摘要:

Analyst, November, 1974, Vol. 99, pp. 739-744 739 A Kinetic Theory of Atomisation for Non-flame Atomic-absorption Spectrometry with a Atomisation for Copper Graphite Furnace. The Kinetics and Mechanism of BY C. W. FULLER (Tioxide International Limited, Billingham, Cleveland) A kinetic approach has been made to the atomisation process in non- flame atomic-absorption spectrometry. Time - absorbance profiles for the determination of copper, using a graphite furnace, have been investigated in the temperature range 1720 to 2220 K and a rate equation derived that describes the variation of the amount of copper atoms in the furnace with time. It has been shown that a slow first-order reaction involving reduction of copper oxide by carbon followed by the rapid vaporisation of the copper formed is the most probable reaction mechanism.The greater sensitivity achieved in the determination of copper using a tantalum-lined graphite furnace has been attributed in part to the greater rate of reduction of copper oxide by tantalum. NON-FLAME atomic-absorption spectrometry has undergone a large growth rate in some areas of application over the last 2 to 3 years. However, the optimisation of instrument design and the extension of the technique to a wider range of analytical problems has suffered from the lack of theoretical investigations into the technique. A popular hypothesis for atomisation in graphite-furnace atomic absorption is the process in which the metal oxide is reduced, by the carbon of the furnace, to the free metal, This process has been suggested as a possible mechanism for the losses of certain elements during pre-atomisation heating periods1 and has since been described in more detail to explain the atomisation process.2 Assuming that reduction of metal oxides by carbon is rapid, if AGO for reaction (1) is negative, Campbell and Ottaway2 proposed that the temperature at which free metal atoms are formed can be determined by calculating the temperature at which AGO for reaction (1) becomes zero. ..- - (1) MO(s) + C(S) + M(g) + CO(g> . . Applying this theory to twenty-seven elements, Campbell and Ottaway found agreement between these theoretically predicted temperatures and the experimentally observed tem- peratures in all but six or seven cases. However, despite this agreement, the thermodynamic approach does not explain the fact that several of the metals form thermodynamically stable carbides at and below the temperatures at which AGO for reaction (1) is zero, e.g., aluminium, calcium, chromium, molybdenum, titanium and vanadium. In addition, the thermodynamic approach cannot give any indication of rates of atomisation and consequently cannot predict absorbance peak shapes.This paper describes a kinetic approach to the atomisation process in flameless atomic- absorption spectrometry. Subsequent work based on this kinetic approach explains the anomalous facts that arise in the thermodynamic approach and gives a quantitative explana- tion of pre-atomisation heating losses. EXPERIMENTAL A Perkin-Elmer, Model 103, atomic-absorption spectrometer was used with an HGA-70 graphite furnace3 and a Model 165 recorder.Argon was used to provide an inert gas atmos- phere within the graphite furnace at a flow-rate of 1-5 1 min-l, controlled by the HGA-70 0 SAC and the author.740 FULLER: A KINETIC THEORY OF ATOMISATION FOR NON-FLAME [A%a&St, VOl. 99 power unit. Oxford “Sampler” micropipettes fitted with disposable plastic sampling tips were used for introducing solutions into the graphite furnace. All atomic-absorption measurements were made, using a single-element copper hollow- cathode lamp, at 324.7 nm with a 0-7-nm band width. A constant scale expansion of 2 was used throughout. Twenty-microlitre aliquots of sample were used in all instances, with a drying time of 40 s and an ashing time of 30 s using Program 4 on the HGA-70 power unit.A standard solution containing 1000 pg ml-l of copper was prepared by dissolving pure copper metal in the minimum volume of 1 + 1 nitric acid and then diluting the solution appropriately with distilled water. The voltage indicated on the HGA-70 power supply was calibrated against the tem- perature attained by the graphite furnace using an optical pyrometer. Good agreement was obtained with the calibration graph supplied by the instrument manufacturers. RESULTS AND DISCUSSION The copper system was chosen for the initial investigational work described here for three reasons: firstly, the production of copper atoms can be conveniently measured at rela- tively low temperatures (1700 to 2200 K) and the results obtained then applied to experi- mental observations at higher and lower temperatures in the graphite furnace ; secondly, copper does not form a stable carbide; and thirdly, copper has relatively low melting and boiling-points, which should simplify the kinetics of atom formation.At the temperatures used in this work, copper(1) oxide (Cu,O) is the most stable species and therefore the most likely to be present in the graphite furnace. All kinetic measurements were made at absor- bance values below 0.15 so as to ensure that under the conditions used the absorbance was directly proportional to concentration of metal atoms in the graphite furnace, i.e., measure- ments were taken on the linear portion of a calibration graph. Atomic-absorption versus time signal profiles were measured in the temperature range 1720 to 2220 K and with absolute amounts of copper in the range 2 x 10b9g to 100 x 10-9g.These conditions ensured that the time constant of the recorder was small in comparison with the rate of change of absor- bance, so that a true time - absorbance profile was obtained, and that measurable time - absorbance profiles were obtained at all temperatures investigated. Measurements were taken with at least three different initial amounts of copper at each temperature. A typical time - absorbance profile for copper is shown in Fig. 1 ; point (a) represents the time at which the automatic timer on the HGA-70 power unit starts the atomisation cycle; (b) represents the time at which the presence of copper atoms is first detected; period (a) to (c) shows the time for the graphite furnace to reach its pre-set temperature; and period (c) to (e) shows the variation in the concentration of copper atoms in the graphite furnace at constant temperature. The period (c) to (e) was confirmed to be a constant-temperature region by the use of an optical pyrometer (Fig.1). L’vov 4 has described a general kinetic example for the production of metal atoms in a graphite cuvette system in which the sample is vaporised under constantly increasing temperature conditions, but his conclusions cannot be applied to the graphite furnace, where steady temperature conditions are attained well before complete atomisation has occurred. The rate equation for the change in amount of copper atoms in the graphite furnace can be written in the form .... where R[Cu] formation represents the rate of formation of copper atoms at time t and which will be afunction of the amount of copper atoms remaining unvaporised at time t. R[Cu] removal represents the rate of removal of copper atoms from the graphite furnace and will be of the type shown in equation (3): R[C~]removal = &[Cu] - - .. .. * (3) The rate of removal will be solely dependent on the amount of copper atoms present in the graphite furnace at time t and on a rate constant, k,, which will be proportional to the rate of flow of inert gas through the graphite furnace. As the volume of gas flowing through the furnace in unit time is proportional to the furnace temperature, then k , will also be proportional to temperature.November, 19741 ATOMIC-ABSORPTION SPECTROMETRY WITH A GRAPHITE FURNACE 741 The values of absorbance against time, in the region (d) to (e) of Fig. 1, for all results were fitted to first, second and third-order rate equations, but were found to agree only with first-order kinetics.Typical graphs of log,, (absorbance) against time at various tem- peratures and initial copper concentrations are shown in Fig. 2, while calculated values for the first-order rate constant, k,, are given in Table I. I 01 40 30 20 10 Time/s 2000 I500 9 !?! --.. 1000 5 E c a) 500 Fig. 1. Variation of the graphite furnace tem- perature against time with the power supply set a t 5.0V, together with the variation of absorbance against time for 20 pl of 1.0 pg ml-l copper solution under identical conditions.Points (a) to (e) are described in the text '1 0 10 20 30 40 Ti me/s Fig. 2. Graphs of log,, (absorbance) versus time for the region (d) to (e) shown in Fig. 1 for various initial amounts of copper, [Cu],, and temperatures, T. (1) T = 2220 K, ~cu],, = 2 x 10-9 g ; (2) T = 2100 K, [CU], = 4 x 10-9 g; (3) T = 1990 K, [cU], = 10 x 10-9 g; (4) T = 1860 K, [CU]~ = 20 x lO-@g In order to obtain first-order kinetics over this region of the time - absorbance profile, it is necessary that the rate of change of concentration of copper atoms with time should be independent of the rate of removal of copper atoms from the furnace and be solely de- pendent on the rate of formation of copper atoms. The rate of formation of copper atoms will follow the first-order kinetics obtained with a rate constant k , equal to the calculated value k.TABLE I EFFECT OF TEMPERATURE AND INITIAL AMOUNT OF COPPER ON THE CALCULATED VALUES OF THE FIRST-ORDER RATE CONSTANT, k,, AND k , AND p k*l S-1 p x 10" Initial amount k,/ TemperaturelK of copper/g x loe S-1 0.34 0.29 0.83 0.78 1.18 1.29 1-47 1.44 1-18 1-84 0.8 0.8 0.9 0.9 1.0 1720 60 0.040 1720 40 0-040 18eO 20 0.072 1860 10 0.073 1990 6 0.151 1990 4 0.166 2100 6 0.238 2100 4 0.239 2220 4 0.329 2220 2 0-3 16742 FULLER: A KINETIC THEORY OF ATOMISATION FOR NON-FLAME [Analyst, Vol. 99 Theref ore , R[CU]forrnation = hi [CUIunvaporised = ki ([Culo - [CUIvaporised) - * (4) where [Cu] represents the initial amount of copper in the graphite furnace and [Cu] vaporised the amount of copper vaporised between t = 0 and t = t.Equation (4) can be integrated to give a value of [CU]vaporised after time t equal to [CU]vaporised = [cU]o (1 - e-kl') which can then be substituted back into equation (4) to give . . - - (5) [ C ~ l f o r m a t l o n = k1 [CU] 0 e-*lt .. Substitution of equations (3) and (5) into equation (2) gives the rate equation for the change in amount of copper atoms in the graphite furnace with time: .. - (6) d[Cul = k, [Cu], e-klt - k , [Cu] . . .. dt The differential equation (6) can be solved for [Cu] to give the final equation (7), which relates the amount of copper atoms in the graphite furnace to time and the initial amount of copper present : Equation (7) can be rewritten in the form of equation (8) in order to relate measured absorbance with time : Absorbance (Cu) = - k1 p [culO (e-klt - e-kzt) .. .. k , - k1 where p is a constant, dependent on temperature, which relates measured absorbance with the amount of copper atoms in the graphite furnace. Values for k, can be obtained, as described above, by plotting log,, (absorbance) against time, while values of k , and p are obtained mathematically by a process of successive approxi- mations. A graph of log,, k , vemm l/T(K) gives a linear relationship and shows that k , can be represented by equation (9) : where A = 1-1 x s-l and E = 13.8 x lo4 J. The best-fit value for k , was found t o be 4.5 x lo-* T , while the variation of p with temperature is shown in Fig. 3. Experimental values of k , and p are also shown in Table I. . . - - (9) k 1 - - Ae-E/RT .. .. . . I0 Temperature/ K Fig. 3. Variation of p with tem- perature, where p relates the measured absorbance with the amount of copper atoms present in the graphite furnace A comparison of the theoretically predicted results, using equation (8) and the calculated values of k,, k, and $, with experimentally observed results is shown in Fig. 4. Apart from the initial few seconds of the production of copper atoms when the temperature of the graphite furnace is still increasing to its pre-set value (time less than zero), the agreement is very good. The first-order rate equation (4) for the production of copper atoms is in agreement with four possible mechanisms of atom formation:November, 19741 ATOMIC-ABSORPTION SPECTROMETRY WITH A GRAPHITE FURNACE Fast (i) cu,o - ZCu(s/l) + 40, .... .. CU(S/l) - C U ( d .. .. .. .. Slow (ii) cu,o + c Fast-+ 2CU(S/l) + co . . ., .. Slow Cu(s/l) -4 Cu(g) .. . . .. .. (iii) cu,o - 2Cu(s/l) + 40, .. .. .. Cu(s/l) -- Cu(g) .. . . .. .. (iv) cu,o + c __+ ZCu(s/l) + co . . .. .. CU(S/l) - C U k ) .. . . .. .. Slow Fast Slow Fast where s, 1 and g represent the solid, liquid and gas phases, respectively. As the concentration of carbon is virtually constant throughout and in large excess over the amounts of copper present, reactions (12) and (15) will reduce to pseudo-first-order kinetics. Timefs Fig. 4. Comparison of the time - absorbance profiles for the atomisation of copper obtained experimentally with those predicted theoretically from equation (8). Solid line, experimental result, 20 x g of Cu, 1860 K; broken line, experimental result, 6 x g of Cu, 2220 K; x , theoretically predicted results, 6 x lO+'g of Cu, 2220 K g of Cu, 1860 K; 0, theoretically predicted results, 20 x In order to differentiate between the four mechanisms, it is necessary to carry out kinetic experiments under identical conditions in the presence of carbon, and in the absence of carbon but with the presence of a different reducing agent.If identical kinetic results are obtained in these two instances, then mechanism (i), (ii) or (iii) will be the most probable; if no copper atomisation occurs or atomisation occurs but with different kinetic results, then mechanisms (ii) and (iv) will be favoured. The modification was achieved by tightly fitting a thin tantalum metal lining in the centre region of the inside of the graphite furnace, but still allowing the introduction of samples through the nomal sample injection hole.Tantalum is itself used as a material for the construction of non-flame atomisers; copper atom formation should therefore still be possible with this furnace configuration. Fig. 5 shows the absorbance - time profiles obtained with the same instrument conditions and the same amount of copper but using the graphite furnace with and without the tantalum liner. The first observation is that the rate of formation of copper atoms is different in the two instances and the second is that the rate of production of copper atoms using the tantalum liner is considerably greater than that with the graphite furnace alone.744 FULLER These facts lead to the following conclusions on the mechanism of formation of copper (a) atomisation does occur through carbon reduction ; and (h) as the kinetic results are similar, then mechanism ( i v ) , i.c., the carbon reduction process, is the most probable rate-controlling step in the production of copper atoms rather than mechanism (ii), in which one would expect identical kinetic results provided that the reduction of the oxide was much faster than the vaporisation of copper for both carbon reduction and tantalum reduction.atoms in the graphite furnace: 0.40 0.30 0.10 0 I I I I 20 10 Timeis L I Fig. 5. Comparison of time - absorbance profiles for the atomisa- tion of copper (A) without and (B) with a tantalum liner fitted to the graphite furnace. Temperature, 1990 K; amount of copper, 4 x 10-9g Additionally, it is clear that a tantalum furnace would give rise to greater sensitivity than the graphite furnace for the determination of copper. This result may be due in part to the greater free energy change involved for the tantalum reduction of copper oxide, but ultimately will depend on a more favourable rate of reaction. This work is published by permission of the Directors of Tioxide International Limited. REFERENCES 1. 2. 3. 4. Fuller, C . W., Analytica Chim. Acta, 1972, 62, 442. Campbell, W. C., and Ottaway, J. M., Talanta, 1974, 21, 837. Manning, D. C., and Fernandez, F. J., Atom. Absorption Newsl., 1970, 9, 65. L'vov, B. V., Pure Appl. Chem.. 1970, 23, 1 1 . Received May 7th, 1974 Accepted July 12th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900739

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99,pp. 745-748 745 Determination of Soap in Refined Vegetable Oils by Atomic-absorption Spectrophotometry BY DINA GEGIOU (Research Department, State Chemical Laboratories, 16 A . Tsoha Street, Athens, Greece) Soap has been determined, as sodium, in alkali-refined vegetable oils by atomic-absorption spectrophotometry. The oil is treated with absolute ethanol, the mixture dissolved in ethyl methyl ketone and the solution then aspirated. Oil solutions have been compared with standards containing virgin olive oil and known amounts of sodium oleate. Concentrations of sodium oleate in the range 3 to 1000 p.p.m. in oil show a linear absorption. The method is rapid and accurate and can also be applied to the detection of adulteration of virgin olive oil with refined olive or other vegetable oils.ANALYSIS of alkali-refined vegetable oils includes the determination of soap residues for the control of refining operations. A variety of methods have been proposed for the determination of soap, among which are ashing,l extraction,2 the tentative AOCS conductivity m e t h ~ d , ~ t i t r a t i ~ n , ~ , ~ ion exchange,6 flame ph~tometry,~ atomic-absorption spectrophotometrys and neutron-activation analysi~.~ Some of these methods are time consuming or are not sufficiently accurate, others do not distinguish between soaps and free alkalinity, or between soaps and other sodium salts naturally present in vegetable oils, and facilities for neutron-activation analysis are not readily available to most analytical laboratories. Consequently, there is still a need for an accurate and fast method.We have studied the quantitative determination of soap, as sodium, by direct aspiration of a solution of the oil into an atomic-absorption spectrophotometer, the result being com- pared with that obtained with standards containing virgin olive oil and known amounts of sodium oleate. The graph of absorption versus concentration of sodium oleate in oil in the range of 3 to 1000 p.p.m. is linear. The proposed method, although also not distinguishing between soaps and other forms of sodium present, is rapid, sensitive and accurate. The determination of soap by such a method, in addition to being useful in the control of refining operations, may be of value in detecting adulteration of virgin olive oil with refined olive or other vegetable oils.EXPERIMENTAL APPARATUS- Standards and samples were analysed on an Instrumental Laboratories Inc., Model-353, atomic-absorption spectrophotometer in an air - acetylene flame. The operating conditions were as described above and in operators’ manuals for use of the instrument for sodium and organic solvent solutions. REAGENTS- Sodium oleate, 99-5 per cent. $we-Applied Science Laboratories Inc. A bsolztte ethanol-Analytical-reagent grade. Ethyl methyl ketone-Analytical-reagent grade. PROCEDURES- All glassware should be rinsed three times with distilled water and dried before use. Stock solution-Weigh into a 500-ml calibrated flask 250 mg of sodium oleate, add 125 ml of absolute ethanol, dissolve the sodium oleate by warming the mixture on a hot-plate and add ethyl methyl ketone to volume.The solution contains 500 p.p.rn. of sodium oleate (37.75 p.p.m. of sodium) and should be freshly prepared. @ SAC and the author.746 GEGIOU : DETERMINATION OF SOAP I N REFINED [Analyst, Vol. 99 Standards-Weigh into 100-ml calibrated flasks 20 g of virgin olive oil, add variable, accurately measured, amounts of the stock solution and dilute the mixtures to volume with a 25 per cent. solution of absolute ethanol in ethyl methyl ketone. Standards should always be freshly prepared. Blank solation-Into a 100-ml calibrated flask, weigh 20 g of virgin olive oil and add a 25 per cent. solution of absolute ethanol in ethyl methyl ketone to volume. Samfiles-Weigh into a 100-ml calibrated flask 20 g of an alkali-refined vegetable oil, add 20 ml of absolute ethanol, dissolve the soap present by warming the mixture gently on a hot-plate for 5 minutes, continuously shaking the flask, then add ethyl methyl ketone to volume.- RESULTS AND DISCUSSION Ethyl methyl ketone, a solvent commonly used in atomic-absorption spectrophotometry, is used to dissolve the vegetable oils to be analysed. Direct aspiration of a vegetable oil into an atomic-absorption spectrophotometer is slow and irregular because of the high viscosity of the oils. Absolute ethanol is used to dissolve soaps that are frequently found mixed with oils, because of their limited solubility in the oils, after alkali-refining operations. Black,' in his flame-photometric method, used ethane-1,2-diol for this purpose.The solubility of sodium oleate in ethane-1,2-diol is about 10 per cent., while in ethanol it is about 5 per cent. Nevertheless, as the former solvent is viscous it has to be used in rather limited weighed amounts, because small variations in viscosity produce large differences in aspiration rates. Ethanol, on the other hand, can be used in various proportions in an oil - ethyl methyl ketone solution without affecting the absorption of the solution (Table I). Thus, ethanol does not need to be accurately measured and can be used in an amount sufficient to facilitate dissolution of the soap. c 4 .k E E - 3 2 2 - Y - 2 5 -' d c1 m a .C TABLE I ABSORPTION OF STANDARD SOLUTION (50 p.p.m.) OF SODIUM OLEATE IN OIL CONTAINING VARIOUS PROPORTIONS OF ETHANOL Concentration of ethanol in oil - ethyl methyl ketone solution, per cent.10 20 30 40 50 Absorption (arbitrary units) 48.5 60 47.6 62.8 60 In Fig. 1 the absorption (in arbitrary units) and the aspiration rates (the volume aspirated, not that entering the flame, per minute) of various oil - ethanol - ethyl methyl ketone mixtures of known oil content are shown. The purpose of this experiment was to provide information about the optimum oil to solvent ratio. However, from the results obtained such an optimum ratio is not evident and an ethanol - ethyl methyl ketone solution containing 20 per cent. of I I 1 I I I I I 0 10 20 30 40 50 Oil, per cent. Fig. 1. The absorption (A, left-hand scale) and aspiration rates (B, right-hand scale) of ethanol - ethyl methyl ketone mixtures containing various proportions of oilNovember, 19741 VEGETABLE OILS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 747 oil was therefore chosen, the viscosity of which, as determined by the aspiration rate, was close to that of water.In considering the volume of oil aspirated per minute vemus the resulting absorption, it is noted that the absorption of the solution containing 10 per cent. of oil is relatively the highest observed absorption. For all the other oil solutions the absorption, compared with that of the 10 per cent. solution, is diminished by 30 to 40 per cent., probably due to the contribution of the solvent to the absorption. It is known that the fraction of the aspirated solution entering the flame depends on the viscosity and the surface tension of the solution.In addition, the organic solvent usually enhances the absorption of the elements by affecting the temperature of the flame.1° Fig. 2 shows that for sodium oleate in the range 3 to 1000 p.p.m. in virgin olive oil (0.226 to 75.5 p.p.m. of sodium) the absorption is linearly dependent on concentration, which is an advantage of the atomic-absorption method over the flame-photometric method, in which sodium emission in oil solutions containing sodium above the level of about 3 to 5 p.p.m. gives a non-proportional luminosity because of the self-absorption of the emitted light. With concentrations of sodium oleate below 3 p.p.m. in oil (levels of sodium that occur naturally in unrefined oils), the precision is poor. 2 4 6 8 10 a Sodium oleate in oil, p.p.m.Fig, 2. Graphs of the absorption of standard solutions versus sodium oleate content : A, 100 p.p.m. ; B, 10 p.p.m.; and C, 1 p.p.m. Table I1 summarises the results obtained from seven samples of alkali-refined vegetable oils, namely olive, maize and cottonseed oils, with soap concentrations ranging from 3 to 100 p.p.m. (0.226 to 7-55 p.p.m. of sodium) analysed on five different days. TABLE I1 INDIRECT DETERMINATION OF SOAP IN OILS BY ATOMIC-ABSORPTION : PRECISION (p.p.m.) FOR SOAP CONTENT OF REFINED VEGETABLE OILS Day 1st 2nd 3rd 3.2 3.3 3.1 4.9 5.7 5.5 8-0 7-4 8.2 13.8 13.2 13.2 23.2 23.8 23.0 57.5 57.0 58.7 97.0 97.0 98.2 4th 5th 2.9 2.6 5.6 5.2 7.9 7.5 13.6 12.9 2 4 4 23.3 68.8 58.8 96.3 95.1 Standard deviation 0.26 0.33 0.33 0.36 0.66 0.85 1.13 Coefficient of variation, per cent.8.7 6.1 4.3 2.7 2.4 1.5 1.2 Precision is satisfactory, reaching *8.7 per cent. for the 3 p.p.m. and *2-7 per cent. for the 13 p.p.m. soap concentrations. The method is standardised with sodium oleate - virgin olive oil mixtures (for the solvent standards, a solution of virgin olive oil was used as a blank), used as described in the experimental section. The last four samples in Table I1 were also analysed by the extraction methodll and, for comparison, the results obtained were 14.0,748 GEGIOU 24.5, 58.5 and 99.0 p.p.m. of soap, respectively. agreement. the soap content of which was found to range between 8 and 25 p.p.m. other vegetable oils. The two sets of values were thus in good The proposed method was applied to a number of commercial samples of refined olive oil, The method can be used to detect adulteration of virgin olive oil with refined olive or The extraction method was standardised with aqueous sodium standards. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. REFERENCES Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” MacMillan Boekenoogen, H. A., Oil Soap, 1941, 18, 8. “Official and Tentative Methods of the American Oil Chemists’ Society,” Volume 1, Third edition, Wolff, J. P., Olt!agineux, 1948, 3, 197. Nelson, R. M., J . Amer. Oil Chem. Sot., 1973, 50, 207. Jenkins, J. \i-\.’,, I b i d . , 1956, 33, 225. Black, L. T., Ibid., 1970, 47, 313. Slavin, W., Appl. Spectrosc., 1966, 20, 281. Buchananan, J. D., and Guinn, V. P., Fd Technol., Champaign, 1963, 17, 17. Riandey, C., 2% Pinta, M., Editor, “Spectrometrie d’Absorption Atomique,” Volume I, Masson et Edmonds, S. M., and Mattikov, M., J . Amer. Oil Chem. SOC., 1958, 35, 680. Received December l’lth, 1973 Amended April 8th, 1974 Accepted May 14th, 1974 and Co. Ltd., London, 1950, p. 548. AOCS, Champaign, Ill., 1964. Cie., Paris, 1971, p. 177.

ISSN:0003-2654    

DOI:10.1039/AN9749900745

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99, pp. 749-754 749 Automatic Analyses of Trace Amounts of 2-Furfuraldehyde in Gas Oil By R. G. LIDZEY AND P. B. STOCKWELL (Department of Iqzdustry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SE19NQ) The design and development of an automatic instrument for the identifi- cation and determination of 2-furfuraldehyde (0 to 6-0 mg kg-l) in gas oil is described. The principle of operation of the instrument, which is in routine use, involves a gas-chromatographic separation followed by a colorimetric determination. HYDROCARBON distillates in the gas oil range (“diesel” or “derv”) are subject to duty when used as a road fuel. Gas oil, which is often identical with diesel oil in hydrocarbon com- position, is exempt from duty when used for stationary machines.In order to prevent its misuse as a road fuel, gas oi11*2 is marked with a mixture of 1,4-dihydroxyanthraquinone (quinizarin), 2-furfuraldehyde (furfural) and a red dye (C.I. Solvent red 24). An automatic method for the extraction, identification and determination of quinizarin in gas oil has been reported3 and an automatic instrument has been in use in this laboratory for 3 years. Con- firmation of the presence of furfural provides further evidence for prosecution cases and the numbers of analyses performed also merits automatic analysis. An automatic instrument has been constructed and installed in the Hydrocarbon Oils Section for routine use, and is described in this paper. A specification for such an automatic instrument can be summarised as follows: (i) the analytical procedure should give a positive, unequivocal identification of the presence of furfural; (ii) furfural should be determined in the range 0 to 6 mg k g l , with a precision of 50.02 mg kg-l (a fully marked gas oil contains 6 mg k g l of furfural); (iii) in order to reduce the development time and to aid serviceability and maintenance, the instrument should be constructed from readily available components and modules wherever possible ; (iv) in order to be able to cope with the numbers of samples analysed annually, the maximum analysis time must not exceed 15 to 20 minutes.Prior to installing the automatic instrument, the standard procedure that had been in use for a number of years involved two separate analyses: determination by an aniline acetate colorimetric method; and identification by thin-layer chromatography.Direct automation of this procedure into a single automatic method was not practicable and several other approaches to the problem were given consideration, for example, modification of the solvent- extraction principles described in the quinizarin system3 so as to enable the furfural to be extracted, and the use of a gas-chromatographic identification or the use of a gas-chromato- graphic method incorporating a trapping system between two columns. Graham4 has described a system similar to the last-mentioned method for the determination of furfural in tobacco-smoke condensates. Both of these design applications were feasible, but it was considered that the identification of a component peak on the basis of a single retention time would provide insufficient evidence for the unequivocal identification required.The advan- tages and disadvantages of the various methods for the determination of furfural have been discussed in more detail el~ewhere.~ A simple and more direct method, which involves the use of a primary gas-chromato- graphic separation combined with a specific colorimetric device, was evaluated against the standard manual technique and proved sufficiently reliable to form the basic design of a routine automatic instrument in this laboratory. @ SAC; Crown Copyright Reserved.750 LIDZEY AND STOCKWELL: AUTOMATIC ANALYSES OF [Ana&St, VOl. 99 CONSTRUCTIONAL DETAILS OF INSTRUMENT The instrument is self-standing and all components are mounted on a specially con- structed moveable bench.All instrument controls are mounted on the front panel of the instrument and the device requires only the supply of the following services : 240-V ax. 50-H~ mains electricity, nitrogen and compressed air. A schematic arrangement of the major components is shown in Fig. 1. The equipment consists of three integral modules: the gas chromatograph with autosampler ; the scrubbing unit ; and the colorimetric detection system. Recorder fi \ 35fC Oven canopy I Injection Scrubber mechanism e u Sample tray Proportional heater and fan \ Peristaltic Pump Fig. 1. Schematic diagram of automatic system for determination furfural in gas oil of GAS CHROMATOGRAPH AND AUTOSAMPLER- A Hewlett-Packard gas chromatograph is used for the primary separation of the com- ponents in gas oil.An automatic unit feeds the chromatograph with samples and a back- flushing unit has been added in order to remove heavy hydrocarbons, which might otherwise choke the column. The carrier gas (nitrogen) is controlled by three electropneumatic valves, as shown in Fig.i2. In the normal mode with valve A open and valves B and C closed, carrier Log amplifier Recorder wiuriiri# Heating coil - I 0.3 ml rnin-' Aniline acetate reagenr Fig. 2. Schematic layout of gas and liquid flow lines. A, B and C are electropneumatic valvesNovember, 19741 TRACE AMOUNTS OF %FURFURALDEHYDE I N GAS OIL 751 gas flows through columns 1 and 2 to the detector. With B and C open and A closed, back- flushing purges column 1 of heavy organic compounds, The system requires only a single pressure regulator, the gas flow being controlled by needle valve restrictors.Two 6 mm diameter columns are used in the analysis, the first being 300 mm long and the second 1.5 m long, packed with Chromosorb G impregnated with Carbowax 20M. The separations are carried out isothermally at 140 “C. The automatic injection device is used in the conventional manner except that an electronic timer with a modified injection controller is used in order to initiate and control the operation of the back-flushing valves. Injections of 10 p1 of gas oil samples are used, these being compatible with the column configurations, so that the colori- metric method must therefore be sensitive to 0.6 mg kg-l of furfural.SCRUBBING UNIT- The scrubbing unit is the most important part of the analytical system and optimisation of its performance is essential for precise operation. The chemical reaction involved produces an unstable colour, and therefore the residence time must be sufficient for colour development to occur but the solution must be transferred rapidly to the detector so as to minimise loss of colour intensity. The unit consists of five component parts whose particular spatial con- figuration has been determined experimentally: the liquid inlet from the peristaltic pump ; the gas inlet from the gas-chromatographic column; the mixing coil, in which the furfural is transferred into the liquid stream and the colour developed; the debubbling unit, which separates nitrogen from the liquid; and the liquid trap from which liquid is transferred to the detector.Fig. 3 shows the detailed configuration of the scrubbing unit. Gas-chromatograph canopy -~ Carrier gas Fig. 3. Detail of scrubbing unit Carrier gas is transferred from the column through a heated metal capillary, which minimises the “dead” volume at the end of the column and prevents condensation of the column effluent prior to its entry into the scrubbing unit. The tube carrying the liquid stream is joined to the gas stream tube, at a T-junction that is joined to the mixing coil by a glass-to-metal seal. Furfural is transferred from the gas stream into the liquid stream and the colour develops, the two phases are then separated by the debubbling unit and the liquid stream is re-sampled through the flow cell of the colorimeter.A Technicon peristaltic pump is used to control a differential flow-rate into and out of the unit such that the liquid trap remains full during the analytical procedure. Tygon pump tubing has been found suitable for this application, while other flow lines are constructed from 1.5 mm i.d. PTFE tubing, both available from Technicon Instrument Company Limited. The flow-rates are indicated in Fig. 2. Bubbles of air are almost entirely eliminated from the colorimeter by752 LIDZEY AND STOCKWELL: AUTOMATIC ANALYSES OF [Analyst, Vol. 99 this procedure; those which are not eliminated appear as sharp spikes that cannot be confused with a chromatographic response on the chart record. The complete scrubbing unit is thermostatically controlled in an air thermostat, con- structed from Perspex, at 35 & 1 "C.In order to respond quickly to changes in the ambient temperature, the thermostat control incorporates an RFL Industries solid-state proportional controller and a fan that circulates air over a bare-element heater. COLORIMETER- It is important to minimise the length of the transfer lines to the colorimeter flow cell and to use readily available components so as to reduce development time. In order to achieve these criteria, the optical section of a Vitatron colorimeter was separated from the electronic controls and placed immediately adjacent to the scrubbing unit, within the housing of the gas chromatograph. The electronic controls are mounted on the front panel of the instrument below the gas chromatograph.Liquid from the scrubbing unit is pumped through the flow cell (path length 20 mm and internal volume 200 pl) fitted with a 520-nm filter. The normal colorimeter response is modified, by using a logarithmic amplifier,* to give an output that is linear with concentration and recorded on a strip-chart recorder. The particular configuration described above produced extremely well defined chromato- graphic peaks without gross peak broadening effects being evident. EXPERIMENTAL REAGENTS- Aniline acetate solution-Stand 1 litre of glacial acetic acid in contact with 20 ml of aniline for 24 hours, recover 90 per cent. by distillation and store in a dark bottle. Distil 100 ml of analytical-reagent grade aniline and store in a dark bottle.For use, mix 20 ml of distilled aniline with 180 ml of distilled acetic acid and cool to ambient temperature. The aniline acetate reagent is unstable, turning yellow with loss of sensitivity, and must therefore be prepared fresh daily. Both the aniline and the acetic acid are distilled in order to prevent background coloration that would produce irregularities in the recorder base-line. Standard furfural solution-From a stock solution containing 0.5 per cent. m/V of furfural (freshly distilled) in unmarked gas oil, prepare a solution containing 10 mg k g l of furfural. OPERATING CONDITIONS OF AUTOMATIC INSTRUMENTS- The needle valves are adjusted so that the flow-rate through the main gas-chromatographic column is 50 ml min-l for normal and back-flush flow.A setting on the programmer causes the injection of a further sample every 12 minutes. Back-flushing is operated after 3 minutes for a period of '7 minutes. The oven and injection block are maintained at 130 and 140 "C, respectively, and the capillary outlet of the gas-chromatographic column is maintained at 80 "C. A recorder span of 30 mV full scale covers the range 0 to 0-2 absorbance unit or 0 to 8.0 mg kg-1 of furfural for a 10-pl injection of gas oil. The aniline acetate reagent is pumped into the scrubber at least 30 minutes before the first injection in order to stabilise the system. Injection of 10 pl of gas oil containhg furfural in the range 0 to 10 mg k g l gives a chromatogram with a single peak. A retention time for the peak of 94 minutes includes the transit time through the scrubber to the optical cell.A negative spike on the recorder trace marks the injection point; no solvent peak is observed with gas oil in this detection system. A typical recorder trace is shown in Fig. 4. A graph of the concentration of furfural in gas oil against peak height was found to be linear. However, drift in sensitivity occurs over a period of several hours owing to a decrease in reagent sensitivity and variation of the flow-rates at the peristaltic pump. In order to compensate for these variations, a standard solution of furfural is injected every fifth test. The furfural concentrations in samples are calculated by calculating the mean peak height of the standard solution before and after the sample set and using a linear interpolation.COMPARISON OF RESULTS WITH THOSE OBTAINED BY THE MANUAL METHOD- In the manual test, the maximum colour intensity of the Schiff's base produced with aniline acetate is measured in situ in the oil, combining the two phases by the addition ofNovember, 19741 TRACE AMOUNTS OF %FURFURALDEHYDE IN GAS OIL 753 2 min H 6 I I I I I Fig. 4. A typical recorder trace showing a range of standard injections in the range 0.6 to 6.0 mg kg-' of furfural. I, injection point; and t, retention time toluene. In order to compare the automatic and manual methods, a series of solutions of furfural in kerosene were examined by both methods, in a randomised sequence. Table I lists the concentrations determined by the two methods and also the known concentrations in the standards. Statistical analysis of the results shows that a small negative bias exists for both the manual and automatic methods, with a maximum of 0.3 mg kg-l at an expected value of 3.0 mg k g l .The repeatability is indicated by the standard deviations of 0.23 and 0.13 mg kg-1 for the manual and automatic methods, respectively. The minimum detectable limit can be regarded as twice these values. TABLE I COMPARISON OF MANUAL AND AUTOMATIC METHODS USING STANDARD SOLUTIONS OF FURFURAL I N KEROSENE Concentration found/mg kg-' Concentration of - 1 Sample furfural added/mg kg-l Manual method Automatic method 0.9, 1.0, 0.9, 0.9 1.0, 0.9, 0.9, 0.8 A 1.1 B 2.4 2.3, 1.9, 2.3, 2.1 2.2, 2-2, 2.2, 1.9 C 0.6 0.4, 0.6, 0.7, 0.6 0-5, 0.5, 0.5, 0.4 A further comparison was made by analysing in duplicate thirty-four samples, collected over a short period, by both the manual and automatic methods.The standard deviations associated with these tests are slightly lower than those obtained by using standard solutions, and are 0-16 and 0.08 mg k g l for the manual and automatic methods, respectively. In use, the automatic method gives significantly more reproducible results than the manual method. DISCUSSION AND CONCLUSIONS Ten other aldehydes were examined by the analytical procedure described above, but only 5-methylfurfuraldehyde gave a colour change with aniline acetate, and this compound would be back-flushed from the column in the automatic sequence as its retention time is much greater than that of furfural on the particular column configuration used. Some inter- ference can occur from a solvent if it has the same retention time as that of furfural.Injection of 10 p1 of toluene, for example, has been shown to produce a peak height equivalent to 0.5 mg kg-l of furfural with a retention time of 3 minutes. The peak results from the change in refractive index when the solvent condenses in the aniline acetate stream and passes through the flow-through cell. In practice, no interference is experienced with the range of samples analysed.764 LIDZEY AND STOCKWELL The automatic instrument meets the design considerations laid down and it has been shown that the results produced are acceptable for reporting. After a 3-month trial period, during which both the manual and automatic systems were operated in parallel, the automatic instrument has been used exclusively and has now become fully integrated into the work of this laboratory. While being designed for a specific analytical problem, the successful coupling of a gas- chromatographic separation with a colorimetric system has obvious potential in solving other analytical problems. The general principle can be further extended to other detection systems, such as a detector sensitive to ultraviolet radiation, where wavelength changes have been used to tune selectively into particular peaks of interest, and fluorescence spectrometry. The authors thank Mr. D. W. L. Callender for help in the practical work, Messrs. B. E. Kent and J. R. Harris for their co-operation and the Government Chemist for permission to publish this paper. REFERENCES 1. “Hydrocarbon Oils (Marking of Oils) Regulations, 1961,” S.I. 1961 No. 861, H.M. Stationery 2. “Hydrocarbon Oils (Marking of Oils) Regulations, 1964,” S.I. 1964 No. 1349, H.M. Stationery 3. Tucker, K. B. E., Sawyer, R., and Stockwell, P. B., Analyst, 1970, 95, 730. 4. Graham, J. F., Beitr. Tabakforsch., 1969, 5 , 43. 6. Lidzey, R. G., Porter, D. G., and Stockwell, P. B., Chronzatogvuphia, in the press. 6. Bunting, W., Sawyer, R., and Stockwell, P. B., Lab. Pvact., 1970, 19, 609. Office, London. Office, London. Received Mavch 21st, 1974 Accepted June 3vd, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900749

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99, p+. 755-758 765 An Improved Method for the Colorimetric Assay of Thiacetazone in Pharmaceutical Preparations BY JACOB S. SHOHET" (Depavtment of Inorganic and Analytical Chemistry. Hebrew Univevsity of Jevusalem, Jerusalem, Israel) An improved colorimetric procedure for the determination of thiacetazone (tibione) has been developed. I t is based on the complete hydrolysis of tibione in a concentrated alkaline medium, after which the hydrolytic product is subjected to diazotisation and coupling with or-naphthylethylenediamine dihydrochloride. The procedure suggested in this paper eliminates some disadvantages of the original method and affords reproducible assay results for tibione in galenical preparations. TABLETS containing thiacetazone (tibione) have already been listed in the B.P.C.1954 and a gravimetric method for the determination of tibione, based on a reaction with silver nitrate, is given there. Among the combinations of tibione with other drugs, that of tibione with isoniazid in tablets is well known. A method for assaying tibione in the presence of isoniazid was reported by Devani and Shishoo.1 These authors suggested the use of non-aqueous titration but their results showed positive errors due to the interference of isoniazid. Deodhar, Shastri and Canatral reported an ultraviolet absorption method for the determination of tibione and a colorimetric reaction with 9-dimethylaminobenzaldehyde for isoniazid. The disadvantage of this method lies in the fact that the absorption technique, generally, does not reveal the presence of decomposition products, With regard to the determination of isoniazid, Deodhar et al.used an inexplicable modification of the method of Kidani, Nakashima, Kochi and KasaharaBs Baggi, Mahajan, Ramma Rao and Bami4 described a specific method for the deter- mination of tibione, whereby tibione is hydrolysed by a dilute acid and the primary aromatic amino group of the hydrolysis product is diazotised and coupled with a-naphthylethylene- diamine dihydrochloride. The absorbance of the resulting coloured product is measured at a wavelength of 545 nm. Simultaneous hydrolysis and colour development of both sample and standard preparation are found to be necessary in order to achieve reproducible results, suggesting that hydrolysis was incomplete and variable under the conditions recommended by these authors.EXPERIMENTAL The equation below represents the possible hydrolysis products of tibione- H2N-(+H=NNHCSNH2 - (D) A and B are the principal products.6 C and D can occur in the event of a partial hydro- lysis. The mechanism and rate of hydrolysis can be studied from the literature on acetanilide and its +substituted derivatives.6-ll Complete hydrolysis of acetanilide and its +-substituted * Present address : ABIC Ltd., Chemical and Pharmaceutical Industries, Ramat-Gan, Israel. @ SAC and the author.756 SHOHET: AN IMPROVED METHOD FOR THE COLORIMETRIC ASSAY [Analyst, Vol. 99 derivatives requires either extreme conditions of acidity and heat (this could be achieved by the addition of 50 per cent.sulphuric acid and refluxing for 15 minutes; however, undesirable sulphonation might occurlO), or extreme conditions of alkalinity and heat, such as would result from the direct addition of potassium hydroxide pellets and refluxing for 8 to 20 minutes. In general, the presence of a substituent in the para-position of acetanilide has only a small effect on the rate of hydrolysis.12 A decrease in the concentration of the acid or the alkali, to a limited extent, can be effected only at the expense of an increase in the reflux If a dilute acidic or alkaline medium (0.1 N) is used, the hydrolysis is very inefficient (10 to 30 per cent.).13 Having studied the various conditions of hydrolysis, and the difficulties encountered by Baggi and co-workers4 in completing it, the following procedure for assaying tibione in the presence of isoniazid has been developed. REAGENTS- Sodium hydroxide solution, 5 N.Dilute hydrochloric acid solution, 0.5 N. Sodium nitrite solution, 1 per cent. m/V. Ammonium sulphamate solution, 10 per cent. m/V. a-Naphthylethylenediamine dihydrochloride solution, 0.5 per cent. m/V. Ethanol, analytical-reagent grade. Thiacetazone (tibione) standard. Isoniazid standard. SOLUTIONS- Preparation of standard solzitio.rz-Weigh accurately 50 mg of tibione standard and 200 mg of isoniazid standard into a 100-ml calibrated flask. Add 50 ml of ethanol and dissolve the standards by warming gently on a water-bath. Cool, and dilute to the mark with distilled water. Preparation of sample solution-Weigh accurately a portion of finely powdered tablets, equivalent to about 50 mg of tibione, into a 100-ml calibrated flask. Add 50 ml of ethanol and heat the mixture for 30 minutes on a water-bath.Cool, and fill the flask to the mark with distilled water. If the solution is not clear, filter it through a Whatman No. 1 filter-paper. PROCEDURE- Transfer, by means of pipettes, 5 ml of both standard and sample solutions into separate round-bottomed flasks. Add to each solution 75 rnl of distilled water and 20 ml of sodium hydroxide solution and reflux for 3 hours. Cool, transfer the contents of the flasks quanti- tatively into 200-ml calibrated flasks and dilute to volume with distilled water, Label these solutions “sample solution A” and “standard solution A.” Transfer by pipette, in triplicate, 2 ml of sample solution A and 2 ml of standard solution A into separate, 15 x 150 mm, ground-glass tubes.To each add 3.0 ml of dilute hydrochloric acid solution and 2.0 ml of distilled water. Into a further tube place 7.0 ml of distilled water to serve as a blank. Add 1.0 ml of sodium nitrite solution and allow the solutions to stand for 5 minutes, then add 1.0 ml of ammonium sulphamate solution, shake the tubes vigorously and allow them to stand for a further 5 minutes. Bubble nitrogen through the solutions in order to remove nitrous fumes, add 1.0 ml of a-naphthylethylene- TABLE I CORRELATION WITH BEER’S LAW Concentration/ Absorption values of samples A B C pgper 1 O m l A \ 5 0-092 0.090 0-094 10 0.186 0.185 0.188 20 0-370 0.367 0.372 30 0.552 0.553 0.550 40 0.727 0-729 0.732 50 0.899 0.903 0.900November, 19741 OF THIACETAZONE IN PHARMACEUTICAL PREPARATIONS 757 diamine &hydrochloride solution, mix well and allow the solutions to stand for 10 minutes.Determine the absorptions of the sample and standard solutions against the blank reagent at a wavelength of 550 nm, using l-cm spectrophotometric cells. RESULTS AND DISCUSSION The conditions of hydrolysis adopted above, namely, concentrations of reactants, heating process and heating time, have been found to be optimal for completing the hydrolysis. When such conditions are adhered to, both precise and accurate results are obtained. The precision of the method has been checked by replicate analyses of three samples. The results obtained are given in Table I.The accuracy of the method has been checked on a series of simulated mixtures composed of various amounts of isoniazid in combination with a constant amount of tibione. Table I1 records the labelled amounts, amounts found and percentage recovery in these samples. TABLE I1 ASSAY RESULTS FOR TIBIONE IN SIMULATED MIXTURES CONTAINING ISONIAZID Labelled amounts/mg C - G Z z d 50 25 50 50 50 100 50 150 50 200 50 250 50 300 Tibione A r \ Mean amount Mean 49.1 98.2 48.6 97.2 50.2 100.4 48.8 97.6 49.3 98.6 49.7 99.4 49.0 99.2 found/mq recovery, per cent. Further, the absorption spectrum of the final product shows a sharp peak at a wavelength of 550nm, in contrast to the broad peak obtained when using a dilute acidic medium. This difference might be ascribed to the fact that in dilute acid both products B and D are obtained to a large extent and, as they absorb in the same wavelength region, a broad peak is formed, as shown in Fig.1. 500 520 540 560 580 600 Wavelengthhm Change in absorbance veysus wavelength: a, in basic medium; and b, in acidic medium Fig. 1. REFERENCES 1. 2. 3. 4. Devani, M. B., and Shishoo, C. J., Indian J . Pharm., 1967, 29, 307. Deodhar, R. D., Shastri, M. R., and Ganatra, J . P., Ibid., 1970, 32, 99. Kidani, I., Nakashima, T., Kochi, Y., and Kasahara, S., Bull. Natn. Hyg. Lab., 1954, 72, 95. Baggi, T. R., Mahajan, S. N., Ramma Rao, G., and Bami, H. L., Indian J . Phavm., 1968,30, 122.758 SHOHET Asahi, Y., Chenz. Pharm. Bull., Tokyo, 1963, 11, 1241. Vogel, A. I., “Practical Organic Chemistry,” Longmans, Green and Co. Ltd., London, 1967, Golovinski, E., and Spasov, A., C. R. Acad. BuLg. Sci., 1962, 15, 507. Stroh, H. H., and Leihr, H., -1. firaht. Chem., 1065, 29, 8. Owen, W. S., Mikroclzinz. Acta, 1963, 19. D~iffy, j. A., and Lcisten, J . A., J . Client. SOC., 1060, 545. Sudo, T., Shimoe, D., and Tsujii, T., Bunseki Kngaku, 1957, 6, 498. Bender, XI. L., and Thomas, R. J., J . Anzer. Chem. SOC., 1961, 83, 4183. Spinkova, V., &lkd Famz., 1967, 16, 138. p. 1075. Received Mavch 121h, 1974 Accepted M a y 29th, 1974 5. 6. 7. 8. 9. 10. 11. 12. 13.

ISSN:0003-2654    

DOI:10.1039/AN9749900755

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

A i d y s t , November, 1974, Vol. 99, pp. 759-764 759 Studies on a Specific Colour Reaction for the Determination of Testosterone BY EMIL FAHMY, DAWOUD A. YASSA ( Reseavch and Control Department, Sociktk Misr pour I’Industrie Pharmaceutique, 92 El Mataria Street, Post El Zeitoun, Cairo, Egyfit) AND NAG1 WAHBA (Biochemistry Department, Faculty of Medicine, A i n Shams University, Cairo, Egypt) The chromogen resulting from preliminary treatment of testosterone with concentrated sulphuric acid gives a stable green colour on further treatment with a saturated aqueous solution of picric acid. The absorbance graph of the green colour shows two maxima, a t 470 and 640nm, and obeys Beer’s law. Under the specified conditions, absorbances a t the two maxima have a fairly constant ratio (2.80 to 2.96), which is characteristic for testosterone and its esters in amounts ranging from 125 to 250 pg.Spectral and structural studies involving a large number of steroids show that the colour reaction is specific for testosterone. THE established methods most extensively used for the determination of testosterone depend upon the activity, either of the A4,S-keto group, e.g., Caro1,l Ercoli, de Giuseppe and de Ruggier2 and Umberge~-,~ or of the 3-keto group, e.g., Ercoli and de Giuseppe4 and Madigan, Zenno and Phea~ant.~ Although highly sensitive, these methods are not specific and are equally applicable to the other steroids that possess the pertinent groups. The B.P. 1968,g for instance, describes a reaction that depends upon the formation of isonicotinyl hydrazone for the assay of injections of nandrolone and testosterone esters as well as progesterone.The U.S.P. XVII17 similarly describes assays for testosterone and medroxyprogesterone esters. The methods described in B.P. 19638 involving spectrophotometric measurement for the assay of testosterone phenylpropionate and semicarbazone formation for the assay of testo- sterone propionate are applicable to progesterone, and the B.P. 1963 method for the assay of progesterone involving the formation of 2,4-dinitrophenylhydrazone has been found to be applicable to testosterone esters after appropriate modification^.^ Both the B.P. 1963 and the U.S.P. XVIII provide identification tests as an aid to verifying the identity of substances taken from labelled containers.As is stated in U.S.P. XVIII,7 such tests, however specific, are not necessarily sufficient to establish proof of identity. A specific reaction is required that can be used for the identification, and possibly the determination, of testosterone without interference from progesterone and oestradiol, which are sometimes mixed with it for simultaneous administration. A few methods have been claimed to be specific for testosterone, e.g., those reported by Koenig,1° Sachsll and Gupta and McCaff erty. l2 All three of these methods depend primarily upon the production of a chromogen by treating testosterone with sulphuric acid (Bernstein and Lenhardl3~1~). For the final colour production potassium guaiacol sulphonate solution containing copper sulphate is used and 50 per cent.V/V sulphuric acid added in the first method, aluminium iron(I1) sulphate and potassium permanganate are included in the concentrated sulphuric acid and 3 per cent. V/V acetic acid is added in the second method, while in the third method, cerium(1V) sulphate and aluminium iron(I1) sulphate are included in the concentrated sulphuric acid and 10 per cent. acetic acid is added. All of these methods therefore involve the use of inorganic catalysts and the results, according to the authors, are variable. This paper describes the development of a new, simple, accurate and specific reaction for testosterone, which can be used for its determination in oily solutions. l 5 0 SAC and the authors.760 FAHMY, Y A S S A AND W-AHBR: STUDIES ON A SPECIFIC COLOUR [AndySt, VOl.99 EXPERIMENTAL On the assumption that the testosterone treated with concentrated sulphuric acid was amenable to a specific type of oxidation, and having regard to the reagents used by Koenig,lo Sachsll and Gupta and McCafferty, l2 which contained inorganic catalysts, it was considered appropriate to attempt to use organic oxidants, e.g., nitro compounds. Preliminary trials with several such compounds confirmed this assumption. CHOICE OF A SUITABLE NITRO COMPOUND- In view of the insolubility in water of most of them and because preliminary experiments showed that ethanol, in which all of them were partially soluble, inhibited colour formation, they were dissolved in the sulphuric acid. A 4 volume (1.0 ml) oi a solution of 25 mg of testosterone propionate in 100 ml of ethanol was placed in each of nine test-tubes and heated to dryness in a boiling water bath.To each tube, still in the bath, 2-0 ml of concentrated sulphuric acid containing 1 per cent. m/V of one of the nine nitro-compounds were added. After 10 minutes the tubes were removed from the bath and cooled in ice. On addition of 2 ml of distilled water, the yellow solutions immedi- ately acquired a green colour. Dilution with 10-0 ml of 50 per cent V/V sulphuric acid and subsequent spectrophotometric examination showed that all of the nitro compounds showed the same maximum absorbance at about 640 nm, but with rising absorptivities in the following order: 2,4-dinitrophenylhydrazine, m-dinitrobenzene, 3,5-dinitrobenzoic acid, 3,5-dinitro- chlorobenzene, 9-nitroaniline, m-nitroaniline, 2,4-dinitrophenol, o-nitroaniline and picric acid.It should be noted that water was essential in the production of the chromogen, which remained remarkably stable and gave consistent readings throughout 6 hours of observation. Picric acid was chosen for further work as it gave the highest absorbance and had the further advantage of being soluble in water. It was observed that addition of picric acid solution after causing the dry hormone to react with concentrated sulphuric acid produced the colour, just as it did when the hormone was heated with the solution of picric acid in concentrated sulphuric acid and water was subsequently added. The use of a separate picric acid solution was therefore adopted for the purpose of fixing the optimum conditions for carrying out the test because separate reagents were presumed to be more stable.OPTIMUM CONDITIONS FOR COLOUR PRODUCTION- The use of concentrated sulphuric acid in the preliminary treatment proved to be essential because if 50 per cent. V/V solution is used at this stage the absorbance at the wavelength of maximum absorbance of 640 nm was reduced to about one-eighth of its value. Other acids, e.g., concentrated hydrochloric acid, phosphoric acid, metaphosphoric acid and glacial acetic acid, failed to produce the desired colour. The optimum reaction time with concentrated sulphuric acid was assessed by taking fourteen pairs of tubes, each pair containing 0.125 and 0.25 mg of testosterone propionate, removing one pair from the boiling water bath and cooling the tubes immediately in ice at 1-minute intervals.The absorbance following the addition of 2.0 ml of 1 per cent. m/V picric acid solution and 10.0 ml of 50 per cent. sulphuric acid rose to a maximum after 9 minutes and remained unchanged until the fourteenth minute. When carrying out a comparison of 0.25,0.5 and 1 per cent. m/V picric acid solutions to develop the green chromogen from the testosterone treated with concentrated sulphuric acid, the highest absorbance was obtained with the 1 per cent. solution following heating of the mixture in the boiling water bath for 3 minutes. Dilution to obtain suitable readings should be carried out with 50 per cent. VjV sulphuric acid; lower concentrations gave rise to either turbidity or reduced absorbance, while higher concentrations conferred no distinct advantage.Nine readily available nitro compounds were used. 1. 2. 3. 4. METHOD The following general method was adopted for the examination of colour stability, Reagents and equi$ment-Sulphuric and picric acids were of analytical-reagent grade, scanning of the absorbance graph, concentration studies and specificity testing.November, 19741 REACTION FOR THE DETERMINATION OF TESTOSTERONE 761 while testosterone, its esters and the other steroids used were obtained from reliable sources and were of comparable purity with official grades, where applicable. Solutions of 12.5 and 25 mg of testosterone in 100.0 ml of absolute ethanol were used. A Carl Zeiss spectrophotometer (PMQ 11) with l-cm cells was used for absorbance measurements.PROCEDURE- Place 1.0 ml of the appropriate solution of the pure testosterone or derivative into a long test-tube (about 1.2 x 20 cm), and dip the bottom of the tube into a boiling water bath in order to evaporate off most of the ethanol, the last trace being blown off by a current of nitrogen. To the dry residue, still dipped into the water-bath, add exactly 2-Om1 of con- centrated sulphuric acid and continue heating with occasional shaking for 10 minutes. Cool the solution in ice. Gradually add 2.0 ml of 1 per cent. picric acid solution while shaking the tube and heat the mixture in a boiling water bath for 3 minutes. To the green-coloured solution add exactly 10.0 ml of 50 per cent. VjV sulphuric acid, mix, and examine the diluted solution spectrophotometrically against a blank containing all of the reagents but no steroid.STABILITY OF THE CHROMOGEN- The absorbance of the solution remained stable for at least 2 days at room temperature and for 2 hours at 40 "C (kept in an incubator) but it declined rapidly when the solution was placed in an oil-bath a t 120 "C. The addition of inorganic catalysts, such as 0.1 per cent. m/V copper sulphate or iron(II1) chloride to the picric acid solution did not accelerate the colour development, increase the stability of the colour or raise the absorbance reading, nor did it change the wavelength of maximum absorbance. Stronger oxidising agents, such as hydrogen peroxide and potassium permanganate, lowered the value of the maximum absorbance, the wavelength for which, however, was maintained.Cool it to room temperature. ABSORBANCE GRAPH- As can be seen in Fig. 1, two maxima were identified, a low, broad absorbance area with a peak at 440 to 470 nm, a minimum at about 516 nm and a sharp, high absorbance band with a peak at about 640 nm. 0 Wavelength/nm Fig, 1. Absorbance graph of the sulphuric acid - picric acid chromogen of testosterone propionate (260 pg)762 FAHMY, YASSA AND WAHBA: STUDIES ON A SPECIFIC COLOUR [Analyst, Vol. 99 CONCENTRATION STUDIES- Standard graphs-The graphs (Fig. 2) of absorbances at both wavelengths, plotted against concentrations, obeyed Beer’s law and passed through the origin. They showed also that the ratios between the absorbances at 640 nm and those at 470nm were fairly fixed and lay between 2.80 and 2.96 for amounts ranging from 125 to 250 pg.Repeated experiments have confirmed that this ratio is, for the purposes of identification, characteristic of testosterone. 1.2 1 Concentration of testosterone propionate/pg Fig. 2. Graphs of absorbance versus concentration of testosterone propionate : -, Amax. 640 nm; and - - -, A,,,. 470 nm iWoZar absorption of testosterone and its esters-The molar extinction coefficients of the chromogens obtained from fixed amounts of testosterone and its propionate, cyclopentyl- propionate, enanthate, and phenylpropionate esters are shown in Table I. TABLE I MOLAR EXTINCTION COEFFICIENTS OF TESTOSTERONE AND SOME OF ITS DERIVATIVES Molar extinction Steroid 1Mr E% coefficient (c) Testosterone . . . . . .288.43 761.6 21 964 526.0 21 371 Testosterone enanthate . . . . 400.61 Testosterone phenylpropionate . . 420.60 504.0 21 205 Calculation of the molar extinction coefficients ( E = aZc, where a is the absorbance, I is the light path (cm) and c is the concentration in moles per litre) gave almost equal values, showing that the esters react according to their testosterone content irrespective of the acid radicals. Testosterone propionate . . . . 344.50 616.0 21 221 Testosterone cyclopentylpropionate 412-62 5 15.0 21 252 SENSITIVITY- It is noted, from the fact that the molar extinction coefficient averages about 21 000, that this reaction is more sensitive than Umberger’s reaction3 using isoniazid, which has a molar extinction coefficient of 12 000, and the ultraviolet absorbance method at 240 nm, which has a coefficient of 17 000.SPECIFICITY- Of the steroids tested the following, or their esters, did not give any colour: progesterone, oestradiol, ethynyloestradiol, methyltestosterone, ethisterone, cortisone, hydrocortisone, prednisone, prednisolone, a-fluoroprednisolone, 16-P-methylprednisone, betamethasone and desoxycort icos terone. Dihydrotestosterone, androstanedione, nortestosterone esters and hydroxyprogesterone n-hexanoate gave blue or green colours with a low, yet appreciable, absorbance at 640 nm in comparison with testosterone, while androstenedione gave a much higher absorbance.November, 19741 REACTION FOR THE DETERMINATION OF TESTOSTERONE 763 Only the colour produced by androstenedione obeyed Beer’s law, but those colours obtained from androstanedione, nortestosterone, hydroxyprogesterone and dihydrotesto- sterone gave inconsistent readings at 640 nm.Despite this inconsistency, the ratio between the intensity of absorbance at 640 and 470 nm was fairly constant for each steroid but widely different from that characteristic of testosterone (Table 11). TABLE I1 VISIBLE LIGHT SPECTRAL RESULTS FOR THE SULPHURIC ACID - PICRIC ACID CHROMOGEN OF VARIOUS STEROIDS Steroid Wavelengths of peaks/nm Ratio* Testosterone . . . . .. 470640 2.9 Dihydrotestosterone . . . . 470640 0.805 Androstenedione . . . . , . 470640 4.3 Androstanedione . . . . . . 3954706-20 1.16 Nortestosterone propionate . . 390 510 0.377 H ydroxyprogesterone n-hexanoate 470 640 0.93 * Ratio of absorbance at 640 nm to absorbance at 470 nm for an amount of 250 pg subjected to the reaction.The complete absorbance graphs (Fig. 3) confirm the difference in spectral nature between the chromophores of these steroids and that of testosterone. This was particularly true of nortestosterone’s chromophore, which showed two peaks at different sites from those given by testosterone, nix., at 390 and 510nm, in addition to an evenly extended general absorbance above 550 nm. Androstenedione, hydroxyprogesterone, androstanedione and dihydrotestosterone, on the other hand, showed the two characteristic peaks at 640 and 470 nm but in altered ratios, as indicated above, androstanedione being further distinguished by an additional high peak at 395nm. DISCUSSION Examination of the structural relationships between testosterone and the other steroids to which the reaction was applied shows that certain groups play an essential, specific part in the formation of the chromogen.Average of four determinations. I -C2’ 1 20 -C- & 0 Numbering system for steroids Testosterone The presence of the secondary alcohol group at C,, is obviously essential, as addition of another group in such a way that it becomes a tertiary alcohol, removes absorbance at 640 nm, as was found with methyltestosterone and ethynyltestosterone. Replacement of the 17-fl-hydroxy group by -CO-CH, or CO-CH,OH, as in proges- terone and desoxycorticosterone, respectively, also removes absorbance at 640 nm, while 17-a-hydroxyprogesterone gives a chromogen that has characteristic absorbance peaks at 470 and 640 nm, which peaks, however, are lower than, and in an altered ratio from, those of testosterone.Androstenedione, on the other hand, which has the 17-P-hydroxy group replaced by a keto group, gives a chromophore with greater absorbance at 640 nm than that of testosterone. The essential part played by the 3,kethylenic bond is shown by the response to the reaction of dihydrotestosterone and androstanedione, which differ from testosterone and androstenedione in lacking the 3,4-ethylenic bond and having much lower absorbances at 470 and 640nm.764 FAHMY, YASSA AND WAHBA Another essential group to the absorbance spectrum is the 19-methyl group; this group is absent in nortestosterone, which yields a coloured solution of different absorbance pattern from testosterone, with low absorbance from 550 nm extending beyond 640 nm.1.21 360 400 440 480 520 560 600 640 680 720 Wavelength/nm Fig. 3. Absorbance graphs of the sulphuric acid - picric acid chromogens of : A, hydroxyprogesterone caproate ; B, nortestosterone phenylpropionate; and C, androstanedione (250 p g ) It is evident from the present discussion that the reaction with sulphuric acid - picric acid is specific for testosterone, involving at least three of its functional groups. The method has further advantages of being simple and reproducible, and of requiring common, readily available and stable reagents. These advantages promoted its trial for use in the determina- tion of oily injection solutions containing testosterone alone and in the presence of proges- terone or oestradiol. Moreover, the fixed ratio between the absorbance intensities at 640 and 470 nm of the steroids that gave positive sulphuric acid - picric acid colour reactions might be of value for the purpose of identification. 1 . 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. REFERENCES Carol, J., J . Ass. 08. Agric. Chew., 1951, 34, 572. Ercoli, A., de Giuseppe, L., and de Kuggier, B., Favmaco Ed. Sci., 1952, 7, 170. Umberger, E. J., Analyt. Chem., 1955, 27, 768. Ercoli, A., and de Giuseppe, L., Farmaco, Ed. Sci., 1951, 6, 702; Chem. Abstr., 1952, 46, 4794. Madigan, J. J., Zenno, E. E., and Pheasant, R., Analyt. Chem., 1951, 23, 1691. “British Pharmacopoeia 1968,” The Pharmaceutical Press, London, 1968, p. 995. United States Pharmacopoeia,” XVIII Revision, Mack Publishing Company, Easton, Pa., 1970, p. 6. “British Pharmacopoeia 1963,” The Pharmaceutical Press, London, 1963, p. 814. Diding, E., Svenska Faravelsfor. Tidskr., 1952, 3, 56; Chem. Abstr., 1952, 46, 6325. Koenig, L. V., J . Riol. Chem., 1941, 141, 487. Sachs, L., Nature, Lond., 1964, 201, 4916. Gupta, D., and McCafferty, E., Steroids, 1966, 8, 459. Bernstein, S., and Lenhard, R. H., J . Org. Chem., 1953, 18, 1146. Fahmy, E., Yassa, D. A., and Wahba, N., in preparation. f , Ibid., 1954, 19, 1269. -- Received March 18th, 1974 Accepted A p r i l 23rd, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900759

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99, pfi. 765-770 765 An Absolute Galvanic Detector for Nitrogen Dioxide BY J. D. ALLEN (British Gas Corporation, Research and Development Division, Watson House, Peterborough Road, London, S W6 3HN) The performance of a modified Hersch and Deuringer galvanic detector for nitrogen dioxide was examined. The results indicate that the response is coulometric a t low flow-rates and low concentrations on the basis of 1 Faraday (96 487 C) per mole of nitrogen dioxide. This response is explained in terms of the electrochemical reduction of nitrogen dioxide to the nitrite ion or nitrous acid. The cell can thus differentiate between nitrogen dioxide and nitrous acid vapour, which might be present in humid air. The detector has an immediate application for calibrating low-concentration mixtures containing nitrogen dioxide, but further work is necessary to investigate the reliability of the selective scrubbers required for monitoring nitrogen dioxide in polluted, ambient air.SIMPLE galvanic cells that are capable of detecting strongly oxidising or reducing gases have been known for many years. Her~chl-~ has described several such cells, including one that was sensitive to nitrogen This cell consisted of a platinum-gauze cathode exposed to a flowing gas stream and a silver or active carbon anode. The electrolyte was a neutral- buffered chloride solution. This type of cell can provide a coulometric output for low con- centrations of halogens3 in the gas stream (k, the output current is related directly by Faraday’s law to the mass flow of halogen, which is reduced to halide ion), whereas the coulometric efficiency for ozone is less than 100 per cent., owing to partial catalytic decom- position on the platinum ~ a t h o d e .~ For nitrogen dioxide, Hersch reported3s4 coulometric efficiencies of up to 35 per cent. based on the assumed cell reaction Pt NO, + H,O + 2e --> NO + 20H . . . . * * (1) 3NOz + H,O -+ 2HNO3 + NO .. .. ’ (2) The low efficiency of the cell was explained by the non-electrochemical reaction of a high proportion of the nitrogen dioxide in the sample with the aqueous electrolyte: We have further examined the performance of a variant of this Hersch cell; it soon became apparent that the cell reaction of nitrogen dioxide was not that assumed by Hersch [reac- tion ( l ) ] and that coulometric efficiencies of 100 per cent.could be achieved. EXPERIMENTAL The cells constructed were of the same basic design as that described by Hersch as a “cavity electrode” cell,3 except that the cathode was vertically orientated (Fig. 1). The glassware was assembled from a Quickfit chromatographic column and a conical flask, and the integral, sintered-glass disc of the former was used to separate the cathode and anode compartments. A small bulb was blown just above the sinter to provide a small electrolyte reservoir. The cathode was made from platinum-wire gauze, 36 mesh per linear inch (Johnson Matthey Metals), to form a cylinder with a closed end, typically 200 mm long by 12 mm in diameter. This gauze was tightly wrapped with glass-fibre paper (Whatman GF2), which acted as a wick, thus ensuring that the cathode was kept moist by a film of electrolyte.The anode was usually prepared as a thick paste of activated charcoal (e.g., Norit SXl) with the electrolyte. A plug of quartz-wool was also used to prevent the contamination of the cathode compartment by carbon particles. The composition of the electrolyte was found not to be critical, but it typically consisted of 2 M potassium chloride in a neutral (pH 7.0) phosphate buffer solution. Usually a high proportion (about 25 per cent. V / V ) of a miscible liquid, such as digol or glycerol, was added to the electrolyte in order to reduce the rate of evaporation of water from the cell during extended periods of operation. @ SAC and the author.766 ALLEN : AN ABSOLUTE GALVANIC DETECTOR [AnaZyst, VOl.99 The galvanic cell current was monitored as the potential drop across a precision resistor (100 or 1000 St, &Om1 per cent.) that was connected between the anode and cathode, being displayed by using a potentiometric chart recorder (Smiths Industries, Servoscribe 2). A higher load resistance caused a significant decrease in the response time of the cell. The voltage drop across the external load resistance had no apparent effect on the response to nitrogen dioxide for values up to 200mV. SamDle inlet + - Sample outlet Glass capillary inlet tube ( ~ 8 m m 0.d.) Layers of glass- fibre paper Bright platinum gauze (36 B.S. mesh) Glass cathode compartment (Qu ic kf i t CR 3 2/2 0) Max. electrolyte level Quartz-wool plug Anode compartment ( Qu ic kf i t FEZ 512) Active charcoal paste Fig.1. Cross-section through “cavity electrode” galvanic cell. Platinum gauze 200 mm long x 12 mm diameter The response of the cell to nitrogen dioxide was assessed by using calibrated mixtures (in bottles) of nitrogen dioxide with nitrogen or air. PTFE sample lines were always em- ployed. These mixtures of gases were initially analysed, and frequently checked, by using a Thermo-Electron 12A chemiluminescence monitor that had been calibrated with mixtures of nitric oxide and nitrogen (standardised against gravimetrically prepared primary standards) and further mixtures prepared dynamically by the electrolysis of nitrosonium hydrogen sulphate in concentrated sulphuric acid.5 These two calibration methods agreed to within the estimated accuracy of the latter method, -+2 per cent.The efficiency of the nitrogen dioxide to nitric oxide converter of the chemiluminescence monitor was found to be close to 100 per cent. for nitrogen dioxide - nitrogen mixtures. For mixtures of nitrogen dioxide with air, the maximum efficiency at temperatures above 600 “C corresponded to the thermo- dynamic equilibrium ratio of nitric oxide to the total oxides of nitrogen (nitric oxide plus nitrogen dioxide) for the reaction for the particular conditions of temperature and pressure in the c~nverter,~,~ i.e., the maximum conversion in air was about 90 per cent. at a temperature of 650 “C and at atmospheric pressure. RESULTS .. . . - - (3) 2N0 + 0, + 2N0, ,. When cells of different cathode area were supplied with the same nitrogen dioxide - air or nitrogen mixture of constant composition, a graph of output current as a function of sample flow-rate (Fig.2) showed the same linear response at low flow-rates. The slope of this linear region was found to correspond, within the limits of experimental error, to half the theoretical response predicted by reference to Faraday’s law from the cathode reaction suggested by Hersch [reaction (l)]. The response was therefore equivalent to 1 Faraday (96 487C) per mole of nitrogen dioxide. The effluent gas from one cell was analysed for oxides of nitrogen by using the chemiluminescence monitor. In the region where the output was proportionalNovember, 19741 FOR NITROGEN DIOXIDE 767 to the sample flow-rate nitrogen dioxide was not detected in the cell effluent, nor was there an increase in the nitric oxide content above that of the original sample gas.When the flow-rate was increased beyond the linear response region, the proportion of unreacted nitrogen dioxide observed in the cell effluent was equal to the fractional loss in the output current predicted from a response of 1 Faraday per mole of nitrogen dioxide. There was, therefore, no evidence of a non-electrochemical reaction of nitrogen dioxide with the aqueous electrolyte [reaction (2)], as was implied by Hersch and Deuringer.2 When a cell was supplied with samples with different concentrations of nitrogen dioxide, the output current in each instance tended towards the predicted value for a response of 1 Faraday per mole of nitrogen dioxide at low sample flow-rates (Fig.3). For each cell there was a maximum flow-rate, above which the predicted output was not achieved for any nitrogen dioxide concentration. For the cell illustrated in Fig. 1, this limiting flow-rate was typically 60 ml min-l. 130 120 110 1 00 Q: 90 2 C 80 !! 3 70 '2 60 m 50 0 40 30 20 + u m - 500 =$' 400 0 - s -.. 5 2 200 0 C 300 + c 2 V C m m .- - 0 100 r I I I I I I I I 0 Sample flow-rate 20 40 60 80 100 120 140 160 Sample flow-rate/mI min-' (10 p.p.m. V/V NO:! in N*)/ml min-' Fig. 2. Graph of galvanic current against sample flow-rate for four cells of different cathode dimensions. (Cell B has the dimensions of the cell illustrated in Fig. 1). Relative cathode areas are: cell A, 4; cell B, 3; cell C, 2; and cell D, 1 Fig.3. Graph of galvanic current against flow-rate of three samples containing nitrogen dioxide in air. (Cell dimensions as shown in Fig. 1). NO, concentration, p.p.m. V/V P Symbol Measured (NO, E 1 e-) Calculated A 1.1 1.06 0 24 23-2 0 92 93 At a flow-rate less than this limiting value, say 50mlmin-l, the output current was found to have a linear relationship with the nitrogen dioxide concentration, in agreement with the predicted value, up to a maximum concentration, typically in excess of 30 p.p.m. VlV (Fig. 4). Not surprisingly, the maximum nitrogen dioxide concentration that gave the pre- dicted output was proportional to the area of the platinum cathode, but it also decreased as the electrolyte film on the cathode evaporated.The addition of digol or, particularly, glycerol, which is hygroscopic, to the electrolyte virtually eliminated the latter effect . The platinum gauze employed was untreated apart from washing with de-ionised water and acetone. Normally the cells gave a very small background current (SIfTO-2 PA) in an atmosphere of nitrogen, and a very small positive response to pure air (less than 0.1 PA). Attempts to increase the activity of the platinum, by cathodic and anodic cycling at high current densities in acidic or alkaline electrolytes, or by washing it in concentrated nitric acid, produced a high, slowly decaying response to oxygen in the sample. This galvanic768 ALLEN : AN ABSOLUTE GALVANIC DETECTOR [Analyst, VOl. 99 response to oxygen was considered to be undesirable.That the catalytic activity of the platinum cathode was not the critical factor in determining the cell performance was indicated by the identical performance of a cell containing a cathode of 20 per cent. gold - palladium alloy gauze. Nitrogen dioxide concentration, p.p.m. V/V Fig. 4. Graph of galvanic current against concentration of nitrogen dioxide in air flowing a t 50 ml min-l (at 20 "C, 1013 mbar). Nitrogen dioxide concentration determined by chemi- luminescence monitor. (Cell dimensions as shown in Fig. 1) Some mixtures containing nitrogen dioxide, prepared by stepwise dilution in aluminium alloy cylinders, were very unstable. The nitrogen dioxide concentration decayed rapidly initially, although an equilibrium level was reached after a period of a few days or weeks.This decay was thought to be due to residual moisture adsorbed on the cylinder walls or in the diluent air. With these mixtures, particularly those containing less than 5 p.p.m. V/V of nitrogen dioxide, the cell response was significantly less than that normally expected from the nitrogen dioxide content that had been indicated by the chemiluminescence analyser ; the unusually sluggish response of the analyser to these gases suggested strong adsorption in the sample lines. The addition of an oxidant tube* to the inlet of the cell increased the response towards the predicted level. This indicated the presence of a component or com- ponents in these mixtures, other than nitrogen dioxide, which could be readily oxidised to nitrogen dioxide and reduced to nitric oxide by the converter of the chemiluminescence analyser.From previous investigations of low concentrations of nitrogen dioxide in moist air,Q it seems probable that the nitrogen dioxide was initially lost from these mixtures by adsorption on the cylinder walls, forming nitrous and nitric acids, and that a small, but significant, fraction of the apparent total oxides of nitrogen in the gas phase was present as nitrous acid vapour. The galvanic cell is not expected to respond to nitrous or nitric acid vapour, as is indicated below. Nitrogen dioxide mixtures prepared in clean, dry cylinders showed no initial decay in concentration and were stable over periods of several months. DISCUSSION The observations regarding the galvanic response of the cell towards nitrogen dioxide cannot be explained in terms of reaction (1).The response appears to be coulometric in terms of 1 Faraday per mole of nitrogen dioxide and the reduction product is not nitric oxide. The most probable explanation for these observations is that the reduction of nitrogen dioxide a t the cathode proceeds according to reaction (4) or ( 5 ) : . . . . * ' (4) NO, + Hf + e-- -+ HNO, . . .. . . * - (5) NO, + e- --f NO2- . .November, 19741 FOR NITROGEN DIOXIDE 769 That nitrite is (electrochemically) a stable product in the cell was confirmed by the addition of a small amount of nitrite solution to the cathode compartment, which resulted only in a small positive pulse, probably caused by the presence of a small amount of dissolved nitrogen dioxide in the nitrite solution added.Similarly, the nitrate ion is stable in this cell. Consequently, the additional components in the unstable, bottled mixtures are most likely to be nitrous and nitric acid vapours, to which the cell should not respond. The ability of the galvanic cell to differentiate between nitrogen dioxide and nitrous acid vapour may have important consequences. The commonly used, wet-chemical tech- n i q u e ~ ~ (e.g., Griess - Saltzman reagents) for the determination of nitrogen dioxide in ambient air are not able to distinguish between nitrogen dioxide and nitrous acid, except by the use of a deliberately inefficient sampling system.1° Similarly, the nitrogen dioxide to nitric oxide converters that are used in the chemiluminescence monitors that are currently available readily convert nitrous acid (and in some instances, nitric acid) to nitric oxide.Although nitric acid vapour is at least as toxic as nitrogen dioxide, the relative toxicities of nitrogen dioxide and nitrous acid have not been established, but their joint presence in polluted air seems probable.lO If nitrous acid is significantly more or less toxic than nitrogen dioxide, information on air quality obtained by most of the commonly used techniques could be mis- leading when interpreted in terms of the toxicity of nitrogen dioxide alone. In addition, the possible formation of nitrous and nitric acid vapours and their ready adsorption from the gas phase could account for some of the discrepancies in the determination of the Saltzman factor,S which is applied in the calibration of Griess - Saltzman reagents.On the basis of the cell reaction proposed in this study, this type of cell has obvious potential as an absolute monitor for nitrogen dioxide. At low flow-rates and low concen- trations, the output current is related to the nitrogen dioxide concentration in the following equation (derived from Faraday's law) : .. - * (8) i = 0.0669 x fC .. .. where i is the galvanic current in pA ;fthe sample flow-rate in ml min-l at 20 "C and 1013 mbar; and C the nitrogen dioxide concentration in p.p.m. V/V. At a flow-rate of 50 ml min-l, this expression represents a sensitivity of 3.34 p 4 per p.p.m. V/V for nitrogen dioxide concentra- tions up to and exceeding those permitted in ambient air and in industrial atmospheres.Cells in regular use for periods in excess of 1 year have shown no change in this sensitivity. The output of the cell at high nitrogen dioxide concentrations appears to be limited by the rate of adsorption and reaction of nitrogen dioxide at the cathode. The total mass of nitrogen dioxide that can be reduced by a cell is limited by the reducing capacity of the carbon anode, the approaching end of its useful lifetime being indicated by an increase in the response time of the cell. The zero stability and signal to noise ratio of the amplified cell output indicate that a detection limit of less than 0.001 p.p.m. V/V is possible. I t is apparent from the graphs in Figs. 2 and 3 that, at very high sample flow-rates, the output becomes independent of flow-rate for a given concentration of nitrogen dioxide.As the coulometric efficiency is very low in this mode of operation, the response is no longer predictable and may vary as a function of temperature, activity and degree of wetting of the cathode, etc. In addition, the high rate of evaporation of the electrolyte at high sample flow-rates necessitates the use of a much larger electrolyte reservoir. Consequently, there is little advantage in operating the cell in this mode, which is independent of flow-rate. Because a coulometric efficiency of 100 per cent. can be readily achieved with this cell, an obvious application is in the standardisation of low-concentration mixtures of nitrogen dioxide with air or nitrogen that are required for the calibration of more specific air-pollution monitors.As has been noted above, bottled mixtures of nitrogen dioxide may be unstable, particularly at very low concentrations. Permeation devices for the preparation of nitrogen dioxide standards require precisely controlled conditions of temperature and 1 iumidity, and are unreliable unless frequently calibrated gravimetricallyll or volumetrically.12 The galvanic cell, however, permits the continuous determination of nitrogen dioxide in a gas stream, regard- less of the method by which it was prepared, with an accuracy limited only by that of current and flow-rate measurements. I t is possible either to measure the concentration by applying equation (8) to the linear region of a graph of galvanic current zleyszts sample flow-rate, as in Fig.3, or, if the response characteristics of the particular cell are known, the galvanic current can be monitored at a constant flowrate. For example, the cell depicted in Fig, 1, the response770 ALLEN characteristics of which are shown in Figs. 3 and 4, exhibits a coulometric efficiency of 100 per cent. at a flow-rate of 50mlmin-1 for concentrations of nitrogen dioxide up to about 60 p.p.m. V/V, i.e., a galvanic current of about 200pA. For monitoring nitrogen dioxide for air pollution purposes, the galvanic cell alone suffers from the disadvantage of poor specificity. Other pollutants that interfere include ozone, sulphur dioxide, hydrogen sulphide, alkyl thiols and halogens, all of which are claimed to be selectively removed by a scrubber consisting of purified silver per0xide.1~ However, any silver carbonate impurity in this scrubber causes excessive absorption of nitrogen dioxide.It is possible that increased specificity to nitrogen dioxide can be achieved by making a modification to the cell itself. Nitric oxide can be monitored specifically by first scrubbing nitrogen dioxide from the sample by using triethanolamine coated on an inert support14 and then oxidising nitric oxide to nitrogen dioxide by means of a solid oxidant based on acidified manganese( IV) oxide, permanganates,* or chromium(V1) oxide.15 However, these oxidants tend to be affected by their moisture content, resulting either in poor oxidation efficiency or in excessive absorption of nitrogen dioxide. Further research is therefore required in order to investigate the application of this galvanic cell to the reliable and selective monitoring of nitrogen dioxide and nitric oxide in ambient air.CONCLUSIONS The galvanic cell described has been shown to respond to low concentrations of nitrogen dioxide, at low sample flow-rates, with a coulometric efficiency of 100 per cent. on the basis of 1 Faraday per mole of nitrogen dioxide. It offers a simple solution to the problem of calibrating low-concentration mixtures containing nitrogen dioxide and has obvious potential as an inexpensive, absolute monitor of nitrogen dioxide in ambient air. The author is grateful to the Director, Research and Development Division, Watson House and the British Gas Corporation for permission to publish this paper.The author is also grateful to Mr. K. A. Goode, London Research Station, British Gas Corporation, for his help in assessing the performance of the cell and for much stimulating discussion. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Hersch, P., and Deuringer, R., Annlyt. Chem., 1963, 35, 897. Hersch, P., in Reilley, C. N., Editor, “Advances in Analytical Chemistry and Instrumentation,” Hersch, P., and Deuringer, R., U.S. Patent 3,314,863, 1967. -- , J . A i r Pollut. Control Ass., 1963, 13, 538. Hodgeson, J . h., Bell, J . P., Rehme, K. A., Krost, K. J., and Stevens, R. I<., AIAX Paper No. 71- 1067, Joint Conference on Sensing of Environmental Pollutants, California, 1971. Breitenbach, L. P., and Shelef, M., J . A i r Pollut. Control Ass., 1973, 23, 128. Hartkamp, H., SchrReihe Landesanst. Immissions- at. Hode.rzizzttzungsschutz Laizdes Nordrhein- Allen, J . D., J . Inst. Fuel, 1973, 46, 123. Nash, T., Aq&. Occup. Hyg., 1968, 11, 235. Scaringelli, F. P., O’Keefe, A. E., Rosenberg, E., and Bell, J. I?., Analyt. Chem., 1970, 42, 871. Saltzman, B. E., Burg, VJ. R., and Ramaswamy, G., Envir. Sci. Technol., 1971, 5, 1121. Harman, J. N., dnu. I . S . A . Conf. Proc., 1971, 26, 554. Levaggi, D. A., Siu, W., Feldstein, M., and Kothny, E. L., Envir. Sci. Technol., 1972, 6, 250. Levaggi, D. -- , , paper presented at Anachem Conference, Detroit, 1963. Volume 3 , John Wiley and Sons, New York, 1964, p. 183. Westf., Essen, 1970, 18, 55. Kothny, E. L., Belsky, T., de Vera, E., and Muellcr, P. K., Ibid., 1974, 8, 348. Received March 7th, 1974 Accepted June 18th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900765

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99, p p . 771-773 771 The Determination of Mercury(I1) by Radiochemical Replacement with Silver- 1 lorn BY H. J. M. BOWEN (Chemistry Department, Reading University, Whiteknights, Reading, Bsrkshive) Mercury(I1) in aqueous solution can be determined by shaking with a dilute solution of silver-1 loin dibutyldithiocarbamate in chloroform, and counting the silver-1 10m extracted into the aqueous layer. Alternatively, the reagent in chloroform can be shaken with up to fifty times its volume of water, and the mercury present measured by the decrease in specific activity of the chloroform layer. The method is rapid, has a minimum sensitivity of 1 ng ~ r n - ~ of mercury and a precision of &5 per cent. Calibration graphs are linear, and the technique would be suitable for use in the field, in order to avoid losses of mercury during the storage or transportation of aqueous samples.Interference by chloride ions can be serious, but is completely overcome by adding potassium cyanide. Gold and palladium are the only metals that are likely to interfere. THE determination of small amounts of mercury in aqueous solutions is of interest to workers who are concerned with pollution. Mercury forms strong bonds with ligands that contain sulphur or nitrogen, which suggests that mercury(I1) might be determined by replacement of a less noble metal from a chelated derivative. For instance, Starg and Kratzerl have shown that the extractability of chelates with diethyldithiocarbamates into carbon tetrachloride follows the order palladium > mercury(I1) > silver(1) > copper(I1) >.. . . Further, Star9 and Burc12 have shown that mercury(I1) in aqueous solutions can replace silver from its diethyldithiocarbamate, when shaken with the latter substance in carbon tetrachloride. Sedivec and Vasak3 have suggested the determination of mercury by exchange with copper diethyldithiocarbamate. It is surprising that while silver diethyldithiocarbamate is relatively insoluble in both chloroform and carbon tetrachloride, the dibutyl derivative is very soluble in chloroform, but sparingly soluble in carbon tetrachloride. It was therefore decided to investigate the heterogeneous replacement reaction “9. chloroform chloroform aq. where D represents dibutyldithiocarbamate and X is a negative ligand, by using radioactive silver-1 10m.EXPERIMENTAL AND RESULTS HgX, + 2AgD = HgD2 + 2AgX .. ’ - (1) The method of Takiguchu, Abe, Kurosaki, Asada and Nakagome4 was used to synthesise labelled silver dibutyldithiocarbamate. Silver nitrate (6.9 g), containing 0.1 to 1 mCi of silver-llOm, was dissolved in 30 cm3 of water in a 200-cm3 round-bottomed flask. Bis(dibuty1- thiocarbamoyl) disulphide (8.3 g), dissolved in 30 cm3 of chloroform, were added and the mixture was heated under reflux for 1 hour. The dark non-aqueous layer was separated from the aqueous layer, washed with 2 M sodium hydroxide solution and water, and then distilled down to about 15 cm3. The remaining oil was poured into about 300 cm3 of diethyl ether, when a reddish powder of silver dibutyldithiocarbamate settled out.This powder was filtered off, washed with ether and dried. The yield was 86.6 per cent. of the theoretical. It was highly coloured, stable in air, insoluble in water and diethyl ether, but very soluble in chloroform and ethanol. Stock solutions were made up in chloroform, but before making up to volume they were washed three or more times with water in order to remove any soluble impurities. Preliminary experiments with 0-002 hi silver dibutyldithiocarbamate and 0.001 nz mercury( 11) nitrate or acetate showed that mercury(I1) quantitatively replaced silver in reaction (1) after shaking for 1 minute (Fig. 1). The blank obtained with distilled water was less than 0.8 nmol of silver, and the exchange was equally rapid in 0.01 M nitric acid and in water.I t was then found that 0.01 M hydrochloric acid largely inhibited the release of radio- active silver to the aqueous layer, although the colour of the reagent diminished in the chloro- 1Q SAC and the author.772 BOWEN : THE DETERMINATION OF MERCURY(II) [Analyst, Vol. 99 Mercury added/pmol Fig. 1. Replacement of silver in a chloroform solution of AgD by shaking with an aqueous solution containing mercury(I1) form layer, showing that a reaction was still taking place. It seems that silver chloride is more soluble in chloroform than it is in water. This solubility had not been measured before, but after shaking solid silver-ll0m chloride with chloroform for 20 minutes at 18 O C , a solubility of 10.3 pg ~ m - ~ of silver chloride in chloroform was found.This solubility is about seven times as high as the solubility of silver chloride in water at this temperature. Experi- ments with very dilute solutions of silver-ll0m chloride chowed that the amount of silver chloride extracted from water into chloroform could be as high as 75 per cent. (this result was not reproducible), that this percentage declined markedly as the amount of silver present in the aqueous layer increased and that chloroform extracted much more silver chloride from water than did either methylene dichloride or carbon tetrachloride, The interference by chloride would appear to reduce the practical value of the technique. The addition of ammonia, in an attempt to make silver chloride dissolve in the aqueous layer, was found to cause difficulties by forming colloidal aminomercury( 11) chloride.However, the addition of potassium cyanide was found to overcome the interference due to chloride. Sufficient cyanide to complex all the silver liberated was added. In practice, 5 mol of cyanide were added for each mole of silver dibutyldithiocarbamate present, and this amount brought the pH to 8 to 9. Cyanide was shown to affect neither the blank nor the rate of equilibration of the system. There may be some advantage in converting all the mercury in the sample into the mercuricyanide anion before analysis, as this anion is not adsorbed strongly on surfaces. At low concentrations, both silver and mercury are strongly adsorbed by poly- ethylene,6 for example, but can be desorbed by soaking in dilute potassium cyanide solution.The main disadvantages are the need to keep samples alkaline and the problem of disposal of waste cyanide solutions. Interference might also be expected from other noble metals such as gold and palladium. Gold(II1) chloride solution was brought to pH 6.8 and made 0.04 iu in potassium cyanide; the final pH was 8-4 and the gold was thought to be present as potassium dicyanoaurate(1). Exchange with silver dibutyldithiocarbamate in chloroform was relatively slow, and was only 60 to 80 per cent. complete after shaking for 5 minutes. Palladium nitrate in 0.01 M potassium cyanide solution exchanged even more slowly, and exchange was only 20 per cent. complete after shaking for 1 minute. Tests were made in order to establish how large a volume of water could be extracted effectively with a small volume of silver dibutyldithiocarbamate in chloroform.A 10-cm3 volume (14.7 g) of 0.0002 M silver-1 10m dibutyldithiocarbamate in chloroform was shaken with volumes of water ranging from 10 cm3 to 1 litre. The lower layer was run off as com- pletely as possible, weighed, made up to 15.0 g and finally counted in order to determine its specific activity. The results (Fig. 2) showed that more than 50 per cent. of the reagent could be recovered with its specific activity unchanged when up to 500 cm3 of water were used. The recovery fell to 20 per cent. when 1 litre of water was used, and the specific activityNovember, 19741 BY RADIOCHEMICAL REPLACEMENT WITH SILVER-1 10m 773 increased. It seems that large volumes of water dissolve more chloroform than they do silver dibutyldithiocarbamate. The recovery of mercury from spiked samples of distilled water, local tap water and very dilute nitric acid was satisfactory when 10 cm3 of 0.0002 M silver-l10m dibutyldithiocarbamate in chloroform was used to extract up to fifty times its volume of aqueous solution.In these experiments the percentage of silver exchanged was determined from the decrease in specific activity in the chloroform layer. When using this technique, the limit of sensitivity was found to be about 1 ng ~ m - ~ of mercury, or about 0-5 pg of mercury per sample, and the over-all standard deviation was &5-4 per cent. of the mean. - 100 250 500 1000 Volume of water/cm3 Fig. 2. Effect of shaking various volumes of water on 10 cm3 of 0.0002 M1lomAgD in chloroform: x = percentage of chloroform layer recovered; o = initial specific activity in chloroform layer, per cent.DISCUSSION Several workers5-7 have noticed that natural waters containing less than 50 ng cm-3 of mercury can readily lose mercury, either by absorption or by volatilisation. This leads to difficulties with the preservation of both samples and standard solutions containing mercury. The present method of analysis has the advantage that it could be used in the field, as it requires a minimum of apparatus. Samples of water of 0.5 litre could be collected directly into a separating funnel containing 5 cm3 of 0.002 M aqueous potassium cyanide and 10 cm3 of silver-l10m dibutyldithiocarbamate in chloroform. After shaking for 1 minute, the aqueous layer could be rejected and the chloroform layer run off into a weighed tube for subsequent weighing and counting. Large volumes of water would not have to be trans- ported to the counting laboratory, as the volume is reduced by 98 per cent., and problems of loss of mercury on storage would be eliminated. No appreciable radiation hazard should arise as the amount of silver-ll0m used per sample is only about 0.1 pCi. REFERENCES 1. 2. 3. 4. 5. 6. 7. Star?, J., and Kratzer, I<., Analytica Chim. Acta, 1968, 40, 93. Star$, J., and Burcl, R., Radiochem. Radioanalyt. Lett., 1971, 7, 235. Sedivec, V., and Vasak, V., Chemicke' Listy., 1951, 45, 435. Takiguchi, T., Abe, M., Kurosaki, K., Asada, E., and Nakagome, M., Kogyo Kagahu Zasslii, 1967, 70, 1182; Cheirt. Abstr., 1968, 68, 22046. Feldman, C., Analyt. Chem., 1974, 46, 99. Lindstrom, O., Ibid., 1959, 31, 461. Chau, Y. K., and Saitoh, H., Envir. Sci. Technol., 1970, 4, 839. Received May 6th, 1974 Accepted June 26th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900771

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

774 Anahst, November, 1974, Vol. 99, $9. 774-781 The Control Analysis of Boron by Measurement of the Transmission of Radioisotope Source Neutrons By T. B. PIERCE, C. R. BOSWELL" AND P. F. PECK (Aeplied Chemistry Division, Atomic Energy Research Establishment, Harwell, Didcot, Oxfordshire, OX1 1 ORA) Boron a t levels of between 0.1 and 10 per cent. has been determined non- destructively in bulk solids such as boron - aluminium alloys by measurement of the intensity of a transmitted neutron beam. The neutrons were produced from a 300-mCi 241Am - Be source and employed in an automatic instrumental system developed for use in a control laboratory at a rate of up to thirty-six samples per hour. The accuracy of such a technique has been shown to be good and the precision is better than f I per cent.The method is best applied to the determination of boron in matrices that have low thermal neutron absorption or scattering cross-sections. THE analytical determination of boron in a variety of matrices by conventional chemical techniques can be time consuming and may make heavy demands on laboratory effort, particularly when large numbers of samples are required to be processed. Consequently, an alternative method is desirable for routine application, which can offer some economy in effort and permits rapid acquisition of analytical data from individual samples. The un- usually high neutron absorption cross-section of boron-10 at low neutron energies is a par- ticular characteristic of boron, which might well be exploited so as to provide the basis of a rapid analytical method, given that the measured indication of neutron absorption by boron can be identified from the information provided by the analytical process.In addition, penetration of neutrons is an attractive concept for some determinations, particularly for the examination of solid materials, as bulk samples can be interrogated, thus simplifying methods of sample preparation and reducing the sensitivity to inhomogeneity and to surface effects. Boron has previously been determined by measuring the intensity of the 0.478-MeV gamma-line emitted during the (1) - (0) transition in the 7Li formed as a result of the reaction 1°B(n,a)7Li following the absorption of reactor neutr0ns.l While meas- urement of the prompt gamma-radiation following neutron capture can provide a satisfactory method for the determination of boron down to levels of the order of a few parts per million, and the contribution due to boron in the total gamma-ray spectrum can be uniquely identified by gamma-ray spectroscopy, suitable reactor facilities are required and the use of appropriate equipment is necessary in order to distinguish the lithium-7 radiation from that of the back- ground and from any others produced in the sample.Clearly, for use in a control laboratory, an alternative method is desirable in which a smaller neutron source and simpler ancillary equipment are employed, yet which provides results that can be simply and rapidly inter- preted in terms of boron concentration. Specifications of instrumentation for control analysis can sometimes be less stringent than those required for more general research pur- poses, as the materials examined are likely to conform to a few relatively well defined com- positions, and the versatility of a general analytical method that can cope with a wide range of compositions is unnecessary.The technique can therefore be designed to suit particular materials of clearly defined specification, and selectivity can therefore be forfeited in order to increase convenience. An alternative method to the determination of boron by measurement of the induced gamma- radiation is offered when the element makes a significant contribution to the macroscopic neutron absorption or scattering cross-section of the sample, as the total attenuation of the neutron beam that passes through the sample can then be measured.Such transmission determinations offer the advantages that only very simple counting equipment is necessary; low-intensity neutron sources are often adequate because the primary source radiation is * Present address : Computer Unit, Massey University, Palmerston North, New Zealand. @ SAC; Crown Copyright Reserved.PIERCE, BOSWELL AND PECK 775 counted; and difficulties that can accrue owing to the long-term use of gamma-ray detectors positioned near to the neutron source are avoided. The use of radioisotope sources for inclusion in control instrumentation is attractive, because their reliability, certainty of performance and small size ensure the availability of neutrons and permit their incorporation into equipment of a size suitable for installation in an analytical laboratory.A major consideration associated with the use of such sources is a need for sufficient shielding to reduce the radiation to acceptable levels, so that source strength should be kept to the minimum compatible with satisfactory analytical response. The energies of neutrons from radioisotope sources are dependent upon the type of emitter and are generally high, so that they must be reduced to those values at which boron has a high absorp- tion cross-section. A neutron beam must be extracted from the moderator in a suitable form so as to permit interrogation of the specific samples to be analysed, but the beam in- tensity must be adequate for the analysis to be completed in a time that satisfies requirements either for the speed of analysis or for the throughput of samples.This paper summarises experience gained with the measurement of boron by neutron transmission using a monitor that was specifically designed to function in a control laboratory and is based upon the use of a radioisotope neutron source. EXPERIMENTAL A 300-mCi americium-241 - beryllium source (total output 6.6 x lo5 neutrons s-1) was chosen to provide neutrons for the transmission system as, although the energy of the neutrons was relatively high, the long half-life of americium-241 (t* = 458 years) ensured that there was little variation in neutron output over long periods of time, and that frequent correction of calibration graphs was unnecessary. The source was housed in a block of polythene so as to moderate the source neutrons and the polythene was surrounded by boron-doped polyester, in order both to reduce the intensity of radiation on the outside of the shielding to below the levels permitted for unclassified workers and to provide a firm housing so as to enable the automatic sample transport system to operate satisfactorily. The top of the source block was covered with a carefully machined layer of boron-doped epoxy resin and contained a suitably shaped aperture to enable a collimated neutron beam to be extracted; a second block of boron-doped epoxy resin was placed on top of the source block to house the neutron detector and to complete the shielding necessary to ensure that the radiation levels around the neutron source were acceptable.Neutrons were detected with a europium-activated lithium iodide scintillator and the output from the detector was processed by Hanvell 6000 series modular electronic units. A scintillator thickness of 2 mm was chosen in order to give 96 per cent. efficiency for the detection of thermal neutrons but with relatively low sensitivity both to energetic neutrons and to high-energy gamma-radiation, which might provide pulse heights similar to those from neutrons. In practice, the detector output, after amplification, showed a clear peak due to the neutrons well above the contri- bution due to the low-energy radiation from americium-241. This permitted a simple discriminator to be set below the neutron peak so as to provide a measure of the neutron count.Samples were in the form of discs 1& inches (2.86 cm) in diameter and were placed for analysis around the periphery of a turntable with positions for thirty-six separate discs. After the completion of each count, the turntable was automatically advanced to feed the next sample into the neutron beam between source and detector so that all samples in the turntable could be interrogated in sequence. The neutron count was printed out after the completion of each measurement together with the contents of an additional scaler, which was increased each time the turntable was advanced so as to maintain a record of the number of samples interrogated. Counting time was pre-set by controls on the timer and was usually 90 s. As the transmission monitor was designed for control analyses, electronic units were placed behind a lockable door and only the minimum functions necessary for routine use were made available on the control panel.These functions included a mains on - off switch, manual turntable advance for loading samples, a means of clearing neutron and sample number scalers prior to the start of the analysis, and start and stop switches. After the electronic units had initially been set to provide suitable counting conditions, these simple controls were adequate for routine use. An additional indication of the progress of the776 PIERCE et al.: THE CONTROL ANALYSIS OF BORON BY MEASUREMENT [Analyst, Vol. 99 analysis was provided by a digital display, which enabled the contents of either scaler to be shown. The neutron transmission monitor is shown in Fig.1 ; the over-all cost of the complete automatic system was of the order of E4500. The standards of known composition were prepared by mixing compounds of known boron content with an appropriate matrix, both of which were in a finely divided form. A homo- geneous mixture was achieved by mixing by mechanical means and the powder produced was then pressed into discs with a small laboratory hydraulic press. Measurement of several discs prepared from the same mixture showed that good distribution of boron through the matrix can be achieved with these powder techniques, and satisfactory calibration graphs were obtained when individual matrices were doped with different amounts of boron. PRINCIPLES OF THE METHOD- The principles of neutron transmission for analytical application have been described in detail elsewhere2 but little effort has been devoted to applying the technique to analytical measurement, although the determination of boron3 and hydrogen4 has been considered.The transmission ( T ) of a uni-directional neutron beam through a sample is given by- where I and I,, are the intensities of the neutron beam with and without the sample in position. The transmission is related to the number of atoms in the beam by the equation- where ni is the number of atoms of the ith kind per square centimetre and ui is the cross-section of the ith atom in square centimetres. When a single element is responsible for virtually all of the neutron absorption equation (2) can be rewritten .. - * (1) .. * * (2) T = r/r, .. .. . . -1nT = C npi . . . . . . . . . . * * (3) . . .. 1 w T M In -=- x No where W is the mass of element per square centimetre, M is the relative atomic mass of the element and N is the Avogadro number. By this means, transmission can be related to the mass of absorbing species per square centimetre. The sensitivity of the analytical technique will therefore depend on the cross- section for removal of neutrons from the beam, which in turn will be a function of neutron energy, so that the greatest sensitivity will be achieved if transmission is measured at energies for which cross-sections are high. While analysis of neutron transmission at selected energies is possible with elaborate instrument systems using high-intensity neutron sources, the flexibility of instruments based on low-output radioisotope sources is more limited and, in order to avoid any instrumental complexity that might detract from the simplicity of the method, the only modification made to the distribution of the energies of the source neutrons in the transmission monitor was to increase the low-energy component by placing the source in a moderator of polythene.Clearly, with a broad distribution of energies in the neutron beam, the effect of neutron absorption by boron will be to harden the spectrum of the neutrons transmitted through the sample. In spite of this variation in distribution of neutron energies, it is desirable for the analytical response of the monitor to follow some convenient relationship such as that given in equation (2), so that interpretation of raw data in terms that are analytically significant is relatively straightforward and neither makes a heavy demand on time nor requires the use of highly skilled laboratory personnel.The response of a laboratory neutron transmission system may be further complicated by the fact that the neutron beam is not uni-directional, because although some collimation of the neutron beam emitted from the source block is possible by judicious use of suitable absorbers and filters, the source block, sample and detector are likely to be placed in close proximity so as to maximise the count-rate obtained from the source of a particular size. If the response of the instrument can be satisfactorily interpreted in terms of equation (3), then the attenuation of a neutron beam by a two-component sample can be written-Fig.1. Neutron transmission monitor. Dimensions : heixht, 47 inches ; width, 26.5 inches; and depth, 37 inches To face page 776November, 19741 OF THE TRANSMISSION OF RADIOISOTOPE SOURCE NEUTRONS 777 where subscripts a and b refer to the two different species contributing to beam attenuation. Equation (4) can be rewritten- where S is the mass of sample in grams per square centimetre. Thus the low levels at which any single element can be determined will depend upon the relative values of the contributions to the absorption due to all of the different components present in the sample, and the effect of the matrix on beam attenuation should be low if good sensitivity is to be achieved.RESULTS AND DISCUSSION In order to identify the proportion of the neutron beam available for boron determinations, the neutron count with the open beam (I,) and with a disc of boron - aluminium alloy con- taining 200 mg cm-2 of boron in the beam (I,) was measured, and the boron ratio (RB) was calculated as follows : . . . . . . . . * - (6) I0 - 1, R, = ___ I , R, was critically dependent upon the positioning of the neutron detector relative to the source block, and generally lay between 3 and 12, depending upon the geometry chosen. A value of approximately 9:l was usually achieved at a convenient setting of the detector that was suitable for automatic operation of the transmission analyser when examining samples in the form of discs. A high count-rate of neutrons that can be absorbed by the boron- aluminium disc ( I , - I,) is desirable, in addition to a high R, value, as the statistics of the over-all count accumulated during the counting period will impose an ultimate limitation upon the analytical precision obtainable, and will govern both counting time and source size that are acceptable, The maximum counting time allowable may be dictated by the sample throughput demanded of the analyser, and increasing the source size to attain a more rapid speed of response is not always satisfactory because the size of the shielding required to reduce the source radiation to acceptable levels will affect the convenience of the instrument ; a typical I , value found with a 300-mCi americium - beryllium neutron source was 310 counts s-l.A counting time of 90 s was usually chosen for boron determinations for control purposes, as it provided adequate precision for measurement while still permitting a high sample through- put of about thirty-seven samples per hour, although longer counts were employed when higher precision was a particular consideration. Machine reproducibility was initially checked by allowing the transmission monitor to run without any samples on the turntable, thus enabling a number of I , counts to be collected but with the turntable advanced between measurements. The standard deviation on 140 determinations, each to about 125 000 counts, was found to be 0.30 per cent. (percentage of the mean) compared with 0.28 per cent. expected from statistics, indicating that the precision of measurement was primarily governed by the number of neutron counts accumulated.Discs pressed from boron - aluminium powders were then loaded on to the turntable in order to confirm that satisfactory reproducibility could be obtained with samples in position, and the machine was allowed to run until each sample had been examined at least seven times. This procedure was repeated until data had been obtained for counting times of 1-5, 5 and 15 minutes; coefficients of variation found are given in Table I. There was some evidence to suggest that poorer reproducibilities resulted from movement of inhomogeneous samples across the holders on the turntable when it was advanced, thus introducing various amounts of boron into the neutron beam, but this type of error could be avoided if discs were a good fit in the sample holders on the turntable.Because of the log-linear dependence of trans- mission upon boron content, the precision for boron determinations will be less than that of the recorded count. Analytical determination of boron in samples of unknown composition was achieved by comparing the neutron transmission through these samples with the transmission through standards. The convenience and accuracy of the technique, therefore, depends to a large extent upon the ability to prepare relatively simply suitable standards of known composition.778 PIERCE 6’t d.: THE CONTROL ANALYSIS OF BORON BY MEASUREMENT [Analyst, VOl. 99 TABLE I COEFFICIENTS OF VARIATION, PER CENT., FOUND FOR REPEAT EXAMINATIONS OF BORON - ALUMINIUM PRESSED DISCS Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Counting timelminutes ------7 1.5 5 15 0.44 0-37 0.28 0.39 0.25 0.14 0.55 0.20 0.24 0.62 0.45 0.23 0.58 0.2 1 0.26 0.55 0.36 0.22 0.74 0.40 0.27 0.94 0.23 0.22 1.15 0.22 0.24 0.96 0.36 0.25 0.73 0.11 0.29 0-85 0.49 0.33 0.85 0.33 0.21 1.02 0.52 0.30 0.95 0.30 0.33 0.77 0.63 0.24 0.79 0.38 0.39 0.80 0.25 0.28 1.00 0.99 0.57 0.59 0.62 0.30 Boron content of sample, per cent. 0.20 0.21 0.4 1 0.42 0.83 0.83 1-24 1.25 1.67 1.68 2-47 2.50 3-31 3.35 3.69 3.79 4.14 4.18 4.57 4.93 Once prepared, standards can be re-used as the transmissions through discs pressed from carefully prepared mixtures of aluminium and a compound of boron of known composition were found to be reproducible; results for counts obtained on several discs pressed from the same powder mixture are given in Table 11.The variation of transmission with boron content through boron - aluminium standards for 5-minute counts is given in Fig. 2 and provides a calibration graph that can be used for the interpolation of transmission from samples of unknown composition. An alternative method of data interpretation is provided by direct calculation of boron contents from raw data, if the cross-sections for removal of neutrons from the beam are known and if the response of the analytical system is given by equation (5). Clearly, however, the neutron beam from the isotope source and simple collimator used in the transmission monitor cannot be expected to provide such satisfactory characteristics for analytical work as a specially designed reactor neutron beam.Computer analysis of the calibration graph given in Fig. 2 showed that data could be fitted to equation (5) over the range of boron contents examined and cross-sections for boron and aluminium were calculated as 624 b and 1-49 b, respectively. These cross-sections are lower than the total cross-sections for neutrons at 2200 m s-l and reflect the limitations of a neutron beam produced from a small source and simple collimator. Deviations from the TABLE I1 BORON RESULTS OBTAINED FROM TEN DISCS OF THE SAME STANDARD MATERIALS BY NEUTRON TRANSMISSION Boron in mixture 1, per cent. 1-07 1.14 1.14 1.06 1.10 1-14 1-15 1-12 1.12 1.09 Mean -f. 1.1 1 f 0.03 standard deviation Boron in mixture 2, per cent.3.62 3-59 3.60 3.50 3.63 3.65 3.62 3-63 3.69 3-63 3-02 f 0.05November, 19741 OF THE TRANSMISSION OF RADIOISOTOPE SOURCE NEUTRONS 779 response given by equation (5) were more marked at high boron levels but the equation was found to be essentially valid over the working range usually chosen for the instrument. 0 10 20 30 Boron/mg cm-* -t Fig. 2 . Calibration graph for the transmission of neutrons through boron - aluminium standards Results obtained for the analysis of a number of boron - aluminium materials of known composition are given in Table I11 and illustrate the use of the neutron transmission system for routine analysis. Samples and standards were placed around the periphery of the turn- table, and a calibration graph similar to that given in Fig.2 was constructed and used to interpret transmissions from the samples in terms of boron content. Good agreement between the neutron transmission results and the known boron content is demonstrated, with the transmission method not requiring samples to be processed chemically in any way. The working range of the transmission technique is governed at low boron concentrations by the need to distinguish between the relative effects of the boron and the matrix on attenuation of beam intensity, while with high boron contents, changes in transmission are relatively insensitive to boron concentration if samples are not thin. TABLE I11 BORON CONCENTRATIONS FOUND BY NEUTRON TRANSMISSION FOR SAMPLES OF KNOWN COMPOSITION Boron (nominal), per cent. 0.5 1.0 1.6 2.0 3-0 4.0 5.0 Boron found, per cent. 0.50 0-98 1.48 2-03 2.93 3.97 4.95 Sample thicknesses were generally chosen so as to provide boron levels varying from 1 to 50 mg cm-2, and the boron content usually lay in the range 0.1 to 10 per cent.The presence of hydrogen, lithium, cadmium and certain of the rare earths in the sample in appreciable amounts will result in high absorption by the matrix, thus raising the lowest level of boron that can conveniently be determined. A number of samples of the same composition but of different thicknesses were examined in order to confirm that beam charac- teristics did not introduce errors when sample thicknesses varied and their boron content was calculated from a knowledge of the boron and aluminium cross-sections. Results presented in Table IV show that over the range of thicknesses examined, no effects that could be attributed to changes in geometry can be identified.780 PIERCE et az.1 THE CONTROL ANALYSIS OF BORON BY MEASUREMENT [Analyst, VOl.99 TABLE IV RESULTS OBTAINED FROM DISCS OF VARIOUS THICKNESSES PREPARED FROM THE SAME BORON - ALUMINIUM COMPOSITION Boron content of disc/g cm-2 0.282 0.404 0.516 0.655 0.775 0.914 1.041 1.166 Mean f standard deviation Boron, per cent. 4.29 4.26 4.18 4.24 4-19 4.1 1 4.0 1 3.99 4.16 f 0.1 1 Extension of the transmission technique to lower levels of boron has been achieved by using a more intense neutron source, in this case 12.5 pg of californium-252, to provide better counting statistics. Over the range 15 to 50 p.p.m. of boron the standard deviation was found to be about 4 p.p.m.The measurement of boron at these low levels is of course highly dependent upon the matrix being well defined, and the presence of any impurity atoms of even modest cross-sections in the bulk matrix could lead to the introduction of substantial errors. Application of neutron transmission to the examination of materials of high boron content has been simplified by mixing the powdered material with a diluent of low cross- section, such as aluminium powder, and pressing the mixture into a pellet containing boron at an appropriate level to fall on the best working range of the transmission monitor. A number of experiments were carried out in order to examine the possibility of analysing powdered materials and so avoid the need to press each material into a pellet prior to examination. The powders were spread evenly over the surface of an aluminium counting tray of dimensions suitable to fit into the sample holders on the turntable, and the analytical procedure was carried out in the same way as for pressed discs.A calibration graph con- structed from loose powders (ferroboron) is shown in Fig. 3 and illustrates that a calibration graph can be constructed which is similar to that for pressed materials, but that the repro- ducibility of sample preparation with loose powders is somewhat poorer. The loose powder method has been utilised for a number of applications, when either the convenience of the method more than compensated for the poorer quality analytical results obtained, or the samples have been unsuitable for pressing. 0 10 20 30 40 50 60 Boron/mg cm-* - Fig. 3. Calibration graph for neutron transmission through aluminium - boron mixturesNovember, 19741 OF THE TRANSMISSION OF RADIOISOTOPE SOURCE NEUTRONS 781 The extension of neutron transmission to the determination of a wider range of com- positions than those described in this paper is clearly possible. More limited measurements of boron in other matrices with low cross-section have been carried out, and the technique has been extended to the determination of certain other elements of high cross-section, such as hydrogen, lithium and cadmium. The authors thank Mr. G. 11. Holmes and IClr. H. L. Giles of London and Scandinavian Metallurgical Co. Ltd. for supplying the analysed standards and for critical evaluation of the technique over an extended period. REFERENCES 1. 2. 3. 4. Garbrah, B. W., and Whitley, J. E., Analyt. Chern., 1967, 39, 345. Taylor, T. I., and Havens, W. W., in Berl, W. G., Editor, ‘‘Physical Methods in Chemical Analysis,” Ljunggren, K., and Christell, R., Atompraxis, 1964, 10, 259. Mott, W. E., and Rhodes, D. F., in “Radioisotope Instruments in Industry and Geophysics,” Received March 25th, 1974 Accepted June llth, 1974 Academic Press Inc., New York, 1956, Volume 111, p. 447. IAEA, Vienna, 1966, Volume 1, p. 347.

ISSN:0003-2654    

DOI:10.1039/AN9749900774

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

[Analyst, VOl. 99 782 Book Reviews LIPID ANALYSIS. ISOLATION, SEPARATION, IDENTIFICATION and STRUCTURAL ANALYSIS OF LIPIDS. Oxford, New York, Toronto, Sydney and The contents of this book are well indicated by the chapter headings : The structure, chemistry and occurrence of lipids ; The isolation of lipids from tissues ; Chromatographic and spectroscopic analysis of lipids, general principles ; The preparation of volatile derivatives of lipids ; The analysis of fatty acids; The analysis of simple lipid classes; The analysis of complex lipids; The analysis of molecular species of lipids; Enzymatic hydrolysis of lipids; The analysis and radioassay of iso- topically-labelled lipids ; and A summary. With the growing interest in lipids shown by workers in many different fields, an increasing number of laboratories are concerned with problems involving lipid analysis.Since many of the procedures to be employed are of recent development it is good to have them collected together in a single text. Analytical techniques are critically reviewed (over 600 references are cited) and the book contains practical details of all the procedures recommended by the author. It is very obvious that the author is himself familiar with the analytical procedures he describes and the book contains many useful practical hints and warnings which will be of help both t o the novice and to the experi- enced in this field. This book is likely to be of great value to many groups of scientists: to the teacher who wants to design experiments for students of chemistry and biochemistry; to the analyst who occasionally has to solve a lipid problem and wants guidance about analytical procedures; and to the increasing number of those who already consider themselves to be expert in some area of lipid analysis.The reviewer recommends Dr. Christie’s book without hesitation and expects his own copy t o be well used. F. D. GUNSTONE By WILLIAM W. CHRISTIE. Braunschweig : Pergamon Press. 1973. Price L6. Pp. xiv + 338. ABSORPTION SPECTRA IN THE ULTRAVIOLET AND VISIBLE REGION. Edited by L. LANG. Volume This volume takes the number of compounds for which spectra have been recorded in full detail in the series from 3213 to 3403. The spectra of aniline and some thirty relatively simple derivatives have been measured. Among other groups of substances well represented are pyridine, pyrimi- dine, quinoline, quinolone and quinoxaline derivatives.The quinones listed include a number of fully substituted complex compounds in the 1,4-benzoquinone group, 1,4-naphthoquinone and -anthraquinone and 1- or 2-halogenoanthraquinones with tetracene-5,12-quinone and pentacene- 6,13-quinone. Other compounds of interest are methyl violet, bromocresol purple, bromophenol blue, methyl yellow, methyl orange, benzyl orange, metanil yellow and tropaeolin 00. Data are supplied on the spectra of a range of complex heterocyclic substances and a number of sulphur- containing complex compounds. There are also curves for various silyl and silazane compounds. As is usual in the series the measurements are shown on one side of a sheet and the absorption curve on the other side.This arrangement has definite advantages to those who use the informa- tion supplied on pure substances in analysing less pure materials showing irrelevant absorption. R. A. MORTON XVIII. Pp. 429 (loose-leaf). Budapest: Akadhmiai Kiad6. 1973. Price k7. LASER MICRO-SPECTROCHEMICAL ANALYSIS. By H. MOENKE and L. MOENKE-BLANKENBURG. Translated by R. AUERBACH. Pp. viii + 253. London: Adam Hilger Ltd. 1973. Price L6. The laser, which can be used to concentrate a great amount of energy into a beam that can be focused to a minute spot, affords a unique “source unit” for spectrochemical excitation on the micro-scale and opens up a variety of new possibilities to the spectroscopist. This volume, although concerned with the subject of laser microspectrochemical analysis generally, specifically details the considerable experience gained by the authors with one particular commercial instrument which originated in Eastern Germany.The book is the English translation of the second German edition, which has been extensively revised and enlarged to take into account recent developments in instrumentation and working methods. Chapter 1 presents a brief review of the principles of operation of solid-state pulsed lasers as energy sources for micro-analysis by optical-emission spectroscopy and of the properties of their radiation. Although it would act as an adequate general introduction to the subject, this reviewNovember, 19741 BOOK REVIEWS 783 is insufficiently detailed or comprehensive to allow those workers contemplating the possible application of laser excitation to a particular problem to make an independent evaluation of the utility of different types of source.Chapter 2 describes the available optical and electronic units and their assembly for laser micro-spectrochemical analysis. Suitable lasers, microscope optics, spark sources and spectrographic equipment are reviewed. Chapter 3 details the procedure of laser microspectrochemical analysis and is specifically concerned with the effect of focal spot size and the extent of the irradiation effect, sample preparation, instrumental conditions, the procedure employed and the interpretation of the spectra. Chapter 4 is concerned with the application of laser microspectrochemical analysis to mineralogy, geology, metallurgy, archaeology, biology, medicine and forensic science.The behaviour of individual elements during laser excitation is then reviewed. The Chapter concludes with consideration of the quantitative aspects of the technique and comments on the possible trends of developments in the technique. The final chapter is devoted to a comparison of the performance of laser microspectrochemical analysis with that of other techniques, such as electron beam microanalysis and conventional spectrochemical analysis. This volunie is well produced and the translation has retained the obvious enthusiasm of the authors for their subject. The book can be recommended to all practising analytical chemists with an interest in spectrochemical analysis.G. F. KIRKBRIGHT HANDBOOK OF MICROMETHODS FOR THE BIOLOGICAL SCIENCES. H. LEDERER. Van Nostrand Reinhold Company. 1974. Price 46-90. By GEOKG KELETI and WILLIAM New York, Cincinnati, Toronto, London and Melbourne: This book, which is cheaply produced and highly priced, lists thirty-one recipes for the pre- paration of biological material for examination and, in a second section, fifty microchemical methods are described. The third section comprises twenty-five methods of “biological charac- terisation, ’, which are mainly immunological techniques. The main emphasis is on analysis of bacterial specimens and their components. The authors’ intention is to provide a stepwise account of each technique “so that the researcher, technician or student can, without need of a library search, analyse a biological preparation.” However, some of the chemical analyses listed could not be carried out quantitatively without reference to original papers.The descriptions of methods are brief and in the form of notes. In many instances a method ends with the instruction to read a colour at a particular wavelength in a spectrophotometer or sometimes, a t dual wavelengths, but no conversion factors are given. It is clearly intended that appropriate standards be em- ployed, but some of the compounds concerned (for example, dideoxysugars) are not easy t o obtain and a list of suppliers would be invaluable. Many methods suffer serious interference from other biological materials, but cautionary notes about this effect are rare, and it would be very easy for the beginner blindly to accept the results of these assays.In several instances when it is essential that commercial products for use in reagents should be purified before use, no mention of this need is made. Some methods emphasise the use of commercial products, which are more expensive and may be less readily available than the components of the reagents. It is ridiculous to omit the composition of reagents such as biuret and ninhydrin from a book of this type. The biological characterisation section suffers from the same defects, and would be confusing to a beginner. In conclusion, the authors’ idea of a methods book for students would be a worthy one if each method were unambiguously stated, with all accessory information included, but the present attempt falls short in a number of respects and could not be recommended even if sold at a realistic price.Reference should be made to more complete and authoritative works, of which several are available. R. A. D. WILLIAMS Pp. xvi + 166. ASCORBINOMETRIC TITRATIONS. By L. ERDEY and G. SVEHLA. Pp. 183. Budapest : Akademiai When one research group has specialised in a particular aspect of analytical development over many years, a vast amount of useful experience is gained and it is helpful if the work, which is necessarily scattered among many publications, can be brought together in a single fext. The advantage of this book is, therefore, that the detailed work of the late Professor Erdey and his co-workers on the applications of ascorbic acid in redox titrimetry is compiled within a single volume, together with other work on this subject.The fact that of 170 references 70 are to the Kiad6. 1973. Price k3-40.784 BOOK REVIEWS [Analyst, VOl. 99 authors’ own work, and that of the other references many are to general texts and reviews, indicates the contribution that the authors have made in this field. The book itself is well written and carefully laid out with sections on the characteristic proper- ties of ascorbic acid solutions, methods of end-point detection in ascorbinometric titrations, direct titrations with ascorbic acid, and indirect determinations that are possible with ascorbic acid. The last two sections contain detailed working procedures for the numerous determinations described. The fact that the authors are discussing their own work means that the text is a compilation of their results and is therefore completely uncritical.For example, different direct and indirect procedures for the same substance are given, with no indication of which is preferable. In reviewing the work in this field, it is disappointing that the amount of effort that has gone into the optimisation of procedures and interference studies has not resulted in more useful practical analytical applications. Many titrations involve slow reactions in which either catalysts or elevated temperatures are used to produce acceptable results, and it would be useful to have detailed kinetic studies to complement the equilibria - redox potential measurements. Despite these deficiencies, this is a useful book and it will be invaluable to analysts using ascorbinometric titrations. J.M. OTTAWAY HANDBOOK OF PROCESS STREAM ANALYSIS. By KENNETH J. CLEVETT. Pp. xviii + 470. Ellis Chichester : Ellis Horwood Ltd. Distributed by It is not surprising, therefore, that genuine efforts are continuously being made to by-pass the laboratory, if by so doing essential data can be made available more cheaply and/or more quickly. Today, on-line analytical instrumentation is a highly desirable feature of any chemical or related process, if optimum efficiencies are to be achieved. The analyst may devise or apply existing instrumental procedures to meet his immediate needs, but i t is usual, and rightly so, for him to hand over the development of so-called “Process Analysis” to someone else. From the brief curriculum vitae on the fly-leaf of the book’s cover, the author is obviously a man of wide experience and a specialist in this field of continuous on-line instrumentation; after a detailed perusal of all that appears inside the cover, congratulations are justified on the choice of layout, chapter headings, and the clarity with which the relevant information is presented. The largest of the book’s seventeen chapters deals with “Vapour-Phase Chromatography,” thereafter, in order of size, chapters occupying about the same number of pages are headed, “Oxygen Measurement, ” “Ultraviolet, Visible and Infrared Absorption,” “Density and Specific Gravity Measurement” and “Water Quality Measurement” ; other chapter titles selected a t random include, “Moisture Measurement,” “pH and Redox Measurement,” “Measurement of Octane Number” and “Sample-Handling Systems.” Each chapter is reinforced with adequate theoretical information to help the reader to make a better appraisal of the underlying principles involved.The final pages immediately before the Index contain a list of manufacturers’ addresses, and a 33-page section on “Analyser Specification Data,” which is arranged to correspond with the chapter headings. Where references are given, they appear a t the end of each chapter. The book is likely t o have the greatest appeal to the analyst involved in industrial quality control, and to the wide range of chemical and electronic engineers who specialise in the making and maintenance of units ’involved in this field of on-line instrumentation. Horzoood Series in Analytical Chemistry. John Wiley & Sons, New York, London, Sydney and Toronto. 1974. Price Q.5. Analytical chemistry is involved in almost every aspect of our technological age. W. T. ELWELL

ISSN:0003-2654    

DOI:10.1039/AN9749900782

出版商:RSC    

年代:1974

数据来源: RSC