摘要:

OCTOBER, 1967 THE ANALYST Gamma-activation Analysis Vol. 92, No. 1099 A Review* BY C. A. BAKER (Analytical Sciences Division, A .E.R.E., Havwell, Bevks.) SUMMARY OF CONTENTS Introduction Types of gamma activation Nuclear considerations for (y,n) reactions Sources of high energy photons Electron accelerators Cyclic machines Linear machines Characterisation of the radiations Standards Determination of light elements, oxygen, nitrogen, carbon and fluorine The use of decay curves and energy thresholds The use of spectrometry and other instrumental methods Destructive analysis by rapid chemical separations Determination of heavier elements of atomic numbers equal to, or greater than, 10 Conclusions ACTIVATION analysis has now become firmly established as an analytical method of wide applicability.The high cross-section of many elements for thermal neutrons has enabled sensitive methods to be devised for trace-element analysis and fast-neutron activation by using 14-MeV neutrons produced by accelerators has proved useful for the measurement of major and minor components in many matrices. Charged-particle activations have also been applied to particular problems where the limited depth o€ penetration is no disadvantage and can even be turned to advantage in the analysis of surfaces. All activation analysis is free from the hazards of reagent blanks, and neutron activation has the additional advantage of homogeneous irradiation and the facility of etching the surface of a sample between activation and processing in order to remove any contamination. In recent years gamma-activation analysis has been exploited, particularly because of the possibility of doing intact analysis on ultra-pure materials for the light elements carbon, nitrogen and oxygen whose nuclear properties are unf avourable for neutron activation.Because of their penetrating nature gamma photons share with neutrons the advantages of homogeneous activation and post-irradiation etching that are of particular value when the metallurgical properties of high purity materials are being investigated. The more general applicability of gamma activation to the analysis of traces, minor and major components in a wide range of materials is now appreciated, and this review is concerned with the whole field of activation analysis with gamma photons.TYPES OF GAMMA ACTIVATION Most of the useful analytical applications of gamma activation require photons with energies in excess of 7 MeV, but, for the sake of completeness, the reactions observed with less energetic photons will be discussed briefly. ( y , ~ ) ACTIVATION : MOSSBAUER EFFECT- This is a very limited form of photon activation in which y-photons of a few hundred keV are emitted by metastable states and can be re-absorbed by other nuclei of the same isotope. There are no analytical methods based on this effect at the present time. For details see Summaries in advertisement pages. 601 Nature of the product nuclides * Reprints of this paper will be available shortly.602 BAKER : GAMMA-ACTIVATION ANALYSIS [AnaZyst, Vol. 92 ( 7 , ~ ’ ) ACTIVATION- Many stable nuclides, particularly of the heavier elements, can exist in a short-lived metastable state that can be produced by the interaction of a y-ray of sufficient energy, for example a reaction of the type 77Se (y,y’) 77mSe, the excess of energy being emitted as a photon 7‘.The cross-sections for such reactions are of the order of microbarns. Kaminishi,l Lukens, Otvos and Wagner2 tabulate isomers from selenium to mercury that can be made in this way with photons from 3 and 6-MeV linear accelerators. The degree of activation increases sharply with increasing energy of the incident gamma rays but, even under the most favourable conditions of 6-MeV energy and 1-mA beam current, the limit of detection is calculated to be one microgram. In the least favourable cases the limit may be several milligrams.(y,n) ACTIVATION IN WHICH NEUTRONS ARE DETECTED- There are two (y,n) reactions with threshold energies lower than the energies available from nuclear disintegrations (Goldstein3). The reaction 9Be (y,n) 8Be (threshold energy E,, 1.67 MeV) has been used to assay for beryllium with Z4Sb (half-life 60 days) and ssY (half-life 87 days) as sources of y-photons. Beryllium concentrations of only 60 p.p.m. can be determined with quite modest y-sources (300 mC) . Deuterium has been determined by the reaction 2D (y,n) lH (Eth 2.23 MeV) with thorium-228 (half-life 1.9 years) and sodium-24 (half-life 15 hours) as y-sources down to 0.02 per cent. v/v of deuterium. (y,n) REACTIONS IN WHICH THE PRODUCT NUCLEUS IS MEASURED- Most of the useful analytical work falls into this category.Basically it is only necessary to irradiate the sample with y-photons of sufficient energy (10 to 25MeV), measure the activity induced in the sample and compare this with a standard in a way that is entirely analogous to thermal neutron activation. (yJ2n) (7,~) (y,pn) J (7,an) REACTIONS- These reactions occur when more energetic photons are used than are necessary for the (y,ii) reactions. They are best regarded as unwanted interferences and avoided if possible. NUCLEAR CONSIDERATIONS FOR (y,n) REACTIONS Most nuclei will react with y-photons of a suitable energy to undergo a (7,n) reaction. The relation between the cross-section for the reaction and energy of the photon will be of the form shown in Fig. 1.Fig. 1. Energy dependence of cross-section for a ( y , n) reaction E, is the threshold energy and E,,,, is the energy of the so-called “giant resonance.” The values of E, and Em,,. vary from nuclide to nuclide, but in general Eth is 10 to 20 MeV for light nuclei, falling steadily to 7 to 10 MeV for heavy nuclei. Similarly, the value ofOctober, 19671 BAKER : GAMMA-ACTIVATION ANALYSIS 603 Ern=. falls with increasing atomic number from 20 to 25 MeV down to about 15 MeV. The cross-section at the peak varies from a few millibarns to a few hundred millibarns ( K ~ c h , ~ Sulin5 and Strauch6). SOURCES OF HIGH ENERGY PHOTONS- An intense source of y-photons is required having an energy in excess of the threshold and preferably in excess of Em,,. No known nuclear disintegration provides such energetic photons, but they can be produced by stopping particles of the required energy (“brems- strahlung” production).The most efficient conversion of energy into bremsstrahlung radiation occurs when a light particle is stopped in a heavy absorber, and this is realised physically by causing a beam of energetic electrons to impinge on a target or “radiator” of platinum, gold or tungsten. Most of the energy of the electron beam is dissipated as heat in the radiator, which may need to be water-cooled, but some energy is radiated in the forward direction as a beam of gamma photons. The energy distribution of photons in the brems- strahlung from a mono-energetic electron beam of energy E will be as shown in Fig. 2. The degree of activation induced by such a photon flux will be proportional to the sum of the products of flux and cross-section in each energy element in the region of overlap.The value of this function increases rapidly with energy, approximately as a 20th power near the threshold and as an 8th power in the region of maximum cross-section. Fig. 2. Energy distribution in a bremsstrahlung ELECTRON ACCELERATORS The two major classes are cyclic and linear. CYCLIC NACHINES- The betatron is a cyclic accelerator in which a pulsed beam of electrons is injected into an evacuated torus located between the poles of a powerful electro-magnet. The magnetic field is initially small but is increased over a period of about 5 milliseconds. This serves both to accelerate the electrons and to constrain them in a circular orbit in the torus.About loG revolutions are made, during which time the electrons are accelerated to perhaps 40 MeV, but the peak energy can be varied at will by varying the current flowing in the magnet. At the peak of each cycle of the power supply, the orbiting electrons are deflected on to a small internal platinum electrode and the bremsstrahlung is generated from that point. The mean current incident on the target is of the order of 10-8 amperes. The bremsstrahlung is radiated tangentially from the torus and the samples are placed at a convenient position in the beam. A considerable increase in intensity over a restricted area has been achieved by Chepel’, Viting and Chapyzhnikov,7 who describe the use of a hollow probe in the torus itself, so that the samples can be introduced at a point immediately in contact with the platinum electrode.A 100-fold increase in flux was observed compared to the best flux available outside the torus, but the sample size was restricted to about 1 mm square.604 BAKER GAMMA-ACTIVATION ANALYSIS [Analyst, Vol. 92 There is one report of the use of a much more powerful cyclic accelerator for gamma- activation analysis. Voigt and Abu-Sumra* used a synchrotron at 70 MeV, but no details are given of the machine or radiator. LINEAR MACHINES- The linear electron accelerator is by far the most useful source of high energy photons for gamma-activation analysis. Electrons are injected into an evacuated tube that consists of several sections of shaped waveguide, each one driven by a klystron amplifier.The electrons are accelerated on the crest of a travelling wave and reach energies of as high as 40 MeV, depending on the size of the machine, and mean electron currents of up to 50 pA have been used, although the heat dissipation can present some difficulties with such currents. The electron beam usually leaves the evacuated volume through a thin steel window and impinges on an air or water-cooled radiator of a heavy metal, such as platinum; the sample can be placed immediately behind the radiator. NATURE OF THE PRODUCT NUCLIDES Most of the products of (y,n) reactions are short-lived and even if a long-lived isotope is formed the cost of operation of the accelerator would normally make it uneconomical to irradiate for a period long enough to obtain a high sensitivity.Furthermore, neutron-deficient isotopes usually decay by positron emission, frequently without any accompanying gamma rays except for the 0-51-MeV annihilation radiation. CHARACTERISATION OF THE RADIATIONS- The most exploited techniques in activation analysis are chemical characterisation through radiochemical separations and gamma spectrometry. These two methods are limited in application to the products of gamma-activation analysis because of the short half-lives and paucity of gamma rays, respectively, but they have been used and supplemented with others as follows. Decay-carve resolzdion-This can range from the simple expedient of allowing the short- lived activities to decay and counting a long-lived tail, to refined computer analysis of up to six components.Energy dependence of cross-sections-By varying the energy of the bremsstrahlung the relative activation of different elements can be altered; in particular, the activation falls to zero for incident energies below the threshold energy. and y-spectrometry-/I-Spectrometers have been used to resolve the energy spectrum of positrons into its components, and several of the product nuclides, particularly those having high atomic weights, have significant contributions from y-decay modes so that y-spectrometry can be used. ChemicaZ se9arations-The separation of activities of individual elements by chemical separation is a very powerful technique. In many cases this has required that a procedure be developed or modified to make it rapid in relation to the half-life of the isotope required.STANDARDS- An essential feature of all activation analysis is the provision of an accurate standard. The gamma fluxes that are available are restricted in area and are heterogeneous, and great care must be taken in the irradiation of standards. Various workers have monitored the flux and energy by instrumental means; measured the degree of activation of sample and standard in separate irradiations by using a flux monitor; or compared sample and standard in the same radiation. DETERMINATION OF THE LIGHT ELEMENTS, OXYGEN, NITROGEN, CARBON AND FLUORINE The pioneer work was carried out in 1954 by Basile, Hur6, L6vi3que and Schuh1,g who determined oxygen in stearol, stearic acid and benzoic acid with a betatron to induce the reaction l60 (y,n) 150.Samples of 2.5 g containing several hundred milligrams of oxygen were irradiated for 4 minutes in a position 20 cm from the radiator. The energy was 18.6 MeV to avoid activation of the carbon, and a quartz monitor was included with each irradiation. By using this model system with the energy regulated so that only one element was activated,October, 19671 BAKER GAMMA-ACTIVATION ANALYSIS 605 it was shown that the initial count-rate of the induced activity was proportional to the oxygen content. The feasibility of intact determination of oxygen in beryllium oxide and alumina was also demonstrated. Beryllium is not activated, and aluminium-26 (half-life 6.5 seconds) can be allowed to decay before counting commences.THE USE OF DECAY CURVES AND ENERGY THRESHOLDS- A great deal of the work reported in the literature relies on the resolution of decay curves into components of half-lives 2, 10 and 20 minutes for the determination of oxygen, nitrogen and carbon, respectively. In addition, Englemann and others have used the energy dependence of the degree of activation with great effect. OXYGEN- AlbertlO and Albert, Engelmann, May and Petitll have reported determinations of oxygen, together with nitrogen and carbon, in beryllium, with the linear electron accelerator at Saclay as a source of photons. The accelerator was operated at 28 to 30 MeV and 50 PA, and the radiator consisted of two platinum discs that were cooled by water passing between them. Oxygen was determined by the reaction 1 6 0 (y,n) 150, the 2-minute component of.the decay curve being attributed to this nuclide. A thorough discussion of the determination of oxygen in this way is contained in three similar papers by Engelmann et aZ.,12713J4 who used the same linear electron accelerator as Albert for the source of photons. A pneumatic transfer system carried specimens from the irradiating position to the counters in 25 seconds. The counters were automatically operated and consisted of two opposed sodium iodide crystals equipped with coincidence circuits so that only 0.51 1-MeV annihilation radiation was detected. Methods are described for the non-destructive charac- isation of the positron activities by using decay-curve resolution and the energy dependence of the activation cross-section.After an irradiation of 5 minutes the component of the decay curve having a half-life of about 2 minutes was attributed by Engelmann to oxygen-15 (half-life 2.1 minutes) from the reaction l60 (y,n) 150 and phosphorus-30 (half-life 2.6 minutes) from the reactions 31P (y,n) 30P and 32S (y,pn) 30P. However, the thresholds of the three reactions are different and by carrying out three radiations, at 28, 18 (below the threshold for sulphur) and 14 MeV when the phosphorus alone is activated, it was possible to identify the individual contributors. Alternatively, the irradiation could be carried out in nominal energies of 27, 20 and 16 MeV, which were not strictly related to the threshold energies but nevertheless gave large differences in the ratio of specific activities for the three reactions.Specimens of “Mylar” film were included with each sample in order to monitor the current and energy of the machine during the irradiations. NITROGEN- AlbertlO suggested that nitrogen could be determined from the reaction 14N (y,n) 13N, and attributed the 10-minute component of the decay curve to this product. The resolution of the 10-minute component was uncertain because of the presence of 2-minute oxygen-15 and 20-minute carbon-1 1, and AlbertlO recommended that the nitrogen should be separated chemically by the Kjeldahl method, for which yields of 70 per cent. were claimed. Alternatively, the irradiation was carried out at 15 MeV, i.e., below the threshold of the interfering reactions. This method carried the penalty of a 500-fold loss in sensitivity.In America, Beard, Johnson and Bradshaw15 have applied gamma-activation analysis to the problem of quality control in ultra-pure beryllium manufacture. The individual activities were recognised by resolving the positron decay curve into three components attributed to oxygen, a mixture of the triad nitrogen, copper and iron, and carbon. Standards were irradiated separately and the beam currents monitored. The linear electron accelerator at Stanford was used and sensitivities of less than 1 pg were claimed. Enge1mannl2 also attributed the 10-minute component to a mixture of nitrogen-13, copper-62 and iron-53, but pointed out that the threshold energies and relative activation at different energies were not favourable for the application of the same methods that he used to discriminate between oxygen-15 and phosphorus-30.606 BAKER GAMMA-ACTIVATION ANALYSIS [Analyst, Vol.92 Engelmann12 therefore determined the copper and iron by some other method and made suitable corrections. The results obtained agreed well with similar irradiations in which pure nitrogen-13 was separated by the Kjeldahl method. CARBON AND-FLUORINE- Carbon is determined by the reaction I2C (y,n) llC, and all workers attribute the 20-minute component of the decay curve to this product. AlbertlO and Albert, Engelmann, May and Petitll pointed out that it is necessary to separate the active carbon in a chemically pure form for samples of aluminium, iron and zirconium in order to eliminate the interfering activity that is induced in the matrix.This was achieved by combusion of the sample in oxygen, then trapping out and counting the carbon dioxide formed. Engelmann12 assessed fluorine by measurements on the 112-minute component due to the reaction 19F (y,n) lSF. Fluorine is subject to interference from the reaction 23Na (y,ncc) 18F, but this can be reduced to negligible proportions by suitably reducing the energy of the irradiation. For materials of the highest purity, when the activities do not interfere with each other, sensitivities of the order of 0.01 to 0.1 pg have been achieved. THE USE OF SPECTROMETRY AND OTHER INSTRUMENTAL METHODS By using the principle of the “pocket” or “well” in the betatron torus to produce a high flux over a small area,’ oxygen has been determined in rubbers by P-spectrometry.l6 This was done by irradiating at 18 MeV, below the threshold energy for carbon activation, but the sensitivity was poor.By operating at 25 MeV a large increase in specific activity was obtained but the carbon activity interfered. To overcome this the positron emission from the thin samples was counted with a plastic scintillator, and pulse height analysis was used to discriminate against the counts from the decay of llC pf 0.99 MeV but retain those from 150 p+ 1.68 MeV. The counter was very inefficient but the over-all sensitivity of the method was increased by a factor of 20 compared with the 18-MeV irradiation. Separate irradiations of boron trioxide were used as standards and solutions of dibutyl phthalate were used to calibrate the method, Carbon has been determined in a piece of steel of archaeological interest by gamma activation and intact gamma spectrometry.A 70-MeV synchrotron was used to provide the bremsstrahlung. The analysis was complicated by the reaction 54Fe (y,pn) 52mMn (half-life 21.3 minutes) in the iron matrix which interfered with the measurement of carbon-11 (half-life 20-3 minutes). Fortunately the manganese-52m has a gamma ray at 1-435 MeV and by measurement of the gamma spectrum induced in pure iron it was possible to correct for this interference (Voigt and Abu Sumras). DESTRUCTIVE ANALYSIS BY RAPID CHEMICAL SEPARATIONS Baker, Pratchett and Williams1’ have used the linear electron accelerator at Hanvell as a source of y-photons to develop gamma-activation analysis methods for carbon and oxygen in materials that cannot be counted without separation of the required nuclide.RAPID SEPARATIONS OF RADIOCHEMICALLY PURE OXYGEN- The determination of oxygen in matrices, which themselves contribute a large short-lived activity ( e g . , iron and molybdenum), requires a chemical separation that is specific, of good chemical yield and very rapid. Baker has achieved this by using an inert-gas fusion in an iron bath at 2000” C, and subsequent purification of the oxygen activity by passing the gas stream through two copper oxide furnaces, one at 600” C and the other at 1100” C. The first furnace has been shown to oxidise the carbon monoxide without loss of radio-oxygen while the second furnace retained more than 99 per cent. of oxygen-15 without contamination by carbon-11 or other activities.The chemical separation was carried out in 4 minutes and the limit of detection was better than 1 pg. Engelmann has recently suggested a similar separation procedure in which gases evolved by fusion in a carbon arc furnace were absorbed on soda lime within the sodium iodide crystal assembly.October, 19671 BAKER : GAMMA-ACTIVATION ANALYSIS 607 CHEMICAL SEPARATION OF CARBON ACTIVITY- Baker has also determined carbon in low alloy and stainless steels by combustion in oxygen and counting the carbon dioxide evolved. Samples were irradiated, together with two standards of graphite of similar area to the sample placed in front of and behind the sample. The mean specific activity of the standards was compared with the activity found in the sample.Excellent agreement with published carbon contents of standard steels was achieved over the range 1 per cent. to 10 p.p.m. and the limit of detection was better than 0.1 pg. The technique has been applied to ultra-pure molybdenum. Ceramic materials such as magnesium oxide and calcium oxide have also been analysed by Baker for carbon by using a flux of borax at 1250" C in a stream of oxygen to dissolve the sample and remove the carbon as carbon dioxide. The high oxygen content of these materials resulted in a nuclear interference, 1 6 0 (y,na) W , that could only be eliminated by irradiating with an energy below the threshold for this reaction, 25.8 MeV. In spite of this the limit of detection was still better than 1 pg.The separation of individual radio-nuclides has the advantage that sources can be counted in a gross gamma well-crystal counter of high efficiency, rather than y - y coincidence counters used by many workers which are necessary to eliminate counts from gamma-emitters in the sample. DETERMINATION OF HEAVIER ELEMENTS, ATOMIC NUMBER EQUAL TO, OR GREATER THAN, 10 Engelmann mentions the possibility of determining zirconium and other elements, and this field is covered fully by Schweikert and Albertls The elements silicon, titanium, chrom- ium, manganese, iron, cobalt, nickel, copper, zinc, arsenic, strontium, zirconium, niobium, molybdenum, silver, cadmium, antimony, hafnium, tantalum, tungsten, platinum, thallium, lead and bismuth were divided into two groups according to the degree of activation induced by 27-MeV gamma photons in a 2-hour irradiation.Methods were indicated for the deter- mination of those elements for which the method is sensitive by decay-curve resolution, energy dependence of activation and gamma spectrometry for non-destructive characterisation of the induced activities. Specific examples were given of analyses that were superior to neutron-activation analysis for the following reasons: the nature of the photon-activation product is more suitable by virtue of sensitivity, gamma spectrum or half-life ; interference is eliminated; and shielding is eliminated in a high cross-section matrix, such as boron, tungsten, manganese, cadmium and hafnium. These advantages occur repeatedly in the following examples.Zirconium can be determined in hafnium with a sensitivity of 1 p.p.m. by irradiating for a short time and immediately counting the 0-59-MeV gamma ray from zirconium-89m (half-life 4.4 minutes) with a y spectrometer.ls Japanese workers report the determination of as little as 10 p.p.m. of zirconium in 100-mg samples of hafnium by y-irradiation at 20 MeV for 1 hour. After 4 days' decay the ratio of the peak heights in the gamma spectrum of zirconium-89 at 0.915 MeV and hafnium-175 at 0.34 MeV was determined. The advantage of using gamma activation rather than neutron activation is that the matrix does not have an enormous cross-section for the activating radiation and does not shield the impurity.lg Titanium has been determined in iron at the p.p.m. level by using the reaction 46Ti (y,n) 45Ti (half-life 3-1 hours).The titanium activity was separated chemically by extraction of the 8-hydroxyquinolinate in the presence of EDTA. The 3-hour half-life titanium-45 was more convenient in this respect than 5.8-minute titanium-51 from neutron activation.ls An interesting application concerns the determination of nickel in copper, which is limited in neutron activation by the reaction 65Cu (n,p) 65Ni, which generates nickel activity from the matrix. Gamma activation produced the nickel isotope 58Ni (y,n) 57Ni (half-life 36 hours), which is not subject to a nuclear interference. A dimethylglyoxime precipitation was used to separate the nickel activity.ls Chepel' and Skemarov20 have measured the chlorine as well as the fluorine content in halogen-containing hydrocarbon polymers.Specimens of PVC, Teflon and mixed polymers, 6 mm in diameter and 1.7 mm thick, were irradiated at 18 MeV with a betatron to produce the reactions 19F (y,n) 18F and 35Cl (y,n) 34Cl without production of interference from carbon-11.608 BAKER : GAMMA-ACTIVATION ANALYSIS [Analyst, Vol. 92 The 0.51-MeV annihilation gamma rays were counted and the decay curve resolved into its two components of half-lives 32 minutes and 1.7 hours. Pellets of lithium fluoride and sodium chloride were activated in separate irradiations to serve as standards; thin silver plates either side of each sample and standard served as beam monitors (silver-106, half-life 24 minutes). More recent work from Russia by Sulin5 and Berezin, Vitozhents, Sulin and Shornikov2l includes a comprehensive survey of the possible use of gamma-activation analysis for the analysis of rocks and minerals. The scope of the papers is limited to intact analysis by using varied energy of activating rays, decay-curve resolution, and gamma spectrometry to characterise the individual components.A useful compilation of nuclear data for most of the elements of the periodic table was included, together with estimates of the sensitivities likely to be achieved (100 to 1000 p.p.m.) in favourable cases with a betatron. Examples were given of successful analyses of model systems for the light elements, oxygen, nitrogen, carbon, phosphorus and sulphur and also for zinc, copper, zirconium by using synthetic mixtures with sample weights of up to 500 g.Mulvey, Cardarelli and MeyerZ2 used the reaction 12'1 (yln) 1261 (half-life 13.3 days) to determine iodine in biological material. Thermal neutron activation of biological materials usually results in high levels of activity due to sodium-24 and chlorine-38 and post-irradiation separation is necessary to isolate 25-minute iodine-128. By using gamma irradiation the analysis was performed directly on 0.9 per cent. sodium chloride media. After 3 days' decay period the only significant activity remaining was iodine-126 which could be measured by y-spectrometry. A limit of detection of 1 pg of iodine was claimed for a 90-minute irradiation at 22 MeV and 250pA. CONCLUSIONS Gamma-activation analysis has become established as a powerful technique for the determination of trace levels of carbon, nitrogen and oxygen in pure materials.It has also been applied to the determination of many other elements at trace levels. In general, the technique should be considered as a complement to neutron activation to be used when the nuclear properties of the impurity or the matrix favour the use of this method, for example, sensitivity, half-life, decay mode, interferences or shielding effect of the matrix. Except in the case of model systems, where the identification of the induced activity is un- equivocal, the characterisation, separation and measurement of activities is likely to be more difficult than with neutron-activation products. Two lines of research would appear to be profitable in this field. NOK-DESTRUCTIVE ANALYSIS- By irradiating at various energies for various times and following the decay of both annihilation radiation and the y-spectrum many results can be collected from which the composition of the sample can be deduced.Modern computers offer the best chance of extracting the maximum amount of information from these results. DESTRUCTIVE ANALYSIS- isotopes. separations could prove very useful in this field. Rapid dissolution procedures and chemical separations are required for the short-lived High temperature chemistry in various molten salt or metal baths, and gas-phase Appendix The sensitivity of any activation analysis method is very much dependent on the conditions of irradiation and counting. In Table I sensitivities have been calculated for ideal systems free from interference, for the following set of conditions which are realistic when using a large linear accelerator.Accelerator parameters : 5-pA, 30-MeV electron beam. 1 -mm tungsten radiator. Sample position 10 cm behind radiator. One half-life or 1 hour, whichever is the shorter. INTRINSIC SENSITIVITY OF GAMMA-ACTIVATION ANALYSIS Irradiation time :October, 19671 BAKER GAMMA-ACTIVATION ANALYSIS 609 Counting time : Counter: Commence immediately after irradiation, count for 1 half-life or 1 hour, whichever is the shorter. Well-type sodium iodide crystal with discriminator set to 100 keV. Limit of detection is defined as that weight of an element that will give a count equal to twice the standard deviation on the background count, which is 120 counts per minute.If a chemical separation is required before counting, the sensitivity must be halved for each half-life delay. Half-lives of less than 10 minutes are italicised. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Nuclide ‘2C 14N 1 6 0 lgF 20Ne 23Na 24Mn 2iA1 28Si 31P 32s 39K 40Ca 46Ti 50Cr “Mn 54Fe 58Ni 63Cu 75As igBr 97Rb 86Sr 99Y gOZr 93Nb 9 2 M ~ lo3Rh 107Ag 1131n 12%b 181Ta 197Au 35c1 59c0 642n 1 2 7 1 Half -life 20.3 minutes 10.05 minutes 2.1 minutes 1.8 7 hours 18 seconds 2.6 years 12 seconds 6.5 seconds 4.2 seconds 2-6 minutes 2-6 seconds 32.4 minutes 7.7 miflutes 0.9 second 3.1 hours 42 minutes 278 days 8.4 minutes 9 hours 37 hours 9-8 minutes 38 minutes 18 days 18.7 days 70 minutes 6.4 minutes 105 days 4.4 minutes 10.2 days 15.5 minutes 24 minutes 14-5 minutes 3.5 minutes 13 days 8.1 hours 9.5 hours 210 days TABLE I SENSITIV~TIES Limit of detection, pg 0.01 0.01 0.04 0.05 0.3 0.3 0.2 0.2 0.05 0.3 0.08 2.0 1.0 0.1 0.1 0.2 0.02 0.2 0.001 0.01 2.0 0.1 3.0 0.03 7 0.01 0.02 0.2 0.02 2.0 0.02 0.4 0.01 0-02 200 20 10 Limit reported J 0.01 pg in Be 0.1 pg chemical separation 1 p.p.m.in Be if pure J 1 p.p.m. in Be if pure -[ 1 pg chemical separation 0.1 pg in Be - - 1 p.p.m. chemical separation - - 1 p.p.m. chemical separation - - 0.1 pg in Hf - - 0.1 pg in Bi - - 1 pg in biological material - REFERENCES Kaminishi, K., J a p . J . A p p l . Phys., 1963, 2, 399. Lukens, H. R., jun., Otvos, J. W., and Wagner, C. D., I n t . J . Appl. Radiat. Isotopes, 1961, 11, 30. Goldstein, G., Analyt. Cham., 1963, 35, 1620. Koch, R. C., “Activation Analysis Handbook,” Academic Press, New York and London, 1960.Sulin, V. V., “Soviet Advances in Nuclear Geophysics,” Consultants Bureau, New York, 1965, Strauch, K., A . Rev. Nucl. Sci., 1953, 2, 105. Chepel’, L. V., Viting, B. I., and Chapyzhnikov, B. A, Instrums Exp. Tech., Pittsburgh, 1962,2,240. Voigt, A. F., Abu-Sumra, A., in “Proceedings of the International Conference on Modern Trends Basile, R., Hur6, J., L6v&que, P., and Schuhl, A., C.R. Hebd. Seanc. Acad. Sci., Paris, 1954,239,422. Albert, Ph., in “Proceedings of the International Conference on Modern Trends in Activation p. 129. in Activation Analysis,” College Station, Texas A and M University, 1965. Analysis,” College Station, Texas A and M University, 1961.610 BAKER 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. Albert, Ph., Engelmann, Ch., May, S., and Petit, J., C.R. Hebd. SBanc.Acad. Sci., Paris, 1962,254, Engelmann, Ch., Commissariat h 1’Energie A tomique, CEA-R-2559, Centre d’Etudes Nucldaires, -, in “Proceedings of the Symposium on Radiochemical Methods of Analysis, Salzburg, 1964,” Engelmann, Ch., Gosset, J., and Loeuillet, M., Bull. Soc. Chim. Fr., 1965, No. 2, 544. Beard, D., Johnson, R. H., and Hradshaw, W. G., Nucleonics, 1959, 17, No. 7, 90. Chepel’, L. V., Chapyzhnikov, B. A., andViting, B. I., J . Analyt. Chem. USSR, 1963, 18, No. 7, 749. Baker, C. A., Pratchett, A. G., and Williams, D. R., U.K. Atomic Energy Research Establishwent Schweikert, E., and Albert, Ph., in “Proceedings of the Symposium on Radiochemical Methods Oka, Y., and Kato, T., J . Chew. SOC. Japan, 1965, 86, 835. Chepel’, L. V., and Skemarov, F. V., Dokl. Akad. Nauk. SSR., 1964, 158, 976. Berezin, A. K., Vitozhents, G. Ch., Sulin, V. V., and Shornikov, S. I., in “Proceedings of the Symposium on Radiochemical Methods of Analysis, Salzburg, 1964,” I.A.E.A., Vienna, 1964, p. 361. Mulvey, P. F., jun., Cardarelli, J. A., and Meyer, R. A., “Symposium of Radioactive Sample Measurement Technique in Medicine and Biology,” I.A.E.A., Vienna, 1965. Received M a y 15th, 1967 119. Saclay, July, 1964. I.A.E.A., Vienna, 1964, p. 361. Report, AERE-R5363, H.M. Stationery Office, London, 1967. of Analysis, Salzburg, 1964,” I.A.E.A., Vienna, 1964, p. 323.

ISSN:0003-2654    

DOI:10.1039/AN9679200601

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, October, 1967, Vol. 92, pp. 611-613 61 1 Improved Determination of Microgram Amounts of Lead in Food with a Radioactive Tracer BY DONALD C. BOGEN AND MICHAEL T. KLEINMAN (Health and Safety Labovatovy, U.S. Atomic Enevgy Commission, New York, New York 10014) Microgram amounts of lead in food are determined by use of a radio- active tracer. The radioactive tracer, lead-2 12, is radiochemically purified from thorium nitrate and is used for chemical yield determination. Food sample dissolution involves freeze drying and wet ashing to minimise con- tamination and eliminate volatilisation. The stable lead is determined spectrophotometrically as the dithizone complex at 510 mp. The chemical recovery for analysis is 94 per cent., and the precision of analysis is better than 10 per cent.THE analysis of low-level lead in environmental samples by existing techniques is sub ect to uncertainty because of contamination and incomplete recovery. Although reagent ( on- tamination can be reduced in some matrices by dry ashing, Gorsuch observed losses of Lead in food dry ashed at ,500" C, but found that by wet ashing in a nitric acid system, excli sive of sulphuric acid, average recoveries of 100 per cent. were obtained.1 A technique has x e n developed that decreases contamination from reagents and corrects for losses in reco tery. This involves speeding up the diqestion procedure by freeze-drying the samples befor: wet ashing and adding a radioactive tracer. This tracer is prepared by a modification of Petrow's procedure,2 with natural thorium nitrate as the source of the radioactivity.The lead is concentrated by solvent extraction of the dihydrogen tetra-iodolead(I1) complex from 5 per cent. hydrochloric acid into 3-methyl-2-butanone.3 The organic phase is wet ashed and lead is extracted as the dithizone complex into chloroform.4 The amount of lead is determined by spectrophotometric measurement of the absorption of this complex at 510mp and the chemical recovery by gamma-scintillation counting of the lead-212 tracer. EXPERIMENTAL APPARATUS-- Spectrophotometric measurements were made with a Beckman Model DU spectrophotometer and radioactivity measurements with a Baird Atomic Model 810 scintillation counter, fitted with a Raird Atomic Model 132 scaler - timer unit. REAGEKTS- solzdion-Prepare a 30 per cent. v/v solution in toluene.A 1-kg sample is freeze-dried, wet ashed and lead-212 tracer is added. Freeze-drying was carried out with a Lab-Line Freeze Dryer. Methyltri-octanoylammonizlm chloride (Aliqzlat-336, General Mills Inc., Kankakee, Illinois) 3-Alefh.d-2- bzdanon e. Diphen-ylthiocarba~one, 0.005 per ce& 7 ~ / v in chloro f o m . Basic h f e r solzction-Mix 30 ml of 10 per cent. w-/v sodium cyanide, 75 ml of 2 per cent. w/v sodium sulphite, 340 ml of concentrated ammonium hydroxide and 605 ml of de-ionised water. Lead- 212 tracer- Prepare by dissolving 10 g of analytical-reagent grade thorium nitrate in 100 ml of de-ionised water and add a few drops of concentrated nitric acid. Evaporate the solution nearly to dryness, dilute to 55 ml and add 45 ml of 6-6 M hydrobromic acid.Lead- 212 was separated from thorium by extraction into 50 ml of 30 per cent. methyltri-octanoyl- ammonium chloride that had been equilibrated with two 50-ml volumes of 1.5 M hydrobromic acid. The organic phase was washed four times with 25-ml portions of 0.1 M hydro- bromic acid and the lead stripped from the organic phase with 25 ml of concentrated hydro- chloric acid. The aqueous phase was washed with 50ml of toluene to remove traces of methyltri-octanoylammonium chloride. The lead-containing solution was evaporated to exactly 25 ml and the radiochemical purity checked by gamma-scintillation counting. As lead-212 has a 10.6-hour half-life, it was necessary to complete the wet ashing of the sample before adding the tracer.612 BOGEN AND KLEINMAN : IMPROVED DETERMINATION OF MICROGRAM [Ana@St, VOl.92 PROCEDURE- One kilogram of the sample was dehydrated by freeze-drying for 24 to 72 hours. Subse- quently, the 30 to 50 per cent. weight-reduced sample was wet ashed with 1 to 2 litres of concentrated nitric acid in new borosilicate glassware. A cover glass placed on top of the ashing beaker provided good reflux action in destroying the organic material. After evaporat- ing the solution until nearly dry, the deposited salts were treated with 100 ml of concentrated nitric acid and 25 ml of 30 per cent. hydrogen peroxide to ensure complete destruction of the organic material. The residue was then dissolved in 100 ml of 6 M nitric acid, and the remaining salts were removed by gravity filtration on ashless filter-paper. The filter-paper containing the salts was placed in a Teflon beaker and the silica expelled by distillation with hydrofluoric and nitric acids, during which treatment the filter-paper was destroyed.A true solution was thus obtained, which was combined with the filtered solution. An amount of lead-212; tracer, sufficient to yield 20,000 counts per minute, was added to the wet-ashed sample and the solution evaporated almost to dryness. The residue was then dissolved in 500ml of 5 per cent. hydrochloric acid and 50ml of 10 per cent. potassium thiocvanate solution added. Iron, tin and other elements that might interfere in subsequent steps were removed by extraction with two 100-ml portions of 3-methyl-2-butanone. About 10 ml of 10 M sodium iodide solution were added and the lead iodide complex extracted with 200 ml of 3-methyl-2-butanone.The organic phase was evaporated to dryness and the residue wet ashed with small amounts of nitric and perchloric acids. After drying, the residue was dissolved in 4 drops of hydrochloric acid and 75 ml of basic buffer solution were added. Lead was extracted with 10-ml portions of dithizone solution until the organic phase remained blue - green for two successive extractions and was then extracted into potassium acid phthalate buffer (pH 3-4), which separated it from any bismuth.j About 75 ml of basic buffer solution were added and the lead again extracted as its dithizone complex. The solution was then transferred into a 50-ml calibrated flask, diluted to volume with chloroform and the lead determined spectrophotometrically at 510 mp.The solution was transferred into a suitable counting vial and counted in a well-type gamma-scintillation counter, after allowing 8 hours to elapse for the ingrowth of lead-212 daughters. RESULTS AND DISCUSSION Table I shows an average recovery of 94 per cent. of the lead that was added to “spiked” Table I1 indicates that the method is repro- samples after correcting for contamination. ducible to within 10 per cent. TABLE I ANALYSIS OF “SPIKED” SAMPLES Milk sample No.* 1 2 3 4 5 6 Lead added, Pg 0 100 200 300 400 500 Lead recovered, PQ 0 88 184 292 372 504 Average lead recovery 94 per cent. * Each of the six milk samples weighed 1 kg. TABLE I1 MEASUREMENT OF SPLIT SAMPLES First analysis, Second analysis, Average percentage Sample Pg Per kQ PQ Per kg deviation from mean Milk A .. .. . . 216 184 200 f 8 Fresh vegetables . . . . 298 244 271 f 10 Canned vegetables . . . . 674 708 691 2 Milk B .. . . . . 186 200 193 & 4October, 19671 613 It is necessary to determine the lead in the nitric acid used in wet ashing as this reagent is a major source of contamination. vITe carried out four determinations on the acid and found a lead concentration of 2 to 5 pg per litre. The procedure was, therefore, designed so that not more than 2 litres of nitric acid were used. The reagent contamination level of the complete technique, including sample preparation, was evaluated by analysis of several blanks and found to total 20 & 5 pg. Kehoe6 has stated a range of lead intake from the diet of 0.10 to 1.0 mg per day.Assuming a total intake of 2.0 kg of food7 the lead concentration in the diet would be in the range of 50 to 500pg per kg. It is apparent, therefore, that contamination presents a serious problem unless large sample aliquots of about 1 kg are used. AMOUNTS OF LEAD I N FOOD WITH A RADIOACTIVE TRACER I I l I l l l I i I I I 1 1 10 20 30 40 50 60 70 80 90 100 110 I20 130 140 I Time, hours 0 Fig. 1. Observed decay curve of lead-212 tracer (10-5-hours half-life) The procedure described for the preparation of the tracer provides radiochemically pure Jead-212. This was verified by gamma counting of prepared solutions of lead-212 over a period of several days. The half-life of lead-212 was determined to be 10.5 hours, as shown in Fig. 1, and this is in good agreement with the established value of 10.6 hours.* As the half-life of lead-212 is 10-6 hours and the sample preparation time is 3 to 6 days, the tracer must be introduced into the solution prepared after wet ashing. 1. 2. 3. 4. 5. 6. 7. 8. REFERENCES Gorsuch, T. T., Analyst, 1959, 84, 137. Petrow, H. G., and Cover, A., Analyt. Chew., 1965, 37, 1659. West, P. W., and Carlton, J . K., Analytica Chim. Acta, 1952, 6, 406. Morrison, G. H., and Freiser, H., “Solvent Extraction in Analytical Chemistry,” John Wiley Talvitie, N. A., and Garcia, W. I., Analyt. Chew., 1965, 37, 851. Kehoe, R. A., Avchs of Envir. Hlth, 1961, 2, 418. Rivera, J., and Harley, J. H., U.S. Atomic Energy Commission Refiovt HASL-147, Health and Strominger, D., Hollander, J. M., and Seaborg, G. T., Rev. Mod. Phys., 1958, 30 (2), Part 2, 788. Received Februavy 27th, 1967 and Sons Inc., New York; Chapman & Hall Ltd., London, 1957, p. 214. Safety Laboratory, New York, 1964.

ISSN:0003-2654    

DOI:10.1039/AN9679200611

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

614 Analyst, October, 1967, Vol. 92, $9. 614-617 The Determination of Aluminium with 8-Hydroxy quinoline Part I. Precipitation in Acetate-buffered Solution BY A. CLAASSEN AND L. BASTINGS (Philips Research Laboratories, N . V . Philips' Gloeilampenfubrieken, Eindhoven, The Netherlands) The conditions for an accurate gravimetric or titrimetric determination of aluminium with 8-hydroxyquinoline in acetate-buffered solution have been established. Precipitation is carried out at a pH higher than 5-2 with an excess of 200 to 300 mg of 8-hydroxyquinoline above the stoicheiometric amount in a final volume of 150 to 200 ml, and the precipitate washed with a neutral 0.02 per cent. solution of 8-hydroxyquinoline. In pure solutions a standard deviation of about 0.01 mg of aluminium has been obtained.ALTHOUGH precipitation of aluminium with 8-hydroxyquinoline has been described in many papers, the conditions given in many textbooks for the determination of aluminium by this method differ. In particular, the instructions as to the amount of reagent required for complete precipitation vary between a slight ~ x c ~ s s , ~ ~ ~ , ~ 15 to 25 per cent. excess415 and an excess of 10 ml of 2.5 per cent. reagent solution.6 As preliminary experiments had shown that the use of a 25 per cent. excess of reagent gave rise to slightly low results it was decided to investigate this method more closely. The procedure chosen was that in which a solution of 8-hydroxyquinoline in dilute acid was added to a slightly acidic aluminium solution, and precipitation effected by the slow addition of ammonium acetate.This method was selected as it furnished a rather coarse, crystalline precipitate, in contrast to that given by the procedure in which the order of addition of reagent and ammonium acetate was reversed, an effect also observed by Stumpf.7 Completeness of precipitation under various conditions of pH and excess of reagent was examined either by titration of the precipitate with bromate or by determination of the amount of unprecipitated aluminium. The latter was carried out by repeated extraction of the filtrate with chloroform and photometric determination of the aluminium as 8-hydroxy- quinolinate, as previously described.8 With a 25 per cent. excess of reagent results were low by 0.15 to 0-20mg in the range from 15 to 30 mg of aluminium, and 1 mg too low with 5 mg of aluminium in a final volume of 150 to 200ml.Complete precipitation was obtained with an excess of S to 15ml of 2.5 per cent. reagent solution; the filtrate then contained 20 to 25 pg of aluminium for amounts of aluminium between 5 and 30 mg. In the final procedure an excess of 10 ml has therefore been prescribed ; much larger amounts lead to crystallisation of 8-hydroxyquinoline during filtration. Miller and Chalmersg also prescribe an excess of 10 ml of 2.5 per cent. reagent for complete precipitation. We cannot, however, confirm their observation that errors of 0.2 per cent. occur when the excess is either 8 or 12 ml. The solubility of the precipitate as a function of pH is given in Table I. TABLE I UNPRECIPITATED ALUMINIUM AS A FUNCTION OF pH IN 150ml OF SOLUTION AT ABOUT 50" c, AFTER PRECIPITATION WITH A 10 ml EXCESS OF 2.5 PER CENT.S-HYDROXYQUINOLINE I N ACETATE-BUFFERED SOLUTION pH . . . . 3-8 4.0 4.3 4.5 4.7 4.8 4.9 5.0 5.3 to 7.0 Aluminium, pg .. >lo00 75 50 41 33 30 28 26 22 to 20 (The results given are the mean of closely agreeing values.) Precipitation can be considered complete at pH values of 5.2 and above. The amount of aluminium escaping precipitation in the pH range of 5.2 to 7 remained at the level of about 20 pg for precipitation at temperatures between 20" and 70" C.CLAASSEN AND BASTINGS 615 Tests performed under the above conditions for complete precipitation were still too low, when the titration with bromate was used, by amounts of 0.05 to 0.08 mg of aluminium, which were much greater than the amounts of unprecipitated aluminium in the filtrate.The cause of these low results was suspected to be the washing procedure of the precipitate. In our experiments the precipitate had been washed with 100 to 150ml of water at about 70" C. Washing with hot water,6 warm water4 and cold water2 have been recommended. We therefore determined the loss of aluminium when washing a precipitate of aluminium 8-hydroxyquinolinate with water at various temperatures. The results are shown in Table 11. TABLE I1 ALUMINIUM IN 100 ml OF FILTRATE WHEN ABOUT 200 mg OF ALUMINIUM 8-HYDROXYQUINOLINATE ARE WASHED WITH WATER AT VARIOUS TEMPERATURES Temperature, "C . . . . . 100 80 60 25 Aluminium, pg . . . . 27 15 7 6 The loss of aluminium is slight and does not account for the heavier loss mentioned above.As it was thought probable that hydrolysis occurred during washing, the amount of 8-hydroxy- quinoline in the wash liquid was also determined photometrically; the results are given in Table 111, together with the equivalent amount of aluminium. TABLE I11 8-HYDROXYQUINOLINE I N 100 ml OF FILTRATE WHEN ABOUT 200 mg OF ALUMINIUM 8-HYDROXYQUINOLINATE ARE WASHED WITH WATER AT VARIOUS TEMPERATURES Temperature, "C . . . . . . 100 80 60 25 8-Hydroxyquinoline, pg . . 1300 700 375 175 Equivalent aluminium, pg . . 81 44 24 11 Comparing Tables I1 and I11 we see that the aluminium 8-hydroxyquinolinate hydrolyses, leaving most of the aluminium as hydroxide on the filter. As the titrimetric determination is based on the determination of the 8-hydroxyquinoline content of the precipitate, it is clear from Table I11 that washing with hot or even warm water leads to considerable loss.This does not occur when washing with a small volume of cold water. We prefer to wash with a neutral 0.02 per cent. solution of 8-hydroxyquinoline. With this solution the amount of aluminium found in 100ml of filtrate is about 5 p g , irrespective of the temperature in the range 20" to 100" C. That the hydrolysis is completely prevented is shown by the accurate results obtained with the recommended procedure. Solutions of 8-hydroxyquinoline in dilute acetic acid or ethanol have been described for the reagent. The latter solvent is unsuitable as the solubility of the precipitate is increased considerably by the presence of ethanol.Thus, in solutions containing 5 per cent v/v of ethanol the solubility increases to about 50 pg and, in solutions containing 10 per cent. V/V, to about 85 pg of aluminium per 150 ml. Most textbooks prescribe a 2-5 or 5 per cent. solution of 8-hydroxyquinoline in 1 to 2 M acetic acid. One disadvantage of these solutions is that, after evaporation to dryness, the residue that has spilt on the wall of the beaker is difficult to dissolve in the test solution. A second disadvantage is that they contain much free acid and it is therefore difficult to reach a sufficiently high pH value. We therefore prefer to use a 2-5 per cent. solution of 8-hydroxy- quinoline in dilute hydrochloric acid. This solution contains practically no free acid (pH about 3) and the dried salt is readily soluble in water and in the test solution.The following procedure is recommended for the gravimetric or titrimetric determination. REAGENTS- S-HydroxyquinoZine, 2.5 per cent.-Dissolve 25 g of 8-hydroxyquinoline in 29 ml of 6 M hydrochloric acid, dilute with water, filter if necessary, and dilute to 1 litre. Ammonium acetate, 20 per cent. w/v-Dilute this solution with water (1 + 9) when it should have a pH of 6-5 to 7.0. If necessary, adjust the pH with ammonia solution or acetic acid. S-Hydroxyquinoline wash solzdion-Dilute 8 ml of 2.5 per cent. 8-hydroxyquinoline solu- tion with water to 500 ml, add 3 drops of bromocresol purple (1 g per litre in 20 per cent. METHOD616 CLAASSEN AND BASTINGS DETERMINATION OF [Analyst, Vol.92 ethanol) and ammonia solution (2 M) until the solution just begins to turn purple (pH about 6), and then dilute to 1 litre. SAMPLE SOLUTION- and 2 to 30mg of aluminium for the gravimetric procedure. The sample solution may contain 2 to 20 mg of aluminium for the titrimetric procedure RECOMMENDED PROCEDURE- Transfer the sample solution into a 400-ml beaker and dilute to between 100 and 125 ml. Add to the cold solution, dropwise, with stirring, 2 M ammonia solution until a slight pre- cipitate of aluminium hydroxide persists. Then add, dropwise, with stirring, 2 M hydrochloric acid until the solution is clear again, followed by an excess of 5 to 10 drops. If the precipitate of aluminium hydroxide cannot be seen clearly, as with small amounts of aluminium, adjust to a pH between 3 and 4, by using pH paper, and add 5 to 10 drops of 2 M hydrochloric acid.Add 0.70ml of 2.5 per cent. 8-hydroxyquinoline for each milligram of aluminium present $,!us an excess of 10 ml. Heat to about 70" C and add slowly, with stirring, 20 ml of ammonium acetate solution. Heat just to boiling, stand the solution on a steam-bath for 30 minutes, and then allow it to cool to about 50" C. (a) Gravimetric determinatiofi-Filter the solution through a weighed No. 4 sintered- glass or quartz crucible, decanting as much as possible. Transfer and wash the precipitate with small portions of warm (50" to 60" C) wash solution; do not use more than about 100 ml in total. Avoid sucking the precipitate dry during washing, otherwise a compact mass is formed that is difficult to wash free from salts. The gravimetric factor for aluminium is 0.05873 and that for aluminium oxide, 0.1110.(b) Titrimetric determination-Filter the solution and wash, as under (a), but use a medium filter-paper. Dissolve the precipitate in a hot mixture of 30ml of concentrated hydrochloric acid (12 M) and 50 ml of water, taken in small portions. Wash the filter with water. Collect the solution in a 300-ml conical, glass-stoppered flask, having a mark at the volume of 150 ml. Dilute to 150 ml and cool to room temperature. Now titrate the solution slowly with 0.1 or 0-2 M potassium bromate solution containing about 20 g of potassium bromide per litre. Shortly before the end-point, add a few drops of 0-1 per cent.methyl red and continue the titration slowly until the colour becomes quite yellow; then add 2 ml of bromate solution in excess. Stopper the flask, mix and wait 2 to 3 minutes. Then add about 2 g of potassium iodide, dissolve it by swirling, and titrate the liberated iodine with 0.1 N sodium thiosulphate solution, adding starch as indicator near the end-point. Subtract the value of the back-titration with thiosulphate from the potassium bromate added. One millilitre of 0.1000 N potassium bromate is equivalent to 2.248 mg of aluminium or 4.248 mg of aluminium oxide. Proceed according to (a) or (b) below. Finally, wash the precipitate twice with 5ml of cold water. Dry to constant weight at 140" to 150°C. NOTE- With this procedure the pH after precipitation lies between 5-2 and 5.8.TEST OF ACCURACY- To test the recommended procedure a standard aluminium solution containing about 1 mg of aluminium per ml was made by dissolving an accurately known weight of high purity aluminium in hydrochloric acid and diluting this solution to a known weight. Weighed aliquots of this solution were used so that the amount of aluminium taken for analysis was known with an accuracy to within 0.001 mg. The results for the titrimetric procedure are given in Table IV, and for the gravimetric procedure, in Table V. The bromate solution was standardised with high purity arsenic(II1) oxide. TABLE IV RESULTS SHOWING THE ACCURACY OF THE TITRIMETRIC PROCEDURE Approximate amount of Number of Main difference from Standard deviation of aluminium taken, determinations the theoretical value, single determination, mg mg mg 2 9 - 0.01, 0.007 10 12 * 0-00 0416 20 10 + 0.00, 0.009October, 19671 ALUMINIUM WITH 8-HYDROXYQUINOLINE. PART I TABLE V RESULTS SHOWING THE ACCURACY OF THE GRAVIMETRIC PROCEDURE 617 Approximate amount of Number of Mean difference from Standard deviation of aluminium taken, determinations the theoretical value, single determination, mg mg mg 2 6 - 0.013 0.005 10* 17 f 0.000 0.007 20 5 + 0.004 0.012 * The results for 10 mg were obtained by four different operators.It follows from these tables that with both procedures the standard deviation for a single determination is about 0.01 mg of aluminium. Excepting the tests with 2 mg, the mean does not differ significantly from the theoretical value.Strictly speaking, results somewhat too low should have been expected, as the solubility of the precipitate corresponds to between 20 and 25 pg of aluminium. EFFECT OF SOME ANIONS- It is known from the literature that the usual inorganic anions, such as chloride, sulphate, nitrate and perchlorate, do not interfere in this determination. As with Wassiljew,lo it was found that small amounts of fluoride (greater than 1 mg) interfere by giving low results. Addition of 1 to 5 g of boric acid decreases the error but is only effective in masking up to 2 mg of fluoride. As in the presence of tartrate no precipitate of aluminium hydroxide appears when neutralising the solution, the neutralisation has to be carried out by using indicator paper at a pH of up to about 3.An extended study of the accuracy in the presence of tartrate has not been made. Citrate and oxalate interfere, as in their presence precipitation is only complete at a pH above 7 to 8. Tartrate does not interfere but gives a much finer precipitate. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Third Edition, Charlot, G., “Les Me‘thodes de la Chimie Analytique. Analyse Quantitative Mintfrale,” Fifth Erdy, L., “Theorie und Praxis dev gravimetrischen Analyse, ” Volume 11, Akaddmiai Kiado, Buda- Vogel, A. I., “A Text-Book of Quantitative Inorganic Analysis including Elementary Instrumental Kolthoff, I. M., Belcher, R., Stenger, V. A., and Matsuyama, G., “Volumetric Analysis,” Volume Wilson, C. L., and Wilson, D. W., Editors, “Comprehensive Analytical Chemistry,” Elsevier Pub- Stumpf, K. E., 2. analyt. Chem., 1953, 138, 30. Claassen, A., Bastings, L., and Visser, J., Analytica Chim. Acta, 1954, 10, 373. Miller, C. C., and Chalmers, R. A., Analyst, 1953, 78, 686. Wassiljew, K. A., Zav. Lab., 1937, 6, 432. Macmillan & Co. Ltd., New York and London, 1952, p. 320. Edition, Masson et Cie, Paris, 1966, p. 584. pest, 1964, pp. 317 and 322. Analysis,” Third Edition, Longmans, Green & Co. Ltd., London, 1961, p. 387. 111, Interscience Publishers Inc., New York and London, 1957, p. 554. lishing Co., Amsterdam, London, New York and Princeton, 1962, Volume I C, p. 106. Received November 18th, 1966

ISSN:0003-2654    

DOI:10.1039/AN9679200614

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

618 Analyst, October, 1967, Vol. 92, $9. 618-621 The Determination of Aluminium with 8-Hydroxy quinoline Part II.* Precipitation in Ainmoniacal Cvanide - EDTA Solution BY A. CLAASSEN, L. BASTINGS AND J. VISSER (Philips Research Laboratories, N . V. Philifis’ Gloeilamfienfabrieken, Eindhoven, The Netherlands) A general procedure is described for the titrimetric determination of 2 to 20 mg of aluminium as the 8-hydroxyquinolinate by precipitation in am- moniacal cyanide - EDTA solution. This method is applicable in the presence of large amounts of a considerable number of elements. Inter- fering elements are beryllium, bismuth, gallium, hafnium, indium, niobium, antimony( 111), scandium, thorium, uranium, vanadium, zirconium and more than 1 mg of fluoride. Interference by chromium and titanium can be prevented by slight changes in the procedure.THE determination of aluminium with 8-hydroxyquinoline in acetate-buff ered solution is very unselective, as it only affords separation from magnesium, beryllium, the alkaline earths and the alkali metals. It can be made more selective by precipitation from ammoniacal cyanide solution, in which many metals are masked as the complex cyanide. The principle of this method is by Lang and Reifer,l and, in particular, by Heczko.2 It has since been applied, in various modifications, for the determination of aluminium in iron and stee1,293Jg6s6 copper al10ys~~~ and zinc al10ys.~ In 1934, we published a photometric methodlo for the determination of aluminium by chloroform extraction of the 8-hydroxyquinoline compound, which was based on a similar procedure ; the selectivity, however, was further enhanced by the addition of EDTA (ethylene- diaminetetra-acetic acid).This principle has since been applied to the macro determination of alumini~m.llsl~s~~ During the course of years an extensive investigation of this method has intermittently been made and has led to the following procedure being recommended for the macro deter- mination of aluminium in complex mixtures. METHOD REAGENTS- 8-Hydroxyqainoline, 2.5 $er cenzt.-Add 25 g of 8-hydroxyquinoline to 29 ml of 6 M hydrochloric acid, dilute with water, filter if necessary, and dilute to 1 litre. 8-Hydroxyqainoline wash solation-Dilute 8 ml of 2.5 per cent. 8-hydroxyquinoline solution to 500 ml, add 3 drops of 0.1 per cent. bromocresol purple (1 g per litre in 20 per cent.ethanol), neutralise with 2 M ammonia solution until the solution just begins to turn purple (pH about 6) and then dilute to 1 litre. Sodium sulphite, anhydrous. Potassium cyanide. Disodium ethylenediaminetetra-acetic acid (Na,EDTA) . Citric acid solution, 20 per cent. w/v. * For details of Part I of this series, see reference list, p. 621.CLAASSEN, BASTINGS AND VISSER 619 RECOMMENDED PROCEDURE- Dilute the aluminium solution containing between 2 and 20 mg of aluminium to about 100 ml, add 5 ml of citric acid solution (Note 1) and then 7 M ammonia solution until present in slight excess. Add 3 g of potassium cyanide and 1 g of anhydrous sodium sulphite (Note 2). Stir until all of the salts have dissolved and dilute to between 150 and 250 ml (Note 2).Heat slowly to a temperature of between 80" and 90" C and maintain the solution at this tem- perature for 2 minutes (Note 3). Then add 1 g of EDTA disodium salt (Note 4) and digest at 80" to 90" C for about 2 minutes. Allow to cool to about 70" C and add, while stirring vigor- ously, the appropriate amount of 2-5 per cent. 8-hydroxyquinoline reagent, add 0.70 ml for each milligram of aluminium, plus an excess of 20 ml. Heat to between 80" and 90" C and maintain the solution at that temperature for about 30 minutes. Allow to cool to about 50" C and filter through a medium filter-paper, retaining as much of the precipitate as possible in the beaker (Note 5). Then wash the precipitate with small portions of warm (50" to 60" C) wash solution, decanting as much as possible; do not use more than about 100 ml.Finally, wash twice with 5 to 10 ml of cold water. Dissolve the precipitate in a hot mixture of 30 ml of concentrated hydrochloric acid (12 M) and 50 ml of water, and titrate with potassium bromate, as described elsewhere (Note 6).14 NOTES- The total amount of citric acid should be 7 to 8 times the weight of tervalent elements, such as iron, rare earths, etc. 2. The amounts of potassium cyanide and sodium sulphite given are the minimum that should always be added. These amounts are adequate for about 0.1 g of elements forming complex cyanides, such as copper, nickel, iron, etc. If these elements are present in larger amounts, the required amounts of cyanide and sulphite should be chosen from Table I, and the volume diluted to that given in the last column.1. Add more citric acid if a precipitate appears when making the solution ammoniacal. TABLE I CONDITIONS FOR PRECIPITATION OF ALUMINIUM Amount of element to be Amount of potassium Amount of sodium complexed by cyanide, cyanide required, sulphite required, Volume after dilution, g g g ml < 0.1 3 1 150 0.1 to 0.25 3 2.5 200 0-25 to 0.5 5 5 200 0.5 to 1.0 10 10 250 If perchlorate is present, potassium cyanide should be replaced by the equivalent amount of sodium cyanide. 3. Do not heat the solution on a hot-plate or flame that is too hot, as local overheating can lead to decomposition of cyanide complexes or formation of organic decomposition products that are carried down with the aluminium 8-hydroxyquinolinate and give rise to increased consumption of bromate.4. The amount of EDTA disodium salt should be a t least 10 times the weight of the elements, cadmium, manganese, lead, zinc and rare earths $us 20 times the weight of alkaline earths and magnesium. 5. The use of quartz beakers is strongly recommended, as glass is attacked by the hot, strongly alkaline solution. Depending on the quality of the glass and the time of heating a t 80" to 90" C, 20 to 80 p g of aluminium can dissolve. 6. A gravimetric finish for this procedure is not advisable, as filter crucibles are heavily attacked by the strongly alkaline solution. A minimum amount of 1 g should always be added. DISCUSSION The reduction of cyanoferrate(II1) can be effected with ~ u l p h i d e ~ , ~ , ~ ~ , ~ ~ or sulphite1~7~*s9 Sulphite proved to be the reducing agent that With the recommended procedure a pH of between 8.5 and 10 is obtained.The aluminium or simply by boiling the alkaline s ~ l u t i o n . ~ ~ ~ ~ ~ gave the best results for cleanness of separation from iron. that escapes precipitation amounts to between 20 and 30pg. NON-INTERFERING ELEMENTS- The determination of 2 to 20mg of aluminium is possible, without interference, in the presence of 0.5 to 1 g of the following elements: silver, arsenic(III), arsenic(V), gold, cadmium,620 CLAASSEN, BASTINGS AND VISSER : DETERMINATION OF [Arta&d, VOl. 92 cerium(III), cerium(IV), cobalt, copper, iron(II), iron(III), germanium, mercury(I), mer- cury( 11) , lanthanum and other rare earths, magnesium, manganese, molybdenum(V1) , nickel, lead, palladium, platinum, antimony(V), selenium(IV), selenium(VI), tin(IV), tellurium(IV), tellurium(VI), thallium(I), thallium(III), tungsten(VI), zinc and the alkaline earths.In these tests, the amount of aluminium found was generally within 20.02 mg of the expected value. Yttrium does not interfere in amounts up to about 50mg; with larger amounts, losses of a few tenths of a milligram of aluminium occur. Chlorides, sulphates, nitrates or perchlorates do not interfere. INTERFERING ELEMENTS- Beryllium, bismuth, gallium, hafnium, indium, niobium, antimony(III), tantalum, thorium, uranium and zirconium should be absent, as they are all precipitated more or less completely as the 8-hydroxyquinolinate.In the presence of scandium the recovery of aluminium is only 80 to 90 per cent. complete. Chromium(V1) precipitates incompletely as an 8-hydroxyquinoline compound, colouring the aluminium 8-hydroxyquinolinate a vivid orange. Up to 20 mg of chromium(V1) scarcely interfere. Interference by chromium(II1) is much stronger, 2 and 10mg of chromium(II1) giving positive errors of about 0.05 and 0.3mg of aluminium, respectively. If the amount of chromium is below 20 mg, interference can therefore be prevented by oxidation to chromium(V1). Larger amounts of chromium can be removed by heating to fumes with perchloric and hydrochloric acids. Interference by chromium can also be prevented by converting it to its EDTA complex by boiling for at least 5 minutes with the necessary amount of EDTA (see Note 4); iron must first be reduced by boiling with sulphur dioxide.Fluoride up to about 1 mg does not interfere; larger amounts give results that are too low, even if a large excess of boric acid is added. It can, however, interfere, if more than about 100 mg of iron are present, by preventing the complete reduction of iron. This results in slight contamination of the precipitate by iron (green colour). Thus 100 mg of iron, with phosphate equivalent to 5 ml of concentrated orthophosphoric acid, showed no interference ; 500 mg of iron with the same amount of orthophosphoric acid, however, caused a positive error of 0.2 mg of aluminium. The positive error caused by 1 mg of vanadium amounts to 0.05 mg of aluminium, and as the vanadium rises to 10 mg, to between 0.2 and 0.3 mg of aluminium.Titanium precipitates completely as the 8-hydroxyquinolinate at a pH of about 9 and below; at a pH above 9, its precipitation is incomplete. Contamination by titanium is shown by the orange discoloration of the precipitate. Determination of aluminium in materials containing amounts of titanium that are not too large can be made by dissolving the contaminated precipitate in hydrochloric acid and removing the titanium by extraction with cupferron - chloroform. Aluminium can be determined in the aqueous phase, as described e1~ewhere.l~ In Table I1 the results are given of the application of the recommended procedure to some standard samples. Orthophosphate does not interfere in principle. Vanadium precipitates incompletely. The interference by vanadium increases below pH 9.TABLE I1 DETERMINATION OF ALUMINIUM IN STANDARD SAMPLES Sample weight, Certified value Sample g and range N.B.S. 93. Borosilicate glass . . .. 1.0 1-03 (1.01 to 1.03) N.B.S. 91. Opal glass . . . . . . 0.5 3.18 (3.16 to 3.31) N.B.S. 94a. Zinc base alloy . . .. 0-25 3.90 (3.88 to 3.95) N.B.S. 62b. Manganese bronze . . . . 1.0 0.97 (0.95 to 0.98) N.B.S. 171. Magnesium base alloy . . 0.5 2.98 (2.96 to 3.00) N.B.S. 106a. Cr - Mo - A1 steel . . . . 1.0 1-08 (1.07 t o 1.12) B.C.S. 179. Manganese brass B . . 0.7, 1.0 1-62 (1.54 to 1.77) B.C.S. 233. Permanent magnet alloy . . 0.15 6.98 (6.93 to 7.08) B.C.S. 312. Permanent magnet alloy . . 0.12 7.87 (7.77 to 7-93) (Ti, 0.79 per cent.) (Ti, 1.21 per cent.; Ta + Nb, 1.35 per cent.; Al, 7.88 per cent.) Aluminium found, per cent.1.02, 1-02 3.19, 3.19 3.91, 3.91 0.97, 0.98, 0.99 2.96, 2.97 1.09. 1.10, 1.10 1.61, 1-81, 1-62 6.95, 6.95 6.96, 6-97 7.85, 7.86October, 19671 ALUMINIUM WITH 8-HYDROXYQUINOLINE. PART I1 REFERENCES 62 1 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Lang, R., and Reifer, J., 2. analyt. Chem., 1933, 93, 161. Heczko, T., Chemikerzeitung, 1934, 58, 1032. Box, F. W., Analyst, 1946, 71, 317. Klinger, P., Arch. EisenhiittWes., 1939-40, 13, 21. Pigott, E. C., J . Soc. Chem. Ind., Lond., 1939, 58, 139. Steele, S. D., and Russell, L., Iron Steel, Lond., 1942, 16, 182. Edwards, W. T., Analyst, 1948, 73, 556. Weidmann, H., 2. Metallk., 1953, 44, 565. Reutel, C., Metall Erz, 1941, 38, 170. Claassen, A., Bastings, L., and Visser, J., Analytica Chim. Acta, 1954, 10, 373. Detmar, D. A., and van Aller, H. C., R e d . Trav. Chim. Pays-Bas, 1956, 75, 1429. Hoekstra, E., and van Dorp, F. C., Chem. WeeRbl., 1955, 51, 895. Mohr, E., Chem. Tech., Bed., 1959, 11, 598. Claassen, A., and Bastings, L., Analyst, 1967, 92, 614. “1964 Book of A.S.T.M. Standards, Part 32, Chemical Analysis of Metals,” American Society for Methods of Analysis Committee, J . Iron Steel Inst., 1954, 176, 263. NOTE-Reference 14 is to Part I of this series. Testing Materials, Philadelphia, p. 258. Received November 18th, 1966

ISSN:0003-2654    

DOI:10.1039/AN9679200618

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

622 Analyst, October, 1967, Vol. 92, pp. 622-626 The Determination of Aluminium in Wool by Atomic-absorption Spectroscopy BY IT;. R. HARTLEY AND A. S. INGLIS victoria, Australia) (Division of Pvotein Chemistry, CSIRO, Wool Research Laboratories, Parkville N.2 (Melbouvne), Atomic-absorption spectroscopy at 3092-7 A in a nitrous oxide - acetylene flame provides a simple and precise method for determining aluminium in wool. The sample is dissolved in constant-boiling hydrochloric acid, and the solution sprayed directly into the flame. The presence of hydrochloric acid partially suppresses the aluminium absorbance, whereas the amino-acids present in the wool hydrolysate enhance the absorbance. A linear calibration graph over the range 0 to 130,xg per ml is obtained from solutions of alu- minium in constant-boiling hydrochloric acid containing the hydrolysed protein.Although the calibration graph has an accuracy of +Z per cent., the heterogeneous nature of wool limits the accuracy of the analysis to f3 per cent. The method, which has been applied successfully down to 0.02 per cent. of aluminium on wool, is equally suitable for other insoluble protein materials, such as hair and hide. IN connection with a study of the properties of wool after treatment with aluminium salts, it became necessary to determine the effect of various conditions on the uptake of aluminium by wool. As only relatively small amounts of aluminium were involved, a method with a high sensitivity was necessary. It was also desirable that the determination should not be affected by the other metal ions present in wool, or by the brown colour of the solutions obtained on dissolving wool.From these considerations it appeared that a procedure involving atomic-absorption spectroscopy with a nitrous oxide - acetylene flame1 would be superior t o methods based on complexometric titrations,2 gravimetric3 or spectrophotometric analy~is.~ In addition, atomic-absorption spectroscopy offered the possibility that the solution of amino-acids obtained on hydrolysis of wool in constant-boiling hydrochloric acid could be used directly, thus eliminating the necessity for filtration, dilution or the addition of further reagents. EXPERIMENTAL APPARATUS- The equipment used in this work was the Techtron AA-3 atomic-absorption spectro- photometer with an aluminium hollow-cathode lamp supplied by Atomic Spectral Lamps.The operating conditions that gave optimum sensitivity were: slit width, 100 p ; wavelength, 3092.7 A; lamp current, 11 mA. These conditions are in agreement with those suggested by Amos and Wi1lis.l A stainless-steel high temperature burner (type AB40), supplied with the Techtron AA-3 spectrophotometer, was used. Aspiration via the Techtron atomiser was by nitrous oxide (pressure, 151b per sq. inch), and the acetylene pressure used was sufficient to produce a red “feather” in the flame ten times the height of the pale blue cone. When more acetylene was present a deposit of carbon formed around the mouth of the burner, which led to an unstable flame and fluctuations in the absorbance reading.When less acetylene was present the aluminium was less completely atomised. The burner was carefully aligned for direction, horizontal position and height to obtain the maximum sensi- tivity. The first two factors were found to be critical, whereas the height could be varied over a range of about 1 cm without altering the absorbance value. The flame was first ignited with only acetylene present. After increasing the acetylene pressure to about 18 lbHARTLEY AND INGLIS 623 per sq. inch, the nitrous oxide was turned on to give a pressure of 15 lb per sq. inch, and the acetylene pressure then reduced to give the desired flame. The whole operation of lighting and adjusting the flame was carried out as rapidly as possible to prevent the formation of a carbon deposit at the mouth of the burner.It was necessary to light the flame at least half an hour before analysing the samples, as variations in readings occurred during this time. REAGENTS- Constant-boiling hydrochloric acid--Prepare by the method described by VogeL5 Aluminium sulfdzate solution*-Prepare a stock solution about 0.1 M by dissolving 31.52 g of aluminium sulphate [A1,(S0,),.16H20] (analytical-reagent grade) in 1 litre of water. Standardise the solution by using an excess of EDTA and back-titrating against a standard solution of zinc chloride, with Eriochrome T as indicator.2 (The standard zinc chloride solution is prepared by dissolving analytical-reagent grade zinc metal in concentrated hydro- chloric acid, analytical-reagent grade, and diluting with water.) Prepare solutions of aluminium sulphate for the calibration graph by diluting the stock solution and adjusting the concentration of hydrochloric acid in each solution to 6.19 N PREPARATION OF CALIBRATION GRAPH- Transfer by pipette suitable volumes of the standardised aluminium sulphate solution to give transmission values in the region of 40 to 95 per cent.Make up to volume with a mixture of the hydrochloric acid and distilled water, into standard flasks, so that the final solution is 6.19 N with respect to hydrochloric acid. Seal 5-ml aliquots of the solutions, together with 0.3 g of a sample of the wool that has not been treated with aluminium salts, in hard-glass tubes and heat in a forced-draught oven at 110" C for 20 hours. After cooling, shake the tube well and cautiously open.Aspirate the solution directly into the flame and record the transmission value. The calibration graph is linear for aluminium concentrations of up to 130 pg per ml (0.0048 M) and should pass through the origin. PREPARATION OF SAMPLES- Weigh accurately samples of wool (about 0-3 g) that contain sufficient aluminium to give a direct reading without dilution. The wool samples should be equilibrated for at least 24 hours in a constant-humidity room before weighing. Seal each sample in a thick-walled tube, together with 5 ml of constant-boiling hydrochloric acid, accurately measured by pipette, and heat in a forced-draught oven at 110" C for 20 hours. After cooling, shake the tube well and open it cautiously. Aspirate the solution directly into the flame and record the trans- mission. A calibration graph must be prepared each time a batch of samples is analysed to eliminate differences that arise in setting up the instrument.After use, the aspirator must be well washed with distilled water to remove any traces of solvent that could cause corrosion if left on the aspirator. RESULTS AND DISCIJSSION EFFECT OF SOLVENT- A series of calibration graphs was prepared by using solutions of aluminium sulphate in water; aluminium sulphate in 6.19 N hydrochloric acid; and aluminium sulphate and hydrolysed protein in 6.19 N hydrochloric acid. Table I shows that all of the systems studied gave linear plots of absorbance against aluminium concentration of up to 130 pg of aluminium per ml (0.0048 M). At higher concentrations the plot curved away from the absorbance axis, indicating that the aluminium was then being less completely atomised. Typical values for the ratio of the absorbance to aluminium concentration, and the sensitivity in the different media, are given in Table 11.The sensitivity in water is a little less than that observed by Amos and Willis,l who detected 1 pg of aluminium per ml at 1 per cent. transmission. This difference is almost certainly caused by minor instrumental differences. It can be seen that the presence of 6.19 N hvdrochloric acid suppresses the aluminium absorbance by 17.6 per cent. In the presence of 0.2 N hydrochloric acid, Amos and Thomas6 found that the aluminium absorbance is suppressed by 11 per cent. ir; an oxygen - nitrogen - acetylene flame, although Amos and Willisl found later that no effect could be detected with 0.5 N hydrochloric acid in a nitrous oxide - acetylene flame.* A standard solution could be prepared with high purity aluminium metal, but this was not readily available in Australia when this work was undertaken.624 [Analyst, Vol. 92 INTERFERENCES- In the presence of the protein hydrolysates studied the aluminium absorbance is enhanced relative to that observed in 6.19 N hydrochloric acid. This enhancement could be caused by evaporation of water vapour when sealing the samples in glass tubes; caused by the trace metals present in the proteins; or by the amino-acids present in the hydrolysates. HARTLEY AND INGLIS : DETERMINATION OF ALUMINIUM Concentration of aluminium, 12-92 25.85 38.77 51.75 77-60 PP Per ml 103.5 129.2 181.1 258.5 TABLE I CALIBRATION POINTS FOR ALUMINIUM IN DIFFERENT MEDIA Medium Water R s t i o o f absorbance to Absorbance concentration 0.0223 0.0458 0.0680 0-0915 0.137 0.182 0.229 0-310 0.426 1.73 1.77 1-75 1-77 1-77 1.76 1-77 1.71 1.67 6.19 N Hydrochloric acid R a t i o o f absorbance to Absorbance concentration 0.0189 0.0373 0.0555 0.0731 0.115 0.152 0.187 0.252 0.347 Medium 1.46 1.44 1.43 1.41 1.48 1.47 1-45 1.39 1.34 6.19 N Hydrochloric akd + scoured wool r Absorbance 0.0200 0.0398 0.0605 0.0809 0.122 0.158 0.201 0.272 0.387 3 Ratio of absorbance to concentration 1-55 1-54 1.56 1.56 1.57 1-53 1.55 1.50 1.49 Concentration of aluminium, 12.92 25.85 38.77 51-75 77-60 p g Per ml 103.5 129.2 181-1 258.5 6.19 N Hydrochloric acid + bleached wool R a t i o o f absorbance to Absorbance concentration 0~0200 1.55 0.0410 1-58 0.0605 1-56 0.0796 1-54 0.119 1.53 0.158 1.53 0.201 1.55 0.272 1-50 0.382 1-47 6-19 N Hydrochloric acid + bovine hide -Ratiof absorbance to Absorbance concentration 0*0200 1-55 0.0410 1.58 0.0595 1.53 0.0809 1-56 0.122 1.57 0.161 1-56 0.201 1.53 0-276 1.52 0.382 1-47 6.19 N Hydrochloric acid + bovine hair - - - - - i a f absorbance to Absorbance concentration 0.0223 0.0458 0.0680 0-0915 0-135 0.181 0.222 0.305 0.420 1.72 1-77 1.75 1-77 1.74 1-75 1.72 1.68 1.62 TABLE I1 DETERMINATION OF ALUMINIUM BY ATOMIC-ABSORPTION SPECTROSCOPY Ratio of Lower limit of absorbance to Sensitivity,? analysis, $ Medium aluminium concentration* pg of aluminium per ml pg of aluminium per ml Water .. . . . . 1.76 f 0.03 2.50 15-3 6-19 N Hydrochloric acid 1-45 & 0.03 3.04 18-5 + scoured wool . . 1.55 f 0.02 2-84 17.3 + bleached wool . . 1.55 f 0.03 2.84 17-3 + bovine hide . . .. 1.55 & 0.03 2.84 17-3 6.19 N Hydrochloric acid 6.19 N Hydrochloric acid 6.19 N Hydrochloric acid 6.19 N Hydrochloric acid + bovine hair . . . . 1-75 f 0.03 2.52 15.4 * Mean ratio of absorbance to concentration in the aluminium concentration range -f Sensitivity necessary to give 1 per cent. transmission, $ Concentration of aluminium in solution necessary to give 94 per cent. transmission, 0 to 130 pg per ml. which is the highest transmission consistent with an accuracy of f 3 per cent.October, 19671 I N WOOL BY ATOMIC-ABSORPTION SPECTROSCOPY 625 The first possibility was eliminated by preparing a calibration graph from solutions of aluminium in 6.19 N hydrochloric acid that were sealed in glass tubes, heated in a forced- draught oven at 110" C for 20 hours, cooled, cautiously opened and aspirated into the flame.The points on this graph were indistinguishable from points obtained by using solutions that had not been sealed. The exact quantitative concentration of each of the trace metals, calcium, magnesium, iron, sodium and potassium, present in wool,' varies from sample to sample, but as the total ash obtained from wool is about 0.4 per cent.,* the amount of each metal present must not be greater than about 0.1 per cent. When each of these metals was added to standard aluminium solutions in concentrations equivalent to 0.1 per cent.and the absorbance measured, only potassium, which enhanced the absorbance by about 4-5 per cent., had any significant effect. Accordingly the proteins were analysed for potassium by using the atomic- absorption spectrometer with an air-coal gas flame and a technique analogous to that described for aluminium. The results indicated that the amount of potassium present in wool and bovine hide is insignificant (about 0.0015 per cent.). Bovine hair was found to contain a larger amount of potassium (0.019 per cent.), which will account for part of the difference in the enhancement of the aluminium absorbance observed with this protein. The ability of potassium to enhance the absorbance of aluminium corresponds to the enhancement by iron observed by Amos and Wi1lis.l The enhancement of the aluminium absorbance by amino-acids was determined by preparing a mixture of amino-acids in about the same proportions as those present in wool (see Table 111).When the mixture of amino-acids (0.3 g) was added to 5 ml of a solution of aluminium in 6.19 N hydrochloric acid (129.3 pg of aluminium per ml), the absorbance was enhanced from 0.187 to 0.204, which was equal, within experimental error, to the en- hancement by wool. These results suggest that the enhancement of the aluminium absorbance by wool and bovine hide is caused by the amino-acids alone. As the amino-acid contents of bovine hideg and woollo are different, it is interesting that both give similar enhancements of the aluminium absorbance. With bovine hair, the potassium present will also contribute to the enhancement observed.TABLE I11 MIXTURE OF AMINO-ACIDS USED TO REPRESENT WOOL HYDROLYSATE Present in 0.3g of wool, Amino-acid mg* Lysine .. . . . . 9 Arginine . . .. .. 20 Aspartic acid . . . . .. 20 Threonine . . . . .. 18.5 Serine . . . . . . . . 29 Glutamic acid . . . . 34 Proline . . .. . . 18 Glycine . . . . . . 27 Alanine . . . . .. 17 Valine . . . . .. . . 18 Methionine . . . . . . 2 Isoleucine . . . . .. 10 Leucine .. . . .. 24 Histidine . . . . . . 2.5 Half-cystine . . .. . . 34 Tyrosine . . .. .. 12 Phenylalanine . . .. 9 * Data from reference 10. Present in mixture, mi? 9 2 17 22 17-5 28 37 21 27 16 34 17 2 9 24 12 8 As the amount of trace metals present and the amino-acid composition will vary from sample to sample, it is necessary to prepare the calibration graph from standard aluminium solutions that contain the hydrolysed protein under investigation in 6.19 N hydrochloric acid. In this way variations caused by the amino-acid composition and metallic contamination of the protein will be eliminated.There is a 3.4 per cent. increase in volume when samples containing 0.3 g are dissolved in 5 ml of constant-boiling hydrochloric acid. However, this does not affect the analysis as the same volume change occurs in the samples used to prepare the calibration graph.626 HARTLEY AND INGLIS REPROD u CI BILITY- The reproducibility of the method was determined by analysing six samples of scoured and bleached wool that had been treated with aluminium sulphate solutions.The results shown in Table IV indicate that the reproducibility is &3 per cent. It was possible to dissolve up to 0.4 g of wool in 5 ml of constant-boiling hydrochloric acid. Thus, from the lower limit of analysis shown in Table 11, the method can be used down to 0.2 mg of aluminium per g of dry wool (0.02 per cent. w/w), with a reproducibility of f3 per cent. As a reproducibility of +2 per cent. is obtained from the calibration graph, presumably the uptake of aluminium is not completely uniform throughout the constituent proteins of wool, and small sampling errors occur, The large increase in uptake of aluminium by bleached wool is not unexpected and may result from the presence of sulphonic acid groups produced by oxidation of disulphide groups during bleaching with hydrogen peroxide.TABLE Iv REPRODUCIBILITY OF ALUMINIUM DETERMINED IN WOOL SAMPLES Scoured wool Bleached wool v L r \ Dry weight, Dry wool, Dry weight, Dry wool, 0.32790 0.0339 0.364 0.24390 0.155 2.05 0.32360 0.0339 0.369 0.2 7 3 98 0-174 2-06 0-29412 0.0327 0.359 0.22769 0-177 2.06 0.28429 0.0315 0.357 0.30368 0.184 1-96 0-31364 0.0339 0.350 0.26847 0.164 1-97 0.37448 0.0410 0.352 0.31177 0.194 2.01 ---- g Absorbance* mg of aluminium per g g Absorbance* mg of aluminium per g Mean .. . . 0.358 -& 0.011 Mean .. . . 2.02 3 0.06 * Sample hydrolysed in 5 ml of constant-boiling acid. SCOPE OF THE METHOD- Although only wool has been studied in detail, the results in Table I indicate that the method should be equally suitable for other insoluble protein materials, such as hair and hide.The great advantage of being able to use the hydrolysate directly will still apply. It is anticipated that such biological materials treated with different metals could also be analysed with advantage by a similar procedure. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Amos, M. D., and Willis, J. B., Spectrochim. Acta, 1966, 22, 1325. Schwarzenbach, G., “Complexometric Titrations,” Methuen & Co. Ltd., London, 1957, p. 88. Trotman, S. R., and Trotman, E. R., “Textile Analysis,” Second Edition, Charles Griffin and Co. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience Vogel, A. I., “A Textbook of Quantitative Inorganic Analysis including Elementary Instrumental Amos, M. D., and Thomas, P. E., Anzalytica Chirn. Acta, 1965, 32, 139. Dusenbury, J. H., in von Bergen, W., Editov, “Wool Handbook,” Third Edition, Interscience Ward, W. H., “Textile Research Institute Summary Report for October 1948 to October 1952,” Veis, A., ip2 Hall, D. A., Editor, “International Review of Connective Tissue Research,” Volume 3, O’Donnell, I. J., and Thompson, E. 0. P., Aust. J . Biol. Sci., 1962, 15, 751. Ltd., London, 1948, p. 255. Publishers Inc., New York and London, 1959, p. 219. Analysis,” Third Edition, Longmans, Green and Co. Ltd., London, 1961, p. 233. Publishers Inc., New York and London, 1963, p. 219. Textile Research Institute, Princeton, New Jersey, 1953, p. 35, Table XI. Academic Press, London, 1965, p. 115. Received Afiril 17th, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200622

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, October, 1967, Vol. 92, p p . 627-633 627 A Simple Quantitative Method for the Determination of Small Amounts of Amino-acids BY J. G. HEATHCOTE AND K. J. WASHINGTON (The University of Salford, Salford 5, Lancashire) h simple method for the quantitative determination of amino-acids has been described. The method depends upon the paper-chromatographic separation of the amino-acids, followed by staining to form a coloured complex with cadmium acetate and ninhydrin. The coloured spots are eluted from the paper with methanol and determined colorimetrically. Amounts within the range 0.1 to 40 pg (0.1 to 25 ,umoles-2) can be determined by this method to within an over-all accuracy of +10 per cent. FEW simple convenient methods are available for the separation and quantitative determina- tion of amino-acids in amounts of 0.1 to 40 pg (0.1 to 25 pmoles-2).Atfield and Morris1 used high voltage electrophoresis to separate mixtures of amino-acids, which were then determined by the cadmium - ninhydrin method of Heilmann, Barollier and Watzke.2 The precision claimed was comparable with that obtained by means of the ion-exchange method of Moore and Stein, and 250 pg of protein were sufficient for a complete analysis. We have developed a similar method for the determination of amino-acids in biological fluids, involving separation of the amino-acids by paper chromatography and dispensing with the need for expensive high voltage electrophoresis equipment. The method is capable of being carried out with simple apparatus such as a chromatographic tank and a colorimeter.It consists in developing chromatograms, uni- or bi-dimensional, in suitable, preferably volatile, solvents and then drawing each of the dried sheets through a standard volume of the staining reagent. In order to reduce the background colour as far as possible, a lower concentration of ninhydrin was present in the reagent than was used by Atfield and Morris.1 Furthermore , to remove the necessity for shaking equipment and subsequent filtration, the coloured complexes were eluted from the paper with methanol. In its present form the method is applicable to the estimation of amino-acids that are present in low concentration in biological fluids, the amino-acids being first separated by paper chromatography. The method has been applied to protein-free filtrates of human gastric juice obtained from normal subjects and from pernicious anaemia patients3; samples of de-salted urine (D.M. Davies, unpublished work); and bacterial culture fluids (M. F. Dutton, unpublished work). EXPERIMENTAL REAGENTS AND APPARATUS- All solvents and the cadmium acetate were of analytical-reagent grade. The concen- trated, reagent grade, ammonium hydroxide was kept in the refrigerator. Chromatographic solvent systems-Although any suitable solvent systems can be used, the most convenient pair for the general separation of the common amino-acids is the propan- 2-01 - water - ammonium hydroxide (7 + 2 + 1 v/v) mixture of Heathcote and Washington,4 and the butanol -acetic acid - water (12 + 3 + 5 v/v) of Smith.5 These are referred to in the text as solvent systems 1 and 2, respectively.For the separation of the leucine - isoleucine mixture the solvent system of Heathcote and Jones6 is recommended. This consists of propan-2-01- formic acid - water (20 + 1 + 5 v/v) as the first solvent, and t-butyl alcohol - ethyl methyl ketone - ammonia solution - water (50 + 30 + 10 + 10 v/v) as the second solvent. Amino-acid standard solutions-Stock solutions of amino-acids, 0.05 M, were made in aqueous propan-2-01, 10 per cent. v/v, with the minimum amount of dilute hydrochloric acid added to effect solution, if necessary. The solutions were kept in the refrigerator and diluted to 0-01 M as required.628 HEATHCOTE AHD WASHINGTON : A SIMPLE QUANTITATIVE METHOD [Analyst, Vol. 92 Amino-acid stainiag reagent-The staining reagent for the separated amino-acids was essentially the same as that used by Atfield and Morris,l but a lower concentration of ninhydrin was used for the following reasons.Opienska-Blauth, Kowalska and Pietru~iewicz~ recom- mended that the concentration of ninhydrin in the cadmium acetate - ninhydrin mixture should be less than 0-5 per cent. w/v in order to ensure negligible values for the blank. Atfield and Morris1 used a concentration of 1 per cent. w/v and although this increases the final intensity of the colour it tends to produce a much higher blank. The reagent was made by dissolving 0.5 g of cadmium acetate in 50 ml of water to which 10 ml of glacial acetic acid had been added. Acetone was then added until the total volume was 500 ml.Portions of this solution were taken before use and sufficient solid ninhydrin was added until the final concentration was 0.2 per cent. w/v. Chromatographic apparatus-Smith's universal apparatus with a 10-inch aluminium frame was chosen, and pre-punched square sheets of Whatman No. 1 chromatographic-paper were used. A polythene dip-tray was used for staining the sheets of paper. Micro-Pipettes-A series of heavy-wall capillary pipettes and precision capillary pipettes were made and each was calibrated to deliver a known volume of liquid. The series of pipettes covered the range 1 to 20 p1. An Agla syringe was found convenient for transferring a series of volumes of the same liquid, such as 2.5, 5, 10 and 20 p1. Miscellaneous apparatus-The papers were handled on glass plates in order to prevent any contamination of the paper surface and supported on glass rods after staining.Volumetric and other apparatus-Beakers of 5 and 10-ml capacity were used for collecting the eluates, and the volume of each was made up to 10 ml in a B grade graduated flask. Elution apparatus-This was based on the publication of Kemble and MacPherson.8 The apparatus used is shown in Figs. 1 and 2. A piece of Whatman No. 1 filter-paper was cut, Fold \ Glass Glass trough- Cork support\ Beaker /Glass I L Fig. 1. Elution apparatus, end view Glass cover / trough, 1-1 Glass plate Fig. 2. Elution apparatus, side viewOctober, 19671 FOR THE DETERMINATION OF SMALL AMOUNTS OF AMINO-ACIDS 629 as shown in Fig. 3, to the shape of a rectangle 30 cm x I cm where I is the length of the trough. The length of trough, 17, 26, or 40 cm, was chosen according to the number of spots to be eluted.The paper was folded along the centre and in an outward direction along a line 3 cm on each side of the central fold. Projections, 10 cm long and 1.5 cm wide, were left with a gap of 1 cm between each projection along each edge. The size of the projection was always slightly larger than the size of the spot, or band, to be eluted, in order to obtain quantitative elution. Fig. 3. Plan of paper cut for elution (not to scale) One or two troughs, semi-circular in cross-section, were accommodated under a rect- angular glass cover. This maintained a solvent-saturated atmosphere in the apparatus. A 5-ml beaker was placed under each 10-cm projection, which was adjusted so that when hanging vertically, it reached one-third of the way down the beaker.METHOD The amino-acids are first resolved into groups by solvent system 1. The groups are then resolved into individual components, if necessary, by the application of solvent system 2 or other suitable systems. APPLICATION OF AMINO-ACIDS AND DEVELOPMENT- For one-dimensional paper chromatograms the papers were marked with the names and concentrations of the amino-acids used. The spacing between each pair of amino-acid spots was at least 3 cm to avoid overlap. The solution of amino-acids to be examined was applied by means of one of the calibrated pipettes, or the Agla syringe, to give a spot of diameter not greater than 0.5 cm. A volume of from 1 to 10 pl of each particular standard solution (0.01 M) of amino-acid was applied.If the volume of solution to be applied was greater than 4 pl, the application was performed in stages with intermittent drying by means of an air blast. The solution was applied as a narrow band along the base-line for one-dimensional paper chromatograms. Development was then carried out with solvent system 1 or solvent system 2, or both, as previously described. The solvent (200 ml) was placed in the aluminium tray. The aluminium frame holding five individual papers was placed in position and the heavy ground-glass lid replaced. Air-tight conditions were ensured by greasing the edges of the lid. The tank was placed on a horizontal surface to ensure a level solvent front and uniform development.The solvent was allowed to run for 11 hours at a temperature of about 21" C. Under these conditions optimum resolution was achieved, the solvent front travelling about 20 cm. The solvent front was best located as a bright line under ultraviolet light. REMOVAL OF SOLVENT- papers in a current of air. any residual smell of ammonia in about 1 hour. When development was complete, the solvent was removed by placing the frame and If solvent system 1 was used, the papers were dry and free from If solvent system 2 was used, even when630 HEATHCOTE AND WASHINGTON : A SIMPLE QUANTITATIVE METHOD [Analyst, Vol. 92 there was no detectable smell of solvent, it was found necessary to continue blowing air for at least 5 minutes over each paper, otherwise any residual acetic acid retarded the development of the colour produced by the staining reagent.Although similar results were obtained whichever solvent system was used first in a two-dimensional chromatogram, it was more convenient to develop first with solvent system 1 because of the speed with which the solvent could be removed from the paper. STAINING- The dry chromatogram was drawn slowly through 15 to 20ml of the staining reagent contained in a polythene dip-tray. After use, the reagent was discarded, the tray dried and fresh reagent poured out for the next sheet; thus concentration of the reagent by evaporation was avoided and consistent results were obtained. Accidental fingerprints were avoided by the use of forceps and by handling above the solvent front. The paper, after hanging until visibly dry, was then carefully transferred to a dark, ammonia-free atmosphere in which the papers were kept overnight. This allowed the amino-acid complexes to attain maximum intensity with minimum development of background colour.The best container was a large desiccator containing concentrated sulphuric acid. In such a container the background of small stained strips remained colourless for many weeks. A good substitute for the desiccator was provided by placing the papers on glass supports inside a well fitting cupboard containing a small beaker of concentrated sulphuric acid. This kept the background colourless for about 3 days, although the papers were best examined 24 hours after staining. The papers developed a pink background colour on standing for several days and this made the location of trace amounts of 1 to 2 pg of amino-acids difficult because of the lack of contrast.Surprisingly enough, the colour of the stained spots of amino-acids remained constant within experimental error. EFFECT OF HEAT ON COLOURED COMPLEXES- Heating caused a deepening in the colour of the complex but tended to render it insoluble in any eluting solvent. Thus heating to 80" C for 1 minute rendered the complex completely insoluble in methanol. Consequently, heating was not used to develop the colour; instead, the papers were allowed to stand in the manner described until the maximum colour had developed. ELUTION OF COLOURED COMPLEXES- The cadmium - ninhydrin complex was eluted from the paper by means of dry, acid-free methanol.Each stained spot was cut out with forceps and stainless-steel scissors and then sewn on to a paper tongue, with a glover's needle (Nos. 3 to 7) and white cotton. The spot and paper strip were held together by cotton, as metal staples cannot be used because these react with the cadmium - ninhydrin c~mplex.~ The spots were sewn on to tongues of paper, 10 cm x 1.5 cm, and the paper placed for elution as shown in Figs. 1 and 2. About 150 ml of methanol were placed in the trough, the cover replaced and the eluates collected in the beakers. It was possible, by using this apparatus, to elute as many as thirty spots at any one time, with a trough of 40-cm length. It was observed that the coloured complexes formed by most of the amino-acids were eluted within 2 hours, only one or two of relatively large amounts, 15 to 25pg, requiring a longer time.It was found convenient to allow the elution to continue overnight, during which time the strips were completely eluted. When the cover was removed, each strip was brought clear of the liquid surface to prevent it from drying out in the atmosphere and thus re-absorbing the eluate. COLORIMETRIC DETERMINATION- The eluate from each spot, which corresponded to a known amount of amino-acid, was made up to a volume of 10ml in a graduated flask. The optical density of the eluate was determined against a blank of pure methanol in a 4-cm glass cell at the position of maximum absorption on a Hilger Uvispek. The maximum wavelength of the red complex was 500 mp and that of the yellow complex of the amino-acids, proline and hydroxyproline, 352 mp.The background reading from the paper was nearly always less than 0.02 in optical density and was therefore negligible. The paper was cut as in Fig. 3.October, 19671 FOR THE DETERMINATION OF SMALL AMOUNTS OF AMINO-ACIDS 631 Graphs of optical density against pmoles-2 of amino-acid were plotted for each amino- acid. Each graph was linear over the range of amounts of amino-acids applied, 0.1 to 25 pmoles-2, but the slope was found to be different for each amino-acid. This was largely to be expected in view of the results of Heilman, Barollier and Watzke,2 who showed that the colour yield of the complex from each amino-acid was different in each case. APPLICATION AND RECOVERY EXPERIMENTS- Various amounts of standard amino-acids between 5 and 50 pmo1es-2 (4.4 to 44.8pg) were applied to sheets of chromatographic paper and the recoveries were determined after chromatography. A known weight, 1.2 mg, of alanylglycine was hydrolysed under nitrogen with 0.30 ml of 5-9 N hydrochloric acid for 20 hours at 105" C in a sealed tube. After removal of excess of acid, the contents were made up to 5 ml and duplicate samples (A and B) of 39.2 p1 each were examined by chromatography.To sample (B) 10 pmoless2 each of alanine and glycine were added and the recoveries ascertained. These conditions of hydrolysis were used for a peptide found in normal human gastric juice, but many biological samples were examined without hydrolysis for the presence of free amino-acids.These included protein-free filtrates of gastric juice shown in Table I and de-salted samples of normal urine and urine from patients with pernicious anaemia, as shown in Table 11. TABLE I FREE AMINO-ACIDS OF NORMAL GASTRIC JUICE Optical Amount of amino-acid, Concentration of amino-acid, Amino-acid density pmoles-2 mg per cent. w/v Alanine . . . . . . 1-455 Aspartic acid . . . . 0,115 Arginine . . . . . . 0.650 Glutamic acid .,. . . 0.280 Glycine . . . I . . 0.225 Leucine . . .. . . 0.74 Valine . . .. . . 0.415 19.1 3-0 7.5 2.7 8-0 9-1 5-1 Amino-acid Alanine . . Aspartic acid Glutamic acid Histidine . . Phen ylalanine Taurine . . Tyrosine . . TABLE I1 FREE AMINO-ACIDS OF URINE Concentration of amino-acids in mg per cent. w/v 1.7 0.4 1.3 0.4 0.6 1.2 0-6 Pernicious anaemia -n r Paper chromatography ion exchange .. 4.7 5.7 . . 0.66 0.93 . . 19.9 16.7 . . 11.4 15.9 . . 1.06 1.73 .. 21.5 19.8 .. 4.3 5.0 Normal Technicon rw ion exchange ion exchange 5.5 1.4 to 4-7 0.48 < 0.6 2-66 0-53 to 2-66 10.5 7.5 to 21-3 3.5 0.6 to 2.06 8-0 5.7 to 19.6 0.99 1.0 to 3.2 RESULTS AND DISCUSSION The median values of optical density that were obtained for varying amounts of 8 control amino-acids are given in Table 111. Table IV illustrates the experimental values obtained for lysine and glycine. These have been chosen because they represent the extreme values of colour yield for a given amount of amino-acid applied to the paper. The method was found to be reproducible if certain variable factors were controlled. The main sources of error arose in the addition by pipette of the standard solutions, and in the control of the amount of reagent coming in contact with each amino-acid; other errors include the sensitivity of the staining system to pH, acid preventing the development of the colour, and the presence, or absence, of background colour.The error from the last can be caused by an increase in intensity due to the presence of ammonia vapour in the atmosphere.632 HEATHCOTE AND WASHINGTON : A SIMPLE QUANTITATIVE METHOD [Analyst, VOl. 92 TABLE I11 TABLE OF MEDIAN VALUES OF OPTICAL DENSITY FOR SEVERAL AMINO-ACIDS Optical density A r pmoles-2 0.1 0.5 1-0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 10.0 12.0 13-0 15.0 17-0 20.0 22.0 25.0 Alanine 0.005 0.027 0.105 0.210 0-257 0.360 0.430 0-504 0-575 0.647 0.711 0.790 0-843 1-092 1.365 1.730 1.745 1-882 Aspartic acid 0.018 0.031 0.035 0-067 0.078 0-152 0.205 0.235 0.260 0.315 0.407 0-485 0.525 0.645 0-695 0-817 0.900 1.065 Arginine 0.015 0.049 0.100 0.188 0.280 0.360 0.450 0.530 0-644 0-653 0.950 1.044 1.215 1.285 1.476 1.796 1.870 1.925 Glutamic acid 0.014 0.052 0.180 0.284 0.300 0.380 0.470 0-543 0-620 0.670 0.860 0.990 1.100 1-242 1.387 1.665 1.860 2.090 G1 ycine 0.002 0.071 0.078 0.110 0-13e 0.234 0.190 0.200 0.205 0.230 0.235 0.267 0.325 0.340 0.390 0.430 0.480 0.540 Leucine 0.012 0.035 0-208 0.300 0.400 0.464 0.500 0-568 0.645 0.680 0.835 0.900 0.990 1-140 1-240 1.420 1-495 1.740 Lysine 0.023 0.048 0.105 0.190 0.268 0.350 0.385 0.51 1 0.635 0.700 0.900 1.050 1-151 1-27 1 1.450 1-745 1.950 2.115 Valine 0.028 0.057 0.105 0.184 0.23 1 0.323 0.41 1 0.453 0.548 0.691 0.738 0.839 0.950 1.100 1.240 1.410 1.582 1.840 The method has given consistent results as can be seen from Table IV.The percentage variation in the results was within +lo per cent., a variation similar to that obtained by Atfield and M0rris.l The method was applicable to the determination of smaller amounts of amino-acids, from 20 pg down to less than 2 pg in certain favourable cases, than the 40 to 10 pg used by Atfield and M0rris.l TABLE IV VARIATIONS I N READINGS OF OPTICAL DENSITY FOR GIVEN AMOUNTS OF GLYCINE AND LYSINE Optical density I 3 Glycine Lysine pmoles-2 & - 1.0 0.080 0.072 0.078 0.108 0.105 0.107 5-0 0.190 0.197 0.190 0.383 0.385 0.400 10.0 0.230 0.220 0.240 0.900 0.910 0.890 20.0 0.405 0.430 0.440 1.750 1.740 1.745 25-0 0.537 0-540 0.540 2.115 2.065 2.185 TABLE V RECOVERY OF VARIOUS AMINO-ACIDS AT DIFFERENT STAGES OF ANALYSIS Amount applied to paper (-Ap, Recovery, Amino-acid pmoles-2 CLg per cent.Procedure Aspartic acid . . 5.0 6.6 97 One-dimensional chromatography Alanine . . .. 5.0 4.4 97 One-dimensional chromatography 10.0 13.3 96 10.0 8.9 93 15.0 13-4 98.5 20.0 17.8 99.5 Alanine . . . . 50.0 44.8 100 Direct staining 50.0 44.8 99.7 Two-dimensional chromatography Arginine . . . . 5.0 8-7 98 One-dimensional chromatography 10.0 17.4 98 15.0 26.1 98-5 20.0 34.5 95October, 19671 FOR THE DETERMINATION OF SMALL AMOUNTS OF AMINO-ACIDS another. be determined at 1 pg (see Table 111, 1 pmole-2 == 1 pg). experiments, were shown by staining to be well defined and compact.97 per cent. or more, of individual amino-acids was obtained (see Table V). 633 The lower limit of quantitative determination varied slightly from one amino-acid to For example, glycine was difficult to determine below 3 pg, but leucine could easily The amino-acid spots, after development in the solvent system used in the present Further, good recovery, Results are given in Table VI for the analysis of the model peptide, alanylglycine. TABLE VI RECOVERY OF AMINO-ACIDS FROM A MODEL PEPTIDE (ALANYLGLYCINE) AFTER HYDROLYSIS Alanine, pmoles-2 Glycine, pmoles-2 - - Sample (A) (*) (A) (B) Theoretical amount . . 6.43 16.43 6.43 16.43 Optical density . . .. 0.500 1.280 0.182 0.280 Experimental amount . . 6.3 16.6 6-1 16.7 Percentage recovery . . 98 101 94 101.5 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Atfield, G. K., and Morris, C. J . 0. R., Biochem. J . , 1961, 81, 606. Heilman, J., Barollier, J., and Watzke, E., Hoppe-Seyler’s 2. physiol. Chem., 1957, 309, 219. Heathcote, J. G., and Washington, R. J., Nature, 1965, 207, 941. -- Chem. & Ind., 1963, 909. Smith, I., Editor, ‘‘Chromatographic and Electrophoretic Techniques,” Second Edition, W. Heine- Heathcote, J. G., and Jones, K., Biochem. J., 1965, 97, 15P. Opienska-Blauth, J ., Kowalska, H., and Pjetrusiewicz, M., Annis Univ. Mariae Curie-Sklodowska, Kemble, A. R., and MacPherson, H. T., Biochern. J . , 1954, 56, 548. Wieland, T., Fortschr. chem. Forsch., 1949, 1, 211. Stein, W. H., J . Biol. Chem., 1953, 201, 45. mann and Co., London, Volume 1, 1960. 1957, Section D . l l , 175. Received March 18th, 1966

ISSN:0003-2654    

DOI:10.1039/AN9679200627

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

634 Analyst, October, 1967, Vol. 92, pp. 634-638 Thin-layer Chromatography of Polyglycerols BY M. S. J. DALLAS AND M. F. STEWART ( Unilever Research Laboratory, The Frythe, Welwyn, Herts.) A simple thin-layer chromatographic technique is described for separating commercial polyglycerol products into their various components. Two slightly different procedures have been developed, the first of which is suitable for carrying out separations according to molecular weight or chain- length, and the second for carrying out separations of branched and cyclic isomers of the lower molecular-weight polyglycerols. POLYGLYCEROLS, which consist of chains of glycerol molecules joined by ether linkages, are prepared commercially by the alkali-catalysed thermal dehydration of glycerol. This process yields a mixture of components with varying chain-lengths, the average chain-length depending on the temperature and duration of heating.The usual methods of production control involve the measurement of a bulk property, e.g., the refractive index or hydroxyl value, and are thus of little value if detailed information about the composition of the mixture is required. The fatty-acid esters of these polyglycerol mixtures constitute a versatile class of emulsi- fiers, covering the entire range of hydrophilic - lipophilic balance. Extensive feeding trials1 ,2 have shown that the esters are completely non-toxic and their use as food additives has been approved by the World Health Organisation and Food and Agricultural Organisation Expert Committee, and is permitted in the United Kingdom3 and the United States4 As the poly- glycerol esters are becoming more widely used, particularly in the food5 and pharmaceutical6 industries, it is desirable to have a method of characterising them.Although the degree of esterification and the nature of the fatty acids present can be determined by using routine procedures, satisfactory methods for the determination of the detailed composition of polyglycerol mixtures are not as readily available. Because primary hydroxyl groups are more reactive than secondary groups, loss of water between the primary hydroxyl groups of two molecules to form a linear polyglycerol, I, may HOCH z.EH(OH) .CH20.CHz.iH(OH).CHzOH CH2OH I CHO.CH~.~H(OH).CHIOH I CHzOH I IIa 7H2OH CHzOH H ~ C /O \~H.CH~OH HzC/’\CH 2 I * HOCHz.HC, ,CH.CHzOH CH-0-LH I I kHz0H bH20H HOCH2.Hi CH z ‘O/ 0 IIb ma m be expected to be the dominant reaction in the commercial preparation of polyglycerols. On the same basis, the formation of IIa (“branched” polyglycerol) should be less likely than that of I, and that of IIb less likely than that of IIa. The cyclic compound, IIIa, could be formed from IIa, but IIIb might be formed from either I or IIb.Before investigating any methods of analysis it was, therefore, necessary to prepare authentic samples of poly- glycerols of the types I, I1 and I11 to act as standards.DALLAS AND STEWART 635 The synthetic routes used in the preparation of the three diglycerols are described later. Type IIb would normally be expected to be formed only in very small amounts and has not yet been synthesised in the pure form.As it is symmetrical there are no isomers of it, optical or otherwise, but it may be mentioned that there are two asymmetric carbon atoms (marked *) in type I, one in type IIa and two in types 111. Upon further condensation with glycerol, the linear diglycerol, I, may be expected to give mainly type I (linear) triglycerol, but diglycerol, IIa, whose primary hydroxyl groups are not equivalent, should give two isomeric branched triglycerols by condensation only of primary hydroxyl groups. Similar addition of a further glycerol unit to IIb, IIIa or IIIb would be straightforward, but addition of a fourth unit would be possible at two different positions. Thus, even when addition by primary alcohol condensation only is considered, the number of possible isomers increases rapidly with increasing number of glycerol units.If the very minor components formed were not disregarded the analysis would be extremely difficult. The objective has therefore been to separate as many as possible of the normal linear polyglycerols, type I, and then most of the isomers of di- and triglycerol. Paper chromatography has been quite extensively applied,' 3899910311 and the method of Zajic8 has given satisfactory results in this laboratory, but the greater speed, resolution and flexibility of thin-layer chromatography has made the latter technique worthy of careful attention. Good separa- tions have also been achieved here by using a gas - liquid chromatographic procedure, which is the subject of a separate publication.12 The thin-layer method of Seherl3 on kieselguhr impregnated with 0.66 per cent.sodium acetate has been found to be good, but not to be easily reproducible. A more reproducible method was therefore sought and has now been developed. The first of two similar procedures described below is designed to give a satisfactory separation of the higher linear polymers, and the second to give a satisfactory separation of the branched and cyclic isomers of the lower polyglycerols. The chromatography of polyglycerols is similar t o that of sugars. ROUTES USED IN THE SYNTHESES OF THE REFERENCE COMPOUNDS- Attention was first directed to the synthesis of the linear polyglycerols, I, because these were expected to be the major components of the polyglycerol mixtures.A survey of the literature revealed that, although various polyglycerol fractions had been isolated from commercial mixtures by fractional distillation of derivatives,l* 9 1 5 9 l 6 3 1 7 the only reported syntheses of individual linear polyglycerols by unambiguous routes were those of diglyceroll* and triglycerol.20 These syntheses involved the hydroxylation of 1-0-ally1 and 1,3-di-0- ally1 glycerol, respectively, and could not be readily extended to the synthesis of linear tetra- and higher polyglycerols. We have found a series of reactions involving the condensation of a suitably protected P-toluenesulphonate with the sodium alcoholate of 1,2-isopropylidene glycerol, IV (R = Na), to be a convenient method of ascending the linear polyglycerol series.Thus, linear 1,l'-diglycerol was synthesised by the condensation of IV (R = Na) with 1,2-isopropylidene-3-(~-toluenesulphonyl)-glycerol, IV (R = @-CH,.C6H,So2-), followed by removal of the protecting isopropylidene groups. This linear diglycerol was then converted into the derivative, V, by successive treatments with trityl chloride, benzyl chloride, hydro- chloric acid (to remove the trityl groups) and $-toluenesulphonyl chloride. The condensation of V (1 mole) with IV (R = Na) (2 moles), followed by removal of the protecting groups, yielded linear tetraglycerol. Linear hexaglycerol, octaglycerol, etc., are being prepared by subjecting the linear tetraglycerol to the same sequence of reactions. Similarly, linear tri- glycerol was synthesised by the condensation of IV (R = Na) (2 moles) with 1,3-di-(p-toluene- sulphonyl-2-0-benzyl glycerol (1 mole), followed by removal of the protecting groups.This series is being extended to the synthesis of linear pentaglycerol, heptaglycerol, etc. I H2C -0 -CH2 CsHs.CH2.OHC CHO.CHz.CsHs I I I I CHzOR p - CH&H4.S02.OHk CH2O.SO&H4.CH3 - p Iv V636 DALLAS AND STEWART : THIN-LAYER [Analyst, Vol. 92 Branched 1,2'-diglycerol, IIa, was prepared by the Woodward and Brutcher hydroxyla- tion21 of 2-0-ally1 glycerol. The cyclic diglycerol, IIIa, which has been isolated from glycerol chlorohydrin preparations,22 was synthesised by using a modification of previously described methods.23 524 A detailed description of the experimental procedures used in the synthetic work will shortly be published elsewhere.EXPERIMENTAL Of the two procedures described below, (b) is recommended for the separation of the higher linear polyglycerols and (c), for the separation of the isomers of the lower polyglycerols (see Fig. 1). PROCEDURE- (a) GeneraLCoat 20 x 20-cm glass plates with about an 0.25-mm layer of the relevant adsorbent, leave them in a horizontal position for about 1 hour, then dry at 110" C for half an hour and store over anhydrous calcium chloride until required. Make up 2 per cent. w/v solutions of the test samples in ethanol and apply 1 or 2-p1 portions as small spots on a line 2 cm from the bottom edge of the chromatoplate. Make up the relevant developing solvent accurately and place a suitable amount in an all-glass tank, which should be lined with chromatographic paper.Develop the chromatograms by ascending technique at room tem- perature and allow 1 hour for the solvent to evaporate from the plate, before spraying with detecting reagent. (b) For separating higher linear polymers-Coat the plates with a slurry of 15 g of kiesel- guhr G (E. Merck, 8129) plus 15 g of silica gel (Whatman SG41) in 60ml of 0-5 per cent. aqueous sodium metabisulphite (Na2S,05) and dry as described in (a) above. Apply the samples as described above and develop the chromatogram in ethyl acetate - isopropyl alcohol - acetone - methanol - water (50 + 15 + 15 + 4 + 16). The development over 15 cm has normally taken 75 to 100 minutes, depending upon the grade of adsorbent and the temperature. A second development in the same direction gives a moderate improvement in the separation of the higher polymers.(c) For separating non-linear from linear isomers of the lower polyglycevols-Coat the plates with a slurry of 15 g of kieselguhr G (E. Merck, 8129) plus 15 g of silica gel (Whatman SG41) in 60 ml of 0.045 M calcium chloride, then dry them and apply the polyglycerol samples as described in (a) above. Develop the chromatogram in ethyl acetate - isopropyl alcohol - water (55 + 30.5 + 14.5). A single development over 15 cm has normally taken 75 to 100 minutes. Three developments in the same direction are worthwhile in this instance (compare plates B and C in Fig. l), the solvent being allowed to evaporate at room temperature after each development. (d) Detection-Spray the plate (from either procedure (b) or (c) above) fairly heavily with a fresh 0.5 per cent.solution of thymol in ethanol - sulphuric acid (95 + 5), then heat in an oven at 120" to 125" C for 1 hour. Polyglycerols (0-5 pg or more) show up initially as purple - black spots on a mauve background that fades on standing. The above method, which was applied to sugars by A d a ~ h i , ~ ~ has been found the best on the basis of sensitivity, reproducibility and permanence. Certain other methods, among many tested, also gave satisfactory results. The spots could also be detected on the calcium chloride impregnated plates at room temperature with either a 0.16 per cent. solution of potassium permanganate in acetone (immediate yellow spots on a mauve background) or with a 1 per cent. solution of lead tetra-acetate in dry benzene (pale orange-yellow spots on a brownish orange background up to 24 hours after with 1,2-diols).On sodium hydrogen sulphite impregnated plates, a modification of the method described by Akita and Ikekawa26 gave clear permanent spots (slate-blue spots on a dull green background after spraying with a solution of 0-5 per cent. potassium permanganate in acetone, followed, after 15 minutes, by a 0.2 per cent. solution of bromophenol blue in ethanol). RESULTS AND DISCUSSION Typical separations obtained with the above procedures are illustrated in Fig. 1. In general, temperature, humidity and the ratio of kieselguhr to silica gel did not appear very critical, but it was found necessary to make up the solvent accurately with analytical-grade1 2 3 4 5 6 7 8 9 1 0 I I 12 13 14 15 A B C Fig.1 . Separation of polyglycerol by thin-layer chromatography on a 1 -k 1 mixture of kieselguhr and silica gel. lYate A, by procedure ( b ) ; plate B, by procedure (c) and with single development; plate C, also by procedure (c), but with three developments in the same direction and with the same solvent. Strips 1, 6, antl 11, 1 , 2’-diglycerol, I I a , (5 pg) +- cyclic diglycerol, IIIa, (10 p g ) ; strips 2, 7, and 12, sample (20 pg) of polyglycerol prcparcd in the laboratory; strips 3, 8, antl 13, glycerol + synthetic linear di, tri and tetraglycerol (5 pg of each); strips 4, 9, and 14, sample (20 pg) of typical commercial polyglyccrol; and strips 5 , l O and 15, pure glycerol (6 pg).Cyclic diglycerol, IIIa, comes well above glycerol on each plate and cyclic diglycerol, IIIb, not illus- trated, separates just below cyclic diglycerol, IIIa, on thc calcium chloride impregnated plates [To face p. 636October, 19671 CHROMATOGRAPHY OF POLYGLYCEROLS 637 reagents and, where relevant, to incorporate the correct amount of calcium chloride into the adsorbent layer. It may be added that there is some indication that the layer thickness is fairly critical for a given load per spot. As with most thin-layer chromatographic separations, care had to be taken to ensure good liquid - vapour equilibrium in the developing tank. Strips 1 to 5 illustrate separations on a sodium hydrogen sulphite impregnated layer by using procedure ( b ) ; in the linear polyglycerol series it may be seen that glycerol itself (see strip 5) has the highest R, value.By using such a layer, separation is mainly according to the number of glycerol units in the molecule, the separation of the acyclic isomers from the linear molecules being much less than on a calcium chloride impregnated layer. The resolution of a typical commercial polyglycerol (strip 4) becomes rapidly worse above what is inferred to be heptaglycerol and is probably caused by the increasing abundance and variety of non-linear isomers as the molecular weight increases ; with 30-cm development (about 5 hours) in tall tanks it was just possible to distinguish a series of ten spots, starting from glycerol. Procedure ( b ) , therefore, gives a good indication of the molecular-weight range in a sample of polyglycerol. As, however, the physical and physiological properties of a polyglycerol mixture are likely to depend also on the relative amounts of non-linear isomers present, procedure (c), too, must be considered necessary for the analysis of a polyglycerol sample.Separations with procedure (c) are illustrated by strips 6 to 15 in Fig. 1. The advantage of the multiple development technique described by Starka and Hamp12’ can be seen by comparing plates B and C. Strips 7 and 12 show a sample of polyglycerol prepared in this laboratory; when the latter strips are compared with strips 9 and 14 (typical commercial polyglycerol), the difference in relative abundance of certain isomers is clearly seen. The identity of the spot visible just above glycerol on strip 12 is still being sought; the spot also on strip 12 and just below the position of cyclic diglycerol, IIIa (strip l l ) , has the same RF value as cyclic diglycerol, IIIb. Of the two spots lying below glycerol, but above linear diglycerol, the lower one has the same RF value as 1,2’-diglycerol, IIa.The rBle of the sodium hydrogen sulphite is not certain; its ability to form addition com- pounds with carbonyls is believed to be the reason for its effectiveness in the chromatography of sugars.25 Nor is the r61e played by the calcium chloride known; the fact that sodium acetate also improves the resolution of the isomers suggests that pH is not responsible, a conclusion supported by results on pH-gradient plates prepared with a gradient spreader.Incorporation of boric acid into the adsorbent has not been found to have any advantage. A single procedure giving good resolution of the isomers, as well as of the higher polymers, has not been found possible. Further research here is now being directed towards the development of a quantitative thin-layer chromatographic procedure and studies will be undertaken to see how the quantitative gas - liquid chromatographic results compare with those obtained by such a procedure. The authors are indebted to Miss M. McMullin, Miss G. Good and Mr. D. Knights for their assistance with the experimental work, and to Dr. I. P. Freeman for providing a sample of the cyclic diglycerol, IIIb. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES Hodansky, M., Herrmann, C.L., and Campbell, K., Biochem. J., 1938, 32, 1938. Babayan, V. K., Kaunitz, H., and Slanetz, C. A., J . Amer. Oil Chem. SOC., 1964, 41, 434. “Emulsifiers and Stabilisers in Food Regulations,” Statutory Instrument No. 720, H.M. Stationery Federal Register, July 2, 1963, p. 6783; March 19, 1963, pp. 2675 and 2833; FdENgng, 1963,35, Nash, N. H., and Babayan, V. K., Bakers’ Dig., 1963, 37, 72. Rabayan, V. K., Kaufman, T. G., Lehman, H., and Tkaczuk, R. J., J . SOC. Cosmet. Chem., 1964, Wurziger, J., and Gebauer, W., Fette Seifen Anstr-Mittel, 1961, 63, 523. Zajic, J., Sb. Vys. s k . Chew.-Technol. Praze, 1962, 6 (3), 179. Wurziger, J., and Gebauer, W., Brot Geback, 1962, 16, 209. Tustanowski, S., Nowicki, R., Nowicka, I., and Zielinski, A. Z., Chemia Analit., 1964, 9, 623. Hartman, L., J . Chrounat., 1964, 16, 223. Barrett, C. B., Sen, N., and Keating, M., J . Gas Chromat., 1967, 5, 269. Seher, A., Fette Seifen Anstr-Mittel, 1964, 66, 371. Office, London, 1962. (91, 28. 15, 473.638 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. DALLAS AND STEWART Rangier, M., C.R. Hebd. Se'anc. Acad. Sci., Paris, 1928, 187, 345. Instin, M., Annls Fac. Sci. Marseille, 1940, 13, 5 ; Chem. Abstr., 1947, 41, 2392. Wright, H. J., and Du Puis, R. N., J . Amer. Chem. SOC., 1946, 68, 446. Wittcoff, H., Roach, J. R., and Miller, S. E., Ibid., 1947, 69, 2655. Siegel, H., Bullock, A. B., and Carter, G. H . , A%aZyt. Chem., 1964, 36, 502. Wittcoff, H., Roach, J. R., and Miller, S. E., J . Amer. Chem. SOC., 1949,71, 2666. Roach, J. R., and Wittcoff, H., Ibid., 1949, 71, 3944. Woodward, R. B., and Brutcher, F. V., Ibid., 1958, 80, 209. Gibson, G. P., J . SOC. Chem. Ind., Lond., 1931, 50, 949. Battegay, M., Buser, H., and Schlager, E., C.R. Hebd. Siaanc. Acad. Sci., Paris, 1929, 188, 796. Summerbell, R. K., and Stephens, J. R., J . Amer. Chem. SOC., 1954, 76, 6401. Adachi, S., J . Chromat., 1965, 17, 295. Akita, E., and Ikekawa, T., Ibid., 1963, 12, 250. Starka, L., and Hampl, R., Ibid., 1963, 12, 347. Received February 24th, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200634

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, October, 1967, Vol. 92, $$. 639-641 639 The Collection of Fractions Separated by Gas - Liquid Chromatography Part II.* The Direct Transfer of the Fraction from the Trap to a Silver Chloride Infrared Cell or a Nuclear Magnetic Resonance Spherical Microcell BY I. A. FOWLIS AND D. WELT1 (Unilevev Research Laboratory, Colworth House, Shavnbrook, Bedfod) Gas-chromatographic fractions are trapped in column packing. They are then eluted directly into silver chloride infrared microcells or nuclear magnetic resonance spherical microcells immersed in liquid nitrogen. A gas flow through the heated trap is caused by the condensation of the argon in the cells. A closed gaseous system is used. THE introduction of commercial silver chloride infrared cells has simplified the technique described by Howlett and Weltil for recovering samples trapped from a gas chromatograph for examination by infrared spectroscopy.The fractions are trapped, as before, in tubes packed with column-packing material, but they are now eluted directly into silver chloride cells cooled in liquid nitrogen. The examination of small amounts (about 50 pg) is easier and more reliable. Alternatively, the fractions are eluted directly into cooled nuclear magnetic resonance spherical microcells. EXPERIMENTAL PROCEDURE- Infrared-The apparatus (Fig. 1 (a)), which consists of a 100-ml round-bottomed flask containing desiccant (5 g of silica gel), a vacuum gauge (0 to 30 inches of mercury) and a stopcock, is evacuated to a pressure of 25 inches of mercury and then re-filled with argon.Although vacuum grease is used in the stopcock and on the flask connection, no grease should be used on the trap connections. A Research and Industrial Instrument Co. silver chloride infrared cell is fitted with a silicon rubber cap, and a trap containing 5 per cent. Apiezon L. on Celite column packing is pushed tightly into a hole bored in the cap. A heater is placed round the trap, which is then attached to the evacuated system. When the trap temperature reaches 200" C, a Dewar flask containing liquid nitrogen is raised up to the cell until the latter is almost completely immersed. Condensation of the argon and air in the cell produces a flow of gas down through the trap. No interfering hydrocarbon impurities will be produced if the trap has been pre- treated at 200" C.The chromatographic fraction is eluted from the trap and re-condensed in the infrared cell. The vacuum in the system reaches equilibrium at a pressure of between 15 and 20 inches of mercury and is allowed to remain at this level for 5 minutes, the liquid nitrogen level being maintained the whole time. The heater is then switched off and the liquid nitrogen allowed to evaporate. The liquid argon evaporates slowly until atmospheric pressure is again attained, the solute remaining on the cold wall of the cell. This technique works equally well with air or argon, but the use of argon reduces the risk of oxidation by liquid oxygen. * For details of Part 1 of this series, see reference list, p. 641.640 FOWLIS AND WELTI: COLLECTION OF FRACTIONS SEPARATED [Anahst, Vol.92 The infrared cells of 0-025-mm path-length hold 1 pl of liquid and the 0-01-mm cells, about 0.2 p1. As the former are easier to fill than the latter, it is preferable to use the 0.025-mm cells and fill them with an infrared-transmitting solvent to avoid the interference fringes caused by a partly empty cell. However, if the sample is small and needs to be examined as a liquid, or is immiscible with infrared solvents, then it is preferable to use a 0-01-mm cell. A = 100-mi flask E = Silver chloride infrared cell B = Vacuum gauge (0 t o 30 F = Nuclear magnetic resonance C = Trap G = Liquid nitrogen D = Heater inches of mercury) micros p he re Fig. 1. Apparatus ( a ) , for recovery into silver chloride infrared cell; ( b ) , for recovery into nuclear magnetic resonance spherical microcell If, after removing the trap from the rubber cap an infrared solvent can be used, the solute is washed down with 1 p1 of the solvent by using a Hamilton syringe.A B5 stopper is fitted into the cap and the cell placed into its holder. The holder and cell are fitted into a polystyrene moulding and spun in a centrifuge for 2 minutes. If no solvent is used, the cap is stoppered and the cell at once spun in a centrifuge. The cell and holder are then placed into the optimum position of the light beam of the spectrometer. Solvent absorption in 0-01 and 0.025-mm cells is generally so weak that no reference solvent cell is needed. Bands like those of carbon tetrachloride a t wavenumbers 700 to 800 cm-1 are shown strongly, but no more information is lost than with a balanced cell system.The cell stoppers must be such as to prevent evaporation of the sample in the light beam. The cells are emptied by inversion in a centrifuge spinning at 4000 r.p.m. for several minutes. Nuclear magnetic resonance-The apparatus is the same as that used for infrared cells. The spherical microcell is attached by a piece of silicone rubber tubing to the underside of a B5 socket that has been drawn out into a tube of the same diameter as the stem of the microcell (Fig. 1 ( b ) ) . The two pieces of glass are made to touch to reduce the amount of exposed rubber to a minimum.October, 19671 BY GAS - LIQUID CHROMATOGRAPHY. PART I1 641 The experimental procedure is similar to that used for the infrared cell, except that with the cell of smaller volume less gas is condensed and, therefore, the vacuum has not exceeded 15 inches of mercury.The cell is not spun in a centrifuge, the sample being washed down into the sphere with 50 p1 of carbon tetrachloride containing a reference standard. The sample is then examined in the way described by Frost, Hall, Green and Leane.2 W c W n 0 3 E I I I I I I I I . I -7- I 650 Wavenumber, cm-1 Fig. 2. Infrared spectra of ethyl hexanoate after collections from gas chro- matograph with 10 : 1 splitter: curve A, 0.2 p1; curve B, 0.1 pl; curve C, 0.05 p1. Silver chloride,cell, path-length, 0.025 mm RESULTS AND DISCUSSION Infrared spectra obtained with a Unicam SP200 from 0.2, 0.1 and 0.05-pl samples of ethyl hexanoate are shown in Fig.2. Recoveries calculated on the basis of the carbonyl band intensity are nearly 100 per cent. for samples injected directly into a trap and about 95 per cent. for samples collected as fractions from a gas chromatograph. Recoveries of 0.2-pl samples of eugenol, either injected directly into a trap or collected from a gas chromatograph, were both 90 per cent. The elution of high boiling alcohols or phenols, such as eugenol, from traps packed with Apiezon L. on Celite is more difficult because of band tailing. It is therefore important that the traps are thoroughly stripped before being used again. Nuclear magnetic resonance spectra of trapped and normal samples of 0-2 p1 of ethyl- benzene have been obtained as an accumulation of 274 scans on a Northern Instrument NS 544 computer of average transcients, with 5 per cent. trifluoroacetic acid as the triggering material on a Perkin-Elmer R10 60m/c spectrometer. The recovery efficiency was about 95 to 100 per cent. The authors thank Mr. D. Frost for his advice on nuclear magnetic resonance spectro- metry. REFERENCES 1. 2. Howlett, M. D. D., and Welti, D., Analyst, 1966, 91, 291. Frost, D., Hall, G. E., Green, M., and Leane, J. B., Chem. & Ind., 1967, 116. NOTE-Reference 1 is to Part I of this series. Received March 31st, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200639

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

642 Analyst, October, 1967, Vol. 92, pp. 642-644 The Determination of 4-Aminobiphenyl in Refined Aniline BY D. A. REILLY (Inzpevial Chemical Industvies Limited, Dyestuffs Division, Blackley, Manchester 9) A method has been developed for the determination of from 2 mg per litre, upwards, of 4-aminobiphenyl in refined aniline. The bulk of the aniline is separated by distilling the sample in the presence of paraffin wax, which retains 75 per cent. of the 4-aminobiphenyl. An acid extract of the distilla- tion residue is diazotised and coupled with sodium 2-naphthol 3, B-disul- phonate (R-salt) . The dyestuffs resulting from 4-aminobiphenyl and residual aniline are separated by circular-paper chromatography, and the former is determined by visual comparison of the corresponding ring on the chromatogram with standards prepared from aniline containing known concentrations of 4-aminobiphenyl.A METHOD has been published by Stagg and Reed1 for the determination of 4-aminobiphenyl in technical diphenylamine. When applied to the determination of this impurity in refined aniline, the procedure was found to be unsatisfactory because the concentration of R-salt was too low in relation to the weight of aniline taken, which resulted in the production of 4-aminobiphenyl from the diazotised aniline at the coupling stage, thus giving falsely high values of up to 1 per cent. for its concentration in the aniline. Norman and Vaughan2 made similar criticisms and modified the method by increasing the concentration of R-salt 10-fold thereby eliminating the production of 4-aminobiphenyl at the coupling stage.Their modified method would detect as little as 20 mg per litre of 4-aminobiphenyl in aniline. EXPERIMENTAL A more sensitive method has been developed in this laboratory, in which a mixture of 100 ml of aniline and 1 g of paraffin wax was distilled until only 0.5 to 1 g of the aniline remained in the distillation flask. Under these conditions about 75 per cent. of any 4-amino- biphenyl present was retained in the distillation residue. The residue was dissolved in toluene and amines were extracted with dilute hydrochloric acid. A portion of the acid extract was diazotised and coupled with R-salt, by using a concentration of R-salt twice that recom- mended by Stagg and Reed for diphenylamine. The coupled solution was subjected to circular- paper chromatography and the intensity of the ring due to 4-aminobiphenyl compared visually with the rings on chromatograms prepared from high quality aniline containing known added concentrations of 4-aminobiphenyl. METHOD APPARATUS- Distillation unit-This consisted of a 250-ml, short-necked, round-bottomed flask, with a B24 joint, sloping recovery bend and a long (about 80 cm) air condenser.100-ml Separating funnels with short stems and ground-glass stoppers. Plates for circular-paper chromatography-Two squares of plate glass about 25 x 25 cm x 0-5 cm thick, one of them with a central hole about 1 cm in diameter. Solution reservoirs for chromatographic plates-Capable of holding 20 ml of liquid. Six-inch straight absorption tubes of the type normally used to hold calcium chloride or soda lime are suitable.The stem of each passes through a rubber bung that fits tightly into the hole in the upper of the pair of chromatographic plates (see Procedure). REAGENTS- Whatman No. 1 jilter-papers, 24 cm in diameter. Parafin wax, melting-point 51" to 54" C. Toluene, laboratory-reagent grade. Hydyochloric acid, N.REILLY 643 4-Amino biphenyl. Sodium nitrite sol.ution, 0.5 N. Sodium carbonate solution, M. R-salt solution, approximately 0.05 N-weigh into a beaker 18.0 & 0.5 g of R-salt (di- sodium-2-naphthol-3,6-disulphonat e, M. W. 348) of analytical-reagent grade, dissolve it in 500 ml of hot water and make the solution slightly alkaline to Brilliant Yellow with sodium carbonate solution (M).Cool to room temperature, filter into a 1-litre standard flask, dilute to the mark with water at 20” C and mix well. Store in an amber bottle. PROCEDURE- Weigh 1.0 * 0.1 g of paraffin wax plus a small piece of porous pot into the distillation flask. Add 100 1 ml of the sample, measured carefully in a clean, dry measuring cylinder. Attach the recovery bend, wrapped in asbestos tape or string, and the air condenser. Heat the flask, preferably electrically, and maintain a fairly rapid rate of distillation until the flow of aniline vapour into the condenser is seen to diminish and slight fuming is visible in the flask. Turn off the heat and allow the flask to cool until it can be held comfortably in the hand, but do not allow the residue to solidify. Detach the flask, add about 20 ml of toluene, swirl it gently and pour into a 100-ml separating funnel.Rinse out the flask with further portions of toluene and add these to the funnel, making a total volume of about 80 ml. Add 15 ml of N hydrochloric acid to the separating funnel, stopper and shake it for 1 minute. Allow the layers to separate and run off the lower aqueous layer into a second 100-ml separating funnel. Repeat the extraction of the toluene layer with two further 15-ml portions of N hydrochloric acid, again running the separated acid layers into the second separating funnel. Add 15ml of toluene to the combined acid layers, shake the mixture for 1 minute and allow it to separate. Run the acid layer into a 50-ml standard flask, dilute to the mark with N hydrochloric acid, and mix well.This is the test solution. Transfer by pipette a suitable portion, not more than 0-5m1, of the test solution into a 6 x 1-inch test-tube and add from a dropping pipette sufficient N hydrochloric acid to make the total volume 5 ml. Add 0.5 ml of sodium nitrite solution, mix by gentle swirling and allow to stand for 3 minutes at room temperature. Pour the contents of the tube into a second 6 x 1-inch test-tube containing 4ml of R-salt solution and 6ml of M sodium carbonate solution, mix by gentle swirling and allow to stand for 10 minutes at room temperature. With a cork borer cut a hole about 1 cm in diameter in the centre of ten sheets of Whatman No. 1 filter-paper, 24 cm in diameter, placed one above the other. Lay the pad of papers on the lower of a pair of chromatographic plates placed horizontally, and place the upper plate on top of the pad so that the hole in it is over the hole in the papers.Place two diametrically opposed weights on the upper plate (either Winchester quart bottles full of water or 5-inch lengths of mild steel bar of 2-inch square cross-section are suitable). Insert the solution reser- voir tightly into the hole in the upper plate. After the 10 minutes’ coupling period, add 10 ml of N sodium hydroxide to the coupled solution and stir until a clear solution is obtained. Pour this into the solution reservoir and allow it to soak completely into the pad of paper. Remove the solution reservoir, rinse free from coloured solution and replace it. Fill the reservoir with N sodium hydroxide and allow the chromatogram to develop until the solution front reaches the edge of the paper.Detach the solution reservoir, remove the paper pad and hang it up to dry. When dry, compare the chromatogram visually with standard chromatograms obtained with mixtures of aniline (free from 4-aminobiphenyl) with known concentrations of 4-aminobiphenyl. The concen- trations to be added will depend on the concentrations expected in the sample, but for an “unknown” sample 2, 5, 10 and 20mg per litre should suffice. RESULTS Three samples of aniline recently manufactured by I.C.I. Ltd., gave no visible pink ring when tested by this method. The lowest concentration of added 4-aminobiphenyl that gave a visible ring was 2 mg per litre. Concentrations of 2, 5, 10 and 20 mg per litre gave rings that could be readily distinguished from each other, and more closely spaced standards644 REILLY could probably be used.Comparison of the rings obtained from known weights of 4-amino- biphenyl dissolved directly in N hydrochloric acid and then diazotised, coupled and chromato- graphed, as described above, with those obtained from equivalent concentrations of 4-amino- biphenyl added to aniline, showed that about 75 per cent. of the 4-aminobiphenyl originally present in the aniline was retained in the distillation residue. Results for the recovery of 4-aminobiphenyl from the distillation residue are as follows- 4-Aminobiphenyl added to aniline, mg per litre . . 22 22 10 10 5 Recovered from distillation residue, equivalent mg per litre . . . . . . . . .. . . 17 to 18 17to 18 7 to 8 7 to 8 slightly less than 5 At the low concentrations present in good quality aniline, visual comparison of the test ring with standards is capable of giving sufficiently precise results. If greater precision at concentrations above 20 mg per litre is needed, then the rings can be eluted from the chromatogram by the technique of Stagg and Reed,l and the dyestuff present in the eluate measured spectrophotometrically. It is evident that no 4-aminobiphenyl is produced at the coupling stage under the conditions used. REFERENCES 1. 2. Stagg, H. E., and Reed, R. H., A+aalyst, 1957, 82, 503. Norman, O., and Vaughan, G. .4., Ibid., 1966, 91, 653. Received March 23rd, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200642

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, October, 1967, Vol. 92, $$. 645-649 645 Spectrographic Determination of Beryllium in its Minerals with a Gas-stabilised Arc BY M. D. MARINKOVIC (The Boris Kidric' Institute of Nuclear Sciences, Belgrade, Yugoslavia) AND A. M. ANTIC-JOVANOVIC (Imtitute of Physical Chemistry, Faculty of Sciences, University of Belgrade, Yugoslavia) A gas-stabilised arc in argon is described for the determination of beryllium in its minerals. The analytical procedure used involves the decom- position of the sample by fusing it with sodium fluoride, dissolving the melt, atomising the solution in an argon stream and excitation of the spray. Experimental variables that may influence the results, and the behaviour of some elements as internal standards, are investigated. When molybdenum was used as the internal standard, the coefficient of variation achieved was about 2.5 per cent.The method was checked on some natural and synthetic beryls. MANY spectrographic techniques for determining beryllium traces in various materials have been reported. However, the number of papers published on the spectrochemical determina- tion of beryllium in its minerals (as beryls, etc.) is rather Conventional chemical methods generally used for this purpose suffer because of the presence of elements with similar chemical properties, e.g., aluminium. This is a source of many difficulties. Emission- spectrographic methods depend much less on chemical properties but give far less accurate results at higher concentrations when, as a rule, they are not satisfactory.Unsatisfactory accuracy is mainly caused by the unsteady operation of the light source and unreproducible volatilisation from the solid sample, which are especially important when powdered ores and minerals are analysed by the d.c. arc technique. By applying a gas-stabilised arc, which has recently been used in different ~ a y s , ~ , ~ ~ ~ 3' 3 8 9 9 and introducing the sample into the analytical gap by atomising the solution in a stabilising stream, conditions have been established for the determination of elements at higher concentrations with satisfactory accuracy. In the present paper a method for the determination of beryllium in beryls is described that has fairly good accuracy and reproducibility. EXPERIMENTAL DESCRIPTION OF THE SOURCE- The first experiments were carried out with the stabilised arc described by MarinkoviC.1° However, this type of source was shown to be unsatisfactory because of the condensation of aerosols inside the chamber and the possibility of contamination with solutions used previously.In our further work we used an improved modification of the stabilised arc (Fig. 1). The source consists of electrically insulated segments, D, E, F, G and H. The arc burns between two graphite electrodes, B and C, the lower one, B, being situated in the chamber. Segments D, E and F are water-cooled and comprise the lower part of the chamber, into which a weak argon stream is introduced to prevent the lower electrodes from being consumed. Segments G and H comprise the upper part of the chamber into which the aerosol of the sample is tangentially introduced.This part of the chamber is not water-cooled646 MARINKOVIC AND ANTIC- JOVANOVIC SPECTROGRAPHIC DETERMINATION [klndyst, VOl. 92 and its temperature while the arc is burning is maintained above 100" C, whereby condensation of aerosol drops is prevented. Between segments D and E, and segments F and G, gaps are left for the outflow of argon and aerosol. The atomising pressure was 2 atmospheres, while the consumption of gas was at the rate of 4.5 litres per minute. The atomising rate of sample solutions was 4 ml per minute, with an efficiency of about 3.5 per cent. The arc was ignited by a high frequency spark. The holder of the lower electrode, A, is shaped so that it can be set in a de Grammon's holder, thus allowing the gas-stabilised arc assembly to be easily mounted on the standard optical bar of the spectrograph. CHOICE OF THE INTERNAL STANDARD AND WORKING CONDITIOXS- The most intense beryllium lines in our source are the spark lines beryllium(I1) 3131.07 and 3130-43 A.The former is used as the analytical line, while the latter interferes with one line of the OH band. It has been ascertained from table+ that only two elements, vanadium and molybdenum, have sufficiently intense lines in this spectral range that could be used as an internal standard. These are the molybdenum(1) 3132.59 and vanadium(I1) 3110.71 A lines. C H, .- ? ;I E' D / cm f Scale Aerosol in - - A - B and C = D, E and F = G and H = Argon in -c- Brass holder Graphite electrodes Water-cooled electrically insulated segments Upper electrically insulated segments into which aerosol is introduced I nsu lators Fig. 1.Diagram of stabilised arc To choose the internal standard and determine the optimum working conditions, a series of experiments has been carried out, and almost all of the experimental conditions that could influence the intensities of the spectral lines were varied. After considering the results of previous work,l0 which showed that the emission of lines with this type of source takes place in relatively narrow zones of the arc column, radial distribution of the emission of the above beryllium, molybdenum and vanadium lines was investigated. Fig. 2 shows a radial distribution of emission coefficients for all three lines, calculated according to Bockasten's method.12 The curves show that the beryllium and vanadium lines emit in close regions of the arc column.It could be expected from this that accidental changes of excitation conditions could be better compensated for if a vanadium line is used as the internal standard instead of a molybdenum line. However, investigationOctober, 19671 OF BERYLLIUM IN ITS MINERALS WITH A GAS-STABILISED ARC 647 I I Radius, mm Fig. 2. Radial distribution of emission coefficients: curve A, for beryllium (11) 3131.07 A line; curve B, for vanadium (11) 3110.71 ii line; and curve C, for molybdenum (I) 3132.59 4 line of the influence of elements with a low ionisation potential showed that this assumption is not correct. Fig. 3 shows the influence of sodium on the intensity ratio of the line pairs beryllium(I1) - molybdenum(1) and beryllium(I1) - vanadium(I1).The influence of aluminium has also been investigated. In the range of concentrations normally found in the samples (up to 15 per cent.) no effect on the intensity ratio of the line pair, beryllium 3131.07 A - molybdenum 3132.59 A, was detected. O 6ol 01 0.2 03 04 05 Sodium, mg per ml Fig. 3. Influence of sodium on the intensity ratio of line pairs: curve A, beryllium (11) - vanadium (11) ; curve B, beryllium (11) - molyb- denum (I) The effect of changes in the arc current is rather small and almost the same on the intensity ratio of both line pairs. When the arc current changes from 5 to 16 amperes, the intensity ratio increases by about 10 per cent. On the other hand, at low current, the in- stability of the burning arc increases. Bearing in mind the above, an arc current of 8 amperes has been chosen as a compromise, with molybdenum as the internal standard.In a further investigation the absence of self-absorption for the chosen lines was established. ANALYTICAL PROCEDURE- Preparation of samples-Beryl samples were decomposed by fusion with sodium fluoride.13 The sample was finely ground in a synthetic corundum mortar, and 100 mg were fused with 500 mg of sodium fluoride in a platinum crucible for 5 minutes. After cooling, concentrated sulphuric acid (2 ml) was added and the contents heated to a gentle melt. The heating was continued until fumes of sulphur trioxide were evolved and then for a further 5 minutes.648 MARINKOVIC AND ANTIC- JOVANOVIC : SPECTROGRAPHIC DETERMINATION [APzaZyst, Vol.92 After cooling, the contents were extracted with water (about 100ml), and the solution was boiled until it was clear and then transferred into a 500-ml calibrated flask. Fifty millilitres of standardised ammonium molybdate solution (5 g of molybdenum per litre) were added to the flask and the solution was diluted to 500 ml with distilled water. This final solution was then atomised into an argon stream. Preparation of standard solutions-Five standard solutions were prepared by diluting a beryllium sulphate solution so that the concentrations of beryllium were 2, 5, 10, 20 and 40mg per litre. Each of these standard solutions also contained ammonium molybdate (500 mg of molybdenum per litre) as an internal standard, and sodium sulphate (550 mg of sodium per litre) and aluminium sulphate (20 mg of aluminium per litre) to match the sample solutions. Aluminium is always present in beryls and sodium is introduced in large amount during the decomposition with sodium fluoride, SPECTROGRAPHIC PROCEDURE- The samples were excited and photographed under the conditions listed below- Spectrograph : Wavelength range : Slit width : Excitation source : Electrode gap : Polarity : Open circuit voltage : Current : Exposure period : External optics : Hilger quartz Lit trow.2450 to 3500A. 0.02 mm. Gas-stabilised arc (Fig. 1). 30 mm. Lower electrode anode. 240 volts. 8 amperes. 15 seconds. The 10-mm section of the arc column, just above segment H (Fig. l), is focused on the spectrograph collimator with lens F1025.The upper electrode is screened off. The densities of the spectral lines were measured on a non-recording Zeiss densitometer and converted to relative intensities by using a Respectra calculator. The analytical graph was constructed by plotting the logarithm of intensity ratio of analytical pair lines against the logarithm of concentration of beryllium (Fig. 4). The background intensity was negligible and no correction was made. I 10 I00 Log concentration of beryllium (pg per ml) Fig. 4. Analytical graph. Beryllium (11) 3131.072 A line and molybdenum (I) 3132-594 A line RESULTS AND DISCUSSION Beryllium was determined, by the method described, in several natural beryls from different Yugoslav bearings.14 The results are given in Table I.The accuracy of the method was checked by determining beryllium in three synthetic beryls prepared by a laboratory procedure developed by RistiC.15 Conditions of synthesis were chosen so as to ensure that the composition accorded with the stoicheiometric formula of the mineral. The results obtained are given in Table 11, and, as can be seen, they show good agreement between the amounts of beryllium found and calculated.October, 19671 OF BERYLLIUM IN ITS MINERALS WITH A GAS-STABILISED ARC 649 TABLE I DETERMINATION OF BERYLLIUM IN SEVERAL BERYLS FROM YUGOSLAV BEARINGS Mean value, Sample Beryllium content, pcr cent. per cent. Cer . . . . .. 4.88, 4.97, 4.90, 4-95 4-92 Prokuplje . . . . 4.72, 4.85, 4.98, 4.74, 4-69, 4.78, 4.73, 4.72 4.78 Kukavica .. . . 4.52, 4.71, 4.78, 4-79, 4.51, 4.74, 4.70, 4.86 4.70 Zheljin . . . . . . 4.87, 4-73, 4.72, 4.67, 4-55, 4.67, 4.58, 4.67 4.68 Bukulja .. .. 4.64, 4.75, 4.56, 4.66 4-65 Juhor . . . . .. 4.80, 4.72, 4.71, 4.69, 5.00, 4.82 4.79 TABLE I1 DETERMINATION OF BERYLLIUM IN SYNTHETIC SAMPLES Number of Mean beryllium content found, Theoretical beryllium content, Sample determinations per cent. per cent. I 4 5-10 5.03 I1 4 5.06 5.03 111 6 5.05 5.03 which 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. The reproducibility of this method is characterised by the coefficient of variation calcu- lated for samples photographed in duplicate on separate photographic plates and for different days. For the concentration of 10mg of beryllium per litre of solution the coefficient of variation is 2.6 per cent.It is noted that by evaluating different plates the analytical graph shows a parallel dis- placement from one determination to another. This displacement can be considerable and it is recommended that at least one standard be photographed on each plate to fix the position of the analytical graph. The method described can also be successfully used for the analysis of other beryllium minerals and ores. It is only necessary to prepare standard solutions, the composition of correspond to those of the samples analysed. REFERENCES Alekseeva, V. M., and Rusanov, A. K., Zh. Analit. Khim., 1957, 12, 23. Greitz, E. B., U.S. Bureau of Mines Report No. 5407, Washington, D.C., 1958. Kehres, P. W., and Poehlman, W. J., Appl. Spectrosc., 1954, 8, 38. Margoshes, M., and Scribner, B. F., Spectrochim. Acta, 1959, 15, 138. Korolev, V. V., and Vainshtein, E. E., Zh. Analit. Khim., Owen, L. E., Appl. Spectrosc., 1961, 15, 150. Collins, A. G., and Pearson, C. A., Analyt. Chem., 1964, 36, 787. Rieman, M., in “Proceedings of the 12th International Colloquium on Spectroscopy,” Hilger, Doerffel, K., and Lichtner, J., Spectrochim. Acta, 1966, 22, 1245. Marinkovid, M., Bull. Boris Kidric‘ Inst. Nucl. Sci., 1965, 16, 65. Meggers, W. F., Corliss, C. H., and Scribner, B. F., “Tables of Spectral-Line Intensities,” Parts I Bockasten, K., J . Opt. SOC. Amer., 1961, 51, 349. Patkor, A. J., and Varde, M. S., Indian J . Chem., 1964, 2, 123. RistiC, S., AntiC-JovanoviC, A., and JeremiC, M., 1st. Symp. Geochem., Beograd, 1965, pp. 409 RistiC, S., “Physikalisch-Chemische Untersuchungen an Beryllen in Zusammenhang mit ihren Received January 17th, 1967 1959, 14, 658. Exeter, 1965, pp. 199 to 204. and 11, Natn. Bur. Stand. Monogr. 32, Washington, D.C., 1961. to 431. Helium,” Doktordissertation, University of Mainz, 1956.

ISSN:0003-2654    

DOI:10.1039/AN9679200645

出版商:RSC    

年代:1967

数据来源: RSC