variety of conditions, e.g., inert, oxidising or reducing atmospheres or under vacuum, and to permit gas analysis to be carried out. The balances used can be grouped into two types. The former incorporates a suitable sensing element that detects any deviation of the balance beam and the application of a restoring force, proportional to the change in weight, to return the beam to its original null-point.This restoring force is then recorded either directly or through a transducer. These null-point type instruments are often readily adaptable to working under vacuum conditions. Deflection instruments, e.g., based on a conventional analytical balance, a helical spring, a cantilever beam, a strain gauge or a torsion balance, involve the conversion of deviations into a record of the weight change. The principles used in detecting and restoring deviations (in null-point balances), and in recording the changes in weight have been fully Furnace design and control is of great importance; it must be designed to provide a suitable smooth input so that it can maintain a linear heating programme or a fixed tem- perature, independent of any changes in external conditions.The comments on temperature control in differential therm a1 analysis3 apply equally to thermogravimetric analysis. Control is generally achieved via a thermocouple situated as close to the furnace winding as possible. They are the null-point and the deflection types of instruments.December, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 913 The appearance of thermal pulsing effects owing to the periodic application of relatively large power increments leads to a ragged weight record,32 and the use of some form of stepless control of power supply to the furnace through a saturable reactor is to be preferred.lO7 Nichrome winding permits a maximum temperature of around 1100" C ; platinum - rhodium winding permits a maximum of around 1450" C.Higher temperatures can be achieved by using a graphite tube furnace, but the associated control and measurement of temperatures raise considerable problems. The recording system should be able to record both temperature and weight continuously and to make a periodic record of the time. The use of "X, - X2"lo5 or side-by-sidelos recording is to be preferred to the use of "X - Y" recording5 since these methods provide an independent record of both temperature and weight. The prospective user has a variety of instruments available over a wide cost range. His choice should reflect the requirements of his particular field of study and the desiderata for good thermobalance design (see, for example, Lukaszewski and Redfern23), namely that- (i) The thermobalance should be capable of continuously registering the weight change of the sample studied as a function of temperature (and time).(ii) The furnace should be capable of reaching the maximum temperature desired. Most commercial instruments are marketed with a maximum of around either 1100" or 1500" C (some thermobalances are available and have been designed to work up to 2200°105 or 2700" Clog). (iii) The rate of heating is truly linear and is reproducible. (iu) The hot zone of the furnace is as near uniform as possible, and that the crucible is always located within this hot zone. This is particularly important in deflection instruments, since the crucible will move in relation to its initial position in the hot zone during the course of the run.(a) variation in heating rate, (b) heating in a dynamic flow of a controlled atmosphere, inert, oxidising or reducing, (c) heating in uacuo, (d) variation in chart speed to aid interpretation and (e) the possibility of carrying out accurate isothermal studies. (vi) Physical effects (e.g., radiation and convection currents, and magnetic effects caused by the winding of the furnace) due to the functioning of the apparatus do not upset the balance mechanism. No interaction should occur between conducting or magnetic samples being studied and the furnace winding. Possible chemical attack, from the gases used or the gases evolved, can be eliminated either by the design of the reaction chamber or by the materials used.The balance mechanism is sufficiently protected from the furnace, so that its sensitivity remains constant throughout the whole of the experimental run. (v) The instrument is as versatile as possible, providing for- (viz) The temperature of the sample is measured as accurately as possible. (viii) The sensitivity of the balance mechanism allows the study of small samples. Several workers have described the construction of relatively simple inexpensive instruments, e g . , those suitable for teaching the principles of the subject or for preliminary studies.22 9110 ?ll1 $12 Micro-recording,l13 Y1l4 vacuum or ~ontrolled-atmosphere1~2 9 1 1 5 9 1 1 6 thermobalances have been described in detail. Recent papers, in which automatic thermo- balancesll7 to 121 are described, include details of models capable of operating up to 40117 and 6O1lS atmospheres.In view of the trends in current instrument design reviewed by East- wood,122 the development of an automatic digital-recording thermobalance is of great interest.123 There is considerable activity in the development of instruments capable of performing both thermogravimetric and differential thermal analysis either on the same sample or under similar c 0 n d i t i o n ~ . ~ 6 J 0 7 ~ ~ ~ ~ 9 ~ ~ ~ 9 ~ ~ 6 However, it should be borne in mind that the conditions for producing meaningful results are not necessarily the same for both techniques, although comparison is often valuable and, in some instances, essential.914 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS [Anallyst, Vol.88 Several workers have described modifications to existing commercial recording balances or thermobalances. Thus the Ainsworth recording balance has been incorporated in a thermobalance design31 ; the Stanton thermobalance has been adapted for work in atmos- pheres of hydrogen12’ and for producing the differential thermogravimetric curve.lO It is possible to measure the sample temperature directly without affecting the per- formance of a conventional balance by means of extremely fine connecting wires (of about 0.001-inch diameter) between the balance arm and a suitably located terminal block. The thermocouple wires then run up the support rod the crucible, and the bead may be located in or close to the sample, depending on the design of the crucibles used.46910s APPLICATIONS ANALYTICAL CHEMISTRY- The widest application of thermogravimetric analysis to date has been in the investigation of analytical procedures : (i) in investigating suitable weighing forms for many elements ; (ii) in testing materials that are actual or potential analytical standards; and (iG) in the direct application of the technique to analytical determinations. Kobayashi128 has reviewed the work of the Japanese School from 1925 to 1940 in the field of gravimetric analysis ; from some twenty-seven references he listed about 300 precipi- tates and gave the recommended drying temperatures. In the first edition of his book,129 Duval reported his studies on over 1000 gravimetric precipitates for nearly 70 elements.He concluded that only about 200 of these are suitable weighing forms for the elements. Erdey’s book130 on gravimetric analysis includes thermogravimetric, differential thermo- gravimetric and differential thermal analysis curves obtained on the Deri~atograph~ for each weighing form discussed. Two reviews have been p~blished.131913~ There has been criticism of some of Duval’s work on a number of grounds, Differences between other published work and that of Duval and his co-workers has been ascribed to different precipitating techniques,% to different washing techniques,l= and to the use of high rates of heating1= Duval’s criterion for rejecting a particular gravimetric precipitate as not being a suitable weighing form was that it did not give a plateau when heated at one particular rate of heating, Newkirk12 refers to the investigation of zinc monosalicylaldoxime in which de Clercq and D ~ v a 1 , l ~ ~ using the method of Flagg and F ~ r m a n l ~ ~ to prepare the precipitate, found no plateau on the thermogravimetric curve corresponding to the anhydrous compound.Detailed work by Rynasiewicz and showed that the anhydrous form was stable from ambient temperature up to 285” C; further, that the plateau depended on the initial water content. Newkirk continues: “In as much as Duval has used a higher heating rate (ca. 380” C per hour), it is perhaps not surprising that he failed to find a plateau. The rejection of this precipitate as an analytical method for the determination of zinc on the grounds that it does not give a stable horizontal in the thermobalance at one particular rate of heating is clearly unwarranted.. . . The results suggest that when the thermobalance is used to study the drying of bulky precipitates that contain considerable water, it would be well to use very slow rates of heating.’’ The use of thermogravimetric data to interpret the best drying temperature (a constant temperature) must be made with a clear understanding of the dynamic nature of the technique. Regard must also be paid to all the factors mentioned in the section on “Techniques.” I t is clear that failure to do this may lead to considerable difficulty and controversy. To sum- marise- (i) The lack of the appearance of a plateau, at one particular heating rate, is insufficient evidence on which to judge the suitability or otherwise of a particular weighing form.(ii) The appearance of a plateau is not conclusive evidence that the weighing form is isothermally stable at all or any of the temperatures that lie on the p l a t e a ~ . ~ (iii) It is evident that the most reliable information will be gained by using several different heating rates or, at least, a slow rate of heating, possibly with preliminary air drying. However, there is much to be gained from a thermogravimetric investigation of a weighing form since the recommended temperatures for some procedures have been arrived at arbitrarily, and have frequently been quoted in the literature without any critical evaluation of the conditions required. Beamish and M~Brydel~~ have shown that certain instructions givenDecember, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 915 for the heating or ignition of materials in some gravimetric determinations were quite inadequate.Recent investigations include the study of : ammonium molybd~phosphate~~ and quino- linium molybd~phosphate~~g as suitable weighing forms for phosphorus ; several su1phideslm and selenides141 to confirm methods of drying and weighing; lanthanum carbonate142 and lanthanum oxalatef4, to find suitable weighing forms for lanthanum ; the use of m-nitrobenzoic acid as a precipitant for ceriumIv 14; and various compounds of plutonium to find suitable weighing forms for it.145?146 Picrates and styphnates of some organic bases, such as guanidine and related compounds, have been examined since these derivatives are constantly used in charact erisat ion and analysis.147 Thermogravimetric studies have been made of many substances of use or of potential use as standards in analysis. Thus a series of papers by Duval has appeared on the use of this technique in conjunction with infrared 9149 Investigations of potassium hydrogen phthalate,150 ethylenediaminetetra-acetic acid (EDTA) and its derivatives,151 some tetra- phenylboron salts of 0xirnes1~~ and 5-substituted barbituric acids153 have been carried out. I t has been shown that soils can be analysed thermogravimetrically for determining hygroscopic moisture, organic matter, and inorganic ~ a r b o n a t e s . l ~ - ~ ~ ~ 3156 Dupuis and Dupuis15' have used thermogravimetry for determining calcium and magnesium in dolomite rock.The use of thermogravimetry in the study of the composition of non-interacting binary mixtures has been 0utlined.2~ Erdey and c o - w ~ r k e r s , ~ ~ ~ 9159 using the Derivatograph, have determined calcium, strontium and barium in a single sample by their precipitation as the mixed hydrated oxalate, using the losses of weight of extraneous water and water of crystal- lisation, and loss of carbon monoxide and dioxide. Berlin and Robinson have used thermo- gravimetry in their determination of magnesium, potassium and lead with dilituric acidl60 and for determining ethylenediamine and quinine,161 with the same precipitant. Fluorine has also been determined thermogravimetrically.162 INORGANIC CHEMISTRY- In recent years the greatest number of papers on thermogravimetric analysis has appeared within the field of inorganic chemistry.In a review of this nature it is not possible to discuss the work in detail; however, an attempt has been made to collate recent work on the basis of the anion of the compound studied. Studies of inorganic compounds are concerned with their stability, decomposition and structure. The types of reaction that have been observed are- Loss of constituent water molecules; such dehydrations may be either single or multi-stage, and may involve the loss of the elements of water, e.g., from hydroxyl groupings. (ii) Decomposition reactions, which may be either of a disproportionation or of a de- gradation type. Bri11,1G3 in what was probably the first publication in the field of thermogravimetry, presented data of this degradation type.He studied the decomposition of magnesium; calcium and barium carbonates- (;) MIICO, + MI10 + CO,. (iii) Degradation reactions specifically involving the atmosphere, for example, oxidative An example is provided by the oxidation in air of trispropylene- degradation. diamine chromium111 chloride- B[Cr(pn),]Cl, --+ Cr203 + v01atiles.l~~ (iv) Loss of constituent volatile ligands from inorganic complexes. An example is provided by the loss of ethylenediamine from trisethylenediamine chromium111 chloride- [Cr(en),]Cl, - [Cr(en),Cl,]Cl + en.95 Though it has been clearly demonstrated that the thermal decomposition of inorganic compounds is dependent on the nature of both anion and cation, there has been little attempt to relate thermal decomposition data to modern theories in inorganic chemistry.Wendlandt1G5 found a linear relationship between the ionic radii of the alkali metal ions and the temperature at which their tetraphenylboron salts began to decompose. Others have proposed reaction schemes to account for the pattern of thermal de~ornposition.~8@~~66~~6~916 COATS AND REDFERN THERMOGRAVIMETRIC ANALYSIS [Analyst, Vol. 88 Anions derived from ekments of Group III-Some borates of lithium and sodium have been studied.lG8 The decomposition of some calcium aluminate hydrates169 and calcium ~arbo-aluminatel~~ has been investigated, that of the former under constant water vapour pressure. Anions derived from elements of Group IV- (i) Carbonates-Freemans1 has studied the kinetics of the decomposition of calcium carbonate, comparing results obtained with results from isothermal studies.Lanthanum carbonate has been investigated as a suitable weighing form for lanthanum142; other lanthanon carbonates have been s t ~ d i e d . ~ ~ ~ ~ ~ ~ ~ The effect of crucible design on the decomposition characteristics of lead carbonate has been studied.173 Other recent studies on carbonates include ammonium scandium carbonate1r4 and certain cobalt111 complex carbonates.175 Thermogravimetric studies of solid-state reactions between cerium oxide, neodymium oxide or samarium oxide with either sodium carbonate or sodium ~ u l p h a t e , l ~ ~ and of the strontium carbonate - zirconium oxide and barium carbonate - zirconium oxide systems have been made.177 Wilburn and co-workers8 have studied the systems sodium carbonate - silica and calcium carbonate - silica, by both thermogravimetric and differential thermal analysis and discussed their behaviour in relation to glass manufacture.(ii) Formates, acetates, oxalates and other oxycarbon anions-Copper11 f0rrnatel7~ and aluminium acetatelso have recently been investigated. However, oxalates have received a great deal of attention, viz.-those of beryllium,lsl magnesium,ls2 calcium5y80 981 zinc,67 cadmium,67 lead,67 mangane~e11,~7 cobalt,173 P~ scandium,lB yttrium,l83 lan- thanum,6s~93~143~183 other l a n t h a n o n ~ , ~ 8 ~ ~ ~ ~ 91849186 thoriumg3 and u r a n i ~ m , ~ ~ , ~ ~ ~ complex cobalt 0xalates,17~ oxalato-niobates1s7 and other complex oxalates.lss The additivity of the de- composition curves of a mixture of oxalates has been demonstrated,ls2 and the thermal decomposition pattern of oxalates in different atmospheres has received a t t e n t i ~ n .~ , ~ ~ ~ Other compounds of oxycarbon anions that have been studied include potassium hydrogen phthalate,lgO caesium propionate, butyrate, and isovaleratelgl and lanthanum and cerium111 palmitate, laurate and stearate.lg2 (iii) Silicates-The dehydration behaviou r of several silicates and silicate minerals has been investigated.lg3 to 198 The use of thermogravimetry combined with infrared studies has been suggested to permit a distinction to be made between constitutional and adsorbed water.199 The effect of alkaline-earth chlorides on the dehydration of silica Xerogel has been studied.200 (iv) Others-Thermogravimetry has been used to determine the conditions under which calcium orthoplumbateIv is formed.201 Anions derived from elements of Group V- (i) Nitrogen-The reaction between boron oxide and sodamide to yield boron nitride202 and the oxidation of aluminium nitride203 have been studied thermogravimetrically. The thermogravimetric behaviour of c e r i ~ m I I 1 , ~ ~ ~ praseodymium, neodymium, samarium,l72 thorium,205 ,206 plutonium146 and nicke1207 nitrates has been investigated.Lead azide has been studied.208 (ii) Phosphorus-The resistance to oxidation of some transition metal phosphides209 has been described. Phosphates of ammonium,2.L0 sodium,211 beryllium,212 stron- tium,214 aluminium,215 y 2 1 6 antimony,217 chromium166 and iron,215 y216 and some halophosphates have been studied.21s y2I9 ,220 The decomposition of disodium dihydrogen pyrophosphate221 and the reactions between magnesium pyrophosphate and strontium oxide or magnesium hydroxide222 have been reported.Gerrard, Mooney and Rothenbury have used thermogravimetry in their investigation of polymers formed from chloroborazoles and phosphorus esters. 223 and arsenates of chromium,58 cobalt225 and nickeP25 have been investigated thermogravimetrically. Ammonium metavanadate,226 some niobates and t a n t a l a t e ~ ~ ~ ' and some 0xalato-niobatesl8~ have also been studied. (iii) Other anions-ArsenicDecember, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 917 Anions derived from elements of Group VI- (i) Oxides, peroxides and related compounds-Thermogravimetric studies of oxides may be sub-divided into two main sections: (a) Dehydration reactions of hydrated oxides and subsequent disproportionation reactions- M,O,.nH,O(,) - M30y(,) + nH20(g)7j228 to 235 (b) Solid-state reactions, which may be either solid - solid reactions involving loss of vola- t i 1 e ~ , 8 , ~ 7 6 , ~ 7 7 y ~ 7 8 y ~ o ~ ~ ~ ~ ~ 9237923*3239 or solid - gas reactions involving loss or gain in weight,76s240j241 have also been reported.The thermal dehydration of copper,242 beryllium,lsl magnesium,54 iron111 243 and nickellg3 hydroxides has been investigated thermogravimetrically . On the basis of thermogravimetric and infrared methods several metal oxyhydrates have been classified into four (ii) Sulphides, sulphates and other anions of szzlphur-The oxidation and thermal decom- position of ~opper,~392~~ zinc,Z45 ~admium,2~~ mercury,245 germanium,246 tin,245 lead,245 and various nickel s ~ l p h i d e s ~ ~ ~ ? ~ ~ ~ p249 has been described, and there have been two studies on p y r i t e ~ .~ ~ 3 ~ ~ Sulphates of the metals listed below have been studied : ~ ~ p p e r , ~ ~ ~ y ~ ~ ~ beryllium,181 mag- nesium,252 calcium,253 mercury,245 indium,255 tin,245 lead,245 nickel,2E6 ~ranium,~~7 lanthanum,258 9259 and other l a n t h a n o n ~ , ~ ~ ~ 9259 as well as several double and com- plex to 263 Solid - solid264 and solid - gas265 reactions involving sulphates have also been reported. Rocchiccioli266 has studied the decomposition of sulphamic acid and some metal sulphamates.(iii) Anions derived from the other elements of Group VI-The selenides of arsenic, rhenium and mercury have been shown to sublime at definite temperatures.141 The decomposition of magnesium selenates has been compared to the corresponding ~ u l p h a t e s . ~ ~ ~ The thermal decomposition of neodymium and basic yttrium chromate has been st~died,~67 as well as some isopolychromates of potassium.268 Tungstic acid,269 ammonium parat~ngstate,~69 several 12-heteropoly t u n g s t a t e ~ , ~ ~ 994 ammonium53 and quinolinium molybdopho~phatel~~ have been investigated thermogravimetrically. Anions derived from elements of Group VII-Thermogravimetric studies of mono-amine dichloro zinc270 and basic aluminium chlorides271 have been made.A study of the application of high temperature thermogravimetry of chlorides and sulphates to soil analysis has been rep0rted.~7~ C ~ T , . X N H , ~ ~ ~ and some tetra-i~tlomercurates~~~ have been investigated. The thermal decomposition of the following compounds involving oxyanions has been reported : sodium barium br0mate~7~ and ~ h l o r a t e ~ ~ ~ and copper11 iodate,276 together with the perchlorates of potassium,48 magnesium,277 calcium277 and b a r i ~ m . ~ 7 ~ 9277 Binary systems of potassium perchlorate with either alkaline or alkaline-earth metal nitrates,278 and the course of the reaction between potassium chlorate and manganese dioxide279 have been followed by using t hermogravimetric techniques. Barium permanganate280 and ammonium perrhenate281 and the compounds AgMn204 and Ag2Mn0Zs2 have been studied by using a thermobalance.Anions derived from elements of Groq5 VIII-Alkali copper,284 magnesium285 and alkaline-earth285 ferrocyanides have been examined by Seifer. Work on ammonium chloro- platinate286 and bromoplatinic has also been reported. Complex inorganic compowds-Complex compounds of cobalt, chromium, nickel and platinum in which the ligand groups are wholly or partly either ammonia, ethylenediamine, propylenediamine or pyridine have been widely studied by Wendlandt and his ~ ~ h 0 o 1 ~ ~ 9 ~ ~ ~ ~ ~ ~ ~ to 292 and also by others.167 92939294. The 8-hydroxyquinolines of many bivalent metal~,~g~ y296 uranium297 and plutoniumf45 have been studied. Bivalent metal anthrani- lates,298 substituted anthranilates of lanthanum and and other related chelate com- plexess8 y3O0 Y3O1 have all been studied thermogravimetrically. Other work on complexes has been carried out by Wendlandt and C O - W O ~ ~ ~ ~ S , ~ ~ ~ ~ ~ ~ ~ ~ ~ Charles305 and Dhar and Basolo.306 Clathrates of the type Ni(CN),.NH, X, where X = benzene, thiophen, pyrrole, furan or have been examined by using a quartz-spring thermobalance.918 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS [Analyst, Vol.88 ORGANIC CHEMISTRY- The application of thermogravimetric analysis to organic compounds is best discussed under two headings. Synthetic organic polymers-These have been much studied in recent years, particularly with the advent of apparatus in which the sample can be heated in an inert atmosphere or under vacuum.Since the breakdown of polymers into the volatile monomer units is nearly always a simple single-stage process, these reactions have been used as standards in non- isothermal kinetics (see “Kinetic Studies”) . 6 4 p 8 2 9 9 6 9 9 7 y g 9 H. C. Anderson has compared differen- tial thermogravimetry with differential thermometry as a method for studying the pyrolysis of p~lymers.~Os Recently Doyle309 has derived equations from which it is possible to determine the isothermal life of polymers from thermogravimetric data. Table I lists some polymers that have been studied by a thermogravimetric technique. TABLE I SYNTHETIC POLYMERS STUDIED BY THERMOGRAVIMETRY References- A I > Polymer Kinetics studied Kinetics not studied Epoxide polymers . . .. . . . . . . Polyamides . . . . . . . . . . . . Polyethylene . . , . . . . . 9 . . . Polymers based on methyl methacrylate . . Polypyromellitimides . . . . . . . . Polystyrene . . .. . . . . . . . . Poly(t-butyl acrylate) . . . . . . . . Polytetrafluoroethylene . . . . . . . . Poly(viny1 alcohol), PVA . . .. . . . . Poly(viny1 chloride), PVC . , .. . . . . Other vinyl polymers . . .. . . . . Phenolic polymers . . . . . . . . . . Urea - formaldehyde resins . . . . . . . . . . Silicones . . . . . . . . . . . . 97 311 64,98 99 64 313 82,96 47,315 316 - 65,310 312 52 63 - - - 310, 314 62, 310 317 318 to 320 310 32 1 - Other organic compounds-Few simple organic compounds have been studied. D ~ v a 1 ~ ~ ~ has investigated the thermal stability of some organic compounds used as analytical standards.Other compounds studied include EDTA,151 barbituric acid and related compounds.153 Some organo-tin compounds have been investigated by therm~gravimetry.~~~ The pyrolysis behaviour of several coals, peats and bitumens has been to 327; their kinetics of decomposition has also received attention.P3 y328 The mechanism of thermal decomposition of some organo-montmorillonites has been in\.e~tigated.~~~ Thermogravimetric and diff eren- tial thermal analytical studies have been made of the pyrolysis of wood and of wood treated with inorganic salts. Some conclusions regarding the action of salts that are flame retardant are d i s c u ~ s e d . ~ ~ , ~ ~ ~ There have been two studies on inclusion or clathrate compounds. McAdiea2 has studied the urea - n-paraffin inclusion compounds, and Gilford and Gordon67 have examined the behaviour of some quinol clathrates under ambient and reduced pressures.COMPLEMENTARY PROCEDURES We have reviewed the various applications of thermogravimetric analysis, but there has been little stress on the fact that, in many of the papers referred to, other techniques have been used to complement the information gained from thermogravimetry. In certain investigations, it is essential to use a complementary technique if the interpretation of the course of a thermal decomposition is not to be pure guess-work; at least, chemical analysis of the solid removed from the thermobalance at an appropriate point on the thermo- gravimetric curve58 or, alternatively of the evolved gas316 should be carried out.Information gained from thermogravimetry becomes more meaningful or can readily be extended by the use of techniques that are applicable to solid-state chemistry.25 I t is not the object of this review, however, to discuss these procedures in detail but to offer a few pertinent comments. DIFFERENTIAL THERMAL ANALYSIS- This method is taken first, owing to its close affinity to thermogravimetry (see Intro- duction, p. 906). Both are dynamic methods, and both are very much dependent on pro- cedural detail^.^ It is not true that what is good technique for thermogravimetric analysisDecember, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 919 is necessarily so for differential thermal analysis. There is no doubt that much useful and meaningful information can be obtained with equipment designed to carry out both thermo- gravimetry and differential thermal analysis at one and the same time on the same sample- the success of the D e r i ~ a t o g r a p h ~ , ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 3 is proof of this.Nevertheless, there are limitations. For example, in differential thermal analysis the sample is generally packed down in the crucible3 so that the energy changes occurring in the course of the reaction produce a “good” peak; hence, again, the use of higher rates of heating in differential thermal analysis. For quantitative thermogravimetry, however (particularly if evaluation of kinetic parameters of the reaction is to be attempted), both of these operations would not be considered good practice.Garn173 has discussed various aspects of this problem. An apparatus, such as that described by Hodgson,l07 offers a compromise, in that both procedures are available to the operator but as completely independent operations. The use of the two procedures permits information to be obtained on all phenomena listed on p. 906. Examination of a material by both procedures permits crystalline transitions, second-order transitions and solid - solid reactions occurring without weight change to be distinguished from a typical decomposition reaction, or indeed any of the phenomena listed on p. 906 that give rise to a weight change. OTHER PROCEDURES- In many of the studies discussed, one of the main uses of thermogravimetry has been the investigation of the thermal stability of the materia1,13,149,322 or to find a suitable drying temperature for the investigation of a weighing form of an element.143 to 146 However by using thermogravimetry together with infrared studies, X-ray diffraction21 or electron-diffraction studies, or magnetic measurements-say, by heating up to a suitable point on the thermo- gravimetric curve (up to a point where a reaction has occurred) and removing the sample for examination by one or more of these methods-information can be gained on what structural changes have occurred during the course of the reaction.Thus a combination of infrared studies and magnetic measurements with a detailed thermogravimetric examina- tion of the dehydration of hexa-aquochromiumIII phosphate and arsenate enabled Lukaszewski and Redfern5* P~ to propose a reaction scheme for the dehydration reactions. A thermogravimetric study, coupled with X-ray diffraction studies and a study of catalytic activity,7 of a hydrated ruthenium dioxide provided preliminary information on some structural changes occurring on heating.CONCLUSION It is evident from the volume of literature now appearing1 that both thermogravimetric and differential thermal analysis have found many varied applications both in analytical chemistry and in other fields. It is likely that the introduction of more sophisticated instru- mentation (with facilities for operating in inert, oxidising or otherwise controlled atmospheres) will lead to further developments, particularly in the fields of organic and inorganic materials and to the use of thermogravimetry directly as an analytical method.It is to be hoped that prospective workers in the field will realise the dynamic nature of the method and of its dependence on the factors outlined in the section on “Techniques.” 1. 2. 3. 4. 5. 6. 7 . 8. 9. 10. 11. 12. 13. 14. REFERENCES Redfern, J. P., Editor, “Thermal Analysis Review,” Stanton Instruments, London (published about Gordon, S., “Encyclopedia of Science and Technology,” McGraw-Hill Book Co. Inc., New York, hlackenzie, R. C., and Mitchell, B. D., Analyst, 1962, 87, 420. Wendlandt, W. W., Franke, P. H. jun., and Smith, J. P., Anal. Chem., 1963, 35, 105. Simons, E. L., and Newkirk, A. E., G.E.C. Report No. 63-RL-3263C, 1963. Haladjian, J., and Carpeni, G., Bull. SOC. Chim. France, 1956, 1679.Keattch, C. J., and Redfern, J. P., J . Less Common Metals, 1962, 4, 460. Wilburn, F. W., and Thomasson, C. V., j. Soc. Glass Tech., 1958, 42, 1 5 8 ~ . Paulik, F., Paulik, J., and Erdey, L., 2. anal. Chem., 1958, 160, 241. Wilburn, F. W., and Hesford, J. R., J , Sci. Instrum., 1963, 40, 91. Guiochon, G., Anal. Chem., 1961, 33, 1124. Newkirk, A. E., Ibid., 1960, 32, 1558. Doyle, C. D., Ibid., 1961, 33, 77. Pellon, J ., paper presented at the International Symposium on Inorganic Polymers, Nottingham, four times a year). Toronto and London, 1960, p. 556. 1961.920 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS [Aadyst, vol. 88 Duval, C., Chim. Anal., 1962, 44, 191. Nernst, W., and Riesenfeld, E. H., Ber., 1903, 36, 2086. Wache, X., Mem. Sci.Rev. Met., 1961, 58, 887. Duval, C., “Inorganic Thermogravimetric Analysis,” Second Edition, Elsevier Publishing Com- Gordon, S., and Campbell, C., AnaZ. Chem., 1960, 32, 2 7 1 ~ . Lewin, S. Z., J . Chem. Educ., 1962, 39, A575. Rocchiccioli, C., Mikrochim. Acta., 1962, 1017. Wilson, C. L., J . Roy. Inst. Chem., 1959, 83, 550. Lukaszewski, G. M., and Redfern, J. P., Lab. Practice, 1961, 10, 469. -- , Ibid., 1961, 10, 552. , Ibid., 1961, 10, 630. , Ibid., 1961, 10, 721. Cleveland, Ohio, April, 1960. Los Angeles, California, April, 1963. pany, Amsterdam, London and New York, 1963. , ~- , -- Otsumo, Y., Bunseki Kagaku, 1962, 11, 999. Symposium on Thermal Methods of Analysis, 137th Meeting of the American Chemical Society, Symposium on Thermal Methods of Analysis, 144th Meeting of the American Chemical Society, Thermal Analysis Review, Abstr.No. 351, Stanton Instruments, London. Garn, P. D., Geith, C. R., and DeBala, S., Rev. Sci. Instrum., 1962, 33, 293. Lukaszewski, G. M., Nature, 1962, 194, 959. Peters, H., and Wiedemann, H. G., 2. anorg. Chem., 1959, 298, 202. Duval, C., Mikrochim. Acta, 1958, 705. Simons, E. L., Newkirk, A. E., and Aliferis, I., Anal. Chem., 1957, 29, 48. Peters, H., and Wiedemann, H. G., Z . anorg. Chem., 1959, 300, 142. Wiedemann, H. G., Ibid., 1960, 306, 84. Richer, A., Inst. Rech. de la Sidei.ugie, 1960, Series A, No. 187. Cueilleron, J., and Hartmanshenn, O., Bull. SOC. Chim. France, 1959, 172. Mauras, H., Ibid., 1960, 260. Saito, H., Proc. Imp. Acad. (Japan), 1926, 2, 58. Longchambon, H., Bull. SOC.franc. Mine‘r., 1936, 59, 145. Razouk, R. I., Farah, M. Y., Mikhail, R. S., and Kolta, G. A., J . Appl. Chem., 1962, 12, 190. Berlin, A., and Robinson, R. J., Anal. Chim. Acta, 1962, 27, 50. Taskaev, N. D., and Korotkov, L. S., Trud. Inst. Energ. i Vodnogo Khoz., Akad. hTauk Kirg. SSR., Reisman, A., Anal. Chem., 1960, 32, 1566. Kaesche-Krischer, B., and Heinrich, H. J., C‘hem.-Ing.-Tech., 1960, 32, 598. Anderson, D. A., and Freeman, E. S., Nature, 1962, 195, 1297. Heron, G. F., “Thermal Degradation of Polymers,” S.C.I. Monograph No. 13, 1961, 475. Paulik, F., Erdey, L., and Gal, S., Bergakadrmie, 1962, 14, 21. Moore, W. R., and Donnelly, E., paper given a t the Conference on Advances in Polymer Science Vassallo, D. A., Anal. Chem., 1961, 33, 1823. Wendlandt, W.W., Anal. Chim. Acta, 1959, 20, 267. Turner, R. C., Hoffman, I., and Chen, D., Canad. J . Chem., 1963, 41, 243. Garn, P. D., and Kessler, J. E., Anal. Chem., 1960, 32, 1563. _ _ _ _ , Ibid., 1960, 32, 1900. Paulik, F., Paulik, J., and Erdey, L., Acta Chim. Acad. Sci. Hung., 1961, 26, 143. Lukaszewski, G. M., and Redfern, J. P., J . Chem. SOC., 1962, 4802. Soulen, J. R., and Mockrin, I., Anal. Chem.. 1961, 33, 1909. Freeman, E. S., Anderson, D. A., and Campisi, J. J., J . Phys. Chem., 1960, 64, 1727. Newkirk, A. E., and Aliferis, I., Anal. Chem., 1958, 30, 982. Gilbert, J. B., Kipling, J. J., McEnaney, B., and Sherwood, J. N., Polymer, 1962, 3, 1. Jones, J . I., Ochynski, F. W., and Rackley, F. A., Chem. & Ind., 1962, 1686. Anderson, D. A., and Freeman, E.S., J . Polym. Sci., 1961, 54, 253. Anderson, H. C., J . Appl, Polynz. Sci., 1962, 6, 484. Leicester, J., and Redman, M. J., J . Appl. Chem., 1962, 12, 357. David, R., Bull. SOC. Chim. France, 1960, 719. Glasner, A., and Steinberg, M., J . Inorg. Nuclear Chem., 1961, 22, 39. Gilford, M., and Gordon, S., U.S. Dept. Com. Office Tech. Serv., P.B. Rept., 145517, 1960, 23 pp. Gorshkov, V. S., and Khmelevskaya, T. A., Sborn. Trud. Vses. Nauchn.-Issled. Inst. Materialov, Jjozhenov, P. I., Izv. Akad. Stroitel i Arkhitekt. S S S R . , 1961, No. 1, 56. Kasaoka, S., and Murata, Y . , J . Chem. SOC. Japan, I n d . Chem. Sect., 1961, 64, 986. Markowitz, M. M., and Boryta, D. A., Anal. Chem., 1961, 33, 949. Longuet, P., Silicates Indust., 1961, 26, 463. Bond, W. D., J . Phys.Chem., 1962, 66, 15‘73. Margotin, P., Stuckens, W., and Durand, R., Compt. Rend., 1961, 252, 4005. Satava, V., Silikdty, 1962, 6, 16. Paidassi, J., Pointud, M. L., Caillat, R., and Darras, R., J . Nucl. Muter., 1961, 3, 162. Freeman, E. S., and Carroll, B., J . Phys. Chem., 1958, 62, 394. Freeman, E. S., Dissertation Abstr., 1961, 22, 92. 1959, 5, 109. and Technology, London, May, 1963. 1961, No. 4, 77. ~- , , J . Chem. Eng. Data, 1962, 7 , 586. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81.December, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 92 1 82.83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. Doyle, C. D., J . Appl. Polym. Sci., 1961, 5, 285. Krevelen, D. W. Van, Heerden, C. Van, and Huntjens, F. J., Fuel, 1951, 30, 253. Schnitzer, M., and Hoffman, I., Chem. 6% Ind., 1961, 1397. Turner, R. C., and Schnitzer, M., Soil Sci., 1962, 93, 225. Anderson, D. A., and Freeman, E. S., J . Appl. Polym. Sci., 1959, 1, 192. Tassel, J. H. Van, and Wendlandt, W. W., J . Amer. Chem. SOL., 1959, 81, 813. Lumme, P., Suomen Kem., 1959, 32B, 198, 237 and 241. Brown, D.H., J . Chem. SOC., 1962, 3189. Brown, D. H., and Mair, J. A., Ibid., 1962, 3946. Jacobs, T., and Fre, M. der, Silicates Indust., 1961, 26, 363. Barrer, R. M., and Bratt, G. C., J . Phys. Chem. Solids, 1959, 12, 130. Padmonbham, V. M., Saraiya, S. C., and Sundaram, A. K., J . Inorg. Nuclear Chem., 1960, 12,356. Brown, D. H., J . Chem. SOC., 1962, 4408. Bear, J. I,., and Wendlandt, W. W., J . Inorg. Nuclear Chem., 1961, 17, 286. Anderson, H. C., Alakromol. Chem., 1962, 51, 233. -, Kolloidzschr., 1962, 184, 26. Smith, D. A,, paper presented a t the Conference on Advances in Polymer Science and Technology, Horowitz, H. H., andMetzger, G., Anal. Chem., 1963, 35, 1464. Coats, A. W., and Redfern, J. P., paper presented by J. P. Redfern a t the SIMA Exhibition, Jacobs, P. W.M., and Tomkins, F. C., in Garner, W. E., Editor, “Chemistryof the Solid State,” Garner, W. E., Editor, op. cit., Soulen, J . R., Anal. Chem., 1962, 34, 136. Technical literature of Robert L. Stone Co., Austin, Texas, U.S.A. Technical literature of Linseis KG. Selb/Bavaria, West Germany. Technical literature of Technical Equipment Corp., Denver, Colorado, U. S. A. Hodgson, A. A., J . Sci. Instrum., 1963, 40, 61. Technical literature of Stanton Instruments, London, England. Woerner, P. F., and Wakefield, G. F., Rev. Sci. Instrum., 1962, 33, 1456. Bottei, R. S., and Gerace, P. L., J . Chem. Educ., 1961, 38, 568. Bretsznajder, S., Kawecki, W., and Raabe, W., Przem. chem., 1961, 40, 443. Wendlandt, W. W., J . Chem. Educ., 1961, 38, 566. Hashimoto, H., and Ito, M., J . Chem.Soc. Japan, I n d . Chem. Sect., 1961, 64, 1515. Madorsky, S. L., Vacuum Microbalance Tech., 1962, 2, 47. Blazek, A., and Halousek, J., Silikdty, 1962, 6, 100. Head, E. L., and Holley, C. E., U.S. A t . Energy Comm. LA-2691, 1962, 20 pp. Baker, E. H., J . Chem. Soc., 1963, 339. Biermann, W. J., and Heinrichs, M., Canad. .J. Chem., 1962, 40, 1361. Gal’perin, L. N., Russ. .J. Phys. Chem., 1962, 36, 1244. Imris, P., Silikcity, 1962, 6, 91. Pascalau, M., Ursu, I., and Bologa, M., Acad. Rep. Populare Rounine, Studii Cercetari Fiz., 1962, Eastwood, F. W., Proc. Roy. Aust. Chem. Inst., 1962, 29, 405. Wendlandt, W. W., Anal. Chem., 1962, 34, 1726. Markowitz, M. M., and Boryta, D. A., Ibid., 1960, 32, 1588. Formanek, Z., and Dykast, J., Silikdty, 1962, 6, 113.Kruger, J. E., and Bryden, J . G., J . Sci. Instrum., 1963, 40, 178. Isensee, H. J., 2. anal. Chem., 1962, 186, 357. Kobayashi, M., Sci. Repts. TGhoku I m p , Univ., 1940, 29, 391. Duval, C., “Inorganic Thermogravimetric Analysis,” First Edition, Elsevier Publishing Co., Amsterdam, London and New York, 1953. Erdey, L., translated by Svehla, G., “Gravimetric Analysis,” Pergamon Press, Oxford, London, New York and Paris, 1963. International series of monographs on Analytical Chemistry, Volume VII, edited by Belcher, R., and Gordon, L. London, May, 1963. Moscow, May, 1963. Butterworths & Co. (Publishers) Ltd., 1955, Chap. 7, p. 184. 13, 349. Szmal, Z., Wiadomdsci Chemi., 1954, 8, 241. Duval, C., Mise au Point, 1958, 6, 19. Wendlandt, W. W., and Brabson, J.A., Aizal. Chem., 1958, 30, 61. Claisse, F., East, F., and Abesque, F., Province of Quebec, Department of Mines, P.R., 305, Clercq, M. de, and Duval, C., Anal. Chim. Acta, 1951, 5, 282. Flagg, J . F., and Furman, N. H., I n d . Eng. Cheun., Anal. Ed., 1940, 12, 663. Rynasiewicz, J., and Flagg, J. F., Anal. Chem., 1954, 26, 1506. Beamish, F. E., and McBryde, W. A. E., in Wilson, C. L., and Wilson, D. W., Editors, “Compre- hensive Analytical Chemistry,” Elsevier Publishing Co., Amsterdam, London and New York, 1959, 455. Wendlandt, W. W., and Hoffman, W. M., Anal. Chem., 1960, 32, 1011. Taimni, I. K., and Tandon, S. N., Anal. Chim. Acta, 1960, 22, 34 and 553. Taimni, I. K., and Rakshpal, R., Ibid., 1961, 25, 438. Sklyarenko, Yu. S., Sklyarenko, I. S., and Chubukova, T.M., Zhur. Anal. Khim., 1961, 16, 417. Liang, S.-C., and Li, I., Sci. Record. (Peking), 1958, 2, 197. Iiang, S.-C., and Mai, H.-C., Ibid., 1959, 3, 301. 1954, 6 pp.922 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. COATS AND REDFERN: THERMOGRAVIMETRIC ANALYSIS [Anabsi!, Vol. 88 Sklyarenko, I. S., and Chubukova, T. M., Zhur. Anal. Khim., 1960, 15, 706. Waterbury, G. R., Douglass, R. M., and Metz, C. F., Anal. Chem., 1961, 33, 1018. Fauth, M. I., Ibid., 1960, 32, 655. Duval, C., and Wadier, C., Anal.Chim. Acta, 1960, 23, 257 and 541. Duval, C., Mikrochim. Acta, 1962, 947. Belcher, R., Erdey, L., Paulik, F., and Liptay, G., Talanta, 1960, 5, 53. Wendlandt, W. W., Anal. Chem., 1960, 32, 848. Wendlandt, W. W., Tassel, J. H. Van, and Horton, G. R., Anal. Chim. Acta, 1960, 23, 332. Berlin, A., Taylor, M. E., and Robinson, R. J., Ibid., 1961, 24, 427. Schnitzer, M., Wright, J. R., and Hoffman, I., Anal. Chem., 1959, 31, 440. Wright, J. R., Hoffman, I., and Schnitzer, PI., J . Sci. Food Agric., 1960, 11, 163. Hoffman, I., Schnitzer, M., and Wright, J . R., Ibid., 1960, 11, 167. Dupuis, T., and Dupuis, J., Mikrochim. Actu, 1958, 186. Erdey, L., Paulik, F., Svehla, G., and Liptay, G., 2. anal. Chem., 1961, 182, 329. Erdey, L., Liptay, G., Svehla, G., and Paulik, F., Talanta, 1962, 9, 489.Berlin, A., and Robinson, R. J., Anal. Chim. Acta, 1961, 24, 224. _ _ - , Ibid., 1961, 24, 319. Kin;, Vu Q., Bull. S O ~ . Chim. France, 1962, 1481. Brill, O., 2. anorg. Chem., 1905, 45, 275. Chong, C. V. Y., Thesis, University of London, 1961. Wendlandt, W. W., Anal. Chem., 1956, 28, 1001. Lukaszewski, G. M., and Redfern, J . P., J . Chem. SOC., 1963, 3790. Chong, C. V. Y., Redfern, J. P., and Salmon, J. E., Ibid., submitted for publication. Bouaziz, R., Bull. SOC. Chim. France, 1962, 1451. Lavanant, F., Compt. Rend., 1962, 255, 1397. Manabe, T., Kawada, N., and Nishiyama, M., Semento Gijzttsa Nempo, 1961, 15, 54. Wendlandt, W. W., and George, T. D., Texas J . Sci., 1962, 13, 316. Wilfong, R. L., Domingnes, L. P., and Furlong, LeRoy R., U.S.Bur. Mines Rept. Invest., No. 6060, Garn, P. D., Anal. Chem., 1961, 33, 1247. Spitsyn, V. I., Komissarova, L. N., Shatskii, V. M., and Pushkina, G. Ya., Zhur. Neorg. Khim., Wendlandt, W. W., George, T. D., and Krishnamurty, R., J . Inorg. Nuclear Chem., 1961, 21, 69. Voskresenskaya, N. K., and Berul, S. I., Zhur. Neorg. Khim., 1962, 7, 850. Keler, E. K., and Kuznetsov, A. K., Zhur. Prikl. Khim., 1961, 34, 2146. Wilburn, F. W., and Thomasson, C. V., Phys. Chem. Glasses, 1960, 1, 52; and 1961, 2, 126. Schuffenecker, R., Trambouze, Y., and Prettre, M., Ann. Chim. (Paris), 1962, 7, 127. Maksimov, V. N., Semenenko, K. N., Naumova, T. N., and Novoselova, A. V., Zhur. Neorg. Khim., Rau, R. C., Adva?e. X-ray Anal., 1961, 4, 19. Wiedemann, H. G., and Nehring, D., 2.anorg. Chem., 1960, 304, 137. Savitskaya, Ya. S., Tvorogov, N. N., Kalabukhova, S. V., and Brykina, L. S., Zhur. Neorg. Khim., Glasner, A., and Steinberg, M., J . Inorg. Nuclear Chem., 1961, 22, 156. Shears, E. C., Trans. Brit. Ceram. SOC., 1962, 61, 225. Bressat, R., Claudel, B., and Trambouze, Y., Bull. SOC. Chim. France, 1963, 464. Vlasov, L. G., Lapiskii, A. V., and Strizhkov, B. V., Vestnik Moskov. Univ. Ser. I T : Khim., 1961, Dollimore, D., and Nicholson, D., J . Inorg. Nuclear Chem., 1963, 25, 739. Dollimore, D., Griffiths, D. L., and Nicholson, D., J . Chem. Soc., 1963, 2617. Newkirk, A. E., and Laware, R., Talanta, 1962, 9, 169. Plyushchev, V. E., Savel’Eva, M. V., and Shakhno, I. V., Russ. J . Inorg. Chem., 1962, 7, 1075. Misra, S. N., Misra, T.N., and Mehrotra, R. C., J . Inorg. Nuclear Chem., 1963, 25, 195. Wodtcke, F., Wencke, K., and Techel, G., 2. anorg. Chem., 1962, 317, 302, Uytterhoeven, J., and Fripiat, J. J., Bull. Soc. Chim. France, 1962, 788. Zhdanov, S. P., Zhur. Prikl. Khim., 1962, 35, 1620. Ushakova, E. N., Mineral Sbornik, L’vov. Geol. Obshch., L’vov. Godudarst. Univ. im. I . Franko, Jorkovsky, R., and Trzil, J., Silikdty, 1962, 6, 36. Gol‘dman, M. M., and Ni, L.P., Izv. Akad. Mauk Kaz. Ser. Met. Obogashchen i Ogneuporov, 1960, Fripiat, J . J., and Uytterhoeven, J., J . Phys. Chem., 1962, 66, 800. Henry, N., and Ross, R. A., J . Chem. SOC., 1962, 4265. Uhlir, Z., Sborn. Ved. Praci. Vysoka, Skola Chem.-Technol. Pardubice, 1960, No. 2, 143. Tagawa, H., and Itouji, O., Bull. Chem.SOC. Japan, 1962, 35, 1536. Guiochon, G., and Jacque, L., Bull. SOC. Chim. France, 1963, 836. Claudel, B., Trambouze, Y., and Veron, J., Ibid., 1963, 409. Claudel, R., and Trambouze, Y., Compt. Rend., 1961, 253, 2950. Busiere, P., Claudel, B., Renouf, J.-P., Trambouze, Y., and Prettre, M., J . Chim. phys., 1961, Weigel, D., Imelik, B., and Laffitte, P., Bull. SOC. Chim. France, 1962, 345. Stammler, M., Abel, J . E., and Kaufman, J. V. R., Nature, 1960, 185, 456. Ripley, R. L., J . Less Common Metals, 1962, 4, 496. 1962, 18 pp. 1960, 5, 2223. 1960, 5, 558. 1962, 7 , 2029. 16, 57. 1959, No. 13, 328. No. 1, 47. 58, 668.December, 19631 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS 923 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224.225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256. 257. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277. 278. 279. Erdey, L., Liptay, G., Gal, S., and Paulik, F., Periodica Polytech., 1961, 5 , 209. Kowalska, E., and Kowalski, W., Przem. Chem., 1962, 41, 73. Boulle, R., and Sallier-Dupin, A. de, Compt. Rend., 1962, 254, 122. Winand, L., and Brasseur, H., Bull. SOC. Chim. France, 1962, 1566. Ropp, R. C., and Aia, M. A., Anal. Chem., 1962, 34, 1288. Yvoire, F. D’, Bull. SOL. Chim. France, 1961, 1762, 2277 and 2283. Arlidge, E. Z., Farmer, V. C., Mitchell, B. D., and Mitchell, W. A., J . Appl. Chem., 1963, 13, 17.Robbins, D., J. Inorg. Nuclear Chem., 1961, 19, 183. _- , paper presented at a Symposium held by the ElectrocRemistry Society, Los Angeles, 1962. Vincent, J. P., Ann. Chim. (Paris), 1960, 5, 579. Kinh, Vu Q., Bull. SOC. Chim. France, 1962, 1486. Ichikawa, A., and Seiyama, T., J . Chem. SOC. Japan, I n d . Chem. Sect., 1962, 65, 1347. Duval, C., Magyar Tudomanyos Akad. Kem. Tud. Oszt. Kozlemen., 1961, 16, 255. Gerrard, W., Mooney, E. F., and Rothenbury, R. A., J . Appl. Chem., 1962, 12, 373. Blazek, A., Caslavska, V., and Frei, V., Coll. Czech. Chem. Commun., 1962, 27, 1014. Charles-Messance, B., and Woerth, A.-M., Bull. SOC. Chim. France, 1962, 574. Trau, J., Roczn. Chem., 1962, 36, 1365. Lapitskii, A. V., Nishanov, D., and Pchelkin, V. A., Vestnik Moskov.Univ. Ser. 11: Khim., 1960, Brindley, G. W., and Choe, J . O., Amer. Min., 1961, 46, 771. Jere, G. V., and Patel, C. C., 2. anorg. Chem., 1962, 319, 175. Kobayashi, I., Rikagaku Kenkyusho Hokoku, 1961, 37, 349. Koz’min, P. A., Zhur. Neorg. Khim., 1961, 6, 2635. Sato, T., Naturwissenschaften., 1961, 48, 693. Ukaji, R., and Minami, F., Nippon Genshiryoku Gakkaishi, 1961, 3, 260. Zemlicka, J., Sb. Vys. S k . Chem.-Technol., Odd. Fak. Anorg. Org. Technol., 1958, 269. Cordfunke, E. H. P., J . Inorg. Nuclear Chem., 1961, 23, 285. Felton, E. J., Juenka, E. F., and Bartram, S. F., Ibid., 1962, 24, 839. Cini, L., Ceramica, Roma, 1961, 16, No. 11, 60. Vydrevich, E. Z., and Galperin, E. L., Zhur. Prikl. Khim., 1961, 34, 1971. Bulgakova, T. I., and Zaitsev, 0. S., Vestnik Moskov.Univ. Ser. 11: Khim., 1961, 16. No. 4, 33. Kent, R. A,, and Eick, H. A., Inorg. Chem., 1962, 1, 956. Schafer, H., Schneidereit, G., and Gerhardt, W., Z. anorg. Chem., 1963, 319, 327. Frei, V., Coll. Czech. Chem. Comm., 1962, 27, 172. Erdey, L., Liptay, G., Gal, S., and Paulik, F., Periodica Polytech., 1961, 5, 287. Cabannes-Ott, C., Ann. Chim. (Paris), 1960, 5, 905. Hoschek, G., Monatsh. Chem., 1962, 93, 826. Galimov, M. D., and Okunev, A. I., Izv. Vysshikh Uchebn. Zavedenii, Tsvetn’ya Met., 1961, Pannetier, G., and Davignon, L., Bull. SOC. Chim. France, 1961, 2131. Pannetier, G., and Abegg, J . L., Compt. Rend., 1961, 252, 1613. Pannetier, G., Abegg, J. L., and Guenot, J., Bull. SOC. Chim. France, 1961, 2126. Blazek, A., Cisar, V., Caslavaska, V., and Caslavsky, J., Silikdty, 1962, 6, 25.Margulis, E. V., Russ. J . Inorg. Chem., 1962, 7, 935. Malard, C., Bull. SOC. Chim. France, 1961, 2296. Piece, R., Schweiz. min. petrogr. Mitt., 1961, 41, 303. Pannetier, G., Bregault, J. M., and Guenot, J., Bull. SOL. Chim. France, 1962, 2158. Deichman, E. N., Zhur. Neorg. Khim., 1961, 6, 1671. Rigault, G., Rend. SOC. Mineral. Ital., 1961, 17, 455. Golovnya, V. A., and Bolotova, G. T., Zhur. Neorg. Khim., 1961, 6, 566. Pannetier, G., and Dereigne, A., Bull. SOC. Chim. France, 1963, 1059. Nathans, M. W., and Wendlandt, W. W., J . Inorg. Nuclear Chem., 1962, 24, 869. Kuzin, I. A., and Galitskaya, I. A., Trud. Lenin. Teknol. Inst. im. Lensoveta, 1961, No. 55, 85. Pfeifer, T. F., Magyar Kbm. Fdly., 1962, 68, 156. Takakura, E., and Yasoshima, Y., J .Chem. SOC. Japan, I n d . Chem. Sect., 1960, 63, 1569. Wendlandt, W. W., and Sturm, E., J . Inorg. Nuclear Chem., 1963, 25, 535. Bosek, I. I., Novoselova, A. V., and Simanov, Yu. P., Zhur. Neorg. Khim., 1961, 6, 2563. Pannetier, G., and Abegg, J . L., Acta Chim. Acad. Sci. Hung., 1962, 30, 127. Rocchiccioli, C., Compt. Rend., 1962, 255, 1942. Schwarz, H., Z. anorg. Chem., 1963, 322, 129, 137. Spitsyn, V. I., Afonskii, N. S., and Tsviel’nikov, V. I., Zhur. Neorg. Khim., 1960, 5, 1505. Wanek, W., Silika’ty, 1962, 6, 70. Block, B. P., Florentine, R. A., Simkin, J., and Barth-Wehrenalp, G., J . Inorg. Nuclear Chem., Water-Levy, L., and Breuil, M., Compt. Rend., 1963, 256, 1286. Schnitzer, M., Wright, J. R., Hoffman, I., Anal. Chim. Acta, 1962, 26, 371.Kirakosyan, A. K., Zhur. Neorg. Khim., 1960, 5, 1806. Heintz, E. A., J . Inorg. Nuclear Chem., 1961, 21, 64. Ishi, G., J. Chem. SOL. Japan, I n d . Chem. Sect., 1962, 65, 1011. Lumme, P., and Lumme, H., Suomen Kem., 1962, 35B, 129. Chudinova, L. I., Izv. Vysshikh. Uchebn. Zavendenii Khim. i Khim. Tekhnol., 1962, 5 , 357. Hogan, V. D., and Gordon, S., J . Chem. Eng. Data, 1961, 6, 572. Miravet, M. del C., and Rocchiccioli, C., Mikrochim. Acta, 1961, 485. 15, 18. 4, 94. 1962, 24, 371.024 COATS AND REDFERN : THERMOGRAVIMETRIC ANALYSIS [AnaZyst, Vol. 88 Hardy, A., Ann. Chim. (Paris), 1962, 7, 281. Gibart, P., Traore, K., and Brenet, J., Compt. Rend., 1963, 256, 1296. Rienacker, G., and Werner, K., 2. anorg. Chem., 1963, 320, 141. Seifer, G. B., Zhur. Neorg.Khim., 1962, 7, 1242. -, Ibid., 1962, 7, 482. -, Ibid., 1962, 7 , 2290. Pannetier, G., Abegg, J. L., and Georavovitch, N., Acta Chim. Acad. Sci. Hung,, 1960, 25, 205. Berg, L. G., Mochalov, K. N., Kurenkova, 1’. A., and Anoshina, N. P., Izvest. Kazansk. Filiala Wendlandt, W. W., and Bear, J. L., J . Phys. Chem., 1961, 65, 1516. Horton, G. R., and Wendlandt, W. W., Texas J . Sci., 1961, 13, 454. Wendlandt, W. W., and Bear, J. L., J . Inorg. Nuclear Chem., 1961, 22, 77. George, T. D., and Wendlandt, W. W., Ibid., 1963, 25, 395. Wendlandt, W. W., Texas J . Sci., 1962, 14, 264. Block, B. P., Ocone, L. R., and Soulen, J . K., J . Inorg. Nuclear Chem., 1960, 15, 76. Block, B. P., Roth, E. S., and Simkin, J., Ibid., 1960, 16, 48. Charles, R. G., Ibid., 1961, 20, 211. Charles, R. G., Perrotto, A., and Dolan, M. A., Ibid., 1963, 25, 45. Bordner, J., and Gordon, L., Talanta, 1962, 9, 1003. Lumme, P., Suomen Kem., 1959, 32B, 253. Dragulescu, C., Simonesu, T., Menessy, I., and Anton, R., Acad. Rep. Populare Romine, Baza Lumme, P., Suomen Kem., 1959, 32B, 261. Thomas, G., and Paris, R. A., Anal. Chim. .4cta, 1961, 25, 159. Morris, M. L., Dunham, R. W., and Wendlandt, W. W., J . Inorg. Nuclear Chem., 1961, 20, 274. Wendlandt, W. W., and Haschke, J., Nature, 1962, 195, 379. Wendlandt, W. W., and Horton, G. R., J . INorg. Nuclear Chem., 1961, 19, 272. Charles, R. G., J . Phys. Chem., 1961, 65, 568. Dhar, S. K., and Basolo, F., J . Inorg. Nuclear Chem., 1963, 25, 37. Leicester, J., Bradley, J. K., and Barr, R. G., Chem. & Ind., 1962, 208. Anderson, H. C., Nature, 1961, 191, 1088. Doyle, C. D., J . Appl. Polym. Sci., 1962, 6, 639. Navord Report O.T.S. 171685. Pied, J. P., Ann. Chim. (Paris), 1960, 5, 469. Paulik, J., Macskasy, H., Paulik, F., and Erdey, L., Plaste u. Kautsclz., 1961, 8, 588. Schaefgen, J . K., and Sarasohn, I. M., J . Polym. Sci., 1962, 58, 1049. Goldfarb, I. J., McHenry, R. J., and Penski, E. C., Ibid., 1962, 58, 1283. Yamaguchi, T., Amagasa, H., and Uuchiyarna, S., Chem. of High Polymers, 1961, 18, 406. Guyot, A., and Benevise, J . P., J . Appl. Polym. Sci., 1962, 6, 489. Winslow, E. C., and Laferriere, A., J . Polym. Sci., 1962, 60, 65. Runavot, Y., and Schneebeli, P., Rech. ae‘ro., 1961, No. 80, 13. Anderson, H. C., SOG. Plastics Engrs. Trans., 1962, 2, 202. Jeffreys, K. D., Brit. Plast. (mould Prod. T r . ) , 1963, 36, 188. Sekine, Y., J. Chem. SOC. Japan, Ind. Chem. Sect., 1960, 63, 1657. Duval, C., Mikrochim. Acta, 1962, 268. Noltes, J . G., and Kerk, G. J . M. van der, Rec. Trav. chim. Pays-Bas, 1962, 81, 41. Abel, O., and Luther, H., Erdol u. Kohle, 1962, 15, 90. Bestougeff, M., Guiochon, G., and Jacque, L., Compt. Rend., 1962, 254, 266. Chantret, F., and Pouget, R., CEA Report No. 2069, 1961, 11 pp. Prabhakaram, P., and Murthy, H. P. S., Natl. Met. Lab. Tech. J . , 1960, 2, 15. Agroskin, A. A,, and Miringof, N. S., Naztch. Trud. Vses. Nauch.-Issled. Inst. Podzemn. Gasifek. Ramachondran, V. S., Garg, S. P., and Kacker, K. P., Chem. & Ind., 1961, 790. Browne, F. L., and Tang, W. K., U.S. Dept. of Agriculture; Forest Service, Report TP-118, Eickner, H. W., Forest Products J., 1962, 194. McAdie, H. G., Canad. J . Chem., 1962, 40, 2195. Akad. Nauk. S S S R , Ser. R h i m . Nauk, 1957, No. 4, 127. Cercetari Stiint. Timisoara, Studii Cercetari Stiinte Chim., 1961, 8, 9. Uglei, 1961, No. 4, 3. 1961, 22 pp. Received J u l y 27th, 1963 280. 281. 282. 283. 284. 285. 286. 287. 288. 289. 290. 291. 292. 293. 294. 295. 296. 297. 298. 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. 310. 311. 312. 313. 314. 315. 316. 317. 318. 319. 320. 321. 322. 323. 324. 325. 326. 327. 328. 329. 330. 331. 332.

ISSN:0003-2654    

DOI:10.1039/AN9638800906

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

December, 19631 ANALYTICAL METHODS COMMITTEE 925 Analytical Methods Committee REPORT OF THE ADDITIVES IN ANIMAL FEEDING STUFFS SUB-COMMITTEE PART 2." REPORT OF THE HORMONES PANEL The Determination of Stilboestrol and Hexoestrol in Compound Feeding Stuffs GENERAL INTRODUCTION THE Hormones Panel was set up under the chairmanship of Dr. R. E. Stuckey, and its membership was: Mr. J. Allen, Mr. L. Brealey, Mr. J. A. Potter and Mr. W. L. Sheppard, with Miss A. M. Parry as Secretary. The Panel was appointed to consider methods for determining synthetic hormones included in animal rations as stimulants. A number of synthetic hormones, both with an anabolic and an oestrogenic action, have been tried in practice as additives in animal feeding stuffs; they are stilboestrol, hexo- estrol, dienoestrol, dianisyl hexane, dianisyl hexene, dianisyl hexadiene and dienoestrol acetate.However, only two-stilboestrol and hexoestrol-were considered by the Panel to be sufficiently important to warrant the detailed investigation necessary; their structures are shown below- Stilboestrol H exo es t rol Although both these compounds possess phenolic groupings and can be assayed, when present in small amounts, by the use of a reagent that reacts with the phenol group, it was considered that some means of differentiating between the two compounds when present in feeding stuffs would be desirable. However, early in the consideration of the assay methods available for these compounds, it was apparent that the methods of choice for stilboestrol and hexoestrol would be different.In fact, although the two compounds are chemically similar, the methods recommended are quite distinct and will be described separately. STILBOESTROL Stilboestrol has been known for many years, and methods for its determination, both in pharmaceuticals and in biological materials, have appeared in the literature; most of the chemical methods are based on reactions involving the phenolic group. Methods proposed include an ultraviolet-absorption method,l Folin's colorimetric methods,2 a bromimetric m e t h ~ d , ~ a nitroso method: the irradiation method5 and the antimony pentachloride method.6 The main problem in the chemical assay of stilboestrol in feeding stuffs lies in the extraction of the stilboestrol in a sufficiently pure form. Cheng and Burroughs,' working on a pre-mix containing 500 mg of stilboestrol per lb of soyabean meal, and a low-potency cattle diet supplement containing 5 mg of stilboestrol per lb in the same meal, described methods for determining stilboestrol in feeds. In these methods the stilboestrol was isolated by chromatography on a column of Celite, and the purified stilboestrol determined by measure- ment of the red colour produced by reaction with antimony pentachloride in ethylene chloride ; alternatively, the extracted stilboestrol was determined by a method involving ultraviolet irradiation of a solution in 80 per cent.acetic acid and measurement of the resulting colour at 410mp. It was found that the colour in both methods obeyed Beer's law over the * Part 1 (Report of the Antibiotics Panel) appeared in The Analyst, 1963, 88, 835.926 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [AndySf, VOl.88 range 5 to 30 pg of stilboestrol, although the authors reported that the irradiation method did not give consistent results on the diet supplements, and the results with the antimony pentachloride method were high, probably because interfering substances in the feeds were not completely removed by chromatography. For concentrations of stilboestrol less than 5 mg per lb, the authors preferred a biological assay-a mouse-uterine weight method.8 In 1956 Hedger and M a n l e ~ , ~ after a study of several chemical assay methods, reported that they preferred the ultraviolet-irradiation method, and described a procedure with which they reported recoveries of 90 to 98 per cent.of known amounts of stilboestrol added to feeds. In 1958 Hanka and Lockhartlo reported that stilboestrol exerted a marked and specific inhibitory effect on the growth of Staphylococcus aureus strain H, and that this effect could be used for determining stilboestrol microbiologically by broth or disc-plate techniques. The Panel finally decided that a chemical assay would be more generally useful than a microbiological or biological method. The irradiation method was in regular use by one of the members of the Panel, and it was decided that this method, modified from that of Hedger and Manley, should first be tried. EXPERIMENTAL AND RESULTS The steps in the procedure used were- (i) mixing of the feed, grinding if necessary, and sampling for analysis; (ii) extraction of stilboestrol from the sample with chloroform ; (iii) washing of the chloroform extract with N sulphuric acid to remove impurities; (iv) extraction of the purified chloroform extract with sodium hydroxide solution ; (v) adjustment of the sodium hydroxide solution to pH 9.0 to 9.5 with 2 N ortho- phosphoric acid, and re-extraction of the stilboestrol with chloroform ; (vi) evaporation of the purified chloroform extract to dryness, dissolution of the residue in glacial acetic acid, and irradiation of the solution with ultraviolet light; (vii) measurement of the optical density at 420 mp and calculation of the stilboestrol present in the original feed by reference to a calibration graph.Collaborative tests by the method outlined above were first carried out on a sample of soyabean meal to which 11 p.p.m.(about 5 mg per lb) of stilboestrol had been added. In the first instance the Panel members who had not previously used the method experi- enced trouble at the irradiation stage. It was found that the wavelength of the light used to irradiate the solutions was critical, the shorter wavelength of 254 mp being satisfactory, whereas the colour of the solution was destroyed by radiation of wavelength 365mp. I t was also found essential to adhere closely to the distances and times specified in the method. After this had been realised and the correct lamp used, no difficulty was experienced in developing the maximum colour. The first results obtained are shown in Table I. TABLE I DETERMINATION OF STILBOESTROL IN SOYABEAN MEAL Sample contained 11 p.p.m.of stilboestrol Laboratory Stilboestrol found, A 8.8, 11.0 B 8.0 C 9.0 D p.p.m. - As an additional difficulty some samples of glacial acetic acid were found to be un- satisfactory, even when the procedure at the colour-development stage was carried out exactly as described in the method. Examination and fractionation of the acid suggested the presence, in some batches, of an inhibiting substance that was extremely difficult to remove. Careful fractionation sometimes produced a satisfactory product, but this did not always remain so on storage. Satisfactory samples of glacial acetic acid were circulated for use by the Panel members, and at this stage it was decided that a special test to assess the suitability of the acid used would have to be included in the reagent specification in the final method.December, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE.PART 2 927 A further collaborative test was carried out on another sample of compound feeding stuff to which 10 p.p.m. of stilboestrol had been added. Considerable difficulty was again experienced in preparing glacial acetic acid satisfactory for the development of the colour by irradiation. At this stage, however, the United States Pharmacopoeia XVI was pub- lished, and it was noted that, in the official assay for stilboestro1,ll phosphate buffer solution instead of glacial acetic acid was used as a solvent at the irradiation stage. A trial of this buffer solution showed that it was much preferable to glacial acetic acid and subject to none of its disadvantages.It was therefore agreed that this solvent should be substituted for glacial acetic acid at the colour-development stage. The results obtained in the collaborative test with the revised method are shown in Table 11. TABLE I1 DETERMINATION OF STILBOESTROL IN A COMPOUND FEEDING STUFF Laboratory A €3 C D Sample contained 10 p.p.m. of stilboestrol Stilboestrol found, Remarks p.p.m. (a) 8.1 8.9 ( b ) 8.8 Phosphate buffer solution used 7.3 (c) 8.8 9.4 9.7 ( a ) 8.4 9.1 ( b ) 8.8 Phosphate buffer solution used 8.4 Phosphate buffer solution used Glacial acetic acid used Phosphate buffer solution used on same extract as in (a) Phosphate buffer solution used (colour developed more quickly) Glacial acetic acid used (with a Hanovia lamp the colour was destroyed if the optimum time was exceeded) In view of the general agreement between the results and the possibility of incomplete incorporation of all the stilboestrol added during the preparation of the compound feeding stuff, it was considered by the Panel that this method was satisfactory and could be recom- mended; the method is described in detail in Appendix I.HEXOESTROL The determination of hexoestrol in pharmaceuticals and in biological materials has been described in the literature. The most common method and that described in the British Pharmaceutical Codex 1959 for determining hexoestrol in tablets depends on the measurement of the blue colour developed with sodium molybdophosphotungstate. The use of hexoestrol in compound feeding stuffs is not as widespread as that of stil- boestrol, and methods for determining hexoestrol specifically in feeding stuffs have not been described in the literature.Measurement of the colour developed with sodium molybdo- phosphotungstate is a sensitive method for determining hexoesterol, but the difficulty in applying it to feeding stuffs lies in isolating the hexoestrol in a sufficiently pure form. After consideration of the methods available, the Panel decided to rely on the experience of one of its members, whose laboratory had devised a method for the chromatographic separation of hexoestrol and its subsequent purification. EXPERIMENTAL AND RESULTS The steps in the method initially tried by the Panel, but which was subsequently modified, were- (i) mixing of the feed, grinding, sifting, and sampling for analysis; (ii) extraction of hexoestrol from the sample with chloroform ; (iii) washing of the chloroform extract with N sulphuric acid to remove impurities; (iv) extraction of the purified chloroform extract with sodium hydroxide solution; (v) adjustment of the sodium hydroxide solution to pH 9.0 to 9-5 with 2 N ortho- (vi) evaporation of a portion of the chloroform extract to dryness, and dissolution of phosphoric acid, and re-extraction of the hexoestrol with chloroform ; the residue in a mixture of tetrahydrofuran, triethylamine and water ;928 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [AnaZyst, vol.88 (vii) removal of phenolic impurities by reversed-phase chromatography on a column of oleated cellulose powder ; (viii) elution of the hexoestrol from the cellulose with ether, and then extraction of the solution with sodium hydroxide solution ; (ix) acidification of the sodium hydroxide solution and re-extraction of the hexoestrol with ether; (x) evaporation of the filtered ethereal extract, reaction with sodium molybdophospho- tungstate reagent, measurement of the optical density of the coloured complex at 750 mp and determination of hexoestrol by comparison with standards.Collaborative tests were carried out by the method outlined above on asample of soyabean In the first test, difficulties with the method were reported; the results obtained are meal to which 11 p.p.m. (about 5 mg per lb) of hexoestrol had been added.shown in Table 111. TABLE I11 DETERMINATION OF HEXOESTROL IN SOYABEAN MEAL Sample contained 11 p.p.m. of hexoestrol Laboratory Hexoestrol found, A 12.0, 10.4 B 7.2 C 8.8 D p.p.m. - Difficulties were experienced owing to the formation of emulsions during the sodium hydroxide - chloroform extraction stage and also to the fact that the final solutions for the colour measurement were often cloudy to different degrees. Incomplete purification, it was found, could also cause high results owing to the presence of extracted substances that reacted with the sodium molybdophosphotungstate. It was decided that, at some stages, the assay procedure ought to be more precisely described, particularly at the stages relating to the extraction and the chromatography, and that the reagent grades ought to be specified.A second collaborative test was carried out on a further sample of cattle feeding stuff to which 10 p.p.m. of hexoestrol had been added. During this second test one laborotory reported high results (about 15 p.p.m.), and it was suspected that fatty material from the chromatographic column might be responsible ; this interference was subsequently minimised by describing in detail each step of the purification stage, by specifying a purified oil for the preparation of the oleated cellulose, and by removing any cloudiness in the final extract by filtration through sand. High results were also attributed to impurities in the solvent, but these could be avoided by distilling the ether before use. Other members reported occasional high results, which, it was considered, would be avoided if the ether were freshly distilled.Further, it was considered that any cloudiness occurring in the final extract should be removed by filtration through sand. The results obtained by the collaborating laboratories in a final collaborative test with the method described in Appendix TI are shown in Table IV. TABLE IV DETERMINATION OF HEXOESTROL IN A COMPOUND FEEDING STUFF Sample contained 10 p.p.m. of hexoestrol Laboratory Hexoestrol found, p.p.m. A 8-3, 8.9, 8.6, 8.5 B 9.0 C 9-3, 9.2 D 10.9 'The results were considered to indicate that the method was satisfactory for determining hexoestrol in compound feeding stuffs, and could be recommended (see Appendix TI).December, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE.PART 2 Appendix I RECOMMENDED METHOD FOR DETERMINING STILBOESTROL IN COMPOUND FEEDING STUFFS PRINCIPLE OF METHOD- 929 The stilboestrol is extracted from the feeding stuff with chloroform, then taken into sodium hydroxide solution, and, after the solution has been adjusted to pH 9.0 to 0.5 with orthophosphoric acid, is re-extracted with chloroform. 'The purified chloroform extract is evaporated, and the residue dissolved in phosphate buffer solution. The optical density is measured at 418 mp before and after irradiation of the solution with ultraviolet light, and from the difference the amount of stilboestrol in the sample is calculated by reference to the optical-density difference when an extract to which a known amount of stilboestrol has been added is similarly treated.APPLICABILITY- (about 5 mg per lb) stilboestrol. The method is suitable for the analysis of feeding stuffs containing about 11 p.p.m. SPECIAL APPARATUS- Extraction upparatus-Any apparatus may be used provided it permits (i) the uniform percolation of the powdered feeding stuff with the extracting solvent and (ii) the regular flow of the solvent vapour around the percolator. A suitable form consists of a glass tube, about 4cm in internal diameter and 15 cm in length, with a piece of coarse filter-paper covered with fine calico firmly tied over the lower flanged end. This percolator rests on a glass spiral inside an outer tube, about 6 cm in internal diameter and about 25 cm in length, the spiral being supported on the shoulder of a B24 standard joint sealed to the lower end of the outer tube. This standard joint fits into a flask of suitable size to contain the solvent, and a reflux condenser is attached to the top of the tube.An alundum crucible of medium porosity and suitable size also forms a satisfactory percolator. Irradiation equipment-This consists of a mercury-discharge tube mounted horizontally in a suitable reflector, together with a holder capable of containing four silica cells, so arranged that the cells are perpendicular to the light source and fixed at 15cm (6 inches) from it. The holder is also positioned so that a line through the liquid centres in the cells is leveI with the centre of the discharged tube, and the cells themselves, spaced about 1 cm apart, are centrally located relative to the ends of the discharge tube.The Phillips Germicidal tube TUV (15 watt), together with a trough fitting, A7003, with the grill removed, is a suitable source of radiation. REAGEXTS- Washed sand. Chloroform-Analytical-reagent grade. The chioroiorm used must comply with the test described below. Shake 35 ml of the chloroform with 70 ml of water, and then allow the layers to separate. To 10 ml of the aqueous extract add 40 ml of distilled water and 2 ml of Nessler's reagent, and set aside in the dark for 15 minutes. No colour or turbidity should be produced. Sulfihuric acid, N. Sodiztm hydroxide, N. Orthophosphoric acid, 2 N. Sodium sdphate, anhydrous. Ethanol, absolute. Potassium phosfihate solution-A 1.8 per cent. w/v solution of dipotassium hydrogen Stock standard stilboestrol solution-Prepare a solution in chloroform to contain exactly Working standard stilboestrol solution-Dilute 10 ml of stock standard stilboestrol solution orthophosphate, K,HPO,, in distilled water. 0.55 mg of stilboestrol B.P.per ml. to 100 ml at 20" C with chloroform; 1 ml -- 55 pg of stilboestrol.$130 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [Analyst, Vol. 88 PROCEDURE EXTRACTION OF SAMPLE- Crush about 1 kg of the sample, mix and grind 100 g of the crushed material until not less than 95 per cent. passes a 30-mesh sieve. Weigh accurately about 40 g of the ground material, and mix with about l o g of washed sand. Transfer the mixture to a stoppered flask, add about 100 ml of chloroform, shake vigorously, and set aside overnight.Transfer the contents of the flask to the percolator of the extraction apparatus, collecting the chloro- form in the flask, and assemble the apparatus. Extract the solid material for 6 hours; use more chloroform if necessary. Filter the extract through a pad of cotton-wool into a 200-ml calibrated flask, washing the extraction flask and filter with chloroform, and finally dilute the combined extract and washings to 200ml with chloroform. PURIFICATION OF EXTRACT- Transfer a 25-ml portion of the solution from the extraction of sample, together with 25 ml of chloroform, to each of two separators; add exactly 1 ml of working standard stil- boestrol solution to the contents of one separator, and treat the contents of each separator as described below. Add 25 ml of N sulphuric acid, swirl gently for 30 seconds, avoiding as far as possible the formation of emulsions, allow the layers to separate for 10 minutes, and transfer the lower chloroform layer to another separator.If emulsions form at this stage, it may be helpful to spin the mixture in a centrifuge. Add 10 ml of chloroform to the aqueous acid layer, shake gently for 10 seconds, still taking care to avoid emulsification, allow the layers to separate and add the lower chloroform layer to the initial chloroform solution. Repeat this last operation with two further successive 10-ml portions of chloroform, shaking vigorously, and then discard the aqueous acid liquid. Shake the combined chloroform solutions carefully for 30 seconds with two successive 10-ml portions of N sodium hydroxide; set aside for 10 minutes each time before running ,off the lower chloroform layer.Combine the sodium hydroxide extracts, add 5 ml of chloroform, shake for 5 seconds, allow the layers to separate, and transfer the lower chloroform layer to the empty separator that previously held a sodium hydroxide extract. Repeat the extraction of the combined sodium hydroxide extracts with two or three successive 5-ml portions of chloroform until a chloroform layer is obtained that, after having been shaken and allowed to separate, is colourless; add each chloroform layer to the first one. Add 5 ml of distilled water to the combined chloroform washings, shake, allow the layers to separate, and run off and discard the chloroform layer. Transfer the washed alkaline solution and the aqueous washings to a 50-ml beaker; rinse the empty separator with successive small portions of distilled water until the washings are free from all alkalinity, and add the rinsings to the contents of the beaker.To the combined alkaline aqueous solution and washings, add 4 ml of 2 N phosphoric acid, and adjust the solution carefully to pH 9.0 to 9.5 with 2 N phosphoric acid; use a pH meter to make the measurements. Return the adjusted solution to the same separator, rinsing the beaker first with two successive 2-ml portions of distilled water and then with 15 ml of chloroform, and add the rinsings to the contents of the separator. Swirl the mixture carefully for 30 seconds taking care to avoid the formation of an emulsion, allow the layers to separate, and transfer the lower chloroform layer to another separator.Add to the chloroform extract 25 ml of distilled water, shake for 5 seconds, allow the layers to separate, and filter the lower chloroform layer through a 1-inch bed of anhydrous sodium sulphate in a sintered-glass funnel; collect the filtrate in a 50-ml calibrated flask. Repeat the extrac- tion of the alkaline aqueous solution with two successive 15-ml portions of chloroform; wash both the chloroform extracts with the same 25 ml of water, and then filter the washed extracts through sodium sulphate as described above. Shake the aqueous washings with successive small portions of chloroform ; use these chloroform extracts to wash the sodium sulphate and the funnel, and then add to the contents of the calibrated flask.Continue this process until the solution in the flask is adjusted to the mark, and mix. DETERMINATION OF STILBOESTROL- Transfer a 25-ml portion of the sample solution from the purification of extract to a 100-ml beaker previously rinsed with chloroform, and evaporate off the solvent with a gentle Discard the chloroform solution.December, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE. PART 2 93 1 current of air, warming the beaker on a water-bath until only the last traces of solvent remain. Dissolve the residue in 5ml of absolute ethanol, warming gently, if necessary, to effect dissolution, and add 5 ml of potassium phosphate solution. Measure the optical density at 418 mp of the solution against water in a 1-cm cell with a spectrophotometer.Transfer about 3 ml of the aqueous ethanolic solution to a silica cell, and irradiate with ultraviolet light for 10 minutes. Measure the optical density at 418mp of the irradiated solution against water in a 1-cm cell with a spectrophotometer. Re-irradiate the solution for successive 1-minute periods until the maximum optical-density reading is obtained. Repeat the irradiation on a further 3-ml portion of the aqueous ethanolic solution, irradiating for the period found previously to give the maximum optical density, and use this figure in the calculation. Treat, in exactly the same manner, a 25-ml portion of the solution from the purification of extract obtained from the extract to which a known amount of stilboestrol has been added. CALCULATION- Calculate the amount of stilboestrol present in the sample from the expression- A - u 55 8 -- x - x (B - b) - (A - a) where A = optical density of sample solution after irradiation, 2 Weight of sample P.P.m” a = optical density of sample solution before irradiation, B = optical density of sample solution with added stilboestrol after irradiation and b = optical density of sample solution with added stilboestrol before irradiation. Appendix I1 RECOMMENDED METHOD FOR DETERMINING HEXOESTROL I N COMPOUND FEEDING STUFFS PRINCIPLE OF METHOD- The hexoestrol is extracted from the feeding stuff with chloroform, then taken into sodium hydroxide solution, and, after the solution has been adjusted to pH 9.0 to 9.5 with orthophos- phoric acid, is re-extracted with chloroform.The purified chloroform extract is evaporated, the residue dissolved in an aqueous solution of tetrahydrofuran and triethylamine, and the hexoestrol separated by column chromatography. The hexoestrol is extracted with ether, the ethereal solution evaporated, the residue treated with a molybdophosphotungstic acid reagent, and the hexoestrol determined absorptiometrically. APPLICABILITY- 5 mg per lb) of hexoestrol. SPECIAL APPARATUS- The method is suitable for the analysis of feeding stuffs containing 10 p.p.m. (about Extraction apparatus-As described in the method for stilboestrol (see Appendix I, p. 929). Chromatographic tube-A glass tube, about 1 cm in internal diameter and about 15 cm in length, drawn out at one end for a length of about 2 cm to terminate in a jet, about 3 mm in internal diameter, is suitable. In such a tube the adsorbent can be supported on a pledget of cotton-wool. REAGENTS- Washed sand.Fine sand. Cellulose powder. Ether, redistilled-Anaesthetic ether B.P., freshly distilled. Arachis oil solution-A 5 per cent. v/v solution of arachis oil R.P., in redistilled ether. C hZoroform-Analytical-reagent grade. Sul$huric acid, N. Sodium hydroxide solution, N.932 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [Analyst, Vol. 88 Orthophosphoric acid, 2 N. Sodium sulphate, anhydrozcs. Tetrahydrojuran - triethylarnine solution-A mixture of 5 volumes of triethylamine, 25 Sulphuric acid, diluted-Mix carefully 1 volume of sulphuric acid, sp.gr. 1.83, with Ethanol, 95 per cent. v / v . Ethanol, diluted-Mix equal volumes of ethanol and distilled water.Hydrochloric acid, diluted-Mix 1 volume of hydrochloric acid, sp.gr. 1.16 to 1.18, with 9 volumes of distilled water. Molybdophosphotungstate reagent-Add 50 g of sodium tungstate, XaW04.2H,0, 12 g of molybdophosphoric acid, H3P04. 12MoO3.24H,O, and 25 ml of orthophosphoric acid (88 per cent. w/w) to about 350 ml of distilled water in a round-bottomed flask. Boil the mixture under reflux for 2 hours, cool, and dilute to 500ml with distilled water. Store in a well- stoppered bottle protected from light. volumes of tetrahydrofuran and 70 volumes of distilled water. 9 volumes of distilled water. Sodium carbonate solution-A 10 per cent. w/v solution in distilled water. Stock standard hexoestrol solution-Dissolve 50-0 mg of hexoestrol in diluted ethanol, Working standard hexoestrol solution-Dilute 5-0 ml of stock standard hexoestrol solution and make up to 100 ml at 20" C with diluted ethanol.to 50 ml at 20" C with diluted ethanol; 1 ml = 50 pg of hexoestrol. PROCEDU KE EXTRACTION OF SAMPLE- Crush about 1 kg of the sample, mix, and grind about 100 g of the crushed material until not less than 95 per cent. passes a 30-mesh sieve. Weigh accurately about 40 g of the ground material, and mix with about 10 g of washed sand. Transfer the mixture to a stoppered flask, add about 100 ml of chloroform, shake vigorously, and set aside overnight. Transfer the con- tents of the flask to the percolator of the extraction apparatus, collecting the chloroform in the flask, and assemble the apparatus.Extract the solid material for 6 hours; use more chloro- form if necessary. Filter the extract through a pad of cotton-wool into a 200-ml calibrated flask, washing the extraction flask and filter with chloroform, and finally dilute the combined extract and washings to 200ml with chloroform. PURIFICATION OF EXTRACT- Transfer a 50-ml portion of the solution from the extraction of sample to a separator, add 25ml of N sulphuric acid, swirl gently for 30 seconds, avoiding as far as possible the formation of emulsions, allow the layers to separate for 10 minutes, and transfer the lower chloroform layer to another separator. If emulsions form at this stage, it may be helpful to spin the mixture in a centrifuge. Add 10ml of chloroform to the aqueous acid layer, shake gently for 10 seconds, still taking care to avoid emulsification, allow the layers to separate, and add the lower chloroform layer to the initial chloroform solution.Repeat this last operation with two further successive 10-ml portions of the chloroform, shaking vigorously, and then discard the aqueous acid liquid. Shake the combined chloroform solutions carefully for 30 seconds with two successive 1O-d portions of N sodium hydroxide; set aside for 10 minutes each time before running off the lower chloroform layer. Combine the sodium hydroxide extracts, add 5 ml of chloroform, shake for 5 seconds, allow the layers to separate, and transfer the lower chloroform layer to the empty separator that previously held a sodium hydroxide extract. Repeat the extraction of the combined sodium hydroxide extracts with two or three successive 5-ml portion of chloroform until a chloroform layer is obtained that, after having been shaken and allowed to separate, is colourless; add each chloroform layer to the first one.Add 5 ml of distilled water to the combined chloroform washings, shake, allow the layers to separate, and run off and discard the chloroform layer. Transfer the washed alkaline solution and the aqueous washings to a 50-ml beaker ; rinse the empty separator with successive small portions of distilled water until the washings are free from all alkalinity, and add the rinsings to the contents of the beaker. Discard the chloroform solution.December, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE. PART 2 933 To the combined alkaline aqueous solution and washings add 4 ml of 2 N phosphoric acid, and adjust the solution carefully to pH 9.0 to 9-5 with 2 N phosphoric acid; use a pH meter to make the measurements.Return the adjusted solution to the same separator, rinsing the beaker first with two successive 2-ml portions of distilled water and then with 15 ml of chloroform, and add the rinsings to the contents of the separator. Swirl the mixture carefully for 30 seconds, taking care to avoid the formation of an emulsion, allow the layers to separate, and transfer the lower chloroform layer to another separator. Add to the chloroform extract 25 ml of distilled water, shake for 5 seconds, allow the layers to separate and filter the lower chloroform layer through a 1-inch bed of anhydrous sodium sulphate in a sintered-glass funnel; collect the filtrate in a 50-ml calibrated flask. Repeat the extraction of the alkaline aqueous solution with two successive 15-ml portions of chloroform ; wash both chloroform extracts with the same 25 ml of water, and then filter the washed extracts through sodium sulphate as described above.Shake the aqueous washings with successive small portions of chloroform; use these chloroform extracts to wash the sodium sulphate and the funnel, and then add them to the contents of the calibrated flask. Continue this process until the solution in the flask is adjusted to the mark, and mix. ISOLATION AND DETERMINATION OF HEXOESTROL- Preparation of adsorptioiz column-Mix 2 g of cellulose powder with 6 ml of arachis oil solution, add more ether, and mix thoroughly, evaporating off the ether during the mixing.Make the residual oleated cellulose into a thin slurry with tetrahydrofuran - triethylamine solution, transfer to the chromatographic tube, and allow to drain. Treatment of test solution-Transfer a 25-ml portion of the chloroform solution from the purification of the extract to a 100-ml beaker, and evaporate off the solvent with a gentle current of air, warming the beaker gently on a water-bath until only the last traces of solvent remain. Dissolve the residue in the minimum amount of tetrahydrofuran - triethylamine solution, and transfer the solution to the top of the prepared adsorption column. Allow the solution to percolate into the column until the surface of the liquid just disappears below the surface of the cellulose, stopper the upper end of the tube, and set aside for 1 hour. Transfer 10 ml of tetrahydrofuran - triethylamine solution to the top of the adsorption column, and allow the liquid to run through the column to remove impurities.When the flow has ceased, remove the residual liquid from the column by applying gentle suction at the bottom of the tube. Transfer the cellulose column from the tube to a separator, wash out the tube with a little ether, and add the washings to the contents of the separator. Add 1 ml of N sulphuric acid and 20 ml of ether, shake, allok the layers to separate, and transfer the upper ethereal layer to another separator. Repeat the extraction of the acidified cellulose suspension with two successive 20-ml portions of ether, adding the ethereal layers to the ethereal extract in the second separator.To the combined ethereal extracts, add 10 ml of N sodium hydroxide, shake, and allow the layers to separate. Transfer the lower aqueous layer to another separator containing 5 ml of ether, shake, allow the layers to separate and transfer the lower aqueous layer to a third separator. Repeat the extraction of the combined ethereal extracts, first with a further 10-ml portion, and then with three successive 5-ml portions, of N sodium hydroxide; wash each aqueous alkaline extract with the same 5 ml of ether, and discard the ethereal solution and washings. Combine the washed aqueous alkaline extracts, acidify with diluted sulphuric acid, add 10 ml of ether, shake, and allow the layers to separate.Transfer the upper ethereal layer to another separator containing 5 ml of distilled water, shake, allow the layers to separate and transfer the upper ethereal layer to a 50-ml flask. Repeat the extraction of the acidified aqueous extracts, first with a further 10-ml portion, and then with three successive 5-ml portions, of ether; wash each ethereal extract with the same 5 ml of water before it is added to the first extract in the flask. Filter the combined ethereal extracts through a 1-cm layer of fine sand, wash the filter with several small portions of ether and evaporate the combined filtrate and washings to a volume of about 5 ml by gentle warming. Transfer the concentrated solution quantitatively to a 14-ml calibrated centrifuge tube with the aid of more ether, and carefully evaporate the ethereal solution to dryness.Dissolve the residue in 0.5 ml of ethanol, add 0.5 ml of distilled water and then 0.4 ml of diluted hydrochloric acid, 0.8 ml of molybdophosphotungstate reagent and 5 ml of distilled water, mix well, and set aside for 10 minutes. Add 3 ml of sodium carbonate solution and934 ANALYTICAL METHODS COMMITTEE [Analyst, 1701. 88 sufficient distilled water to produce 12 ml, mix, set aside for 1 hour, and spin in a centrifuge for 15 to 20 minutes. Measure the optical density at 750 mp of the clear supernatant liquid with a spectrophotometer in a l-cm cell against a blank solution prepared by adding 0.4 ml of dilute hydrochloric acid, 0.8 ml of molybdophosphotungstate reagent and 5 ml of water to 1 ml of diluted ethanol, and continue as described above, commencing at the words “mix well, and set aside for 10 minutes . . .” Calculate the amount of hexoestrol in the solution by comparison of the optical density with that obtained by treating 0.8, 1-0 and 1-2 ml of working standard liexoestrol solution in the same manner, beginning with the addition of 0-4 ml of dilute hydrochloric acid in the previous paragraph; the optical densities of the test solution and of the chosen standard solution should not differ by more than 10 per cent. Hence calculate the amount of hexoestrol in the sample. 1 . 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. REFERENCES Elvidge, W. F., Quart. J , Pharm., 1939, 12, 347. Tubis, M., and Bloom, A., I n d . Eng. Chem.. Anal. E d . , 1942, 14, 309. Sondern, C. W., and Burson, C., Ibid., 1942, 14, 358. Gottlieb, S., J . Amer. Pharm. Ass., Sci. E d . , 1947, 36, 379. Munsey, V. E., J . Ass. Off. Agric. Chem., 1957, 40, 459. Dingemanse, E., Nature, 1940, 145, 825. Cheng, E., and Burroughs, W., J . A s s . Off. Agric. Chem., 1955, 38, 146. Evans, J. S., Varney, R. I?., and Koch, F. C., Endocrinology, 1941, 28, 747. Hedger, F. H., and Manley, D. K., “Proceedings of the Symposium on Medicated Feeds,” Wash- Hanka, L. J., and Lockhart, W. R., J . Bact., 1958, 75, 471. United States Pharmacoepia, XVIth Revision, p. 217. ington, D.C., 1956, p. 150.

ISSN:0003-2654    

DOI:10.1039/AN9638800925

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

December, 19631 ANALYTICAL METHODS COMMITTEE 935 PART 3. REPORT OF THE PROPHYLACTICS PANEL The Determination of Nitrofurazone in Compound Feeding Stuffs INTRODUCTION THE Prophylactics Panel was set up under the chairmanship of Dr. R. F. Phipers, and its membership was: Mr. C. W. Ballard, Mr. K. C. Brown, Dr. H. G. Dickenson, Mr. A. W. Hartley, Mr. A. Holbrook, and Mr. S. G. E. Stevens, with Miss A. M. Parry as Secretary; Mr. J. A. Stubbles and Mr. D. H. Mitchell were co-opted later. The Panel was appointed to consider methods of analysis for determining medicaments against coccidiosis and histo- moniasis included in poultry rations. The Panel turned its attention first to the determination of nitrofurazone, which is widely used as a poultry coccidiostat, and this report describes the results of collaborative tests that led to the recommended method being advanced.There is a considerable volume of published information on the analysis of nitrofurazone in animal feeding stuffs, and it was not expected that there would be any difficulty in making a recommendation. However, the Panel's investigations showed that this was not so, and the method finally evolved contains various recommendations that, if ignored, will lead to inaccurate results being obtained. Ells, McKay and Paul1 described an extraction method for separating nitrofurazone from feeding stuffs, and they compared the optical densities of such extracts with the optical densities of similar solutions in which the nitrofurazone had been reduced by treatment with a4ueous sodium dithionite solution.In recent years the complexity of the components used in the formulation of poultry feeding stuffs has increased; the introduction of grass meals and other vegetable materials yielding highly coloured extracts has rendered valueless simple methods of extraction and reduction. Buzard, Ells and Paul2 attempted to overcome these difficulties by allowing the nitrofurazone extracted from medicated feeding stuffs to react with phenylhydrazine hydrochloride and measuring the red colour produced. Collaborative trials recorded by Puglisi3 showed that this procedure was unreliable. van Zijl and Goosens4 claimed that they could overcome interference from reducing and other materials, by oxidation of the sample solution with potassium permanganate.Tagaki and Uno5 found that the addition of caustic alkali to nitrofurazone solutions led to an intensification of the colour, but the orange-red colour thus produced is unstable, and any determinations involving its use are difficult. Cross, Hendey and Stevens6 took advantage of some investigations by Porter,' who studied the colour reactions of certain nitro-compounds, and they adopted his use of dimethyl- formamide as a solvent. They found that the addition of alkali to solutions of nitrofurazone in dimethylformamide intensified the colour ; this was of obvious value in spectrophotometric work, but they noted that the colour of the resulting solution required stabilisation before it could be used on a quantitative basis. Phenol was found to confer the desired degree of stability. Cross, Hendey and Stevens6 extracted a typical medicated poultry feeding stuff, i.e., one containing about 0.01 per cent.of nitrofurazone, with light petroleum in a Soxhlet apparatus. Nitrofurazone is insoluble in light petroleum, and this first step permitted much interfering material to be removed. This extraction was followed by a similar treatment with carbon tetrachloride, in which solvent nitrofurazone also is essentially insoluble. The nitrofurazone was then extracted with acetone from the pre-extracted meal in a Soxhlet apparatus. After removal of acetone from the extract, the residue was dissolved in dimethyl- formamide. Suitable amounts of phenol, dissolved in dimethylformamide, and then aqueous alkali, were added. Two portions of the solution were taken, and one was treated with sodium dithionite.After centrifugation, the optical density of each solution was measured, and from the difference the amount of nitrofurazone present was obtained by reference to a calibration graph prepared by similarly treating known amounts of nitrofurazone.936 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [Analyst, 1701. 88 EXPERIMENTAL AND RESULTS Information on the above method was given in a personal communication from Mr. S. G. E. Stevens early in the development of the method before the use of sodium dithionite had been introduced. A collaborative test with this early version of the method was carried out by members of the Panel on feeding stuffs of differing complexity. It will be seen from Table I, which gives the results obtained on meals of increasing complexity, that satisfactory results were obtained on the simple Meal A, but on the more complex meals, it is clear that the pre-extraction procedure failed to remove interfering materials.TABLE I EFFECT OF ADDITION OF GRASS MEAL ON THE KECOVERY OF NITROFUKAZONE FROM A FEEDING STUFF Type of feed Nitrofurazone present in sample, Recovery, % Yo Meal A . . .. . . .. .. .. Meal B . . .. . . .. * . .. Meal C . . . . . . . . . . . . Meal A + 10 per cent. of grass meal Meal B f- 10 per cent. of grass meal . . .. 0.100 0-100 0.010 0.010 0.005 94 95 110 135 118 Meal C + 10 per cent. of grass meal .. 0.005 144 In subsequent assays the use of sodium dithionite to provide an empirical blank solution gave somewhat lower results more closely approximating to the theoretical values.It became apparent that, with some complex feeding stuffs, the extractives contained substances having reducing properties. This was demonstrated by measuring, at intervals of time, the optical density of the colour produced from an extract from a complex medicated meal containing nitrofurazone ; the value decreased progressively. The Panel then examined the use of potassium permanganate as described by van Zijl and goo sen^,^ and a modification of the method, in which the permanganate was added drop by drop, yielded the results shown in Table 11. TABLE I1 EFFECT OF TREATMENT WITH POTASSIUM PERMANGANATE ON THE RECOVERY . O F NITROFURAZONE FROM A COMPLEX FEEDING STUFF Recoverv ‘Type of feed Laboratory Nitrofurazone added, (mean of 3 r e h t s ) , % Yo x 1 B With grass meal i ” 0.005 0.010 0.005 0.010 0.005 0.010 0.005 0.010 0-005 0.010 0.005 0.010 0.005 0.010 99 98 78 83 93 94 94 93 99 98 100 92 83 90 During attempts to reconcile the discrepancies that continued to appear in the collabora- tive tests, it became evident that the quality of the dimethylformamide was important; the purity of this reagent should be such that the colour developed in the phenol and sodium hydroxide solutions in the presence of nitrofurazone should be stable for at least 2 hours.It was also established that the sodium dithionite solution must be freshly prepared and that the solid sodium dithionite itself should not be more than 6 months old. Some membersDecember, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE.PART 3 937 also observed that if the meal were medicated by adding to it a portion of an acetone solution of the drug, low recoveries of the nitrofurazone were obtained. This point is illustrated in Table 111. TABLE I r I EFFECT OF METHOD OF ADDING THE DRUG ON THE RECOVERY OF NITROFURAZONE Laboratory Nitrofurazone added, yo A . . . . 0.005 0.010 R . . .. 0.005 0.010 0.010 c . . . . 0.005 0.0 10 D . . . . 0.005 0.010 0-005 0-010 E .. . . 0-005 0.010 0.010 Recovery, % 97, 95 99, 102 100, 97 97, 100 52, 62 62 96, 94 93, 97 87, 86 87, 87 37, 33 39, 39 93 96, 95 99, 94 94 66, 57 62 Method of medication Solid drug added Drug added in solution after extraction with Drug added to feed as solution in acetone carbon tetrachloride 1 Solid drug added Solid drug added Drug added as solution in acetone Solid drug added 'i r Drug added as solution in acetone Destite modifications to the exDerimenta1 techniaues.the Panel continued to obtain I results licking a reasonable degree of ioncordance. examined. The' factors detailed below were therefore (i) The suitability of dimethylformamide as a solvent in the development of colour. (ii) The effect of time of centrifugation. (iii) The effect of potassium permanganate on a solution of nitrofurazone in acetone. (iv) The effect of potassium permanganate on extracts of feeding stuffs medicated with nit rofurazone. (v) The extraction process. Detailed investigation showed that the quality of the reagents and the extraction process were the only significant sources of variation.A series of experiments demonstrated that the intensity of the colour produced was linearly proportional to the concentration of nitrofurazone in the range 0 to 0.0004 per cent. (Beer's law) and also that the calibration curves prepared by individual members of the Panel did not vary by more than $ 3 per cent. By careful definition of the procedure, the Panel members were able to obtain better agreement, although the recoveries were in the range of only 80 to 90 per cent. For medicated feeds produced in bulk, two explanations were considered : (a) electrostatic deposition of the nitrofurazone in the mixing equipment and (b) the possibility of instability of the nitro- furazone in the feeding stuff. Experiments showed that the first of these possible explanations was not true when adequate earthing precautions had been taken.It was still not clear whether the observed losses were due to degradation of nitrofurazone in the feeding stuff itself or during the pre-treatment of the sample. I t was shown that satisfactory results were obtained on fairly old samples examined by members of the Panel when only light petroleum was used for the preliminary extraction. However, if carbon tetrachloride was used as the only pre-extraction solvent, low recoveries were sometimes obtained. Further, it was shown that if such feeding stuffs were extracted with carbon tetrachloride before medication with nitrofurazone then the use of carbon tetrachloride as a pre-extraction solvent caused no losses. These findings appear to indicate that there is some interaction between the nitrofurazone and the materials either extracted from the feed or derived from carbon tetrachloride ; possibly decomposition of nitrofurazone may also result from prior decomposi- tion of carbon tetrachloride itself as a result of contact with other constituents in the feed.The Panel has not conducted an investigation into this degradation; they consider that it is sufficient to draw attention to it.938 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [Analyst, vol. 88 Therefore the original method, in which only light petroleum is used as the pre-extraction solvent, is recommended, and any subsequent use of carbon tetrachloride will be necessary only when results obtained by this method indicate a complexity in the feeding stuff such as to cause undue interference with the optical measurements.The foregoing remarks are only applicable to the analysis of the compound feeding stuffs now commonly marketed and containing only nitrofurazone as the coccidiostat. The analyst should acquaint himself with changes in the composition of the feeding stuff, and he should consider the implications arising from the introduction of new components. RESULTS OF COLLABORATIVE TESTS- Members of the Panel carried out collaborative tests by the recommended method (see Appendix I). Two materials were examined, the first of these being a compound feeding stuff containing 0.0060 per cent. of nitrofurazone, prepared by a member of the Panel TABLE IV RESULTS OBTAINED BY THE RECOMMENDED METHOD ON A MEDICATED POULTRY FEED Sample contained 0.0060 per cent.of nitrofurazone Laboratory Nitrofurazone Recovery, Laboratory Nitrofurazone Recovery, and Analyst found, % % and Analyst found, yo % B1 0-0060 0.0060 0.0060 c1 0-00600 0-00605 0.00595 c 2 0.00565 0.00570 0.00565 D1 0-0063 0.0063 0.006 1 D2 0.0059 0.0057 0.0057 El 0.0057 0.0057 0-0053 100 100 100 100 101 100 94 95 94 105 105 102 98 95 95 95 95 88 TABLE V El 0.0059 0.0065 0.0060 0.0059 E2 0.0056 0-0056 0.0054 0.0057 E3 0.0058 0.0059 0-0063 0.0059 F1 0.00625 0,00645 0.00620 0.00625 0.00605 0.00625 98 108 100 98 93 93 90 95 97 98 105 98 104 108 103 104 101 104 RESULTS OBTAINED BY THE RECOMMENDED METHOD ON MEDICATED POULTRY FEEDS Samples prepared by adding different known amounts of nitrofurazone to an unmedicated poultry feed Laboratory Nitrofurazone added, Nitrofurazone found, Recovery, B1 0*0080 0-00808 101 0.0098 0.00941 96 0*0100 0.01 080 108 c1 0.0052 0.00535 103 0.0060 0.00595 99 0.0062 0.00620 100 422 0.0055 0.00565 103 0,0079 0.008 15 103 0.0087 0.00895 103 D1 0.0060 0.0060 100 0.0074 0.0072 97 0.0076 0.0073 96 D2 0.0073 0.0070 96 0.0080 0.0073 91 0.0089 0.0083 93 El 0.00658 0.00660 100 0-00700 0.00680 97 0.00709 0*00700 99 0.007 12 0.00695 98 and Analyst % % %December, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE.PART 3 939 experienced in feed compounding. The second sample was prepared by each worker from the same unmedicated feeding stuff to which he added a known amount of nitrofurazone within the range 0-005 to 0-01 per cent.The results on these two samples are shown in Tables IV and IT, respectively. Appendix I RECOMMENDED METHOD FOR DETERMINING NITROFURAZONE IN COMPOUND FEEDING STUFFS PRINCIPLE OF METHOD- The nitrofurazone is extracted by a selective solvent-extraction procedure, and then treated with alkali in the presence of phenol. The stabilised colour developed is compared with that produced by a known amount of the drug, similarly treated. APPLICABILITY- The method is applicable to medicated pre-mixes and compound feeding stuffs, of the type marketed at the time this Report was prepared, containing nitrofurazone as the only coccidiostat. APPARATUS- 2071 : 1954). REAGENTS- Soxlzlet extractor-A 100-ml extractor fitted with a B34 socket and a B24 cone (B.S. Extraction thimble-Whatman single, 25 mm x 80 mm.Light petroleum, boiling-range 40" to 60" C. A cetoute-Analytical-reagen t grade. Dimethyl'formamide-N,N-Dimethylformamide. Test the suitability of the reagent by developing the colour from nitrofurazone with solutions of phenol and sodium hydroxide (see "Procedure" below); the colour should remain stable for at least 2 hours. Phenol solution-A 5 per cent, w/v solution in dimethylformamide. Potassium permanganate solutiofa, 0.1 N. Sodium hydroxide solution, N. Sodiwz dithionite solution-A 1 per cent. w/v solution of sodium dithionite, Na,S,O, Prepare Nitrofurazone-Complying with the requirements of the British Veterinary Codex, 1953, (sodium hydrosulphite), (not more than 6 months old) in N sodium hydroxide. this solution immediately before use.p. 240. PRELIMINARY EXTRACTION OF SAMPLE- Weigh accurately an amount of sample containing about 1 mg of nitrofurazone, and transfer it to the extraction thimble; cover the sample with a small pad of cotton-wool. Insert the packed thimble into the extractor, assemble the extraction apparatus, and extract the sample with light petroleum; use an electric pad as the source of heat, so adjusted that the solvent cycles twenty times in about 45 minutes, and sufficient solvent so that the volume in the flask throughout the operation is not less than 25ml. Remove the packed thimble, allow the solvent to drain, and carefully remove any residual solvent in a current of warm air at a temperature not exceeding 60" C. EXTRACTION OF NITROFURAZONE- Transfer the packed thimble to a clean extraction apparatus, and extract the sample with acetone; use a water-bath as the source of heat so that the solvent cycles twenty times in about 1 hour, and sufficient solvent so that the volume in the flask throughout the operation is not less than 25 ml.During this extraction, shield the apparatus from light with a cardboard cylinder containing a small inspection window, or by any other suitable means. When the extraction is complete, rapidly cool the flask containing the extract to 20" C, and add 0.1 N potassium permanganate, drop by drop, until a faint pink colour is obtained PROCEDURE940 ANALYTICAL METHODS COMMITTEE [Analyst, Vol. 88 that is persistent for about 2 seconds (about 4 drops are required). Evaporate the extract on a water-bath to a volume of about 5 ml, shielding the extract from light.It i s important at this stage to avoid evaporating to dryness. Remove the flask from the water-bath, place an externally ribbed conical filter into the neck of the flask, and evaporate off the residual acetone by blowing a current of warm air (temperature not exceeding 60" C) across the top of the funnel in a way such that a slight turbulence is produced on the surface of the liquid in the flask. DETERMINATION OF NITROFURAZONE- Dissolve the residue in dimethylformamide, transfer the solution quantitatively to a 50-ml calibrated flask, suitably shielded from light, and dilute to the mark a t 20" C with dimethylformamide. Transfer a suitable portion, containing about 0.3 mg of nitrofurazone, to each of two 50-ml calibrated flasks containing 5 ml of phenol solution.To the contents of one flask add 2.5 ml of N sodium hydroxide, and dilute to the mark at 20" C with dimethylformamide; this is the sample solution. To the contents of the other flask add 2.5 ml of sodium dithionite solution, and dilute to the mark at 20" C with dimethyl- formamide; this is the blank solution and it should be a pale lemon-yellow colour free from any red or purplish tinge. Spin the solutions in a centrifuge, with a radius of 6 cm, at a speed of not less than 4000 r.p.m. for 2 minutes. Measure the optical density at 530 mp of the clear sample solution against the blank solution in 1-cm cells with a suitable spectrophotometer. Obtain the amount of nitrofurazone present in the sample solution by reference to a calibration curve previously prepared by plotting the optical densities obtained when known amounts of nitrofurazone are treated as described above, beginning at the addition of a suitable portion to each of two 50-ml calibrated flasks containing 5ml of phenol solution. Hence calculate the amount of nitrofurazone in the sample. REFERENCES 1. 2. 3. 4. 5 . 6. 7. Porter, C. C., ,4nnl. Chem., 1955, 27, 805. Ells, V. K., McKay, E. S., and Paul, H. E., J . A s s . OH. Agric. Chem., 1953, 36, 117. Buzard, J . A., Ells, V. R., and Paul, M. F., Ibid., 1956, 39, 512. Puglisi, E., Ibid., 1958, 41, 332. van Zijl, H. J . M., and Goosens, N., Chem. Weekbl., 1956, 52, 624. Tagaki, S., and Uno, T., Jap. J . Pharm. Chem., 1948, 20, 28; Chern. Abstr., 1951, 45, 3294. Cross, A. H. J . , Hendey, R. A., and Stevens, S. G. E., Analyst, 1960, 85, 355.

ISSN:0003-2654    

DOI:10.1039/AN9638800935

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

December, 19631 GOULDEN, GOODWIN AND DAVIES 941 Improvement Of Identif icntion in the Gas - Liquid Chromatographic Analysis of Agricultural Samples for Residues of some Chlorinated Pesticides* Part I. Improvement of Resolution on Single Columns and Application of the Multi-column “Spectrochromatogram” BY R. GOULDEN, E. S. GOODWIN AND L. DAVIES (“SheZ1” Research L t d . , Woodstock Agricultural Research Centre, Sittingboume, K e n t ) Gas - liquid chromatography with detection by electron-capture ionisation has shown itself to be a technique of considerable value in the analysis of crops, soils and animal tissues for residues of the chlorinated pesticides. The chief limitation of the method, however, lies in incomplete certainty of identification of a pesticide based on retention time alone on a single column with the resolving power a t present available.Two methods of increasing the certainty of identification of a pesticide are described in Part I of this paper. The first is by improvement of the resolution of a single packed column; the one described gives complete separation of a t least 11 of the less volatile chlorinated pesticides in a run time of 30 minutes. The second method involves similtaneous gas chromatography on five parallel columns holding stationary phases with differing gas - liquid chromato- graphic characteristics. In this system the eluate is fed to a single electron- capture ionisation detector. The multi-column “spectrochromatogram” produced has an appearance, and shows retention times, that are quite different and therefore highly characteristic for each individual pesticide.By this technique more certain identification is possible, e.g., in such problem analyses as aldrin in interference-containing grain or soil, and dieldrin in admixture with @’-I)DF, in animal tissues. THREE years ago we indicatedl that gas - liquid chromatography with detection by electron- capture ionisation was a technique with considerable potentialities in the field of pesticide- residue analysis, and followed this up by reporting procedures for the rapid identification and determination, on the nanogram scale, of the residues of several chlorinated pesticides in crops,2 soils3 and animal tissues4 by this means. The exceptional sensitivity and selectivity of the electron-capture ionisation detector to such pesticides was also appreciated by Clark5 and later by Watts and Klein,6 by Moore7 and by Taylor,* who successfully applied this technique to residue analyses on crops and animal products, foliage and soil, and animal and avian tissues, respectively.Interest in residue analysis by electron-capture gas - liquid chromatography has since developed quickly in the laboratories of agricultural concerns, public health authorities and instrument manufacturers, particularly in the United States, and has resulted in a rapidly increasing volume of papers either just p~blished,~ to l5 or in the press. There are several advantages of this technique over alternative methods of analysis of residues of chlorinated pesticides. Firstly, the gas - liquid chromatographic step makes possible simultaneous determination of two or more pesticides ; secondly, the exceptional sensitivity of the detector allows analyses to be made, without cencentration of the extract, on samples much smaller than could be used previously; and lastly, the selectivity of the detector to compounds of high electron affinity permits residues of the chlorinated pesticides to be determined in extracts of samples of biological origin with the minimum of prior clean-up. In common with all other analytical methods, however, the technique of electron-capture gas - liquid chromatograph has its own particular limitations.One of these relates to the * -4 summary of this paper was presented at the Vth International Pesticides Congress, London, July 17th to 23rd.1963.942 GOULDEN, GOODWIN AND DAVIES: IDEKTIFICATION IN GAS - LIQUID [Analyst, Vol. 88 characteristics of the electron-capture ionisation detector, whose originator has recently given a timely warning16 of the care needed in its application and use, if the results obtained are to be fully valid. A more general limitation concerns the gas - liquid chromatographic step in which identification of the pesticide being analysed is normally based on the retention time produced on a single gas-chromatographic column. The shortcomings encountered here are that two pesticides may have the same retention time on one particular column or that one pesticide may have the same retention time as a naturally-occurring co-extracted product either having high electron affinity, or being present in massive amount.These failings are of particular relevance in screening analyses on samples of unknown history, when valid control material is not usually available. During the last year our efforts have therefore been directed towards resolving or, at least, alleviating this problem. TABLE I RESOLVING POWER OF GAS - LIQUID CHROMATOGRAPHIC COLUMNS GIVING SATISFACTORY CHROMATOGRAPHY OF A WIDE RANGE OF CHLORINATED PESTICIDES ON THE NANOGRAM SCALE Column length, feet 2 2 4 6 10 4 4 6 2 2 4 4 4 10 10 10 2 2 4 Column bore, inches 0.065 0.095 0.125 0.195 Celite mesh-size, B.S.S. 100 t o 120 120 t o 150 60 t o 72 100 t o 120 60 to 72 60 t o 72 60 t o 72 60 t o 72 100 t o 120 100 to 120 60 t o 72 100 to 120 85 t o 100 72 to 85 100 to 120 100 t o 120 100 to 120 100 t o 120 100 to 120 Stationary phase Aldrin --, Nitrogen retention Silicone Epikote flow-rate, ml .time, oil, yo w/w 1001, yo w/w per minute minutes 2.5 0.25 35 3-25 2.5 0.25 80 4.25 10.0 1.00 100 7.30 2.5 0.25 40 6.90 2.5 0.25 45 7.50 5.0 0.50 120 4.40 2.5 0.25 30 6.50 2.5 0.25 115 4.45 2.5 5.0 2.5 2.5 2.5 0.8 2.5 0.8 0.25 0.50 0.25 0.25 0.25 0.20 0.25 0.20 60 115 100 100 50 100 90 85 4.00 4.70 4.55 6.00 7.10 6-75 34.0 10.0 2.5 0.25 100 3.80 2.0 0.50 100 3.40 0.8 0.20 100 4.50 Aldrin - Telodrin resolution None None 0.75 1.00 1.15 0.95 0.95 1.10 0.55 1.10 1.10 1-30* 1.35 1.70 1-75 2.15 0.851.0.90 0.90 * Illustrated in Fig. 1 R and Fig. 2. t Illustrated in Fig. 1 A. Resolution calculated from the expression difference between retention times sum of peak widths Resolution = 2 x - There are three obvious ways by which certainty of identification of a pesticide can be increased in gas chromatography- two (i) improvement of column resolution, (ii) application of stationary phases with differing characteristics or (iii) complementary use of detectors having dissimilar responses.Each of these approaches has been examined. are reported in Part I of this paper, and on the last in Part 11. The results of our work on the firstDecember, 19631 CHROMATOGRAPHIC ANALYSIS OF AGRICULTURAL SAMPLES. PART I 943 EXPERIMENTAL AND RESULTS IMPROVEMENT OF RESOLUTION ON SINGLE COLUMNS- Consideration of column design parameters-When considering ways in which the resolution of a packed gas-chromatographic column might be improved, attention was paid to the stationary phase, the supporting medium and also to the dimensions and material composi- tion of the column itself.For the stationary phase, no compound was found that would give a basic resolution superior to that of a non-polar silicone oil or elastomer. As supporting medium, plain Celite, probably the most widely used material, was pre- ferred. Although acid-washed Celite or other supports, e g . , Diatoport, can be used with silicone oil alone, endrin may be at least partially decomposed at the nanogram level. With plain Celite, of course, the addition to the silicone of Epikote 1001, in 10 + 1 ratio, is necessary to avoid adsorption or decomposition effects. This small addition has no marked effect on the chromatographic characteristics of the silicone oil.It is cheap, robust and easily worked. It is certainly more reactive than borosilicate glass,17 quartz18 or polytetrafluoro- ethylene, but this disadvantage is obviated in the presence of the Epikote 1001 additive. In developing conditions for improved resolution we imposed a limit of about thirty minutes for the gas - liquid chromatographic run-time, since protracted run-times reduce the attractiveness of the method and, moreover, increase the decomposition problem. Run times can, of course, be reduced by decreasing the amount of stationary phase or increasing the column temperature or the gas flow-rate. The effect of marked reduction of the first would be to increase insecticide decomposition. This, coupled with increased vapour loss of the Epikote would also occur with the second.The gas flow-rate of 100 ml per minute at 163°C was already considered high by normal gas - liquid chromatograhic standards and was likely to be limited by increase in the drop of gas pressure when longer columns were used. The two main variables remaining to be examined in the experiments on packed columns were, therefore, those of column length and bore. Some consideration was also given to the application of capillary columns in this field of work. Such columns have found wide use in the petroleum and petrochemical industries because of their exceptional resolving power in the gas - liquid chromatography of relatively volatile materials, but little has been reportedls on their value in pesticide residue analysis.Evaluation of packed copper columns-Experiments were carried out to determine the resolution obtained, particularly for an aldrin - Telodrin* mixture, on packed copper columns varying in length from 2 to 10 feet and in bore from 0.065 to 0.195 inches. These columns were packed with various concentrations and ratios of silicone oil - Epikote 1001 mixtures as stationary phase, supported on plain Celite, ranging in mesh size between 60 to 72 and 120 to 150 (R.S.S.). The flow rate of nitrogen carrier gas was varied from 30 ml per minute to 120 ml per minute; the maximum inlet pressure used was 45 p.s.i.g. Column temperature was maintained at 163" C throughout. The results obtained in these tests are summarised in Table I. Appraisal of these results and other relevant factors led to the conclusion that, within the run-time limit we imposed, the most suitable column for general use consisted of a &foot length of 0.125-inch bore copper tubing, packed with 2.5 per cent.w/w silicone oil plus 0-25 per cent. w/w Epikote 1001 supported on 100- to 120-mesh plain Celite maintained at 163" C and with a nitrogen flow- rate of 100 to 120 ml per minute. A chromatogram of a mixture of aldrin, Telodrin, dieldrin and endrin run under the above conditions is shown in Fig. 1 B, and may be compared with a chromatogram of the same mixture, shown in Fig. 1 A, run on the standard 2 feet long x 0.195-inch bore type of column we were using2 2 years ago. It will be noted that with the new column, near-baseline resolution is achieved with both pairs of insecticides, though with an increase in run time of about 60 per cent.The greater resolution of the new column permits satisfactory chromatography of a mixture of several more chlorinated pesticides (present only in nanogram amounts) than was * Telodrin is the Shell Trade Mark name for 1,3,4,5,6,7,8,8-octachloro-1,3,3a,4,7,7a-hexahydro-4,7- As a column material, copper has several advantages. methanoisobenzofuran.944 [Analyst, Vol. 85 previously possible. Fig. 2 shows a chromatogram in which thirteen of the less volatile pesticides are fully resolved in all but two instances. Evaluation of copper capillary columns-Most of the work on capillary columns was with copper as column material. The attractions here were, again, cheapness and ease of working. For example, no difficulty was encountered in obtaining gas-tight joints : this was not so when polytetrafluoroethylene was used as capillary tubing.The columns tested had a bore of 0.018 inches and. varied in length from 25 to 125 feet. They were coated with silicone oil, with or without added Epikote 1001, by means of the dynamic method involving a plug of stationary-phase solution. Because of the very low concentration of pesticides present in the extract used, splitting of the 1 to 8 pl sample on injection was not possible. A polytetrafluoroethylene insert was, therefore, located at the inlet of the injection port to reduce its dead volume and thereby minimise diffusion. GOULDEN, GOODWIN AND DAVIES: IDENTIFICATION IN GAS - LIQUID B Aldrin Telodrin I1 II 0 5 F.S.D. 0 Dieldrin 0 5 10 15 20 Time, minutes A: Column dimensions, 2 feet long, 0.195-inch bore B: Column dimensions, 4 feet long, 0.125-inch bore Electron-capture gas chromatograin, showing resolution of chlorinated-pesticide pairs on packed columns Fig.1. A temperature of 163" C was used throughout, with nitrogen carrier-gas flow-rate ranging from 2 to 20 ml per minute. For detection, a planar-source electron-capture ionisation cell of low (0.1 ml) internal volume was used, and was fed with nitrogen scavenger gas to maintain an adequate flow rate (50ml per minute) through the detector. When a mixture of chlorinated pesticides, present in nanogram amounts, was analysed on the gas - liquid chromatographic column, decomposition was found to be appreciable except when Epikote was present.Further, in most experiments, resolution was much inferior to that obtained on normal packed columns.December, 19631 CHROMATOGRAPHIC ANALYSIS OF AGRICULTURAL SAMPLES. PART I 945 The best capillary column prepared was 125 feet long and was coated with a 2 per cent. w/v solution of silicone oil and Epikote 1001 in a 10 + 1 ratio in ethyl acetate. This column, however, had a resolution only slightly better than that of the standard packed column used to produce Fig. 1 A and, moreover, effected no decrease in run time. APPLICATION OF THE MULTI-COLUMN “SPECTROCHROMATOGRAM”- General considerations-In earlier papers2,* we drew attention to two ways in which uncertainty of identification of a pesticide, caused by the presence of interfering material of coincident retention time, might be overcome.These were to effect resolution of the two components by use of a second (polar) column, whose gas - liquid chromatographic character- istics differed from the standard (non-polar) column ; or, alternatively, to remove the interfer- ing component previously by liquid - liquid chromatography. The latter technique is not always suc~essful~~ and is of limited value in screening analyses on samples of unknown history, since loss of other sought-for pesticides or their metabolites may occur during the process. The first-mentioned technique was pursued in the Tunstall Laboratories of “Shell” Research Ltd. by Robinson and Richardson,20 who determined the relative retention of eight chlorinated pesticides on single columns of Apiezon L and Nonidet P40 as well as the silicone elastomer E301 and Epikote 1001 used by ourselves.We therefore examined the usefulness of simultaneous gas - liquid chromatography on multiple parallel columns, each containing stationary phases of differing characteristics, as a method for increasing the certainty of identification of pesticides, particuarly when these suffer interference from other pesticides or from naturally-occurring co-extracted material having coincident or near-coincident retention times on a standard (non-polar) column. The chromatograms so produced we have tentatively called “spectrochromatograms,” because they are indicative of the spectrum of selectivity of the various stationary phases towards individual chlorinated pesticides.iolvent Lindane Heptac h lor I r Ronnel F.S.D. Oe5 r- Heptachlor epoxide r- Chlordane isomer J I I I I 1 I I 0 5 10 15 20 25 30 Time, minutes Fig. 2. Electron-capture gas chromatogram, showing resolution of thirteen chlorinated pesticides on a packed column: 0.4 nanograms of each pesticide946 GOULDEN, GOODWIN AND DAVIES : IDENTIFICATION IN GAS - LIQUID [AIzalySt, VOl. 88 Requirements for, and evaluation of, stationary phases--It was decided to develop a multi- column system having five stationary phases ; this was the maximum number that it was felt could reasonably be handled operationally. These five phases would ideally comprise two or three of proven general applicability in the chlorinated pesticide field, while the others would be present mainly because of their ability to resolve obvious problem pairs, e.g., aldrin and grain interference or dieldrin and p$'-DDE mixtures.The aim in searching for suitable stationary phases was that they should cover the widest possible range of polarity and structure, should be stable at the operating temperature of 163" C and, in particular, exhibit insufficient bleed to affect the highly sensitive electron- capture ionisation detector. In addition to the silicone oil, Epikote 1001 and Apiezon L already referred to, twenty-one materials were evaluated as potential stationary phases on plain Celite for the gas - liquid chromatography of a test mixture of seven chlorinated pesticides. These pesticides were : indane, heptachlor, aldrin, Telodrin, dieldrin, endrin and fi+'-DDT.The stationary phases chosen are grouped below under three headings: 1 2 3 Behenic acid Adipic acid Diethylene glycol succinate Calcium gluconate Araldite Dulcitol D ( +)Glucose Mannitol . Nitrile silicone GE-XE-60 Mannitol hexa-acetate Nitrile silicone XF-1150 Nitrile silicone XF-1112 Pol ygl ycerol Nonidet P40 Salicin Sodium carboxymethylcellulose P.E.G. 4000 D (- ) Sorbi to1 P.E.G. 6000 Zinc stearate P.E.G. 15-20,000 The first group (1) was quite unsatisfactory (giving little or no chromatography); the second group (2) was partially successful (chromatography of 3 or 4 pesticides) ; while the last group (3) was the most successful (chromatography of 5, 6 or 7 pesticides). It was clear from earlier work2y20 that addition of Epikote 1001 to either the silicone or Apeizon stationary phases on plain Celite was essential if adsorption or decomposition of some of the chlorinated pesticides was to be avoided.The effect of this additive on salicin, nitrile silicone GE-XE-60 and diethylene glycol succinate was therefore examined. Only with the latter compound was there significant improvement in chromatography, although even here partial loss of heptachlor and total loss of +p'-DDT still occurred. This work led to a consideration of other ways in which such adsorption - decomposition effects could be overcome. Pre-treatment of the stationary phase and supporting medium with tris-(2-biphenylyl) phosphate (Dow K-1110) was reported by Gunther, Blinn and Kohn,21 but the success of this method was not confirmed by Beckman and Bevenue.l* As an alternative to Epikote 1001, we found that Araldite, a proprietary epoxy resin adhesive, worked as well for the above purpose.The use of such epoxides resulted in experiments on the pre-treatment of plain Celite with epichlorhydrin under reflux. It was found that Celite so treated permitted the test mixture of the seven pesticides to be analysed by gas- liquid chromatography on the nanogram scale without decomposition, when silicone alone was used as stationary phase. Design, composition and operation of the multi-column assembly-A multi-column assembly head was constructed in brass. The injection port led via a short capillary manifold to the inlets of five parallel U-shaped copper columns each 4 feet long with a 0-125-inch bore. The outlets of these five columns were then led via a second short capillary manifold, also located in the assembly head, to the inlet of a single planar-source electron-capture ionisation detector, whose internal volume was about 0.1 ml.All column connections were made with Simplifix fittings . In preliminary tests severe decomposition of some pesticides was traced to contact with the metal of the assembly head. This decomposition was suppressed by washing the head out with a 2 per cent. w/v solution of silicone oil and Epikote 1001 mixture, in 4 + 1 ratio in ethyl acetate, followed by evaporation of the solvent before attaching the columns. The final choice of five stationary phases, based on the requirements and evaluations given earlier, is given below, together with the concentrations used.These were carefully This work has yet to be further investigated.December, 19631 CHROMATOGRAPHIC ANALYSIS OF AGRICULTURAL SAMPLES. PART I 947 chosen after appreciable experimentation to give peak resolution of as many chlorinated pesticides, metabolites and artifacts as possible. 2.5 per cent. w/w silicone oil (May and Baker Ltd.) + 0.25 per cent. w/w Epikote 1001 (Shell Chemical Co. Ltd.). 1.0 per cent. w/w Apiezon L (A.E.I. Ltd.) + 0.2 per cent. w/w Epikote 1001. 1.0 per cent. w/w Epikote 1001. 3.3 per cent. w/w nitrile silicone fluid XF-1112 (General Electric Co., U.S.A.). 2.7 per cent. w/w nitrile silicone fluid GE-XE-60 (Applied Science Labs. Inc.; ( a ) (b) (c) (d) (e) U.K. Agents, W. G. Pye & Co. Ltd., Cambridge).The procedure for residue analysis with the multi-column assembly was essentially the same as that described previously for single column operation, except that injection volumes were increased fivefold to 25 p1 to compensate for splitting of the charge into the parallel columns. The crop, soil or tissue sample was macerated with the minimum volume of II Aldrin II . I Endrin a & c be d Column 1 1 Column 1 I I I a e b d I I I I 1 I I I I. 5 10 I5 0 10 20 30 Time, minutes Time, minutes Fig, 3. Multi-column “spectrochromatograms” of some chlorinated pesticides : injection, 25 pl = 2.5 nanograms of each pesticide948 [Analyst, Vol. 88 redistilled analytical-reagent grade acetone to give a fluid macerate that was filtered through sintered glass.A portion of this acetone extract was then shaken with twice its own volume of redistilled petroleum spirit (boiling-range 62” to 68” C), in the presence of excess aqueous sodium sulphate solution. Column temperature was maintained at 163” C, with the nitrogen carrier-gas flow-rate through each column at the normal rate of 100 ml per minute. The flow-rate was set at this level to keep retention times as short as practicable, although this resulted in a very high combined flow-rate of 500 ml per minute to the detector, which the latter accepted without malfunction. GOULDEN, GOODWIN AND DAVIES: IDENTIFICATION IK GAS - LIQUID The filtrate was made up to two volumes per unit weight of sample. This petroleum-spirit extract was used for the injection. Evaluation of ‘spectrochromatogyams” and multi-column per formance-By using the above- mentioned system, multi-column ‘(spectrochromatograms” were produced for a range of the less volatile chlorinated pesticides. Those for lindane, aldrin, Telodrin, dieldrin, $$’-DDE and endrin are shown together in Fig.3. In general, 3 to 5 distinct peaks are obtained and show retention times which, considered together, provide good evidence for the identification of the compound present. For example, on our standard silicone and Epikote column of two years ago (see Fig. 1 A), aldrin was only partially resolved from Telodrin: in Fig. 3, however, it will be seen that the “spectrochromatogram” of aldrin is quite different both in appearance and indicated retention time from that of Telodrin. Similarly, on the standard column dieldrin was only partially resolved from endrin and not at all from p$’-DDE : Fig.3 shows their “spectrochromatograms” to be appreciably different. Most grain crops (wheat, oats, barley, maize and rice) have been found to contain materials that have the same retention time as aldrin on the standard column. The results of applying the multi-column system to this problem are shown in Fig. 4 in which, despite the presence A B 0-5 I 1 I I I I 5 10 15 20 Time, minutes A: Control oats B: Oats containing 0.25 p.p.m. of Fig. 4. Multi-column “spectrochromatogram” of control and aldrin-treated oats: injection, 25 p1 G 6 rng of crop aldrin ~ ~ ’ - D D E c a e b d -olumn Dieldrin a&c b e d I I ! I I 0 10 20 30 Time, minutes A: Control beef R: Beef containing 1.0 p.p.m.of p$’-DDE and 0.5 p.p.m. of dieldrin Fig. 5. Multi-column “spectrochro- matogram” of control beef and beef containing dieldrin and pp’-DDE : injec- tion, 25 pl L- 6 mg of tissueDecember, 19631 CHROMATOGRAPHIC ANALYSIS OF AGRICULTURAL SAMPLES. PART I 949 of the naturally-occurring interference, peaks whose appearance and retention times are characteristic of the aldrin “spectrochromatogram” are clearly evident in the aldrin-treated sample. Similar results were obtained on interference-containing soil samples. Another problem is the one in which two insecticides that are unresolved on the standard column may well be present together in the test sample. Such an example is that of dieldrin and p$’-DDE in animal tissues. The “spectrochromatogram” produced for this mixture is shown in Fig. 5, in which it can be seen that there is sufficient characterisation to identify these two pesticides.The reproducibility of “spectrochromatogram” retention times was found to be as good as on single columns. Reproducibility of splitting of the injection charge between the five parallel columns was, however, less satisfactory and was thought to be limited by the design of the particular assembly head used. The life of the multi-column assembly was found to be restricted to about two weeks owing to ageing or vapour loss of some of the stationary phases, in particular the nitrile silicone XF-1112 and Epikote 1001. These effects resulted in an increase in pesticide decomposition with time coupled with changes in the retention times shown by the “spectrochromatogram” peaks. In consequence it was found advisable to determine the reference “spectrochromato- grams” at the same period of time as those of the test samples were being run.An indication of the value of the multi-column “spectrochromatogram” technique in screening work was obtained inadvertently when a wide variety of locally purchased “control” crops was being examined for use in work on the complementary detection system to be reported in Part IT of this paper. Clearly-defined “spectrochromatograms” showed traces of identifiable chlorinated pesticide in one or two of these samples. DISCUSSION AND CONCLUSIONS The development work on single packed columns described earlier in this paper produced a useful though not marked gain in resolution in a column having a general use for screening purposes.In consequence the 4 feet x 0.125-inch bore column is to be preferred to the 2 feet x 0.195-inch bore column used previously. It does not, however, greatly improve the certainty of identification of any particular pesticide. Against this may be set the fact that it had to be designed to give rapid chromatography of nanogram amounts of a very wide range of chlorinated pesticides (see Fig. 2). In more specialised circumstances, for example, when longer run times are no detriment, a valuable increase in resolution can be achieved by using 0-125-inch bore columns of much greater length. The brief study of capillary columns proved disappointing, considering the known value of these columns in other fields of work.The main problem here lies in the extremely low dilution of the extract solution being analysed, rendering the stream splitting normally used in capillary work impracticable. There remains, nevertheless, much scope for further development in the application of capillary columns to pesticide-residue analysis, not only on solving the above-mentioned injection problem but in evaluating columns of other bores, lengths and materials and different stationary-phase systems. As a means for improving the certainty of identification of residues of the chlorinated pesticides, the multi-column “spectrochromatogram” technique showed more promise than the others referred to, and the method could well be applied as a supplementary confirmatory technique for those samples that had given an indication on a short primary screening column of the possible presence of pesticide.The multi-column assembly is, of course, more trouble- some to prepare, has a shorter working life than the single column and, at present, is only roughly quantitative. Against this, however, the method is capable of further improvement. Re-design of the assembly head, possibly in stainless steel, aluminium or glass may reduce the tendency for pesticide decomposition inside it, thus avoiding the need for the silicone and Epikote pre-treatment. At the same time, an improvement of the reproducibility of splitting of the charge into the parallel columns might be achieved. This would further increase certainty of pesticide identification by giving more reproducible “spectrochromato- grams,’’ and would also improve the extent to which the technique may be used quantitatively.A further contribution of value would be the introduction of alternative stationary phases ; the two main requirements here are thermally stable and highly polar compounds.950 GOULDEN, GOODWIN AND DAVIES [Analyst, Vol. 88 We thank Mr. D. M. Barnett for much careful experimentation, Mr. G. F. Poulton for the design and construction of the multi-column assembly, and Dr. R. A. E. Galley, Director, and Mr. J. G. Reynolds, Associate Director of the Woodstock Agricultural Research Centre, for their interest and encouragement during the course of this work. 1. 2. 3. 4. 5. 6. 7 . 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. REFERENCES Goodwin, E. S., Goulden, R., Richardson, A,, and Reynolds, J. G., Chem. & Ind., 1960, 1220. Goodwin, E. S., Goulden, R., and Reynolds, J . G., Analyst, 1961, 86, 697. Reynolds, J. G., Commun. Agric. Univ. Res. Sta. State of Ghent, 1961, 26, 1530. Goulden, R., Preprint, Lecture a t the First International Symposium on Methods for the Analysis Clark, S. J., Preprint, Lecture a t the 140th American Chemical Society Meeting, Chicago, September, Watts, J. O., and Klein, A. K., J . Ass. 08. Agric. Chem., 1962, 45, 102. Moore, A. D., J . Econ. Entom., 1962, 55, 271. Taylor, A., Analyst, 1962, 87, 824. Mattick, L. R., Barry, D. L., Antenucci, F. M., and Avens, A. W., J . Agric. Fd. Chem., 1963, Hartmann, H., and Dimick, K. P., American Chemical Society Pesticide Symposium, Los Angeles, Clark, S. J., Ibid., Paper No. 4. Segal, H. S., and Sutherland, M. L., Ibid., Paper No. 7. Beckman, H., and Bevenue, A., Ibid., Paper No. 19. Klein, A. K., Watts, J. O., and Damico, J. N., J . Assoc. OH. Agr. Chem., 1963, 46, 165. Henderson, J. L., Ibid., 1963, 46, 209. Lovelock, J . E., Anal. Chem., 1963, 35, 474. “Aerograph Research Notes,” Wilkens Instrument and Research Inc., Walnut Creek, California, Beckman, H., and Bevenue, A., J . Chromatog., 1963, 10, 231. Patchett, G. G., in Zweig, G., Editor, “Analytical Methods for Pesticides, Plant Growth Regulators and Food Additives,” Academic Press Inc., in the press. Robinson, J., and Richardson, A., Chem. & Ind., 1963, 1460. Gunther, F. A., Blinn, R. C., and Kohn, G. K., Nature, 1962, 193, 573. of Foods, Bordeaux, October, 1962. 1961. 11, 54. April, 1963, Paper No. 3. Summer Issue, 1962. Received JurLe 27th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800941

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

December, 19631 GOULDEN, GOODWIN AND DAVIES 951 Improvement of Identification in the Gas - Liquid Chromatographic Analysis of Agricultural Samples for Residues of some Chlorinated Pesticides* Part 11. A Halogen-sensitive Detector in Complementary or Alternative Use to an Electron-capture Ionisation Detector BY R. GOULDEN, E. S. GOODWIN AND L. DAVIES (“Shell” Research Ltd., Woodstock Agricultural Research Centre, Sittingbourne, Kent) Gas - liquid chromatography with detection by electron-capture ionisation has found considerable use in the analysis of crops, soils and animal tissues for residues of the chlorinated pesticides. The chief limitation of the method, however, lies in incomplete certainty of identification of a pesticide based on retention time alone on a single column of the resolving power at present available.Two methods of increasing certainty of identification, the use of single columns of higher resolving power and the application of the multi-column “spectrochromatogram,” were described in Part I of this paper. In Part 11, a third method based on simultaneous detection by two highly sensitive and selective detectors with dissimilar response characteristics to individual chlorinated pesticides is described. One of these sensors is the electron-capture ionisation detector, and the other is a halogen-sensitive cell of the type used in refrigeration leak detection. The last-mentioned detector has great potential value in residue analysis for the chlorinated pesticides. It is sensitive to the nanogram range, is more selective than the electron- capture ionisation type and has greater linearity of response. Moreover, it is relatively cheap and can be operated with simple circuitry. GAS - LIQUID chromatography with detection by electron-capture ionisation has rapidly proved itself1 9 2 5 3 9 4 to be a technique of particular value for detecting and determining residues of chlorinated pesticides in crops, soils and animal tissues.Its main advantages over alternative methods are greater sensitivity, greater selectivity and its ability to analyse for more than one component at the same time. These advantages result in a method in which extracts of small samples can be analysed rapidly for nanogram amounts of several pesticides without a concentration step, and with the minimum of prior clean-up. Although care is necessary in the proper operation5 of the detector itself, the main limitation of the method lies in the degree of certainty of identifying the individual pesticides, when this identification is based only on retention times obtained on a single column of the resolving power at present available. This factor is most relevant in screening analyses on samples of unknown history, when valid control material is not usually available.In Part I of this paper,6 two methods intended to overcome or reduce this problem were described: the use of single columns of higher resolving power than that of earlier versions, and the production of multi-column “spectrochromatograms” having an appearance and showing retention times highly characteristic of the individual pesticides being analysed.A third approach, using highly sensitive detectors whose responses to the halogenated pesticides differ, is described in Part 11. A detector is needed that has the unusual properties of exceptional sensitivity and high selectivity to the halogenated pesticide, but significantly different response characteristics from the electron-capture ionisation detector, with which it could be used in tandem. Such a device is the halogen-sensitive element used in the Type HA leak-detector made by Associated Electrical Industries Ltd. The idea of applying a halogen-sensitive device of this type to the analysis of pesticide residues is not new, for in 1952. C. A. Reilly, working in the Laboratories of the Shell Develop- ment Company at Emeryville, California, reported on a determined effort to use the General Electric Type H leak-detector of similar design to the Type HA for this purpose.This July 17th to 23rd, 1963. * A summary of this paper was presented a t the Vth International Pesticides Congress, London,952 GOULDEN, GOODWIN AND DAVIES: IDENTIFICATION I N GAS - LIQUID [Analyst, v01. 88 work, however, was not altogether successful, partly because the use of the detector was not preceded by a gas - liquid chromatographic step, and partly because of the electrical or electronic instability of the systems examined. Much more recently the successful use of this type of halogen-sensitive element as a detector in the gas - liquid chromatography of volatile chlorinated hydrocarbons was briefly reported by Cremer, Kraus and Bechtold7 in Germany.In this detailed study, high selectivity to chlorinated compounds was recorded, and such a sensitivity was developed that 0.3 nano- grams of chlorine could be detected under favourable conditions. This lead was followed in the Tunstall Laboratory of "Shelf" Research Ltd. by A. Richard- son, who showed that the A.E.I. halogen-sensitive element already referred to would function as a gas - liquid chromatographic detector, and could provide a relatively high sensitivity with the aid of only simple circuitry. We therefore undertook a detailed evaluation of two versions of this halogen-sensitive device with the object of using it for determining residues of the chlorinated pesticides on the nanogram scale, and of operating it in tandem with an electron-capture ionisation detector as a method for increasing the certainty of identifying these compounds.EXPERIMENTAL AND RESULTS DESCRIPTION AND OPERATION OF THE HALOGEN-SENSITIVE ELEMENT- The A.E.I. Ozotron Type H halogen-sensitive element,* shown diagrammatically in Fig. 1, consists of a pair of concentric platinum cylinders mounted within a protective boro- silicate glass envelope through which, in normal operation, the air to be tested for traces of halogenated compounds is drawn at a rate of about 150 ml per minute. The inner platinum To anode S i I icone-r u b be r - sleeve To heater TO cathode -Glass envelope Brass adapter t From column Fig. 1. Halogen-sensitive detector (A.E.I. type H) cylinder, or anode, which has been sensitised by an alkali treatment, is indirectly heated by an internal platinum filament taking a current of 7.4 amps at 6 volts a.c., and operates at a temperature of about 800" C.An anode-to-cathode potential of 250 volts d.c. is normally used, resulting in the production of a positive-ion standing current that can be suitably amplified and presented. When air, containing halogen vapour, is drawn through the cell there is an apparent increase in the positive-ion current, whose magnitude is indicative of the concentration of halogen vapour present.December, 19631 CHROMATOGRAPHIC ANALYSIS OF AGRICULTURAL SAMPLES. PART I1 953 A newer version of the above detector has recently become available. This model, the Ozotron Type J, is basically the same as the Type H except that the element is mounted more rigidly in a two-piece ceramic envelope from which the platinum heater filament and electrodes are easily removed.DESIGN OF CIRCUITS AND PERFORMANCE OF HALOGEN-SENSITIVE DETECTORS- An Ozotron Type H halogen-sensitive element was connected up as detector to a 4 feet long x 0.125-inch bore silicone - Epikote column of the type described in Part I of this paper. The detector circuitry used is shown diagrammatically in Fig. 2(a). Power or the r-I 240/6L C.V.T. - 1 250 Fz 12KQi 9v Fig. 2 (a). Simple circuit for halogen-sensitive detector heater filament was supplied from a 240 to 6V step-down transformer connected in series with a 1-ohm sliding resistance, and the potential required for the anode was taken from 0 to 120 V high tension batteries.The signal from the cathode was led, via a 1-megohm load in series with a 1500-ohm input resistor (backed off by a 0 to 9 V grid-bias battery controlled by a wire-wound variable resistor), before being led directly into a 1-mV full-scale deflection chart recorder having a maximum input impedance of 1000 ohms and a pen-response time of 2 seconds. An evaluation of detector performance was made under these conditions by injecting 1 to 10 pl volumes of dilute solutions of chlorinated pesticides in light petroleum (boiling- range 62" to 68" C) into the gas - liquid chromatographic apparatus at a column temperature of 163" C, while varying the operational parameters. Sitrogen was used as carrier gas at various flow-rates in the range 100 to 200ml per minute, with and without the admission of air at the inlet of the detector.I t was found preferable to use separate leads to the anode and inner heater terminal instead of a common one as indicated by the manufacturer, as the latter procedure resulted in a higher electrical noise level. Sensitivity, both to chlorinated pesticides and to light-petroleum solvent, increased with increase in heater voltage, but in favour of the solvent. In consequence it was impracticable to work at the highest heater voltage (6 volts), as the large solvent response obtained masked the early part of the chromatogram. Below about 4 volts, however, sensitivity rapidly decreased. Similarly, increase in the potential applied to the anode of the cell resulted in an increase in both sensitivity and electrical noise level.Disproportionate increase of the latter made it desirable to operate at comparatively low potentials, although below about 36 volts d.c. a marked increase in peak tailing seriously affected the chromatograms obtained. The time-constant of the cell was observed to be rather longer than that for the electron- capture ionisation type of detector, and in consequence, higher nitrogen flow-rates (greater954 GOULDEN, GOODWIN AND DAVIES IDENTIFICATION IN GAS - LIQUID [Analyst, VOl. 88 than 100 ml per minute) resulted in improved chromatogram peak shape, but at the expense of some sensitivity. Admission of air to the inlet of the detector to effect oxidative combustion of the pesticides rather than pyrolysis, produced a marked increase in the standing current of the cell and a reduction of sensitivity that became significant at air flow-rates in excess of about 5ml per minute. The level of sensitivity obtainable by using the above-mentioned simple circuitry and with a power input to the heater filament of 4.6 volts a.c.at 6.7 amps, an anode potential of +36 volts d.c. and a nitrogen carrier-gas flow-rate of 100 ml per minute at 163OC, corresponded to a limit of detection of about 2 nanograms for aldrin (approximately twice the noise level). Response of the detector versus pesticide load was observed to be linear up to recorder full- scale deflection, which for aldrin occurred with 100 nanograms. This sensitivity, while encouraging, was not quite sufficient for our purpose, and a more sensitive version of the circuit , shown diagrammatically in Fig.2(b), was therefore developed. Stabilised d. c. supply - - 120v - +-- L 1 Vibron 33B 3ect romete r To recorder Fig. 2 ( b ) . More sensitive circuit for halogen-serisitive detector With this arrangement it was found essential to provide the heater filament with a more stable power supply than was given by the step-down transformer. Replacing the latter by a heavy duty 6 volt accumulator proved satisfactory, except that the high current con- sumption resulted in a gradual drift of the recorder baseline as the accumulator was dis- charged. The latter was therefore replaced by a comparatively inexpensive transistorised Zener-diode stabilised d.c. power supply, specially designed for the purpose.In this more sensitive circuit the signal from the cathode was led through a 1-megohm input resistor in an Electronic Industries Ltd. model A33B current- and resistance-measuring bridge, backed off similarly to the one described above, and then passed into an E.I.L. Vibron model 33B vibrating-reed electrometer, before being fed via appropriate series - parallel matching resistors to the 1-mV recorder. Under the same operating conditions as described above, the limit of detection for aldrin with this circuitry was about 0-1 nanograms at the 100-mV attenuation setting. Again, good linearity of response with increase in pesticide load was observed up to recorder full- scale deflection at the 1000-mV (maximum) attenuation setting, which for aldrin occurred with 30 nanograms.December, 19631 CHROMATOGRAPHIC ANALYSIS OF AGRICULTURAL SAMPLES.PART 11 055 A similar evaluation of the ceramic-protected Ozotron Type J halogen-sensitive element was also undertaken. The performance characteristics observed were similar in some respects to those outlined above. In this work it was found essential to ensure that all signal-carrying leads were fully screened by using co-axial cable, and that electrical connections, particularly the leads to the heater elements] were soundly made, e.g., brazed or silver-soldered. The two least satisfactory features of the Type H and J detectors when used in the manner described, were their general electrical instability and the need for conditioning them appreciably before high sensitivity could be achieved.The tendency towards electrical instability was considered to be caused partly by some lack of rigidity in the location of the concentric platinum electrodes (in the Type H) and partly by shorting of the electrodes] perhaps by carbonaceous deposits resulting from pyrolysis of the samples. Such deposits could easily be removed by cleaning the electrodes in acetone; this was facilitated for Type J by the ease with which the cell could be taken apart. The conditioning of the cells and their maintenance in a state of high sensitivity was the biggest problem encountered, and has not yet been completely solved. A variety of con- ditioning processes were tried with five such cells, including pre-treatment with air, nitrogen or massive loads of chlorinated material, and washing the anode and gentle abrasive cleaning of the cathode in acetone.Of these, pretreatment with air shows promise of being the most effective. The results quoted above were obtained on the most sensitive of the detectors examined and this one (a Type H) was used to produce the halogen-detector chromatograms reproduced in this paper. EVALUATION OF HALOGEN-SENSITIVE AND ELECTRON-CAPTURE IONISATION DETECTORS IN An Ozotron Type H halogen-sensitive element was connected by means of a short brass capillary tube and silicone rubber sleeve on to the outlet port of an electron-capture ionisation detector. This, in turn, was connected to a 4 feet x 0.125-inch bore copper column that was packed with 2.5 per cent. w/w of silicone oil plus 0-25 per cent.w/w Epikote 1001, supported on 100- to 120-mesh plain Celite. The column was operated at a temperature of 163" C , and a nitrogen flow-rate of 100 ml per minute was maintained. For the Ozotron, the more COMPLEMENTARY OPERATION- I- Shandon Universal Amplifier I I-mV potentiometer chart recorder 6' t F rorn column Resistance 1:; 1 1 I -mV potentiometer chart recorder A = Electron-capture detector B = Halogen-sensitive detector Fig. 3. Simultaneous complementary detection system956 [Analyst, Vol. 88 sensitive circuit was used, and the signals from both detectors after attenuation or ampli- fication as appropriate were fed, as shown in Fig. 3, into separate 1-mV recorders running at the same chart speed. By this arrangement simultaneous chromatograms from these two different detection systems could be produced for solutions containing nanogram amounts of the chlorinated pesticides.An example of this is illustrated in Fig. 4, in which the simultaneous halogen-sensitive and electron-capture chromatograms were produced on the injection of 2-3 pl of a petroleum spirit solution, containing 2-3 nanograms each of lindane, ronnel, Telodrin, heptachlor epoxide, endrin, Rhothane and @’-DDT. It will be noted that the relative response of the two detectors is different, particularly for lindane, Telodrin and endrin. This difference in relative response can be more marked when sensitivity to chlorinated and non-chlorinated material is considered. In the top half of Fig. 5 , the electron-capture chromatogram of “control” wheat before and after the addition of 0.25 p.p.m.of aldrin can be seen. As has been previously pointed out,l this control sample, typical of the grain crops, exhibits massive electron-capture interference from a naturally occurring co-extractive, whose retention time is the same as that of aldrin. In the simultaneous halogen-sensitive chromatograms, no such interference is evident. The complementary use of the two detectors could therefore be of much value in increasing the certainty of identification of any pesticide indicated as present in the sample. The value of the difference in relative response of the two detectors to individual pesticides is shown in Fig. 6, in which, under standard operating conditions, reproducible relative responses (as measured by peak area ratios) provide good confirmatory evidence for the GOULDEN, GOODWIN AND DAVIES : IDENTIFICATION IN GAS - LIQUID f I 1 0 5 10 15 20 25 30 Time, minutes A: Electron-capture detector B: Halogen-sensitive detector Fig.4. Simultaneous complementary gas chro- matograms of seven chlorinated pesticides : 2.3 nanograms of each pesticide. Aldrin (0.25 p.p.m) Test Control 1 1 I I I I I I 1 0 5 10 IS 20 25 30 Time, minutes A: Electron-capture detector B: Halogen-sensitive detector Fig. 5 . Simultaneous complementary gas chromatograms of control and aldrin- treated wheat: injection, 5 pl = 6 mg of crop.I>ecember, 1 !%3] CHIZO~I.~TOGKAI>HIC 9 5 i identification of the pesticides lindane and Telodrin, added to spring cabbage to the extm t of 0.6 I1.p.m. In the last-mentioned chromatogram, the area ot tlie halog:en-sensiti\.e detector peak for Teloclrin (clilorinc content, 68.8 per cent.) is greater than that for lindane (chlorine content, 73.1 per cent.), indicating that, for this detector, the rt’sponse obtained is not simply a function of tlie chlorine content of the molecult..Anotlier example of simultaneous complementary halogen-sensitive - electron-capture gas - liquid cliromatography is illustrated in Fig. 7 , which shows the dual chromatogram.s produced by the addition of 1 p.p.m. of either dieldrin or $$’-r)DE to “control” lamb. The retention times o f these two pesticides on a single silicone - Epikote column are nearly coinci- dent, with tlie consequent possibilitv of confusion between them. \Tith the complementary detector system, however, tlie calculated peak area ratios, 1.3 and 1.1 for diel(1rin and fip’-I)I)E rcspectively, conld provide sufficient widener to permit reasonable distinction between t l i t .two pesticides. I) I sc IT s s I 0 N A 9 I) c.< )h.’(-I, c s t 0 K s One of tlie chief merits of the halogen-sensitive detectors described lies in their exceptional selcctivitj- towards the halogenated pesticides compared with natural materials co-extracted fi-om crops, soils or tissues. There is evidence that this selectivity is significantlj- superior to that shown by the clectron-capture ionisation type of detector. This is particularlj- marked in tlie example shown in Fig. 5 . (’ertainly, in the screening o f a wide range of crops with the halogen-sensitive detector, no sample containing non-pesticidal material gave a response greater than that for electron-capture ionisation detection, and there were indicationi that i n general the superiority in selectivity was about tenfold.I , A 0.5 Lindane (0.6 p.p.m.) F.5.D nTelodrin i Test _____^_I J Control 1-- - 1- --1- -1 1 __ i- I 0 5 16 15 20 25 30 Time, minutes A: E:lt.ctron-capture dctcctor (EC) B: Halogcn-sonsitivt detector (HS) Si mi1 I t an PO 11 s c om p 1 c 1 n t x t i - tar)- gas chromatograms ot control and Iindane - Telodrin trcated spring c a l l - bage: injection, 5 pl 6 mg o f crop. Relative peak areas, IlS/EC : lind- danth 1.1 ; Tclodrin 2.5 F1 g . 6 B ! A: Elcctron-capture detcctor (EC ) B: Halogcln-srnsitivcb detector (HS) Fig 7 . Sirnultaneoiis coniplvrnentar\- gas c hromatogranis of control lanil) aiiti lamb containing di~ldriii or p p ’ - 1)1)15: injcction, 5 pl 2.5 Ing o t tissuc Relative peak areas, HS/E:( .dit~lrlrin 1.3 ; $p’- J)I)E 1 * 1958 GOULDEN, GOODWIN AND DAVIES [Analyst, VOl. 88 Another feature in favour of the halogen-sensitive element is that there is linearity of response with increase in load up to at least 100 nanograms. This could be of advantage in quantitative work, by reducing the need for repeated dilution and injection in the analysis of extracts of unknown pesticide concentration. The cost of these detectors (El5 to fll8) is not high, and in tests where the highest sensitivity (0.1 nanogram) is not required, simple and inexpensive circuitry that does not involve the use of either an electrometer or amplifier is adequate.This arrangement could be of immediate value in analyses in the near-residue range, e.g., for chlorinated pesticides in field-strength dusts, fertilisers, wool, wood, plastics and hardboard, and after a tenfold concentration of the sample extract (by partition or evaporation, or both), could be applied to the low residue range. Indeed, it is probable that this concentration step can be eliminated simply by using a larger input resistor, and feeding the signal directly into a high- impedance l-mV recorder. The main disadvantages of these halogen-sensitive detectors, as has been already indi- cated, lie in their indifferent electrical stability under the conditions used, and the difficulty of conditioning the cells rapidly and maintaining them at high sensitivity.I t must be remembered, however, that the detectors have been used in a manner quite unlike the one for which they were designed. I t is possible, therefore, that design and development work on halogen-sensitive elements for use in gas - liquid chromatographic residue analysis could overcome both of these problems and, moreover, effect sufficient increase in sensitivity to make them of even greater value than the electron-capture type. Most of the work reported here was performed with the Ozotron Type H detector, but the Type J shows considerable promise. The unglazed ceramic sheath of the latter, however, may not be as impervious as the glass of the former. Against this, the Type J is easier to clean and, perhaps, to condition since it can be taken apart. Further, it possesses a higher thermal capacity, which makes it somewhat less sensitive to slight fluctuations in the power supply to the heater filament. In conclusion, it is considered that the halogen-sensitive detector when used in com- bination with the electron-capture ionisation detector in the manner described, provides a technique with a potential value at least comparable with that of the multi-column “spectro- chromatogram” (described in Part I of this paper), for improving the certainty of identifying chlorinated pesticides in analysis of residues.Apart from its complementary operation with the electron-capture ionisation type, it has great potential value as an alternative detector in gas - liquid chromatographic analysis of chlorinated pesticides, both on the residue and near-residue scales. We thank Mr. A. Richardson of the Tunstall Laboratory of “Shell” Research Ltd. for focussing our attention on halogen-sensitive detectors, Mr. D. M. Barnett for much careful experimentation, Mr. G. Rlunkell for the design and construction of the stabilised power pack, and Dr. R. A. E. Galley, Director, and Mr. J. G. Reynolds, Associate Director, of the Woodstock Agricultural Research Centre, for their interest and encouragement during the course of this work. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. Goodwin, E. S., Goulden, R., and Reynolds, J. G., Analyst, 1961, 86, 697. Watts, J. O., and Klein, A. K., J . Assoc. Ojf. Agr. Chem., 1962, 45, 102. Taylor, A., Analyst, 1962, 87, 824. “Aerograph Research Notes,” Wilkens Instrument and Research Inc., Walnut Creek, California, Lovelock, J . E., Anal. Chem., 1963, 35, 474. Goulden, R., Goodwin, E. S., and Davies, L., Analyst, 1963, 88, 941. Cremer, E., Kraus, Th., and Bechtold, E., Chew.-1ng.-Tech., 1961, 33, 632. “Leak Detector Type HA,” Instruction book (reference 4092-61. Ed. A .), Associated Electrical Received June 27th, 1963 Summer Issue, 1962. Industries Ltd., Electronic Apparatus Division, Leicester.

ISSN:0003-2654    

DOI:10.1039/AN9638800951

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

The Determination of Tin in Beer K Y I<. ( - . ROONEY ‘4 method is tlescribctl for tleterniining small anioiints 0 1 tin in beer after l’lie metliocl has l ~ ~ l i i made rapid and simple 1 ) ) ~ .A \vet-oxidation o f tlic sample. making use o f tlic atlvaritagvs o f siibtractivc cathotlt.-ra)~ polarography. detection limit o f 0 . 0 0 7 ,ug pcr 1111 of t i n in beer Ilas lwcn o1)tainecl. ,- I HE practice o f supplying beer in tinned cans has increastvl grcatlj- diiring tlie last year or tmo, and this has led to a demand for thc dtdc1rminatic)n of tin in beer samples. I t lias k e n reported that as little as 0.02 p.p.m. of tin can giiyc riw to haze\ in heer,1.233 antl nit11 tin conccntrations greater than 0-1 p.p.m. tlic tlc.grec of tiirhi(1ity is aestheticallj. intolerable I t is necessary to use a method of high senc;iti\ity for cktermining lcss tlian 0.1 1xp.m.of tin, antl, becausc it is likely that s e i ~ r a l otlier metals will be present a t 5imilar concentration levcls, the method should also he specific. Illost of tlie colorimetric methods for dvtermining tin lack the required sensitivity, and :ire prone to interference. 1;or example, according to Sandel14 no colour is produced witli tlitliiol wlien less than 0 . 3 pg of tin per ml is present, and bismutli, copper, nickel, cobalt, mercury, silvcv-, lead, cadmium and some otlicr metal\ inttlrfere. Polarography lias heen used lor (letermining tin in foods ; Godar and Alexander5 de t ern1 i ne d tin aft c L r co- p r eci pi t a t ic )n w i t 1 1 a 1 um in i um h y d r o xi de , an ( 1 Il I ark 1 and an tl S 11 en ton introduced a simpler method involving no scparation.Condliffe and Skrimshirei used ;L single cell cathode-ray polarograpli for detcwnining tin in various foodstuffs, hiit t hej7 did not report any results loLver tlian 20 to 30 p.p.m. I n general, the permissible tin content of foods is quite high, and it is only becaiise of its tlelcterious effect on tlie appearancc of the product that determinations of such liigh scmsitivitJ7 art’ required for beer. I i r w n t develop- ments in polarograpliy liave led to the dcwlopment of instruments witli much tiighcr senyi- tivity, and it was considered likely tliat t l i v use of subtractiL-c polarography8 u70ultl $\,e ;t lower detection limit tlian has previously been attainable.Ex I~El<IMESTAI< 1 lie detection limit for tin in tlic sc.lectetl bwe electrolyte (5 > I~~~drochloric acid) ~ ’ a b determined by measuring the peak obtaincd for tin on ten solutions containing 0.02 pg of tin per ml. The standard deviation, 0, obtaintd corresponded t o 0.0028 pg per ml, which on a 20 hasis corresponds to a detection limit of 0.0056 pg per ml, Tlicsc figures were obtained by r . Sample 1 .ight alc. (hottlvd) Lager (camnctl) . . . . Guinncss (bottled) . . *Mean of 3 tleterininations subtracting base-line effects, etc., 5 N hydrocliloric acid being used in the reference r ,-I1 of the polarograph. The improvement in sensitivity amply justifies the slightly increas-d com- plexity of the procedure. The improvement is illustrated in Figs.1 t o 4.960 ROONEY DETERMINATION O F TIN IN BEER [Analyst, Vol. 88 The wet-oxidation of beer samples was next investigated, and it was found that digestion with nitric, perchloric and sulphuric acids by the proposed procedure gave satisfactory destruction of the sample. Recoveries of added tin were investigated by making additions Fig. 1. Polarogram of beer containing: A, 0.031 pg of tin per ml; B, 0.18 p g of cadmium per ml. (Sensitivity &) Fig. 2. Polarogram of 5 N hydrochloric acid base elec- trolyte, cell (11). (Sensitivity 1) 25 Fig. 3. Subtractive polaro- gram o f beer containing: A, 0.031 pg of tin per ml; B, 0.18 pg of cadmium per ml. (Sensitivity +6) Fig. 4. Subtractive polaro- gram of beer containing: A, 0.031 pg of tin per ml. (Maxi- mum sensitivity) of stannic sulphate solutions to beer samples.obtained, are shown in Table I. recovery, and the method was applied to several beer samples. samples are shown in Table 11. The additions made, and the recoveries It was considered that these results indicated satisfactory The results obtained on these TABLE I1 RESULTS OBTAINED ON VARIOUS BEER SAMPLES Sample Guinness (bottled) . . . . . . . . Light ale (bottled) . . . . . . . . Swiss lager (bottled) . . . . . . Same Swiss lager (canned) .. . . Swiss lager (canned) . . . . . . Swiss lager (bottled) . . . . . . Dutch lager (canned) . . . . .. Same Swiss lager (aluminium can) . . Tin, pg per ml (0.01, <0.01, (0.01 (0.01, <0.01, (0.01 <0-01, <om01 0.066, 0.058 0.044, 0.040 0.10, 0.11 0.031, 0.026 0.056, 0.061, 0.069 It is probable that the limit of detection of the whole analytical procedure will be at a higher concentration than that found for pure tin solutions, and this was found to be so in practice.It has been suggested9 that the standard deviation of the blank value is likely to be the limiting factor in an analytical procedure, and the standard deviation of the blank value was therefore determined by analysing eight blank solutions over a period of 4 days; theI>ecember, 19631 ROOXEY: DETERMINATION 01; TIK I N BEER 96 1 determinations were carried out by a single analyst. The standard deviation was fount1 to be 0.009 pg of tin per ml in the cell, and by using the procedure described below this corres- ponds to 0.0036 pg per ml in the beer sample.The limit of detection of tlie method on a 2~ basis is considered to be about 0-007 pg per ml, and samples in our laboratory which give results indistinguishable from the blank value are reported as < 0 4 1 pg per ml. Stannic chloride has a boiling-point of 114" C,l0 and use has been made of its volatilitj- as a method of separation.ll I t is likely, therefore, that if samples rich in chloride are wet- ashed some tin may be lost by volatilisation; some of the samples investigated in the earl>- part of this work gave low results when charred with nitric acid alone. I t was found that when the sample was first charred with sulphuric acid, as in the procedure described below, satisfactorjy results were obtained. In the 5 N hydrochloric acid base electrolyte used for determinating stannic tin, the only likely interfering substance will be lead.As the lead content of beer is usually low, the bulk of the lead present will probably come from the reagents, and will therefore be deducted with the blank value. Several samples had been investigated by making separate determina- tions of lead in a base electrolyte buffered to pH 4.6, at which tin is non-reducible, showed that the correction to be made for lead content was, in every experiment, negligible. So such correction is therefore incorporated in the proposed procedure, although if desired it can be made by using a separate determination of lead on a second portion of the solution. The statutory limit for the lead content of beer is 0-5 p.p.m., but this limit is seldom reached in practice.If substantial amounts of lead are present, then the shape of the reduction peak obtained will be altered; the tin'' reduction peak is very much sharper than the lead" reduction peak, and little experience is necessary before the slightly more blunt combined tin - lead peak can be readily differentiated from a pure tin peak. Interference from arsenic is prevented, as it will have been oxidised to arsenic", and the presence of germanium", niobium, rhenium or thallium is unlikely. These are the only other elements that will give a r-eduction peak at -0.5 iwlts. M ETH o 11 REAGEKTS- Perclzlovic acid, sp.gr. 1 -54-Lead-free for foodstuffs analysis. AYifvic acid, sj5.p. 1-42-Lead-free for foodstuffs analysis. Sztljdzziric acid, s P . g ~ .1.84-Lead-free for foodstuffs analysis. Hydvochloric a c i d , spp.p. 1.18-Lead-free for foodstuffs analysis. AITAKATI~S- lead blank values. Differential Cat hode-Ray Polarograph. All glassware including cover glasses should be of Pyres or similar glass to minimise tlie The polarograph used for this work was the Southern Analytical ill660 P m c m r RE- Pour the well-shaken beer sample into an open beaker, and leave for about 20 minutes to liberate most of the dissolved gases. If this precaution is not taken, it is difficult to measure tlie sample bj- means of a pipette. 1 ransfer by means of a pipette 25 ml of the sample to a 100-ml beaker, cover the beaker, and evaporate almost to dryness; it will be necessary to heat gently until frothing ceases. Add 2 ml of sulphuric acid, and char the sample by gently heating.,4110~ to cool somewhat, and add 5 ml of nitric acid. Evaporate until fumes of sulphur trioxide appear; thc sample should char again. Allow to cool, and add a further 5 ml of nitric acid and then 3 ml of perchloric acid. Evaporate gently until fumes of perchloric acid appear, and then until fumes of sulphur trioxide appear. Remove the cover from the beaker, and evaporate the solution to dryness. Allow the residue to cool, add 5 ml of hydrochloric acid, cover the beaker, and warm gently to dissolve the residue. Transfer the solution to a 10-ml calibrated flask and make up to the mark with distilled water (see Kote 1). Transfer a portion of this s;olution to a polarograph cell, remove dissolved oxygen with a stream of nitrogen, and measure the tin peak at - 0-5 volts verszI.s the mercury pool anode (see Note 2).I t was found preferable to use 5 s hydrochloric acid in the reference cell of the instrument, for highest sensitivity, <.962 ROONEY: DETERMINATION OF TIN IN BEER [Analyst, Vol. 88 and to measure the difference wave obtained. At least two reagent blank solutions should be processed, omitting the sample of beer, and these blank solutions should also be measured against the 5 N hydrochloric acid reference solution. NOTES- For high sensitivity polarography it has been reported12 9 1 3 that de-ionised water can cause difficulties, and the use of distilled water for base electrolytes is recommended. The polarograms obtained should contain two or perhaps three peaks, caused by copper at -0-25 volt, combined tin - lead at -0.5 volt and possibly cadmium at -0.7 volt.If this simple pattern is not obtained, or if large concentrations of copper or tin appear to be present, the wet-oxidation is probably incomplete and the determination should be repeated. It is necessary for the final evaporation to be carried out slowly, so that perchloric acid condenses on the walls of the beaker. The technique described has been successfully used on samples of lager, light ale and Guinness. 1. 2. 1. 2. 3. 4. 5. 6. 7 . 8. 9. 10. 11. 12. 13. REFERENCES Luers, H., Brau. tech., Wien, No. 4, April, 1928; through Brass. Malt., 1928, 18, 99-106; abstr. in Chabot, L. G., Delie, P., and Gaermynck, G., Bidre Boiss. ferm., 1941, 429. Helm, E., J . Inst. Brewing, 1939, 45, 80. Sandall, E. B., “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience Godar, E. M., and Alexander, 0. R., Ind. Eng. Chem., Anal. Ed., 1946, 18, 681. Markland, J., and Shenton, F. C., Analyst, 1957, 82, 43. Condliffe, W. F., and Skrimshire, A. J. H., J . Polarogra9hic Soc., 1961, No. 1, 10. Davis, H. M., and Seaborn, J. E., “Advances in Polarography,” Pergamon Press, London, New Wilson, A. I., Analyst, 1961, 86, 72. “Handbook of Chemistry and Physics.” Forty-first Edition, Chemical Rubber Publishing Co., Hoffman, J . I., and Lundell, G. E. F., J . Res. Nut. Bur. Stand., 1939, 22, 465. Ferrett, D. J., Milner, G. W. C., and Smales, .4. A,, Analyst, 1954, 79, 731. Rooney, R. C., Ibid., 1958, 83, 83. J , Inst. Brewing, 1928, 34, 455. Publishers Inc., New York and London, 1959, pp. 854 to 867. York and Paris, 1961, Volume 2, pp. 239 to 250. Cleveland, Ohio, 1959, p. 673. Received April 8th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800959

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

December, 19631 NASH ANT) ANSLOW 963 The Polarographic Determination of Lead in Steels and Copper - Zinc Alloys BY J. R. SASH AND G. W. AKSLOiV (/nduslvial Cheunzstry (;voztp, U . K.A. I:.A ., .4 . JV.li I:., '4 Idmnaston, Revkshzve) 1,ead is separated from interfering elements by an ion-exchange medium. The difference in electrical conductivity between the lead-free and lead- containing eluates is used as a variable to effect the separation by automatic control. The separation technique and polarographic tletermination are described. Coefficients o f variation for lead contents of 0.01 per cent. and 0-3 per cent. are &4-25 per cent. and & I - O i per cent., respectively. THE object oi this work was the development of a simple method for determining low concen- trations of lead (about 0.01 per cent.) in steels and copper - zinc alloys.I t was desirable that the method developed should use instrumentation already available and be sufficiently flexible to apply to both plain carbon and alloyed steels. did not appear to meet all these requirements, but Carson8 has described an ion-exchange separation of lead and zinc from cobalt, in 2 M hydrochloric acid solution, and has suggested that this should be applicable to lead in steel. The work of Icraus and N e l ~ o n , ~ with anion-exchange resins, suggested that separation of lead from iron and most of the usual alloying elements in steel, and also copper, should be feasible at hydro- chloric acid concentrations of either 9 M or 2 M. Separation in hydrochloric acid solutions by anion exchange was at tractive because this medium could act as solvent, eluent and polarographic base electrolyte.Further, if 2 III acid was chosen as the initial eluent, the change in electrical conductivity caused by the dilution of the eluting acid could be used to operate an automatic system for isoIating the fraction containing the lead. In view of these considerations, the method chosen involved an anion-exchange separation and determination of lead by linear-sweep cathode-ray polarography. Published papers1 to EXPEKIMENTAL Initial experiments using British Chemical Standard KO. 212 (0.3 per cent. leaded steel) indicated that the ideal volumes and concentrations of eluting acid were 50 ml of 2 M, 100 ml of M and 300 ml of 0.005 M hydrochloric acid. The separation of iron and lead was further improved bv allowing some degree of mixing between each concentration before entering the column.B19 cone 2nd socket 5eparati ng funnel I i e l v i n cable strapping Fig. 1. ETcedcr funnels (350-ml and 100-ml capacity)\ 45 volts -1 I II - -{I Sensitive relay 1 I 4- Fig. 2. Block diagram of electrical circuit Description: for the sensitive relay, a 50-mV electronic recorder with a microswitch on the pen-travel was used. By increasing the resistance of potentiometer A to 500 ohms, a sensitive electro-magnetic relay can be used in place of the recorder, which has a 50 pA operating current on a 2000 ohm coil _.-.- Conductivity Concentration of iron 2. Concentration of lead - - _ _ '. ORDINATE SCALE x I: mgof Pb per25 ml x 100: mg of Fe per 25 ml 2.0 \ I x 20: conductivity units I \ \ \ \ ', '.\ I I \ ' n I n . c L \ ' ' \ 200 300 400 500 I I00 Volume of eluate, ml Fig. 3. Graph relating electrical conductivity, lead concentration and iron concentration to volume of eluate diverting the eluate from waste to sample beaker. The tap was re-set manually between separations. The tap relay was controlled by the sensitive relay (in our case a recorder), which was coupled to the output of the conductivity bridge, so that when the conductivity of the eluate was the same as the bridge setting, the tap relay was energised (see Fig. 2). The apparatus proved to be reliable and the efficiency of the separation is shown in Fig, 3. For this graph, iron was determined colorirnetrically and the lead estimated by using a K 1000 polarograph.December, 19631 OF LEAL) I N STEELS AXD COPPER - ZIKC ALLOYS LI ETH o L) All reagents used should be of analytical-reagent grade. STEELS- 965 Dissolve 1 g of sample in 25 ml of diluted hydrochloric acid (1 1) with tlie minimum of heating, and add nitric acid, sp.gr.1.42, dropwise until no further reaction occurs. Roil the solution free from nitrous fumes, cool, filter off any residue and dilute to 50 ml with water. Regenerate the column with 350 ml of 2 hi hydrochloric acid and introduce the sample solution into the cup at the top of the column. Set the servo tap to waste, and adjust the flow-rate through the column to 8 to 10 ml per minute. 4dcl a further 50 ml of 2 11 hydro- chloric acid to the cup and fill the two eluant feeder funnels with 100 ml of 31 and 350 ml of 0.005 M hydrochloric acid.Open the taps on the feeder funnels, and switch on tlie con- ductivity bridge and associated circuits. LYhen the conductivity of the eluate falls to the bridge setting (see “Bridge Setting”), the servo tap diverts the eluate to a sample-collector and tlie lead-containing fraction of the eluate is collected. Evaporate this solution to approximately 30 ml, add 5 ml of 3 per cent. w/v hydroxylammonium chloride solution while still hot, and boil the solution to complete the reduction of iron. Cool, dilute with water to 50 ml, and record the peak-height on a K1000 polarograph at -0.45 volt, against a standard calomel electrode, using an initial potential of -0.35 volts. A blank experiment on all reagents should be performed, and tlie sample peak-height corrected for the blank value.Convert to percentage lead by reference. to a calibration graph. (’ormzIt - ZISC ALLOYS- liydrochloric acid (1 ant1 proceed as for steels. C XL I 13 RAT I o K - Tlie calibration graph was prepared by analjrsing 1 g quantities of Specpure iron, to whicli were added 0, 1-0, 2.0, 3.0, 4.0 and 5 0 ml of standard lead solution (1.5980 g of Spcc- pure lead nitrate, Pb(SO,),, diluted to 1 litre). The quoted lead additions cover a rang(: from 0 to 0.5 per cent. for a 1 g sample. K n I r x E SETTIXG- Obtain a graph relating conductivity, and iron and lead concentrations to volume of eluate (see Fig. 3 ) by performing an experiment on British Chemical Standard So. 212 (0.3 per cent.leaded steel) and analysing the eluate in 25-ml fractions. From this graph deduce the conductivity for the optimum separation of iron and lead. Set the conductivitv briclgv at this figure. For l c w l contents between 0 and 1 per cent. dissolve 1 g of sample in 30 ml o f diluted 1) and 1 ml of nitric acid, sp.gr. 1-42, with the minimum of heating, AITIACATIOS TO CAKROS STEEI. Six 1-g samples of Specpure iron (Johnson hlatthey & Co. Ltd.) containing 0 . 0 1 per cent. (added) lead impurity, together with a blank sample, were analysed as described in the method. The results obtained after correction for the lead content oithe blank sample showed that the mean value for the recovery of lead was 102 per cent., the standard deviation was 24-34 x Suitable standard steel samples with a lead content between 0.01 per cent.and 045 per cent. were not available. Consequently, we tested the reliability of the method with an actual steel, British Chemical Standard No. 212 (leaded steel, certificate value 0.28 per cent. of lead). Six 1-g samples of this steel, together with a blank sample were analysed. Results showed that the mean value €or the lead content was 0.282 per cent., the standard dmriation was -13.03 x and the coefficient o f variation was *1.07 per cent, lf‘hen British Chemical Standard Yo. 212/l (leaded steel, certificate value 0.22 per cent. of lead) became available, it was also examined. Results showed that the mean value for the lead content was 0.215 per cent., the standard deviation was If3.29 x lW3 and the coefficient of variation was & l a 5 3 per cent. The results show that in carbon steels, lead contents of 0-01 per cent.to 0.3 per cent. can be determined with reasonable accuracy. and the coefficient of variation was +4-25 per cent.966 NASH AND ANSLOW [Analyst, Vol. 88 APPLICATION TO ALLOY STEELS AND COPPER-ZINC ALLOYS- Kraus and Nelsong showed that most interferences are eliminated by the column separa- tion described, which should leave only zinc", cadmium", indium'", antimony''', and, possibly, molybdenum" and tin'' ions in the lead fraction. In polarographic analysis with a dilute hydrochloric acid supporting electrolyte, zinc and cadmium have sufficiently negative half-wave potentials to prevent their interference with the determination of lead.In fact, zinc and cadmium in the lead-containing eluate could be quantitatively determined polaro- graphically. It was most unlikely that indium would be present in steels or copper-zinc alloys, but if present probably would not interfere. Antimony would give a polarographic wave only in strongly acid conditions, but this wave is well removed from that of lead. The presence of molybdenum was shown to have no effect on the recovery of lead. Any tin present in the lead-bearing eluate could seriously interfere with the polarographic determination as the waves of tin and lead in chloride media coalesce. Small amounts of tin tended to remain on the column and were eluted with the lead in the next sample. This problem was solved for 1 g samples containing 0.2 per cent.of tin by using at least 350 ml of regenerant acid between separations (see Table I). TABLE I DETERMINATION OF LEAD IN BRITISH CHEMICAL STANDARD NO. 275 STEEL CONTAINING (Certificate value: 0.04 per cent. of tin, 0.005 per cent. of lead) Tin added, Total tin, Lead found, Recovery, per cent. per cent. per cent. per cent. ADDED TIN Nil 0.20 Nil Nil 0.20 Nil 0.04 0.24 0.04 0-04 0.24 0.04 0.004, 0.005, 0.004, 0.005, 0.005, 0-005, 98 104 98 102 108 100 To determine the effect of tin in a copper - zinc alloy on the recovery of up to 0.2 per cent. of lead, we made a 60+40 copper - zinc alloy and added 0.2 per cent. of lead, and also we made a 60+40 copper - zinc alloy and added 0.2 per cent. of tin. In one series of experiments (6 samples) we added 0.01 per cent. of lead, and found a recovery of between 94 per cent.and 105 per cent., with a mean value of 100 per cent. In the second series of experiments (4 samples), we added 0.2 per cent. of lead, and found a recovery of between 95 per cent. and 100 per cent., with a mean value of 97.5 per cent. These experiments showed that lead, present to the extent of 0.01 per cent., could be determined accurately in carbon and alloy steels, and in copper - zinc alloys, provided that the tin content did not exceed 0.2 per cent. CONCLUSIONS The method has proved reliable for the routine determination of lead contents between 0.005 per cent. and 1 per cent. in a wide variety of steels and copper - zinc alloys, the only limitation being that the tin content for a l-g sample should not exceed 0.2 per cent. We thank the Director, A.W.R.E., Aldermaston, for permission to publish this report. REFERENCES 1. .Milner, G. W. C., and Slee, L. J., Analyst, 1957, 82, 139. 2. Elwell, W. T., and Gidley, J . A. F., Anal. Chinz. Acta, 1961, 24, 71. 3. British Standard 1121 : Part 41 : 1960. 4. Scholes, P. H., Analyst, 1961, 86, 116. 5. Rooney, R. C., Ibid., 1958, 83, 83. 6. Bush, G. H., Ibid., 1954, 79, 697. 7. Bricker, L. G., and Proctor, K. L., I n d . Eng. Chem., Anal. Ed., 1954, 17, 511. 8. Carson, R., Analyst, 1961, 86, 198. 9. Kraus, K. A., and Nelson, F., Proc. Intern. Conf. Peaceful Uses Atomic Energy, United Nations, 1955, Received March 6th. 1963 Volume VII, p. 113.

ISSN:0003-2654    

DOI:10.1039/AN9638800963

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

December, 19631 SHORT PAPERS 967 SHORT PAPERS The Determination of Glucose in the Presence of Maltose and Isomaltose a Stable, Specific Enzymic Reagent BY I. D. FLEMING AND H. I?. PEGLER (Glaxo Research Ltd., Greenford, Middlesex) DURING investigations on the degradation of starch by fungal enzymes, a simple and accurate micro method was required for determining glucose in the presence of other reaction products. The Barfoed reagent of Tauber and Kleinerl is specific for glucose in the presence of maltose, but reproducible results are obtained only if conditions are carefully standardised ; even then, the reagent responds to isomaltose as well as to glucose. Other methods were therefore investigated, and of these a modification of the glucose oxidase procedure was found most satisfactory.All the glucose oxidase enzyme preparations tested, unless prepared in tris buffer as described by Dahlqvist,2 were found to react with maltose and isomaltose as well as with glucose. This enzyme reagent was further tested, and found to be highly specific for glucose, though it suffered from the disadvantage of relative instability. The TABLE I ESTIMATION OF GLUCOSE IN THE PRESENCE OF MALTOSE OR ISOMALTOSE BY GLUCOSE OXIDASE AND BARFOED REAGENT Mixture contained- 50 pg glucose, plus Glucose oxidase Glucose oxidase Barfoed in phosphate in tris buffer reagent 500 pg maltose 65 50 60 500 pg isomaltose 63.5 50 103 Glucose found (pg) by using- colour developed was linearly related to glucose concentration only up to 50 pg of glucose per ml of solution.The addition of up to 40 per cent. w/v glycerol, as recommended by Washko and Rice,3 was found to increase the stability considerably, without altering the specificity of the Dahlqvist enzyme reagent, A number of colour-developing reagents were tested ; of these o-dianisidine hydrochloride, at the specified concentration was found to be the most sensitive, particularly in presence of 5 N sulphuric acid.3 These modifications have been found to increase the life of the enzyme reagent from a few days to over three months a t 4" C, and to increase the linear response up to 100 pg of glucose per ml of test solution. METHOD REAGENTS- Glucose oxidase, 10 wzg-C. F. Boehringer und Soehne, Mannheim, Germany. Horseradish peroxidase, 1 wg-Worthington Biochemical Corporation.o-Dianisidine hydrochloride, 10 nzg-Prepared from o-dianisidine. Tris glycerol buffer, p H 74-Prepare by dissolving 6 1 g tris- (hydroxymethyl) aminomethane in 85 ml of 5 N hydrochloric acid, dilute the solution to 1 litre with water and add 660 ml of analytical- reagent grade glycerol. Adjust the pH to 7-0 if necessary. PROCEDURE-- Mix the solutions and incubate them a t 25" C for 1 hour. After incubation, add 4 ml of 5 N sulphuric acid, mix, and measure the optical density a t 525 mp. (The colour is stable for a t least 12 hours.) Prepare standards from a stock solution containing 2 mg of analytical- reagent grade glucose in saturated aqueous sorbic acid. This solution keeps indefinitely. Dilute stock solution to 1 + 20 for a working standard, and use this standard for blank determinations when performing analyses.Add 2 ml of the reagent to 1 ml of solution containing 0 to 100 pg of glucose. (Include standard glucose solutions.)968 SHORT I?APERS [Analyst, Vol. 88 RESULTS The effect of maltose and isomaltose on the determination of glucose by glucose oxidase Fig. 1 shows the response curves for the various reagents. (with phosphate or tris buffer) and Barfoed reagent is shown in Table 1. 20 eo 60 80 I00 GI ucose, m pg Fig. 1. Calibration curves for glucose: curve A, Dahlqvist reagent; curve B, modified reagent: curve C, Glucostat reagent DISCUSSION As shown in Table I, the Barfoed copper reagent and the Glucostat glucose oxidase reagent, which contains phosphate buffer, are not specific for glucose in the presence of maltose and iso- maltose. When the phosphate buffer is replaced by tris buffer a t pH 7, the maltose and isomaltose activities are inhibited, and the reagent is specilic for glucose. A further disadvantage of the Glucostat reagent is its low sensitivity compared with either Dahlqvist or the proposed reagent (see Fig. 1) both of which are specific for glucose. The proposed reagent although less sensitive than that of Dahlqvist, has a greater range and the additional advantage of being completely stable for a t least six months. REFERENCES 1 . 2 . 3. Tauber, H., and Kleiner, I. S., J . Biol. Chem., 1932, 99, 249. Dahlqvist, A., Biochem. J . , 1961, 80, 547. Washko, M. E., and Rice, E. W., Clipz. Chenz., 1961, 7, 542. Received July 3rd, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800967

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

December, 19631 SHORT PAPERS 969 Improved Paper-chromatographic Separation of Sugar Phosphates by using Borate-impregnated Paper BY D. P. AGARWAL, G. G. SANWAL AND P. S. KRISHNAN (Division of Biochemistry, Lucknow University, Lucknow, U . P., India) THE borate-complexing technique for separating sugar phosphates has been successfully used in ion-exchange chromatography1 g 2 and paper electroph~resis.~ >4 I t has met with limited application in paper chromat~graphy.~~~~~ Cohen and Scott5 used boric acid in neutral ethanol to achieve a sharp separation between aldopentose-5- and 3-phosphates, but the separation among the individual 5-phosphates or the 3-phosphates was not satisfactory. The solvent system failed to effect a satisfactory resolution among the hexose-monophosphates. According to Isherwood,6 incorporation of boric acid in the alkaline n-propanol system readily effected a separation between glucose- 1-phosphate and 6-phosphate.Baron and Brown7 achieved success in separating ribose- 2-, 3- and 5-phosphates by introducing boric acid in the propanol - ammonia solution system. A scheme of paper-chromatographic separation of a mixture, containing essentially hexose- phosphates, by using borate-containing solvents has not been proposed. Nor has, to our knowledge, borate-impregnated paper been used in the separation. A simple and rapid method, in which borate-impregnated paper is used, has been developed by us, and this permits a fairly sharp separa- tion of the components from a mixture of orthophosphate and five sugar phosphates. METHOD PROCEDURE- The samples of sugar phosphate in the form of sodium, potassium, or barium salts, obtained from recognised commercial sources, were de-cationised with a cation-exchange resin, in the hydrogen form, and the solutions neutralised to pH 7 with 0.1 M ammonia solution.The basic borate- free solvent system consisted of 60 volumes of t-butanol (this was superior to n-propanola), 30 volumes of ammonia solution (sp.gr. 0.88) and 10 volumes of water. In the borate-containing solvent system, water was replaced by a solution of sodium tetraborate to give a final concentration of 0.01 M borate. Whatman No. 1 filter-paper strips (12 cm x 38 cm) were used, either without any treatment or after being dipped in borate solution (0-001 M to 0.01 M) and dried a t room tem- perature.Where pre-complexing of esters with borate before application to paper was desired, stock solutions of the phosphate esters were made in 0.01 M borate solution. Portions of the solutions, containing 0.1 to 0-2 pmoles of sugar phosphates, were applied to the paper and dried. One-dimensional descending-solvent chromatograms were developed for about 24 hours at room temperature, the solvent being allowed to overflow and drip from the serrated edge of the paper. The air-dried chromatograms were treated with the spray reagent of Runeckles and Krotkov.* RESULTS AND DISCUSSION Spots were usually round or slightly pear-shaped. The migrations of various esters have been calculated as R, values. This is the distance from the base line to the centre of each phosphate ester spot, as related to the corresponding distance of the orthophosphate spot, the latter being taken as 100 for convenience.Typical R, values, obtained for various sugar phosphates under the influence of borate ions, are given in Table I. There was little separation between the various sugar monophosphates in the non-borate system. The use of borate effected an alteration in the migratory behaviour, and the degree of alteration depended on several factors. The glucose-1-phosphate spot invariably moved faster than the other phosphates under all the conditions used, showing that this ester remained un- complexed with borate. The uncomplexed orthophosphate followed the glucose-1-phosphate under identical conditions. Glucose-1- and 6-phosphates separated from each other after merely introducing borate into the solvent system.This confirmed the observations of Isherwood,a but fructose-6-phosphate and ribose-5-phosphate were not separated from glucose-6-phosphate. The use of borate pre-complexed esters did not offer any significant advantage. Considerable improve- ment in resolution resulted when borate-impregnated paper was used instead of untreated paper. The concentration of borate critically affected the optimum separation. When 0-01 M borate was present, fructose-6-phosphate separated from the other two non-glucosidically phosphorylated monoesters, but ribose-5-phosphate and glucose-6-phosphate did not separate sharply. n’hen970 SHORT PAPERS [Analyst, Vol. 88 the concentration of borate was halved (0.005 M), a separation between glucose-6-phosphate and ribose-5-phosphate tended to occur.At 0.001 M borate-ion concentration, glucose-6-phosphate moved as fast as orthophosphate. Such behaviour is understandable if one assumes, as did Khym and Cohn,l that glucose-6-phosphate is capable of dual behaviour, changing from the highly complexed furanose form to the less complexed pyranose form a t lower borate-ion con- centration. Excellent separation of all the components was found to occur when 0.002 M borate was used for paper impregnation. TABLE I R, VALUES OF SUGAR PHOSPHATES ON PAPER CHROMATOGRAMS UNDER THE INFLUENCE O F BORATE IONS (R, value for orthophosphate = 100) RP values for System Glucose- 1 - phosphate Non-borate system, control . . 120 Borate system, untreated paper : Irrigating solvent containing Esters pre-complexed with 0.01 M 0.01 M borate .. . . .. 121 borate . . . . .. .. 118 0.01 M . . . . . . . . 117 0-005 M . . .. .. .. 125 0.004 M . . .. . . .. 127 0.003 M . . .. .. .. 121 0.002 M . . .. . . . . 125 0.001 M . . . . .. . . 140 Borate-impregnated paper : Glucose-6- phosphate 116 106 87 58 61 65 66 86 97 Fructose-6- phosphate 114 111 83 83 79 63 73 76 86 ~~ Fructose-l,6- diphosphate 59 65 41 40 31 27 32 31 40 -3 ~ ~ _ _ _ _ _ ~ Ribose-5- phosphate 114 107 70 63 55 50 47 41 62 Combination of techniques : Esters pre-complexed with 0-01 M borate, on 0.01 M borate-im- pregnated paper . . .. 114 61 79 40 61 Irrigating solvent containing 0.01 M borate, on 0.005 M borate-impregnated paper . . 108 64 89 58 69 Combinations of these techniques, for example, by using pre-complexed esters with impregnated paper, or borate-containing irrigating solvent with impregnated paper, did not satisfactorily separate all the components.The results reported above for single components were also applicable when the components were present in a mixture. The present method represents a definite improvement on our earlier chromatographic methodJ9 in which borate ions were not used. One of us (D.P.A.) thanks the Scientific Ii-esearch Committee, U.P., Allahabad, and the Council of Scientific and Industrial Research, New Delhi, for financial assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFER:ENCES Khym, J. X., and Cohn, W. E., J . Amer. Chem. Soc., 1953, 75, 1153. Goodman, M., Benson, A. A., and Calvin, M , Ibid., 1955, 77, 4257. Foster, A. B., and Stacy, M., J . Appl. Chem., 1953, 3, 19. Schwimmer, S., Bevenue, A., and Weston, ?V. J., Arch. Biochem. Biophys., 1956, 60, 279. Cohen, S. S., and Scott, D. B. M., Science, 11950, 111, 543. lshenvood, F. A., Brit. Med. Bull., 1954, 10, 202. Baron, F., and Brown, D. M., J . Chem. Soc., 1955, 2855. Runeckles, V. C., and Krotkov, G., Arch. Biiochem. Biophys., 1957, 70, 442. Agarwal, D. P., Sanwal, G. G., and Krishnan, P. S., Analyst, 1962, 87, 758. Received June 4th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800969

出版商:RSC    

年代:1963

数据来源: RSC