摘要:

DECEMBER, 1969 THE ANALYST Vol. 94, No. I 125 Scheme of Silicate Analysis Based on the Lithium Metaborate Fusion Followed by Atomic-absorption Spectrophotometry BY J. C. VAN LOON AND C. M. PARISSIS (Department of Geology, University of Toronto, Toronto 5, Canada) A procedure and results are given for the determination of Si, Al, Ti, Fe, Ca, Mg, K, Na and Mn in twelve standard silicate rocks of widely varying compositions. A comprehensive study has been made of potential inter- ferences in order that the method should be applicable over a wide range of sample compositions. The lithium metaborate - nitric acid dissolution system is used to enable previously developed colorimetric and spectrographic procedures with similar parting solutions to be used for determining important silicate constituents that cannot be determined directly by atomic absorption. ONE of the main goals of silicate analysis over the years has been the development of a procedure that will allow the rapid analysis of large numbers of silicate rocks of widely ranging major element composition while yielding an accuracy acceptable to petrologists.Before the advent of atomic-absorption analysis in its present form, the available techniques usually required time-consuming separation steps or a large library of closely comparable rock standards to ensure the required accuracy. Atomic absorption with its much publicised relatively simpler pattern of interferences offered potential relief from these problems. How- ever, up to the present time no atomic-absorption scheme of analysis exists, to our knowledge, which has proved satisfactory for the determination of a large number of constituents of silicate rocks over a wide range of compositions of major elements.In addition, any such useful scheme, to be rapid, should yield a residual solution that can be used with existing procedures for the determination of those important constituents not advantageously deter- mined by atomic absorption. Several reports have appeared that deal with only a few of the constituents of silicate rocks. These have been summarised briefly in an earlier pub1ication.l In addition, several authors have proposed schemes of silicate analysis of a more comprehensive nature. Abbey233 and Belt,4 by using a hydrofluoric acid - perchloric acid - nitric acid dissolution procedure exposed to air, describe schemes for the determination of iron, magnesium, calcium, sodium, potassium and sodium, and potassium, magnesium, calcium, manganese and iron, respectively.Shapiro6 proposed the determination of silica, aluminium, iron, titanium and phosphorus by solution spectrophotometry, and calcium, magnesium, sodium, potassium and manganese by atomic absorption following a lithium metaborate fusion. No results are given for rock analyses with this procedure, and no investigation of interference in a solution of this kind appears to have been carried out. Silica determination would not be possible following such dissolution. 0 SAC and the authors. 10571058 VAN LOON AND PARISSIS: SCHEME OF SILICATE ANALYSIS [A?Zat?ySt, VOl.94 Recently Bernas6 and Langmyhr and Paus,' p 8 proposed related schemes of silicate analysis involving hydrofluoric acid dissolution in a closed tube. In the former scheme, results were given for a granite, diabase and two tektites, and in the latter for a granite, diabase, nepheline syenite, N.B.S. bauxite, soda feldspar, cement, glass sand and B.C.S. iron ore, basic slag and ferrosilicon. Silica, aluminium, iron, magnesium, sodium, potassium, calcium and titanium were determined in each instance. In addition, vanadium was determined by the former and manganese and chromium by the latter. Both procedures presumably require the tube to be opened and the sample exposed to the air after completion of the dissolution but before the addition of boric acid. No extensive investigation into interferences appears to have been carried out.While Bernas lists no inter-elemental interferences, Langmyhr and Paus record interferences in several instances. Presumably a separate sample solution would be required for important silicate ingredients, such as phosphorus, which are not covered by the proposed procedures. Apparently some difficulty is to be expected with this parting solution in certain instances because of the reported failure to dissolve the precipitated fluorides completely when dealing with clays8 In previous studies statements have been made that imply extensive applicability of the proposed scheme without giving confirmatory results. This has doubtless arisen from the observation that atomic-absorption techniques are less prone to some types of interferences than other techniques.In the following an attempt has been made to define the range over which the proposed method has been proved. While materials containing ingredients, and hence interference concentrations beyond these limits, can often be determined without diffi- culty, further testing is being carried out to obtain proof. Twelve silicate samples covering the range of compositions most often encountered in routine silicate analysis have been analysed. The lithium metaborate - nitric acid dissolution system was chosen for silica as it involves no problems and needs no special equipment. In addition, the use of the proposed reagents makes possible the determination of those constituents not readily determined by atomic absorption by using solution spectrophotometric9 or spectrographiclo procedures previously applied to similar parting solutions.For these reasons, and because the results are in excellent agreement with the preferred values cited, we recommend the following procedure, in which atomic absorption plays a major rhle, as the most versatile yet developed. EXPERIMENTAL REAGENTS- The following Specpure materials (obtainable from Johnson Matthey Ltd.) were used to prepare standard solutions : JM425 SiO,, JM815 Mn,O,, JM126 magnesium metal crystals, JM90 CaCO, and JM340 aluminium metal rods. The titanium metal wire, K5229f, was purchased from Electronic Space Products, Los Angeles, California. Sodium, potassium and iron standards were prepared from Fisher certified NaC1-S-271, KC1-P-217 and iron wire 1-185.Solutions for interference studies were prepared from analytical-reagent grade metals dissolved in a minimum amount of concentrated nitric acid, or from nitrate salts when possible. Lithium metaborate, No. 12598, obtained from K & K Laboratories, was purified.1° The product was heated in a platinum dish at 625" C for 30 minutes to reduce the volume being used for fusions. A 2.24 per cent. solution of lithium metaborate was prepared in dilute nitric acid (1 + 24) for use in the samples and standards. Lanthanum chloride, Alfa Inorganics (L.A. 103), was treated as follows. Dissolve 1OOg of the chloride in 3 litres of water. Precipitate the lanthanum with a 50 per cent. excess of oxalic acid without adjusting the pH.Following digestion of the mixture for 30 minutes on a warm hot-plate, filter it on a large filter-paper. Wash the precipitate with water until all the chloride is removed. Dry the material on the filter-paper at 110" C. Transfer to a platinum dish and ignite a t 1000" C to form the oxide. A 10 per cent. solution of lanthanum was prepared by adding the stoicheiometric amount of hydrochloric acid and diluting. APPARATW S- A Perkin-Elmer 303 atomic-absorption spectrophotometer . was used with either the Boling burner head or the standard nitrous oxide burner. The instrument parameters used are listed in Table I.December, 19691 BASED ON THE LITHIUM METABORATE FUSION TABLE I INSTRUMENT PARAMETERS Element Silicon . . .. Aluminium . . Titanium . . .. Magnesium .. Calcium . . . . Iron . . .. Sodium . . . . Potassium .. Manganese .. Oxidant Nitrous oxide Nitrous oxide Nitrous oxide Air Air Air Air Air Air Wavelength, Slit width, nm mm 251.6 1-0 309.3 0.3 364-3 1.0 285.2 3.0 422.7 1.0 248.3 0.3 589.0 0.3 766-5 3.0 279.5 1.0 Scale expansion usually required x l x l x 5 x l x l x l x l x5 x 5 1059 Lamp type Intensitron Intensitron Intensitron Intensitron Intensitron Intensitron Osram vapour Osram vapour Multi-element PROCEDURE FOR PREPARATION OF STOCK SOLUTIONS AND REFERENCE STANDARDS- Silica-Accurately weigh 0-4 g of the high purity silica powder into a pre-ignited graphite crucible* and mix with 2.8 g of lithium metaborate and 0.2 g of calcium oxide. Fuse the mixture at 925" C for 15 minutes. Swirl the contents gently and pour the melt into 150 ml of nitric acid (1 + 24) in a 250-ml plastic beaker.Introduce a Teflon stirring rod into the beaker and stir the solution magnetically without heating until the sample is completely dissolved. Remove the stirring rod and rinse, collecting the rinsings in the beaker. Wash the contents of the beaker into a 250-ml flask and dilute to volume with nitric acid (1 + 24). Prepare working solutions (containing 280 to 480 p.p.m.) as required by appropriate dilution with nitric acid and lithium metaborate solutions so that the final concentrations are 1 + 24 and 0.56 per cent., respectively. AZumirtiz&m-Accurately weigh 1.0 g of aluminium wire and dissolve it in the minimum amount of nitric acid (1 + 1) with the aid of a drop of mercury. Pass a stream of hydrogen sulphide gas through the solution for 1 hour and then filter it through a No.42 Whatman filter-paper. Boil the solution gently to remove dissolved hydrogen sulphide and transfer the solution to a l-litre calibrated flask. Adjust the nitric acid content so that the final concentration is 1 + 24 and dilute to volume with water. To prepare working solutions (containing 33 to 100 p.p.m.) combine an appropriate aliquot of the standard solution with lithium metaborate and nitric acid solutions so that the final concentrations of lithium metaborate and nitric acid are 0.56 per cent. and 1 + 24, respectively. To prepare working solutions to compare with samples containing more than 100 p.p.m. of calcium, iron or magnesium, add sufficient lanthanum to make its final con- centration 1 per cent.Titartiurn-Accurately weigh 0.5g of titanium wire into a 250-ml flask, add 50ml of sulphuric acid (1 + 1) and heat on a hot-plate until it is completely dissolved. Evaporate the remaining sulphuric acid, fuming nearly to dryness. By using nitric acid (1 + 24) transfer the contents to a l-litre beaker. Heat the sample and stir until completely dissolved by using a hot-plate with a magnetic stirrer (about 15 minutes). Wash the solution into a l-litre calibrated flask and dilute to volume with nitric acid (1 + 24). The solution is standardised gravimet ricall y. Attempts to make more concentrated titanium solutions invariably resulted in the production of insoluble residues. The solution, as prepared above, has a pH below 1.0 and is stable for several months.Dilute working solutions (containing 8 to 60 p.p.m.) can be prepared as required by diluting an appropriate aliquot of titanium standard solution with sufficient aluminium and hydrochloric acid solution to make the final concentration 750 to 1000 p.p.m. and 4 ~ , respectively. IYort-Accurately weigh 1.Og of iron wire and dissolve it in nitric acid solution con- taining 18 ml of concentrated nitric acid and 50 ml of water. Dilute to volume in a l-litre Bask so that the nitric acid concentration is 1 + 24. Prepare working solutions (containing 2 to 13 p.p.m.) by combining an aliquot of iron solution with sufficient lithium metaborate solution and nitric acid to bring the final concentrations to 0.56 per cent. and 1 + 24, respectively.* Graphite crucibles can be manufactured from high purity rod or are available from Ultra Carbon Corporation (A-6108), Bay City, Michigan.1060 [Analyst, Vol. 94 Calcium and magnesizcm-Accurately weigh 2*5g of calcium carbonate and 1-0 g of magnesium metal crystals and dissolve each separately in the minimum amount of 1 + 4 nitric acid. Dilute the solutions to 1 litre with water. Prepare working solutions (for calcium 0.75 to 7 p.p.m. and for magnesium 0.5 to 5 p.p.m.) containing 0.56 per cent. of lithium metaborate, 1 + 24 nitric acid and 1 per cent. of lanthanum after the final dilution. Sodium and potassium-Accurately weigh 2.5g and 1.9g of sodium and potassium chlorides, respectively, dissolve them separately in water and dilute each to 1 litre with water.Prepare working solutions (for sodium 0-25 to 1 p.p.m. and for potassium 0.25 to 1 p.p.m.), to which are added lithium metaborate and nitric acid to match the concentration in the sample solution. Manganese-Accurately weigh 1.4 g of trimanganese tetroxide (Mn,O,) and dissolve it in a minimum amount of concentrated hydrochloric acid. Dilute to volume with water. Working standards (containing 0.2 to 3.5 p.p.m.) are prepared to contain 0.56 per cent. of lithium metaborate and nitric acid at a concentration of 1 + 24. VAN LOON AND PARISSIS: SCHEME OF SILICATE ANALYSIS PROCEDURE FOR DETERMINATION- Pre-ignite an empty graphite crucible for 30 minutes at 950" C and then cool. Take care not to disturb the powdery inside surface. Mix 0-2 g of rock powder with 1.4 g of lithium metaborate in a porcelain crucible, and then transfer to the pre-ignited graphite crucible.Fuse at 900" C in a muffle furnace for exactly 15 minutes. Pour the melt into 100.0 ml of nitric acid (1 + 24) contained in a plastic beaker with a flat bottom, introduce a Teflon-coated plastic stirrer rod (of similar length to the beaker so that it reaches to the bottom of the beaker) and stir. Care should be taken not to exceed the recomended temperature because of possible slight loss of alkalis by volatilisation.1° Wash the solution from the beaker into a 250-ml flask and dilute to volume with nitric acid (1 + 24). Silicon-Aspirate the solution directly into a nitrous oxide flame. Aluminium-When the concentrations of calcium, magnesium and iron are below 100 p.p.m., as is usual, carry out the analysis directly on the sample solution with a nitrous oxide flame.When the concentration of any of these elements exceeds this amount add lanthanum to give a final concentration of 1 per cent. Titanium-Evaporate a 100-ml aliquot (2 x 50 ml) of sample solution in a 100-ml beaker to dryness on a water-bath. Add an amount of aluminium solution to the cake produced to give a final aluminium concentration of between 750 and 1000 p.p.m. Add 10 ml of 4 N hydrochloric acid and evaporate to dryness. Dissolve the residue in 10 ml of 4 N hydrochloric acid, transfer the solution to a 25-ml flask and make up to volume with 4 N hydrochloric acid. Mix the solution well and allow the gelatinous silica to settle at the bottom. Filter through a No.42 Whatman filter-paper and aspirate the filtrate against standards by using 5 or 10 times scale expansion. Iron-Dilute an aliquot of the sample solution, contained in a calibrated flask, with nitric acid (1 + 24) and add sufficient lithium metaborate to bring its final concentration to 0.56 per cent. Calcium-Dilute an appropriate aliquot of sample solution (the final silica content should be less than 160 p.p.m.), contained in a calibrated flask, with nitric acid (1 + 24). Add sufficient lithium metaborate and lanthanum solution to bring their final concentrations to 0.56 and 1 per cent., respectively. Magnesium-Treat the sample solution in a manner similar to calcium, except that no precaution is necessary to maintain the silica content below 160 p.p.m.Sodium-Dilute an appropriate aliquot of stock solution (usually 10 ml diluted to 250 ml) with water. Potassizcm-For potassium use the solution prepared for sodium to conserve stock solution. A scale expansion of 2 is usually necessary. Manganese-Aspirate the sample solution directly. EFFECT OF INTERFERENCE- Extensive tests of interferences were conducted to allow for application of the method over the usual range of variations in composition encountered in silicate analysis. Rock part-December, 19691 BASED ON THE LITHIUM METABORATE FUSION 1061 ing solutions from the fusion contain, in addition to lithium metaborate and nitrate, silicon, aluminium, iron, calcium, magnesium, sodium, potassium, titanium, manganese and phos- phorus as their usual constituents.Phosphorus and titanium were tested up to the maximum usually encountered in normal silicate rocks. The remainder of the rock constituents were tested to indicate their effect if they should constitute 100 per cent. of the original sample. The results of these tests are summarised in Table 11. TABLE I1 INTERFERENCE STUDY Element studied Iron Silicon Aluminium Potassium Sodium Titanium Magnesium Calcium Manganese Concen- tration studied, p.p.m. 10 300 60 1 1 60 1 3 0-5 No interference -it, element p.p.m. up to Ti 20 Ca, Mg, Na, K 100 Si 400 A1 250 P 1 Fe, A1 Ti 30 K, Na, Ca, Mg Fe Ti 60 Si 300 Fe, Al, Ca, Mg 10 Ti 1 Na 5 1 K Ti Fe, K, Na. Ca Ti Fe, K, Na, M g Ti 22 Mg } Si - 100 10 100 100 b 30 Inter- ference - - LiBO, Na above 5 p.p.m. LiBO, LiBO, Complex Si A1 LiBO, Si A1 LiBO, Ti LiBO, Optimum range for the determination with the proposed procedure, Correction p.p.m.- 2 to 10 I Add Li to standards Add Li to standards. Add Na to standards when needed Add Li to standards Adjust A1 con- tent to between 750 and 1000 p.p.m. and acidity to 4 N with respect to HCl 300 to 600 38 to 72 0-5 to 1.5 0-5 to 1.25 8 to 60 Add 1 per cent. of La to sample and standards. Add Li to standards Maintain SiO, 0.75 to 7 content below 150 p.p.m. Add 1 per cent. of La to sample and standards. Add Li to standards Add Li to standards 0.5 to 3 0.2 to 3.5 In addition, 066 per cent. of lithium metaborate, the concentration in the sample solution, does not interfere in the iron and silica determinations. Except for titanium, which was studied in hydrochloric acid, the interferences for the remaining elements were studied in an appropriate nitric acid medium.The acid content of samples and standards must, of course, be approximately comparable.1062 VAN LOON AND PARISSIS RE s u LTS Analyses were made on twelve analysed rock standards (Table 111). Three complete determinations were carried out for each sample and the mean calculated. The deviation represents the difference between our values and those of Ingamells and Suhr,ll Goldich, Ingamells, Suhr and Anderson12 and Roubault, de la Roche and G0~indaraju.l~ The above proposed method has thus been proved for twelve standards covering a wide range of compositions for silicon dioxide, aluminium oxide, titanium dioxide, total iron and calcium, magnesium, potassium, sodium and manganese (11) oxides.Separate samples are required for iron( 11) oxide, water and carbon dioxide determinations. One of us (J. C. Van L) thanks the Department of University Affairs and the Geological Survey of Canada for financial support of this project. Special thanks are given to Mr. C. 0. Ingamells, who kindly sent us the graphite crucible fusion modification of his original decomposition method. Standard WI SI TI GA GH BR G2 GSP-1 AGV-1 BCR-1 PCC-1 DTS-1 Standard WI SI TI GA GH BR G2 AGV-1 PCG-1 GSP-1 BCR-1 DTS-1 TABLE I11 DETERMINATION OF THE MAJOR COMPONENTS OF GEOLOGICAL STANDARDS SiO, per cent. Mean Deviation - 52.7 +0*1 59.8 0.0 63.1 0.0 69.6 -0.1 75.6 0.0 38.5 0.0 69.4 +0*1 67.2 -0.2 59.1 0-0 54.5 0.0 41.9 0.0 40.4 -0.2 A1,0,, per cent. Mean Deviation 14.9 0.0 9.07 +0*06 16.2 -0.1 14.5 -0.1 12.6 0.0 10.3 0.0 15-1 0.0 14.9 -0.1 16.9 0.0 13.4 0.0 -----77 - - - - TiO,, 7 Mean 1.07 0.47 0.55 0.37 2-59 0.46 0.63 1 -02 2.23 - - - per cent.G Z Z n -0.01 -0.01 -0.01 0.0 - 0.02 -0.01 - 0.01 - 0.02 + 0.01 - I - Total Fe, percent. Mean Deviation - 7-78 +0*04 5-80 +0.01 4-10 0.0 1.86 -0.10 0.89 -0.02 9-00 0.0 2.03 + O . l l 2.91 -0.08 4.68 -0.04 9.38 -0.04 5.76 +0.06 6-11 +0*06 MgO, per cent. K,O, per cent. Na,O, per cent. MnO, per cent. Mean Deviation Mean Deviation Mean Deviation Mean Deviation r - - - - - - - - - - - - - Y + V 6.56 4-09 1-84 0.96 - 13.3 0.74 0.96 1.43 3.40 42.9 49.8 + 0.04 + 0-03 + 0.02 - 0.01 +0*1 -0.01 - 0.02 - 0.03 -0.01 - 0.3 +0.1 - 0.64 2.62 1.20 4.12 4.75 1-26 4.56 5.43 2.85 1.73 - - +O*Ol + 0.02 - 0.03 + 0.09 - 0.03 -0.12 + 0.06 - 0.07 - 0.04 +Om15 - - 2.10 3.36 4-36 3.45 3.87 3.07 4.10 2.80 4.28 3.26 - - - 0.05 - 0.02 + 0.04 -0.12 + 0.04 0.0 + 0-05 + 0.02 + 0-07 + 0.03 - - 0-16 0.36 0.07 0.07 0.04 0.19 0.03 0-03 0.08 0.19 0.12 0.11 - 0.01 - 0.05 - 0.03 - 0.02 - 0.01 - 0.01 0.0 - 0.01 - 0.02 0.0 0.0 -0.01 CaO, per cent. & Mean 10.8 10.0 5-09 2-44 0.66 2.02 2.10 4-87 6.94 0-54 0.18 14.0 Deviation - 0.1 -0.1 - 0.03 - 0.04 - 0-02 +0n1 - 0.01 - 0-02 - 0.05 - 0.04 - 0.01 0.0 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES Van Loon, J. C., and Parissis, C., Analyt. Lett., 1968, 1, 519. Abbey, S., Geological Survey of Canada, Paper 67-37, 1967. -, Geological Survey of Canada, Paper 68-20, 1968. Belt, C. B., jun., Analyt. Chenz., 1967, 39, 676. Shapiro, L., U.S. Geological Survey, Paper 575-13, P.B. 187, 1967. Bernas, B., Annlyt. Chenz., 1968, 40, 1682. Langmyhr, F. J., and Paus, P. E., Analytica Chinz. Acta, 1968, 43, 397. Ingamells, C. O., Alzalyt. Chem., 1966, 38, 1228. Suhr, N. H., and Ingamells, C. O., Ibid., 1966, 38, 730. Ingamells, C. O., and Suhr, N. H., Geochim. Cosmochim. Acta, 1963, 27, 897. Goldich, S. S., Ingamells, C . O., Suhr, N. H., and Anderson, D. H., Can. J . Eurth Sci., 1967, 4, 747. Roubault, M., de la Roche, H., and Govindaraju, K., Sciences Terre, 1966, 11, 105. 'Received March 31st, 1969 Accepted June 4th, 1969 2 , Atomic Absorption Newsletter, 1968, 7, 103. -__

ISSN:0003-2654    

DOI:10.1039/AN9699401057

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Analyst, December, 1969, Vol. 94, p p . 1063-1067 1063 Determination of Lithium Oxide in Silic.ate Rocks by Atomic-absorption Spectrophotornetry BY MAURICE STONE AND SUSAN E. CHESHER (Rock Analysis Laboratory, Department of Geology, University of Exeter) Lithium oxide concentrations greater than 0.01 per cent. w/w in silicate rocks are determined directly on solutions prepared after dissolution of the samples in hydrofluoric, nitric and perchloric acids by atomic-absorption spectrophotometry. There are no interferences and the method can be readily incorporated into schemes of rapid silicate analysis. Precision and accuracy are good. THE granites and associated rocks of south-west England commonly contain concentrations of lithium oxide of between 0.01 and 0.5 per cent. w/w, and locally reach 0.8 per cent.As part of a geochemical investigation of some of these rocks, lithium oxide has been determined by atomic-absorption spectrophotometry. However, as there is little published work on the determination of small amounts of lithium oxide in complex materials such as silicates, it was first necessary to investigate the effects of interference by other ions. Recent results obtained by Angino and Billings1 indicate that there is little interference in natural waters from sodium, potassium, calcium, chlorine, sulphate or nitrate, and that the determination of lithium oxide in carbonate rocks is free from interference. Abbey2 describes the depression of lithium oxide absorption by free acid, but observes no interference from the amounts of sodium and potassium likely to be found in most silicates; however, he finds a depressant effect partly caused by total salt concentration and partly by viscosity. A paper by Ohrdorf3 describes the direct determination of lithium oxide in sedimentary rocks by atomic absorptiometry at concentrations lower than those found in the rocks we examined.This paper is concerned with an investigation into the determination of minor amounts of lithium oxide in silicate rocks and its incorporation into schemes of routine rapid silicate analysis. METHOD INSTRUMENTATION AND CONDITIONS- Determinations were made with a Hilger AA2 atomic-absorption spectrophotometer fitted with a 10-cm air - acetylene, water-cooled, laminar-flow burner. The following con- ditions were used: wavelength, 67043nm; slit width, 0-15mm; E.H.T.set on low; flame height on setting 8; air pressure, 2 kg cm-2; air flow-rate, main 8 litres minute-l and by-pass 3 litres minute-l; acetylene pressure, 0-3 to 0.4 kg cm-2; and acetylene flow-rate, 1 litre minu t e-1 . REAGENTS- Lithium standards-Prepare a lithium oxide stock solution, equivalent to 100 p.p.m. of lithium oxide, by dissolving 0.2473 g of Johnson Matthey lithium carbonate in a small volume of dilute nitric or hydrochloric acid and diluting with de-ionised water to 1 litre. Standards, ranging from 15 to 2 p.p.m., are prepared by appropriate dilution of the stock solution. The following analytical-reagent grade chemicals were used. Hydrojhoric acid, 48 per cent. Nitric acid, sp.gr. 1.42. Perchloric acid, 60 per cent.0 SAC and the authors.1064 STONE AND CHESHER: DETERMINATION OF LITHIUM OXIDE IN [A"ySt, VOl. 94 PROCEDURE- The dissolution of silicate rocks and the preparation of solution B according to the methods of Rilep or Shapiro and Brannocks can be followed, although it is best to avoid the use of sulphuric acid if magnesium and calcium are to be determined in these solutions by atomic- absorption spectrophotometry. The method we used is given below, although clearly much variation is possible. Weigh accurately 0-4 g of rock powder (crushed to pass a 120 mesh and dried at 110' C) into a 100-ml PTFE beaker (with its base ground flat). Add 10 ml of hydrofluoric acid and 5ml of concentrated nitric acid. Cover with a plastic lid and digest on a hot-plate until dissolution is complete. Remove the lid and evaporate to dryness, then add 4 ml of concen- trated nitric acid and evaporate to dryness.Add 4ml of perchloric acid and fume almost to dryness, then add another 2ml of perchloric acid and about 50ml of de-ionised water, cover and digest on a hot-plate until dissolution is complete (this may take several hours). Cool, transfer to a 200-ml calibrated flask and make up to the mark with de-ionised water. If a final solution in nitric acid is preferred, the perchloric acid stages can be omitted and nitric acid substituted. This gives a solution in which 20 p.p.m. of any constituent is equiva- lent to 1 per cent. of that constituent in the original rock. Prepare the instrument as indicated in the manufacturer's instructions, allowing a warming-up period of about 1 hour.Prepare a calibration graph from suitable standards, e.g., 8,6,4 and 2 p.p.m. of lithium oxide. Aspirate solution B and a blank solution directly into the flame; use de-ionised water for zero setting. Adjust scale expansion so that the highest standard used gives a reading between 80 and 90 scale divisions (linear on the Hilger AA2); this allows for possible base-line drift (marked at high scale expansions) without the need to re-set the zero setting. One of the standards should be run after groups of three or four samples to correct for any peak height drift that might occur. This graph is linear up to about 6 p.p.m. of lithium oxide. If x is p.p.m. of lithium oxide obtained from the calibration graph and d is the dilution factor (generally d = 1, unless solution B has been diluted to bring lithium oxide within the calibration range) x x d lithium oxide, per cent.w/w = ~ 20 * INTERFERENCE- Qualitative tests by the authors have shown that there is no interference from amounts of calcium, sodium, aluminium, iron, magnesium and rubidium considerably greater than those found in silicate rocks, when these elements are added to standards prepared from lithium carbonate. However, interference from potassium results in a 10 per cent. increase of the lithium oxide absorption result when 50 p.p.m. or more of potassium oxide are presen TABLE I ADDITION OF POTASSIUM OXIDE TO SHALES AND FIREBRICKS Potassium oxide Potassium Lithium i n w n B, oxide added, oxide found, Sample per cent.w/w p.p.m. p.p.m. per cent. w/v SEC 006a 6.03 50 SEC 007a 2.30 23 - 50 0.064 0.084 - 0.085 50 0.087 SEC 007d 6-20 62 - 0-070 50 0.070 B.C.S. firebrick 043 4.3 - 0.099 50 0.101 100 0.103December, 19691 SILICATE ROCKS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 1065 (as potassium sulphate) in the lithium oxide standards. This may be an ionisation effect, although similar interference expected from both sodium and rubidium was not observed. To test this effect in rock samples, known amounts of potassium oxide were added to solution B (prepared as indicated above) of three slates and B.C.S. Firebrick (No. 315). Results, given in Table I, show that the added potassium oxide has little effect on the lithium oxide readings obtained from the slates and causes only a small increase in the apparent lithium oxide content of the firebrick.TABLE I1 RECOVERY OF ADDED LITHIUM OXIDE Sample B.C.S. firebrick.. .. Tourmaline 0003 . . Tourmaline 0015 . . Tourmaline 0027 . . Andesite AGV-I* . . Basalt BCR-l* .. GraniteG2* . . .. Granodiorite GSP-I* Granite porphyry 0013 Granite porphyry 0023 Slate SEC2A . . .. Slate MGB OOlC/B . . Lithium Lithium Potassium oxide oxide hydroxide, present, added, per cent. per cent. per cent. 0-43 0.10 0.20 0.13 0-03 0.10 0-05 0.055 0.20 0-40 0-06 0.467 0.10 2-92 0.003 0-05 1.70 0-004 0-025 4.51 0.009 0.05 5.52 0.008 0.025 9.00 0.029 0.20 0.60 8-33 0.039 0-40 3.68 0.067 0.10 0.20 7.70 0.487 0.05 0.10 * New U.S.G.S. standards. Total lithium oxide recovered, per cent. 0.305 0-131 0-258 0-437 0-562 0.056 0.029 0-061 0.031 0-05 0.619 0.449 0.170 0.274 0.538 0.585 Added lithium oxide recovered, per cent.0-205 0.101 0.203 0.382 0.096 0.053 0.025 0-052 0.023 0.21 0.59 0.4 1 0.103 0.207 0-05 1 0.098 TABLE I11 COMPARISON BETWEEN ATOMIC-ABSORPTION AND FLAME-EMISSION DETERMINATIONS OF LITHIUM OXIDE Sample Tourmalines 000 1 0003 0027 Pelitic rocks OOlC WA3 021c 025a 025c 039b Granites 0035 0038 0039 0044 0045 0046 0047 0048 0056 001 3 0023 Lithium oxide, per cent. w/w f A \ Atomic absorption Flame emission 0-06, 0-06, 0.05 0.07 0.03, 0.03, 0.03 0.06 0.49, 0.48, 0.48 0.45 0-45, 0.45 0.50 0.04, 0.03 0.02 0.63 0-58, 0-64 0.10 0-49 0.49 0.22, 0-23 0.49, 0.47, 0-47 0.19, 0.20, 0.20 0.47, 0.52 0.37, 0.32 0.03, 0-03 0.38, 0.38, 0.38 0.03, 0.03 0.04 0.53 0.04, 0.03 0-04 0.62 0.62, 0.64 0.09 0.53 0.49, 0.50 0.19, 0.22 0.46, 0-53, 0-53 0.19, 0.20 0-52, 0.52 0.39, 0-42 0.05, 0.05 0-37, 0.38, 0.38 0.04, 0-03 0.06, 0-031066 [Autalyst, Vol.94 The results of spiking some silicate mineral and rock solutions (solution B) with known amounts of lithium oxide are given in Table 11. Recoveries are good and the potassium oxide present has no marked effect on the added lithium oxide, although some enhance- ment may be indicated in samples with high potassium oxide and initially low lithium oxide contents. The effects of salt concentration and possibly viscosity indicated by AbbeyZ are unlikely to be important factors here as solution B is ten times more dilute than his sample solutions. PRECISION AND ACCURACY- Precision of the direct determinations of lithium oxide by atomic-absorption spectro- photometry is indicated by replicate determinations of the B.C.S.firebrick. Sixteen deter- minations (made on different solutions determined on different occasions) give an average of 0.10 per cent. of lithium oxide, with a standard deviation of -+_O.Oll. The coefficient of variation is 5.5 per cent. A similar precision is indicated by replicate determinations of other samples shown in Table 111. Results compare favourably with determinations made by flame emission (Table 111) with a simple filter flame photometer (method outlined by Stones). Accuracy is difficult to assess because of the paucity of suitable standards in the range 0.01 to 1.0 per cent. of lithium oxide. The result obtained for the firebrick (0.10 per cent.) is higher than the average obtained from the certificate of analysis (0.085 per cent.of lithium oxide). This may reflect some enhancement by the small amount of potassium oxide present, although the recoveries shown in Table I1 point to little interference and good accuracy. Accuracy at and below the lower end of the range considered here is indicated by analyses of four of the new U.S. Geological Survey standards (Flanagan') given in Table IV. Results obtained from the more basic rocks (andesite, AGV-1 and basalt BCR-1) are close to the U.S.G.S. averages. The siliceous rocks (granodiorite GSP-1 and granite G2) give results that lie well within the large spread of the U.S.G.S. data; this spread is clearly due to marked within laboratory bias (see Flanagan,* Tables 3 and 5).STONE AND CHESHER: DETERMINATION OF LITHIUM OXIDE IN TABLE IV ANALYSIS OF U.S. GEOLOGICAL SURVEY STANDARDS Lithium oxide, p.p.m. r A \ Average Error as Standard U.S.G.S. result 1 standard deviation Results obtained G2 88 29 70, 85, 72, 72 GSP-1 77 17 80, 74, 67, 64 AGV-1 29 5 29, 28 BCR-1 36 8 39, 33 CONCLUSIONS Solutions prepared from the hydrofluoric acid digestion of silicate rocks according to the methods of Riley4 or Shapiro and Brannock: or to the modification described above, can be sprayed into the air - acetylene flame of an atomic-absorption spectrophotometer for the determination of lithium oxide. No previous treatment is necessary. Lithium oxide can be determined with the expenditure of little extra time, as part of the routine analysis of the usual major and minor elements in silicates.The results obtained in this investigation and given in the preceding Tables show that the method has good precision and accuracy in the range 0.01 to 0.6 per cent. of lithium oxide and appears to maintain good results down to 0.003 per cent. (30 p.p.m.) of lithium oxide. Scale expansion and a more concentrated solution would theoretically lower this range down to about 5 p.p.m. of lithium oxide, although the use of a more concentrated solution would probably require calibration against standards and a blank with similar salt concentrations and viscosities. For the determination of samples containing more than 0.6 per cent. of lithium oxide, it is advisable to dilute solution B. It could be argued that the interference observed when potassium oxide is added to the lithium oxide standards is suppressed in the samples by one or more of the other constituents and that if these constituents were absent from any sample, potassium oxideDecember, 19691 SILICATE ROCKS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 1067 would interfere here also.However, except in special instances that would not normally be examined in routine procedures, silicate materials containing more than 0.01 per cent. of lithium oxide are unlikely to contain major constituents outside the ranges examined here. The recoveries shown in Table I1 and the analyses of the U.S.G.S. standards shown in Table IV demonstrate that the method outlined is applicable to a wide range of silicate materials likely to contain more than trace amounts of lithium oxide. A Cook magnetic separator used for the separation of tourmaline samples was purchased with the aid of a grant from N.E.R.C. We thank Mr. R. Ellis for helping with the deter- minat ions. REFERENCES 1. 2. 3. Ohrdorf, R., Geochim. Cosmochim. Acta, 1967, 32, 191. 4. 5. 6. Stone, M., Proc. Ussher Soc., 1963, 1, 50. 7. 8. - , Ibid., 1967, 31, 289. Angino, E. E., and Billings, G. K., “Atomic Absorption Spectrophotometry in Geology,” Elsevier Abbey, S.. Geological Survey Canada, Paper 67-37, 1966. Riley, J. P., Analytica Chim. Acta, 1958, 19, 413. Shapiro. L., and Brannock, W. W., Bull. U.S. Geol. Surv., 1962, No. 1144A. Flanagan, F. J., Geochim. Cosmochim. Acta, 1969, 33, 81. Publishing Company, Amsterdam, London and New York, 1967. Received Mamh 21st. 1969 Accepted June 4th, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699401063

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

1068 Analyst, December, 1969, Vol. 94, pp. 1068-1071 Interferences in the Determination of Iron by Atomic-absorption Spectrophotornetry in an Air = Acetylene Flame BY K. E. CURTIS* (Cresco Fertilizers (".A .) Pty. Limited, P.O. Box 11, Buyswater, Western Australia 6503) Sulphate has been found to interfere in the determination of iron by atomic-absorption spectrophotometry, when using an air - acetylene flame. The effects of varying the sulphate concentration, the nature of the flame and the path of the light beam through the flame were examined. The effects of phosphate, potassium, sodium, ammonium, calcium, magnesium, manganese and aluminium ions were also examined. SLAV IN^ has commented that early workers have shown that the primary advantage of atomic absorption is its freedom from various analytical interferences, especially spectral inter- ference.However, he has also mentioned that certain chemicals in solution, usually anions, can prevent the release of a metal as an atomic vapour and thus cause an analytical error. Allan2 found that no elements of plant origin either depressed or enhanced iron absorption, even when present in the high concentrations obtained by dissolving in 20 ml of the ash from a 1-g plant sample rich in these elements. This finding was later confirmed by David.3 The present work was instigated by the observation that a standard iron solution pre- pared by dissolving iron(I1) sulphate in 0.1 N hydrochloric acid absorbed less strongly than a corresponding standard prepared in the same medium from iron wire, and that both solutions absorbed to an identical extent in the presence of excess of sulphate.EXPERIMENTAL A Techtron, Model AA4, atomic-absorption spectrophotometer was used with the follow- ing conditions: burner path length, 10 cm; lamp current, 10 mA, from an iron hollow-cathode lamp supplied by Atomic Spectral Lamps Pty. Ltd., Melbourne; wavelength, 248.3 nm; slit width, 100 pm; and air pressure, 23 lb inch-2 and flow-rate, 10.0 1 minute-1, giving a nebulising rate of 4ml minute-l. Acetylene was supplied at an external pressure of 12.5 lb inch-2 to a flow-meter control on the instrument, calibrated in arbitrary units of 0 to 10. Flow-rates of 1.2, 16, 1.8, 2.1 and 2.4 1 minute-1 (corresponding to Techtron flow settings of 2.5, 3, 3-45, 4 and 4.5, respec- tively) were used to vary the flame from highly oxidising to reducing conditions.Burner height was measured vertically (in millimetres) from the top of the burner to the centre of the light beam at its point of focus, where the diameter of the beam was 16 mm. Iron solutions used for all experiments were made 0.1 N with respect to hydrochloric acid, and all chemicals used were of analytical-reagent grade. (a) VARIATION OF SULPHATE CONCENTRATION- Forty-six solutions, one half containing 4 p.p.m. and the other 8 p.p.m. of iron were prepared from iron wire, with sulphate concentrations (as sulphuric acid) ranging from 0 to 150 p.p.m. The absorbance of each of these solutions was read, with an acetylene flow-rate of 2.1 1 minute-1, which gave a reducing flame, and a burner height of 0 mm.These conditions allowed scale expansion x5 to be used for the detection of small changes in absorbance. The results of this experiment are shown in Fig. 1. Depressions of absorbance up to 72 per cent. were observed for both the 4 and 8 p.p.m. iron solutions at relatively low sulphate concentrations. However, the analytical conditions used in this experiment were somewhat atypical, and it was necessary to investigate the interference over a range of acetylene flow-rates and burner heights. EFFECT OF SULPHATE * Present address : Department of Chemistry, University of Waterloo, Waterloo, Ontario, Canada. 0 SAC and the author.CURTIS 1069 005 o o l t O t Sulphate (as sulphuric acid), p.p.rn. Fig. 1. Effect of varying sulphate con- centrations on the absorbance of 4 and 8 p.p.m.iron solutions, measured in a fuel-rich (semi-luminous) flame: A, 8 p.p.m. iron solution, no sulphate; B, 4 p.p.m. iron solution, no sulphate; C, 8 p.p.m. iron solution, varying sulphate; and D, 4 p.p.m. iron solution, varying sulphate (b) VARIATION OF FLAME AND BURNER CONDITIONS- The absorbances of two 10 p.p.m. iron solutions prepared from iron(II1) chloride, one free from sulphate and the other containing 100 p.p.m. of sulphur (as sulphuric acid), were recorded over a wide range of analytical conditions. The experiment was repeated four times, and in each instance the results agreed to within 0.5 per cent. transmittance, giving a coefficient of variation of about 1 per cent. on the absorbance scale. The optimum burner height is that which gave the maximum absorbance for the sample containing sulphate.Crucial results are given in Table I. TABLE I EFFECT OF 100 P.P.M. OF SULPHUR (AS SULPHURIC ACID) ON THE ABSORBANCE OF A 10 P.P.M. IRON SOLUTION UNDER VARIOUS FLAME CONDITIONS, AT THE OPTIMUM BURNER HEIGHT FOR EACH FLOW-RATE Acetylene flow-rate, 1 minute-' . . . . 1.2 1.5 1.8 2.1 2-4 Absorbance without subhate . . .. . . 0.347 0.417 0.444 0.408 0.323 Optimum burner height, mm . . * . .. 4 5 7 7 7 -4-4 4-4 +3 - 49 - 77 Change in absorbance 5roduced by sulphate Absorbance without sulphate (expressed as percenGge) This experiment confirmed that the interference prevailed under all of the conditions tested, but was most serious when the flame became luminous (it?,, with acetylene flow-rates of 2.1 and 2.4 1 minute-l). Under normal oxidising conditions, with the light beam passing above the unburnt cone of gases, the interference was quite small, but definitely observable. (c) OTHER OBSERVATIONS- Similar effects were obtained by using the Techtron, Model AA3, atomic-absorption spectrophotometer under similar conditions.The depression of readings was considerably reduced when nebulising rate was decreased. This led to an investigation of the possible effect that 100 p.p.m. of sulphur may have had on the nebulising rate in experiment (b). No detectable change was observed, compared with that of the sulphate-free sample.1070 CURTIS: INTERFERENCES IN THE DETERMINATION OF IRON BY ATOMIC- [Arta&St, VOl. 94 D. J. David (in a personal communication) has confirmed that this effect occurs under reducing conditions with his instrument,* but has found no interference from sulphate under oxidising flame conditions.It is reasonable to assume that the sulphate interference found will be exhibited by other atomic-absorption spectrophotorneters. SINGLE ION EFFECT- Experiments similar to that with sulphate (b) above were carried out with phosphorus (as phosphoric acid), nitrogen (as ammonium chloride), and potassium, sodium, calcium, magnesium, manganese and aluminium (as chlorides). In each instance a solution containing 10 p.p.m. of iron[as iron(II1) chloride] and 100 p.p.m. of the element was compared with a solution containing 10 p.p.m. of iron alone. Again, the ions were observed to have no detectable effect on the nebulising rate.As all of the solutions used were made 0.1 N with respect to hydrochloric acid, correspond- ing to about 3500 p.p.m. of chloride ion, the effect of the additional chloride ion concentration associated with the above cations was considered to be negligible. This was confirmed by the observation that the absorption of the 10 p.p.m. iron solution prepared from iron wire was identical with that of the corresponding standard prepared from iron(II1) chloride under all of the conditions tested. The significant effects for each individual ion are given in Table 11. It was noted that the effect of phosphate was similar to that of the sulphate. Sodium and potassium produced a progressively lesser depression of reading and ammonium an enhance- ment.The effects of calcium, magnesium, manganese and aluminium were nearly identical with those of the ammonium ion. OTHER INTERFERENCES TABLE I1 INDIVIDUAL EFFECT OF OTHER SINGLE IONS ON THE ABSORBANCE OF A 10 P.P.M. IRON SOLUTION UNDER VARIOUS FLAME CONDITIONS, AT THE OPTIMUM BURNER HEIGHT FOR EACH FLOW-RATE Acetylene flow-rate, 1 minute-l . . .. . . 1.2 1-6 1-8 2.1 2.4 Absorbance of 10 p.p.m. of iron alone . . . . 0.347 0-417 0.444 0.408 0.323 Optimum burner height, mm . . .. .. 4 5 7 7 7 +6 +3 +1 - 44 - 67 Change in absorbance produced by phosphate Absorbance without phosphate (expressed as percentage) Change in absorbance produced by sodium Absorbance without sodium (expressed as percentage) +6 +6 +4 -21 - 39 $6 +5 +2 -8 - 10 Change in absorbance produced by potassium Absorbance without potassium (expressed as percentage) f 6 +4 +4 + 14 + 37 Change in absorbance produced by ammonium ion Absorbance without ammonium ion (expressed as percentage) * *Similar effects were observed with calcium, magnesium, manganese and aluminium ions.COMBINED IONS EFFECT- An experiment similar to those to determine the effects of the single ions was carried out with two 10 p.p.m. iron solutions prepared from iron(II1) chloride, one containing iron alone and the other 100 p.p.m. each of sulphur, phosphorus, potassium, sodium, nitrogen, calcium, magnesium, manganese and aluminium, added in the same chemical forms as before. The presence of these ions had no observable effect on the nebulising rate. These results were virtually identical with the effect of the ammonium ion alone.An enhancement was observed under all of the conditions tested, again being minimal under oxidising flame conditions and becoming greater as the reducing nature of the flame increased. DISCUSSION This work has shown that several common ions interfere in the determination of iron by atomic-absorption spectrophotometry, over a considerable range of flame types and light-beam paths through the flame. However, under normal oxidising conditions (fuel-lean flame) ,December, 19691 ABSORPTION SPECTROPHOTOMETRY IN AN AIR - ACETYLENE FLAME 1071 with the light beam passing above the unburnt cone of gases, these interferences are kept to a minimum and, although still observable, they may possibly be considered negligible in work that does not require the highest precision.Seriously misleading results can be obtained under reducing conditions, regardless of the section of the flame used. The similarity of the combined ion effect to that of the ammonium ion alone suggests that the addition of an ammonium “spike” to both standards and samples will produce a consistent enhancement of readings. This should not only reduce further the effects of other ions when using the normal analytical conditions mentioned above, but also enable determinations under other conditions to be performed successfully. There is the possibility that in the determination of other elements this type of interference may also be exhibited over the range of conditions available to the analyst. D. J. David (personal communication) has found an interference effect of molybdenum on manganese absorption, again under reducing conditions close to the base of the flame. The complete experimental results from which those given in Tables I and I1 have been selected will be available on request. I thank Mr. T. C. Shaw for his co-operation and assistance in the investigations with the Techtron, Model AA3, instrument, and for his interest and encouragement during the remainder of the work. I also thank Dr. G. F. Atkinson for his helpful advice, and the Directors of Cresco Fertilizers (W.A.) Pty. Ltd. for permission to publish this paper. REFERENCES 1. Slavin, W., Perkin-Elmer Atomic Absorption Newsletter, 1964, No. 24, 15. 2. Allan, J. E., Spectrochim. Ada, 1959, 15, 800. 3. David, D. J., Perkin-Elmer Atomic Absovption Newsletter, 1962, No. 9, 1. 4. - , Analyst, 1961, 86, 730. Received June 14th, 1968 Accepted May 22nd, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699401068

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

1072 ArcaZyst, December, 1969, Vol. 94, f@. 1072-1074 The Determination of Zinc in Crystalline Insulin and in Certain Insulin Preparations by Atomic-absorption Spectroscopy BY GERALD I. SPIELHOLTZ AND GLORIA C. TORALBALLA (Department of Chemistry, Herbert H . Lehman College of the City University of New Yovk, Bronx, New York 10468, U.S.A.) Atomic-absorption spectroscopy is an excellent technique for the deter- mination of the purity and activity of crystalline and protamine zinc insulins and has been found suitable for the determination of down to 6pg of zinc. The procedure is rapid and direct as preparation involves only weighing and dissolution of the sample. The method has been applied to the determination of zinc in crystalline and protamine insulin samples. The values obtained for zinc compare favourably with those obtained by the dithizone colori- metric method.CRYSTALLINE zinc insulin as well as various zinc insulin preparations are used extensively in therapeutic medicine. Crystalline compounds of insulin containing zinc or other heavy metals have been of value in the X-ray structure studies of insu1in.l An important criterion for identifying and assaying zinc insulin is by its zinc content. This zinc content is especially critical in formulating pharmaceutical and medical standards for the drug industry. The method specified by the United States Pharmacopoeia2 for the determination of zinc in insulin and in insulin preparations is a spectrophotometric one involv- ing the zinc complex of dithizone.2 The Pharmacopoeias of several countries also specify the dithizone method with slight modifications.3p4 As investigators have found this method long, laborious and requiring skilful handling,4~5~6~7 other methods for the determination of zinc in insulin have been developed.Among these methods are polarography,* p9 titration with EDTAJ1O photometric determination of zinc with zincon,sJ1 a colorimetric method in which a complex of zinc with crystal violet is usedJ6 coulometric titration of zinc 8-hydroxyquino- hate7 and a titrimetric method involving the use of the zinc - dithizone complex and standard zinc solutions.l2 All of these methods involve a series of operations consisting of oxidation or precipitation of the protein, complexation, extraction and a final measurement of some physical property. These operations make the method time consuming and subject to relatively greater errors.There exists a definite need for a method that is simpler, fast and accurate. Atomic- absorption spectroscopy meets these requirements. This paper reports the determination of zinc in samples of crystalline insulin and insulin preparations by atomic-absorption spectroscopy. The procedure involves only the dissolution of a weighed sample followed by the spectroscopic measurement. No preliminary treatment is necessary. The method is rapid, direct and sensitive. The accuracy of the method was checked by comparison with one of the standard procedures. MATERIALS AND METHODS INSTRUMENTATION- A Perkin-Elmer Model 303 atomic-absorption spectrophotometer, with the standard pre-mix burner head and platinum - titanium nebuliser, was used.Acetylene was used as the fuel and air as oxidant. The acetylene pressure was 9 p s i . and flow-meter setting 5-5 (2-0 litres minute-1); the air pressure was 30 p.s.i. and flow-meter setting 7.5 (20.0 litres minute-1). The 213-8nm resonance line and slit position 5 giving a slit width of 3mm (2.0 nm spectral slit width) were used for absorbance measurements. Settings with units unstated refer to manufacturers' arbitrary scales on the Perkin-Elmer Model 303 instrument. The visible flame resulting from these acetylene and air flow-rates was about 2.5 inches above, and included a +inch blue reducing zone. 0 SAC and the authors.SPIELHOLTZ AND TORALBALLA 1073 The flow-meter settings, flame height and rate of solution uptake were each adjusted to give maximum absorbance when a solution containing zinc was aspirated.The slit position was as suggested for zinc determination by the manufacturer. A Beckman DU spectrophotometer was used for the absorbance measurements of the zinc - dithizone complex. Matched 1-cm cells were used and measurements made at a wave- length of 535 nm. Measurements of pH were made with a Beckman Expandomatic pH meter and a Beckman No. 39183 probe combination electrode. Weighings were made on a Mettler Type B6 Gram-Atic Balance. REAGENTS- Standard zinc solutions for atomic absorfition-Weigh accurately about 2.5 g of analytical- reagent grade zinc metal, dissolve it in a minimum amount of dilute 6 M hydrochloric acid and transfer the solution to a 1-litre graduated flask, dilute to the mark with water and mix thoroughly.Transfer, by pipette, 4-00ml of this solution into a 500-ml graduated flask, dilute to the mark and mix. The zinc content of this solution is about 0-02 mg ml-l. From this diluted solution prepare a series of standards in the range of 0-2 to 3.0 p.p.m. Thew standards and the 0.02 mg ml-l solution should be prepared every 2 days. No matrix interference was anticipated because the concentration of solids in the solutions tested was 0.001 per cent. or less. This was confirmed by adding 0.001 per cent. of albumin to a set of zinc standards. Identical values for the absorbances were obtained by using standards with or without albumin. 8 Crystalline insulin samples were obtained from Bios Laboratories, New York; Sigma Chemical Company, St.Louis, Missouri; and Mann Research Laboratories, New York. Zinc protamine insulin solutions were obtained from Eli Lilly Company, Indianapolis, Indiana, and Mann Research Laboratories. METHODS PREPARATION OF INSULIN SAMPLES- Weigh accurately about 5 mg of each crystalline sample, then suspend in 10 ml of water, acidify with 1 drop of dilute hydrochloric acid to dissolve the sample and transfer to a 50-ml graduated flask, dilute to the mark with water and mix well. Transfer a 1-ml aliquot of protamine insulin solution to a 50-ml graduated flask. Dilute to the mark with water and mix well. If the preparation is a suspension, acidify with 1 drop of dilute hydrochloric acid to dissolve, dilute to the mark with water and mix well.The protamine insulin sample of Eli Lilly Company required this latter treatment. The absorb- ances of these solutions were measured. ANALYSIS BY ABSORPTION SPECTROPHOTOMETRY- To check the accuracy of the above method several samples were analysed by the colorimetric dithizone pr0~edure.l~ ,l4 This involves the extraction of zinc from an acetate- buffered solution at pH 4.5 to 5-0 with dithizone in carbon tetrachloride. Methyl red indicator is added to the aqueous solution to indicate the required pH, then removed by extraction into carbon tetrachloride. When insulin solutions were extracted with carbon tetrachloride to remove the indicator, emulsification took place and a sharp line of separation slowly appeared. This delay was avoided by using a pH meter to adjust the pH.Preliminary results from dialysis studies in this laboratory indicated that zinc was readily and rapidly dissociated from the insulin molecule, therefore ashing of the sample before zinc determination by the dithizone spectrophotometric method was unnecessary. The zinc was extracted directly from the insulin solution with dithizone in carbon tetrachloride. RESULTS AND DISCUSSION The results are given in Table I. Both methods give values that are in good agreement with each other and with those reported, except for the Mann protamine solution. We suggest that there is an error in the reported value. The precision obtained, considering the relatively small samples taken, is good. The results obtained by atomic absorption are reliable and the short time required for the analysis is most striking.The preparation consists only in weighing and in dissolving the crystalline sample. The insulin preparation, after appropriate dilution, is measured directly.1074 SPIELHOLTZ AND TORALBALLA TABLE I ZINC CONTENT OF VARIOUS INSULIN SAMPLES Dithizone* Sample Atomic absorption* colorimetric method Reported Mann-bovine .. .. . . 0.47 f 0.02 per cent. 0-44 f 0.03 per cent. 0-47 per cent. Mann-porcine .. . . 0.63 f 0.01 per cent. 0.53 & 0-01 per cent. 0-64 per cent. Sigma . . .. .. . . 0.60 f 0.00 per cent. 0-49 f 0.02 per cent. 0-60 per cent. Bios-bovine . . .. . . 0-50 f 0.01 per cent. - No report Lilly protamine solution . . 0-081 f 0-002 mg ml-l - 0.076 mg ml-l Mann protamine solution .. 0.055 f 0.001 mg ml-1 0.069 f 0.001 mg ml-l 0-034 mg ml-1 * Number of trials, three each. Existing spectrophotometric or gravimetric methods for zinc involve many prior separa- tions or extractions, with the accumulation of extraneous zinc present as impurities in the numerous reagents used. In the gravimetric procedure doubt exists as to the exact stoicheio- metric composition of the zinc precipitate in its final weighing form. We thank Mr. Ralph Steinberg and Mr. Harry Miller for technical assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. REFERENCES Harding, M., J. Molec. Biol., 1966, 16, 212. “Pharmacopoeia of the United States, ’’ 16th Revision, United States Pharmacopoeial Convention “The Pharmacopoeia of India,” Government of India Press, Nasik, India, 1966. Libicky, A., Chem. CesR. Farm., 1967, 16, 168. Zak, B., and Cohen, J. S., J. Pharm. Sci., 1963, 52, 912. Shafram, I. G., Zelenova, T. K., and Kazhdan, V. I., Sb. Stat. Vses. Nauchno-lssled. Inst. Khim. Molle, L., and Patriatche, G., J. Pharm. Belg., 1960, 15, 26. Pzibil, R., and Roubal, Z., Colln. Czech. Chem. Commun., Engl. Edn, 1953, 18, 366. Kalinowska, 2. E., and Kowalczyk, J., Farmaceuta Pol., 1963, 19, 493. Aumonier, P., Dulat, C., and Quilichini, R., Bull. Soc. Pharm., Bordeaux, 1963, 102, 279. Amer, M. M., Egyet. Pharm. Bull., 1968, 40, 143. Butts, B. G., Gahler, H., and Mellon, M. G., Metal Finish., 1951, 49, No. 4,60. Kanzelmeyer, J. H., in Kolthoff, I. M., and Elving, P. J., Editors, “Treatise on Analytical Chemis- Received May 9th, 1969 Accepted June 12th. 1969 Inc., Mack Publishing Co., Easton, Pa., 1960. Reactivov and Osobu Christykh Kim Veschesty, 1961, 24, 226. , , Chemickk Listy, 1962, 46, 492. -- try,” Part 11, Volume 3, Interscience Publishers Inc., New York, 1961, p. 157.

ISSN:0003-2654    

DOI:10.1039/AN9699401072

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Analyst, December, 1969, Vol. 94, $9. 1075-1080 1076 The Identification and Determination of Organophosphorus and Carbarnate Insecticides by Thin-layer Chromatography BY M. RAMASAMY (World Health Organisation, Ano+hles Control Research Unit, P.O. Box 603, Kaduna, Nigeria) A simple , rapid and sensitive thin-layer chromatographic procedure has been developed for the identification and simultaneous determination of organophosphorus and carbamate insecticides. The RF values and suitable combinations of adsorbent, developing solvent and chromogenic spray are described in detail for eleven organophosphorus and seven carbamate insecti- cides. The procedures are suitable for determining these insecticides on sprayed surfaces and for the routine screening of food and agricultural products for tolerance limits.The convenient screening range is 0.1 to 20pg on the chromatoplates. THE World Health Organisation Programme for the Evaluation and Testing of New Insecti- cides includes field tests on new materials at the Anopheles Control Research Unit, Kaduna, Nigeria. As part of the over-all assessment of these insecticides, a study of some of the aspects of their fate on treated surfaces was made. The rate and mode of degradation was conveniently determined by use of thin-layer chromatography. This method was applied to eleven new organophosphorus compounds, seven carbamates, some of the parent phenols and other degradation products. The trade and chemical names of these compounds, including code numbers of the manufacturers and the World Health Organisation (W.H.O.), are given in Tables I and 11.These compounds have been used for the control of vectors in public health programmes and for the protection of agricultural crops from pests. TABLE I ORGANOPHOSPHORUS COMPOUNDS Common name Dicapthon .. .. Bromophos . . .. Fenitrothion . . Malathion . . .. Fenthion .. .. Cidial . . .. .. .. - Iodofenphos . . .. .. - .. .. - Dichlorvos .. .. Chemical name 0- (2-chloro-4-nitrophenyl) 00-dimethyl phosphorothioate 0-(4-bromo-2,6-dichlorophenyl) 00-dimethyl phosphorothioate OO-dimethyl 0- (3-methyl-4-nitrophenyl) S-[ 1,2-di(ethoxycarbonyl)ethyl] 00-dimethyl phosphorodithioate 00-dimethyl 0-(3-methyl-4-methylthio- S-a-ethox ycarbon ylbenz yl 00-dimethyl phosphorodithioate O-3-Chloro-4-diethylsulphamoylphenyl 00-dimethyl phosphorothioate 0-( 2,5-dichloro-4-iodophenyl) 00-dimethyl phosphorothioate O-a-Cyanobenzylideneamino 00-diethyl phosphorothioate 0-2-Chloro-a-c yanobenz ylideneamino OO-dieth yl phosphor0 thioa te 2,2-Dichlorovinyl dimethyl phosphate phosphorothioate phenyl) phosphorothioate Manufacturer and code No.American Cyanamid AG 4124 Cela shg-1942 Bayer 41831 American Cyanamid AG-4049 Bayer 29493 Montecatini Edison Sumitomo Chemical Ciba C-9491 L-561 $1349 Bayer 78182 Bayer 77488 Shell, Ciba W.H.O. code No. OMS 214 OMS 668 OMS 43 OMS 1 OMS 2 OMS 1075 OMS 868 OMS 1211 OMS 1197 OMS 1170 OMS 14 0 SAC and the author.1076 RAMASAMY : IDENTIFICATION AND DETERMINATION OF [Artalyst, Vol. 94 TABLE I1 CARBAMATES Manufacturer W.H.O. Common name Chemical name and code No.code No. Mobam .. . . PBenzothienyl methylcarbamate Mobil Chem. MCA-600 OMS 708 Propoxur .. . . 2-Isopropoxylphenyl methylcarbarnate Bayer 39007 OMS 33 Landrin .. . . 3,4,5-Trirnethylphenyl methylcarbamate Shell SD-8530 OMS 597 Carbamult . . . . 3-Methyl-5-isopropylphenyl methylcar- Schering 3461 5 OMS 716 - 2,3,5-Trimethylphenyl methylcarbamate Shell SD-8786 bamate - 3-Isopropylphenyl methylcarbarnate Hercules AC-5727 OMS 15 Carbaryl . . . . 1-Naphthyl methylcarbamate Union Carbide OMS 29 - 2-Cyclopentylphenyl methylcarbarnate Bayer 38799 OMS 1028 Braithwaitel used GibbJs2 reagent for locating organothiophosphates on paper chromato- grams and Miskus, Eldefrawi, Menzel and Svoboda3 used @-nitrobenzenediazonium fluoro- borate solution for the colorimetric determination of some carbamates. Bates4 described the use of 4-methylumbelliferone on paper chromatograms for the detection and identification of organophosphorus residues in foodstuffs. This paper describes a simple general procedure involving the use of these three chromogenic reagents with thin-layer chromatography for the detection, identification and determination of both organophosphorus and carbamate insecticides and their degradation products simultaneously.METHOD APPARATUS- ultraviolet lamp were used. Desaga model thin-layer chromatographic equipment, as designed by Stahl,5 and an A DSORBENTS- Silica gel G and aluminium oxide G were used its adsorbents and were manufactured by Merck according to Stahl's formula, but any equivalent make can be used. DEVELOPING SOLVENTS- 1.Hexane - acetone (3 + 1 v/v). 2. Chloroform - acetone (9 + 1 v/v). 3. Hexane - acetone (5 + 1 v/v). CHROMOGENIC SPRAY REAGENTS- Reagent A-A 0.5 per cent. w/v solution of 2,6-dibromo-fi-benzoquinone4-chlorirnine in cyclohexane. Reagent B-First spray : N ethanolic potassium hydroxide solution ; second spray : a solution prepared by dissolving 25 mg of fi-nitrobenzenediazonium fluoroborate in 10 ml of diethylene glycol and making up to 100ml with ethanol. Reagent C-The chromatoplate is first exposed to bromine fumes for 30 seconds. It is then sprayed with a reagent prepared by the addition of 100 ml of water and 10 ml of 0.1 N ammonia solution to 0.15 g of 4-methylumbelliferone in 100 ml of ethanol. STANDARD SOLUTIONS- organophosphorus compounds, carbamates and their respective phenols. PREPARATION OF CHROMATOPLATES- A 0.25-mm layer of silica gel G or aluminium oxide G is prepared on 20 x 20-cm glass plates: 30 g of silica gel mixed well with 60 ml of water or 25 g of aluminium oxide with 50 ml of water are sufficient for five plates.The preparation of the slurry takes 45 seconds. The prepared slurry is immediately poured into the applicator, which is then drawn over the plates uniformly. The plates are left in position at room temperature for 15 minutes and activated in the oven by drying at 120" C for 30 minutes. They are then stored in a desiccator until ready for use. Standard solutions were prepared in chloroform containing 1 mg ml-l of the various PROCEDURESDecember, 19691 ORGANOPHOSPHORUS AND CARBAMATE INSECTICIDES 1077 PREPARATION OF THE SAMPLE- The insecticides were removed from deposits from the surface of treated mud walls, mud blocks, wooden surfaces and raffia palm mid-ribs, with the aid of cotton-wool swabs moistened with acetone or acetone - chloroform mixture (1 + 1).These swabs are then ex- tracted quantitatively in a Soxhlet apparatus with acetone or methanol. Samples of mud taken from known areas and suitable depths, and thatch from roofs, are also extracted in a similar manner. The solvent is evaporated and the residue dissolved in 5 or 1Oml of chloroform. These extracts are normally satisfactory for thin-layer chromatography as the thin layer of adsorbent cleans up and isolates the insecticide during the development process. An occasional dark-coloured extract is obtained with soot deposits. These are further cleaned up by a simple solvent partition or by column chromatography with alumina, Florisil or magnesium oxide - Celite, depending on the insecticide; details of such clean-up are available in publications by Bates: Anglin and Mckinlep and Mills.' CONTROL SAMPLES- Circular disc-shaped mud blocks of standard size, 8 cm in diameter and 8 mm thick and weighing about 60 g each, were made with the aid of plastic Petri dishes of appropriate size. The area of the sprayable surface was 50 cm2 per block and the volume of clay was suitable for extraction in a Soxhlet apparatus.Each block was sprayed with 2ml of a suspension in water containing 0.5 per cent. w/v of the active ingredient. The suspension was prepared from water-dispersible powder formulations. The micro-sprayer used was constructed in the laboratory.The treated mud blocks were extracted according to the procedure given above, and quantitative determinations conducted immediately after the spraying operations showed recoveries of over 90 per cent. with propoxur, Mobam and dicapthon. DEVELOPMENT AND VIEWING OF THE CHROMATOPLATES- The prepared sample (1 to 10 pl) containing about 1 mg ml-l of the insecticide is spotted on the plate. This amount is sufficient for the detection of major breakdown products, but for the detection of minor breakdown constituents, samples of extract containing more than 1 mg ml-1 of the insecticide may be necessary. Standard amounts of 1 to 10 pg of the pure insecticide and its parent phenol are spotted on the same plate for matching and comparison of the areas and intensities for quantitative determination. As soon as the solvent has evaporated, the plate is inserted into the developing chamber containing developing solvent to a depth of 1 cm.The selection of the appropriate solvent from the three systems given above is made from the R, values given in Tables I11 and IV to suit the sample under examination. The chamber is kept saturated with the developing solvent with the aid of wide strips of filter-paper attached to the walls of the chamber and dipping into the solvent. The development time is about 30 minutes. The solvent-front level is then marked and the plate removed from the chamber and allowed to dry. The plate is sprayed for viewing with the appropriate chromogenic reagent given above. If spray reagent A is used, the plate is examined for any visible spots immediately after spraying.It is then heated at 120" C in an oven for 10 minutes and examined again for spots that had developed during the heating process. With spray reagent B, the plate is sprayed with the first reagent, and allowed to stand for 2 minutes to hydrolyse the insecticide. It is then sprayed with the second reagent when the spots appear immediately. For spray reagent C, the plate is first exposed to bromine fumes in a closed chamber for 30 seconds. It is then removed from the chamber and allowed to stand for 2 minutes to dissipate the excess of bromine vapour. It is then sprayed with the reagent and the plate examined under ultraviolet light.The solvent front is allowed to travel about 12cm. RESULTS The results obtained with the organophosphorus compounds and some of their phenols are given in Table I11 while those of the carbamates, including some phenols, are given in Table IV.1078 RAMASAMY : IDENTIFICATION AND DETERMINATION OF [Ana&st, VOl. 94 TABLE I11 Rp X 100 AND COLOUR OF THIN-LAYER CHROMATOGRAPHIC SPOTS WITH ORGANOPHOSPHORUS COMPOUNDS AND SOME PARENT PHENOLS RF x 100 in solvent systems Sprays vv r A 3 Insecticide 1 2 3 A B C Dicapthon a 60 85 27 Light Br N.R. G b 62 90 55 Light G - Y N.R. Br Phenol from a 18 27 5 Bright Y Y G Bromophos a 55 83 47 G - Br N.R. G b - - 85 G-Br N.R. W Phenol (from a 34 67 22 Bright sky B1 Pu - Br G bromophos) b 28 21 10 Bright sky B1 Pu - Br Br Fenitrothion a 65 77 41 Brappears N.R.G b 73 78 70 Br appears N.R. w Phenol (from a 28 74 11 G turns B1 on Pu G dicapthon b 10 9 8 Bright Y Y Dark Br, but Y to naked eye without heating without heating f enitrothion) standing b 51 51 31 G turns Gr and Pu Br then B1 Malathion a 42 84 26 Light Br N.R. G b 69 88 58 Light Br N.R. W Fenthion a (i) 31 74 (i) 18 (i) Light Gr - G W G b 50 89 79 Light Gr - G Light Pu - Br W Cidial a 52 85 37 Light G W G b 73 96 70 Light G N.R. W OMS 868 a 26 79 16 G W G b 54 80 44 G N.R. W Iodofenphos a (i) 46 (i) 63 56 (i) Light B1 (i) Pi - Br G b - 85 - N.R. Pi - Br G OMS 1170 a 32, 55 88 44 Light Br W G b 74 81 76 G N.R. Br OMS 1197 a 25,48 86 34 Light Br W G b 19, 65 79 68 G N.R. Br Dichlorvos a 30 72 13 W Gr - B1 G b 13, 57 - 83 N.R.Gr - B1 N.R. (ii) 55 (ii) 42 (ii) Br (ii) 68 (ii) 85 (ii) Light Br (ii) Pi - Br a = Silica gel G plate Colour code: B1 = Blue Br = Brown G = Grey Gr = Green b = Aluminium oxide plate Pu = Purple Pi = Pink W = White Y = Yellow N.R. = No reaction All spots with spray reagent C appear as fluorescent or quenched areas, or both. ADSORBENTS- In general, the silica gel G gave greater sensitivity and provided more satisfactory results than aluminium oxide G. Hence for general screening of pesticides, the silica gel plate is preferred to the aluminium oxide plate. However, the aluminium oxide plate proved to be superior in some instances. With the solvent system (1) or (2) the latter gave better resolution of the decomposition products of Mobam and, in addition, the colour of the spot was more distinctive and more suitable for quantitative determination.Further, 4-hydroxy- benzothiophene has a higher R, value than Mobam with the silica gel plate, while the order is reversed with the aluminium oxide plate. The spots are rounded and compact with the silica-gel plate, while some tailing occurs with the aluminium oxide plate. The alkalinity of the aluminium oxide plate probably caused partial decomposition of Mobam during the development. The aluminium oxide plate gives good results with 4-hydroxybenzothiophene, 2-isopro- poxyphenol, 3-isopropylphenol, 3-methyl-5-isopropylphenol and carbaryl. The spots appearDecember, 19691 ORGANOPHOSPHORUS AND CARBAMATE INSECTICIDES 1079 TABLE IV RF X 1oO AND COLOUR OF THIN-LAYER CHROMATOGRAPHIC SPOTS WITH CARBAMATES AND SOME PARENT PHENOLS R F x 100 in solvent systems P Insecticide 1 Mobam a 26 b 34 Phenol (from a 36 b 24 Mobam) Propoxur a 26 b 38 Phenol (from u 49 propoxur b 59 Landrin a 26 b (i) 63 Phenol (from a 34 landrin) b (i) 68 2,3,6-Isomer a 25 of landrin b (i) 58 Carbamult a 27 b 83 (ii) 67 (c) (ii) 68 (c) (ii) 68 (c) Phenol (from a 37 carbamult) b 67 OMS 15 a 25 b 59 Phenol (from a 36 OMS 15) b 64 Carbaryl u (i) 18 (ii) 34 b 44 OMS 1028 a 32 b 65 2 52 81 49 66 52 76 69 74 58 82 52 60 55 76 59 76 55 60 64 75 75 57 67 76 71 76 3 12 21 21 23 14 32 46 52 17 36 24 (i) 30 (ii) 13 17 1 0 30 A B1- G Light B1 Dark G appears without heating B1 on heating G-Br Light G G-Br on heating Light Br Light Y - Br Light Br G-Br Pu - B1 turns dark Pu - B1 turns B1 G Light G (iij 40 (c) 19 G (i) 36 Light G (ii) 51 (c) 28 G 35 19 G - B1 Sky B1 turns G - B1 on heating 35 Light G 28 Dark G - B1 32 10 Intense G - Br Sky B1 turns G - B1 on heating 24 B1- G turns G - Br on heating 20 B1- G (i) 33 Light G (3) 45 (c) B Pu Pu Pu Pu Pu Pu Pu Pu Light Y - Br Light Y - Br Light Y - Br Light Y - Br N.R.Pu - Br Pu - Br Light Br Pu - Br Light Br Pu - Br Orange - Br Pu - Br Orange - Br G Sky blue to B1- G Pu Pu C G W Dark G without ultraviolet Br G Br G N.R. G Light Br G Light Br G W G W G Light Br G Light Br G Light Br G W N.R. W a = Silica gel G plate Colour code: B1 = Blue Br = Brown G = Grey Gr = Green b = Aluminium oxide G plate Pu = Purple W =White Y = Yellow N.R. = No reaction (c) denotes RF value with spray reagent C alone.All spots with spray reagent C appear as fluorescent or quenched areas, or both. immediately after spraying and l-pg amounts are easily detected and determined. The intensities of the spots remain undiminished for at least 2 days and the colour of the spots lasts for several days. The aluminium oxide plate is less sensitive for many carbamates. Mobam gives a satisfactory spot with about 2 pg, while propoxur needs about 5 pg and all other carbarnates more than 10 pg for identification and determination. 2,6-DIBROMO-~-BENZOQUINONE-~-CHLORIMINE (SPRAY REAGENT A)- Less than 5-pg amounts of dicapthon, bromophos, malathion, fenthion, OMS 1170 and OMS 1197 are conveniently determined on the silica gel plate with this reagent but amounts over 10 pg are needed for dicapthon and fenitrothion on the aluminium oxide plate.Further, the aluminium oxide plate needs more than 15-pg amounts of malathion, fenthion, bromophos, OMS 1170 and OMS 1197 for determination. No spots are observed with iodofenphos and dichlorvos.1080 RAM AS AMY The parent phenol of bromophos, 4-bromo-2,5-dichlorophenol, is readily identified by its characteristic sky blue spot. The R, value is low and about 5 pg are needed for detennina- tion. Less than 0.1 pg of 4-hydroxybenzothiophene is detected and determined with this reagent while 2 pg are needed for Mobam. P-NITROBENZENEDIAZONIUM FLUOROBORATE (SPRAY REAGENT B)- P-Nitrobenzenediazonium fluoroborate has proved the most satisfactory spray reagent for all carbamates except landrin.In practice, landrin leaves a faint yellow - brown spot, probably because of the presence of the isomer as impurity in the commercial product. This spray reagent is more sensitive for phenols than the carbamates; less than l-pg amounts of phenols and 1 to 3 pg of the carbamates are needed for quantitative determination, and the spots are stable for more than 1 week. It is equally sensitive for phenols of the organophosphorus compounds but unsuitable for most organophosphorus compounds them- selves as they show no response to this reagent. However, some of these compounds, such as fenthion, Cidial, OMS 868, OMS 1170 and OMS 1197, leave faint white spots. Dichlorvos is unique in that it produces a characteristic greenish blue spot. Carbaryl appears as a grey spot on the silica gel plate and a characteristic sky blue spot on the aluminium oxide plate.The sky blue colour changes to bluish grey in a few hours and fades in about 4 days. Most of the phenols appear as mauve and purple spots but a few are of various shades of brown, as shown in Tables I11 and IV. Dicapthon appears as a bright yellow spot caused by alkaline hydrolysis to the ionic form of 2-chloro-4-nitrophenol. 4-METHYLUMBELLIFERONE (SPRAY REAGENT C)- This reagent is particularly useful for the detection, identification and determination of intermediate products of decomposition. Some of these products, which do not show with spray reagents A and B, are often detected with this reagent. It is far more sensitive than the other two reagents and less than l-pg amounts produce well defined spots, which are fluorescent or show as quenched areas under the ultraviolet lamp, or both.These vary from shiny white to grey and dark brown in colour, as indicated in Tables I11 and IV. DISCUSSION The combination of two adsorbents , three solvent systems and three chromogenic sprays gives a choice of eighteen different procedures. A suitable combination, based on the aims of the assay, can be selected for any particular insecticide by scanning the results given in Tables I11 and IV. The methods described above were found satisfactory for following the fate of an insecticide on sprayed surfaces. These procedures are useful for the routine screening of food and other agricultural products for tolerance limits specified in the Food, Drugs and Cosmetics Acts of various countries. Further, they give reliable results and are useful as confirmatory evidence for clinical and medico-legal cases in which positive un- ambiguous identification and quantitative determination, particularly in the latter category, are essential. The eleven organophosphorus compounds and seven carbamates have widely different R, values and most of these compounds yield characteristic coloured spots with at least one of the reagents and these serve to identify a compound. In general, many of the organo- phosphorus compounds give various shades of brown-to-grey colours with chromogenic reagent A, while most of the phenols and carbamates show a dark to light bluish tinge. The carbamates and phenols are further confirmed by the RF values and the purple-to-mauve coloured spots produced with the chromogenic reagent B. The procedure given above has been used successfully with eighteen compounds and shows promise for a wider range of organophosphorus compounds and carbamates. REFERENCES 1. Braithwaite, D. P., Nature, 1963, 200, 1011. 2. Gibbs, H. D., J . Biol. Chem., 1937, 72, 649. 3. Miskus, R., Eldefrawi, M. E., Menzel, D. B., and Svoboda, W. A., J . Agric. Fd Chem., 1961,9, 190. 4. Bates, J. A. R., Analyst, 1965, 90, 453. 6. Stahl, E., Chemikerzeitung, 1958, 82, 323. 6. Anglin, C., and Mckinley, W. P., J . Agric. Fd Chem., 1960, 8, 186. 7. Mills, P. A,, J . Ass. 08. Agric. Chem., 1959, 42, 734. Received April 22nd, 1969 Accepted June 27th, 1969.

ISSN:0003-2654    

DOI:10.1039/AN9699401075

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Analyst, December, 1969, Vol. 94, $9. 1081-1083 1081 Identification of the Denaturants in Surgical Spirits B.P.C. Using 'Thin-layer Chromatography BY M. J. GLOVER (Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, S.E. 1) As surgical spirits are liable to misuse it is important to ensure that minimum standards of denaturing are maintained. A rapid thin-layer chromatographic technique has been devised as a supplement to the official ultraviolet method for determination of two of the three ingredients. SURGICAL spirit and purple mineralised methylated spirit are the only forms of denatured duty-free ethanol unrestrictedly available to the public. The British Pharmaceutical Codex specifies the following formula- Methyl salicylate . . . . 0.5 per cent.v/v Diethyl phthalate . . Castor oil . . .. . . 2.5 per cent. v/v Although other surgical spirit preparations containing mineral naphtha or brucine in place of diethyl phthalate or methyl salicylate are still occasionally seen, the above formula is now the only one of importance. The combination of the three ingredients, each of which has some denaturant value, is considered to minimise the danger of abuse. Samples of surgical spirit, ready for sale, are submitted by Customs and Excise Officers to the Laboratory of the Government Chemist for checking for conformity with the approved formula. Routine quantitative examination is carried out on these samples by determination of residue after evaporation of the alcohol and by determination of the absorption of samples diluted with ethanol, at 306 and 227 nm, corresponding to absorption peaks for methyl salicylate and diethyl phthalate, respectively.I have observed with certain samples that determinations based on ultraviolet absorption measurements have been subject to error attributable only to the presence of strongly absorbing impurities. Gas - liquid chromato- graphy could be used to check the values for these two constituents but will not enable the presence of castor oil to be confirmed without pre-treatment of the sample. . . 2.0 per cent. v/v in industrial methylated spirit1 } EXPERIMENTAL Glass plates (20 x 20cm) coated with 0.25-mm layers of silica gel, with or without fluorescent indicator, were used as standard. Pre-coated plastic sheet could also be used and was especially suitable for individual samples.Alumina, however, could not be used because the methyl salicylate partially hydrolysed under even slightly alkaline conditions. The sample was applied directly to the plate or as a chloroform solution after evaporation of the alcohol. A standard solution is made up to the above formula and aliquots are applied to the prepared plates. As methyl salicylate is appreciably volatile, the plate must not be heated nor exposed to the atmosphere any longer than necessary. On silica gel the three ingredients migrate in the order methyl salicylate, diethyl phthalate, with castor oil remaining near the origin. A carbon tetrachloride - acetone mixture (19 + 1) was the most satisfactory solvent of many tried. Davies and Tunnicliffez showed that with this solvent mixture, castor oil separates into two spots corresponding to the di- and tri- glycerides.The three substances are colourless and their spots are not visible on the plate. Table I indicates some methods of making them visible and the corresponding sensitivities are given in Table 11. The molybdophosphoric acid method used followed that of Davies and Tunni- cliffe2 and the fluorescent spray technique was as described in the method below. This was developed from the use of mixed fluorescent indicators by Jones, Bowyer, Gresham and Howard. 0 SAC; Crown Copyright Reserved.TABLE I REAGENTS AND CONDITIONS OF USE Method 1 Method 2 Method 3 Method 4 Method 5 Reagent or Iodine vapour Molybdophosphoric Self-fluorescence Fluorescent layer Fluorescent sprayS test acidZ Merck GF254 Examined Daylight Daylight Hg vapour lamp, Hg vapour lamp, Daylight under long wavelength short wavelength Background Buff colour Yellow Violet Green Pink TABLE I1 RF VALUES, SPOT COLOURS AND MINIMUM QUANTITIES DETECTED Methods r 1 4 pl*of 1 2 3 4 5 surgical **--* spirit Colour Limit, Colour Limit, Colour Limit, Colour Limit, Colour Limit, Density contains RF of spot nl of spot nl of spot nl of spot nl of spot nl Bright 1 Bright 1 Red 0.25 Methyl 1.18 20 nl 0.86 Brown 2 - - Diethyl 1-12 80 nl 0.65 Yellow 8 - - - - Dark 4 Red 4 Castor oil 0.96 100nl 0.25 Brown 5 Blue 1 salicyla te blue blue phthalate blue 0.16 Brown - Blue Bright 2.5 - - - - pink - - - - - - - Note-As the standard solution is prepared in percentage v/v the limits are in nanolitres (divide by density for weight in micrograms)."a x * ccDecember, 19691 SPIRITS B.P.C. BY USING THIN-LAYER CHROMATOGRAPHY 1083 METHOD REAGENTS- Standard surgical s@irit-Mix 1.0 ml of methyl salicylate, 4.0 ml of diethyl phthalate and 5.0 ml (4.80 g) of castor oil with industrial methylated spirit and dilute to 200 ml with this solvent. SoZvent-Add 5ml of acetone to 95ml of carbon tetrachloride, place in a thin-layer tank lined with chromatographic paper and shake the tank to saturate its atmosphere. Bromine vapour-A suitable concentration is obtained by pouring bromine vapour, without transfer of liquid, from the reagent bottle into an empty tank. FZuorescent indicator spray-Dissolve 10 mg of Rhodmine B and 40 mg of dichloro- fluorescein in 80ml of acetone and 20ml of water.(The physical action of this spray is markedly affected by the presence of water.) PROCEDURE- Prepare 0.25-mm plates of silica gel - gypsum, activate at 100" C for 10 minutes, then allow to cool by exposure to the atmosphere for half an hour. Apply 1, 2, 3 and 4 pl of stan- dard solution and 4 pl of samples, 1 p1 at a time, to the plates. Develop the plates in the solvent mixture until the solvent front has risen 8 cm. Allow the free solvent to evaporate, treat with bromine vapour for about 1 minute then, when the excess of bromine vapour has volatilised spray lightly, but evenly, with the fluorescent dye mixture. The three constituents are visible in daylight. CONCLUSIONS The method outlined permits positive identification with a possible semi-quantitative estimate of all three denaturants at the same time, requiring about an hour and a half for six samples.Among likely substitutes for the specified ingredients, the other phthalate esters give slight variations in RF values and the same response to the spray. Olive oil and other vegetable oils are readily detectable, giving a series of spots extending the length of the chromatogram, which is quite distinct from that given by castor oil. The above denaturants are used separately in a wide variety of industrial methylated spirit preparations such as toiletry articles. Examples of application of this thin-layer tech- nique include the identification of 1 per cent. of castor oil in hair sprays containing lacquer resins, for which improved characterisation of the castor oil was obtained by increasing the acetone content of the eluting solvent to 10 per cent.; diethyl phthalate in the presence of a considerable excess of castor oil; and the presence of methyl salicylate in an aerosol mouth spray. This paper is published by permission of the Government Chemist. REFERENCES 1. British Pharmaceutical Codex, 1968, The Pharmaceutical Press, London. 2. 3. Davies, J . R., and Tunnicliffe, M. E., J . Chromat., 1967, 30, 125. Jones, D., Bowyer, D. E., Gresham, G. A., and Howard, A. N., Ibid., 1966,23, 172. Received February 20tk, 1969 Accepted June 2nd, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699401081

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

1084 Analyst, December, 1969, Vol. 94, $9. 1084-1089 A Thin-layer Chromatographic Method for the Determination of Capsaicin in Ground Paprika BY P. SPANYAR AND MARTA BLAZOVICH (Centval Food Reseavch Institute, Budapest I I , Herman Ottd Jt 15, Hungary) A method has been developed for the analysis of samples containing capsaicin in amounts above and below 10 mg per 100 g. The same thin- layer procedure is used for all samples, but preliminary extraction varies according to the level of capsaicin in the sample. For levels above 10 mg per 100 g, the sample is extracted with diethyl ether; for lower levels, the ether is removed by distillation, the residue dis- solved in ethanol and the solution shaken with light petroleum to remove colouring matter. The ethanolic solution containing the capsaicin is evapor- ated to dryness and the residue extracted with ether.The extract is transferred on to a Kieselgel G layer and developed with a chloroform - ethanol mixture. The capsaicin is made visible with iron(II1) chloride - potassium ferricyanide reagent. THE Joint Committee of the Pharmaceutical Society and Society for Analytical Chemistry on Methods of Assay of Crude Drugs, formed to resolve the problems related to the deter- mination of capsaicin, suggests in one of its reports1 the use of photometric m e t h ~ d s . ~ ? ~ s ~ These methods are suitable for application on a limited scale only, because they are not sufficiently sensitive. They are not suitable for use with ground paprika containing 10mg or less of capsaicin per 100 g.As the capsaicin content of the highest quality ground paprika is often less than 10 mg per 100 g, and a difference of less than 10 mg may be relevant to the classi- fication of paprika, it seemed necessary to work out a rapid method for the analysis of paprika samples containing 1 to 10mg of capsaicin per 1OOg under the conditions of, and with equipment available in, an industrial laboratory. In all colorimetric methods capsaicin must be separated from colouring matter and other interfering substances. Column-chromatographic and extraction methods are tedious and involve loss of capsaicin. In paper chromatography5 the fat extracted from paprika causes elongation of spots and tailing. In recent years thin-layer chromatography has been used for the purification of cap- ~ a i c i n , ~ 9' s8 many adsorbents and solvent systems being suggested.Heusser7 used this method for its quantitative determination, but with his procedure extracts for thin-layer chromato- graphy must contain at least 100 pg of capsaicin, and it is not possible with samples of low capsaicin content to prepare such concentrated extracts. Our aim was to develop a thin-layer chromatographic method for the determination of 1 to 10 mg of capsaicin in 100 g of material, with a suitable adsorbent and developer and a sufficiently sensitive detection reagent. During the development of the method it became evident that the procedure was suitable for the determination of capsaicin contents higher than 10 mg per 100 g, and the preparation of samples was less complicated, without reducing the accuracy of the method.METHOD MATERIALS- 0-25 mm thick being used throughout. Of three adsorbents examined, Kieselgel G was found to be the most suitable, layers 0 SAC and the authors.SPANYAR AND BLAZOVICH 1085 SOLVENT SYSTEMS- Of the solvent systems described in the literature the following were tested for their capacity for separating capsaicin from colouring matter and other interfering substances. RF of. capsaicin Benzene - ethyl acetate (1 + 1 v/v) . . .. .. .. .. .. . . . . 0-31 Chloroform - cyclohexane - glacial acetic acid - methanol (100 + 100 + 8 + 12 v/v) 0-36 . . - Distilled water saturated with diethyl ether . . .. .. .. .. .. .. Chloroform - ethanol (94 + 6 v/v) . . .. .. .. .. .. .. .. - Chloroform - ethanol (96 + 4 v/v) .. .. .. .. .. .. .. . . 0.60 Chloroform - ethanol (97 + 3 v/v) . . .. . . .. . . .. .. . . 0.42 Chloroform - ethanol (98 + 2 v/v) . . .. .. .. .. .. .. . . 0-34 Chloroform - ethanol (99 + 1 v/v) . . .. .. .. .. .. .. . . 0-26 Cyclohexane - chloroform - acetic acid (70 + 20 + 10 v/v) .. .. .. . . 0.06 Chloroform (running twice) . . . . .. .. .. .. .. .. .. 0.1 Chloroform first, then chloroform - ethanol (97 + 3 v/v) . . .. .. .. .. - Chloroform first, then chloroform - ethanol (99 + 1 v/v) . . .. .. .. . . 0.34 Light petroleum (boiling range 60" to 100" C) first, then chloroform - ethanol (99 + 1 v/v) The best separation was obtained with solvent system chloroform - ethanol (99 + I), - therefore this was used in further experiments. SPRAY REAGENTS- Because of the low capsaicin content of paprika, spectrophotometric determination of the capsaicin spots eluted from thin-layer plates is not practicable.The following spray reagents capable of reacting colonmetrically, thus enabling the spots to be evaluated quanti- tatively, were tested for sensitivity of detection. Sensitivity, pg of capsaicin per spot Sulphanilic acid (0.35 per cent.) - sodium nitrite (0-35 per cent.) (1 + l), then Anisaldehyde, 0-5 per cent. in methanol containing 10 per cent. of concen- 0.1 N sodium hydroxide solution . . .. .. .. .. .. 100 trated acetic acid and 5 per cent. of concentrated sulphuric acid . . .. 100 Benzidine, 0.5 per cent. in ethanol, diazotised . . .. .. .. .. 100 Antimony trichloride, 20 per cent. w/v, and acetic anhydride, 2 per cent.w/v, 4 in chloroform . . .. .. .. .. . . .. .. .. 4 solution . . .. . . .. .. .. .. .. .. .. 2 Sulphanilic acid, diazotised, 0.1 g in 20 ml of 10 per cent. sodium carbonate Rhodamine B, 0.05 per cent., in water . . .. .. .. .. .. Not suitable Iron(II1) chloride (15 per cent. in ethanol) - potassium ferricyanide (0.5 per Iron(II1) chloride (15 per cent. in water) - potassium femcyanide (0.5 per Molybdophosphoric acid, 10 per cent. in ethanol . . .. .. .. 2 because of fluorescence of colouring matters cent. in ethanol) (1 + 1) .. .. .. .. . . .. .. 0.1 cent. in water) (1 + 1) . . .. .. .. .. .. .. .. 0.1 The last two of these detection reagent^,^ which form Berlin blue, are the most sensitive and spraying with the aqueous solution is preferred. The spots obtained by using ethanolic spray solutions are much more diffuse and, therefore, less suitable for evaluation.As the reaction is not specific for capsaicin, and the colouring matters and other inter- fering substances present in paprika give the same blue reaction as capsaicin, the method is applicable only if a highly efficient chromatographic separation has been achieved. DETERMINATION OF CAPSAICIN IN POWDERED PAPRIKA SAMPLES WITH CAPSAICIN CONTENTS ABOVE 10 mg PER 100 g REAGENTS- Ethanol, 57 per cent. Diethy2 ether-Analytical-reagent grade, freed from peroxide with 2 per cent. iron(I1) Chloroform. Standard solution of capsaicin, 0.01 mg ml-I in diethyl ether. Potassium ferricyanide solution, 0.5 per cent ., aqueous. Iron(III) cMoride solution, 15 per cent., aqueous.sulphate.1086 SPANYAR AND BLAZOVICH: A THIN-LAYER CHROMATOGRAPHIC METHOD [Ana&St, VOl. 94 Spyay reagent-Mix equal volumes of the potassium ferricyanide and iron(II1) chloride Kieselgel G layers, 200 x 200 mm and 0.25 mm thick, activated at 105" C for 30 minutes, solutions immediately before use. were used. These should be maintained in a desiccator until required for use. PROCEDURE- Accurately weigh 4.000 g of the ground paprika into a 50-ml glass-stoppered graduated cylinder. Add 3 ml of 57 per cent. v/v ethanol to wet the powder. Extract it by shaking with 30ml of peroxide-free ether for 10 minutes and allow to settle for 5 minutes. Filter the ethereal extract into a round-bottomed flask. Repeat the extraction with a further 30-ml portion of the ether and combine the filtrates.Reduce the volume to about 20 ml by distillation under vacuum in a water-bath at 30" to 40" C. Transfer the concentrated extract to a 25-ml graduated flask and adjust to the mark with peroxide-free ether. Further dilute to give a stock solution containing between 5 and 50 pg ml-l of capsaicin suitable for thin-layer chromatography. To obtain stock solutions suitable for Chromatography, samples with varying capsaicin contents should be diluted as follows. Anticipated capsaicin content of Final dilution for thin-layer chromatography the sample, m g per 100 g 1 to 30 Undiluted 30 to 60 x2 60 to 100 x 6 100 to 160 x 10 160 to 200 x 16 Indicate six equidistant starting points on the prepared Kieselgel G layers. With a 0-1-ml micropipette (graduated in 100 divisions) apply successively the following solutions : 0.02, 0.04 and 0.06 ml of standard solution (0.01 mg ml-l) and 0.02, 0.04 and 0-06 ml of stock solution of paprika.Develop the plate in a prepared chromatographic tank by using chloroform - ethanol (99 + 1) as the solvent system. Continue elution until the solvent reaches the upper edge of the layer. Remove the plate, allow the solvent to evaporate and spray evenly with the freshly prepared iron(II1) chloride - ferricyanide reagent. (Care must be taken to spray evenly as unevenness of background will prevent the colour of the spots from being propor- tional to the amount of capsaicin.) The amount of capsaicin present is obtained by comparing the colour and size of the spot from the stock solution with those from the standard solution.If the colour intensity and size of the spots are not comparable, the procedure is repeated with a more suitable amount. The evaluation must be carried out within 1 to 2 minutes after spraying, because inter- fering spots may occur around the capsaicin, which affect the sensitivity of the determination. DETERMINATION OF CAPSAICIN IN PAPRIKA SAMPLES WITH CAPSAICIN CONTENTS BELOW 10 mg PER 100 g REAGENTS- In addition to the reagents listed in the previous method the following are needed. Light petroleum (boiling range 60" to 100" C). Sodium chloride. Sodium hydroxide solution, 0.1 N. PROCEDURE- Accurately weigh 4.000 g of the ground paprika and proceed exactly as described for samples with capsaicin content above 10mg per 1OOg to sentence ending ".. . combine the filtrates," and continue as follows. Remove the ether by distillation under vacuum in a water-bath at 30" to 40" C. Extract the residue by agitating it with four successive portions each of 5 ml of 57 per cent. v/v ethanol, transferring the extracts into a 100-ml separating funnel. Dissolve the residual pigment in 30 ml of light petroleum (boiling range 60" to 100" C), adding the solution to the same separating funnel. Add about 1 g of sodium chloride to the mixture and shake it for 5 minutes.December, 19691 1087 Allow to separate for 10 minutes and transfer the lower ethanolic layer to a 100-ml frac- tionating flask. Extract the light petroleum phase with a further 10 ml of 57 per cent. v/v ethanol. Allow to separate for 10 minutes and add the lower ethanolic layer to the same fractionating flask.Then add 5 ml of 0.1 N sodium hydroxide solution to the combined yellowish ethanolic solution and remove the ethanol by distillation in a water-bath at 75" to 80" C in a stream of nitrogen. If the alkaline solution foams, stop the distillation and cool the solution while maintaining the stream of nitrogen. Transfer the cold solution to a separating funnel, rinsing the flask with a few millilitres of distilled water. Extract the capsaicin solution thus obtained, first with 30 ml then with 20 ml of ether, shaking the mixture each time for 10 minutes. Separate and filter the ethereal layer through a layer of anhydrous sodium sulphate into a round-bottomed flask. Remove the ether by using a water-jet vacuum pump.Then dissolve the residue in peroxide-free ether, transfer the solution into a 25-ml graduated flask and adjust to the mark with peroxide-free ether. Dilute this stock solution as necessary as in the previous procedure for thin-layer chromatography. FOR THE DETERMINATION OF CAPSAICIN IN GROUND PAPRIKA RESULTS AND CONCLUSIONS The method was tested by nearly 100 experiments. As it seemed possible that during the preparation of samples, particularly those with low capsaicin contents which required several steps, the capsaicin might decompose or fail to be quantitatively extracted, various amounts were added to capsaicin-free ground paprika and the samples taken through the procedure. The results are shown in Table I. TABLE I DETERMINATION OF CAPSAICIN ADDED TO GROUND PAPRIKA IN AMOUNTS BELOW 10 mg PER 100 g Capsaicin, mg per 100 g & Added Recovered 1.0 0.7 1.2 1.0 1.1 1.0 2.0 2.0 2.4 2.3 2.0 2.0 3-0 3.0 2-5 2-6 2-7 2-5 4.0 3.9 4.2 4-2 4.6 4.7 6.0 4.6 5.0 5.6 5-2 5.3 Deviation from the average, mg Per 100 g 0-3 0-2 0.0 0.1 0.0 0.1 0.3 0.2 0.1 0.1 0.3 0.2 0.1 0.0 0-2 0.5 0.1 0.1 0.3 0-4 0-5 0.1 0-5 0.1 0.2 s, mg Per 0-187 100 g 0.200 0.212 0-114 0.300 t 0.000 1.120 3.1691 5.890t 0.746 Added Recovered 6.0 6.0 5.5 5-9 5.7 5.8 7.0 6.3 7.9 7.1 7.5 6.9 8.0 8.2 8.5 8.2 8.1 8.2 9.0 8-8 8-6 8.9 9.1 9.3 10.0 10.5 10.0 9.3 10.1 9.9 Deviation Capsaicin, from the mg per 100 g average, & mgper 100 g 0.2 0.3 0.1 0.1 0.0 0.8 0.8 0.0 0-4 0-2 0.0 0-3 0.0 0.1 0.0 0.1 0.3 0.0 0.2 0.4 0.5 0.0 0.7 0.1 0.1 s, m g Per 100 g 0.200 0.192 0.173 0.0895 0.138 t 2.240 1.166 2.590 0-250 0.000 To establish that the amounts of capsaicin added did not differ to a significant extent from those recovered, the t-test was applied.The results are given in Table I. Even at a probability level of 95 per cent. there is no significant difference between the amounts added and recovered. Investigations were also carried out to determine whether the deviations between the added amounts could be considered to originate from a single distribution. The results of1088 SPANYAR AND BLAZOVICH: A THIN-LAYER CHROMATOGRAPHIC METHOD [ArtahySt, VOl. 94 the ?-test proved that the deviations can be considered to originate from a single distribution, even at the 90 per cent. probability level, thus a common standard deviation could be adopted.Within the given range of concentrations (1 to 10 mg of capsaicin per 100 g) the standard deviation of the method, s, is 0.324 mg per 100 g. However, this is not true and the standard deviation will be different if the natural capsaicin contents of various samples of paprika falling in the same range as those added to a single sample are determined (Table 11). TABLE I1 DETERMINATION OF CAPSAICIN IN SAMPLES OF GROUND PAPRIKA WITH CAPSAICIN CONTENTS BELOW 10 mg PER 100 g Deviation Deviation Capsaicin from the Capsaicin from the content , average, S, content, average, $ 9 Sample mg per 100 g mg per 100 g mg per 100 g Sample mg per 100 g mg per 100 g mg per 100 g 1 10.7 0.7 0.749 4 1.4 0.1 0-282 9.4 0.6 1.1 0.2 10.6 0.6 1.2 0.1 10.2 0.2 1.8 0.5 9.0 0.1 1.2 0.1 11.2 0.6 2.1 0.0 11.8 0.1 2.2 0.1 11.8 0.1 1.7 0.4 12.0 0.3 2.3 0.2 1.3 0.4 6.5 0.6 1.8 0.1 5.4 0.5 1.5 0.2 5-0 0.9 1.7 0.0 6.9 1.0 In this event, standard deviations at the 95 per cent.probability level can be considered to originate from a single distribution, and the standard deviation of the method, s, is 0.499 mg per 100 g. The difference probably results from the variation in the amount and composition of the pigment content, and this strongly affects the separation on the layer, the shape of spots and their evaluation. The behaviour of various paprika types shows marked differences at various stages of the assay, e.g., in the formation of emulsions and foaming. The above calculations were carried out for the method used in the determination of capsaicin contents above 10 mg per 100 g ; 20 or 50 mg of capsaicin per 100 g are added to paprika samples containing no capsaicin and the amount recovered is established by the method described.TABLE I11 DETERMINATION OF CAPSAICIN CONTENT ADDED TO PAPRIKA SAMPLES IN AMOUNTS ABOVE 10 mg PER 100 g 2 11.7 0.0 0-300 5 2.3 0.2 0.244 3 2.0 0.3 0.282 6 5-6 0-3 0-793 Results are given in Table 111. Deviation Deviation Capsaicin, from the Capsaicin, from the mg per 100 g average, s, mg per 100 g average, s, - mgper mgper - mgper mgper Added Recovered 100 g 1OOg t Added Recovered 1OOg 1OOg t 20 18-6 1-36 1.239 0-102 50 46.2 2-63 1-770 2.09 20.9 0-94 49.7 0.87 20.9 0.94 49.7 0-87 20.9 0-94 49-7 0-87 20-9 0.94 45.9 2-93 20.9 0.94 49.9 1-07 18 8 1.16 49.9 1-07 17.9 2.06 50.5 1-67 18-9 1-06 50-0 1-17 20.9 0.94 46.8 2.03 t,, (f= 9) = 2.262.It was shown by the F-test that there was no significant difference between the two The t-test showed standard deviations, while that of the sample was 1.498mg per 1OOg. that there was no significant difference at the 95 per cent. probability level.December, 19691 FOR THE DETERMINATION OF CAPSAICIN IN GROUND PAPRIKA 1089 The shortened method was used to analyse several paprika samples and the results are given in Table IV. The x2-test proved that the standard deviations originate from a single distribution (xzs0 > x2 as calculated), thus a single standard deviation could again be adopted (s is 1.060 mg per 100 g). TABLE IV DETERMINATION OF CAPSAICIN CONTENT AT LEVELS ABOVE 10 mg PER 10 g OF PAPRIKA SAMPLES Deviation Deviation Capsaicin from the Capsaicin from the content, average, S, content average, s, Sample mg per 100 g mg per 100 g mg per 100 g Sample mg per 100 g mg per 100 g mg per 100 g 1 19.5 0.32 0-408 3 34.4 0.36 1.023 19.0 0.18 33.3 1.46 18-7 0.48 35.1 0.34 19.0 0.18 36-1 1-34 19-7 0.52 34-9 0.14 30.3 0.40 43.5 2.30 31.2 0.50 46-0 0.20 29.5 1.20 47.5 1.70 32.3 1-60 45.7 0.10 2 30.2 0-50 1.079 4 46.3 0.50 1-456 It was considered important to compare the results by a standard photometric methodlo with those obtained by using the thin-layer chromatographic method.This comparison could be carried out only with samples containing capsaicin above 10 mg per 100 g, because the photometric method is not suitable for use a t lower levels.Therefore the shortened thin-layer chromatographic method was used for the comparison. Results are given in Table V. TABLE V DETERMINATION OF THE CAPSAICIN CONTENT OF GROUND PAPRIKA SAMPLES BY THIN-LAYER CHROMATOGRAPHY AND PHOTOMETRY Capsaicin, mg per 100 g, by chromato- photo- Sample graphy metry 1 160.0 164.0 2 62.6 69.4 3 19.2 21.9 4 30.7 28-9 zz&?-- - Deviation from the average, mgper 100 g 4-0 6.8 2-7* 1.8 * t88.8 Capsaicin, mg per 100 g, by- chromato- photo- Sample graphy metry 5 34-8 37.5 6 45.8 44.7 7 206.2 194.0 8 180-6 179.5 rthin-layer < t. - Deviation from the average, mg per 100 g 2.7 1.1 12.2 1.1 Each result is the mean of those obtained from five different samples. Apart from one instance, there was no significant difference at the 95 per cent. probability level between the results obtained by the two methods. The authors express their appreciation to Hedwig Nonn-Sas for carrying out the mathematical and statistical calculations. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Report of the Joint Committee of the Pharmaceutical Society and the Society for Analytical Schulte, K. E., and Kruger, H. M., Arch. Pharm., Berl., 1957, 290, 202. SpanyAr, P., Kevei, J., and Kiszel, J., l?lelm. Ipar, 1956, 10, 52. Gibbs, H. D., J . Biol. Chem., 1927, 72, 649. Waldi, D., “Chromatographie,” E. Merck A.G., Darmstadt, 1958. Teichert, K., Mutschler, E., and Rochelmeyer, H., 2. analyt. Chem., 1961, 181, 325. Heusser, D., Planta Med., 1964, 12, 237. Friedrich, H., and Rangoonwala, R., Naturwissenschaften, 1965, 52, 514. TyihAk, E., “A re‘tegkromatogrdfia uj eredme’nyei” (New results of thin layer chromatography), SpanyAr, P., Kevei, J., and Kiszel, J., Konserv. Paprikaipar, 1959, 7, 312. Chemistry, Analyst, 1964, 89, 377. Magyar KCmikusok Egyesiilete BiokCmiai SzakosztAlya Kiadvbnya, Budapest, 1966. Received February 12th, 1969 Accepted June 22n4 1969

ISSN:0003-2654    

DOI:10.1039/AN9699401084

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

1090 Arzalyst, December, 1969, Vol. 94, $$. 1090-1094 The Resolution of Mixtures of Aliphatic and Terpene-type 2,4-Dinitrophenylhydrazones by Thin-layer Chromatography on Silica Gel BY J. H. DHONT AND G. J. C. MULDERS-DIJKMAN (Central Institute for Nutrition and Food Research, TNO Zeist, The Nethedands) Mixtures of aliphatic and terpene-type 2,4-dinitrophenylhydrazones can be separated by thin-layer chromatography on silica gel layers whether or not impregnated by vapour-phase adsorption of nitromethane or acetonitrile. The Galanos and Kapoulas method for RF value correction has been tested on different types of silica gel layers. The method has given consistent results on the silica gel preparation used. IN a previous paper1 a method has been described for the separation of complex mixtures of 2,4-dinitrophenylhydrazones, by means of thin-layer chromatography, into two well defined chemical groups, viz., the derivatives of the aromatic carbonyl compounds and the aliphatic carbonyl compounds, including those of the terpene type.With this method the aliphatic and terpene-type 2,4-dinitrophenylhydrazones, however, do not separate from each other and are always located on the plates as a single group. The separation of mixtures of non-terpene-type aldehyde and methyl ketone 2,4-dinitro- phenylhydrazones according to chain length by partition chromatography on Carbowax2 or 2-phenoxyethanoP has been reported many times, but only two papers that described the separation of several terpene 2,4-dinitrophenylhydrazones have come to our attention.* ~5 In this paper the separation is reported of aliphatic and terpene-type 2,4-dinitrophenyl- hydrazones on plates whether or not impregnated with nitromethane or acetonitrile.R, values have been corrected for plate variations according to Galanos and Kapoulas.6 CHROMATOGRAPHIC TECHNIQUE PREPARATION OF THE CHROMATOPLATES- Twenty-five grams of silica gel G (Merck) were suspended by shaking the gel in an Erlenmeyer flask with 20 ml of water and 40 ml of ethanol. From the resulting slurry, thin layers, about 0.25 mm thick, were prepared by using the standard Desaga equipment. The layers were allowed to equilibrate with the laboratory atmosphere for at least 3 hours. Solu- tions of the 2,4-dinitrophenylhydrazones in ethyl acetate were spotted 1-5 cm from the bottom of the plate.A line was drawn through the layer at a distance of 11.5 cm from the start and the plates were removed from the developing tank as soon as the solvent reached this line. A standard reference mixture was spotted on each plate at three different positions on the starting line to correct for plate differences in the R, determinations (see Determination of R, values). Filter-paper was dipped into the solvent in the usual way to saturate the tank atmosphere. 0 SAC and the authors.DHONT AND MULDERS-DI JKMAN 1091 Three chromatographic systems were investigated, vix., system A, layer silica gel G (equilibrated with the atmosphere), solvent benzene - hexane (1 + l), and number of de- velopments 5 ; system B, layer silica gel G (equilibrated with the atmosphere) impregnated with nitromethane by vapour-phase absorption, solvent hexane, and number of developments 3; and system C, layer silica gel G (equilibrated with the atmosphere) impregnated with acetonitrile by vapour-phase adsorption, solvent hexane, and number of developments 3.PREPARATION OF SYSTEM A- The layer was prepared as described above. The plates were developed by five multiple runs in the same solvent mixture. After each development the adhering solvent was allowed to evaporate by exposing the plate to air for a few minutes. For details of multiple chromato- graphy see references 7, 8, 9 and 10. PREPARATION OF SYSTEM B- The impregnation of silica gel layers with an immobile volatile phase by adsorption of the phase from its vapour was first described by Urbach.3 The immobile phase (methanol in Urbach’s system) was adsorbed from the atmosphere in the developing tank, and was kept saturated with methanol vapour by placing a plug of cotton-wool impregnated with liquid methanol inside the tank. A slightly different procedure was used in our experiments.The silica gel layer of the chromatographic plate was kept saturated with nitromethane by placing a second thin-layer plate impregnated with liquid nitromethane opposite the first one, a t a distance of about 1 mm, in a sandwich arrangement. The sandwiched plates were then placed together in an ordinary chromatographic tank containing hexane as solvent. As normal thin layers are easily damaged when pouring liquid nitromethane on the silica gel surface, special heavy duty “second” plates were prepared for this purpose as follows.Twenty-five grams of Celite 545 and 10 g of gypsum were suspended in a 3 per cent. solution of poly(viny1 alcohol) in water. This slurry was spread on normal glass plates, maximum thickness 1 mm. After drying for about 15 hours a t room temperature the plates are ready for use and last for several months in daily use, but have to be re-impregnated with nitromethane for every new experiment. The silica gel layers for the chromatographic plates were prepared as described under Preparation of the chromatoplates. After drawing the front line and spotting, the Celite layer of the second plate is impregnated by pouring nitromethane on to the plate. The excess of liquid is allowed to drain off.The wet Celite plate is then mounted together with the spotted plate in a sandwich arrangement and, after an equilibration time of 1 to 3 minutes, developed by placing the sandwich in the developing chamber containing hexane as mobile phase. The hexane is evaporated from the plate in a few seconds, by gently blowing over its surface. The plate is then re-mounted together with the impregnated plate in the sandwich arrange- ment. This procedure was repeated until the plate had been developed three times in hexane. As soon as the solvent reaches the front line, the sandwich is dismounted. PREPARATION OF SYSTEM C- The same method has been followed as mentioned under B, with the exception of acetonitrile being used as the adsorbed vapour phase instead of nitromethane.Also in this instance an equilibration time of 1 to 3 minutes was found necessary before the sandwich was placed in the hexane. DETERMINATION OF R, VALUES- As it was our aim to find a method for the identification of small amounts of unknown components with the aid of accurate R, determinations, a procedure that would give R, values with as little variation as possible was necessary. Moreover, the values had to be reproducible, irrespective of the batch of silica gel used in preparing the plates, the variation in room temperature, the atmospheric humidity and the tank saturation. Of the many different methods mentioned in the literature and tested in our laboratory, that described by Galanos and Kapoulas6 for paper chromatography was found to be the best for our purpose.To some extent this procedure was also used by Pataki for thin-layer chromatography,ll whereas Williams12 evolved a similar method from statistical techniques.1092 [Analyst, Vol. 94 Galanos and Kapoulas based their method on the supposition that all variations causing the R, value to deviate from this “true” value are of a linear nature. So the “true” Rp = R$ value can be related to the RF value “found” by the general equation- & = aR, + b. By using two reference Rg values, viz., RFA and RFB, the constants a and b can be found from the following relationships- DHONT AND MULDERS-DIJKMAN : THE RESOLUTION OF MIXTURES OF R , and RFB are the RF values of the reference compounds found on the same chromatogram as the compound of which the R, value has to be determined.As reference compounds the 2,4-dinitrophenylhydrazones of acetaldehyde and of decanal were used in systems A and C. In system B a mixture of the 2,4-dinitrophenylhydrazones of butanal and decanal was used. On each chromatogram a mixture of these reference compounds was always run on at least three different places. The RF values obtained from the first chromatogram were taken as the RF values and all Rp values found on later chromatograms were corrected in accordance with these values. VARIATION OF RF VALUES WITH DIFFERENT TYPES OF SILICA GEL- It is often stated that RF values depend to a great extent on the kind of silica gel used when preparing the layers. It was, therefore, of interest to test the Galanos and Kapoulas procedure in some extreme cases.Experiments were carried out with three different types of silica gel, viz., silica gel G (Merck), silica gel DSF-5 (Camag) and silicic acid for column chromatography (Mallinckrodt). The last product was sieved to less than 60 mesh and mixed with 10 per cent. gypsum (less than 60 mesh) before preparing the slurry. TABLE I INFLUENCE OF THE TYPE OF SILICA GEL ON THE RF VALUE Solvent benzene (two developments) Propanal Benzophenone 2,4-dinitrophenylhydrazone 2,4-dini trophen ylhydrazone A f \ r > A Layer ~RF* Standard deviation 2 R ~ Standard deviation Silica gel G (Merck) . . . . 0.536 0.021 0.724 0.010 Silicic acid (Mall) . . . . 0.768 0.051 0.880 0-036 Silica gel (Camag) . . . . 0.433 0.060 0.580 0.078 * Mean of four experiments. TABLE I1 INFLUENCE OF THE TYPE OF SILICA GEL ON THE R$ VALUE Same conditions as in Table I Standard reference values used were 0.328 for ethanal 2,4-dinitrophenylhydrazone and 0676 for decanal2,4-dinitrophenylhydrazone Layer Propanal Benzophenone 2,4-dinitrophenylhydrazone 2,4-dinitrophenylhydrazone r \ r \ A A 2Rg Standard deviation 2Rg Standard deviation Silica gel G (Merck) .. . . 0.419 Silicic acid (Mall) . . . . 0-428 Silica gel (Camag) . . . . 0.408 0.011 0-004 0.008 0-573 0.578 0.579 0.018 0.003 0.005 Chromatograms were obtained on each preparation with the 2,4-dinitrophenylhydrazones of benzophenone and of propanal as test compounds. The 2,4-dinitrophenylhydrazones of ethanal and decanal were used as reference compounds. In Table I the results of these experi- ments are shown.Each day fresh plates were prepared, the best of each batch being used for the experiment, so that the R, values were influenced by day-to-day variations as wellDecember, 19691 ALIPHATIC AND TERPENE-TYPE 2, 4.DINITROPHENYLHYDRAZONES 1093 as by plate-to-plate variations . According to Galanos and Kapoulas' procedure the R, values were then compared. taking the R, values obtained on silica gel G plates some months before as the standard values . The results are given in Table I1 . It can be seen that. although the R, values in Table I strongly depend on the silica gel used. the corrected values in Table I1 are strictly comparable . TABLE I11 RF VALUES OF SOME ALIPHATIC AND TERPENE-TYPE 2. 'GDINITROPHENYLHYDRAZONES Svstem A Svstem B Svstem C Alkanals- Methanal .. Ethanal .. Propanal . . Butanal .. Pentanal .. Hexanal . . Heptanal . . Octanal .. Nonanal .. Decanal . . Undecanal . . Dodecanal . . Propan-2-one . . Butan-2-one . . Pentan-2-one . . Hexan-2-one . . Heptan-2-one Octan-2-one . . Nonan-2-one . . Undecan-2-one Cydodkanones- Cyclopentanone Cyclohexanone Cycloheptanone Cyclo-octanone a-Ionone . . 8-Ionone . . a-Methylionone 8-Methylionone Carvone . . Carvotane acetone Citronella1 . . Piperitone . . Pseudoionone . . Pulegone . . Isopulegone . . a-Thujone . . P-Thujone . . Dihydrocarvone Citral . . .. Perilla-aldehyde a-Phellandral . . Camphor .. Fenchone . . Methylheptenone Menthone .. lsomenthone . . Alkanones- Terpenes- .. .. .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. * . .... .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . .. .. . . .. .. .. .. .. .. .. .. .. 6 R ~ 0.228 0.217 0.334 0-391 0.436 0.476 0-507 0.527 0.554 0.565 0.587 0.599 0-224 0.366 0.409 0-460 0-499 0-530 0.550 0-604 0.245 0.353 0.394 0-304 0.603 0-626 0.673 0.661 0.618 0.670 0.489 0.388 0-507 0-466 0.520 0.525 0-625 0.568 0.380 0-437 0-642 <O-05 (0.05 0.438 0.513 0.655 'RF 0.016 0.045 0-07 1 0.104 0.148 0.207 0.274 0.368 0.459 0.573 0.659 0.757 0.071 0.129 0-186 0.255 0.342 0.442 0.549 0.734 0.113 0.1 74 0.242 0.317 0.539 0.599 0-676 0-733 0-317 0-485 0.3 19 0.429 0.397 0.571 0.407 0.462 0.457 0.337 0.227 0.198 0-495 0 0 0-277 0.545 0-541 'RF 0-062 0.109 0- 169 0.236 0-306 0-381 0-465 0-557 0.648 0.728 0-780 0.839 0-183 0.279 0-345 0.454 0-532 0.628 0-703 0.824 0.282 0.371 0-469 0.528 0-727 0-756 0.792 0.819 0.537 0-674 0.519 0-637 0-604 0.636 0.561 0.659 0-660 0-566 0.444 0-388 0.677 0 0 0.478 0.734 0.734 DISCUSSION Table I11 shows the R, values obtained in the systems A.B and C . All values were obtained on chromatograms in which the solvent had moved to the front line 11.5 cm above the start for five or three times. depending on the system used . From Fig . 1 is can be seen that this distance is not critical if the Rg values are determined . The closely related isomers. such as cc- and /3-thujone. menthone and isomenthone. are separated in system A. whereas1094 DHONT AND MULDERS-DI JKMAN the ionone isomers are not separated from each other in this system.They are, however, easily separated when using system B. The most useful separations are obtained when two-dimensional chromatography is applied, taking system A for the first direction and system B or C for the second. Fig. 2 shows a chromatogram obtained with these systems. 0.30- 4 e r g I I t I I Distance, s t a r t to frontline, cm Fig. 1. Dependence of Rp and R’F on the distance start - solvent front. Solid lines corres- pond to system B, broken lines t o system A. 0 RF values, 0 RF’ values. 1 and 2, p-methyl- ionone 2,4-dinitrophenylhydrazones ; 3, a-thujone 2,4-dinitrophenylhydrazone ; 4, citronella1 2,4-dini- trophen ylh ydrazone I ’0 Fig. 2, Direction I, system A; direction I1 system B. 1, citral; 2, perilla-aldehyde; 3, carvone; 4, pseudoionone; 5, a-thujone; 6, /3-thujone; 7, car- votane acetone; 8, a-ionone; 9, p-ionone; 10, a-methylionone ; 11, /3-methylionone The values in systems B and C were, each time, obtained by using hexane that was freed from aromatic hydrocarbons by shaking it with sulphuric acid (sp.gr.1.84). Small amounts of aromatics do not have a marked influence on R, if they are corrected according to Galanos and Kapoulas’ procedure. In experiments in which 10 per cent. of benzene was added to the hexane, uncorrected R, values showed a large positive deviation, while Rg values under identical conditions showed a small positive error of 0.02 to 0.04. The purity of the hexane apparently is not of great significance. We thank Chemische Fabriek Naarden for some of the terpene carbonyls, and Dr. H. van Duin for some of the unsaturated aldehydes. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES Dhont, J. H., and Dijkman, G. J. C., Analyst, 1967, 92, 431. Badings. H. T., and Wassink, J. G., Ned. Melk- en Zuiveltijdschr., 1963, 17, 132. Urbach, G., J . Chromat., 1963, 12, 196. Nano, N., and Sancin, P., Annuli Chim., 1963, 53, 677. Vashist, V. N., and Handa, K. L., J . Chromat., 1965, 18, 412. Galanos, D. S., and Kapoulas, V. M., Ibid., 1964, 13, 128. Jeans, A., Wise, C. S . , and Dimler, R. J., Analyt. Chem., 1951, 23, 415. Lenk, H. P., 2. analyt. Chem., 1961, 184, 107. Thoma, J. A., Ibid., 1963, 35, 214. Patah, G., “Dunnschicht Chromatographie in der Aminosaure- und Peptid Chemie,” Walter de Gruyter and Co., Berlin, 1966. Williams, D. A., J . Chromat., 1967, 26, 280. Received March 27th, 1969 Accepted May 27th, 1969 -, J . Chromat., 1963, 12, 441.

ISSN:0003-2654    

DOI:10.1039/AN9699401090

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Analyst, December, 1969, Vol. 94, ee. 1095-1098 1095 Spectrophotometric Determination of the Fungicide Dithianon in Aqueous Formulation Delan-Col with 2,4Dinitrophenylhydrazine BY S. H. YUEN (Imperial Chemical Industries Limited, Agricultural Division, Jealott's Hill Research Station, Bracknell, Berkshire) A simple and specific spectrophotometric method for determining dithianon in Delan-Col, based on reaction between the carbonyl groups of the fungicide and 2,4-dinitrophenylhydrazine, is described. It consists in dissolving the sample in an acetone - chloroform mixture, removing the solvents and refluxing the residue in neutral methanolic 2, Cdinitrophenyl- hydrazine. The resulting solution is made alkaline with methanolic N sodium hydroxide to give an amber colour, the optical density of which is measured a t 550 nm.The error of the method was 2.1 per cent. THE newly established fungicide, dithianon, 2,3-dicyano-lY4-dihydro-l ,4-dithia-anthraquinoneY has shown extensive biocidal properties1 It is particularly effective against scab, Venturia spec., on apples, pears and peaches, and has a low mammalian toxicity.lJ A method, suitable for the analysis of Delan-Col,* a concentrated stable aqueous suspension of dithianon con- taining wetting and dispersing agents, was required. The absorptiometric method described by Goto and It03 for determining dithianon in formulations was based on dissolution of the fungicide in acetone, and reduction with sodium sulphide in methanol to give a green - yellow colour, which was measured at 375 nm.This method was not suitable for formulations containing surfactants, which absorb at 375 nm. The use of 2,4-dinitrophenylhydrazine for isolating and detecting carbonyl compounds as their phenylhydrazones is well In 1951, Lappin and Clark5 demonstrated the general applicability of a colorimetric procedure for determining small amounts of carbonyl compounds. The sample to be analysed was heated in an acidified methanolic solution of 2,4-dinitrophenylhydrazine, made alkaline with potassium hydroxide, and the absorbance of the final solution measured at 480 nm. The reagent has since been widely used.6 t o 12 In these determinations, acidic solutions of 2,4-dinitrophenylhydrazine were used, resulting in a high background absorption. To minimise this effect, the 2,4-dinitrophenylhydrazone formed was extracted into iso-octane,8 pentyl alcohollo or hexane12 before absorptiometric measurement.It was evident that dithianon did not react quantitatively with 2,4-dinitrophenylhydra- zine in acidic solution. In preliminary experiments, neutral methanolic solutions containing dithianon and 2,4-dinitrophenylhydrazine were heated either at 50" C for 30 minutes, or at 100" C for 5 minute^.^ In both instances, inconsistent results were obtained. Subsequently the dithianon solution was evaporated to dryness, the residue refluxed in methanolic 2,4-di- nitrophenylhydrazine and the solution then made alkaline. Under suitable conditions, the reagent blank was low, and intensity of the colour developed was proportional to the amount of dithianon.The coloured product formed is probably an azophenol derivative, as described by Borsche13 for the analogous reaction of p-benzoquinone with 2,4-dinitrophenylhydrazine. EXPERIMENTAL APPARATUS- REAGENTS- Spectrojbhotometer-A Unicam SP600 was used. Aluadwn chips, size 12-Obtained from the Norton Grinding Wheel Co. Limited. 2,4-Dinitrophenylhyd~a~~ne reagent-Shake mechanically an excess of 2,4-dinitrophenyl- hydrazine in methanol for 15 minutes, and filter the solution through a Whatman No. 40 * A registered trade mark of Plant Protection Limited. 0 SAC and the author.1096 YUEN : SPECTROPHOTOMETRIC DETERMINATION OF THE [AnaZyst, Vol. 94 filter-paper. Dilute 20ml of the filtrate to 100ml with methanol. This reagent is stable for at least 3 weeks under normal laboratory conditions. Sodium hydroxide, N, ilz 80 per cent.methanol-Dissolve 20 g of sodium hydroxide in 100 ml of water, add methanol nearly to 500 ml, and dilute the cooled solution to 500 ml with methanol. Standard dithianon solution-Dissolve 0.10 g of pure dithianon (C1,H,02N2S, ; molecular weight, 296-3; m.p. 225" C) in 200 ml of chloroform. This solution should be kept at 5" C when not in use. PROCEDURE- Weigh a portion of well mixed sample, containing about 0.1 g of dithianon, into a 50-ml beaker, add 15 ml of acetone and triturate with a glass rod until a fine suspension is obtained. Transfer to a 250-ml graduated flask, wash the beaker with 150 ml of chloroform, in small portions, and collect the washings in the same flask. Shake the flask vigorously for 1 minute, dilute to the mark with chloroform, and mix; this is solution A.Transfer 5.0 ml of solution A, and 0, 1.0, 2.0, 3.0, 4.0 and 5-0 ml of standard solution, equivalent to 0,0.5, 1.0, 1-5,2.0 and 2.5 mg of dithianon, respectively, to seven dry 1 x 6-inch test-tubes with B19 necks, and adjust the volume of each standard to 5 ml with chloroform. Add an alundum chip to each test-tube, and evaporate the solvents by passing a stream of air into the tube, which is half immersed in a water-bath at about 95" C. Add 5.0 ml of 2,4-dinitrophenylhydrazine reagent, and boil under reflux with the test-tube half immersed in a water-bath at 95" C for 5 minutes. Stopper the test-tube loosely, swirl the contents to dissolve any precipitate adhering to the glass walls, then cool in water.Add 10.0 ml of methanolic sodium hydroxide solution and mix. Within 5 to 20 minutes, measure the optical densities of the solutions at 550 nm in a 1-cm optical cell, against the reagent blank as reference. Draw the calibration graph relating optical densities of standards to concentrations of dithianon in milligrams, and read off the dithianon content of sample solution; alternatively, calculate the dithianon content by interpolation. 5 Y Dithianon content of sample, per cent. = - W 1 ml of solution = 0-5 mg of dithianon. Let this amount be Y mg. where W = weight of sample taken in grams. DISCUSSION PREPARATION OF THE 2,4-DINITROPHENYLHYDRAZINE REAGENT- As 2,4-dinitrophenylhydrazine is sparingly soluble in methanol (0-05 per cent .), its saturated solution is obtained only by prolonged shaking, and the procedure for its preparation requires standardisation.The stock reagent is prepared by mechanical shaking of an excess of the chemical in methanol for 15 minutes, and the working reagent made by diluting the filtered stock reagent with methanol. To find the concentration of 2,4-dinitrophenylhydrazine reagent, which would give a minimum background but adequate sensitivity, four working reagents were prepared by diluting 10, 15, 20 and 30ml of stock reagent to 100ml with methanol. Optical density, when measured 5 minutes after preparing the final solution, was proportional to concentration in the range 0 to 2-5mg of dithianon a t the two higher concentrations of reagent, but graphs relating optical density and concentration showed slight curvature with increasing amounts of dithianon when 10 or 15 ml of stock reagent were used.Both the sensitivity of the method and the reagent blank increased with increased concentration of 2,4-dinitrophenylhydrazine. A dilution of 20 ml of stock reagent to 100 ml with methanol was chosen. With this working reagent, the background absorption was insignificant, while a linear calibration graph was obtained. STABILITY OF COLOUR- The colour developed by refluxing dithianon in 2,4-dinitrophenylhydrazine reagent is strongly yellow, but the final solution is amber. The temperature and period of reflux were not critical provided that dithianon was in solution. When the solution was made alkaline, it immediately darkened but cleared within 1 to 2 minutes to the characteristic amber colourDecember, 19691 FUNGICIDE DITHIANON IN AQUEOUS FORMULATION DELAN-COL I097 attributed to the azophenol.This colour was stable in diffused light for about 20 minutes, then faded slowly. A reduction of 20 per cent. in intensity was evident when the final solution was kept for 12 hours. More rapid fading occurred when the solution was exposed to sunlight. 0 I I I I I 300 350 400 450 500 550 t 0 Wavelength, nm FIG. 1. Absorption spectra (with a Unicam SPSOO) of: A, reagent blank; B, 0.6 mg of dithianon; C, 1 mg of dithianon; and D, 2 mg of dithianon Spectrum curves for 0 to 2 mg of dithianon, prepared by measuring the optical densities of final solutions from 300 to 600 nm, are shown in Fig. 1. Dithianon showed an absorption maximum at 360 nm.At this wavelength the background absorption, attributed to 2,4-dini- trophenylhydrazine, was high, and interference by surfactants was more serious than that at a longer wavelength. Absorption from the reagents alone was less significant above 450 nm; 550nm was selected as the wavelength at which there was a reasonably high absorption from sample and minimum from the reagent blank. The calibration graph of optical density at 550 nm against 0 to 2.5 mg of dithianon is linear and is suitable for analysing Delan-Col. An optical density of 0.75, measured against the reagent blank, was obtained with 2.5 mg of dithianon. EXTRACTION OF DITHIANON- Direct extraction of the fungicide from an aqueous suspension of Delan-Col with chloro- form (solubility of dithianon, 1 per cent.) was not quantitative, because the fungicide was protected by a layer of water and dispersing agents.A recovery of 70 per cent. was achieved by repeated extraction of Delan-Col with chloroform from 6 M calcium chloride, or saturated ammonium nitrate solution, which acted as salting-out agents. Delan-Col is, however, readily dispersed in water-miscible organic solvents, such as acetone or methanol, from which dithianon can be quantitatively extracted into chloroform. Acetone is preferred, its it has a comparatively high solubility for dithianon (1-5 per cent.), and is subsequently completely removable together with chloroform at 95" C, at which temperature the fungicide is stable. INTERFERENCES- Traces of water present in standard dithianon solution, chloroform extract or 2,4-dinitro- phenylhydrazine reagent do not affect the method, but the reagent must be neutral, as an acidic reagent gives rise to a high blank value and an alkaline one to a green tint.Common organic solvents, such as alcohols, hydrocarbons, ethers, chloroform, methylene chloride and light petroleum, did not interfere, but 1,4-dioxan and cyclohexane gave brown colours. Compounds containing carbonyl groups, e.g., aldehydes and ketones, must be absent from the test. Methanol, contaminated with acetone or other carbonyl compounds, requires re- distillation before it is used for preparing the 2,4-dinitrophenylhydraine and sodium hydroxide reagents.1098 YUEN RESULTS To establish the accuracy of the proposed method, pure dithianon equivalent to 50-0 per cent.was added to six blank formulations of Delan-Col containing wetting and dispersing agents, and the mixtures were analysed for dithianon. Results for the percentage of dithianon found are given in Table I. TABLE I DETERMINATION OF DITHIANON IN LABORATORY-MADE DELAN-COL Dithianon added, per cent. 50-0 50.0 50.0 50.0 50.0 50.0 Dithianon found, per cent. 49-6 52-3 51-0 49.7 50-0 51.4 Recovery, per cent. 99-2 104-6 102.0 99.4 100.0 102.8 The dithianon content found averaged 50.7 per cent. with a standard deviation of The error of the method was In routine determination, the proposed method is capable + l a 1 per cent., equivalent to 101.3 percentage recovery. considered to be 2-1 per cent. of giving less than 2 per cent. error among replicates. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES Berker, J., Hierholzer, O., and Mohr, G., in “Proceedings of the Second British Insecticide and Scott, P. D., in op. cit., 1963, p. 367. Goto, S., and Ito, F., Japan Analyst, 1967, 16, 31. Vogel, A. I., “A Text-book of Practical Organic Chemistry,” Longmans, Green and Co., London, Lappin, G. R., and Clark, L. C., Analyt. Chem., 1951, 23, 641. Mendelowitz, A., and Riley, J. P., Analyst. 1953, 78, 704. Banks, T., Vaughn, C., and Marshall, L. M., Analyt. Chem., 1955, 27, 1348. Toren, P. E., and Heinrich, B. J., Ibid., 1955, 27, 1986. Schaffert, R. R., and Kingsley, G. R., J. Biol. Chem., 1955, 212, 69. Sathe, V., Dave, J. B., and Ramakrishnan, C. V., Nature, 1956, 177, 276. Tsao, M. U., Lowrey, G. H., and Graham, E. J., Analyt. Chem., 1959, 31, 311. Lawrence, R. C., Nature, 1965, 205, 1313. Borsche, W., Justus Liebigs Annln Chem., 1907, 357, 180. Fungicides Conference,” 1963, p. 351. 1948, p. 919. Received January lst, 1969 Accepted June loth, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699401095

出版商:RSC    

年代:1969

数据来源: RSC

摘要:

Analyst, December, 1969, Vol. 94, @$. 1099-1105 1099 Ion Exchange in Non-aqueous Media : The Anion-exchange Behaviour of Phenols on De-Acidite FF Resin in Methanolic Media BY D. E. THOMAS AND J. D. R. THOMAS (De9artment of Chemistry, University of Wales Institute of Science and Technohgy, Cardifl, CF1 3NU, Wales) The applications of ion-exchange resins to separations of organic com- pounds in non-aqueous media are briefly surveyed. The technique has been examined for its potential in separating phenols of high pKa and of closely similar pKa values. The latter can be resolved by a rigorous combination of the use of amine - acetic acid buffer solutions in methanol, and gradient elution. The mononitrophenols, cresols and xylenols have been examined and the weakly acidic xylenols are considered to represent the extreme practical limit of the method. BY using non-aqueous and mixed solvent systems, many new and modified ion-exchange separations have been developed in recent years.Such separations of inorganic species have recently been reviewed,l but there are also several instances of the successful application of this separation dimension with organic compounds, thus complementing separations in aqueous media.2 Suitable media include alcohols, dioxan, acetonitrile , acetic acid and benzene. Acetonitrile has been used for the ion exchange of weak bases, such as amides and acet- anilide~,~,' and dioxan and benzene for amh1es.39~96 Sorption rates of p-nitroaniline by Dowex 50, Amberlite IR-120 and Amberlyst 15 have been noted in a variety of solvent^.^,^ The trace components, naphthenic acids, alkylphenols, pyrrolic compounds and nitrogen bases have been separated from petroleum products by using macroporous ion-exchange resins in conjunction with dissolved gases in polar solvents as selective eluent~.~ Of the various solvents examined, methanol has been used the most.In admixture with amines and acetic acid useful separations and fractionations of weak acid systems, such as benzoic acid and hydroxybenzoic acid isomers ,lo brominated salicylanilides,lO chloro- phenolsll ,12 and peat bitumen,13 914 have been obtained. Methanol in admixture with methyl- ene chloride can be used for separating anionic and non-ionic detergentP and with aceto- nitrile for separating amides and acetanilide.3 Methanol alone has been used for separating amides3 and amines6,7 and for the more technological application of separating sulphoxides from petroleum fractions.16 In the early stage, namely the separation of phenolsJl1,l7 and of weak organic acids and basesJ1* *l9 the facility of methanolic and other alcoholic media for overcoming insolubility difficulties was clearly demonstrated.This is a principal reason for interest in the sorption by ion-exchange resins of organic materials from non-aqueous media. The separation of chlorophenols by using the acetate form of a strongly basic anion- exchange resin can be achieved either by graded elution with glacial acetic acid - methanol mixturesl1~l2 or by controlled pH of the eluting medium with triethylamine - acetic acid buffer solutions in methanol.11 These principles have been used in the present investigation to examine the anion-exchange separation possibilities of a wider range of phenols, namely nitrophenols, cresols and xylenols.This represents a series of weak acids with extremes of pK, values extending from 7.1 to 8.4 for the nitrophenols, 10.0 to 10.3 for the cresols and 10.2 to 10.6 for the xylenols. A wider range has been included in a preliminary examination. 0 SAC and the authors.1100 THOMAS AND THOMAS: ION EXCHANGE IN NON-AQUEOUS [Autalyst, Vol. 94 PRELIMINARY EXPERIMENTAL WORK To determine the conditions under which the various phenols could be sorbed from rnethanolic media by the acetate form of De-Acidite FF anion-exchange resin, a series of preliminary experiments was conducted with a resin (greater than 200 mesh) column of dimensions 1 cm diameter x 5-1 cm long. The effluent was monitored by ultraviolet absorp- tion at the wavelength of maximum absorption of the phenol.Table I summarises suitable mine - methanol mixtures for promoting sorption of the phenols; the pKa values of the phenols in aqueous solutions are also given. It should, however, be noted that the same pKa values do not hold for the media of this investigation, for example, the pKa values of cresols and xylenols are greater in methanol by about four units than they are in water,20 but with the order of pKa values being the same as in water,20 thus supporting the use of the more comprehensive aqueous pKa data. TABLE I SUITABLE AMINE - METHANOL MEDIA FOR ION-EXCHANGE SORPTION OF PHENOLS ON THE 5 ml of solution added to column contained 2.5 mg of phenol in each instance pK, of phenol .... .. 4.09* 7-15* 7*22* Per cent. v/v of amine in methanol . . 0. Is also held 0 0 ACETATE FORM OF DE-ACIDITE FF Phenol . . .. .. .. . . 2,4-dinitrophenol P-nitrophenol o -ni trophenol from 4 per cent. v/v of acetic acid in methanol. Phenol . . .. .. .. . . m-nitrophenol 2.4-dichlorophenol m-cresol Per cent. v/v of amine in methanol . . 4 0.2 6 Phenol . . .. .. .. .. p-cresol o-cresol pK, of phenol .. . . .. 8*39* 7.85* lo-lot (triethylamine) (triethylamine) (die thy lamine) pK, of phenol . . .. .. 10.28t 10-33t Per cent. v/v of amine in methanol . . 13 20 (diethylamine) (diet h ylamine) Phenol . . .. .. .. . . 3,5-xylenol 3,4-xylenol 2,S-xylenol pK, of phenol .... .. 10.201. 10.36t 10.40t Per cent. v[v of amine in methanol . . Phenol . . .. .. .. . . 2,3-xylenol 2,4-xylenol 2,4-xylenol pK, of phenol .. .. .. 10.54t l0-60t 10.62t Per cent. v/v of amine in methanol . . 25 per cent. of diethylamine permitted only moderate ion-exchange sorption of these three xylenols To be sorbed on the resin, these xylenols require even more basic conditions than the 25 per cent. of diethylamine in methanol * Reference 21. t Reference 20. Additional experiments with o- and 9-methoxyphenols (pKa = 9.98 and 10.21, respec- tively2l), vanillin (pKa = 7.4O2l), o-vanillin (pKa = 7.9l2l) and isovanillin (pKa = 8*8g2l) indicated that they should fall into the trend according to relative pKa. The mononitrophenols, cresols and 3,5-, 3,4- and 2,5-xylenols were subjected to fuller examination according to the following method.APPARATUS- Ion-exchange column-An ion-exchange column of 1 cm i.d. and of sufficient length to contain a 24-cm column of resin is set up for gradient elution under pressure from a cylinder of nitrogen as shown in Fig. 1. This is based on the design of Skelly12 but, in addition, is provided with a mercury back-pressure trap, T. Fraction collector-A Shandon fraction collector was used to collect the eluate fractions. Specfrophotometer-A Beckman DB ultraviolet - visual spectrophotometer was used to METHOD monitor the eluate fractions. REAGENTS- Acetic acid, glaciad-Analytical-reagent grade. De-acidite FF anion-exchange resin (chloride form)-The material has a mesh size greater than 200, 100 to 200, or 52 to 100 as appropriate.December, 19691 MEDIA: THE ANION-EXCHANGE BEHAVIOUR OF PHENOLS 1101 Sodium acetate solution, 10 per certt.w/v in de-ionised water-Prepare this from analytical- Methanol-Analytical-reagent grade. Diethylamifie-Double distil Hopkin and Williams general-purpose reagent grade at Triethylamifie-Double distil Hopkin and Williams general-purpose reagent grade at reagent grade sodium acetate. 55" c. 89.5" C. PREPARATION OF COLUMN AND GRADIENT ELUTION EQUIPMENT- method of Logiell and finally dried in a vacuum drying apparatus. prepared in methanol. De-Acidite FF resin, as supplied, is converted into the acetate form according to the A slurry of the resin is Nitrogen A, B and C = Ground-glass joints I = Chromatographic tube of I cm i.d.and of sufficient length to hold a 24-cm column of resin M = Mixing flask (250-ml capacity) R = Reservoir flask (I-litre capacity) S = Magnetic stirrer T = Mercury back-pressure trap Fig. 1. Gradient elution apparatus assembly For experiments with the mononitrophenols the column (Fig. 1) is filled with the methan- olic slurry and eluted with 4 per cent. (v/v) triethylamine in methanol, until equilibrium is reached. Before filling the column for experiments with the cresols and xylenols, excess of methanol is decanted from the methanolic slurry. The residue is treated with 20 per cent. v/v di- ethylamine in methanol and, to allow time for the contraction of the resin, is then allowed to stand for 15 minutes with occasional stirring.The column in Fig. 1 is then filled with the slurry, and eluted with 20 per cent. v/v diethylamine in methanol to equilibrium. The column is then charged with 6.0 ml of a solution of the phenol sample in the appro- priate solvent, as defined in Tables 11, I11 and IV, by using, in addition, three 5-ml portions of appropriate solvent to wash the sample into the resin. For experiments with the nitrophenols the mixing flask, M, is filled with 250 ml of 4 per cent. v/v triethylamine in methanol and, for experiments with the cresols and xylenols, with 20 per cent. v/v diethylamine in methanol. The reservoir flask, R, is filled with the appropriate addition eluent, as defined in Tables 11, I11 and IV, for the nitrophenols, cresols and xylenols, respectively. All ground-glass joints are secured and the magnetic stirrer started. Joint C is connected to the chromatographic tube, previously filled with 4 per cent.v/v triethylamine in methanol or 2.0 per cent. v/v diethylamine in methanol, as appropriate, care being taken to exclude air bubbles from tube BC. A slight nitrogen pressure is introduced, and when the eluent emerges from B the joint is made. The elu- tion system thus provides, for the ion exchange, a continuously increasing concentration of the solution in the reservoir flask. The nitrogen pressure is increased to give a flow-rate of about 1 ml minute-l.1102 THOMAS AND THOMAS: ION EXCHANGE IN NON-AQUEOUS [ABaZySt, VOL 94 The chromatogram is obtained by monitoring the 10-ml eluted fractions collected by ultraviolet absorption at the appropriate wavelength as enumerated in Tables 11, I11 and IV.3 Effluent fraction no. (10 ml each) Fig. 2. Separation of a mixture containing 1 mg each of M-, o- and p-nitrophenols on a De-Acidite FF (acetate form; 3 to 5 per cent. DVB; >200 mesh) resin column, elution details as described in Table 11: A, m-nitrophenol; B, o-nitrophenol; C, p-nitrophenol RESULTS A typical gradient elution curve for the nitrophenols is shown in Fig. 2, and Tables 11, 111 and IV summarise the gradient elution patterns obtained for the mononitrophenols, cresols and xylenols, respectively. For convenience in interpreting the tables, the phenols are shown in order of elution. A break in fraction number indicates distinct separation of the phenol from its nearest neighbous during elution, while the appearance of the same fraction number for two phenolic fractions indicates an incomplete separation, although smaller fractions can remedy this.TABLE I1 GRADIENT ELUTION PATTERN OF MONONITROPHENOLS FROM DE-ACIDITE FF COLUMNS (ACETATE FORM) WITH METHANOLIC MEDIA Load applied to column, 6-0 ml of a solution containing 1 mg each of o-, m- and$-nitrophenol in 4 per cent. v/v triethylamine in methanol. Eluent added to 4 per cent. v/v triethylamine in methanol for gradient elution, methanol containing 4 per cent. v/v of M triethylamine acetate in methanol together with 80 ml of M triethylamine in methanol and 10 ml of acetic acid 1-1 (9 ml of glacial acetic acid for those marked *). Size of fraction collected, 10 d.Elution monitored by ultraviolet absorption at 270 nm Fraction no. r \ L Resin cross-linkage m-nitro- o-nitro- p-nitro- (per cent. of DVB) Mesh size phenol phenol phenol 2-3 greater than 200 22-27 49-52 52-55 7-9 greater than 200 22-25 46-52 52-55 2-3 100-200 22-25 4 3 4 6 48-50 3-5* 100-200 23-26 4 6 4 9 49-53 7-9* 100-200 22-28 48-5 1 5 1-54 2-3* 52-100 20-23 46-50 50-53 3-5* 52-100 22-25 47-5 1 51-54 7-9* 52-100 23-26 47-51 51-55 3-6 greater than 200 22-25 4-9 52-54December, 19691 MEDIA : THE ANION-EXCHANGE BEHAVIOUR OF PHENOLS 1103 TABLE I11 GRADIENT ELUTION PATTERN OF CRESOLS FROM DE-ACIDITE FF COLUMNS (ACETATE FORM) Load applied to column, 6.0 ml of a solution containing 1 mg each of o-, m- and 9-cresol in 20 per cent. v/v diethylamine in methanol. Eluent added to 20 per cent. v/v diethylamine in methanol for gradient elution, methanol containing 4 per cent.v/v of M diethylamine acetate in methanol together with 30 per cent. v/v (20 per cent. v/v for those marked *) of M diethylamine in methanol. Size of fraction collected, 1Oml. Elution monitored by ultraviolet absorption at 280 nm WITH METHANOLIC MEDIA Resin cross-linkage (per cent. of DVB) 2-3 3-5 * 7-9* 2-3 3-b* 7-9 2-3 3-5 7-9 r Mesh size greater than 200 greater than 200 greater than 200 100-200 100-200 100-200 52-1 00 52-100 52-100 Fraction no. A o-cresol 9-cresol 20-22 22-24 26-29 29-3 1 22-25 25-27 24-26 26-28 24-26 26-28 & 25-29 26-29 22-27 28-31 31-32 1 m-cresol 26-28 35-37 33-37 29-31 30-33 30-32 32-34 3 1-33 29-32 TABLE IV (ACETATE FORM) WITH METHANOLIC MEDIA Load applied to column, 600ml of a solution containing 1 mg each of 2,5-, 3,4- and 3 , s xylenols in 20 per cent.v/v diethylamine in methanol. Eluent added to 20 per cent. v/v diethylamine in methanol for gradient elution, methanol containing 4 per cent. v/v of M diethylamine acetate in methanol together with 51.4 d of diethylamine litre-l. Size of fraction collected, 10ml. Elution monitored by ultraviolet absorption at 290 nm GRADIENT ELUTION PATTERN OF 2,5-, 3,4- AND 3,S-XYLENOLS FROM DE-ACIDITE FF COLUMNS Fraction no. Resin cross-linkage (per cent. of DVB) 2-3 3-5 7-9 2-3 3-5 7-9 2-3 3-5 7-9 f Mesh size greater than 200 greater than 200 greater than 200 100-200 100-200 100-200 52-100 52-100 52-100 ~ 2,5-xylenol 24-28 26-29 25-28 26-28 2P29* 23-26 23-26 27-29 23-26 3,4-xylenol 28-31 29-3 1 2 8-32 29-31 2&31* 26-29 26-29 26-29 28-31 3,5-xylenol 33-35 3 1-32 31-33 34-35 31-33 31-33 29-32 29-3 1 29-32 * There is poor definition of resolution here and it has been assumed that the absorb- ance (at 290 nm) peak (kom its position) is caused by 2,5-xylenol and that the shoulder is caused by 3,4-xylenol.This situation is best represented by an overlap of fraction numbers. DISCUSSION Separations with ion-exchange resins are dependent on an appropriate interplay of sorption and desorption. With regard to sorption, Webster, Wilson and Franksg mention that physical sorption, rather than the normal ion-exchange reaction, becomes the major factor when polar compounds are applied to resins in non-polar solvents such as hydrocarbons.These physically sorbed materials are regarded as those which are readily eluted with polar solvent^.^ This sorption from the non-polar liquid phase may be understood on the premise1 lU4 THOMAS AND THOMAS: ION EXCHANGE IN NON-AQUEOUS [A?ZU&St, VOI. 94 the polar constituents are attracted to the polar solid resin phase and associate with it in an analogous manner to the principle of “neutral sorption” noted for the high sorbabilities observed for certain inorganic species.’ Elution with polar solvents provides mobile liquid phases in which cations and anions should pair less strongly and provides for the normal ion- exchange reaction. The exact dividing line between the two modes of sorption is not easily defined; it must be dependent on a number of factors, such as solute, eluent and resin form.In the present context, the pattern of behaviour follows that in which substances that are of insufficient basic or acid strength to give sharp end-points during acid - base titrations in aqueous solution can be titrated successfully in a levelling solvent which is able to enhance their acidic or basic properties. Thus phenols, for example, can readily be expected to provide phenate ions in a basic amine-containing medium. It is not unreasonable, therefore, to expect the normal ion-exchange mechanism to be prominent for the phenate ions in the sorption by anion-exchange resins of phenols from amine-containing media, Thus, in a manner similar to that in which it is possible to determine the individual constituents in a mixture of acids (or bases) of different strengths by differentiating titrations carried out in solvents that do not exert a full levelling effect, it should be possible to exert some control over the presence of phenate ions in an ion-exchange system by appropriate adjustment of the basicity of the eluent.Such an effect would permit systematic design in the ion-exchange separations and is evident in the earlier work with phenols11p12 and other weakly acidic sys- t e m ~ ~ ~ J ~ p ~ ~ ; it is further illustrated by the results in TabIe I. In dealing with mixtures of phenols, the levelling effect to promote ionisation, or rather its upsetting, can have serious effects, as Logiell noted when passing $-chlorophenol through a column of De-Acidite FF (acetate form) from a solution containing 0.1 per cent.w/v of 2,4-dichlorophenol and 0.005 per cent. w/v of $-chlorophenol in a 0.2 per cent. w/v solution of triethylamine in methanol. The non-retention of the 9-chlorophenol was attributed to the release of acetate ion brought about by the sorption of 2,4-dichlorophenol giving an effective eluent of triethylamine acetate in methanol.ll In designing procedures, clearly margins have to be allowed for such reversals. One of the aims of the present study was to examine the possibility of extending the method to phenols of pKa greater than that of 9-chlorophenol (pK, = 9.38). The xylenols (pK, = about 10.5) represent the extreme end of the range examined and strongly basic con- ditions (Table I and IV) are necessary for them to be sorbed by the acetate form of De-Acidite FF.Therefore, a pKa of about 10.5 represents the limit to which the method of inducing ionisation by amines in methanol can be extended. However, it is possible that the physical sorption from non-polar media noted above might still be available as a concentration dimension for materials that cannot be isolated by normal ion exchange; the materials after desorption could then be resolved by other means. Of course, macroporous resins or resins that swell in non-polar media are required for efficient sorption in such circumstances. A study was also made of the separation of phenols of closely similar pKa. This had been encountered to a limited degree by Logiell in the separation of 2,4- and 2,6-dichloro- phenols with pKa values in water22 of 7.85 and 6.79, respectively, a difference of about 1.The pKa difference of 0.1 between o-nitrophenol and P-nitrophenol is considerably lower and presents an even larger separation challenge, which is repeated and magnified by the cresols and xylenols. When the pKa values of pairs or groups of phenols are similar, fixed-concentration elution requires a great deal of experimentation, thus gradient elution is more successful.12 However, to find the pH at which the difference in the degree of dissociation of a pair of phenols is greatest, Logiell resorted to controlling the pH of the eluting medium by using triethylamine - acetic acid buffer solutions as eluents for separating the dichlorophenols. As the pK, values of many of the phenols in this investigation were so close, the much more intricate separation problem was solved by a rigorous combination of amine - acetic acid buffer solutions in methanol and gradient elution (Tables 11, I11 and 11’).Triethyl- amine is a suitable amine base for providing appropriate buffer solutions for separating the mononitrophenols (Table IT and Fig. 2), but for the higher pKa cresols and xylenols, the more basic diethylamine must be used (Tables 111 and IV). Variation of the degree of cross-linking of resins generally has little effect on the separation characteristics, but the quality of the separations is usually superior with resins of mesh size greater than 200.December, 19691 MEDIA : THE ANION-EXCHANGE BEHAVIOUR OF PHENOLS 1106 CONCLUSION Quantitative knowledge of polar effects of substituent atoms and groups in organic molecules derived from work on the dissociation constants of organic acids and bases has contributed materially towards an understanding of the fundamentals of many aspects’ of chemistry and has provided the means for accurate forecasting.This is again seen to be the case in the present context where such knowledge, as represented by pK, values, is shown to be a suitable criterion for the preliminary selection of optimum pH required in the non- aqueous ion exchange of phenols as, indeed, it can also be for the ion-exchange separation of other weak acid systems. The gradient elution separation technique described is generally applicable to phenols of closely similar pK, values but, when the pK, values are close, precise experimentation is required to strike the correct balance of medium pH for gradient elution of effective resolution.The Ministry of Social Security is thanked for a research award (to D.E.T.) in aid of this work. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. REFERENCES Moody, G. J., and Thomas, J. D. R., Analyst, 1968,93, 667. Osborn, G. H., Ibid., 1953,78, 220. Cassidy, J. E., and Streuli, C. A., Analytica Chim. Actu, 1964, 31, 86. Gordon, J. E., J . Chromat., 1965, 18, 642. Malarski, Z., and Sobezyk, L., Roczn. Chem., 1963, 37, 871. Gordon, J. E., J . Phys. Chem., 1963, 67, 16. Pietrzyk, D. J., Talunta, 1966, 13, 209. -- , Ibid., 1966, 13, 225. Webkter, P. V., Wilson, J. N., and Franks, M. C., Andytica Chim. Acta, 1967,38, 193. Skelly, N. E., and Crummett, W. B., Anulyt. Chem., 1963,35, 1680. Logie, D., Analyst, 1957, 82, 563. Skelly, N. E., Analyt. Chem., 1961,33, 271. Thomas, J. D. R., Nature, 1962, 193, 975. -- , J . Appl. Chem., 1962, 12, 289. Czoj;, D., and Awerbuch, F., Thruszcze i Srodki Piorace, 1964, 8, 31. Okuno, Z., Latham, D. R., and Haines, W. E., Analyt. Chem., 1967,39, 1830. Anderson, R. E., and Hansen, R. D., Ind. Engng Chem., 1955,47, 71. Carroll, K. K., Nature, 1955, 176, 398. Shelley, R. N., and Umbergo, C. J., Analyt. Chem., 1959,31, 693. Rochester, C. H., Trans. Faraday SOC.. 1966, 62, 356. Korturn, G., Vogel, W., and Andrussow, K., “Dissociation Constants of Organic Acids in Aqwous Solution,” International Union of Pure and Applied Chemistry, Butterworth, London, 1961. Barlin, G. B., and Perrin, D. D., (2. Rev. Chem. SOC., 1966, 20, 76. Received December 2.3~6, 1’968 Accepted July lst, 1969

ISSN:0003-2654    

DOI:10.1039/AN9699401099

出版商:RSC    

年代:1969

数据来源: RSC