摘要:

NOVEMBER, 1967 THE ANALYST Voi. 92, No. I100 Precipitation from Homogeneous Solution A Review’;’ BY P. F, S . CARTWRIGHT (TJie illeta1 Box Company, Acton, London, W.3) E. J. NEWMAN AND D. W. WILSON (Department of Geology, Imperial College, London, S. W.7) (Department of Chemistry, Sir John Cass College, Jewry Street, London, E.C.3) SUMMARY OF CONTENTS Introduction Kale of nucleation in precipitate formation Methods of precipitation Increase in pH Anion rclease Cation release Reagent synthesis Precipitation from mixed solvcnts Valency change Photochemical action Main group elements Transition elements ,4pplications t o gravimetric analysis Co-precipitation and fractional precipitation Industrial applications Miscellaneous applications Conclusions GRAVIMETRIC and titrimetric methods form the basis of classical analytical chemistry.Titrimetry is still very much to the fore, particularly since its scope was extended by the introduction of complexometric methods, but there has been a rapid decline in recent years in gravimetry. New and rapid instrumental methods have, in many instances] superseded the time-consuming gravimetric procedures, with their inherent problems involving pre- cipitates that are, to varying extents, difficult to filter and contaminated by adsorbed and occluded impurities. Nevertheless, there are still occasions when a gravimetric determination is the most useful method of analysis. This is particularly true for single determinations, as opposed to routine repetition, or when a. reference method is required to evaluate a new technique, and soundly based, simple, gravimetric methods may be needed for these purposes for many years to come.The conventional ways of improving the particle size and purity of gravimetric precipi- tates are well known. I t is the usual practice to work with dilute solutions and to add the precipitant slowly, with stirring. When possible, precipitation is carried out in hot solution, and precipitates are usually allowed to “digest” to reduce co-precipitation and occlusion of impurities] and to develop more readily filterable crystals. The economic necessities of a determination, however, impose limits on the effective use of these practices, and even with hot dilute solutions and rapid stirring it is impossible to avoid high concentrations in the locality of the added precipitant.For details see Summaries in advertisement pages. * Reprints of this paper will be available shortly. 663664 CARTWRIGHT, NEWMAN AND WILSON [Analyst, Vol. 92 If these local concentrations could be avoided during the course of precipitation, many of the disadvantages of gravimetry would be overcome. Precipitates would be denser and more easily filtered, occlusion of impurities would be reduced, and larger amounts of material could be handled with improved separation from potentially interfering ions. In many instances this is achieved by the method that has come to be known as precipitation from homogeneous solution, whereby the precipitate is formed at a slow and controlled rate by a uniform change of conditions in an initially homogeneous solution.I t is not altogether unexpected that a method that overcomes a group of difficulties in a striking manner will, itself, introduce other disadvantages that can impose limitations on its use. These are described and discussed under Conclusions. The idea of controlling precipitation to obtain a better product is not new, and much work has been done in the past, particularly in attempting to improve the precipitation and separation of aluminium. Thus Stock1 used a mixture of solutions of potassium iodide and iodate to remove the acid formed by the hydrolysis of aluminium chloride in the precipitation of hydrated aluminium oxide, while Chancel2 used sodium thiosulphate solution for the same purpose. A further method of reducing acidity, to precipitate the 8-hydroxyquinoline complex of aluminium by a bromate - bromide - thiosulphate reaction, was described by Smith,215 who used it to separate milligram amounts of aluminium from gram amounts of magnesium." Dorrington and \Vard3 used potassium cyanate, which, in solution, is hydrolysed to urea and thence to ammonia and carbon dioxide, in an attempt to improve the precipitation of aluminium, chromium and iron.The chromium and iron precipitates were granular, but adhered strongly to the walls of the beaker; the aluminium precipitate was not granular. who, in 1937, described the precipitation of basic aluminium sulphate by a method in wliicli urea was added to the solution of aluminium to yield animonia at a slow, controllable rate by hydrolysis during boiling.The technique of precipitation from homogeneous solution might be considered to date from this time, and it is the progress made since then that forms the main part of this review. Many new methods have been evolved since the early work of Willard and Tang, and an excellent account of some of these methods has been given by Gordon, Salutsky and \$511ardG in the only textbook that has so far been published dealing exclusively with the subject. Various reviews have, froin time to time, dealt with progress in the field.7 $8 $9 ,lo ,11 Two advanced textbooks of analytical chemistry contain comprehensive reviews of the subject,12 9 1 3 and at least one textbook of practical quantitative analysis includes working met110ds.l~ A useful review dealing with the precipitation of metal chelate compounds from homogeneous solution has been published by Firsching.15 In the present review, consideration is given first to the investigation of the rBle of nucleation in precipitate formation, with particular reference to precipitation from homo- geneous solution.The methods of precipitation are then reviewed under six broad classes adopted for the purpose of convenience. In describing the applications to gravimetric analysis, the elements have been divided into main group and transition metals; the latter include copper, silver, gold, zinc, cadmium and mercury. A section is devoted to co-precipita- tion and fractional precipitation, and finally, a short review of industrial applications is given. A significant advance was made by Willard and KoLE OF NUCLEATION IN PRECIPITATE FORMATION The experiments carried out by Von WeimarnlG led to the general conclusion that, for slightly soluble precipitates, the concentration of reagents determined the rate of precipitation and the size of the particles.Particle size increases with increasing solubility and with decreasing concentration of reagents. Although the theory is incomplete, and there are many examples of apparently anomalous behaviour, nevertheless Von Weimarn's observations still furnish the best practical guide to the conditions for the formation of the most satis- factory precipitates . It is now generally agreed that the formation of precipitates involves two processes: the formation and the growth of nuclei. These aspects have been discussed by O'Rourke and Johnson1' in a paper dealing with the mechanism of precipitate formation, and by Klein and Gordonl8 in a general review of nucleation in analytical chemistry.There is disagreement, however, on the steps leading to the formation of the nucleus and on tlie number of ion pairs required for nucleus formation. Turnbullls has concluded that the rate of nucleation is * This was drawn t o our attcntion when the paper was in proof.November, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 665 highly dependent on the extent of supersaturation, that the size of the nucleus is a function of the degree of supersaturation and that the number of ion pairs in a nucleus is of the order of tens for the smallest nuclei and probably hundreds for most nuclei.Christiansen and Nielsen20 and Nielsen,21 however, have developed another approach based on the determination of the induction period, or the time elapsing between the mixing of the reagents and tlie appearance of a precipitate. They conclude that there is a definite critical size for each substance, and that this critical nucleus consists of only a few ion pairs. More recently Klein and Gordon18 and Gordon, Klein and Walnut22 have studied nucleation in analytical chemistry, with particular reference to the precipitation of silver chloride from homogeneous solution, and have concluded that tlie nucleus consists of about 5 ions, thus supporting the Christiansen - Nielsen theory. Fischer23 has developed a theory of nucleation that allows for the induction effects caused by impurities.He considers that the number of available sites, and their distribution with respect to effectiveness, may vary widely with conditions. For crystallisation to start, the solution inust be critically supersaturated with respect to the site, and, during subsequent growth, the site may either become entrained in the crystal or be separated from it. The function of the sites is to accumulate ions from the surrounding solution, ions of one charge attracting those of opposite charge to form groups of lattice ions. Fischer considers that this solid particle represents the type of nucleus envisaged by Turnbull, while the grouping of lattice ions on the site, but not the site itself, is considered to be the critical nucleus of Christiansen and Nielsen.According to Fischer’s theory, nucleation sites or induced nuclea- tion are virtually necessary. Nielsen,24 however, has concluded from studies of tlie pre- cipitation of barium sulphate from homogeneous solutions that in dilute solutions particles start their growth on foreign nuclei, but that at higher concentrations most of the crystals originate from liomogeneously formed nuclei. Fischer has studied the origin of nuclei in precipitation reactions23 and the effect of the form of reagent on particle size of precipitate^.^^ He has expressed doubt that nucleation in precipitation from homogeneous solutions is strictly a homogeneous reaction and is of the opinion that the process is one of direct mixing.26 This view has been challenged by Haber- man and G0rdon,~7 who admit the possibility of direct mixing, but consider that this is not the usual situation.Fischer,28 in reply, has agreed that nucleation in precipitation from homo- geneous solution need not necessarily be a process of direct mixing but considers that it frequently must be. Kleiri and studied nucleation in the precipitation of strontium sulpliate from homogeneous solution by using an electronic technique for counting the number of precipitate particles. Their results supported Nielsen’s theory and the rate of homogeneous nucleation indicated that the critical nucleus consisted of 52 ions. Mealor and Townshend have studied nucleation in barium ~ u l p h a t e , ~ ~ and a series of sparingly soluble salts,31 by a particle-counting method. They found that heteronucleation always occurred, as well as homogeneous nucleation.The onset of heteronucleation varies markedly according to the purity of reagent solutions and their previous history, and the rate of heteronucleation increases only slowly witli increasing precipitant concentrations. Homogeneous nucleation commences only at a critical concentration but, thereafter, increases rapidly with concentration. The authors found, therefore, that plots of particle count against concentration showed sudden breaks caused by the commencement of homogeneous nuclea- tion; these findings are also in accordance with Nielsen’s views. Recently, Thompson and Gordon3* applied the drop method to the study of nucleation of nickel dimethylglyoximate, precipitated from a homogeneous solution containing nickel ions, biacetyl monoxime and hydroxylammonium chloride.Hileman33 used the same method to study nucleation of several nickel and palladium oximates precipitated from homogeneous solution by similar reactions. In this method, the solution is dispersed in a suitable oil in the form of fine drops, and it is supposed that heteronuclei will be confined to some, but not all, of the droplets. The drops were viewed microscopically and those in which lietero- nucleation occurred were distinguished from those in which homogeneous nucleation took place by the different times at which precipitates appeared. Knowing the kinetics of forma- tion of the chelating precipitant, it was possible to calculate tlie concentrations of the metal complexes present before precipitation was seen to commence.The authors of both papers666 CARTWRIGHT, NEWMAN AND WILSON [Analyst, VOl. 92 found that supersaturations of about 1000 times or more of the equilibrium solubilities of the metal complexes were attained before precipitation commenced. The matter remains open to discussion and experiment, and it is likely that more work will be done in this field. I t certainly seems that nucleation sites, particularly on the walls of the vessels used, cannot be ignored, and it is remarkable how many instances have been recorded of precipitation occurring where there are imperfections on the vessel walls and how many of the precipitates formed by the homogeneous process cling tenaciously to the sides of the vessel. Nevertheless, whatever differences may exist between the mechanisms of precipitation from homogeneous solution and of conventional precipitation, there is much evidence available to show that the results of the former technique are usually superior to those of classical methods. I t is not yet known whether this can be attributed entirely to the avoidance of locally high concentrations of the precipitant in the solution.METHODS OF PRECIPITATION INCREASE I N PH- This is the method originally developed by Willard and Tang for the precipitation of basic aluminium salts, as a result of which many of the recognised principles of precipitation from homogeneous solution were established. Precipitation is made to occur by the con- trolled increase of the pH of an initially unsaturated homogeneous solution, usually by the hydrolysis of a suitable substance to produce ammonia iu sitzt.Of the substances available, urea is undoubtedly the most useful and is almost universally used, being cheap and soluble in water, and the solution decomposes on heating to give ammonia and carbon dioxide. The rate of hydrolysis is dependent on the temperature, and can be readily controlled to give the desired rate of precipitation. During the course of their work Willard and Tang first noted the need for the presence of a suitable anion to obtain dense precipitates. The reason for this does not appear to be clearly understood, but it has been suggested6: that the anion may serve as a buffer and control the rate of change of pH; that it may be incorporated in the precipitate forming a basic salt; and that it may reduce the concentration of the cation by complex formation.Although buffer action may play some part, it is certainly not the sole reason for the formation of dense precipitates; for example, Willard and Gordon found that in the precipitation of thorium by the urea method, formate will give dense precipitates but that acetate does not give the desired results.34 During the formation of basic salts some of the “suitable anion” is usually included in the precipitates which are, however, often of variable composition, but nevertheless appear to be uniformly dense. Some evidence of complex formation has been reported, as in the precipitation of basic bismuth f ~ r m a t e ~ ~ in which dense precipitates are obtained.The question is still unanswered, and there is need for further research to determine the conditions that control the type of precipitate formed. Nevertheless, when a suitable anion is present the precipitates obtained by this method are generally dense and readily filterable, but frequently adhere to the walls of the retaining vessel. It is essential in all precipitations from homogeneous solution to work with clean, unscratched beakers. ANION RELEASE- Precipitation by anion release bears certain resemblances to the method of urea hydrolysis, in that the precipitating anions are generated slowly in solutions containing the metal ions to be precipitated, by the controlled hydrolysis of suitable compounds under the conditions necessary for the formation of insoluble precipitates.The nietliod is somewhat restricted by the availability of suitable compounds, e.g., esters and amides, which should preferably, but not necessarily, be water-soluble, and which must hydrolyse at a satisfactory rate in solution. Nevertheless several compounds have been found that yield the more common precipitating anions such as oxalate, phosphate, sulphide and sulphate, and these have been successfully applied to the precipitation of many metals. Great interest has been shown in the precipitation of sulphides, particularly with thio- acetamide and thiourea; the hydrolysis of thioacetamide has been studied in considerable detail36337 938 and Broad and Barnard39 have compiled a bibliography containing 87 references of the use of thioacetamide.I t is, of course, much more pleasant to use for the precipitation of sulphides than hydrogen sulphide, and the usual advantages of precipitation from homogeneousNovember, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 667 solution are obtained. However, it is important to realise that the use of thioacetamide cannot always be substituted for a hydrogen sulphide precipitation. I t is necessary to conform strictly to the conditions derived for its use and consideration should be given to the possibilities of thioacetamide itself combining with metal ions, behaving as a reducing agent or providing buffering action through its decomposition to acetate ions. Thiourea has been used for the precipitation of copper,40 thioformamide proposed for the precipitation of s ~ l p h i d e s , ~ ~ and an apparatus described for the precipitation of sulphides and thiocarbonates with potassium thi~cai-bonate.~~ CATION RELEA4SE- Precipitation from homogeneous solution can be brought about by the controlled release The method of cations from soluble complexes in solutions containing the desired anions.may be conveniently subdivided into- Cation release at constant pH. Cation release by pH change. Cation release by replacement. Cation release at constaizt PH-The method appears to have been developed by MacNevin and D ~ n t o n ~ ~ for the precipitation of hydrated iron oxide. The iron(II1) - EDTA complex was formed and, after thc pH had been adjusted to the required value, the complex was slowly destroyed by boiling with hydrogen peroxide to produce a dense precipitate.Sub- sequent studies44 have shown that the general applicability of the niethod is dependent on complex stability, the degree of insolubility of the precipitate and the possible catalytic breakdown of the hydrogen peroxide by the precipitate particles. The method is not par- ticularly selective as all of the metals present are released into solution, when their complexes are destroyed. Moreover, it is not possible to test for completeness of precipitation, because unprecipitated metal is still present in complexed form. A different method has been reported by Dams and Hoste, who precipitated tungsten by the thermal decomposition of peroxy- tungstate from a nitric acid - hydrogen peroxide solution.45 Other methods may u7ell be proposed for specific reactions, and there is need for further research into the use of other complexing reagents and the use of masking agents for effecting separations of interfering metals.Cation release by pH change-The stabilities of metal complexes are dependent on pH, and precipitation may be brought about by lowering the pH to the point at which the metal complexes begin to break down. The pH of a solution may be decreased by the hydrolysis of a suitable ester to produce an acid, or bv the volatilisation of ammonia from a boiling alkaline solution. Thus ammonium persulphate has been used to precipitate barium from its EDTA complex,46 2-hydroxyethyl acetate to precipitate silver from a silver - ammonia complex,47 and volatilisation used for the precipitation of silver, again from a silver - ammonia complex.48 The method is somewhat limited by the fact that complex stability does not usually decrease sharply at any given pH, but rather over a pH range.Nevertheless, some separations of metals may be achieved by careful pH control; precipitates are usually dense and readily filterable. Cation reZense b)- 1.e~5lcrce~mnt-ITse can be made of the different stabilities of cation complexes to effect precipitation from honiogcneous solution. The metal to be precipitated is first coniplexed, and a solution of another metal that forms a more stable complex is added, usually in a slow dropwise manner, with stirring. The metal forming the less stable complex is slowly released into the solution and is precipitated. For example, barium sulpliate has been precipitated by adding a nickel solution to a solution containing the barium - EDTA complex and sulphate i0ns.4~ REAGENT SYNTHESIS- Increasing attention has been directed in recent years towards the application of methods in which an organic chelating agent is synthesised in solution in the presence of metal ions with which it combines to form a precipitate.The first direct synthesis of a chelating agent in solution to achieve precipitation from homogeneous solution was reported by Tara~evich,~~ who reacted nitrous acid with o-phenyl- enediamine in aqueous solution in the presence of either copper(I1) or silver ions. In this way 1H-benzotriazole was synthesised at a controlled rate and precipitated the insoluble metal compounds.668 CARTWRIGHT, NEWMAN AND WILSON : [AnaZyst, Vol.92 Many other interesting and useful methods have since appeared in the literature, of which the most outstanding are probably the synthesis of dimethylglyoxime in the presence of nickel ions,51 the synthesis of 1-nitroso-2-naphthol in a cobalt solution to produce cobalt nitrosonaphtholate, which can be weighed directly,52 and the precipitation of metal 8-hydroxy- quinolinates via the hydrolysis of 8-acetoxyquinoline to 8-hydroxyq~inoline.~~ These and the several other reported examples of reagent synthesis are dealt with in later sections of this review. PRECIPITA4TION FROM MIXED SOLVENTS- Howick and Jones have described the precipitation of several metal 8-hydroxyquinol- inates from homogeneous solution by a technique to which they refer as precipitation from mixed solvents, and which was first applied to the precipitation of aluminium and mag- n e ~ i u m .~ ~ In this method the original solution is made so as to contain all of the reagents, together with a sufficient amount of a miscible organic solvent that is more volatile than water (usually acetone) to dissolve the metal chelate and so prevent precipitation. The solution is then heated to volatilise the organic solvent preferentially and so cause precipitation. VALENCV CHANGE- The controlled change of the oxidation state of an element has frequently been used to bring about precipitation from homogeneous solution. For example, cerium( 111) iodate in nitric acid solution is oxidised slowly with bromate or persulphate to produce a dense pre- cipitate of ceriuin(1V) iodate, and this has been used to separate cerium from other rare Precipitation of lead chromate from homogeneous solution has been achieved by the slow oxidation of chromium(II1) ions with bromate in a buffered solution, and this forms the basis of a method for the determination of either lead or chromium.56 The reduction of periodate to iodate with ethylene glycol, produced by the slow hydrolysis of 2-hydroxyethyl acetate, has been used, for example, in the precipitation of thorium iodate from homogeneous soluti0n.~7 The reduction of copper(I1) to the copper( I) state with hydroxylammonium chloride occurs slowly in acidic solutions, and this has been applied to the precipitation of copper( I) thiocyanate from homogeneous solution.58 PHOTOCHEMICAL ACTION- Photochemical reactions have been used to alter the conditions of a homogeneous solution to effect precipitation, and the method is interesting in that the alteration can be independent of temperature.Yen and Yang59 generated bromine photochemically to destroy oxalate used as a masking ion in the precipitation of tantalum selenite. Das, Heyn and Hoffman6* used the method more directly in the precipitation of thorium iodate by photochemical reduction of periodate. APPLICATIONS TO GRAVIMETRIC ANALYSIS MAIN GROUP ELEMENTS GROUP 11- A good deal of attention has been given to the precipitation o f certain elements of this group, especially in thc field of co-precipitation; this aspect is, however, dealt with under a separate heading.Beryllium phosphate has been precipitated from an initially acidic solution by using the hydrolysis of urea to raise the pH and thus cause precipitation,'jl the precipitate being dried, ignited and weighed as beryllium pyrophosphate; tlie method was applied to the analysis of beryl. Precipitation of beryllium in the presence of sulphate by hydrolysis of urea has been described.G2 The precipitate, which consisted of fine, amorphous, easily filterable granules containing sulphate that was removed on ignition, was ignited at 1000" C and weighed as beryllium oxide. The method was superior to the conventional ammonia precipitation in giving a precipitate much easier to manipulate and much less contaminated by co-precipita- tion of other metals. The precipitation of magnesium oxalate by the hydrolysis of diethvl oxalate was studied by Gordon and C a l e ~ .~ ~ It was found necessary, however, to carry out tlie reaction in acetic acid solution containing ammonium acetate as a buffer. The precipitate was not weighed directly because of occluded solvent, but the magnesium content was determined by titratingNovember, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 669 the oxalate with permanganate solution. Precipitation of magnesium 8-hydroxyquinolinate has been induced by the hydrolysis of urea,64 by the hydrolysis of 8-acetoxyquinoline to generate the precipitant in s i t ~ 4 , ~ ~ and by precipitation from mixed solvents by using an acetone - water system.66 Calcium, as oxalate, has been precipitated by urea hydrolysi~,~~ and by hydrolysis of dimetliyl oxalate.68 The hydrolysis of diethyl oxalate in 85 per cent.acetic acid solution has been used for the determination of alkaline-earth metals in silicates.69 Bashar and Townshend have made a comparative study of methods for the precipitation of calcium oxalate from homogeneous solution.70 They showed that to precipitate up to 200 mg of calcium in the presence of up to 100 mg of magnesium the urea and diethyl oxalate methods were best, and that if larger amounts of magnesium are present the urea method is preferable because double precipitation is simpler. The precipitation of calcium fluoride by the slow release of calcium irom its EDTA complex has been reported.71 However, according to Morales and W e ~ t , ~ 2 this gives a gelatinous, hydrated precipitate, and these authors preferred to precipitate calcium fluoride by hydrolysis of the tetrafluoroborate ion.This hydrolysis was accomplished by raising the pH by urea hydrolysis in the presence of ammonium chloride to produce buffer conditions. Up to 40 mg of barium could be tolerated in the precipitation of 200 mg of calcium. Elving and Van Atta73 have described the precipitation of strontium, as sulphate, and the separation of barium, strontium and calcium by the hydrolysis of dimethyl sulphate; it was necessary to use methanol solutions. Strontium has also been separated from calcium by the selective precipitation of strontium sulphate ; strontium was replaced from its EDTA complex by the addition of magnesium sulphate solution under alkaline conditions.74 Gordon and F i r ~ c h i n g ~ ~ precipitated barium chromate by urea hydrolysis, and Norwj t z applied the method involving precipitation of barium chromate to the assay of barium compounds.76 The hydrolysis of sulphamic acid for the precipitation of barium sulphate was described by Wagner and W ~ e l l n e r ~ ~ ; strontium was found to interfere with the determination.Two displacement methods have been reported for the precipitation of barium. In the first Firscl~ing~~ displaced barium from its EDTA complex by adding magnesium chloride solution to precipitate barium chromate and separate barium from relatively large amounts of stron- tium and lead. In the second the same author precipitated barium sulphate by the dis- placement of barium from EDTA and tmns-l,2-diaminocyclohexane-NNN"'-tetra-acetic acid (DCTA) complexes.79 Several methods have been proposed for the precipitation of barium. GROUP 111- As mentioned earlier, the precipitation of basic aluminium sulphate was studied by Willard and Tang.4$5 The method of urea hydrolysis has also been applied to the precipitation of aluminium as the 8-hydro~yquinolinate~~ and as phosphate.81 An alternative method of precipitating aluminium as the 8-hydroxyquinolinate by the hydrolysis of 8-acetoxyquinoline has been studied by Marac, Salesin and Gordon,82 and, independently, by Howick and TriggS3; Howick and Jones have used a method involving volatilisation of acetone from an acetone - water solution containing aluminium and 8-hydroxyquinoline for the same purpose.54@ Willard and T;0ggs5 have applied the method of urea hydrolysis to the determination of gallium as basic sulphate; dense precipitates were obtained. Jones, Hileman, Townshend and Gordon were unable to apply successfully the 8-acetoxy- quinoline method to the precipitation of gallium and indium,86 but were able to develop a method for the determination of indium as the 8-hydroxyquinaldate, in which the reagent was generated by hydrolysis of 8-a~etoxyquinaldine.~7 GROUP IV- The precipitation of tin.as basic tin(1V) sulphate by urea hydrolysis was reported by Willard and Gordon.88 Dense precipitates were obtained by carefully controlling the acidity of the solution, but extensive co-precipitation occurred, and the authors recommended the precipitate as being more suitable for use as a carrier than as a method for the determination of tin.Flaschka and Jakobljevichsg have developed a method for precipitating tin as sulphide by the hydrolysis of thioacetamide.670 CARTWRIGHT, NEWMAN AND WILSON : [Analyst, Vol. 92 Thioacetamide has also been used for the precipitation of lead sulphide from acid solu- tiongo; the precipitate was weighed after drying at 110" C. Lead phosphate has been pre- cipitated by urea hydroly~is,~~ but the presence of alkali and ammonium salts generally led to low results. Lead has been determined as sulpliate by the hydrolysis of dimetliyl sulphate in methanol solutiong2 and by the use of sulphamic a~id.~3994 This method was used by Hoffman and Brandt,56 who obtained large, easily handled crystals by generating chromate ions by the oxidation of chromium (111) with bromate in the presence of acetate.GROUP V- The precipitation of arsenic sulphide by the hydrolysis of thioacetamide has received considerable attention, first by Flaschka and Jaltoblje~ich,~~ and later by Butler and Swift The latter studied the rates and mechanism of precipitation of arsenic (111) and arsenic (V) in acidic solution. Thioformamide has also been used for the precipitation of arsenic s ~ l p h i d e , ~ ~ and in the separation of arsenic, antimony and tin.g8 Antimony sulphide has been precipitated in a readily filterable form by the hydrolysis of thioacetamideg9; the precipitate can either be dried and weighed, or dissolved in acid and titrated with bromate.Tliiocarbonic acid has been used for the precipitation of antimonyloO and thioformamide for antimony and tin.101 Flaschka and Kakobljevich studied the precipitation of bismuth sulphide by thioacetamide,lo2 and the same method was used by Stoner and Finstone for the separation of bismuth and uranium.lo3 The technique of urea hydrolysis has been used for the precipitation of basic bismuth formate and the separation of bismuth from lead35; dense precipitates were obtained, but required to be ignited to oxide before weighing. Some evidence was found for complex formation involving bismuth and formate ions. Bismuth phosphate has been precipitated by the hydrolysis of metapliosphoric acid,lo4 and bv cation release at constant pH from a bismuth - EDTA complex in which hydrogen peroxide was used to liberate the Cation release has also been used by Pribil and Ctita for the Precipitation of bismuth hydroxidelo5 by displacing bismuth from its EDTA complex, in ammoniacal solution, by the addition of calcium.One of the best ways of precipitating lead is undoubtedly to form the chromate. Several methods have been proposed for the determination of bismuth. TRANSITION ELEMENTS GROUP I- The precipitation of copper sulphide by thioacetamide has been studied,lo6 and the method developed applied to the quantitative separation of copper, cadmium and zinc.lo7 Thioacetamide has also been used to precipitate copper from a tartrate solution of copper and alurniniumlo8; the copper in the precipitate was determined by titration with EDTA.Copper sulphide has also been precipitated witli thioformamide in acidic solution10g ; the precipitates were dissolved in nitric acid, and copper was determined by titration with potassium cyanide or iodimetrically. Thioformamide has been used for the separation oi copper and arsenic,l1° and the use of trithiocarbonic acid to precipitate copper sulphide has been reported.lll Copper and silver have been precipitated with lH-benzotriazole synthesised i i i sitzi by reacting nitrous acid and o-phenylenediamine.50 Copper has also been precipitated from Iiomogeneous solution with salicylaldoxime (produced by reacting salicylaldehyde and hydr- oxy1aminc1l2) , with cupferron (produced by the reaction between phcnylhydroxylamine and nitrite113) and witli N-benzoyl-X-phenylhydroxylamine (produced by the Iijdrolysis of &V- benzoyl-N-phenylhydroxylammonium acetate114).I t has also been precipitated as the benzoin-x-oxime complex by pH reduction via volatilisation of ammonia15 and as the 8-hyd- roxyquinolinate via volatilisation of ammonia15 and of organic solvent .115 Copper( I) tetra- phenylborate has been precipitated by delayed reduction of copper( 11) ions with ascorbic acid in acidic solution,l16 and precipitation of copper( I I) 8-hydroxyquinaldate has been achieved by hydrolysis of tS-acet~xyquinaldine.~~~ Precipitation of the copper - cupferron complex from homogeneous solution, depending on the synthesis of cupferron in situ, has been des- cribed118 ; the reagent was synt hesised from phenylhydroxylamine and sodium nitrite in ice- cold dilute acetic acid solution containing copper. The precipitate was pale blue and could be filtered off over a sintered-glass crucible of medium porosity, dried at 90" C and weighed directly as copper(I1) - cupferron complex.Those familiar with the conventional methodNovember, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 671 cannot fail to be impressed by this example of what can be achieved by precipitation from homogeneous solution. Moreover, the gravimetric factor is obviously more favourable than when ignition of the precipitate is necessary, and the method is less subject to interferences than the conventional one. Unfortunately, however, it requires the use of freshly prepared phenylhydroxylamine, and it is interesting to note that the authors state that no reaction occurred between phenylhydroxylamine and sodium nitrite in the absence of copper.An excellent method for the precipitation of copper as copper( I) thiocyanate has been described by N e ~ m a n . ~ ~ Hair and Newman described the determination of traces of iron and lead in copper metal and copper(I1) salts after separation of tlie copper by precipitation of copper(1) thiocyanate from homogeneous s01ution.l~~ Silver has been precipitated as silver chloride by cation release from a silver - ammonia complex involving a decrease in pH caused by the hydrolysis of 2-hydroxyethyl a~etate,~7 and as silver halide by cation release caused by volatilisation of ammonia from a silver - ammonia c0rnplex.~8 The volatilisation method has also been applied to the precipitation of silver phosphate.120 GROUP 11- Zinc has been precipitated as oxalate by the hydrolysis of diethyl oxalate,l*l but the authors concluded that the method was of limited use because of interference by several ions.Studies of the precipitation of zinc sulphide have been made by Flaschka,122 Rowersox, Smith and Swift,123 Klein and Swift123 and Hahn and P ~ i n g l e , ~ ~ ~ with thioacetamide as a source of sulphide ions. Ainin126 has also used thioacetamide to precipitate zinc sulphide froin alkaline solution, but determined zinc in the precipitate by titration with EDTX. The use of trithio- carbonic acid in place of thioacetamide has been described,127 and the Separation of zinc from manganese, iron and alkaline earths has been reported.12* Tlie use of thiourea as a source of sulphide ions has also been studied.129 Zinc 8-hydroxyquinolinate has been precipitated by tl ie hydrolysis of 8-acetoxy- q ~ i n o l i n e l ~ ~ and by the ammonia volatilisation method.131 Zinc 8-liydI.oxyquinaldate has been precipitated by the hydrolysis of 8-acetoxvquinaldine.13* Tlie sulphides of the remaining members of the group have been precipitated with tliio- acetamide.The method for cadmium was developed by 1;lasclika and .Jakobljevich,133 and the kinetics of the reaction have been investigated by Bowersox and Swift1S4 and Owens, Swift and Smith,135 and the co-precipitation of zinc sulphide with cadmium sulphide has been studied.136 Precipitation of cadmium 8-hydroxyquinolinate from homogeneous solution has been induced by the volatilisation of arnmonia.l3l Flaschka and Jakobljevicli have precipitated mercury( 11) sulphide with thioacetamidel37; Taylor, Smith and Swift studied the rate of this reaction,13* which is very rapid, and they concluded that at pH 4 and a temperature of 55" C, this precipitation could be used for the quantitative separation of mercury from cadmium, zinc, lead, arsenic and nicliel.Although the sulphide precipitates obtained by these methods are generally denser and more readily filterable than those obtained by classical methods, they are, nevertheless, frequently contaminated, and a titrimetric determitiation of tlie metal is often made. The methods cannot be considered entirely satisfactory, and tliere is still need for alternative procedures for the later members of this group.Precipitation of cadmium 8-hydroxyquinolinate from homogeneous solution has been induced by the volatilisation o f The precipitation of cerium( 1V) iodate from homogeneous solution by the controlled oxidation of cerium(II1) ions has been described under Valency change; this technique is more usefiil for the separation of cerium from other rare earth metals than for its determination.55 GROUP IV- Heyn and Dave have precipitated titanium with cupferron synthesised in sitzt from phenylhydroxylamine and nitrous acid.l13 J~~ Decomposition of titanium - cupferron complex commences at temperatures only slightly higher than normal room temperature, and the precipitate was, therefore, ignited to the oxide. Separations of titanium from aluminium, vanadium and pliospliorus were achieved ; the separation from vanadium is not possible by672 CARTWRIGHT, NEWMAN AND WILSON : [Analyst, Vol.92 the conventional method. Dams and Hoste have described the determination of titanium, niobium and tantalum by thermal decomposition of their soluble peroxy-complexes,~40 and the presence of a 10-fold excess of tungsten caused a positive error of only 2 per cent. in the determination of titanium. The precipitation of basic zirconium succinate and formate has been reported,141 and several early attempts were made to precipitate zirconium phosphate in a satisfactory dense form. Willard and Freund generated phosphate ions by the hydrolysis of triethyl ortho- phosphate142 and Willard and HaEin used trimethyl orthophosphate for the same purpose143; neither method is particularly satisfactory as a long period of boiling is required to effect hydrolysis and the precipitates must be ignited to pyrophosphate.In the separation of zirconium from hafnium, Gump and S h e r ~ o o d l ~ ~ precipitated zir- conium arsenate by oxidising arsenite ions with nitric acid; it was found that the presence of sulphate ions was essential to give dense precipitates. The technique of cation release from a zirconium - EDTA complex was used by Babko and Shtokalo, who established that an acidity of not less than 4~ sulphuric acid was required to cause the complex to dissociate145; zirconium has also been precipitated as the tetra- mandelate via hydrolysis of hydroxypropyl mandelate.146 Thorium has been precipitated as oxalate by several workers; dimethyl oxalate was used to precipitate thorium and rare earths from monazite sand34; thorium was then separated from the rare earths by hydrolysis of urea in the presence of formic acid.Later, however, Kall and Gordonld7 studied the solubility of thorium oxalate and concluded that, whenever possible, its use should be avoided. Banks and Edwards used dimethyl oxalate in the separation of thorium from aluminium,148 while Carron, Skinner and Stevens used oxalic acid in methanol for tlie separation of thorium from the rare earths149; the use of diethyl oxalate has been reported. 15* Gordon, Vanselow and W7illard151 have studied the precipitation of thorium with tetra- chlorophthalic acid. The mechanism of the reaction is not clear, however, and may involve complex formation; the precipitates are of variable composition and must be ignited to oxide for weighing.Stine and Gordon57 precipitated thorium iodate by the reduction of periodate with ethylene glycol to generate iodate ions, the ethylene glycol being produced by hydrolysis of 2-hydroxyethyl acetate; the iodate has also been precipitated by photochemical reduction of periodate.60 Takiyama, Salesin and Gordon152 have precipitated thorium 8-hydroxyquinolinate by the hydrolysis of 8-acetoxyquinoline, and the physical properties of the precipitates were found to be superior to those obtained by the direct addition of 8-hydroxyquinoline; ignition to oxide was found to be a reliable method for the determination of thorium.A comparison has been made of the behaviour during thermogravimetric decomposition of thorium 8- hydroxyquinolinate precipitates formed by direct addition and from homogeneous solution.153 The precipitation of thorium with 8-hydroxyquinaldine, generated by hydrolysis of 8-acetoxyquinaldine, has been described.154 The authors found that the optimum conditions for the precipitation were rather more restricted than those of the conventional method and suggested that the use of 8-hydroxyquinaldine and urea hydrolysis might give a better method. GROUP V- Tantalum and niobium have been precipitated with 3,3’,3’, 5,7-pent ah ydrox yflavone, 155 which is produced from the corresponding flavanone by aerial oxidation on boiling a solution in 6 to 9 N sulphuric acid; the method of precipitation from homogeneous solution overcame co-precipitation of potassium and sulphate ions that occurs in the conventional precipitation, and a procedure was described for the separation of niobium and tantalum.Niobium 8-hydroxyquinolinate has been precipitated from homogeneous solution containing sulphuric and oxalic acids by the urea hydrolysis method.156 The composition of the precipitate was established, and the authors suggested that the method would be suitable for the gravimetric determination of niobium. Dams and Hoste have described the precipitation of titanium, niobium and tantalum by thermal decomposition of the corresponding peroxy-anions in solution,140 J~~ and also described the separation and determination of niobium and tantalum by decomposition of the peroxy- complexes in the presence of tannin and 0xa1ate.l~~ The tantalum complex was decomposedNovember, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 673 in the presence of tannin and oxalate in acidic solution; niobium remained in the filtrate, wliicli was treated with nitric acid and sodium bromate, then neutralised with ammonia solution and treated with tannin to precipitate the niobium.The method has been applied to the analysis of tantalo~olumbites.~~~ GROIJP VI- Flaschka and Jakobljevich were the first to study the precipitation of molybdenum sulphide by the hydrolysis of thioacetamide.160 Later, the method was applied by McNerney and Wagner to the determination of molybdenum in titanium alloys,161 and, more recently, a study has been made of the optimum conditions for precipitation with thioacetamide.162 Trithiocarbonic acid has also been used to precipitate molybdenum ~u1phide.l~~ Dams and Hostea5 have developed a method for the determination of tungsten by thermal decomposition of soluble peroxy-tungstate from a nitric acid - hydrogen peroxide solution ; tungstic acid has also been precipitated by evaporation of chlorotungstate in hydrochloric acid solution.16a The method of generating 8-hydroxyquinoline by the hydrolysis of 8-acetoxyquinoline has been applied to the precipitation of uranium165 and the composition of the precipitate was found to vary according to the pH at which precipitation took place.Uranium 8-hydroxyquinolinate has also been precipj tated by a mixed-solvent method by Howick and Rilis.166 Separation from magnesium, thorium and lead was effected in the presence of EDT-4.The synthesis of 1-nitroso-2-naphthol in situ has been used by Patil to precipitate uranium l-nitros0-2-naphtholate~~~ and the precipitate could be weighed directly. GROUP VII- Flasclika and Abdine have developed a method for the precipitation of manganese sulphide by thioacetamide, followed by titration with EDTA to determine manganese,l68 and trithiocarbonic acid has been used in the determination of manganese and its separation from calcium and magnesium.169 GROUP VIII- One of the first attempts to precipitate iron from homogeneous solution was reported by Willard and Sheldon,170 who used the method of urea hydrolysis to precipitate basic iron(II1) formate; dense precipitates were obtained, and the method was recommended for the removal of iron before determining other metals.The precipitation of basic iron(II1) formate has also been described by Uzumasa, Hayashi, Saito and Eiga,171 who studied the co-precipitation of manganese, nickel and copper, and found that the fraction of metal co-precipitated was dependent on the final pH of the solution and that co-precipitation decreased in the presence of ammonium chloride. Gordon and G i n s b ~ r g l ~ ~ precipitated iron(II1) periodate by the hydrolysis of acetamide in acidic solutions containing iron(II1) and periodic acid; the tech- nique was used in co-precipitation studies of iron, aluminium, yttrium and zinc, by Ginsburg, Miller and G0rd0n.l~~ Iron sulphide has been precipitated by means of trithiocarbonic acid.169 Cation release from the iron(II1) - EDTA cornplcx was used by MacNevin and Dunton to precipitate hydrated iron oxide,43 and a similar technique was used by Nightingale and Uenck,174 who precipitated crystalline iron(II1) oxide by hydrolysing the iron - NN-his(2- hydroxyethy1)glycine complex in acidic solution with urea.Dalziel and Thompson have described a new, accurate method for the gravimetric deter- mination of iron by precipitation of the iron( 111) compound of 3-mercaptopyridjne-N-oxide from slightly acidic solution,175 and this compound could be air-dried at temperatures below 160" C and weighed directly. Precipitation from homogeneous solution via hydrolysis of the water-soluble isothiuronium bromide salt of the reagent was preferred because of the relative instability of the free reagent. The precipitation of cobalt nitrosonaphtholate has been achieved by the synthesis of the reagent i n sit26 by reacting nitrous acid and ,O-naphth~l~~; the precipitate was crystalline and stoicheiometric and could be weighed directly after drying at 115" C.In the conventional method, on the other hand, the precipitate is bulky, difficult to filter and invariably contami- nated with excess of precipitant or its reduction product, and must, therefore, be converted under carefully controlled conditions into a form suitable for weighing (the metal, sulphate or oxide).674 CARTWRIGHT, NEWMAN AND WILSON : [Analyst, Vol. 92 The precipitation of nickel sulphide by the hydrolysis of thioacetamide has been studied in some detai1.176Ji7 The precipitation of nickel dimethylglyoximate was first achieved by Bickerdike and Willard178 by using the method of urea hydrolysis ; later the synthesis of dimethylglyoxime in sitzt by reacting biacetyl and hydroxylamine was studieds1 and applied to the gravimetric determination of n i ~ k e 1 .l ~ ~ ~ ~ ~ ~ In the course of this work evidence was found for the formation of a complex between biacetyl monoxime and nickel. A persistent supersaturation was noted in the system, which was thought to be the reason for tlie inability to precipitate completely small amounts of nickel with dimethylglyoxime. Further studies of the mechanism of this reaction have shown it to be quite comp1icated.l8l Biacetyl and hydroxylamine do not react sufficiently quickly to form dimethylglyoxime to account for the rate of precipitation of nickel dimethylglyoximate and the nickel is thought to form a complex with an intermediate carbinolamine, formed by rapid reaction between biacetyl and hydroxylamine; the dehydration of this nickel complex is thought to give a quicker route to the production of nickel dimetliylglyoxiniate.The authors compared this mechanism with that of the analogous precipitation of palladium dimethylglyoximate and showed how a knowledge of the mechanism assisted in producing conditions that avoided excessive co-preci pi t at ion of dimet hylgly oxime. Precipitation of nickel from homogeneous solution u ith cyclohexane-1,2-dione dioxime has been induced by the hydrolysis of acetamide.l82 A liiglily accurate, gravimetric method for nickel was developed, whereas the conventional method with this reagent generally gives high results because of co-precipitation of reagent ; the precipitate was dense and readily filterable, again in contrast to the heterogeneous method that produces a bulky precipitate, which is difficult to filter and wash, Precipitation of nickel 8-hydroxyqui1iolinate, for which the methods of solvent evapora- tionls3 and volatilisation of ammonia were used, has been reported.lS Three methods of oxime reagent synthesis have been applied to the precipitation of palladium.In the first method furfural and hydroxylamine were used to produce fui-furald- oximela4 ; in the second biacetyl and hydroxylamine, to produce dimet l i y l g l y ~ ~ i n i e ~ ~ ~ ; and in the third, indane-1-one-2-oxime and hydroxylamine, to form indane-1 ,2-dionedioxirne.la6 Gagliardi and Pietsclils7 have proposed the use of thioformamjde for the precipitation of palladium sulphide, and the same authors have studied the use of this reagent for platinumlSs and for rhodium and iridium in the presence of palladium.189 CO-PRECIPITATIOX AND FRACTIONAL PRECIPITATION This section includes not only studies of co-precipitation but also the work that has been carried out on precipitation of the rare earths, as much of this is concerned with co-precipita- tion and fractional precipitation.Heyn and Finstonelgo carried out a tracer study of the separation of sodium and potassium from magnesium, when magnesium was precipitated as the oxalate and 8-hydroxyquinolinate from homogeneous solution.They compared the results with those obtained by conventional methods and concluded that tlie co-preripitntion from homogeneous solution decreased in the order : phosphate (single Precipitation) phosphate (double precipitation) > oxalatc > 8-hydroxyqui~iolinate > quinolinate. Co-precipitation of the alkaline earths has beeii studied by using hydrolysis o f dimethyl sulphate in aqueous and of sulpliamic acid,Ig2 and by precipitation of oxalates.lg3 The fractionation of barium - radium mixtures has been investigated by precipitating barium chromate by urea hydrolysis,lg4 and the co-precipitation of radium with barium sulpliate by hydrolysis of sulphaniic acid has been described.lY5 The precipitation of oxalates has been used for investigations of the rare earths and related elements.FVillard and Gordon34 precipitated the lanthanons by hydrolysis of dimethyl oxalate, Gordon, Brandt, Quill and SalutskylQ6 used the method to fractionate lanthanum - cerium and lanthanum - praseodymium mixtures, Gordon and Shavcrlg7 fractionally precipitated rare-earth oxalates and Weaver198 used dimethyl oxalate for the same purpose. Lanthanum and barium sulphates have been co-precipitated from their EDTA complexes by the cation release, with ammonium persulpliate.199 The heavy lanthanons have beenVovember, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 675 jeparated by urea liydrolysis,200 while the carbonates of lanthanum, neodymium and samarium lave been precipitated by hydrolysis of their trichloroacetates.201 Cerium has been precipitated as iodate by oxidation of soluble cerium(II1) iodate to insoluble cerium(1V) iodate with bromate or per~ulphate.~~ The precipitation of insoluble iodates has also been used to study the separation of thorium from the rare earths and the extent of co-precipitation of lanthanum, praseodymium, promethium, europium, yttrium and scandium.202 The distribution of rare earths with thorium was found to follow the logarithmic or heterogeneous distribution law.F i r ~ c h i n g ~ ~ ~ has applied the method of cation replacement to the fractional precipitation of lanthanum - praseodymium iodates. Actinium oxalate has been precipitated with dimethyl oxalate204 and this procedure has been used to separate actinium from non-radioactive impurities.Gordon, Teicher and €3urtt205 have studied the co-precipitation of manganese with basic tin(1V) sulphate precipitated by urea hydrolysis. INDUSTRIAL APPLICATIONS Few industrial applications have been described, although many of the techniques appear to be suitable for use on a large scale when difficult separations are involved or when a par- ticular form of precipitate is required. Some of the applications in chemical technology have been described by Gordon, Salutsky and 'Cl'illard,6 who have paid particular at tention to separation and co-precipitation processes, and have also mentioned the preparation of special products in the polymer, pigment and catalysis fields. A more recent example of the use of the technique of precipitation from homogeneous solution to obtain a particular precipitate has been described by Burrus,206 and Ranby and B~rrus,~O' who precipitated granular anhydrous calcium hydrogen phosphate for use in lamp phosphors by urea hydrolysis in a solution of diammonium hydrogen phosphate.MISCELLANEOUS APPLICATIONS The formula for niobium 8-hydroxyquinolinate has been established from studies of the precipitate obtained by using a urea hydrolysis method, and a gravimetric method for niobium has been 8-Hydroxyquinolinates precipitated from homogeneous solutions were also found to be useful for thermogravimetric208 and infrared studies.209 The co-crystalli- sation of ultra-micro amounts of iron(T1I) with 8-hydroxyquinoline, formed by hydrolysis of 8-acetoxyquinoline, has been studied,210 and the effects of pH and copper(I1) ions on the rate of hydrolysis of 8-acetoxyquinoline have also been reported.211 The kinetics of the formation of dimethylglyoxime from bjacetyl and hydroxylamine have been investigated, both in the absence and presence of nickel ionslS0; nickel dimethyl- glyoximate obtained by precipitation from homogeneous solution was found to be suitable for electron-microscopy studies.212 The precipitation of manganese(T1) sulpliide from homogeneous solution by the tliio- acetamide procedure has been investigated by the method of 1i.m.r. spectroscopy,213 and the authors were able to show that the solubility product of manganese(I1) sulphide cannot be greater than 6 x 10-l2.It is anticipated that further n.m.r. studies of precipitation from homogeneous solution will yield valuable information concerning the nucleation process. Precipitation from homogeneous solution readily lends itself to the investigation of the course followed by a precipitation reaction; it has been used in this connection to investigate the precipitation of basic bismuth acetate214 and indicates the circumstances in which complex iormation is probable. CONCLUSIONS Disadvantages of precipitation from homogeneous solution include the fairly detailed study of conditions that is needed to develop a successful method, the difficulty of testing for completeness of precipitation and the time required for precipitation, which is usually lengthy by comparison with heterogeneous precipitation. The study of the best conditions is a necessity which is, of course, shared by conventional g-ravimetric methods.The development of a successful method entails a thorough investiga- tion of a11 of the relevant experimental conditions.676 CARTWRIGHT, NEWMAN AND WILSON -Analyst, Vol. 92 With most conventional precipitations, it is relatively easy to work with a sample in which the percentage of species sought is unknown. The precipitant is added under appro- priate conditions until no further precipitate is formed, followed by a pre-determined excess. With precipitation from homogeneous solution, however, all of the precipitant is added initially, before conditions are altered to induce the precipitate to form. I t is, therefore, necessary to know enough about the content of the sample to avoid using insufficient pre- cipitant or an excess so large that other errors are introduced.The same difficulty applies to a lesser extent in regulating the amount of the agent that alters the conditions (excess must be present, although a large excess is not usually detrimental), and in deciding when the change in conditions has proceeded sufficiently to ensure complete precipitation. The latter problem may sometimes be solved by using, e.<q., a pH indicator, but in other instances it must depend on experience gained from invcstigation of the method. The actual manipulative time required for a precipitation from homogeneous solution is seldom significantly more than that required by a conventional method, and may often be considerably less, because the precipi- tate is usually more crystalline and may be filtered and washed much more quickly.An over-all saving of time may also be gained because it is not necessary to leave precipitates obtained from homogeneous solution to “age” or “ripen” to grow into filterable crystals or to reduce co-precipitation errors. One practical disadvantage that is not easy to control is the tendency for precipitates that are produced in boiling homogeneous solutions to “bump.” Also, it is necessary to use unscratched and rigorously cleaned glassware, otherwise the precipitate, particularly if of a basic nature, may adhere tenaciously to the wall of the vessel. The main advantage of precipitation from homogeneous solution, namely, improved crystallinity, far outweighs the disadvantages, and results in the production of precipitates that are easier to manipulate, filter and wash, and that arc purer than those obtained by conventional methods.The technique has, therefore, led to the improvement of many classical gravimetric methods, and an understanding of the principles of precipitation from homogeneous solution has also resulted in the development of new gravimetric methods. The importance of precipitation from homogeneous solution to the analytical chemist lies more than in a revival of interest in gravimetric analysis, however. The higher separation efficiency of the homogeneous technique has led to improvements in fractional crystallisation procedures and to a renewed interest in precipitation as a separation technique.Studies of the mechanisms of precipitation and co-precipitation are facilitated by the controllable procedures of precipitation from homogeneous solution. There is real hope that future research into the kinetics of precipitation from homogeneous solution will lead to an under- standing of the precipitation process, one of the oldest and yet the least explicable of the separation techniques of chemistry. The time factor is not necessarily a serious disadvantage. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Stock, A., B e y . dt. clipm. Grs., 1900, 33, 548. Chancel, G., C.R. Hrbd. Se‘anc. Acad. Sci., Paris, 1858, 46, 987. Dorrington, R. J . F., and Ward, A. M., Analyst, 1930, 55, 625. Urillard, H. H., and Tang, N.K., J . .4mer. Chcwz. SOC., 1937, 59, 1190. Gordon, T>., Salutsky, 51. Id., and LVillard, H. H., “Precipitation from Homogeneous Solution,” Willard, H. H., Analyt. Chem., 1950, 22, 1372. Gordon, L., Ibid., 1952, 24, 459. -, Ibid., 1955, 27, 1704. Dams, R., Meded. Vlaanz. Chem. Vereen., 1961, 23, 65. Williams, M., I n d . Chemist, 1962, 38, 134 and 186. Belcher, R., Wilson, C. L., and West, 1.. S., “New Methods of Analytical Chemistry,” Second Edition, Chapman & Hall Ltd., London, 1964, p. 228. Gordon, L., in Wilson, C. T,., and Wilson, D. \V., Editors, “Coniprchensive Analytical Chemistry,” Volume I A, Elsevier Publishing Co., 4msterdam, London, New York and Princeton, 1959, p. 530. Vogel, A. I., “A Tcxtbook of Quantitative Inorganic Analysis,” Third Edition, T,ongmans, Green and Co.Ltd., London, 1961. Firsching, F. H., Talanta, 1963, 10, 1169. 3 , I n d . Engng Ckem. A n a l y t . Edn, 1937, 9, 357. _ _ _ ~ John Wiley & SonsInc., New York; Chapman & Hall Ltd., London, 1959.November, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 677 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. Von Weimarn, P. P., “Zur Lehre von der Zustanden der Materie,” 1914, Th. Steinkopff, Dresden, O’Rourke, J. D., and Johnson, R. A., Analyt. Chem., 1955, 27, 1699. Klein, D. H., and Gordon, L., Talanta, 1958, 1, 334.Turnbull, D., Acta hletall., 1953, 1, 684. Christiansen, J. A., and Nielsen, A. E., Acta Chem. Scand., 1951, 5, 673. Nielsen, A. E., J . Colloid Sci., 1955, 10, 76. Gordon, L., Klein, D. H., and Walnut, T. H., Talanta, 1959, 3, 177 and 187. Fischer, R. B., Analytica Chim. Acta, 1960, 22, 501. Nielsen, A. E., Acta Chem. Scand., 1961, 15, 441. Fischer, R. B., Analytica Chim. Acta, 1960, 22, 508. -, Analyt. Chem., 1960, 32, 1127. Haberman, N., and Gordon, L., Ibid., 1961, 33, 1801. Fischer, R. B., Ibid., 1961, 33, 1802. Klein, D. H., and Driy, J. A., Talanta, 1966, 13, 289. Mealor, D., and Townshend, A., Ibid., 1966, 13, 1191. -- , Ibid., 1966, 13, 1069. Thompson, S., and Gordon, L., Ibid., 1967, 14, 137. Hileman, 0. E., Ibid., 1967, 14, 139. Willard, H.H., and Gordon, L., Analyt. Chem., 1948, 20, 165. Cartwright, P. F. S., Analyst, 1960, 85, 216. Swift, E. H., and Butler, E. A,, Analyt. Chem., 1956, 28, 146. Butler, E. A., Peters, D. G., and Swift, E. H., Ibid., 1958,30, 1379. Peters, D. G., and Swift, E. H., Talanta, 1958, 1, 30. Broad, W. C., and Barnard, A. J., jun., “Thioacetamide as a Sulphide Precipitant,” J. T. Baker Washizuka, S., Bull. Chem. Soc. Japan, 1954, 27, 76. Antia, M. B., Arora, R. C., and Bhatnagar, R. P., Analyst, 1961, 86, 202. Johri, K. N., Ibid., 1961, 86, 487. MacNevin, W. H., and Dunton, 111. L., d4naZyt. Chem., 1954,26, 1246. Cartwright, P. F. S., Analyst, 1961, 86, 688 and 692; 1962, 87, 163; 1967, 92, 319. Dams, R., and Hostc, J . , Talanta, 1961, 8, 664. Heyn, A.A., and Schupak, E., Analyt. Chem., 1954,26, 1243. Gordon, L., Peterson, J . I., and Burtt, B. P., Ibid., 1955, 27, 1770. Firsching, F. H., Ibid., 1960, 32, 1876. __ , Dissertation, Syracuse University, 1954; in Gordon, L., Salutsky, M.L., and Willard, H.H., Tarasevich, N. I., Vest. Mosk. Gos. Univ., 1955, 10, 111. Salesin, E. D., and Gordon, L., Talanta, 1959, 2, 392. Heyn, A. H. A., and Brauner, P. A., Ibid., 1961, 7, 281. Salesin, E. D., and Gordon, L., Ibid., 1960, 4, 75. Howick, L. C., and Jones, J . L., Ibid., 1961, 8, 445. Willard, H. H., and Yu, S. T., Analyt. Chem., 1953, 25, 1754. Hoffman, W. A., and Brandt, W. W., Ibid., 1956,28, 1487. Stine, C . R., and Gordon, L., Ibid., 1953,25, 1519. Newman, E. J., Analyst, 1963, 88, 500. Yen, J.-Y., and Yang, W., Scientia Sin., 1964, 13, 343.Das, M., Heyn, A. H. A., and Hoffman, M. Z., Talanta, 1967, 14, 439. Patkar, A. J., and Varde, M. S., Indian J . Chem., 1964, 2, 123. Prasad, T. R., and Sastri, M. K., Talanta, 1966, 13, 1517. Gordon, L., and Caley, E. R., Analyt. Chem., 1948, 20, 560. Heyn, A. H. A., and Finston, H. L., f b i d . , 1960, 32, 328. Corkins, J. T., Pietrzak, R. F., and Gordon, L., Talanta, 1962, 9, 49. Howick, L. C., Ford, N. L., and Jones, J. L., Ibid., 1963, 10, 193. Ingols, R. S., and Murray, P. E., AnaZyt. Chem., 1949, 21, 525. Gordon, L., and Wroczynski, A. F., f b i d . , 1952, 24, 896. Elving, P. J., and Chao, P. C., Tbid., 1949, 21, 507. Bashar, A., and Townshend, A., Talanta, 1966, 13, 1123. Shaver, K. J . , and Gordon, Id., in Gordon, L., Salutsky, M.L., and Willard, H.H., op. cit., p. 104. Morales, R., and West, P. W., AnaZytica Chim. Acta, 1966, 35, 526. Elving, P. J., and Van Atta, R. E., Analyt. Chem., 1950, 22, 1375. Berak, L., and Munich, J., Colln Czech. Chem. Commun., 1961, 26, 276. Gordon, L., and Firsching, F. H., Analyt. Chem., 1954, 26, 759. Norwitz, G., Ibid., 1961, 33, 312. Wagner, W. F., and Wuellner, J. A., Ibid., 1952, 24, 1031. Firsching, F. H., Talanla, 1959, 2, 326. -, Analyt. Chem., 1961, 33, 1946. Sturnpf, K. E., 2. analyt. Chem., 1953, 138, 30. Krleza, F., Savic, M., and Kicanovic, J., Bull. SOC. Chimts Rep. Pop. Bosnie Hewegovine, 1956,5, 55. Marec, D. J., Salesin, E. D., and Gordon, L., Talanla, 1961, 8, 293. Howick, L. C., and Trigg, W. W., Analyt. Chem., 19B1,33, 302. Howick, L.C., and Jones, J . L., Talanta, 1962, 9, 1037. Willard, H. H., and Fogg, H. C., J . Amer. Chem. SOL., 1937, 59, 1197. 1925. Chemical Co., New Jcrscy, 1960. o p . cit., p. 103.678 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. CARTWRIGHT, NEWMAN AND WILSON : Jones, J. P., Hileman, 0. E., Townshend, A., and Gordon, L., Talanta, 1964, 11, 855. Jones, J. P., Hileman, 0. E., and Gordon, L., lbid., 1964, 11, 861. Willard, H. H., and Gordon, L., Analyt. Chem., 1953, 25, 170.Flaschka, H., and Jakobljevich, H., Analytica Chim. Acta, 1951, 5, 60. Shu, Chuan Liang, and Kuo, I. Lu, Ibid., 1952, 7, 451, Elving, P. J., and Zook, W. C., Analyt. Chem., 1953, 25, 502. Burriel-Marti, F., and Garati, M. J., Recl. Trav. Chim. Pays-Bas Belg., 1960, 79,1495. Koles, J . E., Shinners, P. A., and Wagner, W. F., Talanta, 1965, 12, 297. Flaschka, H., and Jakobljevich, H., Analytica Chim. Acta, 1950, 4, 486. Butler, E. A., and Swift, E. H., Analyt. Chem., 1957, 29, 419. Gagliardi, E., and Loidl, E., %. analyt. Chem., 1951, 132, 33. Musel. A., Gagliardi. E., and Reischl, K., Ibid., 1953, 140, 342. Flaschka, H., and Jakobljevich, H., Analytica Chim. Acta, 1950, 4, 247. Gagliardi, E., and Pilz, W., Z. analyt. Chem., 1952, 136, 344. Musil, A., Gagliardi, E., and Reischl, K., lbid., 1952, 137, 252.Flashka, H., and Jakobljcvich, H., Analytzca Chim. Acta, 1950, 4, 351. Stoner, G. A., and Finstone, H. L., Analyt. Chem., 1957,29, 570. Ross, H. H., and Hahn, R. B., Ibid., 1960, 32, 1691. Pribil, R., and Ctita, F., Chemickd Listy, 1951, 45, 102. Flaschka, H., and Jakobljcvich, H., Analytica Chim. Acta, 1950, 4,'482. McCurdy, W. H., Van der Heuvel, \V. J . A., and Casazza, A. R., Analyt. Chem., 1959,31, 1413. Amin, A. M., Chemist Analyst, 1965, 44, 66. Gagliardi, E., and Loidl, E., Z. analyt. Chem., 1951, 132, 87. -- , Ibid., 1951, 132, 274. Cagliardi, E., and Pilz, W., M h . Chem., 1962, 83, 54. Pietrzak, R. F., and Gordon, L., Talanta, 1962, 9, 327. Heyn, A. H. A., and Dave, N. G., lbid., 1960, 5, 119.Ellefsen, P. R., Gordon, L., Hclchcr, R., and Jackson, W. G., IDid., 1963, 10, 701. Howick, L. C., and Jones, J . L., Ibid., 1963, 10, 197. Davis, D. G., Analyt. Chem., 1960, 32, 1321. Graham, R. P., Billo, E. J., and Thompson, J . &\., Talanta, 1964, 11, 1641. Heyn, A. H. A., and Dave, N. G., lbid., 1966, 13, 27. Hair, R. P., and Newman, E. J., Analyst, 1964, 89, 42. Firsching, F. H., Analyt. Chem., 1961, 33, 873. Caley, E. R., Gordon, L., and Simmonds, G. A., Ibid., 1950, 22, 1060. Flaschka, H., Chemist Analyst, 1955, 44, 2. Bowersox, D. F., Smith, D. M., and Swift, E. H., Talanta, 1960, 3, 282. Klcin, D. H., and Swift, E. H., Ibid., 1965, 12, 349. Hahn, R. B., and Pringle, D. L., Analytica Chinz. Acta, 1964, 31, 382. Amin, A. M., Chemist Analyst, 1956, 45, 95.Pilz, W7., M h . Chem., 1952, 83, 471. Musil, A., and Pilz, W., 2. analyt. Chem., 1954, 141, 19. Nskano, E., J . Chem. Soc. Japan, I n d . Chem. Sect., 1960, 63, 565. Jones, J . P., Hileman, 0. E., and Gordon, L., Talanta, 1963, 10, 111. Firsching, F. H., and Brewer, J. G., Analyt. Chem., 1963, 35, 1630. Hikime, S., and Gordon, L., Talanta, 1964, 11, 851. Flaschka, H., and Jakobljevich, H., Analytica Chim. Acta, 1950, 4, 602. Bowersox, D. P., and Swift, E. H., AIzalyt. Chem., 1958, 30, 1288. Owens, D. V., Swift, E. H., and Smith, D. M., Talanta, 1964, 11, 1521. Tanigawa, Y., Hasegawa, S., and Takiyana, K., .Japan Analyst, 1962, 11, 1300. Flaschka, H., and Jakobljevich, H., .4naZvtica Chim. A d a , 1951, 5 , 152. Taylor, D. C., Smith, L>. M., and Swift, E.H., -4nalyt. Chem., 1964, 36, 1924. Heyn, A. H. A., and Davc, N. G., Talanta, 1966, 13, 33. Dams, R., and Hoste, J . , Ibid., 1964, 11, 1497. Gordon, L., z t z Gordon, L., Salutsky, M. L., and l\'illard, H. H., op. cif., p, 42. Willard, H. H., and Freund, H., I n d . Engng Chem. Analyt E d n , 1946, 18, 195. Willard, H, H., and Hahn, R. B., Anal-vt. Chem., 1949, 21, 293. Gump, J. R., and Sherwood, G. R., Ihid., 1950, 22, 496. Babko, A. K., and Shtokalo, M. I., Zav. Lab., 1958, 24, 674. Rowe, J, C., Gordon, L., and Jackson, W. G., Talanta, 1965, 12, 101. Kall, H. L., and Gordon, L., Analyt. Chem., 1953, 25, 1256. Banks, C. V., and Edwards, R. E., Ibid., 1955, 27, 947. Carron, M. K., Skinner, D. L., and Stevens, R. E., Ibid., 1955,27, 1058. Hagiwara, Z., Technol. Rep.Tohokzt Univ., 1955, 20, 77. Gordon, Id., Vanselow, C. H., and Willard, H. H., Analyt. Chem., 1949, 21, 1323. Takiyama, K., Salesin, E. D., and Gordon, la., Talanta, 1960, 5 , 231. Crouthamel, C. E., and Johnson, C. E., Ibid., 1961, 8, 377. Billo, E. J., Robertson, B. E., and Graham, R. P., Ibid., 1963, 10, 757. Chan, F. L., Ibid., 1961, 7, 253. Kosta, L., and Dular, M., Ibid., 1961, 8, 265. Dams, R., and Hoste, J., Ibid., 1962, 9, 86. -- , Ibid., 1964, 11, 1599. [Analyst, VOl. 92 3 , lbid., 1950,4, 606. --November, 19671 PRECIPITATION FROM HOMOGENEOUS SOLUTION 679 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196.197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. Dams, R., and Hoste, J., Ibid., 1964, 11, 1605. Flaschka, H., and Jakobljevich, H., Analytica Chim. Acta, 1950, 4, 356. McNerney, W. N., and Wagner, W. F., Analyt. Chem., 1957, 29, 1177. Burriel-Marti, F., and Vidan, A. M., Analytica Chim. Acta, 1962, 26, 163. Gagliardi, E., and Pilz, W., 2. analyt. Chem., 1952, 136, 103. Ku, Y. T., and Wang, S. T., J . Chin. Chem. SOC., 1950, 17, 289. Bordner, J., Salesin, E. D., and Gordon, L., Talanta, 1961, 8, 579. Howick, L. C., and Kihs, T., Ibid., 1964, 11, 667. Patil, S. V., Indian J . Chem., 1964, 2, 505. Flaschka, H., and Abdine, H., Chemist Analyst, 1955, 44, 8. Pilz, W., Mh. Chem., 1963, 84, 471. Willard, H. H., and Sheldon, J. L. Analyt. Chem., 1950, 22, 1162. Uzumasa, Y., Hayashi, K., Saito, M., and Eiga, S., .Japan Analyst, 1962, 11, 322. Gordon, L., and Ginsburg, L., Analyt. Chem., 1957, 29, 38. Ginsburg, L., Miller, K., and Gordon, I,., Ibid., 1957, 29, 46. Nightingale, E. R., and Bcnck, R. F., Ibid., 1960, 32, 566. Dalziel, J . A. W., and Thompson, M., Analyst, 1964, 89, 707. Bowersox, D. F., Smith, D. M., and Swift, E. H., Talanta, 1959, 2, 142. Klein, D. H., Peters, D. G., and Swift, E. H., Ibid., 1965, 12, 357. Bickerdike, E. L., and Willard, H. H., Analyt. Chem., 1952, 24, 1026. Salesin, E. D., and Gordon, L., Talanta, 1960, 5, 81. Salesin, E. D., Abrahamson, E. W., and Gordon, L., Ibid., 1962, 9, 699. Gordon, L., Ellefsen, P. R., Wood, G., and Hileman, 0. E., Ibid., 1966, 13, 551. Jones, P. D., and Newman, E. J., Analyst, 1965, 90, 112. Howick, L. C., and Jones, J. L., Talanta, 1963, 10, 189. Perez, F. P., Burriel-Marti, F., and Conejero, L. M., A n . Fis. Quim., 1959,55, 331. Kanner, L. J., Salesin, E. D., and Gordon, T>., Talanta, 1961, 7, 288. Bark, I-. S., and Brandon, D., Ihid., 1963, 10, 1189. Gagliardi, E., and Pietsch, R., MA. Chcm., 1951, 82, 432. __- , Ibid., 1951, 82, 656. ~ _ _ , Ibid., 1952, 83, 487. Heyn, A. H. A., and Finstone, H. L., Analyt. Chem., 1960,32, 328. Gordon, L., Reirner, C. C., and Burtt, B. P., Ibid., 1954, 26, 842. Cohen, A. T., and Gordon, T,., Talanta, 1961, 7, 195. Nozaki, T., and Kurihara, H., Japan Analyst, 1960, 9, 930. Salutsky, M. L., Stiles, J . G., and Martin, A. W., Analyt. Chem., 1953, 25, 1677. Gordon, L., and Rowley, K., Ibid., 1957, 29, 34. Gordon, L., Brandt, R. A., Quill, L. L., and Salutsky, M. L., Ibid., 1951, 23, 1811. Gordon, L., and Shaver, K. J., Ibid., 1953,25, 784. Weaver, B., Ibid., 1954, 26, 479. Nagao, Ideda, and Hiroshi, Ebihara, Japan Analyst, 1959, 8, 39. Fogg, H. C., and Hess, L., J . Amer. Chem. SOC., 1936, 58, 1751. Salutsky, M. L., and Quill, L. L., Ibid., 1950, 72, 3306. Shaver, K. J., Analyt. Chem., 1956, 28, 2015. Firsching, F. H., Ibid., 1962,34, 1696. Salutsky, M. L., and Kirby, H. W., Ibid., 1956, 28, 1780. Gordon, L., Teicher, H., and Rurtt, B. P., Ibid., 1954, 26, 992. Burrus, H. L., J . Appl. Chem., Lond., 1961, 11, 376. Ranby, P. W., and Burrus, H. L., British Patent Spec., 896,660, May 16th, 1962. Bordner, J., and Gordon, L.. Talanta 1962, 9, 1003. Magee, R. J., and Gordon, L., Ibid., 1963, 10, 851. Weiss, H. V., and Shipman, W. H., Analyt. Chem., 1962, 34, 1010. Elliot, D., Howick, L. C., Hudson, B. G., and Noyce, W. I<., 'I'alnnla, 1962,9, 723. Takiyarna, K., and Gordon, l,., Ibid., 1963, 10, 1165. Causey, R. L., and hlaza, K. M., Analyt. Chem., 1962, 34, 1630. Cartwright, P. F. S., Talanta, 1967, 14, 690. Smith, G. S., .4nalyst, 1939, 64, 577. Receiued May loth, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200663

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

680 Analyst, November, 1967, Vol. 92, p p . 680-684 A Critical Study of 8-Hydroxyquinoline as a Gravimetric Reagent for Aluminium* BY ROBERT A. CHALMERS AND MOHAMMED ABDUL BASIT (Chemistry Department, University of A berdeen, Old A berdeen, Scotland) Two distinct techniques are available for the use of 8-hydroxyquinoline to precipitate aluminium, the solution being heated either before or after adjustment of the pH to that required for complete precipitation. If the heating is carried out first, errors, which can be positive or negative in value, may arise. Critical examination of the methods shows that the positive errors can be attributed to the co-precipitation of 8-hydroxyquinoline, and the negative errors to its loss by volatilisation before precipitation can occur, and to the formation of mixed-ligand polynuclear aluminium species.Heating the solution after addition of all of the reagents gives correct results. ALTHOUGH 8-hydroxyquinoline has been used as a reagent for 40 years, few really critical studies have been made of its efficiency. Moreover, variations on the original methodl have become widely used, although the reasons for their development are obscure. The original method involved addition of all the reagents to a cold solution containing the aluminium, adjustment of the pH to that required for precipitation, and heating to coagulate the precipi- tate. In the later methods the solution was heated to various temperatures (e.g., 56" to 60" C293 and 100" C4) before addition of 8-hydroxyquinoline or adjustment of the pH. For micro-scale determination of aluminium the solution was invariably heated to between 70" and 100" C5,6 before adjustment of the pH.Many standard t e x t ~ ~ y ~ , ~ state that the solution should be heated to 60" C or higher before adjustment of the pH. It has been reportedlo that a negative error occurs when this method is used with too small an excess of 8-hydroxy- quinoline, and a positive error if the excess is too large, and this report was later confirmedll as a result of a statistically planned series of experiments. This paper reports which of the two methods is the better, together with an attempt to find the reasons for the errors that arise when the solution is heated before pH adjustment; for convenience we shall call this the "hot method" and the original one the "cold method." EXPERIMENTAL REAGEKTS- A standard solution of aluminium was made from the Specpure metal (Johnson, Matthey) that had been cleaned with emery paper; the metal was dissolved in acid under reflux to avoid loss by spray.A 2 per cent. w/v solution of 8-hydroxyquinoline in IM acetic acid was prepared accurately from the analytical-grade reagent. PROCEDURE- Cold method-An aliquot of aluminium solution was diluted to about 100 ml in a 250-ml beaker and a carefully measured volume of 8-hydroxyquinoline added. The pH was adjusted to give a final value of 5-2, or more, by addition of about 20ml of 2h1 ammonium acetate, and the solution was heated for about 20 minutes to coagulate the precipitate. The precipitate was filtered off on a porosity 4 sintered-glass crucible and washed with warm water (about 70" C).The crucible and contents were then dried overnight at 150" C, cooled, and weighed against a similarly treated crucible. It was assumed that constant weight would be attained in this period. The crucibles had been conditioned by repeated treatment with acid until they showed no loss in weight when acid solutions were passed through them. Hot method-The conditions were the same as for the cold method, except that after the 8-hydroxyquinoline had been added, the solution was heated to about 90°C and the am- monium acetate added dropwise, with rapid stirring, until a precipitate appeared; the rest was added more rapidly. The precipitate was digested for 15 to 20 minutes before filtration.* Presented at a meeting of the Scottish Section, held in Aberdeen, September 2nd, 1966.CHALMERS AND BASIT 681 The results in Table I clearly show that the cold method gives essentially correct results irrespective of the excess of 8-hydroxyquinoline used, . whereas the hot method gives the pattern of error previously reported.l0Y1l It was observed that in the cold method there was an induction period before any precipitate appeared at all. The implication is that this is a case of precipitation from homogeneous solution (presumably one of the earliest on record, although the name had not been invented at that time) and an explanation of it will be given later. TABLE I PRECIPITATION OF ALUMINIUM WITH EXCESS OF S-HYDROXYQUINOLINE Weight of precipitate, g A r \ Excess of 8-hydroxyquinoline added, ml 2 5 10 15 20 25 2 6 10 15 20 25 2 5 10 15 20 25 2 6 10 15 20 25 Cold method & Theory Found 0.0933 0.0934 0.0936 0.0936 0.093s 0-0936 0.0935 0.1864 0,1864 0.1864 0.1864 0-1SG6 0-1865 0.1868 0.3728 0.3725 0.3728 0.3728 0.3730 0.3730 0.3i3S Hot method +- Theory Found 0.0842 - 0.0846 0.0848 0.0851 0.0859 0.1G84 - 0.1692 0.1694 0.1695 0.1697 0.3368 - 0.2998 0.3372 0.3374 0.33’79 0.4335 0.430G 0.4324 0.4335 0.4339 0.4344 0,4349 The next step was to investigate the reasons for the errors in the hot method.There Nere several possible explanations. The positive errors were almost certainly due to co- x-ecipitation of reagent, a phenomenon that had been reported many times by earlier workers. The negative errors might be due to loss of 8-hydroxyquinoline by volatilisation, precipitation If basic aluminium acetate, or formation of mixed-ligand polynuclear complexes.~OLATILISATION OF 8-HYDROXYQEINOLINE- 8-Hydroxyquinoline solution (10 or 20 ml) was mixed with 20 ml of 2 M ammonium icetate and diluted to volume in a 250-1111 standard flask. Then 50-1nl portions were placed n 250-ml beakers, heated for various periods of time, cooled, diluted to the original volume 50 nil) and the 8-hydroxyquinoline content determined spectrophotometrically. Table I1 ,bows that the loss of 8-hydroxyquinoline is quite rapid at about pH 5 and would still be TABLE I1 VOLATILISATION OF 8-HYDROXYQUINOLINE 8-Hydroxyquinolinc lost, per cent. h Time of heating, 7 minutes 10 ml of 8-hydroxyquinoline 20 ml of 8-hydroxyquinoline 10 24 11 20 29 28 30 54 29 60 77 55682 CHALMERS AND BASIT: A CRITICAL STUDY OF 8-HYDROXYQUINOLINE [Analyst, vol.92 appreciable at pH values between 3 and 5 (assuming that it is the neutral molecule that is lost ; the dissociation constant for protonated 8-hydroxyquinoline is about 10-5 moles per litre). It might be objected that in the presence of aluminium the amount of free 8-hydroxy- quinoline would be reduced because of stepwise complex formation between them ; however, the highest complex that could be formed without precipitation taking place would be Al( Ox)2+, where Ox represents the 8-hydroxyquinolinate anion, and one-third of the stoicheio- metric amount of 8-hydroxyquinoline required, plus the excess, would be available for loss in this way.If the excess was small, there might well be a sub-stoicheiometric amount of 8-hydroxyquinoline left before the pH was raised to that needed for precipitation. However it seemed improbable that the negative errors could be accounted for in this way, especially as a deficit of 8-hydroxyquinoline would be indicated by the absence of the characteristic yellow colour in the supernatant liquid. The loss was, nevertheless, important because it could affect the degree of compensation of error arising from co-precipitation off setting losses, and would also affect (through the common-ion effect) the amount of aluminium left in solution. Because of the uncertainty in the amount of excess of 8-hydroxyquinoline actually present a t precipitation (more would be lost in the digestion) this excess could not be deter- mined, except by heating under reflux.As this would make collection of precipitate difficult, experimentation with a deliberate deficit of 8-hydroxyquinoline in the solution was carried out to study the effect on the nature of the precipitate. SUB-STOICHEIOMETRIC EXPERIMENTS- An experiment with no 8-hydroxyquinoline present showed that basic aluminium acetate was precipitated moderately slowly from hot solution but not from cold, and that basic aluminium acetate, either wet or dry, is insoluble in chloroform. As experiments with a small deficit of 8-hydroxyquinoline present (about 5 per cent.) showed the precipitates obtained by the hot or cold methods were completely soluble in chloroform, basic aluminium acetate could not, therefore, be the cause of the error.TABLE I11 VARIATION IN COMPOSITION OF ALUMINIUM 8-HYDROXYQUINOLINATE PRECIPITATES Excess or deficit Aluminium in precipitate,* of 8-hydroxyquinoline, ml per cent. + I 5.92 0 5.90 -1 5 07 -2 6-25 - 4 6.3 1 - 5 6 69 - 10 6-10? - 12 6-20? - 14 7.20t - 16 7 - l l t - 18 8-8Wf * Theory 5.87 per cent. f Calculated from aluminium left in filtrate. It had been observed that the induction period in the cold method depended on the rate of heating, and was quite long if the solutions were left unheated. Precipitates were therefore obtained from unheated solutions containing various deficits of 8-hydroxyquinoline, and the amount of aluminium left unprecipitated was determined by adding more 8-hydroxy- quinoline and heating the solution.The weights of precipitate obtained in this way failed to add up to the total expected, even though all the aluminium had been precipitated. The individual precipitates were therefore analysed for aluminium content, by destroying the 8-hydroxyquinolin with sulphuric and nitric acids and igniting at 1300" C to yield non- hygroscopic alumina. The method was checked with various alum samples, and corrections were applied for volatilisation of platinum from crucibles, and for other possible errors. Typical results are shown in Table I11 and indicate that the precipitates obtained with a deficit of 8-hydroxyquinoline were not aluminium tris(8-hydroxyquinolinate) . Similar results were obtained for precipitates from the hot method.November, 19671 AS A GRAVIMETRIC REAGENT FOR ALUMINIUM 683 It is well known that aluminium forms polynuclear aquo- or hydroxo-bridged species, especially in the pH region 4 to 6.The obvious explanation of the results is that all the aluminium is initially in the polynuclear form, and that these complexes dissociate to provide the mononuclear species needed for normal 8-hydroxyquinolinate formation. The rate of dissociation is increased by heating the solution, and the equilibrium is shifted as the product is removed by precipitation. The mechanism of formation of aluminium 8-hydroxyquinolinate is, presumably, stepwise displacement of water or hydroxyl ligands by the 8-hydroxyquino- linate anion. In the hot method, the 8-hydroxyquinoline is added in the protonated form and the pH is raised by addition of buffer solution.If insufficient 8-hydroxyquinoline is present, mixed-ligand species (either mononuclear or polynuclear) may be precipitated, and if it is present in excess there may be occlusion of mixed-ligand species when precipitation is too rapid. The most likely polynuclear mixed-ligand species (and their aluminium content) are- AI,(OH),(C,H,0N)4 8-12 per cent.[of aluminium; A1,(OH),(C,H60N)5~ 9-32 per cent. of aluminium. The species that precipitate must be uncharged and have either no polar groups that can interact with the solvent, or so few that the solvation energy is insufficient to cause dissolution. It would be expected that A1,(OH),(CgH,ON), would come into this category and would be readily convertible into Al(C,H,ON), in the presence of an excess of 8-hydroxy- quinoline.If a deficit of 8-hydroxyquinoline were used it would be expected that the precipitate would consist of a mixture of the mononuclear and binuclear species, and that if a very large deficit of 8-hydroxyquinoline were used in cold solution, that the precipitate might also contain some of the higher mixed-ligand polymers. This is borne out by the general trend of the results. In the hot method, dissociation of the polynuclear aquo-species must have already reached an advanced state before precipitation begins (otherwise precipi- tation would not occur so rapidly; cf. the need to heat aluminium solutions to form the EDTA complex) and presumably only mononuclear and binuclear species are present in quantity. The presence of a deficit of reagent would result in a two-component precipitate; a large excess would probably be partly adsorbed on the surface of the precipitate (the heterogeneous precipitation Conditions are ideal for this) and cause a positive error.Occlusion of the binuclear species would cause negative errors. At the particular initial excess of 8-hydroxyquinoline recommended by Miller and Chalmers,lo the two kinds of error either fortuitously cancel or are almost negligible. The different level of excess of 8-hydroxy- quinoline arrived at by Mendelowitzll was almost certainly due to the different conditions used by him for the precipitation. If this account of the precipitation process is correct, then the cold method is an example of precipitation from homogeneous solution, because the mononuclear aluminium species is produced in situ by dissociation of the polynuclear species, and this method also provides a new category of precipitation from homogeneous solution.MASS SPECTROMETRY- The evidence presented so far has been circumstantial in nature. After the work had been completed, it became known that mass spectrometry had been applied to the detection of traces of metal chelates.l2 Specimens of the suspected mixed mononuclear - polynuclear precipitates were therefore prepared, and examined by mass spectrometry by Dr. J . R. Majer at the University of Birmingham (private communication). The mass spectra were compared with those of normal aluminium 8-hydroxyquinolinate precipitates prepared by Dr. Majer. A set of prominent peaks appeared at m/e = 332 to 335 for the “mixed” precipitate but was almost absent from the spectra given by the normal 8-hydroxyquinolinates.An m/e value of 332 would correspond to A12(OH)2(CgH,0N)~+ or to Al(OH),(C,H,ON)$. The latter could be obtained from the former by fission, but if it were supposed to be derived from a mononuclear species, that would have to be Al(0H) (H,0)(CgH60N), to maintain the normal six co-ordjna- tion of aluminium, and a peak at m/e = 350 (or 175 for the doubly charged ion) would also be expected to appear. No significant peak appeared at these values. The peak at m/e = 333 would correspond to A1,(H,O),(C,H,ON)~t or Al(H,O) (CgHGON)$ derived by fission. Such species could arise by fission reactions (accompanied by proton transfer) of the postulated mixed-ligand species. In any event there would appear to be good evidence for the existence of at least binuclear mixed-ligand species.684 CHALMERS AND BASIT CONCLUSIONS We find that Berg’s original method of precipitating aluminium 8-hydroxyquinolinate gives correct results independent of the excess of 8-hydroxyquinolinate used, and that the later variants give either positive or negative errors, which may compensate for each other, depending on the conditions used.The negative errors are probably caused by occlusion of mixed-ligand polynuclear species containing a higher proportion of aluminium than there is in the mononuclear 8-hydroxyquinolinate. The original method constituted a special category of precipitation from homogeneous medium.Additional work on iron and zinc showed that they too were best precipitated by the cold method, but cobalt gave erratic results by this method, often forming a product that melted at temperatures below 100” C to give purple - brown globules that were not a simple cobalt 8-hydroxyquinolinate. We thank the Pakistan Council for Scientific and Industrial Research for granting study leave to M. A. B., and Dr. J. R. Majer for carrying out the mass spectrometry for us. 1. 2. 3. 4. 5. 6. 7. a. 9. 10. 11. 12. REFERENCES Berg. R., 2. analyt. Ckem., 1927, 71, 369. Gadeau, R., Revve Me’tall., Paris, 1935, 398. Kolthoff, I. M., and Sandell, E. B., J . Amer. Chem. SOC., 1928, 50, 1900. Smith, G. S., Analyst, 1939, 64, 577. Benedetti-Yichler, A., Mikrockemie, Pregl Festsckv., 1929, G . Alvarez-Querol, M. C., Mikrochemie, 1952, 39, 121. Vogel, A. I., “A Text Book of Quantitative Tnorganic Analysis,” Third Edition, Longmans, Grcen & Co. Ltd., London, 1961, pp. 387-516. Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Third Edition, Macmillan & Co., Ltd., New York and London, 1952, p. 321. Belcher, R., and Nutten, A. J., “Quantitative Inorganic Analysis,” Second Edition, Butterworth & Co. (Publishers) Ltd., London, 1960, p. 86. Miller, C . C . , and Chalmers, R. A., Analyst, 1953, 78, 686. Mendelowitz, A., Analytica Chim. Acta, 1956, 14, 235. Jenkins, A. E., and Majer, J. R., Talanta, 1967, 14, 777. Received December 30th. 1966

ISSN:0003-2654    

DOI:10.1039/AN9679200680

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, November, 1967, Vol. 92, pp. 685-689 685 Rapid Complexometric Determination of Aluminium and Total Iron in Silicate and other Rock Material BY W. H. EVANS (Ministry of Technology, Laboratory of the Govemment Chemist, Cornwall House, Stamford Street, London, S . E . l ) A method is proposed for the rapid determination of aluminium and total iron that involves no prior separation of interfering elements. The sum of iron, aluminium and titanium (as the peroxy complex) is determined by complexation with trans- 1,2-diaminocyclohexane-NNN’N’-tetra-acetic acid and back-titration of the excess a t a pH of 3.5 to 3.7 with a copper solution, with o-dianisidine-NNN’N’-tetra-acetic acid as a metallofluorescent indicator. The iron is determined in a separate aliquot in a similar manner, but with alu- minium and titanium masked with fluoride.Manganese does not interfere. The titanium present is determined spectrophotomctrically, thus enabling the aluminium present to be calculated. Satisfactory agreement is obtained with existing methods for a wide range of values of both elcmcnts in rock material. A RAPID complexometric method for the determination of aluminium and total iron, as iron(III), in silicate, carbonate and phosphate rocks would compare favourably in time and effort with the existing gravimetric R20, method, and determination of aluminium by differ- ence, and improve the precision for high values of total iron measured spectrophotometrically. Such a scheme would be complicated by the presence of titanium, possibly in moderate amounts, and of smaller amounts of manganese, together with the trace elements normally associated with rock material; the latter are usually at a level at which they may be dis- regarded.Many procedures involving the use of metallochromic indicators have been proposed for the determination of aluminium and iron. They fall into two distinct classes, (a) direct stepwise determination and ( b ) back-titration of excess of complexing agent with a suitable metal ion to give the sum of the two elements. This is followed either by a breakdown of the aluminium chelate with fluoride and titration of the liberated complexing agent, or mask- ing of aluminium before complexation in a further aliquot by one of several reagents to give the figure for iron. Generally, direct determinations of these two elements are not favoured because of the rather slow rate of complex formation under feebly acidic conditions in the presence of other salts.This, however, is an advantage when aluminium, in particular, is reacted with an excess of ethylenedianiinetetra-acetic acid (EDTA) and the excess back-titrated with another metal ion. The inertness of the aluminium chelate, and hence the slowness of its dissociation, enables cations with higher stability constants to be used as back-titrants.l Pi-ibil and Vesely2 proposed trans-l,2-diaminocyclohexane-NNN’N’-tetra-acetic acid (DCTA) as a more suitable chelating agent for aluminium and iron. Its chelates are more stable and dissociate less rapidly; further, when present in excess it has the advantage over EDTA that the process of chelation is largely unaffected by the presence of large amounts of neutral salts.Zinc, copper, lead and thorium solutions have been used extensively for the determination of aluminium, as back-titrants in conjunction with metallochromic indicators sensitive to these elements. Lead has a tendency to form complexes with many anions under weakly acidic conditions, while thorium forms stable fluorides; zinc has been used with dithizone3 or xylenol orange4 as indicators, while 1-(2-pyridy1azo)-2-naphthol5 and pyrocatechol violet6 have been used as indicators in back-titrations with copper. Relcher, Rees and Stephen7 proposed o-dianisidine-NNN’N’- t etra-acet ic acid as a met allofluorescent indicator for the tit rat ion of686 EVANS : RAPID COMPLEXOMETRIC DETERMINATION OF ALUMINIUM [Analyst, Vol.92 copper. In the pH range 4 to 10, copper quenches the fluorescence associated with EDTA. This indicator would enable titrations associated with these determinations to be conducted at a lower pH than hitherto. A method developed for the rapid determination of aluminium and total iron, which involves no time-consuming separations and is not subject to inter- ference in the normal type of rock material, is described in this paper. METHOD REAGENTS- Reagents should be of analytical-reagent grade where available. HydrofEuoric acid, 40 per cent. v / v . Perchloric acid, 60 per cent. v / v . Hydrogen peroxide solution, 20 volume. Sodium acetate solution, 10 per cent. w l v . Ammonium Jlztoride solution, 4 per cent.W / V . o-Dianisidine-NNN‘N‘-tetra-acetic acid, 0.5 per cent. w/w in sodium chloride-10 mg of this indicator mixture are equivalent to 0.01 of 0.01 M copper solution. Copper solution, 0.01 nr-Wash a strip of copper metal with hydrochloric acid ( 1 + 2), rinse with water and dry in an oven. Weigh 6354 mg of the metal and dissolve it by warming in 2 to 3 ml of concentrated hydrochloric acid, to which a few drops of concentrated nitric acid have been added. Dilute to 800 ml and, after adjusting the solution to a pH of 4 with 10 per cent. sodium acetate solution, dilute to 1 litre. Trans-1,2-diaminocyclohexane-NN”N’-tetra-acetic acid (DCTA) , 0.01 5 M-Dissolve 5-24 g of the complexing agent in a few millilitres of 2 N sodium hydroxide solution and adjust to pH 4 with hydrochloric acid (1 + 2) ; dilute to 1 litre and store the solution in a polythene bottle.Standardise against the above copper solution at pH 4. APPARATUS- and is suitably mounted for ease of titration. UZtraviolet lamp-A 125-watt black-glass shielded lamp that emits radiation at 366 mp, PROCEDURE- Decompose 1 g of powdered rock with 20 ml of hydrofluoric acid by digestion overnight in a platinum or PTFE basin. Add 5 ml of perchloric acid and evaporate to dryness on an air-bath. Repeat the evaporation with three further separate 5-ml portions of perchloric acid, to the first of which 2 ml of sulphuric acid ( 1 + 1) have been added, until no further fumes are evolved. Finally, dissolve the residue in 5 to 10 ml of perchloric acid, and dilute the solution to 200 ml.Determine the titanium content of the rock by any of the standard hydrogen peroxide spectrophotometric procedures; that described by Wilsons is satisfactory. Dilute a 10-ml aliquot of the rock solution (equivalent to 50 mg of rock) to 80 ml with distilled water and add 2 ml of 20 volume hydrogen peroxide. Add. by pipette, 20 ml of 0.015 M DCTA (or sufficient excess to give a titration of 5 to 10 ml) and adjust the pH to between 3.5 and 3.7 with 4 to 5 ml of 10 per cent. sodium acetate solution. Warm the solution to a temperature of 50” C, cool, add 50 mg of indicator mixture (more for a highly coloured solution) and titrate the excess of complexjng agent with 0.01 h i copper solution, while stirring, until the blue fluorescence observed by ultraviolet light in a darkened room is quenched. Dilute another 10-ml aliquot of rock solution to 90 ml, add 2 ml of 20-volume hydrogen peroxide, 5 ml of 4 per cent.ammonium fluoride solution, mix thoroughly, and add, by pipette, 10ml of 0.015 M DCTA. Adjust the pH to between 3.5 and 3-7 with 10 per cent. sodium acetate solution, add 50 mg of indicator and titrate the excess of coniplexing agent with 0.01 M copper solution, as before. The complexing agent consumed in this titration, after allowing for the indicator, gives a value for total iron as iron(II1). The first titration gives the sum of iron, aluminium and titanium, from which aluminium may be calculated by difference. If necessary, and particularly if their sum is about 0-05 per cent., the total figure for iron should be corrected for any small amounts of copper, cobalt and nickel present, and the figure for aluminium should be similarly corrected for chromium and zirconium.November, 19671 AND TOTAL IRON I N SILICATE AND OTHER ROCK MATERIAL 687 DISCUSSION The main factor in evaluating the conditions for the determination of aluminium and iron was the need to remove interference from titanium and manganese, without having recourse to time-consuming separations.Manganese would normally be present at low concentrations, 1 0 - 5 ~ (equivalent to 0.20 per cent. of manganese(I1) oxide in 50-mg rock samples) ; it forms a chelate with EDTA that has a smaller stability constant than aluminium. At a concentration of M and pH lower than 4 no complexation with EDTA4 was observed, while at pH higher than 4 the resulting partially formed chelate was labile and readily replaced by copper.DCTA, which forms more stable complexes, behaved in a similar but less pro- nounced manner; at concentrations of M little chelate formation at pH lower than 4 was observed. Titanium cannot be determined complexometrically by back-titration of excess of EDTA with copper or zinc as the stability constants of the chelates of the back-titrants are greater, and the titanium chelate is labile and readily replaced.9 At concentrations approaching M (equivalent to 0-2 per cent. in 50-mg rock samples) and pH 3.5, the titanium could be completely replaced from its chelate, but above this concentration the degree of complexation was uncertain. Attempts to mask the titanium reaction with a 0 .0 5 ~ concentration of phosphate proved unsuccessful ; aluminium complexation was seriously inhibited in the pH range 3.5 to 5-5. Attempts to mask titanium from complexation with several reagents, e.g., ascorbic, mandelic and tannic acids, or Tiron were equally unsuccessful because of incomplete masking, the instability of the iron and aluminium chelates in the presence of the reagents or interference in the end-point indication. In the presence of peroxide, however, the more stable titanium peroxy - EDTA or DCTA chelate is formed, which is stable in the pII range 1 to 6 and is not replaced by copper or zinc on back-titrati~n.~ The strong colour of this complex at titanium concentrations of 1 0 - 4 ~ , reinforced by the strong colour from high concentrations of iron(II1) chelates, made indication difficult. The desirability of keeping the pH below 4 suggested the use of the metallofluorescent indicator o-dianisidinetetra-ace tic acid, which was effective down to pH 3.5, and a copper solution as hack-titrant.This indi- cator is itself a complexing agent that forms chelates with copper of lower stability than EDTA or DCTA and gives an indicator blank. The end-points are exceedingly sharp, z i z . , 0.02 ml of 0.01 M copper solution. The logarithmic conditional stability constant of copper - EDTL4 at pH 3.5 is about 9.2 and that of the DCTA chelate about 10.2. Thus a titration of to M complexing agent should be possible and to within a titration error of 0.1 per cent. for DCTAlo; EDTA might not prove so satisfactory.Previous authors have commented on the disadvantage of using EDTA as excess com- plexing agent for aluminium. Stoicheiometric reaction is difficult to achieve, and further, anomalous results are obtained when iron(II1) forms a complex with EDTA at pH 3.5 and the excess is back-titrated with copper. DCTA reacts quantitatively with both elements over the range 2 to 20 mg and is preferred; there is no reason to suppose that complexing impurities are present in DCTA that would give rise to an incorrect titration. The initial presence of fluoride will inhibit complexation of aluminium; amounts of up to 0.3 mg in the titrating medium did not inhibit the reaction of aluminium with DCTA if the solution was warmed to 60" C. To ensure a fluoride content below this figure it was necessary, during the preparation of the rock solution, to evaporate with a small addition of concentrated sulphuric acid.Sodium acetate was used for pH control and did not interfere up to a concentration of 0.15 NI in the titrating medium ; under the conditions described in the procedure, acetate concentra- tion did not exceed 0.05 hi. Fluoride has been used on many occasions to liberate EDTA from the aluminium chelate and thus enable aluminium to be determined in the presence of other chelated ions. Cimer- mann, Alon and Mashallll have shown this to be possible in the presence of iron(II1) provided that the fluoride concentration is below 0.1 M ; above this concentration the iron(TI1) chelate also tends to dissociate.The DCTA chelate needed to be boiled for some minutes at pH 4 to liberate the complexing agent quantitatively. Unfortunately in the presence of free DCTA, the iron(II1) chelate caused further reaction to take place, giving an apparent decrease in the complexing agent liberated and hence in the aluminium figure, and an increase in the figure for iron(lI1) obtained by difference. The titanium peroxy - DCTA chelate readily liberated complexing agent at pH 3.5 but was somewhat dependent on peroxide concentration. I t was noticed that the corresponding EDTA chelate by comparison was inert to fluoride, the amount688 EVANS RAPID COMPLEXOMETRIC DETERMINATION OF ALUMIXIUM [A%h!&St, VOl. 92 of release decreasing from 90 per cent. at pH 3-5 to 25 per cent. at pH 5.5, with a fluoride concentration of 0.05 M.Trials with rock solutions by this method gave disappointing results and the procedure was abandoned. Pi-ibil and Vesely have suggested salicylic acid in the presence of triethanolamine and alkali12 or ammonium fluoride2J3 as suitable masking reagents for aluminium. Of these two reagents the latter was superior provided that the sodium and fluoride-ion concentrations did not exceed certain levels at which iron(II1) complexation was inhibited by the formation of hexafluoroferrates. The tolerance in their concentration was considerable and sodium acetate could still be used to adjust the pH. The concentration of fluoride should be the minimum necessary and a 0.05 M concentration in the titratirig medium (a 5-fold excess for 10 mg of aluminium) was adequate to mask aluminium and titanium, and to enable the alkaline earth metals to form complexes.It was advisable to add peroxide to both aliquots because rock material high in iron(II), even after decomposition and treatment in an oxidising environment, still contained iron in the unoxidised state. Thus the total iron could be determined in the first aliquot and the sum of iron, aluminium and titanium in the second. Titanium levels normally present in rock material could be determined spectrophotometrically with sufficient accuracy, thus enabling the aluminium to be calculated by difference. Copper, cobalt and nickel form complexes under these conditions and would be included in the iron(II1) figure; similarly, chromium and zirconium react in the absence of fluoride and are included in the aluminium figure.In practice, their sum seldom exceeds 0.04 per cent. ; if desired, the necessary allowance could be made. Where these elements were present in much larger amounts the method was invalidated. RESULTS Because of the varying composition of rock material it was considered not significant to make correlations involving pure solutions of iron( 111) and aluminium. Accordingly, the values obtained by this procedure for about 30 rocks of widely varying composition were compared with those obtained by the gravimetric R,03 and difference method for aluminium, and by the 2,2’-bipyridyl spectrophotometric method for total iron. The maximum differ- ences from the latter were +0-19 to -0.15 per cent.for total iron, as iron(1II) oxide, and +0-16 to -0.22 per cent. for alumina. The correlation coefficient was 1.003 for iron(JI1) oxide and 0-998 for alumina, the relative standard deviation being 0.008 in either instance. Values are shown in Table I covering a range for both elements. TABLE I COMPARISON OF ALUMINA AND IRON(II1) OXIDE VALUES Total iron Alumina determined Alumina determined determined complexometrically complcxometrically, by difference, as iron(II1) oxide, per cent. per cent. per cent. Granite G.l . . 14.06, 14.10, 14.08 14-04*, 14.08t 1-92, 1.88, 1.86 Diabsse W.l . , 16.12, 15.07, 15.06 14,94*, 14.851. 11.09, 11.13, 11.12 Tonalite T.l . . 16-50, 16-43, 16.42 16.46 5.76, 5-83, 5.80 Wollastonite . . 1.56, 1.56 1.56 - Muscovite- tremolite-schist 5.25, 5.27 5.31 2.35, 2-32 Limestone .. 8.91, 8.79 8.80 3-42, 3.42 Alkali-granite . . 13.26, 13-33 13.36 1.92, 1.86 Olivine-basalt . . 15.95, 16.0 16.0 5.92 Basalt . . . . 16.6, 16-6 16.6 9.9, 9.9 Biotitic schist . . 18.3, 18.2, 18.2 18.2 12.8 Dolerite . . 21.3, 21.5 21.5 7.94, 7-92 Montebrasite . . 33.5, 33.5 33.6 0.58, 0.51 * Preferred ~ a l u e . 1 ~ f Preferred value.lS Total iron determined spectrophotometrically as iron(I1T) oxide, per cent. 1-96*, 1-90? 11*16*, 11.lOt 5.83 0.38 2.33 3.45 1.91 5.89 9.9 12.8 7-92 0.51 This paper is published by permission of the Government Chemist and the Director, Institute of Geological Sciences.November, 19671 AND TOTAL IRON IN SILICATE AND OTHER ROCK MATERIAL 689 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Nydahl, F., Talanta, 1960, 4, 141. Piibil, R., and Vesely, V., Ibid., 1962, 9, 23. Wanninen, E., and Ringbom, A., Analytica Chim. Acta, 1955, 12, 308. Korbl, J., and Piibil, R., Chemist Analyst, 1956, 45, 102. Wilkins, D. H., and Hibbs, L. E., Analytica Chim. Acta, 1958, 18, 372. Suk, V., and Malat, M., Chemist Analyst, 1956, 45, 30. Belcher, R., Rees, D. I., and Stephen, W. I., Talanta, 1960, 4, 78. Wilson, A. D., Analyst, 1963, 88, 18. LVilkins, D. H., Analytica Chim. Acta, 1959, 20, 113. Ringbom, A., “Complexation in Analytical Chemistry, Chemical Analysis,” Volume 16, Inter- Cimermann, C., Alon, h., and Mashall, J., Talanta, 1968, 1, 314. Piibil, R., and Vesely, V., Ibid., 1963, 10, 233 and 383. ~- , Tbid., 1965, 12, 385. Fleis’cher, M., and Stevens, R. E., Geochzm. Cosmochim. Acta, 1962, 26, 525. Ingamells, C. O., and Suhr, N. H., Ibid., 1963, 27, 897. science Publishers, New York and London, 1963, p. 77. Received June 20th, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200685

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

690 AnaZyst, November, 1967, Vol. 92, $9. 690-694 The Determination of Small Amounts of Fluorine in Rocks and Minerals BY vCr. H. EVANS AND G. A. SERGEANT (Ministry of Technology, Laboratory of the Goriernment Chemist, Cornwall House, Stamford Street, London, S.E. 1) A shortened method of fluorine determination in rocks and minerals is described. A small, powdered sample is decomposed by alkaline fusion; fluorine is separated by a form of Willard - Winter distillation and collected on anion-exchange resin. After elution, the fluorine is determined colori- metrically by a zirconium - Eriochrome cyanine R method. The method is intended to cover the determination of fluorine in concentrations down to the normal lower trace levcls in rocks. IN the determination of fluorine in rocks and minerals, a preliminary separation of the element by the method of Shell and Craig1 has become well established, and is used in several recent paper^.^,^^^ After decomposition of the sample by fusion with alkali carbonate, silica and interfering cations are removed from the aqueous extract by precipitation with a zinc salt and ammonium carbonate.Fluorine is then isolated by a Willard - Winter di~tillation,~ and determined in the distillate by titrimetry or colorimetry. In the present paper a short- ened method is now proposed in which the preliminary separation of silica is omitted, and the fluorosilicic acid from a single distillation is collected on a column of anion-exchange resin in preparation for the final colorimetric measurement. Interference by silica and aluminium, both of which tend to lower the recovery of fluorine in the distillation, is minimised by restricting the sample weight to not more than 200 mg.EXPE RI 111 E NTA4L For the colorimetric determination of fluoride in aqueous solution, a considerable number of reactions based on the bleaching effect of fluoride ion on various organic dj7e complexes with metals have been described, and Relclier, Leonard and West6” have proposed a direct colour reaction in which fluoride ion forms a ternary complex with cerium(II1) alizarin complexan. After consideration of these possibilities, a method in the former class was finally adopted, based on the zirconium - Eriochrome cyanine R method described by Megregian,s and recently studied in further detail by Sarma.g This has good colour stability and repro- ducibility, and the advantage in the present application, of working in fairly strongly acid solutions without the need for close control of the pH.Under conditions slightly modified from those given by Megregian, a negative linear response was observed in the colorimetric measurement over the range 2 to 54 pg of fluorine per 100 nil of solution, with a molecular extinction coefficient of 10,000. The concentration of fluoride by evaporation in glass vessels, of large volumes of solution such as Willard and Winter distillates, was found by Spechtlo to incur losses either by retention on glass surfaces, or by volatilisation after reaction with borosilicates in the glass. To solve this difficulty, Neilsonll proposed the collection of fluoride on anion-exchange resin.We have used a small column of anion-exchange resin for the quantitative recovery of fluoride up to 600 pg (corresponding to 0.3 per cent. of fluorine in a 200-mg sample). For amounts of fluorine in excess of this, an aliquot of the distillate is taken for the colorimetric measurement, and the resin colunin is not required. The recovery of fluorine by steam-distillation is dependent on such factors as the design of the still, the composition of the acid medium, the distillation temperature, rate of distillation and the presence of interfering elements. Ry using the apparatus shown in Fig. 1 , trials over the range 30 pg to 5 mg of fluorine, added as sodium fluoride to the distillation flask, showedEVANS AND SERGEANT 691 an average recovery of 95 per cent.; relative standard deviation 0.025.The initial content of the distilling flask in these tests was 60 ml of about 50 per cent. v/v sulphuric acid, and the distillation temperature ranged from 145" to 150" C. Aluminium salts and large amounts of gelatinous silica both tend to hold back fluorine during the distillation, but these effects were found to be negligible when working with a small weight of sample, and collecting a sufficiently large volume of distillate. Thus in the presence of 200 mg of silica, added as sodium silicate, and 50 mg of alumina, the average recoveries of 600 pg of fluorine were 97 and 95 per cent., respectively, against 95 per cent. with no additions. Not less than 96 per cent. of the total fluorine collected was found in the first 200 ml of distillate.With an addition of boron, as boric acid (10 mg), large in comparison with that likely to occur in a rock, the average recovery of 300 pg of fluoride was 95 per cent. Also tested were net over-all recoveries of fluoride added to rocks of known fluorine content, by evaporating aliquots of standard fluoride solution, made slightly alkaline, in the platinum crucible used for the fusion. The results, which are given in Table I, again show an average recovery of 95 per cent. ; relative standard deviation 0.026. A reagent blank, not exceeding 3 p g of fluorine, was found experimentally and deducted in all determinations. TABLE I RECOVERY OF FLUORINE ADDED TO SILICATE ROCKS Rock Fluorine added, P6 A 30 B 60 A 150 B 150 A 450 B 450 Recovery, per cent.91.7 97.7 97.6 94.0 92.2 95.4 A = 200 mg of quartz-porphyry (silica, 80 per cent. ; alumina, 13 per cent.; fluorine, 150 p.p.m.). alumina, 24 per cent.: fluorine, 220 p.p.m.). €3 = 200 mg of altered basalt (silica, 32 per cent.; SUPP'Y steam Fig. 1. Distillation apparatus692 APPARATUS- The main part of the distillation apparatus is shown in Fig. 1. Steam is passed into the 100-ml distillation flask, A, connected by a splash head to a condenser. The distillation flask is fitted with an adjustable contact thermometer, T (“HE JU” type, range 0 to 200” C), con- trolling a radiant heater, H (Electrothermal, 300 watts), through a simple transistorised relay, the circuit diagram of which is given in Fig. 2. The distillate is directed by a small plastic EVANS AND SERGEANT: DETERMINATION OF SMALL AMOUNTS [Analyst, Vol.92 METHOD . Heater Contact thermometer Fig. 2. Relay circuit diagram A.C. mains input - funnel into the ion-exchange column, C, a glass tube 20 cm long by 1 cm i.d., closed at the bottom by a retaining plug of cotton-wool over a neoprene bung through which passes a short rigid plastic tube attached to a length of capillary PVC tubing. The rate of drainage of liquid from the column is controlled by adjusting the height of the outflow tip of the PVC tubing. To prepare the column, make 750 mg of moist Ue-Acidite FF resin (Cl- form, 100 to 200 mesh) into a slurry with water, and rinse into the column with more water. Convert the resin into the OH- form by treating it with 5 ml of N sodium hydroxide solution, and then wash it free from excess alkali with about 50 ml of water.Regenerate the resin, which may be used about forty times, in the same way after each fluorine determination. REAGENTS- Sodium cavbonate, anhydrotbs. Sulp Iauric acid, ZN . Sulphuric acid, 1 8 ~ (1 + 1). Sodiziin acetate solzttion, 0.1 M. Standard fEuoride solution-Dissolve 110-6 mg of sodium fluoride in water, and dilute the solution to 500 ml. From it prepare, by dilution with water, a standard fluoride solution containing 2 pg of fluorine per ml. Zirconyl chloride solution-Dissolve 132.5 mg of zirconyl chloride, ZrOC1,.8H,O, in a little water, add 600ml of concentrated hydrochloric acid, and dilute the mixture to 1 litre with water. Erioclzrow cyanine R solution-Dissolve 0-8 g of the solid reagent in 1 litre of water.Different batches of this reagent have been found to contain varying amounts of sodium sulphate impurity. This may be removed by dissolving the dye in methanol, filtering off insoluble material, and evaporating the filtrate to dryness under reduced pressure. The purified reagent should be stored in a desiccator. This solution contains 100 pg of fluorine per ml.November, 19671 OF FLUORINE IN ROCKS AND MINERALS 693 PROCEDURE- Distillation-Fuse 200 mg of sample powder (70 mesh) for 15 to 20 minutes over a Meker burner, in a covered platinum crucible with 1 g of sodium carbonate. After cooling, add 10ml of 2~ sulphuric acid slowly and allow the mixture to digest for 30 minutes, then transfer the contents of the crucible to the distillation flask, rinsing the crucible and lid into the flask with 45 ml of 1 8 ~ sulphuric acid.Set up the apparatus shown in Fig. 1, switch on the radiant heater, and, when the contact thermometer has reached 125" C, begin passing in steam to start the distillation. The distillation temperature should keep within the range 145" to 150" C without further attention at a distillation rate of 6 to 8 ml per minute, set by appropriate adjustment of the supply of steam. Allow 300 to 400ml of distillate to pass through the ion-exchange column, and then elute the fluoride absorbed in the column with 25 ml of 0 . 1 ~ sodium acetate solution, collect- ing the eluate in a 100-ml flask. Colorimetric determination of JEuorine-Take an aliquot of the eluate containing not more than 50 pg of fluorine, and dilute it to 70 ml in a 100-ml graduated flask.Add 10 ml of the zirconyl chloride solution, followed by 10 ml of the Eriochrome cyanine R solution, and dilute to the mark with water. In a similar way, starting with aliquots of dilute standard fluoride solution, prepare a series of colour standards containing from 0 to 50 pg of fluorine. Prepare the colour reference solution by mixing 70 ml of water, 6 ml of concentrated hydro- chloric acid and 10 ml of the Eriochrome cyanine R solution, and diluting to the mark with water in a 100-ml graduated flask. After 30 minutes measure the optical densities of the test and colour standard solutions against the reference solution in 1-cm cells at 525 mp. Deter- mine the fluorine content of the test solution by reference to a calibration graph prepared from measurements on the colour standards.The relationship between fluorine content and optical density is linear over the approximate range 2 to 50pg of fluorine per 100ml of solution. The reference solution should be freshly prepared for each batch of readings, but it should not be necessary to re-calibrate with new fluoride standards on each occasion. ,4s relatively small deviations in fairly high optical densities have to be measured accurately in this procedure, spectrophotometer cells should be covered to prevent volume changes by evaporation. RESULTS The results of applying the method to a variety of silicate and other rock samples are An empirical correction factor has been applied to each to allow for the shown in Table 11.observed 95 per cent. over-all recovery of fluorine in the method. TABLE I1 FLUORINE VALUES FOR SOME ROCKS AND MIXERALS Material analysed Granite G-1 . . . . . . . . Dolerite (diabasc) W-1 . . . . . . Tonalite T-1 . . . . . . . . Oolitic limestone . . . . . . Biotitic green schist . . . . . . Porphyritic basalt . . . . . . . . Hornblende porphyrite . . . . . . Montebrasite . . . . . . . . Carbonate-apatite . . . . .. . . Fluorine found, p.p.m. 610, 630, 620, 630, 620* 230, 240, 230, 220, 230, 220* 460, 450 120, 140, 120, 130 750, 730, 700, 760, 760, 740, 750, 700 1200, 1200 2050, 2000, 2100 1.36, 1.347 3.6, 3.6t * These figures fall satisfactorily within the wide range quoted by Fleischer in t Values givcn as per cent.the most recent summary of results for these standard r0c.ks.l' This paper is published by permission of the Government Chemist, and the Director, Institute of Geological Sciences.694 EVANS AND SERGEANT Appendix I CIRCUIT OF RELAY UNIT R, = 47,000-ohm, ;-watt resistor R, = 47-ohm, &-watt resistor C, = 100-pF capacitor, 25-volt working C , = 500-pF capacitor, 50-volt working C , = 0-l-pF capacitor, 600-volt working A = Relay, 700-ohm (Standard Telephones and Cables, Ltd., Type 4190 ED) D = Selenium rectifier, full-wave, battery-charger type, 12-volt, l-ampere T = Transformer, battery-charger type, 12-volt, l-ampere TR= OC 72 transistor 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES Shell, H. R., and Craig, R. L., Analyt. Chern., 1954, 26, 997. Peck, L. C., and Smith, V. C., Talanta, 1964, 11, 1343. Jeffery, P. G., Geochim. Cosmochirn. Acta, 1962, 26, 1355. Ingamells, C. O., Talanta, 1962, 9, 507. Willard, H. H., and Winter, 0. B., Ind. Engng Clzem. Analyt. Edn, 1933, 5 , 7. Belcher, R., Leonard, M. A,, and West, T. S., .[. Chem. Soc., 1959, 3577. Megregian, S., Analyt. Chem., 1954, 26, 1161. Sarma, P. L., Ibid., 1964, 36, 1684. Specht, R. C., Ibid., 1956, 28, 1015. Niclson, H. M., Ibid., 1958, 30, 1009. Fleischer, M., Geochirn. Cosmochim. Acta, 1965, 29, 1263. , 2 , Talanta, 1959, 2, 92. --- Received May 3rd, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200690

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, November, 1967, Vol. 92, $9. 695-697 695 The Synthesis of the Active Component of Commercial Titan Yellow for Use in the Determination of Magnesium BY H. G. C. KING, G. PRUDEN AND N. I;. JANES (Rothamsted Experimental Station, Harpenden, Herts.) The sodium sulphonate groups of the magnesium-reactive component of a commercial Titan yellow are shown to be in the 7-position on each of the benzothiazole rings, Titan yellow synthesised from 2-($-aminophenyl) -6- methylbenzothiazole-7-sulphonic acid is reliable and sensitive in the determina- tion of magnesium. EARLIER~ we described the chromatographic separation of small amounts of the magnesium- reactive component of commercial Titan yellow, which is a complex mixture.2 We also drew attention to the various formulae ascribed to the dye by earlier workers.Commercial Titan yellow is made by diazocoupling the sulplionated product of dehydro- thio-p-toluidine. I t has been generally assumed that the magnesium-reactive component is a diazo-amino compound with two benzothiazole rings, formed from two molecules of one of the 5 possible sulphonic acid isomers of dehydrothio-$-tohidine. The positions of the sulphonate groups have not been established, although speculative assignments, usually to the 2'- or 3'- positions, have been made.2 y 3 y 4 l5 F~ In thc earliest work on dehydrotliio-P-toluidine sulphonic a ~ i d , ~ , ~ , ~ the position of the sulphonate group was unspecified, although by 1915 a formula for Titan yellow, then called Azidingelb 5G, appeared with sulphonate groups in the 3'- position.8 SchubertlO demonstrated that the commercial azo dye Sirius light yellow RR, where X = - N = N -, was the 7-sulplionate, mixed with a smaller amount of the 5-sul- phonate isomer.By analogy with his structure for Sirius light yellow RR we were led to suspect that the active component of commercial Titan yellow also has two 7-sulphonate groups (X = - N = N - HN -), because the starting material for both dyes is sulphonated dehydrothio-p-toluidine. If this is so, the main reduction product of both commercial Sirius light yellow RR and commercial Titan yellow should be tlie same amino-sulphonic acid. Schubert reduced Sirius light yellow RK, and purified the resulting dehydrothio-9- toluidine-7-sulphonic acid as its ammonium salt, which is insoluble in cold water, a technique discovcred by Green.; When treated similarly, commercial Titan yellow gives a sulphonic acid that can be purified in the same way.The acid is identical with authentic deliydrothio-9-toluidine- 7-sulphonic acid, i.e. 2-(~-aminophenyl)-6-metliylbenzothiazol~-7-sul~lionic acid, in ultra- violet and nuclear magnetic resonance spectra, RF values and chemical properties. Dehydro- thio-~-toluidine-7-sulphonic acid and its ammonium salt are well defined crystalline com- pounds that decompose sharply at high temperatures and, like the sodium salt, are further characterised by their bright blue - violet fluorescence in ultraviolet light. Larger quantities of a Titan yellow, containing negligible amounts of impurities, can be made by diazotising 2-(~-aminoplienyl)-6-metl~ylbenzothiazole-7-sulphonic acid (Aldrich Chemical Co.Inc.) dissolved in slightly more than one equivalent of aqueous sodium hydr- oxide; the diazonium salt is insoluble in the reaction mixture, and can be washed free from acid and salts; it is then coupled with an equal weight of the 7-sulphonic acid dissolved in two equivalents of sodium hydroxide. By washing with hot acetone the product is freed from most of the impurities. The method is reproducible, giving good yields of a product that can be used at a concentration of 0.008 per cent. in determining reliably amounts of magnesium down to 20 pg.696 KING, PRUDEN AND JANES: SYNTHESIS OF THE EXPERIMENTAL [Analyst, Vol. 92 REDUCTION OF COMMERCIAL TITAN YELLOW- A 10-g sample of Titan yellow (Eastman Kodak, P4454) was heated at 60" to 70" C with 250 ml of concentrated hydrochloric acid and 15 g of tin(I1) chloride for 1 hour, then cooled and mixed with 1 litre of water.The flocculent brown precipitate (about 0.45 g) was removed by filtration and the filtrate evaporated to dryness at 40" C in vacuo. The residue was suspended in water and evaporated to dryness several times to remove traces of hydro- chloric acid, then shaken with 50ml of water and filtered. The residue was washed with water until the washings gave no reaction for tin with ammonia solution. The orange prodiict became yellow when dried at 105" C. Yield of crude product, 2-41 g ; m.p. 298" C. CHARACTERISATION OF THE ACID -4s ITS AMMONIUM SALT- The crude reduction product (1.0 g), dissolved in 50 ml of hot 10 per cent.v/v ammonia solution, on cooling deposited crystals of the ammonium salt, m.p. 323" C. The amount of nitrogen found to be present by the Kjeldahl method was 12.08 per cent. and that calculated for C,,HllN,S~S0,NH4 was 12-46 per cent. Two further re-crystallisations from water, containing a trace of ammonia solution, raised the melting-point to 339" C. An ammonium salt prepared in the same way from authentic 2-(~-aminophenyl)-6-methylbenzothiazole- 7-sulphonic acid had m.p. 346" C, and the nitrogen content was found to be 13.40 per cent. The identity of the two samples was proved by- (i) a mixed melting-point, which showed no depression; (ii) identical absorption spectra, Em,,. 335 mp (water), I?:& = 1025 (Unicam SP500 spectrophotometer) ; (iii) chromatography. The two salts were applied to Whatman No.2 papers as 0.05 per cent. w/v solutions in 1 per cent. sodium hydroxide. R, (to centre of spot), in 80 per cent. ethanol,l0.55; in 6 per cent. acetic acid (first way) and s-butyl alcohol - acetic acid - water (14 + 1 + 5 v/v) (second way), 0.29, 0.29. Under ultraviolet light the comet-shaped spot fluoresced bright blue - violet; and The two salts were converted to their sodium salts, dried over phosphorus pentoxide in vacuo and dissolved in deuterium oxide (10 per cent. solution). The two spectra, which were identical, each contained three sets of peaks; a three-proton singlet, upfield of the water signal (aromatic CH,), a two-proton AB quartet ( J = 9 c/s) (two ovtho aromatic H's) and a four-proton AA'BB' system (four aromatic H's of a 9-disubstituted benzene ring), both downfield of the water signal.This spectrum confirms that the sulphonate group is indeed in the 7-position. PREPARATION OF THE ACID- The ammonium salt (1.0 g) in 50 ml of boiling water was acidified with glacial acetic acid. Orange crystals formed on cooling. Acids prepared in this way from the ammonium salts of authentic 2-(~-aminophenyl)-6-metliylbenzothiazole-7-sulphonic acid and from the reduction product of Eastman Kodak's Titan yellow, P4454, gave identical products, melting-point and mixed melting-point 319" C. The chromatographic behaviour of these two purified samples of the acid and of the blue - violet fluorescing component of commercial Titan yellowl was identical.The infrared spectrum of the acid as a Nujol mull showed a doublet in the 3400 cm-l region cliaracteristic of an NH, group (Perkin-Elmer 137 Infracord spectrophotometer). SYNTHESIS OF TITAN YELLOW- A solution of 5 g of 2-(~-aminophenyl)-6-methylbenzothiazole-7-sulphonic acid, 0.68 g of sodium hydroxide and 1.19 g of sodium nitrite in 150 ml of water, was treated dropwise, with stirring, with 30 ml of 20 per cent. w/v hydrochloric acid, and allowed to stand for 1 hour. The precipitated diazonium chloride was filtered off, washed with water and, while still moist, gradually added to a solution of 5 g of 2-(~-aminophenyl)-6-methylbenzothiazole-7-sulphonic acid and 1-36 g of sodium hydroxide in 150 ml of water. After 5 minutes, when the solid had dissolved, the mixture was warmed at 40" C for 30 minutes.The mixture was evaporated to dryness in vacuo, then dried at 105" C to give the crude product (9.8 g), which was purified (iv) nuclear magnetic resonance spectra (Perkin-Elmer R10 spectrometer).November, 19671 ACTIVE COMPONENT OF COMMERCIAL TITAN YELLOW 697 by exhaustive washing with hot acetone in a Soxhlet apparatus. This removed the yellow- fluorescing (ultraviolet) contaminants, i.e., those having small R, values in 80 per cent. ethanol, and most of the unchanged 2-(~-aminophenyl)-6-methylbenzothiazole-7-sulphonate, leaving a product, m.p. not less than 360" C, soluble in water, methanol and ethanol. Alterna- tively, the impurities could be extracted in cold acetone by ultrasonic treatment for 15 minutes.The purified product, identical with Band 2 of the commercial Titan yellow, as shown by paper chromatography and ultraviolet absorption spectrophotometry,l contained a negligible amount of sodium chloride. Its infrared spectrum showed no doublet in the NH, region, confirming that diazo-coupling had attacked the NH, group. In our new synthesis of Titan yellow almost all of the contamination of the commercial sample caused by the presence of sodium salts is avoided, and organic impurities are largely removed by washing the crude product with acetone. The compound most difficult to remove is dehydrothio-~-toluidine-7-sulphonate, which Bradfield6 had previously suggested as a possible component of commercial Titan yellow. The unusual diazo-coupling procedure that we have adopted, ie., adding the solid diazo- salt at room temperature, allows inorganic impurities to be removed, in contrast to the early commercial pr~cedure,~ in which diazotisation was carried out at temperatures up to 40" C and the amino-sulphonic acid, in glacial acetic acid, then coupled at 20" to 30" C for several days.The limited solubility of the sulphonic acid in glacial acetic acid restricts the scale of the preparation, a disadvantage avoided in our method. DETERMINATION OF MAGNESIUM- Synthetic Titan yellow produced a similar calibration curve to Band 2 of the commercial product,l showing a linear response from 0 to 160 pg of magnesium (Fig. 1) (Hilger & Watts Spectrocliem Mk 2 spectrophotometer). w Weight of magnesium, pg Fig. 1. Calibration for magnesium in silicate materials with synthetic Titan yellow We are now using our synthetic Titan yellow for the routine determination of magnesium in silicate minerals by the procedure described in our earlier paper.l 1 . 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES King, H. G. C., and Pruden, G., Analyst, 1967, 92, 83. Hall, R. J., Gray, G. A., and Flynn, L. R., Analyst, 1966, 91, 102. Kolthoff, I. M., Biochem. Z., 1927, 185, 344. -, Chern. Weekbl., 1927, 24, 254. Sandell, E. B., "Colorimetric Determination of Traces of Metals," Third Edition, Interscience Bradfield, E. G., Analytica Cham. Acta, 1962, 27, 262. Grcen, A. G., Bey. dt. chem. Ges., 1889, 22, 968. Ullmann, I?., Enzyklopudie der Technischen Chemie, 1915, 2, 64. Friedlander, I?. , Fortschr. TeerfarbFahr., 1887-1 890, 2, 296. Schubert, M., Justzts I-iebigs Annln Chem., 1946, 558, 23. Publishers Ltd., London, 1959, p. 591. Received June 22nd, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200695

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

698 Analyst, November, 1967, Vol. 92, $9. 698-700 Precise Location of Conductimetric End-points by a Simplified Least Squares Technique BY J. L. LATHAM AND E. C. LAWLEY (Departnzent of Chemistyy and Biology, Harris College, Preston) The least squares technique for the accurate location of conductimetric end-points is greatly simplified by using an odd number of uniformly spaced readings. This modified technique is particularly suitable for titrimetric determinations after correction for dilution has been made.. The over-all accuracy is found to correspond with that of the pipette used. THE instrumental methods that have been widely applied to the location of end-points in titrations are capable of giving a higlier degree of precision than is possible with a visual indicator. In particular, Granl has shown that, providing weight pipettes and burettes are used, it is possible to locate end-points to an accuracy of 0-002 per cent.by a potentiometric method. Higuchi, Rehm and Barnstein2 have developed a spectrophotometric method of end-point detection capable of achieving this degree of precision. However, conductimetric titrations do not appear to have been used for the precise location of end-points, although they are used in many routine analy~es.~ This is mainly because the accurate location of a conductimetric end-point is a laborious process. In the first place, the conductance values must be corrected for the change in volume caused by the addition of titrant. Secondly, if the best lines are estimated by eye after plotting the corrected points on graph paper, considerable errors are introduced.If the normal form of the least squares technique of line fitting is used a tedious calculation is required. It is not generally realised that the least squares method of line fitting takes a particularly simple form if the experimental results consist of an odd number of miformly spaced points. This condition can easily be satisfied in titrations, by running in equal volumes of the titrant, between each reading of conductance. SINPLIFIED LEtlST SQUARES FORMULA The best line of the form y = a -1- bx to fit a set of points (xl, yl), (x2, y2) ..... is given by the solution of the equations .. .. * - (1) . . . . . * (2) C%,?~, - bEx,* - a&, = 0 . . and Cy, - hCx, - 72a = 0 . . where 1.2 is the number of points used.the origin to the centre-point, Cx, = 0, the modified least squares equation becomes If an odd number of points, uniformly spaced along the x-axis is used, then, by changing . . .. - (3) Cy,-na=O .. .. . . .. . . (4). I;x,y, - bCx,2 = 0 . . .. The summations in equation (3) are further simplified if the amount of titrant added is In such a case, if 7 points are used Zx12 = 0 + 2(12 + Z2 + 3') = 28 and Consequently This technique is illustrated by the titration of 0.1 N hydrochloric acid with approxi- Solutions of analytical-reagent grade chemicals in No precautions were taken to exclude carbon exactly 1-00 ml. Ex, y , involves only multiplication by a small integer and simple additions. equations (3) and (4) can be rapidly solved without recourse to a calculating machine.mately 0.1 N sodium hydroxide solution. distilled water were used in the titrations. dioxide from the sodium hydroxide solution in the burette.LATHAM AND LAWLEY 699 Six conductimetric titrations were carried out by using a Wayne-Kerr B221 Universal Conductance Bridge, with platinised platinum e1ectrodes;and a cell constant of approximately unity. The titrations were carried out at room temperature but the change of temperature during the time of the titration was less than 0.1" C and so caused negligible changes in conductance. Observed conductance values were corrected for volume change caused by the addition of titrant, as shown in Table I. TABLE I CONDUCTANCE Before end-point VALUES f Titre 17.0 17.5 18.0 18-5 19.0 19.5 20.0 20.5 21.0 Conductance ------7 Corrected Observed for dilution 10.32 10.17 9.93 9.82 9-56 9.47 9.20 9.12 8.84 8-77 8.47 8-42 8.10 8.06 7.74 7-71 7.37 7.35 Titre 23.0 23.5 24.0 24-5 25.0 25.5 26.0 26-5 27.0 After end-point A \ Conductance I Observed 6.96 7.24 7.50 7.77 8.03 8.31 8-58 8-84 9-09 1 Corrected for dilution 6-98 7-27 7.54 7-82 8-09 8.38 8.67 8.94 9.20 The titrations were carried out by adding 358 ml of distilled water to 20 ml of 0-1 N hydrochloric acid in a 600-ml beaker, the contents of which were magnetically stirred.Nine conductance values were obtained at 0-5 ml intervals before and after the end-point with a Grade A burette. The conductance values for the first titration (expressed in millimhos) are given in Table I. Before the end-point the origin is taken at 19.0m1, then- From equations (3) and (4) Ex2 = 15, ZXY = -10.56, Ey = 78.89, n = 9.a = 8.766 b = -0.704. After the end-point the origin is taken at 25-Om1 then- Ex2 = 15, CXY = 8-35, Cy = 72.89, 9.t == 9. From equations (3) and (4) n = 8.099 b = 0.557. RECONVERSION TO OI~IGINAL AXIS value of x in the least squares lines must be replaced by x - x'. If the abscissa of the centre-point has a value x' on the original co-ordinates, then the Thus the two simplified least squares equations (3) and (4) become- JJ = 8.766 - 0-704(~ - 19) y = 8.099 + 0*657(~ - 25) .. .. . * (5) z.e., y = 22.142 - 0.704% . . .. a.e., y = -5.826 -+ 0.557~ . . .. .. .. . . (6). and Solving equations (5) and (6) gives x = 22.179. This titration was carried out 6 times and the values of the end-points obtained were 22.179, 22.204, 22.192, 22.197, 22.161, 22.181, having a mean value of 22.186 and a standard deviation of 0-013 ml.delivered. of 0.0052 ml. The standard deviation of the 20-ml pipette was found by weighing the quantity of water This was based on 12 values, giving a mean of 19.9179 ml and a standard deviation700 LATHAM AND LAWLEY The standard deviation of the Grade A burette (measured at 22 ml) was found to be These results show that virtually no error is introduced by Let U, = standard deviation in the transfer by pipette procedure, a, = standard devia- 0.011 ml (based on 8 readings). the procedure used for determining the conductimetric end-point . tion in the transfer by burette procedure and titration value volume of the pipette’ R = then the over-all standard deviation u in the combined operations of transfer by pipette and burette is given by Substituting numerical values in (7), u = 0.0124 ml for a normal combined transfer by pipette and burette operation corresponding to the end-point under discussion.This compares with u = 0.013 ml for the conductimetric end-point. These results suggest that if conductimetric titrations are carried out by using weight pipettes and burettes, precision comparable to that obtained by the previously mentioned methods of Gran and Higuclii shoulcl be attainable. If this were done, however, the simplified form of the least squares method could not be used. It should be emphasised that the procedure outlined above is valid only if the conductance of the solution varies linearly with thc volume of titrant added and the two lines obtained in the conductinietric titration intersect exactly at the equivalence point. The first condition requires that the observed conductance value should be corrected for dilution (as in Table I). Even with this correction, the second condition is onlv satisfied for reactions that go virtually to completion, such as the titration of strong acids with sodium hydroxide solution. REFERENCES o2 = R2oP2 + oB2 . . . . .. .. . . (7). 1. 2. 3. Gran, G., Acta Chem. S c a d . , 1950, 4, 559. Higuchi, T., Rehm, C., and Barnstein, C., Analyt. Chem., 1956, 28, 1506. Pungor, E., J . Electroanalyt. Chenz., 1962, 3, 289. Received May loth, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200698

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, November, 1967, Vol. 92, $$. 701-704 701 The Determination of Tertiary Hydroxyl Groups BY M. P. T. BRADLEY AND G. E. PENKETH (Imperial Chemical Industries Limited, Heavy Organic Chemicals Division, Research Department, Organic House, P.O. Box 2, Billingham, Co. Durham) A method of high specificity is described for the quantitative determina- tion of the tertiary hydroxyl group. It involves the use of the reaction with hydriodic acid to form the alkyl iodides, which are extracted into cyclohexane, hydrolysed and the resulting free iodine determined colorimetrically. THE presence of tertiary hydroxyl compounds has always given difficulty in schemes for functional-group analysis, because steric hindrance and a propensity for dehydration combine to give poor results in the general esterification methods for hydroxyl groups.Petrova and Novikoval utilised the dehydration reaction by refluxing with @-toluene- sulphonic acid and determining the resultant water with Fischer reagent, but this method is subject to interierence from many other compounds, and is only of limited applicability. Critchfield2 described a more useful method involving the reaction of hydrobroniic acid in glacial acetic acid to form the alkyl bromide, and titrating the excess of hydrobromic acid with sodium acetate in acetic acid. However, large quantities of water, or other alcohols, or both, interfere because they retard the rate of reaction and tend to be basic to the crystal violet indicator used. Scoggins and MilleI3 allowed the tertiary alcohol to react with hydriodic acid to form the alkyl iodide, forcing the reaction to completion by extraction into cyclohexane.The use of this solvent enables the alkyl iodide to be determined by ultraviolet absorption, and we have confirmed that this simple and elegant method gives good results for individual tertiary alcohols. Primary and secondary alcohols react slowly and do not interiere significantly, but serious interference arises from the presence of compounds that absorb in the ultraviolet region , and from olefinic compounds. Moreover, variations in the extinction coefficients and absorption maxima of the alkyl iodides make it a relatively imprecise method for the func- tional-group analysis of mixtures. A logical extension of this procedure is to hydrolyse the alkyl iodides and determine the liberated iodine, thus giving a stoicheiometric relationship between iodine and tertiary hydroxyl.Our work shows that this step is simple to carry out and has the additional benefit of improving the specificity. METHOD REAGENTS- Cyclohexane, spectroscopic quality. Hydriodic acid, 55 per cent. w / v , aqueous, analytical-reagent grade. Sodium hydroxide solution, 2 per cent. w/v, aqueous. Hydrogen peroxide , 30 $er cent. Potassium hydroxide solution , 30 per cent. w/v, aq~eous. Ethanol , absolute. Szclphuric acid, 10 per cent. v / v , aqueous. Potassium iodate, 4 per cent. w / v , aqueous. Potassium iodide-Weigh accurately 1 g of analytical-reagent grade potassium iodide which has been dried in an oven for 2; hours at 110" C, dissolve it in water and dilute the solution to exactly 1 litre with distilled water.1 ml = 1 mg of potassium iodide.702 BRADLEY AND PENKETH: DETERMINATION OF [Analyst, Vol. 92 PROCEDURE- Dissolve the sample, containing 2 to 200 mg of the tertiary alcohol, in about 20 ml of cycloliexane in a 250-ml separating funnel. Add 3 ml of the 55 per cent. hydriodic acid solution and shake the mixture for 5 minutes. Allow the phases to separate, rinse the stopper and walls of the separating funnel with about 10 ml of water, then discard the aqueous phase. Add 10 ml of 2 per cent. sodium hydroxide and 3 drops of hydrogen peroxide, then shake for 1 minute to destroy the free iodine. When the phases have separated, discard the aqueous phase, then wash the cyclohexane layer with two 10-ml portions of distilled water and discard the washings.Filter the cyclohexane solution through a silicone-treated filter-paper into a 150-ml conical flask. Rinse the separating funnel and filter-paper with two 5 ml-portions of cyclo- hexane. ,4dd to the flask 10ml of 30 per cent. potassium hydroxide solution and 10 ml of ethanol, then heat under reflux for 30 minutes. Allow to cool and wash the condenser with 20 ml of water, running the washings into the flask. Transfer the contents of the flask to a separating funnel, rinsing the flask with distilled water and adding the washings to the flask contents. The volume of the aqueous phase should be a t least 50 ml to ensure good phase separation. Separate the phases and discard the organic layer. Acidify the aqueous layer with 10 per cent. sulphuric acid, cool, and then add 3 ml of 4 per cent.potassium iodate solution. Extract the liberated iodine into 254 ml of cyclohexane and determine it spectrophoto- metrically at 525 mp, with l-cm cells. CALIBRATION PROCEDURE- The method may be calibrated directly with a pure tertiary alcohol or, more simply, by the following procedure with standard potassium iodide solution. Transfer by pipette 1, 2, 3, 4 and 5 ml of standard potassium iodide solution into 250-ml separating funnels and dilute with water to about 80 ml. Acidify with 2 ml of 10 per cent, sulphuric acid and add 20 ml of cyclohexane followed by 3 ml of potassium iodate solution. Shake the mixture for 1 minute to extract the liberated iodine into the cyclohexane phase.Separate the phases and discard the aqueous layer. Transfer the cyclohexane layer to a 25-ml graduated flask,’ rinsing each funnel twice with 2-ml portions of cyclohexane and finally diluting to 25 ml with cyclohexane. Measure the optical density of each iodine solution against the similarly prepared blank solution in 1-cm cells at 525mp. Plot a graph relating mg of potassium iodide to optical density; this should be a straight line passing through the origin. Combine the rinsings with the contents of the flask. 1 mg of potassium iodide -- 0.1024 mg of tert.-OH RESI~LTS Results of recovery experiments are given in Table I for 5 tertiary alcohols, which, for convenience, were dissolved in cyclohexane. For each alcohol at least 10 results were obtained.DETEKMINATIOI~ OF TERTIAKY -4LCOHOLS I N CYCLOHEXANE Concentration range in cyclo- Alcohol hexane, p.p.m. 2-Methylpropan-2-01 . . . . 310 to 2500 2-Methylheptan-2-01 . . . . 490 to 7300 3-Methyloctan-3-01 . . . . 560 to 11200 3-Methylbutan-3-01 . . .. . . 390 to 3900 2,4-Dirnethylpentan-2-01 . . . . 200 to 5000 Average per cent. per cent. Recovery rangc, recovery, ( 99.2 to 100.9 100.1 98.9 to 101.5 100.3 99.3 to 101.0 100.0 98.6 to 101.1 100.0 99.0 to 101.9 100.4 Coefiicicnt )f variation, per cent. 0.71 1-02 0-70 0.76 1-10 These results clearly show that the method is applicable over a wide concentration range with excellent recoveries and good precision. EFFECT OF PRIMARY AND SECONDARY ALCOHOLS- for tertiary alkyl iodides, and similarly the rate of hydrolysis of the iodides is slower.The rate of formation of primary and secondary alkyl iodides is much slower than that TheseNovember, 19671 TERTIARY HYDROXYL GROUPS 703 two factors combine to make interference from primary and secondary alcohols and glycols negligible, as the results in Table I1 show. TABLE I1 EFFECT OF PRIMARY AND SECONDARY ALCOHOLS Alcohol Ethanol . . .. .. . . Propyl alcohol . . .. . . Ethylene glycol . . . . . . 1,2-Propylene glycol . . . . Isopropyl alcohol .. . . 3-Methylcyclohexanol . . . . Pentan-3-01 . . . . .. Hexan-3-01 . . . . . . Octan-4-01 . . . . . . Hexan- 1-01 . . . . .. Octan- 1-01 . . . . . . . . . . . . . . . . . . . . . . . . . . . . Added, mg 1000 1000 588.0 549.5 385.3 407.3 395.3 444.3 404.3 127.5 160.0 Found as mg of equivalent tertiary alcohol Nil Nil Nil Nil Nil 0.5 1.2 Nil 1.9 Nil Nil INTERFERENCES As Scoggins and Miller3 pointed out, the two major sources of interference in their method are substances that absorb in the ultraviolet and are extracted into the cyclohexane, and olefins.In our method, the first of these is automatically removed by the subsequent extraction of the ionic iodide into water, and then of the iodine into cyclohexane. This double extraction also minimises the effect of coloured compounds, which might otherwise interfere with the iodine determination. We examined a variety of unsaturated compounds, and the results are given in Table 111. TABLE I11 REACTION OF UNSATURATED COMPOUNDS Substance Added, mg Oct-l-ene . . . . . . . . .. 609 Hepta-1,6-diene . . ... . . . 126 Hexa-2,4-diene-l-ol . . . . . . 341 Hexa-2,kdiene . . . . . . . . 104 Crotonaldehyde . . . . . . . . 65 Vinylcyclohex-l-ene . . ,. .. 170 Acrylic acid . . .. .. . . 191 * Formed a solid polymer containing iodine. Found as mg of equivalent alcohol 0.5 0.3 > 173* 2.1 0.2 Nil 5 There seems to be virtually no interference from norinal olefinic unsaturation, but conjugated dienes have a small, but significant, effect greatly enhanced by the reactive hexa- 2,4-diene-l-o1. Presumably even this interference could be removed by prior hydrogenation, as suggested by Scoggins and Miller.3 We have confirmed that other oxygenated groups, e.g., carbonyl, acid, ester and ether, do not interfere, a major exception being the tertiary alkyl hydroperoxides, which react similarly to tertiary alcohols and, in fact, can be determined quantitatively by slight modi- fications to the present method.Secondary alkyl hydroperoxides also liberate iodine from hydriodic acid but form only small amounts of alkyl iodide and, therefore, do not interfere seriously. ANALYSIS OF MIXTURES- The method was subjected to the rigorous test of determining small amounts of mixtures of various tertiary alcohols in the presence of substantial excesses of ethanol, isopropyl alcohol and cyclohexane. Results are given in Table IV, and show excellent recoveries.704 BRADLEY AND PENKETH TABLE IV ANALYSIS OF MIXTURES OF TERTIARY ALCOHOLS Components 2-Methylpropan-2-01 . . .. .. 2-Methylbutan-2-01 . . . . . . 2-Methylpentan-2 : 4-diol . . .. 3-Methyloctan-3-01 .. . . .. 2,3-Dirnethylpentan-3-01 . . . . Isopropyl alcohol . . . . .. Cyclohexane . . .. . . . . Calculated tertiary OH value .. Found tertiary OH value . . .. Ethanol . . . . .. .. Per cent. Per cent. 0.062 0.084 0.037 0.050 0.02 - 0.240 - 13.2 9.0 3.9 26-7 to 100 to 100 0.045 0,029 0.044 0.027 W/V w /v - - Per cent. Per cent. 0.017 0.144 0.040 0.167 0.054 0.118 0.01 1 0.095 0-105 - 3.6 4.3 - to 100 to 100 0.022 0.069 0.02 1 0.069 w/v W/V - Per cent. 0.02 1 w/v - - - 0.130 15.9 4.4 to 100 0-024 0.026 CONCLUSIOKS The method fills a distinct gap in the field of functional-group analysis, showing high specificity for the tertiary hydroxyl group and good accuracy over a wide range of concen- trations. Indeed, the range can be readily extended, at the lower end by using ultraviolet spectrophotometry for measuring small amounts of iodine, and at the higher end by using one of the standard titrimetric techniques for iodine or ionic iodide. REFERENCES 1. 2. 3. Petrova, L. N., and Novikova, E. N., Zh. Anulzt. IChim., 1957, 12, 411; Clzenz. &-lbstr., 1958, 52, Critchfield, F. E., “Organic Functional Group Analysis,” Pergamon Press Ltd., Oxford, J>ondon, Scoggins, M. W., and Miller, J. W., Analyt. Chem., 1966, 38, 612. 1860b. New York and Paris, 1963, p. 95. Received May 3rd, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200701

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, November, 1967, Vol. 92, $9. 705-710 705 The Determination of Butter Fat in Margarine Fat by Transesterification and Gas Chromatography BY D. F. WITHINGTON (County Analyst’s Department, County Hall, Durham) Margarine fat is transestcrified to form the ethyl esters of the fatty acids. The reaction mixture is examined by temperature-programmed gas chromato- graphy and the amount of butter fat calculated from the ethyl butyrate formed. THE Food Standards (Butter and Margarine) Regulations, 1955,l do not allow margarine fat to contain more than 10 per cent. of milk fat. Under the Labelling of Food (Amendment) Regulations, 1955,2 when a claim is made that a margarine contains milk, cream or butter, the milk-fat content must be declared in terms of the equivalent amount of butter.The declared percentage is allowed to differ from the actual percentage by a figure of 2. Butter fat is distinguished from most other fats by containing the glycerides of butyric acid. In the analysis of margarine, the presence of butyric acid can be regarded as specific for butter fat. Chemical methods of analysis usually depend upon the determination of butyric acid in one form or another. In this laboratory, the Hydroxamic Acid Jndex method of Bassette and Keeney3 has been adapted as a useful sorting test for the butter-fat content of various fats, but it is not suitable when the amount of butter fat is less than 10 per cent. The determination of butter fat in margarine fat is usually carried out by the method of Bolton, Richmond and R e ~ i s , ~ who used the steam-distillation method, now known as the Reichert - Polenske - Kirschner p r o ~ e s s , ~ to obtain a factor, the “Kirschner Value,” which is largely derived from the butyric acid content.The butter-fat content can be calculated from this value after making small corrections. Modifications of the method of calculation have been suggested, e.g., by Elsdon and Smith,6 and a useful small-scale method has been devised,’ but the original method is probably still the most widely used. I t has been obvious for some time that a gas-chromatographic technique must provide a fundamentally better approach to this problem. The method described below gives results similar to those obtained by the Reichert - Polenske - Kirschner process, but much more quickly and in far less working time.Fats are usually converted to fatty acid esters before examination by gas chromatograpliy, the methyl esters being most often prepared. The methods used are of two types involving- Saponification of the fat, followed by separation of the free fatty acids, which are then Transesterification, in which the fat is treated with tlie appropriate alcohol in tlie presence A special difficulty with butyric acid and its esters is that they are volatile, and to some extent water-soluble, so that methods that involve solvent evaporation or water-washing of reaction mixtures are not suitable. Transesterification has clear advantages for this purpose, as the reaction mixture can be examined directly by gas chromatography. esterified. of a catalyst, resulting in direct conversion into fatty acid esters.TRANSESTERIFICATION- sodium or potassium methoxide,gsl* ,11 sulphuric acid in methanol* and boron trifluorjde.12 Mason and Walker13 and Mason, Eager and Walker14 have described methods involving catalysis by acid and by sodium methoxide, in which dimethoxypropane was used to react with glycerol and take the reaction to completion. In this work a simple transesterification reaction in which ethanolic sodium hydroxide was used was found to be adequate. Transesterification has been carried out with potassium hydroxide in706 WITHINGTON DETERMINATION OF BUTTER FAT I N MARGARINE [AnahSt, VOl. 92 EXPERIMENTAL Preliminary experiments were carried out with butter fat transesterified under conditions similar to those described below, by using solutions of 0-7 per cent.of sodium hydroxide in methanol, 0-7 per cent. of sodium hydroxide in ethanol and 0.35 per cent. of sodium in ethanol. All of the chromatograms obtained were qualitatively satisfactory, but ethyl butyrate was found to separate more clearly than methyl butyrate from the leading ethanol peak. The chromatograms of ethyl esters prepared with sodium hydroxide and ethoxide solutions were virtually identical and all further work was carried out with the sodium hydroxide solution. Methyl hexanoate was chosen as internal standard because it forms a peak midway between ethyl butyrate and ethyl hexanoate. Injection of varying amounts of solutions of methyl hexanoate and ethyl butyrate indicated a linear relationship between peak height and concentrations for both.I t was found, however, that methyl hexanoate is slowly attacked in the cold transesterified solutions, forming ethyl hexanoate. Acidification with acetic acid in ethanol was found to stop the catalytic action and produce a stable solution. METHOD REAGENTS- Sodium hydroxide solution, 0.7 per cent. w / v in ethanol. Acetic acid solution, 6 +er cent. v I v in ethanol. Ethyl butyrate, redistilled and chronzatographically pure. Methyl hexanoate, 1.500 per cent. w / v in ethaizol. E t h a w l , absolute. APPARATUS- about 60 cm long. Test-tubes-25 x 150 mm, fitted with corks and air-condensers of 6-mm glass tubing, Wafer-bath. Gas chromatography-An Aerograph 204, with a linear temperature-programmer, flame- ionisation detector and 5-feet stainless-steel column containing 5 per cent.SE30 on 60 to 80 Cliromosorb W DMCS, was used. Recorder-A Leeds & Northrup "Speedomax W", with full-scale deflection of 1 mV, was used. PROCEDURE- Melt the margarine and pour off the fat through a filter-paper, ensuring that there is complete separation from any aqueous phase. Weigh accurately 3 k0.2 g of margarine fat into the bottom of a dry test-tube, fit the air-condenser loosely and place the tube in the water-bath at 77" to 80" C. After about 1 minute remove the air-condenser, add 3.0 ml of sodium hydroxide solution and replace the air-condenser firmly. Immerse the tube so that the level of the contents is just below the level of water in the bath. Shake it gently for about 1 minute until the phases merge, and leave for 15 minutes. Remove the tube from the bath, cool, and wash the air-condenser with 0.5 ml of ethanol, collecting the washings in the tube.Transfer the contents of the tube, washing the tube with small amounts of ethanol, into a 10-ml calibrated flask containing 0.7 ml of the acetic acid solution. Add exactly 1.0 ml of methyl hexanoate solution and make up to volume. Inject 6 pl into the gas chromatograph under the following conditions : injector port and detector-oven temperature, 200" C; flame- detector setting, 100; and attenuation x 4. Temperature-programme from about 55" C (setting 1-60) to 200" C at 8" C per minute with recorder speed of 15 inches per hour. The first peak, after the leading ethanol peak, is that of ethyl butyrate, followed by methyl hexan- oate.Measure the peak heights in scale-units and calculate the ratio of ethyl butyrate to methyl hexanoate. A typical chromatogram derived from a margarine fat containing 9.8 per cent. of butter fat is shown in Fig. 1. CALIBRATION- make up to volume with ethanol. add exactly 5 ml of methyl hexanoate solution to each and make up to volume. Weigh accurately about 1-3 g of ethyl butyrate into a dry 100-ml calibrated flask and Place 2,4, 6 and 8-ml aliquots into 50-ml calibrated flasks, Inject 6-plNovember, 19671 FAT BY TRANSESTERIFICATION AND GAS CHROMATOGRAPHY 707 9 I. Ethanol 2. Ethyl butyrate 3. Methyl hexanoate 4. Ethyl hexanoate 5. Ethyl octanoate 6. Glycerol Time, minutes 7. Ethyl decanoate 8. Ethyl laurate 9. Ethyl myristate 10.Ethyl palmitate I I . Ethyl stearate and ethyl oleate Fig, 1. Chromatogram of the reaction products formed by transesterification of fat from a margarine fat with a butter-fat content of 9.8 per cent., together with the internal standard, methyl hexanoate aliquots under the conditions described in the method. Calculate the peak-height ratios of ethyl butyrate to methyl hexanoate and plot these values against percentage of ethyl butyrate in the solutions. The points should fall on a straight line, from which the algebraic slope can be calculated. A typical calibration graph is shown in Fig. 2. Ratio, height of ethyl butyrate peak height of methyl hexanoate peak Fig. 2. Relationship between the amount of ethyl butyrate formed and the ratio of the peak height of ethyl butyrate to that of the internal stan- dard, methyl hexanoate708 CALCULATION OF ETHYL BUTYRATE FORMED- WITHINGTON : DETERMINATION OF BUTTER FAT IN MARGARINE [Analyst, Vol.92 Ethyl butyrate formed from margarine fat (as a percentage of the fat) = peak-height ratio x slope of calibration graph 10 weight of fat taken RATE OF REACTION- In an experiment on the rate of reaction, samples of a margarine and butter-fat mixture were taken from the water-bath at various times after the addition of sodium hydroxide solution. The times of reaction and amounts of ethyl butyrate formed, respectively, were : 4 minutes, 0.399 per cent. ; 8 minutes, 0.405 per cent. ; 31 minutes, 0-395 per cent. ; 55 minutes, 0.393 per cent.; and 65 minutes, 0.391 per cent. For practical purposes 15 minutes in the water-bath was found to be a convenient time, although from the above results this is obviously not critical.RELATION OF ETHYL BUTYRATE FORMED TO BUTTER-FAT CONTENT- The amount of butyrate in butter fat is subject to natural variation, and for calculating amounts of unknown butter fats by any method a fixed value must be taken. Nineteen butter fats (8 Danish, 3 New Zealand, 3 English, 1 Finnish, 1 Polish, 1 Australian, 1 French and 1 Irish) were examined to determine such a value for ethyl butyrate formed, The amount of fat taken, and the transesterification conditions used, were exactly as for the above method, but because of the much higher proportion of ethyl butyrate formed, the solutions were diluted appropriately with ethanol before examination by gas chromatography.O' ; Ib 115 ;o :5 Kirschner value 0 Fig. 3. Relationship between ethyl huty'rate formed by transesterification of butter fats and Kirschner values The Reichert - Polenske - Kirschner process was also carried out on these fats and the Kirschner values were plotted against the ethyl butyrate formed (Fig. 3). The ratios of the two values were: 4.97, 4-85, 5-16, 5.02, 4.82, 5.09, 4.96, 5.04, 5-19, 5-00, 5.15, 5.15, 4.85, 4.87, 4-67, 4.94, 4.73, 4.71 and 4.89, the mean ratio being 4-95, corresponding to a mean Kirschner value of 23-7. The value of 4-79 per cent. for ethyl butyrate was used as a fixed value in calculations to represent 100.0 per cent. butter fat. RE su LTS The results of analysis of known mixtures of three different margarine and three different butter fats are shown in Table I, together with results obtained from six different commercial brands of margarine declared to contain 10 per cent.of butter. The butter fat found fromNovember, 19671 FAT BY TRANSESTERIFICATION AND GAS CHROMATOGRAPHY 709 the Reichert - Polenske - Kirschner process was calculated according to the formula of Rolton, Richmond and Revis4 This is not strictly applicable for butter-fat contents above 10 per cent., but the figures have been included for comparative purposes. TABLE I COMPARISON OF BUTTER-FAT CONTENT O F MARGARINE FATS CONTAINING BUTTER FAT, AS DETERMINED B Y TWO METHODS Sample Margarine . . . . . . Margarine and butter . . Margarine and butter . . Margarine . . .. .. Margarine and butter .. Margarine and butter . . Margarine . . . . . . Margarine and butter . . Margarine and butter . . Margarine and butter . . Margarine and butter . , . . . . . . . . . . . . . . . . Margarine (10 per cent. butter Brand A . . .. . . Brand B . . . . . . Brand C . . . . . . Brand D . . . . .. Brand E . . . . . . Brand F . . . . . . declared) Butter fat in fat, per cent. Nil 7.1 10.9 Nil 4.0 11.0 Nil 6.5 8.5 14.7 25.2 Butter fat by gas chromatography, per cent., by using- ethyl butyrate fixcd value for formed from ethyl butyrate Nil Nil 7.3 6.9 11.0 10.5 S i l Xi1 3.7 3.7 10.8 10.9 Nil Nil 5.5 5.1 8.8 8.2 16.2 14.1 25-7 23-8 I n > known butter fat formed 9.2 9.4 10.0 9.5 9.4 10.0 Butter fat found from Reichert - Polenske - Kirschner process, per cent.Nil 7.8 11.3 Nil 4.5 11.8 Nil 6.1 8-9 14.9 23.9 10.2 10.5 9.9 10.0 10.0 11.2 REPEATABILITY- A margarine and butter-fat mixture was prepared containing 9.8 per cent. of butter fat, and 16 analyses were carried out on this mixture during the course of the investigation. The amounts of ethyl butyrate formed were 0.440, 0-448, 0.434, 0.432, 0-439, 0-442, 0-447, 0.434, 0.439, 0.432, 0-427, 0.443, 0.449, 0.452, 0.436 and 0.443 per cent., with a mean of 0.440 per cent., and a standard deviation of 0.007 per cent. (coefficient of variation = 1.6 per cent .) . DJSCUSSION The method gives results as good as those obtained by the Reichert - Polenske - Kirschner process, with a working time of only about 20 minutes per sample, or less, if several samples are examined together.About 45 minutes' running time is required for the gas chromato- grapliy and there was no apparent deterioration in the column, even after about 200 injections. In the examination of margarine a large amount of fat is usually available and, for convenience and accuracy, the method was based on the use of 3 g of fat. I t can easily be scaled down, however, and has been successfully applied to 0-3 g of margarine fat. Rutyrate is measured as ethyl butyrate because the expression of results in terms of butyric acid may cause confusion with results obtained by fully quantitative methods. Although the reaction cannot be completed under these conditions, the amount of conversion of butyrate into ethyl butyrate must be high, as the amount formed from the butter fats, calculated as butyric acid, covers a range of 3-4 to 3.9 per cent., with a mean of 3.6 per cent.; this is close to reported values for butyric acid in butter fat.15 Temperature-programming is not essential for the separation of the early peaks, which have been found to separate well by isotliermal operation at 70" C.I t is necessary, of course, to raise the temperature to about 200" C to clear the column of higher esters within a reasonable time. Although this work has been confined to the determination of butter fat, the chromato- grams obtained contain potentially much information on the constitution of the fat, as most The quantitative aspect of this has not been investigated.7 10 WITH IN GTO N of the fatty acid esters are clearly separated. This is especially true of the esters of the characteristic fatty acids of the coconut-oil class, i.e., those containing C atoms C,, C,, C,, and C12.The octanoic (CJ peak can be easily obtained unobscured by washing with water to remove glycerol. Finally, there seems no reason why the method should not be applied to the determination of butter fat in any mixture of fats, q., from butter confectionery. I thank Mr. F. C. Shenton, Durham County Analyst, for permission to publish this paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Food Standards (Butter and Margarine) Regulations, 1955, No. 1899, H.M. Stationery Office, Labelling of Food (Amendment) Regulations, 1955, No. 1900, H.M. Stationery Office, London. Bassctte, R., and Keeney, M., J . Ass. OH. Agric. Chem., 1956, 39, 469. Bolton, E. R., Richmond, H. D., and Revis, C., Analvst, 1912, 37, 183. Analytical Methods Committee, Ibid., 1936, 61, 404. Elsdon, G. D., and Smith, P., I b i d . , 1925, 50, 53. Dyer, R., Taylor, G., and Hamence, J . H., Ibid., 1941, 66, 355. Wolff, J . P., rinds. Falsif. Fraudes, 1960, 53, 318. Craig, B. M., and Pvlurty, N. L., J . A m e r . Oil Chem. SOC., 1959, 36, 549. Luddy, F. E., Barford, K. .\., and Riemanschneider, K. TV., I b i d . , 1960, 37, 447. tleillan, J , M., J . Dairy Sci., 1964, 47, 546. Mctcalfe, L. P., Schmitz, A. A., Pelka, J . K., Ancllyt. Chem., 1966, 38, 615. Mason, 31. E., and Walker, G. R., Ibid., 1964, 36, 583. Mason, 34. E., Eager, XI. E., and Walker, G. R., Tbid., 1964, 36, 587. Bailey, A. E., “Industrial Oil and Fat Products,” Second Edition, Intersciencc Publishers Inc., Received Tune 2nd, 1967 1,ondon. New York and London, 1951, p. 127.

ISSN:0003-2654    

DOI:10.1039/AN9679200705

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

Analyst, November, 1967, Vol. 92, 99. 711-713 'ill Thin-layer Chromatography of Neutral Drugs BY P. E. HAYWOOD, MARIAN W. HORNER AND H. J. RYLANCE* (Forensic Laboratory, Equine ResearcJh Station, Newmarket, Sujjfolk) A two-way thin-layer system is described for the partial identification of neutral drugs, followed by elution and confirmation by a third RP value. HF values and standard deviations are tabulated, together with sensitivity of detection with ultraviolet light, and chlorine - starch - potassium iodide. IN the screening of body fluids for the presence of drugs, the "neutral" class presents special problems because compounds in this class possess diverse structures without common func- tional groups. Location reagents for use on paper chromatograms that assist in character- isation have been reported,l but they mostly have a destructive effect, which may present ;I serious objection when the sample for analysis is very small.The reaction with clilorine, followed by a starch - potassium iodide spray, is the most useful for location of these drugs, and its use has been reported for a variety of co~npounds.~~~ Tlie disadvantage of this reaction is that the period of time elapsing between chlorination and the treatment with starch - potassium iodide is highly critical for paper chromatograms, and some non-nitrogenous neutral compounds, e.g., mephenesin, do not react on paper systems. In the work reported here we have investigated the use of a two-way thin-layer system on a fluorescent thin-layer powder, and locating ultraviolet-absorbing spots by examination under an ultraviolet source (Hanovia Chromatolite, 254 mp), followed by exposure to chlorine and starch - potassium iodide. Tlie time interval between these two treatments is far less critical than for paper systems.After location bv this method it has been found possible to elute the drugs and re-run in a third solvent system. Partial identification on the two-way system may thus be confirmed by a third R, value and a more specific (though destructive) location reagent.1,4y5 The two-way system used is that reported for tlie identification of corticosteroids in horse urine,6 which are also extracted in tlie neutral group; the method is, therefore, potentially capable of being incorporated into a comprehensive scheme for neutral drugs and corticosteroids.PROCEDURE- Plates (20 x 20 cm) are prepared from Kieselgel G F 254 (Merck), 0-25 mrri thick, and activated for 1 hour at 110" C. Neutral group extracts,l dissolved in methanol, are applied to a two-way plate and run in ethyl acetate (redistilled). The plates are dried for not less than 20 minutes in an air-stream, rotated through 90", and run in dioxan - InethJ'lene chloride - water (1 +2+ 1, allowed to equilibrate, and the aqueous phase discarded) . 7 ,;2fter drying, the plates are examined under ultraviolet light (Hanovia Chromatolite, 254 mp) and absorbing spots marked. Thev are then exposed for a few seconds to chlorine gas, left in air for about 15 minutes and sprayed with 2 per cent. starch solution containing 1 per cent. of potassium iodide.Spots, which appear blue on a white background, are scraped from the plate and eluted into 2 ml of methanol by shaking mechanically for 10 minutes, followed by spinning in a centrifuge. The supernatant solution is separated and evaporated in a vacuum-oven at 40" C, dissolved in methanol and run in chloroform - acetone (941). * Present address : Rheumatic Diseases IJnit, Northern General Hospital, Ferry Road, Edinburgh 5.712 HAYWOOD, HORNER AND RYLANCE: THIN-LAYER [Analyst, VOl. 92 TABLE I R, VALUES AND STANDARD DEVIATIONS FOR NEUTRAL DRUGS RUN IN THREE SOLVENTS Dioxan - methylene Chloroform - acetone Ethyl acetate (redistilled) chloride - water (1 + 2 + 1)t (9 + 1) - Number Mean Drug of runs RF *Amidopyrine . . *Bemegride . . *Benzocaine .. *Bromvaletone. . *Caffeine . . *Carbromal . . Ethyl butyryl- *Ethyl crotonyl- "Glutethimidc . . *l\lephenesin . . *Mephenesin carbamate . . Meprobamate Methylpentynol carbamate . . Methyprylone. . Pentylene tetrazole . . *Phenuronc . . *Sedormid .. Sedulon .. *Styramate . . Urea . . . . urea . . . . urea . . . . 10 25.0 9 62.9 10 69-5 9 62.4 10 15.9 9 68.2 10 55.8 10 53.5 9 68.4 9 47.9 9 50.4 9 48.2 9 65.7 9 39-7 9 22.8 9 53-5 9 63.2 9 31-1 9 50.4 10 1-95 Standard deviation 3.7 3.3 2.7 2.6 2.4 2-3 Number of runs 10 9 10 9 10 9 3-2 10 2.6 10 3.4 9 2.8 10 3.9 9 3-1 9 4.8 9 3-1 9 3.4 10 2.5 10 3.6 10 2.9 9 1.7 9 0.74 10 Mean RF 67.8 92.6 93.6 79.1 56.3 89.4 72.9 73.0 95.6 63.2 58.3 51.3 88.3 72.4 54.6 64.3 80.8 66.6 52.1 1-71 Standard deviation 4-9 5.1 3.2 3.7 3.8 3.6 4.4 3.2 2.0 3-1 7.0 4.6 4.1 1.7 2.7 3.4 5.3 4.7 6.1 0.75 Number of runs 10 9 10 9 10 9 10 10 9 10 9 9 9 9 10 10 10 9 9 10 Mean RP 25.6 40.3 54.4 2 6 4 19.4 47.1 22.8 23.1 55-1 15.7 10.0 6.4 39.6 26.9 12.7 15-2 27.1 22.6 7.8 0.35 Standard deviation 3.2 1.8 2.2 2.2 1.9 2-4 2.1 2.0 4.3 2.9 1-9 1-6 2.9 3.1 1.6 1.6 1-7 3.9 1.6 0.36 * Drugs absorbing ultraviolet light (wavelength 254 mp) .t Allowed to equilibrate and aqueous phase then discarded. Table I shows the R, values and standard deviations determined on the pure solutions of twenty neutral compounds extractable into ether or chloroform, or both, from either acidic or alkaline aqueous solutions. The compounds were each applied to a plate in amounts of 0-5, 1, 3, 5, 10, 15, 25, 50 and 100 pg and developed in ethyl acetate; the smallest of these amounts detected is shown in Table 11.Some variation in sensitivity can be expected according to the size of the spot. TABLE I1 SENSITIVITY OF DETECTION OF NEUTRAL DRUGS AFTER THIN-LAYER CHROMATOGRAPHY Approximate limit of detection,* pg, by- -3 starch - absorption potassium Drug a t 254 mp iodide Drug Approximate limit of detection,* pg, by- chlorine and starch - absorption potassium a t 254 mp iodide Amidop yrine . . 0.5 Bemegride . . . . 10 Benzocaine . . . . 0.5 Bromvaletone . . 10 Caffeine . . . . 0.5 Carbromal . . . . 10 Ethyl butyrylurea . . t Glutcthimide . . 10 Mephenesin . . . . 3 Ethyl crotonylurea 3 0.5 0.5 0.5 0.5 0.5 0.5 0-5 0.5 0.5 3 Pvlephenesin carbamate . . 15 t Meprobamate .. .. Methylpentynol carbamate .. t t Methyprylone .. . . Pentylene tetrazole . . .. Phenurone . . . . .. Sedormid . . .. .. 100 . . t Sedulon . . .. Styramate . . . . .. 15 t Urea . . .. . . .. t5 15 0.5 0.5 0.5 0.5 0-5 0.5 0.5 100 16 * The smallest amount tested was 0.5 pg; detection limits in some instances are probably less. t Not detected in amounts of 100 pg or less.November, 19671 CHROMATOGRAPHY OF NEUTRAL DRUGS 713 R, values of eluates following chlorine and starch - potassium iodide treatment have been determined at least five times for each compound. Some decomposition occurs during this treatment, and eluted compounds run on a third system frequently yield multiple spots. The starting material is generally much stronger than any other spots. The authors thank Mr. M. S. Moss, Director of the Forensic Laboratory, for helpful technical discussions. This work was undertaken as part of a programme of research financed by the Horserace Betting Levy Board, to whom we are grateful. REFERENCES 1. Jackson, J. V., and Moss, M. S., in Smith, I., Editor, “Chromatographic and Electrophoretic Techniques, ” Volume 1, Second Edition, Heinemann Medical Rooks Ltd., London, 1960, pp. 404-40 8. 2. 3. 4. 5. 6. 7. Reindel, F.] and Iloppe, W., Chem. Ber., 1954, 87, 1103. Grieg, C. G., and Leaback, D. H., Nature, 1960, 188, 310. Moss, 31. S., and Jackson, J. V., J. Pharm. Pharmac., 1961, 13, 361. Olesen, 0. V., Acla Pharmac. Tox., 1966, 24, 183. Moss, M. S., and Rylance, H. J., J . Pharm. Pharmac., 1966, 18, 13. Hall, A., Ibid., 1964, 16, 9T. Received February 20th, 1967

ISSN:0003-2654    

DOI:10.1039/AN9679200711

出版商:RSC    

年代:1967

数据来源: RSC

摘要:

714 Analyst, November, 1967, Vol. 92, pp. 714-716 Quantitative Determination of Silyl Derivatives of Glucose by Gas - Liquid Chromatography with Inert Internal Standards BY Y. HALPERN, Y. HOUMINER AND S. PATAI (Department of Organic Chemistry, The Hebrew University, Jerusalem, Israel) Trimethylsilyl derivatives of cc- and P-D-glucose were determined quantitatively, with an internal standard such as terphenyl or triphenyl- ethylene. The results are reproducible to about 5 1 per cent. As glucose is one of the most widely investigated carbohydrates, a rapid and accurate method for its quantitative determination, especially one suitable for simultaneous determination of the two anomers, is of great interest. Most recent methods are based on gas - liquid chromato- graphy of the trimethylsilyl ether d e r i ~ a t i v e s l , ~ , ~ with an internal standard (IS) such as inositol,2 ~ o r b i t o l , ~ or methyl or.-D-galactopyranoside,4 which in itself also undergoes O-tri- methylsilylation.These introduce unnecessary complications and the different silylation rate of a hydroxyl-containing standard from that of the substance being analysed may influence the results, e.g., we found that the low solubility of inositol in pyridine affected strongly the degree of its silylation, giving different peak areas for different reaction times. Hence we investigated the use of inert internal standards that undergo no changes during the entire procedure, and fulfil all the necessary req~irements.~ EXPERIMENTAL REAGENTS- Hexaunet~a~ldisilnzane. TrinzethylchlorosilaIze. Terp heny I .cc-D-Glztcose-Analytical-reagent grade, [cc] gJ = + 110.9". P-u-Glucose-Pure, [a]$ = -1 20.3. Tri~lzeiz_z~leth~11e~ze-Tliis was prepared as described by Adkins and Zartman .G APPARATUS- An Aerograph 600 C gas cliromatograph (Wlkens Instrument and Research Inc.) with a hydrogen flame-ionisation detector, and a Honeywell recorder equipped with a disc-chart integrator, were used. A 5 feet x & inch o.d., coiled stainless-steel column, packed with 5 per cent. SE 30 Chromosorb W, 60 to 80 mesh, (Wilkens Instrument and Research Inc.) was used. OPERATING CONDITIONS- terphenyl, and at 320" C with triphenylethylene. flow-rate of 25 ml per minute. 300ml per minute, Vnder these conditions, total retention time was about 20 minutes. The column temperature was maintained at 190" C, and the injector at 280" C with Helium was used as a carrier gas at a Hvdrogen flow-rate was 25 ml per minute and air flow-rate, Samples of 0-5 to 1.Op1 were injected with a 1+0-,ul Hamilton syringe.HALPERN, HOUMINER AND PATAI 715 SAMPLE PREPARATION- dry 210 was A stock solution of 1.66 g of triphenylethylene or of 1.25 g of terphenyl in 100 ml of pyridine was prepared.A weighed amount of cc- or P-u-glucose, or both (a total of 50 to mg), was added to the pyridine solution of terphenyl or triphenylethylene and silylation carried out according to Sweeley’sl procedure. On addition of the hexamethyldisilazane, or the trimethylchlorosilane, a slight precipitate The experimental results proved that the sample, nevertheless, con- sometimes appeared.tained the full concentration of both the silylated products and the internal standard. CALIBRATION PROCEDURE- D-Glucose, containing various weighed amounts of the anomers in different experiments, was silylated in the presence of varied amounts of terphenyl or triphenylethylene. Calibration curves were obtained by plotting S,/SIS or S,/SIs values against mg,/mg,, or mgp/mgIs, respectively, where S values are the corresponding peak areas determined by the disc chart integrator, and mg values are the weight in milligrams. The straight lines thus obtained were used for quantitative determination of cc- and /I-D-glucose in unknown samples. I t is recommended that for each series of analyses, and at least every day, a new cali- bration curve should be prepared (or at least 4 or 5 points), and the analyses be performed consecutively.This was done for each of the three horizontal sections shown in Table I. TABLE I GAS - LIQUID CHROMATOGRAPHIC ANALYSIS OF a- AND JQ-D-GLUCOSE MIXTURES, BY [JSING EITHER TRIPHENYLETHYLENE OR TERPHENYL AS INTERNAL STANDARD D-Glucose added, mg Pyridine solution of & internal standard, U P ml 30 30 Triphenylethyl- 5 40 40 ene, 5 50 50 5 60 60 5 40 40 10 50 50 10 D-Glucose found, mg -7 0.590 0.467 30.5 28.5 0-766 0.657 39.5 40.0 0.972 0.818 50.0 50.0 1.154 0.965 59.6 59-0 0.393 0-346 40.4 42.2 0.498 0.428 51.3 52.2 .S,/Srs Sp/S1s U B 50 50 Terphenyl, 2 3-050 2.800 50.9 52.4 50 50 4 1.524 1-319 50.8 48.3 50 50 6 1.006 0.9 17 50.0 49.1 50 50 8 0.747 0.682 49.5 49.3 50 50 10 0.629 0.596 51.9 52.7 50 0 Terphenyl, 2 2.747 0.092 51.0 1.5 40 10 2 2-127 0.500 39.4 9.7 35 15 2 1-958 0.770 36.3 15.0 :I 0 20 2 1.654 0.967 30.8 19.1 25 25 2 1.341 1.316 24-3 25.8 20 30 2 1.074 1.529 19.9 30.0 15 35 2 0-902 1.850 18-7 36.3 10 40 2 0.589 2.049 10.8 40.4 0 50 2 0.040 2.476 0.6 48.8 * For each 0.01 g of glucose, 0.1 ml of trimethylchlorosilanc and 0.2 ml of hexamethyldisilazane were used in all procedures. RESULTS AND DISCUSSION Fig.1 shows typical chromatograms of cc- and P-D-glUcose with triphenylethylene and terphenyl as internal standards. The results of the determinations were well reproducible (to about 1 1 per cent.) and the same result was obtained in a second analysis on the same silylated sample, that was carried out after the latter was left standing in a well closed bottle for 3 days.As expected, the best results were obtained when the ratio of the peak areas of the internal standard and of the substance was near unity (see Table I). We believe that mutarotation occurring during the silylation process1 is negligible compared with the other errors.716 HALPERN, HOUMINER AND PATAI Time, minutes Fig. 1. Gas-chromatographic analysis of ct- and p-D-glucose anomers by using (u) triphenylethylene, or (b) terphenyl as internal standards Other polycyclic, high boiling, inert substances, such as chrysene and pyrene, were also successfully used as internal standards in quantitative determinations of D-glucose and may possibly prove to be more suitable than triphenyletliylene or terphenyl for the determination of other sugars. This research has been financed in part by a grant made by the United States Department of Agriculture. REFERENCES 1. 2. 3. 4. 5. 6. Sweeley, C. C., Bentley, R. Makita, M., and Wells, W. LV., J . Amer. Chem. SOC., 1963, 85, 2497. Brower, H. E., Jeffery, J . E., and Folsom, M. W., Analyt. Chem., 1966, 38, 362. Alexander, R. J., Garbutt, J. T., Ibid., 1965, 37, 303. Richey, J. M., Richev, H. G., jun., and Schraer, K., Analyt. Biochem., 1964, 9, 272. Dal Nogare, S., and Juvet, R. S., “Gas - Liquid Chromatography,” Interscience Publishers Inc., Adkins, H., and Zartman, W., i n Blatt, A. H., Editor, “Organic Syntheses,” Collective Volume 11, Received Januayy 23rd, 1967 New York, 1962, p. 256. John Wiley & Sons Inc., New York, 1943, p. 606.

ISSN:0003-2654    

DOI:10.1039/AN9679200714

出版商:RSC    

年代:1967

数据来源: RSC