摘要:

FEBRUARY, 1963 THE ANALYST Vol. 88, No. 1043 EDITORIAL WHEN, in 1924, Council appointed the Standing Committee on Uniformity of Analytical Methods to co-ordinate investigations into methods of analysis, it took-without knowing it -the first step towards the production of a book that the Society is publishing in a few days’ time.* For the Standing Committee was the direct forerunner of the Analytical Methods Committee, and from the careful collaborative investigations of their various Sub-Committees has come the collection of tried and tested methods that form the first part of “Official, Standardised and Recommended Methods of Analysis.” The first of the Sub-committee’s reports appeared in The Analyst in 1927, and it was soon clear that existing methods had to be re-defined and revised, and that often new ones had to be produced, if they were to give results of acceptable precision in the hands of several analysts working in different laboratories.The high regard accorded to the methods resulting from these investigations is reflected in the large proportion that have subsequently been adopted unchanged, or with only minor amendments] as standard methods by such official bodies as the British Standards Institution, the British Pharmacopoeia Commission and the Pharmaceutical Society of Great Britain. Even so, the Society is continually extending the scope of these investigations, both through its own Analytical Methods Committee and in active co-operation with government departments, trade associations and other learned societies.These joint Committees also have recommended methods of analysis which have been published from time to time in The Analyst. One large group of methods, on Trade Effluents, was subsequently re-issued as a book. Both the great popularity of the Trade Effluents book, and the increasing difficulty of referring to methods scattered over 35 volumes of The Analyst, convinced the Committee that a collected edition of the methods would be of considerable value to analysts. This collection forms the first part, extending over 340 pages, of the new publication. The Society acknow- ledges its debt to the Association of British Chemical Manufacturers, the Association of British Manufacturers of Agricultural Chemicals, the Inter-Departmental Advisory Committee on Poisonous Substances Used in Agriculture and Food Storage and the Pharmaceutical Society of Great Britain for agreeing to the inclusion of methods developed by various joint commit tees.But these methods cover no more than a part of the field of analysis. The exhaustive testing of procedures takes time, even when that time is so freely given as it has been by the many members of the score and more Sub-Committees and Panels engaged on the investi- gations. So, shortly after the second World War, a Standard Methods Committee was formed to collect, through a group of specialists, recommendations of reliable published methods for a variety of tests covering a wide range of materials. These compilations, in the form of classified lists of references, were published in 1951 as the Society’s “Bibliography of Standard, Tentative and Recommended or Recognised Methods of Analysis,” a book that for some time now has been unobtainable.The second part of the new publication comprises a completely revised * “Official, Standardised and Recommended Methods of Analysis,” Compiled and Edited for the Analytical Methods Committee of the Society for Analytical Chemistry by S. C. Jolly, B.Pharm., B.Sc., A.R.I.C., M.P.S. Cambridge: W. Heffer & Sons Ltd. for the Society for Analytical Chemistry. Price &6 6s. 7374 PROCEEDINGS [Analyst, Vol. 88 and considerably augmented new edition of this Bibliography. Besides the British and American authorities quoted in the first edition, it includes references to some French and German sources, and to some in other languages for which translations may be readily available. In this one publication an analyst can firid details of procedures on which he can rely, even though he may never have met the particular analytical problems before. The biblio- graphical section gives him references to equally reliable methods that are “official” or have been “standardised,” and also gives him a grea.t deal of the guidance in the sphere of “recom- mended” methods that he could expect from an experienced senior colleague. He will undoubtedly come to share the Society’s gratitude to the members of the investigating Sub-Committees, the compilers of the Bibliographical sections and the book’s editor for their considerable efforts in making this book a reality.

ISSN:0003-2654    

DOI:10.1039/AN9638800073

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

74 PROCEEDINGS [Analyst, Vol. 88 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETING AN Ordinary Meeting of the Society was held at 7 p.m. on Wednesday, February 6th, 1963, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. A. J. Amos, O.B.E., B.Sc., F.R.I.C. The subject of the meeting was “Particle Size Analysis” and the following papers were presented and discussed: “The Size Analysis of Insoluble Drugs,’’ by M. J. Thornton, B.Sc., A.R.I.C.; “Particle Size Analysis in the Formulation of Pesticides,” by C. G. L. Furmidge, BSc., Ph.D., A.R.I.C. ; “Particle Sizing in Aerosol Systems,” by D. A. Blyth, B.Sc., J. M. Creasey and N. W. Wootten, B.Sc., A.1nst.P. MIDLANDS SECTION AND MICROCHEMISTRY GROUP A JOINT Meeting of the Midlands Section and the Microchemistry Group was held at 6.30 p.m.on Friday, December 14th, 1962, in the Chemistry Theatre, The University, Edgbaston, Birmingham, 15. The Chair was taken by the ,Vice-chairman of the Midlands Section, Mr. W. H. Stephenson, F.P.S., D.B.A., F.R.I.C. The following paper was presented and discussed : “The Determination of Carbon and Hydrogen in Organic Materials.” by Miss A. M. G. Macdonald, M.Sc., Ph.D., A.R.I.C. The meeting was preceded by a tour of the new laboratories in the Haworth block of the University. MIDLANDS SECTION A JOINT Meeting of the Section with the Birmingham and Midlands Section of the Royal Institute of Chemistry was held at 7 p.m. on Thursday, January 24th, 1963, at the Technical College, Stoke-on-Trent.The Chair was taken by the Vice-chairman of the Midlands Section, Mr. W. H. Stephenson, F.P.S., D.B.A., F.R..I.C. The following paper was presented and discussed : “Recent Advances in Polarography,” by G. F. Reynolds, M.Sc., Ph.D., M.R.S.H., F.R.I.C. MICROCHEMISTRY GROUP THE thirty-seventh London Discussion Meeting of the Group was held at 6.30 p.m. on Wednesday, December 19th, 1962, at “The Feathers,’’ Tudor Street, London, E.C.4. The Chair was taken by the Chairman of the Group, Mr. C. Whalley, B.Sc., F.R.I.C. The meeting took the form of a RevieTw Meeting, at which any subject related to those already covered in the Discussion Meetings was raised. THE thirty-eighth London Discussion Meeting of the Group was held at 6.30 p.m. on Wednesday, January 23rd, at “The Feathers,” Tudor Street, London, E.C.4.The Chair was taken by the Chairman of the Group, Mr. C:. Whalley, BSc., F.R.I.C. A Discussion on “The Microdetermination of Mercury’’ was opened by R. F. Milton, B.Sc., Ph.D., F.R.I.C., M.I.Bio1.February, 19631 PROCEEDINGS 75 BIOLOGICAL METHODS GROUP THE eighteenth Annual General Meeting of the Group was held at 6.30 p.m. on Thursday, December 13th, 1962, at “The Feathers,” Tudor Street, London, E.C.4. The Chair was taken by the Chairman of the Group, Mr. J. S. Simpson, F.I.M.L.T. .The following Officers and Committee Members were elected for the forthcoming year :-Chairman-Mr. W. A. Broom. Hon. Secretary and Treasurer-Mr. K. L. Smith, Standards Department, Boots Pure Drug Co. Ltd., Nottingham.Members of Com- mittee-Dr. Jillian M. Bond, Mr. D. C. M. Adamson, Mr. L. C. Dutton, Dr. F. W. Norris, Mr. J. S. Simpson and Mr. G. Sykes. Mr. D. M. Freeland and Dr. J. H. Hamence were re-appointed Hon. Auditors. Immediately following the Annual General Meeting a Discussion Meeting on “The Assessment of Vitamins in Feeding Stuffs’’ was opened by A. J. Amos, O.B.E., B.Sc., Ph.D., F.R.I.C. Vice-Chairman-Dr. M. W. Parkes. Hon. Recorder-Miss A. M. Jones. ATOMIC ABSORPTION SPECTROSCOPY DISCUSSION PANEL THE Inaugural Meeting of the Atomic Absorption Spectroscopy Discussion Panel of the Physical Methods Group was held at 6.30 p.m. on Wednesday, December 12th, 1962, in the Meeting Room of the Chemical Society, Burlington House, Piccadilly, London, W. 1. The Chair was taken by the Chairman of the Panel, Mr.W. T. Elwell, F.R.I.C. The following paper was presented and discussed : “Aspects of Atomic Absorption Analysis,” by D. J. David, M.Sc. (see summary below). ASPECTS OF ATOMIC ABSORPTION ANALYSIS MR. D. J. DAVID said that, since the introduction of atomic absorption analysis and the apparatus essential for it by Walsh in 1953, research in Australia and New Zealand had been conducted along three main lines. The study of vaporisation of samples in flames, entailing examination of inter- element effects arising from variations in general composition of samples, interaction between flame gases and sample and the effect of organic solvents. (b) The study of alternative means of vaporisation of samples, notably use of the sputtering chamber for metallic samples. (c) The search for means of increasing sensitivity by choice of absorbing line and modification of apparatus. Recent research in the United States had dealt mainly with the study of increase in length of light path through the vaporised sample as a means of gaining sensitivity and the use of a continuum as source of radiation in place of line sources. The speaker discussed these points and emphasised those upon which further advances might depend. In the discussion that followed, many members of the audience expressed opinions on these These were- (a) points and on other matters of technique that held a general interest.

ISSN:0003-2654    

DOI:10.1039/AN9638800074

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

76 HALL : SPECTROPHOTOMETRIC DETERMINATION OF SUB-MICROGRAM [Analyst, Vol. 88 The Spectrophotometric Determination of Sub-microgram Amounts of Fluorine in Biological Specimens BY R. J. HALL* (Agvicultuval Reseavch Council Institute of Animal Physiology, (Pharmacology Unit), Babraham, Cambridge) A method is described for the accurate determination of sub-microgram amounts of fluorine involving the new direct colour reaction with alizarin complexan. The fluoro-chelate formed with lanthanum alizarin complexanate is extracted into isobutanol containing hydroxylamine hydrochloride. Fluorine is collected from ashed specimens on to magnesium succinate treated filter-paper by diffusion, in sm.all polythene bottles, induced by 50 per cent. perchloric acid containing silver sulphate. Recoveries for 0.1 to 1.0 pg of fluorine are over 95 per cent.Problems associated with determining these small amounts are discussed. The method lends itself well to determining fluorine in water and in biological materials such as blood and urine. THE detection of fluoride with alizarin complexan (3- [ di-(carboxymethyl)aminomethyl]- 1,2-dihydroxyanthraquinone) was first reported by Belcher, Leonard and West.l*2 This new reagent differs from others in that its use appears to give the first direct colour reaction for the fluoride ion. Leonard and West3 later noted that the blue cerium fluoro-chelate of alizarin complexan was soluble in pentanol containing n-tributylamine. Their observation prompted a further study of the conditions for forming the cerium and lanthanum fluoro- complexes, and especially their solubilities in organic solvents containing amines.The use of hydroxylamine hydrochloride in isobutanol as a solvent for the lanthanum fluoro-chelate has made possible the development of a method for determining fluoride in the sub-microgram range. By modifying a diffusion technique for the collection of hydrofluoric acid described in an earlier paper,4 reproducible and quantitative recoveries of 0.1 to 1.0 pg of fluorine have been obtained; the method described below is proposed for the determination of such levels in biological specimens. METHOD REAGENTS- Reagents should be of analytical grade when possible. Magnesium succinate, 0.2 M-Dissolve 5.13 g of fluoride-free magnesium succinate, Mg(C,H,O,),, in water, and dilute to 100 ml; store in a polythene bottle at 2" C.Silver sulphate in perchloric acid-Approximately 0.25 M silver sulphate in 70 per cent. w/v perchloric acid. Mix 4 g of finely powdered silver sulphate with 2 ml of water and 23 ml of 72 per cent. perchloric acid in a 250-ml Erlenmeyer flask, and heat to about 80" C to dissolve the silver sulphate. Add a further 75 ml of 72 per cent. perchloric acid, and cool the solution. Succinate bufer solution, pH 4.6 (0.04 M)-Dissolve 0.4724 g of succinic acid in 96 ml of water (with warming to aid solution). Cool the solution, add 3.2 ml of N sodium hydroxide, and make up to 100ml with water. Best results are obtained when the solution is fresh, but, if necessary, it can be stored at 2" C in a polythene bottle and prepared weekly.Alizarin complexan solution, 0.0005 M--Suspend 38.5 mg of reprecipitated alizarin com- plexan in 2 ml of water in a test-tube; add 0.2 ml of M sodium acetate, and warm the mixture to dissolve the complexan. Transfer the solution to a 200-ml calibrated flask, and make up to the mark with water. Store at 2" C in a dark bottle. The solution is stable for a considerable length of time. To reprecipitate the alizarin complexan (obtainable from Hopkin and Williams Ltd.) dissolve 1 g in about 10 ml of 0.5 M sodiurn acetate, and heat to about 80" C; if necessary, add more sodium acetate solution until the solid has dissolved. Transfer the solution to a The pH should be checked potentiometrically. * Present address : Analytical Laboratories, Kational Agricultural Advisory Service, Ministry of Agriculture, Fisheries and Food, Brooklands Avenue, Cambridge.February, 19631 AMOUNTS OF FLUORINE IN BIOLOGICAL SPECIMENS 77 stoppered 100-ml measuring cylinder, and dilute to 100 ml with water.Carefully add hydro- chloric acid dropwise until a heavy orange precipitate forms. The precipitation is continued for some hours at 0" to 2" C. Remove the supernatant liquid, separate the precipitate from residual liquid by spinning in a centrifuge, wash twice with 20ml of ice-cold water, and dry at 37" C. Lanthanum nitrate, 0.001 M-Dissolve 43.3 mg of lanthanum nitrate, La( N0,),.6H20, in water to make 100ml of solution. Bufered lanthanum alizarin complexanate--By pipette, place 30 ml of the 0.0005 M alizarin complexan, with constant mixing, into 50ml of the succinate buffer solution arid 20 ml of lanthanum nitrate solution.Prepare freshly each day. Hydroxylamine hydrochloride solution-Prepare a molar solution by dissolving 6-95 g of hydroxylamine hydrochloride (recrystallised from 75 per cent. v/v ethanol at 2" C, and dried at 60" C) in water to make 100 ml of solution. Extracting solvent-Shake 3 ml of the hydroxylamine hydrochloride solution with 97 ml of isobutanol until the two liquids have become homogeneous. The solvent will keep for several days, but a slight precipitate tends to form after a time. Stock Jluoride solution, 100 pg of F- per ml-Dissolve 221.05 mg of pure sodium fluoride in water, and dilute to 1OOOml. Standard JEuoride solutions, 10 pg of F- and 2 pg of F- per ml-Dilute a 10-ml and a 2-ml portion of the stock fluoride solution to 100ml.These solutions should be freshly prepared each week. All fluoride solutions should be stored in screw-capped polythene bottles. Magnesium succinate solution, Jluoride- free-First prepare a 5 per cent. w/v solution of succinic acid in water distilled in an all-glass apparatus. Wash 3 g of magnesium turnings with 100 ml of distilled water in a 500-ml Erlenmeyer flask. Discard the washings, and replace by 100 ml of fresh distilled water to which 50 ml of the 5 per cent. succinic acid have been added. After about one minute, carefully pour off the acid, and wash the magnesium twice with 100 ml of water distilled in an all-glass apparatus. Suspend the magnesium in 100 nil of water distilled in an all-glass apparatus and 100ml of 5 per cent. succinic acid.When the main reaction has subsided, evaporate the contents of the flask to about 80 ml and then filter through a washed Whatman No. 42 filter-paper or spin in a centrifuge (radius about 4 inches) at 4000 r.p.m. in cellulose nitrate or polythene tubes. Evaporate to dryness at 100" C overnight, and then grind the white residue to a powder with a glass pestle and mortar. Prepare a 0.2 M solution of the magnesium succinate. Lithium hydroxide solation, JEuoride-free-Dry with filter-paper lumps of lithium metal that have been stored in paraffin spirit. Remove the outer crust with a sharp knife or tin shears, and quickly weigh about 4 g of the clean metal. Cut into small pieces the size of a pea, and drop into 500 ml of water distilled in all-glass apparatus contained in a 1000-ml Erlenmeyer flask.When all the lithium has dissolved, check the normality of the solution by titration against potassium hydrogen phthalate, and then adjust the concentration Then add 25 ml of t-butanol to the mixture. to 1 N. PREPARATION OF SAMPLE- Samples of 1 ml of blood, cerebrospinal fluid, urine, etc., are measured into small platinum crucibles, mixed with 0.3 ml of N lithium hydroxide and 0.2 ml of 0.2 M magnesium succinate and dried at 110" C. Bone, vegetable and soft animal tissues are dried at 110" C for several hours, finely ground, and 0.2 g is mixed with alkali and magnesium succinate as above. The crucibles are covered with lids and placed in a small metal canister having a tightly fitting lid; the canister is then placed in a cold muffle furnace, and the temperature is slowly raised to 400" C and maintained for about 15 hours, usually overnight.The canister is removed from the furnace while still hot. The bone ash is carefully broken up with a small nickel spatula and transferred to a 100-ml calibrated flask; the remaining traces of ash are treated with small portions of 0-5 N perchloric acid to a total of 5 ml, and the dish is washed with about 10 ml of water, all the washings being added to the ash in the flask. One millilitre of 72 per cent. v/v perchloric acid is added, and the flask is gently warmed to about 60" C, when the ash will be completely dissolved apart from a few particles of carbon; the volume in the flask is then adjusted to the mark with water.Samples of 1 ml are taken for the collec- tion of hydrofluoric acid in the diffusion bottles. The ash of blood, cerebrospinal fluid, urine78 HALL : SPECTROPHOTOMETRIC DETERMINATION OF SUB-MICROGRAM [Analyst, Vol. 88 or tissue is similarly broken up and carefully transferred to a polythene diffusion bottle, the platinum dish is washed with 0.4 ml of 0.5 N perchloric acid and then twice with 0.3 ml, and the washings are also placed in the diffusion bottle. Appropriate controls consisting only of the reagents are similarly prepared, as well as. standards containing 0.1, 0-2, 0.5, 0.7 and 1.0 pg of fluoride in 1-ml volumes; when possible all diffusions should be carried out at least in duplicate. DIFFUSION OF HYDROFLUORIC ACID- The procedure already reported4 for collecting fluorine by diffusion as hydrofluoric acid was modified; the changes made were as described below.Rectangles of Whatman No. 541 filter-paper, 3.0 cm x 1.2 cm, previously extracted with hot distilled water, rinsed and dried, are used. These are pushed into the lumen of an adaptor of 5-mm internal diameter polythene tubing inside the well of the bottle cap, thus dispensing with the short length of glass rod originally used. The papers remain reasonably tightly held. The free 2 cm of the paper is impregnated with 15 pl of 0.2 M magnesium succinate placed 0.5 cm from the free edge. A graduated 0.1-ml pipette is useful for this purpose. The sample in a volume of 1 ml is placed in the bottle from a pipette; 2 ml of silver sulphate solution in 70 per cent.w/v perchloric acid is then immediately added from a burette, care being taken not to allow any acid to remain on the inside of the bottle neck. The cap with its filter-paper is screwed tightly into position. The screw threads of the polythene bottle caps and necks tend to wear and cause loss of hydrofluoric acid. This source of error has been overcome by sealing the junction between the cap and the bottle neck with hot wax. A suitable wax is a mixture of equal weights of ceresine and carnauba wax (melting point about 85" C), which is easily applied with a dropping pipette; 1 cm of the end of its stem is bent at an angle of 135" and fitted with a 2-ml rubber bulb. The bottle is tilted on its side at an angle of about 30°, and the wax, which should be very hot (so that it is vaporising), is allowed to run round the junction between the cap and bottle neck.Diffusion is continued for 24 hours at 60" C. DETERMINATION OF FLUORIDE- The papers are removed from the diffusion bottles with forceps and placed in stoppered 5-ml tubes. Lanthanum alizarin complexanate reagent (2-0 ml) is placed by pipette in the tubes, which are then put into a water bath at 60" C for 10 minutes. The colour of the alizarin reagent changes from magenta to blue, according to the amount of fluoride present. The tubes are cooled to room temperature, and the blue complex is extracted by shaking vigorously for 15 seconds each time with a 1.5-ml and then two successive 1-ml portions of the isobutanol solvent. The tubes are spun in a centrifuge (radius about 4 inches) for a few moments at about 2000 r.p.m.after each extraction to separate the solvent layer, and this is removed to another stoppered 5-ml graduated tube; a 2-ml graduated pipette having a finely pointed tip and fitted with a 2-ml rubber bulb is used for the transfer. To obviate the need for a separate pipette for each extract and to prevent contamination, the pipette is filled after each use to the 1-ml mark with the extracting solvent, and the washings are transferred to the extraction tube. After the volumes have been adjusted to 4.3 ml with solvent, the extracts are shaken for 30 seconds with 1 ml of water and then cooled to 0" C or lower. The aqueous phase is separated by spinning in a centrifuge for a short time, and, when necessary, the volume of the extract is adjusted to 4 ml.The extracts are then transferred to glass cells, and their optical densities are measured at 570 mp or photometrically with a red filter. Note that the diffusion bottles should be emptied under running cold water, washed out several times with distilled water only, and dried at 60" C. All tubes should be cleaned by soaking for a few minutes in dilute hydrochloric acid and then washing several times with distilled water. Platinum crucibles should be soaked in 10 per cent. perchloric acid, washed with distilled water, and heated in a bunsen flame until red hot. This procedure ensures that any traces of fluoride are volatilised. In this way contamination is negligible. EXPERIMENTAL RESULTS All experimental observations were carried out in duplicate or triplicate, and a Unicam SP500 spectrophotometer was used for measuring optical densities.February, 19631 AMOUNTS OF FLUORINE IN BIOLOGICAL SPECIMENS 79 SOLVENT FOR FLUORIDE CHELATE- During the preliminary experiments it became evident that, although the blue complex obtained by the reaction of fluoride with the cerium or lanthanum chelate of alizarin complexan was easily extracted into n-pentanol containing tributylamine, the unreacted complexanate was also fairly soluble and there was a direct correlation between ease of extraction and concentration of the amine.Lanthanum alizarin complexanate alone when extracted into 0.08 per cent. v/v tributylamine in n-pentanol gave 30 per cent. of the optical density of the blue complex produced by 1 pg of fluoride in a standard volume of 4 ml.Many amines dissolved in alcohols will extract the fluoride complex with different efficiencies ; these include met hylamine hydrochloride and ethylamines, as well as triben~ylamine.~ Hydroxyl- amine hydrochloride in isobutanol was by far the most satisfactory solvent because of the high solubility of the fluoro-chelate and low solubility of the lanthanum complexanate. Blank values were reduced to an average of 4 per cent. of the optical density of the complex obtained with 1 pg of fluoride in a volume of 4 ml. EFFECT OF BUFFER AND pH ON FORMATION OF FLUORIDE COMPLEX WITH LANTHANUM ALIZARIN All work so far reported on the reaction of fluoride with the cerium or lanthanum chelates of alizarin complexan has involved the use of acetate buffer solutions ranging from pH 4.3 to 5-1.In the work described here it was confirmed that pH 5-0 was optimal for an acetate- buffered reagent,6 but it was also noticed that other buffer systems of the same pH range profoundly affected the reaction. Thus boric acid buffered reagent (final concentration 0.1 M) produced a more intense blue fluoro-chelate than did acetate, but a succinate-buffered reagent, pH 4.6 (final concentration 0.02 M), also markedly increased the speed of reaction to give greater sensitivity and lower blank values. It is suggested that the solubility of the buffer systems in the organic solvent may be a factor governing the efficiency of extraction. Fluoro- chelates produced with borate- or succinate-buffered reagents were much more easily extracted than were those produced with acetate-buffered solutions ; acetate buffers are appreciably soluble in the alcohol solvents and presumably inhibit the effect of the amine.Fig. 1 and Table I show a striking comparison of optical densities of the extracts of fluoro-chelate pro- duced in acetate- and succinate-buffered alizarin complexan reagents. The peak at pH 4.6 in the succinate range appears to be characteristic, irrespective of the method of extraction. CHELATE- TABLE I OPTICAL DENSITIES AT 570 mp PRODUCED BY 1 pg OF FLUORIDE Reaction carried out with lanthanum alizarin complexanate reagents prepared with acetate and succinate buffers, and extracted into 4 ml of solvent. Optical densities corrected for blank value Optical density Acetate Succinate pH (final concentration 0.1 M)* (final concentration 0.02 M ) t 4.0 4.2 4.4 4.6 4.7 4.8 5.0 5.2 0.090 0.099 0.1 19 0.119 0.120 0.122 0.123 0.110 0.159 0.212 0.234 0.264 0.257 0.249 0.253 0.259 * Extracted with 0-08 per cent.v/v of tributylamine in n-pentanol. t Extracted with 0.03 M hydroxylamine hydrochloride in iso-butanol. EFFECT OF TEMPERATURE AND TIME ON REACTION BETWEEN FLUORIDE AND LANTHANUM Results of the influence of temperature and period of reaction on the formation of the Fluoride (1 pg) in 2 ml of the alizarin complexan ALIZARIN COMPLEXANATE- fluoro-chelate are shown in Table 11. reagent was taken. Prolonged immersion at 100" C caused fading of the blue fluoro-chelate. It was decided to adopt 10 minutes a t 60" C as the standard procedure.80 HALL SPECTROPHOTOMETRIC DETERMINATION OF SUB-MICROGRAM [A?ZUlySt, VOl.88 TABLE I1 EFFECT OF TEMPERATURE AND TIME ON REACTION BETWEEN FLUOKIDE AND LANTHANUM ALIZAKIN COMPLEXANATE qb----+-- B --*- -- --c- ~~ Temperature, "C 20 20 20 20 20 60 100 Time, minutes 5 10 30 45 60 10 5 Optical density (4-ml extract) 0.245 0-244 0.245 0-246 0.251 0-266 0.248 COMPARISON OF OPTICAL DENSITY OF DIRECT REACTION MIXTURES AND EXTRACTS- Table I11 compares the optical densities of the lanthanum fluoro-chelate in succinate buffer solution at pH 4.6 read directly and in isobutanol extracts; the superiority of the extraction procedure is clearly indicated. 0*301 0 4.0 4.4 4.8 5.2 O-IDI- Wavelength, mp Fig. 1. Effect of pH on optical density Fig.2. Absorption curve of the fluoro- chelate of lanthanum alizarin complexanate prepared with succinate buffer at pH 4.6: curve PL, chelate; curve B, blank test of extracts of the fluoro-chelate of lanthanum alizarin complexanate prepared with acetate and succinate buffers: curve A, chelate with succinate buffer; curve B, chelate with acetate buffer; curve C, blank test with ace- tate buffer; curve D, blank test with succinate buffer MEASUREMENT OF COLOUR- Fig. 2 shows the absorption curve of the fluoro-chelate in isobutanol - hydroxylamine hydrochloride obtained with the lanthanum-based reagent ; that of the cerium reagent is similar. The general shape of the curve is similar to that obtained with the chelate in aqueous solution, the only difference being a slight shift of the curve to the left; the peak is at 615 mp in aqueous medium.I t can be seen that the optimum wavelength is 570 mp. RECOVERY TESTS- Representative recoveries of fluorine from the diffusion bottles are shown in Table IV and indicate that virtually all the fluorine within the range is collected on the paper wicks under the conditions of the test.February, 19631 AMOUNTS OF FLUORINE IN BIOLOGICAL SPECIMENS 81 TABLE I11 COMPARISON OF OPTICAL DENSITIES OF REACTION MIXTURES AND THEIR The volumes used were 4 ml throughout, and blank values were not subtracted ISOBUTANOL EXTRACTS Optical density Fluorine present, Pg 0 0.1 0.2 0.4 0.6 0.8 1.0 Direct reading on reaction mixture 0.112 0.114 0.128 0.155 0.180 0.205 0.248 at 615 mp 1 Reading on extract at 570 mp 0.008 0.034 0.060 0.111 0.163 0-214 0.265 The method for determining the recovery was to measure the colour obtained from the fluorine collected on the paper against that obtained with the same amount of fluorine added directly by pipette to 2 ml of the lanthanum reagent.A specimen of the filter-paper with 15 pl of the magnesium succinate solution was also placed in the reagent to provide a blank solution. Recovery of fluorine added to blood and urine carried through the entire procedure has been confined to the 0.2- and 0.5-pg levels; recoveries were from 96 to 114 and from 95 to 102 per cent., respectively. The third decimal place shown in the figures is only approximate. TABLE IV RECOVERY OF FLUORINE FROM DIFFUSION BOTTLE Amount of fluoride added, Amount of fluoride found, Pg 0.112 0.089 0.1 0.092 0.096 0.09 1 0.200 0.195 0.193 0.200 0.188 0.487 0.496 0.479 0.486 0.500 1.000 0.965 1.010 0.992 i 1.008 r r 1 0.2 i 0.5 1.0 I Recovery, 112 89 92 96 91 100 97 96 100 94 97 99 96 97 100 100 96.5 101 99 101 % INTERFERENCES- Belcher and West6 reported on substances interfering with the formation of the cerium fluoro-chelate of alizarin complexan. They found that several anions and cations profoundly influence the reaction. Some of these observations have been confirmed in this laboratory and equally apply to the lanthanum chelate.The formation of blue complexes in the succinate-buff ered lanthanum alizarin complexanate reagent with small amounts of cobalt and nickel has been observed in the absence of fluoride.These complexes are to some extent also extractable into the isobutanol solvent. Surprisingly, however, the presence of 3 pg of nickel depresses the optical density of the lanthanum fluoro-chelate from 1 pg of fluoride in the organic solvent by 20 per cent., presumably by inhibiting the extraction. When the cobalt and nickel levels are increased to 30 and lOpg, respectively, the production of the82 HALL SPECTROPHOTOMETRIC DETERMIN.4TION OF SUB-MICROGRAM [Analyst, VOl. 88 fluoro-chelate is almost completely inhibited. Interference of this nature would not be expected in the collection of hydrofluoric acid by diffusion. Although these particular observations were not pursued, further work niay provide more information on the specificity of the reaction for fluoride.DISCUSSION OF THE METHOD The accurate determination of small amounts of fluorine has been a major analytical problem for many years, and the introduction of a direct colour reaction is an important advance in this field. Although the colour reaction of the fluoride ion with the lanthanum and cerium complexes of alizarin complexan appears specific, serious interference may be encountered from several metal ions and radicals,6 and, although most may be masked and removal by extraction is necessary in only a few instances (chiefly coloured ions), such an additional operation will complicate the analysis. All interference is eliminated by the collection of the fluorine by diffusion on to magnesium succinate treated filter-paper. Mag- nesium succinate affects the colour reaction hardly at all, and recoveries from the diffusion bottles of 0.1 to 1.0 pg of fluorine are quantitative for all practical purposes.The inclusion of a silver salt with the concentrated perchloric acid immobilises any chloride ions; a step also adopted by Stegemann and Jung,' who collected the hydrofluoric acid on alkali-soaked poly(viny1 chloride) paper. A more recent application of diffusion methods to the deter- mination of fluoride has been reported by Frere,8 who used plastic Conway units for the collection of hydrofluoric acid and adopted the reaction with lanthanum alizarin complexanate for its subsequent determination. Further studies of the formation of the fluoro-chelates of alizarin complexanates confirmed the findings of Greenhalgh and Rileys that the lanthanum complexanate gave a more intense colour with fluoride ion than did the cerium complexanate. The main advantage of the lanthanum complexanate, however, is that its fluoro-complex is more soluble in the organic phase than is that of cerium.Greenhalgh and Rileyg also reported increased sensitivity by including 12 per cent. of acetone in their lanthanum reagent, and Belcher and West6 extended these findings to include methanol, ethanol and dioxan. In this investigation it was found that a much wider range of solvents apparently enhanced the visual sensitivity of the reaction when present in the mixture at concentrations from 10 to 20 per cent. by volume. However, after these reaction products had been extracted with the isobutanol - hydroxyl- amine hydrochloride mixture the optical densities of the blue fluoro-chelate extracts with one exception were always below that of an extract from an entirely aqueous reagent.The exception was t-butanol, which did not affect the extract, and was included to prevent the formation of precipitates at the junction of the organic and aqueous phases during the ex- traction. In the earlier part of the work traces of the disodium salt of ethylenediaminetetra- acetic acid were employed to prevent these precipitates, but later t-butanol was found to be more effective; it also tended to stabilise the reagent. Increased sensitivity was claimed by Belcher and West1* to be obtainable by reading a t 281 mp the direct reaction in aqueous solution between fluoride and the lanthanum chelate.Although the optical densities obtained with this procedure are much higher, so also is that of the unreacted lanthanum reagent. Increases in the optical density of the lanthanum fluoro-chelate when extracted into the organic solvent were similarly measured at 281 mp, but even the almost colourless blank extract had so high a reading at this wavelength that the procedure was unsuitable for measuring the smallest amounts of fluoride, which requires the blank value to be as near zero as possible. The extraction of the fluoro-complex #of the lanthanum complexanate into isobutanol- hydroxylamine hydrochloride mixture proved to be the best means of measuring the complex formed from less than 0.1 pg of fluoride. Many combinations of organic solvents and amines were tried.Although tributylamine in n-pentanol, mentioned by Leonard and West ,3 easilj. extracts the fluoro-chelate, it also extracts much of the unreacted lanthanum complexanate. The same is true of tribenzylamine in pentanol - s-butanol (30 + 70) suggested by Johnson and L e ~ n a r d . ~ For determining the smallest amounts of fluoride, even with the isobutanol- hydroxylamine hydrochloride solvent, the extract of the reaction mixture still contains sufficient lanthanum complexanate to give a relatively high optical density for the blank solution. Fortunately, almost all of this colour can be removed by washing the extract with 1 ml of water. The test extract is similarly washed, and hence the fluoro-chelate is isolated free from extraneous The fluoro-chelate seems to be reasonably insoluble in water.February, 19631 AMOUNTS OF FLUORINE IN BIOLOGICAL SPECIMENS 53 red colour with an almost colourless blank solution.The advantages of the extraction method are shown clearly in Table 111; in the range of 0.1 to 1.0 pg of fluoride, Beer's law is well obeyed for the extracts of the fluoro-chelate, but erratic correlation between fluoride and optical density was obtained when the direct reaction was measured. These findings are different from those of previous workers who, however, measured the reaction with much larger amounts of fluoride. Some preliminary tests have indicated that the fluoro-chelate in the isobutanol solvent can be re-extracted into a small volume of dilute alkali or buffer in aqueous solution.This method of concentration may lend itself to the measurement of even smaller amounts of fluoride. In applying the method to determining fluorine in biological specimens, some further observations were made. Experience with former conventional methods of ashing, distillation and titration for collecting and measuring small amounts of fluorine makes the analyst aware of the need for carefully controlling potential sources of contamination. The proposed method revealed contamination from fluoride that hitherto well accepted procedures did not show. By the proposed method relatively large amounts of fluorine were found in many common laboratory reagents, frequently several micrograms per millilitre of molar solution ; this particularly applied to salts of calcium, lithium and magnesium.But the most serious contamination of all came from the muffle furnace. Several micrograms of fluoride may be washed from the inside of an empty platinum crucible after being in a muffle furnace for a few hours at 400" C, even if the crucible is covered with a lid. When alkali is present in the crucible this source of contamination becomes greater still and subject to considerable variation. The problem may be overcome by placing the crucibles covered with their lids inside another metal container with a well fitting, but not air-tight, lid. Screw-cap lids are not suitable. Attempts to use crucibles of other materials, such as aluminium, nickel, stainless steel or silica, were without success. Losses were experienced of the small amounts of fluoride placed in the crucibles, possibly by adsorption to the surface of the crucibles-this is a well recognised danger in ashing procedures.In order to fix the fluoride, lithium hydroxide with magnesium succinate was found to be the most suitable agent, and clean ashes were obtained. The method has been successfully used for determining the fluorine content of water and of various animal and plant tissues and body fluids; it is hoped that details will appear later in another publication. One important observation arising from the use of the proposed method concerns the fluoride levels of human and animal blood. Hitherto, most workers using conventional methods have reported figures of about 2 pg of fluoride per ml of whole blood. More recently, Singer and Armstrongll obtained values of 0.12 pg per ml of plasma or serum; they collected the fluorine by means of a special micro still, after ashing with magnesium oxide. Many determinations of fluoride in human, dog and sheep plasma and whole blood have now been carried out in this laboratory, and only rarely has any measurable amount been found in plasma and usually less than 0.1 pg per ml in whole blood. Some of the high figures reported in the past may be connected with the uncertainty of the thorium nitrate titration, which is well known to fluorine chemists. There are many important problems still to be studied on the effects of fluorine on all forms of life, and it is hoped that this technique offers a practical approach to the accurate determination of extremely small amounts of this element and with materials normally available in most analytical laboratories. I thank Dr. Martha Vogt, F.R.S., for all her encouragement and helpful advice and Miss Helen Newport for skilful technical assistance. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Belcher, R., Leonard, M. A., and West, T. S., Talanta, 1959, 2, 92. I _ ~ - , J. Chem. SOL, 1959, 3577. Leon'ard, k A., and West, T. S., Ibid., 1960, 4477. Hall, R. J.. Analyst, 1960, 85, 560. Johnson, C. A,, and Leonard, M. A,, J . Pharm. Pharmacol., 1961, 13, 164~. Belcher, R., and West, T. S., Talanta, 1961, 8, 853. Stegemann, H., and Jung, G. F., Hoppe-Seyl. Z., 1959, 315, 272. Frere, F. J., Anal. Chem., 1961, 33, 644. Greenhalgh, R., and Riley, J. P., Anal. Chim. Acta, 1961, 25, 179. Belcher, R., and West, T. S., Talanta, 1961, 8, 863. Singer, L., and Armstrong, W. D., Anal. Chem., 1959, 31, 105. Received September 21st, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800076

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

84 MEYLING AND MEYLING: A MODIFIED ZIRCONIUM - ALIZARIN [Analyst, Vol. 88 A Modified Zirconium -Alizarin Method for determining Fluoride in Natural Waters BY A. H. MEYLING AND J. MEYLING (Bilharzia Field Unit, South African Council for Scienti$c and Industvial Research, Nelspruit, E. Transvaal) A modified zirconium - alizarin method, accurate to within 0.05 p.p.m. of fluoride, is described for determining fluoride in water without prior distillation. The difficulty of variation in colour was overcome by extracting the alizarin with pentanol. Interference from phosphate and other impurities has been overcome. MOST methods for the colorimetric determination of fluoride are based on the fact that fluoride decreases the colour intensity of some metal complexe~.~ ,2 9 3 , 4 3 5 8 The zirconium - alizarin method7,8$9 is probably the most satisfactory for determining fluoride in natural waters.However, it has th'e disadvantages that- (a) the red and yellow colour mixture produced by the alizarin - zirconium reagent is different for each fluoride content ; (b) most of the impurities found in natural waters influence the results; (c) the methods are accurate to only 0.1 p.p.m. of fluorine, even after distillation. These disadvantages are overcome in the rnodification described here. METHOD REAGENTS- PentanoZ-Redistilled industrial pentanol is satisfactory. Magnesium chloride solution, saturated, aqueous. Sodium $Goride solution-A solution containing 0.1 mg of fluorine per m1.10 Zirconium oxychloride solution-Dissolve 0.177 g of zirconium oxychloride, ZrOC1, .8H20, in about 300 ml of distilled water, add 17 rnl of concentrated sulphuric acid and 50.5 ml of concentrated hydrochloric acid, cool, and make up to 500 ml with distilled water.10 Alizarin solution-Dissolve 0.3752 g of alizarin red S in 500 ml of distilled water.10 Alizarin - zirconium reagent solution-Mix equal portions of the alizarin and zirconium Ferric chloride solution-Dissolve 1 g of ferric chloride, FeC1,.6H20, in 50 ml of distilled Sodium acetate solution, 25 per cent.w/v, aqueozbs. It must be clear and colourless. oxychloride solutions immediately before use. water. PROCEDURE- If the specific conductivity of the sample is below 500 pmho, add 1 ml of magnesium chloride solution per 500 ml of sample. Use three glass-stoppered 100-ml cylirtders (the total volume of each cylinder being at least 140 ml) ; in one cylinder (cylinder 1) place 100 ml of distilled water containing 1 ml of magnesium chloride solution per 500 ml and in the other two (cylinders 2 and 3) place 100-ml portions of sample.Add 1 ml of sodium fluoride solution to the contents of cylinder 3 and 1-ml portions of distilled water to the contents of cylinders 1 and 2. Add 10 ml of alizarin - zirconium reagent solution to the contents of each cylinder, mix well, and allow the colour to develop for one hour in the dkYk. Then add to each 10ml of pentanol, and shake well for one minute. Set aside in the dark until the top layer is clear (usually 15 minutes, but we allow 30 minutes). Transfer with a dropper part of the pentanol layer from each cylinder to a separate 1-cm cuvette.Set a spectrophotometer at wavelength 430 mp, adjust to read zero against the solution from cylinder 1, and then measure the optical densities of the solutions from cylinders 2 and 3.February, 19631 METHOD FOR DETERMINING FLUORIDE IN NATURAL WATERS 85 CALCULATION- E 3 , then the concentration of fluorine in the water is given by the relation- If the optical density of the solution from cylinder 2 is E, and of that from cylinder 3 is p.p.m. E2 E3 - E, DISCUSSION OF THE METHOD By extracting the yellow colour of the free alizarin with pentanol the interference caused by red zirconium lake was avoided, as was that caused by colour and slight turbidity of the water. Pentanol was chosen because alizarin sulphonic acid is soluble in water and alcohol.It was easily available and, as it is relatively insoluble in water, it was an obvious choice. It was shown by adding different known amounts of fluoride to distilled and other waters and measuring the optical densities that the amount of free alizarin was directly proportional to the amount of fluoride in the sample (see Table I ) ; waters of low conductivity were an exception. It was therefore necessary to increase the salt concentration of waters having low specific conductivity. The optical density per 1 p.p.m. of fluorine decreases with increasing salt content in any particular water; this is most marked with aluminium salts (see Table I). Hence the necessity of using a third cylinder containing the sample to be tested, to measure the optical density per 1 p.p.m.of fluorine. TABLE I OPTICAL DENSITY PER 1 p,p.m. OF FLUORIDE FOR DIFFERENT WATERS Sample Fluoride present, Optical density Optical density per 1 p.p.m p.p.m. of fluoride - r 0 0 0.2 1. Distilled water . . "1 0.5 1.0 0 0.2 2. Distilled water containing r I 1.0 r 0 0.5 p.p.m. of aluminium*{ 0.5 0.122 0-255 0.385 0 0.046 0.144 0.386 0 0.61 0.51 0-38 0.23 0.29 0.39 - 0.2 0.098 0.49 0.5 0.140 0.48 I 1.0 0.484 0.48 3. -Artificial watert . . . . 0 0.2 0.5 4. Distilled water containing 1 ml of magnesium chloride per 500 ml . . 0 0.079 0.212 0.443 - 0.40 0.42 0.44 I 0 0 1.0 0.364 0.36 . .{ 2.0 0-702 0-35 3. Buffelspruit water . . 3.0 1.040 0.35 0 0 6. .Artificial water7 containing1 0-2 0.067 0.34 0.5 p.p.m. of aluminium* 7 0.5 0.174 0.35 I 1.0 0.359 0.36 0 0 0-2 0.040 0.20 0.5 0.106 0.2 1 1.0 0.242 0.24 0 0 1 ml of magnesium 1 0.2 0.078 0.39 chloride solution per{ 0.5 0.201 0.40 500 ml and 0-4 p.p.m.of I 1.0 0-41 1 0.41 * Added as KAI(S04),.12H,0. t Distilled water containing 126 p.p.m. each of calcium carbonate, magnesium sulphate and The calcium carbonate was first dissolved in hydrochloric acid. - - r 7 . &Artificial water1 containing J - -1 1 p.p.m. of aluminium* 8. Ihtilled water containing r . . c phosphate .. sodium hydrogen carbonate.86 MEYLING AND MEYLING A MODIFIE-D ZIRCONIUM - ALIZARIN [Analyst, Vol. 88 To show that the optical density of the lblank solution was equal to that of the sample without fluoride, eight waters were made up and examined. The optical density- of the distilled water blank was adjusted to 0-020.The optical densities of waters 1 to 5 (see Table 11) showed no significant differences, and the influence on the final result was less than 0-05 p.p.m. of fluorine. Il'ater No. 1 2 3 4 5 6 7 S \Vater No. TAE~LE I1 INFLUENCE OF VARIOUS SUBSTANCES ox BLASK VALUE Composition Distilled water . . . . .. . . . . . . . . . . Distilled water containing 0-5 p.p.m. of aluminium* . . . . . . Artificial watert . . . . .. . . . . . . . . .. Artificial watert containing 0.5 p.p.m. of aluminium* . . . . . . Distilled water containing 1 ml of magnesium chloride solution per 500ml . . . . . . . . . . . . , . . . .. Distilled water containing 1 ml of magnesium chloride solution per Distilled water containing 1 ml of magnesium chloride solution per Distilled water containing 1 ml of magnesium chloride solution per 500 ml and 0.5 p.p.m.of phosphate . . . . . . . . . . 500 ml and 1.0 p.p.m. of phosphate . . . . . . . . . . 500 ml and 1.5 p.p.m. of phosphate . . . . . . . . . . * Added as K.41(S04),. 12H,O. t For composition, see Table I. TABLE I11 ANALYSIS OF WATERS USED IN EXPERIMENTS Total Colour 1 2 Boundary Creek 5 to 10 3 Buffelspruit . . 10 to 15 4 Nelsrivier . . 40 5 BoreholeNo. 1 . . 5 6 RoreholeNo. 2 . . 5 Nelspruit tap water 0 to 5 Optical density 0.020 0.022 0.010 0.0 18 0.015 0.043 0.063 0.086 Specific pmho conductivity, pH 145 8.5 890 8-4 935 8.1 47 7.7 66 7.3 40 - Alkalinit?-, as CaCO,, p,p.m. 62 500 444 24 30 16 hardness, as CaCO,, p.p.m. 71 536 372 18 20 4.4 Difference + 0.002 -0*010 - 0*002 - 0.005 + 0.023 - 4- 0.043 + 0.066 Sulphate, Chloride, p.p.m.p.p.m. 14 3 46 32 25 59 1 3 1 4.5 4.0 - TABLE I V DETERMINATION OF FLUORIDE I N VARIOUS WATERS Optical density Optical per Optical density Error, so. Natural water 1 p.p.m. density Fluoride Fluoride after Fluoride Actual p.p.m. of of found, added, adding found, fluoride. of fluoride sample p.p.m. p.p.m. fluoridve p.p.m. 0-468 0.032 0.068 0.5 0.262 0.56 0.413 0 0 0.5 0.206 0.50 0.449 0.033 0.073 0.5 0.265 0.59 0.442 0.078 0.18 0-5 0.292 0.66 0.408 0.010 0.025 2.0 0.810 1.98 . . 0.349 0.072 0.21 0.5 0.237 0.68 0.5 0.248 0.66 4 Table I11 5 3 containing 0-5 p.p.m. of 5 containing 0.25 p.p.m. of 6 containing 1.2 p.p.m. of iron* 0.398 0.0121 0.030 2-0 0.810 2.04 3 containing 1 p.p.m.of phos- phatet . . .. . . 0.384 0.08S 0.22 - 3 containing 1 p.p.m. of phos- phate? and 1 p.p.m. of fluorine . . . . . . 0.312 0.380 1.22 - i Compositions as in 0.392 0.077 0-20 0.5 0-254 0.65 6 J aluminium . . aluminium . . . . 0.373 0.06Et 0.18 - - - - * Added as FeC1,.6H,O. 7 Phosphate was removed by precipitation with ferric chloride. p.p.m. fluoride 0.57 -0.01 0.50 0 0.57 $0.02 0.70 -0.05 0.68 -0.02 2.02 -0.04 0.71 -0.03 0.68 -0.02 2.03 i-0-01 0.20 t o - 0 2 1.20 4-0.02February, 19631 METHOD FOR DETERMINING FLUORIDE I N NATURAL WATERS 87 The optical density of the blank solution was, however, influenced by the presence of phosphate (see Table 11). This influence was overcome by adding to the blank solution the same amount of phosphate as was present in the sample.If the amount of phosphate was not known, it was removed by neutralising 500 ml of the water with N hydrochloric acid or sodium hydroxide, with litmus paper as indicator. One millilitre of ferric chloride solution and then 2 ml of sodium acetate solution were added, and the solution was boiled and then filtered hot. When the filtrate was cool it was neutralised with N hydrochloric acid and made up to the original volume with distilled water. Accu RACY- The fluoride contents of 11 different waters were determined; the compositions of these waters are shown in Tables I11 and IV. A known amount of fluoride was added to each of the waters, and the total fluoride content was determined. The difference between the determined and calculated amounts showed that the error was not more than 0.05 p.p.m. of fluorine (see Table IV). We thank the South African Council for Scientific and Industrial Research for permission to publish this paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Steiger, G., J . Amer. Chem. Soc., 1908, 30, 219. Kortum-Seiler, M., Angew. Chem., 1947, 59, 159. Erler, K., 2. anal. Chem., 1950, 131, 103. Ingols, R. S., Shaw, E. H., Eberhardt, VC‘. H., and Hildebrand, J . C., Anal. C k e m , 1950, 22, Nonnier, D., Rusconi, Y., and Wenger, P., Helv. Chim. Acta, 1946, 29, 521. Urech, P., Ibid., 1942, 25, 1115. Megregian, S., and Mayer, F. J., J . Amer. Wat. W k s Ass., 1952, 44, 239. Sanchis, G. M., I n d . Eng. Chem., Anal. Ed., 1934, 6, 134. Lamar, VC’. L., Ibid., 1945, 17, 148. A4merican Public Health Association, “Standard Methods for the Examination of IVater, Sewage Received July 23rd, 1962 799. and Industrial Wastes,” 10th Edition, p. 98.

ISSN:0003-2654    

DOI:10.1039/AN9638800084

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

88 MORRISON AND WILSON THE ABSORPTIOMETRIC [A?’Za[’St, VOl. 88 The Absorptiometric Determination of Silicon in Water Part I. Formation, Stability and Reduction of a- and /3-Molybdosilicic Acids (Central Electricity Research Laboratories, Cleeve Road, Leatherhead, Surrey) The effect of experimental conditions on the formation, stability (in the presence of reagents used for destroying molybdophosphoric acid) and reduc- tion of a- and p-molybdosilicic acids -has been determined. The effectiveness of several reagents for preventing inierference from phosphate has also been investigated. Both cr-molybdosilicic: acid reduced by stannous tin and p-molybdosilicic acid reduced by 1 -amino-2-naphthol-4-sulphonic acid, should be suitable for precise methods of determining “reactive” silicon in water.AT the steam pressures and temperatures used in modern power stations, silicic acid is appreciably soluble in steam, The concentration of silicon in boiler and make-up water is therefore controlled to ensure that the steam contains not more than about 0.02 p.p.m. of silica; above about 0.02 to 0.04 p.p.m. of silica undesirable deposits may be formed on turbine blades. Methods are therefore needed for deteirmining both the “reactive”* and total silicon concentrations in a range of aqueous samples from steam - water circuits in power stations. Thus the methods must be suitable for conde:nsate, feed-water and steam (0-002 to 0-05 p.p.m. of silica), make-up water (0.005 to 0-2 p.p.rrt. of silica) and boiler water (0.2 to 75 p.p.m. of silica). The ratio of phosphate (p.p.m.) to silica (p.p.m.) is usually between 0 and 20, but occasionally ratios as high as 50 may occur.The methods must also be precise, so that they can be used for determining small amounts of “non-reactive” silicon as the difference between the total and “reactive” silicon contents. Absorptiometric methods in which the reduced a- and P-molybdosilicic acids are used appear to be the only sufficiently sensitive means of determining extremely small concen- trations of silicic acid. They can also be used for determining total silicon provided all forms of silicon are converted to silicate or monomeric silicic acid in an earlier step. This paper reports an investigation of the molybdosilicic acids to determine suitable conditions for each method; Parts 112 and 1113 give details of the methods and the results obtained.Two forms of molybdosilicic acid, a- and p-, are of analytical interest and were first distinguished by Stri~kland.~ Each form can be reduced to several different blue compounds with various reducing agents, and Strickland4 has determined the properties of several forms and the conditions for their formation. The relative amounts of ci- and P-molybdosilicic acids produced and their stabilities appear to be affected by factors such as the concentrations of acids, molybdate, neutral salts and reagents such as oxalic or tartaric acid used for des- troying molybdophosphoric acid. However, the literature does not give quantitative values for all these effects, and they have, therefore, been investigated.Boiler waters often also contain added phosphate. EXPERIMENTAL APPARATUS, REAGENTS AND TECHNIQUE- All optical-density measurements were made in 4-cm cuvettes with a Hilger Cvispek spectrophotometer against distilled water in the reference cuvette. Measurements of pH were made with an E.I.L. type 23A pH meter, with glass and saturated-calomel electrodes. A stock solution of sodium silicate was prepared by fusing 1.000 g of Specpure silicon dioxide (ignited at 1100” C) with 5 g of analytical-reagent grade sodium carbonate, dissolving the cooled melt in water, and diluting to 1 litre with distilled water. The silicon content was checked gravimetrically after two dehydrations with perchloric acid. Aliquots of this * “Reactive” silicon-mainly monomeric and dimeric silicic acid1-is defined in this paper as those forms of silicon that react with ammonium molybdate in 10 minutes to form molybdosilicic acid under the conditions of the method given in Part I1 of this series.February, 19631 DETERMINATION OF SILICON IN WATER.PART I 89 solution were diluted with distilled water to give solutions having the desired silicon concen- trations. Distilled water from a Manesty still was used throughout; it contained 0.005 p.p.ni. of silica or less and was stored in a polythene bottle. For factorial experiments, one large uniform batch of water was used in each experiment. The temperature of the laboratory varied between 20" and 25" C during this work. 01-MOLYBDOSILICIC ACID Fo RM ATI o N- Efeect of pH-Strickland reported4 maximum formation of cc-molybdosilicic acid when silicic acid and molybdate were allowed to react a t pH 3.7 to 4.0; formation was incom- plete at higher pH values and /3--molybdosilicic acid was formed a t lower pH values.It was not clear from his paper whether the product in the pH range 3.7 to 4.0 would be entirely a-molybdosilicic acid at lower concentrations of silicon than he used. Preliminary work showed that this point could best be studied by reducing the product with stannous ions and measuring the optical densities of the solutions at the wavelengths of maximum absorption of the reduced 01- and p-forms, ie., 742 and 810 mp, respectively. The absorption spectra in Figs. 3 and 7 (see pp. 91 and 95) show that the optical-density ratio of the molybdosilicic acids ( x t o /3) is 1-09 at 742 mp and 0.414 a t 810 mp.Non-formation of either form would cause a decrease in optical density at both wavelengths. 0-700 2 3 4 PH Fig. 1. Effect of pH on the formation of molybdosilicic acids: 0, 0.014 M MOO,^- (molybdate acidified before mixing with silicate): A, 0.028 M (molybdate acidified before mixing with silicate) ; 0, 0.014 M Mooda- (silicate and molybdate mixed and then slowly acidified); X, 0.014 M (as for 0, but solution then heated in boiling water for 10 minutes); n, 0.014 M (as for 0, but solution then heated in boiling water for 20 minutes). The concentration of silicon was 0.5 p.p.m., as SO,; all concentrations were in the final solution Fig. 1 shows the effect of various ways of combining solutions of ammonium molybdate, sodium silicate and hydrochloric acid.For these experiments, different amounts of am- monium molybdate in 50 ml of water were used; 5 ml of sodium silicate solution (10 p.p.m. of silica) were taken, and the solutions were acidified with dilute hydrochloric acid (1 + 200) to the desired pH. Ten minutes after the solutions had been mixed (except when they were heated), the acidity was increased by adding 15 ml of 17.5 N sulphuric acid, 2 ml of stannous oxalate reducing agent5 were added, the volume was adjusted to 100 ml, and the optical densities were measured at 742 and 810mp. The measurements at 810 mp are most sensitive to changes in the ratio of o( to 18, and Fig. 1 shows that this ratio was markedly dependent on pH. However, the effect of pH90 MORRISON AND WILSON : THE ABSORPTIOMETRIC [Anahst, VOl.88 was considerably decreased if solutions were heated before being reduced. Because of the pH-dependence of the unheated solutions, heating appeared to be a desirable step in an analytical method based on a-molybdosilicic acid, and the necessary heating period was accordingly determined. Efect of period of heating-For these experiments, molybdosilicic acid was formed by adding 5ml of silicate solution (10 p.p.m.. of silica) to 5ml of acidified 5 per cent. am- monium molybdate solution that had been diluted to 60ml in a polythene bottle. The bottles were heated in a water bath for different periods and then cooled; 15ml of 1 7 . 5 ~ sulphuric acid and, 1 minute later, 2 ml of 0.2 per cent. stannous reducing agent (as the chloride) were added.The solutions were then diluted to 100m1, and the optical densities were measured at 742 and 810 mp. The amount of acid added in the first stage was 10 ml of either 0.5 or 0-05 N sulphuric acid; the former probably corresponds to the worst possible conditions for producing a-molybdosilicic acid, as this acidity gives the maximum initial amount of the /3-form (see section on /3-molybdosilicic acid, p. 93). The latter acidity gives an initial pH of about 2.5, which is about the middle of the pH range for constant optical density when the solutions are heated (see Fig. 1 ) . Table I shows the period of heating after .which constant optical densities were obtained at both wavelengths. It is considered probable that the constant optical densities correspond to the molybdosilicic acid being completely in the a - f ~ r m .~ For these tests, the silicon concen- tration used was near the upper concentration of interest ; much larger concentrations may require longer heating periods.6 Small variations in the concentrations of ammonium molybdate are not expected to have any large effect on the rate of conversion of the /3- to the a-form.6 In all further work in which solutions of pure a-molybdosilicic acid were required (except when otherwise stated), they were prepared as described above, 10 ml of 0.05 N sul- phuric acid being used, and heating for 60 minutes in boiling water. TABLE I HEATING PERIOD REQUIRED TO GIVE 100 per cent. a-MOLYBDOSILICIC ACID Final concentrations of silicate and ammonium molybdate were 0.5 p.p.m.of silica and 0-014 M MOO^^-, respectively Period of heating required to give constant optical densities at water bath temperatures of- +- 7 Sulphuric acid added initially, ml N 80" c, 100" c, minutes minutes 10 0.5 60 30 10 0.05 15 Not determined A Stability and absorption curve-The reported stability of a-molybdosilicic acid416 was confirmed; the optical density at 430 mp of unreduced solutions, prepared as described above, was constant to within + 2 per cent. for at least 72 hours. The absorption curve, between 370 and 450 mp, was similar to those reported by other worker^.^*^ EFFECT OF TARTARIC, OXALIC OR HYDROCHLORIC ACID- For these tests, different amounts of the acids were added to identical solutions of a-molybdosilicic acid, prepared by slow acidification of a solution containing silicate and ammonium molybdate (see Fig.l), and the optical densities of the solutions were measured at 430 mp over a period of 60 minutes. The rates of decrease in optical density during the first 15 minutes were approximately constant and are shown in Fig. 2. Because of the method of preparation used, the a-molybdosilicic acid probably contains about 10 per cent. of the p-form, and the curves for the x-form will be somewhat high. REDUCTION- By stannous tin-Milton showed' that stannous chloride could be successfully used for reducing a-molybdosilicic acid in 2 N sulphuric acid. Our experiments showed that reducing a-molybdosilicic acid with stannous tin in the presence of 2.5 N hydrochloric acid gave an extremely unstable product.The optic:al density varied from experiment to experimentFebruary, 19631 DETERMINATION OF SILICON IN WATER. PART I 91 and decreased rapidly when the solution was allowed to stand. By carrying out the reduction in 2.5 N sulphuric acid, with only 0.0033 N chloride present, an intense blue-green colour, stable for an hour, could be produced. The reducing agent was prepared by dissolving 0.1 g of tin foil in 1.5 ml of concentrated hydrochloric acid and diluting to 50 ml with water. The absorption curve of this stable reduced product is shown in Fig. 3. The spectrum below 550 mp was not reproducible and changed with time and the concentration of reducing agent. This part of the spectrum is probably caused by reduction of the excess of molybdate, as a similar curve was also obtained with reagent blank solutions.The absorption spectrum Molar ratio of hydroxy acid to MOO,'- or concentration of hydrochloric acid, N Fig. 2 . Destruction of a- (full lines) and /3- (broken lines) molybdosilicic acids by acids : 0, oxalic acid; A, hydrochloric acid ; 0, tartaric acid. The concentrations of silicon and molybdate in the final solution were 5 p.p.m., as SO,, and 0.014 M respectively Fig. 3. Absorption spectra of cc-molybdosilicic acid reduced by: curve A, stannous tin in 2.5 N sulphuric acid; curve B, l-amino-2- naphthol-4-sulphonic acid ; curve C, reagents only, reduced by stannous tin in 2.5 N sulphuric acid. The concentrations of silicon and molybdate in the final solution were 0.5 p.p.m., as SiO,, and 0.014 M MOO,,-, respectively92 MORRISON AND WILSON : THE ABSORPTIOMETKIC [Analyst, VOl. 88 is similar to that obtained by Strickland4 for the +4e reduced compound, except for the peak at 635 mp, which shows on Strickland’s curve as a flat portion at 650 mp.Strickland’s curve was measured in the absence of excess of molybdate, and the a-molybdosilicic acid was prepared in a different manner from that used in this work. Fig. 3 shows that, for analytical use, the most suitable wavelengths for optical-density measurements are 742 mp and 635 mp. The former is preferred because, although the peak heights are much the same, the optical density of the blank solution is less at this wavelength. The effects of light and the concentraticrns of stannous tin and chloride on the stability of the reduced or-molybdosilicic acid are shown in Figs.4 and 5; the coefficient of variation of the optical densities was about k0.1 per cent. The figures show, therefore, that the final concentration of stannous tin should be 0.34 x M and that of the chloride less than 0-008 M, if the optical densities of the reduced solutions are to be constant for one hour. Further, the solutions should be kept out of direct sunlight. Time after reduction, minutes Effect of stannous tin and chloride on the stability of reduced cc-molybdosilicic acid : e, 0.085 x M Sn2+, 0.0033 M C1-; X, 0.17 x M SnZ+, 0.0033 M C1-; +, 0-25 x 10-3 M Sn2+, 0.0033 M C1-; 0, 0.34 x 10-3 M Sn2+, 0-0033 M C1-; m, 0.68 j< 10-3 M Sn2+, 0.0033 M C1-; m, 0-34 x M Snzf, 0.0083 M C1-; A, 0.34 x M Sn2+, 0.028 M C1-.Fig. 4. All concentrations were in the final solution In the experiments described above, reduction took place in 2.5 N acid. This involved adding a large volume of acid, which raised the temperature of the test solution. Some tests were therefore made to determine whether or not tartaric acid could be used in place of sulphuric acid. When or-molybdosilicic acid was reduced with stannous tin in the presence of 0.053 M tartaric acid, a blue colour resulted that remained stable for 20 minutes and then slowly faded. The absorption spectrum initially resembled that of Strickland’s +5e com- pound,4 but after 90 minutes it had changed to that of the +4e compound. Thereafter, the spectrum remained in this form as the optical density of the solution continued decreasing. Strickland showed that strongly acidified inolybdate instantly oxidised the +5e compound to the +4e form; the +5e form is, therefore, not produced in the presence of excess of acidified molybdate.However, the experiment just described shows that, in the presence of tartaric acid, which complexes the molybdate ion, a considerable amount of +5e form is produced. Thus, the reduction of or-molybdosilicic acid by stannous tin in the presence of tartaric acid is unsuitable for the basis of an analytical method. By l-amin0-2-naphthol-4-s~l$honic acid-Preliminary experiments showed that the optimum concentration of l-amino-2-naphl:hol-4-sulphonic acid for reducing or-molybdosilicicFebruary, 19631 DETERMINATION OF SILICON IN WATER.PART I 93 acid in the presence of 2-5 N sulphuric acid was about 0.004 per cent. The absorption spectrum of this reduced product in the presence of 2.5 N sulphuric acid is shown in Fig. 3. The curve in the presence of 0.053 M tartaric acid had the same shape, but the optical density was about 0.004 to 0-008 lower at most wavelengths. Because of the low absorption, compared with that given by stannous tin, no further work was done with this reducing agent. I: 1 0 Time after reduction, minutes Fig. 5. Effect of light on the stability of reduced a-molybdosilicic acid: 0, liquid kept in flask in thc dark; 0, liquid kept in fiask in a light room out of sunlight; A, liquid kept in flask in sunlight; , liquid kept in a cell in the dark; X, liquid kept in a cell in the light path of a spectrophotometer P-MOLYBDOSILICIC ACID FORMATION- Factors afectiizg the formation of /3-molybdosilicic acid-Strickland showed4 that the ratio of H+ to MOO$- is the most important factor governing formation of /3-molybdo- silicic acid; the optimum ratio lies between 3 and 5.He did not show that his product consisted entirely of the p-form, and possibly there is always some a-molybdosilicic acid present. If the p-form is to be used in an analytical method, it is desirable that the amount produced (from a given concentration of silicic acid) should not depend critically on experimental conditions. Reducing molybdosilicic acid with l-amino-2-naphthol-4-sulphonic acid gives a sensitive indication of the relative amounts of a- and /&forms present in a solution ; at 810 mp, the absorption ratio, a to /3, is 0.2 (see Figs.3 and 7). The effects produced by changes in certain experimental factors were therefore tested. For all these experiments, the H+ to Mo042- ratio was 4.5 and reduction was carried out with 1-amino-2-naphthol-4-sulphonic acid. A ratio of 4 was not used, because an acidified molybdate reagent with such an acidity often develops a blue colour on standing. Although this colour is destroyed by the tartaric acid usually added, it may be avoided by increasing the acidity to give a H+ to 11100,~- ratio of 4-5. In the standard procedure (see Table 11) 2-5 ml of acidified ammonium molybdate solution (8-9 per cent. w/v) were added t o 50 ml of water containing 5 ml of silicate solution containing 15 p.p.m.of silica. After 10 minutes, sufficient tartaric acid solution (28 per cent. w/v) was added to make the molar ratio, tartaric acid to Mo0,2-, 3.6, and 5 minutes later 2 ml of 0.2 per cent. l-amino-2-naphthol-4-sulphonic acid solution (containing 2-4 per cent. of hydrated sodium sulphite and 14 per cent. of anhydrous potassium metabisulphite) were added. The solution was diluted to 100 ml with water, and the optical density was measured at 810nip. The results ofduplicate measurements (corrected for reagent blank values) are shown in Table 11, which indicates that any effects produced by the factors tested were small.94 MORRISON AND WILSOX : THE ABSORPTIOMETRIC [Analyst, Vol. 88 TABLE I1 FORMATION OF /3-MOLYBDOSILICIC: ACID UNDER VARIOUS CONDITIONS All concentrations are in the final solution Level of the variable factol- Standard conditions* .. .. .. . . . . .. . . 0.05 M MOO,,- . . . . . . . . . . . . .. . . 0.05 M Moo,,- and 50 p.p.m. of . . .. .. . . 0.0125 M MOO,,- and 50 p.p.m. of . . . . . . 7.5 . . . . 10 . . . . Initial volume of sodium silicate solution, ml i" 30 . . . . . . . . [:: : * Temperature during formation of #?-molybdosilicic acid Reversed procedure, i.e., 5 ml of sodium silicate added to 2-5 ml of . . 3oo ammonium molybdate reagent, diluted to 50 ml with water Po Optical density compared with that for the standard conditions, per cent. 100.0 (taken as standard) 102.7, 102.0 (mean 102.4) 99.3, 99.5 (mean 99.4) 98.6, 98.4 (mean 98.5) 98.7, 98.3 (mean 98.5) 99.0, 99.2 (mean 99-1) 100.9, 100.2 (mean 100.6) 99.6, 99.8 (mean 99.7) 99.4, 99.5 (mean 99.4) 99.9, 100.2 (mean 100.0) 99.1 (mean 99.1) 99.3, 99.2 (mean 99.2) * Standard conditions: 0.75 p.p.m.of silica; 0.0125 M 0 p.p.m. of phosphate; initial volume 50 ml; temperature 21' C; acidified ammonium molybdate solution added to the sodium silicate solution. 'Time, minutes Fig. 6. Change in absorption of final solution with time allowed for formation of /3-molybdosilicic acid: X, 0.05 M Moo,,-, 0.8 p.p.m. of SiO, and 25 p.p.m. of PO,3-; a, 0-0125 M Moo,,- and 0.01 p.p.m. of SiO,; 0, 0.0125 M and 0.2 p.p.m. of SO,; 0,0.0125 ~MoO,~-and0.44p.p.m. of SiO,; 0 , 0.0125 M Mo0,2-, 0.44 p.p.m. of SiO, and 44 p.p.m. of PO,3-. All concentrations were in the final solution Period required for formation of @zolybdosilicic acid-The length of time required to allow complete formation of /3-molybdosilicic acid was determined by varying the period between adding the molybdate and tartaric acid reagents; all other conditions were similar to those for the standard procedure desci-ibed above, except that the volume of samples plus acidified molybdate reagent had different values, between 75 and 102.5m1, in the various experiments.The experiments were made at different silicon concentrations and, in order to show all the results on one curve, they are given in Fig. 6 as percentages of the maximum optical density for each set of conditions. The figure shows that the optical density had always reached a maximum value in 3 minutes. The figure also confirms the reported4 decrease in absorption caused by conversion of /3-molybdosilicic acid to the a-form.Salts have been reported4 to accelerate the conversion, and this effect was investigated by measuring the optical densities of unreduced /3-molybdosilicicFebruary, 19631 DETERMINATION OF SILICON IN WATER. PART I 95 acid over a period of time; the concentration of ammonium molybdate was 0.028 M in the final solution for these tests. At a silicon concentration of 5 p.p.m., as SO,, the rate of decrease in optical density (at 430 mp) was 0.05 per cent. per minute and 0-06 per cent. per minute in the presence of 0 . 3 4 ~ (20,000 p.p.m.) sodium chloride. Absorption curve-The absorption curve between 370 and 500 mp was similar to those given in the literature4 for /3-molybdosilicic acid.EFFECT OF TARTARIC, OXALIC OR HYDROCHLORIC ACID- a-molybdosilicic acid. to the a-form. Fig. 2 shows these effects, which were determined in a similar manner to those for No correction has been made for the change of JS-molybdosilicic acid Wavelength, rnM Fig. 7. Absorption spectra of p-molybdosilicic acid reduced by : curve A, stannous tin in 2.5 N sulphuric acid; curve B, l-amino-2- naphthol-4-sulphonic acid in 2-5 N sulphuric acid: curve C, l-amino-2- naphthol-4-sulphonic acid in 0.053 M tartaric acid ; curve D, reagents only, reduced by l-amino-2-naphthol-4-sulphonic acid in 0.053 M tartaric acid. The concentrations of silicon and molybdate in the final solution were 0.483 p.p.m., as SiO,, and 0.014 M respectively REDUCTIOX- By 1-amino-2-na~hthol-4-su@honic acid-Fig.7 shows the absorption curves of #3-molyb- dosilicic acid reduced by 1 -amino-2-napht hol-4-sulphonic acid under different conditions. The shapes of the curves are similar to those given in the literature for analogous solutions. The figure shows that the absorption in 2.5 N sulphuric acid is greater than that in tartaric acid. This increased absorption was found whether sulphuric acid was added before or after reducing the molybdosilicic acid and is, therefore, probably caused by increased absorption of light by /3-molybdosilicic acid in solutions of high acidity. Reduced and unreduced a-molybdosilicic acid gave a similar effect . The optical density at 810 mp was constant within f0.14 per cent. over a period between 15 minutes and 24 hours after reduction, when the final concentration of 1-amino-2-naphthol- 4-sulphonic acid was 0.17 or 0.34 x M (0.004 or 0.008 per cent.) ; at lower concentrations than about 0.04 x M the optical density increased slowly over at least an hour and, a t higher concentrations than 0.34 x M, crystals of 1-amino-2-naphthol-4-sulphonic acid formed in the solution. Between 0.34 and 0.042 x M, the maximum optical density in- creased steadily with decrease in concentration; it was 0.8 to 2-4 per cent.more at 0.042 than at 0.34 x 1 0 - 3 ~ , and the results were unaffected by 25 p.p.m. of phosphate. By stannous tin-Fig. 7 shows the absorption spectrum of /3-molybdosilicic acid reduced by stannous tin in 2-5 N sulphuric acid. The small peak at 635 mp was not found96 MORRISON AND WILSON : THE ABSORPTIOMETKIC [Andyst, vol. 88 in 2.5 K hydrochloric acid, nor was it shown on Strickland's curve,4 which was measured in 1 N hydrochloric acid.Fig. 7 also shows that the absorption was greater than that obtained when l-amino-2-naphthol-4-sulphonic acid was used; the difference is not large, but the curves have a slightly different shape. Two possible explanations could account for this difference. Either different reduced states of /3-molybdosilicic acid are produced by stannous tin and l-amino-2-naphthol-4-sulphonic acid or variable amounts (about 20 per cent .) of cc-molybdo- silicic acid are present in the supposedly pure P-form. There is evidence for both possi- bilities, and it should be possible to decide between them by accurately determining the number of electrons, per molecule of molybdosilicic acid, involved in the reduction by the two reagents.In the presence of 0.004 per cent. of stannous tin (the optimum concentration for stability of the reduced /3-molybdosilicic acid) and 2.5 N sulphuric acid, the optical density of the final solution decreased by about 0.4 per cent. in the first hour; the decrease was somewhat erratic from experiment to experiment. This has not been done in the work described here. DESTRUCTION OF MOLYBDOPHOSPHORIC ACID The previous sections have shown that a-molybdosilicic acid is best reduced by 0.004 per cent. stannous tin in the presence of 2.5 N sulphuric acid and P-molybdosilicic acid bji 0-004 per cent. l-amino-2-naphthol-4-sulphuric acid in the presence of tartaric or oxalic acids.The effect of these two sets of conditions on the stability of molybdophosphoric acid was therefore determined. UNDER CONDITIONS FOR PRODUCING ~~-MOLYBDOSILICIC ACID- An 8-ounce polythene bottle (containing 50 ml of water, 5 ml of 5 per cent. ammonium molybdate solution, 10 ml of 0.05 N hydro- chloric acid and the desired amount of phosphate) was placed in boiling water for one hour. The solution was set aside to cool, and then 15 ml of 17.5 N sulphuric acid and, 1 minute later, 2 ml of 0.2 per cent. stannous (as chloride) solution were added. The solution was then cooled to 22" C, diluted to 100 ml, and its optical density measured at 742 mp within one hour. Phosphate was added as ammonium phosphate, prepared by passing 1 N ammonium hydroxide solution through a column of Amberlite IR45 anion-exchange resin in the phosphate form.When this solution was diluted to give a phosphate concentration of 25 p.p.m., the silicon content of the diluted solution was less than 0-001 p.p.m., as 50, (as determined by a /3-molybdosilicic acid method). In the presence of, 25 p.p.m. of phosphate a variable optical density was obtained, equivalent to 0.02 to 0.08 p.p.m. of silica, in the presence or absence of silicon; smaller con- centrations of phosphate had a larger effect per p.p.m. of phosphate. The effect was not reduced by increasing the period of the high-acidity stage, but, when the solutions were not heated, the effect was equivalent to only 0-005 p.p.m. of silica. These results, therefore, show that the substance causing the effect was formed mainly during the heating stage and was not easily destroyed by 2.5 N sulphuric acid.The substance is probably a fine precipitate of molybdophosphoric acid (see next section), and the method is therefore not particularly suitable for samples containing much phosphate. For the maximum effect found, 1 p.p.m. of phosphate was equivalent to 0.008 p.p.m. of silica. The standard procedure was as described below. UNDER CONDITIONS FOR PRODUCING P-MO LYBDOSILICIC ACID- To determine the effect of the concentration of tartaric or oxalic acids, the experiment described below was carried out. To 85 ml of water containing 118 p.p.m. of phosphate (as Na,HP04) were added 4 ml of acidified molybdate solution (ratio H+ to MOO,,- = 4.5). Ten minutes later, 4 ml of either tartaric axid or oxalic acid solution were added, and, after a further 5 minutes, 4 ml of 0.094 per cent.l-amino-2-naphthol-4-sulphonic acid solution were added, and the optical density was then measured at 810 mp. The results (corrected for reagent blank values) are shown in Fig. 8 and were used to choose a suitable value for the molar concentration ratio, hydroxy acid to MOO,^-, to be used in the next experiment. To determine the rate of destruction of molybdophosphoric acid by these two hydroxy acids, some tests were carried out exactly as described above, except that the time between adding the hydroxy acid and the l-amino-2-naphthol-4-sulphonic acid was varied. The results (corrected for reagent blank values) are shown in Fig.9. The figure shows that,February, 19631 DETERMINATION OF SILICON IN WATER. PART I 97 when a suitable amount of hydroxy acid was used, most of the molybdophosphoric acid was destroyed within 4 minutes, although there was a small residual absorption. The residual optical density is apparently caused by a fine yellow precipitate (probably molybdophosphoric acid) , which, under some conditions, is heavy and may cause serious interference. Important factors governing its formation (at a H+ to MOO:- ratio of 4 to 4.6) Molar ratio of hydroxyacid to MOO,’- Effect of acids on molybdophosphoric acid: 0, tartaric acid; 0, oxalic acid. The concentrations of molybdate and phosphate in the final solution were 0.0125 M and 103 p.p.m. of respectively. The interval between adding hydroxy acid and reducing agent was 5 minutes Fig.8. Time, minutes Fig. 9. Variation in residual absorption with time allowed for destruction of molybdophosphoric acid by hydroxy acids: 0, tartaric acid, 103 p.p.m. of Pop3- and 0.0125 M X, oxalic acid, 103 p.p.m. of PO,3- and 0.0125 M MoO,Z-. The ratio of tartaric acid to was 3.5 and of oxalic acid to was 2.0 are the concentrations of phosphate and molybdate and the history of the flasks in which the reaction is carried out (“seeding” effect) ; other factors have not been investigated. “Seeding” effects were detected because successive determinations in the same flask showed increasing interference from phosphate; the effect could be overcome by washing the flasks with 2 N ammonium hydroxide solution. At a molybdate concentration in the final solution of 0-05 M, no seeding effect was detected in the presence of 10 p.p.m.of phosphate (as Na,HPO,) in the final solution after seven98 MORRISON AND WILSON : THE ABSORPTIOMETRIC [APZdJJSt, VOl. 88 determinations in each of two flasks (interference effect, 1 p.p.m. of phosphate = 0.0002 to 0.0019 p.p.m. of silica at 810 mp). With 50 p.p.m. of phosphate, seeding occurred after one determination (initial effect, 1 p.p.m. of phosphate = 0.0008 p.p.m. of silica) and with 100 p.p.m. of phosphate a heavy precipitate formed during the first determination in a flask. When the concentration of molybdate in the final solution was 0-0125 M, 50 p.p.m. of phos- phate did not cause a seeding effect after 5 determinations (interference effect, 1 p.p.m.of phosphate = 0.0000 to 0-0002 p.p.m. of silica at 810 mp) and 100 p.p.m. of phosphate did not cause a precipitate. At 670 mp the interference effects were 2-5 times greater than those given above. It appears, therefore, that for samples containing much phosphate, it is preferable to use 0.0125 M Mo0,2- in the final solution rather than 0.05 M. PREVENTION OF THE R-EDUCTION OF MOLYBDATE When the acidity of ammonium molybdate solutions is increased to 2-5 N, the molybdate apparently changes to a form that is not reduced by stannous tin to a blue compound; this change might be expected to require a finite time. The effect, on the reagent blank value, of the time between adding acid and reducing agent was therefore determined. Experi- mental conditions were the same as the standard conditions given in a previous section for a-molybdosilicic acid.When the initial volumes of solution were 30, 50 and 60 ml, no significant effect, compared with the standard deviation for such determinations (0.0007 optical-density units), was found for times between 15 and 60 seconds. The mean optical density of the reagent blank solutions was 0-003. EFFECT OF TEMPERATURE ON THE ABSORPTION OF LIGHT BY THE REDUCED MOLYBDOSILICIC ACIDS The magnitude of this effect was estimated by placing warm solutions in the spectro- photometer, measuring first the temperature with a mercury thermometer and then the optical density. The optical densities of solutions of wmolybdosjlicic acid reduced by stannous tin increased by about 0.15 per cent.per "C decrease in temperature. The corresponding figure for /I-molybdosilicic acid reduced by l-amino-2-naphthol-4-sulphonic acid was about 0.25 per cent. per "C. DISCUSSION OF THE WORK This was repeated as the solution cooled. The lowest concentration of silicon of interest in samples was about 0.002 p.p.m., as silica. The limit of detection of a method will be determined by the reproducibility of the reagent blank value, and, if the spectrophotometric sensitivity of the absorbing substance is low, this limit may, in practice, be determined by the reproducibility of the spectrophotometer. As the unreduced molybdosilicic acids have much lower spectrophotometric sensitivities4 than the reduced forms, except in the ultraviolet part of the spectrum in which the reagent blank solutions also absorb strongly, only the use of the reduced acids was further considered.The work described above shows that ii method based on reduced a-molybdosilicic acid may be expected to give precise results, because: (1) the acid can easily be produced in constant amount, (2) it is little affected by the high acidity a t which reduction is carried out, (3) the reduced product is stable for at least an hour and (4) there appear t o be no particularly critical factors to be controlled. An advantage of such a method is the reported insensitivity to high concentrations of salts. Its chief disadvantages are: (a) the heating operation, which adds to the necessary time and number of manipulations, (b) measurements must be completed within an hour after adding the reducing agent and (c) the sensitivity to phosphate and chloride interference.Those factors that might make reduced p-molybdosilicic acid unsuitable for use in a precise method have been shown to be less important than expected from the literature. Thus-in the absence of large concentrations of salts at least-/&molybdosilicic acid can easily be prepared in constant amount, and its conversion to the a-form seems insufficiently fast to be a source of large errors. Care is needed in choosing a reagent to destroy molybdophos- phoric acid. When sufficient oxalic or tartaric acid is used, the rate of destruction of molybdo- phosphoric acid and the residual absorption caused by phosphate were similar for both reagents. However, the rate of destruction of molybdosilicic acid is much faster with oxalic than with tartaric acid, and the latter is therefore preferred.As a reducing agent, 1-amino-2-naphthol- 4-sulphonie acid is considered superior to stannous tin, because of the greater stability of theFebruary, 19631 DETERMINATION OF SILICON IN WATER. PART I 99 reduced product ; the slightly greater sensitivity given by stannous tin seems insufficient to warrant its use. When suitable conditions are used, a method based on reduced /3-molybdo- silicic acid should give reproducible results, can be made insensitive to phosphate inter- ference and does not require many operations. The major disadvantage is the reported4 interference of large amounts of neutral salts. The spectrophotometric sensitivities8 of the reduced a- and P-molybdosilicic acids are 0.0030 pg of silica per sq. cm at 742 mp and 0.0026 pg of silica per sq. cm at 810 mp, respec- tively. When a spectrophotometer is not available and a filter giving a peak at about 670 mp (e.g., Ilford No. 608) is used, the sensitivities change to 0.0039 pg of silica per sq. cm and 0.0069 pg of silica per sq. cm for the ct- and p-forms, respectively. CONCLUSIONS Both reduced a- and /I-molybdosilicic acids should be suitable as the bases of precise methods for determining “reactive” silicon in water. Many of the potential sources of error have been evaluated and can be minimised by using appropriate conditions as indicated in this paper. Such methods should be further tested to determine the attainable precision, sensitivity to variations in conditions and liability to interference. These further tests for two methods involving the use of the a-3 and p-forms2 will be described in later papers of this series. We thank Mr. H. C. Bones for pointing out the effect of acidity on the blue colour of acidified molybdate solution. This paper is published by permission of the Central Electricity Generating Board. REFERENCES 1. 2. 3. 4. 5 . Ingamells, C. D., Chenzist Analyst, 1956, 45, 10. 6. Andersson, L. H., Acta Chem. Scand., 1958, 12, 495. 7. 8. Alexander, G. B., J . Amer. Chem. SOC., 1953, 75, 5655. Morrison, I. R., and Wilson, A. L., Analyst, 1963, 88, 100. - I_ , Ibid., in the press. Strickland, J. D. H., J . Amer. Chem. SOC., 1952, 74, 862, 868 and 872. Milton, R. F., J . Appl. Chem., 1951, 1, supplement 2, 126. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience Received October 9th, 1962 Publishers Inc., New York, 1959.

ISSN:0003-2654    

DOI:10.1039/AN9638800088

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

100 MORRISON AND WILSON : THE ABSORPTIOMETRIC [~4?dySt, VOl. 88 The Absorptiometric Determination of Silicon in Water Part 11. Method for Determining “Reactive” Silicon in Power-station Waters BY I. R. MORRISON AND A. L. WILSON (Central Ebctricity Reseavch Laboratories, Cleeve Road, Leatherhead, Suvvey) A method, together with a modilication for obtaining high sensitivity, is described for determining the “reactive” silicon content of water; it is based on the absorptiometric measurement of solutions of reduced 8-molybdosilicic acid. The within-batch coefficient of variation of the optical-density dif- ference, optical density of sample less optical density of reagent blank solution, varied from 7 per cent. to 0-2 per cent. for concentrations of 0.01 and 0.5 p.p.m. of silica, respectively.The limit of detection was about 0.001 p.p.m. of silica. The effects caused by several other substances have been determined. The analysis time is about 14 hours for a batch of 10 samples in duplicate. A METHOD was required for determining “reactive”* silicon in water in the range 0-002 to 75 p.p.m. of silica. The types of samples to be analysed were (a) de-ionised water, feed water, make-up water and steam, which usually contain less than 1 p.p.m. of silica, and ( b ) boiler water, containing phosphate, alkali and between 0-2 and 75 p.p.m. of silica. From the work described in Part I of this series1 it appeared that a suitable method could be based on the absorptiometric measurement of p-molybdosilicic acid, after reduction by l-amino-2-naphthol-4-sulphonic acid.Desirable experimental conditions were therefore chosen, and this paper describes the procedure together with the tests performed to confirm its suitability for use with de-ionised water, feed-water, make-up water and steam. It is likely that the method will also be suitable for boiler-water, but it has not yet been tested sufficiently for a definite recommendation to be made; it is hoped that this part of the work will be reported later. M ETH o D REAGENTS- Silicon in the tartaric acid solution and in reagents added after it has no affect on the final absorption. Other reagents, including water, should be stored in polythene vessels. Water-Water used for preparing reagents and for diluting samples should preferably contain less than about 0.005 p.p.m. of silica, as determined by the high-sensitivity modifica- tion of the method; if the silicon concentration is higher, reagent blank values for the high- sensitivity modification may sometimes cause appreciable bias because of silicon in the 5 ml of water used in this blank solution.Distilled water from an all-metal still, or water that has been passed through a mixed-bed de-ionisation unit, was found to be suitable. AcidiJied molybdate solution-Dissolve 89 g of ammonium molybdate, (NH4),MO,0,,.4H,O, in about 800 ml of water at room temperature. Dilute 62 ml of 98 per cent. sulphuric acid to about 100 ml by adding it cautiously to water, with stirring, and allow to cool. Add the acid to the molybdate solution, and dilute to 1 litre. The reagent may be kept for several months.Any slight blue colour developing in this reagent may be ignored unless the reagent blank values become unduly large. Tartaric acid solution, 28 per cent. w/v--The reagent may be kept for at least 3 months. Redwing agent solution-Dissolve 2.4 g of sodium sulphite, Na2S0,.7H,O, and 0.2 g of Add This All reagents should be of analytical grade unless otherwise stated. l-amino-2-naphthol-4-sulphonic acid (purest grade available) in about 70 ml of water. 14g of potassium metabisulphite, shake well until dissolved, and dilute to 100ml. reagent should be freshly prepared each week. * “Reactive” silicon is defined in this paper as those forms of silicon-mainly monomeric and dimeric silicic acid-that react with ammonium molybclate in 10 minutes under the conditions of the method given in this paper.February, 19631 DETERMINATION OF SILICON IX WATER.PART 11 101 Standard soZutions of silica-Fuse 1.000 g of pure dry silica with 5 g of anhydrous sodium carbonate in a platinum crucible at red heat. When cool, dissolve in water, and dilute to exactly 1 litre. Prepare by dilution a solution containing 10 p.p.m. of silica. The solutions containing 1000 p.p.m. were stable, within t0.5 per cent., for at least 2 years in polythene bottles, and the solutions containing 10 p.p.m. for at least 3 months. The most suitable silica for this purpose is probably transparent Spectrosil rod (Thermal Syndicate Ltd.), which has metallic impurities of less than 1 p.p.m. and is not appreciably hygroscopic. This solution contains 1000 p.p.m.of silica. PROCEDURE- By pipette, place in a 100-ml calibrated flask sufficient sample to give a suitable optical density, and dilute to 80 5 ml with water; in a similar flask for the reagent blank solution place the same amount of water as was used for diluting the sample. Add 2 5 m l of acidified molybdate solution to the contents of both flasks, and mix immediately. Ten minutes $.3 minutes later add 2.5 ml of tartaric acid solution, and mix immediately. Five minutes +1 minute after the tartaric acid has been added, add 2.0ml of reducing agent solution, mix, and dilute to exactly 100ml. Between 15 minutes and 24 hours later (1 hour if the samples contain phosphate), measure the optical densities in a suitable cuvette (preferably 4-cm) against distilled water.Subtract the optical density of the reagent blank solution from that of the sample, and read off the concentration of silicon from the calibration curve. The optical density is best measured at 810 mp (the peak is sharp and should be checked on each instrument); with 4-cm cuvettes, 0.67 p.p.m. of silica in the final solution has an optical density of about 1.0. The optical density can also be measured at about 670mp (e.g., with Ilford No. 608 filters), but the reading is only 40 per cent. of that at 810 mp; 10-cm cuvettes can then be used to increase the sensitivity. PreParation of calibration curve-Treat as samples suitable aliquots of standard silicon solution containing 10 p.p.m. of silica; the blank solution should contain 80 ml of water from the same batch as was used for preparing and diluting the standard silicon solutions.From the results obtained prepare a calibration graph. Compensation for turbidity and/or colour iw a sample-A blank solution often suitable for compensating for the turbidity and/or colour of samples can be prepared by adding to a normally prepared blank solution, after the tartaric acid has been added, the same volume of sample that was used in the determinations. A check should then also be made on whether the turbidity and/or colour is differently affected by the sample and blank procedures. MODIFICATION FOR HIGHEST SENSITIVITY- By pipette, place a 100-ml sample in a 4-ounce polythene bottle; place 5 ml of water in a similar bottle for the reagent blank solution. With the same timing as in the basic procedure, add the same volumes of reagents to the contents of the bottles, except that, after adding tartaric acid to the reagent blank solution, add 95 ml of water to this bottle.Do not dilute the solutions any further before measuring their optical densities, preferably at 810 mp in 10-cni cuvettes, against distilled water. Subtract the optical density of the reagent blank solution from that of the sample, and read off the concentration of silicon from a calibration curve. This is prepared by treating 100-ml aliquots of suitable silicon solutions exactly as for samples. For this curve, the blank determination should be made in exactly the same way as for samples, i.e., 100 ml of water (from the same batch as used for preparing and diluting the silicon solutions) should be taken initially.SOURCES OF ERROR- Reagent blank solutions prepared in borosilicate flasks contained only 0.0005 to 0.0015 p.p.m. more of silica than did those prepared in polythene bottles, provided (1) flasks were reserved for these analyses, (2) old solutions were left in them until just before the next deter- mination and (3) they were rinsed shortly before use. However, two flasks gave systematically high blank values, and flasks should therefore be checked before first being used for analyses. Just before use, pipettes should be rinsed with the solution to be dispensed.102 MORRISON AND WILSON : THE ABSORPTIOMETRIC [A?Z@&St, VOl. 88 If techniques are chosen so that the maximum error involved in each of (i) weighing chemicals, (ii) adjusting reagent solutions to volume and (iii) dispensing the required volume of reagent solution, is less than -+_ 5 per cent., errors from these sources should be extremely small (<1 per cent.). RESULTS VARIATION IN CONCENTRATION OF REAGENTS- The effect of relatively small variations in the concentrations of ammonium molybdate, sulphuric acid and tartaric acid was determined in a factorial experiment.Three concentra- tion levels were used for each reagent: standard, 20 per cent. less than standard and 20 per cent. more than standard. Other conditions were as described in the procedure for the high-sensitivity modification. Single determinations were carried out for each treatment, and the design and analysis of the experiment were the same as those described by Davies.2 The experiment was carried out at a silicon concentration of 0.01 p.p.m., as silica, in the sample and repeated at a concentration of 0.2 p.p.m., as silica.Analysis of the results led to the conclusions: (1) that at 0.01 p,p.m. of silica, variation in the reagents had no significant effect, compared with a standard deviation for random errors of k0.00075 p.p.m. of silica; (2) that at 0.2 p.p.m. of silica, variation in the concen- tration of sulphuric acid had no significant effect, but a k20 per cent. change in the con- centration of either ammonium molybdate or tartaric acid gave a -fO.S per cent. change in optical density; the maximum error from this source was, therefore, &1-2 per cent. The effect of a t-20 per cent.variation in the concentration of the reducing agent was known, from the work reported in Part I of this series, to be no more than k0.2 per cent. LINEARITY OF CALIBRATION CURVE- For the high-sensitivity modification, the calibration curve was linear within & 2 per cent. up to at least 5 p.p.m. of silica in the sample; the slope with 4-cm cuvettes at 810 mp was 0.72 p.p.m. of silica per optical-density unit. REPKODUCIBILITY- To measure the reproducibility of the optical-density reading used for calculating the analytical results ( i e . , sample minus reagent blank optical densities), the procedure described below was used. In one batch, duplicate samples (Sl, S,) of sodium silicate solution and duplicate reagent blank solutions (Bl, B,) were treated and measured as in the high-sensitivity modification of the method, and the differences, S, - B, and S, - R,, were calculated (4-cm cuvettes were used).The procedure was repeated on each of another 9 days, and the differences were analysed by the usual statistical techniques to obtain estimates of the within- and between-batch standard deviations. Table I shows the results expressed as p.p.m. of silica. The within-batch standard deviation of the reagent blank solution is not directly comparable with the other standard deviations, as it is obtained from single figures and not from the difference, sample less blank. Table I also shows the results for reagent blanks analysed over a longer time period. TABLE T REPRODUCIBILITY OF ANALYTICAL RESULTS Silicon in final solution, Period of test p.p.m. of silica 0.47 3.9.60 to 6.10.60 0.19 3.9.60 t o 6.10.60 0.0 1 3.9.60 to 6.10.60 Reagent blank* 3.9.60 t o 6.10.60 Reagent blank* 12.5.60 to 12.1.62 JVithin batch Between batch (---------h---.--, r -7 Standard Degrees Standard Degrees deviation, of deviation, of p.p.m.of silica freedom p.p.m. of silica frecdom 0.001 1 10 0-0005 9 0*0005 10 0-0007 9 0. 000’7 10 n.s. 9 0-0002 10 - 0,0004t 19 - A * Reagent blank figures only and not the difference, sample niiizzts blank. t One extremely bad result excluded. n.s. = Xot significant; variations all assigned to within-batch errors.February, 19631 DETEKMINATION OF SILICON IN WATER. PART 11 103 EFFECT OF OTHER SUBSTANCES- Table I1 shows the effects at 810mp of some substances likely to be found in power- stat ion waters .TABLE I1 EFFECT OF OTHER SUBSTANCES Substance Sodium nitrate . . Magnesium chloride Calcium sulphate Orthophosphate. . Hydrazine sulphate Cyclohexylamine Morpholine . . Octadecylamine Cupric copper . . 17erric iron . . .. . . . . . . . . . . . . . . . . . . Fcrrous iron in the pre- sence of phosphate . . Ferric hydroxide . . 1;luorcscein . Silicon in final solution (as silica), p.p.m. 0, 0.5 0, 0.5 0, 0.01, 0.2 0, 0.01, 0.2 0, 0.01, 0.2 0, 0.01, 0.2 0, 0.01, 0.2 0, 0.01, 0.2 0. 0.01. 0.2 { x.5 Concentration of substance in final solution a t which effect was equivalent to 0.001 p.p.m. of silica, p.p.m. > 1000 > 1000 300 > 1000 > 10 > 10 > 10 10 to 25 0.25 1.7 at 810 and 670 mp > 20 Effect a t higher concentratioiis of substance, p.p.m.of silica per p. p. m . of substance Not determined Not determined Not determined +o*oooo to 0.0002 Not determined Not determined hTot determined -+ 0.0036 + 0.00056 Not determined about + 0.02 about + 0.04 + 0.025 + 0.025 0.36 for 0.5 p.p.m. of PO,3- 0.27 for 1.0 p.p.m. of PO4+ 0.07 for 2.0 p.p.m. of POg3- 0.07 for 5.0 p.p.m. of PO,3- {! 0 0 6 Depends on whether or not the ferric hydroxide is dissolved by the reagents 0, 0.5 > 10 Not determined CORRECTNESS OF REAGENT BLANK VALUE- The correctness of the reagent blank value depends on the assumption that silicon added after the tartaric acid reagent does not form molybdosilicic acid. This assumption was checked by adding 5 ml of the silicon solution containing 10 p.p.m. of silica both with and just after the tartaric acid reagent. In both tests the change in measured silicon content, compared with the normal reagent blank value, was less than 0.001 p.p.m.of silica. APPLICATIONS TO WATEK SAMPLES- The method was used for analysing de-ionised water from many power-stations. The within-batch standard deviation of the results was 0.0012 p.p.m. of silica (duplicate deter- minations made on 29 samples) for silicon concentrations ranging from 0-004 to 0.234 p.p.m. of silica. Tn recovery tests on two samples of power-station de-ionised water and one sample of feed-water, silicon (0.050 p.p.m. of silica) was added to samples and then determined; the maximum discrepancy detected between determined and added amounts was 0-0025 p.p.m. of silica. I>ISC~;SSIOS OF THE METHOI) YIiECISION AND LIMIT OF DETECTIOK- Table I shows that most of the random errors occurred within a batch and that these within-batch errors were equivalent to a coefficient of variation of about +0-2 per cent.at the 0.5 p.p.m. of silica level. This error is low for an absorptiometric method and probably consists largely of spectrophotometric errors. Most of the important chemical factors were, therefore, probably controlled. The limit of detection, defined as 1.65v'5 times the within-batch standard deviation of the reagent blank,3 was 0.0008 p.p.m. of silica (95 per cent. confidence level). As the major source of error seems to be the spectrophotometric measurements, this limit may possibly be improved by using 10-em cuvettes.104 MORKISON AND WILSOY [Analyst, Vol.88 APPLICATION TO POWER-STATION WATERS- Tests at power-stations have indicated that their waters had appreciable absorptions, presumably caused by turbidity and/or coloured materials. Of these two, turbidity seems more important, as we know of no coloured materials (having sufficiently large optical densi- ties) likely to occur in the samples of interest; colour is, therefore, not further considered. As silicon concentrations are often 0.02 p.p.m. of silica or less, optical-density errors as small as 0.003 represent an appreciable bias (about 10 per cent.). The bias caused by turbidity should therefore either be estimated or eliminated by removing the offending materials; a technique for estimating this bias is described under “Procedure,” but may not always be valid.The turbidity could probably be reduced to insignificance by filtering the sample through a Millipore filter; tests would be required to determine a suitable pore-size. Of the substances tested for interference effects only octadecylamine seems likely to be present in some power-station waters in sufficient amounts to cause interference. SPEED OF ANALYSIS- The time of analysis is about 19 hours for a batch of ten samples in duplicate. EXTENSION TO NATURALLY OCCURRING WATERS- The method would probably be suitable for analysing naturally occurring waters, but interference effects would have to be checked for each application. For some samples, e,g,, sea-water, the effect of large concentrations of salts may be important. This effect has not been studied in detail by us, but approximate estimates indicate that the optical density obtained for a given concentration of silicon is decreased by 1 per cent. for salt concentrations of about 3000 to 4000 p.p.m. in the final solution. In analysing such samples, this difficulty can be overcome (a) by using a small aliquot of the sample in order to reduce the salt content of the final solution and (b) by preparing the calibration curve with solutions containing the same concentrations of salts as those in the samples; the precision will probably be reduced in this instance. This paper is published by permission of the Central Electricity Generating Board. REFERENCES 1. 2. 3. Morrison, I. R., and Wilson, A. L., AnaZ,yst, 1963, 88, 88. Davies, 0. L., “The Design and Analysis of Industrial Experiments,” Oliver and Boyd Ltd., TVilson, A. L., Analyst, 1961, 86, 72. Received Octobev 9th, 1962 London, 1954, p. 431.

ISSN:0003-2654    

DOI:10.1039/AN9638800100

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

February, 19631 MOHANRAO AND FOLSOM 105 Gamma-ray Spectrometric Determination of Low Concentrations of Radioactive Caesium in Sea Water by a Nickel Ferrocyanide Method BY G. J. MOHAKRAO (W. ,W. Ir'eck E'ngineeving Lccbovnlories, California Institute of Technology, Pasadena, Califovnia) AND T. R. FOLSOM (Scvipps Iiistitution of Ocennogvaphy, La Jolla, Califonaia) A simple, fairly rapid and reliable method is presented for determining radioactive caesium in sea water, based on a known method of concentration. liadioactive caesium, together with niicrogram amounts of stable caesiuni naturally present in sea water, is co-precipitated with nickel ferrocyanide from a 50- to 200-litre sample. After the precipitate has settled and the supernatant liquid has been syphoned off, the nickel ferrocyanide is trans- ferred to a pint container and counted in a gamma-ray spectrometer equipped with a $inch x 4-inch NaI (Tl) crystal and 256-channel pulse-height analyser.Studies with a caesium-134 tracer helped to determine the most favourable conditions of pH, temperature and concentration of reagent. The procedure has been used aboard ship and in the laboratory since 1960 for oceanographic survey of radioactive caesium in sea water in concentrations ranging between 0-03 and 0.4 pC per litre. THE study of the concentration of radioactive caesium in sea water is becoming a subject of great interest. For example, it has been reported recently1t2 that in human rib bones caesium-137 may sometimes contribute more to the total dose of ionising radiation than in commonly considered to be contributed by strontium-90 in these tissues.-4 great deal of information about the distribution of fallout in man's environment is coining from the studies of the distribution of fission product concentrations in the oceans. Caesium-137 is convenient for this purpose because of its long half-life (27 years) and because of the 0.66-MeV gamma ray it emits through its daughter barium-137. A number of procedures for determining the concentration of caesium in fresh water, fission-product wastes, biological materials and in sea water have been studied. They consist of ion-exchange technique^,^ y 4 co-crystallisation as alum5 and precipitation and co-precipitation as chloroplatinate, perchlorate, cobaltinitrite,6 dipicrylaminate,' ammonium molybdophos- phate,3,839 molybdosi1icate,l0 tungstophosphate,ll tetraphenylb~rate,~?~~$~~ cobaltinitrite and tungstosilicatel* and the ferrocyanides of copper, iron, cobalt and nickel.15 ,l6 9l7* Many investigators have recommended a combination of these techniques to improve the purity and to reduce the size of the caesium sample to be counted.This treatment is essential when the caesium is determined by counting the beta rays. Although several of these methods were reasonably satisfactory, they involved considerable time and delicate chemical processing, In a search for a quick and simple procedure especially suited for shipboard operations, a procedure, in which is used the co-precipitation of caesium by sodium potassium cobalti- nitrite and counting on a gamma-ray spectrometer, has been developed in these laboratories.l8 This procedure, unfortunately, had two difficulties; one was the requirement to keep the temperature below about 19" C and the other was to contend with the significant amount of potassium present, which added to the gamma-ray background.Hence the procedure described below, with nickel ferrocyanide as the concentrating agent (used previously at Kanford Atomic Works to scavenge caesium from fission wastes), was studied as to its usefulness for sea-water analysis. * \Ye acknowledge a personal communication from Mr. Sadakiyo Hori of the Japan Hydrographic Ofice, Tokyo (said to have originated from R. Higano of the same institution) proposing a procedure somewhat similar to that described in this paper, but at a time when preliminary tests were already being made here.106 MOHANRAO AND FOLSOM Y-RAY SPECTROMETRIC DETERMINATION [AndySt, VOl.88 EXPERIMENTAL A known amount of caesium-134 chloride solution was added to 200 litres of sea water, mixed well and distributed into eight 25-litre polythene containers. The influence of the four variables-reagent concentration, pH, temperature and amount of caesium carrier added- was studied. After adjusting the particular variable, solutions of nickel sulphate and potassium ferrocyanide were added, and the contents were stirred for 20 minutes; the reaction appeared to be complete in less than 10 minutes. In all tests the ferrocyanide concentration used was twice the stoicheiometric requirement of the nickel.The precipitate at first is a milky opalescence, gradually increasing in particle size and finally settling down as a distinct layer of amorphous light green sludge; the supernatant liquid is clear and colourless. After the precipitate had been allowed to settle for 24 hours, the supernatant liquid was syphoned off. The precipitate was collected in a commercially available cylindrical polythene container, 10 cm in diameter and 8 cm high. After the precipitate had been allowed to settle for another day, its height was measured and recorded for applying geometrical correction ; the pre- cipitate was then counted in a gamma-ray spectrometer, the container being placed on top of a sodium iodide crystal 5 inches in diameter and 4 inches thick.l* 8okyrl g a 84 5 0 4 8 12 ( C ) h' g o Temperature, OC 8 M - ( a ) h- m - ' 0 20 40 60 in u VI aJ U c .- .- Nickel present, p.p.m.01 I I I I I I I I I I I I 0 4 8 I2 PH ( d ) 0 I I I I I I I I I I I 0 40 80 I 20 Caesium carrier present, p.p.rn. Fig. 1. Influence of four variables on the recovery of caesium-137 (u) Influence of reagent concentration; temperature 24" C; pH 8.3 (b). Influence of pH: 0, 22" C, nickel concentration 6.8 p.p.m.; A, 19.5" C, nickel con- (c) Influence of temperature: 0, pH 8.0, nickel concentration 11.0 p.p.m.; a, pH 8.1, ( d ) Influence of caesium carrier; temperature 22" C; pH 8.4; nickel concentration 11.0 centration 11.0 p.p.m.; @, 20" C, nickel concentration 11.0 p.p.m. nickel concentration 6.8 p.p.m. p.p.m. RESULTS The influence of the four variables on the recovery of radioactive caesium in sea water is shown in Fig.1. The recovery of caesium increased with increasing concentrations of nickel up to 7 p.p.m., becoming essentially constant between 7 and 12 p.p.m. (see Fig. l(a)). From this it was concluded that a nickel concentration of 11 p.p.m. was adequate, and this concentration was used for most of the experiments. Results from three independent experiments are shown in Fig. 1 (b), which indicate that the caesium recovery decreases slightly with decreasing pH down to 2; recovery was unreliable and low below pH 2. This behaviour makes the procedure especially suitable for use with average sea water, which has a pH of 8-2. Fig. l(c), which shows results of two experiments, indicates that the recovery of caesium decreases slightly with increase in temperature up to 30" C and somewhat more significantly above 30" C.It can be seen from Fig. l(d), which shows a study of the effect of variations in the amount of carrier used, that the recovery of caesium-137 was not improved to any significant extent by the addition of caesium chloride carrier up to 20 p.p.m. The recovery was reduced with increase in the amount of carrier above 20 p.p.m.; apparently this wa,s be- cause insufficient nickel ferrocyanide precipitate was found.Februarj-, 19631 OF LOW CONCENTRATIONS OF RADIOACTIVE CAESIUM IN SEA WATER 107 The recovery and reproducibility of the method was studied by precipitating and counting 8 labelled replicate samples of fresh sea water under identical conditions.Three such cxperi- ments were conducted under slightly different conditions. The precipitates from one cxperi- ment were counted in a 3-inch x 3-inch sodium iodide well crystal; the other two were counted on a 5-inch x 4-inch solid crystal. Results from the three sets of experiments are shown in Table I and indicate that the average recovery was between 89 and 90 per cent. ; the coefficient of variation of measurements in each experiment lay between +3-1 and +5-6 per cent. of the average recovery. TABLE I 5:ECOVERY OF CAESIUM-137 FROM LABELLED REPLICATE SAMPLES OF SEA WATER lieplicate No. 1 2 3 4 5 6 7 8 :Iverage . . . . Standard deviation Coefficient of variation .. .. .. Recovery of caesium-137 counted with- h r 7 &inch x 4-inch flat crystal well crystal c, * 04 86.0 S4.0 91.1 97.5 84.0 88.0 82.3 88.6 94.5 84.4 91.1 90.7 92.0 91.1 87.4 93.0 88.6 86.3 96.4 88.6 86.1 89.6 95.9 89.0 90.0 89.0 89.1 5.0 3.9 2.8 5.6 4.4 3-1 F- B, * % % * Solution conditions were- A: 6.8 p.p.m.of Ni2+; pH 8-2; temperature 19" C. B: 1 1 p.p.m. of Ni2+; pH 8.1; temperature 17' C. C: 1 1 p.p.m. of Ni2+; pH 8.2; temperature 22" C. TABLE I1 DETERMINATION OF CAESIUM-137 I N SEA WATER Sample size = 200 litres. Detector = 5" x 4" NaI crystal. Time of counting = 8 hours. Spectrometer = 256 channels (set a t 0 to 1.6 MeV) pH = 8.2 (natural). Temperature = 15" C. Fe(CN);f- = 40 p.p.m. (added) Reagent blank value = 0.015 pC per litre. Standard Replicate No. . . . . . . 1 2 3 4 6 Mean deviation Caesium-137 found, pC per litre 0.133 0.151 0.155 0-151 0.143 0.147 0.01 Coefficient of variation* = 6.8% * The coefficient of variation of about T per cent.might appear fortuitous. However, a graphical smoothing has been carried out on the spectral record so that information equivalent to nine times that in the peak channel was used. Because of this, the coefficient of variation of measurements of this type and concentration should be about 4 per cent. The experimental value is therefore not unreasonable. TEST WITH UNLABELLED SEA WATER- The procedure was applied to five 200-litre replicate samples of sea water collected a t Scripps pier. Table I1 shows that the sea water contained 0.147 2 0.01 pC of caesium-137 per litre. The reagent blank values were no more than about 0.015 pC per litre. Many samples have been collected from the Pacific Ocean since 1960 and analysed by this method; the results are reported e1~ewhere.l~ ,20 STIRRING REQUIREMENTS- Stirring requirements were determined as described below.Four labelled 10-litre samples were each stirred, after the reagent had been added, by an 1800 r.p.m. stirrer having a 3-cm x 3-cm blade for 5, 20, 60 and 60 minutes; the precipitates were then allowed to108 MOHANRAO AND FOLSOM [Analyst, Vol. 88 settle before re-stirring for a further 60 minutes. The recoveries of caesium were 90, 87, 84 and 93 per cent., respectively. From this it would appear that the previously used stirring time of 30 minutes has been adequate. Presumably equilibrium is reached rapidly because of the large area provided by the fineness of the solid suspension. CONCLUSIONS Determination of caesium in sea water by co-precipitation with nickel ferrocyanide has been found to be feasible.Neither adjustment of pH or temperature nor addition of carrier was found necessary. When counted for 8 hours on a 5-inch x 4-inch sodium iodide crystal, 200-litre samples gave results with a standard deviation of t 7 per cent. of the mean. The simplicity of the method makes it convenient for processing samples on board the ship so that precipitates (in pint containers) can be sent home in place of 55-gallon samples of un- processed sea water. Unlike the cobaltinitrite method previously used in this laboratory, the proposed method is useful even at tropical temperatures. The method, unfortunately, requires one day or more for the precipitate to settle in the 55-gallon drum, and two or more days in smaller containers.Studies now being under- taken at these laboratories indicate that the rate of settling can be remarkably increased by application of available coagulants, and practical methods will be reported separately. This work was financed by U.S.A.E.C. under contract KO. AT(11-1)-34, and used the equipment of U.S. ONR under contract No. Noni--233(34). 1 . 2. 3. 4. 0. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. -20. REFERENCES Yamagata, N., and Tamagata, T., United Kations Report No. A/AC. 82jG,lL. 396, New York City, N.Y., 1960. Anderson, R. W., and Gustafson, P. F., Science, 1960, 137, 668. Kahn, B., Smith, D. K., and Straub, C. P., Anal. Chem., 1957, 29, 1210. Caron, H. I,., and Sugihard, T.T., Ibid.. 1962, 34, 1082. Gresky, ,4. T., U.S. Atomic Energy Commission Report AECD-2999, OKXL-742, Oak Ridge, Yamagata, N., Japan Analyst, 1956, 5, 37. Yamagata, N., and Yamagata, T., Anaiyst, 1960, 85, 282. Miyake, Y . , Saruhashi, K., Katsuragi, Y., and Kanazawa, T., J . Radiafion liesearch, Japan, 1961, Yamagata, N., and Matsuda, S., Bull. Chew. SOC. Japan, 1959, 32, 497. Mizzan, E., PDB-128, (1954), Nuclear S c i . Abstvacts, 1955, 9, 879. Glass, G. H., Chemist Analyst, 1953, 42, 50. Handley, T. H., and Burros, C. L., Anal. Chem.., 1959, 31, 332. Schroeder, B. W., and Cherry, R. D., Nature, 1962, 194, 669. Schulz, W. W., and McKenzie, T. R., U.S Atomic Energy Commission Report TID-7517, Washing- De Jonghe, P., Sox. Nucleave, Belgium, 1959, 2, 325. Higano, R., and Shiozaki, M., The Contrzbutions fvom the -?lavine Research Laboratory Hydrographic Folsom, T. K., Mohanrao, G. J., and Winchell, P., Nature, 1960, 187, 480. Folsom, T. R., and Mohanrao, G. J., J . Radiation Research, Japan, 1962, 3-1, 1. _I___ , “Distribution of Cesium-137 in the Pacific and Indian Oceans,” paper presented at Tenn., 1950, p. 11. , , Bull. Chenz. SOC. Japan, 1958, 31, 1063. -- 2-1, 25. ton, D.C., 1956 (Part I), 290. Ofice of Japan, 1960, 1, 137. the -American Geophysical Union Symposium on Radioactivity in the Ocean, April, 1962. Received October l s t , 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800105

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

February, 19631 WILSON AND SERGEANT 109 The Colorimetric Determination of Aluminium in Minerals by Pyrocatechol Violet BY A. D. WILSON* AND G. A. SERGEANT (Department of Scienti$c and Industrial Research, Laboratory of the Government Chemist, Geological Survey and Museum, Exhibition. Road, London, S. W . 7 ) A procedure is described for the colorimetric determination of aluminium in small mineral samples with pyrocatechol violet. Conditions for the forma- tion of this complex and the effects of other cations and anions have been examined. Inter- ference from iron is considerably reduced by the formation of the ferrous - o-phenanthroline complex. The optimum pH in acetate buffer solution is 6.1 to 6.2. T H E determination of microgram amounts of aluminium, as for example in the microanalysis of minerals, is made difficult by the lack of a satisfactory colorimetric reagent for aluminium.The available reagents, such as alizarin, aluminon and Eriochrome cyanine R, are generally characterised by low specificity and a tendency to give high blank values because of their high optical densities at the optimum wavelength. Probably the best of the existing methods is the formation of the oxinate and extraction of this with chloroform to give a yellow-coloured solution of the complex. This method is applicable to many of the common silicate rocks in which there are likely to be few interferences, and where interference due to iron in particular can be almost eliminated by complex formation with o-phenanthroline. The reactions of pyrocatechol violet with a number of metals in solution have been examined by Suk and Malat,l who found that the stability of the blue and yellow complexes formed increased with the ionic charge of the metal ion complexed.In slightly acid solutions, bivalent ions did not form complexes, so that it might be expected that, in addition to the alkaline earths, such metals as manganese, beryllium and uranium, in the uranyl form, would not cause interference. The colour reaction with aluminium is distinctive, as the complex is blue in strong contrast to the yellow colour of pyrocatechol violet in slightly acid solution. A method for the colorimetric determination of aluminium by pyrocatechol violet has recently been described by Anton,2 and it was decided to investigate the reaction further with particular reference to the micro-determination of aluminium in minerals also containing manganese, uranium, beryllium and the rare earths.I I .ooo - x VI C -W m 2 0.500- B 0. I I I I 4.5 5 .O 5.5 6.0 6.5 7 0 Fig. 1. Variation in optical density of the aluminium - pyrocatechol violet complex with pH. Optical densities measured in 1-cin cells against water after 2 hours: curve A, solutions containing 80 pg of aluminium oxide, 1 ml of 0.2 per cent. pyrocatechol violet solution, 50 ml of 10 per cent. ammonium acetate buffer solution and 2 mI of 10 per cent. hydroxylamine hydrochloride solution per 100 ml; curve B, reagents only * Present address : Department of Scientific and Industrial Research, Laboratory of the Government Chemist, Clement’s Inn Passage, Strand, London, W.C.l.110 UTLSON AND SERGEANT : COLORIMETRIC DETERMINATIOS [AnazySt, V O l .88 EXPERIMENTAL pH AND C\ MPLEX FORMATION- Anton formed the aluminium - pyrocatecliol violet complex at pH 5.0 with a pyridine acetate buffer. We have found that the maximum colour is developed at a higher pH, as shown in Fig. 1, from which it can also be seen that above pH 5.9 the optical density rises to a fairly constant level. The optical density of the reagent also increases with pH, at first slowly and then, above pH 6.4, much more rapidly as the colour turns blue. The optimum pH for colorimetry of the aluminium complex lies between 5-9 and 6-4, and a satisfactory working range is pH 6.1 to 6.2 with ammonium acetate - acetic acid buffer. I t becomes necessary in this pH region, which is difficult to stabilise, to adjust the pH to nearly the required value before adding the buffer solution.WAVELENGTH AND OPTICAL DENSITY- The variation in optical density with wavelength of both pyrocatechol violet and the aluminium complex was studied at pH 6-1 to 6.2; the results are shown in Fig. 2, from which it can be seen that there is a maximum for the complex at 580 mp, and this wavelength was chosen for subsequent measurements. Wavelength, rnp Fig. 2. Variation in optical density with wavelength. Optical densities measured in 1-cm cells against water after 2 hours; curve A, solutions containing 80pg of aluminium oxide, l m l of 0.2 per cent. pyrocatechol violet solution, 50ml of 10 per cent. ammonium acetate buffer solution and 2ml of 10 per cent.hydroxylamine hydrochloride in a total volume of 100 ml (pII 8.2) ; curve 13, reagents only TIME OF DEVELOPhlENT OF COLOUR- Under the conditions chosen for the reaction, maxiinurn colour development was found to take place in approximately 1 hour, after which the colour intensity remained virtuall\- constant for several hours. BEER’S LAW AND REAGENT COXCENTRATION- For a range of concentrations up to SO pug of aluminium oxide per 100 ml, the colour intensity follows Beer’s law provided that a sufficient excess of reagent is present (2 ml of 0.15 per cent. w/v aqueous solution). INTEKFEKENCES AND APPLICATIOXS- The determination of aluminium in calcium-bearing minerals presents no difficulties, but, when elements such as iron, beryllium, chromium, etc.(see Table I) are present, two factors must be considered; (a) the precision required of the method and (b) the ratio of aluminium to the interfering metal. Then, if necessa.ry, a correction can be applied or the interferingFebruary, 19631 OF ALUMINIUM I N MINERALS BY PYROCATECHOL VIOLET 111 elements can be removed. For example, a number of strongly interfering cations, such as thorium, zirconium and vanadium but not chromium, are simply removed, together with iron and titanium, by an acid cupferron extraction. In the presence of hydroxylamine and o-phenanthroline, interference by iron is greatly diminished by the formation of the ferrous - o-phenanthroline complex, which absorbs only slightly at 580 mp. A routine addition of 5 ml of 0-15 per cent.w/v aqueous o-phenanthroline is sufficient when the ratio of ferric oxide to alumina does not exceed 4 to 1. The method could be satisfactorily applied in the analysis of an average silicate rock in which the only element likely to be present in sufficient amount to interfere would be titanium, for which a correction could be applied. Both phosphate and fluoride might be expected to interfere as these anions tend to form stable complexes with aluminium. The effect of fluoride is especially important, as fluoride is likely to be present after the decomposition of silicates by mixed hydrofluoric and sulphuric or perchloric acids. The results in Table I show that 50 pg of fluoride per 80 pg of alumina can be tolerated; this is much more than would normally remain as residual fluoride after the usual mixed acids treatment.Phosphate occurs in small concentrations in many silicate rocks and minerals and may be a major constituent, especially of calcium and uranium minerals. Fifty micrograms of phosphate (as P,O,) per 80 pg of alumina were found not to interfere significantly. Larger amounts of phosphate tend to suppress the colour, and in the analysis of phosphatic material it would be necessary to add an equivalent amount of phosphate to the standards, a modification that can also be applied to compensate for other In t edering radicals. Element added .Aluminium Iron . . Uranium . . Heryllium . . Manganese . . Chromium . . Vanadium . . Titanium . . Calcium . . Magnesium Lan thanu in Cerium . . Fluoride . , .4luminium Fluoride .. i\luminiurn Phosphorus .Aluminium Phosphorus .I\ 1 u mi ni 11 in . . . . . . . . . . . . . . . . . . . . . . . . . . 1 :: { . . . . -9 . . ”> TABLE I INTERFERENCE OF VARIOUS IONS Ionic species Concentration, Optical density added ~ 1 3 + l:e3+ uo22+ Re2f Mn2+ Cr0,2 - v0,- Mg2C Ti“+ Cat+ La3+ Ce3+ 1; - F- .w+ . u 3 + 1”0,3- .\P+ . u 3 + I-’O 3 - pg per 100 ml 80 as A1,0, 1000 as Fe,03 50 as Fe,O, 1000 as UO, 1000 as Be0 50 as Be0 1000 as MnO 50 as MnO 50 as CrzO, 50 as V,O, 50 as TiO, 5000 as CaO 5000 as MgO 5000 as T,a,O, 5000 as Ce,O, 200 as F 80 as X1,03 50 as F 80 as .41z03 200 as P205 80 as A1,03 50 as P,O, 80 as N,03 1.060 0-086 0.004 0.014 0.078 0.004 0.034 0.002 0.109 0.046 0.188 0.002 0.002 0.080 0.050 1.022 1.080 1.035 1-059 In tcrference, % - t 8 .1 + 0.4 -t 1.3 + 7.4 + 0.4 + 3.2 -1 0.2 +- 10.3 t 4.3 + 17.8 + 0.2 + 0.2 + 7.5 -+ 4.7 - 3.6 0.0 - 2.5 - 0.1 METHOD REAGENTS- Pyrocafechol violet solutioja, 0-15 per cent. wlv, aqueous. Hydvoxylamiize hydrochloride solutiogz, 10 per cent. w/v, aqateozts. o-Phetzanthroline solution, 0-15 per ce9it. w l v , aqzceous. Bzifer solution-Prepare a 10 per cent. w/v aqueous solution of ammonium acetate, and adjust the pH to 6-2 with acetic acid (approximately 0-4 per cent. w/v of acetic acid is required). Cztp fewon solutioiz, 1 per cent. in chloro form-Prepare by acidifying an aqueous solution of cupferron and extracting into chloroform.112 WILSON AND SERGEANT jA.tzalyst, Vol. 88 PROCEDURE- To approximately 20 ml of slightly acid test solution containing not more than 40 pg of aluminium, add 2 ml each of hydroxylamine hydrochloride solution, o-phenantliroline solution and pyrocatechol violet solution and then 5ml of buffer solution.Mix well, and adjust to pH 6.1 to 6.2 by careful additions of dilute ammonia solution and dilute acetic acid, care being taken not to allow the solution at any time to become distinctly alkaline. Transfer the solution to a 100-ml calibrated flask, and add a further 50 ml of buffer solution as rinsings. Dilute to the mark, mix the solution, set aside for 2 hours, and measure the optical density against distilled water in l-cm cells at 580 mp. Apply a correction for the reagent blank value. EXTRACTION PROCEDURE- When necessary, precede the colour development by a cupferron extraction. Add 2 ml of 12 N hydrochloric acid to about 20 ml of solution contained in a small Extract interfering elements with successive 5-ml portions of a 1 per Wash the solution twice with chloroform and Continue with the procedure described, omitting the addition separating funnel.cent. solution of cupferron in chloroform. once with light petroleum. of o-phenant hroline. TABLE I1 COMPA4RISON OF RESULTS Solutions were prepared from 1 -000-g samples of rock powder ; portions containing between 300 and 500 pg were used in the determinations Sample laboratory NO. 1543 1771 1821 1822 1823 1836 1816* Interfering elements Total ferric oxide, Titanium dioxide, 11.2 1.1 17.1 2.0 19.6 2.2 15.9 1.9 11.3 1-3 12.7 1.3 A I 7 0’ Yo /0 1.3 0-06 Alumina found by- r gravimetric method, 15-0 12.2 21.3 23.8 19.6 15.1 18.2 % 3 proposed method, 15.1, 14.9 12.8, 12.9 21.4, 21.8 23.8, 24.8 19-9, 19.3 154, 15.0 18.1, 18.6, 18.0, 17.7 Yo * Re-ground sample. RE s ULTS The proposed method is intended for the rapid evaluation of aluminium in milligram samples of minerals. However, in order to test the procedure, the method was applied to a selection of previously analysed silicate rocks from the Geological Survey collection, Since rock specimens, in contrast to mineral specimens, are not homogeneous, large amounts were taken for the purpose of this test to avoid sampling errors; the necessary concentrations for the procedure were obtained by suitable dilution. Corrections were applied for iron and titanium. The results are shown in Table 11; the reproducibility is only moderate but is sufficiently good for the intended applications. In obtaining these results, no analytical separations, other than the removal of silica, were made. This paper is published by permission of the Government Chemist and the Director of the Geological Survey, Department of Scientific and Industrial Research. REFERENCES 1. 2. Suk, V., and Makit, M., Chemist Analyst, 1956, 45, 30. Anton, A., Anal. Chem., 1960, 32, 725. Received April 171k, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800109

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

Febrriary, 19631 ASH H K O 0 K 113 The Determination of Scandium in Uranium Compounds by Solvent Extraction with 2-Thenoyltrif luoroace tone BY ALLAS W. ASHBROOK (Iirwarch &- Dezrelopimnt Dzvision, Eldomdo AVlznzng & Refining Limited, Ottawa, Ontarao, Canada) A method is described for determining scandium in uranium compounds. In the proposed procedure scandium is separated from interfering elements by extraction into a 0-2 M solution of 2-thenoyltrifluoroacetone in xylene. The scandium is then separated from the organic phase with nitric acid and determined spectrophotometrically as the red-brown coloured scandium - alizarin complex. Of the elements usually found in uranium compounds only copper interferes. EXISTIXG methods for determining scandium in the microgram range, especially in the presence of thorium, are time-consuming and not well suited to routine analysis.A method suitable for determining scandium in uranium compounds on a routine basis was not available. Volumetric and gravimetric met hods were too insensitive, and solvent extraction appeared to offer the best approach to the problem. Y-Benzoylphenylhydroxylamine has been used for separating scandium,' but in the presence of thorium, uranium and other elements no ready separation of scandium could be achieved. Broido2 has used 2-theno3dtrifluoroacetone (TTA) for extracting scandium for separation and purification purposes. Preliminary work with this reagent showed promise, and has resulted in the development of a rapid and accurate method for determining scandium in the microgram range in uranium compounds.EXPERIMENTAL The use of alizarin sulphonic acid as the chromotropic reagent for the spectrophotometric determination of scandium was found to provide the required sensitivity, and this reagent has been used throughout this work. The scandium - alizarin complex was developed in an acetate - acetic acid buffer a t pH 3-5. Beer's law was obeyed over the range 0 to 200 pg of scandium in a volume of 100 ml. EXTRACTIOX OF SCAKDIUM- Scandium was extracted from solutions adjusted to different pH values in the range 1 to 6 by shaking for 3 minutes with a 0.2 hi TTA - xylene solution, and separated from the organic phase with 2 M nitric acid. The scandium was then determined with alizarin. Quantitative extraction was given over the pH range 2.5 to 6 (see Fig.1). At about pH 2 and above, uranium precipitates from solution, especially when amounts of the order of 1 g in 50 to 100ml are present. In order to keep the uranium in solution above pH 2 several complexing agents were tried. Tartaric acid was satisfactory for the purpose, but small amounts were found to bleach the colour of the scandium - alizarin complex. Consequently it was necessary to remove all traces of tartaric acid from the organic phase before separating the scandium. Washing the organic phase with water tended to produce emulsions, and the presence of an electrolyte was found necessary to inhibit emulsion forma- tion. A 10 per cent. w/v solution of sodium nitrate was found to be suitable for this purpose. In the presence of tartaric acid, the pH range for complete extraction of scandium was decreased from 2-5 to 6 to 5 to 6 (see Fig.1). INTERFERING ELEMEXTS- Elements found to give a colour with alizarin at pH 3.5 were ironII1, vanadiumv, thorium, aluminium, copper 11, zirconium, titaniumII1, titaniumIv, molybdenum, tin11 and uraniumv1. Elements that did not give a colour and consequently did not interfere were lead, bismuth, chromiumV1, zinc, cobalt, arsenicv, antimonyv, magnesium, calcium, nickel, sodium and114 [Analyst, Vol. 88 potassium. Of the elements giving a colour with alizarin, thorium, vanadiumv, titaniumIII, titaniumIV and copper11 were found to interfere after extraction with TTA - xylene solution. Thorium-Thorium is extracted into TTA - xylene solution above about pH 0.5 (see Fig.1). In the presence of tartaric acid, the range for quantitative extraction is pH 3 to 4 (see Fig. 1). At pH 5, about 50 per cent. is extracted, and a t pH 6 the amount extracted drops to about 25 per cent. These results were given when 2 M nitric acid was used to separate the thorium from the TTA - xylene solution. In the absence of tartaric acid, thorium is quantitatively extracted into TTA - xylene solution a t pH 1 (see Fig. 1) and will not, therefore, be separated from the organic phase with nitric acid at pH 1. Scandium is extracted into TTA - xylene solution to the extent of about 20 per cent. from solutions at pH 1, and can be separated from the organic phase with nitric acid solution at pH 2. ASHBROOK: DETERMINATION OF SCANDIUM IN URANIUM COMPOUNDS The amount of each element added was 10 mg.100- 80. s. S 60, I= ro &I u1 40 20 I 1 I 0 I 2 3 4 5 6 PH Fig. 1. Extraction of scandium and thorium: curve A, thorium (no tartaric acid); curve B, scandium (no tartaric acid); curve C, thorium in presence of tartaric acid; curve D, scandium in presence of tartaric acid was acid A solution containing 67 pg of scandium and 100 p g of thorium, adjusted to pH 5.5, shaken with 0-2 M TTA - xylene solution. After washing the organic phase with tartaric solution adjusted to pH 5 with ammonium hydroxide, and then with a 10 per cent. solution of sodium nitrate, the scandium was separated from the organic phase with nitric acid solution at pH 1. Complete extraction of scandium was given, with no interference from thorium.Tests carried out with up to 1560 pg of thorium and 67 pg of scandium showed that less than 20pg of thorium was extracted by the procedure describedabove; thoron was used to determine the thorium- Thorium added, pg . . . . . . 260 320 1040 1560 As shown in Table I , 20 pg of thorium gives a colour with alizarin having an optical density equivalent to 2 pg of scandium. Vanadium and titanium-Addition of hydrogen peroxide to the sample solution before extraction eliminated interference from both vanadiumv and titanium111 and titaniumIv. Addition of hydrogen peroxide below pH 2 to 3 caused precipitation of uranium, presumably as UO,; addition at pH 4 to 5 was satisfactory. Co$@er--Amounts of copper greater than 2 mg caused precipitation of a copper - alizarin complex at pH 3.5.Only extremely small amounts of copper ((50 pg) can be tolerated. Thorium extracted, pg . . . . . . 10 7 18 14February, 19631 BY SOLVENT EXTRACTION WITH 2-THENOYLTRIFLUOROACETONE 115 Extraction with dithizone, or sulphide precipitation if sufficient copper is present, may be used to remove the copper. Interfering anions-Fluoride and phosphate interfere in the procedure and must be absent. Up to 1 g each of sodium sulphate and sodium chloride do not interfere. TABLE I EFFECT OF THORIUM ON THE DETEKMINATION OF SCANDIUM Thorium added, pg 5 10 20 50 100 1000 Optical density at 530 mp 0.00 1 0.00 1 0.002 0.010 0.015 0.125 Scandium equivalent to optical density, Pg 1 1 2 12 19 158 METHOD RE.4GENTS- 2-Tlzenoyltrifiztoroacetone, 0.2 M-Dissolve 25 g of 2-t henoyltrifluoroacetone in 500 nil Tartaric acid, 20 per cent.w/v. Tartaric acid wash solution, 5 per cent.-Adjust to pH 5 with ammonium hydroxide. Sodium nitrate, 10 per cent. w/v. Nitric acid, pH 1-0-Adjust about 1 litre of water to pH 1.0 with nitric acid. Hydrogen peroxide, 30 per cent. Bufer solzction-Adjust a 10 per cent. w/v solution of sodium acetate to pH 3.5 with Alizarin sulphonic acid, 0.05 per cent., aqueous. of xylene. acetic acid. PROCEDURE- For solid samples, weigh a suitable portion into a platinum crucible, add about 5 g of peroxide, and mix thoroughly; sinter at 400" to 450" C for 1 hour. Cool, dissolve the melt in water, just acidify with nitric acid, and boil for several minutes. Cool the solution, and dilute to a suitable volume.Transfer a portion of solution containing up to 200 pg of scandium to a 100-ml beaker. Add 10 ml of 20 per cent. w/v tartaric acid solution, and, with use of a pH meter, adjust the pH to about 5 with ammonium hydroxide. If necessary, cool the solution to room tem- perature. Add 2 ml of 30 per cent. hydrogen peroxide, and adjust the pH to 5 to 5.5. Transfer the solution to a 125-ml separating funnel, add 20 ml of 0.2 M TTA - xylene solution, and shake vigorously for 3 minutes. Allow the phases to separate, and discard the aqueous phase. Shake the organic phase with 10-ml portions of the tartaric acid wash solution until the aqueous phase is colourless, and discard the washings. Shake the organic phase twice with 10-ml portions of 10 per cent. sodium nitrate solution, shaking for 1 minute at each extraction.Discard the aqueous phase. Separate the scandium from the organic phase twice with 10-mZ portions of nitric acid solution at pH 1.0, and run the aqueous phase into a 100-ml calibrated flask. By pipette, transfer 5 ml of the alizarin solution to the flask. Add dilute ammonium hydroxide to the solution until the colour of the solution just changes from yellow to red. Add by pipette 5 ml of the buffer solution, dilute to the mark with water, and mix. Measure the optical density of the solution in 1-cm cells at 530 mp, the spectrophotometer having been set to zero against a blank solution. Calculate the amount of scandium in the sample from a previously prepared calibration graph. PREPARATION OF CALIBRATION GRAPH- and dilute to 1 litre.scandium per nil, Dissolve 0.1534 g of pure scandium oxide (99.9 pzr cent.) in dilute hydrochloric acid, Dilute this solution five-fold to give a solution containing 20 pg of116 ASHBKOOK [Analyst, Vol. 88 Measure from a burette 0-, 1-, 3-, 5-, 7- and 10-ml portions of the scandium solution into 100-ml beakers and continue as described under “Procedure.” Plot a graph of optical density against pg of scandium. TABLE I1 The solutions contained 0.5 g of sodium diuranate DETERMIXATION OF SCANDIUM I N THE PRESENCE OF THORIUM AND URANIUM Scandium added, CLQ 65 65 130 130 195 195 260 260 Thorium added, Pg 500 1000 500 1000 500 1000 501) 1000 Scandium recovered, 100 101 102 105 99 102 102 100 O/O TABLE I11 COMPARISON OF PROPOSED METHOD \VITH XEUTRON-ACTIVATION ANALYSIS Amount of scandium found by- 7 - proposed method, neutron activation, Sample p.p.m. p.p.m. A 265, 270 257 R 262, 258 258 C 80, 82 - RESULTS AND CONCI.USIONS Known amounts of scandium and thorium were added to solutions of sodium diuranate. The results shown in Table I1 indicate that the recovery of scandium was good. In Table I11 are shown results obtained on uranium compoimds, together with results by neutron-activation analysis. The procedure has been found suitable for determining small amounts of scandium in uranium compounds and also in lime samples. Dilution to a final volume other than 100 ml may be used; for example, if traces of scandium are to be determined, a final dilution of 50 ml or the use of 5-cm cells will increase the sensitivity of the procedure. REFERENCES 1. 3. Alimarin, I. P., and Yung-Schaing, T., I’alanta, 1961, 8, 317. Rroido, A., U.S. Atomic Energy Commission Report r1ECD-2616, 1947; declassified June 1949. Received August 20th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800113

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

February, 19631 RA JZMAN 117 The Quantitative Micro-determination of Biphenyl in Citrus Fruit BY ASNA RA JZMAN { Departnaent of Fruit and Vegetable Stovage, Natio?zal and Usiversity Institute 01 Agriculture, Hehovot, Israel) h simple and rapid colorimetric method is described for the micro- determination of biphenyl in citrus h i t ; it is based on the specific colour reaction given by biphenyl with sulphuric acid, traces of formaldehyde and ferric iron. A sample of the ground fruit is steam-distilled, the biphenyl in the distillate is extracted with chloroform, and the extract is purified with 95.2 per cent. w/w sulphuric acid. The biphenyl is then determined colori- metrically. The method is suitable for routine use and permits the deter- mination of small amounts of biphenyl of the order of 0.06, 1-0 and 0.3 p.p.m.in the fruit pulp, peel and whole fruit, respectively; the accuracy is within the limits of experimental error. BIPHENYL is used as a fungistat for protecting citrus fruit from certain diseases that develop during storage and transport. In the past few years several countries have set limits for biphenyl residues in citrus fruit, and thus it became necessary to be able to discern untreated fruit and to determine minute amounts of biphenyl in treated fruit, The various colorimetric methods1 y 2 v3 described are rather laborious, and the accuracy of spectrophotometric method~**~.~ 9’ y8-despite the great efforts made to eliminate inter- fering substancess-is limited by a variable background or apparent biphenyl value.Further, these methods usually require costly apparatus that is not available in most laboratories. It is based on a blue colour reaction given by biphenyl with sulphuric acid, traces of formaldehyde and ferric iron.g The reaction is specific and sensitive ; it makes possible visual identification or quantitative determination of extremely small amounts of biphenyl in fruit. The proposed method is suitable for routine use. The method proposed here is both simple and rapid. Fig. 1 . Distillation apparatus METHOD GENERAL DESCRIPTION- Biphenyl is steam-distilled from a sample of ground fruit and extracted from the distillate with chloroform ; the extract is then treated with sulphuric acid to eliminate interfering substances. One volume of purified extract is mixed with nine volumes of acetic acid con- taining formaldehyde ; sulphuric acid containing ferric iron is then added.The optical density of the blue colour developed is measured a t 610mp. The colour reaction and the deter- mination of biphenyl in chloroform solution have been described in detail in a previous paper.g118 RA JZMAN : QUANTITATIVE MICRO-DETERMINATION [Analyst, VOI. 88 APPARATUS- Electric blender. Shaking machine. A bsorptiometer-A Hilger absorptiometer type H810, with 6-inch x &inch tubes, was Distillation apparatus-A round-bottomed S-litre flask, condenser and connection tube, Beakers, 250 ml. Separating funnels, 100 and 200 ml. Calibrated $asks, 50 ml, with ground-glass stoppers. Erlenmeyer $asks, 100 ml. Graduated tubes-Tubes, 10 ml, graduated at 0.1-ml intervals, with ground-glass stoppers Test-tubes, 150 mm x 16 mm.Pipettes, 1, 5 and 10 ml. Micropipettes-Curved tip 0-1- and 0.2-ml pipettes, graduated at 0.001-ml intervals. Filter paper, Whatman No. 42. used. with ground-glass joints (see Fig. 1). and adaptor for concentration (see Fig. 2). Fig. 2. Graduated 10-rnl tube with adaptor for concentration of extract: A, inlet; B, connection to vacuum line REAGENTS- All reagents should be of recognised analytical grade. Standard bijbhenyl solation-Dissolve 80 mg of pure biphenyl in chloroform, and make up to 100ml. Acetic acid reagent-Mix 1 ml of 37 pel- cent. formaldehyde solution with glacial acetic acid to make 100 ml. Determine colorimetrically the amount of formaldehyde impurity present in the glacial acetic acid, and then add the necessary amount of the prepared solution to more acetic acid to obtain a reagent containing 16.6 mg of formaldehyde per litre.Check the formaldehyde content of the reagent as described below. Prepare a standard solution containing 16-6mg of formaldehyde per litre of water by mixing 1 ml of 37 per cent. formaldehyde with distilled water to make 100 ml; then dilute a 4-5-ml portion of this solution to 1 litre. Place 0.0, 0.05, 0.10 and 0.15 ml of the standard solution and 0.05, 0.10 and 0.15 ml of acetic acid reagent in separate test-tubes; add 5 ml of chromotropic acid reagent-500 mg of chromotropic acid in 100 ml of diluted sulphuric acid (4 + %B)--to the contents of each test-tube. Place the test-tubes in boiling water forFebruary, 19631 O F BIPHENYL I N CITRUS FRUIT 119 15 minutes, and then cool, and measure the optical densities a t 610 mp.Set the absorpti- ometer to zero against the test-tube containing no standard formaldehyde solution. Equal amounts of acetic acid reagent and standard solution should show the same optical densities. Sulphuric acid reagent-Prepare a stock solution by dissolving 50 g of ferric sulphate in 950 ml of diluted sulphuric acid (1 + 8-5). For use, mix 5 ml of stock solution with 95 ml of 98 per cent. w/w sulphuric acid; prepare daily. Sulphuric ucid, 95-2 per cent. w/u-Dilute 98 per cent. w/w sulphuric acid (1 + 0-053). Chloroform (stabilised by ethanol). Bon,e charcoal powder (not adsorbing biphenyl), PREPARATION OF STANDARD GRAPH- Mix 1 ml of standard biphenyl solution with 9 ml of acetic acid reagent in a 10-ml graduated tube.(Each 0.1 ml of mixture contains 8 ,ug of biphenyl.) Place in separate test-tubes 0-05-, 0-lo-, 0.15-, 0-20-, 0.25- and 0-30-ml portions of this biphenyl mixture. Gradually add 5ml of sulphuric acid reagent to the contents of each test-tube. When a small amount of sulphuric acid reagent has been added, mix vigorously until full colour intensity is obtained; then add the rest of the reagent with thorough mixing. The colour obtained should be clear; any secondary colour, such as grey or violet, indicates failure of the test. Wait a few minutes until gas bubbles have disappeared, and then measure the optical densities at 610mp. Set the absorptiometer to zero against the sulphuric acid reagent.Beer - Lambert’s law is obeyed up to 14 pg of biphenyl in 5 ml of reagent. The standard graph is reproducible and is valid for solutions containing between 25 and 800 pg of biphenyl per ml of chloroform. Other solutions have to be suitably concentrated or diluted with chloroform. PRELIMINARY PROCEDURE- Biphenyl is determined in the peel and in the pulp of fruit and calculated for the whole fruit. Sampling-From an average sample of fruit, prepare 125 g of peel and about 600 g of pulp. Weigh, for example, 10 fruits, and remove their peels carefully to avoid contaminating the pulp with traces of biphenyl from the peel; put aside from each fruit an average sample of 12-5 g of the peel, making altogether 125 g. Weigh the pulps, and calculate the weight of the total peels.If the pulps weight more than 600 g, take from each pulp an aliquot part. Place the chopped peels in the electric blender, rinse the plate with 400ml of distilled water, add the rinsing to the peels, and grind to an impalpable purke; this is essential for the complete extraction of biphenyl. Transfer the purbe quantitatively to the distillation flask, rinse the blender with 400ml of distilled water, and add the rinsing to the purke. Preparation of pulp sample-Grind the chopped pulp in the electric blender. Transfer the resulting purke quantitatively to the distillation flask, rinse the blender with 200ml of distilled water, and add the rinsing to the pur4e. Distillation-Connect the distillation flask to the condenser (see Fig. l ) , which dips into a 250-ml beaker containing 20 ml of distilled water.Heat the flask with a gas burner, and distil. To ensure a clear distillate, boil gently, and avoid production of foam. Collect the first 100 ml of distillate. Use 10 ml of chloroform to rinse carefully the inner tube and the tip of the condenser, and collect the rinsings in the distillate. Rinse the beaker with 5 ml of chloroform, add the rinsings to the contents of the funnel, and shake for 2 to 3 minutes. Carefully decant off the chloroform layer, and extract three more times with 10-ml portions of chloroform. Collect the chloroform layers in a 50-ml calibrated flask, make up to the mark with chloroform, and insert the stopper tightly in the flask.” (These extracts can, if necessary, be stored for some time in the dark.) Clean-wp of interfering substances-Place about 35 ml of the chloroform extract into a 100-ml separating funnel.Add 10 ml of 95.2 per cent. w/w sulphuric acid, and shake for 10 minutes. Decant off the acid layer carefully, and repeat the operation several times. Transfer the purified extract (see “Preliminary Determination”) to a 100-ml Erlenmeyer Biphenyl will be determined per ml of extract and calculated per 50 ml. Preparation of $eel sample-Chop the peels on a plate. Extraction-Transfer the distillate to a 200-ml separating funnel. * Avoid all changes in the concentration of the extract.120 RA JZMAN QUANTITATIVE MICRO-DETERMINATION [A naZySt, YOI. 88 flask, add a little bone charcoal powder, mix, and filter through Whatman No. 42 filter-paper. Collect the filtrate in a 50-ml graduated cylinder, and close carefully. (Record the exact volume to allow correction for eventual evaporation of chloroform.) PreLiminary determination-Certain extracts are sufficiently pure for determination, even though the chloroform layer and sulphuric acid extract are intensely coloured; others, especially those from lemon peels or from peels and pulp containing small amounts of biphenyl, require more careful purification, even if the chloroform layer and the sulphuric acid extract are colourless. After four or five treatments with sulphuric acid (six or seven for the less satisfactory extracts), check whether or not the extract is sufficiently pure and simultaneously determine its approximate biphenyl concentration.Place 0-01-ml portions of the chloroform layer in three separate test-tubes, and add to each 0.09, 0.14 or 0.19 ml of acetic acid reagent.Add to the contents of each tube 5 ml of sulphuric acid reagent, as described for the prepara- tion of the standard graph. Measure the optical densities, (a) If the colours produced are not clear blue, purify the extract an additional two or three times, and repeat the preliminary determination. (b) If the colours of the solutions in the three test-tubes are blue and their optical densities are about the same, the extract is sufficiently pure and its biphenyl concentration is appropriate for the determination. (c) If the colours produced are intense-blue or turning from violet to blue---and the optical densities increase for the larger amounts of acetic acid reagent, then the extract is sufficiently pure, but needs to be diluted.Take less than 0.01 ml of extract, e.g., 0.005 or 0.0025 ml, until the optical densities obtained with increasing amounts of acetic acid reagent are about the same. (d) If the colours are not discernible visually or are only faintly blue or violet, the extract needs to be concentrated and the remaining impurities interfering with the colour reaction eliminated. (Extracts purified four or five times need to be purified a further two or three times before concentration.) Place 1 ml of the chloroform layer in a 10-ml graduated tube, add 1 ml of acetic acid reagent, and concentrate to 0.1 ml as described under “Procedure for determining Biphenyl.” Add a further 1 ml of acetic acid reagent, and concentrate once more to 0.1 ml.Adjust the volume with acetic acid reagent to 0.5 ml for pulp extracts or 2 ml for peel extracts. Place in two separate test-tubes 0.10-ml portions of the mixture (0.30 ml if the colour is extremely faint), add to the contents of one tube 0.10 ml of acetic acid reagent, and then introduce into each tube 5 ml of sulphuric acid reagent as described for the preparation of the standard graph. If the colour is not clear, proceed as indicated under (a) above. Measure the optical densities (see (b) and (c)), and then calculate the volume of extract to be concentrated in order to make 5 or 10 ml of mixture. Sote the required degree of dilution. PROCEDURE FOR DETERMINING BIPHENYL-- According to the approximate concentration of biphenyl in the extract determined under ( b ) , (c) or (d) above, proceed appropriately as described below.(6) Place 1 ml of extract in a 10-ml graduated tube, add 9 ml of acetic acid reagent, insert the stopper in the tube, and mix. Measure the optical densities corresponding to 0.05, 0.10 and 0-15ml of the mixture (0.10, 0.20 and 0.30ml if the colours are extremely faint) as described for the preparation of the standard graph. The colours obtained must be clear, and the optical densities should rise proportionally with the amounts of the mixture taken for the determination. Calculate the amount of biphenyl in 0.10 ml of the mixture. (c) Place in a 10-ml graduated tube the required volume of extract, e.g., 0-50 or 0.25 ml, adjust the volume to 1 ml with chloroform, add 9 ml of acetic acid reagent, insert the stopper in the tube, and mix; continue as under (b).(d) Introduce by pipette into a 10-ml graduated tube a maximum of 5 ml of extract. Add 1 ml of acetic acid reagent, and insert the concentration adaptor (see Fig. 2) in the tube. Place the tube in a beaker containing a depth of 2 to 3 cm of water maintained at 50” to 60” C. Connect the adaptor to a vacuum pump, and carefully reduce the volume to about 1.0ml. Close the connection to the vacuum pump, and add through the inlet a maximum of 5 ml of extract. Rinse the inlet with a few drops of chloroform, insert the stopper, and again reduce the volume to 1.0 ml. Continue this process until the required volume of extract has been introduced. Add 1 ml of acetic acid reagent, and concentrate once again to 0.1 ml.Remove the adaptor from the tube, and use 1 or 2 ml of acetic acid Concentrate finally to 0.1 ml.February, 19631 OF BIPHENYL IN CITRUS FRUIT 121 reagent to rinse the inside of the adaptor and the capillary; collect the rinsings in the graduated tube. Add 0.5 or 1-Om1 of chloroform, and adjust the volume with acetic acid reagent to 5 or 10 ml, respectively (see (d) under “Preliminary Determination”). Insert the stopper in the tube, and mix; continue as under (b). CALCULATIOK- Calculate the biphenyl content in the peel and in the pulp from the relation- p.p.m. a x 10 x b x 50 c x d Biphenyl content = where n -- micrograms of biphenyl in 0.10 ml of mixture, b = volume of prepared mixture in millilitres, c = volume of extract in millilitres taken for the preparation of the mixture and d = weight of peel or pulp in grams taken for determination.Calculate the biphenyl content of the whole fruit. DISCUSSIOY OF THE METHOD SEXSITKIT\-- The method permits the determination of 0.06, 1.0 and about 0.3 p.p.m. of biphenyl in the pulp, peel and the whole fruit, respectively, The sensitivity can be increased by preparing from the concentrated extract a smaller amount of mixture or by using a larger sample of peel or pulp. ACCURACI-- Biphenyl was determined in essential oils, peel and pulp of untreated fruit to which known amounts of biphenyl had been added (see Table I ) ; in several samples originating from the same mixture of peels of fruit treated with biphenyl (see Table 11) ; in several average samples of peels successively removed from the same batch of ten biphenyl-treated fruits (see Table 111).The results show that deviations are slight and remain within the limits of experimental error. TABLE I RECOVERY OF BIPHENYL Sample Orange essential oil Orange essential oil Orange essential oil Orange essential oil Orange essential oil Orange peel . . Orange peel . . Orange peel . . Lemon peel . . Lemon peel . . Lemon peel .. Grapefruit peel . . Grapefruit peel . . Grapefruit peel . . Orange pulp . . Orange pulp . . Lemonpulp . . Lemon pulp . . Grapefruit pulp . . Grapefruit pulp . . . . .. .. . . . . . . .. . . .. . . . . .. . . .. .. .. .. .. .. .. .. . . . . . . . . . . .. .. .. .. .. . . .. .. .. .. .. .. .. .. Biphenyl added, mg 0.00 4.00 17.90 20.40 55-70 0.00 8.00 50.00 0.00 5.00 8.00 0.00 8-00 20.00 0.00 0-80 0.00 0-50 0.00 0.40 Biphenyl found, mg 0.00 3.80 17.20 20.00 55.00 0.00 8.15 50.00 o*oo 4.85 7-80 0.00 7-90 20.00 0.00 0.80 0.00 0.48 0.00 0-41 Difference, mg 0.00 - 0.20 - 0.70 - 0.40 - 0.70 0.00 +Om15 0.00 0.00 -0.15 - 0.20 0.00 - 0.10 0.00 0.00 0.00 0.00 - 0.02 0.00 +0*01 Error, 0.0 - 5.0 - 4.0 - 1.9 - 1-2 0.0 + 1.9 0.0 0.0 - 3.0 - 2.5 0.0 - 1.2 0-0 0.0 0.0 0.0 - 4.0 0.0 + 2.5 % VALIDITY- and water was found in the first 25 ml of distillate (see Table IV).from 300 to 2000 ml, does not affect the results. DistiZZution-&lost of the biphenyl collected during distillation of a mixture of biphenyl The amount of water,122 RA JZMAN : QUANTITATIVE MICRO-DETERMINATION [Analyst, Vol.88 Moreover, it was observed that the amounts of biphenyl liable to be found in fruit samples are completely collected in the first 100 ml of distillate (see Table V). The times for complete distillation of peel and pulp samples are about 20 and 40 minutes, respectively. TABLE I1 DETERMINATION OF BIPHENYL I X TKEATED ORANGE FEELS Four samples taken from the same mixture of peels Sample No. 1 2 3 4 Average Biphenyl found, mg . . . . . . 40.50 41.13 39.50 40.50 40.4 1 Deviation from average value, mg . . -toe09 + 0-72 -0.91 $- 0.09 Deviation from average value, yo . . +0*22 +1.77 -2.22 + 0.22 TABLE III IIETERMINATION OF BIPHENYL IN TREATED CITRUS PEELS Three average samples taken from the same batches of 10 fruits Biphenyl found in- 7 -l Sample No.orange peels, lemon peels, grapefruit peels, mg mg mg 1 17.48 34.8 11.2 2 17.48 26.3 11.2 3 17.48 26.7 10.8 TABLE 11' RECOVERY OF BIPHENYL IN SUCCESSIVE %-ml FRACTIONS OF DISTILLATE Distillation of a mixture of water and biphenyl Biphenyl found in fraction No. - 7 Biphenyl r~--------- -~ added, I, 11, 111, IV, v, v1, total. mg mg mg mg mg mg 1% mg 20.0 18-80 0.94 0.12 0.00 0.00 0.00 19-86 50.0 48-75 1.16 0.28 0.00 0.00 0.00 50.19 lC)O*O 78.75 16.25 5-00 1.00 0.08 0.00 101-ox Yurijcation with 95.2 per cent. w/w sul~lzz~ric acid-Conflicting opinions exist among workers1~4~5~7~s~10 about the eventual loss of biphenyl owing to the action of the concentrated sulphuric acid used to remove interfering substances. According to Koether,lo biphenyl is lost owing to sulphonation, and the loss increases with time of shaking.Rajzmang noted that biphenyl, detectable by the colour reaction, disappeared rapidly when a large amount of sulphuric acid was added to the biphenyl dissolved in a small amount of acetic acid or ethanol; this was attributed to sulphonation. On the other hand, during this investigation no loss of biphenyl was observed after rigorous purification of three volumes of chloroform extract by one volume of concentrated sulphuric acid. That no biphenyl was lost seemed surprising; assuming that this Lvas due to particular conditions empirically established for the clean-up procedure, the effect of 98 per cent. w/w sulphuric acid acting under various conditions on biphenyl dissolved in chloroform was studied. I t was found that some biphenyl was lost, and a number of conclusions were drawn.(1) Loss of biphenyl depends on the raho between the volumes of concentrated sulphuric acid and chloroform; it increases with an increase in the ratio (see Table "1). (2) Loss of biphenyl increases with shaking time and is accelerated by repeatedly replacing the concentrated sulphuric acid by fresh acid (see Table hT1I). The influence of shaking time on the loss of biphenyl was not constant. (3) Loss of biphenyl depends on the concentration of the sulphui-ic acid, and decreases with increased dilution (see Table VIII).February, 19631 OF BIPHENYL IN CITRUS FRUIT TABLE V RECOVERY OF BIPHENYL IN SUCCESSIVE 100-ml FRACTIONS OF DISTILLATE 1Tntreated f y u i t - Orange pulp Lemon pulp Grapefruit pulp Orange peel Orange peel Lemon peel Lemon peel Grapefruit peel Grapefruit peel Tjreated fmit--- Orange pulp Lemon pulp Grapefruit pulp Orange peel Orange peel Lemon peel Lemon peel Grapefruit peel Grapefruit peel .... .. .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biphenyl found in fraction Xo.- Biphenyl ,- - added, 1, 11, 111, total, mg mg "g mg mg 0.40 0.40 0.80 10.00 20.00 1-60 4-00 1.60 20.00 0.40 0.38 0.78 9.85 80.00 1.66 4.00 1-60 19-90 0.45 0.28 ' 0.12 32-60 38.44 3-36 10.15 5.04 6.18 0.0 0.0 0.0 0.0 0.06 0.0 0.0 0.0 0-04 0.0 0.0 0.0 0.09 0.0 ? 0.0 0.0 0-02 0-0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0-0 0.40 0.38 0.78 9.85 20.06 1.66 4-00 1-60 19-94 0.45 0.28 0.12 32-69 35.44 3.36 10.15 5-04 6-20 123 TABLE VI EFFECT OF THE RATIO BY VOLCME OF SULPHURIC ACID TO CHLOROFORM ON LOSS OF BIPHENYL Biphenyl (300 mg per ml) in chloroform shaken for 15 minutes with 98 per cent.w/w sulphuric acid Ratio . . . . . . s o t 0 1 2 0 t o 1 1 0 t o l 5 t o 1 2 t o 1 1 t o 1 1 to 3 Loss of biphenyl, :& . . 100.0 100.0 95.0 90.0 85-0 5.0 0 4 TABLE VII EFFECT OF SHAKING TIME ON LOSS OF BIPHENYL Biphenj-1 (10 mg) in 30 ml of chloroform shaken with 10 ml of 98 per cent. w/w sulphuric acid replaced every 10 minutes Shaking time, hours . . 4 8 12 16 J-oss of biphenyl, ol0 . . 0 20 40 50 TABLE VIII EFFECT OF WATEK CONTENT OF SULPHURIC ACID ON LOSS OF I~IPHENI'I, Biphenyl (10 mg) in 30 ml of chloroform shaken with 10 ml of sulphuric acid replaced every 10 minutes; shaking time 12 hours Concentration of sulphuric acid, w;w. .98.0 97.5 97-0 96.4 95.8 95.2 Loss of biphenyl, yo . . . . . . 40 25 17 13 0 f l The 95-2 per cent. acid chosen lends itself well to the clean-up procedure. Less con- centrated acids can be used, but the more dilute the acid the more highlj. coloured beconies the chloroform layer. It is possible that the differences of opinion mentioned above are due, in part at least, to the fact that the sulphuric acids used by various workers, although concentrated, $\-ere not exactly similar. Elimination of the remainiizg iwter feri?zg substances-In some extracts, particularly from lemon peels, interfering substances reacting in part with formaldehyde were found to persist124 RA JZMAN [Aizalyst, Vol.88 despite prolonged treatment with sulphuric acid, washing with sodium hydroxide solution and distilled water, and treatment with potassium permanganate. This difficulty of eliminating interfering substances has also been reported by Gunther and Blinnll in the spectrophotometric determination of biphenyls ; despite a rigorous clean-up treatment, they found that a slight and variable interference resulted from the presence of p-cymene from the citrus oil.ll It is possible that the impurities interfering with the colour reaction are also formed, in part at least, by p-cymene, of which they possess certain properties. I t has been noted that these impurities dissolved in a small amount of acetic acid reagent (or acetic acid) are eliminated by evaporating the solvent. No loss of biphenyl was found to take place during the evaporation of acetic acid reagent as long as a small amount remained. Presence of o-phenyZphenoZ-It should be noted that o-phenylphenol, which is used as a fungicide, may be found in the extract. With sulphuric acid, traces of formaldehyde and ferric iron, o-phenylphenol gives a pink c o l ~ u r . ~ Nevertheless, its presence in the extract does not interfere with the determination of biphenyl, since it disappears quickly during the treatment with sulphuric acid. This work was sponsored by the Citrus Marketing Board of Israel and was carried out with the technical assistance of Mrs. H. Heller, M. Braha and H. Levy. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Tomkins, R. G., and Isherwood, F. -4., Analyst, 1945, 70, 330. Haecke, F., and Cats, H., Cham. Weekbl., 1957, 53, 609. Bohme, H., and Hofmann, G., 2. LebensmnittUntersuch., 1961, 114, 96. Cox, H. E., Analyst, 1945, 70, 373. Stern, A. P., and Rosselet, I?., Ibid., 1949, 74, 89. Bohme, H., and Bertling, L., 2. LebenswiittUntevsuch., 1957, 105, 311. Ihloff, M. J., and Kalitzki, M., Mitt. GDCH Fachgr. Lebensmitt., 1957, 11, 129. Gunther, F. A., Blinn, E C., and Barkley, L. H., University of California, 1959. Rajzman, A., Analyst, 1960, 85, 116. Koether, B., 2. Lebensmitt Untersuch., 1958, 108, 158. Butz, W. H., and Noebels, H. J., Editors, “Instrumental Methods for the Analysis of Food Additives,” Interscience Publishers Inc., New York and London, 1960, p. 129. Received Afiril 16th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800117

出版商:RSC    

年代:1963

数据来源: RSC