ISSN:0003-2654    

DOI:10.1039/AN95782FX050

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BX052

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782FP157

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BP163

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

DECEMBER, 1957 Vol. 82, No. 981 THE ANALYST PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETING AN Ordinary Meeting of the Society was held at 6.30 p.m. on Wednesday, December 4th, 1957, in the meeting room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by Dr. J. R. Nicholls, C.B.E., F.R.I.C., Past President. The meeting took the form of a discussion on “Standardisation” and was opened by R. C. Chirnside, F.R.I.C., L. S. Theobald, M.Sc., A.R.C.S., F.R.I.C., J. Haslam, D.Sc., F.R.I.C., and G. Ingram, A.R.I.C. Owing to Mr. Theobald’s absence through illness his contribution was read by Mr. Chirnside. NEW MEMBERS ORDINARY MEMBERS Ernest Vernon Browett, BSc. (Lond.), A.R.I.C. ; David Maxwell Brown, BSc. (Edin.), F.P.S.; Arthur Clark, BSc.(Sheff.); Raymond Thomas Clark; Clive Kear Colwell, M.A. (Cantab.) ; Eric Leslie Crooks; Ernest George Cummins, B.Sc. (Lond.) ; Christine Sylvia Delves, BSc. (Reading) ; William Ernest Elstow, BSc., Ph.D. (Lond.), A.R.I.C. ; Frank Joseph Green, M.Sc. (Lond.), F.R.I.C. ; Dennis Francis Harris; George Frederick James Hart; Francis Henry Hillyard, BSc. (Birm.) ; Sidney Louis Kidman, A.R.I.C. ; John Weale Macaulay, A.M.C.T.; Fasir Mahmood, B.Sc. (Punjab) ; Bernard Whitley Elliott Minifie, F.R.I.C.; John Park, BSc. (Edin.), Dipl.Tech.Chem., F.R.I.C.; Robert Gray Reid, A.R.I.C. ; Brian James Rushton, BSc. (Manc.) ; Rosemary Sales, BSc. (Lond.) ; Ian Robert Scholes, B.Sc. (Lond.), A.R.I.C. ; Derek William Skidmore; Jack Thomas; James Raymond Townley ; Edmond Albert Underwood; John Oswald Lance Wrigley.JUNIOR MEMBERS Keith Edward Bicknell, A.I.M.L.T. ; Derek Alfred Day; Douglas McDonald Dick, BSc. (Aber.) ; James Alistair Grant; David Ramsden. SCOTTISH SECTION AN Ordinary Meeting of the Section was held at 7.15 p.m. on Wednesday, October 30th, 1957, in the Central Hotel, Glasgow. The Chair was taken by the Chairman of the Section, Dr. Magnus Pyke, BSc., F.R.I.C., F.R.S.E. A discussion on “The Estimation of Additives to Bread and Flour” was opened by J. Sword, M.A., BSc., Ph.D., F.R.I.C. A. N. Harrow, A.H.-W.C., F.R.I.C., and H. C. Moir, B.Sc., F.R.I.C., also contributed. The meeting then continued with a general and informal discussion. WESTERN SECTION AND MICROCHEMISTRY GROUP A JOINT Meeting of the Western Section and Microchemistry Group, together with the South- Western Counties Section of the Royal Institute of Chemistry, was held on Friday and Saturday, September 27th and 28th, 1957, in Exeter.On Friday afternoon a visit was paid to the Silverton Paper Mills of Messrs. Reed & Smith Ltd. At 5.15 p.m. there was a meeting in the Washington Singer Laboratories of the University of Exeter on “Some Applications of Microchemistry.” The Chair was taken by Mr. D. F. Phillips, F.R.I.C., Chairman of the Microchemistry Group, who thanked Mr. E. Bishop, BSc., A.R.T.C., A.R.I.C., of the University of Exeter, for his work in the organisation 777778 PROCEEDINGS [Vol. 82 of the meeting. The following papers were presented and discussed: “Applications to Paints and Pigments,” by C.Whalley, B.Sc., F.R.I.C. (see summary below) ; “Applications to Soils and Fertilisers,” by B. M. Dougall, MSc., F.G.S., A.R.I.C. After a social programme on Saturday, September 28th, there was a Discussion Meeting on “The Use and Abuse of Microchemistry,” which was introduced by C. L. Wilson, D.Sc., Ph.D., F.R.I.C., and S. Bance, BSc., A.R.I.C. SOME APPLICATIONS OF MICROCHEMISTRY : APPLICATION TO PAINTS AND PIGMENTS MR. C. WHALLEY selected three topics for detailed discussion, as follows- (i) the semi-micro qualitative analysis of pigment systems; (ii) the analysis of small amounts of gases; (iii) the techniques used to examined the various types of blemishes that occasionally In the first example a new scheme for the qualitative analysis of white pigments and extenders on the milligram scale was described.Classical analytical schemes were not easily applicable to this type of system owing to the intractable nature of the materials present, and further information was required about the composition of the system than could be obtained by finding the elements present. The new scheme avoided fusions, relying on the limited solubility of the materials in acids of increasing strength. The materials were divided into groups based upon their solubilities in dilute nitric, concen- trated hydrochloric, concentrated sulphuric acid and concentrated phosphoric acids and sodium hydroxide solution. Sensitive colour tests were used to detect the various elements after limited separations. The whole scheme was set in the form of a working board, and mixtures up to and including sixteen components could readily be analysed.The second example was illustrated by an improved apparatus for the analysis of about 0.3 ml of gas samples. The method used was based upon classical techniques, but solid bead reagents were used as absorbers. All the gas holders, reagent bead holders, sparking wires and transfer pipettes were securely mounted, but could easily be brought into action in the gas chamber and measuring burette. The whole apparatus had been designed to be very simple and foolproof in operation. Finally, the techniques developed to examine the various stains, inclusions, blooms, crystals and so on that occasionally appear on painted surfaces were described. These could cause deterioration in the appearance of the paint film and might even cause the adhesion to fail.The amount of material available for this detective work was usually very small, and the whole of the work was carried out under the microscope at a magnification of 50 x with the help of a pair of micro dissector units. A variety of special cutting tools, needles, remote operated forceps, and so on, for removing the material for investigation were described, together with a micro fusion device and “wet box” for carrying out chemical tests. These latter were performed in ground depressions on a microscope slide, in hair capillaries, on treated papers and on the surface of single beads of ion-exchange resins suitably treated. Most of the work was qualitative and tests were described for the identification of the common types of blemishes, appear on painted surfaces.WESTERN SECTION AND PHYSICAL METHODS GROUP A JOINT Meeting of the Western Section and the Physical Methods Group was held at 6.30 p.m. on Friday, October 25th, 1957, in the Kings Head Hotel, Newport, Monmouthshire. Mr. P. J. C. Haywood, BSc., F.R.I.C., Chairman of the Western Section, opened the meeting, welcoming the Physical Methods Group to Newport and then invited Dr. J. E. Page, F.R.I.C., Chairman of the Physical Methods Group, to take the Chair. The following papers on “Flame Photometry” were presented and discussed: “Atomic Absorption Spectroscopy,” by A. C. Menzies, M.A., DSc. (see summary below) ; “Recording Flame Photometry,” by L. Brealey, BSc. (see summary below). The meeting was preceded at 2.15 p.m.by a visit to the factory of Monsanto Chemicals Ltd. ATOMIC ABSORPTION SPECTROSCOPY DR. A. C. MENZIES said that, since Wollaston’s discovery of the absorption lines, known as Fraunhofer lines, in the spectrum of the sun, little use had been made of theDec. 19571 PROCEEDINGS 779 phenomenon for analytical purposes. A. Walsh in Australia had drawn attention to atomic absorption as a means of analysis, had made appropriate apparatus, and had used it more especially for the measurement of oscillator strengths, of great importance in the theory of atomic emission and absorption processes. At first it had been hoped that the method would be generally applicable, but conditions had not so far been found for that. Elements fell broadly into three classes: Some, like magnesium, zinc and copper, were very sensitive; others, like iron and chromium, were only moderately sensitive; some, like aluminium, were extremely insensitive.The explanation of this was not straightforward, and there might be a number of causes. A second point of interest was the curvature of the working graphs produced by plotting the optical density of the flame against concentration of the metal being sprayed into the flame. Near the origin, i.e., for low concentrations, the graph was linear, but it changed slope eventually, tending to become more parallel to the axis of concentration. These matters were being worked upon, and the speaker stated the latest views on them. He also gave some account of uses to which the equipment could be put.RECORDING FLAME PHOTOMETRY MR. L. BREALEY said that flame photometry had been established as one of the most useful of analytical techniques, and, as hotter flames were used, the number of elements that could be determined was increasing. As spectra became more complex, recording instruments could be used with great advantage, both for developing analytical methods and for routine use. Such instru- ments were quite simple, and any good flame spectrophotometer could be easily adapted for recording purposes. One of the most useful applications had been a study of the changes in flame background that occurred with variation in sample constitution. Many of the reported cationic interferences could be attributed to this change in background, and by using recorded spectra it was a simple matter to eliminate this variable. MIDLANDS SECTION Ordiiiary Meeting of the Section was held at 7 p.m. on Thursday, October 24th, 1957, The Chair was taken by the Vice-chairman of the The following paper was presented and discussed : “The Analytical Chemistry of Morphine in the Gas Showrooms, Nottingham. Section, Dr. S. H. Jenkins, F.R.I.C., F.1nst.S.P. Poisoning,” by A. S. Curry, M.&4., Ph.D. MICROCHEMISTRY GROUP THE eleventh London Discussion Meeting of the Group was held at 6.30 p.m. on Wednesday, October 30th, 1957, in “The Feathers,” Tudor Street, London, E.C.4. Dr. G. F. Hodsman, A.Inst.P., took the Chair. A discussion on “British Standards in Microchemistry” was opened by C. Meredith and G. Ingram, A.R.I.C.

ISSN:0003-2654    

DOI:10.1039/AN9578200777

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

780 POLLARD, MCOMIE, BANISTER AND NICKLESS [Vol. 82 Quantitative Inorganic Chromatography Part III.* The Separation and Determination of Ferrous and Ferric Iron BY F. H. POLLARD, J. F. W. McOMIE, A. J. BANISTER AND G. NICKLESS The method of selecting a solvent system suitable for the separation of various metals in different valency states is described. The optimum composition of this solvent system for the separation of ferrous and ferric iron has been investigated and a detailed study made of the experimental vari- ables that affect this separation. A comparative study has also been made of methods for the quantitative determination of ferrous and ferric iron after separation by paper chromato- graphy. The preferred method involves extraction of the metal from the paper and colorimetric measurement after complexing with 2-nitroso-l- naphthol-4-sulphonic acid.The ratios of ironu to ironIu investigated were between 25 to 1 and 1 to 49, and the method was found to be applicable to amounts ranging between 25 and 1250 pg of ironII and between 25 and 2450 pg of ironIU. The maximum error expected for any single determination was between 1 and 4 per cent., depending upon the amounts present and the ratio of iroiP to ironIU. A new type of graduated capillary dropper that delivers, almost instan- taneously, minute amounts of liquid (about 1 to 3 p1) is described. The application of the new chromatographic technique to the determination of iron in mixtures with chromium, manganese, cobalt, nickel and copper is indicated. Salicylaldehyde is used as a sensitive spray reagent for the detection of manganese.Part 1. The Selection of a Solvent for the Optimum Separation of Ferrous and Ferric Iron by Paper Chromatography .~LTHOUGH much work has been done in the last few years on the paper-chromatographic separation of mixtures of different metal cations, the separation of the ions of a metal according to their valency states has received much less detailed attention. Separations of various ionic species of individual polyvalent metals have, however, been reported for antimony,l arsenic,l,293 chromium,lr4 cobalt,l copper,lJ’v6 i r 0 n , ~ ~ 7 ~ * ~ ~ mercury,l1517~9~10,11 molyb- denum ,1,9 P P J 4 platinum: plutonium,16 t h a l l i ~ m , ~ ? l 6 > l ~ uranium**17 and vanadium.1 Part 1 of this paper describes the selection of a general solvent for the separation, on the same chromatogram, of the valency states of many metals as chlorides,l the modification of this solvent (by a semi-graphical method) to give the optimum separation of ferrous and ferric ammonium sulphates, and a study of the experimental variables that affect this separation.SELECTION OF GENERAL SOLVENT The components of the chromatographic solvent system, ether, methanol, water and Iiydrochloric acid, were selected for the following reasons. Aqueous hydrochloric acid was chosen because (a) as is well known, the anion and cation of a salt may move independently of each other when eluted on a chromatograni, depending on the nature of the mineral acid present in the solvent. Doubling of the spot is liable to occur if the acid of the solvent is much weaker than the acid from which the solute is derived, e.g., ferric chloride produces two spots in the solvent system n-butanol - glacial acetic acid - water (5 : 4 : 1) on acid-washed paper.ls Also (b) the sensitivity of Fe2+ and u4+ to atmospheric oxidation rapidly diminishes with decreasing pH, and (c) ferric chloride and uranyl chloride tend to show high RF values in chromatographic solvents containing high concentrations of hydrochloric acid together with ether, lower alcohols or ketones, For ironII1, this is due to the solubility of the complex formed between ferric chloride and the excess of hydrochloric acid.* For particulars of Part I1 of this series, see reference list, p. 799.Dec. 19571 QUANTITATIVE INORGANIC CHROMATOGRAPHY. PART I11 781 Methanol and ether were chosen (a) because the low viscosity of methanol assisted in rapid separation and (b) because, by the use of solvents prepared by mixing a polar solvent with various proportions of a non-polar solvent, the RF values of ions and the quality of separations could be systematically varied.The best ratio of methanol to ether was found by adding various amounts of methanol (between 10 and 50 nil) to a mixture of 1 ml each of water and concentrated hydrochloric acid, sp.gr. 1.18, and 50 ml of ether, and then plotting graphs of the RF values for each metal and valency state against the ratio by volume of methanol to ether. In general, RF values rose steadily as the methanol - ether ratio increased, and the widest RF separations occurred at ratios between 1 to 1.6 and 1 to 2.(The solvent system ether - acetone - water - concen- trated hydrochloric acid was also investigated, but the iron111 spot showed appreciable tailing.) With 30 ml of methanol and 50 ml of ether, the volumes of water and concentrated hydro- chloric acid were varied as in the example shown in Fig. 2. The solvent finally selected for the separation of the valency states of iron, chromium, molybdenum, uranium and other metals was composed of ether, methanol, water and concentrated hydrochloric acid in the ratio 50 : 30 : 15 : 4.l3 THE SELECTION OF A MODIFIED SOLVENT BY A SEMI-GRAPHICAL METHOD The general solvent was modified after an examination of the results of the experiments with the general solvent discussed above.Fig. 1 shows the saturation curve of 50 ml of ether plus 30 ml of methanol with aqueous hydrochloric acid at 18" C; any point within the area bounded by the axes and the curve corresponds to a homogeneous mixture, while points outside this area correspond to mixtures that give two phases. -0 L - --Q '$lo<; 8 5 .; :s 05 5 - E 2 2x s= 0 10 5 lo 15 20 25 Volume of water, ml Fig. 1. Saturation curve for 50 ml of ether and 30 ml of methanol EXPERIMENTAL CHROMATOGRAMS- EJects of water and acid concentrations-The effects on the chromatogram of variations in water and acid concentrations are shown in Fig. 2. In each experiment the volume of ether was 50 ml and that of methanol was 30 ml, and all mixtures were homogeneous. Iron was applied as 1.40-pl spots of a solution of ammonium ferrous or ferric sulphate in 0.25 or 0.5 N sulphuric acid, respectively, and a 1 + 1 mixture of the two solutions (all solutions containing 5 pg of iron per pl) ; these spots were placed 2 cm apart on the starting line of each sheet of unwashed Whatman No.1 filter-paper. The times of jar equilibration before elution were 70 minutes 2 1 minute and the times of run were 2 hours 30 minutes i 2 minutes at 16.8" & 0.5" C. All other details of the ascending elution technique were as described later for the investigation of factors affecting the separation of ferrous and ferric iron (p. 783). When the chromatograms were dry, the solvent front, indicated by a bluish white fluorescence in ultra-violet light, was marked on each chromatogram, and the spots were revealed by spraying with 0.5 per cent.aqueous potassium ferricyanide.8 Since the purpose of the investigation was to find the solvent that gave the most rapid and complete separation of ferrous and ferric iron, the chromatograms were run for a fixed time and the distance moved by the spots was plotted against acid concentration at 4, 8 and 12 ml of water (Fig. 2). This figure shows that the best separation took place with 8 ml of water and 6 ml of acid. A similar conclusion was also reached when the results were plotted in terms of R, instead of distance ,7 82 I" 16- 14- E, 2 12- v 10- E 8 - 8 2 6- 6 POLLARD, MCOMIE, BANISTER AND NICKLESS : [Vol. 82 \ solvent front J"" ironmspot I I j tailing I : ! 4- 0 1 I I I I I I I I l l 4 4 4 4 8 8 8 8 12 12 12Water 2 4 6 8 2 4 6 8 2 4 6 Concentrated HCI Composition, ml Variation of distance moved in 2& hours by iron spots with change in volumes of water and acid; crosses denote the centre of gravity of each spot and dotted lines indicate forward or backward tailing of the spots Fig.2. i I 1 1 - 30 40 50 60 70 Volume of ether, ml Fig. 3. Variation of distance Fig. 4. Variation of distance moved and RF value with volume of methanol; volume of ether, 50 ml In Figs. 3 and 4: volume of water, 8 ml; volume of concentrated hydrochloric acid, moved and RF value with volume of ether; volume of methanol, 30 ml 6 ml; x FcII = distance moved by the x FeIII = distance moved by the centre of gravity of the Rr Fell1 = (Rpvalue of ironIII) x 10; centre of gravity of the iron11 spot; iron111 spot; Itp pe11 = (HF value of ironII) x 10; Ax = x FeIII - x Fe"; AH* = RF Fern - RF FeII Efects of ether and methanol concentrations-With 8 ml of water and 6 ml of concentrated hydrochloric acid, the volumes of ether and methanol were varied as shown in Figs.3 and 4. Methanol contents below about 26 ml could not be investigated, since 26 ml of methanol liesDec. 19571 QUANTITATIVE IKORGANIC CHROMATOGRAPHY. PART I11 783 just above the minimum quantity that produces a homogeneous mixture with 50 ml of ether plus 8 m l of water plus 6 m l of concentrated hydrochloric acid. Except for the time of equilibration, which was 1 hour 40 minutes 5 minutes, and the temperature, which was 17.4" 1- 0.3" C, the experimental conditions were those used in investigating variations in water and acid content.RESULTS- Choice of volume of ether-The curves of composition against distance moved (x) and Iiy (Fig. 3) indicate that the widest separation occurred at 60 and 40 ml of ether, respectively. If the permitted length of run were unlimited, the peak in the Ax curve would have been chosen as giving the best solvent, but since the small size of the apparatus restricted the length of run, 50 ml of ether was chosen as a suitable compromise between these two results. If a solvent with maximum RF separation is required, graphs may be plotted of composition against RF for chromatograms with constant length of run. Choice of volume of methanol-The curves of composition against RF and x (Fig. 4) indicate that the best separation occurred close to 35 ml of methanol, but the maximum separation (in terms of x ) of the adjacent extremities of the spots occurred a t 30 ml of meth- anol.Hence the modified solvent as finally chosen was composed of ether - methanol - water - concentrated hydrochloric acid in the ratio 50 : 30 : 8 : 6 v/v. This method of designing solvents for quantitative separations in which rapidjty is stressed together with maximum RF separation has been applied to other metal valency states.lS The semi-graphical method has also been found useful in selecting solvents for separating complex mixtures of closely related compounds, e g . , the series of sodium silver thiosulphate complexesaZ0 INVESTIGATION OF FACTORS AFFECTIKG THE SEPAKATIOS OF THE FERROUS AND FERRIC IROK EXPERIMENTAL PROCEDURE- Solvent components were mixed in the order: water, acid, ether, alcohol, in glass- stoppered bottles to prevent evaporation, cooled after the additions of the acid and the alcohol, and finally allowed to stand for about 1 hour to attain room temperature.All solvents were freshly prepared. After about 36 hours, the composition of the solvent had altered, probably by esterification, and this resulted in poorer separations. The following standard solutions of ferrous and ferric iron were prepared and used- Ferrous chloride solution (4.95 pg of iron per p1)-FeC1,.4H2O dissolved in 5 per cent. Ferric chloride solution (4.80 pg of iron per pZ)-FeCl, dissolved in 5 per cent. v/v aqueous v/v aqueous hydrochloric acid.hydrochloric acid. Mixed chloride solution-A 1 + 1 mixture of the ferrous chloride and ferric chloride solutions. Ammonium ferrous suL$hate solution (5.0 pg of iron per pL)-(NH4) ,S0,.FeS0,.6H20 dissolved in 0.25 A' sulphuric acid. Ammonium ferric sulplzate solution (5.0 pg of iron per pl)-(NH4) ,S0,.Fe,(S0,),.24H20 dissolved in 0.5 N sulphuric acid. Mixed sulphate solutioH-A 1 + 1 mixture of the ammonium ferrous sulphate and ammonium ferric sulphate solutions. The chloride solutions were used with the general (unmodified) solvent and were standard- ised colorirnetrically with 2-nitroso-1-naphthol-4-sulphonic acid according to the method described in Part 2, p. 795. The sulphate solutions were used with the modified solvent. Strips of Whatman KO. 1 filter-paper (30.5 x 6.5 cm), cut perpendicular to the machine direction, were marked with a starting line 2 cm from one end, and 1.40-pl spots of the appropriate solutions were applied 2 cm apart on this line from a calibrated capillary dropper (see p.789). The apparatus consisted of a gas-jar 30 cm high with an internal diameter of 7-5 cm, two pieces of glass, 14 x 5 cm, ground flat on the 14-cm edges, a cover plate 14 x 9 cm and two elastic bands. The top of the gas-jar was greased and half-covered with one of the pieces of glass, 100 ml of solvent mixture were poured into the jar and the second piece of glass was pressed on to the rim to complete the seal. The cover plate was placed on top and held firmly on the glass slides by the two elastic bands. The liquid was swirled round the sides784 POLLARD, MCOMIE, BANISTER AND NICKLESS : [Vol.52 of the jar to saturate the atmosphere with solvent vapour, care being taken to prevent the solvent from reaching the grease at the top of the jar. After 1 hour's equilibration of the gas-jar, the cover plate was removed and the glass slides were pulled about 1 mm apart. The filter-paper strip was inserted through the gap until the lower edge of the paper just touched the solvent surface. The cover plate was placed over the protruding upper end of the strip and the elastic bands were passed over the ends of the glass slides and cover plate.21 The paper strip was left in position for the required length of time or until the solvent front had reached the position marked on the edges of the strip.I t was then removed and dried. Both the solvent front and the acid front (which could also be detected with indicator) were visible in ultra-violet light on the unsprayed chromatograms as a fluorescent band and a thin line, respectively. The iron spots were detected by spraying the chromatogram with 0-5 per cent. aqueous potassium ferricyanide.s RF values were measured for the positions of highest concentration, gauged visually, the probable error being By using this procedure, except where specifically stated otherwise, the following factors affecting the efficiency of the separation were studied: (a) volume of solvent in the gas-jar; (b) time of equilibration of the gas-jar atmosphere; (c) type and treatment of the paper; ( d ) equilibration of the paper; (e) position of the spots with respect to the edge of the paper; (f) other substances present with the solutes (and hence the drying time) ; (g) distance from solvent level to starting line; (h) length of run; (i) temperature.(a) VOLUME OF SOLVENT IN THE GAS-JAR- Conditions-Modified solvent and sulphate solutions; time of gas-jar equilibration, 65 minutes & 2 minutes; spots dried for 5 to 10 minutes; time of elution, 2 hours 40 minutes Experimertt-Chromatograms were run with 25, 50, 75, 100, 125 and 150 ml of solvent in 0.01 on a 15-cm run. 2 minutes; temperature, T = 17.7" & 0.5" C. the gas-jars. The results are plotted in Fig. 5 . 0 . 9 7 25 50 75 100 125 150 Volume of solvent, ml Variation of RP value with volume of solvent in the gas-jar Fig. 5 .The decrease in RF value observed with increase in volume of solvent was probably due to (i) improved equilibration and (ii) the relatively smaller changes in composition of the solvent caused by loss into the paper during elution and in saturating the atmosphere of the gas- j ar . (b) TIME OF EQUILIBRATION OF THE GAS-JAR ATMOSPHERE- Conditions-As for (a), but T = 19-1" & 0.5" C and 100 ml of the modified solvent were used. Experiment-The equilibration time was varied between 3 and 60 minutes. On increas- ing the equilibration time from 10 to 60 minutes, the RF values decreased slightly, the total decrease over this range (0.02 for iron11 and 0.03 for ironII1) being almost within the experi- mental error. With equilibration times of less than 10 minutes, RF values were variable-Dec.39571 QUANTITATIVE INORGANIC CHROMATOGRAPHS. PART I11 785 depending upon how well the jar was shaken. An equilibration time of 30 to 60 minutes is recommended. The sensitivity of RF values to poor equilibration appears to be due to evaporation from the chromatogram during elution. This applies particularly to the region between the acid and solvent fronts, which contains a high proportion of ether (see p. 787). When the top of the gas-jar was adequately greased, ironII and iron111 had RF values of 0.14 and 0.72, respec- tively, but these rose to 0.20 and 0.87 when the top was ungreased. Therefore, although the separation was little affected, the RF values substantially increased with leakage of solvent vapour from the gas-jar. One of the important features of the ascending-elution apparatus used is that the chrom- atogram can be admitted to the gas-jar with L'ery little disturbance of the atmosphere of the jar.This is not true for the apparatus normally used for descending elution, and to obtain reproducible RF values it is necessary to equilibrate the tank further, and hence the paper, before admitting the solvent to begin elution. Such a procedure is highly undesirable for the separation of iron11 and iron111 owing to their instability (particularly when dry) when in contact with the paper. (c) TYPE AND TREATMENT OF THE PAPER- Coizditions-General solvent and chloride solutions ; spots dried for 10 minutes ; Exfiei+nent-Whatman filter-papers Nos. 1, 3MX, 31, 54, 540 and 541 and acid-washed No.1 (AW.1)21 were investigated with regard to R, reproducibility, speed of running and amount of iron impurity. For constant lengths of run, no significant differences in iron11 and iron111 R, values were observed. 32 > 54 > 3MM > AW.1 > 541 > 1 > 540. KO. 31 paper required 70 minutes for a 15-cm run past the starting line and No. 540 paper, 4 hours 45 minutes. Iron impurities in the paper were detected as a black band at the iron111 position by spraying with oxineZ3 after elution. Visual comparison of the iron bands showed the order of increasing iron content to be- AW.1 < 540 < 54 < 541 < 1 < 3RIM < 31. .4 comparison was made of RF reproducibility of unwashed and acid-washed No. 1 paper. Variations of R, value, especially for ironIII, from one strip to another (cut from the same sheet) were found to be appreciably smaller with sheets of untreated paper than with the acid-washed sheets (particularly when they were freshly washed and dried).A 1-month old batch of washed paper approximated the most closely to untreated paper, and so was used in experiment (d). ( d ) EQUILIBRATION OF THE PAPER- Conditions-General solvent and chloride solutions; time of elution, 3 hours 30 minutes ? 2 minutes; T = 16.7" Experiment-Series of chromatograms were run with 1 hour's equilibration and with no equilibration of the paper both on unwashed and acid-washed Whatman No. 1 filter- paper. Paper equilibration was effected by hanging the spotted and dried strips in covered gas-jars so that the bottom of the strip did not quite touch the solvent surface.After equili- bration, the strips were lowered to start the elution. T = 17.5" & 0.5" C. The speed of running was in the order- 0.5" C. TABLE I EFFECT OF EQUILIBRATION OF THE PAPER ON RF VALUES FOR FERROUS AND FERRIC IRON IN MIXED CHLORIDE SOLUTION The errors quoted show Time of Paper equilibration Unwashed . . . . 0 Unwashed . . .. 1 hour rlcid-washed . . 0 Acid-washed . . 1 hour the maximum deviations from the mean Mean RF value for- , ferric iron ARp acid front ferrous iron 0.876 i: 0.005 0.270 0.020 0.770 k 0.010 0.500 0.830 i: 0.028 0.250 S 0.020 0.720 0.025 0.470 0.870 0.020 0.270 i: 0.020 0.780 0.020 0-490 0.815 i 0.010 0,260 & 0.015 0.725 f 0.015 0.465786 POLLARD, MCOMIE, BANISTER AND NICKLESS : [Vol. 82 The results, which are shown in Table I, are the average values from series of four chromatograms run under identical conditions.Although pure ferric chloride was placed on the centre of the starting line, all chromatograms that had had 1 hour's equilibration of the paper also showed faint spots for ferrous chloride in this position. This is believed to have been due to the reduction of the ferric to ferrous chloride by the paper,Z although it has been suggested that the action of light has this effect.24 In the experiments with initially unequilibrated paper, equilibration must occur during elution and the results in Table ~I suggest a decrease in R, value with increasing rate of elution-other factors, however, also affect the R, value for different lengths of run. The results showed that (i) there was no major difference between the treated and untreated paper, (ii) the best separations occurred on unequilibrated unwashed paper and (iii) a decrease in RF values and separation of ironII and iron111 when the paper was equili- brated; the reproducibility of the results on unwashed paper was not improved.I t was for these reasons that unequilibrated paper was used in the search for the solvent that was to be used for quantitative work on acid-washed paper. (e) POSITION OF THE SPOTS WITH RESPECT TO THE EDGE OF THE PAPER- On runs very close to the edge of the paper, larger RF values were obtained, particularly for iron111 (see Fig. 6). Trumbore and Rogersz5 and Almassy and Dazsoz6 have also stated that RF values may vary with the width of the chromatographic strip.(f) OTHER SUBSTANCES PRESENT WITH THE SAMPLE TO BE CHROMATOGRAPHED- Incomplete drying of the spot on the starting line resulted in increased iron11 RF values and decreased iron111 R, values. This was almost certainly due to the modification of the solvent's properties as it passed over the damp spot on the starting line, as an increase in water concentration in the solvent was found experimentally to have a similar effect. (g) DISTANCE FROM SOLVENT LEVEL TO STARTING LINE- Various i n v e ~ t i g a t o r s , ~ ~ ~ ~ ~ 7 2 8 ~ 3 0 ~ 3 1 particularly in organic chromatography, have demon- strated that R, values vary with the distance between the solvent source and the spot of solute on the starting line.This has usually been attributed to effects of solvent-composition gradients on the chromatogram, particularly of water, and the work now described supports this conclusion. Conditions-Modified solvent and sulphate solutions ; time of gas-jar equilibration, 60 minutes 2 minutes; spots dried for 25 to 30 minutes; d varied between 1 and 10 cm; 1 was 12.0 i 0.2 cm; T = 17.7" Experiment--In these chromatograms, d is the distance from the starting line to the lower end of the strip and 1 is the length of run from starting line to solvent front. The chromatograms were placed in the gas-jars so that the tip of the paper just touched the solvent surface. During elution, the chromatogram swelled so that, at the end of a 15-cm run, the tips of the strips were about 5 mm below the solvent level.The distance, d, is therefore marked off before elution. Very faint backward tailing of iron111 increased in extent and intensity with increasing values of d, so that at d = 8 cm it reached the iron11 RF value, where a definite spot could be discerned. As d increased beyond 4 cm, the points for ironII and ironII1 approached linearity. At d = 1 cm, the R, value of iron111 was not easily reproducible to within the assessed experi- mental error of & 0.01 to 042. For minimum tailing, but maximum separation and repro- ducibility, the best distance from starting line to solvent level was 3.0 cm. However, in quantitative work, where stability of the solutes was more important than high reproduci- bility, d = 2 cm had to be adopted. Descending elution, when d is normally about 3.5 to 5.5 cm, was therefore unsuitable for this work.It is significant that throughout these runs, and also those at different lengths of run, the iron111 spot remained at the acid front. This suggested that little movement of ferric iron took place while the acid-free region (zone 1) ahead of the acid front passed over the iron spot (SO that after any movement in zone 1, the spot was very soon overtaken by the acid front), i.e., in a solvent with the composition of zone 1, the RF value of iron111 is very small. In fact, the RF values of iron11 and ironII1 were shown to be zero in a number of ways- (i) A chromatogram was spotted with ammonium ferrous and ferric sulphates, d was 15 cm and The acid front lay behind the starting line throughout the run and only zone 1 No movement took place.06°C. Results are plotted in Fig. 7. This tailing was probably due to reduction of iron111 to iron11 by the paper. I was marked a t 4 cm. passed over the spots.Fig. 6. Quantitativc chrornatograms photographed in ultra-violet light; starting line marked 2cm from the end of each chromatogram; a fluorescent band reveals the position of each solvent front: A, 25 pg of iron11 25 pg of iron111 (double strip) B, 27.8 pg of iron" + 27.8 pg of ironIIr (triple strip) C, 350 pg of iron" + 250 pg of iron111 D, 100 pg of iron11 7 600 pg of iron111 E, 50 pg of iron11 - 1200 pg of iron111 F, 500 pg of iron" + 100 pg of ironlI1 G, 1250 pg of iron" + 50 pg of ironI1I [To face p. 706Dec. 19571 QUANTITATIVE INORGANIC CHROMATOGRAPHY.PART I11 787 (ii) On another strip (d = 6 cm; I = 3 cm), elution by zone 1 produced no displacement of ferric chloride or ferric sulphate, but a spot of ferric chloride dissolved in a solution containing 5 per ccnt. v/v of hydrochloric acid moved to RF 0 4 . This indicated that neither ferric chloride nor ferric sulphate moved until they were converted to the hydrochloride complex of ferric chloride by the hydrochloric acid in zone 2. (iii) As a final check, 20 g of washed and dried cellulose powder were packed i n a glass column of 2.9 cm internal diameter; the height of the powder in the column was 17.7 cm. Phenol red was dissolved in a portion of the modified solvent mixture to reveal thc position of the acid front during elution. The solvent was poured through the dry column and fractions were collected a t intervals.The indicator was pink in the acid zone and yellow in the neutral zone. The composition of these small fractions (each 0.5 to 1.5 ml) was determined by gas chromatography.a2 The water concentration in zone 1 was found to depend to some extent on the pre-treatment of the cellulose powder. A solvent composed of ether - methanol - water (59 : 33 : 8). as given by the analysis of the eluate immediately before the arrival of zone 2, represented the solvent from zone 1 most likely to result in the movement of iron on a chromatogram (through having the highest methanol and water concentrations), but when a solvent of this composition was investigated, it produced no movement of ammonium ferrous or ferric sulphates (but only in the absence of hydrochloric acid from the original spots).(it) LENGTH OF RUN- In conditions similar to those described in (g) above, RF values were obtained for lengths of run 4 to 20 cm and d = 2 cm (Fig. 8). As discussed in (g), these variations in R, are principally the effect of solvent-composition gradients on the chromatogram. 0- 2 4 6 8 1 0 Distance between solvent level and starting line, crn Fig. 7. Variation of RF value with distance, d , between the solvent level and the starting line; 1 = 12 & 0.2 cm 0.7 - 0.6 - 0.5 0.4 - 0.3 - 0.2 - - cc 0- 4 8 12 16 20 Length of run, crn Fig. 8. Variation of RR~. value with length of run, I ; d = 2 cm In Figs. 7 and 8: A RF = ( I i p value of ironII1) - (Rp value of ironI1) (i) TEMPERATURE- error.Fluctuations in R, over the temperature range 15' to 20" C lay within the experimental THE POSITIONS OF THE AMMONIUM AND SULPHATE IONS By spraying the chromatogram with lead cobalt nitrate - sodium nitrite reagent,% the ammonium ion was found a t R, 0.12, and by spraying with 0.2 N aqueous barium chloride followed by aqueous rhodizonic acid (sodium salt), a white spot appearing on a pink back- ground showed the sulphate ion to be at RF 0.72. On these chromatograms iron11 was at RF 0.17 and iron111 at RF 0.82. With small spots of iron111 (e.g., 10 pg) very faint backward tailing reached its maximum concentration at the position of the sulphate ion. This suggested that the iron111 was giving double spots owing to the competition between the two anions (SO:- and C1-) for the cation (Fe3+), but this conclusion was shown to be incorrect when the replacement of ammonium ferric sulphate by ferric chloride to avoid such competition did not eliminate the phenomenon.788 POLLARD, MCOMIE, BANISTER AND NICKLESS : r ~ 0 1 .82 Since this eK'ect was not serious and was not evident in the quantitative work, the cause was not further investigated. Coxcr.r;srO~s Under the optimum conditions-I00 ml of freshly prepared solvent, ether - methanol - water - concentrated hydrochloric acid (50 : 30 : 8 : 6 v/v) at 17' & 2" C, equilibrated in the apparatus for 60 minutes & 5 minutes, and with spots of ammonium ferrous and ferric sulphates in 0.25 and 0.5 N sulphuric acid, respectively (1.40 pl of solution con- taining 5 pg iron per pl), dried in air for 5 to 10 minutes on strips cut from a single sheet of Whatman No.1 filter-paper, no paper equilibration, d = 2 cm and 1 = 15 cm--the R, values of ferrous and ferric iron were found to be 0.17 and 0.82, respectively. The spots lay within the R, limits 0.08 to 0.22 (ironII) and 0.77 to 0.85 (ironII1). On any one sheet of paper, under these conditions, R, values were easily reproducible to within 0.02, but quite large differences (e.g., within a range -+_ 0.05) were found between chromatograms run under similar conditions, but on different occasions and with different batches of paper. This is evident on comparing the points for similar conditions in the figures showing the results of investigations. When these investigations were repeated, the shapes of these graphs and the conclusions drawn from them remained fully reproducible.Hence, in an investigation into the effect of any experimental variable on a chromatographic separation, it is strongly advisable to cut the strips from the same sheet of paper and to run the chromatograms concurrently. The detailed examination of the solvent under wide ranges of experimental conditions showed its suitability for quantitative determinations of unknown mixtures of ferrous and ferric iron. This examination also showed the unsuitability of descending elution for such work. Part 2. The Colorimetric Determination of Iron with 2-Nitroso-1-naphthol-4sulphonic Acid The ferrous and ferric iron were applied to the paper strip as a band of mixed ammonium ferrous and ammonium ferric sulphates.The time required for elution rarely exceeded 2fr hours and often shorter times of about 1 hour were sufficient. In previous quantitative work on other metaW4 some difficulty had been encountered in designing a chromatogram in which the pilot strip gave a reliable indication of the position of the metal on the strip used for the determination. To avoid any such difficulties, the positions of the iron bands on a quantitative chromatogram were established by holding the chromatogram in a gas-jar of ammonia vapour immediately after elution. Green ferrous hydroxide rapidly turned brown by atmospheric oxidation, and the dark hands of each valency state showed up clearly in ultra-violet light. The iron in the excised bands was extracted with hydrochloric acid and determined colorimetrically with 2-nitroso-I-naphthol- 4-sulphonic acid, sodium salt.Ten to fifteen samples of ferrous plus ferric iron could Ixt completely determined in 5 to 6 hours. The analytical procedure adopted was as follows. The separation was effected by ascending chromatography. EXPERIMENTAL PRE-TREATMEKT OF THE PAPER- Ten sheets of Whatman No. 1 filter-paper, 56 x 23 cm, were steeped in 2& litres of diluted AnalaR hydrochloric acid (1 + 4, V/V) for 1 week, and the process was then repeated once. The inside sheets from the first washing were arranged on the outside for the second. Regular agitation each day, e.g., for approximately ti minutes every few hours, improved the extraction. PREPARATION OF THE SOLUTIONS FOR THE CHROMATOGRAMS- Stock solution of ammonium ferrous sulfilzate (25 mg of iron per ml)-43.890 g of AnalaK ammonium ferrous sulphate, (NH,) ,S04.FeS0,.6H 20, were dissolved in 250 ml of dilute sulphuric acid (containing 25ml of 50 per cent.w/w sulphuric acid), Stock solution of ammonium ferric sulphate (12.5 mg of iron per m1)-Some difficulty was experienced in dissolving sufficient AnaIaR ammonium ferric sulphate,Dec. 19571 QUANTITATIVE INORGANIC CHROMATOGRAPHY. PART I11 789 (NH4)2S0,.Fe2(S0,),.24H,0, to produce a concentration of 25 mg of iron per ml. Therefore, 26.986 g were dissolved in 250 ml of dilute sulphuric acid (containing 25 ml of 50 per cent. w/w sulphuric acid) to give a solution of half the desired concentration. Gravimetric standardisation of this solution by precipitation as ferric hydroxide and ignition to Fe,O, gave 12.50mg of iron per ml.Freshly redistilled water was used in the preparation of both ferrous and ferric stock solutions. Mixed solutions-Standard solutions containing both ferrous and ferric iron were pre- pared by putting suitable volumes of the stock solutions into 60-ml stoppered bottles and diluting to 50 ml with freshly redistilled water. The ratios of ferrous to ferric iron investi- gated and the volumes of the stock solutions required for each standard mixed solution are shown in Table 11. The amount of iron present in 0.1 ml of mixed solution is tabulated for TABLE I1 RATIO OF FERROUS TO FERRIC IRON IN STANDARD MIXED SOLUTIONS FeII in Volume of FeIII in Volume of Volume of mixed ,-A-, 25 mg per ml 12.5 mg per nil solution put on Ratio of 0.1 ml, 50 ml, solution, 0.1 ml, 50 ml, solution, chromatograms, F e I I to Fe"' r g mg ml r g mg ml ml 1 to 1 100 50 2 100 50 4 0.025, 0.5, 0.1 500 250 10 500 250 20 0.1 I to 5 100 50 2 500 250 20 0.1 1 to 10 100 50 2 1000 500 40 0.1 1 to 24 50 25 1 1200 600 48 0.1 1 to 49 50* 12.5 0.5 2450* 612.5 49 0.2 3 to 1 500 250 10 100 50 4 0.1 10 to 1 1000 500 20 100 50 4 0.1 "5 to 1 1250 G25 25 50 25 2 0.1 * In 0.2 ml.each valency state, since this was the volume applied to all chromatograms (except at ratio 1 to 49, when 0.2 ml was used). It would have been possible to have prepared stronger solutions at some of the ratios close to 1 to 1 so that a smaller volume could have been taken for the chromatogram, but this would have unnecessarily introduced another variable.The ferrous and ferric iron were applied to the chromatograms from mixed solutions and not in separate applications in order to simulate the conditions of an actual investigation. The 1 + 1 solution used for the investigation with the capillary dropper contained 5 ml of stock FeII solution plus 10 ml of stock FeIII solution diluted to 20 ml, i.e., 6.25 pg of each valency state per pl. Biftary 1 $- 10 mixtures of iron witla chromium, manganese, cobalt, nickel or copper- Mixtures of iron with a ten-fold excess of a second metal were prepared from the stock standard ammonium ferric sulphate solution and weighed quantities of AnalaR chromic potassium sulphate, Cr2(S04),.K,S0,.24H,0, manganese sulphate, MnS0,.4H20, cobalt sulphate, CoS0,.7H,O, nickel sulphate, NiS0,.7H20 and copper sulphate, CuSO4.5H,0.The iron concentration in each solution was 100 p g of iron per 0.1 ml. DESIGN AND CALIBRATION OF CAPILLARY DROPPERS- I t consisted of a capillary tube of 0.6 mm internal diameter and 7 mm external diameter, joined to a stem of glass tubing 10 cm long, of 5.5 mm internal diameter and 7 mm external diameter. The height to which the solution rose in the capillary was measured and the capillary tube was cut so that a length shorter than the measured rise remained attached to the stem; this length depended upon the capacity required. The tip of the capillary tube was ground to as fine a point as possible. The slight difference between the results by the two methods was attributed to the difference in the shapes of water and mercury menisci and so the result by the first method was used for aqueous solutions. Aqueous calibration-The surface of the standard ammonium ferrous sulphate solution (containing 25mg of iron per ml) was just touched with the tip of the freshly cleaned and dried dropper.The liquid inside the capillary immediately rose to the top and the liquid The design of the capillary dropper has been described elsewhere.2a The capillary dropper was calibrated by an aqueous and a mercury method.790 POLLARD, MCOMIE, BANISTER AND NICKLESS : [Vol. 82 outside the capillary rose a little above the tip, which was then withdrawn slowly from the solution; this minimised the quantity of liquid adhering to the outside of the dropper.Because the tip was ground, the solution on its outside rapidly evaporated and consequently did not interfere with the calibration. The filled dropper was then placed in contact with a piece of acid-washed Whatman No. 1 filter-paper, of area 1 sq. cm, for about 5 seconds to ensure com- plete drainage, Provided there was no entrapped bubble of air in the capillary column, the liquid flowed completely into the paper immediately on contact. This procedure was repeated on three more pieces of paper. The amount of iron on each piece of paper was determined colorimetrically by the procedure described on p. 795, and the results were 56.2, 55.9, 54.9 and 55.5 pg of iron (mean, 55.6 pg). Since the original solution contained 25 pg of iron per pl, the volume of the dropper was calculated to be 2.22 p1 PREPARATION OF THE SOLVENT- The components of the chromatographic solvent-water, concentrated hydrochloric acid, ether and methanol (8 : 6 : 50 : 30)-were added in that order to a glass-stoppered bottle, which was cooled after the addition of the acid and after the final addition of alcohol.Immediately after mixing the temperature usually rose to 28" to 30" C, making it necessary to prevent evaporation by cooling to room temperature in a closely stoppered bottle. Between 75 and 100 ml of solvent were poured into each gas-jar, which was left for 30 to 60 minutes to come to equilibrium before the chromatogram was put in. No temperature control was necessary between 15" and 20" C, though above 20" C forward tailing of ferric iron at the edges of the chromatogram started to be troublesome.0.76 per cent. cHROMATOGlL4NS- Chromatograms of three designs were used in the quantitative work (see Fig. 6). (i) A paper strip, 30.5 x 6.6 cm, was marked with a starting line 2 cm from one end. Starting 1 cm from this end, two slots, 15 x 0.3 cm, were cut out of the paper so that for 15 cm it was equally divided into three strips, 15 x 2 cm. This type of triple-strip chromzto- gram was used for work in which the spots were placed on the starting line from a capillary dropper. Two of the strips were used for a separation and the third strip was used as a blank. (ii) A paper strip was marked and cut in the same way except that only one control slot TI as cut, leaving two strips, 15 x 3.15 cm.This type of double-strip chromatogram was used for separations of solutes contained in less than 0.05 ml. One strip was used for the separa- tion and the other for the blank. (iii) A paper strip, 30.5 x 6.5 cm, with the starting line marked 2 cm from one end. This type of chromatogram was used for all separations involving volumes of solution from 0.05 to 0.2 ml. The blank was run on a separate strip. METHODS OF DETERMINING AND EXTRACTING SUBSTANCES FROM FILTER-PAPER CHROMATOGRAMS There are two main procedures used for the determination of substances after elution in paper chromatography : ( 2 ) extraction techniques, followed by standard microanalytical procedures and (ii) in situ determinations. Extraction techniques are generally accounted to be the more accurate, although some in situ light-absorption methods have an accuracy approaching extraction procedure^.^^ Lacourt, Sommereyns and Wantie96 quote 4 per cent.as the expected accuracy of the spectrophotometric determination, with disodium cdtechol-3 : 5-disulphonate, of 10 pg of iron that has been chromatographically separated from other metals on untreated Whatman No. 1 filter-paper and extracted for 90 minutes with 0.25 per cent. w/v sulphuric acid. On account of the accuracy required in the work described in this paper, an extraction method was chosen out of the many that were investigated, EXPERIMEKTAL- For the preliminary investigation into the efficiency of extraction, the iron was deter- mined by measuring the red colour produced by ferric iron in alkaline medium of pH 7 to 10 with disodium catechol-3 : 5-di~ulphonate.~' The disadvantages of this method were (a) its low sensitivity and (b) the need for oxidation of iron in the ferrous state to ferric.For these reasons, attention was paid to the reagent 2-nitroso-1-naphthol-4-sulphonic acid, described in the next section on the colorimetric determination of iron. Various aliquots of a solution containing 25 to 250 pg of ferric iron, delivered from a inicrometer syringe, were absorbed into sections of acid-washed Whatman No. 1 filterpaper,Dec. 19571 QUANTITATIVE INORGANIC CHROMATOGRAPHY. PART I11 791 of area 25 sq. cm, and allowed just to dry. The paper was then treated by the extraction method under investigation. Cold leaching in 15 ml of 1 per cent.v/v acid (but using AnalaR hydrochloric acid instead of formic acid) and washing with an equal volume of water after the method of Martin38 gave very low results-approximately 50 per cent. recovery on 100-pg samples. Micro Soxhlet extraction3e was investigated, glass crucibles fitted at one end with Jena No. 1 glass frits being used in an all-glass apparatus, since previous work had established the presence and difficult removal of many heavy metals in trace quantities in micro-extraction thimbles. Extraction with 1 per cent. v/v hydrochloric acid or 2.5 per cent. v/v sulphuric acid for up to 3 hours gave low results. Complete wet digestion of the paper by conventional means for destroying organic materia140 usually gave quantitative recovery. A disadvantage of the wet-ashing technique is the long time required to complete the digestion (about 3 to 4 hours) in order to obtain white ash residue. Ashing the paper in a crucible was not tried, since it has been reported that this leads to low results.Kolier and R i b a ~ d o , ~ ~ when determining iron by paper chromatography, used an elevated-temperature leaching procedure, which we investigated, modified and adopted. This modified procedure was used in conjunction with both colorimetric methods of analysis (that with disodium catechol-3 : 5-disulphonate and that with 2-nitroso-1-naphthol-4-sulphonic acid) and the recovery was 99 to 101 per cent., even on 25-pg samples. With each batch of determinations, an equal area of similarly purified paper was used in a blank determination.The order of magnitude of the iron left in the paper after this leaching process was O-lpg per sq. cm. THE COLORIMETRIC DETERMINATION OF IRON The grouping -CO-C(=NOH)- as contained in the monoximes of diketo compounds is of considerable analytical importance, since it readily forms inner complex salts with many heavy-metal salts.42 These complexes are generally only partially soluble in water, and extraction into an immiscible organic solvent is necessary before colorimetric determination. When the grouping is incorporated into an aromatic ring system, it still retains the ability to form inner complex salts, and when sulphonic acid groups are substituted into the aromatic ring of such a complexing agent, the product so formed (and its metal complexes) is generally more water-soluble.Nitroso-R salt (disodium 1-nitroso-2-hydroxynaphthalene-3 : 6-disul- phonate) is an important reagent for the spectrophotometric determination of cobalt,43144 and has also been used for the colorimetric determination of ferrous iron.45946 The reaction between ferrous iron and nitroso-R salt was critically studied by Griffing and Mell~n?~ who found that the sensitivity was much greater than with 1 : 10-phenanthroline, but that the method suffered from three main disadvantages: (i) the intense yellow-green colour of the reagent; (ii) its sensitivity to changes of pH and (iii) the time required for full development of the green colour of the complex. SPECTROPHOTOMETRIC STUDY OF THE REACTION OF FERROUS ASD FERRIC IRON This method was therefore rejected.WITH 2-NITROSO-l-NAPHTHOL-4-SULPHONIC ACID Many coloured complexes are formed in reactions between iron, cobalt, nickel and copper salts with the various nitrosonaphtholmonosulphonic acids.48 ,49,50. One of these acids, 2-nitroso-1-naphthol-4-sulphonic acid, was found to give a green coloured complex with iron (Naphthol green G). Sarver51 reported that 2-nitroso-1-naphthol-4-sulphonic acid formed complexes with ferrous and ferric iron, and he recommended the reagent for the spot-test detection of iron, cobalt, nickel and copper. Its use as a chromatographic spray reagent for a number of cations, including iron, has recently been described.= The reagent has been used in the spectrophotometric determination of cobalt,52 and two complexes are reported to be formed with nickel, depending upon the pH of the solution.53 No reference to work on the spectrophotometric determination of iron with this reagent has been found.Cronheim,u@ while studying complex formation by o-nitrosophenol and its substituted derivatives, found that grass-green and brown complexes were formed with ferrous and femc iron, respectively, but the formation of the ferric complex was not quantitative. In the study to be described, both ferrous and ferric iron gave deeply coloured complexes with 2-nitroso-1-naphthol-4-sulphonic acid, in confirmation of Sarver’s observation, but the intensity of the colour with ferric iron was found to depend on the volume of complexing792 POLLARD, MCOMIE, BANISTER AND NICKLESS: [Vol. 82 agent present in excess of the theoretical requirement.Consequently, before the reagent was added, all iron was reduced to the ferrous state with hydroxylamine in acid solution. This produced no deterioration in sensitivity, and the procedure given later (p. 795) remains general for the spectrophotometric determination of iron. Further advantages of using the lower valency state complex for the determination are that (i) the ferrous extract from the chromatogram does not require oxidation, while reduction to the ferrous state can be carried out in the cold, and (ii) the reducing “atmosphere” of the solution ensures the stability of the reagent. In the following sub-sections are included details of the reaction between 2-nitroso-l- naphthol-4-sulphonic acid and both ferric and ferrous iron, and the effects on the recom- mended procedure of pH, reagent concentration, the nature of the anion, concentration of the reducing agent, time of colour development and the iron concentration at the stage of determination.The molecular ratio of complexing agent to ferrous iron has also been determined by two methods. EXPERIMENTAL- with 10-mm glass cells fitted with lids. SOLUTIONS USED- Standard solution of ferric nitrate (100 pg of irolz per m1)-Pure iron wire was dissolved in dilute nitric acid by warming, and the resulting solution was filtered and standardised gravimetrically by precipitation of the hydroxide and ignition to ferric oxide. An appropriate aliquot of this solution was diluted to 1 litre in a calibrated flask. Standard solution of ammonium ferric sulphate (100 pg of iron per ml)--0.8634g of AnalaR ammonium ferric sulphate, (NH,) 2S0,.Fe2(S0,) ,.24H20, was dissolved and made up to 1 litre in water containing 1 ml of concentrated sulphuric acid, sp.gr.1.84. All optical-density measurements were made with a Unicam SP500 spectrophotometer, Hydroxylamine hydrochloride-A 5 per cent. w/v aqueous solution. Acetic acid, 0.20 M. Sodium acetate solution, 0.20 M. Sodium hydroxide solution, 2 M . Complexing reagent-0.247 g of sodium 2-nitroso-1-naphthol-4-sulphonate was dissolved in a mixture of methanol and water (1 + 4) and made up to 100 ml in a calibrated flask. A very slight brown residue remained, which was filtered off in a Jena No. 3 sintered-glass crucible. EFFECT OF pH ox REAGENT AND COMPLEX- With a 1.0-ml pipette, 1.0 ml of complexing reagent was transferred to a 25-ml calibrated flask and 10.0 ml of buffer were added to produce the required pH; at pH 4, 4.5, 5.0 and 5.5, the volumes of 0.20 M acetic acid and 0.20 M sodium acetate solution added were 8.0 and 2.0, 6.0 and 4.0, 3.0 and 7.0, and 1.0 and 9.0, respectively. The mixture was diluted to 25 ml with distilled water and the optical density was measured against a blank of distilled water.The optical density at peak absorption rose sharply with increase in pH and so, for buffer solutions at pH greater than 7 (Table 111), only 0.5 ml of the reagent solution was added. TABLE I11 BUFFER SOLUTIONS FOR PRODUCING pH VALUES GREATER THAN 7 PH Buffer 7.4 9.0 ml of 0.20 M boric acid and 1.0 ml of 0.05 M sodium tetraborate solution 9.5 10.0 ml of a solution containing 20.0 g of sodium hydrogen carbonate and 11.0 10.0 ml of a 3 + 7 mixture of AnalaR ammonia solution, sp.gr. 0.880, and water &laximum optical density occurs at 425 i. 5mp, and the absorption at this wavelength rises to a maximum at approximately pH 8. The optical density is almost zero after 520mp, so that the colour of the excess of reagent has very little effect on the determination, since maximum absorption of the complex occurs at 700 mp. 10.0 g of anhydrous sodium carbonate per litre The results are shown in Fig. 9.Dec. 19571 QUAKTITATIVE INORGANIC CHROMATOGRAPHY. P A m III 793 0.9 ~ 0 . 8 - .- Lo C W 0 U CI Q 0.7 .- 0.6- Fig. 9. Absorption spectra of reagent and of iron complex a t various pH values: curves A, B, C and D, reagent at pH 4.0, 5.0, 6.5 and 9.5, respectively; curves E, F and G, 40 pg of iron, as nitrate, a t pH 6.0, 9.5 and 11, respectively; curve H, 40 p g of iron, as sulphate, at pH 5.0 , VARIATION OF ABSORPTION SPECTRA OF THE FERROUS COMPLEX WITH VARIATION I N pH- A 1.0-ml portion of standard ferric nitrate solution, 2.0ml of 5 per cent.hydroxyl- amine hydrochloride solution, 1.0 ml of complexing reagent solution and 10 ml of buffer were diluted to 25 ml in a calibrated flask with distilled water. The components were added in that order and the final solution was deep green: the buffer solutions used are listed above. The optical densities of the solutions were measured at various wavelengths against a similarly prepared reagent blank.lOmp., and at pH 5.0 & 0.5 (Fig. 9). The maximum optical density occurred at 700[Vol. 82 794 POLLARD, MCOMIE, BANISTER AND NICKLESS: EFFECT OF REAGENT CONCENTRATION- Provided sufficient reagent was present to complex all the ferrous iron, excess of reagent produced only a very slight rise in optical density. Various volumes of the reagent solution were added to a standard quantity of iron (100 pg in 25 ml of ammonium ferric sulphate solution) ; this concentration was the upper concentration limit in the determinations. Molar concentration of reagent . . 3.6 x 7.2 x 10.8 x 10-4 Optical density a t 700 mp . . . . 1.720 1.725 1,725 EFFECT OF TIME OF DEVELOPMENT- instantaneous and no alteration in optical density was noted over 24 hours.The development of the colour formed by ferrous iron and the reagent appeared to be EFFECT OF HYDROXYLAMINE HYDROCHLORIDE COXCENTRL4TION- Little effect was produced by a reasonable excess of the reducing agent (5 ml of 5 per cent. solution), but a t very high concentrations (10 ml of 20 per cent. solution) results were low. With the volume used in the recommended procedure, the excess of hydroxylamine produced no noticeable effect. ORDER OF ADDITION OF REAGENTS- The rate of colour development was slower when the reagent solution was added after the buffer solution, so that the recommended order for the addition of reagents to a weakly acid solution of iron is reducing agent, reagent solution and buffer, followed by dilution to the correct volume. The iron extracts from the paper chromatograms were highly acid and had to be neutralised to approximately the correct pH before addition of the buffer.Recovery was quantitative if, after the addition of hydroxylamine hyrochloride solution and a drop of 0.05 per cent. w/v methyl orange indicator solution, 2 M sodium hydroxide was added dropwise from a capillary pipette until the solution just turned yellow (pH 5 ) . Fig. 11. Mole-ratio method applied t o the iron complex at 700mp STRUCTURE OF THE COMPLEX- The nature of the complex was initially determined by the mole ratio method of Yoe and Jones.37 In this method, the variation of optical density at a constant wavelength is observed for a series of solutions containing different molecular ratios of metal to reagent. The iron concentration chosen was 7.17 X Various volumes of 2- nitroso-1-naphthol-4-sulphonic acid solution were added to the iron solution, by means of an Agla micrometer syringe, to cover the range of ratios 0 to 5 moles of reagent per mole of iron.The optical densities of these solutions were measured a t 700 mp against a similarly prepared reagent blank. The results are shown in Fig. 11, which indicates that in the complex 1 mole of iron combines with 3 moles of reagent. The same result was obtained by the slope- ratio method.S6 (100 pg of iron per ml).Dec. 19571 QUANTITATIVE INORGANIC CHROMATOGRAPHY. PART I11 i 9 5 EFFECT OF THE ANION AND THE VALENCY STATE OF THE IRON- When the absorption spectrum of the complex was determined with ammonium ferric sulphate solution (containing 100 pg iron in 25 ml) and the colorimetric procedure described earlier under “Effect of pH on reagent and complex,” the peak height was appreciably above that found for ferric nitrate solution (Fig.9). This was thought to be due to the presence of excess of nitrate in the latter solution and its subsequent interference in the colorimetric determination. Provided excess of the complexing reagent was present, the absorption spectrum of the complex formed with ferric iron was the same as that obtained with a similar concentration of ferrous iron. When sufficient complexing reagent for a 1 to 3 complex was added to ammonium ferric sulphate, the optical density of the mixture did not reach the value that was produced with a corresponding concentration of ferrous iron.When the amount of reagent added was sufficient for a ratio of 1 to 4, full colour developed in about 10 minutes. At 1 to 5 or lower ratios, the colour developed almost immediately. Since excess of complexing reagent gives complexes with both ammonium ferrous and ammonium ferric sulphates at pH 5 that have identical absorption spectra at similar concentra- tions of iron, it suggests that the valency state of the central metal atom is the same. During the study of the reactions between ferrous and ferric iron and isonitrosodimethyldihydro- resorcinol (isonitrosodimedone), similar observations were made by Sh0me.~7,58 Cr~nheirnM?~~ found that o-nitrosophenol reduced ferric to ferrous iron in strong light. Because of the excess of 2-nitroso-1-naphthol-4-sulphonic acid required for full colour development with ferric iron, it may be that this oxidation - reduction reaction occurs here also.In the deter- minations of ferric iron described later, hydroxylamine was added to the solution immediately before the addition of the reagent. This minimised the excess of reagent required for full colour development and hence reduced the blank reading. PREPARATION OF CALIBRATION GRAPH- Various volumes of the ammonium ferric sulphate solution were transferred to 25-ml calibrated flasks, so that they contained between 0.0 and 4-0 pg of iron per ml. To each was added 2.0 ml of 5 per cent. hydroxylamine hydrochloride solution, 1.0 ml of complexing reagent solution, 3-0 ml of 0.20 M acetic acid and 7.0 ml of 0.20 M sodium acetate solution.The solutions were made up to volume, and their optical densities were measured at 700 mp against a similarly prepared reagent blank. Beer’s law was obeyed over the range studied, and the slope of the calibration graph was 0.428 optical-density units per pg of iron per ml. Ten determinations on 4.0 pg of iron per ml and 0.50 pg of iron per ml gave an average of 3.98 pg and 0.50 pg of iron per ml, respec- tively. On comparing the sensitivities of the reactions of 2-nitroso-1-naphthol-4-sulphonic acid, nitroso-R salt, 1 : 10-phenanthroline and disodium catechol-3 : 5-disulphonate by the method of Woods and M e l l ~ n , ~ ~ we have- 2-Nitroso- l-naphthol- Disodium Nitroso-R 4-sulphonic 1 : 10-Phen- catechol-3 : 5- salt acid anthroline disulphonate Iron required for 50 per cent. transmission for I-cm cells, using maximum absorption of curve, p.p.m.. . .. * . 0.67 0.71 1.7 2.9 While not quite so sensitive as nitroso-R salt, 2-nitroso-1-naphthol-4-sulphonic acid has the advantages that the reagent does not absorb at the wavelength used for the determination of the iron and that colour development is instantaneous. Since the chromatographic separation gave pure iron solutions, a study of the effects of other ions on the reagent was not undertaken. A short spectrophotometric study of the reaction between the complexing reagent and cupric copper was carried out, but the slope of the calibration graph under the optimum conditions for absorption was only 0.181 optical- density units per pg of copper per ml. PROCEDURE At least two chromatographic separations were carried out on each of the following mixtures: (a) 0.025 ml of the 1 to 1 solution described in Table I1 on a double-strip chromato- gram (i.e., 25 pg of each valency state); (b) 0.00445 ml (2 spots from the 2-225-pl capillary796 POLLARD, hlCOMIE, BANISTER AND NICKLESS : [Vol.81 dropper) of the 1 + 1 solution used for the investigation with the capillary dropper (see p. 789) on a triple-strip chromatogram (27.8 pg of each valency state); (c) 0.05 ml of the 1 to 1 solution described in Table I1 on a single-strip chromatogram (50 pg of each valency state) and (d) 0.1 ml of all solutions described in Table I1 (except that 0.2 ml of the 1 to 49 solution was used) on single-strip chromatograms. All volumes of 0.025 ml or greater were placed on the chromatograms from an Agla micrometer syringe.Volumes of 0.1 ml were placed on the starting line in two applications of 0.05 ml. The first band was allowed to dry for 10 minutes before the second was applied. It was found that the R, for ferrous iron increased and the R, for ferric iron decreased with the time of drying of the second band. The importance of time of drying in quantitative work has also been stressed by Lacourt, Sommereyns, Stadler-Denis and Wantier in the separation of chromium and molybdenum.BO The minimum times of drying necessary for complete separation, shown in Table V, were found by experience and are only intended as a rough guide. An upper limit to the time of drying was also found to exist owing to the decreased stability of each valency state on dry paper.This limit was not critical except at extreme ratios of ferrous to ferric iron. On the double-strip chromatograms (25 pg of each valency state), 0.025 ml of the mixed solution was placed on the starting line in two applications of equal volume; each band was allowed to dry for 3 to 4 minutes. On the triple-strip chromatograms, the first spot from the capillary dropper was dried for 2 to 3 minutes and the second spot for 3 to 4 minutes. At the ratio 1 to 49, 0.2 ml of solution was applied to the starting line in three applica- tions: 0.08 ml, then 13 minutes’ drying; 0.08 ml, then 10 minutes’ drying; andfinally0.04 ml. Elution was begun 15 minutes after this final application. The separations were usually carried out in batches of about eight with one blank chrom- atogram.If possible, each chromatogram in a batch was cut from the same sheet of acid- washed paper. Lengths of paper 6.5 cm wide, but between only 20 and 30 cm long, could still be used after being joined with adhesive tape to unwashed paper. The chromatograms were placed in the ascending-elution apparatus21 and removed after a length of run of between 10 and 18cm. The most suitable lengths of run, as found by experience, are shown in Table V. Eluted chromatograms were immediately held in a gas- jar charged with a few millilitres of ammonia solution, sp.gr. 0.880. This caused the iron to be deposited as hydroxide; ferrous hydroxide rapidly oxidised to ferric hydroxide. The positions of the iron bands were marked in ultra-violet illumination (see Fig.6). In any one batch in which similar or slightly differing ratios were being investigated, the sizes of the iron bands were measured and, for each valency state, a similarly sized strip of paper was cut from the quantitative chromatogram and from the blank with a new pair of heavily plated scissors. The excised bands were cut into 4 or 6 pieces so as to fit easily into 100-ml beakers, These beakers, covered with watch-glasses, were then ready for the acid extraction. CHROMATOGRAMS OF IRON WITH OTHER METALS- The general procedure was as for the valency-state separations. Two 0.05-ml aliquots of the mixed solution were applied to the chromatogram to give a total of 0.1 ml, each band being allowed to dry for 10 minutes.The length of run each time was about 15 cm in 90 minutes. After the eluted chromatogram had been held in ammonia vapour, the dark bands of chromium, iron, cobalt and copper showed up clearly in ultra-violet light. With each metal, a band of paper 6.5 x 5-0 cm cut from the chromatogram was sufficient to remove the iron for determination. After removal of iron, the nickel was detected by a rubeanic acid spray. The usual reagents for manganese, e.g., oxine, are not very satisfactory, so a new spray reagent was sought. The best found so far consisted of 4 per cent. v/v of salicylalde- hyde in 1 + 1 aqueous ethanol. After the spraying, the paper was held in ammonia. The band for manganese appeared brown on a yellow background, but was clearer in ultra-violet illumination as dark brown on a brightly fluorescent blue-green background.Other metals (about 10 pg in a 7-mm spot) that gave reactions differing from that of ammonia alone were copper (yellow-green), silver (light brown), lead and nickel (yellow), vanadium as V I V or V V and manganese (grey-brown), chromium (faint yellow-green), ferrous and ferric iron (different shades of brown) and cobalt (pale yellow-brown). In ultra-violet illumination all these metals except silver and chromium appeared as dark spots on the fluorescent background.Dec. 19571 QUANTITATIVE IXORGAKIC CHROMATOGRAPHY. PART I11 TABLE IV RESULTS OF DETERMIXATIONS ON A SINGLE BATCH OF ACID-WASHED PAPER Ratio of FeII to FeII' I t 0 1 1; 1 t d 5 1 to 10 1 to 24 1 to 49 Iron present as 7- FeII, FeIII, PLg I-Lg 25* 25* 27.87 27.8t 50 50 100 100 500 500 100 500 100 1000 50 1200 50 2450 FeII, PLg 24.6 27.5 50.8 99.0 499.0 100.5 99.4 49.5 49.7 FeIII, Ilg 24.8 28.0 49.5 99.8 501.5 494.0 1002.0 1217.0 2401.0 Iron found as FeII, FeIII, Pg Pg h 24.6 24.3 27.7 28.3 51.0 49.9 99.4 100.3 497.0 497.0 101.0 497.0 100.6 1005.0 49.7 1221.0 51.0 2440.0 FeII, rg - - 49.9 99.5 504.0 101.2 99.3 51.0 - 5 t o l 500 100 512.0 101.0 505.0 101.0 495.0 l O t o l 1000 100 994.0 101.4 982.0 103.0 997.0 25to 1 1250 50 1238.0 51.0 1238.0 51.5 1230.0 * Double-strip method. t Using a capillary dropper and the triple-strip method.FeIiI, PLg - - 49.2 99.3 504.0 490.0 995.0 1225.0 101.0 101.8 51.8 - 797 Mean errors 1.6 1.8 0.72 1.3 1.3 0.93 0.70 0.40 0.53 0.57 0.90 1.3 0.63 0.40 1.2 1.8 1.3 1.2 1.5 1.0 0.90 2.1 1.2 2.9 When analyses of unknowns are carried out, the ferrous to ferric ratio is often known approximately, and Table V suggests experimental conditions that will result in a satisfactory separation.TABLE V RECOMMENDED EXPERIMENTAL CONDITIONS FOR VARIOUS RATIOS OF FEII TO FEIII Amount of iron, Pg 25 50 100 500 1000 1250 25 50 100 500 1000 1200 2450 Ratio 1 to 24 1 to 49 5 to 1 10 to 1 25 to 1 1 to 1 1 to 1 25 to 1 1 to 1 5 to 1 10 to 1 1 to 1 1 to 5 1 to 10 1 to 24 1 to 49 i Minimum time of drying of applied band for separation, minutes 2 6 6 10 (0.2 ml) 6 6 6 8 8 8 8 2 6 8 6 8 8 8 6 6 6 10 (0.2 ml) Minimum length of run, cm 10 15 17 18 15 16 16 17 17 17 17 10 15 17 15 17 17 17 16 16 17 18 Maximum RP of leading edge of FeI1* 0.40 0.35 0.35 0.40 0.40 0.40 0.40 0.40 0.45 0.40 0.45 Minimum RF of rear of FeIII 0.56 0.65 0.50 0.55 0.50 0.50 0.45 0.45 0.45 0.40 0.40 Minimum size of band to be cut out,? cm 5.0 6.0 6.0 6.5 6.5 6.5 6.5 6.5 7.0 7.0 7.5 3.0 3.0 5.5 4.5 5.0 6.0 7.0 5.0 8.5 8.5 10.5 * At the recommended minimum time of drying and length of run.t I n all cases (except 25 pg to 25 pg, where width of paper is 3 cm), the width of the paper is 6.5 cm, and so the area of paper cut out is 6.5 times the value given in this column. Except in a few instances (as shown in Table 11), the volume of solution applied t o the starting line was 0.1 ml.798 POLLARD, MCOMIE, BANISTER AND NICKLESS : [Vol. 82 EXTRACTIOX- To each iron-bearing section of the chromatogram was added 10 ml of 50 per cent. v/v AnalaR hydrochloric acid, and the solution was boiled for 1 minute on an electric hot-plate.This extract was poured into another beaker; the extraction was repeated with 10ml of 25 per cent. v/v hydrochloric acid, followed by 10 ml of distilled water. The three extracts were combined, filtered through a Jena No. 3 sintered-glass funnel and evaporated to a few drops. It was then ready for spectrophotometric determination. SPECTROPHOTOMETRIC DETERMINATION OF THE EXTRACTS- The few drops of solution were transferred to a 25-ml calibrated flask, together with the washings, and 2.0 ml of 5 per cent. hydroxylamine hydrochloride solution were added. Then 1 drop of 0.05 per cent. w/v aqueous methyl orange was added and 2 M sodium hydroxide was run in dropwise from a capillary pipette until the indicator turned yellow.Finally, 1.0 ml of 0.247 per cent. w/v 2-nitroso-1-naphthol-4-sulphonic acid in 1 + 4 methanol - water mixture, 3.0 ml of 0.20 M acetic acid and 7.0 ml of 0.20 M aqueous sodium acetate solution were added. The solution was made up to volume and the optical density was measured at 700 mp against a similarly prepared chromatogram blank. RESULTS The errors tabulated are an indication of the total error from the whole method-the solutions, the elution, impurities in the paper and reagents, removal of the iron, the extraction and colori- metric determination. One blank was run with each batch of chromatograms, and Table VI gives the blank readings in the ferrous and ferric positions for ten consecutive sets of runs. For the first five Table IV shows the results obtained on a single batch of acid-washed paper.TABLE VI BLANK READINGS ON ACID-WASHED WHATMAN No. 1 FILTER-PAPER FeII r h I Width Area of paper Total iron of strip, cut out, in blank, cm sq. cm Pg 2 5.4 3.9 17.05 5.8 13.95 3.7 14-55 4.8 4.9 1.2 3.1 2.0 4.9 39.0 0.88 45.5 3.7 6.5 FeIII r A > Area of paper Total iron cut out, in blank, sq. cm pg 6.0 0.76 9.3 5.6 10.85 5.7 9.3 4.8 7.75 2.5 32.5 2.6 32.5 3.2 39.0 2.9 45.5 1.1 32.5 3.1 TABLE VII DETERMINATION ON OTHER TYPES OF PAPER Ferrous iron Ferric iron > c r A h \ Whatman Length Width Blank Width Blank filter- of FeII of reading, FeIII of reading, paper run, Time of found, Error, Blank, band,* pg per found, Error, Blank, band,* pg per S O . cm elution pg % pg cm sq.cm pg % pg cm sq. cm 1 16.2 2h. 30m. 240.0 4.0 23.0 4.1 0.86 232.5 7.0 28.0 4.2 1.03 54 16.1 2 h . 30m. 242.5 3.0 14.8 5.0 0.46 2486 0.6 16.4 3.7 0.68 540 17.6 211. 55m. 248.5 0.6 6.7 5.0 0.22 247.5 1.0 8.9 4.4 0.31 * Width of the band cut from the paper after elution. The “length” of the band is 6.5 cm (i.e., the “width” of the chromatogram is 6.5 cm). runs (top five lines of results), the average is 0.39 and 0.45 pg of iron per sq. cm in the ferrous and ferric bands, respectively, and the second five runs average 0.062 and 0.071 pg of iron per sq. cm. These blank readings include the blank of the colorimetric method and so, whenDec. 19571 QUANTITATIVE INORGANIC CHROMATOGRAPHY. PART I11 799 expressed in pg of iron per sq. cm of paper, the values decrease with increasing area of paper cut from the chromatogram. The work described in Part 1 had indicated that Whatman Xos.54 and 540 filter-papers were the most free from iron, and so a separation of 250 pg of each valency state was carried out and a comparison made with Whatman No. 1 filter-paper. The volume of solution applied to the starting line was 0.05 ml and the time of drying of the band was about 6 minutes (band appeared just damp). From the results in Table VII it can be seen that, when an accuracy of within 1 or 2 per cent. is acceptable, Whatman No. 540 filter-paper is satisfactory for the separation of about 250 pg of each valency state. SEPARATION OF 100 pg OF FERRIC IRON FROM 1 mg OF ANOTHER METAL- Table VIII shows the amount of iron found by this procedure in mixtures of 100 pg of ferric iron with 1 mg of chromium, manganese, cobalt, nickel or copper.Also shown are the approximate R, value for the leading edge of the second metal and the R, separation between this leading edge and the rear edge of the ferric iron band. A blank was run with each type of paper. TABLE VIII SEPARATION OF IRON AND OTHER METALS FeI** found (100 pg taken), Second metal PLg Cobalt . . .. . . 99.5, 100.2 Chromium , . . . 100.0, 100.6 Manganese . . . . 99.0, 100.3 Nickel . , .. .. 100.5, 100.3 Copper . , . . . . 100.0, 100.4 RF of leading edge of second metal (k 0425) 0.25 0.375 0.275 0.26 0,425 R p separation between second metal and iron 0.20 0.075 0.175 0.20 0.025 CONCLUSION By using a simple ascending-elution technique, it was found possible to determine both ferrous and ferric iron in ratios between I to 49 and 25 to 1.The method showed maximum accuracy at ratios close to 1 to 1 and 100 to 500 pg of each valency state, and for a single application of 0.05 ml of the unknown solution to a 6.5-cm wide chromatogram. Two of the authors (A. J.B. and G.N.) thank the Department of Scientific and Industrial Research for Maintenance Grants. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. REFERENCE Pollard, F. H., McOmie, J. F. W., and Banister, A. J., Chem. G. Ind., 1955, 1598. Pollard, F. H., McOmie, J. F. W., and Elbeih, I . I . M., J . Chem. Soc., 1951, 470. Harasawa, S., and Sakamoto, T., J . Chenz. SOC. Japan (Pure Chem. Section), 1952,73, 611. Bighi, C., Ann.Chim., 1955, 45, 1087. - , Ibid., 1955, 45, 532. Harasawa, S., and Sakamoto, T., J . C h e w SOC. Japan (Pure Chem. Section), 1952, 73, 300. ____ , Ibid., 1952, 73, 240. Polldrd, F. H., Brit. Med. Bull., 1954, 10, 187. Anderson, J. R., and Martin, E. C., Anal. Chim. Acta, 1955, 13, 253. Elbeih, I. 1. M., McOmie, J. F. W., and Pollard, F. H., Disc. Faraday SOC., 1949, 7, 183. Pollard, F. H., McOmie, J. F. W., and Elbeih, I. I . M., Nature, 1949, 163, 292. Nicholas, D. J. D., and Stevens, H. M., Ibid., 1955, 176, 1066. Candela, M. I., Hewitt, E. J., and Stevens, H. M., Anal. Chim. Acta, 1956, 14, 66. Stevens, H. M., Ibid., 1956, 14, 126. Cowan, M. R. (Miss), and Foreman, J. K., Chem. 6 Ind., 1954, 1583. Lederer, M., Anal. Chim. Acta, 1951, 5, 185. - , Ibid., 1950, 4, 629. Pollard, F. H., McOmie, J. F. W., Martin, J. V., and Hardy, C. J., J . Chem. Soc., 1955, 4332. Pollard, F. H., McOmie, J. F. W., and Banister, A. J., unpublished work. Miller, M. C., B.Sc. Thesis, University of Bristol, 1956. Pollard, F. H., and Banister, A. J., Anal. Chim. Acta, 1956, 14, 70. Heisig, G. B., and Pollard, F. H., Ibid., 1956, 16, 234. Pollard, F. H., and McOmie, J. F. W., “Chromatographic Methods of Inorganic Analysis: with Special Reference t o Paper Chromatography,” Butterworths Scientific Publications Ltd., London, 1953, p. 53. Tamura, Z., and Ashikawa, R., Japan Analyst, 1954,3, 475.800 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. CLINCH AND GUY: THE EXTRACTION AND ABSORPTIOMETRIC p o l . 82 Trumbore, C. N., and Rogers, H. E., J . Chem. Educ., 1952, 29, 404. Almissy, G., and Dazso, I., Magyar Kim Folydirat, 1955, 61, 158. Burma, D. P., J . Indian Chem. Soc., 1951, 28, 631. Kowkabany, G. N., and Cassidy, H. G., Anal. Chem., 1952,24, 643. Zimmerman, G., Z. anal. Chem., 1953, 138, 321. Tewari, S. N., Naturwissanschaften, 1954, 41, 229. Batt, C. W., and Kowkabany, G. X., J . Chem. Educ., 1955,32, 353. Gay, T. B., B.Sc. Thesis, University of Bristol, 1956. Pollard, F. H., McOmie, J . F. W., and Stevens, H. M., J . Chem. Soc., 1951, 771. Pollard, F. H., Nickless, G., and Banister, A. J., Analyst, 1956, 81, 577. Vaeck, S. V., Nature, 1953, 172, 213. Lacourt, A,, Sommereyns, G., and Wantier, G., Analyst, 1952, 77, 943. Yoe, J. H., and Jones, A. L., Ind. Eng. Chem., Anal. Ed., 1944, 16, 111. Martin, J. V., Ph.D. Thesis, University of Bristol, 1955. Pollard, F. H., hlcOmie, J. F. W., Stevens, H. M., and Maddock, J. G., J . Chem. Soc., 1953, 1338. Middleton, G., and Stuckey, R. E., Analyst, 1954, 79, 138. Kolier, I., and Ribaudo, C., Anal. Chem., 1954, 26, 1546. Whiteley, M. A,, J . Chem. Soc., 1903, 44. Marston, H. R., and Dewey, D. W., Australian J . ExfiI. Bid. Med. Sci., 1940, 18, 343. Young, R. S., Pinkney, E. T., and Dick, R., Ind. Eng. Chem., Anal. Ed., 1946, 18, 474. Sideris, C. P., Ibid., 1942, 14, 756. Sideris, C. P., and Chun, H. H. Q., Ibid., 1944, 16, 276. Griffing, &I,, and Mellon, M. G., Anal. Chem., 1947, 19, 1014. Hoffman, O., Ber., 1885, 18, 46. -, Ibid., 1891, 24, 3741. Baeyer, D., Ibid., 1887, 20, 1426. Saner, L. G., Ind. Eng. Chem., Anal. Ed., 1938, 10, 378. W’ise, W. M., and Brandt, W. W., Anal. Chem., 1954, 26, 693. Tolmachev, V. N., and Korobka, L. A., Zhur. Anal. Khim., 1954,9, 134. Cronheim, G., and Wick, W., Ind. Eng. Chem., Anal. Ed., 1942, 14, 447. Cronheim, G., J . Org. Chem., 1947, 12, 1. Harvey, E. A,, and Manning, D. L., J . Aruter. Chem. Soc., 1950, 72, 4488. Shome, C., Curr. Sci., 1946, 15, 107. -, Anal. Chem., 1948, 20, 1206. Woods, J. T., and Mellon, M. G., Ind. Eng. Chenz., Anal. Ed., 1941, 13, 551. Lacourt, A., Sommereyns, G., Stadler-Denis, A,, and Wantier, G., Mikrochem. Microchint. Acta, 1953, 40, 268. NOTE-Reference 34 is to Part 11 of this series. THE UNIVERSITY DEPARTMENT O F IXORGANIC AND PHYSICAL CHEMISTRY BRISTOL, 8 July 9th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9578200780

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

800 CLINCH AND GUY: THE EXTRACTION AND ABSORPTIOMETRIC p o l . 82 The Extraction and Absorptiometric Determination of Uranium as Thiocyanate BY J. CLINCH* AND MARGARET J. GUY A method is proposed for the determination of uranium in the presence of many other elements. The uranium is extracted as thiocyanate from a solu- tion containing ethylenediaminetetra-acetic acid at a pH of 3.5 t o 3.9 into 32.5 per cent. v/v tributyl phosphate in carbon tetrachloride. The optical density of the dried organic extract is compared with that of extracts containing no uranium at 350 mp in a spectrophotometer. The method has a coefficient of variation of about 0.6 per cent. Application of the method to the deter- mination of uranium in low-grade ores and to the determination of uranium in thorium oxide are given.BECAUSE of a need for the determination of uranium in thorium-containing materials, an examination has been made of the various methods available for the determination of micro- gram amounts of uranium. Although several sensitive methods are available, they generally demand that the uranium should be in as pure a state as possible for determination. The most sensitive method depends on the fluorescence of uranium in fused sodium fluoride me1ts.l This method gives errors of about The uranium must, however, be separated from most other elements, because of their quenching effect. Apparatus to measure the 10 per cent. * Present address : Fisons Limited, Avonmouth, Bristol.Dec. 19571 DETERMINATION OF URANIUM AS THIOCYANATE 801 fluorescence is also required; without the apparatus, and with use of visual standards, errors are somewhat greater than 10 per cent.Polarographic methods also require purification of the uranium? The oxine method of Silverman, Moudy and Hawley3 requires rigid pH control and even in the presence of ethylenediamine- tetra-acetic acid (EDTA), which is added to suppress the interference of elements such as iron, high and variable blanks have been found. Although the diethyldithiocarbamate method of Jones4 includes a step in which the uranium is purified by making use of the insolu- bility of uranium diethyldithiocarbamate in carbon tetrachloride, it is still unable to deal with more than small amounts of interfering elements. The aqueous thiocyanate method of Currah and Beamish6 has often proved to be the most useful colorimetric method, despite the low sensitivity. Ferric iron is reduced to the non- interfering ferrous state with stannous chloride, but molybdenum and rhenium are major interfering elements.Because of the acid conditions of the method, thorium can be tolerated, even in large amounts. An objection to the method in practice has been the production of a spurious yellow colour on the addition of stannous chloride to the sample. The method now proposed was suggested by the observations of Freiser and his associate^,^^^ who found that ferric and cupric thiocyanates could be extracted into tributyl phosphate or mixtures of tributyl phosphate and carbon tetrachloride. Experiments showed that uranium thiocyanate could also be extracted into mixtures of tributyl phosphate and carbon tetrachloride to give, depending on the amount of uranium present, colourless to yellow solutions.To prevent the extraction of ferric thiocyanate, a solution of EDTA was added to the aqueous layer before extraction. Uranium was still found to be extracted, but ferric ion was retained in the aqueous layer. WAVELENGTH OF ABSORPTION- Twenty-five millilitres of a solution of uranium nitrate containing 0.01364 g of uranium per litre were mixed with 5 g of ammonium thiocyanate and 1 g of the disodium salt of EDTA. The solution was diluted to 50 ml in a 100-ml graduated separating funnel and 10 ml of a 25 per cent. v/v solution of tributyl phosphate in carbon tetrachloride were added. After the mixture had been shaken for 30 seconds, the layers were allowed to separate and the lower organic layer was withdrawn into a 25-ml flask containing about 0.5 g of anhydrous sodium sulphate.After the organic layer had been set aside for 5 minutes to allow it to dry, its optical densities were compared at wavelengths from 300 to 380 mp with a solution prepared in a similar fashion, but with the uranium omitted. Graphs of the same form as those shown in Fig. 1 were obtained. A notable feature of the absorption was the peak at 350 mp instead of the expected gradual increase in absorption as the wavelength decreases, a feature character- istic of the aqueous thiocyanate method.6 Subsequent measurements were always made at 350 mp. Because of the width of the absorption peak, it was found possible to increase the slit width used in these experiments from 0.22 mm (corresponding to a 0.5-mp band pass) to 0.4 mm, thereby increasing the sensitivity of the readings.The previous work of Freiser and his associates suggested that the extraction might depend on the ratio of tributyl phos- phate to carbon tetrachloride, the pH of the extracted solution and the concentration of thiocyanate. RATIO OF TRIBUTYL PHOSPHATE TO CARBON TETRACHLORIDE- Ten millilitres of organic solvent were used and 0.341 mg of uranium was extracted from 50 ml of a solution containing 5 g of ammonium thiocyanate and 0.5 g of the disodium salt of EDTA. The maximum absorption appeared to be reached at a concentration of 32.5 per cent. v/v of tributyl phosphate in carbon tetrachloride; a solvent having this composition was used in subsequent work.CONCENTRATION OF AMMONIUM THIOCYANATE- Ten millilitres of 32.5 per cent. v/v tributyl phosphate in carbon tetrachloride were used and 0.341 mg of uranium was extracted from 50 ml of solution containing 0.5 g of the disodium salt of EDTA and various volumes of 50 per cent. v/v ammonium thiocyanate solution. The results are shown in Fig. 3. Several colorimetric methods have been investigated. EXPERIMENTAL The results are shown in Fig. 2.802 CLINCH AND GUY: THE EXTRACTION AND ABSORPTIOMETRIC [Vol. 82 I - 1, 0 300 320 340 360 380 400 Wavelength, rnp Variation of optical density with wavelength: curve A, solution containing 0.341 mg gf uranium measured against blank solution; curve B, blank solution containing no uranium measured against carbon tetrachloride..1 32.5 per cent. vjv mixture of tributyl phos- phate in carbon tetrachloride was used as the solvent for the preparation of both curves Fig. 1. Concentration of tributyl phosphate in tributyl phosphate - carbon tetrachloride mixture, % Fig. 2. Variation of optical density of a solution containing 0.341 mg of uranium with percentage of tributyl phos- phate in tributyl phosphate - carbon tetra- chloride mixture 0.21 , , , 1 0 0 4 8 12 16 20 Volume of 50 per cent. ammonium thiocyanate in 50 ml of aqueous layer, ml 4 5 o'0 0 I 2 PH 3 Fig. 4. Variation of optical density with pH Fig. 3. Variation of optical density of a solution containing 0.341 mg of uranium with concentration of ammonium thiocyanate in the aqueous layer The results plotted show that the optical density increases to a maximum with increasing thiocyanate concentration. This is an advantage over previous uranium thiocyanate methods, in which the optical density always increases with the thiocyanate concentration and no maximum is attained.Hence, 10 ml of 50 per cent. w/v ammonium thiocyanate solution were used in subsequent work. EFFECT OF pH- If the pH of the aqueous layer is between 1.9 and 3.9, the optical density of the extract is constant (Fig. 4). Above pH values of 3.9 the extraction of uranium rapidly diminishes, and on shaking solutions of pH values of less than 1.9 the organic layer becomes pink and its optical density higher. Tests with various interfering elements (vide infra) showed that interference was often most marked at pH values of less than 3.For this reason we have usedDec. 19573 DETERMINATION OF URANIUM AS THIOCYANATE 803 the narrower pH range of 3.5 to 3.9, in which there is very little interference from those cations that react with thiocyanate. The organic extract may be kept for several hours without change if prepared in the correct pH range. The solution should be kept away from ground-glass surfaces, since it readily extracts traces of iron. APPARATUS- A Cam- bridge Instrument Co. bench-type direct-reading pH meter was used for the pH measurements. REAGENTS- n-Tributyl phosphate-Wash with 10 per cent. w/v sodium carbonate solution and redistil under reduced pressure (30 mm of mercury) to remove water, butanol and mono- and dibutyl phosphates.Carbon tetrachloride. Tributyl phosphate in carbon tetrachloride, 32.5 per cent. v/v-Dilute 325 ml of tributyl phosphate to 1 litre with carbon tetrachloride. Ammonium thiocyanate solution, 50 per cent. w/v-Dissolve the contents of a 500-g bottle of ammonium thiocyanate in warm water, filter through a paper-pulp pad and dilute to 1 litre with water. Care should be taken over the selection of a suitable grade of ammonium thiocyanate; inferior grades turn yellow and a precipitate forms in the solution within a few days. EDTA solution, 10 per cent. w/v-Dissolve 100 g of the disodium salt of ethylenediamine- tetra-acetic acid in 900 ml of boiling water, filter, cool and dilute to 1 litre. Ammonia solution, sp.gr. 0.880. Ammonia solution, M.Hydrochloric acid, 11-4 M. Hydrochloric acid, M. METHOD Optical-density measurements were made on a Uvispek spectrophotometer. The AnalaR grade has proved satisfactory. PROCEDURE- Add 20 ml of 10 per cent, w/v EDTA solution and adjust the pH to between 3.5 and 3.9 with M ammonia solution or M hydrochloric acid. Add 10ml of 50 per cent. w/v ammonium thiocyanate solution, mix, re-check the pH and transfer the solution to a 100-ml graduated separating funnel, Dilute to 50 ml with water and add 10.0 ml of 326 per cent. v/v tributyl phosphate in carbon tetrachloride. Shake the separating funnel vigorously for 30 seconds, allow the organic layer to separate and run it off into a small flask containing about 0.5 g of anhydrous sodium sulphate. Dry the organic layer by swirling the contents of the flask.Measure the optical density of the solution in 1-cm cells against a blank prepared in a similar fashion, at a wavelength of 350 mp and with a slit width of 0.4 mm. The uranium may be in nitrate, chloride, sulphate or perchlorate solution. RESULTS Extraction of uranium into the organic layer appears to be extremely rapid, a 5-second shaking period being sufficient for pure uranium solutions. A 30-second shaking period was adopted. The optical density obeys Beer's law up to a value of 1.5 (the maximum tested) ; the molecular extinction coefficient is 5310 and the precision (u) is 1.7 pg of uranium. The molecular extinction coefficient compares favourably with the values by previous published methods- Temperature effects were not investigated. hlolecular extinction coefficient Peroxide (depending on method)s.... .. .. .. 1OOOto 1500 Diethyldithi~carbamate~ . . . . . . . . .. .. 4650 . . .. .. 3640 to 6550 Oxine (depending on ~ a v e l e n g t h ) ~ . . DibenzoylmethaneQ .. . . .. . . .. . . 21,000 Aqueous thiocyanate5 . . .. .. .. .. . . 3440 Aqueous acetone - thiocyanatelO . . .. .. . . .. 3850 Since many metallic ions give colours or precipitates with thiocyanate ion in water, an Besides the well extensive survey was undertaken to determine their effect on the method.804 CLINCH AND GUY: THE EXTRACTION AND ABSORPTIOMETRIC [Vol. 82 known colours given by ferric iron, rhenium, molybdenum and cobalt ions, lead and nickel thiocyanates absorb in the ultra-violet region. Experiments were therefore performed with solutions of various pH values from 2.5 to 3.9 containing 50 mg of metal.It was generally found that interference, if any, was much less at the higher pH value. More experiments were therefore made at pH values of about 3.8, with 20, 50, 100 and 200 mg of metal being used. The following metallic ions in the given valency states were found to give no inter- ference at all at pH 3.8 and with up to 200mg of metal: TlI, AgI, CuII, NiII, MnII, CdII, PbII, SnII, BiIII, FeIII, CrIII, SbIII, light LnIII, (heavy Ln + Y)III, ScIII (60 mg was the most tried), AlIII, TiIv (100 mg was the most tried), ThIv, VV (80 mg was the most tried) and ReVII. Elements that have caused difficulties are- (i) Molybdenzm and tungsten-Molybdenum (as ammonium molybdate) interferes considerably at pH values less than 3.6.At a pH of 3.85, 700 mg of molybdenum gave an optical density of 0.032 in the absence of uranium. In the presence of uranium, molybdenum causes considerable diminution of the optical density. The effect is shown by the following values found in the presence of 0.341 mg of uranium at a pH of 3.8- Molybdenum taken, mg . . * . 0 14 70 140 350 Optical density at 350 mp . . .. 0.762 0.713 0,669 0.582 0.396 Tungsten (as ammonium tungstate) interferes similarly, but the effect is less pro- nounced, as shown by the following results, 0.341 mg of uranium again being present- Tungsten taken, mg .. .. 0 20 50 200 Optical density a t 350 mp . . .. 0.764 0.742 0.708 0.582 Molybdenum should therefore be absent, but tungsten may be tolerated in amounts up to 10mg.(ii) Cobalt-Although cobalt appears to be largely complexed by EDTA, some extraction does occur and, with increasing cobalt content, the organic layers become tinged with blue. The effect is shown by the optical densities at 350 mp in the presence of 0.341 mg of uranium- Cobalt taken, mg . . .. .. 0 20 50 100 200 Optical density a t 350 mp . . .. 0.761 0.779 0.805 0,822 0.876 Small amounts of cobalt may therefore be tolerated, but amounts larger than 10 mg should be avoided. (iii) Zirconium, t i d y and ceriumIV-These elements were precipitated in the presence of EDTA a t pH values of 3.8, and some of the uranium appeared to be co- precipitated. The effect is shown by the optical densities of various solutions containing zirconium and 0-341 mg of uranium, as follows- Zirconium taken, mg .. * . 0 20 50 100 200 Optical density a t 350 mp . . .. 0.762 0.585 0.663 0.552 0.513 TinIv and ceriumIV interfere similarly, but by reducing them to tin11 and ceriumII1, interference is avoided. (iv) Gold and $latinum-In tests, 50 mg of gold as chloroauric acid appeared to be completely extracted into the organic layer and to colour it a deep orange. The solution absorbed strongly at 350 mp, 10 mg of gold giving an optical density of 0.229 in 1-cm cells. The solution does not appear to obey Beer’s law and the absorptionis tooinsensitive for the method to have any practical use in the determination of gold. The effect of gold can be completely suppressed by the addition of sufficient potassium cyanide almost to decolorise the solution. Platinum (added as chloroplatinic acid) interferes similarly, but the interference cannot be completely eliminated with cyanide.( v ) Mercury-Mercurous and mercuric ions are extracted and absorb strongly in the presence and absence of uranium. Anions that cause no interference in amounts up to 5 g added as the ammonium salts The effects of other groups are considered (i) Nitrites and nitrogen dioxide-These must be absent because of the formation of Care should be exercised even when nitrates Pure neutral nitrates appear to be Mercury ions must therefore be absent. include chloride, nitrate, sulphate and perchlorate. below- a yellow or red colour, which is extractable. are used, since these may contain nitrogen dioxide.Dec.19571 DETERMINATION OF URANIUM AS THIOCYANATE 805 quite safe; up to l o g of ammonium nitrate has been used in an extraction without alteration of the final result. Small amounts of nitrogen dioxide can be eliminated by warming with 0.5 g of urea before extraction, but larger amounts are conveniently removed by heating to fumes with perchloric acid. (ii) Fluoride-Interference with the extraction is found at very low levels, as shown by the following results, 0.341 mg of uranium being present- Fluoride taken, mg .. .. 0 2 5 10 20 Optical density a t 350 mp . . .. 0.762 0.729 0.613 0.426 0.115 The fluoride can be eliminated by heating t o fumes with perchloric acid with or with- out the addition of boric acid, or complexed by adding aluminium nitrate to the sample.(iii) Phosphate-Uranium is precipitated by phosphate under the pH conditions used in the method. Precipitation is, however, slow and, if acid solutions are rapidly neutralised and extracted, the correct result is obtained. Uranium precipitated in the presence of phosphate may be brought into solution by adding acid. The rate of pre- cipitation of the phosphate is diminished the lower the pH of the solution. Up to 200 mg of P,O, have been satisfactorily dealt with in this way. (zv) Organic acids-The effect of acetic, citric and tartaric acids has been examined. The absorption is diminished a t amounts greater than about 1 g, but can be neglected for 0.5-g amounts. All the acids can be eliminated by heating to fumes with nitric and perchloric acids.( v ) Dichromate and permanganate-Dichromate ions give an extractable brown colour with thiocyanate, which interferes strongly. Reduction t o the chromic state with sulphur dioxide removes the interference. Permanganate is reduced by EDTA and is non-interfering. Arsenate behaves similarly. APPLICATIONS OF THE METHOD The method has been successfully used for examining the distribution of uranium in a thorium process from the starting ore containing about 0.1 per cent. of U,O, and 6 per cent. of Tho, down to the final pure thorium oxide containing less than 0.2 p.p.m. of U,O,. Puri- fication from gross amounts of impurities is easily effected by using the standard methods involving extraction with ether, either continuous solvent extractorsll or extraction from cellulose columns.12 Two examples of these techniques are given- (a) Determination of uranium in ores-Uranium has been determined in a variety of ores by the cellulose column method.12 After removal of the ether, the aqueous remainder was heated with a few millilitres of Derchloric acid and evaDorated almost to drvness.The proposed method was then applied to the residue or anLaliquot. The result; are shown in Table I. TABLE I Ore Monazite . . . . Pyrochlore Siliceous ore Columbite Monazite . THE DETERMINATION OF URANIUM IN ORES g % Weight taken, Aliquot taken U,O, found, 0.198 0.271 0.223 0,561 0.477 0.549 0.488 0.264 0,203 0.471 0.653 0.565 0.205 0.253 0.287 0.325 0.229 10/50 10j50 10/50 10/50 10/50 10/50 0.381 0.378 0.360 0.0273 0,0273 0.0269 0.325 0.358 0.343 0.0366 0,0372 0.0370 0.114 0.114 0.113 0.114 0.111 U,O, pyent,I1 /o 0.37 - 0.025 0.32 0.036 - 0.12 - -806 CLINCH AND GUY: THE EXTRACTION AND ABSORPTIOMETRIC [Vol.82 ( b ) Determination of uranium in thorium oxide-Some preliminary experiments were made to see whether uranium could be quantitatively extracted from acidified thorium nitrate solutions into tributyl phosphate - carbon tetrachloride mixtures as uranyl nitrate. The tributyl phosphate - carbon tetrachloride mixture could then be removed, treated with EDTA and thiocyanate and the optical density measured. Some difficulties were found in this work. This necessitates use of a buffering agent such as sodium acetate when the organic layer is removed and treated with thiocyanate. Also, 100 per cent. extraction was not obtained in one shaking.This meant that several portions of tributyl phosphate - carbon tetrachloride mixture would have to be used, mixed and diluted to a known volume. The sensitivity of the method would thereby be considerably decreased. The separation of the phases, too, was very poor. It was therefore decided to use a continuous ether extractor and determine the uranium in the extract. REAGENTS- Nitric acid is extracted into the organic layer. Nitric acid, concentrated, 16 N . Sodium juoride solution-Dissolve 10 g of pure sodium fluoride in water and dilute to Ammonium nitrate. Saturated ammonium nitrate solution-Dissolve sufficient solid ammonium nitrate in warm water to form a saturated solution and cool to room temperature. Perchloric acid, 12 N .PROCEDURE- Weigh 30 g of thorium oxide into a 400-ml beaker, and add 10 ml of sodium fluoride solution and 50 ml of concentrated nitric acid. Cover with a clock-glass and warm gently to boiling. Boil until the thorium oxide is dissolved, adding more nitric acid if necessary, and evaporate the solution until crystallisation commences. Cool somewhat, and add 10 ml of water and 5 g of ammonium nitrate. Cool the mixture to room temperature and dilute with water until all the solid is dissolved. Transfer the solution to the extraction vessel of a con- tinuous ether-extraction apparatus, using saturated ammonium nitrate solution for washing out the beaker. Adjust the level of liquid in the extraction vessel to 2 cm below the side- arm. Extract with a mixture of 50 ml of water and 100 ml of ether in the boiler flask for 18 hours.Distil off the ether from the boiler flask and transfer the aqueous solution to a 250-ml beaker. Add 5 ml of concentrated nitric acid and 5 ml of perchloric acid and evaporate gently to fumes of perchloric acid. A small ashless filter-paper added to the evaporating solution often prevents bumping. Evaporate the solution nearly to dryness and transfer to a 100-ml beaker, using the minimum of water. Determine the uranium present by applying the proposed method, but using 15 ml of 32.5 per cent. v/v tributyl phosphate in carbon tetrachloride and measuring the optical density in 4-cm cells. RESULTS A standard graph was prepared by extracting amounts of uranium equivalent to up to 200 pg of U,O, into 15 ml of 32.5 per cent.tributyl phosphate - carbon tetrachloride mixture and measuring the optical density in 4-cm cells. An optical density of 1 was equivalent to 199.5 pg of Tj,O,. A pure thorium oxide, with and without added uranium, was examined by this method. The results are shown in Table 11. 500 ml. Cool the extraction vessel by means of an ice jacket. Determine the uranium present in parts per million. The standard deviation was 1.5 pg of U,O,. TABLE I1 DETERMINATION OF URANIUM IN THORIUM OXIDE Weight of Uranium oxide Uranium oxide Mean uranium Standard thorium oxide taken, (U,O,) added, (U,O,) found, oxide (U,O,), deviation, g Pg I*8 Pg Pg 30 nil 8.0, 5.4, 2.4, 5.1 2.3 7.2, 6.2, 2.8 30 40 44.7, 41.3, 42.0 43.3, 38.7 2.6Dec. 19571 DETERMINATION O F URANIUM AS THIOCYAKATE 807 The result for no added uranium corresponds to a uranium content in the sample of 0.17 p.p.m. of U,O,, with a standard deviation of 0.08 p.p.m. We are grateful to the Directors of Thorium Limited for permission to publish this paper and to Messrs. W. P. Kemp and K. W. Ponting for the preparation of the tributyl phosphate used in the work. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Price, G. R., Ferretti, R. J., and Schwartz, S., Anal. Chem., 1953, 25, 322. Shalgosky, H. I., Analyst, 1956, 81, 512. Silverman, L., Moudy, L., and Hawley, D. W., Anal. Chem., 1953, 25, 1369. Rodden, C. J., “Analytical Chemistry of the Manhattan Project,” McGraw-Hill Book Co. Inc., Currah, J. E., and Beamish, F. E., Anal. Chem., 1947, 19, 609. Melnick, L., Freiser, H., and Beeghly, H. F., Ibid., 1953, 25, 856. Melnick, L. M., and Freiser, H., Ibid., 1955, 27, 462. Rodden, C. J., 09. cit., p. 82. Yoe, J. H., Will, F., 111, and Black, R. A., Anal. Chem., 1953, 25, 1200. Crouthamel, C. E., and Johnson, C. E., Ibid., 1952, 24, 1780. “Handbook of Chemical Methods for the Determination of Uranium in Minerals and Ores,” H.M. Stationery Office, London, 19.50. Burstall, F. H., and Wells, R. A,, Analyst, 1951, 76, 396. New York, 1950, p. 113. ANALYTICAL LABORATORIES THORIUM LIMITED UPHALL ROAD ILFORD, ESSEX March 29th, 1957

ISSN:0003-2654    

DOI:10.1039/AN9578200800

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

Dec. 19571 DETERMINATION OF URANIUM AS THIOCYAKATE 807 The Determination of Thorium in Ores with APANS - rnesorartaric Acid Reagent after a Shortened Chromatographic Separation BY D. A. EVEREST AND J. V. MARTIN A rapid procedure is described for the determination of thorium in medium-grade and low-grade ores containing as much as 50 per cent. of zirconia. Thorium is separated from gross amounts of other elements by elution on a cellulose - alumina column and determined by a selective absorp- tiometric finish with APANS [ l-(o-arsonophenylazo)-2-naphthol-3 : 6-di- sulphonic acid], mesotartaric acid being used as a masking agent for zirconium. THE determination of thorium in complex and low-grade ores by classical precipitation methods is extremely tedious and uncertain. In recent years attempts have been made t o shorten the operational time involved by the introduction of selective solvent-extraction and chromatographic methods. Mesityl oxide1JJ and tri-n-butyl phosphate have been the princi- pal solvents used for the extraction of thorium, but such methods suffer from the disadvantage that thorium must be separated from zirconium by precipitation of thorium as oxalate or fluoride.1~3 The chromatographic procedure developed by Williams4 has been used for the determina- tion of thorium in a wide variety of minerals and ores.Thorium is separated from gross amounts of other elements by elution from a compound cellulose - alumina column with ether containing 12 per cent. v/v of nitric acid and is then precipitated as thorium oxalate and finally weighed as Tho,.When the amount of thorium in the sample is low and the amount of zirconium is high, a double-column separation, i.e., thorium is twice eluted on cellulose - alumina columns, is required. Alternatively, thorium has been separated from most of the zirconium present by a preliminary precipitation as fluoride and finally eluted from a cellulose - alumina column with ether - nitric acid s01vent.~ Small amounts of thorium thus separated have been determined by spectrographic6 or colorimetric methods.537 The purpose of this work was to develop a method for the determination of thorium in which only one separation on a column would be required, and which should have an absorptio- metric finish in which small amounts of impurities (particularly zirconium) present in the808 eluate could be tolerated.Certain authors have criticised the original method of Williams on the grounds that recovery values for small amounts of thorium are lo^.^^^ Guest7 suggested that thorium is not quantitatively precipitated as hydroxide from aqueous solution after the first column separation, because of the presence of organic material extracted from the cellu- lose by the eluting solvent. Steele6 claimed that thorium may be retained on the column by co-precipitation with titanium and zirconium in insoluble products of hydrolysis and in phosphates insoluble in nitric acid. In view of the above-mentioned criticisms, the chromatographic procedure of Williams was examined and it was found that, for a single-column separation, no losses of thorium were encountered, even when the zirconium content of the sample was high.The method des- cribed in this paper is much more rapid than previous chromatographic procedures in that no lengthy chemical precipitations are needed and no double-column separation is required. The chromatographically separated thorium is determined absorptiometrically with APANS [l-(o-arsonophenylazo)-2-naphthol-3 : 6-disulphonic acid], mesotartaric acid being used as a masking agent for zirconium. APANS is a highly selective reagent for thorium and is now widely ~sed.2~3~8~~JO Zirconium is one of the few metals that interferes appreciably, and Grimaldi and Fletcher have recently found that its interference may be reduced by using D(+)-tartaric acid as a masking agent.ll These authors also made brief mention of meso- tartaric acid, which has been found in this work to mask satisfactorily any zirconium eluted from a cellulose - alumina column. Since the disodium salt of ethylenediaminetetra-acetic acid (EDTA) is known to form a very stable complex with zirconium, it also was examined as a possible masking agent, but was found to be unsatisfactory.The procedure was developed principally because of a need to determine thorium in experiments on the amenability of certain thorite-bearing tailings, containing approximately 6 per cent. of Tho,, to treatment with mineral acid. The tailings contain 50 to 70 per cent. of zircon and the residue for analysis contains almost all the zircon and only small amounts (less than 0.5 per cent.) of thorium.Because, however, the accuracy attainable with APANS (j0.5 to 1.0 per cent.) is comparable with that by other methods at present used for thorium, applica- tion has also been made to the determination of thorium in materials containing as much as 10 per cent. of Tho,. Titration of thorium with EDTAI2 after elution from a cellulose - alumina column has also been examined as a rapid method for the determination of the metal in medium-grade ores. The results have been good for some ore samples, but the titration is liable to inter- ference by traces of impurities, e.g., zirconium and phosphate, eluted from the column. Because of its limited application, the titration method is inferior to the APANS - mesa- tartaric acid procedure. EVEREST AND MARTIN: THE DETERMINATION OF THORIUM I N [Vol.82 EXPERIMENTAL CHROMATOGRAPHIC SEPARATION- The method used for the breakdown of the ore and for the chromatographic elution is that described by Williams? starting at p. 301, line 16, and finishing at p. 302, line 17, of his paper. Certain operations in the method were found to be of particular importance for samples of high zirconium content. The sodium phosphate added to the solution con- taining the thorium should be thoroughly stirred into the warmed solution, otherwise much will remain undissolved in the gelatinous mass formed by precipitated zirconium phosphate. Sufficient alumina should be added to this solution to give a dry wad, so that the gelatinous mass of zirconium phosphate is broken up and well distributed in the alumina.RECOVERY OF MACRO AMOUNTS OF THORIUM FROM EXCESS OF ZIRCONIUM BY THE COLUMN METHOD- Because of the criticism by certain authors of the cellulose - alumina column procedure for thorium, the recovery of thorium in the presence of excess of zirconium was first checked for macro amounts of thorium (25 mg of Tho,). After a single-column chromatographic separation, thorium was precipitated as oxalate, ignited and weighed as Tho,. The recovery of thorium from a mixture with zirconium (containing 350 mg of 25-0,) was 99 per cent., which suggested that, for the single-column procedure at least, there are no serious losses of thorium. The recovery of smaller amounts of thorium is discussed subsequently.Dec. 19571 ORES WITH APANS - MeSOTARTARIC ACID REAGENT 809 COLORIMETRIC DETERMINATION OF THORIUM WITH APANS WITH USE OF A MASKING AGENT As has already been mentioned, the masking effect of EDTA and of mesotartaric acid on the characteristic interference of zirconium with the absorptiometric procedure with APANS for thorium was studied. To the stated amount of thorium were added by pipette the masking agent, 1 ml of concentrated hydrochloric acid, 5 ml of 0.1 per cent.w/v aqueous APANS solution and 20 ml of absolute ethanol, in that order, and the solution was made up to 50 ml. The blank solution contained all the reagents except the masking agent. FOR ZIRCONIUM- The results are shown in Table I. TABLE I REAGENTS FOR MASKING ZIRCONIUM IN THE ABSORPTIOMETRIC DETERMINATION OF THORIUM WITH APANS Composition of mixture r A , Optical density at 550 mp Masking agent Tho,, mg ZrO,, mg No masking agent .. .. . . . . 0.302 0.417 0.302 0.2 0.538 0.2 ml of 0.025 M EDTA solution . . . . 0.302 0.418 0.302 0.2 0.375 1 ml of 5 per cent. w/v mesotartaric acid solution 0.302 0.418 0.302 0.2 0,409 0.302 1 0.409 0,302 2 0.414 With no masking agent present zirconium causes a high positive interference. In the system containing EDTA, zirconium induces a small negative interference similar to that observed by Grimaldi and Fletcherll with D( +)-tartaric acid. With mesotartaric acid present as masking agent, a similar negative interference, which is however negligibly small, is observed, and as much as 2 mg of ZrO, can be successfully masked. When larger amounts of zirconium are present, the red precipitate of the zirconium - APANS complex is slowly formed and optical-density measurements are impossible.This precipitation is not prevented by increasing the concentration of mesotartraric acid by a factor of 2.5. nzesoTartaric acid, however, appeared to be a very satisfactory reagent for masking zirconium and was therefore used in the determination of thorium with APANS in subsequent work. The relatively slight interference of ferric iron in this determination is not eliminated with mesotartaric acid, and hydroxylamine hydrochloride was used to reduce iron to the non- interfering ferrous state. ABSORPTIOMETRIC DETERMIKATION OF THORIUM \VITH APANS - MCSOTARTARIC ACID REAGENT AFTER CHROMATOGRAPHIC SEPARATION- The absorptiometric method with APANS and with mesotartaric acid as masking agent for zirconium was applied to the determination of thorium separated by the cellulose - alumina column procedure. A high degree of accuracy was achieved for the determination of as little as 0.2 mg of Tho, in as much as 500 mg of ZrO,.Thorium was also determined satisfactorily in a number of low-grade and medium-grade ores. REAGENTS- METHOD Hydroxylamine hydrochloride solutiom-A 10 per cent. w/v aqueous solution. mesorartaric acid solution-A 5 per cent. aqueous solution of the monohydrate. APANS solution-A 0.1 per cent. w/v aqueous solution of the sodium salt of 1-(0-arsono- phenylazo)-2-naphtho1-3 : 6-disulphonic acid. PROCEDURE- After the chromatographic separation of thorium, add 150 ml of water and several glass beads to the nitric acid - ether eluate, and evaporate the ether by warming on a steam-bath. Boil the remaining aqueous solution to small bulk, transfer it to a calibrated flask and make up to the mark with 25 per cent.v/v nitric acid (see Kote 1). By pipette, put a suitable aliquot (see Note 2) into a 100-ml beaker and evaporate to dryness. Add 2 ml of a mixture of 70 per810 EVEREST AND MARTIN: THE DETERMINATION OF THORIUM IN cent, perchloric acid, concentrated nitric acid and water (2 + 1 + 1) and heat to dryness. Add by pipette I ml of concentrated hydrochloric acid, cover the beaker with a watch-glass, and warm gently to dissolve the residue. Dilute to about 10 ml, add 1 drop of hydroxylamine hydrochloride solution and heat to a point just short of boiling to reduce ferric iron.Cool the solution and transfer it to a 50-ml calibrated flask. Add by pipette 1 ml of mesotartaric acid solution, 5 ml of APANS solution and 20 ml of absolute ethanol in that order; allow to cool and make up to the mark. Measure the optical density in 1-cm glass cells at 550 mp with a Unicam spectrophotometer (red cell) against a blank solution containing all reagents. (See Note 3.) I t is advisable to prepare a separate calibration graph for each batch of stock APANS solution made. [Vol. 82 Beer's law is obeyed between 0 and 0.5 mg of Tho,. NOTES- Thorium may form an insoluble precipitate with organic matter resulting from degradation of the cellulose if the details of this step are not carefully followed. Not more than 0.4 of the total volume should be taken.With larger aliquots sufficient zirconium may be present to cause immediate precipitation with APANS. Optical-density readings are made 1 hour after the preparation of the thorium - APANS solutions except for solutions from low-grade samples, containing much zirconium. For these, optical-density readings are made after 15 to 20 minutes, since precipitation of the zirconium - APANS complex may occur if the solutions are set aside for a longer period. 1. 2. 3. RESULTS DETERMINATION OF SMALL AMOUNTS OF THORIUM IN ARTIFICIAL MIXTURES WITH ZIRCONIUM- Solutions containing thorium and zirconium were prepared from solutions of the metal nitrates. The mixed solutions were evaporated just to dryness, the residue was dissolved in 20 ml of 25 per cent.v/v nitric acid and the normal single-column procedure was followed. Thorium in the eluate was determined as described above. The results for thorium in the presence of 500 mg of 21-0, were as follows- Thorium present (Tho,), mg . . 1.01 0.97 0.97 0.20 0.20 Thorium found (Tho,), mg . . 1.01 0.98 0.98 0.22 0.22 PURIFICATION OF ZIRCONIUM- I n preliminary experiments zirconium was not purified and the recovery values were high in the determination of small amounts of thorium. This apparent error was subsequently found to be due to the presence in the zirconium of traces of thorium as an impurity. The zirconium was therefore purified by the following method, which is based on the procedure used by Steele6 for the determination of thorium in zirconium- Twenty-five grams of zirconium nitrate were warmed with 50ml of concentrated hydrofluoric acid in a polythene beaker for 1 hour.Then 100 ml of water were added and the solution was warmed on a steam-bath for 2 hours to complete the dissolution of the zirconium. The solution was diluted to 200ml with water and cooled, Next, 4 m l of lanthanum nitrate solution (containing 10mg of La,O, per ml), 80ml of mer- curous nitrate solution (a 0.95 per cent. w/v Hg,(NO,), in 0.1 per cent. v/v nitric acid) and 8 ml of 7 per cent. v/v hydrochloric acid were added in that order. The solution was set aside overnight, a small amount of cellulose pulp was added and the mixture was filtered through a Whatman No. 540 filter-paper in a polythene funnel into a large platinum dish. The filtrate was evaporated to dryness and the residue was heated to fumes several times with sulphuric acid.The residue was dissolved in water and the solution was diluted to 500 ml; zirconium was precipitated as the hydroxide with ammonia solution. The precipitate was collected in a Buchner funnel and dissolved in nitric acid, and zirconium hydroxide was re-precipitated five times to remove sulphate ions, The hydroxide was dissolved in 100 ml of concentrated nitric acid and the solution was diluted to 500ml and filtered through Whatman No. 42 filter-paper. To the filtrate 275 ml of concentrated nitric acid were added, and the solution was cooled and made up to 1 litre with water. The zirconium content was found gravimetrically by evaporat- ing an aliquot of the solution and igniting to the oxide, ZrO,.Thorium was determined in a number of residues from the mineral-acid leaching (with The zircon content was between 50 DETERMIXATION OF THORIUM IN LOW-GRADE AND MEDIUM-GRADE ORES- nitric and sulphuric acids) of thorite-bearing tailings.Dec. 19571 ORES WITH APANS - WeSOTARTARIC ACID REAGENT 811 and 70 per cent. in all the residues. Each analysis was performed in duplicate on I-g samples and the results were confirmed by spiking duplicate samples with a known amount of thorium, approximately equal to that already present in the ore. The ore was spiked after the extrac- tion of the potassium hydroxide melt with nitric acid by the addition of a known volume of standard thorium nitrate solution. Residue No. . . . . . . . . 1 2 3 4 5 Thorium (Tho,) found by direct Thorium (Tho,) found by deter- The results were as follows- determination, % ... . 0.26, 0.26 0.35, 0.36 0.44, 0.45 0.30, 0.30 0.22, 0.22 mination after spiking, yo . . 0.26, 0.26 0.36, 0.35 0.41, 0.45 0.29, 0.28 0.21, 0.21 Thorium was also determined in medium-grade ores and the results are compared in Table I1 with those obtained by the cellulose - alumina column procedure with an oxalate finisha4 TABLE I1 DETERMINATION OF THORIUM IN MEDIUM-GRADE ORES Sample Indian monazite . . .. .. African monazite .. .. Thorite . . .. . . .. Monazite . . . . . . . . African monazite concentrate . . Thorium containing gold residue Asian monazite . . .. . . Thorium (Tho,) found by APANS - mesotartaric acid finish,* 9.85 6.40 6.55 7.60 15.50 12.50 6.95 Yo Thorium (Tho,) found by oxalate finish,* 9.85 6.40 6.55 7.45 15.50 12.20 7.00 % * Mean values are given for a number of determinations. APPENDIX Since the completion of this work, Grimaldi, Jenkins and Fletcher have published papers13>14 about the determination of thorium in silicate rocks and in ores by using a system containing APANS and mesotartaric acid after preliminary precipitation of thorium as iodate.We thank Mr. G. H. Smith and other members of the Analytical Survey Group for supplying certain of the analytical results. This work is published by permission of the Director of the Chemical Research Laboratory. Levine, H., and Grimaldi, F. S., United States Atomic Energy Commission Report AECD-3188, Banks, C. V., and Byrd, C. H., Anal. Chem., 1953, 25, 416. Cuttitta, F., United States Department of Interior, Geological Survey, Trace Elements Investi- Williams, A. F., Analyst, 1952, 77, 297. Steele, T. \V., Government Metallurgical Laboratory Progress Report (South Africa), Analytical Kingsbury, G. W. J., and Temple, R. B. F., Analyst, 1952, 77, 307. Guest, R. J., Canadian Department of Mines and Technical Surveys, Mines Branch, Technical Thomason, P. F., Perry, M. A,, and Byerly, W. %I., Anal. Chem., 1949, 21, 1239. hlayer, A,, and Bradshaw, G., Analyst, 1952, 77, 154. Clinch, J., Anal. Chim. Acta, 1956, 14, 162. Grimaldi, F. S., and Fletcher, M. H., Anal. Chem., 1956, 28, 812. Ford, J . J., and Fritz, J. S., United States Atomic Energy Commission, ISC-520. Grimaldi, F. S., Jenkins, L. B., and Fletcher, M. H., Anal. Chew,., 1957, 29, 848. Fletcher, M. H., Grimaldi, F. S., and Jenkins, L. B., Ibid., 1957, 29, 963. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 1950. gation, Report No. 498, 1955. No. 23, 1955. Paper So. 1, 1953. CHEMICAL RESEARCH LABORATORY TEDDINGTON MIDDLESEX July loth, 1957

ISSN:0003-2654    

DOI:10.1039/AN9578200807

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

812 PATTERSON AND SAVAGE : DETERMINATION OF DEXTROSE AND [Vol. 82 The Use of Selective Desorption from Carbon Columns for the Determination of Dextrose and Maltose in Starch Conversion Products BY STELLA J. PATTERSON AND R. I. SAVAGE A method is proposed for determining the mono- and disaccharides in starch conversion products, based on selective desorption followed by elution from a carbon column. Recovery is quantitative provided the most suitable temperature conditions for recovery from each batch of carbon are established before it is put into routine use. The method of testing the carbon in order to obtain these conditions is described. THE classical methods available for determining individual sugars in a mixture such as occurs in starch conversion products depend on reducing power towards cupric salts under various conditions and on optical rotation.In these methods inaccuracies arise because of the effect of the other sugars present on the sugar being determined, and it is not easy to make allowance for this effect. A method that physically separates the sugars one from another is therefore desirable. Selective desorption from a carbon column provides such a method, and has been used by a number of workers. A mixture of sugars is adsorbed on a column of carbon, and the individual sugars are then removed selectively by washing successively with water and with aqueous ethanol of increasing concentrations. In this way, monosaccharides are separated first, then disaccharides, trisaccharides and so on. Most of the methods published, however, are inconvenient for routine quantitative use, many of them depending on gradient elution of the sugars from the column or on forcing the liquid through the column by pressure from a gas cylinder, and in much of the work done the American carbon Darco G 60 has been used.Further, some of the procedures described did not yield quantitative recovery of the sugar^.^^^^^ I n this paper a test for the suitability of a carbon for use in the columns is described. Of the carbons we have investigated, so far only one, B.D.H. (British Drug Houses Ltd.) “Activated Charcoal, Acid Washed,” has been found to give quantitative recovery of sugars when used as supplied. The capacity of this carbon for retaining sugars is such that, in order to be certain of recovering the separated sugars quantitatively in sharp bands uncontaminated by other sugars, the concentration of sugar in the eluates has to be too low for accurate determination by the usual macro methods, although the semi-micro method of Somogyi4f5 is suitable.We found it of advantage to modify Somogyi’s improved reagent by reducing the recommended amount of sodium sulphate in order t o avoid the crystallisation that took place at the temperatures normally prevailing in this country. This modified reagent has been satisfactory for determining sugars in the eluates, and we have had no evidence that its keeping qualities have been impaired. During very cold weather, crystallisation of sodium sulphate occasionally occurs overnight with this modified reagent, which is saturated with sodium sulphate a t approximately 15” C.Warming the solution to redissolve the crystals has not affected its reduction equivalent, and we have thought it undesirable to reduce any further the amount of sodium sulphate present. For quantitative work, we have found this carbon-column method more convenient than elution of the sugars from paper chromatograms. The immediate object of our work has been to provide a method for determining the amounts of mono- and disaccharides in liquid glucose and similar starch conversion products; by making use of the same principles, it should be possible t o extend the method to include the determination of higher saccharides and other mixtures of sugars. It was decided that it would be of advantage if we could use a “blind” method, that is, one in which we knew that the dextrose was all being eluted into one fraction and the maltose all into another, instead of collecting a number of fractions to make sure of obtaining all of one sugar and none of another.The first batch of B.D.H. charcoal that we tried worked satisfactorily under the con- ditions described at laboratory temperature, but we had difficulty with subsequent batches. As shown later in the paper, this can be overcome by eluting at a suitable temperature.Dec. 19571 MALTOSE I N STARCH CONVERSION PRODUCTS 813 METHOD APPAUATUS- The assembly of the carbon column is shown in Fig. 1. A 6-inch x 2-inch test-tube has a hole blown in its end, and to this is fused a short length of 4-mm glass tubing. This forms the adsorption column.The rubber bung, A, fits either a standard 100 to 110-ml sugar or Reichert - Polenske flask or a 6-inch x ;-inch test-tube, which can be changed by slowly opening tap B to the air and rapidly substituting the replacement. + Water at constant temperature 3 Fig. 1. Assembly of carbon column Some device for maintaining the column a t the required temperature is necessary. We use a Liebig condenser type of jacket surrounding the column (as shown in Fig. l), through which water is circulated from a coil immersed in a beaker of water and heated or cooled as required. For the Somogyi determinations, a series of glass-stoppered boiling-tubes, or ordinary boiling-tubes with fitted corks, is required. To hold the boiling-tubes rigidly in a boiling- water bath, we use a rack of Terry’s clips that can be secured on either side of the bath.REAGENTS- Charcoal Powder for Decolorising Purposes. Washed with Acid.” Activated carbon-The product supplied by the British Drug Houses Ltd., “Activated Kieselguhr-We used Metasil A. Other brands have been found to be equally suitable. Aqueous ethanol, approximately 7 per cent. v/v-Dilute 75 ml of aldehyde-free 95 per cent. ethanol to 1 litre with water. Modified Somogyi micro-copper reagent-Dissolve 30 g of potassium sodium tartrate and 30 g of anhydrous sodium carbonate in about 200 ml of hot water and add 40 ml of N sodium hydroxide solution and, with stirring, 80 ml of a 10 per cent. w/v solution of copper sulphate, CuS04.5H,0. Boil the solution to expel air. Dissolve 290g of sodium sulphate, Na,SO,.lOH,O, in about 300ml of water and boil the solution to expel air; then add it to the copper solution contained in a 1-litre calibrated flask.Add 8 g of potassium iodide dissolved in a few millilitres of water, and then 10.0 ml of N potassium iodate. Dilute to the mark with air-free water. If the solution is not perfectly clear, filter it. Sulphuric acid, approximately 6 N-Dilute 17 ml of concentrated sulphuric acid to 1 litre with water. Sodium thiosulphate solution, approximately 0.005 N.814 [Vol. 82 PROCEDURE FOR ROUTISE DETERMINATIONS- Place a small plug of cotton-wool inside the column, press it well down, and moisten it thoroughly with water. Make a slurry of about 0.1 g of kieselguhr with water, pour it into the tube, and allow it to drain.Nake a slurry of 7 g of a mixture of equal parts by weight of carbon and kieselguhr with water, taking care that every particle is wetted, and pour it into the tube. Apply gentle suction. (From this time onwards the top of the carbon column must never be allowed to become dry.) When only a 1-mm layer of supernatant liquid remains, apply a circle of filter- paper to the top of the column and press down gently. Pour 100 ml of water at the required temperature into the separating funnel, and apply suction, keeping about a 1-cm head of liquid on top of the column. The column is ready for use when the water has passed through. Prepare a solution of the sample of such a concentration that 5 ml will contain approxi- mately 10 mg of anhydrous maltose; for commercial liquid glucose a 2 per cent.solution is suitable, Place an empty 100-ml calibrated flask in position, and put 5 ml of the solution by pipette on to the column. Apply suction and, just before the last of the solution dis- appears into the column, wash down the walls of the tube with a few millilitres of water; then elute the column with water at the required temperature until 95 to 100 ml of eluate containing the dextrose have been collected in the receiver. Release the suction, and substitute an empty 100-ml calibrated flask as receiver. Remove the surplus water from the top of the column with a pipette, and substitute 7 per cent. ethanol at the required temperature for water in the separating funnel. Apply suction again, and collect 95 to 100 ml of ethanolic eluate con- taining the maltose.The rate of flow of liquid through the column does not appear to influence the elution of the sugars; we have found that a rate of 100 ml in 40 to 50 minutes is convenient. Dilute the aqueous and ethanolic eluates to volume in the 100-ml flasks and determine the sugars by Somogyi’s method as follows. By pipette put duplicate 5-ml portions of (a) water, ( b ) 7 per cent. ethanol, (c) the aqueous and ( d ) the ethanolic eluates from the column into eight 6-inch x 1-inch boiling-tubes. Then also by pipette put 5-ml portions of the micro-copper reagent into each tube. Mix thoroughly, cover the mouths of the tubes with glass bulbs, and immerse the tubes in a bath of briskly boiling water. Withdraw the four tubes containing the aqueous eluates after exactly 10 minutes and the four containing the ethanolic eluates after a further 10 minutes, and cool them to room temperature in a bath of cold water.To each tube add approximately 1 ml of 6 N sulphuric acid, drawing the glass bulb aside just sufficiently to admit the jet of the pipette. The acid should be rapidly squirted, rather than permitted to flow into the test-tube, so that the entire contents of the tube are mixed and acidified a t once. A teat pipette is convenient for this purpose. Immediately swirl the tube vigorously until no trace of cuprous oxide remains undissolved. Set the tube aside for 2 minutes, swirl again, wash down the glass bulb with water into the tube, mix well, and titrate slowly with 0,005 N sodium thiosulphate, mixing well during the early stages of the titration after each addition of 0.5 ml of sodium thiosulphate.Add freshly prepared starch indicator near the end-point, and shake thoroughly just before the final drop is added. Read the burette to 0.01 ml (we find a 25-ml grade A burette suitable). (In theory it would be no doubt more correct to carry out blank determinations with water and ethanol that had passed through the column, but in practice we have found that results determined in this way are the same as those given when the water and ethanol are used direct.) Standardise the sodium thiosulphate against 0.005 N potassium iodate. Calculate the differences between the titres of the sugar solutions and the corresponding blanks in terms of 0.005 N sodium thiosulphate, and read off the corresponding weights of sugar from the relevant graphs. It is preferable for each worker to construct his own graph showing the relation between volume of 0.005 N sodium thiosulphate in ml and amount of sugar in mg.The results shown below were obtained from several closely agreeing determinations made in this laboratory- PATTERSON AND SAVAGE : DETERMINATION OF DEXTROSE AND Duplicates should agree within 0.02 ml. Anhydrous dextrose (10 minutes’ heating) mg . . 1.33 1.0 0,667 0.5 0.4 0.2 Volume of 0.005 N sodium thiosulphate, ml . . 8.75 6.47 4.27 3.12 2.45 1.15 Anhydrous maltose (20 minutes’ heating), mg . . 2.0 1.5 1.0 0.667 0.4 0.2 Volume of 0.005 N sodium thiosulphate, ml . . 7.51 5.58 3-63 2.36 1.365 0,635 The purity of the sugars used in constructing the graphs was tested by paper chromato- graphy and polarimetry, and the appropriate corrections were made for moisture.Dec.19571 MALTOSE IN STARCH CONVERSION PRODUCTS 815 TESTING OF THE METHOD Mixtures prepared from anhydrous dextrose and maltose were quantitatively separated by the proposed method; for a mixture containing 10.0 mg of each sugar, 10.1 mg of each were found, and similarly for a mixture containing 14-0 mg of each sugar, 14.2 mg of each were found. Further, analyses of samples of liquid glucose gave reproducible figures that were in agreement with the results obtained by quantitative paper chromatography. In another experiment the aqueous and the ethanolic eluates from the column were collected in successive 5-ml portions, each of which was separately analysed for sugar.The results are illustrated in Fig. 2. Volume of eluate, ml 5 10 I S 20 25 30 35 40 45 50 ss 60 Dextrose (aqueous eluate) 5 1 0 15 20 25 30 35 40 45 SO 55 60 65 70 75 80 85 90 95 100 Maltose (7% ethanolic eluate) Fig. 2. Histogram for the recovery of dextrose and maltose from liquid glucose with use of batch No. 1 B.D.H. charcoal a t room temperature It will be seen that all the dextrose was contained in the first 50 ml of water and all the maltose in the first 75 ml of 7 per cent. ethanol. The 100 ml of each prescribed in the method therefore allowed adequate "clearance." illoreover, there is no indication that the later ethanolic fractions were eluting any maltotriose. After one of the routine determinations of dextrose and maltose in liquid glucose, 100 ml of 95 per cent.ethanol were passed through the column and collected separately. The remain- ing 90 ml of the aqueous and the 7 per cent. ethanolic eluates, and the 95 per cent. ethanol, were separately evaporated to dryness, and the residues were taken up in 1 drop of water and submitted to paper chromatography. The aqueous portion was found to contain dextrose with possibly a minute trace (less than 0.005 mg) of maltose, the 7 per cent. ethanolic portion contained only maltose and the concentrated ethanol portion, which contained triose, tetrose and so on, gave no trace of either dextrose or maltose. (It was necessary to evaporate the maltose portion in a partial vacuum at 70" C, since evaporation to dryness a t 100" C caused transformation of the maltose into a complex mixture of sugars.) We recommend that the amount of maltose put on to the column should be about 10 mg.I t is possible to put on considerably more than this amount before the column is so overloaded that some of the maltose appears in the 100-ml aqueous eluate with the dextrose, but con- tinued washing with large volumes of water after all the dextrose has been eluted will slowly remove maltose from a heavily over-loaded column. By restricting the amount of maltose put on to the column to about 10 mg, we have ensured that none of it is eluted by the first 200ml of water, i.e., by 100ml in excess of the amount of water specified for eluting the dextrose. A liquid glucose was specially prepared without the addition of sulphur dioxide, a considerable excess of sulphur dioxide over the amount normally present was added, and the dextrose and maltose were determined a t intervals during several months. The presence of the sulphur dioxide had no effect on the accuracy of the method. Most commercial liquid glucose contains a small amount of sulphur dioxide. VARIATIONS BETWEEN DIFFERENT BATCHES OF CARBON It was found on using a second batch of B.D.H.charcoal of the same description as the first, that, although the sugars could be recovered quantitatively as before, the recommended 95 to 100 ml of 7 per cent. ethanol were barely sufficient for complete elution of the whole of the maltose. The histogram shown in Fig. 3, which was obtained by collecting successive 10-ml fractions of aqueous and ethanolic eluates containing dextrose and maltose, illustrates this.AND O L L o u 816 PATTERSON AND SAVAGE : DETERMINATION OF DEXTROSE ~.~ Fraction No. I 2 3 .A 1 2 3 4 5 6 [Vol. 82 h, 7 8 9 10 II 10 20 30 40 50 60 70 80 90 100 110 Volume of eluate, ml 10 20 30 40 50 60 70 80 90 100 110 Dextrose (aqueous eluate) Maltose (7% ethanolic eluate) Fig. 3 . Histogram for recovery of dextrose and maltose from liquid glucose with use of batch No. 2 B.D.H. charcoal at room temperature A third batch of B.D.H. charcoal of the same description was found to yield maltose quantitatively to 7 per cent. ethanol more readily than either of the first two batches, as shown by the histogram, Fig. 4. Volume of eluate, ml 1 0 20 30 40 50 60 70 80 90 100 110 10 20 30 40 50 60 70 80 90 100 110 Dextrose (aqueous eluate) Maltose (7% ethanolic eluate) Fig.4. Histogram for recovery of dextrose and maltose from liquid glucose with use of batch No. 3 B.D.H. charcoal at room temperature I n view of this variability from batch to batch of carbon, it appears advisable to test each batch by collecting the eluates in fractions initially before putting it into routine use, to determine the conditions most suitable for ensuring a clean separation of sugars in convenient volumes of eluate. An initial check of this type also has the advantage of providing a ready test for types of carbon other than the one we recommend, and the same principle could be extended to the elution of higher sugars with other concentrations of aqueous ethanol.We found that the volume of 7 per cent. ethanol required for complete removal of maltose from a column is considerably influenced by temperature, as the histograms in Figs. 5, 6, 7 and 8 show. h L 18090l00l10120130 --- --- Fraction No. I 2 3 4 5 6 7 8 9 10 I 2 3 4 5 6 7 8 9 1 0 Fig. 5. Histogram for recovery of Fig. 6. Histogram for maltose from batch No. 2 B.D.H. char- recovery of maltose from coal at 40' C batch No. 2 B.D.H. charcoal at 3OoCDec. 19571 MALTOSE IN STARCH CONVERSION PRODUCTS Volume of 7% ethanol, 2.0 :LA I .o I0 2 0 3 0 4 0 5 0 60 70 RO90100110120130 817 --- --- Fraction No. I 2 3 4 5 6 7 8 9 10 I 2 3 4 5 6 7 8 9 1 0 Fig. 7. Histogram for recovery of Fig. 8. Histogram for maltose from batch No. 2 B.D.H. recovery of maltose from charcoal a t 2OoC batch No.2 B.D.H. charcoal a t 10°C It is largely a matter of individual choice whether the temperature or the volume of the eluting liquid is adjusted in order to obtain satisfactory results from any one carbon. These histograms show that if a 100-ml flask is to be used for collecting maltose washed out from charcoal No. 2, a working temperature of 30" C is only just sufficient to give quantitative recovery; if the temperature is significantly below this, the results will be low. Volume of Maltose Triose - Fig. 9. Histogram for recovery of maltose from batch No. 3 B.D.H. charcoal at 30' C 7% ethanol, m l 1 0 20 304( -+ Fraction No. I 2 1 i,,,,,. 360 70 80 901W110120130 4 3 4 5 6 7 8 9 1 0 2.0 :k I .o 10 20 30 1 -+ 1 2 - L 3 68 3 c 70 80 90 IW 110 120 13C 4 5 6 7 8 9 1 0 Fig.10. Histogram for recovery of Fig. 11. Histogram for maltose from batch No. 3 B.D.H. char- recovery of maltose from coal at 16OC batch No. 3 B.D.H. charcoal at 10°C818 [Vol. 82 We have not experienced any difficulty in recovering dextrose quantitatively from a carbon column in 100 ml of water. Recovery is speeded up slightly at increased tempera- tures, but the effect is very much less marked than with maltose. Histograms for the elution of maltose from the third batch of B.D.H. charcoal at 30°, 16" and 10°C are shown in SLATER: DETERMINATION OF SMALL AMOUNTS OF LACTOSE I N Figs. 9, 10 and 11. When this carbon from batch No. 3 was used, all the maltose appears to have been eluted into the first 40 to 60 ml at 30" C, and was followed immediately by ihe triose.Clearly, 30" C is too high a working temperature for this carbon, the most suitable being the normal room temperature of 1.5' to 18" C. PROCEDURE FOR TESTING A BATCH OF CARBON The histograms at various temperatures were obtained by preparing a 7-g column as usual with equal parts of the carbon under test and kieselguhr, and washing with 100 ml of water. By pipette, 5 ml of a 2 per cent. solution of liquid glucose were put on to the column, and the dextrose was removed, either into 100 ml of water or into the required fractions, numbered 6-inch x %-inch test-tubes graduated at 10 ml and 20 ml being used. Then surplus water was removed from the top of the column, 7 per cent. ethanol was substituted for water in the tap-funnel, and the ethanol fractions were collected.All solutions were heated or cooled to the working temperature before being passed through the column. The usual Somogyi sugar determinations were carried out on 5 ml of each fraction, and the ordinates of the histograms are differences between titration and blank values. Differences in titration of less than 0.1 ml of 0.005 N sodium thiosulphate have been disregarded, Quantitative recovery cannot be expected when the eluate is split up into many fractions; the histograms merely show the position at which the sugars are recovered. When a new carbon is used, it is also necessary to confirm, by putting through known amounts of pure sugars, collecting 100 ml each of aqueous and ethanolic eluates and carrying out a quantitative determination at the selected temperature, that recovery is in fact quantitative. We are grateful to members of the Sub-committee of Subject 26 (Starch-Conversion Products) of the British National Committee of the International Commission for Uniform Methods of Sugar Analysis for the considerable amount of work they have done in testing the proposed method, to Mr. J. L. Buchan, MSc., for his interest and advice, and to the Govern- ment Chemist, Dr. G. M. Bennett, C.B., F.R.S., for permission to publish this paper. REFERENCES 1. 2. 3. 4. 5. CUSTOM HOUSE Whistler, R. L., and Durso, D. F., J . Amer. Chem. Sac., 1952, 72, 677. McDonald, E. J., and Perry, R. E., J . Res. Nut. Bur. Stand., 1951, 47, 363. Jermyn, M. A., Aust. J . Chem., 1957, 10, 55. Somogyi, M., J . Bid. Chem., 1945, 160, 61. - , Zbid., 1952, 195, 19. THE LABORATORY LONDON, E.C.3 July 3rd, 1957

ISSN:0003-2654    

DOI:10.1039/AN9578200812

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

818 SLATER: DETERMINATION OF SMALL AMOUNTS OF LACTOSE I N [Vol. 82 A Rapid Method for the Determination of Small Amounts of Lactose in Milk and Tissue Suspensions of Mammary Gland BY T. F. SLATER A method is described for the determination of lactose in milk and tissue suspensions of mammary gland. The method is based on the development of a green colour on heating lactose, orcinol and ferric ions in concentrated acid solution. LACTOSE is the only carbohydrate occurring in normal cow and rat milk in other than trace concentration^.^^^^^^^ Lactose can, therefore, be deterniiiied in them by a non-specific method of carbohydrate determination. The normal routine methods used for the deter- mination of lactose in milk, for instance, reduction of cupric salts in alkaline solution and The method is applicable to microgram amounts of lactose.Dec. 19571 MILK AND TISSUE SUSPEKSIONS OF MAMMARY GLAND 819 the chloramine-T and polarimetric methods, are not suitable for the determination of ex- tremely small concentrations of lactose.Colori- metric methods described for the determination of lactose and applicable to the rapid analysis of numerous samples are either not simple enough for routine practice or have involved critical stages that must be carefully controlled, for example, the ferric chloride - sodium carbonate method of Mitra and Roy,6 the tetrazolium method of Mattson and Jensene and the methyl- amine method of Malpress and Rilorrison.7 Similar difficulties arise with determinations of lactose on tissue suspensions of mammary gland. Rat mammary gland tissue contains a variable content of milk retained in the ducts and alveoli of the gland.In full lactation, the retained milk may account for up to 60 per cent. of the gross wet weight of the gland.* This retained milk contributes appreciably to the total nitrogen content of the whole tissue homogenate, hence invalidating the direct usage of total nitrogen values as a measure of the tissue present in the ample.^ It is therefore of importance, in quantitative experiments involving the mammary gland, to correct analyses on rat mammary tissue for the milk retained in excised gland. Again, normal methods are not suitable for determinations of lactose in small samples of rat mammary tissue suspensions in which the amount of lactose present may be less than 100 pg per sample.Orcinol (3 : 5-dihydroxytoluene) reacts with pentoses in the presence of ferric ions and concentrated hydrochloric acid to give an intensely green chromogen.10 This is the basis of the standard method for estimating ribonucleic acid in animal tissues.ll ,12 Orcinol also reacts with other carbohydrates under the conditions outlined. HexoseP and disaccharides, for instance, give brownish hazy solutions and, as the concentration of the sugar is increased, a heavy brown deposit forms, which is soluble in ethanol to give a clear yellow-brown solution. The orcinol reaction has been adapted in this investigation to provide a quantitative method for determining lactose in the absence of other carbohydrates. The colorimetric method described below enables 5 to 200 pg of lactose to be determined rapidly and accurately in milk and mammary gland suspensions; further, a number of determinations can readily be performed concurrently.Further, the methods are time-consuming. EXPERIMENTAL An investigation of the variables in the lactose - orcinol reaction was carried out in order to find the optimal conditions for the reaction. VARIABLES IN THE LACTOSE - ORCINOL REACTION- Absorption c.urve-One millilitre of a lactose solution (500 pg per ml in water) was heated on a boiling-water bath for 50 minutes with a 0.1 per cent. solution of ferric chloride in con- centrated hydrochloric acid containing 0.5 per cent. of orcinol. After the solution had cooled, ethanol was added to clear the haze and the absorption spectrum was determined by means of a Beckmann SP500 spectrophotometer, 1-cm cells being used.The absorption spectrum is shown in Fig. 1 (a). Stability of the colow-The stability of the colour formed is illustrated by the fact that the extinction a t 670 m p decreased by approximately 4 per cent. in 24 hours and the extinc- tion at 440 mp increased by approximately 6 per cent. in 24 hours. Boiling time-Mixtures of 3 ml of a 0.1 per cent. solution of ferric chloride in concentrated hydrochloric acid, 0.3 ml of a 10 per cent. solution of orcinol in ethanol and 1 ml of a solution of lactose (approximately 300 pg per ml in water) were boiled for different times in a water bath. Fig. 1 (b) shows how the absorptions a t 440 m p and 670 m p increased with boiling time.I t can be seen that the green chromogen (wavelength maximum 670 mp) was formed more rapidly than the brown chromogen (wavelength maximum 440 mp) and reached a plateau after 10 minutes’ boiling. The positions of the absorption maxima did not vary with boiling time. Concentration of orcinol-Mixtures of 3 ml of a 0.1 per cent. solution of ferric chloride in concentrated hydrochloric acid, 1 ml of a solution of lactose (200 pg per ml in water), x ml of orcinol solution of various concentrations in 95 per cent. ethanol and (1 - x ) ml of water were boiled for 15 minutes in a water bath. After the solutions had cooled, the absorptions a t 670mp were measured. The extinctions of the solutions increased as the concentration of orcinol increased and reached a plateau for final orcinol concentrations over 1 per cent.820 Concentration of ferric ions-The influence of ferric ions on the lactose - orcinol reaction is shown in Fig.1 (c). Ferric chloride solutions of various concentrations in concentrated hydrochloric acid were prepared and 3 ml of each solution were boiled for 15 minutes with 0.5 ml of a 10 per cent. solution of orcinol and 1 ml of a solution of lactose (200 pg per ml in water). After the solution had cooled, 2 ml of ethanol were added and the absorption a t 670mp was measured. SLATER: DETERMINATION OF SMALL AMOUNTS OF LACTOSE I N [Vol. a2 I 420 460 500 540 580 620 660 700 Wavelength, rnl 0.4 0.2 .- 7?rl LE .- oo 0.05 0.10 0.15 0.20 Concentration of ferric ions, % Wavelength, mH 20.61 n 5 '0 1.0 2.0 3.0 4.0 5.0 6.0 Final concentration of hydrochloric acid, N Y 0 0 0 I I I I 0 10 20 30 40 50 60 70 Time of boiling, minutes 1.6 1.4 - 1.2 - :: 1.0- Y 0.8- .- 06- - 0 .- Y Y X w / .- 014 02Yi 0 0 200 400 600 800 1000 Lactose added,pg Fig.1. Variables in the lactose - orcinol reaction Fig. l(a). Absorption spectrum of the solution obtained by heating lactose, orcinol, ferric chloride Fig. l ( b ) . Effect of boiling time on the extinction: curve A, a t 440 mp; curve B, a t 670 mp Fig. l(c). Effect of the concentration of ferric ions on the extinction at 670mp. The con- centration of ferric ions refers to the concentration of ferric chloride in the 3 ml of ferric chloride - hydrochloric acid solution added Effect of 2 : 2'dpyridyl on the absorption spectrum of a solution of lactose, orcinol and concentrated hydrochloric acid Effect of acid concentration on the extinction at 670 mp: curve A, sucrose; curve B, lactose Calibration curve for 0 to 200 pg of lactose with the correlation line shown Deviation from linearity of the calibration curve for high concentrations of lactose.and concentrated hydrochloric acid Fig. l ( d ) . Fig. l ( e ) . Fig. l(f). Fig. l ( g ) . The straight-line part of the curve is the correlation line from Fig. l(f)Dec. 19573 MILK AND TISSUE SUSPENSIONS OF MAMMARY GLAND 82 1 It can be seen that the presence of ferric ions is essential for the production of the green chromogen, whereas the brown chromogen is apparently formed in the absence of ferric ions. This effect is more clearly seen in Fig. 1 (d), which gives the absorption spectrum obtained after heating for 15 minutes a mixture of 3 m l of the concentrated hydrochloric acid, 0.5 ml of a 10 per cent.solution of orcinol, 1 ml of a solution of lactose (100 pg per ml in water) and 0.1 ml of a 0.1 per cent. solution of 2 : 2'-dipyridyl. The position of the maxi- mum in the absence of ferric ions appears to be somewhat displaced from that of the corresponding peak obtained in their presence. Concentration of acid-The green chromogen is formed only in the presence of con- centrated acid. Mixtures of 0.5 ml of a 1 per cent. solution of ferric chloride in water, x ml of concentrated hydrochloric acid, (3 - x ) ml of water, 0.5 ml of a 10 per cent. solution of orcinol and 1 ml of a solution of lactose (100 pg per ml) were boiled for 6 minutes in a water bath.After the solution had cooled, 2 ml of ethanol were added and the extinctions a t 670 m p were measured. Fig. 1 (e) shows the results obtained with the lactose solution and also with a solution of sucrose (1000 pg per ml). Optimal conditions-The conditions adopted for the determination of lactose in protein- free solutions were heating for 10 minutes in a boiling-water bath 3 ml of a 0.1 per cent. solution of ferric chloride in concentrated hydrochloric acid, 1 ml of the sample (containing 0 to 200pg of lactose per ml.) and 0 6 m l of a 10 per cent. solution of orcinol in absolute ethanol. After cooling, 2 ml of absolute ethanol were added and the absorption a t 670 m p was measured. A typical calibration curve is shown in Fig.I(/); at higher concentrations of lactose than mentioned above, the response deviates from linearity, as shown in Fig. l(g). With the procedure described, two levels of lactose were used to determine the variability of the method; 6 samples at each level of lactose were taken. The mean values, with their standard errors, were 106 The standard errors quoted are less than 3 per cent. of the corresponding mean values. 0.6 pg per ml and 56 & 1.2 pg per ml. APPLICATION OF THE METHOD TO MILK AND TO MAMMARY TISSUE SUSPE~SIONS- The determination of lactose in milk or in suspensions of mammary gland in general involves the prior precipitation of protein. Ferric hydroxide has often been used as a protein precipitating agentsJ4 and was used in the early stages of this present work.Samples of whole cow or rat milk (obtained after prior injection of 5i.u. of oxytocin) were treated with colloidal ferric hydroxide to precipitate the protein. Lactose was deter- mined in aliquots of the filtrate. Two methods of determination were used; first, the orcinol method described here and, secondly, the chloramine-T method of Hinton and Macara,15 as modified by Folley and Greenbaum.* In general the two methods gave closely similar results (see Table I), but occasionally, even on clear milk filtrates, higher lactose titres were obtained by the chloramine-T method than by the orcinol method. Typical results are shown in Fig. 2, which illustrates the variability between the two methods. TABLE I COMPARISON OF THE ORCINOL METHOD WITH OTHER METHODS FOR DETERMINING LACTOSE Lactose found by orcinol method with ferric hydroxide Lactose found by alternative method with ferric hydroxide used as Sample used as precipitant for protein, precipitant for protein, Pg Per ml Pg Per ml Cows' milk 100.0 100.0, 108.0, 103.5, 100.0 by chloramine-T method 100.0 101.6,* 99.6,b 99.5C by orcinol method 100.0 2.8* 97.4 i.0.66 by chloramine-T method," 99.4 0.66 by Benedict's method* 109.0, 105.0, 102.0, 96.0 by methylamine method Rat mammary gland 100.0 100.0 152.4 8.9 by chloramine-T method? * Mean and standard error of 6 determinations. t Mean and standard error of 26 determinations. a Zinc hydroxide used as precipitant for protein. b Trichloroacetic acid used as precipitant for protein. 0 Tungstophosphoric acid used as precipitant for protein.822 SLATER: DETERMINATION OF SMALL AMOUNTS OF LACTOSE IN [Vol.82 In a similar series of experiments, 1 in 10 homogenates of rat or rabbit mammary gland were prepared by adding 9 volumes of ice-cold glass-distilled water to 1 volume of gland and homogenising in a top-driven Waring Blendor.16 The suspension was strained through muslin to remove intractable connective tissue. Aliquots of the suspension were deproteinised with colloidal ferric hydroxide and lactose was determined in the filtrate by both the orcinol and the chloramine-T methods. From Table I it can be seen that almost invariably the chloramine-T method gave appreciably higher results. The ratio of the results by the two methods was variable from animal to animal, indicating that some factor in the methods was not under strict control.The precipitation stage seemed the obvious cause of the discrepancies. 10001 r' / 0 I 1 , 1 0 200 400 600 800 I( ~~. Lactose found by method B, yg per ml of filtrate Fig. 2. Correlation between results by the chloramine-T method (method A) and the orcinol method (method B). The broken line represents the theoretical agreement between the two methods s $ / 400 ," 0 40 80 120 160 200 240 280 Protein present, pg per rnl of filtrate Fig. 3(a). Effect of protein on the determination of lactose in milk filtrates: curve A, chloramine-T method; curve B, orcinol method 5 01-1 2 0 10 80 120 160 200 240 Protein present, pg per ml of filtrate Fig. 3 ( b ) . Effect of protein on the determination of lactose in tissue sus- pensions of rat mammary gland : curve A, chloramine-T method; curve B, orcinol method It was thought probable that the higher values obtained by the chloramine-T method were the result of incomplete precipitation of protein from the milk or mammary gland sus- pensions by ferric hydroxide, the soluble protein remaining in the filtrate affecting the chlor- amine-T method more than the orcinol method. This was considered likely from the study of the effects of protein on both methods of lactose determination.The protein in both mammary tissue suspensions and in milk was precipitated to various degrees by adding increasing amounts of colloidal ferric hydroxide. After the protein had been removed by filtration, a series of more or less opalescent solutions was obtained, in which the protein was determined by the colorimetric method of Lowry, Rosebrough, Farr and RandaU.17 Determinations of lactose by both the chloramine-T and the orcinol methods were carriedDec.19571 MILK AND TISSUE SUSPENSIONS OF MAXMARY GLAND 823 out on aliquots of these suspensions. Fig. 3 (a) shows that for milk the chloramine-T method was affected by the presence’ of protein; increasing concentrations of protein increased the “apparent lactose content.” The orcinol method did not appear to be affected to any appreciable degree by small concentrations of protein. It therefore appeared that the difference between the two methods of determination was due to incomplete precipitation of protein, possibly albumin, by ferric hydroxide.When protein is completely removed from milk filtrates, or when the methods are compared on aqueous solutions of lactose, the two methods agree, as can be seen from Table I. However, although the complete removal of protein from mammary gland suspensions does decrease the difference between the two methods of determination, even with protein- free filtrates of mammary gland the methods gave different results (see Figure 3(b) and Table I). The chloramine-T method gave consistently higher results than the orcinol method. It seems probable that this was the result of some reducing component of mammary sus- pensions interfering under the mild conditions of the chloramine-T reaction and leading to falsely high results, rather than of a component of the mammary gland inhibiting the orcinol reaction, which involves hydrolysis by boiling concentrated hydrochloric acid.The same discrepancy between the chloramine-T and orcinol methods occurred with precipitants other than ferric hydroxide, e.g., zinc hydroxide, although usually the disagreement was not so great, as the filtrates were virtually protein-free. However, it can be seen from Table I that the methylamine method of Malpress and Morrison7 gave results similar to those by the orcinol method when used on rat mammary gland suspensions. Various other protein precipitating agents were used in lactose determinations by the orcinol method on milk or mammary gland suspensions. Both the zinc hydroxide reagent of Letonoff’s and 5 per cent, trichloroacetic acid were found to be satisfactory.However, a disadvantage in the routine use of zinc hydroxide is that relatively large amounts are required to deproteinise milk or mammary gland suspensions completely. Trichloroacetic acid (5 per cent.) is a satisfactory precipitant for both milk and mammary gland suspensions and does not interfere with the orcinol reaction. Determinations were carried out on aqueous solutions of lactose with and without the addition of trichloroacetic acid to a final concentration of 10 per cent. The results show that, compared with a value of 100 pg per ml of lactose obtained in the absence of trichloro- acetic acid, the mean of 8 determinations of lactose in the presence of trichloroacetic acid was 99.9 j: 1.1 pg per ml. Complete recovery of added lactose from milk or mammary tissue suspensions was obtained by using ferric hydroxide or zinc hydroxide as precipitant, and the orcinol method for the determination (see Table 11).TABLE I1 RECOVERY OF ADDED LACTOSE FROM MILK AND TISSUE SUSPENSIONS OF MAMMARY GLAND Lactose recovered by Lactose recovered by chloramine-T method orcinol method r * > /G % 30.08 97.0 56.5 54.9a 97.2 60.08 106.0 84.7 79.2a 93.4 85.08 100.3 141.2 136.08 96.3 135.58 96.0 38-1 38.0,s 99.7, 38.0b 99.7 76.0 81.0,& 106.6, 72.0b 94.7 61.9 113.0 7 Sample Lactose added, tG Pg Cows’ milk 30.9 29.9a 97.0 Mean 98.1 Standard error 1.4 a Ferric hydroxide used as precipitant for protein. b Zinc hydroxide used as precipitant for protein. Rat mammary gland 65.08 105.0 112.0” 99.2 100.5 1.6 Procedure foy determining lactose-A I-ml sample of milk was deproteinised by adding trichloroacetic acid to a final concentration of 5 per cent.and the volume was made up to824 [Vol. 82 250 ml with water. The solution was filtered and lactose was determined in 1-ml portions of the clear filtrate as described under “Optimal conditions,” p. 821. The volumes stated can be scaled down proportionately as required. For tissue suspensions of rat mammary gland, 2 t o 5 ml of a 1 in 10 water homogenate were deproteinised by adding trichloroacetic acid to a final concentration of 5 per cent. and the volume was made up to 25 ml with water. The solution was filtered and 1-ml portions of the filtrate were taken for the determination of lactose as before. ABSORPTION CURVES OF OTHER CARBOHYDRATES- Various 1-ml samples of other carbohydrates were substituted for lactose in the procedure given under “Optimal conditions,” p.821. The absorption spectra of these solutions are shown in Fig. 4. It can be seen that all the carbohydrates tested, with the exception of xylose, gave similar absorption spectra. The orcinol method therefore cannot be used to determine lactose in the presence of other carbohydrates, except when they are present in trace amounts. SLATER: DETERMINATION OF SMALL AMOUNTS OF LACTOSE IN Xylose showed no peak near 540 mp. t Wavelength, my Absorption spectra: curve A, 1OO.Opg of xylose per ml; curve B, 203.8 p g of galactose per ml; curve C, 207.4 p g of glucose per ml; curve D, 190.6 pg of maltoseper ml; curve E, 196.3 p g of sucrose per ml.Extinction values were measured with a Unicam SP500 spectrophotometer in 1-cm cells Fig. 4. DISCUSSION The method described is suitable for the rapid determination of microgram amounts of lactose in milk or in mammary tissue suspensions. An advantage of the method is that it is insensitive to traces of protein, in contrast to the chloramine-T method. The orcinol method readily determines 10 pg of lactose. Briicknerlg has described the reaction of disaccharides with orcinol in concentrated sulphuric acid, but his work is not suitable for the routine determination of lactose, as the timeDec. 19571 MILK AND TISSUE SUSPENSIONS OF MAMMARY GLAND 825 of heating is short and critical. The advantage of the orcinol reaction described here is that none of the reaction variables is used in a critical region.I express my gratitude to Dr. A. L. Greenbaum for his help and encouragement of this work and to Mr. L. McCollum and Mr. D. Planterose for skilled assistance. I also express my gratitude both to the Agricultural Research Council for a grant during the earlystages of this work and to the Beit Memorial Trust for the award of a Fellowship. J . 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. REFERENCES Davies, W. L., “The Chemistry of Milk,” Chapman & Hall Ltd., London, 1936 Caputto, R., and Trucco, R. E., Natuve, 1952, 169, 1061. Malpress, F. H., and Morrison, A. B., Ibid., 1952, 169, 1103. Roberts, H. R., Pettinati, J. D., and Bucek, W., J . Dairy Sci., 1954, 37, 538. Mitra, S. N., and Roy, J. K., J . Indian Chem. Soc., I n d . Ed., 1955, 18, 34. Mattson, A. M., and Jensen, C. O., Anal. Chem., 1950, 22, 182. Malpress, F. H., and Morrison, A. B., Biochem. j . , 1949, 45, 455. Folley, S. J., and Greenbaum, A. L., Ibid., 1947, 41, 261. Greenbaum, A. L., and Slater, T. F., Ibid., 1957, 66, 148. Bial, M., Dtsch. Med. Wschr., 1902, 28, 253. Mejbaum, W., 2. Physiol. Chem., 1939, 258, 117. Dische, 2.. in Chargaff, E., and Davidson, J. N., Editovs, “The Nucleic Acids,” Academic Press Inc., Brown, A. H., Arch. Biochem. Biophys., 1946, 11, 269. McDowell, A. K. R., J . Dairy Res., 1941, 12, 131. Hinton, C. L., and Macara, T., Analyst, 1927, 52, 668. Folley, S. J., and Watson, S. C., Biochem. J . , 1948, 42, 204. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J., J . Bid. Chem., 1951, 193, 265. Letonoff, T. V., Ibid., 1934, 106, 693. Briickner, J.. Biochem. J . , 1955, 60, 200. New York. 1955, Volume I. DEPARTMENT OF BIOCHEMISTRY UNIVERSITY COLLEGE LONDON April 1861a. 1935

ISSN:0003-2654    

DOI:10.1039/AN9578200818

出版商:RSC    

年代:1957

数据来源: RSC