ISSN:0003-2654    

DOI:10.1039/AN95378FX055

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BX057

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BP153

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

DECEMBER, 1953 Vol. 78, No. 933 THE ANALYST PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS AN Ordinary Meeting of the Society was held at 7 p.m. on Wednesday, October 7tli, 1953, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. D. W. Kent-Jones, F.R.I.C., and about 170 members and visitors were present. The subject was introduced by Dr. G. Roche Lynch, O.B.E., M.B., B.S., D.P.H., F.C.G.I., L.M.S.S.A., F.R.I.C., and the following papers were presented and discussed: “The Preparation of 13iological Material for the Determination of Trace Metals. Part 11. A Method for the Destruction of Organic Matter in Biological Material,” by G. Middleton, B.Sc., F.R.I.C., and R. E. Stuckey, E.Sc., Ph.D., F.R.I.C., Ph.C.; “Determination of Lead in Foodstuffs,” by H. C. Lockwood, Ph.D., F.R.I.C. At this meeting there was a discussion on “Destruction of Organic Matter.” NEW MEMBERS Raymond Leslie Bass, R.Sc. (Lond.), A.R.I.C. ; David Richmond Brown, B.Sc. (Lond.), A.R.I.C. ; James Dennis Burton, B.Sc. (Lond.), A.R.I.C.; Denys Irvine Coomber, B.Sc., Ph.D. (Lond.), A.R.I.C. ; John Mervyn Davies, B.Sc. (Wales) ; Guillermo V. Freile Gagliardo, B.Biol.Sci. (Rocafuerte), D.Chem. (Guayaquil) ; Marian Norden Gibbons, BSc. (Lond.), A.R.I.C.; Cecil Alfred Johnson, B.Pharm., BSc. (Lond.), Ph.C., A.R.I.C.; John Edwin Morrison Moxley, B.A. (Oxon.) ; Stanley Newton, R.Sc. (Lond.) ; Peter Stross, B.Sc. (Lond.). DEATH \Ye regret to record the death of Williani C,harles Hughes

ISSN:0003-2654    

DOI:10.1039/AN9537800685

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

686 MILLER AND CHALMERS: MICRO-ANALYSIS OF SILICATE ROCKS [Vol. 78 Micro-analysis of Silicate Rocks Part IV. The Determination of Alumina Ru (XRTS‘I’TNA (1. MIT,LER AND ROBERT A. CHALMEKS (I”rP.sci.ztrd at the meeting of the Society ifz Glasgow oft Wednesduy, May 6th, 195.3) A new procedure is prescribed for the separation and direct determination of alumina in 5-mg samples of silicate rocks. Silica is volatilised by heating with hydrofluoric and sulphuric acids, and the residue is fused with potassium bisulphate and extracted with N hydrochloric acid. Iron, titanium, vanadium and zirconium are removed together by precipitation with cupferron and extraction with o-dichlorobenzene. Acetylacetone is added to the aqueous phase, and aluminium and beryllium acetylacetonates are extracted from the buffered solution a t a pH of 6 to 7 by means of diethyl ether.From the ether extract aluminium and beryllium are withdrawn into 6 N hydrochloric acid, and the aluminium alone is precipitated and weighed as aluminium 8-hydroxy- qui nolinate. The method has been applied in the presence of all the elements commonly found in silicate rocks. In the analysis of rocks, the results obtained are relatively 1 per cent. lower than those based on the classical procedure. IN the analysis of silicate rocks by the classical procedure, the component determined with the least certainty, and only after a considerable expenditure of time, is alumina. Its indirect determination depends on the weight of seldom fewer than five oxides (“mixed oxides”) and the separate determination of all except alumina, which is then found by difference.On the micro-scale the inherent difficulties and uncertainties of the method are increased. Slow-filtering hydroxide precipitates can be very troublesome and the hygroscopicity of alumina, and the many high-temperature ignitions, which significantly affect the weight of platinum crucibles, are sources of error. In microchemical work especially, one would like to abolish the hydroxides precipitation and also determine alumina more directly. Hitherto, a potent reason for retaining the precipitation of the hydroxides has been that their ability to carry down some residual silica has enabled the determination of silica to be completed. We have, however, recently proposed a method for the direct determination of total silica in a separate portion of r0ck.l Precipitation of the hydroxides is not essential for the sub- sequent determination of iron and titanium, which are commonly determined at this stage although they are easily determined separately, but it facilitates the determination of calcium and magnesium. Early attempts to avoid the precipitation of the hydroxides in micro-analyses of silicate rocks were made by Schoklitsch2 and by H e ~ h t , ~ both of whom freed the rocks from silica, fused the residues with potassium bisulphate, precipitated together the 8-hydroxyquinolinates of iron, aluminium and titanium, and from them obtained, after destruction of organic matter, a solution of metal chlorides.Schoklitsch then precipitated and removed by filtration the cupferron complexes of iron and titanium, destroyed organic matter in the filtrate and precipitated aluminium with 8-hydroxyquinoline.Hecht, on the other hand, precipitated iron as iron sulphide in an ammoniacal tartrate solution, thus separating it from aluminium and titanium, which were then precipitated together and weighed as the ,8-hydroxyquinoline complexes. After a somewhat complicated procedure, he then determined titanium as titanium dioxide and found alumina by difference. 8-Hydroxyquinoline can certainly be used to simplify the determination of alumina in silicate rocks of low complexity. McLennan4 jointly determined iron aluminium and titanium (TiO,, 2 per cent .) gravimetrically as their 8-hydroxyquinolinates with a positive error of less than 0.5 per cent. One to 2 per cent.of calcium oxide and magnesium oxide, 0.5 per cent. of phosphoric oxide and 0.2 per cent. of manganous oxide were without influence at pH 5, but larger amounts of manganous oxide divided, and large (10 per cent.) amounts of calcium and magnesium oxides caused significant positive errors. As various minor constituents would accompany aluminium, one may conclude that the substitution of 8- hydroxyquinoline for ammonia is not in itself the solution of the problem when rocks of greater complexity are under consideration,Dcc., 19531 PART IV. THE DETERMINATION OF ALUMINA 687 The object of our investigation was therefore to devise a method for the direct determina- tion of alumina in 5-mg samples of more complex silicate rocks, preferably one that would not prevent the subsequent determination of calcium and magnesium, and perhaps manganese, in the same sample.After volatilising silica by heating with hydrofluoric and sulpliuric acids, and fusing the residue with potassium bisulphate, we have quantitatively removed iron and titanium and some other elements from a N hydrochloric acid solution of the melt by means of cupierron and o-dichlorobenzene. Aluminium has then been converted into the acetylacetone complex at pH 6 to 7 and extracted quantitatively with diethyl ether, from which it has been with- drawn into 6 N hydrochloric acid and, after suitable adjustment of tlie conditions, precipitated arid weighed as aluminium 8-hydroxyquinolinate. The method has been applied to some silicate rocks.QUANTITATIVE SEPARATION OF IRON, TITANIUM AND SIMILAR ELEMENTS FROM ALUMINIUM Cupferron is frequently used in conjunction with an organic solvent for removing iro11I'~ and titanium1" from mineral acid solutions before determining small amounts of aluminium. When aluminium is a major component, care is required because aluminium cupferronate is extractable from solutions of low acidity. Hence the acid concentration should be reason- ably high and the concentration of cupferron and the volume of organic solvent should be the minimum that permits the complete removal of iron, titanium and similar elements. The efficiency of the common solvents, diethyl ether, carbon tetrachloride, chloroform and benzene for separating iron and aluminium in N hydrochloric acid solution was barely adequate, and superior results were obtained with o-dichlorobenzene.In order to facilitate separation of the liquid phases and to minimise washings, specially designed stoppered centrifuge tubes (Fig. lA), made water-repellent by treatment with a silicone solution, were used. The wide necks were required to prevent loss when the silicone-treated stoppers were rinsed into the tubes. As it was impossible on the micro-scale to judge when the formation of cupferronates was complete, and slight loss of aluminium occurred when more than 2 to 3 mg of cupferron in excess of theoretical requirements was used, it was necessary for maximum accuracy to know the approximate amounts of iron and titanium present, so that the quantity of cupferron added could be controlled.For maximum extraction of iron a shaking time of 2 minutes was required to overcome a slight holding effect of potassium bisulphate. Under tlie conditions prescribed below, about 2 parts per 1000 of ironIJ1 were left in the aqueous phase. As ironlI was only 85 per cent. extracted, and some is formed when hydrochloric acid solutions of bisulphate melts containing ironIII are held in platinum, it was necessary to limit the period of contact at 100" C to 5 minutes, or else to re-oxidise the solution before extracting. Titanium (TiO,, 250 pg), zirconium (ZrO,, 50 pg) and vanadium (V,O,, 25 pg) were completely transferred to the organic phase, whereas calcium (CaO, 500 pg), magnesium (MgO, 500 pg), manganese (MnO, 50 pg), beryllium and cerium (Be0 or Ce,O,, 25 pg) and nickel, cobalt, chromium and platinum (NiO, COO, Cr,O, or PtO,, 300 pg) were completely held in the aqueous layer.The presence of phosphate (P,O,, l50pg) had no significant effect on the results. The following method has been evolved. EXPERIMENTAL BY MEANS OF CUPFERRON- QUANTITATIVE SEPARATION OF ALUMINIUM FROM CALCIUM, MAGNESIUM AND SOME OTHER ELEMENTS BY MEANS OF ACETYLACETONE- In the analysis of potable waters, Stene5 used acetylacetone and extraction with carbon tetrachloride or chloroform to separate aluminium, iron, beryllium, cerium and copper from calcium, magnesium, manganese, titanium, cobalt, nickel, zinc, uranium and so on, which were not extracted at a pH of 4.5 to 7.5. Abrahamczik6 similarly separated iron, aluminium and manganese in weakly alkaline solution from magnesium and calcium. If, in silicate rock analyses, iron, titanium, vanadium and zirconium were initially removed by means of cup- ferron, then it seemed that aluminium might be separable from calcium and magnesium, which were the main elements to be considered at this stage, and also, by suitable control of pH, from manganese and various other subsidiary elements, excluding beryllium.Some experiments were first made on the precipitation of aluminium with other diketones and extraction with other solvents. A reagent superior to acetylacetone was not found,688 MILLER AND CHALMERS: MICRO-ANALYSIS OF SILICATE ROCKS [Vol. 78 but diethyl ether proved to be a more efficient solvent. Mechanical loss was again minimised by the use of appropriately designed extraction tubes, as shown in Fig. 1B.Water-repellent coatings were not permissible because of the solvent action of ether, and therefore the necks of the tubes could be narrower than before. The effect of variations in pH on the extraction of aluminium acetylacetonate, under conditions comparable to those prescribed on p. 691, was examined for 1-mg amounts of alumina in the presence of 100 mg of potassium bisulphate. After the complex had been given time to form and the pH had been measured with a pH meter, the three extractions with ether were made and the residual aqueous layer was tested by means of aurine- tricarboxylic acid. The results, which are shown in Fig. 2, indicate that extraction is essentially complete over the pH range 6 to 7.U J W B C I I I ( 1 ) Centimetre scale Fig. 1. Extraction tubes The behaviour of several other elements in the procedure was investigated at a pH of 6 to 7. Calcium and magnesium (CaO or MgO, 500 pg), manganese (MnO, 50 pg), nickel and platinum (NiO or PtO,, 250 pg) and cerium (Ce,O,, 25 pg) were not significantly removed by acetylacetone and ether. About 3 per cent. of cobalt and chromium (COO or Cr,O,, 300 pg) were extracted by ether. These elements are minor components of rocks, and would not be expected to interfere. Beryllium accompanied aluminium. The presence of phosphate (P,O,, 150 pg) caused no obvious precipitation with any of the above elements except cerium and chromium. No loss of aluminium by co-precipitation with phosphate precipitates need be feared with the small amounts of these elements likely to be present in silicate rocks.FINAL DETERMINATION OF ALUMINIUM AS ALUMINIUM 8-HYDROXYQUINOLINATE- Attempts made to determine aluminium (in absence of beryllium) as aluminium acetyl- acetonate simply by evaporating the ethereal extract containing it and then heating the residue at 120" C were not a success, apparently because slow volatilisation of the metal complex occurred. The diketonate had to be broken down by treating the ethereal extracts with acid and restoring the aluminium to the aqueous phase for determination. If it were precipitated with 8-hydroxyquinoline in a solution containing acetic acid, no interferenceDec., 19531 PART IV. THE DETERMINATION OF ALUMINA 689 from beryllium would result. A reagent superior to 8-hydroxyquinoline for the determina- tion of aluminium was not found.Quantitative disruption of aluminium acetylacetonate was effected by shaking the ethereal solution vigorously for 6 minutes with 6 N hydrochloric acid in extraction tubes of the special type shown in Fig. 1C. Thereafter it was immaterial whether the ether was removed by evaporation in a current of air or by direct withdrawal after separation of %he phases by centrifugation. Much has been published on the determination of aluminium by means of 8-hydroxy- quinoline. Its determination in faintly acid solution was required, so that, if beryllium were present, or small amounts of calcium and magnesium inadvertently accompanied aluminium, Fig. 2. Effect of pH value on the extraction of aluminium acetylacetonate with ether they would be prevented from interfering. Got6' gave the range of complete precipitation of aluminium 8-hydroxyquinolinate as pH 4-15 to 9.80.Barrels showed that precipitation began at pH 3-85 and was quantitative at pH 4.7. The Chemical Analysis Sub-committee of the British Ceramic Research Association9 obtained consistent results in the determination of aluminium over the pH range 4-80 to 5-14, and results distinctly lower below pH 4.48. As positive errors were prevalent in our early work, it was necessary to examine the conditions of the precipitation somewhat fully. It is customary to add an excess of an acetic acid solution of 8-hydroxyquinoline to the slightly acid solution containing aluminium and to effect precipitation in the heated solution by adding an excess of ammonium acetate to raise the pH appropriately.In experiments made with 1.3 to 57 mg of alumina in a final volume of 120 ml, the pH after precipitation could be varied from 4-4 to 6.7 without significant effect, and it mattered not whether the solution was filtered hot or cold. Too great a rate of addition of ammonium acetate solution caused significant positive errors, the percentage errors for the same amount of added 8-hydroxyquinoline being greater for smaller amounts of alumina. When ammonium acetate solution was added very slowly, dropwise with constant stirring, the precision of the results was greatly improved, as has also been shown by Stumpf.lo The main feature of the results was a roughly linear relationship, as shown in Fig.3, between the weight of the excess of 8-hydroxyquinoline used and the percentage error in the weight of precipitate, from which it was deduced that a 250-mg excess was required. On the micro-scale, one-thirtieth of all quantities was required and an excess of 8 mg of 8-hydroxyquinoline was indicated. Despite the fact that the ammonium acetate solution was added extremely slowly from a fine-tipped horizontal micro-burettell while the solution was vigorously stirred by a magnetic stirrer, an average error of +0.2 per cent. was incurred690 MILLER AND CHALMERS: MICRO-ANALYSIS OF SILICATE ROCKS [Vol. 78 in determining alumina. Reduction of the excess of 8-hydroxyquinoline to 6 mg eliminated the error, under the conditions of precipitation described on p.691. The average pH of the solution after precipitation was about 4-7. The presence of a little acetylacetone did not affect the results. APPARATUS- METHOD Use Pyrex or similar glassware throughout. Reagents should as far as possible be of recognised analytical quality. o-Dichlorobenzene-Distil technical grade o-dichlorobenzene before use and collect the DiethyZ ether-Free the ether from peroxide and distil. REAGENTS- fraction that boils between 177" and 178°C. Excess of 8-hydroxyquinoline, mg Fig. 3. Relationship between weight excess of S-hydroxy- quinoline and percentage error Acetylacetone solution-Distil acetylacetone and collect the fraction boiling between 136" Cu+ fevron solution-Prepare freshly a cold, 5 per cent.w/v aqueous solution. and 138" C. Mix 10 ml with 20 ml of ethanol and 70 ml of water. 8-Hydroxyquinoliize solution-Prepare a 2.5 per cent. w/v solution in 4 per cent. v/v acetic acid. Ammonium hydroxide, 6 N-Prepare from cylinder ammonia and store in a polythene bottle. HydvochZovic acid, 6 N-Prepare by distillation from the concentrated acid. PROCEDURE FOR ROCKS- With the aid of a stoppered weighing stick, weigh into a l-ml platinum crucible about 5 mg of the dried material. Add 0.15 ml of 40 per cent. hydrofluoric acid and then, 5 minutes later, 0.15 ml of 10 N sulphuric acid. Evaporate to dryness on a steam-bath and cautiously expel sulphuric acid over a micro-bunsen flame. Next fuse the residue with 100mg of potassium bisulphate, cool, place the crucible in a small porcelain capsule, and dissolve the melt in 1 ml of N hydrochloric acid.Heat on a steam-bath for not moYe than 5 minutes and transfer the solution, by means of a fine capillary attached to a suction apparatus, to a silicone- coated centrifuge tube of the type shown in Fig. 1A. Wash both capsule and crucible four times with 0.25-ml portions of N hydrochloric acid and finally rinse the end of the capillary with a drop of acid. Centrifuge the contents of the tube and proceed with the cupferron treatment. Add to the solution 0.05 ml more cupferron solution than is theoretically required for removal of iron, titanium, vanadium and zirconium (0.1 ml = 0-8 mg of Fe,O,) and 1 ml of o-dichlorobenzene. Shake the stoppered tube for 2 minutes at the rate of four shakes per second, then unstopper it, centrifuge and transfer the aqueous layer via a micro-filter and capillary to the appropriate centrifuge tube, Fig.1B. The filter removes a slight opalescence that sometimes appears in the aqueous phase. Rinse the stopper into the extrac- tion tube with 0 6 m l of N hydrochloric acid, insert it, and shake the tube for 15 secondsDec., 19531 PART IV. THE DETERMINATION OF ALUMISA 681 before centrifuging and removing the upper layer as before. Rinse the stopper again with 0.5 ml of acid, but do not shake the tube. Finally wash the walls with 0.5 ml of acid, centrifuge and remove the acid. Evaporate the solution containing the aluminium in a current of air, while heating the centrifuge tube in a bath of boiling water, until the volume is about 0.5 ml.Shake the tube to dissolve separated solids and add 0-5 ml of acetylacetone solution, followed by 0.5 ml of 5 N ammonium acetate solution and bromocresol purple as indicator. Mix, and then add 6 N ammonium hydroxide until the indicator shows blue-green (in daylight), or grey-green if some aluminium acetylacetonate precipitates. Rinse the neck of the tube with 0.1 ml of water, allow 5 minutes for complex formation and then readjust the colour of the indicator, if necessary. Add 3 ml of ether, moisten the stopper of the tube with a drop of water and insert it. Shake the tube for 3 minutes as before, unstopper it, centrifuge and transfer the upper layer by slow suction via a dry capillary to the third centrifuge tube of the type shown in Fig.1C. Rinse the stopper into the extraction tube with 3 ml of ether and proceed as before. Repeat with 1 ml of ether and shaking for 30 seconds and then with 0.25 ml and no shaking. To the combined ether extracts add 1 ml of 6 N hydrochloric acid and moisten the stopper of the tube with a drop of acid. Shake the stoppered tube as usual for 6 minutes and then remove and rinse the stopper with a little water into a 6-ml beaker. Rinse down the neck of the extraction tube with 6 drops of 6 N hydrochloric acid, centrifuge and discard the ether layer. Place the tube in a bath of cold water and gradually heat to boiling, while a slow current of air plays on the surface of the acid and removes residual ether. Centrifuge and transfer the contents by capillary to the 6-ml beaker.Wash the tube three times with 0-75-ml portions of water, always centrifuging briefly before transferring the liquid to the beaker. Evaporate the solution in the beaker to about 1 ml, and adjust the pH to about 3 by means of 6 N ammonium hydroxide and dilute hydrochloric acid, using bromophenol blue as indicator. Add 0.25 ml more of the 8-hydroxyquinoline solution than is theoretically required for aluminium (0.5 ml fi 1.5 mg of A1203) and, if a precipitate forms, just redissolve it with a minimum of 0.2 N hydrochloric acid. Dilute the solution to about 4 ml., heat to 80" to 90" C and, while vigorously stirring it by means of a magnetic stirrer, add 1 ml of 2 Ar ammonium acetate solution very slowly from a horizontal burette. Next digest the precipitate on a steam-bath for 5 minutes, and then filter it hot on a sintered-glass filter stick of No.4 porosity. Wash the beaker, filter, stirrer and precipitate with six I-ml portions of hot water. A similar beaker brought up to weight by the addition of pieces of Pyrex-glass rod should be available as a tare. Wipe both beakers with damp flannelette and then with chamois leathers that have been kept in a desiccator containing saturated calcium nitrate solution. Heat both in an oven at 160" C for 1 hour, cool in aluminium blocks, using no desiccant, and weigh. Repeat the heating until the weight is constant; then dissolve the precipitate in chloroform, withdraw the solution, and dry and reweigh the beaker, filter stick and stirrer. The precipitate is Al(C,H,ON),; multiply the weight of the precipitate by 11-10 and divide by the weight of the sample to get the percentage of alumina (A1203).Centrifuge and remove the acid. RE su LTS In all the results quoted, correction has been made, where necessary, for aluminium in the reagents. ALUMINIUM AS ALUMINIUM 8-HYDROXYQUINOLINATE- A standard solution of aluminium chloride in N hydrochloric acid was prepared from Hilger and Watts's "H.H.P." aluminium (99.98 per cent.) by dissolving it in hydrochloric acid. Alumina was determined in weighed portions by the procedure given for silicate rocks, except that in half of the experiments an 8-mg excess of 8-hydroxyquinoline was added instead of 6 mg. The smaller amount gives better results, as shown- Al,O, taken, pg . . . . '494 534 526 1170 1216 1919 1987 Error, pg .. . . .. 3.1 0 +1-5 1-4.5 + 6 +4 +4 8-wzg Excess of 8-hydyoxyquinoline- 6-mg Excess of 8-hydvoxyquinoline- A1,0, taken, pg . . .. 649 533 1041 1400 1878 2059 Error, pg . . . . . . $1.5 0 -1 0 $ 2 0692 MILLER AND CHALMERS : MICRO-ANALYSIS O F SILICATE ROCKS [Vol. 78 I t is evident that, for accuracy to 2 or 3 parts per 1000 in the determination of alumina, the amount of 8-hydroxyquinoline added must be under control to the nearest 2 mg, that is, the alumina content of a 5-mg sample must be known to the nearest 4 per cent. ALUMINA IN THE AQUEOUS PHASE FROM THE CUPFERRON - DICHLOROBENZENE TREATMENT- Weighed aliquots of the standard solution of aluminium chloride, to which were added 100mg of potassium bisulphate, were submitted to treatment with 0.05ml of 5 per cent. aqueous cupferron solution and dichlorobenzene, as in the analysis of rocks, and the alumina content of the aqueous phase was determined by precipitating aluminium 8-hydroxy- quinolinate.The results were as follows- A1,0, taken, pg . . 546 530 512 551 1916 1897 1899 2139 Error, pg . . . . -2 -1 +1 + 3 +2 +5 +4 -3 ALUMINA EXTRACTED BY ACETYLACETONE AND DIETHYL ETHER- Weighed aliquots of the standard solution of aluminium chloride, to which were added 100 mg of potassium bisulphate, were suitably prepared and treated with acetylacetone and diethyl ether. The ether extracts were shaken with 6 N hydrochloric acid and the aluminium was determined with 8-hydroxyquinoline, all as described for rock analysis. The following results were obtained- A1,0, taken, pg .. .. 525 579 1169 1154 1888 1937 Error, pg . . . . . . - 0.5 - 0.5 +I - 4.5 0 +2 ALUMINA IN PRESENCE OF OTHER ELEMENTS- As satisfactory results were obtained in the three preceding steps, experiments were next made on the determination of various amounts of alumina in synthetic mixtures con- taining selected combinations of the elements that might be found in silicate rocks. All the test solutions contained 100 mg of potassium bisulphate and the procedure was that prescribed for rocks, except that the initial treatment with hydrofluoric acid and the fusion were omitted. The results given below indicate that in a 5-mg sample of rock it should be possible to determine 0 to 40 per cent. of alumina, with a reasonable degree of accuracy, in the presence of a con- siderable amount of all the usual components and small amounts of several others.(a) Other elements preseTzt: CaO, 500 pg; MgO, 500 p g ; P,O,, 150 p g ; MnO, 50 p g ; BeO, 25 p g ; Ce203, 25 pg- A1,0, taken, pg .. . . .. 502 542 1969 1893 Other elements present: Fez03, 1000 p g ; TiO,, 250 p g ; P,O,, 150 p g ; ZrO,, 50 p g ; V,O,, 25 pg- Error, pg . . .. .. .. 4- 2.5 +l - 0.5 - 3.5 Al,03 taken, pg . . .. .. 548 635 1953 1910 Error, pg . . .. .. .. + 0.5 + 4* +2 - 7" (b) * These results are corrected for 1 pg of Fe,O, in the precipitate. ALUMINA I N SILICATE ROCKS-- Finally, a number of rock samples were submitted to analysis through the whole procedure. As the method is primarily intended for use in the analysis of more complex materials, where the number and the amount of other components of the "mixed oxides" are large, and the standard samples available were lacking in this respect, it was necessary to include two synthetic samples in the series of rocks analysed.They were prepared by adding accurately weighed portions of U.S. Bureau of Standards Feldspar No. 99 (dried) to the ignited residues from the evaporation of appropriate amounts of solutions containing the elements required. The feldspars used were as issued by the U.S. Bureau of Standards. The flint clay was from the part that had passed through a 300-mesh sieve. As the percentage of silica found in this part agreed excellently with the certificate value, it was assumed that comparison with the certificate value would be valid for alumina.The olivine-basalt was part of the original sample used by Guthrie and Miller12 and was less finely divided. All the samples, save the flint clay, which was dried at 140" C, were dried at 105" to 110" C before use. In general, the results are about 1 per cent. lower than those shown in the last column. An attempt was made to obtain figures for comparison by other than the classical procedure. :Is the feldspars contained insignificant amounts of iron, titanium, manganese, calcium and magnesium, it was practicable to omit the use of cupferron and acetylacetone, precipitate The results of the analyses are shown in Table I.Dec., 19631 PART IV. THE DETERMIXATION OF ALUMIK\‘A 693 together the 8-hydroxyquinolinates of aluminium, iron and titanium, and then correct for the iron and titanium content of the precipitates on the assumption that Fe(C,H,ON), and TABLE I DETERMINATION OF ALUMINA I N SILICATE ROCKS A129 by classical Approxi- method matc Amounts of certain components A120,by or weight A -- micro- calcula- I Silicate taken, SiO,, Fe203, TiO,, P,O,, CaO, MgO, PvlnO, method, tion, m g % % % 70 % % O / /O % % Feldspar No. 70” .. 4.0 67 0.03 <0.01 0.01 0.07 0.01 <0.01 17-86 18.031. Feldspar No. 99* . . 4.8 69 0.07 0-02 0.14 0.36 0.05 (0.01 18.94 19.067 Flint clay No. 97* . . 4.6 43 0.98 2-38 0.08 0.10 0.26 <0.0111 38.23 38.77t 6 3 17.71 6-3 18.77 4-5 38.43 Synthetic rock A . . “4.7” 40 15-0 5.0 3.0 10.0 6.0 - 11.13 11.135 “4.5” 10.59 10.75s Synthetic rock B . . “4.9” 40 10.0 3-0 1-0 6.0 6.0 - 25.43 25.69s “4.7” 25.31 25.559 Olivine-basalt .. 4.2 44 13.0 2.4 0.5 10.0 13.0 0.2 12.59 12-59: 3.7 12-68 * U S . Bureau of Standards sample. t U.S. Bureau of Standards certificate value. Guthrie and Miller’s figure for semi-micro analyses.12 5 Calculated from the weight of Feldspar No. 99 (A1,0, = 19.06 per cent.) in A, and the weight of 11 Also ZrO,, 0.23; Cr,O,, 0.08; and V20,, 0.04 per cent. feldspar (“-2.8 mg) plus additional A1,0, in B. TiO(C,H,OX), were present (cf. McLennan4). the results obtained were- In experiments with 70 to 90-mg samples, Feldspar No. 70-(17.60), 17-78, 17-79 and 17-81 per cent. Feldspar No. 99-18-82 and 18-85 per cent. which seem to suggest that the figures obtained by the classical procedure tend to be rather high. It is of interest that Guthrie and Miller,12 working with the same sample of Feldspar No.70, found 17.94 per cent. of alumina by a semi-micro classical procedure. The method described above was inapplicable to the flint clay on account of the complexity of the “mixed oxides. ” If these figures for the feldspars are accepted as being correct, and the composition of the synthetic samples is recalculated accordingly, then the errors for the four materials, calculated as weight of alumina, are as follows- Silicate Feldspar No. 70 Feldspar No. 99 Synthetic “A” Synthetic “R” (9 (ii) (i) (ii) (9 (ii) 0) (ii) A1203 calculated, pg . . 714 1116 908 1196 476 515.6 1185 1262.5 A1,0, found, pg . . 717 1111 913 1191 474 521.0 1180 1256.0 Error, pg . . .. + 3 -5 +5 -5 - 2 3-5.5 -5 - 6.6 bringing them into line with those found for the determination of alumina in complex mixtures.Should, however, the certificate figures be correct and a negative error of 1 per cent. be genuine, it is still possible to say that the errors in terms of the rock total do not greatly exceed the limits normally allowed for duplicate analyses of gram samples by the same analyst .13 Moreover, they are considerably less than those evident in the recent co-operative investigation of precision and accuracy in the chemical analysis of silicate and rather less than those shown in a later co-operative analysis of a synthetic glass of known composition.16 In the new micro-method, it is almost certain that the results are for alumina only and, as far as we can judge without actual experimentation, it should be possible to determine calcium and magnesium by a standard procedure in the aqueous extract left after the removal of aluminium by means of acetylacetone and ether.We gratefully acknowledge a maintenance grant to one of us (R. A. C.) from the Depart- ment of Scientific and Industrial Research, and grants from Imperial Chemical Industries Limited and the Trustees of the Ritchie Bequest. We are indebted to Dr. M. Borrel of Lyons University for a copy of his thesis.694 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. MILLER A S D CHALMEKS : MICRO-AKALYSIS OF SILICATE ROCKS REFERENCES [Vol. 78 Miller, C. C., and Chalmers, K. A., Analysf, 1953, 78, 24. Schoklitsch, K., Mikrochemie, 1936, 20, 247. Hecht, F., Mikrochim. Acta, 1937, 2, 188.McLennan, I. C. , Thesis, Edinburgh University, 1940. Stene, S., Chem. Zentralbl., 1930, 110 (l), 3433. Abrahamczik, E., Mikrochemie, 1948, 33, 209. Got6, H., Sci. Rep. Tbhoku Imp. Univ., 1937, 26, 391. Borrel, M., Thesis, Universit6 de Lyons, 1952. Chemical Analysis Sub-committee of the British Ceramic Research Association, Tmns. Bvit. Stumpf, K. E., 2. anal. Chem., 1953, 138, 30. Lacourt, A,, Metallurgia, 1948, 38, 355. Guthrie, W. C. A,, and Miller, C. C., Min. Mag. Lo?& 1933, 23, 405. Groves, A. W., “Silicate Analysis,” Second Edition, George -411en and Unwin Ltd., London, 1951. Schlecht. W. G., Anal. Chenz., 1951, 23, 1568. Fairbairn, H. W., and Schairer, J . F., Amer. Min., 1952, 37, 744. Ceram. Soc., 1952, 51, 438. p. 228. XOTE-Reference 1 is to part I1 of this series.CHEMISTRY DEPARTMENT THE UNIVERSITY, EDINBURGH, 9 THE UNIVERSITY SCIENCE LABORATORIES SOUTH ROAD, nURHA4M June 3rd, 1953 DISCUSSIOX MR. R. C. CHIRNSIDE said that he had followed with interest this paper by Dr. Miller and Mr. Chalmers, as indeed he had all of lh-. Miller’s papers on the subject. Those concerned with silicate and rock analysis would know that the techniques had changed very little since the time of Berzelius, and attempts to determine some of the constituents directly, as instanced in this paper, were therefore very welcome. Some work along the same lines had been done by the Analytical Committee of the British Ceramic Research Association, and he would like to ask Dr. Miller if she was acquainted with this work, which h a d been published last year in the Jozw??nl of the C r v a ~ i i c Society.In this connection, he wished to mention that, with certain refractories where free alumina in the form of corundum was known to occur, i t had been found impossible to get complete decomposition of the sample except by using a flux of sodium carbonate and borax. He wondered whether some of the low results to which Dr. Miller had referred could have arisen from incomplete decomposition of the sample. He also wanted to mention that during the co-operative work carried out by the B.C.R.A. Committee, the cause of some apparently low results for alumina was traced to dilution of the sample with silica removed from agate mortars during the grinding of the sample. MR. CHALMERS replied that they had read the account of the work done by the Analytical Committee of the British Ceramic Research Association, and had considered the use of sodium carbonate for the fusion of the residue left after the removal of silica by means of hydrofluoric acid, but had rejected it, partly because of the risk of loss by spattering in the subsequent dissolution in acid and partly because the bisulphate fusion was easier and seemed quite satisfactory. It was possible that the presence of free alumina might account for the apparent negative error, but it was not likely.There seemed no reason why other fluxes should not be used instead of potassium bisulphate. MR. A. F. WILLIAMS mentioned that sodium peroxide was in universal use for breaking down ores. The technique was devised by Rafter, of New Zealand, and could be carried out in a platinum crucible a t a temperature of about 400” C. The method was rapid and there was no difficulty caused by spraying. There might be difficulty in getting the sodium peroxide frec from alumina, but suitable brands should be available. MR. CHALMERS said that they were acquainted with Rafter’s paper and thought that the considerations given in the previous reply would apply here also. DR. MILLER added, in a written reply to Mr. Chirnside’s comments, that the methods described by the British Ceramic Research Association did not fully lend themselves to their purpose, which was to determine alumina in association with all the elements, including some that were commonly neglected, that might be present in igneous silicate roclts, and afterwards to determine calcium and magnesium in the same sample. With reference to the apparent negative errors recorded for the rocks, some chemical procedures tended to give biased results, and alumina determinations, as normally made, gave results that were more often too high than too low. It might be of significance that, in the co-operative analysis of the synthetic glass of known composition, the average percentage of alumina found was 16.19 instead of 15.78- a difference of 0.4. All their results for the standard samples were lower than the certificate values, the average deviation being 0.3. The standard samples were not further ground and could not therefore be diluted with silica.

ISSN:0003-2654    

DOI:10.1039/AN9537800686

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Dec., 19531 COOKSON, COPPOCK AXD SCHNURMANN 695 Ultra-violet Spectrophotometric Estimation of the Quality of Mineral Oils Extracted from Bread BY M. A. COOKSON, J. B. M. COPPOCK AND R. SCHNURMANN (Presented at the meeting of the Society on Weducesday, May 20th, 1953) A method has been devised for determining in bread the degree of refining of a mineral oil that has been absorbed by the dough during bread-making. The method consists basically in sulphating, under controlled conditions, the unsaponifiable fraction of the total oils extracted from the bread, so that the natural saponifiable matter is destroyed without completely removing the unsaturated hydrocarbon constituents inherent in mineral oils refined to different degrees. The recovered mineral oil is then examined by ultra- violet absorption spectrophotometry to determine its quality.A spectro- photometric criterion, based on the absorption intefisity of the sulphated oils at 2600 A, is suggested for the quality of mineral oils that can be regarded as satisfactory for the lubrication of plant used in bread production. Some properties of the naturally occurring unsaponifiable oils of bread are also described. IN recent times bread-making has become a highly mechanised process. The nature of bread dough is such that lubrication is necessary at various stages of production to prevent adherence of the dough to metallic equipment. This is particularly so at the dough divider, where the mass of dough is separated into unit pieces, and also when the dough pieces are tinned before baking. White oils are amongst the lubricants that have been used for these purposes. Since mineral oil may also be absorbed by doughs as a result of accidental con- tamination from faulty machinery, it is important to maintain a reasonable standard of quality for the oil.Mineral oils absorbed by doughs can be extracted by a solvent from the finished bread, but the naturally occurring flour-oil and other edible fats and oils added as bread ingredients are extracted at the same time. It was recommended1 in 1949 that mineral oils should be permitted as lubricants and greasing-aids in bread-making provided that (a) the quantity absorbed did not exceed 0-2 per cent. of the weight of the bread (this quantity is now the legally permitted limit2) and (h) the degree of refining of the oil was not less than that of liquid paraffin, B.P.A method for determining the quantity of mineral oil in bread has been publi~hed,~ and it was seen to be desirable to form some estimate of the original quality of the oil, i.e., its quality before it was absorbed by the dough. The development of a suitable method was complicated by (i) the small amounts of oil available for analysis, unless very large quantities of bread were laboriously extracted, (ii) the miscibility of mineral oils with the unsaponifiable fraction of bread-oil (i.e., the natural oils in bread), which makes separation difficult, and (iii) the lack of suitable methods for deter- mining mineral oil quality. The sulphuric acid test of the British Pharmacopoeia for aromatic hydrocarbons in medicinal liquid paraffin is only suitable for fairly large amounts of white oils, and the con- ventional interpretation of its results has been criticised.4 Ultra-violet absorption spectro- photometry, however, appears to be a sufficiently discriminating technique for determining the various ranges of quality of mineral oils after separating the unsaponifiable matter from bread, and it was the method chosen for this work.EXPERIMENTAL Five mineral oils (see Table I) of different and known degrees of refining were used in most of the experimental work. For spectrophotometric examination, all the specimens were dissolved in purified zso- octane and the absorption characteristics of the solutions studied in the wavelength region 2100 to 4500 A (particularly 2400 to 3400 A).Most of the spectra were recorded by means698 COOKSON, COPPOCK AND SCHNURMANN : ESTIMATION OF THE QUALITY [VOl. 78 of a medium Hilger quartz spectrograph, fitted with a Speklter photometer, at the Physics Department of the Manchester Oil Refinery Ltd. The materials examined were prepared at the British Baking Industries Research Association, the experimental procedure being as follows. Bread doughs, to which small quantities (0.1 to 0.5 per cent.) of the mineral oils listed in Table I were added during the dough-making, were baked, care being taken that they were not otherwise contaminated with extraneous mineral oil from tins or machinery. About 18 hours after baking, the cooled bread was crumbled and digested with carbon tetrachloride for 24 hours at room TABLE I THE CHARACTERISTICS OF THE MINERAL OILS USED IN THE EXPERIMENTS Oil Liquid paraffin, B.P.. . Technical white oil . . .. Transformer oil . . .. An acid and clay treated Edeleanu extract of a lubricating-oil distillation fraction Spindle oil* .. .. E:Zm values at 2600 A r A -l After treatment with 96 per cent. w/w Appearance As refined H,SO, at 100" C Colourless (white oil) 0.030 0.051 Colourless (white oil) 0.270 0.435 Fluorescent, straw colour 37.0 17-4 Fluorescent, brown colour 49.0 14-2 Fluorescent, ruby colour 220.0 21.5 * This term is not connected with the spindle dough moulder; it is derived from spindling in the manufacture of textiles. temperature to extract the combined mineral and natural oils. These mixed oils were saponified and the unsaponifiable matter was extracted by standard procedure^.^ As the oil from bread to .which no mineral oil has been added contains about 10 per cent.of unsaponifiable matter, further separation is required to isolate the mineral oil as free as possible from other unsaponifiable material. Two methods were tried for this purpose : (a) percolation of a light petroleum solution of the unsaponifiable matter through a column of activated alumina and ( b ) heating the unsaponifiable matter with sulphuric acid to remove sterols, tocopherols and hydrocarbon compounds other than those constituting the mineral oil addition, as described below. Both methods suffer the disadvantage that some of the constituents of the less refined mineral oils are removed and part of the natural unsaponifiable oil is retained in the end product.However, the working conditions for method (b) have been so adjusted that a sufficiently satisfactory separation can be effected. The separation of the oil-About 0.5 g of the unsaponifiable matter is dissolved in ether and transferred to a Babcock milk-test bottle. After removal of the solvent, 5 ml of carefully standardised 85 per cent. w/w sulphuric acid are added and the bottle is vigorously shaken for 30 minutes in a water-bath at 50" C. After cooling, sufficient 85 per cent. sulphuric acid is added to bring the liquid level into the graduated neck of the bottle, which is then centrifuged until the oil layer separates. The oil is then removed with a fine pipette for spect ropho t ome tric examinat ion.RE s u LTS Absorption values for the five oils studied are shown in the third column of Table I. I t was necessary to ensure that decomposition, particularly of liquid paraffin and technical white oil, did not occur under baking conditions, thereby making some adventitious contribu- tion to the spectra of the total oils extracted from bread containing mineral oil. This was done by heating liquid paraffin and technical white oil at baking oven temperatures (about 250" C; apart fxom the crust, bread itself during baking rarely exceeds a temperature of 100" C) for baking times (about 30 minutes). I t was found that even in an atmosphere of air, not of carbon dioxide and steam as in bread baking, no marked change in absorption spectra for these two oils was observed.MINERAL OILS-Dec., 19531 OF MINERAL OILS EXTRACTED FROM BREAD 697 As previously stated, about 10 per cent. of the oil that can be extracted froni bread, made from flour, yeast, salt and water only, is unsaponifiable. Most, if not all, of the bread-oil is derived from the flour, modified during fermentation and baking, and possibly containing traces of esters and other organic compounds produced in bread-making. It is well known that flour is a variable material, depending on the wheats from which it is made (k, the grist), the condition of the wheats when milled, the extraction rate of the flour and the age of the flour when used. Moreover, the method of fermentation and baking produces marked differences in the nature of the resultant bread.I t is not surprising, therefore, that the unsaponifiable fraction of bread-oil has been found to bc of variable composition. In Fig. 1 the two unbroken lines show the extremes of the range of absorption spectra observed over a period of 4 years from unsaponifiable oils prepared from breads produced NATURAL UNSAPONIFIABLE OILS-- Wavelength, A Fig. 1. Absorption spectra of (i) the unsaponifiable oils of breads made from flours of different extraction rate; (ii) the unsaponifiable oils from bread containing 0.26 per cent. of liquid paraffin, and of the chromatographed extract of this oil; and (iii) liquid paraffin-the basis for comparison. - Unsaponifiable oils from National breads (lower flour extraction rates, 80 to 85 per cent.). _ _ - - Unsaponifiable oil from National bread containing 0.26 per cent.of liquid paraffin. - . . - . . Chromatographed extract of the unsaponifiable oil from National bread containing 0-26 per cent. of liquid paraffin. . . . . . . Liquid paraffin - Unsaponifiable oil from a wholemeal bread (highest flour extraction rate). from National flour, bleached and treated, of 81 or 85 per cent. extraction, by the same formula and baking technique. Fig. 1 also shows the spectrum of the unsaponifiable oil from a wholemeal bread; the effect of the higher extraction rate of the flour is apparent from the higher absorption intensity of the oil from this flour. A comparison of these absorption characteristics with those in Table I for the five mineral oils studied reveals that the light absorption in the region of the ultra-violet explored is much698 COOKSON, COPPOCK AND SCHNURMANN: ESTIMATION OF THE QUALITY [Vol.78 greater for these unsaponifiable bread-oils than for white mineral oils, and that these naturally occurring unsaponifiable oils have absorption intensities similar to those of some mineral oils that have not undergone the same degree of refining as white oils. As compared with liquid paraffin, B.P., these less refined oils show 100 to 10,000 times as much absorption. The variations in the intensity of absorption of unsaponifiable oil from different batches of flour might be attributed to the presence of various amounts of a highly absorbing component in the unsaponifiable part of the natural oil. Fractionation of the natural unsaponifiable oil on a column of activated alumina has indicated the existence of several components, some of which, separated with light petroleum - benzene mixtures as eluants, absorb much more strongly than those shown in Fig.1 , e.g., at 2600 A one fraction shows a maximum Ei2m value of 480, and a second fraction crosses this peak and rises to an E:Tm value of 750 at 2400 A. It is hoped that further details about this particular phenomenon will be published later. UNSAPONIFIABLE OILS CONTAINING MINERAL OILS- The maximum permitted quantity2 of mineral oil that may be present in a food as a result of its use as a processing aid is 0.2 per cent. by weight of the food, but the desirable quality of the mineral oil is not legally defined. The method of oil extraction described in the experimental section, when applied to a bread containing no mineral oil, yields approximately 0.4 per cent.of oil, expressed on the bread weight, and approximately 0.53 per cent. when a bread containing 0.2 per cent. of mineral oil is extracted. After saponification of these oils the remaining unsaponifiablc oils are approximately 0-04 and 0.11 per cent., respectively, expressed on the bread weight, i.e., about two-thirds of the latter is attributable to mineral oil. I t is, therefore, in the light of the various spectra discussed, and having regard to the high intensity of the bread’s natural unsaponifiable oil, clearly not to be expected that the quality of the mineral oil originally used as a lubricant can always be ascertained by merely examining the spectrum of the total unsaponifiable oil; the interference would be too great.The spectra in Fig. 1 of liquid paraffin and of the unsaponifiable oils from National breads with and without liquid paraffin, i.e., the first, third and fifth lines from the abscissa, illustrate this point. However, when the mineral oil is of rather low quality, a detectable increase in absorption intensity of the combined unsaponifiable matter might be expected. This has been found to occur in practice, provided there is present about 0.2 per cent. of oil of a quality approximately equivalent to an acid and clay-treated Edeleanu extract of a lubricating oil distillation fraction, but if less than 0.1 per cent. is present, it is very difficult to detect. A further factor to be taken into consideration is that the proportion of natural to mineral oil in the total unsaponifiable oil will not be constant, so that, in the absence of any characteristic band structure connected with the mineral oil in the combined oil, quantitative measurements would be difficult to apply in estimating the quality of the mineral oil.It is seen, therefore, that because of the relatively large absorption and heterogeneity of the naturally occurring unsaponifiable matter, and also the variation in the proportion of mineral oils in the unsaponifiable oil, a separation of the mineral oil fraction of the total unsaponifiable oil must be achieved in order to appraise the quality of the rnineral oil in the mixture. THE ATTEMPTED SEPARATION OF MINERAL OIL FROM UNSAPONIFIABLE MATTER BY CHROMATO- GlL4I’HIC METHODS- The use of adsorption columns of activated alumina for the separation of hydrocarbons has been known for some time.This technique lias been suggested as a way of treating the unsaponifiable fraction in a general method for determining mineral oil in saponifiahle oils,5 and. particularly for determining mineral oil in bread.6 However, it was reported in our paper6 that a small fraction of the natural unsaponifiable matter was eluted immediately without being adsorbed on the separation column. If mineral oil is also present in the unsaponifiable material, it is possible that not only is the same natural non-adsorbed fraction (probably hydrocarbon in nature) eluted, but that further fractions of the natural oil may be eluted at the same time by a solvent action of the mineral oil.I t has been feud that the eluates containing mineral oil obtained by this method are still sufficiently variable and intense in absorption to make difficult the discernment of the quality of the mineral oil. Although the absorption intensities of these chromatographedDec., 19531 OF MINERAL OILS EXTRACTED FROM BREAD 699 fractions are not so great as those exhibited by the total unsaponifiable matters containing mineral oils, they are still appreciably greater than those of the original mineral oils. The second and third lines from the abscissa in Fig. 1 show typical results given by the unsaponifiable matter from bread oil mixed with liquid paraffin, B.P., before and after removal of some of the absorbing materials by chromatographic separation.The liquid paraffin was present at a concentration of 0.26 per cent. in the bread. More efficient separation than that provided by this particular chromatographic technique would therefore be necessary to determine the original quality or degree of refining of a mineral oil that has been extracted from a loaf of bread. Although greater efficiency might be attained by varying the adsorbent in the chromatographic column and the eluting solvent, separation of the mineral fraction by sulphation appeared to hold more promise. SEPARATION OF MINERAL OIL BY SULPHATION OF UNSAPONIFIABLE MATTER- The A.O.A.C. method' for determining mineral oil in fats is based on heating the unsaponifiable fraction with concentrated sulphuric acid of not less than 94 per cent.w/w a t 100" C for 30 minutes with occasional shaking. The apparatus used for this digestion is a Babcock milk-test bottle into which the fat is introduced in ether solution, the ether then being evaporated before the acid is added. After sulphation, further acid is added to the cooled digest until the liquid rises well into the graduated neck of the bottle, which is then centrifuged to bring the oil layer to the surface, where its volume is determined. This technique has been recommended3 for determining mineral oil in the unsaponifiable matter from bread oil, but it was suggested that the results might be unreliable if a low grade of mineral oil was present, owing to removal of impurities in the oil. In a preliminary experiment, the five mineral oils of various degrees of refining were treated as described above, and the ultra-violet absorption spectra of the sulphated oils were examined.The results in the fourth column of Table I indicate appreciable differences from the spectra of the corresponding materials before acid treatment. Although the more impure oils showed considerable reaction, as evidenced by loss of colour and appreciable reduction in absorption intensity, the white oils exhibited an increased absorption, particularly in the wavelength region below 2600 A, the Ei&, values at 2400 A being 0.19 for liquid paraffin, B.P., and 0.9 for a technical white oil. The corresponding values for the non-sulphated oils were 0.05 and 0.65. The increase in absorption intensity may be caused by the formation of sulphonic acids and the dehydrogenation of reactive naphthenes by this intense treatment with sulphuric acid; an alcohol extraction and clay treatment would probably be necessary to remove these absorbing compounds.Less drastic conditions of sulphation of the unsaponifiable matter from the total oil extracted from bread containing mineral oil were investigated. A temperature of 50" C was arbitrarily chosen for the reaction, and the acid concentration reduced until separation of the mineral oil either no longer occurred or was difficult. This was found to occur with 80 per cent. sulpliuric acid. The use of 85 per cent. acid, however, permitted a separation of oil after the reaction. Under these conditions the less refined oils usually still retained some of their colour after treatment and this in itself allowed a rapid distinction to be drawn between them and white oils.If the layer above the acid in the Babcock bottle is solid or semi-solid in reasonable quantity and not a thin black sludge, as occurs when no mineral oil is present, it is safe to assume that a paraffin wax or petroleum jelly has been absorbed into the bread dough. Fig. 2 shows the spectra given by separated mineral oils after sulphation of the total unsaponifiable matters from breads containing these oils with 85 per cent. w/w sulphuric acid at 50" C for 30 minutes. These results have been obtained in duplicate by operatives working in different laboratories, and in one case using a Uvispek absorption spectrophoto- meter. The spectra of the extracted mineral oils after this sulphation treatment closely approxi- mated to tklose of the original oils, except for some fall in the extinction values for the least refined oil.The main difficulty with this technique is that the spectra of the recovered two white oils are not readily distinguishable, although the figures in Ta,ble I show that in relation to each other imeasurablc differences exist. The reason for this may, in part, lie in a reaction between sulphuric acid and quite higlily refined oils, with the formation of substances having differeqt absorption characteristics.700 COOKSON, COPPOCK AND SCHNURMANN: ESTIMATION OF THE QUALITY [Vd. 78 A further series of experiments was made with 90 per cent. instead of 85 per cent. sulphuric acid, and although the resultant spectra of the extracted oils still permitted distinction hetween oils of different purities, they were not as sensitive as those obtained aiter treatment with 85 per cent.acid. The wavelength region between 2100 and 2 4 0 0 ~ was examined in the hope that dis- crimination would be possible between two white oils of different quality after extraction from bread. Liquid paraffin and a technical white oil were examined, but the spectra of these oils after extraction from bread could not be readily distinguished a t these lower wavelengths. 200 \ \ 150- .o p ; 100 s1 i j 50 2400 2600 2800 3000 3200 3400 Wavelength, A Fig. 2. Absorption spectra of mineral oils separated from the unsaponifiablc matter of bread oils with 85 per cent.sulphuric acid at 50" C. . . . . . . Liquid paraffin. - _ - - Technical white oil. -. ._. . Transformer oil. --__ An acid and clay treated Edeleanu extract: of a lubricating oil distillation fraction. _- - Spindle oil The results recorded and used for correlation purposes in the original quantitative work oil mineral oils in bread3 have not been amended by this new sulpliation technique, as the re-determinations would be very tedious and would not significantly alter or increase the sensitivity of the quantitative method. None the less, the technique now described is preferable for routine use if a quality determination is also required. CONCLUSIONS For the lubricating oils used in bread-making, the question arises whether liquid paraffin, B.P., and some other white oils are so significantly different in physical characteristics that only the former, as has previously been maintained, should be used.It is suggested that Table 11, which shows the E::m values at 2 6 0 0 ~ for the various oils examined before use and after extraction from bread, will supply the answer to this question. From theDec., 19531 OF MINERAL OILS EXTRACTED FROM BREAD 701 figures shown in this table, it is clear that medicinal liquid parafin, B.Y., and soine other TABLE I1 E:?~ at 2600 A ,-h--------\ Mineral oil Refore use Extracted from bread Liquid parafiin, B.P. .. . . . . .. 0.030 2.6 Technical white oil . . . . .. .. .. 0.270 4.0 Transformer oil . . .. .. .. . I 37-0 33.0 An acid-treated Edeleanu extract of a lubricating- oil distillation fraction .. .. .. .. 49.0 44.0 Spindle oil . . . . .. .. .. .. 220-0 173.0 wliite oils, even though they may not be of E.P. quality, fall into a related group when compared with the less refined colour-bearing mineral oils (see Fig. 2). The absorption intensities of the mineral oils when compared with the natural unsaponifiable oils from bread, are (i) lower for the white oils, (ii) usually greater for the colour-bearing minerai oils and (iii) lower for all the mineral oils examined when compared with some of the com- ponents that have been isolated from natural unsaponifiable matter of the oil in bread. Mineral oils can be recovered from bread by the suggested procedure comparatively unchanged spectroscopically, and the relationship of their absorption intensities to that of the natural unsaponifiable oil of bread remains as sumniarised here.Provided, as is desirable on both pharmacological8 and technical grounds, that the viscosity of the oil is not abnormally low, then, it seems to us for the above reasons that mineral oils for which the E!Fm value at 2 6 0 0 ~ does not exceed 0.5. before use, or 5.0 after extraction from bread by the procedure suggested above, could be regarded as satisfactory for the lubrication of plant used in bread production. The authors express their thanks to Mr. W. F. Maddams, MSc., Miss P. M. Martin, bS.Sc., and Mr. R. Mayoh, BSc., for their assistance with the spectrographic work. 1. 2 . 3. 4. 5. 6. 7. 8. REFERENCES Coppock, J. U. M., and Cookson, M. A., Brit. Med. J., 1949, 73. Statutory Instrument, 1949, No.614, The Mineral Oil in Food Order, H.M. Stationery Office, Cookson, M. A., and Coppock, J. B. M., J . Sci. Food Agric., 1951, 2, 434. Schnurmann, R., Martin, P. M., and Maddams, W. F., J . Pharm. Yharmacol., 1951, 3, 298. Williams, I<. A,, J . Ass. Off. Agric. Chem., 1949, 32, 668; 14nalyst, 1946, 71, 261. Coppock, J. B. M., and Cookson, M. A., J . SOC. Chern. Ind., “Official Methods of Analysis of the Association of Official Agricultural Chemists,” The Association Coppock, J. B. M., Brit. J . Nutr., 1951, 5, 383. 1949. 1949, 68, 274. of Official Agricultural Analysts, Washington, D.C., 1950, p. 209. BRITISH BAKING INDUSTRIES RESEARCH ASSOCIATION PHYSICS DEPARTMENT M ANCHESTER CHORLEYWOOD, HERTS. MANCHESTER O ~ L REFINERY LIMITED First submitted, illarch IBth, 1963 Amended, June Wc, 1953 DISCUSSION THE PRESIDENT, Dr.D. W. Kent-Jones, said he thought that many of the chemists who specialised in the examination of oils might have important observations to make, as they probably had expcriencc of such matters in other connections. As many analysts did not possess spectrophotometers, he wished to know if it was possible to use the ordinary Spekker absorptiometer with the ultra-violet light attachment. He was intercsted to hear that wholemeal bread, as opposed to white bread, apparently contained something in the unsaponifiable portion of the bread oil that had very high absorption in the ultra-violet spectrum, and he wondered if the authors had any information as to what these substances were.MR. COOKSON replied that preliminary experiments indicated that mineral oils of various qualities could be differentiated with a Spekker absorptiometer adapted to measure the transmission of ultra-violet light through the oil. With suitable filters to exclude as much as possible of the visible light, increasingly large-scale deflections were obtained as the quality of the mineral oil decreased. However, more recent work had indicated that further experimentation was required beforc such x technique inight be found t o be successful.702 COOKSON, COPPOCK AND SCHNURMANN: ESTIMATIOK OF THE QUALITY [Vol. 78 The compounds naturally present in the unsaponifiable fraction of bread oils and responsible for high absorption in the ultra-violet spectrum had not as yet been identified.Adsorption chromatography had enabled a separation of some constituents of unsaponifiable flour and bread oils to be made, and the similarity between these oils from various sources had been shown by ultra-violet spectrophotomctry. The reason for the greater absorption intensity of the unsaponifiable oil from wholemeal bread was unknown, but it might be due to the presence in this oil of compounds that are removed when flours are milled to a lower extraction rate. The very high extinction coefficients found a t wavelengths below 3000 .4 were, however, of some interest. DR. J. R. NICHOLLS said that white petroleum oils were prepared commercially by the somewhat drastic treatment of coloured oils with sulphuric acid, whereby many other impurities were removed with the colour.The better grades were used for pharmaceutical purposes and were required by the British Pharmacopoeia to conform to a maximum limit of carbonisable substances. From the absorption curves illustrated it appeared clear that technical white oils contained only slightly greater amounts of impurities than liquid paraffin, B.P., and that the proposed method was adequate for distinguishing white oils froin cruder products. The only difficulty in permitting all grades of white oils to be used for lubrication of plant in bread- making was the possibility that some grades might contain carcinogens. Bearing in mind the purification that all white oils had undergone and the maximum amount of oil that was allowed to be present in the finished bread, any danger appeared to be small.But it was desirable that adequatc tests should bc made. MR. COOKSON said that extensive surveys of the literature and discussions with authorities on this subject had failed to produce the slightest evidence of carcinogenicity from the oils included in their specification. DR. K. A. WILLIAMS said he understood that both white and medicinal oils were made from naphthenic- base stocks by treatment first with fuming sulphuric acid and then with alcohol. The hydrocarbons present included a fairly high proportion of cycloparaffins mainly derived from cyclopentanc and cyclohexanc. MR. N. L. ALLPORT referred to a sample of medicinal liquid paraffin sent to him because it had been reported against as failing the British Pharmacopoeia's sulphuric acid test.In point of €act it quite safely passed the B.P. requirement, and in the course of subsequent discussion it transpired that the first analyst had used concentrated sulphuric acid of 98 to 99 per cent., whereas the Pharmacopoeia specified a special reagent described in the First Appendix as consisting 01 96 per cent. by weight of nitrogen-free sulphuric acid. When presenting the paper Dr. Coppock had interposed that with certain unsaponifiable fractions the extinction coefficient as measured by a photo-electric type of spectrophotometer was different from that recorded by a photographic type of instrument. He asked if the authors could say which instrument gave the true value. DR. SCIINURMANN replied that in the wavelength region between 2400 and 3000 A it did not matter whether a photographic plate instrument or a photo-electric instrument was used.With the same type of prism in either of these instruments, results were almost identical. Difficulties might arise, however, a t wavelengths shorter than 2400 A when photo-electric instruments were used because a t these shorter wavelengths the emission intensity of the hydrogen discharge tube fell off, the sensitivity of the photo-cell decreased, the reflectivity of the mirrors became poorer, and the proportion of scattered radiation increased : hence with photo-electric instruments the indicated absorption intensity might appear to decrease in the region between 2400 and 2 0 0 0 ~ on approaching the shorter wavelength end even for samples for which i t could be proved with the help of photographic plate instruments that the absorption intensity increased towards shorter wavelengths.The photographic plate instrument, in addition to supplying a permanent record, had the advantage that i t could be safely used over the whole region over which the photographic plate was sensitive. DR. W. W. MYDDLETON, in a written contribution, said he wondered what iise was to be made of the results described in the paper, and asked whether there was to be a drive in favour of using medicinal paraffin or technical while oils and, if so, on what grounds. He asked if it was claimed that poorly refined oils endowed thc bread with an objectionable taste, smell or colour, because if so, the results of the spectrophotometric method could but confirm the evidence of the senses.He said that at one point in the presentation of the paper they were left to infer that, because medicinal paraffin and a technical white oil could be heated above 200" C for some time without any effect on the ultra-violet absorption spectrum, a similar state of affairs would exist in the baking of the loaf. He found this inference difficult to draw, for the oil in the bread was probably in thin films and was in contact with many different components of the bread and, in addition, some of i t was in contact with metallic and metallic oxide surfaces during baking. There was therefore a probability of catalysed reactions occurring and involving, perhaps, dehydrogenation, polymerisation or pyrolysis. If after being baked the product was cleaned by extraction and controlled sulphation, it was doubtful what was in fact produccd as a result of heating the mineral oil, or even medicinal paraffin.There was 110 evidence; it had not been looked for and, in fact, it may have been destroyed.Dec., 19531 OF MINERAL OILS EXTRACTED FROM BREAD 703 As an example of the variable behaviour of mineral oils under different conditions, Dr. Myddleton referred to his own experience of the changes that occurred in a technical white oil, similar to the one heated to 200" C by the authors, when exposed to sunlight for three months ( J . Inst. Pet., 1951, 37, 57). The maximum temperature was about 62" C for about 1 hour on bright days. The minimum on the ultra-violet absorption curve was raised by 64 per cent.and the maximum by 14.5 per cent. The ratio of the maximum t o the minimum was lowered from 1.63 to 1.18 and the position of the maximum was shifted from 2710 to 26%0 A . He therefore felt diffident in accepting the assumption that even highly refined white oils remained unchanged in the process of bread baking. DR. COPPOCK, in a written reply, said he wished to emphasise that the main purpose of the paper was to describe a mcthod whereby the original quality of oils used for lubrication in bread-making could be determined after use. In reply to the first question raised by Dr. Myddleton, he pointed out that the Mineral Oil in Food Order restricted the quantity of mineral oils present in foods as a result of processing operations to not more than 0.2 per cent.by weight, but did not define the quality of oil regarded as desirable. Thc authors had advocated previously that the quality of mineral oil used for lubricating bread-making machinery should be that of liquid paraffin, B.P., but in view of the spectrophotometric evidence presented in this paper they now suggested that certain white oils might also be used. The pharmacology of mineral oils had naturally been taken into account, and there was no evidence that the oils included in the proposed specification could in any way be harmful when used for this purpose. Secondly, the smell, colour or taste of poorly refined oils could not be detected a t levels of 0.2 per cent. in bread, and therefore more scientific methods would be required to show their presence.Thirdly, the heating of white oils to 200" C in air caused negligible change in the oils as determined by their ultra-violet absorption spectra; Dr. Myddleton's experience of the effect of sunlight on mineral oils would appear irrelevant to a consideration of the baking process, but the above conditions of heating were considerably more drastic than would be found in bread-baking, for which the temperature, except at the crust, did not exceed 100" C and the atmosphere was preponderantly carbon dioxide. Further, none of the components of bread would be expected to decompose mincral oil, even if the oil was dispersed in thin films. The possibility of decomposition at the baking-tin surface could not be ignored, but it was doubtful whether the conditions were sufficient for this to occur; it should also be noted that the commercial use of mineral oils for tin greasing was almost negligible.In any event, a vegetable oil could be just as prone, and possibly even more prone, to adverse reactions ; e.g., their breakdown products under certain conditions of heating were believed to be carcinogenic (see, e.g., Peacock, I-'. R., and Beck, S., Nature, 1948, 162, 252) and there were inherent dangers should vegetable oils become rancid (Frazer, A. C., Trans. Roy. SOG. Trop. Med. Hyg., 1952, 46, 576). Fourthly, it was extremely unlikely that the extraction procedure removed any decomposition products fornied in mineral oils on baking in bread and, as just stated, the authors did not believe that any decom- position occurred. It should be remembered that white oils were refined by far more drastic sulphuric acid treatments than were used to separate the mineral oil from the natural unsaponifiable oil in bread, and the sulphation conditions were selected to be just sufficient to achieve this. The similarity of the ultra-violet absorption spectra of the mineral oils of different quality, before use and after recovery from bread, were striking, and did not suggest that any marked changes in the mineral oils had occurred, or that the controlled sulphation was sufficiently vigorous to destroy any decomposition products that may have formed. The much more pronounced absorption intensities of the natural unsaponifiable oils in bread compared with those of white oils should also be noted in this connection. Finally, with regard to Dr. Myddleton's statement of the effect of daylight exposure of a technicaI white oil which by spectrophotometric evidence was not stable, expcriments recorded ekewhere (Schnurniann, R., paper presented at the Rkunion Internationale de Spectroscopie Moleculaire, Paris, July, 1953) have shown that medicinal liquid paraffins that obeyed the spectrophotometric quality and stability criterion showed no significant changes of their spectra after exposure to an ultra-violet light soiirce pifovided that the oils were contained in a glass vessel.

ISSN:0003-2654    

DOI:10.1039/AN9537800695

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

704 MENDELOWITZ AND RILEY : THE SPECTROPHOTOMETRIC DETERMINATION [kT0l. 7 8 The Spectrophotometric Determination of Long-chain Fatty Acids Containing Ketonic Groups With particular reference to Licanic Acid BY A. MENDELOWITZ AND J. P. RILEY A spectrophotometric procedure, based on the alkaline 2 : 4-dinitrophenyl- hydrazone method of Lappin and Clark, is described for the determination of ketonic groups in long-chain compounds. The effect of a number of variables on the final intensity of the colour has been investigated. The method has been applied to the determination of licanic acid in the presence of other saturated and unsaturated fatty acids and shows a standard deviation of 0.8 per cent. IN the course of a study1 of oils containing licanic acid (4-oxo-octadeca-9:11:13-trienoic acid), a method was required for its determination in the presence of elaeostearic acid (octadeca-9 : 11 : 13-trienoic acid), with which it is usually associated.As the ultra-violet absorption spectra of the two acids are almost identical,2 measurement of the absorption at 270.5nip gives only a measure of the total conjugated trienoic unsaturation. On the other hand, licanic acid can be estimated by determination of the ketone value. Several volumetric procedures have been described for the estimation of ketonic groups in long-chain fatty-acid derivative^,^^^^^^^ mainly adaptations of the hydroxylamine method originally introduced by Brochet and Cambrier,6 or its modification in which pyridine is used to ensure complete r e a ~ t i o n . ~ These methods suffer from the disadvantage that the end-point is always somewhat indistinct, particularly in the presence of pyridine.A number of workers have used phenylhydrazine and its derivatives for the gravimetric8 or v o l ~ m e t r i c ~ J ~ ~ ~ ~ determination of ketonic compounds. These procedures depend on either the quantitative recovery of the phenylhydrazone of the ketone or the determination of unreacted phenylhydrazine. All were found to be inapplicable to long-chain fatty-acid derivatives owing to the difficulty of separating the phenylhydrazones in a pure state; for instance, attempts to prepare the 2 :4-dinitrophenylhydrazone of methyl 12-oxostearate resulted in the production of an oily product containing much unreacted 2 :4-dinitrophenyl- hydrazine, which could only be removed with great difficulty.Observations by Bamberger12 and Gnehm and Benda13 that iiitropheiiylhydrazones form dark-coloured water-soluble compounds on treatment with alkali have been made the basis of colorimetric procedures for the estimation of ketonic compounds. The earlier applications of this method14 7 1 5 7 l 6 ,l7 7 l 8 were to water-soluble ketonic compounds, and they are therefore unsuitable for use with the higher fatty acids and their derivatives. Lappin and ClarklS have reported an extension of the method to the determination of traces of aldehydes and ketones in water, organic solvents or organic reaction products. They reported that both the wavelength of maximum absorption and the molecular extinction were inde- pendent of the structure of the ketonic compound.Poole and Kloose20 have described a syectrophotometric procedure for the estimation of monoketonic compounds in rancid foods. The 2 :4-dinitrophenylhydrazones of the ketones were formed in benzene solution in a column of activated alumina. Alcoholic potassium hydroxide was added to the eluate from the column and the intensity of the resultant red colour was determined at 435mp. The colour was not stable, however, and faded rather rapidly. In view of the poor results that were attained with the hydroxylamine method for the determination of methyl 12-oxostearate, the procedure of Lappin and Clark seemed to offer the advantages of accuracy and simplicity. These workerslg treated 1 nil of a methanolic solution of the ketone with 1 ml of a 0.06 per cent. solution of 2:4-dinitrophenylhydrazine in methanol and 1 drop of concentrated hydrochloric acid.This was heated in a loosely stoppered test tube at 100" C for 5 minutes and then 5 ml of 10 per cent. methanolic potassium hydroxide were added. The absorption of the solution was measured at 480 mp against a blank determination made with 1 ml of methanol instead of the ketone solution.Dec., 19631 OF LOSG-CHAIK FATTY ACIDS CONTAINING KETOSIC GROUPS 706 Preliminary experiments made with cyclohexanone and methyl 12-oxostearate revealed five defects in their procedure- (;) Much evaporation took place during the heating a t 100" C, and often no methano1 remained at the end of the reaction period. In order to reduce evaporation to a minimum, a temperature of 60" C was used for the condensation in all subsequent work and ethanol was substituted for methanol.(ii) Small variations in the amount of hydrochloric acid added as catalyst in the condensation reaction produced large differences in the intensity of the colour. 61 350 450 550 Wavelength, rnp 3 Fig. 1. Comparison of absorption curves of solution and blank solution. Curve A, methyl oxostearate ; curve B, reagent blank; curve C, methyl oxostearate less reagent blank (iii) Potassium chloride was precipitated after the addition of alcoholic potassium hydroxide and it was difficult to obtain absolutely clear solutions. (iv) The wavelength of maximum absorption was not 480mp, but varied with the nature of the ketone. Long-chain aliphatic ketones and CycZohexanone exhibited maximum absorption at 435mp (with a secondary maximum at 535mp, see Fig.I), whereas with cinnamaldehyde the maximum absorption was at 486 mp. (v) Owing partly to the difference in structure of the dinitrophenylhydrazones and partly to the equilibrium nature of the condensation reactions, the molecular extinction at the wavelength of maximum absorption depended markedly on the nature of the ketonic compound. For example, cyclohexanone and methyl oxostearate gave E (at 435mp) of 17,900 and 11,400, respectively. A sample of pure cyclohexanone-2 :4-dinitrophenylhydrazone in the same strength of alcoholic potassium hydroxide had an E of 19,530 at 435 mp, showing that the condensation of cyclohexanone with the reagent had only proceeded to the extent of about 92 per cent.During preliminary work it was found that high reagent blanks resulted when measure- Fig. 1 shows the absorption curves for a determination with ments were made at 435 mp.706 MENDELOWITZ AND RILEY: THE SPECTROPHOTOMETRIC DETERMINATION [Vol. 78 methyl oxostearate and for the reagent blank, both measured against water in the compensator cell. The rather high absorption of the reagent blank at 435 mp was almost entirely due to the shoulder of the absorption band at about 395 mp of alkaline 2 :4-dinitrophenylhydra~ine.~l On the other hand, at the secondary absorption maximum at 535 mp, the absorption of the blank was quite low and remained fairly constant over a considerable range of wavelengths; for these reasons all further measurements were made a t this wavelength.The effect of variation in hydrochloric acid concentration on the condensation reaction was examined. Some l-ml samples of a 0.0065 per cent. ethanolic solution of methyl oxo- stesrate were treated with l-ml portions of dinitrophenylhydrazine reagent containing different strengths of acid (0.06 g of dinitrophenylhydrazine with 1 ml of hydrochloric acid made up to 100 ml with ethanol). The mixture was heated at 60" C for 50 minutes; then, after cooling, 5 ml of 10 per cent. alcoholic potassium hydroxide were added and the solution was diluted to 10 ml with water. The results are shown below- Normality of acid . . .. .. .. 2 3 4.6 6 E:& at 535 mp* . . .. .. .. 227 344 352 361 * After deduction of reagent blank. To minimise variations caused by changes of acid concentration, all subsequent work was performed with a reagent containing 1 ml of constant-boiling hydrochloric acid per 100 ml.With this reagent, the time for attainment of equilibrium in the initial reaction was determined for licanic acid and oxostearic acid and its methyl ester, with the results shown in Table I. In each determination the condensation was complete within 50 minutes at 60" C. TABLE I RATE OF REACTION OF LONG-CHAIN KETONIC COMPOUNDS WITH 2 :4-DINITROPHENYLHYDRAAZINE Time, minutes 30 35 40 45 50 56 60 E:?m a t 535 mp for oxostearic methyl licanic acid oxostearate acid 36 1 346 - 370 355 - 376 359 - 378 362 265 378 362 270 - 268 377 362 269 7 L > - Small changes in the amount of alcoholic potassium hydroxide solution added produced little alteration in the intensity of the red colour; for instance, use of 7.5 ml of this reagent instead of 5 m l increased the final optical density by only 1-5 per cent.The substitution of water for alcohol in the dilution to the final volume had only a slight effect on the final colour and prevented the precipitation of potassium chloride that had been encountered when the method of Lappin and Clark was used. With the modified conditions it was found that, although highly reproducible results were attained (for methyl oxostearate) if the same sample of alcohol was used throughout, fresh batches of alcohol gave rather different values. That this was due to the presence during the condensation reaction of variable amounts of water in the alcohol was shown by the addition of known amounts of water to the alcohol used as solvent for the ketonic compound.Water in alcohol, per cent. . . * . 0 0.1 0.5 1.0 .. .. .. 377 377 362 366 El% 1cm a t 535 mp . . To overcome this difficulty a control test must be made simultaneously on the pure ketone being determined, or if this is not stable, on some ketonic compound of a similar nature that has been calibrated against the unstable ketonic compound. In the present work, oxostearic acid was used as a control for licanic acid and for methyl oxostearate, with the results shown in Table 11. The constancy of the extinction coefficients of both methyl oxostearate and licanic acid, when related to an arbitrary fixed value of 360 for the extinction coefficient for oxostearic acid, is proof of the validity of this procedure.Dec., 1953) OF LONG-CHAIN FATTY ACIDS CONTAINING KETONIC GROUPS TABLE 11 707 E:L I'ALUES FOR SIMULTANEOUS RUNS OF OXOSTEARIC ACID WITH METHYL OXOSTEARATE AND WITH LICANIC ACID Methyl oxostearate - calculated to Oxostearic oxostearic acid found acid = 360 370 362 352 368 360 352 364 368 354 367 360 353 Mean 353 Licanic acid P calculated to Oxostearic oxostearic acid found acid = 360 358 270 272 373 280 273 363 275 270 Mean 272 It will be noted that the value found for licanic acid is very much lower than those of either oxostearic acid or its methyl ester.This emphasises the fact that the method can only be applied to a particular ketonic compound after standardisation with that compound. Unsuccessful attempts were made to replace the alcohol used in the condensation reaction by other solvents, such as benzene or cyclohexane, which are more readily obtained free from ketonic compounds.The red colour, if it was produced at all, was very weak and ephemeral. METHOD STANDARD SUBSTANCES- Methyl 12-oxostearate-Methyl ricinoleate (purity 99-4 per cent .)22 was hydrogenated with Kaney-nickel catalyst .23 The resulting methyl hydroxystearate was treated with a 100 per cent. excess of a 10 per cent. solution of chromium trioxide in glacial acetic acid. After being allowed to stand at room temperature for 2 hours the mixture was poured into water and filtered. The precipitate was dissolved in ether and the ethereal solution was washed with 5 per cent. potassium carbonate solution and with water.The crude ester, recovered by evaporation of the solvent, was crystallised three times from acetone at 0" C. The purified methyl 12-oxostearate consisted of white plates, m.p. 46.8" to 47.5" C; previous workers have reported 45" to 46" C24 and 44.5" to 45" C.25 12-Oxostearic acid-Crude methyl oxostearate was saponified with alcoholic potassium hydroxide. The recovered acid was crystallised three times from acetone at 0°C. The pure acid had m.p. 81.5" to 82" C; previously recorded values are 80" to 81" C24 and 81" C.26 Licunic acid-The mixed acids obtained from a commercial oiticica oil were dissolved in 20 ml of boiling light petroleum, boiling range 40" to 60" C, per gram of mixed acids. The boiling solution was filtered through a warm funnel to remove small amounts of resinous oxidation products.On cooling, the filtrate deposited crude licanic acid, which was re- crystallised five times from light petroleum. The purified licanic acid was pressed well on the filter and dried in a high vacuum for not longer than 10 minutes. The products from a number of preparations all melted in the range 74.5" to 75-5" C, and gave the following values of at 270.5 mp: 1711, 1703, 1724, 1702, 1725, 1729, mean 1716; Rose and Jamieson2 record 74.5" to 75-5" C and 1784. REAGENTS- Ketone- free ethanoZ-Distil absolute alcohol from zinc dust and potassium hydroxide. Heat 1-5 litres of the distillate under reflux with 10 g of 2:4-dinitrophenylhydrazine and 10 ml of concentrated sulphuric acid. Distil the ketone-free ethanol after 2 hours and store it in a tightly stoppered bottle in the dark.The alcohol, which is slightly coloured yellow owing to the volatility of 2 :4-dinitrophenylhydrazine, is stable for about a week. 2 :4-DinitrophenyZhydraxine yeagent-Add 1 ml of constant-boiling hydrochloric acid to 0.06 g of 2:4-dinitrophenylhydrazine, add 80 ml of ketone-free ethanol and warm until all is dissolved. The reagent is stable for at least 3 days if stored in the dark. AZcohoZic potassium hydroxide-Dissolve 10 g of potassium hydroxide in 20 ml of water and dilute to 100 ml with ethanol (which need not be ketone-free). Cool and dilute to 100 ml with ketone-free ethanol.708 MENDELOWITZ AND RILEY : THE SPECTROPHOTOMETKIC DETEKMINATION [Vol. 78 PROCEDURE- Weigh about 0.1 g of the substance to be examined, dissolve it and make the solution up to 100 ml with ketone-free ethanol. Dilute this solution accurately with the same alcohol to give a solution approximately 0.001 per cent.with respect to the ketonic group (>C=O). Place 1 ml of the diluted solution in a 10-ml calibrated flask and add 1 ml of dinitrophenyl- hydrazine reagent. Stopper the flask loosely and heat it to 60" C in a thermostatically controlled bath. After 50 minutes, cool the flask in water for 2 minutes. Add 5 ml of alcoholic potassium hydroxide rapidly and dilute to 10ml with water. Measure the optical density of the solution in a 1-cm cell at the secondary absorption maximum (535 mp for 12-oxostearic and licanic acids), using water in the compensating cell.Determine the reagent blank in the same manner with 1 nil of purified ethanol instead of the solution of the ketone. Carry out control estima- tions with the pure ketone simultaneously with the determination. If the ketonic compound to be determined is unstable, e.g., licanic acid, make simultaneous calibration tests with a related, but stable, ketone, e g . , oxostearic acid, and with the unstable compound. Use the stable ketone in place of the unstable compound as a control in all determinations. After 1 minute press the stopper home to minimise evaporation. CALCULATION FOR UNSTABLE KETONES- Calculate the extinction coefficients (E:?J of the sample and standards from the optical densities of the solutions after deduction of the reagent blanks. Let a and b be the extinction coefficients, determined simultaneously for the pure unstable ketone and the stable standard, respectively. If x and y are the extinction coefficients of the sample containing the unstable ketone and of the stable ketone respectively, then the percentage of the unstable ketone in the sample is- b x x x 100.DETERMINATION OF I~ETOSES IS THE PRESENCE OF FATTY xms- Known mixtures of ketones with non-ketonic compounds were analysed with the results shown in Table IIJ. The non-ketonic compounds used included the mixed acids of tune; oil, in order to ensure that such acids would not interfere in the determination of licanic acid in oils such as oiticica. Each result is the mean of four determinations showing an average standard deviation of 0-8 per cent.TABLE I11 ANALYSIS OF MIXTURES Ketone Non-ketonic E;:m of of compound mixture control Methyl oxostearate Methyl stearatc 154 350* Methyl oxostearate Palmitic acid 183 350* - Palmitic acid 0 350* Oxostearic acid Tung mixed acids 176 3787 - Tung mixed acids 2.4 378t Licanic acid Tung mixed acids 164 3737 * Methyl oxostearate. 7 Oxostearic acid. As licanic acid, but probably due to oxidation products. Ketone in mixture I Found, 44.0 52.3 0 0.8: 46.5 58.6 % 1 Calculated, 43.6 52.5 0 0 46-3 58.1 % C o N CL u s I o N s- The method described is by no means limited to the fatty acids, but could be applied after suitable standardisation to many other types of ketonic compounds. It presents the advantages of rapidity and accuracy, and can be very readily adapted for use with micro-samples.The authors wish to thank Professor T. P. Hilditch, C.B.E., F.R.S., for his interest and valued criticism.L)eC., 19531 OF LONG-CHAIS FATTY ACIDS CONTAINING KETOSIC GKOUPS REFERENCES 70'3 1. 3. 4. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 9 d. 3. Mendelowitz, A, unpublished work. Rose, W. G., and Jamieson, G. S., Oil and Soap, 1943, 20, 227. Leithe, VC'., Fette ?(. Seif., 1938, 45, 615. Kaufmann, lit. P., Funke, S., and Liu, F. Y., Ibid., 1938, 45, 616. Feuell, -g. J., and Skellon, J. H., Analyst, 1953, 78, 135. Brochet, -I., and Cambrier, R., Comfit. Rend., 1895, 120, 449. Bryant, W. 31. D., and Smith, D. M., J . Amer. Chcm. SOC., 1936, 57, 57. Iddles, H. X., and Jackson, C. A., Ind. Eng. Chem., '4nal. Ed., 1934, 6, 454. Strache, H., Monatsh. Chem., 1891, 12, 514. Schoniger, W., and Lieb, H., Mikrochemie, 1951, 38, 165. Ardagh, E. G. Ii., and Williams, J. G., J . Amer. Chem. Soc., 1925, 47, 2976. Bamberger, E., Ber., 1893, 26, 1306. Gnehm, R., and Benda, L., Ibid., 1896, 29, 2017; Annalen, 1898, 299, 101. Xeuberg, C., and Strauss, E., Arch. Biochem., 1945, 7, 211. Neuberg, C., and Hobel, M., Biochem. Z., 1928, 203, 467; 1929, 210, 466; 1929, 216, 493; 1930, Case, E. M., Biochem. J., 1932, 26, 753. Lu, G. D., Ibid., 1939, 33, 249. Friedemann, T. E., and Haugen, G. E., J . B i d . Chenz., 1943, 147, 416. Lappin, G. R., and Clark, L. C., Anal. Chem., 1951, 23, 541. Poole, M. F., and Kloose, A. A., J . Amer. Oil Chem. SOC., 1951, 28, 215. Roberts, J. D., and Green, C., J . Amer. Chem. Soc., 1946, 68, 214. Riley, J. P., Analyst, 1951, 76, 40. Adkins, H., and Billica, H. R., J . Amer. Chem. Soc., 1948, 70, 695. Grummitt, O., and Siedschlag, K. G., J . Auner. Oil Chenz. SOL, 1940, 26, 690. Perotte, R., Compt. Rend., 1934, 199, 358. 219, 490; 1930, 229, 256; 1932, 256, 481; Ber., 1930, 63, 1986. DEPARTMENT OF OCEANOGRAPHY UNIVERSITY OF LIVERPOOL LIVERPOOL, 3 June 3 4 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800704

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

L)ec., 1Dfi3] OF LONG-CHAIS FATTY ACIDS CONTAINING KETOSIC GKOUPS 70‘3 Design and Operating Technique of a Vacuum Drying Oven Part I. Design of the Oven BY S. D. GARDINER A syininetrical vacuum drying oven of circular plan, large heat capacity and uniform temperature distribution has been designed. It has a shallow, hollow- but thick metal base, which holds six sample dishes with the minimum of free space. Controlled dry air bleed is incorporated. A STANDARD reference method for determining water in sugar products, as distinct from quick routine methods, requires the use of a vacuum oven free from the many faults in design inherent in almost all of the ovens currently available, such as those mentioned in the Proceedings of the International Commission f o r Uniform Methods of Sugar Analysis, 1949.To overcome the difficulties in drying such labile products as sugar syrups and molasses containing fructose, the oven described here was designed and constructed in the Tate and Lyle Research Laboratories. Its use need not be limited to the products mentioned-it can be more generally utilised whenever exact and reproducible results of experiments on drying losses from many other materials are required. The design incorporates features of the oven reported to the author by a colleague, who visited the National Gas Board Laboratories. Attention has also been paid to the recommendations made by Iles and Shaman ,1 although these refer to atmospheric pressure ovens. GENERAL DESCRIPTION- base, which holds six sample dishes with minimum free space. Fig.I shows that the oven consists primarily of a shallow, hollow, but thick, metal The lid is shown resting[Vol. 78 in the off position (at the rear of the figure), where it is out of the way and suspended so that it can be kept clean and undamaged. The vacuum seal is produced with the aid of a wing- nut that clamps down the heavy lid over the central control spigot, the raised annulus of the base sealing the oven by pressin5 on to the grease-free silicone-rubber ring, inserted into the complementary circular groove in the lid. At the samc time, provided sufficient pressure is applied, good thermal contact is made with the base. Vacuum sealing at the spigot is 7 10 GARDINEK : DESIGN AND OPERATING TECHNIQUE - I Vacuum seal Annular I / ' / I * 'I / recess J 1 , I I 0 ? 1 *Air exit inches I Fig.3. Design of vacuum oven ensured by a copper washer under the wing nut. The washer is held in position by a circular projection in the lid and is normally lubricated with a little grease. To conform with the shape of the oven the air flow has been arranged symmetrically. Fig. 2 shows the under-side of the oven with the air-distributing leads from the central supply connected by thick rubber tubes to six entry ports situated around the perimeter of the base of the oven. Fig. 3 shows that these ports, which are inclined at an angle of 45", project the air up against the oven lid immediately over the sample dishes, from whence it passes over their surfaces to be drawn out through six exit ports arranged symmetrically in the lid near the centre.These exit ports inclined at 30" force the air upwards through the centre hole enclosing the spigot, and so out of the oven. The angles are chosen to facilitate removal of water vapour from above the sample. The air flow is adjustable to between 0.5 to 100 ml per minute by means of an Edwards needle-control valve (type LBl), after the air has been dried through activated alumina (6 to 8 mesh) and over barium oxide lumps. The alumina is arranged in a long narrow tube (Fig. 1) in order to reduce the water content of the air to 0.005 mg per litre; theFig. 1. General view of the ovenFig. 2. View of the underside of the oven and the air distributing leadsDec., 19531 OF A VACUUM DRYING OVEN 71 1 lumps further reduce the water to 0.0007 mg per litre. The drying train is attached to the oven and moved with it.Iles and Sharman state that in the absence of the steaming effect (catalysis of glucose and amino-acid condensation) about 2 ml of air per minute per gram of water are required. The new vacuum oven is certainly free from this effect. Six samples each holding a maximum of 2.5 g of water (including added water) would have a maximum total of 15.0 g of water to be removed, which according to Iles and Sharman would need a minimum of 30 ml of air per minute, but as this amount is well within the limit of control by the needle valve, the steaming effect is safeguarded against on two counts. A simple liquid- detergent film air-flow meter may be suspended from the vertical drying tube.Fryd and Kiff2 found experimentally that the heat from the air flow contributes only slowly to the rise in temperature of the sample; conduction and radiation supply most of the heat required. The oven described here gives the same degree of heat transfer by radiation and conduction at all points in its interior. As the surface of the recess is ground flat, heat transfer is facilitated to such a degree that it is essential to commence heating at a low temperature and to heat to operational temperature by resetting the thermostatic control. This procedure assists in preventing crust formation. Uniform temperature distribution is attained by using twelve electric heaters inserted radially around the base and penetrating to near the centre. The total wattage is 80, each heater being deliberately run at one-tenth of its normal rating.An indicator lamp assists in setting the thermostatic control. One oven was controlled to within k0.5" C by a Sunvic adjustable bimetal thermostat (T.S.l) in an oil-filled pocket, in conjunction with a Sunvic hot-wire vacuum relay (F.102-3M). Another oven was regulated to within 1.0" C by an electronic capacity-switch, operating off the 0" to 110" C thenno- meter. The former method is cheaper and, in the long run, simpler to maintain. Temperatures are read on an ordinary 0" to 110" C thermometer, inserted in an oil-filled pocket drilled into the base. Two thermometers diagonally situated give identical temperatures ; these comply with the suggestion of Hayes3 and show agreement with the surface temperature of the interior of the oven.Quadruplicate drying tests with cane molasses agree to within less than &0-05 per cent. loss, and with sucrose solution to less than &0.02 per cent. The oven is designed to operate efficiently from 50" to 90" C. For drying at atmospheric pressure the temperature may be raised to 105" C as long as the air flow is sufficient to prevent rusting. As an additional precaution the exposed metal surfaces can be phosphate treated. Once an efficient oven has been constructed there remain other problems in the deter- mination of water in thermally sensitive materials. Choice of absorbent, the effect of extra water, sample to absorbent ratio, particle size, deliberate formation of degradation products are all being investigated. Sugar syrups and molasses containing fructose and organic non-sugars require especial care. Tests are being made in drying golden syrup and molasses, and the results are being utilised to find a correction for refractometer solids based on ash content, after correction has been made for the effect of invert sugar. The refractometer is so useful in the sugar industry that the oven recommended here is not intended to replace it for routine use, especially as the time required for drying is too long for this purpose, although it is tonsiderablyshorter than has been needed for other vacuum ovens of less suitable design. The results of the work, particularly on cane molasses, will be dealt with in further papers. Two methods of temperature control were attempted. REFERENCES 1. 2. 3. Hayes, F. W., in Proc. I.C.U.M.S.A., 1949, 42. Iles, G., and Sharman, C. F., J. SOC. Chern. Ind.. 1949, 68, 174. Fryd, C. F. H., and Kiff, P. R., Analyst, 1951, 76, 25. TATE AND LYLE RESEARCH LABORATORY KESTON, KENT WESTERHAM ROAD A p d 27th, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800709

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

712 HUTT, WEATHERALL AXD CULSHAW : THE DETERMINATION OF [Vol. 78 The Determination of Total Phosphatide in Commercial Lecithin BY H. H. HUTT, H. WEATHERALL AND T. CULSHAW In a previous publication it has been suggested that the difference between acetone-insoluble matter and petrol-insoluble matter can be made the basis of a means for assessing the phosphatide content of commercial lecithin for works-control purposes. A modified method, which has improved precision, is now described for the determination of acetone-insoluble matter; it is compared experimentally with previously published methods. The ratio of phosphatide to non-phosphatide removed at various stages of washing is examined. The difference between acetone-insoluble matter determined by the method proposed and petrol-insoluble matter is suggested as giving a suitable measure of phosphatide content.TWO general methods for the determination of the total phosphatide content of commercial lecithin1 are (i) by determination of the phosphorus content, the found value being multiplied by an arbitrary factor (usually 26.31 for commercial purposes) and (ii) by determination of the difference between acetone-insoluble content and the amount of matter insoluble in petroleum spirit of boiling range 40" to 60" C. The rapid works-control method for the determination of acetone-insoluble matter by direct elution previously describedl was based mainly upon experience with commercial ground-nut lecithin. Examination of a large number of specimens from various botanical sources has subsequently afforded us the opportunity of modifying the method and of making it more acceptable as a means of comparatively assessing the total phosphatide content of all types of material, when this is defined as acetone-insoluble content minus petrol-insoluble content.Results recorded in the literature for phosphatides of different botanical origins show that the phosphorus contents of carefully prepared petrol-soluble acetone-insoluble materials are lower than expected for di-oleyl glycerophosphoryl choline (3.94 per cent.) or di-oleyl glycerophosphoryl ethanolamine (4.16 per cent .) . Typical specimens of total phosphatide mixtures from soya-bean lecithin have been reported2 as containing 3.4 per cent. of phosphorus, and from cotton-seed le~ithin,~ 2.9 per cent.of phosphorus, and we have found 3.0 per cent. of phosphorus in total phosphatides of ground-nut lecithin prepared in the manner indicated. The lower phosphorus content found in the oil-free preparations from the naturally occurring phosphatides, as compared with the theoretical contents of the chemical compounds mentioned above, is caused mainly by combined and closely associated carbohydrate^.^^^ 7 4 9 5 As these carbohydrates appear always to be present in the natural substances and as it is probable that their hydroxyl groups have a function in most of the industrial purposes for which commercial phosphatides are used, it seems permissible to include them in any com- parative assessments of total-phosphatides. They in no way constitute any deleterious addition to, or indicate any incomplete purification of, the material under examination.Moreover, phosphorus determination is a lengthy and somewhat expensive procedure. We therefore suggest that a method that will give reasonably reproducible results for matter insoluble in acetone can be used as a basis for assessing total phosphatide. Of methods proposed for this determination, other than the one already mentioned, that of dissolving the material in light petroleum and precipitating and eluting with acetone6 is probably the most widely used. The comparative merits of this method and the proposed modified method are reported upon below. Error in the method formerly proposed for production control analysis1 was largely caused by entrainment of insoluble matter during decantation-this was particularly noticeable with cotton-seed lecithin-and by the solubility of phosphatide in the oil-acetone solution on the first wash-observed especially with soya-bean lecithin.Preliminary centrifugation eliminates the error caused by entrained material and the solubility error can be minimised by first washing the sample in the cold.Dec., 19531 TOTAL PHOSPHATIDE I N COMMEKCIAL LECITHIN 713 METHOD PROCEDURE- Weigh accurately 3.0 * 0.5 g of sample into a centrifuge tube and add 40 ml of acetone. Stir thoroughly, kneading the lecithin in the acetone, and then setting the tube aside for 10 minutes at room temperature. Centrifuge for 3 to 5 minutes at a relative centrifugal force of approximately l O O O g and decant the supernatant liquid.Extract twice more with boiling acetone and cool to 0" C before centrifuging. For soya-bean lecithin, only one of these further extractions is required. After completion of the washing, remove the bulk of the residual acetone from the insoluble material on a water-bath in a current of air and dry the residue in a vacuum oven at 100" C and at a pressure of approximately 40 cm of mercury. EXPERIMENTAL Acetone conforming to analytical reagent standards was used for all determinations. Table I shows a comparison of results by the earlier methodl with those by the same procedure corrected for matter recovered from the decanted solutions by centrifugation and with those found by the proposed modified method. TABLE I ACETONE-INSOLUBLE CONTENTS OF COMMERCIAL LECITHINS Type of lecithin Unrefined ground-nut ..Refined ground-nut . . .. Soya bean .. .. .. Cotton-seed . . * . . I Acetone-insoluble material r A I Former method,l % .. .. 50-7 43-6 48.3 49.9 53.6 .. .. 63.4 66-1 66.4 66.0 67.5 .. .. 53.9 61.5 54.3 45.2 23.5 .. .. 49.8 49.4 51.3 47.6 47.3 Former method' corrected for material recovered by centrifugation, 53-4 47-0 50.3 53.0 56.2 65.1 67.6 68.2 66.3 68.3 56-2 63.5 62.6 62.6 33.5 52.5 51.2 54.7 (52.6 (50.3 % Proposed method, 53.5 47.7 50.7 54.6 56.8 65.6 68.4 67.8 67.5 69.1 61.0 66.6 66.2 66.1 35.1 53-6 52.5 55.4 52.2 51-2 % NOTE-Results are the means of duplicate determinations. Table I1 compares results obtained by this modified elution method with those obtained by the method of precipitation and elution.6 TABLE I1 COMPOSITION OF COMMERCIAL LECITHIKS I Lecithin Unrefined ground-nut (L 3) .. Refined ground-nut (L 49) Bleached soya bean (D 41) . . Unbleached soya bean (D 84). . Cotton-seed (L 50) . . .. Acetone-insoluble content* Proposed A.O.C.S. method, rnethod,s % % 57.8, 57.9 59.4, 60.0 69.8, 69.9 71.1, 71.6 66.1, 66-2 65-5, 65.7 66.2, 65.9 67.0, 66.6 56.7, 56-8 60-9, 60.8 A Benzene- insoluble Moisture, content, % % 2.50 0-29 0.73 0.39 1.30 trace 0.32 trace 0.58 0.14 Petrol- insoluble content, 0-47 0.76 trace trace 0.26 % * No correction was made for petrol-insoluble or benzene-insoluble matter.7 14 HUTT, WEATHERALL AND CULSHAW : THE DETERMINATION OF ivoi. 78 Phosphatides, which are known to be slightly soluble in acetone in the presence of fatty acids,' have been found to be slightly soluble in pure acetone (see Table IV).To assess and compare the solubility errors, acetone-soluble matter was weighed separately in individual washes (see Table 111) obtained by the proposed method and the A.O.C.S. method, phosphorus being determined in these individual extracts (see Table IV). TABLE I n REMOVAL OF ACETONE-SOLUBLE MATTER BY SUCCESSIVE WASHING Amount of acetone-soluble material removed by wash number A 7 Lecithin Method 1, 2, 3, 4, 5, 6, 7, 8, % Yo % Yo % % Yo % Unrefined ground-nut Proposed 29.8 7.83 2.50 1.00 0.76 0.46 0.54 0.62 (L 3) {A.O.C.S. 33.3 4.46 1.26 0-81 0.73 0.42 0.43 0.52 Refined ground-nut Bleached soya bean Unbleached soya bean Cottonseed (L 50) Proposed 22-7 5.30 1-45 1-14 Proposed 27.8 6.27 2.04 1.20 1-10 1-09 0.78 0.75 A.O.C.S.29.7 3.49 1-29 1-01 0.92 0.90 0.60 0.85 Proposed 27-8 6-59 2-38 1-36 Proposed 32.1 8-05 2.12 1.36 A.O.C.S. 24.5 3.40 0.94 - (L 49) (D 41) (D 84) A.O.C.S. 30.2 2.69 1.48 - A.O.C.S. 34.5 4.00 1-05 - TABLE IV PHOSPHATIDE IN ACETOKE-SOLUBLE MATTER REMOVED BY SUCCESSIVE WASHING Amount of phosphatide* removed by wash number Lecithin I A Method 1, 2, 3, 4, 5, 6, 7, 8,' % % % Yo % % y!! % Proposed 0-40 0.35 0-22 0.21 0.20 0.19 0-20 0.23 A.O.C.S. 0.48 0.16 0.11 0.32 0.32 0.15 0-18 0.23 Proposed 0.41 0.36 0.26 0.26 Proposed 0.69 0.87 0-60 0.37' 0.40 0.45 0.37 0.36 A.O.C.S. 0.66 0-74 0.43 0.37 0.38 0.40 0.28 0.43 Proposed 0.74 1-27 0-91 0.68 Unrefined ground-nut Refined ground-nut Bleached soya bean Unbleached soya bean (L 3) (L 49) (D 41) (D 84) A.O.C.S.0.29 0.19 0.15 - { { A.O.C.S. 1.31 0.35 0.32 - Proposed 0.43 0.48 0.29 0.40 A.O.C.S. 0.69 0.16 0.09 - Cotton-seed (L 50) * Calculated by multiplying the figurc found for phosphorus by the factor 26-31. For the phosphorus determinations, the acetone-soluble matter was saponified and ignited to ash; the ash was dissolved in water and the phosphorus measured on a Spekker absorptiometer after development of a blue colour by an adaptation of Zinzadze's m e t h ~ d , * ~ ~ J ~ by means of molybdic anhydride reduced with molybdenum metal. This was calculated by multiplying the figure for phosphorus by the arbitrary factor 26.31, which is at present used in the commercial assessment of phosphatide on the basis of phosphorus determination. As pointed out earlier, the factor necessary for converting phosphorus to phosphatide would be higher than this for the total mixed phosphatides and it is probably much higher for phosphatides removed in the fourth and subsequent washes.Nevertheless, this variation does not greatly alter the interpretation of the figures. A more important consideration is whether or not the phosphorus-containing material removed during the acetone washing is definitely phosphatide. It is evident that it is because (i) the phosphorus content of the acetone washes can be suppressed by using acetone previously saturated with oil-free phosphatide (the addition being allowed for), (ii) the material removed in the later washes cannot be redissolved in a smaller volume of acetone and (iii) if larger amounts of acetone are used, correspondingly larger amounts of phosphorus-containing material are dissolved.Further, as only minor amounts of water-soluble phosphorus were The results shown in Table IV are expressed as the percentage of phosphatide.Dec., 19531 TOTAL PHOSPHATIDE I N COMMERCIAL LECITHIN 715 present in the acetone washes, the acetone-soluble phosphorus derivatives could not be free phosphoric or glycerophosphoric acids. There does not appear to be any other suitable fat solvent in which the phosphatides are less soluble generally than acetone. In attempts to find ways of suppressing the existing solubility, other than by temperature control, it was found that acetone saturated with phosphatide or magnesium chloride gives a lowering that is insufficient to warrant its use.Examination of the existing techniques was therefore continued. To assess the balance between phosphatide removed from the sample by acetone and oleaginous non-phosphatide left in, at given stages of the washing, the calculations summarised in Table V were made. BALANCE OF ERRORS IN Lecithin Unrefined ground-nut (L 3) Unrefined (L 3) (L 49) (L 49) (D 41) (D 41) (U 84) (D 84) (L 50) (L 50) ground-nut Refined ground-nut Refined ground-nut Meached soya bean Bleached soya bean soya bean soya bean 7Jnbleached Unbleached Cotton-seed Cotton-seed Mcthod Proposed A.O.C.S. Proposed A.O.C.S. Proposed A.O.C.S. Proposed A.O.C.S. Proposed A.O.C.S. Factor * I I1 I11 I I1 I11 I I1 I11 I I1 111 I I1 111 I I1 I11 I I1 111 P I1 111 I 11 I11 I I1 I11 TABLE V ESTIMATING ACETONE-INSOLUBLE MATERIAL 7 *I Wash number \ 2, % 0.40 7-48 + 7.08 0.48 4.3 1 + 3-83 0.4 1 4.94 -t 4-53 0.29 3.21 -I- 2.92 0-69 5.40 0.66 2.75 + 2-09 0-74 5.32 + 4.58 1-3 1 2-34 -+ 1-03 0-43 7-57 + 6.94 0.69 3.84 t 3 .1 5 $- 4.7 1 3, YO 0.75 2.28 + 1-53 0.63 1-15 + 0-52 0.77 1-20 + 0.83 0.48 0.79 +0*31 1-66 1.44 -0.12 1.40 0.86 - 0.54 2.01 1.47 - 0.54 1-66 1-16 - 0.50 0.91 1.83 + 0-92 0-85 0.96 + O - l l 4, % 0.97 0.79 0.74 0.49 - 0.25 1.02 0.88 -0.14 -0.18 2.16 0.83 - 1.33 1-83 0.64 - 1.19 2-92 0.68 - 2.24 1.20 0.96 - 0.24 5, 6, % Y O 1.18 1.38 0.56 0.27 1.06 1.38 0.41 0.27 -0.62 -1.11 -0.65 -1.11 2.53 2-93 0.70 0.64 2.20 2.58 0.54 0.50 - 1.83 -2.29 -1.66 -2.08 7, 8, % % 1-57 1.77 0.34 0.39 1.53 1-71 0.25 0.29 -1.23 -1.38 -1.28 -1.42 3-38 3.75 0-41 0.39 2-98 3.16 0-32 0.42 -2.97 -3.36 -2.66 -2.74 * I : Total phosphatide removed by preceding washes. 11 : Non-phosphatide removed by wash enumerated.111: Difference between I and 11, after previous wash. DISCUSSION OF RESULTS The errors in these calculations, which are caused by using the factor 26.31 for converting phosphorus to phosphatide, are small compared with the amounts involved and, further, their direction is such as to contribute to the arguments in their favour. This being so, certain useful deductions can be made. The most important of these is that the least error is incurred in either method by washing no further after the second operation for soya-bean lecithin and after the third for the other lecithins. If a higher conversion factor were used, the estimates of phosphatide removed (I) would be higher and of non-phosphatide removed (11) correspondingly lower.The increased dis- parity would be such as to require an earlier discontinuation of the washing. The diminishing amounts of oleaginous non-phosphatide removed, however, suggest that this would be undesirable. The persistent removal of what appeared to be non-phosphatide in the fourth and sub- sequent washes is probably caused by a different type of phosphatide with a lower phosphorus content, but this, at present, is not certain. If, however, these substances were glycerides, the amount removed would diminish at a greater rate.716 HUTT, WEATHERALL AND CULSHAW [Vol. ‘is As already shown in Table I, centrifugation reduces the entrainment almost completely.It is also evident that a cold initial wash reduces the amount of phosphatide dissolved in the oil - acetone solution. Statistical treatment of the duplicate analyses, of which averages are given in Table I, shows that the standard deviation of the modified method is significantly lower than that of the former method (see Table VI). From a limited number of results (five pairs, representative of different types of lecithiii- see Table 11) it seems that the standard deviation (0.114) of the proposed method, calculated on the same range of duplicates, is significantly lower than that (0.286) of the elution method. The improved precision in measuring acetone-insoluble matter is still reflected in the ultimate estimation of the difference between acetone-insoluble matter and petrol-insoluble matter when the standard deviation of the method of determining the latter (0.0186) is taken into account.The standard deviation of the difference is 0.164 for the proposed method compared with 0.321 for the earlier method,l even when the results for soya bean lecithin, which has been shown to have a significantly higher variance than other types, are excluded. TABLE VI PRECISION OF METHODS OF DETERMINING ACETONE-INSOLUBLE MATTER Type of lecithin Standard deviation r > Former methodl Proposed method A Unrefined ground-nut . . .. * . 0-366 Refined ground-nut . . . . .. 0-265 Soya bean . . .. .. .. 0.975 Cotton-seed . . .. .. .. 0.326 .. 0.321 * (Over-all) . . .. .. .. 0.134 0.190 0.118 0.195 0.163 * Excluding soya-bean lecithin, for which variance was shown to be significantly greater than it was for other lecithins.In the light of the results presented and discussed, we suggest that the difference between acetone-insoluble matter and petrol-insoluble matter gives an adequate measure of the phosphatide content of commercial lecithin, acetone-insoluble matter being determined by the method described in this paper and petrol-insoluble matter being determined as described previously.1 Other criteria may need to be referred to occasionally, and, as the modified method gives higher results than the earlier procedure, standards will need to be revised for given qualities, when those standards have been based upon acetone-insoluble content. We are indebted to other colleagues of our Analytical Department for considerable assistance, to Mr. J. W. Lord for statistical analysis of results and to J. Bibby & Sons, Ltd. for permission to publish this work. REFERENCES 1. 2. 3. Olcott, H. S., Science, 1944, 100, 226. 4. 5, 6. 7. Linteris, L., and Handschumaker, E., Ibid., 1950, 27, 260. 8. Zinzadze, C.. Ind. Eng. Chern., Anal. Ed., 1935, 7, 227. 9. Gerritz, H. W., J . Ass. Off. Agric. Chew.., 1940, 23, 321. 10. Schricker, J. A., and Dawson, P. R., Ibid., 1939, 22, 167. Hutt, H. H., and Weatherall, H., Analyst, 1944, 69, 39. Scholfield, C. R., Dutton, H. J., Tanner, F. W., jun., and Cowan, J. C., J . Amer. Oil Chern. SOL, Hutt, H. H., Malkin, T., Poole, A. G., and Watt, P. K., Nature, 1950, 165, 314. Scholfield, C. K., Dutton, H. J., and Dimler, R. J., J . Amer. Oil Chem. SOL, 1952, 29, 293. American Oil Chemists Society’s Committee on Analysis of Commercial Fats and Oils, Ibid., 1948, 25, 368. 1947, 24, 7 7 . J. BIBBY AND SONS LIMITED KING EDWARD STREET LIVERPOOL, 3 June 16th, 1053

ISSN:0003-2654    

DOI:10.1039/AN9537800712

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Dec., 19531 SILVERMAN AXD TREGO 717 The Determination of Small Amounts of Potassium, Calcium and Magnesium in Sodium and its Compounds BY LOUIS SILVERMAN AND K. TKEGO Preliminary concentration is necessary before potassium can be deter- mined in the presence of large amounts of sodium ion. Metallic sodium, sodium hydroxide and sodium salts are converted to chloride. Measured, cold, saturated aqueous chloride solutions are treated with dry hydrogen chloride gas, which separates the bulk of the sodium chloride. The residual potassium, calcium and magnesium can then be deter- mined by the usual methods. Calcium and magnesium may also be determined with or without preliminary sodium chloride separation. THE potassium content of commercial sodium metal is about 0.005 per cent. This is below the limits for direct quantitative determination of potassium in sodium1 y 2 j 3 y 4 and necessitates preliminary separation of sodium or concentration of potassium.In the preliminary separation as sodium chloride, the initial weight of test sample is reduced from 5 to 50 g of sodium chloride to 0.2 to 0.5 g of concentrate, containing all of the potassium, calcium and magnesium with incidental sodium chloride. This concentrate may contain 0.5 mg or more of potassium perchlorate, which can be conveniently determined by the ethyl acetate extraction pr~cedure.~ This paper describes the precipitation of the bulk of the sodium chloride from aqueous solution by hydrogen chloride gas5y6 and results are presented to demonstrate the effectiveness of the separation of the potassium from the bulk of the sodium chloride.Potassium (as well as calcium and magnesium) is not retained by the sodium chloride precipitate. A method for the direct determination of calcium7y8 in sodium chloride has been reported, but none has been described for magnesium.9 These direct determinations may be used as check methods against the pre-concentration procedure for the determination of calcium and magnesium. METHOD KE AGENTS- of glacial acetic acid in a beaker. solution and dilute to 200 ml with water. of ethyl or methyl alcohol. Acetic acid solution of 8-hydroxyqzcinoline-Place 5 g of 8-hydroxyquinoline and 10 ml Warm to 50" to 60" C for several minutes to complete Alcoholic solution of 8-hydroxyquinoline-Dissolve 5 g of 8-hydroxyquinoline in 200 ml Ammonium oxalate solution-Dissolve 10 g of the salt in 100 ml of water.Saturated hydrochloric acid solution-Saturate hydrochloric acid, sp.gr. 1.2, with dry Wash water, pH 10-Add ammonium hydroxide, sp.gr. 0.880, to distilled water until Discard after 2 weeks. Discard after 2 weeks. hydrogen chloride gas, at 0°C. the pH is 10. PROCEDURE FOR DIRECT DETERMINATION OF POTASSIUM- Open the container and transfer it and the sample to a weighed glass beaker. Treat the sample with discrete portions of anhydrous methyl or ethyl alcohol until the metallic sodium has been converted to the alcoholate. Small portions of water may be added cautiously to accelerate the action. Remove the container and wash, dry and weigh it in order to find the amount of metallic sodium taken.Add water to the alcoholate solution to increase its volume by a quarter and mix thoroughly. The sodium content if as low as 1 to 3 g may also be determined at this point Metallic sodium-Weigh the sodium metal plus container.718 SILVERMAN AND TREGO: THE DETERMINATION OF SMALL [Vol. 78 by titration with N hydrochloric acid solution. Add concentrated hydrochloric acid to the alcoholate solution until the solution turns litmus paper red, and add about 5 ml of hydro- chloric acid in excess. Evaporate the solution nearly to dryness, then bake the residue for 4 to 12 hours in an oven at 110" to 130" C. Cool and weigh. The gain in weight is equivalent to the amount of sodium chloride. This weight of sodium chloride may be used to calculate or check the weight of metallic sodium taken for analysis.Divide the weight of sodium chloride in grains by 0.35 (the solubility of sodium chloride in water is 35 per cent.) and add this calculated amount of water to the beaker containing the sodium chloride. Warm the solution to dissolve the salt completely. Cool the beaker in the refrigerator to 0" C and weigh the beaker and contents cold. Sodium hydroxide-Weigh the sodium hydroxide and transfer the sample to a beaker, weighed to the nearest tenth of a gram, containing five times as much water as sodium hydroxide by weight. Add a measured amount of hydrochloric acid to make the solution just acid. Cool. Calculate the weight of sodium chloride formed. Evaporate the solution until the beaker contains a saturated solution of sodium chloride.Sodium chloride-Weigh the sodium chloride, transfer the sample to a weighed beaker and add sufficient water to make a saturated solution of the sodium chloride. Separation of salt-Pass dry hydrogen chloride gas through a train into the cold solution. If a tank of gas is available, successively pass the gas through a wash bottle of sulphuric acid, a safety bottle and an inverted funnel or thistle tube into the aqueous sodium chloride solution. The funnel or thistle tube should be well below the surface of the salt solution. Set the cold sodium chloride solution in an ice-bath and maintain the ice in the bath throughout the action. Pass in dry hydrogen chloride gas slowly at first, then increase the flow of gas until bubbles pass through the sodium chloride solution.If the amount of sample is large the salt solution may become hot, but it will cool when the absorption of gas is complete. The volume will increase by about 30 per cent. When the absorption seems complete and the liquid is fuming, remove the beaker from the system and weigh it. The action is complete when the salt solution has increased in weight by 47 to 51 per cent.-roughly, twice the weight of the dry sodium chloride. Set the covered beaker in the refrigerator until it is ready for filtration. Decant the cold fuming solution through a cold dry 500-ml sintered-glass crucible, transfer the solid to the crucible and wash it with 20 ml of cold hydrochloric acid saturated with hydrogen chloride. Evaporate the filtrate to a small volume, transfer the solution to a 150-ml beaker weighed to the nearest hundredth of a gram and evaporate the liquid nearly to dryness; dry the residue in an oven at 110" to 130" C.Cool and weigh. The gain in weight (about 0.2 g) represents sample concentrate, and the amount of ethyl acetate to be used will depend on this weight. Separation of potassium from sodium-Add 5 ml of nitric acid, sp.gr. 1.4, and 5 ml of 72 per cent. perchloric acid and evaporate nearly to dryness, leaving only sufficient perchloric acid to prevent formation of insoluble ferric oxide. Add 6ml of anhydrous ethyl acetate (5 ml will dissolve the equivalent of 0.3 g of sodium chloride) and heat. Complete the potassium determination by the procedure of Willard and Smith.4 Discard the residue. Cool to room temperature.PROCEDURE FOR SEQUENCE DETERMINATION OF CALCIUM, MAGNESIUM AND POTASSIUM- Prelimi9zary separation from sodium-Proceed as described for direct potassium determina- tion with sodium metal, sodium hydroxide or sodium chloride to the point where the dried sample concentrate would be treated with nitric and perchloric acids. Instead add 50 ml of water to the 150-ml beaker containing the sample concentrate of sodium chloride. Warm to dissolve the salts. Separation of calcium and magnesium from potassium and sodium-Add 15 ml of the alcoholic solution of 8-hydroxyquinoline and warm to 60" to 80" C. Slowly add ammonium hydroxide, sp.gr. 0.880, until the pH is raised to between 9.5 and 10.5. A pH meter or thymol- phthalein indicator may be used. Heat to incipient boiling and then cool to room temperature.Check the pH reading, adding ammonium hydroxide, if necessary. Filter the solution through a 12.5-cm Whatman No. 40 filter-paper into a 250-ml beaker. Wash the beaker with an aqueous solution of ammonium hydroxide adjusted to a p1-I value of 10. Wash the precipitate and paper several times with the wash-water. Reserve the filtrate for potassium determination.*Dec., 19531 AMOUNTS OF POTASSIUM, CALCIUM AND MAGNESIUM 519 Calcium determination-Return the paper and its contents to the beaker and add 15 ml of nitric acid, sp.gr. 1.4, and 8ml of 72 per cent. perchloric acid. Mix and evaporate the solution to heavy fumes of perchloric acid. Dilute to 50 ml with water and boil for 1 minute. Precipitate the calcium as oxalate, filter, reserve the filtrate for magnesium determination, and complete the calcium determination by titrating with 0.02 N potassium permanganate solution in the usual manner.Magnesium determination-Heat the filtrate from the calcium oxalate precipitation to boiling and slowly add 15 ml of the solution of 8-hydroxyquinoline in acetic acid. If iron is present a black precipitate forms and the supernatant liquid should be yellow. Additional reagent is added, if necessary, to make the solution yellow. Stir the solution until the black precipitate coagulates, then filter off the iron precipitate on a Whatman No. 40 filter-paper containing paper pulp and collect the filtrate in a 400-ml beaker. Do not allow the black particles to creep above the paper. Wash the paper and precipitate with 0.2 per cent.v/v acetic acid solution. If the solution is not yellow, add 5 to 10ml of alcoholic 8-hydroxyquinoline reagent to the filtrate and then warm to 60" to 80" C. Complete the magnesium determination by the procedure of Bergg Potassium detevmination---To the reserved potassium solution, add 15 ml of nitric acid, sp.gr. 1.4, and 8 ml of 72 per cent. perchioric acid, evaporate to heavy fumes of perchloric acid and continue heating until nearly all the perchloric acid has been evaporated. Heat the beaker over an uncovered bunsen burner flame to volatilise ammonium salts. Complete the potassium determination by the procedure of Willard and Smith.4 PROCEDURE FOR DIRECT DETERMINATION OF APPRECIABLE AMOUNTS OF CALCIUM AND Convert the metallic sodium, sodium hydroxide or other sodium samples to sodium chloride. ,4djust the volume of solution so that the salt content is 20 to 25 per cent. and proceed according to the section on separating calcium and magnesium from potassium and sodium (p.718). DISCUSSION OF RESULTS The first column shows the weight of sodium chloride (weighed or calculated), the second the Cool. If the solution appears cloudy, filter, wash and discard the residue. Discard the paper and the iron precipitate. MAGNESIUM~- The effective separation of the bulk of the sodium chloride is shown in Table I. TABLE I SEPAHATION OF SODIUM CHLORIDE BY HYDROGEN CHLORIDE GAS NaC1, g 38.8 47.8 69-7 116.5 117.2 122.3 149.9 69.45 61.8 22-45 57.35 10.75 4.85 9.75 84.95 56.65 70-5 54.2 44-65 40.0* 40*0* 40*0* Water added, HCl absorbed, ml g 110.9 37-8 136.8 80.5 194.3 57.8 329.0 62.1 334-7 180.5 349.4 240.0 428.3 285.0 198.4 138-3 176.7 112.8 64.1 40.5 163-9 105.3 30-7 19-2 13.9 9.0 27.9 18.1 242.7 156.3 161.9 100.0 201.4 135-4 154-9 98.5 127-6 87-4 114-0 72-2 114.0 74.1 114.0 71-9 * Synthetic sample.Gain in weight from HCl, 25 44 22 14 40 51 60 52 48 47 48 47 48 48 48 46 50 47 51 47 48 47 % Weight of sample concentrate, g 0.80 0.70 0.50 14.8 3.3 0.45 0.62 0.24 0.22 0.12 0-20 0.10 0.09 0.05 0.30 0.24 0.19 0.18 0.15 0.15 0.15 0.14720 SILVERMAN AND TREGO: THE DETERMINATIOS OF SMALL [Vol. 78 amount of water added and the third the weight of hydrogen chloride gas absorbed. The fourth column shows the percentage gain in weight of the solution and the fifth shows that all but a small amount of salt is precipitated from solution.It is evident from Table I that a sample weight of 15 to 20 g of sodium metal (40 g of sodium chloride) can be used, and that all but about 0.2 g of salt can be precipitated from solution when the salt solution has absorbed nearly 50 per cent. of its weight of hydrogen chloride. For the results in Table 11, a 40-g sample of sodium chloride was used, and blank deter- minations of potassium, calcium and magnesium were made. After this, known amounts of the three elements were added to 40-g samples of the same batch of C.P. sodium chloride, and the three elements were determined. It should be noted that the C.P. sodium chloride sample itself contains 0.002 per cent. of potassium, 0.0002 per cent.of calcium and 0-00016 per cent. of magnesium, and these are the blank values that must be subtracted, as mentioned in a footnote to Table 11. TABLE I1 SYNTHETIC SOLUTIONS OF POTASSIUM, CALCIUM AND MAGXESIUM I N 40 g OF SODIUM CHLORIDE By preliminary hydrogen chloride separation Potassium Calcium Magnesium & & 1 mg mg % mg mg O ' / O mg mg % I A 1 Elements added Added, Found, Added, Found, Added, Found, None. C.P. sample - 0.80 0.0020 - 0.082 0.00020 - 0.065 0.00016 used for blank values { -- 0-84* 0.0021 - 0.10 0.00025 - K 0.3 0.37 0.00075 K, Ca 0.30 0.40*1 0.0010 0.40 0.48t 0.0012 Ca 0.30 0.257 0.0006 Ca, Mg 0.40 0.367 0.0009 Ca, Mg 0.40 0.43t 0.0011 By direct 8-hydroxyquinoline separation r L -l Calcium Magnesium r L \ f A 3 Elements added Added, Found, Added, Found, mg mg % mg mg % None.C.P. sample - 0.08 0.0002 - 0.05 0.000 13 used for blank values K K, Ca Ca Ca, Mg 0.20 0.207 0.0005 0-40 0.36f. 0~0009 Ca, Mg 0.40 0.35t 0~0009 0.40 0-36f. 0.0009 * No potassium found in re-worked precipitate. t Corrected for blank values. The solubility of the potassium, calcium and magnesium in the mixture of cold saturated hydrochloric acid and sodium chloride solution is important. The calcium and magnesium form anionic complexes that are soluble. Only the potassium has limited solubility, but in the technique described no error need be expected. For example, consider the synthetic sample in Table I. For 40 g of sodium chloride, 114 ml of water are used, and after the addition of hydrogen chloride the volume has increased by 30 ml.From this precipitation 0.15 g of sample concentrate was left in solution. From the point of view of the common ion effect, this residuum could be divided as 80 per cent. of sodium and 20 per cent. of potassium, ie., 30 mg of potassium chloride (15 mg of potassium) would remain in solution. Since the obvious application of this analytical technique is to separate small amounts, or at the most 5 mg, of potassium from large amounts of sodium, the technique is well-founded. This is shown experimentally in Table 11. Here, the precipitated sodium chloride was filtered from the sample concentrate, dissolved in water, precipitated as before with hydrogen chloride gas and a second sample concentrate was collected. No potassium was found in this second It is also necessary to prove that no potassium is occluded by sodium chloride.k c ., 19531 AMOUNTS OF POTASSIUM, CALCIUM AND MAGNESIUM 721 sample concentrate, and it is then stated that no potassium is found in the reworked precipitate. TABLE I11 DETERMINATION OF POTASSIUM IN SODIUM METAL Number of experiment * 1 1 1 2 2 3 3 3 4 4 5 5 5 Type of metal As received As received As received Distilledt Distilledf Treated: Treated:: Treated: Treated1 Treated: Treated 1 Treated $ Treated$ Amount of metal, g 33.4 39.8 31-5 15-7 17.2 23-75 21-95 17-7 29.8 29.4 38-8 24.5 22-4 Potassium found, mg 2.9 4.7 7.0 1.4 1.4 0.72 0.41 0.51 0.68 1.53 0.3 0.2 0.6 Potassium, 0.009 0.012 0.022 0.009 0.008 0.003 0.002 0.003 0.002 0.004 0.001 0.001 0.003 % * Samples with the same number are considered as nearly duplicate samples. tloftness, Ruebsamen and Coultas.’O $ Ruebsamen.ll This procedure has particular application to the determination of small changes of potassium metal content in treated and untreated sodium metal (see Table 111).Samples received from the manufacturer contain 0.01 to 0-02 per cent. of potassium. Vacuum distillation reduces the figure, but treatment with graphite chips4 reduces the amount of potassium to about 0.001 per cent. The sets of treated samples are considered as duplicates from the same operation. To check the results for calcium and magnesium the method described for direct 8-hydroxyquinoline separation (p. 719) was used, as shown in Table 11. It was thought that the results obtained from the preliminary hydrogen chloride separation (p. 718) would be higher, but this was not so. This report is based on studies conducted for the Atomic Energy Commission under contract AT-40--l-GEN-1064. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Adams, M. F., and St. John, J. L., Ind. Eng. Chem., Anal. Ed., 1945, 17, 435. Adie, R. H., and Wood, T. B., J. Chem. SOC., 1900, 77, 1076. Kolthoff, I. M., and Bendix, G. H., Ind. Eng. Chem., Anal. Ed., 1939, 11, 94. Willard, H. H., and Smith, G. F., J. Amer. Chem. Soc., 1922, 44, 2819; 1923, 45, 286. Mellor, J. W., “Inorganic and Theoretical Chemistry,” Volume 2, Longmans, Green & Co. Ltd., Seidell, A., “Solubilities,” Volume I, D. Van Nostrand Co. Inc., New York, 1940. Berg, R. , “Die andytische Verwendung uon O-OxyqinoZine und seinev Deviuate,” Ferdinand Enke, Rynasiewicz, J., and Polley, M. E., Anal. Chem., 1949, 21, 1398. Berg, R., op. cit., p. 30. Loftness, R. L., Ruebsamen, W. C., and Coultas, T., NAA-SR-126, November 20th. 1951. Ruebsamen, W. C., “The Removal of Small Amounts of Potassium from Sodium,” NAA-SR-139, London, 1946, p. 542. Stuttgart, Germany, Second Edition, 1938, p. 40. July 27th, 1951. ATOMIC ENERGY RESEARCH DEPARTMENT NORTH AMERICAN AVIATION INC. DOWNEY, CALIFORNIA, U.S.A. First submitted, August 25th, 1952 Amended, June 22nd, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800717

出版商:RSC    

年代:1953

数据来源: RSC