ISSN:0003-2654    

DOI:10.1039/AN95378FX047

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BX049

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BP123

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

OCTOBER, 1953 THE ANALYST Vol. 78, No. 931 PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS HONORARY ASSISTANT SECRETARYSHIP THE Council has appointed Mr. N. L. Allport, Member of Council, to be Honorary Assistant Secretary of the Society. In this capacity, Mr. Allport will act as a link between the Council of the Society and the Committees of the Sections and Groups. SECRETARYSHIP OF THE SOCIETY WE have pleasure in announcing the marriage of Miss D. V. Wilson, Secretary of the Society, to Mr. J. P. Hicks. NORTH OF ENGLAND SECTION AN Ordinary Meeting of the Section was held at 2 p.m. on Saturday, May 2nd, 1953, at the City Laboratories, Mount Pleasant, Liverpool, 3. Mr. A. A. D. Comrie, B.Sc., F.R.I.C., presided over an attendance of fifty-eight .Mr. W. Gordon Carey, F.R.I.C., and Mr. J. G. Sherratt, B.Sc., F.R.I.C., introduced a comprehensive discussion on “The Analysis of Waters, Sewages and Effluents” to which a large number of members contributed. Tm Sixteenth Summer Meeting of the Section was held at the Imperial Hotel, Llandudno, from Friday, June 12th, to Monday, June 15th, 1953. The Vice-Chairman, Mr. J. R. Walmsley, A.M.C.T., F.R.I.C., Ph.C., presided and thirty- eight members attended, including the President, Dr. TI. W. Kent- Jones, F.R.I.C., and Mrs. Kent- Jones. On the morning of Saturday, June 23th, a paper was presented by Mr. C. A. Adams, C.E.E., B.Sc., F.R.I.C., on “Random Reflections on Food Legislation’’ (see below). A tour of the district was made by motor coach on the Sunday afternoon. SOME RANDOM REFLECTIOXS ON FOOD LEGISLATION Mr.Adams said that the new Food and Drugs Bill now being drafted for presentation to Parliament was a matter of outstanding importance to Public Analysts and chemists engaged in food manufacture, but it was also of importance to the general public, as buyers and consumers of food. On the interest shown by the general public much of the success of the Bill would depend, and it was to be hoped that members of the Society and the general public would do their utmost to stimulate this interest-for legislation seldom ran ahead of public opinion. The outstanding interest in the new Bill would be in the powers sought to transfer temporary war-time legislation into permanent forin and in any additional powers that might be necessary to deal with the addition of chemicals to food, whether intentional in the course of manufacture, or adventitious in the form of pesticide residues or in any other way.These matters had receutly been the subject of official enquiry in the United States of America. They had also been dealt with at length by Dr. J. R. Nicholls in his recent Presidential address, and were summed up from the official angle in the Annual Report of the Advisory Council on Scientific Policy for 1950-51 to the Lord President of the Council. Mi-. Adams said that in his view the main value of a change in the law would be in relation to the use of chemicals newly introduced into the food industry. 569570 OBITUARY [Vol. 78 If the Orders under the Defence (Sale of Food) Regulations, 1943, were included in the new Bill, it would make them unalterable without appeal to Parliament. Instead, fuller powers to make such Orders would be sought in the Bill, and the Orders could then be re- enacted. Amongst these Orders the speaker allotted pride of place to the Labelling of Food Orders and the Food Standards Orders. All these Orders had been enforced by the Food and Drugs Authorities as though they were part of the Food and Drugs Act. As the result of the experience in the working of these Authorities and food manu- facturers, he was of the opinion that we had seen some welcome progress in what might be called the “co-operative administration of food legislation.” If the range of the Food and Drugs Act was to be extended to cope with problems arising from the use of chemicals, pesticide residues and the sale of sub-standard products, now was the time to make recommendations. Once the Bill became an Act, further changes might be long delayed.

ISSN:0003-2654    

DOI:10.1039/AN9537800569

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

570 OBITUARY [Vol. 78 Obituary JOHN ROBERT STUBBS JOHK ROBERT STUBSS, M.Sc., F.R.I.C., who passed away on April 17th, 1953, in his 73rd year, was educated at Warton Schools, Winsford and Witton Grammar School, Northwich. From Witton he proceeded by scholarship to Liverpool University, where he obtained his BSc. degree in 1900 and his M.Sc. degree in 1903. He became an assistant in the laboratory of the late Dr. Campbell Brown, then Public Analyst for the County of kancaster, in 1901. Under a succession of County Analysts, he continued to serve faithfully the Lancashire County Council and was eventually himself appointed County Analyst and Official Agricultural Analyst in 1938-a promotion he richly merited. He was a most conscientious and painstaking analyst and these qualities are well exemplified in the many papers, some in collaboration with the late Dr.Elsdon, that he contributed to The Analyst. He will always be remembered for his fundamental work in connection with the deter- mination of extraneous water in milk by means of the freezing-point, work that the writer has good reason to know involved considerable physical effort and concentration in the early stages. The pages of The Analyst also contain examples of his researches in other fields of analytical work. During the first world war he served in the K.A.O.C. for three years in charge of one of the two Schools of Ammunition in France, returning with the rank of Captain. In 1929 he became Honorary Secretary of the North of England Section of the Society, the Section being then in its infancy.He continued in that office until 1940, when he was appointed Chairman, a position he occupied for three years. Both the Section and the Society must be ever deeply indebted to him for the wholehearted and splendid service he rendered during this period of 14 years. This service included the initiation and organisa- tion of the highly successful pre-war Summer Meetings of the North of England Section of the Society, which began at Scarborough in 1930 and have contributed, by their sociable nature, so much to the personal friendships of professional colleagues. Although what were known, in the County Laboratory, as “ersatz” Summer Meetings were held in Manchester during the late war, it is a matter for regret that Stubbs’ retirement from the Chairmanship prevented him from presiding at a normal Summer Meeting, to the success of which he had so largely contributed. The Society was represented by the writer. ARNOLD LEES He was laid to rest at Weaver Cemetery on April 22nd, 1953.

ISSN:0003-2654    

DOI:10.1039/AN9537800570

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

October, 19531 IRVING AND BUTLER 57 1 A Reversion Method for the Absorptiometric Determination of Traces of Lead with Dithizone BY H. M. IRVING AND E. J. BUTLER (Presented at the meeting of the Society o n Wednesday, A p r i l Ist, 1953) A method is described for the determination of lead in small samples of biological materials. The procedure involves wet ashing of the sample, extraction of all the lead (and any bismuth present) with dithizone, a preliminary separation from the bulk of the bismuth, and a final absorptio- metric determination of lead by the ‘‘reversion’’ technique. Quantities of lead in the range 0 to 2 pg can be determined satisfactorily in the presence of up to 20 pg of bismuth and at least 1OOpg of copper and zinc. ALTHOUGH it seems to have no particular function, lead is invariably present in traces in most fluids and tissues as an “inevitable consequence of life on a lead-bearing planet.”l In slightly larger amounts it is toxic, perhaps by competing with essential trace metals for the metal-binding components of enzyme systems.It has been suggested that lead may be a factor in the etiology of certain degenerative diseases of the nervous system, such as disseminated s c l e r o ~ i s , ~ ~ ~ and in the course of investigating this hypothesis the need was felt for a more sensitive, specific and convenient analytical method than any that was already available, particularly for samples that are necessarily limited in quantity, such as cerebro- spinal fluid, specialised neurological tissue and biopsy specimens.The use of dithizone for the absorptiometric determination of lead has been extensively studied, and even in the presence of other metals, e.g., zinc, copper, iron and so on, the procedure can be made almost specific if the dithizone complex is extracted from an alkaline citrate buffer containing cyanide. Bismuth, tin’’ and thallium are to some extent co- extracted and estimated together with the lead.4,5 If a one-colour procedure is used for the subsequent absorptiometric determination of the red coloured lead - dithizone complex, Y ~ D z , , ~ for which E = 68,600 in carbon tetrachloride at An;ax. of 520 mp,7 errors may arise if the “stripping” of excess dithizone (which absorbs appreciably at this wavelength) is incomplete, or is carried out so efficiently as to cause partial decomposition of the lead dithizonate.8 The same criticisms apply to procedures in which the absorptiometric measurement is actually made on the equivalent amount of dithizone liberated from this stripped lead dithizonate by treatment with dilute a~id.~,lO If, however, the determination of lead is made to depend upon measurements of the absorbancy of lead dithizonate a t 520 mp, or on that of unreacted dithizone, a known excess amount having been added originally, the initial concentration of this dithizone (E = 34,600 in carbon tetrachloride at A,,,.620 mp’), the pH of extraction and the phase ratio, which determine the extent of its partition into the aqueous phase, must be rigorously controlled. The difficulty of preserving solutions of dithizone of constant strength is well known, and although such variations in concentration are of no importance in the procedure described by Kozellta and Kluchesky,ll in which absorbancies are measured at two selected wavelengths, an enormous amount of work is involved in preparing the necessary families of calibration curves.Moreover, these curves (like those for the other procedures mentioned above) would be invalidated by the co-extraction of other metal dithizonates or other coloured materials from the biological digest; the latter often appear on basification when nitric acid has been used to oxidise samples containing much fat. Most of these objectionable features are removed by the process of “reversion.”6 If A, is the absorbancy measured in a 1-cm cell at any wavelength of a mixed-colour extract containing molecular concentrations C, of lead dithizonate, PbDz,, and C, of excess dithizone, HDz, and Ci of coloured impurities, of which the molecular extinction coefficients are E,, eT and ei, respectively, we have- ..(1) A, = E,C, + E,C, + I=E& . . .. .. ..572 IRVING AND BUTLER: A REVERSION METHOD FOR THE ABSORPTIOMETRIC [Vol. ’is If now a part or all of this organic phase is shaken with an aqueous reagent that decomposes only the lead dithizonate, permitting a return of lead ions to the aqueous phase while an equivalent amount, 2C,, of dithizone appears in the organic layer (reversion), the absorbancy measured at the same wavelength will increase to a new value, A,, given by- A,. = E,C, + 2 4 , + CciCi .. . . The increase in absorbancy, termed the reversion value, R, is then given by- R = A, - A, = (26, - E,)C, - . .. . . .. .. (‘1 The concentration of lead is thus proportional to the reversion value, and the sensitivity of the method will be greatest at that wavelength where (26, - E ~ ) is greatest; for the lead - dithizone system 620 mp is appropriate. Although significant amounts of thallium are not normally present in biological fluids and tissues, bismuth is a common constituent of medicinal preparations, some of which may be prescribed for the treatment of disseminated sclerosis, and interference from this element may be serious in the analysis of specimens from diseased subjects. When deter- mined under equilibrium conditions, the curve relating the percentage extraction of bismuth to the pH value of the aqueous phase lies to the left of that determined for lead under com- parable conditions of reagent concentration and buffer composition, i.e., its extraction is quantitative at a lower pH value.As the concentration of the dithizone solution is increased, the separation of these curves increases, since the extractability of lead depends on the square, whilst that of bismuth depends on the cube of the excess dithizone concentration.12713 Various workers have attempted to separate bismuth from lead by carrying out extractions from acid media (in the pH range of 2 to 4), when bismuth is preferentially extracted,*7l2 but inevitably some lead is co-extracted and lost. If, however, extraction is made from an alkaline citrate - cyanide buffer under optimum conditions for the removal of lead, all of the bismuth is co-extracted.Although treatment of the organic phase with a strongly acid solution restores both lead and bismuth to the aqueous phase, Bambach and Burkey14 have shown that lead is preferentially reverted by a phthalate buffer at a pH of 3 4 , and Kozelka and Kluchesky,ll working with much stronger solutions of dithizone, advocate nitric acid at a pH of 2.3 to 2.5. Table I shows the extent of reversion of bismuth on shaking a solution of metal dithizonate in carbon tetrachloride (initial concentration of dithizone 12.5 mg per litre) with various reagent solutions (phase ratio 3 to 4) for different lengths of time. TABLE I REVERSION OF BISMUTH Bismuth taken, PP 2.0 3.0 2.0 1.0 1.0 Lead taken, Reversion mixture PP 0.0 5 N sulphuric acid 0.0 3-5 per cent.sodium acetate - sulphuric acid buffer of pH 3.4 0.0 0.03 per cent. v/v nitric acid of pH 2-40 0.0 37 1.0 33 Time, “Lead” miniites found, PP 10 2.71 10 1.71 1 0.33 1 0.00 1 1.00 It can be seen that the reversion of 1 pg of lead is quantitative on shaking with dilute nitric acid of pH 2.4 for 1 minute, and that, although 1 pg of bismuth does not interfere, 2 pg is reverted to the extent of about 12 per cent. In the procedure described below lead is first removed completely from the biological digest by successive extractions with portions of dithizone (25 mg per litre in carbon tetrachloride), and a preliminary separation from bismuth is achieved by reversion with dilute nitric acid (pH 2.4).The small amount of bismuth remaining in this acid extract does not affect the precision of the subsequent deter- mination of lead, which is carried out by extracting from an alkaline citrate - cyanide buffer under standard conditions with a more dilute solution of dithizone and subsequently determining the reversion value under standard conditions (see Tables I11 and IV). Tin does not interfere as it is oxidised to the stannic state during digestion.October, 19531 REAGENTS- DETERIVIINATIOX OF TRACES OF LEAD UTTH DITI-IIZONE METHOD 573 All reagents should be of recognised analytical purity. IWetaZ- free water-Redistil laboratory distilled water from an all-glass Pyrex still. Carbon tetrachloride-Redistil reagent grade carbon tetrachloride in all-glass Pyrex Concentrated sulph uric and nitric acids-Redistil these acids in all-glass Pyrex apparatus PerchZoric acid-A 70 per cent.w/v solution redistilled in all-glass Pyrex apparatus Hydrochloric acid-A redistilled constant-boiling mixture. Nitric acid, 0.03, 1 and 2 per cent. v/v solutions-Dilute 0-3, 10 or 20 ml of the redistilled concentrated acid to 1 litre with metal-free water. Ammonium hydroxide, q5proximately 4 N-Absorb the ammonia given off by heating reagent grade ammonium hydroxide, sp.gr. 0.880, in metal-free water cooled in an ice - salt mixture. AmmoniacaZ citratt: solution A--A 50 per cent. w/v monohydric citric acid solution. Dissolve 500 g of citric acid monohydrate in approximately 500 ml of metal-free water and add about 1 ml of phenol red indicator (0.4 per cent.w/v). Add portions of ammonium hydroxide, sp.gr. 0.880, with continuous stirring until the colour of the indicator changes t o red (pH about 8.5). Allow the solution to cool after each addition and finally make up the volume to 1 litre with metal-free water. Purify the solution by shaking it with portions of a solution of dithizone in redistilled carbon tetrachloride (approximately 50 mg per litre) until two successive extracts show the unchanged green colour of the reagent. Shake the citrate solution with two small portions of any redistilled carbon tetrachloride to extract any dissolved dithizone. This procedure was adopted since similar solutions prepared from analytical grades of ammonium citrate were found to be highly contaminated by traces of heavy metals.Ammoniucal citrate solution J3--,4 20 per cent. w/v solution of monohydric citric acid. Dilute 400 ml of solution A to 1 litre with metal-free water. Potassium cyanide solution, 5 per cent. w/v-Dissolve 50 g of potassium cyanide in about 100 ml of metal-free water and purify the solution as described for ammoniacal citrate solution -1. Finally make the solution up to 1 litre with metal-free water. Hydroxylamine hydrochloride solution, 20 per ccnt. w/v-Dissolve 200 g of hydroxylamine hydrochloride in about 600 ml of metal-free water and neutralise the solution with ammonium hydroxide, sp.gr. 0.880. Purify as described €or ammoniacal citrate solution A and make up to 1 litre with metal-free water. Dithizone solutions-Prepare a stock solution by dissolving 100 mg of good quality commercial dithizone reagent in 1 litre of redistilled carbon tetrachloride and dilute as required.Use Pyrex bottles covered with black paper and keep in a cold store. Standard lead solutions-Dissolve lead nitrate in 1 per cent. v/v nitric acid to give a solution containing 1 nig of lead per ml. Dilute this stock solution with 1 per cent. v/v nitric acid to give a standard solution containing 1 or 2 pg of lead per ml. This shows no appreciable change in concentration over six months. apparatus. under reduced pressure. under reduced pressure. All-glass Pyrex apparatus must be used. APPARATUS- All-glass apparatus must be made of Pyrex glass and it should be reserved for the estimation of lead or other trace metals. I t must be specially cleaned with strong sodium hydroxide solution (about 50 per cent .), concentrated nitric acid (boiling where possible) and then rinsed thoroughly with distilled and metal-free water.Separating funnels should be washed with 50 per cent. nitric acid and then rinsed with water before use. The efficiency of the cleaning process should always be tested with a drop of dilute dithizone solution. Stainless steel instruments should be boiled with dilute acetic acid and then rinsed with water as above. The tops of reagent bottles and standard flasks should be covered with small beakers or boiling tubes to exclude dust. Pipettes may be stored in a covered %litre measuring cylinder and should be rinsed with metal-free water before and after use. Pyrex syringe-pipettes made from 10-ml pipettes Corks and rubber bungs must not be used.574 IRVING AND BUTLER: A KEVERSION METHOD FOR THE ABSOKPTIOMETRIC [VOl.’is and discarded syringes with broken nozzles are convenient for transferring concentrated acids, ammonium hydroxide and potassium cyanide solution. The pipettes must not touch the bench. Terry spring-clips are convenient for holding separating funnels on racks and on the mechanical shaker. White Vaseline petroleum jelly is a satisfactory lubricant for the taps. The digestion of samples is best carried out in micro-Kjeldahl-type tubes of about 30-mZ capacity on an electrically-heated rack with a rheostat control. The funnels should have short stems. COLLECTION OF SPECIMENS- Dissolve any solids by adding hydrochloric acid.Take blood by veni-puncture with an ungreased Pyrex syringe and a stainless steel needle, and expel the blood into a tube provided with a stopper and containing a small amount of purified solid ammonium citrate as an anti-coagulant. Collect cerebrospinal fluid through a stainless steel needle in a tube provided with a stopper. Cover the tops of bottles and tubes with filter-paper secured with elastic bands when not in actual use, to avoid contamination from dust. Take a large specimen of soft tissue in the normal way and then remove samples from the interior with specially cleaned stainless steel or glass instruments. Dry to constant weight in covered dishes by vacuum desiccation and heating at 110” C. Collect urine directly in a 2- or 3-litre bottle with the aid of a funnel if necessary. PROCEU u RE- Digestion-Heat the sample (20 in1 of urine, 5 ml of blood, 5 to 10 ml of cerebrospinal fluid, or 100 to 500 mg of dry soft tissue) with 1.0 ml of concentrated sulphuric acid, 2.0 ml of concentrated nitric acid and 0.5 ml of 70 per cent.w/v perchloric acid until a colourless solution remains. Include a small piece of Pyrex glass to promote even boiling. Heat gently at first to avoid excessive frothing and finally raise the temperature until fumes of sulphuric acid appear. Dissolve dry bone (about.1 g) in 10 ml of 50 per cent. v/v nitric acid with gentle heating. Take an aliquot containing 50 to 100mg of bone and digest with 1-Om1 of concentrated nitric acid and 1.0 ml of constant-boiling hydrochloric acid solution until the nitric acid has distilled away. Preliminary extraction-When the solution has cooled, add the appropriate volume of ammoniacal citrate solution A (containing phenol red): 5 ml for blood or soft tissue and 10ml for urine or bone.Then add ammonium hydroxide until the colour of the indicator changes to red (pH about 8.5) and dissolve any solids by gentle heating. Allow the solution to cool, re-adjust the pH with ammonium hydroxide if necessary, and transfer to a 60-ml separating funnel. Wash the digestion tube three times with metal- free water to make the total volume about 40 ml. Test the completeness of the transference by swirling a drop of dilute dithizone solution in the digestion tube; repeat the washing if this solution does not remain green.Add the washings to the contents of the separating funnel. Add 3.0 ml of 5 per cent. w/v potassium cyanide solution and, for blood and liver, 1.0 ml of 20 per cent. w/v hydroxylamine hydrochloride solution. Dry the stem of the funnel with a roll of “ashless” filter-paper and plug it with a small piece of cotton wool freed from metals by dithizone extraction. Extract the lead (plus bismuth) by shaking manually with 2-ml portions of a solution of dithizone in carbon tetrachloride (25 mg per litre) until two successive extracts show the unchanged green colour of the reagent. Note the volume of the dithizone solution required; this serves as an indication of the total amount of lead (plus bismuth) present. Four extractions are usually sufficient for a normal specimen.Run the organic extracts into a 20-ml separating funnel, having previously dried the stem of this funnel and the bore of the tap with a roll of filter-paper. Do not allow the aqueous phase to enter the bore of the tap of the 60-ml funnel and take care that no drops of the alkaline aqueous phase accompany the organic extracts, as they will raise the pH of the unbuffered reversion reagent and may lead to incomplete reversion of lead. Shake the organic extracts mechanically for 1 minute (200 shakes) with 5 ml of 0.03 per cent. v/v nitric acid. Separate the organic phase carefully, without allowing the aqueous Add further portions of nitric acid if necessary.October, 19531 DETERMINATION OF TRACES OF LEAD WITH DITHIZONE 575 phase to enter the bore of the tap, and discard it unless the bismuth content is to be deter- mined.Take an appropriate aliquot of the aqueous phase for the determination of lead if it is judged that the total amount of the metal present will not lie on the linear portion of the calibration graph. Determination of Lead by reversion-Add 3 ml of a solution of dithizone in carbon tetra- chloride at the appropriate concentration. This should be about 12.5 mg per litre (A, 21. 0.5) if a Spekker absorptiometer model H546 is used to record absorbancy readings in the most sensitive region of the logarithmic scale. With the new version of this instrument (H760) the concentration is limited principally by the deviation from Beer’s law, and it may be increased to about 20 mg per litre (A, N 0.7). Then add 1.0ml of ammoniacal citrate solution B, 1.5 ml of 5 per cent.w/v potassium cyanide solution and 2.5 ml of 2 per cent. v//v nitric acid solution and shake mechanically for 1 minute (200 shakes). The pH value of the aqueous phase should be between 8 and 9. Dry the stem of the funnel if necessary and plug it with a small piece of purified cotton wool to prevent small droplets of the aqueous phase from being carried through into the optical cell (where they would make variable contributions to the absorbancy) or into the reversion funnel (where they may cause incomplete reversion of lead). Discarding the first few drops, which serve to wash out the stem, run about 1.5 ml of the mixed colour extract into a 10-ml separating funnel (with a dry, plugged stem) calibrated at 1.5 ml.Add 2 ml of 0.03 per cent. v/v nitric acid solution and shake mechanically for 1 minute (200 shakes). Determine the absorbancy of the organic phases at 620mp before and after reversion (L4, and A,, respectively), using a Spekker absorptiometer with I-cm micro-cells of 0.5 ml capacity and the tungsten-filament lamp and Ilford No. 607 orange filter combination. Wash the cell with a small amount of the organic phase and cover with a glass slide to prevent evaporation. Calculate the reversion value, R, by subtracting A, from A, and correct this for the reversion value for the blank, which must be included with each batch of samples and treated in exactly the same way. Calculate the amount of lead present in the original sample by reference to a previously determined standard curve.Prepare a standard curve by mixing x ml of a standard solution of lead in 1 per cent. v/v nitric acid (see below) with (5 - x) ml of 1 per cent. v/v nitric acid solution. Add 1.5 ml of 5 per cent. w/v potassium cyanide solution, 1.0 ml of ammoniacal citrate solution B, 2.5 ml of metal-free water and 3 ml of dithizone solution (see below). Shake for 1 minute (200 shakes) and measure the absorbancy before and after reversion as described above (p. 574). If a model H546 Spekker absorptiometer is used, the concentration of the standard lead solution should be 1 pg per ml and that of the dithizone solution about 12.5 mg per litre (A, N 0.5). The standard graph obtained was linear for up to 3.5 pg of lead and in this range the reversion value was 0.075 per pg of lead.With a model H760 Spekker absorptiometer, a standard lead solution containing 2 pg per ml and a dithizone solution containing about 20 mg per litre (A, N 0.7) was used, when the standard graph was linear over the range 0 to 7 pg of lead and the reversion value was 0-070 per pg of lead. Consistent readings could not be obtained with the model H546 absorptiometer when working in a dark room and using the technique recommended by the makers. Hence the lamp-house shutter was fixed permanently in a fully open position and the light switched on and off for each measurement. The position of the cell carriage was also found to be critical, and to ensure that it is reproduced exactly each operator must follow strictly his personal technique. These difficulties appear to have been eliminated by mechanical improvements in the new model (H760).DISCUSSION OF RESULTS Satisfactory recovery of lead from small biological samples is indicated by the typical results given in Table I1 for blood, dry bone and two different samples of urine. Wet oxidation was preferred to dry method^,^ which have given much lower recoveries in the hands of some workers owing to volatilisation or loss of lead by fusion into the surface of the vessel used for ignition15; fictitious high lead figures will result if ignition is carried out in porcelain crucibles fired with a lead glaze. Any precipitate of calcium phosphate obtained on basification, or calcium sulphate formed when sulphuric acid has been used in the oxidative digestion mixture, carries lead down with it and must therefore be redissolved before extraction.Although this can be effected by warming with alkaline citrate solution,576 IRVING AND BUTLER: A REVERSION METHOD FOR THE ABSORPTIOMETRIC [Vol. 78 the over-all concentration of citrate required (approximately 6 per cent. for blood and soft tissues, 12 per cent. for urine and considerably higher for bone after digestion with a sulphuric - nitric acid mixture) reduces the efficiency with which lead is later extracted. For example, a graph relating the percentage of lead extracted by 3 ml of dithizone solution (12.5 mg per litre in carbon tetrachloride) on shaking for 1 minute with 10ml of alkaline citrate buffer TABLE I1 RECOVERY OF LEAD FROM BIOLOGICAL SPECIMENS Lead Specimen added, tLg 18 ml of urine .. . . . . 0-0 0.0 1.0 1.0 18 ml of urine . . .. . . 0.0 0.0 2.0 2.0 2 ml of blood . . .. . . 0.0 0.0 1.0 1.0 0.0 1.0 1-0 27.4 mg of dry tibia . . . . 0.0 Rcorr.” 0.020 0-025 0-092 0.094 0.033 0.035 0.186 0,177 0.059 0-064 0-138 0.129 0-020 0.014 0.096 0.093 Lead Average lead found, found, Recovery, Pg Pg % 0.30 94 1-25 0.45 2-36 0.79 0.82 1-72 0.23 103 1-24 99 96 * Rcorr. is the reversion value corrected for the blank. containing 2 pg of lead to the pH was constructed in the usual way. Extraction was found to be quantitative over the pH range 8 to 9 when 2 per cent. citrate was used; but on increasing the concentration to 5 per cent., the extraction curve had a flat maximum covering the same pH range, and only 75 per cent. of the lead was extracted.For this reason the extraction of all biological digests was carried out with successive portions of dithizone until the colour of the last extract was a clear green. To minimise the precipitation of calcium salts at any TABLE I11 SEPARATION OF LEAD FROM BISMUTH BY REVERSION WITH 0.03 PER CENT. w/v NITRIC ACID Lead Bismuth & & taken, taken, (A) (B) (4 (B) Rcorr. “Lead” found, pg CLg CLg 0.0 2.0 0*000 o*ooo 0.00 0.00 0.0 5.0 0.051 0.000 0.73 0.00 0.0 10.0 0.079 0~000 1.13 0.00 0.0 15.0 0.091 0~000 1-30 0.00 0.0 20.0 0.096 0.000 1.37 0.00 2.0 0.0 0.138 0.142 1.97 2-03 2.0 1.0 0.139 0.138 1-99 1-97 2.0 2.0 0.144 0-136 2.06 1.94 2.0 5.0 0.140 0.143 2.00 2.04 2.0 10.0 0.123* 0.136 1-76* 1.94 2.0 15.0 0-155* 0.139 1.64* 1.99 2.0 20.0 0.113” 0.140 1.61* 2.00 NOTE-The initial concentration of dithizone was 20 mg per litre.Measurements under the heading €3 refer to mixtures of lead and bismuth that have been subjected to the full procedure (p. 574) of two cycles of extraction and reversion. Results under heading A were obtained on mixtures subjected to the second stage only; the fixed amount of dithizone was here insufficient for the complete extraction of lead in the presence of large amounts of bismuth, which accounted for the low recoveries indicated by asterisks. * Low recoveries.October, 19531 l>ETERMINATION OF TRACES OF LEAD WITH DITHIZONE 577 stage, the citric acid was invariably added before basification, and a nitric - hydrochloric acid mixture was used when bone samples were digested; this had the added advantage of pre- venting the separation of solids and any consequential bumping.In the analysis of blood and liver, excessive oxidation of dithizone by ferricyanide derived from the high iron content was prevented by the hydroxylamine. Table I11 illustrates the efficiency with which lead can be separated from bismuth. The final absorptiometric determination by the reversion technique deals effectively with lead in the presence of 5 pg of bismuth (column A); the two-stage process described in the full procedure (p. 574) is effective in the presence of at least 20 pg of bismuth. Since the extractability of bismuth depends on the cube and that of lead upon the square of the con- centration of excess dithizone,13 the retention of bismuth in the organic phase when reverting lead with dilute nitric acid is strongly favoured by using a dithizone concentration of 20mg per litre (Table 111) rather than 12-5mg per litre (Table I).The efficient separation and recovery of lead from biological fluids containing bismuth and other metals is illustrated by the data in Table IV. TABLE IV RECOVERY OF LEAD FKOM 18-ml SPECIMENS OF URINE IN THE PRESENCE OF ADDED BISMUTH, COPPER AKD ZIKC Metal added Pb. PLg 0.0 1.0 1.0 1.0 1.0 Bi, CUP PLg Pg 0.0 0.0 0.0 0.0 2.0 0.0 5.0 0.0 10.0 0.0 7 Lead Zn, RCOIT. found, Recovery, PF, tLg % 0.0 0.032 0.43 - 0.0 0-104 1.39 96 0.0 0.108 1.44 101 0.0 0.104 1.39 96 0.0 0.111 1.48 105 0.0 0.0 0.0 0.0 0.02 1 0.28 - 1.0 0.0 0.0 0.0 0.101 1.35 107 0.0 0.0 100.0 100.0 0-024 0.32 - 1.0 0.0 100.0 100.0 0.096 1.28 96 An abnormal amount of bismuth in a biological digest is readily detected, as its dithizone extract (clear orange) appears before that of the lead (full red).The bismuth is retained in the organic phases from the initial and the final reversion stages whence it can be removed by shaking with 1 per cent. nitric acid solution. It may then be determined by extraction with dithizone from an ammonium hydroxide - cyanide buff er4 solution followed by reversion with N sulphuric acid.6 Apart from the effective determination of lead in the presence of bismuth and other metals we would emphasise that the principle advantages of the procedure described above are flexibility of conditions and economy of time and materials. Any good commercial sample of dithizone can be used without purification, as any coloured impurities initially present or subsequently introduced do not give rise to errors, provided that they are un- affected by the reversion process.Day to day changes in the concentration of dithizone solutions are without significance. For conventional mixed-colour methods it is necessary to remove all dithizone remaining in cyanide and citrate solutions after purification ; this tedious operation need not be carried to completion in the proposed procedure. The adjust- ment of pH is not critical provided it remains between 8 and 9. Finally, the seiisitivity of the reversion method (R per pg of Pb 21 0-075) is about twice that of the mixed-colour method. The above procedure has been in frequent use over the past three years for the determina- tion of lead in fluid and tissue specimens from patients with chronic neurological diseases16 and lead poisoning, and has proved satisfactory in the hands of laboratory technicians.It should be possible to apply it to other materials with little modification. Our thanks are due to Mr. J. R. P. O’Brien for his interest in this work and for providing laboratory facilities. \Ye are also indebted to Mr. G. E. Newman for skilled technical assistance and to Dr. E. A. Stocken of the Biochemistry Department, Oxford, for the gift of a digestion apparatus.578 IRVING AND BUTLER: A REVERSION METHOD FOR THE ABSOKL’TIOMETKLC [vol. 78 REFERENCES Kehoe, R. A., Thamann, F., and Cholak, T., J . I n d . Hyg., 1933, 15, 257. Cone, W., Russell, C., and Harwood, R.IJ., Arch. Nezkrol. Psychiat., 1934, 31, 236. Campbell, A. M. G., Herdan, G., Tatlow, W. F. T., and Whittle, E. C., Brain, 1950, 73, 52. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” First Edition, Interscience Hart, H. V., Analyst, 1951, 76, 692. Irving, H., Risdon, E. J., and Andrew, G., Nature, 1948, 161, 805; J . Chern. Soc., 1949, 537. Cooper, S. S., and Sullivan, M. L., Anal. Chern., 1951, 23, 613. Clifford, P. A., and Wichmann, H. J., J . Ass. 08. Agric. Chern., 1936, 19, 130. Morton, F., Analyst, 1936, 61, 465. Liebhafsky, H. A., and Winslow, E. A., J . AYner. Chew. SOC., 1937, 59, 1966. Kozelka, F. L., and Kluchesky, E. F., I n d . Eng. Chem., Anal. Ed., 1941, 13, 484 and 492. Risdon, E. J., D.Phi1. Thesis, Oxford, 1945.Irving, H., and Williams, R. J. P., J . Chenz. Soc., 1949, 1841. Bambach, K., and Burkey, R. E., I n d . Eng. Chem., Anal. Ed., 1942, 14, 904. Allport, N. L., and Garratt, D. C., J . SOC. Chem. Ind., 1948, 67, 382. Butler, E. J., J . Ncurol. Psychiat., 1952, 15, 119. Publishers Inc., New York, 1944, pp. 83 and 279. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. THE INORGANIC CHEMISTRY LABORATORY SOUTH PARKS ROAD OXFORD THE DEPARTMENTS OF BIOCHEMISTRY AND NEUROLOGY THE RADCLIFFE INFIRMARY OXFORD July 30th, 1952 DISCUSSIOX MR. N. L. ALLPORT, in congratulating the authors on their work, recalled that it was just 21 years ago that, in association with G. H. Skrimshire, he had presented a paper in that same room showing how lead might be extracted from an alkaline aqueous phase containing citrate and cyanide by means of a chloroformic solution of dithizone. It was then pointed out that bismuth interfered, but there did not sccm to be any way of overcoming the difficulty.Since that day a great deal of work had been published on the use of dithizone for the isolation and determination of trace metals and in the intervening years great advances had been made in the design and construction of spectrophotometers. It was therefore of particular interest to learn that the problem of bismuth interference had a t last been overcome. DR. Il. C. GARRRTT said that, as the cause of the irrelevant absorption was unknown and hence it was impossible to say that it was unaffected by the extraction, he questioned whether the authors’ correction could be considered valid.MR. R. C. CHIRNSIDE said that a number of people appeared to be worried as to the possible effect of foreign ions the identity and amounts of which were not known. DR. IRVING said that Dr. Garratt’s and Mr. Chirnside’s questions referred to the rather ill-defined term ZciCi of equations (1) and (2). I t must be remembered that the reversion technique was applied to the quantitative determination of lead only after preliminary stages of extraction (p. 574) and reversion (p. 575) under carefully controlled conditions designed to concentrate all the lead, and essentially only the lead, in an aqueous acid. Here it would be accompanied possibly by some bismuth, but by very little else. Other water-soluble impurities would have remained almost exclusively in the aqueous phase rejected after the initial extraction, whilst organic-soluble impurities would have passed into the chloroform extract and remained there after treatment with the 0.03 per cent.nitric acid. The irrelevant absorption in the final dithizone extract used for measuring A, was therefore not really unknown, for it would be due primarily to the bismuth - dithizone complex. MR. A. A. SMALES said he thought that the doubt that had arisen about the reversion stage in the authors’ procedure was best expressed by saying that the specificity of hydrogen ion for reverting lead but not bismuth a t this stage was open to question. He suggested that this specificity might well be improved by the use of ethylenediaminetetra-acetic acid or one of the other “complexones” described by Schwarzenbach, which formed strong complexes with lead but very weak ones with bismuth.DR. IRVING agreed that the hydrogen ion did not have the specific character theoretically postulated for a reversion agent (cf. reference 6). A t best it was a “selective” reagent ( c j . Irving, H. M., and Williams, R. J. P., Analyst, 1952, 77, 513), and in the proposed procedure its use inevitably led to a degree of com- promise. The principle of the method described in this paper was in fact devised by the senior author in 1946, i.e., in pre-complexone days, and there was no doubt that considerable improvements in the choice of reversion agents for lead should now be possible. DR. E. C . WOOD said that the authors had stressed at the beginning of the paper that there was no pH a t which a quantitative extraction of lead from dithizone could be made without a t the same time extracting a t least some of any bismuth present; yet the theory of the reversion technique seemed to require that this should be done in between the two optical measurements.He asked Dr. Irving to clarify this point.October, 1‘3531 DETERMINATION OF TRACES OF LEAD WITH ~ITHIZOKE 579 DR. IRVING said that in order to save space they had not included in their paper details of the very extensive preliminary work ( c j . reference 12) involving studies of (a) the variation of percentage extraction with pH for various aqueous salt media, and (b) the effect of variations in pH on (2) the rate of extraction of lead and of bismuth by a solution of dithizone in chloroform and (ii) the rate of reversion of chloroform solutions of various compositions. I n the procedure described only a small fraction of any bismuth extracted as dithizonate was decomposed by the acid reversion mixture, partly because of the choice of pH, partly because the time of shaking specified did not permit this reaction to reach the equilibrium appropriate to the pH chosen, although the relatively more rapid decomposition of lead dithizonate did so, and partly because the dithizone set free during the reversion of the lead dithizonate favoured the retention of the bismuth as the dithizone complex in the organic phase (p.577; cf. reference 15 and also Irving, H., Bell, C. F., and Williams, K.J. P., J . Chem. Soc., 1952, 356). In short, conditions of acidity were so chosen that theoretically (i.e., under equilibrium conditions) the minimum amount of bismuth should be reverted along with all the lead, and then the time of shaking was further chosen so that non-equilibrium conditions prevailed and the kinetics of reaction were made to weight the balance still more heavily in favour of extracting lead in preference to bismuth. That the procedure was successful in practice followed from the figures shown in Tables I11 and IV. DR. J. HASLAM asked if there was much difference in the extinction coefficients of different samples of so-called “purified” dithizone. He was worried about the effect of this variation on the extinction term for dithizone in the eqdation R = L4r- A, = etc.DR. l R T i m c said that there were considerable variations in the molecular extinction coefiicients of different samples of so-called “purified” dithizone. In the customary “two-colour” methods such differences between samples could easily make nonsense of even the most carefully constructed calibration curves (cf. reference 11 and the reply to Mr. Wyatt, below). But it was generally agreed that one of the best ways of determining the molecular extinction coefficient of a pure sample o f dithizone (Cooper, S. S., and Sullivan, Sister M. L., Anal. Chem., 1951, 23, 613), or of determining the content of pure reagent in an impure sample (Irving, H., and Bell, C. F., ./. Chem. Soc., in the press), was to make use of the fact that dithizone, HDz, formed metal complexes with bivalent metals, MI1, of exactly stoicheiometric com- position MDz,, and that these dithizonates contained two equivalents per molecule of absolutely pure dithizone.The change in absorbancy when a concentration, C,, of pure metal dithizonate was “reverted” to give a concentration, 2C,, of perfectly pure dithizone of molecular extinction coefficient E , was thus Cc(2c, - EJ, as given in equation (3), where E , and E , referred to perfectly pure substances. Of course, they had stated that one of the advantages of the reversion method was that the use of highly purified dithizone (or of dithizone of invariable purity) was unnecessary. If one distinguished between reagent dithizone of extinction coefficient E‘, and the dithizone fiuriss.of extinction Coefficient cr liberated by reversion, equations (1) and (2) of the paper could be rewritten as follows- A, = E’$, + cCCc 4- X E & ~ . . .. .. .. . . (la) and A, = E’,C, + ~ E , . C , -b Z E ~ C ~ . . . . . . .. . . (2a) from which equation (3) followed. Hence the reversion value was iiidepcndent of the absolute purity of the reagent. MR. C. H. PRICE asked if there was not a risk that some of the excess of dithizone might be oxidised between the determinations of A,,, and A,, i.e., between the final two readings of absorbancy, and if it would not be an advantage to use some sulphurous acid or a reducing buffer. DR. IRVING replied that they had found no signs of oxidation during the few minutes that separated the determinations of A,,, and A, and there need be no fear of error from this cause.hlR. E. E. ARCHER asked whether a buffer solution could be used to extract lead in one stage directly from the first dithizone extract, so eliminating one stage from the procedure. DR. IRVING replied that this simplification could be justified in practice only if the method was to he applied to routine samples of the same type and approximately the same lead content in which the presence of interfering elements, especially bismuth, was known to be constant and very small. The two-stage procedure described in the paper allowed for a considerable degree of variation from sample to sample and this flexibility was gained a t the expense of only one additional stage. MR. P. F. WYATT said that he had tried the reversion method soon after its publication in the Journal of the Chemical Society in 1949, but had found a disadvantage in that the standard curve for dithizone when the Ilford orange 607 filter was used was only rectilinear below an optical density of about 0.5; above this value there was an increasing deviation from Beer’s law, and he had had difficulty in getting reproducible results, although he agreed that for a limited low range of lead concentrations results were good.For this reason he had developed a “mixed colour” method, which he preferred to use, whereby the optical density of the chloroform extract containing the lead dithizonate plus an excess of free dithizone was measured a t two wavebands corresponding to the Ilford orange 607 and green 604 filters.The absorp- tion due to lead dithizonate with the orange filter was very small, so that thc optical density measured with an Ilford 607 filter could be assumed to give a measure of the excess of dithizone alone, E (Dz). From appropriate standard curves for free dithizone, the absorption in the waveband passed by the Ilford 604 filter that was due to the excess of dithizone could be calculated and deducted from the total optical680 IRVING AND BUTLER [Vol. 78 density measured with the 604 filter. concentration, E (Pb). containing 35 to 40mg per litre being used. dilution; 1-cm cell. The difference gave the optical density corresponding to the lead Typical values for the calibration graphs were as follows, a dithizone solution 1. Standard curves for dithizone in chloroform-Approximately 37 mg of dithizone per litre; 25-ml c Dithizone solution, ml .. 1 2 3 4 5 6 8 E (Dz) with 607 filter . . 0.170 0-330 0.485 0.626 0.75 0.876 0.975 1.066 E (Dz) with 604 filter . . - 0.084 - 0.165 - 0.251 - 0.329 2. Standard cuwe for lead-10-ml dilution; l-cm cell. Volume of dithizone ml 8 2-5 16 4.0 24 6.0 32 7.0 40 9.0 Lead, solution used, Observed optical density r Ilford 607 Ilford 604 ’ filter filter 0.155 0-265 0.165 0.50 0.23 0.735 0.17 0.955 0.24 1-30 E (Dz) with 604 filter calculated from E with 607 filter 0-038 0.04 1 0.058 0.042 0.060 Net reading E (Pb) with 604 filter 0.227 0.459 0.677 0.913 1.140 A graph of the net reading, E (Pb), against weight of lead taken was a straight line over the whole range. With regard to bismuth, Mr. Wyatt agreed that quantitative separation of bismuth from lead by extraction from a solution buffered a t about pH 3 was difficult. A slight increase in pH caused some lead t o be lost by extraction with the bismuth, while a slight drop in pH led to incomplete removal of bismuth. He preferred to remove bismuth by extraction with diethylammonium diethyldithiocarbamate from a solution of high acidity, 4 to 6 N, depending on the relative amounts of sulphuric and hydrochloric acids present (Strafford, N., Wyatt, I?. F., and Kershaw, F. G., Analyst, 1945, 70, 232), and then to isolate the lead by extraction with the same reagent at a much lower acidity. The extract containing the lead was then decomposed and the determination completed by the dithizone “mixed colour” method already outlined. DR. IRVING said that he was not surprised that Mr. Wyatt had had trouble when working with the more concentrated solutions of dithizone and lead dithizonates to which Beer’s law no longer applied. But this was scarcely a reflection on the reversion method and he was glad to learn that Mr. Wyatt had had good results from it in the low lead range where it could most legitimately and profitably be used. He asked whether, in Mr. Wyatt’s problems, there was any objection to using a smaller sample, or to using a smaller aliquot of the digest from a large sample if there was any danger of segregation. In the “mixed colour” method he described, variations in the quality of the dithizone used would, of course, affect the calibration. Unlike the reversion procedure it did not discriminate between lead and bismuth and could, therefore, only form the final determination step after some previous stages of quantitative separation. MR. SMALES said he wondered whether the brilliant colours of the dithizone complexes were not, in fact, dazzling us, and whether it might not have been better if the reagent and the complexes had been colourless. More use might then have been made of dithizone for its remarkable separating powers, the final determination being by some alternative to absorptiometry, such as polarography. Polarographic instruments were available with a sensitivity of 0.02 pg of lead per ml of solution, more than adequate for the class of determination being considered.

ISSN:0003-2654    

DOI:10.1039/AN9537800571

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

October, 19533 ASHTON, FOSTER AND FATHERLEY 581 A Colorimetric Determination of Dihydrostreptomycin BY G. C. ASHTON, M. C. FOSTER AND M. FATHERLEY A colorimetric method for the determination of dihydrostreptoniycin is presented. It is based on a reaction between guanido materials and diacetyl, alkali and a-naphthol. The colour formed is not subject to inter- ference at high salt concentrations and can be used in the routine analysis of factory samples. Penicillin and its compounds do not interfere with the reaction, so the method can be applied to the determination of dihydro- streptomycin in mixtures with penicillin. The standard error of analyses made by the method is k1.3 per cent., which compares favourably with that found in microbiological assay. SEVERAL chemical methods are available €or determining dihydrostreptomycin.In general they are either tedious or subject to interference from substances other than dihydro- streptomycin in the samples submitted for assay. We required a simple method for determining dihydrostreptomycin in a variety of samples ranging from process control samples to samples of the pure antibiotic alone and in admixture with other antibiotics. Oxidimetric, spectrophotometric and colorimetric methods have been published for the determination of dihydrostreptomycin. The oxidimetric procedures depend on the oxidation of dihydrostreptomycin by sodium metaperiodatel or periodic acid.2 9 3 Formaldehyde is produced, and, after distillation under carefully controlled conditions, the formaldehyde is determined with chromotropic acid.The necessity for distillation was overcome by Vail and Bri~ker,~ who removed the interfering periodic and iodic acids by lead acetate precipitation and developed the colour of chromotropic acid with formaldehyde in the clear centrifugate. Methods based on periodate oxidation suffer irom the disadvantage that streptomycin produces different amounts of formaldehyde under different conditions. In our experience such careful control of this procedure is required as to make it inconvenient for routine use. A method based on the spectrophotometric absorption of dihydrostreptomycin at 265 mp after acid hydrolysis was suggested by Hiscox4 and is suitable for pure dihydrostreptomycin, but it cannot be applied to mixtures of antibiotics without the use of appropriate correction factors, which reduce its precision.As the dihydrostreptomycin and streptomycin molecules are identical except that the aldehyde grouping of the streptose moiety is hydrogenated in dihydrostreptomycin and is therefore not reactive, colorimetric methods for streptomycin based on the streptidine or X-methylglucosamine moieties are potentially applicable to the determination of dihydro- streptomycin. Hence, the streptidine assay developed by Sullivan and Hilmer5 for streptomycin has been applied to dihydrostreptomycin by Monastero.6 The streptomycin assay described by Scudi, Boxer and Jelinek' is based on the Elson - Morgan reaction €or glucosamine and, similarly, can be used for dihydrostrept omycin estimation on the nearly pure substance, but dihydrostreptomycin gives only one twenty-fourth of the colour given by streptomycin with the same technique and reagents.Our objective being to develop for dihydrostreptomycin an assay that could be applied routinely to a variety of samples, as mentioned above, we investigated a number of reactions dependent on the streptidine moiety of the molecule. Such a reaction is not specific for dihydrostreptomycin ; it is also given by streptomycin and mannosidostreptomycin. Strepto- mycin and dihydrostreptomycin occur together at certain stages during dihydrostreptomycin manufacture, and in assaying these samples for dihydrostreptomycin by a streptidine assay it is necessary to allow for the streptomycin present. This can be done conveniently with the ferric maltol assay,s which determines streptomycin but not dihydrostreptomycin.Mannosidostreptomycin is not usually present in the streptomycin used for the manufacture of dihydrostreptomycin, but if any were present it would be estimated by the ferric maltol assay. It is unusual to find streptomycin and dihydrostreptomycin associated in samples of mixed antibiotics. Hence, for practical purposes a streptidine assay can be used as a measure of dihydrostreptomycin content in mixed antibiotic samples. EXPERIMENTAL Several possible methods based on reactions of the guanidine groups of the streptidine moiety were examined. The method of Monasteros with the oxidised nitroprusside reagent582 ASHTON, FOSTER AND FATHERLEY: A COLORIMETRIC [Vol. 78 of Weberg is subject to interference by salts.A method for aromatic amidineslO is applicable to dihydrostreptomycin, but was found also to be subject to salt interference. We also investigated extensively the Sakaguchi reaction,ll using combinations of a-naphthol or 8-hydroxyquinoline with hypochloritep2 or 1~ypobr0rnite.l~ We found the resulting colours to be unstable and could not obtain reproducible results with any combination of the reagents. A recent paper by Halliday14 described a spray reagent €or streptomycin chromatograms. The reaction involved is attributed to Voges-Proskauer, and the spray reagent used is based on Barritt's15 procedure for guanidine materials, involving the use of diacetyl, a-naphthol and potassium hydroxide. This reaction was finally selected as the basis of our method.Halliday14 said that the diacetyl colour with streptomycin developed slowly and faded after reaching a maximum optical clerisity. We have confirmed this, although the fading rate and time at which the maximum optical density occurs depend on the relative concentrations of the three reagents. ESTABLISHMENT OF OPTIMUM CONCEXTRATIONS- The original instr~ctionsl~ for applying the diacetyl reaction to guanidine materials were as follows. To 1 in1 of guanidine solution add 1 ml of a 0.1 per cent. solution of diacetyl in water, 2.25 ml of water, 0.5 rril of 5 per cent. ethnnolic a-naphthol and 0-25 ml of 40 per cent. alcoholic potassium hydrosick. These amounts of reagent proved suitable for aqueous solutions of dihydrostreptomycin, but from solutions containing much salt the salts were precipitated owing to the high alcohol concentration of the final solution.This was prevented by preparing the caustic potash solution in water rather than in methanol, but on addition of the a-naphthol solution this reagent was itself precipitated. If, however, the potash was added before the a-naphthol, no precipitation occurred. Changes in reagent volumes were then made, so that 2 ml of sample were diluted with 15 ml of water and 1 ml of each of the three reagents was added. As already mentioned, the time of maximum colour development and the amount of colour formed depend on the relative concentrations of the three reagents. Experiments with different combinations of 0.1, 0.2 or 0.4 per cent. diacetyl, 10, 20 or 40 per cent.aqueous potassium hydroxide and 2.5, 5 or 10 per cent. alcoholic a-naphthol were carried out. The optical density of each solution after standing for 30, 60 and 90 minutes was measured. From these results it appeared that 0.4 per cent. diacetyl, 20 per cent. potassium hydroxide and 10 per cent. a-naphthol solutions gave maximum colour formation. This combination of reagents gave a colour that remained at its maximum value for 10 to 15 minutes. WAVELENGTHS OF MAXIMUM ABSORPTION- were determined; A,,,. ranged from 505 mp to 530 mp. finally adopted, the maximuni absorption was at 525mp. The wavelengths a t which the solutions from the first experiment absorbed most strongly With the combination of reagents EFFECTS OF TIME AND TEMPERATURE- A solution of dihydrostreptomycin in water was treated with 0.4 pcr cent.diacetyl, 20 per cent. potassium hydroxide and 10 per cent. a-naphthol solutions at 21", 23", 25" or 27" C. Aliquots were removed a t 5-minute intervals and the optical densities measured at 525 mp against water. Both time and temperature have a marked effect on the rate at which the colour develops, but the final colour is always maximal after about 40 minutes and is fully developed between 23" and 27" C. It is, therefore, recom- mended that the temperature at which colour development is carried out should be not less than 23" C, although satisfactory results can be obtained at lower temperatures. Further, it is necessary to measure the developed colours at a fixed time after adding the reagents; we have found 40 minutes to be suitable.The results are shown in Fig. 1. BLANK I'ALrE OF THE REAGENTS- The reagents themselves give rise to a coloured solution, the intensity of the colour increasing with time (see Fig. 1). Therefore, if a series of samples is being assayed at timed intervals it is not practicable t o use the same reagent blank value. Rather than develop a separate reagent blank €or each sample, we find it convenient to measure the developedOctober, 19531 DETERMINATION OF DIHYDROSTREPTOMYCIN 583 sample colour against water, and subtract from this measurement the optical density value of a separately developed and similarly timed reagent blank measured against water. CALIBRATION GRAPH- An apparently linear calibration graph is obtained with dihydrostreptomycin solutions containing up to 400 units per nil when the final coloured solution is measured in a 1-cm cell a t 525 mp against water.This line passes through a point on the ordinate corresponding to the blank value of the reagent. Calibration is also linear with a Speltker absorptiometer and No. 604 filters. 0.6 05 0'4 x v) c aJ V u & .- - 0.3 .- 0.2 0" 0. I Time of development, minutes Effect of time and temperature on develop- ment of colour between dihydrostreptomycin and the diacetyl - a-naphthol reagent. Curve A, 27" C; curve B, 25°C; curve C, 23°C; curve D, 21°C; curve E, reagent blank Fig. 1. STABILITY OF REAGENTS- From experiments with combinations of reagents of different ages, it appears that the diacetyl and potassium hydroxide solutions are stable for at least 2 weeks.The slope of the graph relating optical density to concentration increases slightly as the cr,-naphthol solution ages, an effect not S O ~ C ~ Y due to an increased blank value, but good results have been obtained with anaphthol solutions 2 weeks old. EFFECT OF COSTAMINANTS ON COLOUR- The eflect on the diacetyl determination of substances that might be encountered in pharmaceutical preparations of mixed antibiotics, or during the production of dihydro- streptomycin, was determined. The substances chosen were added to a solution containing 250 units of dihydrostreptomycin per ml. No colour suppression or intensification was obvious when the dihydrostreptomycin solutions contained 10 per cent. w/v of anhydrous sodium sulphate, 5 per cent.w/v of anhydrous sodium citrate, 1 per cent. w/v of calcium chloride (CaC12.6H20), 0.2 per cent. w/v of sodium benzylpenicillin, 0.03 per cent. w/v of procaine penicillin or 0.03 per cent. w/v of penethamatc hydriodide. The effect on the diacetyl reaction of substances other than those mentioned has not been investigated. Caution should be exercised in applying the method to pharmaceutical preparations containing vegetable oils, ointment bases or unknown excipients. The diacetyl reaction is also given by streptomycin and mannosidostreptomycin, streptidine, arginine and other guanido derivatives.584 ASHTON, FOSTER AND FATHERLEY: A COLORIMETRIC [Vol. 78 METHOD REAGENTS- Diacetyl solution-A 0.4 per cent. w/v solution of diacetyl in distilled water. Potassium hydroxide solution-A 20 per cent.w/v solution of potassium hydroxide in distilled water. a-Naphthol solution-A 10 per cent. w/v solution of a-naphthol in absolute ethanol. PROCEDURE- Either dissolve the sample in distilled water or dilute liquid samples to give a solution containing between 100 and 400 units of dihydrostreptomycin per ml (0.1 to 0.4 mg per ml of dihydrostreptomycin base). Transfer 2 ml of the sample solution to a 6-inch x l-inch test tube, add 15 ml of water and then 1 ml of diacetyl solution, 1 ml of potassium hydroxide solution and 1 ml of a-naphthol solution, in that order. Mix the contents of the tube by inversion after each addition. Start timing when the diacetyl solution is added. Precisely 40 minutes later determine the optical density at 525 mp in a l-cm cell against water.Alternatively, measure the optical density on an absorptiometer with a suitable filter, e.g., a Spekker absorptiometer with No. 604 filters. Prepare a standard graph from suitable dilutions of a standard dihydrostreptomycin solution, using the same procedure for colour development as described for the sample. Determine the potency of the diluted sample solution from the standard graph. A standard graph should be prepared for each determination to allow for differences in room temperature, reagents, and so on. RESULTS AND DISCUSSION PRECISION- An examination of the results from replicate determinations on different samples a t one concentration level showed the apparent standard error of a determination to be & 0.6 per cent.However, observations on replicate determinations at several concentrations of one sample showed the standard error to be greater than this. It seems likely, therefore, that the graph of optical density and concentration has a slight curvature over the range 0 to 400 units per ml. To determine the precision attainable over the recommended range, three different dihydrostreptomycin samples were dissolved in water to give solutions containing approxi- mately 400 to 450 units per ml. Each solution was then diluted accurately with water to give solutions of concentrations 75, 50 and 25 per cent. of the original solutions. The three original solutions and their nine dilutions were assayed in duplicate by the method described. The potencies of the twelve solutions were determined from a standard graph prepared from a standard dihydrostreptomycin material.The results of this experiment are shown in Table I. TABLE I POTENCIES OF DIHYDROSTREPTOMYCIN SOLUTIONS AT FOUR CONCENTRATIONS Dilution level, Sample A Sample B Sample C Level average o/ /O 25 50 75 100 423.1 431.9 424.5 418.7 420.1 424.1 421.6 419.4 426.1 429.0 417.2 414.3 417.2 419.2 426.7 423.8 427.02 415.25 416.55 420-02 Sample average 422.93 421.69 414.50 419.71 An examination of the results showed that the major part of the error is due to the non- agreement between results from different levels on any one sample, caused by slight curvature of the graph of optical density and concentration. The error between duplicate detennina- tions at one level on one sample is considerablv less than the major error due to curvature.October, 19531 DETERMINATION OF DIHYDROSTREPTOMYCIN 585 For routine purposes, however, this curvature error is insufficient to warrant special precautions (such as restricting values to the linear portion of the graph or making allowances for the curvature).The standard error of a pair of duplicate determinations, including the error due to curvature, is about & 1.3 per cent. COMPARISON WITH BIO-ASSAY- A number of dihydrostreptomycin sulphate samples and various dihydrostreptomycin solutions were assayed both microbiologically by the Klebsiella @iwzmoniae plate assay and chemically by the method described. The results are shown in Table 11. TABLE I1 COMPARISON BETWEEN PROPOSED METHOD AND MICROBIOLOGICAL ASSAY Solid samples Solutions A A r > r \ By diacetyl By microbiological By diacetyl By microbiological method, method, method, method, units per mg units per mg units per ml units per ml 765 778 770 765 750 780 780 767 343 34 1 354 318 326 326 365 340 350 340 335 330 There are no significant differences between the chemical and microbiological results. APPLICATION TO STREPTOMYCIN- It is clear that the method described is also applicable to streptomycin and gives results similar to those by the ferric maltol method,8 both with simple streptomycin solutions and with mixed antibiotic samples.The method cannot, however, be used for the determination of the streptomycin in fermentation broths, as it gives a measure of the total guanidine- reacting material present in the broth.We believe the chief value of the method lies in its application to dihydrostreptomycin, for which the more specific ferric maltol method cannot be used. We are indebted to Mr. J. P. R. Tootill not only for carrying out the statistical analysis of the results, but also for useful criticism and helpful suggestions in the planning of experi- ments during the development of the method. We wish also to thank Mr. K. A. Lees for carrying out the microbiological assays. Note added in proof.-After this paper had been accepted for publication, a paper appeared describing the application of the Voges - Proskauer reaction to the quantitative assay of streptomycin (Szafiv, J. J., and Bennett, E. O., Science, 1953, 117, 717). REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. ' 12. 13. 14. 16. Garlock, A. E., jun., and Grove, D. C., J . Clin. Invest., 1949, 28, 843. Colon, A., Herpich, G. E., Johl, R. G., Neuss, J. D., and Frediani, H. A., J . Amer. Plaarm. Ass., Vail, W. A., and Bricker, C. E., Anal. Chew., 1952, 24, 976. Hiscox, D. J., Ibid., 1951, 23, 923. Sullivan, M. X., and Hilmer, P., Abstract of the 109th Meeting of the American Chemical Society, Monastero, F., J . Amer. Pharm. Ass., ScZ'. Ed., 1952, 41, 322. Scudi, J. V., Boxer, G. E.. and Jelinek, V. C., Science, 1946, 104, 486. Boxer, G. E., Jelinek, V. C., and Leghorn, P. M., J . Biol. Chem., 1947, 169, 153. Weber, C. J., Ibid., 1928, 78, 465. Trought, H., Ashton, G. C., and Baker, R. G., Analyst, 1950, 75, 437. Sakaguchi, S., J . Biochem., Tokyo, 1925, 5, 25. Albanese, A. A., and Frankston, J. E., J . Biol. Chem., 1945, 159, 185. Vincent, D., and Brygoo, P., Bull. SOL Chiun. Biol., 1946, 28, 43. Halliday, W. J., Nature, 1952, 169, 335. Barritt, M. M., J . Path. Bact., 1936, 42, 441. Sci. Ed., 1950, 39, 335. April, 1946, p. 4 ~ . GLAXO LABORATORIES LTD. SEFTON PARK STOKE POGES, BUCKS. February 24dh, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800581

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

586 WILLIAMS: A SYSTEMATIC APPROACH TO THE CHOICE [Vol. 78 A Systematic Approach to the Choice of Organic Reagents for Metal Ions BY R. J. P. WILLIAMS The choice of an organic reagent for any particular metal ion is limited by the nature of the free energy change occurring on the formation of the complex of the reagent and the metal ion in aqueous solution. A broad general division of such reactions is made that is based upon whether the entropy change or the heat change of the reaction is the more important term in the free energy change. Small cations of large charge (usually ions of low electronegativity) are divided from the larger and more electro- negative cations in this way. The choice of the type of reagent suitable for selective reaction with a particular metal ion can be based on these principles.Minor factors, such as the nature of the available orbitals, also afTect the free energy of formation of complexes with particular metal ions. Many examples of the successful use of organic reagents are examined and some suggestions are made for the design of further reagents. IN their search for new organic reagents for metal ions, analytical chemists generally prepare an organic compound without regard to the requirements of any special metal ion, unless the compound happens to be a simple modification of an organic reagent already in use for the determination of a particular cation. The testing of the reagent then proceeds by a series of experiments of a trial and error nature. Often the new compound is rapidly shown to be to some extent selective but in no way specific.It is proposed to give, in this paper, some general approach to the problem of the selectivity of organic reagents. The formation of a complex in aqueous solution and its subsequent transfer to a second solvent or its subsequent precipitation can generally be represented by the series of equilibria- f? K - ML (precipitate) . . . . * * (1) M + L + ML{ K, .. ML (in a second immiscible solvent) M and L represent a metal ion and a reagent, respectively, charges are omitted for convenience, and the constants K, K,, K,, are the equilibrium constants for the three possible stages: formation of the complex, its partition and its precipitation. K, is simply related to the partition coefficient and K, to the solubility of the anhydrous complex, as defined by Irving and Wi1liams.l Previous discussions of these equilibria have been based on the equilibrium constant K, its variation from metal to metal and from one ligand to another1 and on the pH control in the formation of precipitates or in the partition of the complexes.2 However, a better understanding of the equilibria can be obtained by resolving the free energy changes, AG, into their component entropy and heat changes, AS and AH, respectively. First, it is important to realise the kind of change that can occur in AS and AH.This is not possible in terms of the above equilibria (l), as they misrepresent the reactions that take place, in so far as they onlit the water molecules involved in the reactions. The reactions are- AG = -2.303 RT log,,K = AH - TAS.+ ML (precipitate) -+ dH,O + ML (seconz phase) + fH,O M(H,O), + L(H,O)b + ML(H,O), + (a i- b - c)H,O K, .. * - (2) *‘ K . . The three reactions are accompanied by considerable changes in the number of water molecules bound to the reactants and therefore in the total number of molecules “free” in the solution. This means that an entropy change takes place. If L is an anion and M is a cation, the number of “free” molecules in the solution will increase upon complex formation because of the release of some of the molecules of water co-ordinated to the ions concomitant withOctober, 19531 OF ORGANIC REAGENTS FOR METAL IONS 587 the neutralisation of charge. Table I shows some examples of entropy changes on complex formation, which illustrate the large changes that favour the complex formation when a neutralisation of charge 0ccu1-s.~ y 4 Many further examples will be published later.5 The MOLAR HEATS AND Reaction Sn"-OH' Sn' *-Cl' Sn"-Br' Mg"-SO," Ba' '-S20," 2n"-CH,(CO,)," Mg"-CH2(COz)z" TABLE I ENTROPIES OF FORMATION AG, AH, K cals.K cals. - 17.2 -2.1 - 16.0 - 1.2 - 6.9 + 7.5 - 2.0 + 8.5 - 0.8 $6.1 - 13.2 - 0.3 - 14.0 - 1.0 - 2.0 + 5.0 - 17.0 - 10.0 - 1.6 + 2.6 - 1.0 + 1.4 - 3.6 + 5.7 - 4.0 + 3.2 -3.1 + 2.6 - 5.0 +3.1 OF SOME COMPLEXES AT 25°C AS, cals. per + 50.5 + 50.0 + 40.0 + 35.0 + 23.0 + 43.0 + 43.3 + 23.3 + 23.0 + 14.0 + 8.0 + 31.0 + 24.0 + 19.0 + 27.0 Reference Nat. Bur. Stand. Circulars Evans and Uri3 " C 39 Y3 n I? Nat. Bur. Stand. Circulars Vanderzee' 39 7) 7) Evans and Monk* Davies and Wyattg Evans and Monk* 99 decrease in hydration must depend to some extent upon the initial degree of hydration of the reacting ions.Table I1 gives some examples of the entropy of hydration of anions TABLE I1 ENTROPIES OF HYDRATION OF U"" Fe"' Hydration entropy . . .. -78 -70 Entropy change on formation of hydroxide . . . . 3-50 +50 Entropy change on formation + 35 of chloride.. .. .. - F' Entropy of hydration . . - 2.3 Entropy change on ferric Entropy change on stannous complex formation . . + 49 complex formation . . - IONS EXPRESSED AS CALORIES PER " c Cr"' Cd" Sn" Mg" Ca" Sr" Ea" -73 -14.6 -9 -28 -13 -9 $3 - - - 3-43 - 3-23 - - - - $23 $8 f14 - 013' C1' Br' I' - 2.5 + 13.2 + 19.2 +26.1 + 36 + 23 - + 50 + 23 + 14 +8 - and cations. Latimerlo has shown that these values can be expressed in a general formula- 2702 3 so =-R.log 2 M + 37 --, Y2 where M, z and Y are the molecular weight, charge and effective radius of the ion.Thus it is to be expected that the entropy change upon complex formation should be related to z/r2. Definite proof of such an exact relationship is not available as yet, but the data in Tabie I indicate that in the reaction between one cation and a series of anions, the entropy change is related to the inverse of the anionic radius, F' >, OH' > C1' > Br'. Again, in the reactions of a series of cations with one anion, e.g., chloride or hydroxide in Table I, the order of the entropy changes is U"" = Fe"' > Cr"' > Cd" = Sn", an order that coincides with that of the order of the ionic potentials Z / Y .If further water molecules are released when the complex is precipitated or extracted, the entropy changes will again follow the same sequence. The general parallel between precipitation and complex formation will be discussed in more detail on p. 589. No specificity of reaction can be expected from the entropy changes on complexforina- tion, but it is clear that these changes can result in a high degree of selectivity. Organic reagents that are anions will react preferentially with highly charged small cations on this basis. Furthermore, as it is known that the reaction of cations with neutral molecules involves only small, or even opposed, entropy changes on complex formation (see Table 111),588 WILLIAMS: A SYSTEMATIC APPROACH TO THE CHOICE [Vol.78 highly charged small cations are to be expected to react rather with anions than with neutral molecules, provided that the heat of reaction is not overwhelmingly large for the latter. TABLE I11 HEATS AND ENTROPIES OF FORMATION OF COMPLEXES BETWEEN CATIONS AND NEUTRAL MOLECULES Reaction AG, AH, AS, Reference K cals. K cals. cals. per O C Mg"-2NH3 - 0.3 - 1.2 - 3-0 Nat. Bur. Stand. Circular8 Hg"-4NH3 - 26.2 - 28.5 - 7.7 Fyfel' CO' '--5N H3 - 42.4 - 49.4 - 23.3 Nat. Bur. Stand. Circulars Cu"-En - 14.2 - 18.9 - 15.0 W illiamslZ Fe"-3Dipy - 24.5 - 24.5 0.0 Williams5, re Cu"-4NH, - 16.6 - 19.7 - 10.0 >> Ag;'-2NH3 - 10.0 - 13.3 - 11.0 39 2n"-En - 7.7 - 9.8 - 7.0 3) Ag'-Py - 2.8 - 4.7 - 6.3 39 NOTES-1. In this table "En" represents ethylenediamine, "Dipy" dipyridyl and "Py" pyridine.2. Many other examples will be given in a further publication.5 HEAT CHANGES ON FORMATION OF COMPLEXES A consideration of reaction (2) shows that the loss of free energy in the formation of a complex, ML, must be considered in conjunction with the gain in free energy consequent upon the decrease in hydration. In the same way as the entropy changes in the reaction are dependent upon the entropy of hydration of the reactants, so is it to be expected that the heat of reaction will be dependent upon the heat of hydration of the ions and molecules involved. However, whereas the entropy change in the reaction is more favourable for reactants of greater hydration, the heat change will be more unfavo~rable.~ Hence ions that form mainly ionic complexes, such as aluminium,111 thoriumIV and magnesium"-- all ions of low electronegativity and weak acceptors-form complexes with anions of high electronegativity, i.e., weak donors, only through the agency of large entropy changes that are opposed by the heat of reaction.Table I contains many examples of complexes formed in this way; several other examples are discussed el~ewhere.~ It is only to be expected that the heats of formation of complexes of organic reagents that are anions and weak donors, and which, in addition, are large molecules, will be even further opposed to complex formation (by these ions), as the "interference volume" of the ligands around the cations will be large; large repulsion terms between the co-ordinated residual water molecules and the combined reagent will be inevitable.This repulsion term will be much smaller if the reagent is a chelating agent; much of the stability of chelate complexes arises from the small interference volume of these co-ordinated molecules.5 Metal ions that are weak acceptors will have little tendency to bind electropositive neutral molecules that are good donors. Furthermore, the entropy changes in these reactions are not favourable to complex formation. The over-all result is that these cations of high ionic potential and which are also poor acceptors will always form hydroxide complexes (with the aid of large favourable entropy changes) at a much lower pH value than they will forin complexes with neutral molecules. It is now possible to discuss the type of organic molecule that will be suitable as a reagent for these cations.CATIONS OF LOW ELECTRONEGATIVITY AND HIGH IONIC POTENTIALS- The extreme members of this first group of cations are the very small quadrivalent ions of the metals of group IVA of the periodic table, namely, titanium, zirconium, hafnium and thorium. A suitable organic reagent for these very small ions must be anionic and, preferably, a dibasic acid, so that two molecules of the reagent can completely neutralise the cation charge (four molecules of a monobasic acid would have a much larger interference volume). The reagent must co-ordinate through oxygen atoms, weak donor groups, for otherwise it will react with the larger, more electronegative cations, such as thallium"'. Finally, the insolubility of the complex produced should not depend upon the formation of a continuous lattice, for small cations are not easily built into such structures because of the radius ratio e f f e ~ t .~ An ideal reagent for these cations is an organic arsenite, such as phenyl- arsonic acid, C,H,.AsO(OH),. The sulphiiiic acids are also suitable, but a disulphinic acidOctober, 19531 OF ORGANIC REAGENTS FOR METAL IONS 589 might be even better. As the co-ordinating groups are made more electropositive, i.e., better donors, the selectivity of the reagent is reduced and the conditions for reaction with certain tervalent cations coincide with those for the reactions of the quadrivalent ions. This is already apparent with the reactions of the lake-forming phenols, such as purpurin, I, and titan yellow, 11.Ferric, gallium and chromic ions react with these reagents under the HO.SO, HO 0 OH HO ()yyoH 0 OH 'Y ' SO,.OH 'OH I I1 same conditions as zirconium. If the co-ordinating groups are greatly increased in volume, the region of overlapping reactions is still further extended. The stability of the ethylene- diaminetetra-acetic acid complex of titaniumIv, log,, K = 17-3, is much lower than that of the ferric complex, log,, K = 25.3, and is not much greater than that of lanthanum", log,, K = 15.4. The two nitrogen atoms in ethylenediaminetetra-acetic acid, 111, are good donors. HOOC.CH, )N.CH,.CH2.N~2*cooH HOOC.CH, CH,.COOH I11 When devising a reagent for a tervalent ion that is a poor acceptor, such as aluminium, scandium, or a rare earth ion, maximum selectivity from the reactions of bivalent ions can again be achieved by making use of the larger entropy changes of the reactions of the tervalent ions.Thus carboxylic acids, phenols and carbonyl groups, which can form chelates with the cations, make excellent reagents. Their reactions selectively depend upon the size of the cation. Typical reagents of this kind are alizarin, IV, aurin tricarboxylic acid and salicylic acid, V. The size of the ligand is not so critical as it is with quadrivalent cations, but the 0 OH OH 0 IV V reagent still must not be a good donor. an overlap of reactions with the more electronegative metal ions, such as cupric. reason oxine, VI, discriminates poorly between bi- and tervalent ions. is a good donor. An increase in donating power immediately introduces For this Its nitrogen group OH VI For a bivalent metal ion that is also a poor acceptor, such as magnesium, calcium, strontium or barium, it is clear that, as the importance of the entropy of complex formation is now smaller, it will not be possible to separate it from other, larger, bivalent ions that are good acceptors.The procedure is now to find reagents for the good acceptors instead (see p. 591). It is still interesting to know if it is possible to separate the four ions of group IIA from one another by means of organic reagents. Table IV illustrates the principles on which the analytical chemist should act in approaching this problem. The greatest difference between the stabilities of the magnesium and the calcium complexes occurs with glycine. A much smaller difference is found for the oxalate complexes and, in the lattice of these salts, calcium is more stable, as shown by the much greater solubility of the magnesium complex.A reagent for magnesium that will not react with calciuin590 WILLIAMS: A SYSTEMATIC APPROACH TO THE CHOICE [Vol. 78 must, clearly, have a small interference volume and, also, must not form a continuous lattice, such as that found in oxalates.* Oxine is one such a reagent; it is selective for magnesium in the presence of calcium. Other reagents that also suggest themselves are compounds such as VIII. Such a molecule would, in some respects, resemble chlorophyll, which is a very selective magnesium reagent.13 Many examples are given elsewhere, most of them from the work of Schwarzenbach.4 "Magneson," VII, is a similar reagent.Selectivity for calcium must be based on the size of the ion. TABLE IV THE LOGARITHM OF STABILITY CONSTANTS OF SOME ALKALI - EARTH METAL COMPLEXIS Metal ion Reagent Mg Ca Sr Glycine . . .. .. .. 3.44 1-43 Oxalate .. * . .. .. 3.43 3.00 2.54 (4.07) (8.64) (7.25) Tartrate . . . . * . .. 1-36 1.80 1.65 Thiosulphate . . .. .. .. 1-84 1.92 2.04 (sol.) (3.00) Nitrate .. .. .. .. 0.00 0-28 N(CH,.COOH), . . .. .. 5-41 6-41 4.94 (sol.) (sol.) (4.64) (6.55) Sulphate . . .. .. .. 2.15 2-28 NoTE-The numbers in parenthesis are logarithms of solubility products. Ba. 0.77 4.82 2-33 (6-96) 1-62 2.33 (4.00) 0.92 (2.35) (10.00) The data in Table IV show that the stabilities of the complexes of the strong acid anions follow the sequence of the radii: Ba" > Sr" > Ca" >Mg".This sequence can only arise in one way. The heat of formation of these complexes so strongly opposes the reaction that it controls the order of the stabilities despite the contrary influence of the entropy changes4 This only occurs with the strong acid anions. It must result from the variation in the hydration of the cations. The difference between the reactions of the various ions is much more noticeable in the solubility products and these differences too can be traced to the large differences in the heats of solution for the various cation salt^.^,^ If an organic reagent is to be prepared for barium, it must be similar in character to the sulphate anion. I t must be a derivative of a dibasic and, preferably, strong acid, it must form a continuous CH,.CH.COOH.CH, \ N / r \ /<N )=/ No2-/-\ C N - O H \=/-N=N\J \L/ \ CH,.CH.COOH.CH, lattice and it must be a large molecule.One or two reagents of this type are known, e.g., rhodizonic acid, IX. Molecules such as those of the sugar acids, which also contain large / OH VII VIII 0 0 IX numbers of hydroxyl groups, would also seem suitable. Note the way in which hydroxyl groups stabilise the larger ions on passing from the oxalates to the tartrates in Table IV. Highly fluorinated compounds, such as X and XI, or nitrated compounds, might also serve. 7 2 , CF,.COOH HO.HC CF.COOH I CF.COOH I I CH.OH I HO.HC CF2.COOH 'CF,' X XIOctober, 19531 OF ORGANIC REAGENTS FOR METAL IONS 59 1 The principles on which a search should be made for organic reagents for the larger group IA metals should be related to the above, but there should be only one ionising group in any molecule chosen and as many other co-ordinating groups of a highly polar character as possible.An excellent example of such a reagent is compound XI1 for potassium. Reagents for magnesium should always be tried as reagents for lithium. NO2--(-~H-->>O2 Y O 2 \ - \ NO, NO, XI1 ELECTRONEGATIVE METAL IONS- Molecules such as ammonia and ethylenediamine, cyanide ions, iodide ions and bromide ions react with metal ions such as niercury", platinum", platinum", thallium'II, gold' and silver1 with a considerable evolution of heat,5 and with but small entropy changes. Not all these groups can be introduced into organic reagents without modifying their activity to such an extent as to make them unreactive, e.g., iodide.However, both nitrogen and sulphur co-ordinating groups are easily prepared in many organic compounds. As specific reagents for the electronegative metals such compounds are ideal. A great number are already known, amongst which are dithizone, XIII, thionalide, XIV, substituted thioureas, /r\ >J /N=N \NH-NH s=c XI11 XIV XV, and substituted nitroso-anilines, XVI. As the metal ions are large it is an advantage if the combining ligand molecule occupies a large volume around the cation so that. steric factors act to increase the selectivity of the reagent for these large ions. I t is noteworthy that the selectivity of these reagents is of a different order from that discussed in the previous section.For example, apart from osmium and ruthenium, few metal ions are able to react xv XVI with tri- and tetra-ethyl thioureas, even at high pH values. As has been pointed out elsewhere, the selectivity of a reagent such as dithizone differs considerably from that of oxime.1 The most electronegative cations react with many of the reagents in the above group even in strong mineral acid solutions. Between this group of cations and the cations in the large first group come a large number of ions with somewhat similar properties; these vary from cupric ion, which is almost to be classed with the good acceptors, and manganous ion, which is not much different from a poor acceptor such as magnesium. For example, along the series of bivalent ions, manganese, iron, cobalt, nickel and copper, the acceptor properties increase steadily, and selectivity of reaction will depend largely upon a careful choice of conditions.The problem has been discussed in detail in an earlier paper.1 In general, the larger the ion and the higher its electronegativity as measured by its ionisation potential, the more easily it will react with compounds that are good donors. Furthermore, the lower the charge on a subgroup-B metal, the more easily does it form complexes with strong donors. These and other trends amongst the stability constants of such metal ions have been discussed a1ready.l There are two other methods of obtaining selective reactions, which are more important in the present discussion : a change of valency and a change of ground-state of the ion involved in the reaction. Often the two changes are interdependent.592 WILLIAMS: A SYSTEMATIC APPROACH TO THE CHOICE [vol.78 CHANGE OF VALENCY- In general, if there is a lower valency state readily available, a change to this state from the stable valency of the metal in aqueous solution will take place on the replacement of water by a more electropositive ligand provided that no radical electron rearrangement is involved. A change to a higher valency state will be brought about by a change of ligand to a more electronegative group than water, again provided there are no radical rearrange- ments of the non-valency electrons. High-valency states are usually found in oxide (0") complexes ; manganesevI1, chromiumV1, molybdenumV1 and ironlIi are amongst many others.Low-valency states are usually found with cyanides, e.g., with such cations as nickel1, palladium1, manganese1 and molybdenum". Such changes of valency have great importance for the analyst, because a cation can be made to react selectively first in one valency state and then in another. The reduction of cupric ions to cuprous with iodide and the reduction of both ferric and cupric ions with thiols are well-known examples. Auric and thallic ions are most stable when they are bound to oxygcn in anionic complexes, whereas the lower valency states of gold and thallium are most stable when bound to cyanide. There are very few organic compounds that bring about these changes and little effort seems to have been made to find them.Reagents that increase the valency of the metal ion must be oxidising agents that form strong complexes through oxygen atoms. Hydrogen peroxide is a clear example. It is used in the preparation of peroxy-acids of chromium from a chromic salt solution. Hence it seems that other per-acids of organic carboxylic acids should produce the same effect. The efficacy of reducing agents is well known, the best examples being with thio-compounds such as thiourea and thioglycollic acid. Moreover, dithiocatechol is a common reagent for molybdenumn. There might be phosphorous compounds that would also bring about such reactions. ORBITAL CHANGES- Many ions of the transition metals are able to change their electronegativity by a change of electronic structure.The greater the d-orbital character in the bonds of these ions the more electronegative they become. These changes have often been suggested as a possible source of specific reactions. The ferrous state is stabilised relative to the ferric state by complex formation with dipyridyl, o-phenanthroline or cyanide, but not by complex formation with hydroxyl, chloride, ethylenediaminetetra-acetic acid or oxine. The former are the more electropositive reagents, but they also form complexes with ferrous ions that are diamagnetic. In these complexes ferrous ion binds the ligands with d2sp3-orbitals. A selective valency change that is different from the one discussed above, and is opposed by the tendency of the lower valency state to combine with the good donor, can be brought about if on losing an electron the ion reaches an upper state in which the possible orbital combinations are much stronger acceptors.A well-known example is the reaction of cobaltous ion with cyanide. The reaction of this ion with a-nitroso-P-naphthol also depends upon the change in electronegativity in the tervalent state. Other selective reactions caused by orbital changes are those of nickel and palladium with dimethylglyoxime. The problems of whether nickel can be made ter- or quadrivalent in solution have yet to be tackled and it may be possible to stabilise manganese in a quadrivalent state. The stabilisation of such valency states may require very powerful donor ligands such as arsenic or phosphorus compounds. Orbital changes often involve changes in stereochemistry, and specificity can arise through the choice of the reagent to fit the steric requirements of the cation.For this reason 2:2'- diquinolyl reacts with cuprous (tetrahedral) ion, but not with ferrous (octahedral) i0n.l No systematic search has been made for other examples of this kind, but one or two suggest themselves. The heat of hydration is hardly different from that of the cupric ion. The reason €or this stability is not known, although it is either due to the Stark splitting of the d-energy levels in the octahedral field or else to the interaction with the incompletely filled 3d band. The d-orbitals can only supply additional stabilisation if the nickel ion is surrounded by six octahedrally placed molecules or four planar molecules.Now, a ligand can be chosen that is too large to fit anything but a tetrahedral arrangement. This complex will be unusually unstable relative to the nickel hydrate. Some examples Some clear examples are known. The hydrate of nickel is peculiarly stable.October, 19531 O F ORGANIC REAGENTS FOR METAL IONS 593 are known. Nickel chloride is less stable than either cobaltous or cupric chloride; the substituted 2-methyl oxine complex of nickel is less stable than the corresponding cobaltous HN-CO XVII complex or, at least, it is no more stable than the latter despite the considerably greater stability of the nickel oxinate. Surely, if much larger substituents are used this effect must be greatly enhanced. If Orgel’s calculations are correct, it might well be possible to reduce the stability of nickel complexes by a factor of 10 logarithm units relative to the corre- sponding zinc complex.13 The dithizonates may be a case in point.CH i-Pr I i-Pr XVIII XIX The difference in hydration can be used in other metal complexes. The reaction of argentous ions with ammonia is somewhat sensitive to the substitution of alkyl groups for nitrogen. However, this ion will form amine complexes even with tertiary amines. Although cupric and nickel ions form amines of a stability greater than that of argentous ions, the stability of their complexes is considerably reduced by substitution of the ammonia molecule. Tertiary amine complexes have not been reported. At the same time, a specific reagent for argentous ion in the presence of large amounts of either nickel or cupric ions is a tertiary amine of formula XVII.There must be other instances of complex formation being made deliberately to favour a large ion of slight hydration as opposed to a small ion of large hydration energy. The selectivity of dithiols must be greatly increased if large groups are substituted ortho with respect to the thiol groups. Such a reagent as that shown in XVIII might be of use for stannous or plumbous ions and might not react with the other group elements of the analysis tables, viz., coppern and cadmium. The bias in the design of the above reagents has been in the favour of the large ions. Surely there must be ways of designing molecules so that the co-ordinating groups are too close together to react with large ions. The porphyrins are natural compounds of this kind, but thiophene derivatives of the same kind (XIX) might be more selective. The approach to the design of organic reagents can be greatly improved and it is hoped that this article has shown a few possibilities. In conclusion, it must be pointed out that it has only been possible in this article to quote a few examples from the reagents commonly used for different classes of cations. Reference to standard works on the subject of the selectivity of organic reagents15 will immediately supply many more examples that will confirm some of the general remarks made. REFERENCES 1. Irving, H., and Williams, R. J. P., AnaZyst, 1952, 77, 813. 2. -,- , Nature, 1948, 162, 764. 3. Evans, M. G., and Uri, N., in “Society for Experimental Biology, Symposium No. 5,” Cambridge University Press, 1951, p. 130. 4. Williams, R. J. P., J . Chem. SOL, 1952, 3770. 5. - , J . Phys. Chem., in the press. 6. “Selected Values of Chemical Thermodynamic Properties,” National Bureau of Standards, 7. Vanderzee, C. E., J . Amer. Chem. SOC., 1952, 74, 3552 and 4806. 8. Evans, I. J., and Monk, C. B., Trans. Faraday SOC., 1952, 48, 934. Circular 500, Washington, D.C., 1952.594 ROGINA AND DUBRAVCIC : MICRO-DETERMINATION OF IODIDES BY [Vol. 78 9. 10. 11. 12. 13. 14. 15. Davies, C. W., and Wyatt, P. A. H., Trans. Faraday SOC., 1949, 45, 770. Latimer, W. M., and Powell, R. E., J . Chern. Phys., 1951, 19, 1139. Fyfe, W. S., J . Chern. SOC., 1952, 2018 and 2023. Williams, R. J. I?., Biol. Rev., 1953, in the press. Orgel, L. E., J . Chern. SOG., 1952, 4756. Williams, R. J. P., I b i d . , 1953, in the press. Feigl, F., “The Chemistry of Specific, Selective and Sensitive Reactions,” New York, Academic OXFORD March 3rd, 1053 Press Inc., 1949. MERTON COLLEGE

ISSN:0003-2654    

DOI:10.1039/AN9537800586

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

594 ROGINA AND DUBRAVCIC : MICRO-DETERMINATION OF IODIDES BY Micro-determination of Iodides Catalytic Reduction of Arresting Ceric Ions [Vol. 78 the BY B. ROGINA AXTI M. DUERAVCIC A method is described for the determination of small amounts, 0.01 t o 1.00 pg, of iodine as iodide. The method is based on the catalytic effect of iodides on the reduction of a ceric salt by arsenious acid, The rate of reduction is readily determined by arresting the reaction a t a given time by the addition of ferrous and thiocyanate solutions and by measuring the resulting red colour of ferric thiocyanate by means of a photometer. The method offers advantages over that described by Chaney in which the rate of reduction is measured while the reaction is in progress. Results are quoted t o show the accuracy of the proposed method.A reproducibility of &0-002 pg has been attained in the range from 0.01 to 0.10 p g of iodide in 8 ml of aqueous solution. IT has been known for some time that iodides have a catalytic effect on the reduction c!f ceric sulpliate by arsenious acid in sulphuric acid solution. The reaction takes place in the following way- ... 2Ce... + As....' 2Ce"" + As The gradual progress of the reaction can be followed by the steady disappearance of the yellow colour of ceric ions. Sandell and Kolthoff1y2 found the speed of the reaction to be nearly proportional to the iodide concentration, and they used this reaction for the deterrnina- tion of small amounts of iodides. The conditions of the reaction (concentration of reagents, acidity and temperature) were fixed, and iodides were determined by measuring the tiine required for the complete disappearance of the yellow colour.He preferred to measure the concentration of the remaining non-reduced ceric ions after a fixed time, instead of waiting for complete decoloration (reduction). It has been observed that some inevitable technical difficulties occur in measuring the concentration of ceric ions while the reaction is in progress. The catalytic reaction t:l.kes place in a thermostatically controlled enclosure, whereas the coiicentration of the remaining ceric ions is measured elsewhere with a photo-electric absorptiometer. Because of instability of the electricity supply, frequent adjustments to the photometer are necessary and the measurerLents are therefore protracted.It is difficult to carry out the measurement quickly enough to avoid inaccuracies arising both from the reaction being still in progress and from changes of temperature. We found that, after a suitable time had elapsed, the progress of the catalytic reaction could be arrested. By addition of an excess of ferrous ions at a given moment the remaining ceric ions are immediately reduced as follows- A more convenient method based on the same reaction was developed by This results in errors in the determination of iodides. Ce"" + Fe" -+ Ce"' + Fe"' The resulting amount of ferric ions is equivalent to that of ceric ions present immediately It is also proportional to the intensity of the red The depth of this colour is The before the catalytic reaction was arrested. colour that appears after addition of a thiocyanate solution.inversely related to the iodide concentration and can be determined photometrically. amount of iodide is deduced from a previously prepared calibration graph.October, 19531 AI~RESTING THE CATALYTIC REDUCTION OF CERIC IONS 595 By the method described, the colour can be measured with a colour comparator or a I n the visual type of photometer without undue haste and therefore more accurately. method described, a Zeiss - Pulfrich photometer with a green filter has been used. EXI’ERIMENTAL ICATIO OF EXTINCTION *ro C E I ~ 10s CONCEXTRATICN- As the first step in the development o f the method, the relation hetween the rate of reduction o€ ccric salt and the photometer druin readings was examined.To solutions of ceric salt of different concentrations, ferrous and thiocyanate solutions were added and the optical densities were measured. I t was found that a 2 x N Concentration of ceric salt produced a considerable extinction (about 2) which, however, could be measured with sufficient accuracy. N of ceric ions was chosen as the most suitable initial (maximum) concentre t’ ion. To 10 rnl of different concentrations of ceric salt in solutions N in sulphuric acid, 1 ml of 1.5 per cent. ferrous solution and Z ml of 4 per ccnt. thiocyanate solution were added. The extinctions of the red colour so produced were measured in 0.5-cm cells with the green filter S 50 (maximum transmission at 488 mp). These extinctions, plotted against the concentration of ceric ions, gave a straight line (Fig.1, graph *4). Therefore, a concentration of 2 >= Fig. 1. Extinction values of ferric thiocyanate solutions plotted against concentration of ceric ions. Measure- ment in 0-5-cm cells with a green filter of maximum transmission at 488 mp Considerably higher initial concentrations of cei-ic salt were used by the authors quotedm2 9 3 For comparison, solutions of ceric salt of various concentrations to a maximum o€ 2 x AT, made 1.2 N in respect of sulphuric acid, were measured directly. As shown by Fig. 1, graph B, the extinction values are considerably lower and there is not such a good straight line relation as was found with the lower concentrations. THE CATALYTIC REACTION AX 1) CALIBRATION GRAPHS- The amount of iodide could only be determined on the basis of the rate of the reduction of ceric ions if all the conditions under which the catalytic reaction takes place were fixed and adhered to.In laying down the conditions we availed ourselves partly of the experiences of other author^^,^ after taking into consideration our particular circumstances. For the reasons mentioned above, a reaction mixture 2 x A solution 5 x For a satisfactory rate of reduction, the temperature and duration of the reaction should be adapted to the range in which lie the iodide concentrations that have to be determined. N in ceric ions was used. N in arsenious acid, and N in sulphuric acid was found to be suitable.596 ROGINA AND DUBRAVCIC : MICRO-DETERMINATION OF IODIDES BY [Vol. 78 The most suitable conditions for the “high range,” i.e., for amounts of iodide between 0.1 and 1.0 pg, were found to be a temperature of 20” C and 8 minutes as the limit for the duration of the reaction.For the “low range,” i.e., for 0.01 to O-lOpg of iodides, a temperature of 30” C and a reaction time of 20 minutes were used. With the above conditions and the technique described in the procedure, calibration graphs as shown in Figs. 2 and 3 were plotted. Fig. 2. Calibration graph for 0.01 t o 0.10 pg of iodide (“low range”). Temperature, 30” C; reaction time, 20 minutes Iodide. p g Calibrationgraph for 0.1 to 1.0 pg of iodide (“high range”). Temperature, 20” C; reaction time, 8 minutes Fig. 3. Interferences with the catalytic effect of iodides by other ions were not especially studied in this work; this point is covered in the literat~re.~,~,~j6 It need only be mentioned here that chlorides, bromides and, especially, osmium ions have a catalytic effect on the reduction of ceric salt, and cyanides, thiocyanates, mercuric and silver ions have an inhibitory e f f e ~ t .~ ~ ~ Indifferent ions, such as the sodium ion, decrease the speed of the catalytic reaction when present in high concentrations. If such ions are present, the reference standards must contain the same amount of indifferent ion, and a corresponding calibration graph must be drawn. STABILITY OF THE COLOUR- Tempera- ture has niost influence on the colour. The rate of fading of ferric thiocyanate increases with increasing temperature. To minimise any differences in temperature, the solutions must be set aside at room temperature for 45 minutes after the red colour has been developcd.The measurement can be made at any time after this waiting period has elapsed, in contrast to the usual practice in iron determinations. Owing to the lower conccntration of potassium thiocyanate, the colour of the ferric thiocyanate solution is stable enough and does not fade, provided the room temperature remains substantially constant. The temperature of measurement should not differ widely from that at which the calibra- tion curve was recorded. Readings taken at different temperatures can nevertheless be compared if reference standards or numerical corrections are used. Ovenston and Parker6 found that the extinction of ferric thiocyanate solutions decreased by about 1 per cent.for every degree centigrade rise of temperature. We found such a correction to be useful. It is more accurate, however, to make in every set of determinations a few measurements of known amounts of iodides in order to obtain reliable reference values. The stability of the colour of ferric thiocyanate was found to be satisfactory. METHOD APPARATUS- A rack with test tubes of about 2 cm in diameter and 18 cm in length. Pipettes.October, 19531 ARRESTING THE CATALYTIC REDUCTION OF CERIC IONS 597 A water-bath controlled at 20" and 30" C (FO.2" C), fitted with a transparent front, A Pulfrich plzotometer with green filter No. S 50 (with maximum transmission at 488 mp) A stop-watch. NoTE-A~~ glassware should be soaked in a strong sulphuric acid - bichromate mixture and then thoroughly rinsed, first with tap water and then with distilled water.Sulphuric acid, 60 per cent. w/w. Arsenious acid-A 0.1 N solution in about 0.01 N sulphuric acid. large enough to contain the rack with the test tubes and the bottle of ceric salt solution. and 0 6 c m cells. REAGENTS- Dissolve by heating 4.946 g of arsenic trioxide in about 500 ml of water acidified with 10 drops of sulphuric acid. Dilute to 1 litre. Ceric ammonium szdphate solution-A 0.02 N solution in 1.6 N sulphuric acid. Dissolve 13-38 g of Ce(S0,),.2(NH4),SO,.4H,O in water, add 44 ml of sulphuric acid, sp.gr. 1434, and dilute to 1 litre. Ferrous ammonium sulphate solution-A 1.5 per cent. w/v solution of FeS04. (NH,) S0,.6H20 in 0.6 per cent. sulphuric acid.Potassium thiocyanate solution, 4 per cent. w/v. Standard potassium iodide solzztions-Dissolve 261.6 mg of potassium iodide in 1 litre of water to give a. solution containing 200 pg of iodine per ml. Dilute the solution to 1 in 200 and 1 in 2000 to make "high range" (1.0 pg per ml) and "low range" (0.1 pg per ml) standard solutions. NOTE-Stock solutions must be stored in a dark place and kept tightly closed. PROCEDURE- If the approximate concentration of the iodide solutions is unknown, make a preliminary test to determine what dilution should be prepared to obtain concentrations corresponding either to the "high range" (0.1 to 1.0 pg in 8 ml of sample solution) or to the "low range" (0.01 to 0.10 pg in 8 ml of sample solution). The "high range" concentrations give at 20" C an appropriate rate of reduction after 8 minutes.The conditions hid down should be scrupulously observed (20.2" C and &5 seconds), both in the preparation of the calibration graph and for the determination of iodides in samples. For lower concentrations, higher temperatures and longer periods of time should be used. For the "low range," use a water-bath temperature of 30" C & 0.2" C and a reaction time of 20 minutes. Several samples can be anslysed at intervals of 1 minute in one set of determinations. As the reaction lasts 8 or 20 minutes, depending on the range used, and as 1 minute is required for the necessary laboratory work with thc test tube, a maximum of 8 or 20 test tubes, respectively, should be used. To increase the accuracy, every set of determinations should include from two to four appropriate reference standards. These standards will indicate any difference, however small, between the conditions used in determination and those used in the preparation of the calibration graph, and a correction to the photometer readings can be made, if necessary.Place in each test tube 8.0 ml of iodide solution, 0.5 ml of arsenious acid and 0.5 ml of sulphuric acid. If less then 8 ml of iodide solution is taken because the concentration is above the normal range, add water to make the total to 8ml. Shake each tube thoroughly to mix the contents, rinse the inside walls of the tubes by rotation, and place the rack with the tubes in the water-bath at the appropriate temperature (20" or 30" C).After about 20 minutes, the temperature of the tube contents will be that of the bath. Add to the first tube 1 ml of ceric ammonium sulphate solution (at the same temperature), start the stop-watch, mix the contents of the tube by shaking and put it back on the rack. Repeat the procedure with the second tube, timing the addition of the ceric ammonium sulphate to coincide with a reading of 60 seconds on the stop-watch. At 1-minute intervals repeat the addition of the ceric ammonium sulphate solution to the other tubes. After 8 or 20 minutes, according to the range, arrest the catalytic reaction in the first tube by adding 1 ml of ferrous ammonium sulphate solution and mixing. The remaining598 ROGINA AND DUBRAVCIC MICRO-DETERMINATION OF IODIDES BY [VOl. 78 yellow colour disappears suddenly. Add 1 ml of potassium thiocyanate solution, which will cause the red colour of ferric thiocyanate complex to appear.Place the tube on a second rack beside the water-bath. Arrest the reaction in the other test tubes at 1-minute intervals, and develop the red colour as described. Place the rack with red coloured solutions near the Pulfrich photometer and allow the solutions to attain room temperature. After 45 minutes, or later, measure the extinction of the solutions with a green filter, S 50, and O-5-cm cells, with water in the reference cell. From the measured extinctions evaluate the iodide concentrations from a calibration graph prepared at a temperature not differing greatly from the prevailing room temperature. If the values for the standard solutions included in the set do not exactly fit in the calibration graph, draw a parallel line through the new standard points, and read the iodide concentration from this. RESULTS The reproducibility of determinations of iodide by the method described is shown in Four different amounts of iodides (two at “low range’’ and two at “high range”) The calibration The results show that at the “low range” the amount Table I.were each determined 12 times, the determinations being spread over 3 days. graph was prepared on another day. of iodide found is within &0.002 pg of the amount taken. is greater, but it does not exceed 4-2 per cent. of the iodide taken. At the “high range” TABLE I KEPKODUCIBILITY OF RESULTS First . . Second . . Third . .First . . Second . . Third . . First . . Second . . Third . . First . . Second . . Third . . .. .. . . . . .. .. .. .. .. . . . I .. Iodine taken, Pg 0.030 77 7 9 0.080 77 7) 0.300 97 77 0.800 77 77 Iodine found, 0.032 0.029 0.031 0.029 0.028 0.030 0.029 0.028 0.030 0-031 0.03 1 0.029 0.081 0.082 0.079 0.078 0.078 0.078 0.079 0.079 0.080 0.080 0.082 0.080 0.310 0.307 0.304 0.304 0-295 0.295 0.296 0.307 0-304 0.295 0.296 0.300 0-81 7 0.817 0.817 0.808 0.753 0.834 0.800 0-766 0.800 0.792 0.800 0.808 Pg I I I Range of errors, Pg & 0.002 * 0.002 + 0.010 - 0.005 f 0.034 the error Error, /O & 6.7 0 1 f 2.5 + 3.3 - 1.2 -+ 4.2 DISCUSSION OF RESULTS It is reasonable to suppose that Chaney, too, met with difficulties in measuring the concentration of the remaining ceric ions while the reaction was in progress.In his later publication4 he proposed a technical improvement of the photometer. The catalytic reduction was carried out in a special photo-electric absorptiomcter with a built-in thermostat. A vacuum- type photo-tube, an electronically regulated power supply and a continuous recording device were used to improve sensitivity. Chaney claimed that the ultimate sensitivity of this method of measurenient ~7as 0.001 pg of iodine in 5 ml of reaction mixture. His repro- ducibility figure of +0.001 pg was attained in the range of 0.05 to 0.10 pg. He did. not state whether the same reproducibility could be attained for amounts below 0.05 pg. Table I shows that, in the range of 0.01 to 0*1Oyg of iodides, we attained in a much simpler way a reproducibility of +0.002 yg in 10 ml of reaction mixture.For less than 0.01 pg of iodide the accuracy was not so great because of the variable catalytic effect of the lowest iodide concentrations. The effect is often upset by the traces of other elements present as impurities. These concentrations of iodides could, therefore, be considered a “region of uncertain reaction”.’ The red colour of the ferric thiocyanate complex has advantages for measurement as compared with the yellow colour of ceric salt solutions. Although the latter show the most significant absorption in the violet and ultra-violet region of the spectrum, solutions of ferric Sometimes there is a slight catalytic effect, as shown in Fig. 2.October, 19531 ARRESTING THE CATALYTIC REDUCTION OF CERIC IONS 599 thiocyanate have greater absorption in the blue and green regions.As the sensitivity of the human eye and the sensitivity of most photo-electric cells rises towards yellow-green, it is more convenient to measure a red colourcd solution than a yellow one. In addition, the absorption of the ferric thiocyanate solutions is greater and covers a larger scale of extinction values than does the absorption of ceric solutions alone (Fig. I), even though the latter are of a higher concentration. It was found that the solution of ferrous ammonium sulphate, acidified with sulphuric acid, was sufficiently stable for some weeks. ,4 slight increase in extinction values, up to 0.10, caused by the gradual oxidation of this reagent by atmospheric oxygen must be taken into account. Standard solutions of potassium iodide, prepared several months before, showed the same catalytic effect as freshly prepared solutions. No change in activity of the otlier reagents was observed. REFBREXCES 1. 2. -,-- , Mikrochern. Acta, 1937, 1, 9. 3. 4. __ , Anal. C h e w , 1950, 22, 939. 5 . 6. 7. Sandell, E. B., and Kolthoff, I. M., J . Anzcr. Chern. Soc., 1934, 56, 1426. Chaney, A. L., Ind. Eng. Chern., Anal. Ed., 1940, 12, 179. Moore, J., and Anderson, R., J . Anzer. Chern. Soc., 1944, 66, 1476. Ovenston, T. C. J., and Parker, C. A., Anal. Chinz. Acta, 19413, 3, 277. Feigl, F., “Chemistry of Specific, Selective and Sensitive Reactions,” Academic Press Inc., New York, 1949, p. 14. CENTRAL INSTITUTE OF HYGIENE ZAGREB, YUGOSLAVIA January lDth, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800594

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

October, 19531 ARRESTING THE CATALYTIC REDUCTION OF CERIC IONS 599 The Determination of Potassium and Traces of Sodium in Some Potassium Salts BY C. JACKSON For determining sodium in certain potassium salts of weak acids the sample is titrated with 0.2 N perchloric acid in glacial acetic acid, a small controlled excess being added. After the precipitated potassium perchlorate has been removed by filtration, the filtrate is evaporated to dryness and the sodium is determined as sodium zinc uranyl acetate. It is thus possible to estimate volunietrically the total equivalent alkali metal and gravimetrically both iodium and potassium on one sample. The method can be applied to the determination of sodium in most potassium salts. THE determination of sodium as sodium zinc uranyl acetate has received much attenti0n.l *2 9 3 Although the method is applicable in the presence of appreciable amounts of potas~ium,~s~ it cannot be applied when the potassium is in large excess, as, for example, in the determination of sodium in potassium salts, owing to precipitation of potassium zinc uranyl acetate.More- over, the reagents commonly used for precipitating potassium produce a filtrate unsuitable for sodium determinations ; sodium cobaltinitrite is obviously ruled out and perchloric acid in an alcoholic medium leaves a filtrate that cannot be evaporated without danger. Flame phot0metry6~7~~~~ is rapid and convenient, but in many laboratories the necessary facilities for it are not available. In these laboratories sodium and potassium salts of weak acids, such as acetic acid, are usually titrated with perchloric acid in glacial acetic acid.l0 During the titrations a heavy precipitate formed in the presence of potassium, whereas with sodium salts the titrated liquid remained clear. The solubility of potassium perchlorate has been shownl1,l2 to be less in glacial acetic acid than in the conventional alcoholic medium, the values being 0.027 g per litre in the former and 0.15g per litre in the latter.Preliminary experiments showed that potassium could be quantitatively precipitated by the use of a small measured excess of perchloric acid and that the filtrate from the600 JACKSON THE DETERMINATION OF POTASSIUM AND [Vol. 78 precipitated potassium perchlorate could be evaporated to dryness without danger.zinc uranyl acetate determination of sodium in the residue presented no difficulty. The METHOD REAGENTS- Perchloric acid, 0.2 N T o 1 litre of glacial acetic acid add 46 ml of 60 per cent. perchloric acid (analytical reagent grade) and then 500 ml of acetic anhydride (analytical reagent grade), stirring and cooling the solution continuously. Dilute to 2 litres with glacial acetic acid. Standardise this solution against anhydrous sodium carbonate dissolved in acetic acid, using methyl violet as indicator. Methyl violet-A 0.2 per cent. solution in ethanol. Ethanol snturated with potassium perchlorate. Zinc uranyl acetate solution-Prepare this as described in the British Pharmacopoeia, Ethanol saturated with sodium ziitc uranyl acetate. A cetone-Dry. 1953, p. 68'7.PROCEDURE- Weigh sufficient of the potassium salt to give a titration of about 40 ml of 0.2 N perchloric acid, transfer it to a dry beaker and dissolve it in 50 ml of glacial acetic acid. Add 6 drops of methyl violet solution and titrate with the standard perchloric acid until the indicator becomes emerald green. If a volumetric assay is required, note the volume added; then add a further 0.5 ml. Filter and collect the precipitated potassium perchlorate on a tared sintered-glass crucible of porosity 3, transferring and washing the precipitate with glacial acetic acid (a flexible polythene wash-bottle is useful for this). Finally wash the precipitate with a little ethanol saturated with potassium perchlorate, collecting the ethanolic washes separately and rejecting them.Dry the precipitate at 160" C and weigh it. 1 g of potassium perchlorate zz 0.28217 g of potassium. TABLE I RECOVERY OF SODIUM AND POTASSIUM Potassium carbonate found,*- taken, found, 0.5892 0.5883 99-85 0.5913 0.5890 99-76 0.5876 0.5886 100.17 0.5895 0-5910 100.25 0.5899 0.5899 100~00 0.5899 0.5907 100.13 0.5898 0.5895 100.00 0.5903 0-5909 100.10 0.5900 0.5912 100.20 g g % Sodium carbonate r A taken, found, recovered, F; g g - 0.0029 - 0.0033 - - 0.0029 - - 0.0029 - 0.0292 0.0319 0.0289 0.0292 0.0310 0.0280 0.0293 0.0321 0.0291 0.0057 0.0087 0.0057 0.0057 0.0089 0.0059 - Mean (9) . . . . 100-05 Mean (6) . . .. .. .. recovered, % - - 99.0 95-9 99-3 100.0 103.5 99.5 Standard error of mean . . &0.05 Standard error of mean . . .. . . &l-2 * The potassium perchlorate was dried at 105" C.Evaporate the filtrate to dryness on an electric hot-plate in a fume cupboard. Add about 5 ml of water and, with the aid of a rubber-tipped rod, loosen the slightly tarry residue. (The use of redistilled acetic acid and acetic anhydride slightly reduces the amount of this residue, but analytical grade reagents are satisfactory and usually free from sodium). Filter the solution through a Whatman No. 41 filter-paper into a small beaker, washing the original beaker and the filter-paper thoroughly with water. Evaporate the filtrate to about 2 ml, add 25 ml of zinc uranyl acetate reagent and set the beaker aside for 30 minutes. Filter on a tared sintered-glass crucible of porosity 3, transferring the precipitate with the aid of a small amount of reagent.Wash with three 2-ml portions of ethanol s;ttixratedOctober, 19531 TRACES OF SODIUM I N SOME POTASSIUM SALTS 601 with sodium zinc uranyl acetate and then three 2-ml portions of dry acetone. Dry the precipitate at 100" C for 30 minutes, cool and weigh it. 1 g of sodium zinc uranyl acetate = 0.01495 g of sodium. RESULTS The method was tested on potassium carbonate and sodium carbonate of analytical reagent gradc. The dried carbonates were dissolved in a weighed amount of glacial acetic acid, weighed aliquots being taken for test. Recoveries of potassium were erratic and higher than was consistent with the presence of the amounts of sodium found. The high results appeared to be due to residual acetic acid, as the precipitates had been dried at 105" C, and a higher temperature was subsequently found necessary to remove the last traces.The tests were therefore repeated with a drying temperature of 160" C ; as shown in Table I1 (first five results), recoveries were then more consistent and in agreement with the presence of traces of sodium. Further replicate tests were also made (Table 11, second five results), and the figures found for sodium agree with the value of 0.2 per cent. (0.46 per cent. as sodium carbonate) found by flame photometry. The recoveries are shown in Table I. TABLE I1 RECOVERY OF SODIUM AND POTASSIUM Potassium perchlorate dried at 160" C Potassium Carbonate Sodium carbonate A r A > I \ taken, found, found, taken, found, found, g g % g g % 0.6877 0.6853 99-65 0.6879 0.6859 99.71 0,6876 0.6856 99.69 0.6880 0.6856 99.65 0.6874 0.6859 99.78 Test for sodium omitted 0.5907 0-5872 99.41 nil 0.0028 0.47 0-5900 0.5868 99-46 nil 0.0028 0-47 0.6923 0.5896 99-54 nil 0.0026 0-44 0.5930 0.5895 99.41 nil 0*0030 0.51 0.5926 0.5891 99-41 nil 0.0028 0.47 Mean (10) .. 99.57 Mean (5) .. 0-47 - Standard error of mean . . &O*Od Standard error of mean . . &0.011 SCOPE OF THE METIIOD If it is required to determine sodium alone in a neutral potassium salt, this can be done by adding the calculatcd amount of perchloric acid to precipitate, say, 98 per cent. of the anticipated amount of potassium. Most of the potassium can then be removed by filtration, and the sodium can be determined in the filtrate as described above. If the potassium salt is difficult to dissolvc in glacial acetic acid, the sample can be dissolved in a small amount of water before the acetic acid is added.After the perchloric acid solution has been added, sufficient acetic anhydride is added to convert the water into acetic acid. Heat is evolved in tile course of this addition and the solution should be cooled before filtering. With this technique, tests of which the results arc shown in Table TI1 were carried out on potassium nitrate (analytical reagent grade) with additions of sodium nitrate and on potassium chloride (analytical reagent grade) with additions of sodium chloride. The results show that the method could be extended to the determination of sodium in neutral potassium salts, but the greatest advantage appears in determinations of sodium in the potassium salts of weak acids.I t is not considered that the distribution of results as between sodium nitrate and sodium chloride is indicative of a difference in behaviour between the two salts. The chances against the two samples of nitrate and the two samples of chloride giving results significantly different from 100.0 per cent. and in opposite directions602 JACKSON [Vol. 78 are only 8 to 1, and the discrepancies in any event depend upon differences of not more than 0.5 to 1-0 mg. TABLE I11 RECOVERIES OF SODIUM FROM NEUTRAL POTASSIUM SALTS Potassium salt taken, g 0.7980 0.8012 0.801 2 0.7986 Potassium nitvate- Recovery of Sodium salt sodium salt P /-----h- added, found, g g g % nil 0*0004 Blank experiment nil 0.0004 Blank experiment 0-0341 0-0341 0.0337 98.8 0.0372 0.0369 0.0365 98.1 Potassium chloyide- 0.5970 nil 0.0007 Blank experiment 0.5990 nil 0.0007 Blank experiment 0.5988 0.0322 0.0340 0.0333 103-4 0.5978 0.0274 0.0287 0.0280 102.2 Mean (4) . .100.6 The novel points of this technique are the use of the acetic acid medium for precipitating potassium perchlorate and of a standard solution of perchloric acid. It is to be expected that the potassium and sodium estimations will still be subject to the usual interferences; for example, ammonium salts will form an insoluble perchlorate and lithium will give an insoluble lithium zinc uranyl acetate; phosphate or other anions that react with zinc or uranyl salts must be absent. Therefore a volumetric assay and a gravimetric potassium determination could probably be made on potassium phosphates, of which all the potassium is titratable in the medium specified.No sodium determination would be possible here. The full procedure can be used on potassium carbonate, bicarbonate, hydroxide, acetate and, it would seem probable, other salts of organic acids. Although we have not tried a wide range of compounds, the technique appears suitable for sodium or potassium estimations on a variety of potassium salts. I am indebted to British Drug Houses Ltd. for determining by flame photometry the sodium in the sample of Potassium carbonate used for the results shown in Table 11. 1. 3. 4. 9 Y. 5. 6. 7. 8. 9. 10. 11. 12. REFEREXCES Kolthoff, I. M., and Barber, H. H., J . Amer. Chew. SOL, 1928, 50, 1625. Haslam, J., and Eeeley, J . , Analyst, 1941, 66, 185. 12elcher, K., and Nutten, A. J., Anal. Chim. A d a , 1950, 4, 595. Cumming, A. C., and Kay, S. A., “Quantitative Chemical Analysis,” Ninth Edition, Oliver and Vogel, A. 1 ,, “A Textbook of Quantitative Inorganic Analysis,” First Edition, Longmans, Green Parks, T. D., Johnson, H. O., and Lykken, L., Anal. Chem., 1948, 20, 822. West, P. W., Folse, I>., and Montgomery, D., Ibid., 1950, 22, 667. Osborn, G. H., and Johns, H., Analyst, 1951, 76, 410. Leyton, L., Ibid., 1951, 76, 723. Seaman, W., and Allen, E., Anal. Chem., 1951, 23, 592. Seidell, A, “Solubilities of Inorganic and Metal Organic Compounds,” Third Edition, D. Van -, op. cit., p. 79.5. Boyd Ltd., Edinburgh, 1945, p. 367. and Co. Ltd., London, 1948, p. 568. Nostrand Co. Jnc., Sew York, 1940, p. 793. GLAXO LABORATORIES LIMITED ULVERSTON, LANCASHIRE January 29th, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800599

出版商:RSC    

年代:1953

数据来源: RSC