摘要:

AUGUST, 1963 THE ANALYST Vol. 88, No. 1049 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY SECRETARYSHIP OF THE SOCIETY THE Council has appointed Miss P. E. Hutchinson to be Secretary of the Society. Miss Hutchinson has been Assistant Secretary since 1955. SECRETARIAT OF THE ANALYTICAL METHODS COMMITTEE THE Council of the Society has appointed Mr. P. W. Shallis to be Secretary of the Analytical Methods Committee in place of Dr. C. H. Tinker (resigned). Mr. Shallis has been Assistant Editor of The Analyst since 1957. NEW MEMBERS ORDINARY MEMBERS Roy Thomas Brittain, B.Pharm., M.P.S., M.I.Bio1. ; John Ivor Dennis; David John Dimmock; Geoffrey Edwin Downes, B.Sc.(Lond.), A.R.I.C. ; John Esson, B.A. , B.Sc.(Oxon.) ; Ralph Waldo France, B.Sc., Dip.Ed., Ph.D., F.1nst.P. ; Derek Alfred Greenwood; George Rumbold Jamieson, B.Sc.(Lond.), F.R.I.C.; Geoffrey David Ratcliffe Jarrett ; Allan Raymond Lister; Enzo Mannucci, Doc.Pharm. (Siena) ; John Michael Murphy, BSc. (Manc.) ; David John Pannett, B.Sc.(Sheff .) ; John Anthony Potter, A.R.I.C., Dip.App.Chem. ; Lina Raffa; Ivor Smith, B.Sc., Ph.D.(Lond.), F.R.I.C., M.I.Bio1. JUNIOR MEMBER Brian Thomas Ashurst, A.C.T. (Liv.), A.R.I.C. DEATH William Branch Pollard. WE record with regret the death of NORTH OF ENGLAND SECTION THE twenty-sixth Summer Meeting of the Section was held at the Savoy Hotel, Blackpool, from Friday, June 14th, to Monday, June 17th, 1963. The Chairman of the Section, Mr. C. J. House, B.Sc., A.R.C.S., F.R.I.C., presided over an Ordinary Meeting at 10.30 a.m. on Saturday, June 15th, at which H.Pritchard, M.Sc., F.R.I.C., gave a lecture entitled “The Private Analyst and Public.’’ On the Saturday evening the party saw the Morecombe and Wise Show, and on the Sunday afternoon made a coach tour to Windermere, taking tea en route. BIOLOGICAL METHODS GROUP THE Summer Meeting of the Group was held on Thursday, June 13th, 1963, and took the form of a visit to the Control Laboratories of the United Dairies Ltd., Wood Lane, London, w.12. The morning session was conducted under the Chairmanship of the Vice-chairman of the Biological Methods Group, Dr. M. W. Parkes, BSc. After the Group had been welcomed by R. J. MacWalter, BSc., Ph.D.(Lond.), A.M.I.Chem.E., F.R.I.C., the following papers were read : “Aspects of the Natural Composition and Hygienic Quality Schemes,” by R. C. Wright, B.Sc., Ph.D. ; “Antibiotics in Milk,” by J. Tramer, B.Sc., Ph.D. After lunch the Group was conducted around the laboratories. The thanks of the Group were expressed to the Company by Mr. W. A. Broom, B.Sc., 57 1 F.R.I.C., Chairman of the Biological Methods Group.

ISSN:0003-2654    

DOI:10.1039/AN9638800571

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

572 JAMES: METHODS OF SEPARATION OF [Analyst, Vol. 88 Methods of Separation of Long-chnin Unsaturated Fatty Acids A Review* BY A. T. JAMES ( Unilevev Reseavch Labovatory, Colworth House, Shavnbvook, Beds.) The isolation of virtually any unsaturated fatty acid can be accomplished on the gram scale by counter-current distribution involving use of the silver- ion containing solvent systems described by Dutton, Scholfield and Jones. The technique is, however, too slow for analysis and too cumbersome for amounts less than a gram. In this range, chromatographic procedures, especially liquid - liquid columns, thin-layer plates with the silver-ion corn- plexing systems and gas - liquid chromatography are the methods of choice. Accurate and rapid analyses are best obtained by gas - liquid chromato- graphy, but the combination of this technique with that of thin-layer chromatography is probably superior to all others.MODERN analytical techniques have shown how complex are the mixtures of naturally occurring fatty acids. Indeed, the list is now so extensive that it almost rivals that of the petroleum hydrocarbons. The complexity of these mixtures places great demands on analytical and separation techniques. The unique property of chromatographic systems lies in their ability to exploit relatively small differences in chemical structure between solute molecules. Unlike the older techniques of fractional distillation, crystallisation, etc., the chromatogram cannot conveniently deal with large quantities, but it can provide a quantitative yield of pure substances and is a better analytical tool.For these reasons this Review will be limited to counter-current distribution and the various chromatographic techniques. It is hoped to show the limitations of each of the individual techniques and suggest how they may be used to best advantage. COUNTER-CURRENT DISTRIBUTION This technique, developed originally by Craig, is capable of achieving purifications similar to those obtainable by chromatography. Its major disadvantages are cumbersomeness and slowness of operation. Its advantage lies in the large volumes of solvent that can be easily handled so that amounts of a few grams of a pure substance may be obtained. The plate efficiency of the counter-current machine is a direct function of the number of transfers employed, and the time of passage through a single plate is much slower than in a chromato- graphic column.In an early paper Ahrens and Craig1 described the separation of oleic, linoleic and lino- lenic acids with a solvent system of heptane, methanol, formamide and acetic acid. In this system, the distribution coefficient is altered to the same extent by the introduction of a double bond as by the removal of two CH, groups from the chain. Consequently, oleic and palmitic acids, and linoleic and myristic acids cannot be separated. A new type of solvent system has recently been introduced by Dutton, Scholfield and Jones,2 who incorporate silver salts into the more aqueous of the two phases. A reversible complexing between the double bonds of unsaturated fatty acids and silver ions occurs, producing a large change in the distribution coefficient of the unsaturated fatty acids.This complexing occurs more readily with cis than with trans double bonds and hence affords a means of separating such isomers (see Figs. 1, 2 and 3). Many such procedures could be used to exploit more fully the capabilities of counter-current distribution. LIQUID - LIQUID COLUMNS- that the phases used possess a small dielectric constant. only by having a relatively non-polar stationary phase. CHROMATOGRAPHIC SYSTEMS The non-polar nature of the backbone of the long-chain fatty acids makes it necessary Suitable columns can be made This approach was first exploited * Reprints of this paper will be available shortly. For details, please see p.658.August, 19631 LONG-CHAIN UNSATURATED FATTY ACIDS 600. ? 5 0 a - - 400 L 0 L U M .- ; 200 573 - r C ? - L I cis, 700 800 900 1000 I100 Fig. 1. Counter-current distribution in a 200- tube automatic instrument of methyl elaidate and methyl oleate between light petroleum and 0.2 M silver nitrate in 90 per cent. methanol. Sample was re-cycled t o Ti00 transfers, and fractions wen moved by the single withdrawal procedure (after Dutton, Scholfield and Jones2) Transfer number *0° I Trans, Tronr 50 5 00 600 700 800 Transfer number Fig. 2 . Counter-current distribution of selenium- isomerised methyl linoleate between light petroleum and 0.2 M silver nitrate in 90 per cent. methanol. Sample was recycled to 450 transfers, and fractions werc rcmoved by the single withdrawal procedure (after Dutton, Scholfield and Jones2) 7 njugated -7- U nconj ugated Cis, Cis 200 3 00 400 Transfer number Fig.3. C,ounter-currentid stribution of alkali-isomerised methyl linoleate between light petroleum and 0.2 M silver nitrate in 90 per cent. methanol. Fractions removed by the single withdrawal procedure (after Dutton, Scholfield and Jones2)574 JAMES: METHODS OF SEPARATION OF [Analyst, Vol. 88 by Howard and Martin,3 who used liquid paraffin and aqueous acetone or methanol as the phase pairs, giving rise to the term “reversed phase” column, since all the earlier liquid - liquid chromatograms had the most polar liquid as the stationary phase. High molecular weight substances swellable by aqueous organic solvents can act not only as stationary hydrocarbon-like liquids, but also as the mechanical support, thus obviating the need for keiselguhr.Boldingh4 used powdered rubber and Hirsch5 used a powdered solid polymerised vegetable oil (Factice). In essence all these columns function in the same way and suffer from the disadvantage of the same overlap fractions described under counter-current dis- tribution. The silver-ion solvent systems described by Dutton et aL2 would, when suitably modified, also give excellent separation of the unsaturated acids. Paper chromatograms prepared with the same types of solvent system are subject to the same disadvantages; for a recent review see Viswanathan and Meera Rai.6 Schlenk et al.’ showed that chromatograms made at low temperatures gave improved separation of the saturated and unsaturated fatty acids owing to a greater change in distribution coefficient of the unsaturated acids.I Cis , 15 30 45 60 Volume of eluate, ml Fig. 4. Column chromatography of 10 mg each of a mixture of methyl stearate, elaidate and oleate. Adsorbent, 2 g of silver nitrate on silica; column height, 11 cm; column diameter, 8 mm; eluent, benzene - light petroleum (10 + 90 to 30 + 70) ; rate, 0.5 ml per minute (after de Vrieslo) LIQUID - LIQUID CHROMATOGRAPHIC TECHNIQUES- Thin-layer chromatography and silica gel columns give similar types of separation, the former giving higher resolution. Such columns are capable of resolving cholesterol esters of saturated and unsaturated acids, e.g., saturated acids are eluted as a group followed by oleate, linoleate and arachidonate.8 Simple esters could be similarly resolved.Thin-layer chromatography of fatty acid methyl esters after formation of the acetoxy- mercuri-methoxy derivatives (produced by reaction between the unsaturated acids and mercuric acetate) gives a group separation of saturated and unsaturated acids with a light petroleum - diethyl ether (4 + 1 v/v) moving phasesg After the plate has been dried a second solvent flowing in the same direction (n-propanol- glacial acetic acid, 100 + 1 v/v) and gives a separation of monoenoic acids (R, 0.85), dienoic acids (a double spot, R, 0.45 and 0.55) and trienoic acids (R, 0.15), whereas more highly unsaturated acids remain at the origin. Larger amounts can be obtained by streaking a sample across the plate.The use of silver ion complexes of the unsaturated acids has greatly improved separations. de VrieslO described good separations of methyl esters of unsaturated acids and glycerides on silver nitrate impregnated silicic acid columns (see Figs. 4 and 5). Such complexes have also been used by Morrisll on thin-layer plates, showing excellent sey aration of cis-trans isomers and of unsaturated acids of increasing degree of unsaturation (see Figs. 6 and 7).August, 19631 LONG-CHAIN UNSATURATED FATTY ACIDS 575 Such thin-layer and column techniques in conjunction with gas - liquid chromatography represent the most powerful combination suitable for the isolation of milligram amounts of unsaturated fatty acids. GAS-CHROMATOGRAPHIC SEPARATIONS- One of the major advantages of the gas - liquid chromatogram is that the moving phase is a permanent gas whose interactions with the solutes can be ignored.The solubility of the moving phase in the stationary phase is so small that this can also be ignored, and the only interaction forces that need be considered are interactions between solute and solvent. It is for this reason that the relation between chromatographic behaviour and structure of solute and solvent are more readily understandable in gas - liquid than in other types of chromatography. SOLUTION EFFECTS IN THE LIQUID PHASE- There are two major classes of solute - solvent interactions that need be considered. First, London Dispersion Forces, which occur in all solvents. These interactions are due primarily to the vibration of bonds setting up oscillating dipoles and so giving rise to weak 20 - E I5 L a, Q E" 6 10 0 .- Y L u i ! 5 V 0 I00 I50 200 25 0 Volume of eluate, mI Column chromatography of 30 mg each of a mixture of methyl oleate, linoleate and linolenate.Adsorbent, 10 g of silver nitrate an silica; column height, 23 cm; column diameter, 14 mm; eluent, benzene - light petroleum (40 + 60 to 100 + 0) ; rate, 0-5 ml per minute (after de VrieslO) Fig. 5. interaction forces at low intermolecular distances. For a given molecular weight solvent the forces increase with increasing molecular weight of solute and decrease with chain branching. With cyclic structures large entropy effects become manifest, such that the apparent inter- action energy is greater for a cyclic structure than for its straight-chain isomer.In this simplified treatment, which is confined to observations on the relative rates of movement of saturated and unsaturated fatty acids, such entropy effects will be ignored. The total London Dispersion Interaction of a long-chain saturated acid in a non-polar solvent, such as the Apiezon greases, is much greater than the total of the polar interaction forces considered in the next section. It is unusual in gas chromatography to deal with formally charged solutes, since these have too small a partial pressure. It is also unusual to deal with molecules possessing large formal dipoles. The major class of interaction forces between polar or polarisable solutes and solvents are of the electron donor-acceptor type. In these interactions, as would be expected from the name, one molecule partially donates and shares a bonding electron with an acceptor molecule, thus setting up a very weak reversible bond.The electron transitions involved in such complexes would be expected to bear some relation to the transition state in chemical reactions. How- ever, no bond re-arrangement occurs, but the known effects of substituent groups in controlling Secondly, there are polar interaction forces.576 JAMES: METHODS OF SEPARATION OF [Analyst, Vol. 88 chemical reactivity would be expected to facilitate or obstruct the formation of such com- plexes. The double bonds of unsaturated fatty acids show electron donor - acceptor interactions with the ester groups of the polyester stationary phases. The interaction is specific but not very large.The total interaction energy of a fatty acid methyl ester with a polyester is less than with a hydrocarbon stationary phase (see Table I). (The interaction energy is a function of the corrected retention volume.) In the next section a simplified theoretical treatment of the interaction of solute and The hydrogen bond is one of the best known of such donor - acceptor bonds. solvent has been attempted. I . . . . . . . 1 2 3 4 5 6 7 I Fig. 6. Thin-layer chromatogram of methyl esters on silver nitrate impregnated silica gel: 1, stearate; 2, oleate; 3, elaidate; 4, petroselinate; 5, petroselaidate ; 6, linoleate ; 7, mixture of 1, 2, 3 and 6 . Developing solvent, diethyl ether - hexane (10 + 90). Spots were located in ultraviolet light after the plates had been sprayed with 2', 7'-dichlorofluorescein, and were reproduced by tracing (after Morrisll) A 0 0 0 .. . . . . . , 1 2 3 4 5 6 7 8 0 0 0 0 0 0 0 0 . . . . . . . . 1 2 3 4 5 6 7 8 Fig. 7. Thin-layer chromatogram of methyl esters on silica gel (A) and silver nitrate impregnated silica gel (B) : 1, cis-9,1 O-epoxyoctadecanoate ; 2, cis-9,10-epoxyoctadec-12-enoate ; 3, cis-12,13-epoxy- octadec-9-enoate ; 4, 12-hydroxyoctadecanoate ; 5, 12-hydroxyoctadec-9-enoate ; 6, 12-hydroxyoctadec- 9-enoate : 7, 9-hydroxyoctadecanoate ; 8, 9-hydroxy- octadec-12-enoate. Developing solvent, diethyl ether - hexane (40 + 60). Spots were located in ultra- violet light after the plates had been sprayed with 2', 7'-dichlorofluorescein, and were reproduced by tracing (after Morrisll) TABLE I SPECIFIC RETENTION VOLUMES OF METHYL MYRISTATE PER GRAM OF STATIONARY PHASE AT COLUMN TEMPERATURE Stationary phase v,, ml hpiezon L a t 197" C .. . . . . . . 1320 Polyethylene glycol adipate a t 180" C . . 507 . . SEPARATION FACTORS AND CHEMICAL STRUCTURE- The rate of movement of any substance along a column is governed by its distribution coefficient. The latter is controlled by the free energy required to transfer a molecule of solute from solution in the column stationary phase to the gas phase. Provided the concentration is small enough for the distribution coefficient to be inde- pendent of concentration ( i e . , the necessary condition for symmetrical peaks) then Raoult's law can be assumed.577 August, 1963 J LONG-CHAIN UNSATURATED FATTY ACIDS Then- p i = py + RT Ln hT2 (derived by Martin12), where p l is the chemical potential of substance A in either phase, p? is its chemical potential in the same standard state and Ni is the mole fraction of A in the phase in question, If the gas phase and liquid phase are in equilibrium then- p: - /& = 0, which = py - p."," + RT Ln NC, - RT Ln NL, .If py - py = ApA. NL Then ApA = RT Ln -*- But A= K, the partition coefficient. NC, NL N: Then Ln K - --, where ApA is the free energy required to transport 1 mole of A from A-RT the liquid phase to the gas phase. 2.0 I -~ 2.0 6 8 10 12 14 16 18 20 22 No. of carbon atoms Relation between chain length and the logarithm of the relative retention time on saturated fatty acids in: A, a non-polar medium (Apiezon L); B, a polar medium (poly- ethylene glycol adipate) Fig.8. If a molecule, B, differs from A by possession of a substituent group, X , then- Ln K - AP -2 B - ~ ~ 1 RT and Ln K, - Ln K, = --? (ApB - ApA) where A,ux is the free energy necessary to move the group X from one phase to the other. Now KA = C, (V:). where C, is the column constant and V p is the corrected retention volume (volume of gas required to elute substance A from the column).578 JAMES: METHODS OF SEPARATION OF Then- [Analyst, Vol. 88 = Ln retention volume of substance B relative to substance A. only on the temperature and the chemical nature of the stationary phase. Thus log,, (retention volume) of B relative to A This is dependent -- APX - 2-3 RT * If the group X can be added to molecule A without causing any perturbation of electronic configuration in atoms other than that to which X is attached, then addition of a further X group should give the same increase in relative retention volume.I L - 1.0 - 0.2 - I 2 3 4 5 6 No. of double bonds Fig. 9. Relation between the logarithm of the relative retention volume and the number of double bonds in fatty acids having: A, 22 carbon atoms; B, 20 carbon atoms; C, 18 carbon atoms; D, 16 carbon atoms EFFECT OF INCREASE I N CHAIN LENGTH- Increasing the length of a chain by a -CH2- group would thus be expected to give the same relative increase in retention volume irrespective of chain length. This is demonstrated in Fig. 8, where chain length is plotted against log,, (relative retention volume) for long-chain saturated acid methyl esters in two stationary phases.The largest -CH,- separation factor is given by the non-polar stationary phase (Apiezon L grease) in which the London Dispersion Interaction is at its maximum. The -CH2- separation factor is purely a London Disper- sion effect, since the group is unable to show any polar interaction. In the polyester (ethylene- glycol adipate) stationary phase the -CH2- separation factor is lower since the total London Dispersion Interaction in this phase is less than in the non-polar phase. EFFECT OF INSERTION OF DOUBLE BONDS- In the hydrocarbon non-polar stationary phases, introduction of a double bond decreases the retention volume compared with the corresponding saturated fatty acid. This would be expected since the maximum interaction will occur between -CH,- groups of solute and solvent, and -CH = CH- groups will have a smaller interaction.August, 1963 LONG-CHAIN UNSATURATED FATTY ACIDS 579 In a stationary phase such as a polyester, which shows an interaction between its ester groups and the double bonds of the solute, introduction of the double bond into the solute molecule causes an increase of retention volume.This effect is opposed by the London Dispersion Interaction, which is working in the opposite direction. A plot of log (retention volume) against the number of double bonds in molecules of different chain length is shown in Fig. 9, demonstrating that introduction of each double bond gives approximately the same increase in relative retention volume.This is true only for double bonds separated by a CH2 group and not for conjugated systems. The separation factors for introduction of both CH, groups and double bonds are shown in Table 11. TABLE I1 INCREASES I N RELATIVE RETENTION VOLUME BROUGHT ABOUT BE' INTKODUCTION OF -CH2- AND -CH=CH- GROUPS Increases expressed as retention volume relative to the parent saturated acid Apiezon 1, Xpiezon M Polyethylene glycol Polyethylene glycol Reoplex 400 Group a t 197" C at 197" C succinate a t 180" C adipate at 180" C a t 197" C -CH,- 2.36 2.39 1.66 1-97 1.8 -CH = CH- 0.9 0.89 1-17 1.18 1.13 EFFECT OF CHANGE OF DOUBLE BOND POSITION- In non-polar stationary phases there is a small but detectable difference in retention volume between the isomeric monoenoic acids (see Table 111).Thus the A4- and A6-octa- decenoic acids have slightly lower retention volumes than has the A9- acid. The effect with polyester stationary phases is much smaller. With dienoic acids, both types of stationary phase are capable of differentiating between positional isomers (see Table IV). With trienoic acids on the other hand, change of position of the double bonds does not appear to alter the retention volume (see Table V). With tetraenoic acids there is again a small positional effect (see Table VI). TABLE I11 EFFECT ON RETENTION VOLUME O F DOUBLE BOND POSITION I N MONOENOIC ACIDS Retention Apiezon L Fatty acid a t 197" C 16: 17 cis - 16 : 18 cis - 16: 19 cis 0.90 18: 19 cis 2.03 18 : 16 cis - 18: 1 4 cis - volumes relative to methyl palmitate Apiezon M PEGA PDEGS Reoplex 400 at 197" C a t 180" C a t 200" C at 197" C - 1.15 1.18 1.13 0.89 - 0.89 1-15 1.18 1.13 2.08 2.2 1 2.00 - 2.05 2.22 - - 2.05 - - - - - TABLE IV EFFECT ON RETENTION VOLUME OF DOUBLE BOND POSITION IN DIENOIC FATTY ACIDS Apiezon M PEGA Reoplex 400 Acid a t 197" C a t 184.5" C a t 197" C 16 : 26,s 0.86 1.45 1-28 16 : 2 9 ~ 2 0.89 1.45 1.39 18 : 26t9 1.81 2.76 2.56 18 : 2'1l2 1.88 2.60 2.32 20 : 28911 4-17 4.22 (197" C) - 20 : 211~4 4-38 4.46 (197" C) - BOND CONJUGATION- Relatively little work has been carried out with conjugated fatty acids of known structure.Beerthuis et a l l 3 have studied the behaviour of conjugated C,, dienoic and trienoic acids on an Apiezon L column. The relative retention volumes they obtained are shown in Table VII and suggest that double bond position has less effect than conjugation and that cis-trans structures overlap trans-cis structures.The trienoic acids are well separated from the corre- sponding dienoic acids, unlike the unconjugated acids.580 JAMES: METHODS OF SEPARATION OF TABLE V EFFECT ON RETENTION VOLUME OF DOUBLE BOND POSITION IN TAnaZyst, Vol. 88 TRIENOIC FATTY ACIDS Apiezon M PEGX Acid a t 197" C at 184.5" C 16: 36,9212 0.77 1-72 16 37,10,13 0.77 1-72 18: 3 6 , 9 , 1 2 1.88 - 18: 39,12,15 1.88 - 20 : 35 98 ,I1 3-89 - 20 ; 38 311 914 3.89 - TABLE VI EFFECT ON RETENTION VOLUME OF DOUBLE BOND POSITION IS TETRAENOIC AND PENTAENOIC ACIDS Apiezon ill Acid a t 197" C Reoplex 410 20: 45,8,11,14 3.48 5.28 20 : 48 ,I1 914 ,I7 3.48 5.35 2 2 : 57,10,13,16,19 7.82 11.7 2 2 : 55,8,11,14,17 7.88 12.2 EFFECT OF cis-trans ISOMERISM ON RETEXTION VOLUME- In non-polar stationary phases the trans acids, since they possess a more extended configuration than the corresponding cis acids, show a greater interaction and have a retention volume greater by a factor of approximately 1.05 in monoenes, 1.2 in conjugated dienes and 1.17 in conjugated trienes (see Table VIII).In the polyester stationary phases the major interaction is between the double bond and the ester groups. A change of configuration of the hydrocarbon chains about the double bond does not alter the retention volume of the monoenes. Litchfield, Isbell and Reiser14 have recently published a study of the gas chromatographic behaviour of geometrical isomers of Ag ~12-octadecadienoic acid.Their results indicate the expected retardation of the cis-trans compound relative to the cis-cis acid, but an unexpected overlap of the trans-cis and trans-trans acids on an Apiezon L capillary column. On a diethylene glycol succinate polyester column the cis-cis and cis-tram acids overlap, but the trans-cis and trans-trans acids are separable. By using both columns all the isomers can be resolved, the results expressed as carbon numbers15 are shown in Table IX. TABLE VII RETEETION VOLUMES OF CONJUGATED ACIDS ON APIEZON L AT 197" C Retention volume relative Acid to methyl palmitate 18 : 2'>12 C ~ S - C ~ S . . . . . . . . . . . . 1.90 18:O . . . . . . . . . . . . . . . . 2.36 18 : 2*0,12 trans-cis, 18 : 29111 czs-trans . . . . . . 2-56 18 : 2lo3l2 tvans-trans, 18 : Z9$l1 trans-trans .. . . 3.04 18 : 3' Jl1 J13 cis-tram-tvans . . . . . . . . . . 3.82 18 : 39 9 1 3 tvans-trans-trans . . . . . . . . 4.43 OVERLAPPING OF SATURATED AND UNSATURATED FATTY ACIDS- saturated acids in one pass through a gas - liquid column. columns show some overlaps. These are listed in Table X. It would be too much to hope for a separation of all the possible saturated and un- Both polyester and Apiezon STABILITY OF UNSATURATED FATTY ACIDS ON GAS - LIQUID CHROMATOGRAMS- Some comparisons have been made between the results obtained from Apiezon L columns and the alkali isomerisation techniques. No indications have been found of any breakdown or loss of acids containing up to 4 double bonds on these columns. Condensation of the peaks emerging and subsequent infrared and ultraviolet examination has not disclosed anyAugust, 19631 LONG-CHAITU' UNSATURATED FATTY ACIDS 581 significant bond re-arrangement.The same is true for polyester stationary phases, but here it is necessary to work at as low a temperature as is possible consonant with the required speed of analysis, since it has been shown that there is a progressive loss of slower moving acids when column temperatures exceed 200" C. TABLE VIII EFFECT OF GEOMETRICAL ISOMERISM OF THE DOUBLE BONDS ON RELATIVE RETENTION VOLUME (METHYL PALMITATE = 1) Apiezon L Apiezon M PEGA Apiezon L DEGS Acid at 197" C of 197" C a t 180" C a t 150" C a t 120" C 16: l9 cis . . .. 0.90 16 : l9 trans . . 0.92 18:19 cis . . . . 2.03 1 s : l9 trans .. 2.12 is: 14 cis . . .. 18 : l4 trans .. 10: 12 cis . . .. 10 : 12 trans . . 10: 13 cis . . . . - - - - - * Figures published by Lennarz et al.16 relative to methyl decanoate. TABLE IX (AFTER LITCHFIELD et aZ.I4) CARBON NUMBERS OF THE GEOMETRICAL ISOMERS OF METHYL LINOLEATE Apiezon L DEGS Compound a t 200" C a t 175" C 1l2 cis-cis-octadecadienoate . . . . . . 17-48 19.46 112 cis-tvans-octadecadienoate . . . . . . 17.59 19.46 ,12 trans-cis-octadecadienoate . . . . . . 17.64 19.55 9 912 trans-trans-octadecadienoate .. . . 17.64 19.35 TABLE X SOME UNRESOLVED MIXTURES Acid Rpiezon Polyesters * * ) Overlap Resolved Iso-branched saturated of x carbon atoms Monoenoic acids of x carbon atoms .. . . .. . . Dienoic and trienoic, x carbon atoms . . . . . . . . Overlap Resolved Tetra- and pentaenoic, x carbon atoms .. . . . . . . Overlap Resolved Cis and tvans isomers of monoenoic acids . . . . . . Resolved Overlap Trienoic acids of x carbon atoms and saturated acids of x + 2 carbon atoms . . . . . . . . . . .. . . Resolved Partially resolved CHOICE OF STATIONARY PHASE- I t may be seen from the above that neither the polyester nor the hydrocarbon phases are capable of separating all the possible saturated and unsaturated fatty acids. Whenever possible, both types of column should be used. If this cannot be done, then the type of column capable of giving the maximum information for the particular fatty acid mixture should be used. For example, when cis-tvans isomers are likely to exist, the Apiezon column would be best; when separations of mono-, di-, tri- and more highly unsaturated acids are required, then the polyester phases would be best.A wide variety of different polyesters have been used, but the major ones in use now are- Polyethylene glycol adipate (PEGA) . Polyethylene glycol succinate (PEGS). Polydiethylene glycol succinate (polybutanediol succinate) (PDEGS) . When there is overlap on one column the components can almost always be resolved on the other type of column. Thus linolenic and arachidic acids incompletely separated on most polyester columns can be cleanly resolved on an Apiezon column. The same is true in reverse, the di- and trienoic, and tetra- and pentaenoic acids overlap on Apiezon columns. When doubt exists as to the structure of an acid from its chromatographic position alone,582 JAMES [Analyst, Vol.88 then measurement of its relative retention volume on the two types of column will give information as to both the number of carbon atoms and the number of double bonds in the molecule.17 Positional isomers of unsaturated acids have very similar relative retention volumes, so that chromatographic information alone is not sufficient conclusively to define structure. This is best done by collection of the substances from a gas chromatogram, with subsequent oxidative degradation and identification of the fragments by gas chromato- CAPILLARY COLUMNS- Capillary columns, when operated in such amanner as to produce very high plate numbers, can sometimes help in identification, because of the better separation obtained with small separation factors.Since, however, the small internal diameter columns can carry only small charges, they are not so useful when isolated fractions are required. Lipsky, Lovelock and Landowne21 have described high efficiency Apiezon L capillary columns showing an almost complete separation of methyl oleate and methyl elaidate. I thank Dr. H. J. Dutton, Dr. B. de Vries, Dr. L. J. Morris and the Editors and Publishers of Chemistry and Indastry for permission to reproduce diagrams. graphy.18 9 9 920 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. REFERENCES Ahrens, E. H., jun., and Craig, L. C., J . Biol. Chem., 1952, 195, 299. Dutton, H. J., Scholfield, C. R., and Jones, E. P., Chem. & Ind., 1961, 1874. Howard, G. A., and Martin, A. J. P., Biochem. J , , 1950, 46, 532. Boldingh, J., Rec. Trav. Chim., 1950, 69, 247. Hirsch, J ., in “Digestion, Absorption Intestinale et Transport des Glycerides chez les Animaux Viswanathan, C. V., and Meera Bai, B., J . Chromatog., 1962, 7, 507. Schlenk, H., Gellerman, J. L., Tillotson, J . A., and Mangoid, H. K., J . Amer. Oil Chem. Soc., Klein, P. D., and Janssen, E. T., J . Biol. Chem., 1959, 234, 1417. Mangold, H. K., and Kammereck, R., Chem. & Ind., 1961, 1032. de Vries, B., paper presented a t the VIth Congress of the International Society for Fat Research, Morris, L. J., Chew. G. Ind., 1962, 1238. Martin, A. J. P., in “Partition Chromatography,” BiochemicaE Society Symposia No. 3, Cambridge Beerthuis, R. K., Dikstra, G. D., Keppler, J. G., and Recourt, J. H., Ann. N.Y. Acad. Sci., 1959, Litchfield, C., Isbell, A. F., and Reiser, R., J . Amer. Oil Chem. SOL, 1962, 39, 330. Woodford, F. P., in Garattini, S., and Paoletti, R., Editors, “Drugs Affecting Lipid Metabolism,” Lannarz, W. J., Light, R. J., and Bloch, K., Proc. Nut. Acad. Sci., 1962, 48, 840. James, A. T., in Glick, D., Editor, “Methods of Biochemical Analysis,” Interscience Publishers Inc., Stoffel, W., and Ahrens, E. H., jun., J . Amer. Chem. SOL, 1958, 80, 6606. James, A. T., and Webb, J. P. W., Biochem. .I., 1957, 66, 515. Von Rudloff, E., J . Amer. Oil Chem. Soc., 1956, 33, 126. Lipsky, S. R., Lovelock, J . C., and Landowne, K. A., J . Amer. Chem. Soc., 1959, 81, 1010. Superieurs,” Marseille, July 18th to 20th, 1960. 1957, 34, 377. London, April 9th-l3th, 1962. University Press, 1950, p. 4. 72, 616. Elsevier Publishing Co., New York and Amsterdam, 1962, p. 189. New York and London, 1960, Volume VIII, p. 1. Received April 18th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800572

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

August, 19631 ANALYTICAL METHODS COMMITTEE 583 Analytical Methods Commit tee REPORT PREPARED BY THE MEAT PRODUCTS SUB-COMMITTEE Nitrogen Factors for Chicken THE Analytical Methods Committee has received the following report from its Meat Products Sub-committee. The report has been approved by the Analytical Methods Committee and its publication has been authorised by the Council. REPORT The Meat Products Sub-committee of the Analytical Methods Committee responsible for the preparation of this Report was constituted as follows : Dr. S. M. Herschdoerfer (Chair- man), Mr. s. Back, Mr. P. 0. Dennis, Mr. J. R. Fraser, Mr. H. C. Hornsey, Dr. A. J. Kidney, Mr. T. McLachlan, Dr. R. A. Lawrie, Dr. A. McM. Taylor and Mr. E. F. Williams, with Dr. C . H. Tinker as Secretary. As in its reports on pork1 and beef,2 the Sub-committee agreed to base on Stubbs and More’s method its recommendations for the determination of the raw fresh meat content in products made from chicken meat.A nitrogen factor of 3.5 is generally used for such determinations, and the Sub-committee decided to test the validity of this factor and, if necessary, to recommend the use of alternative factors. Again, the Sub-committee was fortunate in obtaining the collaboration of several laboratories of meat-product manufacturers (see below), who agreed to analyse chicken carcases according to the requirements of the Sub-committee. The Sub-committee recommended that older heavier birds should be examined, as it is unusual for young light birds to be used for manufacturing purposes. It was also arranged that the breast meat should be analysed separately from the other meat, as breast meat and dark meat are sometimes used separately for manufacturing purposes.The instructions listed below were issued to the collaborating laboratories. PKOCEDURE FOR THE PREPARATION OF SAMPLES 1. The neck, head and feet should be removed from the already eviscerated and plucked chicken. 2. The breast meat should be dissected completely, weighed, and then finely minced and mixed. 3. All other flesh and skin (exclusive only of any adhering intestinal fat), should then be completely dissected from the bones, weighed, finely minced and then mixed. 4. Portions from 2 and 3 should then be weighed out for moisture, fat, ash and nitrogen determinations. 5. All these operations should be performed with a minimum of delay, to minimise evaporation losses. 6.When possible, the following details should be supplied- (a) Weight and age of bird. (b) Weights of “breast meat” and “other meat.” (c) Percentages of moisture, fat, ash, nitrogen and nitrogen expressed on the fat-free, on both the “breast meat” and the “other meat.” RESULTS Those marked with an asterisk were obtained on samples taken by the procedure outlined above, and the values thus marked in the “Whole Carcase” section of the chart have been calculated from the values for the component parts. In addition, the values marked with a dagger in this section were obtained on composite samples prepared by dissecting and comminuting all flesh and skin from the entire carcase.In other tests sampling was not carried out by the prescribed The results collected by the Sub-committee are shown in Fig. 1.584 ANALYTICAL METHODS COMMITTEE [AnaZyst, Vol. 88 method, since the analyses were completed before the instructions had been issued. The resulting figures may be to some extent affected by this variation; it would, for instance, be reasonable to assume that the average figure for leg meat found by Laboratory C would have been higher had the skin not been removed. The Sub-Committee was of the opinion that it should allow for the different use of breast meat, and that it should recommend different nitrogen factors for breast meat and all the other meat. When no information is available on the kind of chicken meat used, a general factor would have to be applied. RECOMMENDATION The Sub-Committee recommends the following average figures for use in the analyses of chicken products: breast meat, 3.9; dark meat, 3-6; whole carcase 3.7. ACKNOWLEDGMENT The Sub-Committee thanks those listed below for their help and communications- Brand & Co. Ltd. Crosse & Blackwell Ltd. N.V.H. Hartog’s Fabrieken Oss, Holland. Harvey’s Belgravia Foods Ltd. J. Lyons & Co. Ltd. The Official Norwegian Quality Control Institute for Canned Fish Products. S. J. Palmer (N/C) Ltd. J. Sainsbury Ltd. C. Shippam Ltd. Unilever Ltd. REFERENCES 1. 2. - , Ibid., 1963, 88, 422. Analytical Methods Committee, Analyst, 1961, 86, 557.

ISSN:0003-2654    

DOI:10.1039/AN9638800583

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

August, 19631 WILSON AND LEWIS 585 n-Propyl Gallate as a Gravimetric Reagent for Bismuth and Antimony BY A. D. WILSON AND D. T. LEWIS (Department of Scientific and Industrial Research, Laboratoyy of the Government Chemist, Clement's I n n Passage, Strand, London, W.C.2) The use of n-propyl gallate as a reagent for separating and determining bismuth and antimony in the presence of interfering cations has been examined. Quantitative recovery of bismuth is achieved from 0.01 N nitric acid solution and of antimony from 0.5 N hydrochloric acid solution. A proposed structure for these compounds is suggested, based on the results of infrared analysis. Applications of the propyl gallate gravimetric method to the analysis of alloys and pharmaceuticals are discussed. POLYHYDRIC phenolic compounds have found wide application in analytical chemistry as gravimetric and colorimetric reagents.Pyrogallol and gallic acid have been proposed as precipitants for bismuth and antimony,fJ~3~4 but it is well known that solutions of pyrogallol are unstable. Kieft and Chandlee4 have reported that gallic acid decomposes in aqueous solution above 85" C. A polarographic study by Nash, Skauen and Purdy5 of the anodic oxidation of phenolic compounds at a wax-impregnated graphite electrode showed that n- propyl gallate was more resistant to oxidation than was gallic acid, which in turn was more stable than pyrogallol. Our qualitative observations on aqueous solutions of these reagents are in accordance with this work and suggest that n-propyl gallate is to be preferred to gallic acid and pyrogallol as an analytical reagent.n-Propyl gallate is available in the form of hydrated crystals, m.p. 65" C, or as an anhydrous powder, m.p. 147" C. It dissolves readily in warm water and gives, with various inorganic cations, precipitates whose solubilities are dependent on pH. Table I presents those reactions observed for the group-2 cations in nitric acid solutions and in buffered acetic acid at pH 5 . TABLE I PRECIPITATION OF METALS BY n-PRoPYL GALLATE The concentration of the metals was 0.2 mg per ml In acetic acid - acetate buffer at pH 5.0 In N nitric acid In 0.01 N nitric acid No precipitate No precipitate No precipitate No precipitate No precipitate No precipitate No precipitate White precipitate (forms slowly) No precipitate No precipitate No precipitate No precipitate No precipitate No precipitate No precipitate No precipitate Yellow precipitate Yellow precipitate White precipitate No precipitate Opalescence Opalescence No precipitate No precipitate White precipitate Brown precipitate Yellow precipitate Yellow precipitate Yellow precipitate White precipitate No precipitate White gelatinous precipitate White gelatinous precipitate * Forms precipitate with n-propyl gallate in neutral solution. Bismuth and antimony have been found to yield flocculent precipitates that can readily be collected by filtration and can be dried in sintered-glass or porous-porcelain Gooch crucibles at 110" C to give stoicheiometric compounds of the generic formula C,,H,,O,.M, which could obviously be represented by either the monohydrate structure (I) or the subgallate structure (11).Steric considerations would tend to favour structure 11, but more conclusive evidence has been furnished by infrared spectrographic observations in Nuj ol mulls prepared from the dried propyl gallate salts of bismuth and antimony. The absorption spectra586 WILSON AND LEWIS: n-PROPYL GALLATE AS A GRAVIMETRIC [Analyst, VOl. 88 (see summarised results in Table 11) confirm the absence of combined water and indicate the presence of free phenolic groups ; this evidence favours structure 11. The spectroscopic results also suggest that hydrogen bonding is present. This again is a most probable mani- festation c;f the characteristics of structure 11. COOC,H, COOC,H, I I EXPERIMENTAL Examination of Table I suggests that propyl gallate will afford a ready means for separating and determining bismuth and tervalent antimony in alloys, pharmaceuticals, etc., containing zinc, cadmium, lead, copper and arsenic as impurities.Mercuric salts interfere, but mercurous salts give soluble gallate complexes. Bismuth may be precipitated as the n-propyl gallate complex from boiling 0.01 N nitric acid solutions; the yellow precipitate coagulates rapidly and permits immediate collection on a filter. The weight of the dried precipitate corresponds to the empirical formula C,,H,,O,.Bi. Under more strongly acid conditions complete recovery may not be obtained. Precipitates formed in the cold are finer and tend to clog the porous porcelain crucibles used for collection; precipitation from hot 0.01 N nitric acid solution is preferred because the precipitate can be rapidly collected, and this procedure has been consistently adopted for bismuth.TABLE I1 INFRARED SPECTRAL DATA FOR BISMUTH AND ANTIMONY PROPYL GALLATES Wave number of peaks 3650 I 3450 3255 Bands between 2850 and 2950 2680 2600 1686 1676 Bands between 1375 and 1640 1342 1348 1253 1253 1218 1220 1156 1153 1087 1092 1053 1073, 1035 Bands between 687 and 998 Comments Free -OH stretching Hydrogen bonded -OH stretching CH,--, CH,- stretching (oil) Chelate -OH stretching ? Ester C= 0 stretching Aromatic C-0 stretching Phenolic C,-0 stretching -OH deformation CH,--, CH,- deformation (oil) Phenolic --OH deformation C-0 stretching Ester C--0 stretching Phenolic -OH deformation C-0 stretching Ester C-0 stretching Ester C-0 stretching ? -OH deformation ? 1,3,4,5 substituted benzene, out-of-plane deformation ? The effectiveness of this method for separating bismuth from several metals commonly associated with it in alloys has been examined. Table XI1 indicates that the separation of bismuth from zinc, cadmium and copper is excellent, and spectrographic analysis of the ignited precipitates completely failed to detect cadmium and showed only a trace of copper.Lead, when present in a five-fold excess, does not significantly interfere with the determination of bismuth, but when present in fifty- or five-hundred-fold excess causes high results, and spectrographic analysis confirmed the presence of some lead in precipitates. Both antimony and tin are often present with bismuth in alloys and these elements interfere in the method, antimony111 forming an insoluble complex with propyl gallate and tinIV and antimonyv hydrolysing to form insoluble hydrated oxides.Prior chemical separation of these cations is therefore necessary.August, 19631 REAGENT FOR BISMUTH AND ANTIMONY TABLE I11 SEPARATION OF BISMUTH FROM OTHER CATIONS Precipitations were carried out from solutions containing about 200 pg of bismuth per ml in 0.01 N nitric acid. Bismuth was weighed as C,,H,,O,.Bi Weight of g Interfering element interfering element, f 2.5 2.5 1.0 Copper Zinc Cadmium 2.5 1.0 0.2 1.0 r 2.5 587 Bismuth added, mg 49.76 50.00 49.88 20.00 21.13 50.10 50.00 20.00 20-12 50.10 21-31 20.00 19-92 50-12 20.00 19.99 50.61 Bismuth recovered, mg 49.46 50.10 50.00 20.08 21.13 50.01 49.98 20.06 20.20 50.20 21.30 20.04 19-98 50.52 20.18 20.06 50.3 1 Error, / O - 0.6 + 0.2 + 0.2 + 0.4 0 - 0.2 0 + 0.3 + 0.4 4- 0.2 0 +- 0-2 + 0.3 $- 0.8 + 0.9 + 0.3 - 0.6 0' Conditions for the precipitation of antimony111 with propyl gallate are different from those for bismuth, as the antimony precipitate appears to be more soluble in hot solutions.Precipitation of antimony was therefore carried out at room temperature. Hydrochloric acid was chosen as the medium in this instance; it is to be preferred to nitric acid as a solvent for antimony trioxide because the formation of antimonyv is precluded. In 0.5 N hydro- chloric acid a coarse white crystalline precipitate of the antimony III - propyl gallate complex slowly forms after an initial inhibition period. When set aside overnight complete recovery of antimony is obtained, the weight of the dried precipitate corresponding to C,,H,,O,.Sb.The separation of antimony from bismuth and arsenic111 has been examined and, as shown in Table IV, arsenic did not interfere. In the presence of bismuth, interference was some- times observed, bismuth being detected spectrographically in the precipitate ; the error sometimes is as high as 1 per cent. TABLE IV SEPARATION OF ANTIMONY FROM OTHER CATIONS Precipitations were carried out from solutions containing about 200 pg of antimony per ml in 0-5 N hydrochloric acid Interfering Weight of element interfering element, Antimony added, Antimony recovered, Error, - - 20-00 20.06 + 0.3 100 20.00 20.12 + 0-6 200 40.52 40.58 + 0.1 100 20.00 20.20 + 1.0 200 39-42 39.42 0.0 mg mg mg % - - 20.00 19.94 - 0.3 Bismuth Arsenic i APPLICATIONS OF THE METHOD In the practical application of this gravimetric procedure to determining bismuth in materials, it is often necessary to reduce the excess of acid by neutralisation with alkali.Bismuth is only metastable in solution at the acidity of precipitation (0.01 N nitric acid) and, on standing, slow hydrolysis takes place with deposition of bismuth compounds from solution as a fine white precipitate. We have found that the most satisfactory way of588 WILSON AND LEWIS: n-PROPYL GALLATE AS A GRAVIMETRIC [ATZabSt, VOl. 88 preventing complications owing to the precipitation of bismuth oxy-salts at low acid concen- tration is to condition the earlier stages of the analysis so that the final solution is approxi- mately 0.2 N in nitric acid.The required acidity is then attained by twenty-fold dilution with water, the propyl gallate reagent then being added immediately to the boiling solution. This procedure has potential applications in pharmaceutical and certain metallurgical fields of analysis. In the determination of bismuth and antimony in pharmaceuticals, organic matter can be destroyed by gentle ignition in an electric furnace, and the resulting trioxides can be dissolved in 2.5 N nitric acid for bismuth or 5.0 N hydrochloric acid for antimony before dilution to the appropriate acidity and precipitation of the propyl gallate complexes. The bismuth content of a pharmaceutical whose base was predominantly magnesium carbonate - sodium hydrogen carbonate was found by the proposed method to be 0.65 per cent., compared with 0.59 per cent.by the normal phosphate method. The method could, no doubt, be successfully applied to determining bismuth in alloys of the Woods metal type. These alloys contain tin, which must be removed by the procedure described below; similar alloys not containing tin or antimony may be dissolved in con- centrated nitric acid. The excess of acid may then be neutralised by alkali in the manner described, or be removed by evaporation to dryness. When evaporation is used the residue is then dissolved in N nitric acid, diluted one-hundred-fold, and the bismuth is precipitated as its propyl gallate complex. Tin and antimony frequently occur with bismuth in alloys, and when nitric acid is used to attack these alloys tinIv and antimonyv are precipitated as their hydrated oxides.As tin is completely oxidised to the stannic form its removal from solution is quantitative; antimony, in the presence of a ten-fold excess of tin, is also completely removed from solution. This might serve as a method of separating these elements from bismuth, but we found that these hydrated oxides occluded appreciable amounts of bismuth. An alternative approach is to attack these alloys by a hydrobromic acid - bromine mixture and, after the addition of perchloric acid, to expel the volatile bromides of tin and antimony by evaporation to fumes of perchloric acid. I t is necessary to repeat this process by the addition of hydro- bromic acid alone, with subsequent evaporation, until all tin and antimony are removed, which is indicated by the final solution remaining clear.When this procedure is adopted, recovery of bismuth is good (see Table V). TABLE V SEPARATION OF TIN AND ANTIMONY FROM BISMUTH BY VOLATILISATION OF BROMIDES The metals were dissolved in 10ml of hydrobromic acid containing 15 per cent. v/v of bromine, 5 ml of perchloric acid were added, and the mixture was evaporated to fumes of perchloric acid. Then 5ml of hydrobromic acid were added, and the mixture was again evaporated to fumes of perchloric acid. This operation was repeated until a clear solution was obtained. Traces of bromide were destroyed by the addition of a few drops of concentrated nitric acid and evaporation to fumes of perchloric acid Interfering Weight of element interfering element, Bismuth added, Bismuth recovered, Error, 50.17 50.14 - 0.1 50.00 49.80 - 0.4 500 50.00 49.80 - 0.4 500 50-00 49-88 - 0.2 Antimony 500 50.00 49.85 - 0.3 mg mg mg Yo - - - - Tin { METHOD FOR DETERMINING BISMUTH REAGENTS- n-Pro$yZ gaZZate solzhon-Dissolve 1 g of the reagent in 100 ml of warm water.Wash soZ.ution-Nitric acid, 0.01 N. PROCEDURE- Bring 250 ml of 0.01 N nitric or perchloric acid solution containing about 50 mg of bismuth to the boil. Add slowly, with constant stirring, 25 ml of n-propyl gallate solution, bring the solution once more to the boil, and maintain at the boiling point for about one minute.August, 19631 REAGENT FOR BISMUTH AND ANTIMONY 589 Filter the hot solution immediately, and collect the precipitate on a sintered-glass or porous- porcelain Gooch crucible.Wash the precipitate several times with the hot wash solution and finally once or twice with ethanol. Dry at 110" C for periods of one hour to constant weight. 100.0 mg of bismuth propyl gallate = 47.91 mg of bismuth. METHOD FOR DETERMINING ANTIMONY n-Propyl gallate solution-Dissolve 1 g of the reagent in 100 ml of warm water. Wash solution-Hydrochloric acid, 0-5 N. PROCEDURE- To a solution of antimony (about 40 mg) in 0.5 N hydrochloric acid add 25 ml of propyl gallate solution. Filter the cold solution through a porous-porcelain or sintered-glass Gooch crucible, and wash the precipitate several times with the cold wash solution, and finally once or twice with ethanol. 100.0 mg of antimony propyl gallate = 34-89 mg of antimony. REFERENCES REAGENTS- Stir, and set aside at room temperature overnight. Dry the precipitate at 110" C for periods of one hour to constant weight. 1. 2. 3. 4. 5. Feigl, F., 2. anal. Chem., 1924, 64, 41. Feigl, F., and Ordelt, H., Ibid., 1925, 65, 448. Gomez, J. O., and Romero, J. G., Te'cnica met. (Barcelona), 1948, 4, 461. Kieft, L., and Chandlee, G. C., Ind. Eng. Chem., Anal. Ed., 1936, 8, 392. Nash, R. A,, Skauen, D. M., and Purdy, W. C., J . Amer. Pharm. Ass., 1958, 47, 433. Received November 13th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800585

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

590 HEADRIDGE AND TAYLOR: VOLUMETRIC DETERMINATION OF [Aflal'ySt, VOl. 88 The Volumetric Determination of Iron, Molybdenum and Tungsten in Fluoride Solutions BY J. B. HEADRIDGE AND M. S. TAYLOR (Department of Chemistry, The University, Shefield 10) Both tungstenVI and molybdenumVI in 2 M hydrochloric acid - 0.5 M hydrofluoric acid are quantitatively reduced to the tervalent state on a Jones reductor amalgamated to the extent of 0.75 per cent. w/w. They are deter- mined, after collection in ammonium ferric sulphate solution, by titration with standard dichromate solution with barium diphenylamine sulphonate as indicator. Neither molybdenumV1 nor tungstenvl in 2 M hydrochloric acid - 0.5 M hydrofluoric acid is at all reduced on a silver reductor a t loo C, although iron111 is quantitatively reduced to ironII under these conditions.-4t 60" C in 0.2 M hydrofluoric acid - 1.5 to 2.0 M hydrochloric acid, molybdenumVI is reduced quantitatively to molybdenumv and in 0.2 M hydrofluoric acid - > 4 M hydrochloric acid quantitatively to molybdenumII1. No reduction of tungstenV1 in 0.2 M hydrofluoric acid - hydrochloric acid occurs on a hot silver reductor a t a hydrochloric acid concentration below 5.5 M. Solutions containing ironII1, molybdenumv1 and tungstenV1 and no other species capable of reduction have, therefore, been analysed for all three metals by using the Jones and silver reductors. The methods have been applied to the analysis of synthetic mixtures, a standard ferro-molybdenum alloy and a standard ferro-tungsten alloy. The mean determined compositions for the molybdenum and tungsten contents of the alloys were identical to the certificate compositions.The maximum deviation of any result from the mean determined compositions was 0-2 per cent. IT has already been reported1 that tungstenVI in 2 M hydrochloric acid - 0.5 M hydrofluoric acid is quantitatively reduced to tungsten111 on passage through a Jones reductor (38 cm long, 1.8 cm internal diameter, packed with 16- to 30-mesh zinc shot amalgamated to the extent of 0.75 per cent. w/w). The tungstenIII solution was collected in ammonium ferric sulphate solution under oxygen-free nitrogen, and the equivalent amount of iron11 thus produced was titrated with standard potassium dichromate solution in the presence of phosphoric acid with barium diphenylamine sulphonate as indicator.Up to 18.4mg of tungsten trioxide were determined with a mean error for 16 determinations of -0.00(3) mg and a standard deviation from the mean error of 0-02(1) mg. For amounts of tungsten trioxide in excess of 18.4 mg the reduction is not quantitative. This incomplete reduction is certainly associated with the fact that the tungsten111 species is fairly rapidly oxidised by hydrogen i0n.l In an effort to extend the upper limit for quantita- tive reduction of tungstenvl, the effect of different degrees of amalgamation of the zinc was investigated. A plot of oxidation state of the tungsten in the effluent from the reductor against amount of tungsten trioxide taken is shown in Fig. 1. It is obvious that an amalgamation of 0-75 per cent. w/w, which was originally chosen, produces the most satisfactory conditions for reduction.As expected, iron111 and molybdenumV1 in 2 M hydrochloric acid - 0.5 M hydrofluoric acid are quantitatively reduced to iron11 and molybdenum111 under the same conditions as used for tungstenV1. Up to 15.7 mg of ironlI1 were determined with a mean error for 11 determinations of -0.00(3) mg and a standard deviation from the mean error of 0-02(0) mg. Up to 20.3 mg of molybdenum trioxide were determined with a mean error for 13 deter- minations of -0.00(5) mg and a standard deviation from the mean error of O.Ol(8) mg.1 In view of the fact that molybdenumIII, in contrast to tungstenII1, is much less rapidly oxidised by hydrogen ion,l it was surprising to find that the upper limit for the quantitative determination of molybdenum trioxide was 20.3 mg.On further investigation, it was ascertained that low results for amounts of molybdenum trioxide in excess of 20.3 mg were caused not by incomplete reduction to molybdenumII1, but by premature end-points in theAugust, 19631 IRON, MOLYBDENUM AND TUNGSTEN IN FLUORIDE SOLUTIONS 591 subsequent titration. Molybdenum111 from the reductor reacts with iron111 in the collecting vessel to produce iron11 and molybdenumv. The iron11 reacts rapidly with dichromate in the titration, but molybdenumv reacts only sluggishly, and a premature end-point is obtained for amounts of molybdenum trioxide in excess of 20.3 mg. For these larger amounts of molybdenum the indicator is oxidised before all of the molybdenumv is titrated. By using the E.E.L.titrator with an Ilford No. 601 filter for photometric end-point detection, it was established that the molybdenumV1 in 2 M hydrochloric acid - 0.5 M hydrofluoric acid is, in fact, quantitatively reduced to molybdenum111 for amounts up to at least 50.8 mg. The photometric titrations of iron11 plus molybdenumv were made with standard dichromate solution in the absence of indicator. The end-point occurred at the intersection of two ex- trapolated straight lines on the plot of relative optical density against volume of titrant added, but the fact that the galvanometer needle took some time to fall to a steady value in the vicinity of the end-point verified that the reaction between molybdenumv and dichromate was indeed slow.D Tungsten trioxide taken, mg Fig. 1. Oxidation state of tungsten in the effluent from the Jones reductor as a function of the weight of tungsten trioxide taken for degrees of amalgamation of: curve A, 0-1 per cent. w/w; curve B, 0.5 per cent. w/w; curve C, 0.75 per cent. wlw; curve D, 1.0 per cent. w/w; curve E, 1.5 per cent. wlw Although amounts of molybdenum trioxide between 20.3 and 50.8 mg could certainly be determined quantitatively by photometric titration, the method is not recommended because it is much slower than a visual titration. Conditions are easily arranged so that no more than 15 to 20mg of molybdenum trioxide are passed through the reductor. Incidentally, the low recoveries for amounts of tungsten trioxide in excess of 18.4mg are associated with the side-reaction of tungsten111 and hydrogen ion in the reductor, and not with the titration.A photometric titration with dichromate of amounts of tungsten trioxide in excess of 18.4mg after reduction and collection in iron111 solution, and in the absence of indicator, also produced low results. Iron111 and tungsten111 appear to react to give iron11 and tungstenv1, and iron11 is titrated quantitatively with dichromate before the indicator is oxidised. Although the method described above for the volumetric determination of tungsten is completely satisfactory, tungsten is often associated with iron and molybdenum, and it was felt that the full potentialities of the method would only be realised if suitable volumetric procedures that would lead to a quantitative determination of all three elements in mixtures could be developed.592 HEADRIDGE AND TAYLOR VOLUMETRIC DETERMINATION OF [Analyst, Vol.88 I t was therefore decided to investigate the reduction of ironII1, molybdenumv1 and tung- stenVI in fluoride solutions, on the silver reductor. I t was soon established that iron111 in 2 M hydrochloric acid - 0.5 M hydrofluoric acid, at about 10" C, is quantitatively reduced to iron11 on the silver reductor, whereas molybdenumVI and tungstenV1 are unaffected. The reduced solution was again collected in ammonium ferric sulphate solution under nitrogen, and the iron11 was titrated with dichromate in the presence of phosphoric acid and barium diphenylamine sulphonate. Up to 22.0 mg of ironIII, alone and in the presence of molyb- denumVI and tungstenv1, have been determined with a maximum error of 0.04mg.The fact that molybdenumV1 is not reduced at all in this medium is of considerable interest, because molybdenumv' in 2 M hydrochloric acid is partly reduced to molybdenumv at about 10" C on passage through the silver reductor (extent of reduction approximately 55 per cent.). It is evident that the molybdenumV1 is stabilised in the presence of hydro- fluoric acid and is less easily reduced. MolybdenumVI can, of course, be reduced quantita- tively in solutions of 2 M hydrochloric acid at 60" to 80" C on passage through a silver reductor,2 0 2.0 4.0 6.0 8 Concentration of hydrochloric acid, M Fig. 2. Oxidation state of molybdenum and tungsten in the effluent from the hot silver reductor as a function of hydro- chloric acid concentration : curve A, molybdenum; curve B, tungsten.and experiments were undertaken with a hot silver reductor to see if molybdenumv1 in hydro- chloric - hydrofluoric acid solutions could be reduced to a definite oxidation state. Because hydrofluoric acid inhibits the reduction of molybdenumv1, the concentration of this acid was reduced to 0-2 M, which is still sufficient to retain tungstenV1 in true solution. The extents of reduction of both molybdenumV1 and tungstenV1 in 0.2 M hydrofluoric acid containing 0.5 to 7 M hydrochloric acid are shown in Fig. 2. As can be seen, molybdenumV1 is quantitatively reduced to molybdenumv in 0.2 M hydrofluoric acid - 1.5 to 2.0 M hydrochloric acid, and to molybdenum111 in 0.2 M hydrofluoric acid - >4 M hydrochloric acid.TungstenVI is not reduced until the hydrochloric acid concentration exceeds 5-5 M. By using synthetic solutions, it was established that up to 13.2mg of molybdenum trioxide in 4.5 M hydrochloric acid - 0.2 M hydrofluoric acid, both in the presence and absence of tungstenv1, could be determined on the hot silver reductor with a maximum error of 0.04 mg. IronIII, as expected, is quantitatively reduced to iron11 under these conditions. The scheme outlined below is, therefore, available for determining mixtures of ironIII, molybdenumV1 and tungstenv1. IronIII, alone, is reduced on the cold silver reductor in 2 M hydrochloric acid - 0-5 M hydrofluoric acid. Iron111 and molybdenumV1 are reduced The effluent was also 0.2 M in hydrofluoric acid.August, 19631 IROX, MOLYBDENUM AND TUNGSTEN IN FLUORIDE SOLUTIONS 593 together on the hot silver reductor in 4.5 M hydrochloric acid - 0.2 M hydrofluoric acid.IronIII, molybdenumV1 and tungstenvl are all quantitatively reduced on the Jones reductor in 2 M hydrochloric acid - 0.5 M hydrofluoric acid. The results obtained for the analysis of synthetic mixtures of ironIII, molybdenumVI and tungstenvl, a ferro-molybdenum alloy and a ferro-tungsten alloy are reported below. METHOD APPARATUS- Because of the corrosive nature of hydrofluoric acid solutions towards glassware, the apparatus used was constructed entirely in polythene. The Jones reductor was prepared from heavy-gauge polythene tubing (length 38 cm; internal diameter 1-80 cm) narrowing rapidly a t the base to a short piece of narrow-bore tubing, 4 mm in diameter.A polythene funnel 18 inches in diameter at the maximum width, welded to the base of the reductor, served this purpose. The zinc shot was supported by a perforated disc of diameter 6 mm made from $-inch polythene sheet and inserted into the wider end of the polythene funnel stem. Polythene tubing (6mm bore) was wrapped round the column from top to bottom in the form of a spiral, for cooling purposes. The collecting vessel was a 250-ml polythene bottle fitted with a polythene screw-cap having three holes in the top to accommodate the column delivery tube, and inlet and outlet tubes for nitrogen. The inlet tube reaches almost to the bottom of the bottle. A solution reservoir at the top of the reductor was made by welding a polythene funnel (3& inches in diameter at the maximum width), from which the stem had been removed at the appropriate position, on to the top of the column.A screw-clamp at the base of the reductor prevented leakage when the column was not in use. The column of amalgamated zinc was 38 cm long. The silver reductor was prepared from polythene tubing (length 38 cm; internal diameter 1.0 cm) narrowing rapidly at the base to a short piece of narrow-bore tubing, 5 mm in diameter, just long enough to protrude into the collecting vessel. The central portion (28 cm long) of the reductor tube was encased in heavy-gauge polythene tubing (1.8 cm internal diameter) sealed on to the reductor at each end, with an inlet tube at the bottom and an outlet tube at the top, in the form of a Liebig condenser.This permitted the temperature of the reductor to be controlled as required. The solution reservoir at the top of the reductor, the per- forated disc for supporting the silver, and the collecting bottle were constructed and fitted to the column as for the Jones reductor. REAGENTS- The column of silver was 34 cm long. HydrofZuoric acid and other acids-These were of analytical-reagent grade. Nitrogen-This was oxygen-free “spot” nitrogen obtained from the British Oxygen Co. Ltd. Potassium dichromate solution-An exactly 0.1000 N ( ~ / 6 0 ) solution of potassium dichromate was prepared from the analytical-reagent grade salt. Oxygen- free distilzed water-This was prepared by boiling distilled water vigorously for 15 minutes and allowing it to cool under liquid paraffin.The water was stored under liquid paraffin. Standard solutions of iron, molybdenum and tzcngsten-( 1) Iron-This was prepared from AnalaR ammonium ferric sulphate that had been analysed for iron by using the usual gravimetric method of hydroxide precipitation, and then filtration and ignition to ferric oxide. (2) Molybdenum and tungsten-These were prepared from Specpure trioxides, obtained from Johnson, Matthey and Co. Ltd., by dissolving accurately weighed amounts of the trioxides in boiling ammonium hydroxide, sp.gr. 0.88, in 100-ml polytetrafluoroethylene beakers. When dissolution was complete the solutions were evaporated to dryness to remove the excess of ammonia. The resulting solids were dissolved by heating gently with 10 ml of a solution of diluted hydrofluoric acid (2 + 3), adjusted to be 0.5 M in hydrofluoric acid and stored in screw-cap polythene bottles. The concentration of the element was arranged to be approximately 3 to 4 mg per g of solution.PREPARATION OF AMALGAMATED ZINC SHOT- AnalaR zinc shot, 16 to 30 mesh (batch Nos. 53324 and 58066), was obtained from Hopkin and Williams Ltd. The zinc shot was amalgamated to the extent of 0-75 per cent. w/w by vigorously shaking 500 g of shot-previously washed with dilute hydrochloric acid594 HEADRIDGE AND TAYLOR VOLUMETRIC DETERMINATION OF [Analyst, Vol. 88 (2 + 98) and then with distilled water-with 250 ml of 2 per cent. w/v mercuric chloride soliztion in a stoppered flask for 45 to 60 seconds.The supernatant solution was discarded, and the shot was washed several times with distilled water and then once with dilute sulphuric acid (2 + 98). The shot was finally washed with distilled water and stored under dilute hydrochloric acid (1 + 99). Zinc shot amalgamated to the extent of 0-1 per cent., 0.5 per cent., 1.0 per cent. or 1.5 per cent. w/w was prepared in the same way by shaking, in each instance, 500 g of shot with the appropriate amount of mercuric chloride solution. PREPARATION OF SILVER FOR THE SILVER REDUCTOR- The silver was prepared from AnalaR silver nitrate by dissolving 40 g of the salt in 250 ml of distilled water containing 5 ml of nitric acid, sp.gr. 1.42. This solution was then agitated with a sheet of electrolytic copper until deposition of the silver was complete. The precipitated silver was washed thoroughly with dilute sulphuric acid (1 + 99) until free from copper11 ions, and then with distilled water until free from sulphuric acid.The silver was stored under dilute hydrochloric acid (1 + 99). PREPARATION OF THE FERRO ALLOYS FOR ANALYSIS- Dissolve 0.1 to 0.2 g of alloy in 5 ml of 40 per cent. w/w hydrofluoric acid and 1 ml of 36 per cent. w/w hydrochloric acid in a 100-ml polytetrafluoroethylene beaker. Add 1 ml of 100-volume hydrogen peroxide, and heat gently until dissolution is complete. Evaporate the solution gently to dryness, and redissolve the solid in 5 ml of 40 per cent. w/w hydrofluoric acid. Repeat this operation twice to destroy excess of hydrogen peroxide.Redissolve the solid in 5 ml of diluted hydrofluoric acid (2 + 3), and transfer the solution to a dry, pre-weighed, 100-ml polythene bottle fitted with a polythene screw-cap, washing the beaker several times with oxygen-free distilled water. Transfer the washings to the bottle. Place 5 ml of dilute hydrofluoric acid (1 + 10) in the beaker, and boil gently for a few minutes. Transfer the solution and subsequent beaker washings to the bottle, and dilute to a suitable volume with oxygen-free distilled water. Cool the bottle and solution to room temperature, and re-weigh. PROCEDURES FOR DETERMINING IRONIII, MOLYBDENUMVI AND TUNGSTENVL (a) Iron-Transfer a suitable weighed portion of solution containing not more than 22-0 mg of iron to a 65-ml (2-02) polythene bottle calibrated at 50 ml.Add sufficient 36 per cent. w/w hydrochloric acid and dilute hydrofluoric acid (1 + 9) so that the final composition of the solution, when diluted to 50 ml with oxygen-free distilled water, is 2.0 M in hydrochloric acid and 0-5 M in hydrofluoric acid. Connect the water-jacket of the silver reductor to a cold-water tap, and allow water to pass through freely. Then open the screw clamp on the bottom of the reductor column. Allow the solution level inside the reductor to fall almost to the silver, and add 25 ml of 2 M hydrochloric acid - 0-5 M hydrofluoric acid wash solution to the reductor reservoir. Pass this solution through the reductor at a flow rate of 5 m l per minute. When the solution level again almost reaches the silver add a further 25-ml portion.Repeat this operation until 100ml of wash solution have been passed through the reductor. At no time should the solution level fall below the level of the silver. When the final 25-ml portion of wash solution has been added to the reservoir, start the flow of nitrogen to the inlet tube of the collecting bottle, and adjust the flow-rate to approximately 250 ml per minute. Immediately before the level of solution in the reductor reaches the silver, attach t o the base of the column a 250-ml polythene bottle containing 10 ml of 0.1 M ammonium ferric sulphate solution and 30ml of phosphoric acid, sp.gr. 1.75. Introduce the iron111 solution into the top of the column in 5- to 10-ml portions, and maintain a constant depth of solution above the silver. Take 50 ml of 2 M hydrochloric acid - 0.5 M hydrofluoric acid wash solution, and rinse the 2-02 polythene bottle with three 5-ml portions.Pour these rinsings on to the column, allowing the level of solution in the column to fall almost to the level of the silver before adding the next portion. Add the remaining 35 ml of wash solution in a similar manner. When the level of the final 5-ml portion has almost reached the silver, unscrew the collecting vessel, and rinse the nitrogen-flow tube with oxygen-free distilled water during removal of the bottle. Dilute to the 50-ml mark.August, 19631 IRON, MOLYBDENUM AND TUNGSTEN IN FLUORIDE SOLUTIONS 595 Add 30ml of phosphoric acid, sp.gr. 1.75 (see Note l), and 10ml of sulphuric acid, sp.gr. 1.84. Add exactly 5 drops (0.25 ml) of an aqueous 0.2 per cent.w/v solution of barium diphenylamine sulphonate as indicator, and titrate the solution with 0.1000 N potassium dichromate to a permanent purple end-point. Add titrant from a 6- or 10-ml calibrated grade-A micro- burette. Process a duplicate portion of iron solution of approximately half or double the weight of the first portion. Dilute the solution to 230 to 240 ml with oxygen-free distilled water. Subtract from the titre a blank value (s3e Kote 2 (i)). 1.000 ml of 0.1000 N potassium dichromate == 5.585 mg of iron. (b) Iron and molybdenum--Transfer a suitable weighed portion of solution containing not more than 22-0 mg of iron and 13-2 mg of molybdenum trioxide (8.8 mg of molybdenum) to a 65-ml (2-02) polythene bottle calibrated at 50 ml.Add sufficient 36 per cent. w/w hydrochloric acid and dilute hydrofluoric acid (1 + 49) so that the final composition of the solution, when diluted to 50 ml, is 4-5 M hydrochloric acid - 0-2 M hydrofluoric acid. Dilute the solution to approximately 40 ml with distilled water, and pass a steady stream of oxygen- free nitrogen through it for 15 minutes (see Note 2 (ii)). Rinse the nitrogen-flow tube with oxygen-free distilled water during removal of the bottle, and dilute the solution to the 50-ml mark. Connect the water-jacket of the silver reductor to a supply of water at 70" C (see Note 3j, and allow the temperature of the reductor to become constant. Then open the screw-clamp at the base of the reductor. When the level of solution in the reductor column has almost fallen to the silver, add 25 ml of 4.5 M hydrochloric acid - 0.2 M hydrofluoric acid wash solution, and wash the column in the same way as with the iron solution in procedure (a), but use 4.5 M hydrochloric acid - 0.2 M hydrofluoric acid wash solution.When the final 25-ml portion of wash solution has been added to the reservoir, heat the sample solution by immersing its container in a small reservoir into which is flowing the waste water from the reductor heater. The reservoir was constructed from a 250-ml polythene bottle, from which the top part had been removed at the appropriate position. A side-arm was fitted to the reservoir inch from the top as an overflow pipe. When the level of the final 25-ml portion of wash solution has fallen almost to the level of the silver, attach a collecting bottle containing 10 ml of 0.1 M ammonium ferric sulphate solution and 30ml of phosphoric acid, sp.gr.1.75, to the base of the reductor. Introduce the hot solution containing the iron and molybdenum into the top of the reductor in 5- to 10-ml portions, and process as in procedure ( a ) , but use 4.5 M hydrochloric acid - 0.2 M hydrofluoric acid wash solution. Rinse the nitrogen-flow tube with oxygen-free distilled water during the removal of the collecting bottle. Add 30 ml of phosphoric acid, sp.gr. 1.75 (see Note l), and dilute the solu- tion t o 230 to 240 ml with oxygen-free distilled water. Add indicator, and titrate the solution as in procedure ( a ) . Subtract from the titre a blank value (see Note 2 (ii)) and a volume of 0.1000 N potassium dichromate equivalent to the amount of iron present in the portion.Process a duplicate portion of approximately half or double the weight of the first. 1.000 ml of 0.1000 N potassium dichromate = 4.798 mg of molybdenum trioxide = 3.198 mg of molybdenum. (c) I ~ o n , molybdenum and tungsten-Transfer a suitable weighed portion of solution containing not more than 15.7 mg of iron, 20.3 mg of molybdenum trioxide (13.5 mg of molybdenum) and 18.4 mg of tungsten trioxide (14-6 mg of tungsten) to a 65-ml (2-02) polythene bottle calibrated at 50 ml. Add sufficient 36 per cent. w/w hydrochloric acid and dilute hydrofluoric acid (1 + 9) so that the final composition of the solution, when diluted to 50 ml with oxygen-free distilled water, is 2 M hydrochloric acid - 0-5 M hydrofluoric acid.Dilute to the 50-ml mark. Connect the cooling-spiral of the Jones reductor to a cold-water tap, and run the water for some minutes. Then open the screw-clamp at the base of the reductor. Drain the solution in the column until the level almost reaches the zinc shot, and wash the column with four 25-ml portions of 2 M hydrochloric acid - 0.5 M hydrofluoric acid wash solution. Process the sample solution in exactly the same way as described in procedure (a), but use 35 ml of wash solution instead of 50 ml to wash the sample solution through the reductor.596 HEADRIDGE AND TAYLOR VOLUMETRIC DETERMINATION OF [Anal-yst, Vol. 88 Rinse the nitrogen-flow tube with oxygen-free distilled water during removal of the collecting vessel.Add 30ml of phosphoric acid, sp.gr. 1.75 (see Note l ) , and 10 ml of sulphuric acid, sp.gr. 1-84. Add indicator solution, and titrate exactly as in procedure ( a ) . Subtract from the titre a blank value (see Note 2 (iii)) and calculated volumes of 0-1000 N potassium dichromate equivalent to the amounts of iron and molybdenum present in the portion of sample solution. 1.000 ml of 0.1000 N potassium dichromate = 7-730 mg of tungsten trioxide = 6,131 mg of tungsten. NOTES Dilute the solution to 230 to 240ml with oxygen-free distilled water. 1. INTERFERENCE FROM TUNGSTENVI- I t has already been reported3 that the end-point as shown by the colour change of barium diphenylamine sulphonate indicator in the titration of iron11 with potassium dichromate solution, is retarded in the presence of chloride ion.This effect increases with increasing concentration of chloride ion in the titrand, but is constant for a fixed concentration of chloride ion. The chloride ion concentration is therefore partly responsible for the magnitude of the blank value. However, the blank value was found to vary when ironT-I, in the presence of a fixed concentration of chloride ion but increasing amounts of tungstenv1, was titrated with potassium dichromate with barium diphenylamine sulphonate as indicator. The blank value decreased as the concentration of tungstenv' increased. TungstenVI appears to have a catalytic effect on the colour change of the indicator. Why chloride ion and tungstenv' should influence the oxidation mechanism of the indicator is not obvious, but the variable effect produced by different amounts of tungstenV1 can be overcome by adding 30ml of phosphoric acid, sp.gr.1.75 (making 60 ml in all), before titrating the iron11 with dichromate. Under these conditions the blank value is constant in the presence of a fixed amount of chloride ion and 0 to 60 mg of tungsten trioxide. 2. DETERMINATION OF BLANK VALUES- (i) The cold silver redwtor-There are four sources of error, three positive and one negative. These are (a) impurities present in the reagents, (b) chloride ion concentration, (c) theoretical indicator blank value-all positive eff ects-and (d) hydrogen peroxide pro- duced in the silver reductofl by reduction of dissolved oxygen present in the solutions- a negative effect.The positive effects are constant for a fixed chloride ion concentration, but the negative effect may differ from one solution to another according to oxygen content. Because a sample solution, whose iron content is to be determined, may have a different oxygen content from that of a standard iron solution, it is not advisable to apply to the sample solution a correction for a blank value determined for a standard solution. The blank value should be obtained by processing two portions of sample solution, which should be fairly different in weight, and applying the equation- w,v, - w,v, w1 - w, b = where b is the blank value in millilitres, W, and W, are weights of sample solution taken and V, and V, are the titres in millilitres for W, and W,, respectively. (ii) The hot silver reductor-Although the blank value is reasonably constant for various amounts of standard iron111 solution passed through a cold silver reductor, this is not so for a hot silver reductor.The blank value is generally lower, presumably because more hydrogen peroxide is produced, and is more variable. Since different portions of a sample solution, after adjustment to a volume of 50 ml with hydrochloric and hydrofluoric acids, are unlikely to have identical oxygen contents, it was considered advisable to remove the oxygen from these solutions by passing oxygen-free nitrogen through them for 15 minutes before processing them on the reductor. Various amounts of iron111 solution, which had been gassed out with nitrogen and passed through the hot reductor, showed, as expected, a higher and much more consistent blank value.The blank values are again determined by using the equation shown in Note 2 (i).August, 19631 IRON, MOLYBDENUM AND TUNGSTEN IN FLUORIDE SOLUTIONS 597 (iii) The Jones reductor-No difficulty from hydrogen peroxide is encountered in using the Jones reductor, since any oxygen present in the solution being passed through the reductor is completely reduced to water.5 To determine the blank value for the Jones reductor, process a series of various standard amounts of iron111 in 2 M hydrochloric acid - 0.5 M hydro- fluoric acid. In each determination calculate the volume of 0.1000 N potassium dichromate theoretically required to titrate the iron11 produced by the reductor, and subtract this volume from the experimental titre.None of the blank values should differ from the mean blank value by more than 0.004ml. A typical blank value for the Jones reductor for an iron111 solution was 0.105 ml. 3. HOT WATER SUPPLY- The temperature of the water supply to the silver reductor was maintained at 70" C by passing warm water at 40" C through two Liebig condensers connected in series, up the centre of which steam was passing. The temperature of the water supply to the reductor could then be adjusted by altering either the rate of flow of the water or the pressure of steam. The fall in temperature from the bottom to the top of the reductor column is 3" to 4" C. The temperature of the effluent containing the reduced species under these conditions is 58" to 60" C.(MolybdenumVI or tungstenVI can also be used, but iron111 is cheapest.) RESULTS TABLE I ANALYSIS OF SOLUTIONS OF PREPARED MIXTURES OF IKON, MOLYBDENUM AND TUNGSTEN Iron content Solution -- i - 7 number Actual, Determined, mg mg 1 32.5 32.6 2 30.6 30.6 3 46.2 46.2 4 68.0 68.1 5 63.1 63.0 Molybdenum trioxide content r-----h--- Actual, Determined, mg mg 74.2 74.2 30.9 30.7 47-7 47.8 35.8 35.9 74.7 74.9 T I Tungsten trioxide content Actual, Determined, 7 h-- mg mg 20.5 20.6 73-4 73.5 82.3 52.3 74.1 74.0 18.2 18.2 Solutions were prepared from stock solutions of the three elements; the total weight Suitable fractions were analysed by the methods of each solution was approximately 30 g. described above, and the results are shown in Table I. TABLE I1 RESULTS FOR THE ANALYSIS OF FERRO ALLOYS Alloy A Determined composition r 7 , B.C.S.Certificate Sample r-- _____- number composition, number Iron, Molybdenum, Tungsten, % % Yo % 231/2 Molybdenum 1 28.8, 28.9 70.2, 70.0, 70*2,* 70*0,* 69*9,* 70*2* - 70.1 { 2 28.9, 29.0 70.0, 70.0, 70-1,* 70*2,* 69.9* - 242/1 Tungsten 1 17.5, 17.5 0*3(2), 0-3(2) 82.0, 82.0 82.0 { 2 17.4, 17.6 0*3(5), 0.3(5) 82.0, 81.9 * Obtained by using the Jones reductor; the other molybdenum values were obtained by using the silver reductor. DISCUSSION Although the maximum amount of molybdenumV1 in 4-5 M hydrochloric acid - 0.2 M hydrofluoric acid, determined by using the hot silver reductor, was 13.2 mg expressed as molybdenum trioxide (8.8 mg of molybdenum) , it is extremely likely that, as with the Jones reductor, amounts of molybdenum trioxide up to 20.3 mg (13.5 mg of molybdenum) could be determined, by using a visual titration with potassium dichromate solution and barium diphenylamine sulphonate as indicator.598 HEADRIDGE AND TAYLOR [Analyst, 1'01.88 The maximum amount of iron111 determined in 2 M hydrochloric acid - 0.5 M hydrofluoric acid by using the Jones reductor was 15-7 mg, but much larger amounts are likely to be quantitatively reduced on this reductor. It is evident from Table 11 that the methods just described should be useful for deter- mining iron, molybdenum and tungsten in ferro-molybdenum, ferro-tungsten and certain other alloys, preferably after a qualitative spectrographic analysis to make absolutely certain that the alloys contain no appreciable amounts of other elements that could interfere in the methods by being reduced on the silver or Jones reductors. The behaviour of a few elements that can be quantitatively reduced to lower oxidation states on metal reductors from solutions free from fluoride was investigated for the fluoride systems employed for iron, molybdenum and tungsten. In all instances the maximum amount of element passed through the reductor was 14.3 mg. The procedures used for the reductions and titrations were identical to those employed for iron, molybdenum and tungsten. The results are shown in Table 111. TABLE I11 OXIDATION STATES OF VANADIUM, TITANIUM AND CHROMIUM AFTER PASSAGE THROUGH VARIOUS REDUCTORS Oxidation states of the species in the effluent from the reductor 7 A 1 Species Cold silver Hot silver Jones Vanadiumv . . .. .. + 4-00 + 4-00 + 2.00 TitaniumIV . . . . .. + 4.00 + 4.00 + 3-00 Chromium111 . . . . + 3.00 + 3.00 + 2.83 Chromium111 is not reduced to a definite oxidation state on the Jones reductor in cold 2 M hydrochloric acid - 0.5 M hydrofluoric acid. Vanadium11 reacts with iron111 in the collecting vessel to produce iron11 and vanadium1v. In the titration of such a solution with barium diphenylamine sulphonate as indicator, the indicator changes colour before any of the vanadiumlv is oxidised. VanadiumIV can, however, be determined by photometric titration with dichromate in the absence of the indicator. We gratefully acknowledge the receipt of a research grant from the B.S.A. Educational Trust Fund to maintain one of us (M.S.T.). 1. 2. 3. 4. 6. REFERENCES Headridge, J. B., and Taylor, M. S., Proceedings of the Feigl Anniversary Symposium, 1962, Birnbaum, N., and Walden, G. H., J . Amcr. Chern. Soc., 1938, 60, 64. Headridge, J. B., and Taylor, M. S., Analyst, 1962, 87, 43. Miller, C. C., and Chalmers, R. A., Ibid., 1952, 77, 2. Lundell, G. E. F., and Knowles, H. B., I:nd. Eng. Chew., 1924, 16, 723. in the press. Received March 12th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800590

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

August, 19631 BRADSHAW E 599 A The Volumetric Determination of Antimony in Antimony - Lead and Antimony - Tin - Lead Alloys BY G. BRADSHAW (Capper Pass & Sons Ltd., Melton Works, Ai'orth Ferriby, Yorkshire) A method is described for determining antimony by titration with bromate involving the use of the dead-stop end-point. The proposed method has advantages over the indicator method and has been successfully applied to the determination of antimony in lead - antimony and tin - lead - antimony alloys. Up to 10 mg of copper and 50 mg of iron can be tolerated without seriously affecting the results. POTASSIUM bromate is widely used for the titration of a n t i m ~ n y , ~ ? ~ ~ ~ but the methyl orange indicator commonly employed is non-reversible, is subject to fading and introduces a blank titre that varies with the amount of indicator added. Reversible indicators, which are an improvement over methyl orange, have been reported by Bel~her.*,~ An amperometric titration method has been described by Harris and Lindsey6 for determining antimony with potassium bromate, a vibrating platinum micro-electrode being used.This paper deals with the satisfactory application of the dead-stop end-point to the titration of antimony. First described by Foulk and Bawden,' the dead-stop end-point has since found many applications. In the proposed method a potential is applied across a pair of platinum electrodes immersed in the solution to be titrated, and a microammeter measures the current flowing between the electrodes. When tervalent antimony is titrated in this apparatus, the cell is polarised and the current flowing remains at a low level until the whole of the antimony is oxidised to the quinquivalent state; a slight excess of bromate, which acts as a depolariser, is then sufficient to cause a large deflection of the microammeter.Details of the circuit are shown in Fig. 1.600 BRADSHAW : VOLUMETRIC DETERMINATION OF ANTIMONY IN CHOICE OF POLARISING POTENTIAL- To determine the optimum polarising potential, solutions containing 5 ml of 0.1 N antimony solution, 5 ml of concentrated sulphuric acid and 40 ml of concentrated hydro- chloric acid were diluted to 300 ml, and the stirred solutions were titrated at 80" C with 0-1 N potassium bromate, various applied potentials being used. Results are shown graphically in Fig.2. Changes in the applied potential between 0.2 and 0.6 volt cause little change in sensitivity, but the residual current increases rapidly with potential. For this reason a potential of 0.2 volt was chosen. WARNING OF THE APPROACH OF THE END-POINT- In the earlier work with the method when the electrodes were positioned well into the solution there appeared to be no warning of the approach of the end-point, and an excess of titrant was easily added unless the titration was carried out dropwise. It was found, [Analyst, Vol. 88 EXPERIMENTAL 0.1 N Potassium bromate added, ml Fig. 2. Determination of optimum polar- ising potential: curve A, 0-6 volts; curve B, 0.4 volts; curve C, 0.2 volts however, that a warning could be obtained by positioning the electrodes close to the surface of the solution, with the cathode near to the point of entry of the titrant.At the start of the titration the bromate is consumed rapidly, and the needle of the microammeter remains almost stationary. As the titration proceeds, local concentrations of bromate build up near the point of entry of the titrant, depolarise the cathode and cause oscillations of the micro- ammeter needle, which become more violent as the end-point is neared. At the end-point, when a slight excess of bromate exists throughout the solution, a permanent deflection is obtained. EFFECT OF TEMPERATURE- When titrations are carried out at room temperature the response of the meter is slow, and consequently the end-point is easily overshot. At intermediate temperatures, i.e., about 40" C, the microammeter needle oscillates with each addition of bromate, making itAugust, 19631 ANTIMONY - LEAD AND ANTIMONY - TIN - LEAD ALLOYS 601 difficult to judge when the titration is nearing completion.If, however, the solution is titrated between 80" and 90" C only slight oscillations take place until within a few millilitres of the end-point ; then the oscillations increase in amplitude reaching almost full-scale deflection within a few drops of the end-point. EFFECT OF VARIOUS IONS- To test the effect of various ions they were added to a solution of known antimony content, After the addition of 75 ml of concentrated hydrochloric acid and 1 ml of saturated sulphur dioxide solution the mixture was evaporated, with vigorous boiling, to 60 5m1, diluted to 300 ml, and titrated at a temperature between 80" and 90" C with 0.1 N potassium bromate.Tin and Zead-To test the effect of tin and lead, the metals were dissolved by heating in sulphuric acid. Up to 2-5 g of tin and 5 g of lead did not cause any interference, although it should be noted that with amounts of lead greater than about 2 g crystallisation occurs during the evaporation, and this gives rise to serious superheating unless the beaker is continually agitated. Co@er-Because of the possibility of the method being used for determining antimony in white metals, the effect of copper was tested. Table I shows the results obtained in the presence of copper. Amounts above 10 mg cause serious interference and, therefore, if more than 10mg of copper are present a separation is necessary.TABLE I EFFECT OF COPPER Copper added, mg 10 50 100 150 50 100 150 Antimony added, mg 60.9 60.9 60.9 60.9 150.2 150.2 150.2 Antimony found, mg 60.3 59.2 58.1 57.5 147.8 146.2 145.6 Arsenic-The evaporation stage in the method satisfactorily removes up to 10mg of arsenic, but, as expected, any arsenic remaining is titrated by the bromate and gives high results. The response of arsenic, titrated under the same conditions as outlined under "Method," is similar to that of antimony, and the proposed method has been used in this laboratory for determining arsenic in certain materials. Iron-In lead - antimony alloys only extremely small amounts of iron would be normally encountered, but the effect of iron, added in the ferric form, was tested because it is an element likely to cause difficulty when an indicator is used.The effect of iron on the titration is to produce a high residual current that drops rapidly just before the end-point is reached. Thus, at the beginning of the titration a solution containing 100 mg of iron had a residual current of 60pA and, when near to the end-point, this fell to 10 pA. Results in Table I1 show that iron causes slightly low results, but amounts of up to 50 mg have no significant effect. TABLE I1 Iron added, mg 25 50 100 25 50 100 EFFECT OF IRON Antimony added, mg 60.9 60.9 60.9 150.2 150.2 150.2 Antimony found, mg 60.6 60.4 59.5 149.5 149.0 148.2 Other ions-Five grams of ammonium or potassium sulphate, used to facilitate the dissolu- tion of the sample by elevating the boiling-point of the sulphuric acid, caused no interference.602 BRADSHAW METHOD [Analyst, Vol.88 REAGENTS- Sulphuric acid, concentrated. Potassium sulphate. Hydrochlorzc acid, concentrated. Sulphur dioxide, saturated solution. Potassium bromate, 0-1 N, aqueous-Dissolve 2.784 g of potassium bromate in 1 litre of water, and standardise against pure antimony. PROCEDURE- Weigh accurately 0-25 to 2 g of sample according to the antimony content, transfer to a 400-ml beaker, and add 10 ml of concentrated sulphuric acid and 5 g of potassium sulphate. Heat gently to dissolve the sample, and then heat over an open flame to expel any sulphur. Cool, and then add 10 ml of water, 75 ml of concentrated hydrochloric acid and 1 ml of satura- ted sulphur dioxide solution.Place a few glass beads in the beaker, and evaporate the solution, with vigorous boiling, to 60 Dilute to 300ml with water, and heat the solution to a temperature between 80” and 90” C. Set the polarising voltage to 0.2 volts, and adjust the electrodes so that they are just below the surface of the solution. Titrate with 0.1 N potassium bromate until the meter shows a permanent deflection. RESULTS- This method has been in use in our laboratory for several years for determining antimony in antimony - lead alloys with an antimony content from 10 to 90 per cent. and in tin - lead solders with an antimony content between 0.2 and 5 per cent. The results obtained by this method agree with those obtained by using methyl orange, and the considerable number of results exchanged with other laboratories on these types of materials have been in good agreement. To check the reproducibility of the method an antimony - lead alloy was analysed by several different operators. An average of 11.23 per cent. of antimony was obtained with a standard deviation of 0.05 per cent. I thank the Managing Director of Capper Pass & Son Limited for permission to publish this paper. 5ml. REFERENCES 1. 2. 3. 4. 5. - , Ibid., 1951, 5, 30. 6. 7. “A.S.T.M. Methods for Chemical Analysis of Metals,’’ American Society for Testing Materials, Philadelphia, U.S.A., 1956, p. 483. Furman, N. H., Editor, “Standard Methods of Chemical Analysis,” Sixth Edition, D. Van Nostrand Co. Inc., New York, 1962, Volume I, p. 92. Kolthoff, I. M., and Belcher, R., “Volumetric Analysis,’’ Interscience Publishers Inc., New York and London, 1957, Volume 111, p. 514. Belcher, R., Anal. Chim. Acta, 1949, 3, 578. Harris, E. D., and Lindsey, A. J., Analyst, 1951, 76, 650. Foulk, C. W., and Bawden, A. T., J . Arne?.. Chew. SOG., 1926, 48, 2045. Received January 28th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800599

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

August , 19631 PIERCE AND PECK 603 A Rapid Solvent-extraction Sampling Technique for Neutron-activation Analysis BY T. B. PIERCE AND P. F. PECK (Analytical Chemistry Branch, A .E.R.E., Harwell, Didcot, Berks.) A solvent-extraction sampling technique has been used for comparing the specific activity of an element separated from an irradiated sample with that of an irradiated standard after both have been diluted with inactive carrier. The method has been applied to provide a rapid separation and determination of silver, the activity of the 2.3-minute silver-108 nuclide being measured. PRECIPITATION techniques have been widely applied in neutron-activation analysis, as a means of assessing chemical yield, when separation is required to obtain the element to be determined in a radiochemically pure form.When the activity of short-lived nuclides is to be measured, however, the time needed for precipitation is often inconveniently long, and other methods, for example, spectrophoto- metry after liquid counting,l y 2 have been employed for determining the yield. These methods are usually designed to assess the total yield of elements, thus permitting the specific activity to be calculated after the radioactivity has been assayed. The amount of element in the original sample can then be found by comparing this specific activity with that of an irradiated standard, inactive carrier having been added to both sample and standard at the start of processing. An alternative procedure would be to measure the radioactivity of the same amount, rather than total yield, of element derived from sample and from standard, after carrier has been added; simple comparison of activities will then give the ratio of the specific activities in the two cases.Although some decrease in sensitivity will result from measuring only part of the final yield, this might well be off-set by the expediency of the method when measurement of radioactivity soon after completion of irradiation is essential. Recently, a solvent-extraction sampling method has been proposed for a tracer dilution technique3 in which a solution of constant metal concentration is prepared by virtually complete conversion of a known amount of complexing agent, HL, to the corresponding complex, ML,, with the metal to be determined Mn+, ML, being extracted into a known volume of organic phase.This type of sampling should be applicable to activation analysis, and further, since HL will form a complex with Ma+ in preference to other metals present in the system giving weaker complexes than ML,, the solvent-extraction step will impose some restriction on the number of elements likely to interfere. We report here investigations carried out to assess the possibility of applying solvent-extraction sampling to neutron- activation procedures with particular reference to a method devised for the rapid determina- tion of silver in which the activity of the short-lived 2.3-minute silver-108 nuclide is measured. A solution of diphenylthiocarbazone (dithizone) in carbon tetrachloride was used to extract the silver. EXPERIMENTAL Commercially available dithizone (Hopkin and Williams Ltd.) was dissolved in carbon tetrachloride, extracted into dilute aqueous ammonia solution, and the aqueous phase washed with fresh organic solvent.After the solution had been acidified with hydrochloric acid, the precipitated dithizone was dissolved in carbon tetrachloride and its purity assessed by measurement of the ratios of the optical densities at the absorption maxima at 620 and 450 mp and the minimum at 510 mp. Purification by extraction was repeated, if necessary, until satisfactory values were obtained. The concentration of the dithizone solutions was calculated from the absorption at 620 mp of a sample (diluted if necessary), assuming a molecular extinction coefficient* of 34.6 x lo3. All glassware was made of Pyrex glass, and for experiments in which tracer was required 253-day silver-1 10m was used.Irradiations necessary for activation analyses were carried out by sealing samples and standards in thin-wall 6-mm silica tubes and irradiating in a flux of approximately604 PIERCE AND PECK: A RAPID SOLVENT-EXTRACTION [Analyst, Vol. 88 10l2 neutrons per sq. cm per second, a pneumatic-transfer system being used. “Doped” samples of known silver content were prepared by weighing out about 0.1 g of iron or steel found to have no detectable silver content into 6-mm silica tubes, and adding known amounts of silver in a few microlitres of solution. The silica tubes were then sealed and irradiated in the usual way. Chemical techniques and gamma liquid-counting methods involving use of a well-type scintillation counter were perfectly standard and will not be mentioned further; an M6H type liquid beta counter was used for beta counting.Kemp and Smalesl have described a con- venient method for liquid counting the beta activity of sample and standard alternately. The activation analysis procedure is described and discussed in the section dealing with the determination of silver, p. 606. GENERAL CONSIDERATIONS- The solvent-extraction sampling technique may be most conveniently applied when reagent HL is virtually completely converted to ML, in the presence of excess of M,+ over a wide range of conditions. Dithizone was therefore chosen as extractant for silver, since RESULTS AND DISCUSSION Active silver added, pg Fig.1. Dependency of count rate of silver dithizonate phase (cz) on amount of active silver added the reagent is known to form a strong primary silver complex in acid solution, which is soluble in carbon tetrachloride, but nearly insoluble in water.5 The reliable stoicheiometry of the 1 + 1 silver - dithizone complex formed in acid solution, also suggests that it should be possible to prepare a standard silver dithizonate solution from a solution of dithizone in carbon tetrachloride of known concentration in the presence of excess of silver under suitable conditions. If the specific activity, S,, of an element under standard counting conditions is given by S, = cJm, counts per minute per g, where c1 is the absolute count rate in counts per minute of m, grams of material, then when m, grams of the active element are added to M grams of inactive carrier, after equilibration the new specific activity S, will be given by S, = c,/(m, + M) counts per minute per g.Thus the count rate c, of X grams of diluted element will be- When activation analysis is used for determining trace elements, the amount of active element isolated from a sample (m,) is likelyto be smallcomparedwith carrier (M), and therefore, if X , S, and M are kept constant, there should be a linear dependency of c, upon m,. This was confirmed by shaking 5ml of a solution of dithizone in carbon tetrachloride with a 0.5 M sulphuric acid phase, containing small amounts of active silver and 750 pg of inactiveAugust, 19631 SAMPLING TECHNIQUE FOR NEUTRON-ACTIVATION ANALYSIS 605 silver carrier, there being enough dithizone present to convert only 375 pg of silver into the complex.The count of 4ml of the organic phase as a function of the active silver added is shown in Fig. 1, from which it can be seen that the predicted linear relationship is found. INTERFERENCES- Since solvent-extraction sampling assumes complete conversion of reagent HL to complex ML,, extraction of elements, other than the one to be determined, will result in the measure- ment of the activity of smaller amounts of Mn+ than theoretically expected, thereby intro- ducing error. Further, when solvent-extraction sampling is being used in activation analysis procedures, although the amounts of elements extracted together with ML, may not be large enough to affect, appreciably, the amount of Mn+ whose activity is measured, there still may be sufficient impurity present in the organic phase to interfere seriously with the radiochemical purity of the solution of ML, to be measured. If reaction between dithizone (here represented as the monobasic acid HL) and two metals MIn+ and MZP+, is given by- Mln+ +nHL + M,L,+ nH+ .. . . . . . . (2) and M2"+ + pHL + M2L, + $H+ . . . . .. - * (3) where K, and K, are equilibrium constants for equations (2) and (3), and the subscript, o, refers to species present in the organic phase. Elements likely to interfere when solutions of dithizone are used for solvent-extraction sampling have already been considered with particular reference to the tracer dilution technique3 and therefore will not be re-considered here.It can be seen from equation (4), however, that a high [M,Ln],/[M,L,], value is favoured by a larger ratio of equilibrium constant K,/K,, and interference to the extraction of silver would be expected from mercury, palladium and gold-elements known to form extremely st able dithizonat es. To assess interference with the extraction of silver caused by the presence of other elements, approximately N sulphuric acid phases were prepared containing 750 pg of inactive silver, 4-5 pg of active silver tracer and known amounts of other inactive elements. Aqueous phases were shaken with 5 ml of a solution of dithizone, containing sufficient reagent to extract a total of 375 pg of silver, and the radioactivity of 5 ml of the organic phase was counted.Results shown in Table I indicate that mercury, palladium and gold do interfere, but the presence of large amounts of other elements known to form well defined dithizonates does not decrease the count from the silver found in the organic phase. TABLE I INTERFERENCE WITH THE EXTRACTION OF SILVER CAUSED BY THE ADDITION One milligram of each element was added OF INACTIVE ELEMENTS Count rate of 4 ml counts per minute Element added of organic phase, Silver only present . . lron . . . . .. Cobalt . . .. . . Nickel . . . . . . Zinc . . . . . . Cadmium . . . . Indium . . . . . . Manganese . . . . 15,366 15,229 15,292 15,060 15,143 15,175 15,280 15,119 Count rate of 4 ml counts per minute Element added of organic phase, Tin11 . . . . .. 15,168 Lead . . . . . . 14,912 Copper .. . . . . 15,520 15,350 Bismuth . . . . . . Gold . . . . . . 64 Mercury . . .. . . 1 Palladium . . . . 407 Separation of interfering elements from the silver could be carried out by conventional chemical separation procedures. As, however, it was hoped to devise a neutron-activation method based on the measurement of the short-lived 2.3-minute silver-108 nuclide, the time available for any processing was severely limited. Mercury, palladium and gold may be removed by a preliminary extraction with dithizone, If the expected total amount of these metals is small compared with the silver carrier added, the dithizone necessary for their extraction will cause only a small loss of silver from the606 PIERCE AND PECK: A RAPID SOLVENT-EXTRACTION [Ancdyd, VOl.88 system, even in the extreme instance when the interfering metals are not present. Any decrease in sensitivity owing to the extraction of some silver at this stage could well be off-set by the short time needed to complete the separation, which is likely to be less than that required to separate the 2.3-minute nuclide by conventional means. A series of experi- ments was carried out, similar to those described above, which provided the results in Table I, except that the aqueous phase was first shaken with 10 ml of M dithizone in carbon tetrachloride. Under these conditions it was found that the amounts of these elements added (up to a total of 30 pg) caused no decrease in the amount of silver finally extracted. U S E OF SOLVENT-EXTRACTION SAMPLING FOR DETERMINING SILVER BY NEUTRON ACTIVATION In order to investigate the possibility of determining silver by assay of the radioactivity of the 2-3-minute silver-108 nuclide, the silver contents of a number of “doped” samples were measured.After irradiation, the silica tube containing the sample was opened and dropped, with its contents, into 10 ml of diluted nitric acid (1 $- 1) containing 1000 pg of silver carrier. When the sample had dissolved, the resulting acid solution was diluted to about 20 ml with distilled water, and hydroxides were precipitated by addition of ammonia solution to provide a preliminary separation of many elements from silver. After heat had been applied for a few seconds, the precipitate was removed on a Whatman No. 541 filter- paper, and the filtrate was collected in a beaker cooled by ice, made approximately 0.5 N in sulphuric acid, and 2 ml of 20 per cent.w/v hydroxylamine sulphate were added. The acid silver solution was then transferred to a separating funnel containing 10 ml of l o - 4 ~ dithizone in carbon tetrachloride, the funnel and its contents were shaken for about 20 seconds (longer if palladium was present), and the phases were separated by spinning in a centrifuge. This step not only removed interfering elements but also pre-saturated the aqueous phase with organic solvent. The aqueous phase was transferred to a second separating funnel containing 5 ml of 7.4 x M dithizone in carbon tetrachloride, experiments with inactive silver and active tracer having shown that a yield of a t least 50 per cent.could be expected a t this stage. The funnel and contents were shaken for about 20 seconds, the phases separated by centrifugation, and 4 ml of the organic phase was placed by pipette in a polythene container. When convenient, the silver standard was added to 1000 pg of inactive carrier in 0.5 N sulphuric acid, warmed to achieve chemical equilibrium, cooled, and 2 ml of 20 per cent. W/V hydroxylamine sulphate solution were added. The aqueous phase was extracted with two dithizone solutions in the same way as described above in the procedure for the sample, and 4 ml of silver dithizonate solution were again transferred by pipette to a polythene container. Sample and standard were counted alternately to assess the radioactivity of the solutions, and the decay curves of the nuclides were plotted.In some experiments it was found that the silver was contaminated with activity from small amounts of copper. Although silver will replace copper from dithizone solution, and in fact copper dithizonate may be used as an absorptiometric reagent for silver, it is clear from equation (4) that some copper must be extracted with the silver into the dithizone phase. Usually, the amount is small and has a negligible effect (e.g., see Table I), but the great sensitivity of many elements to neutron activation may result in minute amounts of extracted elements leading to appreciable radiochemical interference. The contribution of copper to the radioactivity of the sample could be found by resolution of the decay curve obtained, but ethylenediaminetetra-acetic acid (EDTA) has been used as masking agent for small amounts of copper when silver is being extracted by a solution of dithizone.6 Consequently, when copper was present in the sample, EDTA was added to the final aqueous phase, the conditions used being similar to those already proposed by Friedeberg.6 For a number of samples beta- rather than gamma-counting was used, since the higher sensitivity to silver activity and lower background of beta-counting were preferred t o the greater convenience of gamma-counting.For beta-counting 2000 pg of silver carrier were added to sample and standard, 15 ml of dithizone solution were used for the final extraction, instead of 5 ml, and the activity of 10 ml was counted.August, 19631 SAMPLING TECHNIQUE FOR NEUTRON-ACTIVATION ANALYSIS 607 From equation (1) it can be seen that the count rates cZ1 and cz2 of silver dithizonate derived from m, grams of silver standard and m2 grams of carrier, will be- The specific activity, S,, will be the same for both if sample and standard are irradiated in a similar neutron flux, and no self-shielding occurs; thus if M is large compared with m1 and m2, then- CZ1/C,Z = m,/mr Results obtained for the analysis of several “doped” samples by measuring the 2-3-minute activity of the silver-108 nuclide are shown in Table 11, together with the type of counter used.TABLE I1 ANALYSIS OF “DOPED” SAMPLES FOR SILVER BY MEASUREMENT OF THE ACTIVITY OF THE 2.Q-MINUTE losAg NUCLIDE Inactive silver added, Pg 12.0 7.9 5 . 2 1-06 0.74 3.77 1.25 1.24 0.38 Silver content calculated, p.p.m. 102.5 75.2 42.7 7-07 4.95 36.0 10.3 7-6 3.42 Silver content, found, Counter used p.p.m.108.0 Y 75.0 Y 36.3 Y 9.84 Y 5.72 Y 34-2 P 10-5 P 7-08 B 3-21 rs The time elapsing between receipt of the sample in the laboratory and the start of counting was usually about 8; minutes; a further 2 minutes were required for transferring the sample and standard from reactor to laboratory, so that counting usually began about 5 half-lives after the completion of irradiation. Under these conditions and with the counting techniques used, the lower limits of determination given by count rates of 600 gamma-counts per minute and 200 beta-counts per minute above background at the start of counting were found to be -0.4 pg and -0-06 pg, respectively. I t should, however, be noted that the samples could be completely dissolved in less than 1 minute, and a longer time is likely to delay the start of counting, thus reducing the sensitivity of the method. The sensitivity achieved by measuring the silver-108 activity 10 minutes after the completion of irradiation is not as high as that estimated for measurement of the activity of the 253-day silver-ll0m nuclide (5 x g after 1 month’s irradiation7). However, for the determination of more than g of silver, the short irradiation time and rapid completion of analysis when the shorter-lived nuclide is used for measurement does present advantages. (Since this paper was submitted for publication an articles has appeared in which are considered some theoretical aspects of the solvent-extraction sampling technique as applied to activation analysis.) REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. Kemp, D. M., and Smales, A. A,, Anal. Chim. Acta, 1960, 23, 397. Steele, E. L., and Meinke, W. SV., Ibid., 1962, 26, 269. RbiiCka, J., and Star$, J., Talanta, 1961, 8, 228. Cooper, S. S., and Sullivan, M. L., Anal. Chem., 1951, 23, 613. Iwantscheff, G., “Das Dithizon und seine Anwendung in der Mikro und Spurenanalyse,” Verlag Friedeberg, H., Anal. Chem., 1955, 27, 305. Jenkins, E. N., and Smales, A. A., Quart. Rev., 1956, 10, 83. KbiiCka, J., and Star$, J . , Talanta, 1963, 10, 287. Chemie, G.m.b.H., Weinheim, 1958. Received March 21st, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800603

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

608 STANTON AND MCDONALD : APPLICATION OF THE AUTOANALYZER [Analyst, Vol. 88 Application of the Auto Analyzer to the Determination G H . of Zinc in Soils and Sediments 0-32 ml per mini ~ 3.40 ml per mini i Hydrochloric BY R. E. STANTON AND ALISON J. McDONALD (Department of Geology, Imperial College of Science and Technology, London, S . W.7) The AutoAnalyzer is adapted to the determination of zinc with dithizone. A solvent-extraction technique, which requires only simple modifications of the system, is applied to the analysis of solutions from soil and sediment samples containing as little as 0.1 pg of zinc per ml. Samples are analysed a t the rate of 250 per day, and a t this speed the results obtained are within & 5 to 10 per cent. a t the 95 per cent. confidence level.A FIELD method for determining zinc in soils, suitable for geochemical prospecting, has been developed by Lakin, Stevens and A1mond.l The procedure2 in current use at the Geochemical Prospecting Research Centre, Imperial College, was derived from this method ; \ / / / - - \ I \ I \ Sampler I I 1 Proportioning pump 0.32 ml per m i n r - 1 ... J//' \ I I \I, ' I -- 2.50 ml Der min' I D \hf--.-.- /I id,, V V d L C l .-\, 12.50 mi per mini f E Water - I I I I I . . . acid 5-1 it re bottle 1 1 of di t h izone L--I I solution Range tiometer \1 To waste Fig. 1. Arrangement of apparatus it consists of the visual comparison of sample solutions with a standard series and makes use of the mixed colour reaction of zinc with dithizone in benzene solution. Concentrations over the range 5 to 350 p.p.m.are determined from a portion representing 10 mg of the sample. This is typical of many methods now available for geochemical field work, and permits the analysis of 100 samples by one person in an 8-hour day. That this rate of analysis can be achieved is due largely to the fact that results within +25 per cent. at the 95 per cent. confidence level are adequate for most prospecting purposes. The Technicon AutoAnalyzer was conceived by Skegg~,~ and the mechanics and principles have been discussed by Ferrari, Russo-Alesi and Kelly.* The geochemical field method for zinc has been adapted to this system, resulting in an improvement in both analytical precision and the lower limit of determination, together with an increased speed of analysis.August, 19631 TO THE DETERMINATION OF ZINC IN SOILS AND SEDIMENTS DESCRIPTION OF APPARATUS 609 Proportioning $ump-The manifold assembly is shown diagrammatically in Fig.1, the pumping action being such that solutions progress from right to left. Sampler-The Technicon “Super-sampler” consists of a circular aluminium rack that holds test-tubes (19 mm x 150 mm) in four concentric circular paths of 50 tubes each. With the pump in action, a solenoid-operated polythene sampling tube drops vertically into a test- tube and aspirates sample solution for 1 minute (Fig. 1, tube A) ; a t the end of this period the tube rises to aspirate air, and on completion of a total cycle time of 2 minutes the rack rotates sufficiently to present the next sample, and the tube falls again.A modification has been made to the arm supporting the sampling tube so that when this tube rises to aspirate air, a second tube, identical in bore and length, falls into 0.5 N hydrochloric acid (Fig. 1, tube B). The alternate intake of sample solution and acid ensures a constant volume of air and aqueous phase, which is essential for the subsequent phase-separation stage. The total cycle and sampling times are pre-set by means of two indeDendent controls, each of which extends to 35 minutes with sub-divisions mechanically impossible to operate with a sample time modification of the equipment. Mixed .- Open to of 30 seconds.* However, it is shorter than 1 minute without phases Benzene phase to absorptiometer Fig. 2. Modified Y-tube Solvent extraction-Since the Tygon pump tubing is not resistant to benzene, dithizone dissolved in this solvent is introduced into the system by displacement with water, which is pumped through tubes D and E (Fig.1) into a 5-litre darkened bottle (at F) containing the dithizone. At G, the aqueous solutions and air (tubes A, B and C) pass through a coil of glass tubing of 2-mm bore with 28 turns approximately 1.5 cm in external diameter, in order to mix sufficiently before meeting the dithizone stream at H. This junction is made with a glass h-piece containing a constriction, which has the effect of distributing the benzene phase as a series of small globules throughout the aqueous solution, and the combined streams then circulate through seven more mixing coils at I. The mixing coil assembly lies hori- zontally, and zinc dithizonate is extracted into the benzene phase by repeated inversion of the mixed phases as they move along.An adequate extraction is achieved by the time the stream enters (at J) the modified Y-tube (see Fig. 2), where the phases separate and the benzene solution passes into the flow cuvette of the absorptiometer, the aqueous phase running to waste. After passage through the cell, the solution runs to waste through a tube at the bottom of the absorptiometer. The drainage tube supplied with the instrument is made of material unsuitable for use with benzene, and it must be replaced; a glass condenser adapter is a convenient substitute. Range expander-The photocell response is received by this unit, and it may be mag- nified by a factor of 2, 4 or 10 before transmission to the recorder.Recorder-This is operated with a chart speed of 10 inches per hour. Absorptiometer-A flow cell of 1-cm light path is used with 538-mp filters.610 REAGENTS- STANTON AND MCDONALD : APPLICATION OF THE AUTOANALYZER [Analyst, Vol. 88 METHOD Potassium hydrogen sulphate, fused, powder. Hydrochloric acid, N and 0.5 N-Prepared from concentrated acid of analytical-reagent grade. Dithizone solz&on-A 0.01 per cent. w/v solution in carbon tetrachloride, and a 0.001 per cent. w/v solution in benzene prepared each day from a 0.02 per cent. w/v stock solution in benzene. Use analytical-reagent grade dithizone for preparing these solutions. Carbon tetrachloride-Analytical-reagent grade. Benzene-Crystallisable grade.Acetate bufer solution-Dissolve 500 g of sodium acetate trihydrate, 125 g of sodium thiosulphate pentahydrate and 5 g of sodium fluoride in about 1.5 litres of water, add 15 ml of glacial acetic acid, and extract with 0.01 per cent. dithizone in carbon tetrachloride until free from zinc. Remove the excess of dithizone by extraction with carbon tetrachloride, and then dilute to 2 litres with water. Standard zinc solutions-Dissolve 0.2199 g of analytical-reagent grade zinc sulphate heptahydrate in 0.5 N hydrochloric acid, and dilute with the same acid to 500 ml in a cali- brated flask. From this solution (containing 100 pg of zinc per ml) prepare as required, by dilution with 0.5 N hydrochloric acid, solutions containing from 0.1 to 10 pg of zinc per ml.PROCEDURE- Weigh 0-1 g of the sifted sample, place in a numbered 19-mm x 150-mm borosilicate- glass test-tube, mix with 0-5 g of potassium hydrogen sulphate powder, and fuse. Cool, leach the residue with 5 ml of N hydrochloric acid, dilute to 10 ml with water, and mix. Assemble the manifold as indicated in Fig. 1, and connect to the appropriate reagent solutions. Insert 538-mp filters and a 1-cm flow cell, set the range expander at 1, bring the pump into operation, and allow the supplementary sample-line to aspirate 0.5 N hydrochloric acid continuously, Adjust the recorder and absorptiometer to read zero transmission when no light is passing through the flow-cell; then select a convenient magnification factor on the range expander, and obtain an arbitrary base-level for 100 per cent.transmission. Load the sampler rack with sample solutions, several solutions for the determination of reagent blank values, standard solutions at intervals of 1 in 20, control samples at a frequency of 1 in 40, and test-tubes containing only 0.5 N hydrochloric acid at an incidence of 1 in 40. Set the sampler unit in operation with a total cycle of 2 minutes and a sampling time of 1 minute. After the last solution has been sampled, allow the supplementary line to aspirate 0.5 N hydrochloric acid for a further 15 minutes, until the final peak has been traced. Switch off the system, isolate the bottle containing the solution of dithizone in benzene, place the reagent lines in water, pump for a further 5 minutes, and then release the roller-head from the platen of the proportioning pump.Remove the recorder chart, and draw a baseline through all the zero levels obtained from the 0.5 N acid solutions and the initial setting. Measure the optical density directly from this line by using a ruler calibrated with the appropriate logarithmic scale. Make a graphical plot of the results obtained for standard solutions and interpolate the zinc concentrations of the sample solutions. When the zinc concentration is outside the range covered by the magnification factor selected on the range expander, adjust this unit accordingly and repeat the determination. DISCUSSION OF RESULTS For prospecting purposes, it is only necessary to take into account cobalt, copper, lead and nickel when considering the effect on the determination of zinc of other metals that react with dithizone.The aqueous phase will have a pH value in the range 5.6 to 6-0, within which there is no appreciable reaction by nickel. The reactions of the other three metals are not completely prevented by sodium thiosulphate, and their dithizonates have absorption maxima too close to that of zinc dithizonate for the optical filters to discriminate; from solutions each containing 100 pg of cobalt, copper and lead per ml, apparent zinc values of approximately 10, 3 and 1 pg per ml will be obtained, respectively. These degrees of inter- ference are tolerable for the proposed application.August, 19631 TO THE DETERMINATION OF ZINC IN SOILS AND SEDIMENTS 611 Fluoride is necessary in the buffer solution to mask aluminium, which would otherwise inhibit either the formation or the extraction of zinc dithizonate (or both) when present at a concentration greater than 150 times that of zinc.Slight decomposition of the dithizone solution may occur, causing a drift of the baseline. With a magnification factor of 4 or 10, the fall in base-level should not exceed 0.02 or 0.04 optical-density units, respectively, over an 8-hour period ; this is equivalent to about 0.05 pg of zinc per ml in each instance. When there is a greater change in the baseline, either the dithizone is too unstable to be used or else the drift is indicative of some instrumental fault. Alternative means of decomposing the sample are not precluded, such as digestion with perchloric acid or aqua regia, or the preliminary removal of silica by treatment with hydro- fluoric acid.The sample attack is merely restricted by its subsequent chemical effect and the need to obtain a final aqueous phase with the same pH value and concentrations of fluoride and thiosulphate as those obtained in the procedure described here. It is convenient to adopt a method of attack that can be carried out in test-tubes, since these are necessary to contain the solutions in the sampler unit. Sample solutions contain an appreciable amount of suspended solid matter (mainly siliceous) that is carried into the system with the sample stream and collects at the aqueous - benzene interface in the phase-separation cell. In early experiments, air was introduced continuously to promote mixing of the benzene and aqueous layers, but when escaping at the phase-separation stage it caused some solids to be carried over into the flow cell of the absorptiometer, which gave rise to erratic tracing on the recorder chart; the air intake was minimised accordingly. Optimum conditions for solvent extraction were established experimentally with up to twelve mixing coils.Centrifugation occurred with eight or more coils, causing coalescence of the previously segmented benzene stream and thereby spasmodic overflow into the absorp- tiometer cell, which resulted in a chart tracing of poor quality. Zinc dithizonate is extracted less rapidly from sample than from standard solutions, and care was taken to ensure that the same degree of extraction from both was achieved. Mostly, they are within 15 per cent. of their nominal value, but it has been found possible to deviate seriously from the optimum proportions of sample and reagent solutions by taking the tubes at their face value.Results on control samples distributed among 550 routine samples have been used for calculating standard deviations; the same solu- tions were also analysed successively at as many amplification factors as possible. A comparison of results obtained manually and by the AutoAnalyzer is shown in Table 11. It is essential to calibrate the pump tubes before assembling the manifold. Some figures for precision are shown in Table I. TABLE I REPRODUCIBILITY OF RESULTS FOR ZINC KO. of Mean value, Standard Range expander Sample determinations p g per ml deviation setting A 6 0-28 0.019 10 A 8 0.30 0.02 1 4 B 6 0.12 0.011 10 C 6 1-23 0.053 4 C 5 1.16 0.023 2 D 6 3.01 0.13 2 A 5 0.27 0.009 10 A 7 0.29 0.006 4 B 7 0.1 1 0.003 10 C 7 1-18 0.054 4 C 7 1.13 0.050 2 C 7 1.28 0.025 1 D 7 3-00 0.092 2 D 7 3-00 0.105 1 From ruutixe opeyatiun- From consecutive analyses- The results in Table I1 were obtained on a set of control samples prepared according to Craven’s recommendation^.^ Two bulk samples were obtained, one (h) containing about 380 p.p.m.of zinc and the other ( I ) about 35 p.p.m. of zinc; a series of twelve samples was612 STANTON AND MCDONALD APPLICATION OF THE AUTOANALYZER [AnaZyst, vol. 88 TABLE I1 COMPARISON OF RESULTS OBTAINED MANUALLY AND WITH THE AUTOANALYZER Zinc found- r- Sample No. manually, p.p.m. 1 200 2 400 3 250 4 200 5 380 6 190 automatically, p.p.m.213 385 268 163 363 208 Zinc found- 7-A- Sample No. manually, automatically, 7 140 125 8 40 40 9 10 308 10 30 43 11 50 55 12 210 21 3 p.p.m. p.p.m. prepared by mixing different amounts of the two bulk samples. For each sample in this series, the weight fraction of the high component is designated 8 and that of the low com- ponent A, so that 0 = h/(h + I ) and A = Z/(h + I ) for any given sample. When the twelve samples had been analysed, values for the high and low components (H and L , respectively) were calculated from- CABZA2 - ZAAZBA H=-- C82ZA2 - (ZBA)2 ze2w - (Cm)2 ZAAZO2 - CIAOCBh and L = where A is the-analytical result found for a sample. sample, from A = HB + LA, and the standard deviation was obtained from- 0 = dZ (A - A)”(n - 2).By Craven’s definition the percentage mean accuracy is obtained from (100 x 2o)/&(H + L ) . The results in Table I1 when treated in this manner give- A value was then calculated for each H L Mean accuracy, % Manual . . . . 383 34 22.7 Automatic . . . . 385 36 9.0 From this statistical treatment it may be presumed that there is good agreement between the two methods, but that the results from automatic analysis are more mutually consistent than are those obtained manually. TABLE I11 CONCENTRATION RANGE FOR EACH RANGE EXPANDER SETTING Expansion (----- -3 Range of zinc concentrations, pg per mi factor Linear Non-linear 1 0-25 t o 4.00 4.0 to 10.0 2 0.10 t o 2-00 2.0 t o 4.0 4 0.05 t o 1.00 1.0 t o 2.0 10 0.01 t o 0.50 0.50 t o 0.75 The calibration graph for each setting of the range expander is not linear over its full extent; the ranges are shown in Table 111.Interpolation from the curved section of the calibration graph has proved to be satisfactory, but for a large number of samples it is likely to be quicker to repeat at a lower sensitivity. By using a sample solution, the effect of variations in the timing of the sampler unit operation was studied to discover how much the precision and the lower limit of the deter- mination might be improved at the expense of the speed of analysis (see Table IV). By increasing the sampling time to 2 minutes, the sensitivity of the determination is improved by 15 per cent., but a further improvement of only 4 per cent, is obtained with a sampling period of 3 minutes.The start-up procedure and the final flushing out of the system may be accelerated by use of a two-speed proportioning pump. This can operate at a fast pumping rate for 3 or 4 minutes and gives an additional 30 to 40 minutes of analytical time. Without this advantage, the equipment has been left unattended for 2 or 3 hours at a time and allowed toAugust, 19631 TO THE DETERMINATION OF ZINC IN SOILS AND SEDIMENTS 613 run during the lunch-hour, and one person can thus analyse 250 samples in an 8-hour working day, in addition to control samples and standard solutions. One sample takes about 15 minutes to be processed by the system. TABLE IV REPRODUCIBILITY AT DIFFERENT TOTAL CYCLE AND SAMPLING TIMES Zinc concentration Total cycle, Sampling time, minutes minutes 2 1 3 1 4 1 3 2 4 2 5 3 30. of determinations 19 10 9 10 10 10 Mean, PLg per 1.97 1.99 2-02 2.00 2.02 2.01 S tandari deviation 0.038 0.032 0.026 0.049 0-045 0-028 Although no sample solutions with less than 0.1 pg of zinc per ml were analysed, a lower limit of 0.01 pg per ml can be attained. When this is necessary, it would be advisable to use a total cycle of 5 minutes with a sampling time of 3 minutes. The work described formed part of the programme of the Geochemical Prospecting Research Centre under Professor J. S. Webb and was assisted by a special grant from the Department of Scientific and Industrial Research. REFERENCES 1 . 2 . 3. 4. 5 . Lakin, H. W., Stevens, R. E., and Almond, H., Econ. Geol., 1949, 44, 296, Stanton, R. E., Imperial College of Science and Technology, London, Geochemical Prospecting Research Centre, Technical Communication No. 19, 1962. Skeggs, L. T., Amer. J . Clin. Path., 1957, 28, 311. Ferrari, A, Russo-Alesi, F. M., and Kelly, J . M., Anal. Chem., 1959, 31, 1710. Craven, C. A. U., Tyans. Inst. Man. Metall., Lond., 1953/1954, 63, 551. First received October 1 ‘ith, 1962 Amended, Mavch 22nd, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800608

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

614 BUSH : POLAROGRAPHIC DETERMINATION OF COPPER, CADMIUM, [Analyst, VOl. 88 The Polarographic Determination of Copper, Cadmium, Thallium, Lead, Tellurium and Iron in Selenium BY E. L. BUSH (Standard Telecotnmunication Laboratories Lid., London Road, Harlow) A polarographic method has been developed for determining copper, cadmium, thallium, lead, tellurium and iron in high purity selenium used for making rectifiers. The selenium matrix is removed completely by dis- tillation as selenium bromide from hydrobromic acid solution. The residue is dissolved in specially purified oxalic acid solution, which acts as the electrolyte. On a 2-g sample of selenium, 0.3 p.p.m. of copper, 0.02 p.p.m. of lead, 0.0015 p.p.m. of tellurium, 0.01 p.p.m. of cadmium, 0.05 p.p.m. of thallium and 0.4 p.p.m.of iron can be determined in 1 ml of solution. The limits of de- tection for copper and iron are poor owing to relatively high blank values of 0.1 p g and 0.2 pg, respectively. ELEMENTAL selenium is used for making rectifiers, the electrical characteristics of which are affected by the presence of metallic impurities. Relatively little has been published on the determination of trace impurities in se1enium.l ,2 9 3 7 4 Peterson and Currier1 spectrographically determined 1 to 100 p.p.m. of aluminium, bismuth, copper, lead, magnesium, nickel, silver and zinc in selenium, Itsuki and Kaji2 determined copper and lead in high purity selenium with a square wave polarograph. They employed perchloric acid as electrolyte, after removing the selenium with sulphuric and nitric acids.P0hl3 determined thallium, iron, copper, cadmium, bismuth and lead polarographically in selenium, He dissolved the selenium in nitric acid, removed the selenium dioxide by volatilisation in a muffle furnace, and separated the iron and thallium by extraction with di-isopropyl ether from 6 r ~ ’ hydrochloric acid. After the solution had been evaporated to dryness and the residues dissolved in ammonium tartrate, the other metals could be deter- mined. Jones4 determined tellurium, lead, copper, nickel and cobalt polarographically in selenium. He removed the selenium dioxide by volatilisation from hot concentrated sulphuric acid solution. The tellurium was separated from the other elements by precipitation with sulphur dioxide. The copper and lead were determined in 0.1 N hydrochloric acid - 0.1 N potassium chloride solution after reduction of the iron to the ferrous state with hydroxylamine hydro- chloride. The cobalt and nickel were determined in pyridine solution, and the tellurium determined in ammonium hydroxide - ammonium chloride electrolyte.An advantage of the polarographic method is that more than one element can readily be determined on the same prepared sample solution. However, selenium in most electrolytes produces large reduction and catalytic waves, which would seriously interfere with the determination of many other elements. It is therefore necessary to remove selenium before making the final determination. Selenium may be removed by precipitation, but many other elements are co-precipitated,j and so this method was not used.Distillation from concentrated sulphuric acid solution has been r e p ~ r t e d , ~ but it is necessary to purify the sulphuric acid before use to limit the blank values for the whole procedure. A more convenient process involves distillation as the bromide from hydrobromic acid solution.6 With this technique most elements remain as involatile bromides. Briefly, the method described here involves dissolution of the sample in a mixture of nitric and hydrochloric acids, removal of the excess of nitric acid, distillation of selenium bromide, and finally dissolution of the residue in the appropriate electrolyte, which is examined polarographically. A simple but extremely effective heater assembly is used for evaporating liquids in beakers, which has a provision for minimising contamination by airborne impurities (see Fig.1).August, 19631 THALLIUM, LEAD, TELLURIUM AND IRON IN SELENIUM 615 METHOD REAGENTS- Hydrochloric acid, 6 to 7 N-Distil 7 N hydrochloric acid in quartz apparatus. Nitric acid-Distil analytical-reagent grade concentrated nitric acid in quartz apparatus. Hydrobromic acid-Distil 48 per cent. hydrobromic acid, of analytical-reagent grade, in quartz apparatus. Oxalic acid, 0.2 M-Dissolve 25 g of analytical-reagent grade oxalic acid in 1 litre of water, and pass the solution through an ion-exchange column packed with analytical-grade Amberlite 2 M Ammonium acetate - 2 M acetic acid - 0.004 M EDTA solution-Prepare a solution of 1.5 g of analytical-reagent grade disodium ethylenediaminetetra-acetate in 230 ml of redis- tilled analytical-reagent grade acetic acid and 145 ml of redistilled analytical-reagent grade ammonia solution.IR-l20(H). Dilute to 1 litre with water distilled in quartz apparatus. 3 " 64 - silic +' thick aluminium lass- or silica-wool 4 Pyrex-glass beaker Fig. 1. 50-ml beaker heater with down-draught assembly SEPARATION PROCEDURE- Place 2 g of sample in a 50-ml quartz beaker, and dissolve in 5 ml of (1 + 1) hydrochloric acid and 10 ml of concentrated nitric acid. When the reaction has subsided place the beaker in the heater assembly provided with a down-draught of filtered air (see Fig. l ) , and evaporate the solution to dryness. After allowing the residue to cool slightly, carefully add 20 ml of hydrobromic acid, replace the mixture in the heater assembly, and evaporate off the selenium as the tetrabromide.I t is important at this stage that the temperature of the heater is not above 155" C (measured with a thermocouple between the beaker and heater wall) otherwise some tellurium will be lost as tellurium bromide, boiling-point 421" C. After the solution has been evaporated to dryness and the residue cooled slightly, add a further 1 ml of nitric acid to re-convert into selenious acid any elemental selenium that may have been formed. Transfer the solution obtained to a 10-ml calibrated flask, and make up to the mark with distilled water. Transfer two separate 4-ml portions to 50-ml quartz beakers, and evaporate to dryness as before. Add to the contents of each beaker 1 ml of hydrobromic acid, and remove the remaining selenium as described above.The residues thus obtained are polarographically free from selenium and contain the trace-impurity metals as bromides.616 POLAROGRAPHIC PROCEDURE- BUSH : POLAROGRAPHIC DETERMINATION OF COPPER, CADMIUM, [Analyst, VOl. 88 A Southern Instruments’ A1660 Davis differential cathode-ray polarograph was used. The approximate reduction potentials of the elements in oxalic acid at pH 2 are- Element . . . . . . Iron111 Copper Titanium Lead Thallium Cadmium Tellurium Reduction potential, volts . . 0.15 0.25 0.60 0.65 0.65 0’7.5 0.95 DETERMINATION OF IRON, COPPER, CADMIUM, TELLURIUM AND COMBINED LEAD AND Dissolve the residue from one of the beakers (see “Separation Procedure”) in 2 drops of 0.1 N hydrochloric acid and then 0.5 ml of 0.2 h i oxalic acid.Transfer the solution to a l-ml calibrated flask by means of a glass dropping tube. Rinse the beaker with water, and transfer the rinsings to the l-ml flask. Add 1 drop of 0.5 N ammonium hydroxide to the solution in the flask, in order to neutralise the hydrochloric acid. Make the solution up to the mark with distilled water, and transfer the solution to a semi-micro polarographic cell. Prepare a reagent blank solution treated in a similar manner to the sample, and place it in the other cell of the differential polarograph. De-oxygenate the oxalic acid solutions in the polarographic cells with pure nitrogen for 1 to 3 minutes. Switch on both cells, and measure the reduction waves for iron, copper, cadmium, tellurium and the combined lead and thallium.(The iron and copper waves can be measured either on forward or reverse sweep.) Determine the iron, copper, cadmium and tellurium contents by comparison with a standard solution containing known amounts of these elements. If a reduction wave for lead or thallium is obtained, then these elements may be determined by using 1 M ammonium acetate - 1 M acetic acid - 0.002 M EDTA electrolyte. THALLIUM- DETERMINATION OF LEAD AND THALLIUM- EDTA are- Approximate reduction potentials in 1 M ammonium acetate - 1 M acetic acid - 0.002 M Element . . . . . . Iron111 Copper Thdlium Titanium Bismuth Tellurium Lead Reduction potential, volts . . 0.15 0.3 0.45 0.5 0.7 1.0 1.0 To the residue in the second beaker (see “Separation Procedure”) add 0-5 ml of 2 M Transfer the solution to a Transfer the solution Prepare a reagent blank solution in a similar Switch on both cells, The thallium wave is best measured Determine the lead and ammonium acetate - 2 M acetic acid - 0-004 M EDTA solution.l-mt calibrated flask, and make up to the mark with distilled water. obtained to a semi-micro polarographic cell. manner to the sample, and place it in the other cell of the differential polarograph. and measure the reduction waves for thallium and lead. on reverse sweep to prevent interference from the copper wave. De-oxygenate the solutions with pure nitrogen for 1 to 3 minutes. TABLE I TYPICAL RESULTS Recoveries were made with the heater at 155” C (22 volts) Copper Cadmium Lead 7+ 7- & Added, Found, Added, Found, Added, Found, PLg tLg PP Pg P-Lg Pg Reagent blank value - 0.1 0.0 1 0.0 1 Recovery .. . . 1.2 1.2 0.54 0.54 1.7 1.7 - - Reagent blank value Recovery . . . . Tellurium Thallium Iron -7 -7 -7 Added, Found, Added, Found, Added, Found, CLg PLg Pg Pg PFLg Pg 0.2 - <0.0015 - < 0.005 - 1.5 1.5 1.1 1.1 1-6 1.6August, 19631 THALLIUM, LEAD, TELLURIUM AND IRON IN SELENIUM 617 thallium contents by comparison with a standard solution containing known amounts of these elements. If no thallium is present, then the lead can be determined from the results obtained on the oxalic acid solutions. SENSITIVITY OF THE METHOD- If the impurity level is extremely low then the final volume of solution used for polaro- graphy can be reduced from 1 to 0.20ml. The increased sensitivities will, however, be accompanied by a consequent reduction in accuracy, owing to the inevitable errors in measur- ing and handling this volume.I t is, however, thought that this sacrifice is warranted in some circumstances. RESULTS Tables I and I1 list some typical results obtained with the methods described. The necessity of avoiding excessive temperatures is indicated by the results shown below for the recovery of 1.5 pg of tellurium at different heater temperatures- Temperature of heater, "C . . 258 (30 volts) 155 (22 volts) 100 (19 volts) Tellurium found, pg . . 0.35 1.5 1.5 TABLE I1 TYPICAL RESULTS Sample Copper Cadmium Lead Tellurium Thallium Iron number found, p.p.m. found, p.p.m. found, p.p.m. found, p.p.m. found, p.p.m. found, p.p.m. * * 0.16 1 0.007 1.3 2 * 0.09 0.68 0.014 0.16 * 3 0.09 0.48 0.022 0.16 * * * * Equivalent t o the reagent blank value. DISCUSSION OF THE METHOD In the methods described, oxalic acid was used as electrolyte mainly because it can readily be purified by passage through ion-exchange resin. However, in different electrolytes it is possible to determine other elements having involatile bromides. For example, 0.2 M lithium chloride electrolyte was used for determining barium in the three samples described. The barium reduction wave occurs at -2.0 volts in neutral lithium chloride, and barium contents of 0.1 to 1.5 p.p.m. have been measured. REFERENCES 1. 2. 3. 4. 5. ti. Luke, C. L., Anal. Chew%., 1949, 21, 1369. Peterson, G. E., and Currier, E. W., Appl. Spectroscopy, 1956, 10, 1 . Itsuki, K., and Kaji, T., Japan Analyst, 1956, 8, 703. I'ohl, F. A., J . Polavographic Soc., 1958, 1, 8. Jones, R. H., Analyst, 1946, 71, 60. Bode, H., 2. anal. Chem., 1956, 153, 335. First received August 30th, 1963 Amended, February 28th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800614

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

618 SCAIFE : PRECISE MICRO-DETERMINATION OF ZINC AND [Analyst, VOl. 88 Precise Micro-determination of Zinc and Cadmium by Photometric Titration with Disodium Ethylenediaminetetra-acetate BY D. B. SCAIFE (Department of Applied Chemistry, Northampton College of A duanced Technology, St. John Street, London, E.C.l) A method is described for determining zinc and cadmium by weight titration with disodium ethylenediaminetetra-acetate, Eriochrome black T being used as indicator and a photometric end-point detection technique being employed. By this method, 0.6- and 0.06-mg amounts of zinc can be determined with standard errors of about 0.02 and 0.1 per cent., respectively; similar amounts of cadmium can be determined with the same precision. A METHOD of high precision ( k 0 - 1 per cent.or better) for determining micro amounts of zinc and cadmium was required in connection with an investigation of the stabilities of certain halide complexes of these metal ions in aqueous solution. Complexometric titrations are suitable for determining small amounts of metal ions, and methods for zinc and cadmium have been estab1ished.l For extremely low concentra- tions of metal, photometric end-point detection techniques have been used. Sweetser and Bricker2 determined zinc and cadmium in solution at pH 10 by photometric titration with disodium ethylenediaminetetra-acetate (EDTA;i at 222 to 228 mp. Ten- and 1-mg amounts of cadmium were determined with standard errors of about 0.3 per cent. and 0.5 per cent., respectively; similar amounts of zinc were determined with slightly better precision.Hunter and Miller3 determined zinc by photometric titration with EDTA and Eriochrome black T as indicator. The results for the titration of 5- and 0.5-mg amounts of zinc in the presence of various other metals showed standard deviations of 0-3 per cent. and 1.6 per cent., respectively. In the investigation described here, a photlometric method similar to that of Hunter and Miller3 has been developed for the precise determination of amounts of both zinc and cadmium less than 1 mg. An important innovation was the use of a weight titration technique, and special attention was given to the design of the titration cell. Many different types of titration cell have been designed for use with various com- mercial spectrophotometers in which the normal cell compartments have been suitably modified so that titrations can be carried out in sitzc.(An excellent review of photometric titration techniques and apparatus is given by Headridge.4) A disadvantage of most of these designs is that, after the initial adjustment of the optical-density scale (usually the untitrated solution is taken as zero), further checks on the reference point cannot be made. Serious end-point extrapolation errors may result from fluctuations in the reference level, particularly if these occur during the final stages of the titration. This difficulty has been resolved in the design of the titration cell described below. APPARATUS- I t was constructed from &-inch and $-inch Perspex sheet and had a titration volpme of about 25 1x1, the light path being 2 cm in length.The cell was designed to stand on the normal cell-holder carriage of a Unicam SP500 spectro- photometer. Incorporation of a reference cell allowed zero checks to be made before each optical-density reading. Increments of titrant were added with the cell outside the cell compartment, and effective mixing was achieved by tilting the cell several times. This titration procedure was chosen in preference to an in situ method for two main reasons. First, no modifications had to be made to the cell compartment of the SP500 spectrophoto- meter. Secondly, since only relatively few points near the end-point were considered once the shape of a titration curve had been established, the in situ method offered no real advantages with respect to time-saving. EXPERIMENTAL The titration cell is shown in Fig.1.August, 19631 CADMIUM BY PHOTOMETRIC TITRATION WITH EDTA 619 REAGENTS- All solutions were made up with de-ionised water and stored in polythene bottles. Standard metal solzttions-Solutions containing more than 0.15 mg of metal per milli- litre were prepared by dissolving a known weight of pure metal (purity greater than 99.99 per cent.) in a minimum amount of analytical-reagent grade nitric acid and making up to a known weight of solution (about 2000 g) with pure water. Solutions containing less than 0.1 mg of metal per millilitre were made up by weight dilution of more concentrated solutions. EDTA sol.ution-A known weight of analytical-reagent grade disodium ethylenediamine- tetra-acetate, Na2H2C,,H,20,N2.2H20 (previously dried a t 80" C for 4 days) was dissolved in pure water and made up to a known weight of solution.Buffer solution, pH 10-Analytical-reagent grade ammonium chloride (35 g) was dissolved in 285 ml of analytical-reagent grade ammonia solution, sp.gr. 0.880, and the solution was made up to 500ml with pure water. Eriochrome black T-The indicator (0.1 g) was dissolved in a mixture of 15 ml of tri- ethanolamine and 5ml of ethanol. DETERMINATION OF WAVELENGTH FOR TITRATION- A Perkin - Elmer 137UV recording spectrophotometer was used to determine the visible spectrum of a solution containing 19 ml of water, 1 ml of buffer solution, 6 drops of 0-001 M EDTA solution and 6 drops of Eriochrome black T indicator solution.Similar experiments Scale cm 0 5 / / Reference cell Titration space Fig. 1. Titration cell: (a) side view; ( b ) plan view were carried out on (a) a solution containing 19 ml of water, 1 ml of buffer solution, 6 drops of 0.001 M cadmium solution and 6 drops of indicator solution and (b) a solution made up as for (a), but containing 6 drops of 0.001 M zinc solution instead of the cadmium solution. Comparison of the spectra showed that a maximum difference between the optical densities of the free indicator and the metal - indicator complex occurred at 640 mp for zinc and at 660mp for cadmium. PHOTOMETRIC TITRATIONS- A known weight (about 4 g) of the metal solution to be titrated was dispensed into the titration cell from a weight burette. The burette was made by drawing out the stem of a 25-ml dropping funnel to a fine jet and attaching a metal hook to the neck to facilitate weigh- ing. One millilitre of buffer solution at pH 10 and 2 drops of Eriochrome black T indicator solution were added, together with a volume of pure water calculated to give a total titration volume of 20 ml.The indicator concentration was such that optical-density values at the end-point fell in the range 0.4 to 0.5, i.e., where relative errors in optical-density readings are a minimum. EDTA solution was then added until the red solution began to turn blue. At this stage, the burette was re-weighed, and the optical density of the solution after it had been thoroughly mixed was measured at the appropriate wavelength with pure water as the reference solution.Small increments of titrant were then added, and, after each addition, the optical density of the solution was measured and the burette re-weighed. Several additions were made after the end-point had been reached. It was possible to transfer increments of titrant weighing as little as 0-005 g by allowing a small drop to form on the These wavelengths were used for the titrations.620 SCAIFE PRECISE MICRO-DETERMINATION OF ZINC AND [AnaZYjst, VOI. 88 tip of the burette and bringing this into contact with the surface of the solution in the titration cell. A graph of optical-density readings against the weight of EDTA solution added was constructed, and the end-point was determined by extrapolation in the usual way. Optical- density readings were not corrected for volume changes, since these were negligible in the extrapolated region of the titration curves.Four series of titrations were carried out on (i) l-mg portions of cadmium, (ii) 0.1-mg portions of cadmium, (iii) 0-6-mg portions of :zinc and (iv) 0-06-mg portions of zinc. The reagent blank value was determined by titrating a solution containing 1 ml of buffer solution, 2 drops of indicator solution and 19 ml of water. RESULTS AN:D DISCUSSION The reagent blank value was found to be 0.022 g of the EDTA solution used in the metal titrations. This value was the mean of six determinations, viz., 0-022 g, 0.023 g, 0.022 g, 0.022 g, 0.022 g and 0-021 g of EDTA solution. In the first of the two series of eight titrations of cadmium solutions (concentration 2-2452 mmoles per kg) the average weight of cadmium taken was about 0-92 mg.By taking the EDTA solution as standard, i.e., assuming the exact composition of the sample of de- hydrated disodium salt to be Na2H2C,,H,20,N,.2H,0,5 the weights of cadmium found were consistently higher by about 0.2 per cent. than the weights of cadmium taken (cadmium metal = 100 per cent.). It was considered that this discrepancy was likely to be due to the non-stoicheiometry of the EDTA salt. The results of this series of titrations were used to standardise the EDTA solution a procedure that was fully justified by the results of the subsequent metal determinations. The concentration of the EDTA solution (cadmium = 100 per cent.) was found to be 0.0012609 moles per kg, the standard deviation from the mean being 0.02, per cent.The results of the other series of titrations of cadmium and zinc are summarised in Table I. TABLE I THE RESULTS OF THREE SERIES OF CADMIUM AND ZINC TITRATIONS Concentration Number Approximate Approximate Standard error of titrand, of deter- weight of weight of in weight of Titrand mmoles minations metal taken, titrant required, metal found, O/ Per kg mg g / O Cadmium solution . . 0.26697 8 0.13 0.9 0.1, Zinc solution . . . . 2.0146 8 0.61 7.5 0.02, Zinc solution . . . . 0.19626 8 0.057 0.7 0.1, The main source of error in the method was found to arise from the graphical extrapola- tion used to locate the end-point of the titratiolns. This can be demonstrated by considering the titration curves in the region near the end-point. Figs.2 (a) and 2 (b) are typical curves for the titration of 1 mg of cadmium and 0.6 mg of zinc, respectively; similar curves were obtained in the low concentration range (0.1 mg of cadmium, 0.06 mg of zinc). In all titra- tions the weight of titrant required could be determined to within 0.001 g of solution. Errors arising from this effect could therefore be as much as 0.01 per cent. for determinations of high concentrations of metal (when the weight of titrant was about 7 g) and approximately 0.1 per cent. for determinations of low concentrations of metal (weight of titrant about 0.8 g). These errors are of the same order as the standard errors observed in the respective concen- tration ranges. An attempt to improve the precision of the determinations of low concentrations of metal by using a more dilute EDTA solution was not successful; although a higher weight of titrant was required, the dilution had the effect of producing a lower end-point slope and thereby increasing the error in extrapolation. A feature of interest in Figs.2 (a) and 2 ( [ I ) is the difference in the shape of the zinc and the cadmium curves. For the cadmium titration, the slope near the end-point increases with increasing weight of titrant added, whereas for zinc the slope decreases. An attempt to correlate this effect with the theoretical work of Reilley and Schmid6 was prevented by lack of stability constant data for the metal - Eriochrome black T complexes,.August, 19631 CADMIUM BY PHOTOMETRIC TITRATION WITH EDTA 62 1 Application of the method to even smaller amounts of metal, of the order of lOpg, should be feasible, although errors of about 2 per cent, would be expected, assuming the same precision in end-point location. However, Reilley and Schmid6 have shown that, as the concentration of indicator becomes similar to that of the metal, serious extrapolation difficulties arise; such an effect might lead to errors well in excess of 2 per cent. 0‘490 I 0.460 .$ 0.440 C a, T I e U - .- ”a 0.420 0 6.30 6.32 6’34 6‘36 EDTA solution added, g 7.55 7‘57 7’59 EDTA solution added, g Fig. 2. Titrations of: (a) approximately 1 mg of cadmium; (b) approximately 0-6 mg of zinc REFERENCES 1, 2. 3. 4. 5. 6. Schwarzenbach, G., “Complexometric Titrations,” translated by H. Irving, Methuen & Co. Ltd., Sweetser, P. B., and Bricker, C. E., Anal. Chem., 1954, 26, 195. Hunter, J . A., and Miller, C. C., Analyst, 1956, 81, 79. Headridge, J . R., “Photometric Titrations,” Pergamon Press, Oxford, London, New York and Blaedel, W. J., and Knight, H. T., A~zal. Chem., 1954, 26, 741. Reilley, C. N., and Schmid, R. W., Ibid., 1959, 31, 887. London, 1957, p. 83. Paris, 1961. Received February 28th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800618

出版商:RSC    

年代:1963

数据来源: RSC