ISSN:0003-2654    

DOI:10.1039/AN95479FX047

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479BX049

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479FP137

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479BP143

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

OCTOBER, I954 Vol. 79, No. 943 THE ANALYST PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETING AN Ordinary Meeting of the Society was held at 6.45 p.m. on Wednesday, October 6th, 1954, in the meeting room of the Royal Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. D. W. Kent-Jones, F.R.I.C. The following papers were presented and discussed : “The Theoretical Basis of ‘Sensitivity Tests’ and their Application to some Potential Organic Reagents for Metals,” by H. M. N. H. Irving, M.A., D.Phil., F.R.I.C., L.R.A.M., and Mrs. H. S. Rossotti, B.A., B.Sc.; “An Investigation of 5-Nitroso-oxine as an Analytical Reagent,” by H. M. N. H. Irving, M.A., D.Phil., F.R.I.C., L.R.A.M., and R. G. W. Hollingshead, M.A. NEW MEMBERS ORDINARY MEMBERS Percy Hugh Bond, B.Sc.(Lond.); Frederick Mason Brewer, M.B.E., B.Sc., M.A. (Oxon.), Ph.D. (Cornell), F.R.I.C. ; William Dennis Brighton, B.Sc. (Dunelm.) ; William Burns, A.R.I.C. ; Chwee Leong Chia, M.Sc. (Malaya), A.R.I.C. ; Jenkyn William Davies, B.Sc. (Birm.), A.R.I.C. ; Ernest Lancelot Deeley, F.R.I.C. ; Michael Christie Foster, BSc. (N.Z.) ; John Gareth Fraser; Jacob Jackson, B.Sc., Ph.D. (Lond.), A.R.I.C. ; Ramprasad Kaushal, D.Sc. (Agra), Ph.D. (Manc.), F.R.I.C. ; Ervin George Muller, BSc. (Manc.), A.R.I.C., F.R.S.S. ; Mervyn Desmond Rogan, F.R.I.C. ; William Kenneth Leslie Thomas, M.Sc. (Lond.), B.Sc. (Wales), D.I.C. JUNIOR MEMBERS Sylvia Rosemary Benson, B.Sc. (Lond.) ; Ronald Joseph Garwood; Stanley Graham, A.R.I.C. ; Frank Hobson, B.Sc. (Lond.) ; Zdenek Hybs, B.Sc., Ph.D. (Birm.), A.R.I.C. ; David Thomas Simmonds, B.Sc. (Lond.), A.R.I.C. ; Garth John Smart, B.Sc. (Sheff.) ; Derek Edward Tutt, B.Sc. (Lond.), A.R.I.C. DEATHS WE regret to record the deaths of Harry Hurst James Waywell. 593

ISSN:0003-2654    

DOI:10.1039/AN9547900593

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

594 BARNARD : THE MASS SPECTROMETER [Vol. 79 The Mass Spectrometer as an Analytical Instrument BY G. P. BARNARD (Presented at the meeting of the Physical Methods Grozlp on Tuesday, April 6th, 1954) A brief description of some modern iristruments , techniques in operation and analytical procedures is given. Some examples of quantitative organic and inorganic analysis are considered. In respect of accuracy and speed in these analytical applications, mass spectrometry is shown to compare favourably with other physical methods of analysis. The use of the isotope dilution method for the determination of trace elements in solids is described. Other uses of the mass spectrometer for the determination of impurities are noted briefly. DURING the last decade the mass spectromete L: has become an indispensable analytical instrument in much technological progress in many industrial spheres.However, many physical methods laboratories in this country are not yet equipped with mass spectrometers; and for analytical chemists in these laboratories it. may be difficult, without direct experience, to assess the value to them in their work of this new, apparently complex and rather costly machine. Solely on grounds of simplicity and initial cost, the mass spectrometer must compare unfavourably with radiation spectrometers, so that a careful scrutiny is required of the claims of experienced mass spectrometrists as to accuracy, speed, versatility and saving of man-hours in a wide range of analytical work. In this paper an attempt is made to survey, without a great deal of detail, the principles and modern techniques of operation and some recent applications of mass spectrometers.THE PRINCIPLES AND MODERN TECHNIQUES OF OPERATION The treatment of the theme is defined throughout by the requirements of an analytical chemist who has no practical experience of mass spectrometry, but who seeks a general notion of this modem instrument and its capabilities in applications to some analytical problems. It is assumed that, for this purpose, much technical detail would be unwelcome. It may be self-evident, but it is, nevertheless, worth emphasising here, that mass is the key to the identity of any particle under observation. Hence, if a complex mixture is introduced into the apparatus, the mass spectrometer must carry out automatically a sorting process, so that identification according to mass number of all the different mass species is possible.The first problem in design is how to achieve the required degree of mass separation without losing in the process too many particles of all mass numbers. There is bound to be a considerable loss, however good the design; the more so, if a high mass resolution is required. Moreover, imperfections in this respect signify more than just some loss of sensitivity, because the percentage lost at each mass number is not the same. This preferential treatment by the machine of each mass species is known as mass discrimination, and all instruments suffer from this defect to various extents. It follows that when the mass sorting process is completed, a simple counting procedure cannot, without reference to known standards, give a precise indication of the true relative amounts of the various mass species.Hence, with mass spectrometry, as with so many other laboratory methods, comparative measurements are essential for quantitative analysis. Although an absolute first-order single-focusing instrument must be regarded as practically unattainable, it is important for work to continue so that a better understanding can be gained of the various sources of mass discrimination, the causes of instabilities in the system and of aberrations in the particle trajectories through the apparatus. New knowledge gained from such studies makes improved instrumentation possible for new and more difficult applications.The larger manufacturing firms maintain substantial research staffs for this purpose. It has been emphasised that accurate measurements depend mainly on the operator’s ability to calibrate the mass spectrometer with known reference standards. In routine studies, calibration with reference standards is accepted as a normal feature of a planned programme of work; but in studies of a non-routine character, the need for calibration canOctober, 19541 AS AN ANALYTICAL INSTRUMENT 595 sometimes be overlooked, particularly in new applications. Some confusion may then arise, because the mass spectrometer always provides an answer without calibration. If calibration is not undertaken, it is up to the laboratory workers to decide what the answer means. In isotopic work, calibration has been achieved by the use of known synthetic mixtures of previously separated pure isotopes. By this means, Nierl has been able to determine for 10 common elements isotopic abundance ratios that are believed to be accurate; probably the work will be extended in due course to many other elements.In all analytical applications, pure elements or compounds or known synthetic mixtures of compounds and elements are used for calibration purposes. However, in spite of a need to undertake, sometimes at intervals of only a few weeks, the extra operations involved in calibration, the mass spectrometer’ compares favourably in respect of speed with all other existing methods in analytical work. For example, the analysis of a C, to C5 hydrocarbon gas stream, containing paraffins, olefins, hydrogen, oxygen, nitrogen, carbon monoxide and carbon dioxide, by means of Podbielniak low-temperature distillation methods in conjunction with chemical methods and infra-red spectroscopy (and perhaps Raman and ultra-violet spectroscopy) might take up to 12 hours, whereas with the mass spectrometer a complete analysis of the same mixture could be undertaken in less than 2 hours with modern com- puting aids.This gain in speed of analysis has been used to great advantage in many large oil refineries throughout the world. For refinery control, both gas and liquid samples are fully examined for an average of 16 components at an average rate of about 700 samples per month per machine. Apart, however, from purely routine work of this nature, machines are used as extensively for analyses of samples arising from process research studies.I t is interesting to note that with the experience gained in the operation of standard commercial machines, it has been found possible in at least one organisation to build up a library of calibration data in association with known settings of the instrument controls, and from this information to adjust the output characteristics of the machine to coincide with a previously determined set of calibration data. In this way full recalibration was found unnecessary for a period of 2 years. During this period only spot checks with a specific mixture were made. In the petroleum industry, the mass spectrometer has become, therefore, the first choice as a routine gas-analysis instrument, although it cannot supplant entirely the older equipment.In spite of the great speed attained in the use of the mass spectrometer, the results are not appreciably less accurate than those of the best of the older methods. For example, with certain exceptions, determinations of the concentration of a component in a complex mixture to 0-3 mole per cent. would be regarded as a good average performance for routine work. In the first class, analytical and isotope-ratio mass spectrometers are combined in one machine for a mass range of about 2 to 100 mass units. Analyses are limited to specimens that exert a minimum vapour pressure at room temperature of about 200 microns absolute pressure. Within these limitations in respect of mass number and vapour pressure, inorganic or organic compounds of almost any class or chemical nature can be analysed.As an isotope-ratio instrument, the ratio of the concentration of two isotopes in a gas, from 0 to 1.0, can be measured directly by the use of dual collectors for simultaneous collection of the two isotopic beams and a balancing electronic network. For isotopic ratios above 0.1 the ratio measure- ment can be made to six decimal places, with a useful limit in sensitivity of a few parts in 106. For example, successive determinations of carbon-13 at the level of natural abundance (ratio of 13C to 12C = 0.0110) on a routine basis would show reproducibility to well within h0.3 per cent. of the ratio. With well stabilised controls, high repetition accuracy is thEs a satisfactory feature of these machines.However, as has been emphasised, absolute accuracy depends on the operator’s ability to calibrate with known standards. Finally, it may be added that for a single sample introduction with a molecular flow leak, only about 150 litre-microns of gas are required. The normal gas consumption of a continuously pumped spectrometer is about lusecs (litre-microns per second), so that measurements should be conducted on a rigid time schedule to avoid errors arising from gradual fractionation of the sample with time. The second class of machine is used mainly for quantitative or qualitative analyses of mixtures of inorganic or organic gases and liquids, and again the scope of the work that can be undertaken is limited mainly by sample volatility.Components for analysis must be Modern commercial machines can be divided roughly into three classes.596 BARNARD THE MASS SPECTROMETER [vol. 79 such as to provide at room temperature a vapour-phase sample at about 50 microns absolute pressure. The maximum mass resolution available lies in the range of 250 to 400 mass units, depending on the variation of design from one manufacturer to another. On one American instrument,2 it is possible, with maximum resolution adjustment, to separate mass 44.004 of carbon dioxide from mass 44.076 of propane. The high sensitivity available is such that, in favourable mixtures, 0.001 mole per cent. could be detected with certainty, as for example, in the detection of argon in neon.. The third class of machine, which has been developed recently by modification of the second class, permits accurate scanning of the mass scale for molecular ranges of up to 1000 mass units.This high-mass instrument was designed primarily for studies of lubricating oils, waxes and greases containing up to 31 carbon atoms per molecule. The very low vapour pressures exerted at room temperature by these materials prohibit normal handling, so that provision must be made for heating the sample inlet system. The inlet system3 used is shown in Fig. 1; this is continuously heated at temperatures up to 400” C to get these solids into the vapour phase. The secret lies in the use of molten gallium, which boils at 2000” C, as a valve and as a seal over a sintered-glass disc. To pump t ,-Molten gallium Iron Molte A ~ ~ - S i n t e r e d - g l a s s disc Magnet -M :n piungerw gallium i To mass spectrometer _c Fig.1. Heated inlet system for sample introduction So far, little has been said about the internal working of the orthodox mass spectrometer. To a large extent this is mainly the province of the physicist and instrument designer; but the analytical chemist should have at least a general notion of the mechanism, if only so that he is able to tell the physicist in no uncertain terms what new modifications are required to meet special needs in further development. Operational training is normally given to an intending purchaser, and superficial experience for routine working is quickly gained. It is apparent from earlier comment that the sample for analysis is introduced in the gaseous or vapour phase into an evacuated enclosure.At the point of entry the gas or vapour enters an ionisation chamber in which a group of positive ions representative of the original material is produced. Normally the ions are produced by bombarding the gas with a stable beam of electrons of closely controlled energy. Some of these ions are with- drawn from the chamber and then accelerated as an ion beam of narrow divergence by a system of collimating electrodes. The ion beam is projected into a magnetic field in a direction at right angles to the magnetic lines of force. Th.e mass sorting process occurs in the magnetic field. The term “mass spectrometer” is really a misnomer, for, essentially, the instrument produces a momentum spectrum, and it is only by the application of various artifices and devices that the momentum spectrum becomes, to all intents and purposes, a mass spectrum.If the group of ions produced in the ionisation chamber varied in random fashion in both mass and energy, so that no constant relationship existed between the velocity of a particle and its mass, the arrangement would, without additional devices, be useless as a mass spectro- meter . The ion-source system in a conventional spectrometer is arranged, therefore, to produce a group of practically mono-energetic ions. Suppose now these ions acquire their velocity by falling through an electrostatic potential difference, V. We can write- .. .. * * (1) +Mv2=eV .. .. .. ..October, 19541 AS AN ANALYTICAL INSTRUMENT 597 where M represents the mass of a particle, v its velocity, and the charge on an ion is measured in terms of a single electronic charge, e.As Mv2 is constant by arrangement, it is apparent that a means of mass selection is presented immediately because a characteristic velocity is associated with each mass number. This is the basis of the radio-frequency mass spectro- meter, in which radio-frequency fields are used for velocity selection and hence mass selection. However, these instruments have not yet been applied industrially, and much development work is still needed. In a conventional machine the beam of mono-energetic particles is sent into a magnetic field, H , in a direction a t right angles to the magnetic lines of force. From elementary mechanics, it is seen that a particle experiences a centrifugal force Mv2/r, where Y is the radius of curvature of the particle.For equilibrium this must be balanced by the force, Hev, due to the field, or- Mv2/r = Hev, so that Y = Mv/eH . . .. It is seen that the radius of the trajectory of each particle is proportional to Mv, and the instrument yields a momentum spectrum. However, by the selected terms of reference Fig. 2. Schematic diagram of Dempster’s mass spectro- meter. The dimensions are those given in Dempster’s paper4 (Mv2 constant), r becomes proportional to M4, and the momentum spectrum is converted into a mass spectrum with a definite velocity associated with each mass. Equations (1) and (2) can be combined and ZI eliminated to give the mass spectrometer equation- .. ..* * (3) M/e = y2H2/2V . . .. .. The mass spectrometer differs from the mass spectrograph in two main respects. First, it is a constant deviation spectrometer in that it has fixed entrance and exit slits and all particles have essentially the same trajectories. Either I‘ or H or both are adjusted to bring the beams to the fixed exit slit. Secondly, the beam intensities are measured electrically. There are also other differences. From the three equations given, it is apparent that others can be derived to give expressions for dispersion and mass resolution; but it should not be forgotten that there are many sources of aberration and discrimination in the instrument, so that such expressions really only represent design ideals. For all applications the mass spectrometer should be used as a comparator and should be calibrated in some suitable way.The various masses describing paths that have different radii in accordance with equation (3) are shown in Fig. 2. It will be noticed that refocusing of the initially diverging beams occurs in the plane of the exit slit, S,. It is apparent that the modern mass spectrometer depends upon the existence of this refocusing property of magnetic fields. However, for a first-order direction-focusing machine, this refocusing is not perfect , and the focusing error or spherical aberration is found to be ra2, where u represents the semi-angle of divergence of the beam from the source. Fig. 3 shows the ion paths for a mass spectrometer having a 60” sector field. Studies of such trajectories provide working rules for positioning the entrance and exit slits with respect to the magnetic field so as to secure maximum mass resolution. In sector-field instruments, some compensation for fringing flux disturbances698 BARNARD : THE MASS SPECTROMETER [Vol.79 is necessary; usually the set-up is made in accordance with ideal field theory and simply withdrawn, as shown, from the field to compensate for the effect of fringing flux. The analytical chemist will want to know the relative advantages and disadvantages of the sector-field instrument and the 180" instrument, apart from obvious considerations relating to size and cost of magnet for a particular design. Theoretically, the angle of deviation in the magnetic field in these first-order focusing machines is immaterial from the point of view of resolution.In practice it is found that a certain minimum and by no means inconsiderable magnetic field is required in the region of the source, not only to align the electron beam for impact ionisation, but also to produce satisfactory collimation (small a) of the ion beam. It would be expected, therefore, that in the 180" instrument, in which the region of the source is almost fully immersed in the main magnetic field, greater mass discrimination would occur, but at the same time there would be better collimation of the ion beam with a gain in resolution; this appears to be so. I y' Fig. 3. Magnetic mass selection and focusing of ion beam It may be helpful at this stage to consider the various component parts to be assembled and correctly inter-related as a complete mass spectrometer (see Fig. 4).It is seen that the complete machine houses a considerable amount of auxiliary apparatus : gas-inlet plant, high-vacuum pumping equipment, stabilised supplies for electron and ion-beam acceleration and for magnet current control, and a sensitive d.c. amplifier, together with autorratic recording equipment. A large analytical machine of the 90" sector type, made by Metropolitan- Vickers, is shown in Fig. 5. Reference was made earlier to the introduction of the sample into an evacuated enclosure. It is apparent that high-vacuum conditions must be established to avoid scattering of the ion beam by many collisions between gas molecules and ions. The working pressure sought in the main vacuum in the analyser is 10-6mm of mercury or less.In the region of the ion source, with its relatively small dimensions', a total gas pressure of up to 10-4mm of mercury on introduction of the sample can be permitted. Many requirements have to be met in the introduction of a complex gas mixture consisting of many components. Perhaps the most important of these are the ensuring of {a) a known relationship between the partial pressures of the components in the sample and the corresponding partial pressures of the components in the ionisation chamber, (b) a constant composition of the mixture in the gas reservoir during the analysis and (c) no interference between components in transit from sample reservoir to ionisation chamber. The last requirement is essential to ensure that a mixture ion beam, consisting of contributions of ionised mass fragments all having the same mass number but arising from different components, should be made up of linear super- positions of individual ion-current intensities, i e ., at any mass number the machine should be linearly additive.October, 19541 AS AN AXALYTICAL INSTRUMENT 599 As an ion beam of a particular mass number is swept across the exit slit of the spectro- meter, the current to the collector behind the exit slit will first rise, then attain a maximum, and finally fall as the beam leaves the slit. The shape of the curve depends on the relationship between exit-slit width and ion-beam width. A mercury isotope spectrum is shown in Fig. 6. K \ stabilised high voltage unit and ior accelerating voltage control N i o n r source regior source mag Fig.4. Schematic relationship of component parts of mass spectrometer assembly It can be seen that if there are as many peaks as this to deal with, it can be very laborious to determine them all by individual measurements. A considerable strain is put also on the stability of all the controls. The next stage is clearly to have an automatic recorder with automatic range-changing to deal with peaks covering a wide range of heights. This is shown in Fig. 7.5 With a complex sample, many measurements of peak height may have to be made, and the most recent development is a machine that can make a direct conversion of analogue information-in the form of varying voltages-to digital information, i.e., t o numerical form.The recorder then becomes unnecessary. An analytical mass spectrometer made by the Consolidated Engineering Corporation, U.S.A., with analogue-to-digital converter. is shown in Fig. 8.600 BARNARD THE MASS SPECTROMETER [Vol. 79 When a complex molecule is bombarded by lelectrons of energies of from 50 to 100 eV, the molecule is found to disintegrate into all conceivable fragments, with simultaneous ionisation of the fragments. No theoretical predictions can be made as to the abundances of the fragments. To some extent the cracking pattern depends on the operating conditions in the instrument, owing to mass discrimination. However, for practical analytical work with a chosen set of experimental conditions, it is most important to establish the dis- integration scheme for each molecular species and to ensure that the intensities of the 202 2270 2280 2290 2300 2310 :2320 2330 2340 2350 2369 lo n-acce le rat i ng potential, vo Its Fig.6. Mercury isotope spectrum with narrow slits. Pressure 1.4 x mm of mercury ion fragments relative to the intensity of the parent molecule ion, or to the intensity of the most abundant ion fragment, or to the total ion current summed for all mass numbers, are constants that are characteristic of the species. Characteristic fragmentation for each molecular species permits identification of different molecular species having almost identical mass numbers. This is shown in Table I, in which are given the cracking patterns for nitrogen, carbon monoxide and ethylene. TABLE K CRACKIKG PATTERNS OF NITROGEN, CARBON MONOXIDE AU i D ETHYLENE M/e 1 2 12 13 13.5 14 15 16 24 26 26 27 28 29 30 Nitrogen - 100.0 0.742 0.005i Carbon monoxide - 4.7 1 0.05 0.75d 1.67 - - - - 100.0 1-16i 0.222 Ethylene 4.11 0.51 2.14 3.52 0-42d 6-31 0.56 3.7 1 - 11-7 62.3 64-8 100.0 2.22i 0-04i d = doubly-charged ion.i = isotope peak. Generally, the way in which a complex molecule disintegrates under electron bombard- ment is that expected from general chemical experience, although some particular aspectsFig. 6 . Metropolitan-Vickers general purpose mass spectrometer, M.S.2. Sample inlet system on right; control panel in centre; recorder on leftFig. 8. General purpose analytical mass spectrometer with automatic analogue-to-digital converter, A peak selector is also incorporated for made by the Consolidated Engineering Corporation, U.S8.A.pre-programming a run. Measurement of individual peaks on a record is unnecessaryOctober, 19541 AS AN ANALYTICAL INSTRUMENT 601 of bond rupture are not clearly understood. Consider, for example, the structural formulae of- H H H H H H H 45 I I I I I I I I H-C-C-C-C-H H H H H 12-butane and I I I, I H-C-C--C-H i4 I H H-C-H H I isobu tane It would be expected that these isomers would disintegrate in different ways. In n-butane, by the rupture of a single carbon-to-carbon bond, the molecule can split into two equal parts, forming two C,H, fragments of mass 29. This cannot happen in isobutane. Hence, in the cracking patterns for n-butane and isobutane (see Table II), the mass 29 should be more abundant for n-butane than for isobutane. However, some isomers have similar cracking patterns and present a large number of ill-conditioned equations for computation.This is particularly true for the individual Molecules of CO, 7 '2c'60'80 Fig. 7. Automatic record of analysis of carbon dioxide with mass spectrometer, M.S.2 butenes, which generally cannot be separated accurately. This is not a serious limitation, because often knowledge of the total butene content only is wanted. Table I1 has been included to give some indication of the computations involved. In the calibration cracking patterns for these five compounds, the relative peak intensities are referred t o the most abundant ion fragment as 100. At the bottom of the table, sensitivity factors, s, are given.The sensitivity of the reference peak is expressed as the height of the peak in so many divisions for unit pressure of calibration gas in the inlet sample bottle. The requirements for satisfactory analysis are (a) constant cracking patterns and s values over a reasonably long period and (b) linear superposition of the spectral contribution of each component in the composite spectrum for any mixture of compounds. Suppose now these five compounds are proportioned in some way in a mixture and it is required to find how much of each is present. Let the partial pressures of each be PI, pa, p3, 9, and p5, respectively, and sl, s2, s3, s4 and s5 be the corresponding sensitivity602 BARNARD : THE MASS SPECTROMETER [Vol. 79 coefficients. There is one such equation for each mass number.For mass 15, for example, the equation is- The five unknowns, jbl to p,, are to be found. It will be observed that increasing numbers of gaps occur in the rows as the mass number increases. It is possible to start at the high-mass end and eliminate the high-mass com- ponents first, by choosing for the solution the set of five equations for M / e = 58, 57, 44, Then a set of simultaneous equations can be written down directly. 5.30 s1 el + 6.41 s2 p , + 6.19 s3 p3 + 4.57 s4P4 + 85.6 S , 9, = 167.3 This is a particularly simple example. M/e 15 16 26 27 28 29 30 31 38 39 40 41 42 43 44 45 50 51 52 53 54 55 56 57 58 59 s TABLE I1 CALIBRATION DATA AND MIXTURE PEAK-HEIGHTS FOR DEPROPANISER OVERHEAD SAMPLE n-Butane 5.30 0.12 6.17 37.1 32.6 44.2 0.98 - 1.89 1-63 12.5 27.8 12-2 100.0 3.33 0.05 1-29 1-05 0-26 0-74 0.19 0.93 0.72 2.42 0.54 0.396 12-3 isoBu tane 6.41 0.18 2.36 2.62 6-16 0.13 27.8 - 2.77 2.37 16.5 38-1 33.5 100.0 3.33 0-03 0.89 0.74 0.15 0-50 0.07 0.42 0.34 3.00 2.73 0.1 1 0.464 Propane 6.19 0.15 8.59 39.4 59-1 100.0 2:20 - 5-29 2.52 5.82 17-0 12.7 22.8 29.0 0.88 - - - - - - - - - - 0.335 Mixture, peak heights 167.3 134-2 14G.1 365.6 757.8 613.1 130.1 2.4 29.0 98.0 14.7 94-1 49.9 207-4 145.4 4.6 0-6 0-5 0.1 0.6 0.1 0.7 0-5 2.5 5-0 0.2 30 and 16. There are some problems in procedure here.For example, is the chosen set the best? Which set of five equations should be chosen? Or would it be better to use all the available data and obtain a solution by a weighted least-squares method? There is a slightly different answer each time, How should the best solution be defined? These are some of the problems studied in recent publications.6,7 The computations can become quite complicated in some circumstances, and modern computational aids are necessary to reduce the labour involved.If a typical mixture is presented many times for analysis, i.e., if the same components recur in various concentrations and if the calibration data can be taken as constant, then it is not necessary to undertake the full computation each time. The procedure is to set up an inverse matrix, valid for the duration of the work, and to substitute each time the new mixture peak-heights. APPLICATIONS OF THE MASS SPECTROMETER The mass spectrometer has been applied suecessfully to a wide range of work, including the determination not only of hydrocarbons, but also of oxygenated compounds, for example, alcohols, ethers and esters, of halogenated compounds-alkyl chlorides and iodides-and of many types of sulphur compounds.It has been suggested also that gas analysers of this type could be used as primary elements in closed feed-back loops to give, with servo-operated equipment, complete automatic control of a process. No gas analyser has been used as yetOctober, 19541 AS AN ANALYTICAL INSTRUMENT 603 for automatic process control. But as a first step towards this goal, mass spectrometers have been used for continuous monitoring of a process stream to provide information whereby an operator can detect off-specification conditions and then take manual corrective action to restore the plant to the specified state.s The mass spectrometer has no rival for this purpose.In the chemical industry, for example, mass spectrometers have been used for furnace control in the production of acetylene by the Wulff process, the first step in producing the newest synthetic fibres, Dacron (Terylene) and D ~ n e l . ~ A number of recent publications6y10911 give much wider surveys of routine work than is possible here. It is proposed, therefore, to deal only with a few recent applications with special research instruments to indicate the trend of future development. ANALYSIS OF STAINLESS STEELS An application of the mass spectrometer in a different sphere is its use for the analysis of stainless steels.12 The method has not yet been applied on a routine basis for the analysis of alloys, such as steels, but it is already apparent that the mass spectra of metals and their alloys are much less difficult to interpret than, say, the emission spectra of the metals. There is, therefore, considerable industrial interest in this development for routine work.The best method of producing the required positive ions from a solid material of this nature is by use of the high-frequency vacuum-spark source developed by Dempster in 1934, The material to be analysed forms one electrode of a high-voltage spark gap and is usually in the form of a rod of 0.025 inch diameter. When the necessary potential is applied across the gap to cause electrical break-down, some of the electrode material is evapqrated into the space between the electrodes, where it is ionised by the electrons in the discharge.The discharge is maintained between the sample rod and a tantalum disc by a pulsed radio- frequency supply, obtained either from a Tesla coil or pulsed electronic oscillator. This type of source has the advantage that it has no blind spots; that is, all elements present are ionised with roughly the same efficiency. There are several disadvantages with a source of this type, and these for a long time restricted its use to special researches with mass spectrographs. Quite recently, however, it has been demonstrated that, by suitable design, the source can be used for the analysis of solids by mass-spectrometric methods. Analyses of six samples of stainless steel by mass-spectrometric methods and by existing established analytical methods revealed excellent agreement. This is shown in Table 111.TABLE I11 ANALYSES OF SAMPLES OF STAINLESS STEEL Chromium r- Mass spectro- Sample Chemical, meter,* X3534 23.7 23.4 0.4 X3380 18-2 19-3 F 0.3 X3275 13.4 13.7 2 0.2 x3532 9.1 9.1 +_ 0.2 x3533 5.5 5.5 2 0.1 % % X3522 2-95 2.97 & 0-05 Nickel & Mass spectro- Chemical, meter,* % % 13.6 13.7 k 0.5 8-6 8.7 & 0-2 0.37 0.25 & 0.01 0-57 0.56 2 0.02 6.8 7.1 0.1 25.7 25-6 & 0.3 Chemical, 61.6 69.8 83-0 85.5 84.3 70-0 % Iron - Mass spectro- meter,* 61.8 k 0.4 68.6 0.3 82-8 k 0.3 86.5 f 0.3 84-0 f 0-1 70.1 0.1 % Other elements by chemical methods, 1.1 3.4 3-2 4-8 3.4 1.3 % * Mean value and average deviation of 4 or 5 determinations.The main disadvantages to be overcome in the use of the vacuum-spark source in mass spectrometry are that (a) ions are produced with a very large energy spread, so that a con- ventional single-focusing mass spectrometer cannot be used, (b) the intensities of the ion beams fluctuate so wildly that electrical measurement by normal methods becomes impossible and (c) electrical interference between the spark source and sensitive electrometer amplifiers necessitates elaborate screening precautions. The large energy spread makes the source suitable for use only with the rather more complex double-focusing mass spectrometer of, say, the Mattauch or Dempster type, in which the beam dispersion in the magnetic field caused by velocity spread is compensated exactly by an equal and opposite dispersion in a preceding electrostatic field.Double focusing The instrument used is shown in Fig. 9.604 BARNARD : THE MASS SPECTROMETER [Vol. 79 in respect of velocity and direction is thus achieved. However, the source is still erratic and subject to wild-fluctuations; and this calls for some ingenuity in method. Here slit S , is used not only as a slit but also in part as a monitoring collector. A constant fraction of the total ion current emitted by the source is collected on the edges of S , and fed into a d.c. amplifier. In the usual way an ion beam of any particular mass number is focused on to S, and collected. The current collected by the plate behind S, is fed into a second amplifier, and the outputs of the two amplifiers are compared on an automatic recorder, Hence, although the ion beam from the spark source is fluctuating, the outputs of the two amplifiers fluctuate in unison, so that the ratio of the two signals is independent of source fluctuations.When the magnetic field is scanned, the ratio of the selected beam current to the total beam Fig. 9. Dempster-type double- focusing mass spectrometer Collector ' s5 60 56 52 48 Mass number Fig. 10. Mass spectrum of stain- less steel sample No. X3380, 7/24/50 current varies according to the intensity of the selected mass peak, and the recorder chart plots a mass spectrum similar to that obtained with a conventional instrument (Fig. 10). Here again the mass spectrometer is used as comparator and is first calibrated with known synthetic samples to obviate errors due to mass discrimination in the instrument.To determine the chemical composition, corrections have to be made for the known isotopic distributions, and the analysis shows the relative distribution of the components in the original material. In the results shown in Table I11 everything was referred to iron, which was used, in effect, as an internal standard. DETERMINATION OF TRACE: ELEMENTS IN SOLIDS In orthodox mass spectrometry with the blest mass analysers now in use, the limit of sensitivity is about But there are to-day many analytical problems in which much higher sensitivities are required. For example , consider the germanium rectifier and transistor, in which more than 1 part per million of arsenic as impurity is sufficient to reduce the maximum back-voltage for negligible current flow (or turnover voltage) from about 100 volts for the purest specimens to less than a few volts.Therefore much more sensitive methods of analysis are required. The isotope dilution method has been known for a number of years, but has been used only comparatively recently for impurity work:.13 The method was developed originally by Rittenbergl, for the quantitative analysis of complex mixtures (particularly amino-acids) encountered in protein studies, and it depends on the fact that the usual chemical procedures do not result in significant fractionation of isotopes. It is not proposed to discuss its use in biological work, but to indicate how it has been applied, with essentially similar principles, to trace element determinations in solids.For this work it has become a practical analytical method with sensitivities for many elementsOctober, 19541 AS AN ANALYTICAL INSTRUMENT 605 far surpassing anything previously known. This may, perhaps, be illustrated best by con- sidering a hypothetical example. Suppose it is required to determine the amount of an element, E,, present as a trace impurity in a solid, D. Suppose also that element E , has two stable isotopes, I , and I,, with a normal abundance ratio of Ro = IJI,. Now the atom per cent. concentration of isotope Il in element E , can be defined as- Number of I , atoms x 100 Number of I2 atoms + Number of I , atoms’ or 100 Atom per cent. of I - - ’- R,+ 1’ Suppose now a small quantity of the pure element E is available enriched in respect of isotope I , to give an isotopic abundance ratio of R = 12/Il.Then the atom per cent. excess concentration of I , over the normal is given by- 100 100 C =F Atom per cent. excess of I - - R T 1 - R ? l ’ The experimental procedure is as follows. To a measured weight, Mg, of solid, D, add a small quantity, Wg, of the pure element E enriched in respect of isotope I1. These are now brought into solution, so that the tracer element E mixes with any normal element Eo present in solid D to form an inseparable mixture so far as the usual chemical procedures are con- cerned. From the mixture isolate a small quantity of pure element E’ and measure the isotopic abundance ratio, 12/I,. Clearly, if R’ is not equal to R, the normal element E , is present in solid D, and dilution of the tracer element E has occurred. The new atom per cent.excess concentration of isotope I , over the normal is given by- 100 Suppose that this ratio is R’. 100 -- C’ = - R’+ 1 R,+ 1’ Let W’ g be the weight of the normal element Eo present in weight Mg of solid D, and let A , be the atomic weight of the normal element E, and A the atomic weight of the enriched element E. The relation between C and C‘ is given by- W A [,” z ] c x - = C ’ - + - , C - C ‘ A W’- - - x-OXW C’ .A or or on substitution for C and C’. In practice, except when the tracer element E is very appreciably enriched in respect of isotope I,, A,/A may be taken as unity to a first approxima- tion. For the normal element Eo, let the atom per cent.of I , be 20-0, and for the tracer element E , let it be 40.0. Then Ro = 4-0000 and R’ = 1.5000. As a difference of 0.0001 in relative abundance ratio can be detected with certainty, it is apparent that an amount W’ as low as 0.0001W can be detected. If initially M = 1 g and W = 1 pg, then an amount W’ of 10-lo g can be detected. In fact, it has been found that many of the elements can be detected with as little as 10-12g of material present, while some elements, for example, potassium, can be detected easily at a concentra- tion of about lO-14g. These high sensitivities have been attained so far only with special research instruments ; but it should be remembered that the method was developed originally for use with an ordinary gas machine. Trace element determinations in a lower sensitivity range, say 0.01 to 10p.p.m., by use of this method would appear to be possible with con- ventional machines.The advantages of this method are that (a) the final result does not depend on the complete quantitative recovery of the element; for, in the example considered, if a certain percentage of Il had been lost in the processing, the same percentage of I , would also have been lost, (b) the method is absolute and (c) the method is extremely sensitive to trace A hypothetical example can now be considered.606 BARNARD : THE MASS SPECTROMETER [Vol. 79 impurities. The first is that only one element can be dealt with at a time. Secondly, the separated isotopic tracers must have isotopic compositions considerably different from normal, and some restrictions exist on their use by intending purchasers.However, sometimes high sensitivity can be attained with quite a small degree of enrichment, as indicated by the above example. If, therefore, an analytical laboratory is solely interested in a particular impurity determination, it may well be possible for such a laboratory with its own facilities to build a small plant15 to provide the tracer material. A third disadvantage is that some elements do not have a second stable isotope that can be used as the diluent. However, sixty-seven elements can be determined in this way.. If, for example, 1 milligram of ordinary dust became mixed with the sample at any time during the chemical processing, considerable errors might arise. OTHER IMPURITY PROBLEMS There are, of course, some disadvan1:ages.A fourth disadvantage results from problems of contamination. The detrimental effect of 1 p.p.m. of arsenic as impurity in germanium has already been noted. Unfortunately, arsenic has no second stable isotope, so that the isotope dilution method cannot be used. It may be noted that the chemist is also interested in trace quantities of boron in germanium, and for its determination the isotope dilution method can be applied. The hope of using the mass spectrometer for determinations of traces of arsenic has not, however, been abandoned, and attention h.as been directed to physical enrichment processes,l6 whereby the impurities can be drawn out of the sample and concentrated in very small zones, so that the mass spectrometer has a more concentrated sample for analysis.This has been carried out in a gradient furnace. The raw material is melted in a carbon crucible of high purity and is then permitted to cool gradually while the crucible is lowered slowly out of the furnace. Impurities that lower the melting point of germanium concentrate in the top zones of the ingot, which are the last to solidify; the main body is purified, and impurities that raise the melting point of germanium are concentrated in the bottom. It has been found that the concentration of impurities in the top may be increased a hundred-fold in a single operation, and it is believed that routine analyses of arsenic in germanium should soon be possible to within 1 part in lo8. Another impurity problem of quite a different character arises in water pollution abatement work, particularly in the determination of minute quantities of contaminants- mainly hydrocarbons-in drinking water.Recently a method for applying the mass spectro- meter t o this work was described.17 Briefly it consists in taking a 1.5-litre sample of water and stripping the volatile compounds from the boiling water by passing purified hydrogen gas through it. Hydrocarbons boiling below 400” F are carried over and condensed in a liquid-nitrogen trap. The method is now being used to determine gaseous and liquid hydrocarbons, boiling up to at least 400” F, in the range 0.01 to 100 p.p.m. in water. The sensitivity depends, of course, on the volume of sample-water taken. The condensate is then analysed in the.mass spectrometer. CONCLUSION It is apparent that, in respect of speed, accuracy and versatility, the mass spectrometer has rapidly become a powerful analytical instrument. Much development work continues and new applications are being reported frequently. The main deterrent in acquiring such an instrument is often one of cost in comparison with the older physical instruments for chemical analysis. For a particular analytical 1 aboratory, the main basis of assessment must be on the saving of man-hours without loss of accuracy and the greater facilities afforded for a wider range of work. This paper is published with the permission of the Director of the National Physical Laboratory. REFERENCES 1. 2. 3. 4. 5. Nier, A. O., “Mass Spectroscopy in Physics Research,” N.B.S. Circular 522, US. Government Consolidated Engineering Corp., U.S.A. Analytical mass spectrometer, model No. 21-103. O’Neal, M. J., “Mass Spectroscopy in Physics Research,” N.B.S. Circular 522, US. Government Dempster, A. J., Phys. Rev., 1918, 11, 316. Metropolitan-Vickers Electrical Co. Ltd., Mass spectrometer, M.S.2. Printing Office, Washington, 1853, p. 131. Printing Office, Washington, 1953, p. 217.October , 19541 AS AN ANALYTICAL INSTRUMENT 607 6. 7. 3. 9. 10. 1 1 . IS. 13. 14. 15. 16. 17. Barnard, G. P., “Modern Mass Spectrometry,” Institute of Physics, London, 1953. Barnard, G. P., and Fox, L., Conference on Applied Mass Spectrometry, The Institute of Petroleum, Lanneau, K. P., Conference on Applied Mass Spectrometry, The Institute of Petroleum, October, C.E.C. Recordings, 7, KO. 4, December, 1953. Conference on Applied Mass Spectrometry, The Institute of Petroleum, 1953, t o be published, 1951. Conference on Mass Spectrometry, The Institute of Petroleum, 1950, published 1952. Gorman, J . G.. Jones, E. J., and Hipple, J. A., Anal. Chem., 1951, 23, 438. Inghram, M. G., J . Phys. Chew., 1953, 57, 809. Rittenberg, D., J . A p p l . Phys., 1942, 13, 561. Urey, H. C., Ibid., 1941, 12, 270. Honig, R. E., Anal. Chew., 1953, 25, 1530. Melpolder, F. W., Warfield, C. W., and Headington, C. E., Ibid., 1953, 25, 1453. Paper No. 18, October, 1953. 1953. PHYSICS DIVISION NATIONAL PHYSICAL LABORATORY TEDDINGTON, MIDDLESEX Apvil 28th, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900594

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

October , 19541 607 An Automatic Coulometric Titrimeter BY N. BETT, W. NOCK AND G. MORRIS (Presented at the meeting of the Physical Methods Group on Tuesday, February 9th, 1954) Coulometric titrimetry in which the titrant was generated externally was examined by manual methods, and the experimental error was shown to be about 0.1 per cent. for titrations of 25 ml of 0.01 N acid or alkali. For the titration of 0.1 N solutions, a modified form of electrolysis cell and a specially stabilised current supply were devised. As coulometric analysis depends on the flow of an electric current, it is particularly suitable for automatic operation. An automatic titrimeter was constructed and by its use 25 ml of 0.1 N acid or alkali were titrated in about 43 minutes; the standard deviation was less than 0.2 per cent.When an instrument capable of recording a quantity of electricity by integration is available, there is no need to stabilise the electrolysis current, and a different speed from that of the titration can be used in the approach to the end-point. An automatic titrimeter incorporating such an integrator is described in detail. Titrations involving the generation of iodine are described, and the results were satisfactory. When the automatic titrimeter was used in a routine analytical laboratory, the results were found t o be reproducible to within kO.2 per cent. TK recent years an analytical technique has been developed in which titrations are performed with an electrolytically generated titran t. This procedure is known as coulometric analysis and has been used as a manual method, particularly when a few micro-equivalents of material had to be determined.This paper records an attempt to apply coulometric analysis in routine factory analytical work. Coulometric analysis has certain advantages over conventional volumetric analysis, especially if it is required that the titration result be recorded. In addition there is no burette to be filled and adjusted, and standard solutions are unnecessary. To compete with volumetric methods the electrolytic generation must be sufficiently rapid to titrate 25 ml of decinonnal acid or alkali in 5 minutes or less. One of the earliest uses of coulometric methods was made by Szebelledy and Somogyi,l who standardised volumetric solutions by means of a silver coulometer. More recently Furman, Cooke and Reilley2 have shown how to produce ceric ions quantitatively, and micro-determinations have been made of ~anadium,~ thalli~rn,~ iron,5 ar~enic,~ manganese,6 silver7 and many other substances.710911 ,12 slO,ll ~ 1 2 ferrous ions,6s13 hydrogen ions14 and hydroxyl i 0 n s ~ 4 , ~ ~ have been produced in small quantities and in measured amounts by an electric current. The development of sensitive end-point proceduresl6 has made coulometric titration a valuable asset in micro analysis. Internal generation was used exclusively in the above processes, but DeFord, In addition to ceric ions, iodine,s608 BETT, KNOCK AND MORRIS: AN AUTOMATIC [Vol. 79 Pitts and Johns17 have demonstrated the production of acid, alkali and iodine in quantitative amounts in an external cell and have shown that automatic titrationsls can be performed on this basis.This paper is solely concerned with external generation. PRELIMINARY EXPERIMENTS As the quantities of titrant to be handled were much greater than usual, preliminary experiments were made to ascertain whether the necessary reproducibility would be main- tained. A cell similar to that used by DeFord, Pitts and Johns17 was fed with a 1 per cent. solution of sodium sulphate in distilled water, the pH being adjusted to 4.0 (that of the chosen end-point); the solution was conveyed to the cell by a short length of rubber tubing provided with spring and screw clips for flow comtrol. The sample beaker was supported by a Baird and Tatlock magnetic stirrer in such ;a.way that it received the discharge from one of the delivery arms of the cell, the contents of the other arm being drained to the sink. It was essential that the delivery tip of the cell should not come into contact with the sample in the beaker, as this would affect the pH-electrode indication. The progress of the titration was indicated by measurement of the pH of the sample with a glass electrode. The electrolysis current was taken from a current-stabilised source and was measured by the potential drop across a standard resistor. The time was measured with a stopwatch. MANUAL TITRATIONS AT A CONCENTRATION OF 0.01 N- It was found that, when the efficiency of the cell was 99-6 per cent., 25 ml of 0.01 N sulphuric acid could be titrated to the end-point at pH 4.0 in 174 seconds by a current of 139 mA.The efficiency was remarkably high. In all the work reported here the instrument was standardised against chemical standards so that the absolute value of its efficiency is not jlmportant; its reproducibility is essential for accurate titration. The standard deviation of 12 determinations was 0-08 per cent. DEVELOPMENT OF A CELL FOR TITRATIONS WITH 0.1 N SOLUTIONS- In order to titrate decinonnal samples in a reasonable time, it was necessary to increase the generating capacity of the cell. A current sta.biliser that could supply current over the range from 200 mA to 1 amp. was constructed. Several designs of cell were tried in a attempt to increase the conductivity and improve the arrangements for disposal of the gas evolved during electrolysis. These cells were no better than the cell of DeFord, Pitts and Johns.With this cell, the electrolysis current was increased until the efficiency of the cell began to drop; the highest useful current was found to be about 300 mA. A cell to carry 1 amp. was then made by joining three of these cells together. When the cells were operated with a current of 300 mA each, the concentration of the electrolyte had to be increased to reduce the cell resistance and so reduce heat generation. The concentration of the electrolyte was finally increased to 1OOg of sodium sulphate per litre of water, The heat generated during a titration was the greatest obstacle in cell design, as it determined the minimum rate of electrolyte flow permissible if a serious rise in tempera- ture was to be avoided.It was also important that the space around the platinum electrodes should be kept small to prevent large bubbles of gas accumulating and to reduce the amount of “dead liquid” to be washed out at the end of a titration. It was found that the conductivity of the cell was almost independent of the rate of flow of electrolyte and therefore, while this could not be reduced much below 5 ml per minute at each delivery arm, there was no great advantage in increasing it. AUTOMATIC TITRATION- Some of these use calibrated motor-driven syringes, while others use conventional burettes with electrically operated taps. Titrations with the former take an unacceptably long time, while of the great variety of taps proposed, not all are acceptable.In addition, the use of a conventional burette requires it to be set up initially and the remaining volume read, an operation that requires some skill and cannot, at present, be recorded at a distance. Many workers have described automatic devices for volumetric titrations.October, 19541 COULOMETRIC TITRIMETER 609 From the point of view of automatic titration, coulometric methods have the advantage that the titre is recorded as a quantity of electricity; this is easier to control and record than a volume of solution. A first model automatic titrimeter was built in which the current from a stabilised source was passed through the triple cell until the pH, measured by a glass electrode and calomel cell, reached a value preset so that the final pH of the titration when the cell was flushed with electrolyte after the current was cut off was the required end-point.The indicator of the pH meter was fitted with a modified form of a Fielden Tektor so that the electrolysis current was started manually and was stopped by the Tektor relay. The time of current flow was recorded by a stopwatch electrically operated by the electrolysis current. In this form the titrimeter was used in a routine analytical laboratory for some time, and it was exhibited at the International Congress on Analytical Chemistry at Oxford in September, 1952. When a current of 923 mA was passed, 25 ml of 0.1 N sulphuric acid, which had been standardised with bromophenol blue as indicator, were titrated automatically in 275 seconds.Fig. 1. Section of cell in its final form The standard deviation of the results from a typical set of 11 determinations was 0.13 per cent. This was not quite so small as for the manual experimnts, but was fully satisfactory for routine laboratory tests. The efficiency of the cell was only 95.6 per cent., and the rise in temperature in the titration sample averaged 14" C. The eZectroZysis cell-Considerable trouble was experienced owing to clogging of the electrolysis cell by small particles of glass-wool and so on. The temperature rise in this cell was excessive. A cell was constructed in which the glass-wool used in previous models was replaced by two sintered-glass discs of porosity 1. The platinum electrodes were flat spirals, 8 inch in diameter, that lay on the outside of the partitions.The cell in its final form, which was made to our specification by Messrs. Mullard Ltd., is shown in Fig. 1. It was large enough to pass a current of 1 amp. without a serious rise in temperature, the heat dissipation being about 60 watts. The essential features of the design were the small end- volumes, i.e., hold-up volumes on the outside of the sintered plates, and the elevation of the exit tubes from the electrode chambers. The latter feature caused the gas bubbles to move away rapidly into the narrow delivery arms where the speed of the flow prevented large gas bubbles from forming. It also allowed the electrodes to be completely surrounded by electrolyte at all times. The integrating motor-A device that can accurately integrate the generating current in a coulometric titration will greatly simplify the design of a titrimeter.A short time after the first titrimeter was constructed, an integrating motor made by Electro Methods Ltd. became available. This apparatus consists essentially of a d.c. motor in which friction and heat loss have been reduced to a minimum. The rotation of the armature shaft is geared to drive a light610 BETT, KNOCK AND MORRIS: AN AUTOMATIC [Vol. 79 mechanical counter. The integrating property of the motor depends on the linear relationship between the speed of rotation of the shaft and the applied voltage. If this voltage is derived from the potential drop across a standard resistor through which the electrolysis current is passing, the counter reading will be proportional to the quantity of electricity that has passed through the resistor and hence through the cell.The resistor, being only about 1 ohm, provides adequate damping for the motor and ensures good response to rapid changes of input voltage. -- -- -- Fig. 3. Power supply An integrating motor for 1-5-volt operation was obtained and its performance was investigated. The accuracy of integration proved to be h 0 . l per cent. over the range 0.25 to 1-5 volts. When tested for a continuous run of 400 hours, the motor behaved satis- fact orily . Two major improvements in design could now be made: (i) the electrolysis current no longer required to be stabilised; and (ii) the electrolysis current could be decreased near the end of a titration to allow a slower approach to the end-point.THE AUTOMATIC COULOMETRIC TITRIMETER A commercial pH meter was used, as it was considered to be an added advantage to have the pH meter as a separate unit. Fig. 2 is a photograph of the complete appar,atus, which is seen to consist of four parts: the power supply, control unit, pH meter and titration stand. The power sz@$dy-The power supply, the circuit diagram of which is shown in Fig. 3, consists of two independent units, each with its own mains switch. The smaller unit provides hgh-tension current at 250 volts d.c., two supplies of 300 volts a.c. in antiphase, 230 volts a.c. from the mains and 6.3 volts a.c. for the valve heaters. The larger unit supplies the titration The automatic titrimeter will now be described in detail.Fig.2. The complete automatic coulometric titrimeterOctober, 19541 COULOMETRIC TITRIMETER 61 1 current at two levels, each level being capable of independent variation by means of the toroidal rheostats, R1 and R,. The circuit is designed to deliver a maximum electrolysis current of 1.5 amp. at 100 volts. 2 I I The control unit (Fig. 4)-This instrument is designed to monitor the input voltage from the pH meter (derived from the socket marked RECORDER) and to close the relays A and B in turn at two pre-determined voltage levels. This is achieved by comparing the input voltage with the voltages on the sliders of two helical potentiometers, R, and R,, making use of a Carpenter relay operating at 50 cycles per second. Hence from each side contact612 BETT, KNOCK AND MORRIS: AN AUTOMATIC [Vol.79 of the Carpenter relay a square wave is obtained, the amplitude and phase of which is deter- mined by the input voltage and the setting of the appropriate helical potentiometer. The two waveforms are fed into two separate but identical phase-sensitive amplifiers that operate their respective relays when the input waveform c:hanges phase. To ensure that the response of the thyratrons, V, and V,, during a phase change in the input signal is sharp, it is necessary to provide from 4 to 6 volts negative bias (obtained across R16). The circuit will detect changes in the input voltage of about 1.5 mV; this represents a change in pH value of 0-05 in the input to the pH meter. Relay A selects the low value of titration current when energised, while relay B in closing shuts off the titration current altogether.These operations are indicated by three dial lamps, green, yellow and red. A third relay, G, serves to prolong the stirring and flow of Solenoid- winding Fig. 5. Solenoid valve Fig. 6. Automatic burette tap electrolyte for about 8 seconds after the titration current has ceased, for reasons previously explained. This relay also breaks the titration current line so that, once stopped, the titration can only be started again by pressing the push-button marked START. The switch, S,, reverses the polarity of the electrolysis cell so that samples of alkali can be titrated as well as acid; S, also reverses the phase of the Carpenter relay so that the switching sequence of the electrolysis current is reversed.The front panel of the unit carries three lamps. When the instrument is ready for use or is running at full current, the green lamp is lit. When the instrument is operating at reduced current, the yellow lamp shines, while the red light indicates that titration is complete. The pH meter-This is of the direct-reading type made by Electronic Instruments Ltd., having a scale with pH values from 0 to 14. A constant current output of 20 pA per pH unit is available at the socket marked RECORDER; this is an adequate signal with which to operate the control unit. Provision is made for temperature compensaticm because of the large rise in temperature observed in a sample during titration. The ordinary type of temperature compensation is unsuitable because it does not compensate for zero errors but merely provides scale correc- tion.A special compensating unit was obtained from the makers, and it could provide temperature compensation at a fixed pH. The titration stand-A double-posted retort stand is used to support the various com- ponents, the uppermost of which is a sintered-glass filter of porosity 2, introduced to protect the sinters of the cell from clogging. The magnetic flow valve is a glass type of simple The delay network, R,, and C,,, provides the necessary time lag.G1.3 construction, as shown in Fig. 3. 'The coil consists of 5000 turns of 36 S.W.G. enarnellecl copper wire and requires a potential of about 40 volts a.c. for operation. Below the Val\-e and joined to it by a ground-glass socket is the clectrolysis cell, which is supported on a light hrass platform.The electrical connections to the valve and cell terminate in Belling-Lce plugs and sockets on the base of the stand and these are wired underneath to a Jones connector that receives the output from the control unit. Tlie apparatus is easily dismantled for cleaning purposes. OPERATIOS OF THE TITRIMETER- Settig!g ztj?~-To set up the end-point controls of the titrimeter, the pH meter reading is tirst adjusted to the end-point pH l:y meails of the knobs marked sm BUFI-ER, and t l i i i provides the control unit with a signal equal to that a t the end-point. Tlie knob marked E~I)-TJOIST (K3, Fig. 4) on the front panel of the control unit is adjusted until thv light just clianges from yellow to red.The pH meter reading is then adjusted to the pH at which TABLE I Octobcr, 1934: co I LON 1.1'12 I(: '1 ITKIRII.,TEII AVTOMATIC TITRATIOXS WITH THE COULOMETRIC TITRIMETER In each titration, the sample was 25 ml of 0.1 N sulphuric acid (factor 1.008) Counter Deviations Initial Final reading from mean End pH temperature, temperature^ " C " C 110.20 110.20 110.30 1 10.35 110.20 110.15 110*10 10990 109.90 109.90 110-10 110.10 1 10-05 109.70 109.90 '0.15 -+ 0.15 7 0.25 L 0.20 + 0-15 +0.10 + 0.05 -0.15 -0.15 -0.15 + 0-05 -f 0.05 0.00 - 0.35 -0.15 4.00 4.05 4.05 4.08 4.0 1 4-01 4.00 4.00 4-01 4.00 4-01 4.02 4.01 4-01 4.01 19 20 21 2 0 21 21 21 2 0 21 20 20 21 21 21 21 34.5 37.6 37.0 36-5 36.5 37.5 37.0 37.0 37.5 37.0 37.5 38.0 37.0 37.0 37.0 Standard deviation = 0.15 per cent.the titration rate is desired to slow down and an adjustment is made t o the L o w - c c R m A ' r ~ T A K T knob (R5, Fig. 4) until the light just changes from green to yellow. The pH meter is then standardised with a buffer solution and the instrument is ready for use. Titration-To carry out a titration, the beaker containing the sample is placed in the operating position, the counter reading is noted, and then the START button (S,) is pressed. \\'ith the cell described, a titration current of 1.3 amp. can be used and this is reduced to 2550 mA near the end-point. The minimum electrolyte flow that should he used is 10 in1 per minute. \$'hen the red lamp is illuminated, the titration is complete, and the reading on the counter is noted. The change in the counter reading is a measure of the titre.The counter reading is calibrated by titrating a known amount of acid or alkali and the concentration of an unknown sample is found by proportion. By this procedure the counter indication is directly related to the chemical standards of the laboratory : absolute efficiency of electrolysis does not affect the performance of the titrimeter. PERFORMASCE OF THE AUTOMATIC TITRIMETER- The instrument was tested in the routine analytical laboratory where titrations were performed coulometrically for several days. Some of the results are shown in Table I. The pH of the sample after titration has been recorded to show the efficacy of the apparatus in detecting the end-point. The initial and final temperatures of the sample are also recorded. It was found that standardisation once daily was adequate and the control settings required only occasional checking.Rapid operation requiring a minimum of sldl con- sistently gave results that were reproducible to within k0-:! per cent. The efficiency of the614 BETT, KNOCK AND MORRIS: AN AUTOMATIC [Vol. 79 cell was found to be 97.5 per cent. and was measured by performing some titrations with an accurately stabilised current and measuring the titration time. IODINE TITRATIONS WITH THE AUTOMATIC TITRIMETER- coulometrically in the cell. iodide, which dissolved the iodine as it was produced at the anode of the cell. 'some experiments were carried out to determine whether iodine could be produced The electrolyte was a. 5 per cent. aqueous solution of potassium Titrations TABLE 11 TITRATION OF SODIUM THIOSULPHATE BY C~OULOMETRICALLY GENERATED IODINE In each titration the sample was 10 ml of 0.1 N sodium thiosulphate and starch solution was used as indicator Counter reading Deviation from mean 36-76 + 0.05 37-00 + 0.30 36.70 0.00 36.96 + 0.26 37.20 + 0.50 36.90 + 0-20 36.80 +0.10 36.80 +0.10 36.30 - 0.40 36.80 + O s l o 36.30 - 0.40 36-60 - 0.20 36.60 -0.10 36.40 - 0.30 36.40 - 0.30 Mean 36-70 Standard deviation = 0.26 per cent.were performed in which the sample taken was 10 ml of 0.1 N sodium thiosulphate solution, and the results are shown in Table 11. The end-point indication was based on an ampero- metric principle described by Knowles and Lowden,lS who used a platinum indicator and calomel reference electrode.The electrodes could be connected directly to the terminals of the control unit provided that the 1500-ohm resistor (R, in Fig. 4) was removed. In the presence of excess of sodium thiosulphate, there was practically no e.m.f. across the electrodes, but an excess of iodine corresponding t o 0.2 mg in 100 ml was sufficient to produce an e.m.f. of 250mV. The control unit was so adjusted that an input of 200mV terminated the titration, and 100 mV was a sufficient signal to start titration at the slow rate, For successful titration, two modifications had to be made in the control unit- (i) the relay A in Fig. 4 was made to seal down when energised so that a transient signal of 100 mV caused the titration rate to drop permanently until the end-point ; and (ii) the flow and titration current were stopped simultaneously so that if, after the end- point has been reached, continued stirring shows that more iodine is needed, i t will be added dropwise.The titrimeter with these alterations was found to switch on and off repeatedly near the end-point until equilibrium was reached. It was unnecessary to wash out the cell between successive titrations, as there was practically nlo tendency for the iodine in the delivery arm to diffuse through the sinter. As the main source of error in these titration:; lies in the determination of the end-point, the standard deviation in the titration of 10-ml samples of 0.1 N sodium thiosulphate (Table 11) should not be very different from that found in titrating larger samples. There may, however, be a slight and inconsistent loss of efficiency in the cell owing to small amounts of oxygen being liberated with the iodine.RECENT DEVELOPMENTS Recent experience has shown that the electronic end-point detecting circuits of the automatic titrimeter can be usefully applied to volumetric titrations wherever electrodes canOctober, 19543 COULOMETRIC TITRIMETER 6 1 3 be made to yield a sufiiciently sharp voltage change a t the end-point. For example, with a glass electrode and conventional pH meter or metal electrode, e.g., tungsten or antimony, directly coupled to the apparatus, it can record changes of pH, whilst with the circuit described above for use in the determination of iodine it can be made to detect an amperornetric end- point. As the signals for “approach to” and “arrival at” the end-point take the form of sharp clianges of coloured lamps, these changes are more easily and accurately detected than with either a meter reading or with a colourcd indicator.Several special instruments Have been built €or routine applications of this kind. here are many analytical reagents that cannot be quantitatively generated by an electric current. For these an electricallji controlled burette having two speeds of flow has been made; this is shown in Fig. 6, and it consists of a conventional burette to which is attached by ilieans of a ground-glass joint a double valve, each part of which is identical with the electromagnetic valve described on p. 612. When both valves are open, liquid flows from the burette at a rate determined by the upper tap.When the upper magnetic valve is shut, the rate of delivery of the burette is determined by the glass tap in the by-pass arm. The rate of dropping should be such that between anj- two successive drops equilibrium is cstahlisfied in the titrating vessel. Trials of this apparatus have so far been limited to a few conventional titrations in which it has been successful, but there is no reason why it should not be used to perform any titration for which a suitable electrode system €or tlie detection of the end-point can be found. For rapid titration there must be adequate warning of the approach of the end-point; good stirring is czlso necessary to ensure an acciirate end-point. m The authors thank those of their colleagues in I.C.I.Ltd., who have helped with the analytical applications of the apparatus described above. APPENDIX LIST OF COMPOKEKTS USED I 3 THE COKSTRUCTION OF THE COULOMETRIC TITRIMETER CoZrmioL UNIT (FIG. 4)- I<, = 1600-ohn~, wire-wound resistance. 1x2, &, R,, = 100,000-ohm, O.Fi-watt, carbon resistance. 133, R, = 5000-ohni, helical potentiometer (Colvern Ltd.). L i = 100,000-ohm, wire-wound resistance. R = 33,000-ohm, I-watt, carbon resistance. K,, R,,, li,,, H,, == 220,000-ohm, OeS-watt, carbon resistance. it,, R2, = 1.5-megohm, O-T,-watt, carbon resistance. R,,, K,, = 1000-ohrn, @;‘,-watt, carbon resistance. R1,, It,,, R,,, R,, = 0.47-megohm, @!%watt, carbon resistance. it,,, I<,, = 47,000-ohn3, 0.5-watt, carbon resistance. 4 s = 108),000-ohm, -\vim-wound potentiometer.1123 == 3300-ohm, 0.5-vlr&tt, c arl )on resistance. K,, = 1.2-ol-im, precision r c tor, L.5 amp. rating jCroydoi1 Precision Instruments [{I, 2x17 - - 20,000-ohm, wire-wound resistamx. Iltd .) * = 10,000-ohm, 0*,5-\vatt, carboii resistance. == 100-pF, 12-volt working, elcLirolytic condenser. = O.l-pIt;, 350-volt u orking, paper condenser. , C,, C,, = l-pl?, 350-volt working, paper condenser. - - TrO-pLF, 50-volt working, electrd,lytic cmdensei . = 0-01 -pF, 350-vo1i working, piper condenser. C,,, C,,, C,, =; 0-05-1~F, 350-voit woritiag, paper condenser. =I- G4-p1?, 4%0-~011 working, electrolytic condenser. = EE91 valve. = 86A2 iicoii stabihscr i a l v e . , p2, 1’3 = Pilot lights. G-volt, 0.2 amp. (Arcolectric). = Push ?illtton \\vltch (Lucas). = Double-pole, two-way, wafer switch.= Iiiteqraiing motor, 3.5 volts (Llecti-o Methods Ltd.). T\7L, \-3 = 6D21 thyratroii v a l ~ c . I< =- Carperi‘cer polariscd relay, Type 3 . -1 = Post Ofice relay 10,000 ohm. 3 = Post Office relay 10,000 ohm. G = I’ost (Iffice relav 10,000 ohm. Sl K = Metal rectifier, Type RMI (Standard Te!ephoiie Co. Ltd.}.wo1. 79 616 BETT, KNOCK AND MORRIS = 50-ohm, 100-watt, Toroidal potentiometer. = 500-ohm, 50-watt, Toroidal potentiometer. = 2000-ohm, 50-watt, Toroidal potentiometer. = 1500-ohm, 50-watt, Toroidal potentiometer. = 100,000-ohm, O.5-watt, carbon resistor. = 96-pF. 500-volt working, 150(l mA ripple current, electrolytic condenser. = 32-pF, 450-volt working, electrolytic condenser. = Mains transformer : primary windings, 10-0-220-240 volts ; secondary windings, 220-0-220 volts at 1-6 amp.= Mains transformer : primary windings, 10-0-220-240 volts; secondary windings, 300-0-300 volts at 100 mA, 6.3 volts at 3 amp. = Low-frequency choke, 5 henries, d.c. resistance 20 ohms, rating 1.5 amp. = Low-frequency choke, 20 henries, 20 mA. = Selenium rectifier, 220 volts, full-wave, rating 1.5 amp. POWER SUPPLY (FIG. 3)- Rl R2 R3 R4 R5 c, CZ TI TZ L, L2 MR M1 P1 P2 B1 B2 F1 F2 F3 Vl s, s2 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Ammeter, moving coil, 0 to 5 amp. Neon indicator, Type F (G.E.C.). Pilot light, 6-volt, 0.2 amp. (Arcolectric). 3-pin plug (Bulgin) . 12-pin Jones connector (Painton). = Fuses, 3 amp. = Fuses, 5 amp. = Fuses, 250mA. = EZ40, rectifier valve. = Double-pole switch (Painton). = Double-pole switch (Painton). REFERENCE s Szebelledy, L., and Somogyi, 2.. 2. anal. Chem., 1938, 112, 313. Furman, N. H., Cooke, W. D., and Reilley, C. N., Anal. Chem., 1951, 23, 946. Carson, W. N., jun., and KO, R., Ibid., 1953, 25, 226. Buck,'R. P., Farrington, P. S., and Swift, E. H., Ibid., 1952, 24, 1195. MacNevin, W. M., and Baker, B. B , Ibid., 19512, 24, 986. Cooke, W. D., Reilley, C. N., and Furman, N. IFF., Ibid., 1952, 24, 205. Lord, S. S., jun., O'Neill, R. C., and Rogers, L. B., Ibid., 1952, 24, 209. Ramsey, W. J., Farrington, P. S., and Swift, E. H., Ibid., 1950, 22, 332. Sease, J. W., Niemann, C., and Swift, E. H., Ibid., 1947, 19, 197. Farrington, P. S., and Swift, E. H., Ibid., 1950, $2, 889. Shaffer, P. A., jun., Briglio, A., jun., and Brock.man, J . A., jun.$ Ibid., 1948, 20, 1008. Lingane, J . J., and Small, L. A., Ibid., 1949, 2Y., 1119. Cooke, W. D., and Furman, N. H., Ibid., 1950, 22, 896. Carson, W. N., jun., and KO, R., Ibid., 1951, 23, 1019. Epstein, J., Sober, H. A., and Silver, S. D., Ibid., 1947, 19, 675. Cooke, W. D., Reilley, C. N., Furman, N. H., Itlid., 1951, 23, 1662. DeFord, D. D., Fitts, J . N., and Johns, C. J., Arbid., 1951, 23, 938. Knowles, G., and Lowden, G. F., Analyst, 1953, 78, 159. , , Ibid., 1951, 23, 941. --- RESEARCH DEPARTMENT NOBEL DIVISION IMPERIAL CHEMICAL INDUSTRIES LTD. STEVENSTON, AYRSHIRE March 25th. 1964

ISSN:0003-2654    

DOI:10.1039/AN9547900607

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

October, 19541 EDISBT-RI-, GILLOW ASD TAYLOR 617 The Determination of Total Tocopl.ae-rcs1 BY J. R. EDISBURY, MIss JEAN GILLOW AXD R. J. TAYLOR -4 modified Emmerie and Engel test has been evolved to avoid nori- linearity, high or variable blanks and the use of corrosive solvents. A solution of the saponified material in light petroleum is purified by chromato- graphy, the tocopherol being adsorbed on mildly alkaline alumina, washed I\ itli diethyl ether - light petroleum solvent and subsequently eluted with chloroform. The eluate is concentrated by evaporation and treatccl with chlorolorm solutions of ferric chloride and 2 :2’-dipyridyl. -4fter the sohtion h , s stood for 5 minutes, a measured volume of water is added and the optical density of the aqueous extract is determined a t 520 mp; altcrnati\dy the depth of colour, which is stable for some hours, may be assessed visually, e.g., by comparison with a standard cobalt sulphate solution.The only interference encountered has been from highly oxidised caroteiioids, for which correction is possible. Sensitivity is five times that of the original Emmerie and Enqel test and precision is about 5 per cent. r. 1 OCOPHEKOL (fat-soluble vitamin E) is credited with a number of valuable physiological functions ranging from the relief of cardiac distress and the cure of “Crazy chick” disease to the general maintenance of fertility, and it is important chemically as an anti-oxidant in oils and fats. Four distinct tocopherols and their esters have been identified in the natural state and have since been synthesised.The commonest and physiologically most active form, cx-tocopherol, is the one naturally chosen for large-scale synthesis, the other three forms being much less active, f&tocopherol about one-third, y-tocopherol one-fifth and &tocopherol one -hundredth. They can be separated by paper chromatography (brown,l Russel Eggitt and Ml’ard2) and if a sensitive colour test is used, this may lead to a convenient method of individual assay of the four types of tocopherol. Meanwhile there is a need for a simple method of determining total tocopherol, particularly when a series of assays on similar matcrials is required. The method described in this paper was developed for the determination of tocopherol in feeding stuffs, with which the difficulties of isolating tocopherol effectively are probably greater than with any other suhstancc.. It is, however, suitable for general application and has the particular advantage of substantially increased stability of the i’ :2’-dipyridy1 complex with toco$herol.The magnitude of the general problem is indicated by the number and variety o f published tests that 1 1 a ~ ~ xisen through dissatisfaction with previous method:,. Competent reviews of these are given by 13acharach3 and Eden and B o ~ t h , ~ and useful informatic,n is provided b>- Tosic and 3loore.j Our iiiet’nod is a develop- ment of their original technique, wliich was based on the f x t that in the presence of 2:2’- dipyridyl one molecule of free tocopherol reduces two atoms of ferric iron and the resulting ferrous irnii forms a 3 to P chelated complex ion with tlic 2:2’-dipyriciv!.Hence each molecule of tcmpherol is stoicheiometrically equivalent to six molecules of 2 :Z’-dipyridyl. The complex is red and under ideal conditions givcs a pink solution o ~ i n g to absorption of light in the green region of the spectrum. Emmerie and Engel6 described the first quantitative approach. IAI?SOKPTIO?i SPECTRUM OF THE COMPLEX- The pink solution is conventionally described as showing a broad absorption maximam at 51 6 nip. Investigation with a visual spcctrophotQmeter showed that this broad region of absorption results from two overktpping bands, one at 484 inp and the other at 531 1 3 , ~ (see Fig-. 1). Measurements of optical density can he made with almost equal validity at any point within this range ; we have selected 520 mp as a suitable value with any commercial instrument ( w p have used Beckman DU, Unicam SPSOO with tungsten or hydrogen lamps and SPCiOO), and a spectral slit-width of LIP to 20 mp can be used without introduction of CSTOT618 EDISBURY, GILLOW AND TAYL0:R : THE DETERMINATION [Vol.79 MODIFICATION OF THE EMMERIE AND ENGEL METHOD- Of the several published modifications of the Emmerie and Engel test that we have tried, only those in which glacial acetic acid is used as !;ol~ent.'~*~~~~'' have given a reaction colour that is both stable and in linear relation to the amount of tocopherol present ; but the corrosive acid vapour could seriously damage a photo-electric instrument and we cannot recommend the routine use of these methods. Ethanol has been widely used as solvent, but it has certain limitations : the 2 :2'-dipyridyl solution is relatively unstable and the intensity of colour 520 /----k-, \ / / a \ \ / 512 / / / F E F - I 450 500 550 Wavelength, mp Fig.1. Spectra of relevant absorption for the determination of tocopherol: curve A, uncompensated test solution; curve B, test solution after allowance for blank; curve C, uncompensated blank; curve D, blank compensated against distilled water; curve E, contri- bution of excess of ferric chloride to curves A, C and D; curve F, contribution of excess of 2: 2'-dipyridy'.l t o curves A, C and D; curves G and G', spectroscopic components of curve B; curve H, cobalt sulphate. Length a, optical density of uncompensated test solution at 520mp; length b, optical density of uncompensated blank at 520 mp; and length G = a - b development is non-linear; the blanks are high anld the precision is not always certain.The interest that other workers have shown in the use of acetic acid is indicative of the difficulties with ethanolic solutions. Chloroform, with its faintly acidic bias, has proved to be as con- trollable as acetic acid, but the blanks were again high and the reaction colour was seldom pink. It was found, however, that the colour was transferable to water, and further investiga- tion showed that the transfer of the complex was quantitative. The aqueous solution had a clear pink colour, stable for some hours and suitable for spectrophotometric or other measurement, and the blanks were reasonably low (see curves C and D in Fig.1). Before the reaction is attempted, the tocopherol must be freed from other reducing substances, otherwise any substance that reduces ferric iron will be recorded as tocopherol. Emmerie and Engel themselves recognised this, but of the adsorbents suggested by them or by subsequent investigators, none that we have tried has been wholly satisfactory when dealing specifically with feeding stuffs. In the double column chromatographic assay of vitamin A (Boldingh and Drostll), the lower column of alkaline alumina retains the tocopherol and similar substances, but while tocopherol is undoubtedly retained it cannot afterwardsOctober, 19541 OF TOTAL TOCOPHEROL 619 be recovered; as tocopherol is markedly unstable in strong alkali, it is probably destroyed. Trials with alumina of different degrees of alkalinity led us to select alumina containing 0.7 to 0.8 per cent.w/w of sodium hydroxide, and at this concentration free tocopherol is quantitatively adsorbed from light petroleum solution and is firmly retained while most carotenoids, \itamin A and probably some other contaminants are removed by washing with dietliyl ether - light petroleum solvent; it can then be eluted with chloroform so that the impurities remain in a tenacious grey zone at the top of the column. I t is recommended that all tocopherol preparations, even concentrates, slioulcl be treated in this way before test. Pure x-tocopherol is recovered quantitatively and so can be used to assess the suitability of each batch of alkaline alumina.Some of the more highly oxidised carotenoids may also be retained with the tocopherol, and in such circumstances allowance can be made by following the Tosic and Moore5 technique of correction for p-carotene. They multiplied the amount of p-carotene, in pg per g, by a factor lying between 3.3 and 2.5 and subtracted the resultant value from the total apparent amount of tocopherol, in pg per g ; Kaunitz and Beaver8 used a factor of 4.8 under different conditions; for oxidised carotenoids we use the value 3.0. Although the correction is only approximate, it is generally small. Ideally the chloroform eluate is practically colourless and no correction is required, but if a correction is necessary, then- carotenoid content, in pg per ml = E at 450 mp x 4, where E at 450 mp is the optical density measured in a 1-cm cuvette.This can be calculated directly to concentration of carotene, in pg per g, in the original material and so, by the factor of 3, to the required tocopherol correction. METHOD RE AGENTS- reagent grade in distilled water and dilute to 100ml. grade in chloroform and dilute to 100ml. and dilute to 100 ml. Potassiziuz Jzj)droxide soZzitio7b-Dissolve 60 g of potassium hydroxide pellets of analytical- Ferric chloride solzitioir-Dissolve 0.2 g of crystalline ferric chloride of analytical-reagent 3 :~'-l)i~~'rir~~Z-Dissolve 1.0 g of 2 2'-dipyridyl of analytical-reagent grade in chloroform. a-Tocophrvol, pure. Gallic acid or pyrogallol. Chl o YO f ov MZ , 1) we. EtlzaTi ol, n bsolute.Light prtrolezm, boiliqig range 40" to 60" C. DietJz2ll ether-Freshly distilled over sodium hydroxide pellets. A ceto 11 e--Analyt ical-reagent grade. SztpP1~~ of carbon dioxide. A c t i m nluniiiza, staPzdard tj'pe--Prepare it by heating crystalline alumina trihydrate a t 800" C for 7 hours, and use the fraction that passes through a 150-mesh B.S. sieve. Actizle nlknliize alumina-To 10 g of activated alumina in a 50-ml round-bottomed flask, add 10 ml of a 0-75 per cent. w/v aqueous solution of sodium hydroxide, shake the mixture well, attach the flask to a water pump and leave it under reduced pressure (3 to 4 cm) for 1 hour. While the flask is still under reduced pressure, place it for a further 45 minutes in an oil-bath maintained at 135" C.Cool the alumina under reduced pressure and when cold transfer it to a small well-corked bottle. PROCEDC-RE FOIi EXTRACTION OF UKSAPONIFIABLE FATTY MATTER- Sarnjblc.7 coirtniizing more than 40 pg per g of tocopherol-Weigh up to 5 g of sample into a 200-ml flat-bottomed flask, add 5 ml of potassium hydroxide solution, 25 ml of ethanol and 40 to 50 mg of gallic acid or pyrogallol, and boil the mixture under a reflux condenser for 30 minutes. Transfer it to a 500-nil separator, using two 25-ml quantities of distilled water, and extract it once with 100 ml and then three times with 50-ml portions of diethyl ether. \!'ash the combined extracts successively with four 50-ml portions of lukewarm distilled water. Remove the ether by distillation in an inert atmosphere, add 2 ml of ethanol and heat the flask while passing a stream of carbon dioxide through it to dehydrate the620 EDISBURY, GILLOW AND TAYLOR THE DETERMINATION pol. 79 extract and to free it from solvent.Dissolve the residue in light petroleum and dilute to 5 ml. Samples containing less than 40 pg per g of tocopherol-Weigh 10 to 20 g of sample, add 100 mg of gallic acid or pyrogallol and extract the mixture for 2 hours in a Tait thimble with pcroxide-free diethyl ether or acetone. Remove the solvent by distillation in an inert atmosphere and immediately saponify the treated sample as described above. CHROMATOGRAPHY- Set up the chromatographic tube, shown in Fig. 2, fit a small wad of cotton-wool in the lower tip and place the tip on a piece of rubber to prevent the liquid from running through.Air pressJre of 65 Calibration 5 ml above alumina Calibration to show top level of alumina 5.0 mm internal diameter P1u.g of cotton-wool Fig. 2 . Apparatus used for isolation of tocopherol by chromatography. The column contains 1 g of alumina; when the three-way tap is in position 1, the air pressure is applied, but in position 2 the pressure is released through the side-arm, which is open to the atmosphere at A and is attached to the column for support Fill the narrow section of the tube with light petroleum and add 1 g of alkaline alumina, tapping the side of the tube to make the column compact. Remove the tip of the tube from the piece of rubber and allow the solvent to run through under a slight air pressure. When th(\ upper solvent level is just above the alumina, run in 1 ml of the solution of unsaponifiable matter and repeat the procedure.Then add 5 ml of a 36 per cent. v/v solution of diethyl ether in light petroleum. Now replace the receiver by a small clean flask and pass 5 ml of chloroform through the column. COLOUR REACTION- Evaporate the chloroform solution to approximately 0.5 ml in an inert atmosphere. Cool it and add 0.2 ml of ferric chloride solution and then 0.5 ml of 2:2’-dipyridyl solution. Allow the solution to stand for 5 minutes, then add 3.6 ml of distilled water and shake the mixture vigorously. Transfer it to a centrifuge tube and spin it at a moderate speed in a small hand centrifuge for 10 seconds. With a pipette, remove the lower (chloroform) layer and spin it again more vigorously.Make a blank determination with the reagents.October, 19541 OF TOTAL TOCOPHEROL 63 1 MEASUREMENTS- measure its optical density by a spectrophotometer at 520 mp. solution. The tocopherol content, in pg per g, of the meal may now be calculated directly. A bsorptiometric-An absorptiometer fitted with a 520-mp filter is first calibrated in terms of amount of tocopherol, in pg, by means of tests with a series of solutions of pure cc-tocopherol in chloroform. Then if R, and R, are the values from calibration results- S+ectrophotometric-Transfer the clear pink aqueous solution to a 1-cm cuvette and Repeat this with the blank Then if E, is the optical density of the test solution and Eb that of the blank- 100(E, - E,) = tocopherol, in pg, in aliquot taken for test.R, - R, = tocopherol, in pg, in aliquot taken for test. From this value the total amount of tocopherol, in pg per g, can be calculated. CORRECTION FOR CAROTENOIDS- Most carotenoids are eliminated during the chromatographic stage, but occasionally oxidised products that remain with the tocopherol are present ; this is indicated by the orange- yellow colour of the chloroform solution. Treat an independent aliquot of the solution by passing it through a column of alkaline alumina as before, then remove the chloroform by evaporation in carbon dioxide, dilute the residue to a suitable volume with light petroleum and measure the optical density of this solution at 450 mp. , Let E at 450 mp = optical density. V, = volume of light petroleum solution, in ml.E at 450mp x 4 = carotenoid content of solution, in pg per ml. E at 450 mp x 12 = equivalent tocopherol content of solution, in pg per ml. E at 450 mp x V, x 12 = equivalent tocopherol content of aliquot taken for test, This gives the value of the equivalent tocopherol, in pg per g, of the interfering carotenoids in the meal, and so the necessary correction can be made. in pg. EXAMPLES- (a) A 10-g sample of a meal was extracted and the extract was saponified. The un- saponifiable matter was diluted to 5 ml with light petroleum and 1-ml aliquots were taken for test ; the mean E at 520 mp reading of the 2 :2’-dipyridyl complex was 0.44. The correction from a blank test was 0-16. . Then nett E a t 520 mp value attributable to tocopherol = 0.28. Tocopherol per ml of aliquot = 100 x 0-28 = 28 pg.Tocopherol content of meal = 14 pg per g. (b) The same meal as in (a) was fortified with 40 pg per g of pure tocopherol. A 3-g sample was saponified directly and the unsaponifiable matter was diluted to 3 ml with light petroleum; 0.5-ml aliquots were taken for test. The mean E at 520 mp reading was 0.44. The nett E at 520 mp reading attributable to tocopherol was 0.28. Tocopherol per 0-5-ml aliquot = 100 x 0.28 = 28 pg. Total tocopherol content of meal = 56 pg per g (compared with a nominal 54 pg per g). (c) A 5-g sample of another meal was saponified directly and the unsaponifiable matter Carotenoids The was diluted to 5 ml with light petroleum; 0.75-ml aliquots were taken for test. were present in the chloroform eluate.correction from a blank test was 0.18. Then nett E at 520 mp reading = 0.38. Apparent tocopherol content of meal = 60.6 pg per g. residue was dissolved in 3.6 ml of light petroleum. The mean E a t 520 mp reading was 0.56. Carotene correction-A 0.75 ml aliquot of the eluate was treated as described above, and the The E a t 450 mp was 0.168. 0.168 x 3.6 x 12 = 9.6 pg per g. 0.75 Equivalent of tocopherol in meal = True tocopherol content of meal = 41 pg per g.622 EDISBURY, GILLOW AND TAYLOR : THE DETERMINATION [Vol. 79 DISCUSSIOX This modification of the Emmerie and Engel test requires only about one-fifth of the amount of tocopherol per reaction as the original test. It may be that this low level is one of the reasons for closer adherence to linearity. There are, however, probably several other factors involved in what is undoubtedly a complex reaction.In the reaction of tocopherol with ferric chloride and 2 :2’-dipyridyl, the results have been best when sufficient iron was present to oxidise, theoretically, 320 pg of tocopherol and sufficient 2 :2’-dipyridyl was present to oxidise 2400 pg, although only 20 to 70 pg of tocopherol are present in a typical reaction mixture; unless a11 these amounts of reagent are present in a total of 1.5 ml or less of solution, the reaction is not complete. (This is similar to the reaction of vitamin A with antimony trichloride, when a molecular ratio of lo5 to 1 is required, but is of no avail unless the antimony trichloride solution is nearly saturated.) As it is the tocopherol that is being determined and not the iron or 2:2’-dipyridyl, it is to be expected that both reagents should be in excess; but the general behaviour is unusual.Below a molecular ratio of iron to tocopherol of 10 to 1, .i.e., an excess of five times the nominal stoicheiometric requirements, there is a marked dimmution of colour even when 2 :2‘-dipyridyl is present in excess, and the diminution can be partly arrested by further additions of 2:2’- dipyridyl. With about 20 times the stoicheiometric amount of iron, there is again a diminution of colour no matter how much 2:2’-dipyridyl is present. In the presence of a reasonable excess of iron (e.g., 10 to l ) , 2:2’-dipyridyl requires to be in much greater excess for full colour development. With amounts of iron up to about 50 times the nominal quantity, the effect of progressively increasing the amount of 2 :2’-dipyridyl can be discerned ; thereafter the rate of increase of colour is undetectable.Under our conditions of test there is an excess of iron of between 4.5 and 16 to 1 and of 2:2’-dipyridyl of between 34 and 120 to 1 over the stoicheiometric requirements, with usual values of 10 to 1 and 75 to 1, respectively, repre- senting molecular ratios of tocopherol to iron to 2:2’-dipyridyl of 1 to 20 to 450, in place of the theoretical 1 to 2 to 6. The experimental requirement that 20 to 70 pg of tocopherol should be present in the reaction mixture makes it generally expedient to try one or two pilot tests on the solution of the extract before proceeding t o the chromatogiraphic stage. For these tests an aliquot of the test solution is taken and freed from light petroleum, and it is then treated directly with ferric chloride and 2 :2’-dipyridyl in the prescribed manner.The complex is transferred to water and its absorption at 520 mp is determined. This will yield a result for the tocopherol that may be 100 per cent. too high, but the test is quickly made and it will enable a correct choice of aliquot to be made for the more exact analysis. Under our conditions development of colour for less than 3 minutes usually gives low results; over 7 minutes may give slightly high results with a correspondingly high but less predictable blank. The choice of 5 minutes is a workable compromise. Measurement of the time of development has been used by Stern and BaxterlO to distinguish between the various forms of tocopherol ; in glacial acetic acid the a-form produces colours most rapidly and the 8-form least rapidly. The method is not likely to be reliable, but should not be overlooked.More promising as a means of differentia- tion is paper chromatography of a novel kind devised by Brown1 and modified by Russell Eggitt and Ward.2 Filter-paper thinly pre-treated with petroleum jelly or liquid paraffin, B.P., is used, with 75 per cent. aqueous ethanol as the developing solvent. The developed spots are identified on one strip and eluted from ccrresponding locations on a parallel strip. This method might well be linked with our own test. The factor of 100, which we use t o calculate the amount of tocopherol in the test solution, in pg, is applicable only when the total volume of the aqueous solution is 3-6 ml and a 1-cm cuvette is used for measurement.It is based on an E:Fm value of 360 at 520mp for the 2:2’-dipyridyl complex in terms of tocopherol. As the complex is really one of reduced iron, this value is conventional, but it can be used for the purpose of calculation when volumetric conditions of test other than those here described are chosen. If a tocopherol standard is not available but the alkali-treated alumina is known to be efficient, the colour due to the tocopherol can be visually matched against an aqueous solution of cobalt sulphate. In Fig. 1 the absorption specttum of the iron - 2:2’-dipyridyl complex is compared with that of cobalt sulphate, which shows a maximum at about 512 mp and, apart from divergence below 450 mp, follows the over-all shape of the complex curve more The time factor is important but not critical.Octobcr, 1%4] OF TOTAL TOCOI’HEIIOL f;23 closely than most of the stanclards tlmt find ready acceptance in abridged spectrophotomctrv. This divergence imparts a blue tinge that can be corrected by addition of ;i small amount of ferric chloride to the cobalt solution, so securing an almost pcrfcct visual match. -1s ;t. first approximation, a 7 per cent. w/v solution of rccrystallised cobalt sulpliate r;in bc taken as visually equivalent to 5 pg per ml of tocopherol. X series of six narrow test tubes containing graded concentrations of c o l d t sulphate from about 3 to 1 per cent. w/v at intei-vals of about 0.4 per cent., antl a constant 0-2 pcr ccnt. w/v of crystalline ferric chloride irrespective of cobalt concentration, covers a uselul 1-mge suited to the conditions of test already outlined. Aqueous extracts of measured volume froin known reaction mixtures can bc directly assessed by visual comparison to within probably 10 to 15 per cent. ; alternatively, a galvanometer instrument can be roughly calibrated with cobxlt sulpliate, or cobalt sulpliate can be used on one side of a dip-tube colorimeter. TS ILF~Y ER IX~I ITED i’orir SV’NLIGHT Cr I ES 1-1 I RE

ISSN:0003-2654    

DOI:10.1039/AN9547900617

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Octobcr, l%4] Volumetric Determination of Pectin as Calcium Pectate X inethotl is described lor the direct (letermination o f tlie calcium coilten t ol precipitated calcium pectatc. b y titration with e t h ~ leiiecli;t1uiiictctr;l-;1~~.ti~ acid. l‘he pectate ions do not in terfere. The values so tleterrninetl are useful ;is ;L measure of the calcium pectate yield of thc sample from which t h e precipitate is derived. ‘I‘he calcium contents of precipitates containing A t o 24 mg of calcium (as c:~lcium carbonate) were proportion;Ll to tht: qmntities o f citrus pectin from \irhicli the precipitates came. The procedure described lor s~tI)onific,Ltioii ant1 precipitation of pectin c;111 be completed within 2 hours. GKIYIME ~ R I C : determination as calcium pectate has been widely used for measirring Hintonl used the phrase “calciiiin pectate yield” to denote tlie nv3glit of calcium pectate produced from tlic sample expressed as a percentage o f tlie original saniple.Thc. same author found the criterion uscful in assessing the purity of pectins cxtrat tcd from fruits. 130th lie and Branfoot6 f o m t l tlic precipitate to be of fairl). const;~nt coinpobitioii. Despite its cmpil-icism, the method is widely used and is particuhrly valuahlc. for plirpses of comparison. ’The significance of the 1-esiilts aiicl tlie nimei-ous modi-fications tli<tt t l i 1 2 method has undergone are cliscussetl by Kertex2 Tlie essential stages in all metliotls for the determination of c:tlcium pc.ct:ite ~;ield ;ire extraction, purification, saponification, acidifimt ion, precipitation, filtratioii, J ~ X ~ , I ~ in ancl weighing.I t is not proposed to consiclci- in this pap‘cr tlic estrnition ; ~ n d purification stages, and it is assumed throughout t h a t a siiitably purified sollition of y c t ;TI 15 liciiig examined. For details of the methods o f purification and extraction, rcfcwncc shouicl be nlciclc to the work o f Emmet ancl (’arrk3 rtntl to that of Kertcl;z.2 Table I illiiitratcs the conditions used in some typical proccclures recommended for the determination of tlie c;:Iciiim pectate yield.624 HOLT : VOLUMETRIC DETERMINATION OF [Vol. 79 TABLE I CONDITIONS USED FOR THE DETERMINATION OF CALCIUM PECTATE YIELD Normality of NaOH used for Time of Method saponification saponification, hours Carr6 and Haynes4 .. .. 0.025 Overnight Hintonl (“rapid method”) . . 0-2 0.5 Kertesz2 . . .. .. 0.1 1 King5 . . .. .. 0.02 1 Interval between precipitation Time of and filtration, drying, hours hours 1 12 1 Not specified 1 2 None Overnight The total time taken for the shortest procedure, that of Hinton, would be at least 4 hours. The object of the present work was to replace the drying and weighing of the calcium pectate precipitate by a titration of the calcium, and also to reduce the time of saponification and the interval between precipitation and filtration to a minimum. The method evolved by Schwarzenbach, Biedennann and Bangerter7?* was adopted as a basis for the determina- tion of calcium. EXPERIMENTAL An outline of the methods used will be given first to explain the various conditions that were studied.The full recommended procedure is described below. MODIFIED CARRI~ AND HAYNES PROCEDURE FOR PRECIPITATION- An aliquot of the citrus pectin solution was measured into a conical flask and made up to 300ml with water, 100ml of 0.1 N sodium hydroxide were added, and the solution was left to stand overnight; 50 ml of M calcium chloride were then added and, after 1 hour, the contents of the flask were boiled for a few minutes and then filtered. The flask and filter-paper were washed twice with boiling water. The precipitate was washed back into the original flask and boiled for 5 minutes with 200 to 300 ml of water. The contents of the flask were filtered through the same filter-paper a s before and washed with boiling water until chloride was absent from the filtrate, as shown by testing it with silver nitrate.It should be noted that the washing technique is somewhat different from that originally proposed by Carrk and Haynes. It is shorter and has been found satisfactory by later workers, e.g., Kertesz2 SHORTENED PROCEDURE FOR PRECIPITATION- An aliquot of the citrus pectin solution was measured into a conical flask and a quarter of its volume of 0.1 N sodium hydroxide was added. After 5 minutes, 50 ml of A; acetic acid and sufficient water to make the total volume about 350 ml were added. The addition of calcium chloride, filtration and washing were as in the modified Carrk and Haynes procedure The interval between precipitation and filtration was varied as described below. DETERMISATION OF CALCIUM IN THE ASHED PRECIPITATE- The ash was dissolved in 1 ml of concentrated hydrochloric acid and this solution was evaporated to dryness.The residue was dissolved in water and the calcium in this solution was titrated with sodium ethyleoediaminetetra-acetate solution (EDTA solution), Eriochrome Black T being used as indicator. DIRECT DETERMINATION OF CALCIUM IN THE PRECIPITATE- The precipitate was taken up in excess of buffered EDTA solution and its calcium content was evaluated by titrating the excess of E.DTA with standard calcium solution. By means of these methods, the following experiments were performed in an attempt to shorten the over-all procedure. All the results are calculated for 100 ml of pectin solution and expressed as mg of calcium carbonate.COMPARISON OF DETERMINATION OF CALCIUM IN THE PRECIPITATE DIRECTLY AND AFTER It was necessary first to determine whether the titration of calcium by EDTA was Four 100-ml aliquots of pectin solution were ASHIXG- applicable in the presence of pectate ions.October, 1!354] PECTIN AS CALCIUM PECTATE 62,; treated by the modified Carrk and Haynes procedure, and 12.2 and 12.3 mg of calcium were found in the ashes from two of tlie precipitates. The calcium contents determined by direct titration of tlie other two precipitates were 12.1 and 14.2 mg. This agreement was satis- factory. TIME OF SAPOSIFICATIOS- The interval between precipitation and filtration was 1 hour. The calcium found after ashing two of tlie precipitates was 12.3 mg in each instance.By direct determination the calcium contents of the other two precipitates were found to be 12.1 and 124mg. The shortened saponification procedure appeared to be suitable. \Vhen the means are considered, the agreement in these experiments hetween titration of calcium directly and after ashing is good, but the duplicate results of the direct determina- tions do riot agree with each other. Eight more determinations by the direct method over tlie range 60 to 200ml of pectin solution were therefore made. The results, together with the two previously determined, are shown in Table 11. Four 100-rnl aliquots of pectin solution were treated by the shortened procedure. TABLE I1 REPRODUCIBILITY OF RESULTS BY SHORTESED PKOCEI)C‘RI-1 CaCO, ,per 100 ml of pectin solution, Pectin solution taken, ml mfi 5 0 12.2, 12.1 150 12.4, 12.3 200 12.2, 12.1 100 12.1, 12.4, 12.3, l-L*:< IATTERVAL BETIVEES PRECIPITATIOS ASI) FILTRATIOX- The time of standing between precipitation and filtration was first reduced to 10 minutes without any apparent effect on the filtration or the results.A series of results determined with filtration immediately after precipitation were also similar. In Table 111 these results arc sumniarised and compared with those in Table 11. Xeither tlie time of saponification nor the interval between precipitation and filtration had any effect on the quantity of water or the time required to wash the precipitate free from chloride. TABLE I11 EFFECT OF TIXE ISTEIII’AL BETWEEX PKECIPITATIOS AS11 FILTR.ITIOS Time of Xuniber of Standard standing, tests Mean, d e ~ iation, In 1 n ti t eb mg per 100 nil mg per 100 in1 0 6 12-1 0.18 10 6 12.3 0.12 60 10 ] 2.2 0.12 The means of the samples allowed to staiid 0, 10 and 60 minutes were not significantly different.METHOD REAGESTS- Sodium hydroxide, 0.5 K. Acetic acid, 1.0 ic’. C’alcium chloride, 1-0 11. Ljiifler solzifioii-Dissolve 40 g of borax, 10 g of sodium hydroxide and 5 g of sodium sulphide in distilled water and make up to 1 litre. Staizdnrd c a l c i t m .solutioii--To 0.5 g of pure calcium carbonate, add 2-5 ml of diluted hydrochloric acid (1 + 1 ) . Allo\v tlie solution to stand until the calcium carbonate is dissolved. EJt~yZe~zenicil.)iiiietetra-acetate solirtion-Dissolve 4.0 g of the disodium salt in 43 ml of 0.5 AY sodium hydroxide, makc up to 1 litre and standardise this solution peri~dically,~ as described below, against the standard calcium solution.Make up to 1 litre in a calibrated flask.626 HOLT : VOLUMETRIC DETERMINATION OF [Vol. 79 Eriochrome Black T indicator solution-To 1-10 ml of N sodium carbonate and 30 ml of Shake the solution and make it up to 100 ml with distilled water add 1.0 g of solid indicator. isopropanol. PROCEDURE- Stavzdardisation of the EDTA solution-To 130 ml of water and 20 ml of EDTA solution, add 10 ml of buffer solution and 3 or 4 drops of indicator. Titrate with the standard calcium solution until the first pink tinge appears in the solution, i.e., the colour changes from blue to purple. Determination of pectin-Measure into a 500l-ml conical flask an aliquot part of the solution under test.It should be less than 200 ml in volume and contain preferably 30 to 120 mg of pectin (calculated as calcium pectate). Add from a graduated cylinder one-quarter of the volume of 0.5 N sodium hydroxide and rotate the flask. Set the mixture aside for 5 minutes, add 50 ml of N acetic acid, mix the contents of the flask, add water to make the total volume about 350ml and again mix. From a burette or otherwise in a thin stream, add 50 ml of M calcium chloride solution while vigorously rotating the flask. Allow the flask to stand for 10 minutes. Boil the contents of the flask for 2 minutes and filter through a hard, fast, 15-cm filter- paper (Whatman No. 541 is suitable). Wash the flask and the filter-paper twice with boiling water. Return the precipitate to the original flask and boil it for 10 minutes with 200 to 300 ml of water.Pour the contents of the flask through the same filter-paper. Filter and continue to wash with very hot water until the filtrate shows no reaction for chloride when tested with silver nitrate. Wash the whole of the filtrate back into the original flask with boiling water, making the volume roughly 150ml. Add 10ml of buffer solution and an excess of EDTA solution. Heat the flask until the contents are boiling or almost boiling, then cork it and shake it, cautiously at first, until the precipitate dissolves. Cool the flask until it can be handled com- fortably. Add 3 or 4 drops of indicator and titrate with standard calcium solution to the same colour change as in standardising the EDTA solution.The quantity of EDTA to be added to the suspension of calcium pectate is problematical. It is convenient to have an excess of about lOmE., but the result does not depend on the exact amount. If the approximate pectin content of the solution under test is known, use can be made of the fact that 1 ml of EDTA solution is approximately equivalent to 5 mg of pectin. Even with a completely unknown solution, an estimate of the amount present can, with a little practice, be made from the appearance of the precipitate. Failing these devices it will be necessary to add the EDTA solution in, say, 10-ml increments, shaking between additions until the precipitate dissolves. CaZcuZation and expression of res.uZts-The theoretical calcium content of the calcium salt of pectic acid is 10.2 per cent.The average calcium content of the precipitates, as deter- mined by most workers, is about 7.6 per cent. On this basis 1 mg of calcium carbonate is equivalent to 5.3 mg of calcium pectate. For such purposes as comparing a series of samples or extracts from similar sources, it will probably ble most convenient to express the results as mg of calcium carbonate per g of material or per 100ml of extract. Meanwhile wash the filter-paper once or twice with boiling water. DISCUSSION AND CONCLUSIONS The factor that is most likely to affect the reliability of the gravimetric yield is co- precipitation of araban and galactan. Various quantities of these substances are present in extracted pectins, and they are extremely difficult to remove either by precipitation with miscible solvents or by precipitation of pectic acid as its salts.2~10 In determining calcium pectate yields, Peynaudll found arabans present in his precipitates.The results were 8 to 10 per cent. higher than those determined by titrating the pectic acid with an alkali. Calcium pectate yields determined by titration of the calcium in the precipitate will probably not be affected to the same extent as the gravimetric results by co-precipitation of araban and galactan. No appreciable amount of calcium would be expected to combine with these substances, although they might hinder the combination of calcium with the pectic acid. There are other factors that might affect the ratio of calcium to pectic acid in the precipitate, but in extracts from similar sources, i%t least, these are not likely to detract seriously from the utility of the method for rapid comparative purposes.October, 19541 PECTlN AS CALCIUM PECTATE 627 REFERENCES 1. 2. 3. 4. 5. 6. 7 . 8. 9. 10. 11. Hinton, C. L., “Fruit Pectins: Their Chemical Behaviour and Jellying Properties,” D.S.I.R. Food Kertesz, 2. I., “The Pectic Substances,” Interscience Publishers Inc., Sew York, 1951, p. 225 et seq. Emmet, A. M., and Card, M. H. l’iochem. J., 1926, 20, 6. Card, M. H., and Haynes, D., Ibid., 1922, 16, 60. King, J., Analyst, 1925, 50, 371. Branfoot, M. H., “A Critical and Historical Study of the Pectic Substances of Plai~ts,” D.S.I.R. Schwarzcnbach, G., Biedermann, W., and Bangerter, F., Irlelv. Clziiiz. Ada, 1946, 29, 811. Biederman, W., and Schwarzenbach, G., Chiwzia, 1948, 2 , 56. Goetz, C. A,, Loomis, T. C., and Diehl, H., Anal. Chevn., 19.50, 22, 798. Hirst, E. L., J . Chew. SOC., 1942, 70. Peynaud, E., Industv. Agric. Ahwent., 1951, 68, 609. Tnvestigation, Special Report Number 48, H.M. Stationery Office, 1939. Food Investigation, Special Report Slimher 33, H.M. Stationcry Ofhcc, 1929. THE FRUIT AND VEGETABLE CANNING BND QUICK FREEZING RESEARCH ASSOCIATION CHIPPING CAMPDEN, GLOS. r l p ~ i l Sth, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900623

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

October, 19541 627 The Biological Assay of Adrenaline with the Hexamethonium-treated Cat BY G. F. SOMERS A method is described for the assay of adreiialine by means of the hesaniethonium-treated cat. I t obviates the need for spiiialising and the results by this method compare fa vourably with those by other methods. Hesamethoniclm is a potent hypotensive drug that blocks the sympathetic aiid parasympathetic ganglia, so providing a simple alternative procedure for preparing the cat, and the results by this method are compared ivitli those by other methods. XUMEKOUS methods have been described for the biological estimation of adrenaline, hut few of them are suitable for routine assay purposes. They have been reviewed by M’est,l Gaddum2 and Burn, Finney and Goodwin.3 The method used should be amenable to suitable design and statistical analysis, so as to permit the precision of the estimates to be evaluated and to provide evidence that the actions of the preparation under test and of the standard preparation do not differ qualitatively (Emmens4).Methods based on the vasopressor action of adrenaline on mammalian blood pressure are preferred to those utilising isolated tissue preparations, because the responses are more regular and the preparation is not so susceptible to the presence of interfering substances such as preservatives. Elliott’s5 method, involving use of the “spinal” cat, is the commonest in this country; the official method in the United States6 uses the atropinised dog. The technique of preparing a spinal cat is relatively simple in experienced hands, but it does involve a radical surgical procedure, haemorrhage may be severe and casualties can occur.The cat is spinalised in order to give a low blood pressure and to destroy the parasympathetic compensatory mechanisms, so enhancing its sensitivity to adrenaline. The use of hexamethonium,7 a potent hypotensive drug that blocks the sympathetic and parasympathetic ganglia, provides a simple alternative procedure for preparing the cat. The use of this drug in adrenaline assays is described and the results by this method are compared with those by other methods. METHOD The cat is anaethetised by direct intravenous injection of chloralose (60mg per kg), with the addition of pentobarbitone sodium (6 mg per kg).” The blood pressure is recorded from the carotid artery and injections are made into the femoral vein.A dose of 50 mg per kg of hexamethonium bromide is given subcutaneously, and the blood pressure is allowed t o fall to a steady base-line, usually about 70mm of mercury. * As the proprietary preparation Xembutal. The assay procedure is as follows.628 SOMERS : THE BIOLOGICAL ASSAY OF ADRENALINE [Vol. 79 Occasionally cats have been found to be refractory to even larger doses of hexamethonium, and they have had to be spinalised. If the base-‘line is unsteady, the injection of 1 mg per kg of atropine often stabilises it. The standard is a sample of adrenaline as a 1 in 1000 w/v solution prepared according to the directions of the British Pharmacopoeia for Solution of Adrenaline Hydrochloride.s Dilutions of this (1 ml in 40 ml) are prepared in normal saline immediately before the assay and stabilised by the addition of 100 mg of ascorbic acid.Fresh dilutions are prepared each hour. Similar dilutions of the test preparation are prepared. A dose of the standard that produces a submaximal effect is determined, and this is designated the high dose of the standard, S,. A dose of the test is then determined with about the same effect, Tl. The low doses of the standard, S,, a:nd of the test, T,, are usually half the high doses, and it is important that the ratios should be the same. Suitable doses having been selected, they are given at &minute intervals in a randomised order according to a Latin Square design. Each assay therefore consists of sixteen responses, but additional responses may be obtained to suit the accuracy required.RESULrS The results of a typical assay are shown in Table I and the statistical analysis in Table 11. The calculation of potency ratio and limits of error were made by the methods described by Schildg and Holton.lo The test solution was art 80 per cent. dilution of the standard, the strength of which was unknown to the analyst. The calculated potency was 81.6 per cent. with fiducial limits of error (P = 0.05) from 75.7 to 87-9 per cent. The slope b was 56-4 and the index of precision s / b or A, 0.027. Hence the method is accurate. From 18 assays the mean values for b, s and s / b were: b = 53-3 with a standard deviation of 13.0; s = 2.07 with a standard deviation of 0.70; and s/b = 0.04 with a standard deviation of 0.019.TABLE I INCREASE I N BLOOD PRESSURE, IN MILLIMETRES OF MERCURY, WHEN ADRENALINE IS ADMIKISTERED TO A HEXAMETHONIUM-TREATED CAT Groups Preparation \ and dose Dose, 1 2 3 4 0.4 60 62 60 62 0.2 40 42 47 46 0.4 55 55 56 57 T2 0.2 38 38 39 41 sum 193 197 202 206 s, s2 TI TABLE I1 STATISTICAL ANALYSIS OF RESULTS IN TABLE I Degrees of Source of variation freedom Groups . . . . . . .. 3 Regression .. . . . . 1 Deviation from parallelism . . 1 Standard and unknown . . .. 1 Residual error . . . . . . 9 Total . . . . . . .. 15 Sum of square Variance F 24 8 3.31 100 100 41.39 1156 1156 478 0.25 0-25 0.1 21-75 2.416 - 130,” Sum 244 175 223 156 795 P > 0.05 < 0.001 <0.001 > 0.06 - Sm = 0.0143. b = 56.4. s = 1.55.s/b = 0.027. For 9 degrees of freedom, t = 2.262 (P = 0.05). The fiducial limits for M = jS, x t = 0.0323. As M = 1.9115, fiducial limits are 1.9438 and i.:3792. The potency of T is therefore 81.6 per cent. of S with fiducial limits (P = 0.05) from 75.7 to 87.86 pcr cent. These results compare favourably with those published by Noelll for the atropinised dog (see Table 111), but the slope, b, is not quite so steep. A limited number of comparative assays were carried out on the spinal cat; their results (see Table 111) suggest that this is the most accurate method of all. The use of hexamethonium can also be applied to the dog, with results similar to those for the hexamethonium-treated cat. The cat method has alsoOctober, 1954; IVITH THE HEXAMETHONI~M-TREATED CAT 629 been compared with the better known isolated organ preparations in which the same exyeri- mental design was used. The results, shown in Table 111, confirm the superiority of the blood pressure methods.COMPARISOK OF METHODS OF ASSAYISG ADRENALISE Me tliod Author I , Dog blood prcssurc . . . . Koel’l 7 3 Cat blood pressure, hexamethonium] r 53 I 31 Cat blood pressure, spinal . . * * I Rabbit gut . . . . . . . . I I 79 Rat colon, acetylcholine . . (12s Dog blood prcssure, hexamethonium } Soniers { 41 Rat colon, Carbachol . . 1: iGaCldum Rat uterus, Carbachol . . . . and J Lembick12 1 * One experiment only S 2-41 2.09 1.4.‘ 1-77 5.00 7-55 :‘.O(j* - Kumber of assays 2 i 19 4 2 4 !I !) #- REFERENCES 1. 2. 3. 4. 5. 6. United States Pharmacopoeia, Fourteenth Revision, 1950. 7. 8. British Pharmacopoeia, 1953, p. 21. 9. IYest, G. B., “Hormone Assay,” Acadeniic l’ress, New York, 1930, p. 91. Gadclum, J . II., “Akthods in Medical Iicitai-cli,” >’ear Book Publishers, Chicago, 19.50, 3, p. 116. Burn, J. H., Finney, D. J., and Gooclwin, L. G., “Biological Standardisation,” Second Edition, Emmens, C. IV., “Hormones: A Survcy oi their Properties and Uses,” The Pharmaceutical Press, Elliott, T. R., J . Physiol., 1912, 44, 374. l’aton, W. D. M., and Zaimis, E. J., Brit. J . Phnr.tnacol., 1951, 6, 155. Schiid, H. O., J . Physiol., 1912, 101, 116. 10. Holton, P., B y i t . J . Phurmacol., 1948, 3, 328. 11. Soel, Zi. H., J. Pl~nrn~ncol., 1915, 84, 278. 13. Oxford University Press, 1950, p. 218. London, 1961, p. 113. Gadrlum, J . H., and Lembick, I?., UYit. J . YJiavmacol., 1949, 4, 401. SCHOOL O F I’H.4K1RIA4CY l T ~ ~ ~ ~ ~ ~ OF LOXDOX

ISSN:0003-2654    

DOI:10.1039/AN9547900627

出版商:RSC    

年代:1954

数据来源: RSC