ISSN:0003-2654    

DOI:10.1039/AN95378FX005

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BX007

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BP013

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

FEBRUARY, 1953 Vol. 78, No. 923 THE ANALYST EDITORIAL A REPRINT OF THE ANALYST FROM 1876 TO 1951 FOR some years past the demand for back numbers of The Bndyst has been continuous and insistent. The known locations of complete sets are but few; long runs appear but seldom in the second-hand market, and the stock of complete volumes for recent years remaining in the hands of our publishcrs is almost completely exhausted. The reason for this lamentable position is not far to seek; the ever increasing appreciation of the importance of analytical chemistry in industry, public health, agriculture and in human activities in general has, during the last thirty years, led to the installation of many more chemical libraries in works and research laboratories, colleges , schools of chemistry and in private possession.To all of these a file of The AmzZ?ist is a necessity, and all are finding it increasingly difficult to fill their needs. Plans that were in existence, many years ago, for reprinting the early volumes had perforce to be set aside on thc outbreak of war; since 1945, and until recently, it has been impossible to obtain supplies of paper in excess of current requirements, or to plan ahead in face of ever rising costs. It is therefore with no little pleasure that wc are now able to announce a project for the reprinting of our back numbers up to 1951 and to know that it will be possible for the many gaps in chemical libraries to be filled. On the last page of this issue there will be found a prospectus of the proposed reprinting, in which the 76 volumes to be dealt with have been divided into groups, some of which, for recent years, will also be sold as separate volumes. This grouping of the earlier volumes has been made in order to reduce the cost to subscribers to the lowest possible figure. The reprinting must necessarily depend upon the response to our publisher’s prospectus and order form, supplied with this issue; for it will be easily understood that a work of this magnitude cannot be undertaken unless a reasonable proportion of its cost can be guaranteed in advance. The financial responsibility for this undertaking, which is far beyond the resources of the Society, will be borne by our publishers, W, Heffer gL Sons, to whom all analytical chemists will be grateful for making The A~znlyst more readily available. -9 lo

ISSN:0003-2654    

DOI:10.1039/AN9537800073

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

74 CAMA AND MORTON: SPECTROSCOPIC PROPERTIES OF VITAMIN A,: [VOI. 78 Spectroscopic Properties of Vitamin A2 Application to the Assay of Cod-Liver Oil BY H. R. CAMA AND R. A. MORTON (Presented at the meeting of the Society on Wedrtesday, October lst, 1952) Vitamin A,, made by reducing retinene, with lithium aluminium hydride, shows a main ultra-violet maximum near 361 mp (Eik about 14-00) and a secondary peak near 287mp (E:& about 750). The blue solution (with antimony trichloride reagent) shows Amx. at 693 mp, E:k about 3900. The absorption intensities at different wavelengths and for different solvents have been measured and expressed as fractions of the maximum intensities. The biological potency of vitamin A, is taken to be 1.33 x l@ i.u. per g, that of vitamin A, being 3.33 x 1V i.u.per g. Fish-liver oils m general contain much more vitamin A, than A,, depending on the species. In cod-liver oils vitamin A, may account for about one- seventh of the total vitamin A (molecule for molecule). By determining Et:, at 693 mp (A,) and at 620 mp (A,) in the antimony trichloride colour test (applied to the unsaponifiable fraction), and the E:'X, at 326 mp, 361 mp and 286 m p in the ultra-violet, oils can be tested for both vitamins. The 693-mp absorption measures vitamin A, directly and from it the vitamin-A, contributions to ultra-violet absorption at 351 mp and 327 mp can be calculated. A conversion factor is given for calculating the probable vitamin-A, contribution to the potency. A cod-liver oil typical of those studied by spectrophotometric methods, corrected for all irrelevant absorption, gave an estimated vitamin-A potency about 6-5 per cent.lower than the estimate that included the possible vitamin-A, contribution. VITAMIN A, is detected in fish-liver oils, by means of an absorption band near 693 mp, in the blue solution produced by the interaction of the vitamin and the Carr- Price reagent {anhydrous antimony trichloride in chloroform). In low-potency oils where vitamin A, predominates over A,, the 693-mp band can only be seen when the unsaponifiable fraction is used, as the colour test is substantially inhibited when tests are made on the whole oil.Feb., 19531 API’I-ICATION TO THE ASSAY OF COI>-I,IVER 011, 75 The ultra-violet absorption spectrum shown by most liver oils and unsaponifiable fractions is mainly due to vitamin A, (Amax.at 326 to 328 mp), but the absorption curvc will be distorted to a greater or lesser extent when vitamin A, (Amax. at 351 and at 286 mp) is present. Fish-liver oils and extracts therefrom also exhibit irrelevant absorption from substances either unrelated to vitamins A or derived from them by oxidation. It is a matter of practical TABLE I INTENSITIES OF ABSORPTION EXPRESSED AS FRACTIONS OF Em,,. FOR VITAMIN-A2 ALCOHOL ON DIFFT, 4 KENT SOLVESTS Solvent r---.-..--. -------A ------- Wavelengtli, cycZoHexane isoPropano1 Ethanol Light petroleum mP 250 0.193 0.170 0.163 0.163 260 0.228 0.219 0.209 0.208 266 0.274 0.280 0.272 0,266 270 0.327 0-335 0.322 0,317 275 0-432 0.438 0.435 0.422 280 0-472 0.443 0-435 0.427 2 85 0-529 0.549 0-541 0-630 - - 0.540 286 - 287 0.555 0.549 0.544 - 290 0.529 0.477 0.482 0.482 295 0.4 18 0.39 1 0.393 0.396 297 0-406 0-39 1 0.393 - 300 0.410 0.408 0-408 0.412 305 0.459 0.477 0-477 0-482 310 0.540 0.567 0.560 0.567 315 0.624 0.646 0.638 0.648 320 0.704 0-733 0.728 0.738 325 0-786 0.821 0.8 10 0.819 330 0-864 0.888 0.880 0.890 335 0.927 0.943 0.938 0.944 340 0.964 0-971 0-970 0.9’73 345 0.990 0.994 0.995 0.99 1 -- - 1-000 348 - 350 0.998 1.000 1.000 0.997 352 1.000 355 0.994 0.980 0.988 0.973 360 0.950 0.923 0.935 0-9 18 365 0.900 0.865 0.878 0.86 1 370 0.838 0.797 0.813 0.793 375 0-757 0-700 0-715 0.695 380 0.643 0-579 0.590 0.569 385 0.533 0.478 - 0.470 390 0.448 0.400 0-419 0.380 400 0.286 0-225 0-241 0.220 410 0.111 0.073 0.083 0.072 420 0.035 0.025 0.032 0.023 7 - - - importance to the analyst to be able to determine the vitamin-A, content by studying spectrophotometrically the antimony trichloride colour and then translating the result into the equivalent observed ultra-violet absorption contribution of vitamin A,.By sub- tracting the curve for the vitamin A, moiety from the observed curve (preferablJ? measured on the unsaponifiable fraction) it is possible to assess the extent to which the residual absorp- tion curve is distorted by adventitious interfering substames. At presept the analyst is handicapped by a lack of precise information to enable him to calculate the ultra-violet absorption of vitamin A, from the colour test measurements at 693mp. Vitamin A, has not been isolated as a pure crystalline substance and the best preparations hitherto obtained1 have been rather inadequately characterised. However, retinene,, the aldehyde of vitamin A,, has been obtained crystalline and apparently pure.3 The well-known reducing agent lithium aliiminium hydride (LiAlH,) effects a fairly smooth conversion of retinene, to vitamin A, under appropriate exFerimenta1 conditions.Although the process is not quantitative, the reduced material can be purified by chromatography and the vitamin-A2 fraction can be studied both in respect of its ultra-violet absorption and by the antimony76 CAMA AND MOKTON : SPECTROSCOPIC PROPEliTIES OF VITAMIN A,: [VOl. 78 trichloride colour test. With known values of E:tm at 693 mp for the colour test and of E;$m at 351 mp for the ultra-violet absorption test, it will be possible to calculate the ultra- violet absorption due to vitamin A, at 326 to 328 mp (the maximum for vitamin A,) from the colour test reading. At this stage it is necessary to know the relative intensities of absorption at wavelengths over the range 250 to 400 mp for vitamin A, in different solvents, and the most convenient specification is in terms of Emax.= 1.0 (k, at 350 to 352 mp). EXPERIMENTAL Crystalline retinene, (m.p. 77" to 78" C, E:2m at 385 mp = 1460 in cyclohexane) was used (Cama et ds). Lithium aluminium hydride (0.2 g) was finely ground and dissolved in 20 ml of anhydrous diethyl ether and 0.05 g of retinene, was dissolved in a further 50 ml of anhydrous ether. The reagent solution was added slowly, with stirring, to the retinene, solution protected against light.When the solution Both solutions were cooled to 0°C. TABLE I1 SPECTROSCOPIC PROPERTIES OF VITAMIN A, Workers 1 2 3 3 3 3 3 1 2 3 Solvent Ethanol . . Y t .. Y> .. Light petroleum cycloHexane . . isoPropano1 . . Chloroform . . SKI, colour test > > 9 ) 1. Shantz.1 . . .. .. . . . . . . . . . . .. . . Emax. mP .. 352 287 .. 352 288 .. 351 286 . . 348 256 . . 351.5 287 . . 351 288 . . 356-6 291-5 , . 693 . . 693 . . 693 E a t 620 (no peak) El" 1 cm 1460 820 1330 678 1410 698 1390 750 1320 733 1370 753 1280 704 4100 3700 3870 1580 E a t smaller peak E a t main Inflection, peak None 0.66 None None 0.61 None 277 0.54 (5) 277 0-54 my. % 275 0.55(5) 277 0.55 282 0.55 2. Farrar, Hamlet, Henbest and Jones.3 3.Cama and Morton; work described here. had become colourless (4 to 5 minutes), the excess of lithium aluminium hydride was decom- posed by adding ice-cold water dropwise to the mixture. If the reaction was allowed to continue for more than 5 minutes the mixture turned violet and the yield of vitamin A, fell progressively with time. The mixture was extracted with ether (redistilled over reduced iron) and the combined extracts were washed with water and dried with sodium sulphate. The solvent was removed under reduced pressure. The residue, which weighed 0.03 g and had an E:Tm at 351 mp of 970, was mainly vitamin A,. It was chromatographed on alumina* weakened by stirring in water (10 per cent. w/w) under light petroleum. The light petroleum carried through the column a small amount of material showing A,,,. at 350 and at 285 mp- possibly an isomer of the normal vitamin A,; this was followed by a small fraction showing maxima at 320 and 345 to 350 mp.The main vitamin A, fraction was eluted with a mixture of light petroleum (95 volumes) and ether (5 volumes). By means of a two-fold increase in the proportion of diethyl ether, a little anhydrovitamin A, and some oxidised material was eluted. The main fraction was chromatographed again on a fresh column. The portion carried through by the mixture of liglit petroleum and diethyl ether (95 + 5) seemed to be quite homogeneous (Amax. at 350 and at 285 mp, with a slight inflection at 275 mp). The absorption spectrum of this material was measured in different solvents, The results are shown in Tables I and 11.* Supplied by P. Spence and Co., LM.Feb., 19531 APPLIcxrIoN TO THE ASSAY OF COD-LIVEK OIL 77 DISCUSSION OF RESULTS The best preparations of vitamin A, recorded in the literature show the characteristics recorded in Table 11. The absolute values of E::, obtained in the present work are not particularly stressed because somewhat higher values are not unlikely to be found when vitamin A, can be crystallised. The agreement is nevertheless good and the relative values at 693 and 351 mp are sufficiently trustworthy to meet all practical needs. The absorption curves in the ultra- violet plotted on a scale relative to Em,- = 1.0 agree well with the corresponding curves obtained by Salah4 on Nile fishes; some of these fish yield liver oils showing vitamin A over- whelmingly preponderant in vitamin A,.The small inflection near 275 mp observed by US Wavelength, mp Absorption spectrum of unsaponifiable fraction of cod-liver oil. Curve A, gross absorption; curve B, unsaponifiable (gross) - vitamin A,; curve C, vitamin-A, con- tribution; curve D, vitamin-A, contribution is regularly recorded by Salah, but it may be due to the presence of a small amount of an isomer of all-trans vitamin A,. Incidentally, there is as yet no proof that natural vitamin A, or even synthetic vitamin A, is mainly all-trans. Other properties of vitamin A, and its chemical constitution have been discussed in detail by Cama, Dalvi, Morton and Salah.5 The practical use of the information given in this paper is illustrated below.APPLICATION OF RESULTS TO A SAMPLE OF COD-LIVER OIL- A cod-liver oil tested on its unsaponifiable fraction by ultra-violet absorption gave E:"X, at 327 mp = 1.16 (uncorrected; solvent cyclohexane); by the colour test, E:%, at 693 mp = 0.384 and at 620 my = 2.80. From the properties of vitamin A, (see Table 11)- Fig. 1. 0.384 13'0 = 0.13 (in cyclohexane). 3860 E:& at 351 mp= The contribution of vitamin A, to the observed ultra-violet absorption is obtained by The resulting multiplying all the figures in column 1 of Table I by the factor 0.13 (see Fig. 1). curve is subtracted from that obtained for the unsaponifiable fraction. Now EiF, at 327 mp (for vitamin A,) = 0.13 X 0.8 (see Table I) = 0.104, and the intensity of absorption at 327mp corrected for vitamin A, = 1.16 - 0.104 = 1-056 (approximately 1-06].In Fig. 1, the vitamin-Al curve is calculated on the basis of Emx. -= 0.53 for all-trans vit;trniii A, (0.5 per cent. solution) 117 midtiplying by 0 . S all thc figiires for a cyclohexane78 CAMA AND MORTON: SPECTROSCOPIC PROPERTIES OF VITAMIN A, [-vOl. 78 solution of pure all-trans vitamin A, as observed by Cama, Collins and Morton.6 The colour test for vitamin A, shows absorption at 620 mp equal to that at 693 mp divided by 2.45 (see Table 11). 0.384 The corrected colour test for vitamin A, = 2.80 -- - 2.45 = 2.64. 2.64 2.92 The equivalent ultra-violet absorption = - = 0.91.” Shantz and Brinkman’ have shown that the potency of vitamin A, is 40 per cent. of that of vitamin A, (3-33 x 106i.u.per g), i.e., 3-33 x lo6 x 0.4 or 1.33 x 106i.u. per g. The conversion factor, therefore, for E:tm at 352 mp (cyclohexane) is 1100 and for the colour test at 693 mp is 345. The potency of vitamin A, in the oil is 136 (0.136 x 1000) i.u. per g or 132 (0.38 x 345) i.u. per g. In the past it was found useful to multiply the gross value on “total unsaponifiable” by the factor 1600. It is now necessary to multiply the corrected E:2m value (on the unsaponifiable fraction) by the factor 1900; on applying these procedures to the above sample we have 1.15 x 1600 = 1856 i.u. per g, and 0.9’3 x 1900 = 1881 i.u. per g. The earlier empirical conversion factor was well chosen and is still useful. A great many fish-liver oils contain a mixture of neovitamin A and all-trans vitamin A, in the ratio of 1 to 3, and a conversion factor that reflects the present state of knowledge concerning such a mixture is 1800; E:m at 327 mp (corrected for vitamin A,) = 1.056. Therefore potency = 1.056 x 1800 = 1900 i.u.per g. arises it is probable that for most purposes sufficient accuracy has been reached. Three approaches lead to a figure of about 1870 and unless some entirely new problem We are indebted to the Research Committee of the Royal Liverpool United Hospital for financial assistaizce, which enabled one of us (H. R. C.) to participate in this work. 1. 3. 3. 4. 5. 6. 7. HE FE KE N c‘ E S Shaiitz, E. M., Sciencc, 1948, 108, 417. Farrer, I<. li., Hamlet, J. C., Henbest, H. U., and Jones, E. R. ti., C h m . C Iizd., 1951, 49. Cama, H. R., nalvi, 1’.D., Rforton, 12. A., Salali, 11. K., Steinberg, G. I<., and Stubbs, A. L., Salah, R l . K., 1951, private communication. Cama, 13. R., Dalvi, 1’. D., Morton, R. A., and Salali, M. K., Biochetn. J . , 1953, 52, 540. Cama, H. K., Collins, F. D., and Morton, R. A., Ibid., 1961, 50, 48. Shantz, E. M., and Brinkman, H., J. Riol. Clzenz., 1950, 183, 467. Biocliewi. J., 1952, 52, 535. UEPARTMENT OF UIOCHEMISTKY ~JNIVERSITY O F LIVERPOOL July 17th, 1952 DISCUSSION MR. S. A. REED asked why a low value for vitamin A, was given on applying a conversion factor to the reading a t 620mp. PROFESSOR MORTON replied that when the colour test was applied to cocl-liver oils (unsaponified) A,. was a t 600 to 610 mp, depending on the sample. IVhen the colour test was carried out on the un- saponifiable fractions of the same oils, A,,a,.was at 617 to 820 mp and E,,,,. was much greater. The partial inhibition that occurred when oils were tested had not been fully cxplained; i t certainly invalidated the test for quantitative purposes. The colour test given with unsaponifiable extracts was less variable in the amount of inhibition from sample to sample and showed a good correlation with the corrected ultra- violet absorption a t 327 mp. Nevertheless, taking the properties of all-trans vitamin-A alcohol as standard, v 1750 -- 5070 for all-trans vitamin-A alcohol. In this calculation the inhibition of the colour by other unsaponifiable constituents is assumed to compensate roughly for the marc intense colour test shown by neo-vitamin-h ;~lcoliol, wliicli is usiially prcsent.Thc approxiinntion will ht. I * at 620mp onsidcrc.d fiirtlicr in ;I f u titrr. 1~11)cr.Feb., 19531 APPLICATION TO THE ASSAY OF COD-LIVER OIL 79 the colour test tended to give rather lower results than the ultra-violet absorption. I t was not certain that the difference was due to residual colour test inhibition. Nevertheless, for a series of cod-liver oils, the colour test on the unsaponifiable fraction was a good measure of relative potencies. DR. D. C. GARRATT asked whether the factors suggested by Professor Morton for conversion with the different techniques were sufficiently close to be used for all cod-liver oils or if they would be much in error for non-typical oils. Some low-potency oils exhibited absorption due to conjugated poly-ene acids, which exceeded in intensity that due to vitamin A.It was possible, however, to treat as normal the unsaponifiable fraction or vitamin fraction obtained by chromatography. If an oil was untypical because it had deteriorated and contained an unusual amount of free vitamin A or anhydrovitamin A, recourse should be had to chromatography, because correction procedures were then of doubtful validity. DR. E. C. WOOD, in supplementing the previous question, asked how much variation there was between cod-liver oils from different sources, e.g., between those from Norway and Newfoundland. PROFESSOR MORTON said that he had too little information about the origin of the oils he had tested to say whether there were consistent differences between the oils from different lishing grounds.DR. J. E. PAGE enquired about the possibility of determining vitamins A, and A, in natural products by infra-red absorption spectroscopy. He said that the infra-red absorption spectrum of vitamin A, was significantly different from that of vitamin A, and that it should be possible to determine the two vitamins in simple mixtures. PROFESSOR MORTON said in reply that without trial it was unwise to give a categorical answer, but it was unlikely that any standard infra-red absorption technique was as yet sufficiently quantitative to give E values for the accurate determination of vitamins A, and 9, in oils or unsaponifiable extracts. DR, R. E. STUCKEY said that certain cod-liver oils gave much irrelevant absorption on the unsaponifiable matter.He asked if Professor Morton would agree that a chromatographic proccdure was necessary for these oils and what percentage of correction it would be reasonable to apply. PROFESSOR MORTON said that it was often worth while to chromatograph anomalous oils (with or without saponification). Straight chromatography on weakened alumina and development with purified light petroleum often gave a vitamin-A ester fraction that after saponification proved to be spectroscopically more nearly normal. Corrections exceeding 20 per cent. of the gross value on the unsaponifiable matter were perhaps too great to be accepted uncritically. MR. H. E. MONK asked if the different forms of vitamin A served any different biological function and whether richer oils, such as halibut-liver oil, contained amounts of neo-A and A, similar to those in cod-liver oil. He also asked if the green colour with vitamin A, was as fleeting as the blue with vitamin A, and whether there was any method of dealing with these transient colours. PROFESSOR MORTON replied that there was at present no indication that vitamins A, and A, or their isomers had distinct physiological functions. Most fish-liver oils contained neovitamin A, but, although the all-trans form was usually predominant, the all-trans to neo ratio was a matter for measurement. In general, fish-liver oils contained 10 to 20 times as much *\, as A,, but for a few species of marine fishes the ratio of A, to A, could be as low as 5. The antimony trichloride colour test absorption was transient both for -4, and A, and no method of prolonging the full intensity had been deviscd. PROFESSOR MORTON said that it was difficult to generalisc about non-typical cod-liver oils.

ISSN:0003-2654    

DOI:10.1039/AN953780074b

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

80 FERNELL AND KING: THE SIMULTANEOUS DETERMINATION OF [Vol. 78 The Simultaneous Determination of Pentose and Hexose in Mixtures of Sugars BY W. R. FERNELL AND H. K. KING (Presented at the meeting of the Society on Wednesday, October lst, 1952) Pentose and hexose can be estimated when present together in mixtures by heating them with orcinol and acid either alone or with ferric chloride. The colours produced are measured a t suitable wavelengths and the results calculated directly from the colorirneter readings by means of a nomogram. DURING investigations into the chemical composition of bacterial cell material it became necessary to determine pentoses and hexoses when present together. An approximate answer, at least, was needed even when the exact nature of the sugars present was not known, Colorimetric methods based on the orcinol reaction proved suitable and required only small amounts of material.REACTION OF ORCINOL WITH ACID DECOMPOSITION PRODUCTS OF SUGARS- When orcinol is heated with sugars in the presence of sulphuric or hydrochloric acid the product has an absorption band in the blue end of the spectrum. Fig. 1 shows the absorp- tion spectra obtained when the reaction is carried out under the conditions prescribed on p. 81. Pentoses and hexoses both give a peak at 425 mp, but the absorption with the hexoses is the 560 600 Wavelength, rnp Fig. 1. Absorption spectra of material obtained on heating sugars with the orcinol- sulphuric acid reagent. Solutions containing 24 p g of the sugars per ml were treated with the reagent as described in the text.The absorption spectra of the solutions in a 1-cm cuvette were read on the Unicam spectrometer. Xylose , ribose - - - - -, fructose 0 0 0 0 0 0 , glucose 0-0-0-0-0, galactose x-x-x-x-x so0 Wavelength. rnv 7w Fig. 2. Absorption spectra of material obtained on heating sugars with the orcinol- ferric chloride reagent. Solutions containing 24 pg of the two pentoses per ml or 72 p g of the hexoses per ml were treated with the orcinol- ferric chloride reagent as described in the text. The absorption spectra of the solutions in a l-cm cuvette were read on the Unicam spectro- meter. Xylose , ribose # fructose 0 0 0 0 0 0 , galactose x-x-x-x-x glucose 0-0-0-0-0, lesser. If the colour is read on an “EEL” calorimeter* incorporating the appropriate filter, the colour ratio of pentose to glucose is about 1.3 to 1, or 1.1 to 1 if molar concentrations are considered (see Table I).This is the basis of the Tillmans - Phillippi method for assaying total carbohydrate.lP2 If, however, the reaction is carried out in the presence of ferric chloride, * Made by Evans Eleetroselenium Ltd., Harlow, Essex.Feb., 19531 PENTOSE AND HEXOSE I N MIXTURES OF SUGARS 81 yentoses give, in addition, a strong absorption band in the red, which is not given by hexoses. (See Fig. 2, which is essentially the same as that given by Brown.3 Note that the hexoses were used at a concentration thrice that of the pentoses in Fig. 2, to permit accurate measure- ment of the small absorptions at the greater wavelengths.) The resulting green colour is used in the Bial test for pentoses, which has been adapted for quantitative use by Mejbaum4 and Brown.3 At the peak (660 mp), xylose, ribose and arabinose (not shown in Fig.2) give the same absorption, viz., about 15 times that shown by fructose and 30 times that of glucose or galactose. Hence this reaction alone allows for estimation of pentose when an excess of hexose is not present. Brown3 allowed for hexose by measuring the absorption at two wavelengths, at the pentose peak of 660 mp, and also at 520 mp where the difference between pentose and hexose was less marked. Although in principle his method could be used for determining both pentose and hexose present together, it has its disadvantages. The reagent blank at 520 mp is considerable and variable.The large blank absorption at the blue end of the spectrum also limits the usefulness of measurements at what would otherwise be the most suitable wavelength, vix., 425mp, when the absorption for both yentose and hexose is at a maximum and of the same order of magnitude. In the method recommended here the mixture of sugars is heated with orcinol and acid, alone or with ferric chloride, and the absorption is read at the most suitable wavelengths, vix., 425mp and 660mp, respectively. The concentrations of both pentose and hexose can then be read from a nomogram. TABLE I COLORIMETER READINGS WITH VARIOUS SUGARS Sugar Glucose . . .. .. .. Galactose .. .. .. Fructose . . .. .. .. Sucrose . . .. Maltose (C,,H,,d,;.H,Oj * . . Ribose . . .. .. .. Xylose . . .... .. Readings with acid Readings with acid orcinol- ferric orcinol reagent and chloride reagent and 621 (blue) filter 608 (red) filter 19 2 20 5 2 2 13 17 2 26 78 26 78 64 METHOD PREPARATION OF MATERIAL FOR ASSAY- When only monoses or oligosaccharides are present, no preliminary preparation is required except for deproteinisation ( e g . , with trichloro-acetic acid) when necessary. Polysaccharides and combined sugars are hydrolysed with N hydrochloric acid at 100" C for 4 to 6 hours in a sealed tube. REAGENT- Orcinol-Traces of impurities are liable to yield derivatives similar to, but more deeply coloured than, those of orcinol itself. The blank does not provide full correction, and if the colour value much exceeds the normal range the linear relationship between carbohydrate present and colour produced no longer holds.The orcinol must therefore be freshly re- crystallised from hot water with a little charcoal. This yields the monohydrate, which should be converted to the anhydrous material by drying in a vacuum-desiccator. If orcinol containing water is used for making the orcinol- sulphuric acid reagent a high blank may result. ORCINOL - SULPHURIC ACID REACTION^- Dissolve dry orcinol (0.2 g) in 100 ml of diluted sulphuric acid (2 + 1) without heating. The reagent is stable for 2 days at 0" C, but must be discarded if it shows any discoloration. This is particularly likely to happen if freshly-recrystallised orcinol is not used. Add 10 ml of the reagent to 1 ml of the test sample (containing 5 to 25 pg of carbohydrate) in a 6 x 9-inch test tube.After mixing, cover the mouth of the tube with a glass bulb, heat the tube for 15 minutes at 100" C and cool rapidly to room temperature. Read the colour after 30 minutes, either in an "EEL" colorimeter with a 621 (dark blue) gelatin filter* or in a * Transmission maximum at 450 mp.82 FERNELL AND KING: THE SIMULTANEOUS DETERMINATION OF [Vol. 78 spectrophotometer at 425 mp. and correct the absorption value found. of impurities in the reagent and the results should be rejected. The colour is stable for 6 hours. Make a blank determination If the blank value is abnormally high it is a sign ORCINOL - FERRIC CHLORIDE REACTION3- Dissolve 0.15 g of ferric chloride (FeC1,.6H20) and 0.4 g of orcinol in 10 ml of cold water and make up to 200 ml with 30 per cent.w/v hydrochloric acid. Add 6 ml of this A B C D Fig. 3. Nomogram for calculating quantities of pentose and hexose from colori- meter readings. A, total carbohydrate (pentose plus hexose), p g per ml; B, colori- meter readings, orcinol- sulphuric acid method ; C, colorimeter readings, orcinol- ferric chloride method; D, pentose, pg per ml reagent to 2 ml of the sample (containing 5 to 25 pg of carbohydrate per ml) in a 6 x #-inch test tube. After mixing, cover the mouth of the tube with a glass bulb, heat the tube at 100” C for 20 minutes, and then cool rapidly to room temperature. Read the colour after 30 minutes either in an “EEL” colorimeter with a 608 (red) filter* or in a spectrophotometer at 660 mp. Make a blank determination and correct the zero setting of the colorimeter.PRECAUTIONS- The colour is stable for 2 hours. Careful attention must be paid to the following points- (1) The test solutions must not be diluted until shortly before the determination, or they may suffer appreciable decomposition. (2) A large water-bath should be used, as time and temperature of heating are critical. (3) The test tubes must be covered with loose glass bulbs or stoppers during heating so that the sulphuric acid is not diluted by absorption of steam or the hydrochloric acid by loss of hydrogen chloride. CALCULATION OF RESULTS- The amounts of hexose and pentose present can be calculated from the colorimeter readings by means of a nomogram (Fig. 3). The scales B and C correspond to the colorimeter readings for the acid-orcinol and the acid-orcinol - ferric chloride reactions, respectively.A line, e.g., bc, is drawn connecting the two readings for the test sample. This line is extended * Absorbs all wavelengths shorter than about 640 mp.Feb., 19531 PENTOSE AND HEXOSE IN MIXTURES OF SUGARS 83 at both ends to cut the other two scales at a and d. The total carbohydrate present (hexose plus pentose) is given at a, and d gives the pentose. The hexose is obtained by difference. The nomogram assumes that the hexose present is glucose. If a large proportion of fructose, (or sucrose) is present a substantial error will be introduced. If fructose is known to be the principle hexose present a special nomogram can be constructed. Construction ofthe nomogram-Scales B and C are drawn to a ratio of 1 to 3, this being the ratio of colorimeter readings when the same amount of pentose is subjected to the acid- orcinol and acid-orcinol- ferric chloride reactions, respectively.Scales A and D are identical and give the concentrations of pentose (pg per ml) corresponding to the readings on either scale B or C. The distance between the four scales is such that AD/CD is the ratio of the Fig. 4. Simultaneous estimation of pentose and hexose. Solutions were prepared containing glucose and ribose in various proportions. The total amount of sugar was always the same viz., 24 pg per ml. The two sugars were estimated. Dots, pentose found ; circles, hexose found. The two lines represent the theoretical readings sensitivity for pentose to hexose in the acid-orcinol- ferric chloride method, and AD/BD in the acid-orcinol method. These ratios vary slightly with the filters and instruments used and each worker should construct his own nomogram.Although Fig. 3 applies to determina- tions carried out with an “EEL” colorimeter the same principles are used in constructing the nomogram for any other type of instrument, e.g., a Beckmann spectrophotometer. RE s u LTS Table I shows the colorimeter readings with solutions containing 24 pg per ml of various sugars and treated with the two reagents as described above. The disaccharides were not hydrolysed before estimation. Ribonucleic acid (yeast) recorded 40 per cent. of the theoretical amount of pentose before hydrolysis and slightly more after hydrolysis. This is in accordance with expectation, since the pyrimidine nucleotides resist hydrolysis.5 Deoxyribosenucleic acid (thymus) gave no colour with either reagent. In Fig. 3, abcd is the experimental line for a mixture containing 9 pg of ribose and 15 pg per ml; atbrcrdr for 24 pg per ml of ribose only, and attbt’ct’dtt for 24 pg per ml of glucose only. Fig. 4 shows the concentrations of hexose and pentose found in a series of solutions containing various amounts of glucose and ribose. We are grateful to Professor R. A. Morton, F.R.S., for his interest in this work, which was performed during the tenure by one of us (W. R. F.) of a grant from the Ministry of Education under the Further Education and Training Scheme. REFERENCES 1. 2. 3. 4. 5 . Tillmans, J., and l’hillippi, I<., 13iochem. Z., 1929, 215, 36. Pirie, K. W., Brit. J . Ex?. Path., 1936, 17, 260. Brown, A. H., Arch. Biochem., 1946, 11, 269. Mejbaum, W. L., Hoppe-Seyl. 2.. 1939, 258, 117. Cori, G. T., and Cori, G. F., J . Biol. Chem., 1945, 158, 321. BIOCHEMISTRY DEPARTMENT JOHNSTON LABORATORIES UNIVERSITY OF LIVERPOOL June 13th, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800080

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

84 BIGGS : THE CONTROL OF ANTICOAGULANT THERAPY [vol. 78 The Control of Anticoagulant Therapy BY ROSEMARY BIGGS (Presented at the meeting of the Biological Methods Group on Friday, March 14th, 1952) The one-stage prothrombin test is the best available method for the control of anticoagulant therapy. This test does not necessarily measure prothrombin; its results are affected by the presence of factors V and VII. Patients treated with the dicoumarin derivative tromexan Iack factor VII, and it is this deficiency that controls the results of the test in these patients. A method that probably gives a true measure of prothrombin is described; the results of this test confirm the belief that the amount of prothrombin is only slightly reduced by anticoagulant therapy with the dicoumarin group of drugs.This method is of no value in the control of anticoagulant therapy. DICOUMARIN and its derivative tromexan have an anticoagulant effect on the blood by inter€ering with the normal coagulation mechanism. To measure the anticoagulant effect of the dicoumarin group of drugs it is necessary to measure the extent of the clotting abnormality. It has been found that the so-called one-stage prothrombin test gives a measure of the abnormality and that if the anticoagulant is administered in doses sufficient to maintain the clotting time by this test within certain fairly definite limits, haemorrhage from overdosage is unlikely and it is probable that the anticoagulant has a therapeutic effect. This problem is technical and much has already been written on the subject, so this paper is restricted to a consideration of the reasons why the one-stage test is a reliable guide to therapy with dicoumarin.The process of clotting, although superficially simple, is complex. A chain of preliminary reactions precedes the appearance of fibrin. Clotting can be studied only by observing fibrin formation; hence information about the earlier stages of the process is derived from indirect evidence. Because of this difficulty it has become customary to observe some phenomenon associated with coagulation and then to deduce the existence of a coagulation “factor” to account for the observation. When proposed, most of these factors have little claim to reality. As they appear the factors are ignored, adversely criticised, or studied by opponents; and gradually, with the multiplication of experiments, a few factors emerge as being more probable than others.These more probable factors gradually become accepted because the hypothesis of their existence explains the observations of several groups of workers. Naturally these several groups of workers have themselves given names to the factors, so that even the most acceptable factors proposed in the past 50 years are known by many different names. I t is against this mobile back- ground of elusive hypothetical substances that an attempt is made to measure coagulation defects in terms of isolated substances. Prothrombin, the substance long thought to be deficient in the plasma of patients treated with dicoumarin, is a widely recognised factor.It has survived more than 50 years of experiments and it now has only one name in common usage. Prothrombin is a substance occurring in plasma, and although not itself a coagulant of fibrinogen, it can, in suitable circumstances, be converted into a coagulant, thrombin. This conversion is greatly accelerated by tissue extracts called thromboplastins. According to the classical theory of blood coagulation, which was generally accepted during the first 40 years of this century, the reactions of blood could be written- Thromboplastin Prot hrombin -+ Thrombin Calcium In practice this problem is not difficult. In work on blood coagulation, theory has always been important. In this way a great many factors have been proposed. Thrombin The last of these two reactions, the thrombin-fibrinogen reaction, has been studied in detail, and it is now well established that when thrombin is added to fibrinogen, the clotting Fibrinogen -+ FibrinFeb., 19531 BIGGS : THE CONTROL OF ANTICOAGULANT THERAPY a!j time of the fibrinogen is inversely proportional to the thrombin concentration.Hence the clotting time of fibrinogen can be used as a measure of thrombin. If the classical theory of blood clotting is accepted and knowledge of the thrombin - fibrinogen reaction is applied, it becomes possible to devise methods for measuring prothrombin. These methods have proved useful, but they have given rise to endless practical and theoretical difficulties. They are known as the one-stage and the two-stage prothrombin tests. THE ONE-STAGE “PROTHROMBIN” TEST- The one-stage test consists in adding 0.1 ml of tissue extract to 0.1 ml of citrated plasma and then adding 0.1 ml of calcium chloride.The clotting time of the mixture is recorded. According to theory, the speed of thrombin formation in the mixture depends on the amount of prothrombin present, and the speed of clotting will depend on the speed of thrombin formation and the amount of thrombin formed. Thus, the clotting time is a measure of prothrombin. Even if this theory is accepted, there are some obvious disadvantages in the method. Con- ditions such as the presence of heparin or reduction of fibrinogen below 100 mg per 100 g will influence the clotting time regardless of the prothrombin concentration. Moreover, endless difficulty has arisen over the expression of results.A common method is to test dilutions of plasma by the one-stage test and draw a dilution curve relating clotting time to so-called concentration of prothrombin. When this is done it is found that the shape of the curve depends greatly on the diluent used. In addition to this method of expressing the results, there is a second method in which the clotting times are expressed as a percentage. This method bears no relation to the previous method, but in written communications the two methods are often not differentiated and so a great deal of unnecessary confusion has arisen. The first difficulty that arose from these methods was that the results from the two tests on any one sample of plasma seldom agreed. This in itself suggests that there is some- thing wrong with the theory. When these tests were being developed a haemorrhagic disease of Canadian cattle was investigated and was found to be caused by the cattle eating spoiled sweet-clover hay.By a series of brilliant experiments (summarised by Link1 in 1944) the haemorrhagic agent dicouinarin was isolated, and it was found that haemorrhage was due to reduction in prothrombin as measured by both one-stage and two-stage tests. Dicoumarin has been, and is, used widely as a therapeutic agent in various thrombosing diseases, and most workers believe that the plasma of patients treated with dicoumarin lacks prothrombin. This belief rested primarily on an acceptance of the classical theory of blood clotting, which is now known to be incomplete.There are at least two other factors that must be included in the theory, which should now probably be mitten2- Thromboplastin preparations do not all behave in exactly the same way. The two-stage test is more complicated and will be discussed later. Calcium Tissue extract + Factor VII -- -+ Thromboplastin Thromboplast in Calcium Prothrombin + Factor V -+ Thrombin Thrombin Fibrinogen -+ Fibrin Factors V and VII are essentially accelerators of blood clotting that affect the speed of thrombin formation, and a deficiency in either factor will influence the time for the one-stage prothrombin test. Factor V is a generally accepted factor but still has a number of alternative names, such as accelerator gl~bulin,~ prothrombin accelerator4 and a~celerin.~ Factor VIP is not at present generally accepted, but it appears to be an essential factor and its existence will account for observations of various authors who postulate the following substances: co-throm boplast in ,’ serum prot hrombin conversion accelerat or8 and convertin .6 A fuller discussion of this problem has been given elsewhere.2 Fortunately there are simple tests for deficiencies of Factors V and VII.Plasma treated with aluminium hydroxide, barium sulphate or calcium phosphate lacks prothrombin and Factor VII but contains Factor V. Thus, if a plasma sample has a long clotting time by the one-stage method and if the clotting time is greatly shortened by the addition of86 BIGGS: THE CONTROL OF ANTICOAGULANT THERAPY [Vol. 78 10 per cent. of plasma that has been treated with aluminium hydroxide, then Factor V is deficient in the plasma sample.Normal serum contains Factor VII in large amounts, small amounts of Factor V and little prothrombin. If a plasma sample has a long clotting time by the one-stage test and this time is not shortened by adding 10 per cent. of plasma treated with aluminium hydroxide but i s shortened by adding 10 per cent. of normal serum, then the plasma sample lacks Factor VII. By these tests it is found that the plasma of patients treated with tromexan lacks Factor 1711. The one-stage clotting time is not reduced by the presence of 10 per cent. of plasma treated with aluminium hydroxide, but is reduced, often to normal, by 10 per cent. of normal serum. THE MEASUREMENT OF PROTHROMBXN- This finding throws doubt on the usual conception that the plasma of patients treated with tromexan lacks prothrombin.To decide whether or not a sample lacks prothrombin it is necessary to have a reliable measure of prothrombin. If the one-stage test is not a reliable measure of prothrombin, the question arises whether the two-stage test is more so. c a .- E 2 Time in,minutes 0 I 2 3 4 Time in minutes Fig. 1. The formation of throm- Fig. 2. The formation of thrombin in a bin and its neutralisation in normal normal and a prothrombin deficient patient’s plasma by the two-stage method Curve plasma tested by the two-stage method. A, normal; curve B, prothrombin deficient In the two-stage test, plasma and tissue extract are mixed and calcium chloride is added. At intervals, samples are removed from this incubation mixture, which is forming thrombin, and added to fibrinogen.The clotting times of the fibrinogen solutions give a measure of the amount of thrombin present at the time the sample was removed. In this way the progress of thrombin formation is recorded. From Fig. 1 it will be seen that thrombin formation proceeds rapidly until a maximum is reached and then the level of thrombin declines. The level of thrombin is not maintained, because tlirombin in plasma is neutralised by an inhibitor, antithrombin. The usual method of calculating a relative percentage of prothrombin by the two-stage method is to compare the maximum level of thrombin formed in the pathological plasma with that in a normal plasma and to express the result as a percentage. In the example shown in Fig.2 the patient would be said to have about 33 per cent. of prothrombin. Another example shown in Fig. 3 will suggest that this method has disadvantages. In this illustration two examples of pathological plasma are compared. One curve is derived from the plasma of a patient who had uncomplicated prothrombin deficiency and the other is from the plasma of a patient treated with tromexan. The curve for normal plasma is not shown in the figure,Feb., 19531 BIGGS: THE CONTROL OF ANTICOAGULANT THERAPY 87 but from the levels of thrombin reached, the two plasma samples would be said to contain about the same amount of prothrombin. Yet it is clear that the plasma of the patient treated with tromexan had a much greater ability to form thrombin than that of the prothrombin-deficient patient.If it is taken that prothrombin is the precursor of thrombin, then the plasma of the patient treated with tromexan must have contained more prothrombin than the plasma of the other patient. It would seem that the duration of thrombin formation should be taken into account in assessing the amount of prothrombin present in the plasma. The essential problem of this technique is that the level of thrombin is not maintained because thrombin is neutralised by antithrombin. Most workers have assumed that, for this reason, the test should be carried out on greatly diluted plasma because the effects of antithrombin are less obvious in diluted plasma. But although the effects of antithrombin are less obvious in diluted plasma, they are still present.2~~ Moreover, some pathological ._ 3 3{ Time in minutes Fig. 3.The formation of thrombin in a pro- thrombin deficient patient’s plasma (curve A) and in the plasma of a patient treated with tromexan (curve B) plasma samples cannot be diluted because the clotting times become unduly long. If the problem of antithrombin in the two-stage method has to be taken into account, it is advisable to use undiluted plasma, because in undiluted plasma the effects of antithrombin can be studied in a few minutes, whereas in greatly diluted plasma the study would be very time- consuming. From the curves shown in Fig. 3 it seems that a reasonable method of assessing the amount of prothrombin in these two plasma samples might be to compute the area beneath the curves as a measure of prothrombin.But it was difficult to know whether or not this procedure was likely to give reliable results. A method of testing the hypothesis was to study some theoretical system similar in general pattern to that of the thrombin generation system in plasma. Suppose that a substance A is converted to B by a first-order reaction and that as B is formed it is converted to C, also by a first-order reaction; then the reaction is expressed in the curves of Fig. 4a. In the blood coagulation system A would be prothrombin, B thrombin, and C a neutralised thrombin - antithrombin association. In the two-stage test it is B that is measured, but the test must be designed to give a relative measure of A. The problem is how best to measure B.The factors that influence the shape of curve B are the amount of A present and the speeds of formation and destruction of B. If the speeds of formation and destruction of B are constant but the amount of A varies, then the curves shown in Fig. 4b will result. A study of these curves shows that the level of B achieved and the areas enclosed by the curves would give a proportional measure of A. If the amount of A is constant and the speed of destruction of B is constant but the speed of formation of B is varied, then the curves of Fig. 4c result. From these curves it is obvious that the levels of B achieved are now no guide to the amount of A present, but it is found that the areas enclosed by the curves are still proportional to the amount of A present.If the amount of A is constant and the speed of formation of B is constant but the speed of destruction of B varies, then neither the levels of B achieved nor the areas enclosed by the curves can give a measure of A (Fig. 44. On applying these results to the coagulation system it appears that the measurement of the area enclosed by the curves will give a relative measure of prothrombin provided that the antithrombin activity is constant. It is important to know how closely the reactions of coagulation conform to the theoretical system and apparently the agreement is good.88 BIGGS : THE CONTROL OF ANTICOAGULANT THERAPY [Vol. 78 7i 5 L 6 - - v) 2 4- 3- 2- 1 - I I I I I 1 2 3 4 5 6 7 8 Time I 3 4 5 6 Time (4 Fig. 4. ( a ) (b) (c) (d) Calculated curves for a theoretical react-on in which substance A is converted to B by a first-order reaction and simultaneously B is converted to C.Curves to show the disappearance of A and the formation of B and C. The concentrations of B at different times when formed from three different initial concentrations of A. The concentrations of B at different times assuming three different speeds of formation of B. Curves to show the effect on the concentration of B of varying the speed of conversion of B to C, the initial amount of A and the rate of formation of R bcing constant The initial concentration of A and the speed of neutralisation of €3 are constant.Feb., 19531 BIGGS : THE CONTROL OF ANTICOAGULANT THERAPY 89 In Fig. 5 the average curve for 15 normal subjects is compared with a theoretical curve.In Fig. 6 curves obtained by varying the amount of prothrombin are compared with theoretical curves. In Fig. 7 curves in which the speed of thrombin formation varies are compared with theoretical curves. The prothrombin content of normal and prothrombin-deficient plasma can be measured by the method and the prothrombin content of mixtures of the two can be measured. The theoretical prothrombin content of One more test can be applied to this method. Time in minutes The formation of throm- bin in normal plasma testcd by the two-stage method (continuous line) compared with a theoretical curve (broken line) Fig. 5 . the mixtures can be calculated from a knowledge of the prothrombin in the normal and abnormal plasma samples. When this is done it is found that there is a reasonable agreement between the theoretical and observed areas (Fig.8). I t is clear that the general pattern of the formation of thrombin agrees with the theoretical system. It is true that apparent agreements of this sort have no great significance and certainly do not mean that the reactions of blood clotting can be explained in terms of this simple pattern. But the deduction made from the agreement is empirical: it is suggested that in a system that follows this general pattern the measurement of area is likely to give a reasonable proportional measure of prothrombin. Provided that the experimental con- ditions are rigidly uniform, it is probable that this method of measuring prothrombin is more generally applicable than any other.If this method of measuring prothrombin is applied to the plasma of patients undergoing treatment with tromexan, it is found that prothrombin is not greatly reduced, and in no instance has less than 50 per cent. of prothrombiii been found.1° There was no correlation between the results of the one-stage and two-stage tests. This finding is not surprising, as the one-stage test does not record a change in clotting time for the reduction of a clotting factor by 50 per cent. Thus the reduction in prothrombin would be unlikely to influence the one-stage test. Morcover, the one-stage clotting time is shortened nearly to normal by 10 per cent. of normal serum, and from this evidence alone the coagulation defect cannot be attributed to pro- thrombin deficiency, because serum diluted to one-tenth of its normal strength provides insignificant amounts of prothrombin.From this clrork it appears that the main abnormality in the plasma of patients treated with tromexan is a deficiency of Factor VII. This deficiency is measured by the one-stage method, which, consequently, is useful. It may be thought that an attempt to define the90 ro- 9 - 8- 7- 6 - 'E 5 - 3 4n c, BIGGS : THE CONTROL OF ANTICOAGULANT THERAPY 7j 2- I - 2 4 6 8 1 0 1 2 (4 Time in minutes Fig. 6. Experimental observations in varied compared with the theoretical curves 4 h [Vol. 78 which Time in minutes (b) the amount of prothrombin in plasma was (a) Theoretical curves ( b ) Experimental results. Different amounts of prothrombin prepared by adsorption were added to plasma of a patient with uncomplicated prothrombin deficiency and thrombin formation was followed by the two-stage method Time Fig.7. (u) Theoretical curves (b) Experimental observations in which the speed of thrombin formation was varied. Comparison with the theoretical curves Experimental results. Curve A, thrombin formation in tromexan plasma; curve B, thrombin formation in tromexan plasma to which 10 per cent. of normal serum was added ; curve C, thrombin formation in normal plasmaFeb., 19531 BIGGS : THE CONTROL OF ANTICOAGULANT THERAPY 91 deticiency in the plasma of patients treated with tromexan is unnecessary because empirically the one-stage test has been found to be adequate. But although the one-stage test, carefully standardised, is valuable, much difficulty has arisen in practice from modifications of the technique and from deductions based on the theory that the test gives a measure of ./ I200 - 1000 - $ 800- a 0 200 400 6do 800 I000 I2W Observed area Fig.8. The area enclosed by the two- stage curves was measured in normal plasma, abnormal plasma and in mixtures of the two in known proportions. The observed area in the mixtures is compared with the area calculated for the proportions of normal and abnormal plasma mixed prothrombin. A clearer conception of the significance of the results of the test in particular coagulation defects may, in the end, help in the understanding of the normal coagulation mechanism. The experimental work on which this paper was based was carried out in collaboration with Dr.A. S. Douglas and will be reported fully elsewhere.ll All the experimental techniques used are described in detail by Biggs and Macfarlane.2 REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. (3. Biggs, R., “Prothrombin Deficiency,” Blackwell’s Scientific Publications, Oxford, 1951. 10. 11. Link, K. P., Hawey Lect., 1944, 39, 162. Biggs, R., and Macfarlane, R. G., “Human Blood Coagulation and its Disorders,’’ Blackwell’s Scientific Publications, Oxford, 1953. Ware, A. G., and Seegers, W. H., J. Riol. Chem., 1948, 172, 699. Fantl, P., and Nance, M., Med. J. Aust., 1948, 1, 128. Owren, F. A., “Proceedings of the Third Congress of the International Society of Haematology Koller, F., Loeliger, A., and Duckert, F., Acta Haematol., 1951, 6, 1. Mann, F. D., and Hurn, M., Azizer.J. Physiol., 1951, 164, 105. de Vries, A., Alexander, B., and Goldstein, R., Blood, 1949, 4, 247. Douglas, A. S., unpublished data. Biggs, R., and Douglas, A. S., J . Cliiz. Path., 1953, in the press. 1950,” Grune & Stratton, p. 379. Jtinc 6th, 1952 PATHOLOGY DEPARTMEKT RADCLIFFE INFIRM-IRY OxFoRn DISCUSSION MR. K. L. SMITH suggested that the areas enclosed by the curves might be computed approximately DR. H. 0. J. COLLIER enquired about the use of the test in practice and asked how the results were as the product of the mean of all the points multiplied by the length of base. related to medical findings.92 HASLAM, GROSSMAN, SQUIRRELL AND LOVEDAY: THE DETECTION [VOl. 78 DR. BIGGS said that a start had been made in the routine use of the test, but not enough work had yet been done to say how the results would compare with clinical findings.In reply to a question by Mr. A. L. Bacharach she said that Factor V was an additional factor rather than an intermediate product in the formation of thrombin. DR. G. E. FOSTER enquired whether the findings were the same with animal as with human plasma. DR. BIGGS replied that she had no experience with ox plasma, but she preferred to use human plasma, which was, for her, more important. As a source of thromboplastin she used human rather than rabbit brain. DR. R. M. HARDISTY asked if Dr. Biggs’ “Factor VII” was the same as “Factor VII” of Koller and the “proconvertin” of Owren. If so, what was the evidencc for including it in the first stagc of the coagulation process as combining with brain extract to produce thromboplastin. DR. BIGGS, replying, referred to the work of R. F. Jacox ( J . Clin. Invest., 1949, 28, 492) and of Owren (reference 5 of paper), and said that Factor VII was essentially an accelerator. She mentioned also a recent paper by F. D. Mann and M. H. Hurn (Proc. SOC. Exp. Biol. N.Y., 1952,79, 19) dealing with species specificity of thromboplastin. DR. E. BRASTED enquired whether the one-stage prothrombin test could be used as a measure of fibrinogen. DR. BIGGS said that the one-stage prothrombin test could be used as a measure of fibrinogen deficiency, but it would give a very inaccurate and unreliable measure of fibrinogen; other and much better methods for measuring fibrinogen were in routine use. DR. L. ELLIS asked whether i t was possible to measure prothrombin in terms of an independent prothrombin rather than as a percentage of normal human prothrombin content. DR. BIGGS said that prothrombin could not be isolated from plasma in a form freed from all other coagulation factors ; an absolute measure of prothrombin was therefore impossible.

ISSN:0003-2654    

DOI:10.1039/AN9537800084

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

92 HASLAM, GROSSMAN, SQUIRRELL AND LOVEDAY THE DETECTION [Vol. ‘78 The Detection and Determination of Ultra-Violet Absorbers and Other Additives in Polymethyl Methacrylate and Methyl Methacrylate -. Ethyl Acrylate Co-Polymers BY J. HASLAM, S. GROSSMAN, D. C. M. SQUIRRELL AND S. F. LOVEDAY Chemical and spectroscopic methods are described for the detection and determination of small amounts, 1 per cent. or less, of additives in poly- methyl methacrylate and methyl methacrylate - ethyl acrylate co-polymers. These methods deal with the determination of lauryl mercaptan and total sulphur in such polymers, as well as the qualitative and quantitative detection and determination, both chemically and spectroscopically, of such ultra- violet absorbers as methyl salicylate, phenyl salicylate, 2 :Cdihydroxy- benzophenone, stilbene and resorcinol monobenzoate. Details are given of a direct spectroscopic method for the determination of small amounts of dibutyl phthalate, and observations are made on the behaviour of cresyl esters and hydroquinone and catechol monobenzoates in the appropriate tests.WITHIN recent months it has become necessary to detect and determine small amounts of additives used in the preparation of polymethyl methacrylate and in co-polymers of methyl methacrylate with other monomers. It is proposed to describe here some of the methods that have been found most useful in connection with this work. The additives, which may be included either in the course of, or a t the conclusion of, the polymerisation, are usually only present at small concentrations (less than 1 per cent.).Although there are many of these additives, it is proposed to describe the detection and determination only of those that have been encountered in day-to-day work in the plastics industry. The particular substances dealt with are: dibutyl phthalate, lauryl rnercaptan, methyl salicylate, phenyl salicylate, 2 :4-dihydroxybenzophenone, stilbene and resorcinol mono- benzoate. Lauryl mercaptan, it is understood, is an extremely useful substance because itsFeb., 19531 AND DETERMINATION OF ULTRA-VIOLET ABSORBERS IN POLYMERS 93 addition to a polymethyl methacrylate polymer slows down the depolymerisation that occurs on heating, and polyners incorporating this substance have, in general, a much higher softening point than those from which the mercaptan is absent.Phenyl salicylate, methyl salicylate, 2 :4-dihydroxybenzophenone, stilbene and resorcinol monobenzoate can Wavelength. mp Fig. 1. Transmission curve of Q-inch sheet poly- methyl methacrylate. Curve A, before exposure ; curve B, after exposure be added to an acrylic polymer as ultra-violet absorbers. Their inclusion in the polymer sheet tends to reduce the “yellowing” that may occur on exposure to sunlight. This phenomenon may be more readily understood by reference to the ultra-violet exposure test diagrams shown in Figs. 1, 2 and 3, which show, as examples, the transmission curves of &-inch sheets of polymethyl methacrylate (a) unplasticised, (b) containing 1 a 0 per cent. of 250 300 350 400 450 500 Wavelength, mp Fig.2. Transmission curve of *-inch sheet of poly- methyl methacrylate containing 1.0 per cent. of phenyl salicylate. Curve A, before exposure; curve B, after exposure phenyl salicylate and (c) containing 0.44 per cent. of stilbene, before and after exposure to ult ra-violet radiation. The samples were exposed to ultra-violet radiation under fixed conditions for 50 hours, i.e., the equivalent of 50 hours of mid-summer sunshine in Great Britain.04 HASLAM, GROSSMAN, SQUIRRELL AND LOVEDAY: THE DETECTION [VOl. 78 A comparison of the relative colour formation or “yellowing” of the samples after exposure, assessed in the visible region between the wavelengths 400 and 500 mp and indicated by the shaded portions in the diagrams, is as follows- Unplasticised Phenyl salicylate, 1*00/, Stilbene, 0.440,: 8 2.5 1 This, apart from showing the effect of the inclusion of an ultra-violet absorber in poly- methyl methacrylate sheet, clearly demonstrates the difference in efficienq- between the two absorbers used for the tests.THE DETERMINATION OF MERCAPTANS AND TOTAL SULPHUR IS POLYMERS OF THE POLYMETHYL METHACRYLATE TYPE The determination of mercaptans and total sulphur in polymethyl methacrylate-type polymers forms rather a different problem from that of the determination of dibutvl phthalate and ultra-violet absorbers, and will be dealt with separately. Lauryl mercaptan in polymethyl methacrylate and methyl methacrylate c o-polymers can be determined by a modification of the method originally developed by Kolthoff and Wavelength, rnp Fig.3. Transmission curve of *-inch sheet poly- methyl methacrylate containing 0.44 per cent. of stilbene. Curve A, before exposure; curve B, after exposure Harris1 for the amperometric titration of mercaptans. In the test the polymer is dissolved directly in acetone ; ammonium hydroxide and ammonium nitrate are added and the mercaptan is titrated amperometrically with standard silver solution. Full details of the test are given below. THE 1) E TE R M I N AT1 ON 0 F ME KC APT AX S L4PPARATU S- The apparatus is similar to that of Kolthoff and Harris,l namely, a rotating platinum- wire indicator electrode, and a reference half-cell containing mercury in contact with a potassium iodide, mercuric iodide and potassium chloride solution, connected through the test solution by means of a salt-bridge and short-circuited through a micro-ammeter.REAGENTS- L4 cetone-Analytical reagent quality. ,4mmo9aiiim hydroxide-A 5 M solution. Ammoniiwn nitrate-A M solution. SiZver nitrate-A 0.005 N standard solution.Feb., 19531 AND DETERMINATION OF ULTRA-VIOLET ABSORBERS IN POLYMERS 95 PROCEDURE- Accurately weigh 2 g of the sample into a 250-ml beaker and stir magnetically with 100 ml of acetone until solution is complete (this usually takes about 1 hour). Add 5 ml of the ammonium hydroxide solution and then 5 ml of the ammonium nitrate solution and stir for a further 15 to 20 minutes, during which time any polymer precipitated by the addition of the aqueous solutions redissolves. Transfer the beaker and its contents to the titration apparatus and titrate the mercaptan in the solution amperometrically with the 0.005 N silver nitrate solution by the general procedure used by Kolthoff and Harris1 The accuracy of the test is adduced in Table I from results of experiments that involved the amperometric titration of 100-ml portions of (a) solutions of known amounts of lauryl mercaptan in acetone, (b) solutions of known amounts of lauryl mercaptan in a mixture of equal volumes of acetone and alcohol and (c) solutions containing, in addition to known amounts of laurvl mercaptan, 2 g of dissolved polymethyl methacrylate. TABLE I DETERMINATION OF LAURYL MEKCAPTAN Lauryl mercaptan Lauryl mercaptan mg mg Acetone .. .. .. . . . . 4.83 4.82 9.66 9-67 Ethyl alcohol - acetone mixture (1 + 1) 4.83 4-96 9-66 9-77 Medium added, found, (High results owing to small blank on alcohol) Acetone containing 2 per cent.of dis- 0.00 0.00 solved polymer. . .. .. .. 4.83 4.82 4.83 4.86 It may often be desirable to determine the total sulphur content of the polymer sample, for although a certain amount of a substance such as lauryl mercaptan may have been added initially in the preparation of the polymer, only a part of this may exist as lauryl mercaptan in the final product. Again such substances as thio-ethers may have been used and these would not respond to the amperometric test for mercaptans. For these reasons it is desirable to have available a method for determining the total sulphur content (0 to 0.3 per cent.) in these polymer preparations; the following method has been found suitable.THE DETERMINATION OF TOTAL SULPHUR APPARATUS- A Mahler - Cook bomb was used for combustion of the sample in a rapid and quantitative way; this forms part of the Mahler - Cook New Quick-seal bomb calorimeter ordinarily used for the determination of the calorific value of coal.* REAGENTS- Hydrogen jkroxide-A 5-volume solution. Sdphuric acid-A 0.005 N solution. Hydrochloric acid-A N solution. Precipitatin,g reagent-Solution A. Dissolve 0.2 g of peptone in 50 ml of 1 per cent. Buffer to a pH of 5.0 with 0.02 N hydrochloric Heat Solution B. Dissolve 0.4 g of ground gum ghatti in 200 ml of distilled water by warming Store solutions A and B separately and prepare the final precipitating reagent just * The Mahler - Cook New Quick-seal bomb is manufactured by Messrs.Chas. W. Cook & Sons Limited, barium chloride (IEaC1,.2H2O) solution. acid solution; add l o g of sodium chloride (analytical grade) and dilute to 100ml. in a water-bath for 10 minutes and add a few drops of chloroform. slightly. before use by diluting 10ml of solution A to 100ml with solution B. Birmingham. When solution is complete add 2.0 g of barium chloride (BaC1,.2H20).96 HASLAM, GROSSMAN, SQUIRKBLL AND LOVEDAY: THE DETECTIOK [VOl. 78 PROCEDURE- Accurately weigh 0.5 g of the sample into a glass crucible through which passes the platinum resistance wire, connected to the firing terminals. Add 10 ml of distilled water to the bomb and place the crucible and sealing cap in position. Seal the bomb and introduce oxygen to give a pressure of 25 atmospheres.Immerse the bomb in water, iire it, and after 10 minutes release the gases by bubbling them through 50 ml of the 5-volume hydrogen peroxide solution, using a sintered-glass distribution plate. Bulk the washings and scrubbing solution together in a 250-ml beaker, add 15 ml of N hydrochloric acid solution and boil down to a volume of approximately 25 ml. Filter the cooled solution into a 50-ml calibrated flask and add 5 ml of the prepared precipitating reagent; dilute to 50 ml, shake well and set aside for 30 minutes. Measure the optical density of the resulting solution in a Speklter absorptiometer using a l-cm cell and Spectrum red 608 filters. Prepare a calibration curve by putting known aliquots of 0-005 N sulphuric acid solution (2 to 16ml cover a suitable range) through the above precipitation procedure and plot a graph relating the number of millilitres of 0.005 N sulphuric acid solution per 50 ml of final solution to the absorptiometer indicator drum reading. From this graph calculate the sulphur content of the test sample.The method was tested by determining sulphur in samples of polymethyl methacrylate containing known amounts of lauryl mercaptan or dioctyl thio-ether, which had been included in the polymerisation process. Dismantle the bomb and wash it thoroughly with distilled water. The results were as shown in Table 11. TABLE I1 DETERMINATIOK OF SULPI-JUR IN PREPARED SAMPLES OF POLYMETHYL METHACRYLATE Sulphur added as lauryl mercaptan, Sulphur found, nil nil 0.06 0.05 0.06 0.05 0.09 0.09 YO % THE QUALITATIVE DETECTION AND OTHER Sulphur added as lauryl mercaptan and dioctyl thio-ether, Sulphur found, Y O YO 0.12 0.13 OF ULTRA-VIOLET ABSORBERS ADDITIVES Our experience has been principally of the ultra-violet absorbers phenyl salicylate, methyl salicylate, 2 :4-dihydroxybenzophenone, stilbene and, to a lesser extent, resorcinol monobenzoate.* It has been shown, for instance, that qualitative evidence of phenyl salicylate, methyl salicylate, 2 :4-dihydro~ybenzophenone and resorcinol monobenzoate can be obtained by applying comparatively simple tests to the polymer or co-polymer.With each, the sample is dissolved in acetone and the polymer precipitated with alcohol and water. The polymer is filtered sff and the appropriate tests applied to the acetone - alcohol - water solution containing the ultra-violet absorber.The addition of alkali to this solution and the immediate production of a yellow colour is presumptive evidence of the presence of 2 :4-dihydroxybenzophenone. The tests for phenyl salicylate, methyl salicylate and resorcinol monobenzoate are based on treating the solution with a known amount of alkali, neutralising the resulting solution and applying three tests : a 2 :6-dibromoquinonechlorimide test for phenols, Millon’s tests for phenols and the ferric chloride test for salicylates. Phenyl salicylate, methyl salicylate and resorcinol monobenzoate behave differently in these tests, the details of which are given below. CHEMICAL EXAMINATION REAGENTS FOR SOLUTION O F SAMPLE- Acetone-Analytical reagent quality. Absolute alcohol, 99.9 per cent.* Preliminary tests under an ultra-violet lamp often give useful pointers, particularly if samples containing known additives are available for comparison.Feb., 19531 AND DETERMINATION OF ULTRA-VIOLET ABSORBERS, ETC. 97 PREPARATION OF SAMPLE SOLUTIOX- Weigh 1 g of the finely divided sample into a 150-ml conical flask, add from a pipette 10ml of acetone, stopper, and set aside overnight to ensure complete solution. Place the flask on a magnetic stirrer and allow the glass-enclosed metal stirrer to revolve for a few minutes to homogenise the solution before adding 45 ml of absolute alcohol dropwise from a burette to precipitate the polymer. Set the solution aside for 10 minutes and then add 48 ml of distilled water from a burette in the same manner.After setting aside for a further period of 15 minutes, filter the solution through a Whatman No. 1 filter-paper and retain the filtrate. Notes-Certaiu yolymethyl methacrylate samples, when treated in this way, yield highly viscous solutions in acetone, which on treatment with alcohol form insoluble gels. In such circumstances it is desirable to first dissolve the polymer in a mixture of 10ml of acetone and 10 ml of alcohol, stirring constantly to maintain mobility. The polymer is precipitated finally by adding 36 ml of absolute alcohol and 48 ml of distilled water. The ternary mixture of 10 ml of acetone, 45 ml of absolute alcohol and 48 ml of distilled water gives a final volume of 100 ml. IXEAGENTS FOR QITALITATIVE TESTS- Sodium hydroxide solution, N.Nitric acid solution, N. 2 :6-Dibrornoqz~i~zo~zechlorinzide sol.ution--l>issolve 0-1 g of 2 :6-dibromoquinonechlorimide in 25 ml of 95 per cent. ethyl alcohol just before the solution is required for use. Ferric chloride solution-A 10 per cent. w/v solution of ferric chloride (FeC1,.6H20) in distilled water. Millon's reagent-Dissolve 2 ml of mercury in 20 ml of analytical reagent grade con- centrated nitric acid. Allow to cool and add 35 ml of distilled water; then add 10 per cent. w/v sodium hydroxide solution dropwise until a faint permanent turbidity is formed and finally 5 ml of 20 per cent. v/v nitric acid solution to clear the solution. Borate bufw solzition--I>issolve 23.4 g of sodium borate (Na,B,O,.lOH,O) and 3.27 g of sodium hydroxide pellets in distilled water and dilute to 1 litre.(2 u ALITATIVE TB STS- 1. (a) Add 2-0 ml of N sodium hydroxide solution to 50 ml of the prepared solution. The immediate formation of a yellow colour, which is destroyed by the addition of 2-2 ml of N nitric acid solution, is indicative of 2 :4-dihydroxybenzophenone. (b) If no yellow colour forms immediately, bring the alkaline solution to the boiland continue to boil for 15 minutes. The formation of a yellow-green colour is indicative of resorcinol monobenzoate. (N.B. The formation of a pale straw-yellow colour at this stage should be ignored.) 2. If test 1 (a) is negative, transfer a fresh 20-ml aliquot of the prepared sample solution to a SO-ml calibrated flask containing 10 ml of distilled water.Add 4.0 ml of N sodium hydroxide solution and stand the flask in a thermostat maintained at 25" C for 4 hours, after which time add 4.2 ml of N nitric acid solution. To this faintly acid solution apply the following tests- (a) To a 10-ml aliquot contained in a 50-ml Nessler cylinder, add 5 ml of Millon's reagent. A pink colour, developing after 5 minutes and becoming deeper over a period of 30 minutes, is given by phenyl esters such as phenyl salicylate. (b) To a 1-ml aliquot contained in a 6 x Q-inch test tube add 2 ml of the borate buffer solution. Mix well and add 5 drops of the 2 :6-dibromoquinonechlorimide reagent. In this test, resorcinol monobenzoate yields an immediate violet colouration, and compounds such as phenyl salicylate give a blue colour that develops within 5 minutes.To the remainder of the slightly acid solution, add 2 drops of the ferric chloride reagent, and dilute to 50ml with distilled water. The immediate formation of a violet-coloured ferric salicylate complex is indicative of a salicylate in the hydrolysis products. (c)98 HASLAM, GROSSMAN, SQUIRRELL AND LOVEDAY: THE DETECTION [VOl. 78 From the results of these tests the probable presence of the undernoted ultra-violet absorbers is inferred- (;> (ii) (iii) (iv) Positive reaction in test 1 ( a ) . . . . Probably 2 :4-dihydroxybenzophenone Positive reaction in test 2 (c) . . > > 2 (a) . . : :} Probably phenyl salicylate. Blue colour in test 2 (b) . . . . Positive reaction in test 2 (c) Negative >> (a) and (bj * } Probably methyl salicylate.Immediate violet positive reaction in Xegative reaction in test 2 (c) Positive reaction in test 1 (b) . . Probably resorcinol monobenzoate. . . .. . . test 2 (b) . . . . .. . . Note on test for salicylates-Under these test conditions we have found that methyl salicylate gives a more intense colour than an equal weight of phenyl salicylate, owing, no doubt, to the more complete hydrolysis of methyl salicylate by aqueous alkali. It is always desirable to supplement this qualitative chemical information by the corre- sponding qualitative spectroscopic information obtained by applying the ultra-violet absorption tests detailed below. SPECTROSCOPIC EXAMINATION When the sample submitted for analysis is in the form of clear sheet, useful information can often be obtained by a straightforward examination of the light transmission of the sheet itself.Ultra-violet absorbers have a bathochromic effect, i.e., with polymethyl methacrylate sheet, they cause a shift towards the red of the straight-line vertical portion of the transmission curve. For example, experience has shown that in the examination of the light transmission of &-inch sheet of polymethyl methacrylate, when the wavelength at 50 per cent. transmission is greater than 310mp the sample contains an ultra-violet absorber. This can be seen from the data given in Table 111, taken from the transmission curves shown in Figs. 1, 2 and 3. TABLE I11 BATIIOCHROMIC EFFECT OF STILBENE AND PHENYL SALICYLATE ON &-inch POLYMER SHEET Wavelength for 50 per cent.transmission, mtL Unplasticised polymethyl mcthacrylate . . .. 284 Polymethyl methacrylate containing 0.44 per cent. of stilbene . . . . .. . I .. .. 345 Polymethyl methacrylate containing 1.0 per cent. of phenyl salicylate . . .. .. .. .. 35 1 A plasticiser such as dibutyl phthalate exhibits an effect somewhat similar to that of ultra-violet absorbers, but to a much smaller extent. It is always desirable, however, to carry out a general qualitative examination for additives such as dibutyl phthalate and the ultra-violet absorbers phenyl salicylate, methyl salicylate, 2 :4-dihydroxybenzophenone, stilbene and resorcinol monobenzoate in the following way. PREPARATION OF TEST SOLUTION- Weigh 0-3 to 0-4g of sample in the form of small chips or fine drillings into a 100-ml calibrated flask, add about 60 in1 of chloroform B.P.and shake at once to swell the polymer. Shake mechanically until the solution clears; 30 minutes is usually adequate. Make up to 100 ml with chloroform and shake for a few minutes until homogeneous. This constitutes the test solution. PROCEDURE- Transfer a portion of this test solution to a 1-cm cell, cover with a ground-glass lid and examine spectrophotometrically in the region 265 to 340 mp against a paired cell containingFeb., 19531 AND DETERMINATION OF ULTRA-VIOLET ABSORBERS IN POLYMERS 99 chloroform from the same batch as that used in the preparation of the solution. It may be necessary with an unknown sample to dilute the solution at this stage. Under the conditions of the test, solutions of polymethyl methacrylate, without additives, are relatively transparent in the region examined.The probable presence of the additives under investigation , namely, dibutyl phthalate, methyl salicylate, phenyl salicylate, 2 :4- dihydroxybenzophenone, stilbene and resorcinol monobenzoate, is denoted by the characteristic absorption maxima in polymer solution shown in Table IV. TABLE IV ABSORPTION DATA OF CERTAIN ADDITIVES -4 dditive hmax., hmin., m P mP 274 - Resorcinol monobenzoate .. .. .. 276 - 2 : 4-Dihydroxybenzophenone . . .. .. 289, 324 313 Stilbene . . . . .. .. .. .. 299, 310 306 .. 308 L 312 - Dibutyl phthalate . . .. .. .. Methyl salicylate . . .. .. . . Phenyl salicylate . . .. .. .. .. Note-As a further check, it is current practice in this laboratory to plot the absorption spectra of the sample on transparent paper and compare this with the spectra of samples of known composition accumulated over a period of time. THE QUANTITATIVE DETERMINATION OF ULTRA-I'IOLET ABSORBERS AND OTHER ADDITlVES The qualitative chemical tests supplemented by the corresponding qualitative spectro- scopic tests will have given sound evidence of the presence or absence of the various additives such as methyl salicylate, 2 :4-dihydroxybenzophenoneJ dibutyl phthalate, and so on.It will often be desirable to carry out quantitative determinations of the various substances of which qualitative evidence has been obtained. The chemical and spectroscopic methods we have found most useful in this connection are indicated below.CHEMICAL EXAMINATION The sample solution is prepared exactly as described in the qualitative tests from an After precipitation of the polymer the additives accurately weighed l-g portion of the sample. are contained in 100 ml of solvent. REAGENTS FOR THE DETERMINATION OF NETI-IYL SALICYLATE- Sodium hydroxide solution, N. Hydrochloric acid solution, N. Ferric chloride solution-A 10 per cent. w/v solution of ferric chloride (FeC1,.6H2O) in distilled water. Filter the solution immediately before use. PROCEDURE FOR THE DETERMINATION OF RIETHYL SALICYLATE- Transfer 10 ml of the prepared sample solution by means of a pipette to a 50-ml calibrated flask containing 5 ml of distilled water. Add 2.0 ml of N sodium hydroxide solution and stand the flask in a thermostatically controlled water-bath at 25" C for 4 hours.Remove the flask from the bath and add 2.2 ml of N hydrochloric acid followed by 2 drops of the ferric chloride sohition. Mix the solution well and dilute to the 50-ml mark and set aside for 15 minutes before filtering through a Whatman No. 1 filter-paper into a 4-cm cell. Measure the optical density of the solution in this cell, against a blank solution of the reagents used, in a Spekker absorptiometer, using Ilford green 604 filters and Calorex heat absorbers. Prepare a calibration curve covering the range from 0 to 1-0mg of methyl salicylate per 50 ml of final coloured solution (or 0 to 1.0 per cent. working on 1 g of sample) by the following procedure. Dissolve 0.1 g of methyl salicylate in 100 ml of acetone contained in a l-litre flask.Add 450 ml of absolute alcohol and dilute the solution to 1 litre with distilled water; 1 ml of100 HASLAM, GROSSMAN, SQUIRRELL AND LOVEDAY-: THE DETECTION [VOl. 78 this solution is equivalent to 0.1 mg of compound. Into ten 50-ml calibrated flasks place 1, 2, 3 . . . 10 ml of this solution and dilute to 10 ml with a mixture containing 10 parts by volume of acetone, 45 parts by volume of absolute alcohol and 48 parts by volume of water. Add 6 ml of water to each flask followed by 2 ml of N sodium hydroxide solution and proceed as in the method described. Draw a graph relating milligrams of compound per 50 ml of final coloured solution to the Spekker indicator drum reading. Calculate the methyl salicylate content of the original sample from this calibration graph, which, in our experience, is a straight line passing through the following points- Methyl salicylate corresponding to 50 ml of coloured solution, mg .. .. . . 0.0, 0.2, 0.4, 0.6, 0.8, 1-0 Indicator drum reading . . * . . . 0-0, 0.074, 0.146, 0.220, 0.295, 0.380 DETERMINATION OF PHENYL SALICYLATE- For samples containing up to 0.5 per cent. of methyl salicylate, we have found it sufficient to carry out one precipitation of the polymer, but for samples containing phenyl salicylate, a second precipitation should be carried out by the following procedure. Procedwe-Precipitate the polymer, as already described in the method for the deter- mination of methyl salicylate, from 10ml of acetone by the addition of 45ml of absolute alcohol and 48 ml of distilled water.Set aside for 15 minutes and then decant the supernatant solution through a Whatman No. 1 filter-paper, leaving as much of the polymer as possible in the flask. Collect the filtrate in a 100-ml calibrated flask and make up to the mark with a mixture of acetone (10 parts by volume), absolute alcohol (45 parts) and water (48 parts) by washing through the filter-paper. Determine the salicylate content of this solution by the procedure described for methyl salicylate, using a calibration curve prepared with a standard solution of phenyl salicylate. Transfer the filter-paper and funnel to the original flask and puncture the filter-paper. Wash the polymer on the paper through the hole into the flask with 10ml of acetone and stir for about 2 hours, during which time the polymer will redissolve or become well penetrated by the solvent.After this digestion, re-precipitate the polymer by adding 45 ml of absolute alcohol and 48 ml of distilled water before determining the salicylate in this, now much more dilute, solution by the procedure already described. Calculate the total phenyl salicylate content of the sample from the sum of these two determinations. The calibration graph for phenyl salicylate is, in our experience, a smooth curve passing through the following points- Phenyl salicylate corresponding to 50 ml of final solution, mg . . .. .. . . 0.0, 0.2, 0.4, 0.6, 0.8, 1.0 Indicator drum reading . . .. . . 0.0, 0.060, 0.118, 0.178, 0-237, 0.296 The need for a second precipitation may be made clear by noting that a sample made to contain 1.0 per cent.of phenyl salicylate gave a figure of only of the order of 0.9 per cent. when the determination was carried out on the filtrate from a single precipitation only. By the double precipitation method described above the figure was 0.97 per cent. DETERMINATION OF 2 :4-DIHYDROXYBENZOPHENONE- Procedure-To 50 ml of the prepared sample solution contained in a 50-in1 calibrated flask add 2.0 ml of N sodium hydroxide solution; shake well and set aside for 10 minutes. Filter the solution until it is optically clear through a Whatman No. 1 filter-paper into a 4-cm cell and measure the optical density against a blank of the reagents used in a Spekker absorptiometer, using Spectrum violet 601 filters and Calorex heat absorbers. Prepare a calibration curve as follows- Dissolve 0.1 g of 2 :4-dihydroxybenzophenone in 100 ml of acetone in a l-litre calibrated flask and dilute the solution to 1 litre by the addition of 450 ml of absolute alcohol and the balance of distilled water.Run 0, 5, 10 and so on up to 30 ml of this solution into seven 50-ml calibrated flasks and dilute to the marks by adding a 10 to 45 to 48 mixture by volume of acetone, alcohol and water. Add 2.0ml of N sodium hydroxide solution to each flask and after well mixing and setting aside for 10 minutes, measure the optical density exactly as described above. From this calibration curve relating optical density to mg of 2 :4-dihydroxybenzophenorie per 50 ml of final solution and covering the range 0 to 3 mg (Le., 0 to 0.6 per cent.of 2:4- dihydroxybenzophenone for 1 g of sample) , calculate the 2 :4-dihydroxybenzophenone content of the original sample.Feb., 19531 AND DETERMINATION OF ULTRA-VIOLET ABSORBERS IN POLYMERS 101 The calibration graph that we obtained by the above method is a smooth curve passing through the following points- 2 : 4-Dihydroxybenzophenone per 50 ml of final solution, mg . . 0.0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0 Indicator drum reading . . . . 0.0, 0.045, 0.083, 0.120, 0.150, 0.180, 0.208 SPECTROSCOPIC EXAMINATION DElEIUvII?r’ATIOY OF METHYL SALICYLATE- Preparation of test solution-Accurately weigh 0.25 to 0-35 g of finely divided sample into a 100-ml calibrated flask, add 60 ml of chloroform B.P. and shake at once to swell the polymer. Agitate mechanically until the solution clears ; 30 minutes is usually adequate.Make up the solution to 100ml with chloroform B.P. and shake for a few minutes until homogeneous. Procedzwe-Transfer a portion of the test solution to a l-cm cell, cover with a ground- glass lid and measure the density at A,,,. = 308 mp against a paired cell containing chloroform from the same batch as that used in the preparation of the solution. Take triplicate measurements. Clean the cell thoroughly with chloroforni, dry in a stream of filtered air and repeat the measurements on a fresh portion of the sample solution. From the mean value of the density readings obtained, calculate the corresponding Ei:m value by the following relationship- The resulting clear solution is used for the test. where D = optical density, and 1 = cell length in centimetres.Read off the methyl salicylate content in the original sample from a calibration graph relating EiFm a t 308 mp to percentage of methyl salicylate in polymethyl methacrylate covering the range 0 to 1 per cent. prepared as described below. Weigh 3-0 g of polymethyl methacrylate drillings, transfer into a 500-ml calibrated flask and dissolve in 300ml of chloroform by mechanical shaking. Make up the volume to 500 ml. To 25-ml aliquots of this solution contained in eleven 100-ml calibrated flasks add 0, 1.0, 2.0, 3.0 . . . 10.0 ml of a standard solution of methyl salicylate in chloroform B.P. The concentration of this solution should be such that 1 ml is equivalent to 0.15 mg of methyl salicylate. Make up all the solutions to 100ml with chloroform.Measure the density of each of these solutions at 308 mp and calculate the Ei!m values exactly as described in the method above. Plot a calibration curve relating E:2m at 308mp to percentage of methyl salicylate calculated as a percentage of the total polymethyl methacrylate plus methyl salicylate. We have found that under these conditions the calibration graph is a smooth curve passing through the following points- c = concentration of sample in g per 100 ml of solution Methyl salicylate, per cent. . . * . . . 0.0, 0.20, 0.40, 0.60, 0-80, 1.00 E:?m at 308 mp . . .. .. . . 0.02, 0-57, 1-12, 1.67, 2-22, 2.77 DETERMIXATION OF PHENYL SALICYLATE- Procedure-Carry out the procedure for the determination of phenyl salicylate by following, in general, the lines adopted for the corresponding determination of methyl salicylate except that- Prepare a test solution containing 0.15 to 0.20 g of accurately weighed sample per 100 ml of final solution.Measure the density of the test solution at 312 mp and calculate the corresponding value of E:Tm. Read off the phenyl salicylate content in the original sample from a prepared calibration graph relating Ei?m at 312mp to percentage of phenyl salicylate in polyme t h yl met hacrylate. Prepare a calibration curve relating Ei:m a t 312 mp to percentage of phenyl salicylate in polymethyl methacrylate to cover the range 0 to 2 per cent. Use a standard solution of phenyl salicylate such that 1 ml = 0.30 mg of phenyl salicylate.102 through the following points- HASLAM, GROSSMAN, SQUIRRELL AND LOVEDAY : THE DETECTION [Vol.78 The calibration graph for phenyl salicylate is, in our experience, a smooth curve passing Phenyl salicylate, per cent. . . .. . . 0.0, 0.40, 0.80, 1.20, 1.60. 2.00 E:zm at 312 mp . . .. .. . . 0.02, 0.98, 1.93, 2-89, 3.85, 4-82 DETERMINATION OF 2 :4-DIHYDROXYBENZOPIIENONE- Procedure-Carry out the procedure for the determination of 2 :4dihydroxybenzophenone by following, in general, the lines adopted for the determination of methyl salicylate, except (i) Prepare a test solution containing 0.15 to 0.20g of accurately weighed sample per 100 ml of final solution. (ii) Measure the density of the test solution at 289 and 324mp and calculate the corresponding values of E:2m at these wavelengths.(iii) Read off the 2:4-dihydroxybenzophenone content in the original sample from a prepared calibration graph relating E:Fm at 289 mp to percentage of 2:4-dihydroxy- benzophenone in polymethyl methacrylate. Check this result in a similar way by using at 324 mp. (iz) Prepare calibration curves relating EiZm at 289 and 324mp to percentage of 2:4- dihydroxybenzophenone in polymethyl methacrylate to cover the range 0 to 0.5 per cent. Use a standard solution of 2:4-dihydroxybenzophenone such that 1 ml is equivalent to 0.075 mg of 2 :4-dihydroxybenzophenone. The calibration graph for 2 :4-dihydroxybenzophenone is, in our experience, a smootli that- curve passing through the following points- 2:4-Dihydroxybenzophenoiie, per cent. . . 0.0, 0.10, 0.20, 0.30, 0.40, 0.50 .... . . 0.02, 0.62, 1.25, 1.86, 2-48, 3.09 E:Fm at 324 mp . . .. . . . . 0.01. 0.43, 0.86, 1-28, 1.70, 2.12 El% lcm at 289 m p . , DETERMINATION OF STILBENE- general, the lines adopted for the determination of methyl salicylate, except that- Proccdzwe-Carry out the procedure for the determination of stilbene by following, in (i) Prepare a test solution containing 0.20 to 0.25 g of accurately weighed sample per 250 ml of final solution. (ii) Measure the density of the test solution at 299 and 310mp and calculate the corresponding values of EiFm at these wavelengths. (iii) Read off the stilbene content in the original sample from a prepared calibration graph relating E:5m at 299 mp to percentage of stilbene in polymethyl methacrylate. Check this result in a similar way by using E:&, at 310mp.(iv) Prepare calibration curves relating E:2m a t 299 and 310mp to percentage of stilbene in polymethyl methacrylate to cover the range 0 to 0.5 per cent. Use a standard solution of stilbene (symdiphenylethylene) such that 1 ml is equivalent to 0.075 mg of stilbene. Measure standard synthetic solutions containing 0 to 0.25 per cent. and 0.25 to 0-50 per cent. in 1.0 and 0.5-cm cells, respectively. The calibration graph for stilbene is, in our experience, a smooth curve passing through the following points- Stilbene, per cent. . . . . . . . . 0.0, 0.10, 0.20, 0.30, 0.40, 0.50 E:& at 299 mp . . . . .. . . 0.02, 1.47, 2.92, 4.37, 5.82, 7-27 E:2m at 310 mp . . . . . . . . 0.02, 1.45, 2.88, 4.31, 5.74, 7.17 DETERMINATION OF DIBUTYL PHTHALATE IN POLYMETHYL METHACKYLATE SHEET- Particular attention has been directed in this laboratory to the determination of small amounts, i.e., less than 1 per cent., of dibutyl phthalate.The determination of dibutyl phthalate plasticiser, normally present to the extent of about 5 per cent. in plasticised polymethyl methacrylate, has been previously described by Haslam and Soppet.2 The principle of the method involves solution of the sample in acetone, with subsequent precipitation of the polymer with light petroleum (b.p. 40" to 60" C) , filtration and recovery of the dibutyl phthalate by evaporation oi the light petroleum solution. With samples containing small amounts of plasticiser and particularly with methylFeb., 19531 AND DETERMINATION OF ULTRA-VIOLET ABSORBERS IN POLYMERS 103 methacrylate - ethyl acrylate interpolymer compositions, the result obtained for the small percentage of dibutyl phthalate present may not be as accurate as is desirable because of the presence in the recovered dibutyl phthalate (i) of low molecular weight polymer and (ii) of small amounts of other additives in the polymer.For these reasons a direct ultra- violet absorption method of determining dibutyl phthalate in polymethyl methacrylate sheet has been worked out, and details of this method are described below. Preparation of test solutiow-Accurately weigh 0.40 to 0.50 g of finely divided sample into a 100-ml calibrated flask, add 60 ml of chloroform B.P. and shake a t once to swell the polymer. Transfer the flask to a mechanical shaker and agitate until a clear solution is obtained; 30 minutes is usually adequate. Make up the solution to 100 ml with chloroform and shake for a few minutes until homogeneous.The resulting solution is used for the test. Procedwe-Transfer a portion of the test solution to a 2-cm cell, cover with a ground- glass lid and measure the density on a spectrophotometer at A,,,x. = 276 mp against a paired cell containing chloroform from the same batch as that used in the preparation of the solution. Measure the optical density in triplicate. Clean the cell thoroughly with chloroform, dry in a stream of filtered air and repeat the above measurements on a fresh portion of the sample solution. From the mean of the density readings, calculate the corresponding E:2m value by the following relationship- D e x 1 E:Zm ==- where D = optical density, and 1 == cell length in centimetres.Read off the dibutyl phthalate content in the original sample from a calibration graph relating Ei2m at 276 mp to percentage of dibutyl phthalate in polymethyl methacrylate covering the range 0 to 1 per cent. prepared as described below. Weigh 5.0 g of drillings of polymethyl methacrylate sheet, transfer to a 500-ml calibrated flask and dissolve in 300 ml of chloroform B.P. by shaking mechanically. Make up the c == concentration of sample in g per 100ml of solution TABLE V ANALYSIS OF POLYMER SOLUTIONS CONTAINING ULTRA-VIOLET ABSORBERS AND DIBUTYL PHTHALATE Compound added Methyl salicylate . . . . 79 . . . . Stilbene . . .. .. 99 . . .. . . Phenyl salicylate .. . . 99 .. .. 2 : 4-Dihydroxybenzophenone Resorcinol monobenzoate . . 77 77 Dibutyl phthalate . . . . 99 . . .. By chemical tests Added, Found, % % . . 0.58 0-63 .. 1.00 1.02 . . . . . . . . . . . . . . .. . . .. - 0.49 0.89 0.30 0.40 0.50 0.50 0.50 - 0.53 0.83 0-29 0.36 0.45 Qualitatively detected Qualitatively detected - By spectroscopic tests Added, Found, % % 0.39 0.38 1.08 1.10 0.29 0.29 0.25 0.24 0.55 0.53 0.99 0.99 0.25 0.24 0.34 0-30 Qualitatively detected 0.64 0.64 - - 1.06 1.03 volume to 500 ml and shake until homogeneous. To 50-ml aliquots of this solution contained in eleven 100-ml calibrated flasks add 0, 1.0, 2-0, 3-0 . . . 10.0 ml of a standard solution of dibutyl phthalate in chloroform B.P. The concentration of this solution should be such that 1 ml is equivalent to 0.5 mg of dibutyl phthalate.Make up each solution to 100 ml with chloroform and shake well. Measure the optical density of each of these solutions in a 2-cm cell and calculate the EiZm value exactly as described above. Plot a calibration graph relating E:Zm a t 276mp to percentage of dibutyl phthalate calculated as a percentage of the total polymethyl methacrylate plus dibutyl phthalate.104 HzZSLL4M, GROSSMAN, SQUIRRELL AND LOVEDAY: THE DETECTIOK [VOl. 78 TABLE VI ANALYSIS OF POLYMERS COSTAIXING ULTRA-VIOLET ABSORBERS AKD PLASTICISERS Intended compositions of polymers TJnplasticised polymethyl meth- acrylate containing no ultra- violet absorber Polymethyl methacrylate con- taining 0.5 per cent. of dibutyl phthalate Unplasticised polymethyl meth- acrylate containing 0.5 per cent.of methyl salicylatc Polymethyl methacrylate con- taining 0.5 per cent. of methyl salicylate and 0.5 per cent. o f dibutyl phthalate Unplasticised polymethyl meth- acrylate containing 0.5 per cent. of phenyl salicylate Polymethyl methacrylate con- taining 0.5 per cent. of phenyl salicylate and 0-5 per cent. of dibutyl phthalate Unplasticised polymethyl meth- acrylate containing 0.5 per cent. of 2 :4-dihydroxybenzophenone Unplasticised polymethyl meth- acrylate containing 0.5 per cent. of stilbene Sample 1 2 3 4" 6 6 7* 8, 9*, lo*, 11 12 Chemical results No evidence of salicylate, 2:4- dihydroxybenzophenone or re- sorcinol monobenzoatc Phthalate detected 0.50 per cent. inethyl salicylate 0.49 per cent. incthyl sslicylate 0.47 pcr cent.phcnyl salicylatc. 0.46 per ccnt. phcnyl salicylate 0.48 per cent. 2 :4-dihydroxy- benzophenone No evidence of salicylate, 2:4- dihydroxybenzophenone or re- sorcinol inonobenzoate Spectroscopic results No evidence of plasticiser or ultra-violet absorber 0.50 per cent. tlibutyl phthalate 0.49 per cent. inethyl salicylate 0.49 per cent. methyl salicylate 0.43 per cent. phenyl salicylate 0.43 per cent. phenyl alicylate 0.49 per cent. 2 :4-dihydrosy- bemophenone 0.44 per cent. stilbene TABLE VII ANALYSIS OF COMMERCIAL ANT) UNICNOWN SAMPLES Chemical Evidence of salicylate but not phenyl or cresyl. 0.49 per cent. as methyl salicylate No evidence of salicylates, 2 :4-dihydroxy- benzophenone or resorcinol mono- benzoate Evidence of phenyl salicylate 0-98 per cent.Evidence of salicylate but not phenyl or cresyl. 0-50 per cent. as methyl salic ylate Evidence of 2 :4-dihydroxybenzophenone 0.26 per cent. No evidence of salicylates, resorcinol monobenzoate or 2 :4-dihydroxybenzo- phenone, but strong evidence of presence of a phenolic or cresylic con- stituent. Suspected surface contamina- tion-later proved Evidence of salicylate but not phenyl or cresyl. 0.52 per cent. as methyl salic ylate No evidence of salicylates, 2 :4-dihydroxy- benzophenone or resorcinol mono- benzoate Evidence of 2 :4-dihydroxybenzophenone 0.12 per cent. Spectroscopic Characteristic of methyl salicylate, 0.49 per cent. Xo evidence of dibutyl phthalate or ultra- violet absorbers Characteristic of phenyl salicylate, 0-0s Characteristic of methyl salicylate, 0.50 per cent.per cent. Characteristic of 2 :4-dihydroxybenzo- phenone, 0.30 per cent. Small amounts of unknown absorbing material detected. Later shown con- ducive to presence of phenolic or cresylic constituent Characteristic of methyl salicylate, 0.50 per cent. No evidence of dibutyl phthalate or ultra- violet absorber Characteristic of 2 :4-dihydroxybenzo- phenone, 0.11 per cent. * These samples are methyl methacrylate - ethyl acrylate co-polymers.Feb., 19531 AND DETERMINATION OF ULTRA-VIOLET ABSOKHEKS IN POLYMERS 105 Under these conditions we have found that the calibration graph is a smooth curve passing through the following points-- Dibutyl phthalate, per cent. . . . . 0.0, 0.20, 0.40, 0.60, 0.80, 1.00 E:Fm at 276 mp . . .. . . . . 0.02, 0.11, 0.19, 0-28, 0.37, 0.46 RESULTS The methods described for the determination of dibutyl phthalate and ultra-violet absorbers in polymethyl methacrylate-type polymers have been tested by- (i) Analysis of synthetic solutions containing known weights of pure polyrnethyl methacrylate and ultra-violet absorbers or dibutyl plithalate dissolved in acetone for the chemical tests and chloroform B.P. for the spectroscopic tests. The results shown ill Table V were obtained. (ii) Analysis of polymers prepared with the object of the final compositions containing known amounts of ultra-violet absorber and plasticiser. The intended compositions of the polymers, together with the results obtained, are shown in Table 1’1. (iii) Analysis of commercial and unknown samples, the results of which indicate the agreement between the chcmical and physical figures, as shown in Table VIT.NOTES ON THE TESTS- methyl methacrylate - ethyl acrylate co-polymers. are being taken; for example, methyl salicylate is quite distinctive. to the Millon’s and indophenol tests as shown in Table VIII. The methods described above have been applied with equal success to the analysis of It is always desirable to pay particular attention to the odour of the sample when drillings When cresyl esters are suspected it may be noted that they differ in their reactions TABLE 1’111 REACTIONS OF CRESYL ESTERS Millon’s 2 :6-Di bromoquinonechloriniidc Homologue test test Phenol . . .. .. . . Pink colour deepening in 5 to 10 Deep blue developing in 5 minutes minutes o-Cresol . . .. . . . . Pale yellow-pink deepening to Immediate pronounced blue yellow-brown changing to deep purple p-Cresol .. . . . . Immediate pink deepening to deep No colour claret In-Cresol . . . . . . Claret colour deepening Pronounced blue Such substances as hydroquinone monobenzoate and catechol monobenzoate behave as shown in Table IX in the appropriate Millon’s and indophenol tests, as compared with resorcinol monobenzoate. TABLE IX REACTIONS OF MONORESZOATI~S OF HYUROQUINONE AND CATJ3CI-IO1, COMPAliED WITH THOSE OF RESORCINOL MONO13ENZOATE Millon’s 2 :6-Dibromoquinonechlorimide test test Resorcinol monobenzoate . . Immediate yellow precipitate Immediate violet-deepening Hydroquinone monobenzoate Iinmediate yellow precipitate Fairly rapid reduction to brown- Catechol monobenzoate . . Immediate yellow precipitate As for hydroquinone, but much blac k---darkening more slowly The absorption maxima of laboratory prepared specimens of these compounds in ethyl alcohol solution are as follows: catecliol monobenzoate, Amax. = 273 mp; resorcinol mono- benzoate, Amax. = 274 mp; hydroquinone monobenzoate, Amax. = 375 mp.106 HUNTER AND HALL: THE DETERMINATION OF [Vol. 78 Although we have no experience of the application of the method, it seems certain that the method described by Nicholls3 for the determination of small amounts of benzoic acid by conversion to salicylic acid with hydrogen peroxide and ferric chloride could be used for the determination of resorcinol monobenzoate after an appropriate extraction and hydrolysis. All absorption spectra measurements were made with the Hilger Uvispek spectro- photometer. REFERENCES 1. 2. 3. Kolthoff, I. M., and Harris, W. E., I.lzd. Eng. Chenz., Anal. Ed., 1946, 18, 161. Haslam, J., and Soppet, W. W., Analyst, 1950, 75, 63. Nicholls, J. R., Ibid., 1927, 52, 19. IMPERX AL CHEMICAL INDUSTRIES LIMITED PLASTICS DIVISION WELWYN GARDEN CITY, HERTS. June loth, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800092

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

106 HUNTER AND HALL: THE DETERMINATION OF [Vol. 78 The Determination of Calcium in Plants and Soils BY J. G. HUNTER AND A. HALL The method is based on the turbidity formed on adding a precipitating reagent to a sodium acetate - acetic acid buffer solution containing the calcium. The precipitating reagent consists of a solution of ammonium oxalate and citric acid in water, ethanol and butan-1-01. The reagent is stable for 8 hours, and the turbidity, which is determined by means of a photo-electric absorptiometer, is constant from 5 to 60 minutes after adding the reagent. The method, which was devised for calcium in plant and soil extracts, determines from 0-05 to 0-50mg of calcium with an accuracy of within rf 5 per cent. Permissible concentration limits of certain interfering ions are given, and examination of the amounts of these ions in plant and soil extracts indicates that significant interference will seldom be encountered. METHODS previously published for the determination of calcium 011 a semi-micro scale by a turbidimetric procedure have been speedy, but have low accuracy; in our experience: a 20 per cent.error is not exceptional. The simplicity of the turbidimetric procedure is a great asset, and investigation showed that the accuracy could be considerably increased. The accuracy of the method now proposed is satisfactory (within &5 per cent.), and the manipula- tive simplicity inherent in the turbidimetric procedure is retained. METHOD PHEF-\RATION OF THE SAMPLE- The object of the preliminary treatment of the sample is to produce a solution containing no excessive amounts of organic matter or interfering ions and one in which the calcium concentration is such that 0.05 to 0.50 mg of calcium is contained in easily handled aliquots.In addition, the calcium extractant should be easily removable so that it can be replaced by Morgan’s reagent,l the composition of which is given below (p. 107); if Morgan’s reagent is itself the extractant, the required aliquot should be not more than 5ml. PZmt tissues-Ash the sample and dissolve it in dilute hydrochloric acid ; transfer a convenient aliquot to the reaction flask, evaporate to dryness and dissolve the residue in exactly 5 ml of Morgan’s reagent. Plant tissues that have been extracted with Morgan’s reagent are used without treatment other than removal of excessive amounts of organic matter by activated carbon (Hester2). The carbon must contain no extractable calcium; purify ordinary grades by washing with acid.Soils-Use solutions prepared by extracting the soil with Morgan’s reagent, described by Peech and Engli~h,~ as indicated above.Feb., 19531 CALCIUM I N PLANTS AND SOILS 107 If the extract is prepared with neutral N ammonium acetate (Pipe+) or 0-5 N acetic acid (Williams5 and Williams and Stewart6), evaporate to dryness and dissolve the residue in a convenient volume of dilute hydrochloric acid; take the aliquot from this solution. Alternatively, evaporate a convenient aliquot of the initial extract to dryness in the reaction flask and dissolve the residue in 5 ml of Morgan's reagent.With either technique, destroy excessive amounts of organic matter by adding small amounts of 20-volume hydrogen peroxide as the evaporaiion nears completion. REAGENTS- Morgan's reagent1-Dissolve 100 g of hydrated sodium acetate and 30 ml of glacial acetic acid in water and dilute to 1 litre. Precipitating reagent-A mixture of 95 per cent. ethanol (rectified spirits), butan-1-01 (technical n-butyl alcohol*), 0-5 per cent. w/v ammonium oxalate solution and 25 per cent. w/v citric acid solution. Mix 375ml of 95 per cent. ethanol with 125ml of butan-1-01. After 5 minutes, add 500 ml of 0-5 per cent. ammonium oxalate solution rapidly to the mixture and stir thoroughly. After a further 5 minutes, pour this mixture rapidly into 40ml of 25 per cent. citric acid solution and mix.The reagent is ready for use 1 hour after preparation and is stable for a further S hours. Calcium stock reagent-Prepare a solution containing 2000 parts per million of calcium by adding 4.994 g of calcium carbonate to approximately 100 ml of water in a covered beaker and then add approximately 5 ml of glacial acetic acid. Warm the mixture and add approxi- mately 500 ml of Morgan's reagent. Heat and stir the mixture until the carbonate dissolves. Cool and dilute accurately to 1 litre with Morgan's reagent. Calcium standard solution-Prepare a solution containing 100 parts of calcium per million by diluting accurately 50 ml of calcium stock reagent to 1 litre with Morgan's reagent. PROCEDURE- Place an aliquot containing from 0.05 to 0.50 mg of calcium in the 100-ml conical tlask and treat it as described above for preparing the sample, so that ultimately the calcium is dissolved in 5 ml of Morgan's reagent.Add 25ml of the precipitating reagent to the solution whilst shaking the flask and contents. Set the mixture aside for not less than 5 minutes and not more than 1 hour, and during that period determine the turbidity; a Hilger Spekker photo-electric absorptiometer with 4-cm cells and glass turbidity filters is suitable for the determination. Construct the graph for converting the absorptiometer readings to concentrations, by diluting 0, 1 , 2, 3, 4 and 5 ml of the calcium standard solution to 5 ml with Morgan's reagent, adding the precipitating reagent, and determining the intensity of the turbidity produced under the conditions described above. Noms ON THE METHOD- The calibration readings vary owing to minor variations in the preparation of the precipitating reagent, For example, 0.30 mg of calcium may give an absorptiometer reading between 0.37 and 0.27, and sometimes the reading for 0.50 mg of calcium is unusually low and slightly off-scale, i.e., less than 0.This effect, however, is related to the precipitating reagent, and is compensa.ted for by making a calibration graph for each batch of precipitating reagent prepared. The effect of temperature (from 10" to 30" C) on the results was found to be negligible. Typical calibration readings are as follows- Calcium, mg . . . . Nil 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Spekker reading . . 0.99 0-92 0.81 0.67 0.53 0.41 0.31 0.21 0-12 0.06 0.02 INTERFERENCE FROM OTHER IONS The effect of certain ions on the method was investigated, and it was found that some of these influenced the results when present in amounts outside the tolerance limits given in Table I.From Table I1 it is seen that high calcium values result from excessive amounts of magnesium, iron, manganese oi- aluminium in the aliquot. Within the established limits * Supplied by British Ind.ustria1 Solvents, Carshalton, Surrey.10s HUNTER AND HALL: THE DETERMINATION OF [Vol. 78 iron and manganese can be controlled by citric acid, which, in addition, contributes to the precision of the method. TABLE I ION TOLERANCE LIMITS Permissible amounts (milligrams) o f ions in the aliquot taken for analysis Calcium .. . . 0.05- 0-30 Magnesium . . . . 0-0.75 Sodium . . . . 0-5.0" Iron . . . . . . 0-2.0 Chloride . . . . 0-5.0" Manganese . . . . 0-0-25 Nitrate . . . . 0--5.0 * Aluminium . . . . 0-2.0 Phosphate . . . . 0-5-0" Ammonium . . . . 0--5.0* Sulphate . . . . 0 -5.0" Potassium . . . . 0 -5o* *: The effect of greater amounts was not investigated. TABLE I1 ION INTERFERENCE Effect oi magnesium, iron, manganese and aluminium on the determination of calcium Calcium present, mg 0.300 0.300 0.300 0.300 0.300 0-300 0.300 0.300 0-300 0.300 0.300 0.300 Magnesium Iron Manganese added, added, added, mg mg mg 0.50 0-75 1 so0 1.00 1.50 2.00 3.00 0.10 0.25 0.50 Aluminium added, mg 2.00 3.00 Calcium found , mg 0.300 0.310 0-320 0.300 0.310 0.305 0.330 0.300 0.310 0.330 0.310 0.325 Percentage error 0 3.3 6.7 0 3.3 1.7 10.0 0 3.3 10.0 3.3 8-3 The tolerance limits are not often exceeded in solutions as usually prepared for analysis.More calcium than magnesium is present in most plants or plant organs, other than seeds (Cooper, Paden and Garman'), and in most soils (Russells) ; interference from magnesium is therefore unlikely. Hunterg summarises the position with regard to interference from iron, manganese and aluminium in the determination of magnesium, and concludes that only in exceptional plant analyses will the tolerance limits be exceeded; as the tolerance limits of these elements in the magnesium determination are equal to or less than those in the calcium, and the amounts of calcium present are usually greater than the amounts of magnesium, the possibility of interference from these elements in the determination of calcium in plants is remote.Hunter9 points out that in soil analysis the concentration of interfering ions in an extract will depend not only on their content and form in the sample, but also on the method used to prepare the extract. He tabulates the calcium, iron, manganese and aluminium con- centrations in extracts of twelve soils, in some of which the calcium concentrations were low and interference therefore liable to occur; the extracting reagents used were (i) neutral N ammonium acetate solution (Pipe+), (ii) Morgan's reagent, pH 4.8 (used as described by Peech and English3 but with the extraction period increased to 2 hours), and (iii) 0.5 N acetic acid, pH 2-5 (Williams6).In none of these extracts is the concentration of magnesium, iron or aluminium high enough to interfere with determinations of calcium by the proposed method, and in only one acetic acid extract is the concentration of manganese excessive. Hence it can be taken that significant interference seldom arises in soil analysis.Feb., 191531 CALCIUM I N PLANTS AND SOILS 109 ACCURACY To determine the accuracy of the method, solutions containing known amounts of calcium were prepared and their calcium contents determined twenty times. The calcium contents of the hydrochloric acid extract of hay ash and a Morgan's reagent extract of a soil were also determined twenty times. The results of a statistical analysis of the values obtained are shown in Table 111.TABLE I11 Calcium per determina- tion, mg 0.05 0.10 0.20 0.30 0.40 0.45 0.50 ACCURACY OF THE METHOD Statistical results for replicate determinations Xumbcr of determina- tions 20 20 20 20 20 20 20 Standard deviation ot 0.00134 0.00138 0.002 16 0.00351 0-00'300 0.0156 Standard error of the mean 0 0.000300 0.000309 0.000483 0-000785 0~002012 0-003488 1 Percentage error* (mean of 20 3bservations) 0 0.86 0-44 0.46 0.56 1.28 2.00 Percentage error* (mean of 2 observations) 0 2.71 1.40 1.46 1.75 4.05 6.3 1 294: 20 0.896 0.200 0.19 0.62 536: 20 6.85 1-53 0.82 2.59 * This error will not be exceeded 99 times out of 100. Results identical. $ Milligrams of calcium per 100 g of plant dry matter or dry soil. Type of solution Standard 9 9 Plant extract Soil extract I t will be seen that the standard deviation and standard error of the mean for each set of standard determinations increase with the amount of calcium present, indicating that greater variations can be expected with larger amounts of calcium.In order to estimate on a percentage basis the degree of accuracy at different calcium concentrations, the product of each standard error and the appropriate t value is expressed as a percentage of the mean determined. In this way the percentage accuracy that will be obtained 99 times out of 100 was estimated and is recorded in Table I11 for the means of both 20 and 2 determinations. The values show that, for a mean of 20 determinations, 99 times out of 100 the result will be within 1 per cent. of the true value with 0.05 to 0-40 mg of calcium, and within 3 per cent.with 0.40 to 0.50 mg of calcium. The results for the plant and soil extracts also indicate a high measure of reproducibility. The calcium contents of numerous samples of plant tissue (ash extracted by hydro- chloric acid) and soil (extracted by 0.5 N acetic acid) were determined by the method described. A comparison of these results with those obtained by an established macro- volumetric method (Piper4) and flame photometry* (MitchelllO) showed good agreement. Results by the proposed method and Piper's method are compared in Tables I V and V. TABLE IV DETERMINATION OF CALCIUM I N PLANTS Calcium per 100 g of dry matter (-------.-.--.. 7 By macro-volumetric By turbidimetric Material determination, determination, mg mg Tomato leaves .. .. .. 5340 5366 .. .. .. 367 350 .. .. .. 1831 1873 Swede bulbs. Swede leaves . . . . .. 341 325 Oat straw . . .. .. .. 309 300 Hay Oat grain . . .. .. .. 109.2 110.1 Percentage difference + 0.49 - 4.63 + 2-29 - 4.69 -2.91 + 0.82 * These results were supplied by Dr. R. L. Mitchell, Department of Spectrochemistry, The Macaulay Institute for Soil Research.110 Soil PH 4.9 4.6 4.3 4.9 4.6 4.8 5-3 7.0 6.8 HARVEY : MICRO-DETERMINATION OF TABLE V DETERMINATION OF CALCIUM IN SOILS 0.5 N acetic acid extraction Calcium per 100 g of soil A I \ Percentage loss By macro-volumetric By turbidimetric on ignition determination, determination, mg mg 4 12.3 12.9 8 70.1 68.1 8 122.4 116.8 7 131.2 130.6 54 272.5 275.0 21 56-3 58.8 26 69-8 67-8 9 640.0 639.0 10 182.6 187.5 [Vol. 78 Percentage difference + 4.88 - 2.85 - 4.58 - 0.46 + 0.92 + 4.44 - 2.87 -0.16 + 2.68 The degree of accuracy of the method can therefore be considered satisfactory by the standards usually recognised in this type of work. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Morgan, M. F., Conn. Agric. Exp. Sta. Bull., 1941, No. 450, 579. Hester, J. B., J . Amer. SOC. Agron., 1940, 32, 549. Peech, M., and English, L., Soil Sci., 1944, 57, 167. Piper, C. S., “Soil and Plant Analysis,” The University, Adelaide, 1944. Williams, R., J. Agric. Sci., 1928, 18, 439. Williams, E. G., and Stewart, A. B., J . SOG. Chem. Ind., 1941, 60, 291. Cooper, H. P., Paden, W. R., and Garman, W. H., Soil Sci., 1947, 63, 27. Russell, E. J., “Soil Conditions and Plant Growth,” Eighth Edition, Longmans, Green & Co., Hunter, J. G., Analyst, 1950, 75, 91. Mitchell, R. L., Spectrochim. Acta, 1950, 4, 62. CRAIGIEBUCKLER London, 1950. THE MACAULAY INSTITUTE FOR SOIL RESEARCH ABERDEEN June 19th, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800106

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

110 HARVEY : MICRO-DETERMINATION OF [Vol. 78 Micro-Determination of Phosphorus in Biological Material BY H. W. HARVEY An absorptiometric method for determining microgram amounts of phosphorus in small samples of algal growths has been devised and tested. It is based on the conversion of the phosphorus to phosphomolybdic acid and reduction to molybdenum blue by stannous chloride. The range of the method is from 1 to 70 pg of phosphorus. IN connection with another research it was found necessary to determine phosphorus in uni-cellular algae, which had been separated by centrifugation and contained a few micrograms of phosphorus. The method proposed is suitable for 1 to 70-pg amounts of phosphorus. I t consists in decomposing the organic material with sulphuric acid and hydrogen peroxide, heating with 5 N acid to convert the pyrophosphate to orthophosphate, decomposing residual hydrogen peroxide with sulphite and converting the orthophosphate to phosphomolybdic acid.The phosphorus is then determined by the molybdenum-blue method after controlled reduction by stannous chloride. REAGENTS- METHOD SuZphuric acid-A 50 per cent. v/v solution. Hydrogen peroxide-A 100-volume solution.Feb., 19531 PHOSPHORUS IN BIOLOGICAL MATERIAL 111 Sodium sulphite-An aqueous solution containing 33 g of Na2S0,.7H20 per 100 ml. Acid molybdate solution-Dissolve 10 g of ammonium molybdate in 100 ml of water and Hydrochloric acid-Dilute 5 ml of concentrated hydrochloric acid to 100 rnl with water. Iodine solution-A 0.033 N solution of iodine in dilute potassium iodide solution.Stannous chloride (stock solution)-Dissolve 40 g of SnC12.2H,0 (clean unoxidised crystals) (This solution is reasonably Stannous chloride (dilute solution)-Add about 0.7 ml of stock solution to 20 ml of 5 per This provides 2mg of stannous ions in about 0.75m1, and this The stock solution is conveniently In order to ascertain the volume of dilute solution containing (This solution only add to 130 ml of concentrated sulphuric acid diluted with 170 ml of water. in 50 ml of concentrated hydrochloric acid and 50 ml of water. st able.) cent. hydrochloric acid. amount decolorises 1 ml oi 0.033 N iodine solution. added from a blood pipette. 2 mg of tin, it is necessary to titrate it against 1 ml of iodine solution. retains its titre for a few h0urs.l) APPARATUS- Hard-glass test tubes or centrifuge tubes-These must be not less than 12 cm long.For the determination of less than 1 pg of phosphorus, transparent silica tubes are preferable.2 Hard-glass $asks-Capacity 120 to 200 ml, with beakers to cover them. (In order to rid new tubes and flasks of phosphate, moisten with sulphuric acid and heat in an oven before washing with distilled water.) Pipettes-Two graduated pipettes of capacity 1 ml, provided with rubber teats (blood pipettes), and one 2-ml pipette, preferably of the syringe type.3 Thin quill tube-Drawn out and cut off. PROCEDURE- Add 0.2 ml of 50 per cent. v/v sulphuric acid to the organic matter in a hard-glass centrifuge tube or test tube, heat until charring starts, cool, add 1 drop of hydrogen peroxide and heat over a micro burner until fumes of sulphuric acid just fill the lower part of the tube.If the liquid is not then colourless add another drop of hydrogen peroxide and heat again. The resulting liquid contains orthophosphoric acid with a small amount of pyrophosphoric acid. ,4dd 0.6 ml of water to reduce the acid concentration to 5 N and heat to boiling twice. This treatment converts almost all of the phosphate to the ortho state (see p. 112). Transfer the contents of the tube to 100 ml of distilled water contained in a wide-necked flask protected from dust by an inverted beaker, add 0-75 rnl of sulphite solution, and then 2 m l of acid molybdate. The molybdate solution is conveniently added from a syringe pipette., Mix, and then, not less than 3 minutes after adding the acid molybdate, add a volume of the dilute stannous chloride solution containing 2 rng of tin from a blood pipette and mix at once.Between 5 and 7 minutes after adding the stannous chloride solution measure the optical density of the blue solution in a cuvette or vessel of suitable length, through a red filter (Chance OR1 or Ilford Spectrum red). The optical density per centimetre of light path through the liquid (Elem) bears an nlinost linear relation to the phosphate content. One or two reagent blanks are included in the series in order to correct for traces of extraneous phosphate. One or two tubes containing a known quantity of phosphate can be included in the series in order to ascertain the value of Elcm per pg of phosphorus, or this value may be determined independently. It remains constant for the reagent6 (after correcting for temperature), provided that the same amount of stannous ion is added and that the amounts of 50 per cent.sulphuric acid and of acid molybdate solution are reasonably constant (see p. 113). EXPERIMENTAL COMBUSTIOX- the acid fumes condense in the lower half of the tube. manner for a long time there was no loss. part of the tube is decomposed when the contents are washed into the sulphite solution. Marked to deliver about 0.2 ml. During the wet combustion there is no loss of phosphoric acid vapour, since all When phosphate was heated in this Any hydrogen peroxide condensing in the upper112 HARVEY: MICRO-DETERMINATION O F [Vol. 78 CONVERSION OF PYROPHOSPHORIC TO ORTHOPHOSPHORIC ACID- When orthophosphate was heated with acid and then (without conversion) washed into water, about 2 per cent.remained as the pyro acid, but when pyrophosphate was heated with acid, diluted with 0.6 ml of water and boiled, about 7 per cent. remained as pyrophosphate. Hence in the recommended procedure any loss of phosphorus by incomplete conversion of pyro to ortho acid is not likely to exceed 2 per cent. of 7 per cent., or, say, 0.2 per cent. ACCLJRACY OF DETERMINATION OF ORTHOPHOSPHATE-- When a solution containing phosphomolybdic acid was reduced by stannous chloride, the blue colour increased to a maximum and then slowly faded. As the concentration of phosphorus was increased, the maximum colour was reached in a shorter time and fading started sooner.In the recommended procedure, the intensity of the colour is measured after a fixed reaction time (5 to 7 minutes). In this manner a plateau value is reached when the concentration of phosphorus lies between 15 and 90 pg per 100 ml, but the intensity has not quite risen to its maximum with small (1 to 2-pg) concentrations. In addition to the effect of time, the intensity of the blue colour depends on the con- centration of tin and, when phosphate concentration is high, on the ratio of tin to phosphate. In consequence, the intensity of the colour developed with 2-0 mg of tin, after a reaction time of 5 to 7 minutes, departs slightly from strict linearity in the range 1 to 80 pg of phosphorus. However, neither this departure from linearity nor an increase in reaction time to 10 minutes after adding the stannous chloride appreciably affect the results.This was shown by the following experiment. A series of solutions containing different amounts of phosphate in distilled water was treated with sulphuric acid, sodium sulphite, molybdate reagent and stannous chloride as described in the procedure (p. 111). The observed optical densities and calculations are shown in Table I. In the last column the values found for additions of 20 pg of phosphorus to 100 ml of water are used as a basis for calculating the “phosphate found” at the other dilutions. The difference between quantity added and found includes volumetric and instru- mental errors. TABLE I EFFECT OF VARIATIONS IN REACTION TIME AT Phosphorus added to Internal 100 ml of length of water, cuvette, PE: cm Reaction time, 5 minutes- 0 15.0 1 Y> 2 ,, 4 ¶ > 20 2.0 20 7 9 48 1.0 80 >> Reaction time, 10 minutes- 0 15.0 1 >> 2 7, 4 7 7 E observed * 0-02 1 0.128 0.220 0.417 0.267 0.268 0.313 0.510 0.024 0.131 0,229 0.426 E1cm- (reagent Elcm blank) 0.0014 - 0-00854 0.00714 0.0 1467 0.01327 0.0278 0.0264 0.1335 0.1340 0.313 0.3116 0.510 0.5086 (reagent blank) “o::;;:> 0-0016 - 0.00873 0.007 13 0.0146 0.0130 0.0284 0.0268 (reagent blank) 13.5” C E1cm- (reagent blank) Per Pg of phosphorus - 0.007 14 0.00663 0.00660 0.00662 0.00649 0.00636 * The difference between the optical densities of the solution and distilled water. Phosphorus found, CLg - 1.07 2.0 3.9 Basis 47.1 76.8 - 1.07 1.97 4-04 The intensity of blue depends also on the concentration of stannous tin.The effect of varying the concentration of stannous ion on different concentrations of phosphate is shown in Table 11. It can be seen that the error caused by small differences in the concentration of stannous chloride is insignificant.Feb., 19531 PHOSPHORUS I N BIOLOGICAL MATERIAL TABLE I1 EFFECT OF VARIATIONS IN THE CONCENTRATION OF STANNOUS CHLORIDE FOR 2 TO 8 p g OF PHOSPHORUS 113 Phosphorus in 100 ml of reaction Internal length mixture, of cuvette, Pg cm About 2 15-0 3, 2 15.0 > ¶ 2 15-0 7' 20 2.0 >) 20 2.0 7) 20 2-0 93 80 1.0 '9 80 1.0 3) 80 1.0 Tin added, mg 1.7 2.0 2-3 1.7 2-0 2.3 1-7 2.0 2.3 Optical density 0.182 0.187, 0.186 0.188 0.257 0.259, 0.259 0.260 0.398 0.403, 0.404 0.408 The optical density is also dependent upon the ratio of acid to molybdate in the reaction Table I11 shows the effect of variation in the volume of 50 per cent. sulphuric The greater the volume added the less the optical density after 5 minutes.mixture. acid added. TABLE I11 EFFECT OF ACIDITY ON OPTICAL DENSITY Amount of 50 per cent. v/v sulphuric acid added, ml 0.1 0.15 0.2 0.25 0.3 Optical density of reaction mixture containing 20 pg of phosphorus, 6 to 7 minutes after adding 2 mg of tin, corrected for blank 0-272 0.267 0-260 0.258 0.250 As the addition of 0.2 in1 of 50 per cent. sulphuric acid can be made to within 0.01 ml, the error due to variations in the volume is small. In addition, there is some loss of acid by reduction of the sulphuric acid to sulphurous acid during the wet combustion of the sample.If a relatively large quantity of organic matter, e.g., 3 mg of sugar, is decomposed, and all the oxygen were derived from the sulphuric acid, the loss of sulphuric acid from this cause would amount to 0.022 ml; it would, in fact, be less, for some oxygen is supplied by the hydrogen peroxide. The combined error due to variations in the volume taken and loss of sulphuric acid during combustion does not seem likely to exceed about 1 per cent. Variations in the quantity of acid molybdate added from a syringe pipette were insignificant. The effect of temperature on colour development has been investigated2 34; it aniounts to an increase of 0.96 per cent. per 1°C. EFFECT OF INTERFERING SUBSTANCES- But in the proposed procedure in which the inorganic residue from 1 to 3mg of organic matter is contained in 100 ml, none of these is likely to be present at a concentration sufficient to cause significant error, with the possible exception of copper.The body fluids of several species of marine animals containing haemocyanin are relatively rich in copper, which has an outstandingly potent effect on the formation of molybdenum blue. The depression in optical density caused by additions of copper to a solution containing 20 pg of phosphorus is as follows- The formation of molybdenum blue is known to be hindered by several Copper added (as sulphate), pg . . .. .. .. 10 5 2 Optical density. Expressed as a percentage of that given by 20 pg of phosphorus . . .. . . 87.5 93.5 97.0114 COULSON AND HALES: THE DETERMINATION OF [Vol. 78 CONTAMINATION- Many thousand duplicate determinations of phosphate and of organic phosphorus in sea water made in this laboratory, and involving 1-pg amounts of phosphorus, have shown that, provided simple precautions are taken, the danger of contamination is slight. Contamination of the laboratory air by the phosphorus pentoxide evolved on igniting a lucifer match is sufficient to vitiate a determination. For samples containing more than 1 to 2 pg of phosphorus the error due to solution of phosphate from hot acid-washed Pyrex or Hysil glass is insignificant. With samples containing less than this amount, silica tubes are best (p. 111). The flasks required for this work should be washed thoroughly with water immediately after use. REFERENCES 1 . Armstrong, F. A. J., J . Mar. Biol. Ass., 1949, 28, 701. 2. Harvey, H. \V., Ibid., 1948, 27, 337. 3. __ , Analyst, 1951, 76, 657. 4. Kalle, K., Ann. d . Hydrog., 1934, 96. 5. Buch, K., and Ursin, M., Murcntutskimuslaifokscn Julknrsu Nari.sfol.skiiinjisinstituts Skvift, 19-18, U. Kalle, K., An)%. d. Hydrog., 1935, 58. CITADEL HILL, PLYMOUTH No. 140. THE LABORATORY First received, May 5th, 1952 Amended, . ] d y 172h, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800110

出版商:RSC    

年代:1953

数据来源: RSC