摘要:

628 STREET : DETERMINATION OF GLlJCOSE I N BIOLOGICAL FLUIDS [Vol. 83 Determination of Glucose in Biological Fluids with Ethylenediaminetetra-acetic Acid BY HAROLD ‘V. STREET (Biochemistry Laboratory, Department of Pathology, Crumpsall Hospital, Manchester, 8 ) The titration of bivalent copper in the presence of cuprous oxide with ethylenediaminetetra-acetic acid has been investigated. A method has been developed for determining glucose by titirating the unreduced copper (without removal of cuprous oxide) after the sample has been heated with an alkaline copper tartrate reagent. The method can be applied to blood, cerebrospinal fluid or urine and requires no more than 0.2 ml of sample. Recovery experiments indicate that the method has the same degree of accuracy as methods of the Schaffer - Xartmann type, but it is less time- consuming and the solutions are stable.THE reducing action of glucose on alkaline solutions of many metallic salts, e.g., those of copper, silver, mercury and bismuth, has been known for many years. Of these solutions, only those of copper are in general use for quantitative determination of glucose. As early as 1844, Barreswill described the use of tartrate to increase the stability of alkaline copper sulphate solutions, but it was Fehling2 who, in 1848, evaluated the proportions of copper sulphate, alkali and tartrate in the reagent that is still known by his name. Since Fehling’s time, many modifications of his reagent have been proposed. In nearly all these quantitative methods, cuprous oxide is produced by the action of glucose on alkaline copper tartrate mixture.The methods after formation of cuprous oxide may be classified into two groups as follows-. (i) Volumetric methods in which glucose solution is titrated against a known volume of boiling standardised copper solution until reduction is complete. (ii) Methods in which an excess of copper reagent is heated for a fixed time with the glucose solution and either reduced or unreduced copper is determined. The cuprous oxide formed may or may not be removed by filtration before the final determination. I have criticised3 the use of certain methods for determining reducing carbohydrates in presence of starch, and a study of the history of copper-reduction methods suggested the possibility of determining the excess of unreduced copper in presence of a finely divided precipitate of cuprous oxide by direct titration with ethylenediaminetetra-acetic acid (EDTA).The general theory of the use of EDTA for determining metals in solution has been adequately described by Schwarzenbach* and by Welcher.6 Flaschka6 and Amin7 have described the direct titration of copper with rnurexide as indicator. Eriochrome black T has been used as indicator in the determination of copper by back-titration of excess of EDTA with a standard solution of a manganous salt.8 Other indicators that have been proposed for copper include pyrocatechol violet,* :l-(2-pyridy1azo)-2-naphthol9 and Fast Sulphon black F.I0 EXPERIMENTAL PREPARATION AND TITRATION OF ALKALINE CONPPER REAGENT- Several different copper reagents were tried, and that finally selected has been stored in a polythene bottle for over 12 months without deterioration.The reagent includes tartrate as auxiliary complexing agent and is essentially Somogyi’sll high alkalinity reagent, but the iodate and iodide are omitted, It is prepared as described under “Reagents” on p. 631. The pH was followed during titration of the alkaline copper reagent solution with 0.005 M EDTA solution. Ten millilitres of the reagent solution to which 10 ml of distilled water had been added were used, and the pH was measured with a glass electrode. The pH remained constant a t 10.4 throughout the addition of 50 ml of the EDTA solution. The pH was also measured during titration with 0.005 M EDTA solution of a mixture of 1 ml of distilled water, 8 ml of Folin’s sodiuni sulphate - sodium tungstate solution,12 1 ml of 0.33 N sulphuric acid and 10 ml of alkaline copper reagent solution.The pH remainedNov., 19581 WITH ETHYLENEDIAMINETETRA-ACETIC ACID 629 constant a t 10.0 during the addition of 50 ml of EDTA solution. IlIurexide was added to the mixture during each of these titrations. The pH of the alkaline copper reagent solution is not decreased by addition of EDTA in amounts sufficient to chelate all the copper present, and, further, the presence of tungstate in the mixture has no significant effect on the titration curve. Hence, the tartrate in the reagent efficiently buffers the mixture under the conditions proposed for the final procedure. The adequacy of the tartrate as a buffer in this mixture is also apparent from theoretical considerations, as the alkaline copper reagent solution is approximately 0.1 M with respect to potassium sodium tartrate, and 1 ml of this solution is mixed with a maximum volume of 5 ml of 0.005 M EDTA solution. In other words, the pH remains sufficiently high after addition of tungstic acid for the alkaline copper reagent solution still to be suitable for oxida- tion of glucose.It is also apparent that the pH of the tungstic acid - alkaline copper reagent mixture is suitable for titration of the copper with 0.005 M EDTA solution when murexide is used as indicator and that the presence of tungstate ions has no effect on the colour change of murexide a t the end-point for copper. CONCENTRATION OF EDTA AND VOLUME OF REAGENT SOLUTION- The concentration of EDTA solution used for titration of the copper depends to some extent on the volume of alkaline copper reagent solution to be heated with a glucose solu- tion.Several concentrations of EDTA solution were tried, ranging from 0.1 M (the solution being added from an Agla micrometer-syringe pipette) to 0.0025 M . The volumes of alkaline copper reagent solution were vaned between 0.2 and 2-0 ml and titrations were carried out both with and without dilution of the mixtures. The best results were obtained when 1 ml of alkaline copper reagent solution was diluted with 1 ml of distilled water and the mixture was titrated with 0.005 M EDTA solution, 2 drops of a freshly prepared saturated aqueous solution of murexide being used as indicator.CHOICE OF INDICATOR- Of the indicators tried, murexide is undoubtedly the best for the titration of the copper in the reagent solution. Belcher, Close and Westlo have found that Fast Sulphon black F gives good results with bivalent copper, but I have found that, in the presence of tartrate (an auxiliary complexing agent), this indicator is particularly sluggish near the end-point. Pyrocatechol violet has also been tried, but the pH has to be adjusted to about 6 to obtain a good colour change a t the end-point. The colour change with murexide (yellow to violet) is most striking and occurs a t about pH 10, which is the pH of the alkaline copper reagent - tungstic acid mixture. RELATIONSHIP BETWEEN EDTA TITRE AND GLUCOSE PRESENT IN AQUEOUS SOLUTIONS- Into each of a series of 5-inch x #-inch boiling-tubes were measured 1 ml of alkaline copper reagent solution and 1 ml of a freshly prepared aqueous solution of glucose.The con- centration of the glucose solution was varied between 0.5 and 50 mg per 100 ml. A mixture of 1 ml each of distilled water and alkaline copper reagent solution was used as a blank. The tubes were heated in a boiling-water bath for 5 minutes and then cooled in cold water. After the addition of 2 drops of murexide solution, each mixture was titrated with 0.005 M EDTA solution. At the end-point, there is a marked colour change from yellow or yellowish green to violet. In presence of large amounts of cuprous oxide, the final colour is reddish violet. Two curves have been used for con- venience of presentation and it can be seen that the relationship between EDTA titre and glucose concentration over the range 0 to 50mg per 100ml is practically linear.This would give a range of 0 to 500 mg of glucose per 100 ml for a sample that had been diluted (1 + 9). Identical results were obtained with glucose solutions prepared in saturated aqueous solutions of benzoic acid instead of water. EFFECT OF VARIATION IN HEATING TIME- The previous experiment was repeated with mixtures containing 10, 20, 30 and 50 mg of glucose per 100ml. Three sets of such mixtures were taken and the times of heating in the boiling-water bath were 5, 10 and 20 minutes, respectively, for each set. The results are shown in Fig. 2, from which it can be seen that there is a slight divergence from linearity Fig.1 shows the results of this experiment.0 X - h ._ 7 i-1 i-1 L 30- - w 20- 1 2 a ; w 4- ._ Y Concentration of glucose. mg per 100 ml Fig. 1. Relationship between EDTA titre and concentration of glucose in aqueous solutions APPLICATIONS OF THE METHOD WHOLE BLOOD- It was originally intended to use the isotonic sodium sulphate - copper sulphate solution of King, Pillai and Beall13 to receive the blood, and to precipitate proteins by adding sodium tungstate, as in IGng’s14 procedure. This wou:ld permit the mixture to be kept, without loss of glucose, for 1 or 2 days before the determination is carried out. Also, as the cells would be carried down intact with the protein precipitate, “true” glucose values would be obtained. However, it was subsequently decided that use of a solution containing copper in a method depending on the titration of an excess of copper was undesirable.Accordingly, other protein-precipitating agents were investigated, e.nd that of Folin12 was finally selected, In this procedure, 1 volume of blood is added to 8 volumes of an isotonic solution containing 15 g of anhydrous sodium sulphate and 6 g of :,odium tungstate (Na2W0,.2H20) per litre. After the mixture has been set aside for 5 minute:;, 1 volume of 0.33 N sulphuric acid is added slowly with gentle mixing. In this mixture, the cells are carried down with the protein precipitate so that “true” glucose values are obtained. According to Varley,ls no loss of glucose occurs after the proteins have been precipitated, so that mixtures of Folin’s solution and blood could, for example, be dispatched by post and the glucose determined without loss the following day.CEREBROSPINAL FLUID- Cerebrospinal fluid is treated in exactly the same way as blood. / /./ 10-Nov., 19581 WITH ETHYLENEDIAMINETETRA-ACETIC ACID 631 URINE- If much phosphate is present during titration with EDTA solution, the end-point is drawn-out and indefinite when murexide is used as indicator. This phenomenon is observed when equal volumes of urine (from a healthy subject) and alkaline copper reagent solution are titrated with EDTA solution in presence of murexide. Dilution of the urine to one-tenth of its original concentration does not completely prevent this effect, but a t dilutions of (1 -+ 24) or more the end-point is the same as for mixtures of the reagent solution and water, Fortunately, however, determinations of glucose in urine are generally of little clinical importance when the glucose level is below about 0.1 per cent.; this allows dilutions of (1 + 49), (1 + 99) or more to be made.Concentration of glucose, mg per 100 ml Fig. 2. Relationship between EDTA titre and concentration of glucose for different times of heating with alkaline copper reagent: A, 6 minutes; X, 10 minutes; 0, 20 minutes METHOD REAGENTS- The water used must be free from polyvalent metals, and, if there is doubt as to its purity, de-ionised water should be used instead. Sodium sulphate - sodium tungstate reagent solution-Dissolve 15 g of anhydrous sodium sulphate and 6 g of sodium tungstate, Na,W04.2H,0, in distilled water, and dilute the solution to 1 litre.(This is the isotonic solution recommended by Folin.12) Sulphuric acid, 0.33 N-Pour 10 ml of sulphuric acid, sp.gr. 1.84, carefully into about 900 ml of distilled water. Cool the solution to room temperature, and dilute to 1 litre with distilled water. Standardise against analytical-reagent grade anhydrous sodium carbonate (use methyl orange as indicator) and then dilute the solution until it is 0.33 N . Check the normality of the acid after dilution. (Details of the titration are given by King and Woottod6.) Alkaline copper reagent solution-Dissolve 25 g each of anhydrous sodium carbonate and potassium sodium tartrate in about 600 ml of distilled water containing 40 ml of 1.0 N sodium hydroxide.Dissolve 6 g of cupric sulphate, CuS04.5H,0, in about 100 ml of distilled water, pour this solution into the alkaline tartrate solution, and dilute the mixture to 1 litre with distilled water. EDTA stock solution, 0.1 M-Dissolve about 38 g of disodium ethylenediaminetetra- acetate dihydrate in 1 litre of distilled water. Standardise this approximately 0.1 M solution against analytical-reagent grade calcium carbonate by the procedure of Patton and Reeder.17 EDTA working solution, 0.005 M-Place 25 ml of 0.1 M EDTA solution in a 500-ml calibrated flask and dilute to the mark with distilled water. Murexide indicator solution-Shake about 0-5 g of powdered murexide with distilled water, spin in a centrifuge at slow speed and use the supernatant liquid as indicator.Pour off and discard any unused supernatant liquid each day. Shake the residue with distilled water, and spin in a centrifuge as before to prepare fresh solution. Glucose stock solution, 100 mg per 100 ml-Dissolve 500 mg of analytical-reagent grade anhydrous glucose, which has been dried in a desiccator, in 500 ml of a saturated (0.3 per cent.) aqueous solution of benzoic acid. All solutions should be kept in polythene bottles. This solution contains approximately 1.5mg of copper per mi.632 STREET: DETERMINATION OF GLlJCOSE I N BIOLOGICAL FLUIDS [Vol. 83 Standard glucose solution, 10 mg per 100 w.1-Dilute 10 ml of glucose stock solution to 100 ml with a saturated aqueous solution of benzoic acid. PROCEDURE- For urine, dilute the sample with distilled water (1 + 4) and (1 $- 9), or, if the glucose concentration is greater than 5 g per 100m1, dilute the sample (1 + 99).Add 0.2 ml of whole blood, cerebrospinal fluid or diluted urine to 1.6 ml of sodium sulphate - sodium tungstate reagent solution in a conical centrifuge-tube. Set the mixture aside for 5 minutes and then add 0.2 ml of 0.33 hr sulphuric acid. Mix the solutions gently but thoroughly, and spin in a centrifuge. Add 1 ml of the supernatant liquid to 1 ml of alkaline copper reagent solution in a 5-inch x 8-inch boiling-tube. Prepare a blank solution by mixing 1-ml portions of distilled water and alkaline copper reagent solution and also a standard solution (equivalent to 100 mg of glucose per 100 ml of blood, cerebrospinal fluid or diluted urine) by mixing 1-ml portions of standard glucose and alkaline copper reagent solutions.Thoroughly mix the contents of each tube, plug the tubes lightly with cotton-wool, place in a boiling-water bath for exactly 5 minutes and then cool in cold water. Add 2 drops of murexide indicator solution to each tube, and titrate with 0.005 M EDTA solution until the colour changes to a definite violet. Light from a 60-watt lamp placed behind the tube greatly facilitates end-point detection. The procedure is suitable for blood, cerebrospinal fluid and urine. Calculate the amount of glucose in the sample by using the following expression- x 100 Blank titre - sample titre Blank titre - standard titre Amount of glucose present, mg per 100 ml = (For urine, the result must be multiplied by a further factor of 5 or 10, depending on the initial dilution.) Note that- (i) Just before the end-point is reached, the yellow colour changes to pale pink, which sharply changes to purple a t the end-point. In the absence of cuprous oxide, the purple colour is stable for many hours, but, when cuprous oxide is present, the colour disappears at a rate dependent to some extent on the amount of oxide present.(The disappearance of colour is caused by gradual oxidation of cuprous oxide to the cupric state, which changes the indicator to its yellow form.) However, even when relatively large amounts of cuprous oxide are present, the disappearance of the purple colour is not sufficiently rapid to affect end-point detection. (ii) If the concentration of glucose in the sample is greater than about 600 mg per 100 ml, the alkaline copper reagent solution will be completely reduced.I n these circum- stances, use a 0.2-ml portion of the supernatant liquid after centrifugation and dilute with 0.8 ml of distilled water before mixing with 1 ml of alkaline copper reagent solution. (This applies only to blood and cerebrospinal fluid; urine has already been diluted.) The result must be multiplied by 5. RESULTS RECOVERY OF GLUCOSE ADDED TO BLOOD- A solution containing 20 mg of glucose per 100ml was made up in Folin’s isotonic solution, and a series of mixtures with the compositions shown in Table IA was prepared. The mixtures were spun in a centrifuge, and a 1-ml portion of each clear protein-free supernatant liquid was heated with alkaline copper reagent solution.When cool, the solutions were titrated with 0.005 M EDTA solution, a freshly prepared saturated aqueous solution of murexide being used as indicator. The results are shown in Table IB, from which it can be seen that amounts of added glucose from 20 to 160pg can be recovered from whole blood with an error of &5 per cent. RECOVERY OF GLUCOSE ADDED TO URINE- A solution containing 50 mg of glucose per 100 ml was made up in Folin’s isotonic solution, and a series of mixtures with the cornpositions shown in Table IIA was prepared, the urine first being diluted (1 + 4). The mixtures were spun in a centrifuge (whether theNov., 19581 WITH ETHYLESEDIAMINETETRA-ACETIC ACID 633 urine contained protein or not), and a 1-ml portion of each protein-free supernatant liquid was heated in a boiling-water bath for 5 minutes with 1 ml of alkaline copper reagent solution.When cool, the solutions were titrated with 0.005 M EDTA solution, murexide being used as indicator. The results are shown in Table IB, from which it can be seen that amounts of added glucose up to 400 pg can be recovered from urine with an error of less than i-2 per cent. TABLE IA MIXTURES FOR RECOVERY OF ADDED GLUCOSE FROM BLOOD Each mixture contained 0.4 ml of blood and 0.4 ml of 0.33 N sulphuric acid Amount of Amount of Amount of Folin’s isotonic solution A* glucose added, Mixture solution present, present, mg per 100 ml S O . ml ml of blood 1 3.2 0.0 0 2 2.8 0.4 20 3 2.4 0.8 40 4 2.0 1.2 60 5 1.2 2.0 100 6 0.0 3.2 160 * A solution containing 20 mg of glucose per 100 ml of Folin’s isotonic solution.TABLE IB RECOVERY OF ADDED GLUCOSE FROM BLOOD Amount of glucose Amount of Amount of found corrected for Mixture glucose added, glucose found, original blood-glucose, Recovery No. Pg Plg Pg % - 1 0 56 0 2 20 i 5 19 95 3 40 95 39 97.5 4 60 119 63 105 5 100 159 103 103 6 160 224 168 105 TABLE IIA MIXTURES FOR RECOVERY OF ADDED GLUCOSE FROM URINE Each mixture contained 1.0 ml of diluted urine (1 + 4) and 1.0 ml of 0.33 N sulphuric acid Amount of Amount of Amount of Folin’s isotonic solution B* glucose added, Mixture solution present, present, pg per ml of No. ml ml mixture 1 8.0 0-0 0 2 7.0 1.0 50 3 6.0 2.0 100 4 5.0 3.0 150 5 3.0 5.0 250 6 0.0 8.0 400 * A solution containing 50 mg of glucose per 100 ml of Folin’s isotonic solution.T.4BLE I I B RECOVERY OF ADDED GLUCOSE FROM URIXE Amount of glucose -4mount of Amount of found corrected for Mixture glucose added, glucose found, original urine-glucose, Recover! No. llg t% Pg % 1 2 3 4 A 6 0 50 100 150 2.50 400 4 53 104 155 252 406 0 49 100 151 248 402 - 98 100 09.4 99.2 100.5634 CORNISH : THE PRACTICAL APPLICATION OF CHROMATOGRAPHIC THEORY TO [VOl. 83 I thank Dr. J. Davson for his helpful criticism of the manuscript, Dr. F. L. Rose, F.R.S., for his kind permission to use the library of Imperial Chemical Industries Ltd., Pharma- ceuticals Division, and Miss M. Rider, librarian of the I.C.I. biological library, for her invaluable help. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. REFERENCES Barreswil, L. C. A,, J . Pharm. Chim., 1844, 6, 301. Fehling, H., Annalen, 1849, 72, 106. Street, H. V., Clin. Chinz. Acta, 1958, 3, in the press. Schwarzenbach, G., “Complexometric Titrations,” Methuen & Co. Ltd., London, 1957. Welcher, F. J., “The Analytical Uses of Ethylenediaminetetraacetic Acid,” D. Van Nostrand Co. Flaschka, H., Mikrochem. Mikrochim. A d a , 19ij2, 39, 38. Amin, A. M., Chemist Analyst, 1955, 44, 17. Kinnunen, J., and Wennerstrand, B., Ibid., 1955, 44, 33. Barnard, A. J., Broad, W. C., and Flaschka, H., Ibid., 1956,45, 86 and 111; 1957,46, 18,46 and 76. Belcher, R., Close, R. A., and West, T. S., Chem. & Ind., 1957, 1647. Somogyi, M., J . Biol. Chem., 1938, 125, 399. Folin, O., Ibid., 1930, 86, 173. King, E. J., Pillai, S. S., and Beall, D., Lancet, 1941, i, 310. King, E. J., “Micro-analysis in Medical Biochemistry,” Second Edition, J. & A. Churchill Ltd., Varley, H., “Practical Clinical Biochemistry, ” William Heinemann Ltd., London, 1954. King, E. J., and Wootton, I. D. P., “Micro-,analysis in Medical Biochemistry,” Third Edition, Patton, J., and Reeder, W., Anal. Chem., 1956, 28, 1026. Inc., New York and London, 1958. London, 195 1. J, & A. Churchill Ltd., London, 1956. Received May 20th, 1958

ISSN:0003-2654    

DOI:10.1039/AN9588300628

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

634 CORNISH : THE PRACTICAL APPLICATION OF CHROMATOGRAPHIC THEORY TO [VOl. 83 The Practical Application of Chromatographic Theory to Analytical and Preparative Separations by Ion-exchange Elution BY F. W. CORNISH (Analytical Chemistry Group, A .E.R. E., Harwell, nr. Didcot, Berks.) Data are given from which adequate operating conditions can be deduced for the separation of ionic species by ion-exchange elution. ION-EXCHANGE elution is but one of the available chromatographic techniques1 p2s3; it is, however, a very valuable method, since, under the correct conditions, it can result in the complete separation of species and can be used from the sub-microgram region up to at least 1-g amounts. The procedure has been exl.ensively used in analytical and preparative work and is now accepted as a standard techniq~e.~ Some of the more spectacular achieve- ments include the separation of the lanthanides, of zirconium and hafnium and of the actinides; details of conditions suitable for these difficult separations are a~ailable.49~ 8 Frequently, the chemist is faced with the separation of more common elements for which either little or no detail is available or conditions are described that yield products the purity of which is far in excess of requirements.The selection of column operating conditions is then often a matter of guesswork, and usually results in “over-separation,” thus taking up more time than is really necessary. This paper outlines a general approach to the selection of operating conditions for the separation of any pair of eleinents to any desired degree of purity.In the elution, method the species to be separated are absorbed from a small volume of solution to form a narrow band at the top of a. column. They are then moved down the column by the action of a continuous flow of an eluting solution. A typical result showing the concentration of each species in the solution leaving the column as a function of total volume passed through is shown in Fig. 1. (Note that the total ionic concentration of the solution leaving the column is equal to that of the eluting solution; Fig. 1 shows only the concentration history of the solutes undergoing separation.) The cross-contamination between the two peaks depends on (i) the voluine between the peak maxima, and (ii) theNOV., 19581 ANALYTICAL AND PREPARATIVE SEPARATIONS BY IOS-EXCHANGE ELUTION 635 shape and width of each peak.it is possible to consider each separately. As these two factors are almost completely independent, Solute I I I I Concn.=m, e q u i n I I I I I Width of peak, w,,/ I I \ i a t t = 0 . 3 6 8 of /J- maximuin 'Effluent split into two parts I so t h a t A m , = A m * = ? -- m2 mi n n c n . = m 2 Solute 2 equiv I ! \ I * Volume Fig. 1. Idealised elution peaks showing notation: volume v is the residual liquid in the column before the start of elution SEPARATION OF PEAK MAXIMA The volume required to elute a solute is governed by the rate a t which the solute band moves down the column, and this, in turn, depends on the distribution of the solute between the resin and the eluting solution.In general, this distribution is determined by many factors, but, under the best conditions for elution chromatography, the solute occupies less than about 3 per cent. of the resin capacity, and the equilibrium distribution coefficient, K,, should then be independent of the absolute solute concentration.' The distribution coefficient still depends on the nature of the resin and the composition of the eluting solution, and is given by the following expression- (amount of solute on resin ) x (volume of solution) - (grams of resin) x (amount of solute in solution) It has been shown* that the volume of eluting agent, V, required to elute the maximum * * (1) .- K - * ' of a peak is given by- where V has the significance shown in Fig. 1. Increase in the ratio B, to TI, i.e., the ratio of the distribution coefficients, gives rise to greater separation of the species.This ratio, the separation factor, u, is the most important parameter in elution chromatography. The nature of the reagent used in the eluting solution can exert considerable influence on u. From the point of view of the species undergoing separation three situations can arise. (a) If the solutes do not readily interact with reagents in solution (solutes such as alkali-metal ions or simple anions), their separation factor is almost independent of the nature of the reagent. A solution of any simple acid, base or salt can be used as the eluting agent; the mineral acids are popular, as the solutes can be easily recovered. ( b ) If one or both solutes readily enters into soluble complexes, their separation factor may vary from reagent to reagent ; hence, the eluting positions of ferric iron and aluminium vary with the anion in the eluting solution.(c) If the solutes are chemically similar ( e g . , the rare earths, or niobium and tantalum), it is necessary to find special complexing agents, for example, citric acid, which provide a separa- tion factor as far from unity as possible. Many of the investigations into the separation of the lanthanides and actinides have dealt essentially with the evaluation of reagents in this r e ~ p e c t . ~ The resin may also contribute to an improvement of the separation factor, as selective absorption increases with cross-linking.1° If the separation is essentially due to resin affinity V = K , X (mass of resin in the column) .. .. * * (2)636 CORNISH : THE PRACTICAL APPLICATION OF CHROMATOGRAPHIC THEORY TO [Vol. 83 alone, then a highly cross-linked resin should be used (e.g., one containing 10 to 16 per cent. of divinylbenzene). If differential complexing in the solution provides most of the separation effect, a resin of low cross-linkage (e.g., one containing 4 to 8 per cent. of divinylbenzene) is acceptable and may even be preferable, as difiusion rates are thereby increased. The first practical step towards establishing an elution method is to ascertain the distri- bution coefficients of the materials to be separated. In many instances, sufficient information is available from previous w0rk.47~9 l1 to l6 Alternatively, distribution coefficients can be determined experimentally by measuring the distribution of a small amount of a solute (less than 0.1 milli-equivalent per g of resin) between a weighed amount of resin and a known volume of the proposed eluting solution after equilibration in a flask or bottle (see equation 1).A suitable eluting solution, i.e., one that gives a good separation factor and distribution coefficients in the range 2 to 30, can be easily selected after a few equilibrium experiments of this type. PEAK SHAPE AND WIDTH Symmetrical peaks are desirable, as any elongation of leading or trailing edges increases the cross-contamination between neighbouring peaks. The well known symmetrical Gaussian elution curve is obtained only if the distribution coefficient of the solute is independent of concentration.This is achieved if the column loading is of the order of 50 mg per sq. cm or less. The diameter of the column required for separations that involve continuous elution with one eluting agent can be determined from this limit. If the column is overloaded in this respect, the trailing edge of the peak is drawn cut; it is noteworthy that this trailing can be overcome or minimised by gradient elution, that is, by continuously increasing the eluting power of the eluting agent during the column run, although the separation factor is thereby reduced. The width of an eluted peak is related to the number of theoretical plates, N, for that particular elution by the following equation1J7 ,18- 8 (v $- V ) 2 L .. .. .. . . N=---- _ - w:: h where v is the first volume collected, ie., the residual liquid in the column before the start of elution, V is the volume of eluting agent required to elute the maximum of the peak, w is the width of the peak at 0.368 of the peak height (see Fig.l), L is the length of the resin bed and h is the height equivalent to a theoretical plate. The height equivalent to a theoretical plate is that length of column from which the solution leaving the exit end is in equilibrium with the resin a t the entrance end. A decrease in the height equivalent to a theoretical plate leads to improved separation in that each peak is narrower and there is less overlap. SELECTION OF COLIJMN PARAMETERS In practice, the separation factor (K, the ratio of distribution coefficients) is usually known, and the problem is to select conditions of resin size, flow rate and length of column that will yield the required separation. Glueckaufl8 has published curves relating the cross- contamination of two peaks as a function of iche number of theoretical plates for various values of the ratio a 2 to a, (see Fig.2), which is. approximately equal to the ratio (Kd2 + 1) to (Kdl + 1). The cross-contamination, 7, is given by the ratio Am, to m, (see Fig. l), where m, and m, are the equivalents of the species being separated. 7 is calculated on the assumption that the effluent from the column is divided into two portions to give products The percentage purity of each solute is then 100 (1 - 7 ) . of equal purity, i e . , -2 = __ m1 m2 The parameter used as abscissa in Fig.2 is dz'kmA; this parameter is easily derived Am 2mm9 from q (the permissible cross-contamination) and m, and m2. It should be noted that, in theory, the purity of each product increases as the ratio m, to m2 deviates from unity; under these conditions, the division of the effluent should be made not half-way between the peaks, but nearer the peak involving the least solute (see Fig. 1). Hence, from a knowledge of the distribution coefficients and the equivalents of the solutes to be separated, the number of plates required to yield the desired purity can be found from Fig. 2.Nov., 19581 ANALYTICAL AND PREPARATIVE SEPARATIOKS BY ION-EXCHANGE ELUTION 637 The operating conditions necessary to give rise to the required number of plates can be derived from Fig.3. The height equivalent to a theoretical plate is read off for the resin size available and the chosen flow rate; the required length of resin, L, is given by L = N x (height equivalent to a theoretical plate). Three points must be made in a2 a1 - I 7 m,Z) 2 m , m2 Fig. 2. Purity of separated products as a fnnction of 3, number of theor- a1 m etical plates, N, and ratio of equivalents, -' m2 connection with Fig. 3. First, many unknown variables (e.g., exact shape of the resin particles, exact size-distribution between two mesh limits, cross-linking and efficiency of column packing) are involved, so that the height equivalent to a theoretical plate obtained from these curves may differ from practical results by a factor of up to about 3 (see Table I). The curves638 CORNISH: THE PRBCTICAL APPLICATION OF CHROMATOGRAPHIC THEORY TO [VOl.83 serve essentially as a guide; the actual column must be slightly over-designed to ensure adequate separation, or, alternatively, a trial run must be made. Second, the curves will lead to an adequate combination of variables, but not necessarily to optimum conditions.1 Third, the curves may be considerably in errlor when applied to anion resins used with strongly acid solutions (see Table I). Mesh size 3 Particle radius, rnrn Fig. 3. Empirical relationship between the height equiva- lent to a theoretical plate and particle size for ion-exchange columns a t different rates of flow: curve ,4, 5ml per sq. cm per minute; curve B, 2ml per sq. cm per minute; curve C, 1 ml per sq.cm per minute; curve D, 0.3 ml per s q . cm per minute; curve E, 0.1 ml per sq. cm per minute SEPARATION OF SODIUM AND POTASSIUM An extremely good separation of about 9 mg of potassium and about 3 mg of sodium was required in order to recover the potassium as free from sodium as possible. A purity of 99.9999 per cent. was desirable for the potassium. It was known from earlier workl9320121J2 that hydrochloric acid was a satisfactory eluting agent ; the resin selected was a 12 per cent. cross-linked specimen of Dowex 50 (200 to 400-mesh). Distribution experiments showed that, with 0.6 M hydrochloric acid, the respective distribution coefficients of sodium and potassium were, approximately, 15 and 47. The separation was performed in 1956 with a column 20 cm long and 9 mm in diameter, which was thought to be adequate.The elution curve is shown in Fig. 4. Concentrations were measured on approximately 0.5-ml samples by radiochemical methods. (Analysis of samples in volumes greater than about 5 ml would make the peaks look wider than they really are.) The peaks are almost sym- metrical; the width of the sodium peak is equivalent to about 580 plates, which gives a height equivalent to a theoretical plate approximately 0.35 mm, and that of the potassium peak to about 710 plates, which gives a height equivalent to a theoretical plate approximatelyTABLE I EXAMPLES OF EXPERIMENTAL HEIGHTS EQUIVALENT TO A THEORETICAL PLATE Z 0 c Y i9 equivalent equivalent u 00 Height Height to a to a Flow rate, theoretical theoretical ml per Distribu- plate plate Column Column sq.cm per Particle Mesh tion (by experi- (from * 1: 4 Reference 2 resin cm mm mm ( K d ) mm mm 0 0-11 Betts, R. H., et aLZ0 F 0.7 M * F Eluting agent Ion-exchange length, diameter, minute radius, size Elements coefficient, ment), Fig. 3). Hydrochloric acid, Dowex 50 102 9.0 0.22 0.0125 - Na -10 0.11 Hydrochloric 0.6 M 20 9.0 0-97 - 200 to400 {F acid, Dowex 50 containing 12 per cent. of divinylbenzene Hydrochloric acid, Amberlite 40 10-0 0.5 - 80to 120 {p Hydrochloric acid, Wofatit KS 110 10.0 6.0 0.137 - Na and K Uramildiacetic acid Amberlite Hydrochloric acid, Wofatit KPS 100 10.0 7.0 0.06 - Li Hydrochloric acid, Dowex 50 4 16.0 0.37 - 200 to400 {F 0.1 M IR-100 0.1 M solution 120 6 10.0 3.0 0.067 - {P 0.1 M 200 0.26 M containing 12 per cent.of divinylbenzene Hydrochloric acid, Amberlite IR-1 - 0.1 M 140 10.0 1-4 Ammonium sul- Zirconium - 14 8.0 2.0 Hydrochloric acid, Dowex 50 6-3 27.0 1-3 12-0 0.17 -0.025 Hydrochloric acid, Zeo-Karb 225 21 Citric acid, 0.5 M, Amberlite 24 12.0 0.2 4 . 1 2 5 phate, 0.2 M, phosphate a t 77" C - 1.2 M 2.5 M plus ammonia IR-lOO(H) solution to pH 4.25 * Private communication from R. K. Webster and J. Morgan. Z u +d hl -15 0-35 0.35 Fig. 4 -47 0.28 0.35 18 2-7 0.9 Kayas, G.21 $ 42 1.2 0.9 95 !i a -36 10.0 11.0 Wickbold, R.22 M -10 1.6 -40 1-1 N 1.5 6.5 3.5 Jentzsch, D., et a1.26 w * -70 0.5 0.3 Arons, W. L., et aLZ6 2 v) ::: Buser, w.24 -210 0.4 0-3 0 2 B 4 8 - 6 0.3 0.9 Brooksbank, -26 0.45 1: t:: W. A.s7 100to200 {g -15 0-4 - 6 5.6 3.0 & -22 5-6 3-0 5 x cl 3.0 Burgess, J.S., -15 7.0 60to80 {F 3.5 3.0 et al.a8 X 2.0 Campbell, D. N., 5 M 50 to 100 Mg -14 6.0 et aLZO -12 0.5 0-25 Webster, R. K., ,?! -17 0.4 0-25 et al.* C 0 et a1.t z - Sr - 7 0-65 1.0 Smales, A. A., 2 aa W a t Private communication from A. A. Smales and A. J. Wood.TABLE I-continued 0.3 k P 0 Eluting agent Ion-exchange rcsin Ethylcnediamine- Ambcrlite tetra-acetic acid, IR-IOO(H) 0.1 M . plus am- monium acetate, 0.5 M , to pH 5.35 Citric acid, 5 per Amberlite cent., PLUS am- IR-l monia solution to p1-I 2.64 Citric acid, 0.0125 M, plus am- Dowex A1 monia solution to pH2.1 Hydrochloric acid, Dowex 1 12 M {Dowex 1 Hydrochloric acid, Dowex 2 9 M Hydrochloric acid, De-Acidite FF 9 M , and hydrofluoric acid, 0.1 M 1.0 M . and oxalic acid, 0.01 M 0.5 M , and hydro- fluoric acid, 0.5 M 0.1 M Hydrochloric acid, Dowex 1 Hydrochloric acid, Dowex 1 Sodium oxalate, Dowex 1 Sodium nitrate, Dowex 1 2.0 M Column length, cm 24.0 0-4 14.9 40.0 18.6 11.0 12.0 85.0 27.0 10.5 6-7 0 Height Height equivalent equivalent 0 Flow rate, theoretical theoretical 5 Column sq.cm per Particle Mcsh tion (by experi- (from .. diameter, minute radius, size Elements coefficient, ment), Fig. 3), Reference c) i.2 - 6 0.58 1.0 Smales, A. A., 'd et a1.t z 2 to a to a z x v) ml per Distribu- plate plate mm mm ( K d ) mm mm 12.0 0-2 -0.125 Sr - 4 Y - f 250 to 325 Y -GO 0.7 0.3 <, F % 170to200 Y -GO 1.05 0-45 Ketelle, B. H., 'd 50 to 60 Y -GO 2.0 1-5 et aL30 - 30 to40 Y -60 3.3 3-0 250 to 500 1:: -100 0.7 0-3 -90 0-7 0.3 Huffman.E. H., : et al.31 2 =i 12 0.35 - 3,2 0.3 z - 7.2 0.3 200 to230 Sc - 1 5.0 3.3 Kraus, K. A., 0 8-0 0.3 - 200 to230 V - 4 5.8 3.3 et al.32 crl - 6.5 0.5 -55 6.0 1.1 Huffman, E. H., -170 4.5 1.1 et aL33 2 60 to 100 {F: 3'0 -36 3.8 -80 3.0 3.0 Cabell, M. J., el aZ.= 5 F: 12.0 8.0 - 120 to 200 { p 5 % 2 6.0 0.4 - 80 to200 Nb -10 2.8 0.6 Wacker, R. E., et al.35 cl 1.7 0.3 200 to230 Nb -45 0.85 0.35 Krsus, K. A., c) - i3 plus el al.36 Zr n % 4 2.2 0.35 - 200to230 Sb -19 0.8 0.4 Smith, G. W., et a1.37 c) 20.0 3.4 100to200 I -10 2.0 1-5 DeGciso, R. C., 0 - - et aLa8 t Private communication from A. A. Smales and A. J. Wood. 03 WKov., 19581 ANALYTICAL AND PREPARATIVE SEPARATIONS BY IOX-EXCHANGE ELUTION 641 0.28 mm ; this change in the height equivalent to a theoretical plate with distribution coefficient is expected from theoretical considerations.1923 The separation is clearly more than adequate even for the high purity demanded.10 Volume of eluting agent, ml Fig. 4. Separation of 9 mg of potassium from 3 mg of sodium on a column, 20cm x 9mm, of 200 t o 400-mesh Dowex 50 containing 12 per cent. of divinylbenzene with 0.6M hydrochloric acid as eluting agent a t a rate of flow of 0.97 ml per sq. cm per minute a t room temperature (23°C) ; curve A, sodium; curve B, potassium It is of interest to study the column parameters that would have been obtained from the considerations in this paper. The total mass to be separated is 12 mg, so a reasonable column area is given by 12 mg per 50 mg per sq. cm, i.e., 0.24 sq. cm, thus giving a column diameter of 5.3 mm.For The actual column was slightly over-designed in this respect. a required purity of 99.9999 per cent., q = Hence, T(m12 + m22) is 1.7 x 10-4. From 2m1m2 -~ a 48 a, 16 the equilibrium distribution data, 3 is -, i.e., 3. From Fig. 2, the number of plates required is about 45. With 200 to 400-mesh resin and a flow rate of about 1 ml per sq. cm per minute, the height equivalent to a theoretical plate is about 0.35 mm and the required length of column is 45 x 0.35, i.e., 16 mm. This could be over-designed to give a bed length of about 3 cm. In fact, the experimental height equivalent to a theoretical plate was about 0.35 mm and the experimental bed length was about seven times longer than necessary. DERIVATIOK OF HEIGHT OF A THEORETICAL PLATE been defined by Gl~eckauf,l9~7 according to the following equation- The height equivalent to a theoretical plate, h, as a function of column variables has where Y is the particle radius, K b is the bed distribution coefficient (approximately 0.3 Kd), f i s the fractional column void (about 0.4), F is the linear flow rate in millilitres per unit area per unit time, Ds is the diffusion coefficient of solute in resin phase (of the order of sq.cm per second) and D is the diffusion coefficient in solution phase (of the order of sq. cm per second). The first term allows for finite particle size and for irregular flow and packing in the column: the second term allows for the finite rate of equilibration in the resin phase caused by diffusion, and the third term allows for the finite rate of equilibration across the liquid film surrounding the resin particle.Values of h were calculated from equation (3), the parameters of several published practical elution curves being used. It quickly became evident that practical values of h were larger than those calculated, especially at small particle sizes, presumably owing to such642 NOTES [Vol. 83 factors as irregular packing, irregular particle shapes and channelling. The curves in Fig. 3 were calculated from equation (3) by using the values of 2.8 for K , and 10-6 and sq. cm per second, respectively, for D, and D. The value of the first term was varied to be lor, Sr, 7r, 6r, 5r and 4r at particle radii of 0.01, 0.04, 0.06, 0.10, 0.20 and 0.30 mm, respectively. These values for the first term were selected in order to give a reasonable fit with published experimental data (see Table I).The curves ar,e therefore completely empirical, and, since much of the height equivalent to a theoretical plate arises from the first term, its variation with K,, D, and D is a second order effect. The initiation of this paper was fostered by Mr. A. A. Smales, to whom I express my thanks. I also thank Dr. E. Glueckauf for his stimulating papers and discussions. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. RE FE RE PI[ CES Glueckauf, E., “Ion Exchange and its Applications,” Society of Chemical Industry, London, Martin, A. J. P., and James, A. T., “Fundamentals of Chromatography,” Elsevier Publishing Co., Keulemans, A.I. &I., in Verver, C. G., Editor, “Gas Chromatography,” Reinhold Publishing Samuelson, O., “Ion Exchangers in Analytical Chemistry,” Almquist and Wiksell, Stockholm, Amphlett, C. B., Metallurgical Rev., 1956, 1, 419. Seaborg, G. T., and Ghiorso, A., ScientQic Amcrican, 1956, 195, 66. Tompkins, E. R., and Mayer, S. W., J . Amer. Chem. Soc., 1947, 69, 2859. Mayer, S. W., and Tompkins, E. R., Ibzd., 1947, 69, 2866. Stewart, D. C., International Conference on Peaceful Uses of Atomic Energy, Geneva, 1955, Kitchener, J. A., “Ion Exchange Resins,” -Methuen & Co. Ltd., London, 1957. Kraus, A., and Nelson, F., International Conference on Peaceful Uses of Atomic Energy, Geneva, Finston, H.L., and Miskel, J., Ann. Rev. Nuclear Sci., 1955, 5, 269. Frysinger, G. R., and Thomas, H. C., Ann. Rev. Phys. Chem., 1956, 7, 137. Osborn, G. H., Analyst, 1953, 78, 221. Schindewolf, U., Angew, Chem., 1957, 69, 226. Iiraus, K. A., and Nelson, F., Ann. Rev. Nuclear Sci., 1957, 7, 31. Desty, D. H., Editor, “Vapour Phase Chromatography,” Butterworths Scientific Publications, Glueckauf, E., Trans. Faraday Soc., 1955, 51, 34. Brooksbank, W. A., and Leddicote, G. W., J . Phys. Chem., 1953, 57, 819. Betts, R. H., Harris, W. E., and Stevenson, M . D., Canad. J . Chem., 1956, 34, 65. Kayas, G., J . Chim. Phys., 1950, 47, 408. Wickbold, H., 2. anal. Chem., 1951, 132, 401. Cornish, F. W., Phillips, G., and Thomas, A., Canad. J . Chem., 1956, 34, 1471. Buser, W., Helv. Chinz. Acta, 1951, 34, 1635. Jentzsch, D., and Frotscher, I., 2. anal. Chem., 1955, 144, 1. Arons, W. L., and Solomon, A. K., U.S. Atomic Energy Commission Report AECU-2661, 1952. Brooksbank, W. A., U.S. Atomic Energy Commission Report ORNL-867, 1950. Burgess, J. S., Maynard, J. C., and Peacegood, J. A., Atomic Energy Research Establishment Campbell, D. N., and Kenner, C. T., Anal. Chem., 1954, 26, 560. Ketelle, B. H., and Boyd, G. E., J . Amer. Chenz. Soc., 1947, 69, 2800. Huffman, E. H., and Oswalt, R. L., Ibid., 19580, 72, 3323. Kraus, K. A., Nelson, F., and Smith, G. W., J . Phys. Chem., 1954, 58, 11. Huffman, E. H., Iddings, G. M., and Lilley, R. C., J . Amer. Chem. Soc., 1951, 73, 4474. Cabell, M. J., and Milner, I., Anal. Chim. Acta, 1955, 13, 258. Wacker, R. E., and Baldwin, W. I%, U.S. Atomic Energy Commission Report ORNL-637, 1950. Kraus, K. A., and Moore, G. E., J . Amer. Chem. SOC., 1951, 73, 9. Smith, G. W., and Reynolds, S. A., Anal. Chim. Acta, 1955, 12, 151. DeGeiso, R. C., Rieman, W., 111, and Lindenbaum, S., Anal. Chem., 1954, 26, 1840. 1954, p. 34. Amsterdam, 1957. Corporation, New York, 1957. John Wiley & Sons Inc., New Yorls, and Chapman & Hall Ltd., London, 1953. Paper No. 729. 1955, Paper No. 837. London, 1957, pp. xiii and 31. Report RCC/R 80, 1957. Received April 2nd, 1958

ISSN:0003-2654    

DOI:10.1039/AN9588300634

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

642 NOTES [Vol. 83 Note's THE DETERMINATION OF NIOBIUM AND TANTALUM I N TITANIUM DIOXIDE PIGMENTS THE suitability of titanium dioxide pigment for enamel production is known to be considerably influenced by certain minor constituents,1 and, in the absence of any additions to the pigment after it has been processed, niobium often causes a usually undesirable blue tinge in rutile-opacified white enamel.2Nov., 19581 NOTES 643 As might be expected from the close chemical similarity between niobium, titanium and tantalum (and as found by applying the procedures described), the sulphuric acid process does not alter the relative proportions of these elements, so that this effect varies in strength according to the geographical origin of the ore from which the pigment is produced.To predict the effect, suitable determination procedures for both niobium and tantalum were sought. With ilmenite, which is essentially ferrous titanate with small amounts of manganese, phosphorus, chromium, vanadium and tungsten, a preliminary concentration of earth oxides was achieved by potassium pyrosulphate fusion and then hot aqueous leaching of the melt, any dissolved titanium, niobium or tantalum being recovered by precipitation with ammonium hydroxide after addition of am- monium chloride and ethylenediaminetetra-acetic acid to the solution (alternatively, the modified ethylenediaminetetra-acetic acid - tannic acid procedure of Sankar Das, Venkateswarlu and Athavale3 can be used). The resulting concentrate was then treated in an identical fashion to titanium dioxide pigment.The final separation of niobium and tantalum from titanium and residual phosphorus and tungsten could not be achieved by the classical tannic acid precipitation, which is unsatisfactory a t low concentrations.4 The method eventually chosen consisted in separation on a cellulose column, as suggested by Mercer and Wells,6 and then determination either by paper-strip chromato- graphy,6 for the lower values encountered, or absorptiometrically for higher values. EXPERIMENTAL 3'0 information was available as to the values likely to be found, and a preliminary survey of ilmenite ores of various origins was carried out by a paper-strip method8 designed for geo- chemical prospecting. This proved to be well adapted to the purpose, although often insufficiently sensitive.It indicated that the niobium value was generally.much in excess of that of tantalum. For recovery of a pure niobium - tantalum concentrate, the cellulose-column technique,6 in which a 1-g sample of pigment that has been converted to a fluoride solution is eluted with a hydrofluoric acid - ethyl methyl ketone - water solvent, was found to be satisfactory; spectroscopic examination of this concentrate showed that contaminants were of too small a concentration to interfere in any absorptiometric finish. TABLE 1 NIOBIUM AND TANTALUM CONTENTS OF SOME TITANIUM ORES Niobium and tantalum were separated from each sample by the method of RIercer and Wells6 Origin of ore Canada . . . . Transvaal . . . . Norway . . . . Sierra Leone . . Finland .. . . Spain.. . . . . Egypt . . . . Australia . . . . Senegal , . . . Brazil.. , . . . Travancore . . . . Malaya . . . . Portugal . . South Africa '(;utile) Australia (rutile) . . Amount of Nb,O, found- p.p.m. % - 25 80 - I50 - 150 - 150 - - 0.02 - 0-06 - 0.15 - 0.15 - 0.21 - 0.15 - 0.185 - 0.21 - 0.28 - 0.11 Amount of Ta,O, found- - <5 <5 <5 <5 < 5 - 0.004 5 - 0.008 - 0.008 - 0.006 - 0.013 - 0.035 0.08 - 0.008 - 0.03 p.p.m. % - - - - - - - Method used for determination + Hunt et al.B Marzyss A pure concentrate having been obtained, it wds decided to try two methods, both applicable to a single solution of the concentrate, for the individual determination of niobium and tantalum. The methods were those of Hunt and Wells,7 in which pyrogallol is used as reagent for both metals, and of Marzys,s who used pyrogallol for tantalum, but a thiocyanate - acetone reagent for niobium.Both methods were used and found to give good agreement, but the latter was adopted for routine use, as it was less subject to interference by titanium, traces of which can be easily picked up by accident in a laboratory where pigment is handled in bulk. As the maximum weight of concentrate recovered from the cellulose column was about 5 mg and its tantalum content was relatively small,644 NOTES [Vol. 83 it was necessary to modify Marzys's procedure by carefully fusing the concentrate with 0.5 g of powdered potassium pyrosulphate and dissolving the melt in 5 ml of hot 15 per cent. tartaric acid solution, which was made up to 10 ml with water in a calibrated flask. For niobium, the suggested sample aliquot and flask size are then suitable, but, for tantalum, the use of a 5-ml sample aliquot and a 25-ml flask is necessary to obtain sufficient sensitivity.The niobium and tantalum content of the pigment was sometimes too low to permit use of this procedure, and recovery of the concentrate after elution required a collector. I t was found that the use of a 2-g sample on a cellulose column of the usual length resulted in a small break- through of titanium, which served this purpose. In these instances, the concentrate was used to prepare a series of dilutions, which were then applied to paper strips, the two elements being determined by exactly the same procedure as was used in the preliminary survey.RESULTS By applying the proposed procedure, results have been found for the niobium and tantalum contents of titanium ores from various commerc.ia1 deposits. Representative results are shown in Table I , together with an indication of the appropriate methods of determination. I thank the Research Director, Laporte Titanium Ltd., for permission to publish this Note. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. LAPORTE TITANIUM LTD. Johnson, G., and Weyl, W. A,, J . Amer. Coram. SOC., 1949, 32, 398. Laporte Titanium Ltd., unpublished data. Sankar Das, M., Venkateswarlu, Ch., and Athavale, V. T., Analyst, 1956, 81, 239. Milner, G. W. C., and Smales, A. A., Ibid., 1954, 79, 315. Mercer, R. A , , and Wells, R. A,, Ibid., 1954, 79, 339. Hunt, E. C., North, A. A., and Wells, R.A , , Ibid., 1955, 80, 172. Hunt, E. C., and Wells, R. A,, Ibid., 1964, 79, 345. Marzys, A. E. O., Ibid., 1954, 79, 327; 1955, 80, 194. RESEARCH LABORATORIES LUTON, BEDS. (PRESENT ADDRESS: ELDORADO MINING AND REFINIWG LTD. METALLURGICAL LABORATORIES TUNNEY'S PASTURE OTTAWA, CANADA) L. G. STONHILL Received April 15tk, 1958 THE USE OF NITROSOPHENOL COMPLEXES IN THE DETECTION AND DETERMINATION OF NITRITES THE classical starch - iodide test for nitrites, although inapplicable in the presence of oxidising agents, has been adapted as a spot test by Wilson.' Details of the use of benzidine as a highly sensitive reagent for nitrites have been published.2 Feig13 has devised a test with 1 : 8-naphthylene- diamine, which is claimed to be sensitive to 0,,1 pg of nitrite and specific except for selenites.The Griess test,4 which has been used with various modifications since 1879, is still probably the most sensitive. The proposed methods were devised after thls investigation of various nitrosoresorcinol - metal complexes, for example, the complexes of nitrosoresorcinol monomethyl ether.5 In the proposed methods, the nitrite is used to nitrosate a phenol, which then forms a coloured complex with a metal. SEMI-MICRO QUALITATIVE METHOD- The semi-micro qualitative method consists in warming approximately 5 ml of solution with a little sodium carbonate and then removing ariy copper or nickel carbonates by centrifugation, Four drops of a solution of 1 g of resorcinol in :LOO ml of N acetic acid are added, the solution is acidified with acetic acid and then 4 drops of a solution of 1 g of ammonium ferrous sulphate in 100 ml of N acetic acid are added.SPOT-TEST TECHNIQUE- EXPERIMENTAL A green colour develops in presence of nitrite. A spot-test method for detecting nitrites was devised, which is carried out as follows. A resorcinol reagent and a ferrous reagent are prepared by dissolving 5-g portions of resorcinol and ammonium ferrous sulphate in separate 5-ml portions of glacial acetic acid and then diluting each solution to 100 ml with water. One drop of a nitrite solution is placedNov., 19581 NOTES 645 on a white tile and 1 drop of resorcinol reagent and then 1 drop of ferrous reagent are added; 10 minutes are allowed for development of the colour. The spot test was carried out on 1 drop of a solution containing 0.1 g of nitrite, as NO2-, per litre, i.e., 1 drop contained approximately 5 pg of nitrite; a satisfactory green colour was obtained.The solution was diluted tenfold and the green colour was just discernible when the spot test was carried out on 1 drop, thereby indicating the detection of approximately 0.5 pg of nitrite. This is not so sensitive as the Griess test, which is alleged to be sensitive to 0.01 pg of nitrite. The reagents for the Griess test, a solution of 1 g of sulphanilic acid in 100 ml of 30 per cent. acetic acid and a solution of 0.03 g of a-naphthylamine in 70 ml of water, boiled, decanted and mixed with 30 ml of glacial acetic acid, were prepared. The same spot-test procedure as for the resorcinol test was used; a good pink colour was given with 0.5 pg of nitrite, and, on tenfold dilution, a pink colour was discernible with 0.05 pg of nitrite.A comparison was therefore made with the Griess test. . - volume of test solution, ml x 106 limit of identification. UP. Dilution limit-The dilution limit was taken to be ; the I V resorcinol test gave a value of 1 x lo5 as against a published value of 1 x lo6 for the Griess test. Although not, therefore, as sensitive as the Griess test, the ease of application of the test warranted further investigation into its use. QUANTITATIVE METHOD- On the addition of resorcinol, acidification and addition of a ferrous salt, green solutions having reproducible optical densities are obtained with very dilute nitrite solutions. Red-brown insoluble complexes are formed when cobalt salts are substituted for iron.The ferrous and the cobalt complexes are soluble in isoamyl alcohol, and can be separated from aqueous media by this solvent. The wavelengths of maximum absorption of the ferrous and the cobalt complexes were determined in 1-cm cells with a Unicam SP500 spectrophotometer, a tungsten-filament lamp being used as the source of illumination. Ten millilitres of a solution containing 0.1 g of nitrite, as NO,-, were diluted to 95 ml with water, and 0.2 g of resorcinol was dissolved in the solution; then 1 ml of glacial acetic acid, followed immediately by 0.2 g of ammonium ferrous sulphate, were added. The solution was made up to 100 ml with water, mixed well and set aside for 20 minutes. With use of a red-sensitive photocell, optical-density readings were taken a t different wavelengths, and a curve relating optical density to wavelength was prepared ; a rather flat-topped curve was produced, which was usable between 685 and 700 mp.A similar procedure was carried out with cobalt sulphate instead of the ferrous salt. After development of the colour, 50 ml of the coloured solution were extracted with 50 ml of isoamyl alcohol, the optical densities of the isoamyl alcohol extract after filtration being measured with use of a blue-sensitive photocell. The sharply peaked curve of optical density against wave- length showed that optical density was at a maximum at 396.2 mp. The pH values of the solutions prepared as described mere determined by means of a Cambridge pH meter with the Morton electrode system; both solutions were found to have a pH of 3.06.The effect of variation in pH was investigated. Solutions containing equal amounts of nitrite were buffered so that, after development of the iron colour, the final pH values were 2.9, 3.10, 3.17, 3.33 and 3.54. These solutions gave optical-density readings of 0.43, 0.45, 0.45, 0.44 and 0.32, respectively. Preparation of calibration curves and determination of nitrite-All solutions were prepared with tap-water, as the method is principally applicable to water testing. A stock solution containing 0.1 g of nitrite, as NO,-, was prepared and portions of 0.5, 1, 1.5, 2, 2.5, 3, 4, 5 and 6 ml were diluted to about 49 ml in 50-ml calibrated flasks. To each were added 0.1 g of resorcinol, 0.5 ml of glacial acetic acid and 0.1 g of ammonium ferrous sulphate, and then the contents of the flasks were mixed and each was made up to the mark with water.Maximum colour intensity occurred in 20 minutes and did not subsequently alter for a t least 24 hours. The optical densities of the solutions were measured a t 690 mp, the instrument being set to zero against a solution of the reagents in tap-water in the absence of nitrite. From the results, a calibration curve of optical density against concentration of nitrite was prepared. The cobalt-complex method was carried out by adding to each portion of the nitrite solution 0.1 g of cobalt sulphate instead of 0.1 g of ammonium ferrous sulphate; 10 ml of each coloured solution were then extracted with 10 ml of isoamyl alcohol and the optical densities were measured a t 396.2 mp.646 XOTES [Vol.83 Similar reproducible calibration curves ar.d results were obtained with either a Spekker absorptiometer or a Unicam SP400 DG spectrophotometer for the ferrous-complex method. With the Spekker absorptiometer, measurements were made against water in I-cm cells with a water setting of 1.0; the source of illumination was i t tungsten-filament lamp and Ilford No. 608 red filters and H503 heat filters were used. The spectrophotometer was operated a t 690mp and measurements were made in 1.75-cm tubes against a blank solution of resorcinol, acetic acid and ferrous iron in water. For the determination of the nitrite content of water, 50 ml of sample were taken, the reagents were added as described, the optical density was measured a t the appropriate wavelength and the nitrite content was read from the calibration curve.INTERFERENCES IN THE QUANTITATIVE METHOD- Although large excesses of anions, notably sulphate, chloride and nitrate, as the sodium salts, and magnesium sulphate and calcium hydroxide produced no variation in optical density, nickel, copper and ferric salts gave brown complexes, which interfered with the determination, particularly by the ferrous-complex method. Nickel salts could be tolerated up to ten times the nitrite concentration, as could ferric salts, but copper salts present in concentrations more than one-half the nitrite concentration caused interference. Copper could be tolerated up to three times the nitrite concentration if 0.1 g of anhydrous sodium carbonate was added to 50 rnl of the solution, the solution was heated nearly to the boiling-point, cooled, filtered through a Whatman No.541 filter-paper and then treated as before, but with 1.5 ml of acetic acid instead of 0.5 ml. As stated earlier, hard water, approximately 20 degrees total hardness, was used in the tests; the applicability of precipitation in soft water was therefore investigated. With distilled water, no visible precipitate was produced by sodium carbonate, it being obvious that the hardness in the tap-water acts as a precipitation carrier. However, it was found that the addition of a small amount of magnesium sulphate co-precipitated the interfering ions when sodium carbonate was used, there being no interference from the magnesium ions.After precipitation of interfering ions, the iron colour should not be left for more than 2 hours before measurement of optical density. SENSITIVITY OF THE QUANTITATIVE METHOD- with resorcinol can be used down to a concentration of approximately 5 x 100 mi of solution. of solution when a Unicam SP500 spectrophotoineter is used at 525 mp. weight, namely, resorcinol monomethyl ether (m.-methoxyphenol) was investigated. It was found that, when a Unicam SP500 s:?ectrophotometer is used at 700 mp, the method g of nitrite per The Griess method is sensitivla to approximately 5 x 10-6 g of nitrite per 100 ml In an attempt to increase the sensitivity of the method, the use of a phenol of higher molecular USE O F RESORCINOL MONOMETHYL ETHER- A solution of 1 g of resorcinol monomethyl ether in 10 ml of glacial acetic acid was prepared, To 50 ml of sample solution was added 1 ml of resorcinol monomethyl ether solution, followed immediately by 0.1 g of ammonium ferrous sulphate, which was dissolved by shaking.A satis- factory green colour developed in less than 30 minutes and could easily be extracted with isoamyl alcohol to concentrate the colour. Zellera 9' has described a n-butyl alcohol extraction method for concentrating the colour in the determination of nitrite in pond water by the Griess test. The green colour produced by using resorcinol monomethyl ether as described had an absorp- tion maximum at 705 to 710 mp, and a concentration down to 2 x 10-6 g of nitrite per 100 ml of solution could be determined in 1-cm cells with a Unicam SP500 spectrophotometer. The instrument was set to zero against a solution of the reagents in tap-water.When 50 ml of the coloured solution were extracted with a 10-ml portion of isoamyl alcohol and the optical density of the extract was measured at the same wavelength, the test was sensitive to 1 x 10-6 g of nitrite per 100 ml of solution. In this test, the instrument was set to zero against the isoamyl alcohol extract of a blank solution of the reagents in t,xp-water, but without nitrite. CONCLUSIONS The methods described appear to offer reasonably sensitive qualitative and quantitative tests for nitrites, which are notable for their ease of application, as compared with the usual methods, in which the nitrite is used for diazotising and subsequent coupling to form a dye.Nov., 1958; NOTES 647 REFERENCES 1 .2. 3. 4. 5. 6. 7. RUGBY COLLEGE O F TECHNOLOGY AND ARTS S. M. PEACH Wilson, C. L., Chem. G. Ind., 1940, 378. Tananaeff, K. A., and Schapowalenko, A. &I., 2. anal. Chem., 1935, 100, 360. Feigl, F., “Spot Tests,” Elsevier Publishing Co. Inc., New York, 1954, Volume I. Griess, I>., Ber., 1879, 12, 427. Pcach, S. >I., Analyst, 1956, 81, 371. Zeller, H. D., Trans. Amer. Fish. Soc., 1952, 82, 281. -, Analyst, 1955, 80, 632. nEPARTMENT O F SCIENCE AND &IATHE>I.\TICS EASTLAKDS, RUGBY Received March 5th, 1958 EXPERIENCES WITH THE PLATE ASSAY OF L-PHENYLALANINE WITH A MUTANT OF Escherichin coli MUTANTS of Eschevichia colil requiring amino acids were utilised for the assay of methionine by Ruebnerl and for phenylalanine assay by Dickinson and Th~mpson,~ dilution assays being used.A plate method has, however, many advantages, and has been used to compare phenylalanine concentrations in extracts derived from Kreb’s ascites tumour cells. Although its use, so far, has been to compare widely different values, it appears to offer promise for assay purposes a t levels between 2 and 10 pg per ml; the curve flattens out between 10 and 20 pg per ml. For this reason, the method and some of the difficulties encountered are described. E. coli 1<12/58-278 was kindly supplied by Dr. B. D. Davis; a valine-resistant mutant was easily obtained and mas maintained on nutrient agar slopes. The medium for the tests consisted of 2.4 g of lactic acid, 5 g of sodium chloride, 1 g of diammonium hydrogen orthophosphate, 1 g of potassium dihydrogen orthophosphate and 0.2 g of magnesium sulphate heptahydrate dissolved in distilled water, the solution being diluted to 1 litre.The pH was adjusted to 7.0, and 100 ml of the medium were heated in an autoclave a t a pressure of 15 lb per sq. inch for 15 minutes. To this stock, 1.5 g of Difco agar were added and the medium was again heated in an autoclave. When required, the stock was melted and 0.2 pg of biotin per 100 ml was added. The inoculum (10 ml per 100 ml of agar) was prepared from a 24-hour culture in Lemco broth, washed three times in saline and made up to a standard turbidity in the lactate medium. (The optimal value is best determined by trial and error; a just visible turbidity of the inoculum produces better zones than a heavy inoculum.) Cups were filled, and the Petri plates were incubated for 48 hours a t 37’ C.Diffusion before incubation did not improve the zones, which were then more difficult to read. Plates could not be kept longer than 2 days after preparation and were normally used on the same day. The following points should be noted- (i) After the medium plus agar has been heated in an autoclave, a precipitate may develop, Steaming was not satisfactory. (ii) When L-phenylalanine was assayed in the presence of 0.2 per cent. of casein hydrolysate, This did not interfere (iii) Phenylalanylglycine and glycylphenylalanine gave zones equivalent to their phenyl- (iv) D-Phenylalanine only gave zones a t much higher concentrations (about 50 pg per ml).(v) Against the same standards, in distilled water, the plate assays of cell extracts always gave lower results than dilution assays with the E. coli mutant in lactate medium, the factor varying between 4 and 10. This discrepancy did not affect the work on cell extracts, since the order of a series of very different extracts was the same (e.g., values of 100 pg and over, as against less than 10 pg) by each method. However, the factors responsible for this discrepancy obviously need to be investigated. The plate assay is less sensitive than the dilution test, which can detect 0.1 to 0.2 pg of L-phenylalanine, but it is much less laborious, especially when a large series of extracts is to be compared.REFERENCES but this does not interfere with the reading of the zones. the zones were smaller than those of standards in distilled water. with the assay of the extracts, but may be troublesome in other work. alanine content. 1. 2. 3. BOOTS PURE DRUG Co. LTD. LOIS DICKINSON Grey, C., and Tatum, E. L., Proc. Nut. Acad. Sci., 1944, 30, 404. Ruebner, B., J . Lab. Clzn. Med., 1956, 47, 140. Dickinson, L., and Thompson, M. J., Bvit. J . Pharmacol., 1957, 12, 66. RESEARCH DEPARTMENT, BIOLOGY DIVISION NOTTINGHAM Received March 3 4 1968648 NOTES [Vol. 83 WITH the medium described by Dr. Dickinson and a newly revived freeze-dried culture of the mutant, indistinct zones were obtained. Additions to the basal medium of salts were made and extra biotin was added; other media's2 were tried, .but no improvement in the clarity of the zones resulted until a casein hydrolysate low in phenylalanine was added.Difficulties were encountered with the inoculum for the plates; a 24-hour culture grown a t 37O C often gave hazy zones, but an inoculum grown for 48 hours a t the same temperature always gave sharp zones. After 6 months of weekly subculturing on the same slant medium,l the 24-hour inoculum could nearly always be relied upon to give sharp zones. Assays carried out on the original medium now gave sharper zones than before, but the sharpest zones were still obtained with basal mledium containing casein hydrolysate. In no instance did refrigeration of the plates before incubation increase the clarity of the zones, All the preliminary work was done on Petri plates, but, later, large 12-inch x 12-inch plates were used for 8 x 8 or 6 x 6 assays with a (2 + 2) d e ~ i g n .~ ? ~ Stock cultures of the mutant were maintained on a medium similar to that of Harrison, Lees and Wood,' and were subcultured weekly. The inoculum was prepared from a 24-hour growth in peptone water,l washed twice with sterile saline and re-suspended in saline to give an opacity equivalent to that of Brown's tube No. 6 ; 10 ml of this suspension were used to inoculate 130 ml of medium a t 48' C. The assay medium contained 0.8 g of lactic acid, 0.33 g each of ammonium phosphate and potassium dihydrogen orthophosphate, 0 4 7 g of magnesium sulphate heptahydrate, 1.2 g of phenylalanine-low casein hydrolysate, 10 pg of biotin, 5 ml of a solution containing 25 g each of dipotassium hydrogen and potassium dihydrogen orthophosphates diluted with distilled water to 250 ml and 5 ml of a solution containing 10 g (of magnesium sulphate heptahydrate, 0.5 g of manganese sulphate tetrahydrate and 0.03 g of ferric chloride (dissolved in 5 drops of concentrated hydrochloric acid) diluted with distilled water to 250 ml.These amounts of substances were diluted to 1 litre with distilled water to give the assay medium finally used. The pH was adjusted to 7.0, 1-5 per cent. of 0x0 agar was added, and the medium was sterilised by steaming for 20 minutes. Plates were prepared in the usual manner, and the required number of holes was cut; when filled, the plates were incubated overnight at 37' C. The storage life of the prepared plates is not more than 2 days.The levels in the (2 + 2) assay were 12.5 and 50.0 pg per ml of L-phenylalanine. The limits of error (P = 0.05) for the 6 x 6, 8 x 8 and Petri-plate ;assays were, respectively, 88.6 to 113.1 per cent., 84.0 to 119.0 per cent. and 82.4 to 121.4 per cent. for phenylalanine and 89.4 to 111.9 per cent., 84.4 to 118.8 per cent. and 81.0 to 123.0 per cent. for casein hydrolysate. Recoveries of L-phenyl- alanine added to casein hydrolysate were 85 to 115 per cent. when the proportion (w/w) of L-phenylalanine to hydrolysate was (1 + l ) , (1 + 4), (1 + 50), (1 + 100) or (1 + 200). Several samples were assayed by both the pla.te method with Escherzchia coli and the tube method with Leuconstoc wesenteraides P605; good agreement was obtained with hydrolysates containing from 0.3 to 6.0 per cent.of L-phenylalanine. The casein samples were hydrolysed by heating under reflux for 24 hours with 5 N hydrochloric acid and were then further treated to give different phenylalanine contents ; commercial casein hydrolysates were assayed without further treatment. It must be emphasised that only casein hydrolysates or phenylalanine have been assayed by this plate method, and its possibilities for othe:r substances have not been investigated. Chromatograms of the hydrolysed caseins run nith n-butyl alcohol - ethanol - water6 gave only one spot, R, 0.52, as a bioautographic result on the basal medium seeded with the mutant. On the same papers, 1000 p g per ml of L-phenylalanine gave a similar value.We are grateful to Beecham Research Labora.tories Ltd., Brockham Park, Betchworth, for the preparation of the casein hydrolysates, and we thank the Directors of Beecham Maclean Ltd. for permission to publish this Note. REFEREXES 1. 2. 3. 4. Harrison, E., Lees, K. A., and Wood, F., Analyst, 1951, 76, 696. Burkholder, P. R., Science, 1951, 114, 459. Lees, K. A., and Tootill, J. P. R., Amdyst, 1955, 80, 95. Simpson, J. S., and Lees, K. A,, Ibid., 1956, 81, 562.Nov., 19581 NOTES 649 6. 6. Horn, M. J., Jones, D. B., and Blum, A. E., United States Department of Agriculture, Miscellaneous Hardy, T. L., Holland, D. O., and Nayler, J. H. C., Anal. Chem., 1955, 27, 971. Publication No. 696, 1950. BEECHAM MACLEAN LTD. PRODUCT RESEARCH LABORATORIES LUCOZADE ANNEXE, GRE.4T WEST ROAD BRENTFORD, MIDDLESEX A.JONES SHELAGH M. Received February BURNS 24th, 1958 DETERMINATION O F SORBITOL IN THE PRESENCE OF CARBOHYDRATE A PROCEDURE for the determination of sorbitol in the presence of sugars was described in a recent article by Adcock.1 Boiling with dilute sodium carbonate was recommended for the destruction of interfering sugars. The acids so produced are removed from the solution by ion-exchange treatment, and sorbitol is determined by periodate oxidation. The procedure is suggested for use with foods and biological materials. Adcock stated that the method removes all carbohydrates from a sorbitol - carbohydrate mixture, and that, contrary to other methods, the procedure does not require knowledge of the carbohydrates present.We have found, in a study preliminary to the use of Adcock’s method on honey, that all carbohydrates are not removed from a solution by this treatment. Reducing sugars are quanti- tatively destroyed, as would be expected, but non-reducing sugars are not destroyed and subse- quently interfere in the periodate determination of sorbitol. The resistance of sucrose, for example, to alkaline treatment is well known. A mixture was made of 350 mg of D-glucose, 350 mg of D-fructose, 6 mg of aa-trehalose, 22 mg of raffinose, 20 mg of melezitose, 52 mg of maltose and 26 mg of sucrose in 1 ml of water. This was heated under reflux for 4 hours with 79 ml of 0.1 N sodium carbonate. The solution was cooled and treated batchwise with Dowex 50*; the pH changed from 10.5 to 2.33. The solution was then passed through a column of Duolite A-4*; the pH of the effluent was 6.4.It was evaporated to dryness and fractionated on a charcoal column,2 the monosaccharide, disac- aaride and higher sugar fractions being collected. These were evaporated and subjected to paper chromatography, together with a portion of the original mixture. No reducing sugars were found, but sucrose, raffinose, trehalose and melezitose were present at concentrations approximating to those of the original solution. The procedure used would have detected the presence of 20 pg of reducing sugar remaining in tbe alkali-treated solution. Mixtures A and B contained 100 mg each of D-glUCOSe, sucrose and sorbitol. Mixture C contained 100 mg each of D-glucose and sorbitol. Mixture B was hydrolysed by heating on a steam-bath for 20 minutes with 5 ml of 0.1 N hydro- chloric acid, and then 25 ml of 0.1 N sodium carbonate were added. Mixtures A and C were each dissolved in 20 ml of 0.1 N sodium carbonate, The three solutions were heated under reflux for 4 hours and then subjected to the above-described ion-exchange treatment. Sorbitol was determined in the solutions by the acid periodate procedure desclhed by Adcock. Solution A appeared to contain twice as much sorbitol (195 mg) as the average of the other two solutions (96 mg), although each in fact contained the same amount. It is obvious that the identity of the carbohydrates present in a solution to be analysed for sorbitol by Adcock’s procedure must be taken into account, certainly when they are non-reducing sugars. Interference by such sugars can be avoided by the inclusion of a suitable hydrolytic step before the alkali treatment. Three carbohydrate mixtures were prepared. REFERENCES 1. 2. Adcock, L. H., Analyst, 1957, 82, 427. Whistler, R. L., and Durso, D. F., J . Amer. Chem. SOC., 1950, 72, 677. EASTERN REGION.4L RESEARCH LABORATORY EASTERN UTILIZATION RESEARCH AKD DEVELOPMENT DIVISION, AGRICULTURAL RESEARCH SERVICE U.S. DEPARTMENT OF AGRICULTURE JONATHAN W. WHITE, JUN. Received May 5th, 1958 * Mention of trade names does not constitute endorsement by the Department over others of a similar PHILADELPHIA 18, PENNSYLVANIA nature not named.

ISSN:0003-2654    

DOI:10.1039/AN9588300642

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

650 BOOK REVIEWS [Vol. 83 Book Reviews METHODS OF BIOCHEMICAL ANALYSIS. Volume VI. Edited by DAVID GLICK. Pp. x + 358. J . M. Webb and H. B. Levy of Bethesda, Maryland, review new developments (i.e., since 1953) in the chemical determination of nucleic acids. The analyses may rest on phosphorus content, sugar content, or purine and pyrimidine content, determined often by spectrophotometry. Methods based on ribose and deoxyribose can dispense with separating ribonucleic acid and deoxyribonucleic acid; the mixed nucleic acids can be hydrolysed and the sugars determined by colour tests. Two methods of separating deox:yribonucleic acid and ribonucleic acid are con- sidered and an interesting use of ultra-violet abs,orption is described, whereby de-polymerisation effected by acid hydrolysis is measured.This article is an interim report on an advancing subject, H. K. Miller of New York deals with the :microbiological assay of nucleic acids and their derivatives: “setting up a microbiological assay is somewhat laborious, as a rule, but once it is in operation large numbers of samples can be run with a minimum of effort.” Miller describes the assay of nucleic acid components by several micro-organisms, e g . , Lactobacillus acidophilus R-26 has an absolute requirement for deoxynucleosides or deoxynucleotides, and an E. coli T(-)mutant requires thymine, thymidine or thymidylic acid, and thus lends itself to the assay of deoxyribonucleic acid. L. Helveticus may be used for the assay of uracil or thymine and L. bveuis for the assay of uracil plus cytosine.A rapid assa.y for thymine is described, in which Stveptococcus faecalis is used, and Neurosfiova is used for the d.etermination of pyrimidine ribonucleosides. W. R. Frisell and C. G. Mackenzie of Denver, Colorado, discuss the determination of formalde- hyde and serine in biological systems. The problem arises mainly out of the biochemistry of “one-carbon” compounds. An improved periodate method for serine is described, and formalde- hyde is measured colorimetrically via the chromotropic acid - formaldehyde interaction. Alterna- tively, free formaldehyde or periodate-liberated formaldehyde may be isolated as the dimedone derivative, which can be weighed. F. Bergman and S. Dikstein of Jerusalem discuss the purification and separation of purine with mercury compounds.Mercury combines with purines by salt formation, e g . , mercuric urate or thyminate, or by complex formation, e.g., caffeine - mercuric chloride. Uric acid may be determined by quantitative precipitation as me:rcuric urate from plasma or urine. Use can be made of an ion-exchange column loaded with Hg2+ to adsorb purines, which can be later be removed by hydrochloric acid. Methods of recognising purines on paper chromatograms are discussed. S. Udenfriend, H. Weissbach and B. B. Brodie of Bethesda, Maryland, contribute an essay on serotonin and related metabolites, enzymes and drugs. Serotonin (5-hydroxytryptamine), the vasoconstrictor substance of serum, can be studied by bio-assay procedures, but these are not specific enough and it has been necessary to work out methods based on ultra-violet absorption, colorimetry, fluorescence and chromatography.This article is of special interest in its description of the use of spectrophotofluorimetry for the analysis of serotonin and other indoles a t below microgram levels. They describe, firstly, methods based on the determination of amino acids by a variety of procedures, and, secondly, methods based on the determination of cc-keto acids by chromatographic, chemical and spectrophotometric methods. The chemical (:gasometric) method, based on a ceric - sulphuric acid decarboxylation, is stoicheiometric for most cc-keto acids. The saccharide component is measured by a colorimetric anthrone method .with or without heterogeneous or homogeneous hydrolysis. Methods are given for the determination of specific glycolipids, e g ., cerebrosides or sulphatides in brain. 0. Mickelsen and R. S. Yamamoto of Bethesda, Maryland, contribute a long article on the determination of thiamine by animal, microbiological and enzyme assays, as well as by chemical and physico-chemical methods. A. Kolin of Los Angeles gives an account of ra:pid electrophoresis in density gradients combined with pH and conductivity gradients or conductivity gradients alone. This is an interesting de- velopment particularly suitable for separating different types of suspended particles (bacteria, viruses, mitochondria, cell nuclei, etc.). The emphasis of this article is on the new technique. S. Gardell of Stockholm discusses in detail the determination of hexosamines.All the articles are valuable and authoritative, although the newer work must be a little tentative. This volume maintains the standard of the series, New York and London: Interscience Publishers Inc. 1958. Price $8.50; 68s. A. J. Aspen and A. Meister of Boston deal with the determination of transaminase. N. S. Radin of Chicago deals with the determination of glycolipide. R. A. MORTONNov., 19581 PUBLICATIOKS RECEIVED 651 CLINICAL BIOCHEMICAL METHODS. By A. L. TARNOKY, RSc. Tech., Ph.D. Pp. x + 239. The methods given in this volume are those in most regular demand in clinical laboratories, and the author claims that those chosen are to be recommended on the grounds that the procedures are accurate and rapid and the apparatus is simple. The methods are dcscribed as for a laboratory working sheet, and the laboratory worker can rapidly follow the procedures as outlined. This method of presentation makes it possible for very junior workers to carry out routine tests under adequate supervision.No attempt a t clinical interpretation of results is given, although normal levels are outlined. Main author references are with the title of the method, but more comprehensive lists of authorities are given a t the end of the book. In clinical biochemistry, many methods chosen are the subject of the users’ personal preference, and it is an exception when most workers choose a particular technique. For this reason, the methods given in this book should not be criticised against the reviewer’s personal prejudices. Most of the methods given are sound, and the scope is wider than is usually necessary in most clinical laboratories. I t is to be recommended as a benchcom- panion rather than for use as a reference book. London: Hilger & Watts Ltd. 1958. Price 50s. The author is a biochemist a t the Royal Berkshire Hospital, Reading. R. F. MILTON

ISSN:0003-2654    

DOI:10.1039/AN9588300650

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

Nov., 19581 PUBLICATIOKS RECEIVED 651 Publications Received THE B.D.H. BOOK OF ORGANIC REAGENTS. Tenth Edition. Pp. x4 -+ 182. Poole, Dorset: ‘The Eritish Drug Houses Ltd., B.D.H. 1-nboratory Chemicals Division. 1958. Price 18s. T A L m T A . I<clitecl by Dr. C. L. \YILSON, D.Sc., F.R.I.C. Volume I, Nos. 1/2, July, 1958. Pp. 196. London, New York, Paris and Tlos Angeles : Pergamon Press 1,td. Subscription A (normal) 120s. ; $17.00 per volume; subscription I3 (for individual subscriber’s personal use) 105s. ; $15.00 per annum. Prepared by The Commission on Codification, Ciphering, and Punched Card Techniques of The International Union of Pure and Applied Chemistry. Tentative Version; subject to revision. Pp. viii + 165. London, New York and Toronto: Longmans, Green & Co.Ltd. 1958. Price 35s. Santiago de Compostela, Spain : Seminario Conciliar. 1). T3’. G. RALLKSTYNE, I3.Sc., and L. E. $2. WALKER, A.R.C.S. Chapman & Hall Ltd. 1958. Price 30s. 13y N. H. JONES, C.B.E. Third Edition. Pp. 72. Sheffield: The United Steel Companies Ltd. Free of charge on request from the Welfare Office, The United Steel Com- panies Limited, 17 T’l’estbourne Roacl, Sheffield, 10. Pp. vi + 391. New York: Reinhold Publishing Corporation ; London : Chapman & Hall Ltd. 1958. Price $10.00; 80s. A surve-y of the scope of flavour and associated reseavch, compiled f r o m papers presented By G. I. BROWN, B . A . , J*ondon, New York and Toronto: Longmans, Green & Co. 1,td. TECHNOLOGY OF CoLuninIuhi (NIOBIUM). Edited by R. W. GONSER and E. M.SHERWOOD. 1958. Papers pilesenled at the Symposium on Colirmhium (Niobium) of The Electrothermics and Metallurgy Dirision of The Electrochemical Society, May 15th and 16th, 1958, Washing- ton, D.C. Pp. viii New York: Reinhold Publishing Corporation; London: Chapman & Hall Ltd. A PROPOSEI) INTERNATIONAL CHEMICAL XOTATION. ($lhlICA ANALITICA CUANTITATIVA. A L~CTIOXARY OF NAMED EFFECTS AND LAWS I N CHEMISTRY, PHYSICS AND MATHEMATICS. By Prof. Dr. FRAXCISCO BERMEJO MARTINEZ. Pp. XVi + 1079. By 1958. Price 600 Ptas. Pp. vi + 205. London: L‘ACTORIES ACTS 1937 &. 1948: A CONCISE SUMMARY APPLICABLE TO IKON AND STEEL WORKS. 1958. FLAVOR RESEARCH AND FOOD ACCEPTANCE. Sponsored by Arthur D. Little, Inc. in a series of symposia given in 1956-57. AN INTRODUCTION TO ELECTRONIC THEORIES OF ORGANICHEMISTRY. 13.S~. 1958. Price 15s. Pp. viii + 120. Price $7.00; 56s. Yp. viii + 209. New York: John W’iley & Sons Inc.; London: Chapman & Hall Ltd. THE ENCYCLOPEDIA OF CHEMISTRY (SUPPLEMENT). Editor-in-Chief GEORGE L. CLARK. + 323. 1958. Price $10.00; 80s.

ISSN:0003-2654    

DOI:10.1039/AN9588300651

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

652 REPORTS OF THE ANALYTICAL METHODS COMMITTEE REPORTS OF THE ANALYTICAL METHODS COMMITTEE OBTAINABLE FROM THE SECRETARY The Iieports of the Analytical Methods Committee listed below may be obtained direct from the Secretary, The Society for Analytical Chemistry, 14 Bclgrave Square, London, S.W.l (not through Trade Agents), a t the price of 1s. 6d. t o members of the Society, and 2s. 6d. to non-members. Remittances must accompany orders and be made payable to “Society for Analytical Chemistry..” Sub-Committee on Dirt in Milk. Essential Oils Sub-committee: Report. Determination of Dirt in hlilk. Report No. 1. Estimation of Cineole in Essential Oils. Report No. 2. Physical Constants (1). Report No. 3. Physical Constants (2). Out of print. Report KO. 4. Intcrim Report on the Determination of Acetylisablc Constitucnts in Ikcntial Oils.Rcport No. 5. Dctermination of Phenols in Essential Oils. Report KO. 6. Detcrmination of Citral in Leimn Oil. Report No. 7. Determination of Solubilities. Report No. 8. Determination of Cineole in Essential Oils. (2) Camphor Oil. (3) Other Oils. Out Rcport S o . 9. Determination of Carvone and hlenthone. Rcport No. 10. Determination of Citroncllal. Out of print. Report No. 11. Determination of Aldehydes other than Citroncllal. Report No. 12. Determination of Ascaridole. Report No. 13. Determination of Esters. Report No. 14. Solubility Test for Ceylon Citronella Oil. (Gratis.) Report No. 15. Determination of Linalol in Esssntial Oils. Report No. 4. Determination of Zinc. Determination of Lead in Foodstuffs: Tentativc Method.(1) Cajuput and Eucalyptus Oils. Out of p r i n f . of prirzt. Oirt of print. (Addcndum to Report No. 13, Gratis.) Metallic Impurities in Foodstuffs Sub-committee : Sub-committee on the Determination of Unsaponifiable Matter in Oils and Fats and of Un- saponified Fat in Soaps: Report KO. 1. Report So. 2. Report No. 3. Report No. 4. Report KO. 6. Rcport No. 6. Dctermination of Unsaponifiable Matter in Oils and Fats. Determination of Unsaponificd ]?at in Soap. Out o j P Y i i t f . Determination of Free Alkali in Soaps. Determination of Free Alkali and Silica in Silicatcd Soaps. Determination of Rosin in Soaps. Determination of Phenols in Soaps. Poisons Sub-committee appointed to investigate Methods of Assay for Various Substances appearing in the Poisons Schedules of the Poisons Regulations, 1935: Report KO.1. Report No. 2. Rcport KO. 3. Report S o . 4. Report No. 5 . Report No. 6. Report on the Determination of Fluorine in Foods. Report on the Microbiological Assay of Riboflavine and Nicotinic Acid. The Determination of Carotene in Green-Leaf iMaterial. The Determination of Carotene in Green-Leaf Material. The Chemical Assay of Aneurine [Thiamine] in Foodstuffs. The Microbiological Determination of Thiamine. The Estimation of Vitamin Bit. Report No. 1. Report No. 2. Examination of Detergent Preparations. Assay of Lobclia (Lobelia I ? z ~ N / o ) . Assay of Gelscmium. Assay of Aconite. Assay of Yohimba. Assay of Jaborandi. Assay of Ephcdra and of Ephedrine in Sasal Sprays. Fluorine in Foods Sub-committee: Sub-committee on Vitamin Estimations : (Addendum to this Report, Gratis.) Part 1. €;resli Grass. Part 2 . Green-Leaf hlaterials other than Grass. (Gratis.) Tragacanth Sub-Committee: Evaluation of Powdered Tragacmth. Evaluation of Flake Tragacanth. Soapless Detergents Sub-committee : Meat Extract Sub-committee: Analysis of Meat Extract. Determination of Gelatin in Meat Extract and Meat Stocks: Interim Report. Pesticides Residues in Foodstuffs Sub-committee : Determination of Small Amounts of Total Organic Chlorine in Solvent Extracts of Vegetable Material.

ISSN:0003-2654    

DOI:10.1039/AN9588300652

出版商:RSC    

年代:1958

数据来源: RSC