ISSN:0003-2654    

DOI:10.1039/AN95176FX033

出版商:RSC    

年代:1951

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95176BX035

出版商:RSC    

年代:1951

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95176FP077

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

THE ANALYST xiOSEPH LUCAS LIMITED require Engineers to fill theJfollowing vacancies in the factory which they propose toopen in Liverpool for the manufacture of Fuel Injectionequipment.METALLURGICAL ASSISTANTS and experience in a t least oneof the following:-1. Production Heat Treatment Checking.2. Physical Testing of Raw Materials.3. Pyrometry and Radiography.1.2. Analysis of Raw Materials.3. Analysis of Plating Solutions.CHEMICAL LABORATORY ASSISTANTS and experience in oneProcess Control (Plating, Blacking, etc.).Selected applicants will be expected to spend a short perioda t the Company’s Laboratories in Birmingham prior totaking up their positions in Liverpool. The Company willpay the expenses in connection with such visits.Application, giving full details of previous experience,should be made in writing to the Personnel Manager, JosephLucas (Gas Turbine Equipment) Limited, Shaftmmr Lane,Hall Green, Birmingham, 28.or more of the following:-UNLOP have two vacancies for assistant chemists atDthe Research Centre, Birmingham.Applicants shouldbe aged 20-26, with a degree hi chemistry. One post isconcerned with the development of new analyticalmethods inthe field of rubber and plastics, the other with microchemicalanalysis. Previous analytical experience, though desirable,is not.essentia1. Salary according to age, qualifications andexperience. Applications in writing, quoting ref. A.S 100,To: Personnel Manager, Dunlop Rubber Co. Ltd., FortDunlop, Erdington, Birmingham.24.ROCHE PRODUCTS LIMITED have an opening for assist-ants in their Analytical Department, age about twenty-five years, of B.Sc. or A.R.I.C. standard, who are trained orwish to do chemical analysis. Write stating qualifications,experience and salary required to the Secretary, RocheProducts Limited, Welwyn Garden City, Herb.T H E Research and Development Lkpt. of the DistillersGI. Ltd. have a vacancy for a chemist to work on bio-chemical analysis connected with antibiotic fermentationresearch. Age under 28. Applicants must possess anHonours Degree in Chemistry or equivalent, and experiencein biochemical analysis is essential. The vacancy affordsgood opportunities for any chemist wishing to enter theantibiotic research field, with good prospects of advancement.Salary will depend on qualifications and experience.Non-contributory pension scheme. Apply: Staff Manager, 21,St. James’s Square, London, S.W. 1.HIS MAJESTY’S COLONIAL SERVICEFEDERATION OF MALAYAVACANCY exists for a RESEARCH OFFICER (ForestAchemist) in the Forestry Department, Federation ofMalaya. The duties of the post cover all aspects of woodchemistry with special reference to the manufacture of paperpulp and fibre-board from mixed tropical hardwoods. Thepost is pensionable after three years’ probation. The salaryscale, including pensionable expatriation allowance, is $500per month to $1,180 per month (L700 per annum to L1,662 perannum, current sterling equivalent a t one Malayan dollar to2/4d.) plus a variable cost of living allowance subject to thefollowing maxima: for a single officer $240 per month (1;336per annum).for a married officer kthout children $430per month \&SO2 per annum); for a mamed office; withchildren, $506 per month (L707 per annum).Government quarters if available are a t a nominal rentand free first-class passages on fir& appointment and onleave are provided for the officer his wife and children under10 years, not exceeding four perlons besides himself. A tourof service of three or four years earns approximately fivemonths’ paid leave in the United Kingdom. Income tax isa t local rates, which are very much lower than those in theUnited Kingdom. Candidates should preferably be under30 and must possess an Honours degree in Chemistryand Associate membership of the Royal Institute ofChemistry.Preference will be given to candidates who havedone post-graduate research in wood-chemistry or who havesome experience in wood-pulp and fibre-board manufactureand some knowledge of forest products research in general.The selected candidate will be required to pass examinationsin Malay Standard I, General Orders and Identification ofTimbers in due course.Write giving brief details of age, qualifications and experi-ence to the Director of Recruitment (Colonial Service)Sanctuary Buildings, Great Smith Street, London, S.W.1:quoting reference No. 27106/45.OLONIAL PRODUCTS ADVISORY BUREAU (PLANTThe Civil Service Commissioners invite applications for twopermanent appointments as Senior Experimental Officer.Candidates must have been born on or before 1st August,1916, and must have obtained the Higher School Certificatewith Mathematics or a science subject as a principal subject,or an equivalent qualification.They must also have hadconsiderable technical experience of the raw material resourcesof the Colonial territories which are of plant or animal origin.Inclusive London salary scales :MEN WOMENRates for posts outside London are somewhat lower.Starting salary will be determined on an assessment of thesuccessful candidates’ qualifications and experience.Further particulars and application forms from CivilService Commission, Scientific Branch, Trinidad House, OldBurlington Street, London, W.l, quoting No.S.4067/51.Completed application forms must be returned by 4thOctober, 1951.CAN, ANIMAL) : SENIOR ExPERImNTAL OFFICER.L780-L1,000 L655-&900XPERIENCED Senior ANALYTICAL CHEMIST withEdegree required for research laboratory in NorthernRhodesia specialising on non-ferrous metals. Excellentopportunity for advancement in chemical phases of research.Basic salary L1,050 per annum plus fluctuating cost of livingallowance amounting now to about ,47 per month. LifeAssurance contributory Pension Scheme. In addition CashBonus equivalent a t present rate to about 50 per cent. onbasic salary. Reply by airmail to P.O. Box 172, Kitwe,Northern Rhodesia.ROCESS CONTROL CHEMISTS required for rapidlyPexpanding Chemical Works in East Anglia.The postsvacant are for men, not necessarily qualified, but of a t leastInter. B.Sc. standard, with experience in Heavy Chemical andAllied Industry. and having a preference for plant control andmanagement. Age preferably 25-35. Excellent prospects.Salary ~450-L600 per annum according to qualifications andexperience. Write (quoting Ad. 616), stating age and fullparticulars of experience to Box No. 3781, THE ANALYST,47, Gresham Street, London, E.C.2.~~HIFT ANALYSTS required for routine process analysis.SSome experience desirable. Five-day week, LankroChemicals Ltd., Eccles, Manchester.OVIBOND TINTOMETER. Type: White Light Cabinet,Lcomplete with all colour slides, etc. Nearly new condition,with full directions for use. Canbe seen a t 19, Manor Way, Purley, Surrey, or ’phone Uplands0887 for particulars.Bargain, €50 or near offer.HEFFER’Sof Cambridgepublish from time t o timecatalogues and lists o f booksin various subjects, andannouncements of individualnew books of special im-portance. Let us add yourname t o our mailing list.W. H E F F E R & S O N SLIMITED3 & 4 PETTY CURY, CAMBRIDG

ISSN:0003-2654    

DOI:10.1039/AN95176BP085

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

SEPTEMBER, I95 I THE ANALYST Vol. 76, No. 906 PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS THE AMERICAN CHEMICAL SOCIETY On Wednesday, September 5th, in New York, the President, Dr. J. R. Nicholls, C.B.E., presented the following address to the American Chemical Society on the occasion of the celebration of the 75th Anniversary of its foundation- FROM THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS TO THE AMERICAN CHEMICAL SOCIETY On the occasion of the Celebration in New York, in September, 1951, of the Seventy- Fifth Anniversary of the founding of the American Chemical Society, the President, Council and Members of the Society of Public Analysts and Other Analytical Chemists send Greetings to the Officers and Members of the American Chemical Society.The Society sincerely welcomes the opportunity thus given to it in its own seventy-eighth year to express the friendship and goodwill of all its members to their American confreres celebrating the seventy-fifth anniversary of their Society, and offers its best wishes for the continued prosperity of the American Chemical Society . (Signed) J. R. NICHOLLS (Presideizt). J. H. HAMENCE (Honorary Treas24rer). K. A. WILLIAMS (Honorary Secretary). Dated this Third day of September Nineteen Hundred and Fifty One. 0 Seal of the Society of Public Analysts and Other Analytical Chemists XIIth INTERNATIONAL CONGRESS OF PURE AND APPLIED CHEMISTRY, SEPTEMBER 10th TO 13th, 1951 THE Council of the Society of Public Analysts and Other Analytical Chemists has appointed Dr, J.R. Nicholls, C.B.E., Mr. R. C. Chirnside and Dr. K. A. Williams as delegates to the XIIth International Congress of Pure and Applied Chemistry. BIOLOGICAL METHODS GROUP THE Summer Meeting of the Group was held at 2 p.m. on Friday, June lst, 1951, at the National Institute for Medical Research, The Ridgeway, Mill Hill, London, N. W.7. The following papers were presented and discussed: “Errors of the Mouse Insulin Assay,’’ by P. A. Young and G. A. Stewart; “Simplification in Statistical Computation,” by D. C. Gilles; “Cup-plate Assays of Biotin and Nicotinic Acid,” by S. Morris and A. Jones. 503504 JOHNSON AND KING: THE DETERMINATION OF [Vol. 76 From 4 p.m. until 6 p.m. the following demonstrations were presented: “Demonstrations of Some of the Less Well-known Methods of Biological Assay in use at the National Institute for Medical Research,” by W. L. M. Perry; “Assay of Digitalis,” by G. F. Somers; “Assay of Vitamin B,, with EugZena gracilis,” by P. Waterhouse; “Microbiological Assay of Crude Extracts Containing Several Growth Factors,” by F. A. Robinson and B. W. Williams; ‘‘Apparatus for Automatic Recording of Blood-clotting Times,” by S. S. Randall.

ISSN:0003-2654    

DOI:10.1039/AN9517600503

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

504 JOHNSON AND KING: THE DETERMINATION OF [Vol. 76 The Determination of the Acidity of Milk BY E. I. JOHNSON AND J. KING (Presented at the meeting of the Society on Wednesday, April 4th, 1951) A review is given of the various mlethods that have been used in deter- mining the acidity of milk. Measurements of the increase of colour with increase of pH during the addition of al.kali when phenolphthalein is used as indicator have shown that the comparative method, using a fixed concen- tration of rosaniline acetate in the milk as a colour standard, is sound in principle and gives concordant results. THE acidity of milk is of fundamental importance to the dairy industry and many attempts have been made to standardise a method for its determination so that observers in different laboratories shall obtain concordant results.The difficulty in so doing arises mainly because milk is an excellent buffer from its normal pH to values well on the alkaline side, so that the usual sudden change in colour observable when. a strong alkali is added to a strong acid in the presence of a suitable indicator does not occur. The gradual increase in colour with addition of alkali raises doubts in the mind of the operator as to whether there is any true end-point. The observed end-point usually depends not only on the acidity of the sample but also on the operator’s acuity of colour perception. This depends on whether the operator has normal or abnormal colour vision, on his state of fatigue and on the method of lighting. To minimise these effects it is customary in many laboratories to match the onset of pink colour to that of a standard amount of rosaniline acetate solution added to a given amount of the same milk.l12 Alternatively, the amount of 0.1 N sodium hydroxide required to alter the pH of, say, 10 ml of the milk to a standard pH, usually 8.3, is recorded as the “acidity,” the measurement of pH being performed elect rometrically and not colorimetrically.The effect of adding increasing amounts of phenolphthalein has also been studied, as it is now well known that, with decreasing amounts of the indicator, the pH to which it is necessary to titrate the milk before a pink colour is given rises from about 8.3 with an indicator con- centration of 0.05 g per 100 ml of milk to 8-9 when the concentration is only 0.005 g per 100ml.2,3 The addition of the higher amount of indicator in alcoholic solution to milk at its normal pH leads to precipitation of some of the phenolphthalein in an extremely finely divided form, and this acts as a “reservoir” that increases the amount of the soluble (coloured) form of indicator in the solution as the pH rises.The dilution of the milk by different normalities of alkali used in the titration displaces the end-point. Sommer and Minos,” in a review of the literature on the subject and from their own experimental work, conclude that tri-calcium phosphate is precipitated on the addition of alkali, dilution decreasing the amount affected and hence the titratable acidity. Dilution also decreases the “salt effect’’ and “protein effect” on the colour of the indicator.Kruisheer5 has described six methods, which make use of various volumes of milk, normalities of alkali and concentrations of phenolphthalein, and has compared the results given by some of them in the determination of the acidity of samples of the same milk. In 1931 the Commission Internationale de 1’Eclairage (C.I.E.) adopted certain resolutions for the measure- ment and specification of colour.6 The Lovibond - Schofield colorirneter is a convenient The measurement of colour has been intensively studied in recent years.Sept., 19511 THE ACIDITY OF MILK 505 instrument for measuring colour in terms of units that are readily convertible to the C.I.E. ~ystem.~ The conversion graph with this instrument is based on Judd’s Uniform Chromaticity System,8 which enables colour differences to be expressed in terms of numbers of “least perceptible differences.” This system has been adopted in the following work, which was carried out to ascertain the theoretical basis for the “comparative” method of determining the acidity of milk.EXPERIMENTAL APPARATUS- The Lovibond - Schofield colorimeter was used throughout, measurements being trans- ferred to the graph of equal chromaticities described by S~hofield.~ One of the blocks of magnesium carbonate was removed and a glass cell, which completely filled the aperture, was fixed in its place. The milk to which the indicator or rosaniline acetate had been added was poured into this cell and its colour measured in the normal manner. Direct comparisons without the aid of the colorimeter were also made between a milk to which the standard rosaniline acetate had been added and the milk at various stages of titration in a divided Petri dish, constructed as follows.Molten paraffin wax was poured into a Petri dish of 6.8 cm internal diameter and 1.2 cm depth until the remaining volume was about 30 ml. Before the wax had set, a thin slip of glass measuring about 6-8 cm by 1.1 cm was held centrally in the dish to divide it into two equal semi-circular cells. The wax was then allowed to solidify and the cells were made water-tight by sealing at the ends of the partition with paraffin wax. Molten paraffin wax was then poured into each com- partment until each volume was about 12 ml. For the electrometric determination of pH the meter used was of normal construction with a glass electrode; it could be read to 0.02 pH.The readings of the meter were checked at a range of pH 8.0 to 9-0 with standard buffer solutions. METHODS- For direct comparisons, 10-ml portions of milk were delivered into both halves of the divided Petri dish. To the contents of one cell was added 1 ml of a 0.5 per cent. solution of phenolphthalein in 50 per cent. v/v alcohol and to the other 1 ml of a freshly-prepared 040024 per cent. solution of rosaniline acetate in 50 per cent. v/v alcohol, slightly acidified with acetic acid. A 0.1 N solution of sodium hydroxide was then added t o the cell containing phenolphthalein, while stirring until the colour matched that of the contents of the rosaniline- acetate cell.With a little manipulation it was possible to bring the surfaces of the liquids to the top of the glass dividing partition by dropping in glass beads or by further addition of paraffin wax so that the two fields were practically touching each other with a sharp dividing line. I t is possible to judge a colour match more accurately in this way than by comparing the contents of separate dishes and also more accurately than in the small fields of the colorimeter. It is therefore possible to discern the first slight change of colour at a point somewhat before that necessary to record “one least perceptible difference” by measure- ment in the colorimeter. For measurement of the colours by the Lovibond - Schofield colorimeter, 50 ml of milk were used, the amount of the phenolphthalein and rosaniline acetate solutions being increased correspondingly.This bulk was necessary to provide sufficient material for a viewing cell of sufficient size and for the simultaneous measurement of pH as the titration proceeded. A 0.1 N solution of sodium hydroxide was added in increments of about 0.4ml at a time and the colour and the pH were measured as rapidly as possible, because the colour fades on standing, especially as the pH approaches 9. It was found to be very difficult to measure the pink colour at a pH below 8.2 with accuracy by means of the instrument, but differences in the Petri cells were observable down to a pH of about 8.0. The milk containing the rosaniline acetate was then poured into the cell and its colour measured in the same way as described for the milk plus phenolphthalein.The readings from the Lovibond - Schofield colorimeter were transferred to a scale of equal chromaticity and are shown in Fig. 1. The corresponding pH of the milk for each colour reading is also shown. The lines joining the central point to “Blue,” “Yellow” and “Red” represent the chromaticities of the blue, yellow and red Lovibond glasses on the equal chromaticity scale. The line on the right represents the locus of saturated colours,506 JOHNSON AND KING: THE DETERMINATION OF [Vol. 76 i.e., of monochromatic wavelengths from about Ei65 mp to 585 mp. The colour of the milk before adding indicator (or rosaniline) is represented by a point on the line joining the central point to yellow, showing that this milk is on the yellow side of white by about two “least perceptible differences.” As the alkali is added to the milk containing phenolphthalein, the milk becomes at first slightly pink and then a deeper red with increasing amounts of blue, the final colour being a red - magenta.Starting from the yellow line it will be seen that an increase of pH at first brings about very little change in chromaticity, so that there is little perceptible colour change. From about: a pH of 8.4, however, there is a rapidly increasing colour change with increasing pH. This rate of colour change has been plotted in Fig. 2 to a scale of least perceptible differences, with the corresponding pH and additions of 0.1 N sodium hydroxide. The colour of the milk containing rosaniline is also shown on the same figure, and it is obvious from an inspection of the curves that it occurs at a point where the rate of change of colour with increasing pH is rapidly increasing.Below this pH, 0 5 10 15 -1 Scale of least Perceptibk Differences 570 mp Rosanillne Colour 1 8.08 8.16 8.3 I 836 848 8.56 8.74 8.84 1 Y CHOW Red 580 mi Fig. 1. The dotted line shows alteration of colour with changing pH. Titration of 50 ml Of milk containing Ei ml of 0-5 per cent w/v phenolphthalein. Locations of colours measured on Schofield’s Uniform chromacity chart. The colour changes slowly a t first with changing pH, from a pale yellow to pale pink, the red colour rapidly increasing with increasing pH. the colour change is so gradual that the operator may be in doubt when to cease adding alkali, but at higher pH values the rate of change of colour is so great that a very small addition makes an appreciable colour change.It. is therefore easy to obtain a colour match at the “rosaniline point.” DISCUSSION OF RESULTS It has been a matter of dispute among laboratories as to whether more consistent results are obtained by comparison with the rosaniline acetate tinted milk or by judging the first onset of pink colour. A number of laboratories recently collaborated in a comparative study of the two methods in connection with the formulation of a standard method to be issued with a British Standard, and Mr. H. B. Hawley undertook a statistical analysis of the results. On the whole there was a decided preference for the comparative method. Barkworth and Evans9 also reached this conclusion, but their statistical analysis referred only to results obtained by one operator working always in the same laboratory. In order that an observer shall be in no doubt as to the “end-point” of a titration, it is essential that the rate of change of A study of the diagrams shows why this should be so.Sept., 19511 THE ACIDITY OF MILK 507 colour shall be reasonably great compared with the rate of change of pH.In titrating strong acids with strong bases the rate of change is very great, but when the rate of change of pH is small and uniform with addition of alkali, as occurs with milk (Fig. 2), the circumstances are quite different. Here the rate of change of colour follows an asymptotic curve and does not reach a point where a change is readily observable until a pH of about 8.3 is reached.This point is approximately that of the rosaniline tinted milk on the colour curve, and this fact explains why replicates between observers and laboratories are more easily obtained by this method. I t is the nearest point to the first perceptible colour change at which a rapid change of colour is observed on the addition of alkali and is probably the best point that could have been chosen for the average observer. The milk recorded in Figs. 1 and 2 is that of a bulk-delivery milk issued in the London area by large dairy companies. Milks from Jersey cows are distinctly yellow and the curve 100 mi 9.0 ml 8.0 ml 7.0 ml PH Fig. 2. Rate of colour change (Fig. 1) indicating colour change with change of pH for these milks will be approximately parallel to that shown, but appreciably to the right.I t is unfortunate that acidity as given by the rosaniline standard is slightly higher than that judged from a direct titration, but it seems certain that greater uniformity among laboratories would be achieved by the rosaniline comparative method. In view of the fact that the “titratable acidity” of milk is an empirical value, having no exact equivalent in terms of a given acid (although usually reported as lactic acid), the use of the comparative rosaniline method. is justified. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Davis, J. G., and Sadek, G. M., Milk Ind., 1942, 22, 33. Barkworth, H., Dairy Ind., 1944, 9, 20. Pizer, N. N., Chem. and Ind., 1936, 55, 708.Sommer, H. H., and Minos, J., J . Dairy Sci., 1931, 14, 136. Kruisheer, C. I., V I I C0ng.p.. intern. indus. agr. Paris, 1948, Quest. II-H. Smith, T., and Guild, J., Trans. Opt. SOC., 1931, 33, 73. Schofield, R. K., J . Sci. Instr., 1939, 16, 74. Judd, D. B., J . Res. Nat. Bur. Stand., 1935, 14, 41. Barkworth, H., and Evans, E. M., Daivy Ind., 1944, 9, 640. GOVERNMENT LABORATORY STRAND, LONDON, W.C.2508 JOHNSON AKD KING: THE DETERMINATION OF [Vol. 76 DISCUSSION THE PRESIDENT remarked that this was an important paper. The acidity of milk had often been regarded as an easy determination to make and i t was usually carried out by the most junior staff of a food laboratory. But anyone who had had experience of the determination knew how relatively uncertain was the end-point and how i t was dependent on the amount of indicator used.While comparative results could be obtained by the prescribed technique of any one laboratory, the results from different laboratories might differ considerably; so much so that one wondered whether there was such a thing as “acidity of milk” that could be measured by titration involving the use of indicators. The different proportions of buffering salts in different samples of milk might mean that titratable acidity had little significance. It was a good thing that the process had been submitted to critical review. MR. KING in reply stated that any method relying on a change of pH was purely empirical and was by no means a measure of lactic acid. If, however, universally uniform results were to be attained, a carefully standardised method was essential.He welcomed any further suggestions on this point. DR. J. G. DAVIS complimented the authors on the way in which they had scientifically investigated the interesting and important phenomenon of the colour change of phenolphthalein in milk. His interest in the problem of the determination of the acidity of milk arose on visiting a creamery many years ago when he had been told that brine-cooling of the milk resulted in a fall in acidity. The basis of this statement was that the acidities measured by the creamery receiving the milk as compared with those measured before despatch were consistently lower to the extent of about 0.02 per cent. of “lactic acid.” Investigation of the technique used in the two laboratories showed that the first laboratory used only three drops of phenolphthalein, which may be described as the classical method in the dairy industry, while the receiving creamery used 1 ml.This had prompted him to study the effect of concentration of indicator and it had been easy to demonstrate that the pH of the combined milk and indicator, a t the point where the first faint pink colour was perceptible, varied appreciab1.y according to the concentration of the indicator used.1 They had therefore recommended the standard use of 1 ml of a 0.5 per cent. solution of phenol- phthalein for every 10 ml of milk as a reasonable compromise. It was true that this did not settle the vexed question of the colour end-point. A variety of methods had been tried, including the use of pink stirring rods, pink tiles and a special Tintometer glass, and the best of these appeared to be the use of the Tintometer glass, perhaps because the glass could be adjusted to match the end-point colour most accurately.Un- fortunately this method was very expensive and for this reason alone it had not been recommended. In practice, experienced dairy laboratory analysts worked to a very early end-point. Thus, in titrating milk of average quality, a distinct greyness 01 characteristic change in colour was perceptible when the titration was a t about 0-12 per cent. of “lactic acid” and the first perceptible pink colour appeared a t about 0.14 per cent. The inexperienced analyst nearly always worked to a later end-point (more intense pink) and so returned a somewhat higher figure for acidity.Obviously from the scientific point of view a colour standard was required, and the rosaniline acetate method, while giving a reasonably good match, required a special solution to be available, which was a further complication in the test; for this reason i t had not been generally adopted. Dr. Davis emphasised that the test really was a coinplete anomaly in that the value usually stated for fresh milk (0.14 per cent.) was not really lactic acid at all, although this expression was always used. The test really measured the buffer value of the milk from its initial pH (about 6.6) to the pH of the visually detected end-point with phenolphthalein (about 8.4). If the test really measured lactic acid it might be worthwhile to bring out a more elaborate and accuIate form of the test, but for farmers’ milks he suggested that the test should be abolished.Mastitis, which affected about one-third of the cows in this country, resulted in the lowering of the acidity of milk to a variable extent, while a high proportion of solids or feeding on certain types of land could result in a high acidity. As a measure of souring or hygienic quality for herd milks the test was, therefore, not reliable. What was wanted was a specific test for the amount of lactic acid formed in the milk, and work was proceeding on these lines. From the point of view of dairy control i t would have to be a very quick test. The industry urgently wanted a still quicker test that would give in half or a quarter of a minute a reasonably accurate measure of the degree of souring.These objections did not, of course, apply to bulk milks, e.g., as from a 3000-gallon tanker, for the bulking levelled out the variations of acidity in individual herd milks. The titratable acidity test, therefore, was a reliable measure of souring for bulk milks provided that a standardised technique could be developed. Whatever test was recommended would have to be simple and of a type that could be used in the ordinary dairy control laboratory. MR. KING, in his reply, agreed with Dr. Davis on the completely empirical nature of the acidity determination by any method in which phenolphthalein was used as an indicator and further agreed that a rapid and specific determination of lactic acid would afford far more information as to the condition of milk.Until phenolphthalein should cease to be used in determining milk acidity i t was most desirable to standardise the procedure on a sound basis. DR. EGAN said that i t might be that some mixed indicators increased the number of least perceptible differences that accompanied the indicator colour change. Kolthoff and Stenger2 have described the use The present official rejection test, the rapid resazurin test, took 10 minutes.Sept., 19511 THE ACIDITY OF MILK 509 of phenolphthalein in conjunction with a dye, methyl green, the colour of which was complementary to magenta, for the titration of milk: unfortunately the pH a t which the best colour change occurs is 9.0. Dr. Davis had described an effect that may be observed in favourable conditions without the addition of a dye or a second indicator; namely, that the yellow colour of the milk being titrated gave way to a chalky neutral tint immediately before the phenolphthalein colour appeared. Had Mr. King any experience of the colour changes or any comments to make on them? Assuming that the average milk had a spectral reflection curve that was preferential in the yellow part of the spectrum, the superimposition of a slight magenta colour-which was complementary to yellow and given by the first change of phenolphthalein-would result in a chalky-white neutral colour, immediately before the pink - magenta given by a further addition of alkali to the milk. He agreed that the range of pH covered by Kolthoff’s suggested mixed indicator was far too great to give agreement with the usually accepted figures for the acidity of milk. REFERENCES TO DISCUSSION Davis, J. G., and Zadek, G. M., Milk Ind., 1942, 22, 33. Kolthoff, I. M., and Stenger, V. A., “Volumetric Analysis,” Volume 11, Interscience Publishers MR. KING replied that he thought there was a sound basis for Dr. Davis’s observations. 1. 2. Inc., New York, 1947, p. 57.

ISSN:0003-2654    

DOI:10.1039/AN9517600504

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

Sept., 19511 THE ACIDITY OF MILK 509 An Improved Volumetric Method for the Determination of Hydrogen Sulphide and Soluble Sulphides BY J. A. KITCHENER, A. LIBERMAN AND D. A. SPRATT (Presented at the meeting of the Society on Wednesday, April 4th, 1951) The usual iodimetric method of determining hydrogen sulphide or sulphur in iron and steel is not satisfactory when accurate results are required. The use of alkaline sodium hypochlorite as combined absorbing and oxidising reagent has been investigated and found to offer many advantages. The reagent is stable to boiling and gives quantitative oxidation of sulphide to sulphate. The recommended procedure has been shown to give results correct to within 0.3 per cent. of the sulphur content. IN the course of researches into the thermodynamical activity of sulphur in liquid iron,l it was necessary to make a large number of determinations of (a) hydrogen sulphide in mixtures of hydrogen sulphide and hydrogen and (b) sulphur in iron.The quantity of sulphur in the samples was geflerally between 1 and 10 mg, and it was required to determine it to within IfIl per cent. A method of analysis was therefore needed that would be (i) reliable, both in reproducibility and absolute accuracy, (ii) sensitive to 0.01 mg and (iii) reasonably rapid. A review of the literature suggested that, of the many and various methods that have been described for the determination of hydrogen sulphide and sulphur in iron, the commonly used volumetric meth0d~9~ would be the most suitable for the purpose. In this method, the hydrogen sulphide in a gas mixture, or resulting from the dissolution of a sample of iron in acid, is absorbed and then titrated by iodimetry.There are many modifications for the absorbing medium and method of titration, several of which are in common use in steelworks. Absorbing solutions include ammoniacal or buffered cadmium chloride or acetate or zinc sulphate, while in the determination of hydrogen sulphide in gases, sodium hydroxide is often used. The resulting sulphide suspension or solution may be acidified and titrated immediately with iodine or iodide - iodate mixture, but the most common procedure is to add the sulphide to an excess of acidified iodine and then titrate the residual iodine. The basic reaction is- The usual evolution method3 is generally considered satisfactory for routine analysis of plain carbon steels, but when it was applied to a number of the hydrogen sulphide - hydrogen and pure iron - sulphur samples arising from the present work the results were not S” + I, -+ s J.+ 21’.510 KITCHENER, LIBERMAN AND SPRATT IMPROVED VOLUMETRIC METHOD FOR [VOl. 76 sufficiently concordant, although several of the usual absorbents were tried. Closer examina- tion of the ordinary volumetric method shows that it is really semi-empirical rather than accurately stoicheiometric. It is reliable only for a carefully standardised set of conditions and any variation of the conditions may lead to unexpected errors from a variety of causes. Examples of errors up to 17 per cent. arising merely from variation in the size of the sample used in the analysis have been quoted by Lundell, Hoffman and Bright.2 In the present work it was often necessary to use very small specimens of iron because of the abnormally large proportion of sulphur in many of them (e.g., about 1 per cent.of sulphur instead of the usual range of 0.01 to 0.1 per cent.). In these circiimstances it would clearly be unsatisfactory to take the ordinary volumetric, method on trust when the research required an absolute accuracy of 1 per cent., as the results might easily be subject to large unknown errors. SOURCES OF ERRORS IN THE ORDl NARY VOLUMETRIC METHOD Consideration was given to the possibility of developing a more reliable volumetric method, and the potential sources of error in the existing methods were first examined. The following errors have been recognised- Incomplete evolution of the sulphur as hydrogen sulphide during dissolution of the iron- This is a well known difficulty with certain cast irons and alloy steels, but since it does not apply to pure iron - sulphur samples or hydrogen sulphide - hydrogen mixtures it need not be discussed further here.Incomplete absorption of the hydrogen sulphide--Rapid dissolution of the sample is usually recommended as this is found empirically to give better results.2 Theoretically it should make no difference; that it is found to do so is another indication of the existence of weakness in the method. If the gas contains much hydrogen sulphide some of it may escape absorption and two absorption bulbs in series must then be Loss of hydrogen sulphide was noticed with some of the high-sulphur samples when tested in conjunction with the usual absorbents.Alkali was found to be the most effective absorbent. Errors in the titration procedures-(a) If cadmium solutions are -used, the cadmium sulphide must be protected from sunlight since it is photo-~ensitive.~ (b) With the cadmium or zinc back-titration method, the end-point is not altogether satisfactory, the colour changing from a reddish-brown to a deep blue colour. Attempts to increase the sensitivity of the titration by working with more dilute solutions gave dis- couraging results. (c) If the sulphide solution is first acidified and then titrated directly with iodine, some hydrogen sulphide may be lost before the titration is completed.(d) If the back-titration procedure is used there is a possibility of occlusion of iodine by the colloidal sulphur that is formed.6 (e) If sodium hydroxide is used as the absorbing solution, atmospheric oxygen is apt to cause rapid oxidation of the alkali sulphide in solution. (f) If alkali absorbent is mixed with excess of acidified iodine there is a possibility of volatilisation of iodine owing to the rise of temperat~re.~ To avoid errors (d) and (f) the solution is usually diluted considerably,8 but, with alkali sulphides, this emphasises the danger of (e). The above list gives the chief potential sources of error in the volumetric methods; certain others have been noted by Etheridge.g Although they do not necessarily vitiate the determinations, they indicate the importance of very careful control of conditions.Shawlo and Treadwell and Hall11 have pointed out that differences of technique, dilution of the solution, rate of addition, and amount of potassium iodide present may produce large variations in the results. There is clearly a need for a better method, giving greater reliability and sensitivity. In developing such a method the following points deserve attention- (i) The reactions employed in the titration should be accurately stoicheiometric (no side-reactions) . (ii) Absorption should preferably be by alkali hydroxide, as this minimises the danger of loss of hydrogen sulphide. (iii) If alkali is to be used, there must be complete exclusion of air from the apparatus and the solutions. Further, transference or dilution of the sodium hydroxide - sodiumSept., 19511 51 1 sulphide solution or addition of large volumes of reagent solutions containing dissolved air should be avoided.(iv) The neutralisation of large amounts of alkali should be avoided to minimise loss of iodine. (v) If an oxidising agent could be found that would oxidise sulphide to sulphate instead of to colloidal sulphur, it would improve the sensitivity fourfold, avoid occlusion of iodine, and improve the end-point. (vi) For accurate work an all-glass apparatus should be used, so avoiding rubber, which absorbs hydrogen sulphide. DETERMINATION OF HYDROGEK SULPHIDE AND SOLUBLE SULPHIDES USE OF ALK.4LINE HYPOCHLORITE FOR ABSORPTION AND OXIDATION All the above desiderata were satisfied by the introduction of an alkaline solution of Kolthoff and Sandell12 state that sulphides are quantitatively oxidised to sulphates sodium hypochlorite as a combined absorbing and oxidising reagent.by calcium hypochlorite in alkaline media according to the reaction- but they do not specify the conditions necessary to secure complete oxidation. The use of hypobromites and hypochlorites for the determination of sulphides was proposed by Willard and Cake.13 Hypochlorites are the more stable, but it was found that at least 4 N sodium hydroxide was needed to effect oxidation at room temperature, whereas 2.5 N sodium hydroxide was sufficient with sodium hypobromite. Willard and Cake therefore proposed absorbing hydrogen sulphide in 2.5 N sodium hydroxide and then washing this into 0.3 N sodium hypobromite.The method to be described below avoids transferring alkali sulphide solutions (cf. point (iii) above) by using a combined absorbing and oxidising reagent. The more stable hypochlorite is used, at a concentration of 0.1 N instead of 0.3 N to give greater sensitivity. The alkali is 0.4 N instead of 2 N to reduce the heat of neutralisation (cf. point (iv) above). The use of dilute calcium hypochlorite solutions for the approximate deter- mination of microgram quantities of sulphides has been described recently by P e p k 0 ~ i t z . l ~ Solutions of pure hypochlorites are remarkably stable if kept free from organic matter (dust). They can be boiled without decomposition.16 They are more stable still in the presence of free alkali,16J7 and such solutions can be kept for many months without appreciable change of titre, provided they are stored in dark bottles to avoid photochemical decomposition by sunlight.When attempts were made to absorb hydrogen sulphide directly into alkaline hypo- chlorite it was found that some colloidal sulphur was always formed, even with fairly con- centrated sodium hydroxide and a considerable excess of sodium hypochlorite. Experiments were therefore made to find a means of securing quantitative oxidation to sulphate. It was not considered desirable to use very concentrated alkali for reasons already given. The use of higher temperatures was therefore considered. Willard and Cake reported more nearly complete oxidation with hypochlorite at 45” C. I n the present work it was found that any colloidal sulphur formed during absorption at room temperature could be removed by subsequently heating the solution to about 70” C.How- ever, warming just sufficiently to clear the solution of sulphur did not lead to concordant results. There must certainly be intermediates between sulphide and sulphate, and apparently some stages of the reactions are sluggish, This procedure led to concordant results and stoicheiometric relations corresponding to formation of sulphate. As, so far as is known, the boiling of alkaline hypochlorite solutions is a novel step in volumetric analysis the following evidence is presented to demonstrate its reliability. The validity of the procedure is also supported by the accuracy of the results of the new method as recorded below.Solutions approximately 0.4 N with respect to sodium hydroxide and 0-1 N with respect to sodium hypochlorite were made by passing chlorine from a cylinder into the alkali until the required titre was reached. Provided there is a considerable excess of alkali and the solutions are cool and dilute, the product obtained is almost entirely hypochlorite, no appreciable amount of chlorate being formed.18319 It is then satisfactory to standardise the solution by the iodimetric method (see Kolthoff and Sandell,12 pp. 587 and 640). The s” + 4 OC1’ 3 SO,’’ + 4 Cl’, Pepkowitz recommended 80” to 90” C. Finally, it was found satisfactory to boil the solution.512 KITCHENER, LIBERMAN AND SPRATT: IMPROVED VOLUMETRIC METHOD FOR [VOl. 76 solution is just acidified with hydrochloric acid, excess of potassium iodide is added, and the iodine liberated is titrated with sodium thiosulpha.te, using starch at the end-point. Aliquot 25-ml portions of such solutions were transferred by pipette into conical flasks and boiled briskly for different times.The flasks .were then cooled under the tap, and the contents titrated as described above. Table I shows the results (in millilitres of 0.1 N sodium thiosulphate) of independent tests by two of the authors on three solutions prepared at different times. TABLE I EFFECT OF BOILING ON SOLUTIONS OF SODIUM HYPOCHLORITE Titre of solution A Titre of solution B ml ml ml Titre of solution C Duration of boiling (A.L.), (A.L.), (D.A.S.), 0 (cold) 1 min. 2 min. 5 min. 10 min.25.45 25.25 25-20 25-19 25-20 30.00 29.85 29-80 29.80 29-85 47.05 46.84 46.86 46-85 46.81 It is seen that during the first 2 minutes’ boiling there was a small loss of titre amounting to about 0.7 per cent. in each experiment. Thereafter the titre remained constant within the accuracy of titration (about kO.1 per cent.) during a further 8 minutes’ boiling. The initial loss is believed to be due to reaction with traces of dust. It is quite clear that alkaline hypochlorite alone is perfectly stable to boiling, and that the titre reached after, say, 5 minutes’ boiling is a highly reliable figure.* It is essential, of course, to protect the solutions from bright sunlight during treatment. Experiments in which duplicate samples of sodium sulphide solutions were added to the alkaline hypochlorite and boiled for 5 minutes gave similarly reproducible results.The rate of addition of the sodium sulphide was found to be immaterial, whereas the amount of sulphur initially precipitated and the titre in the cold were very dependent on the mixing conditions. Since sulphate could be detected after the reaction, and since all possible oxidation reactions had evidently finished after 5 minutes’ boiling, it was concluded that the treatment effected complete oxidation of sulphide to sulphate irrespective of the intermediates formed. Since also the starting and final hypochlorite solutions were stable it was evident that the process could form the basis of a reliable and stoicheiometric method of determining hydrogen sulphide or any sulphides from which hydrogen sulphide could be completely evolved.Proof of the absolute accuracy of the hypochlorite titration is not easy, as there is no convenient substance that can be taken as a primary standard for hydrogen sulphide. Even sulphides such as those of zinc and cadmium, which are sometimes used in the gravimetric determinations of those metals, cannot be prepared completely pure and of stoicheiometric composition.22 It has therefore been necessary to resort to indirect checks, which are described later in this paper. APPARATUS- The complete apparatus for the determination of soluble sulphides or sulphur in iron and steel is shown in Fig. 1. The chief parts, preferably made of Pyrex glass, are the evolution flask, A, which has a wide ground-glass joint carrying a small dropping funnel, B, and a spray trap and condenser (conveniently combined, C), and the absorption flask, E.Connection is made by the tube, D, which has ground cones at each end. If desired, the ground joint to flask E may be replaced by a rubber stopper. If the evolution apparatus is likely to be used in bright daylight, the absorption flask must be screened by a box of wood or cardboard. The dimensions of the apparatus are not critical. METHOI) REAGENTS- Sodium hypochlorite, 0.1 N , in sodium hydroxide, 0.4 N-Prepare by passing chlorine from a cylinder into a 0-5 N solution of sodium hydroxide until the required concentration * It is worth recording that “Chloramine T,” which is a very convenient substitute for hypochlorite for titrations in the cold,20J1 does not behave in the same way to boiling.Instead, there is a progressive loss of titre, no doubt due to reaction with the organic part of the compound.Sept., 19511 DETERMIXATION OF Hl7DR0GEK SULPHIDE AKD SOLUBLE SULPHIDES 513 is reached (see above). For 2 litres of solution, this takes between 5 and 7 minutes at a moderate rate of bubbling. Sodium thiosulphate, 0-05 N-Prepare by dissolving 25 g of sodium thiosulphate and 7.6 g of borax (as preservative) in 2 litres of water. It is most conveniently standardised against a solution of potassium iodate (1.7 g of potassium iodate dried at 180" C for 2 hours). Hydrochloric acid, diluted (1 + 1)-Prepare an air-free solution by diluting concentrated hydrochloric acid with its own volume of boiled, cooled water.Store in a Winchester bottle fitted with a siphon and protect the contents from oxidation by means of a bubbler containing alkaline pyrogallol. Potassium iodide-A freshly made approximately 10 per cent. solution. Hydrochloric acid, approximately 2 N . Starch solzttion-A freshly made approximately 0.5 per cent, solution. [L L Fig. 1. Details of apparatus L E f J PROCEDURE- To determine, for example, sulphur in iron, first wash the whole apparatus with water, rinse it with alcohol and dry with warm air. Weigh the sample to k0.5 mg and place it in the evolution flask. Assemble the apparatus as shown in Fig. 1, the ground joints being very lightly greased, Transfer 25 ml of standard alkaline hypochlorite solution by means of a pipette into the absorption flask and add 25 ml of water.The absorption flask must be screened from bright light. Pass a steady stream of hydrogen from a cylinder through the apparatus for 10 minutes to sweep out all air. Place a small piece of lead-acetate paper and a small piece of starch - iodide paper in the exit tube of the absorption flask to test the completeness of absorption. Admit about 20 ml of diluted hydrochloric acid (1 + 1) into the evolution flask through the tap-funnel, keeping the amount of air introduced with it to a minimum. Place a small flame or, better, an electric hot-plate under the evolution flask and maintain a continuous, slow stream of hydrogen through the apparatus.. With high-sulphur irons the rate of dissolution must not be too great.514 KITCHENER, LIBERMAN AND SPRATT: IMPROVED VOLUMETRIC METHOD FOR [VOl.76 When the sample has finally dissolved (as shown for iron by means of a magnet placed near any residual particles of carbon, etc.), boil the solution in the evolution flask, A, for 5 minutes to ensure that the hydrogen sulphide has been expelled. The trap and condenser prevent hydrochloric acid from passing over into the absorption flask, E. Then disconnect the absorption flask and boil its contents steadily for 5 minutes fl minute. The white colloidal sulphur that appears during the absorption is completely oxidised to sulphate and the solution becomes quite clear. Remove the tubes from the absorption flask:, E, and rinse them, collecting the rinsings in the flask. Immediately cool the flask thoroughly, e.g., by covering the neck with a small inverted beaker and standing the flask under a irunning tap.The solutions are then ready for titration. Add 10 ml of 10 per cent. potassium iodide solution and then 20 ml of 2 N hydrochloric acid. Titrate the liberated iodine against 0.05 N sodium thiosulphate, and add 5 ml of starch solution as indicator when the end-point is nearly reached. Determine the thiosulphate equivalent of the hypochlorite reagent in duplicate by boiling 25 ml and titrating in the same way as in a sulpliur determination. The difference between this figure and that given by a sulphur determination is the volume of thiosulphate solution that is equivalent to the sulphur in the sample- 1 ml of 0.05 N sodium thiosulphate = 0.200 mg of sulphur. CONFIRMATION OF THE ABSOLUTE ACCURACY OF THE METHOD In the absence of any satisfactory primary standard for hydrogen sulphide or soluble sulphides, it was necessary to resort to indirect t’ests.The sulphur contents of a number of materials were therefore determined by the hypochlorite method described above and by a standard method, and the results compared. CoMPARISON WITH DIRECT WEIGHING METHOD FOR HYDROGEN SULPHIDE- Sherman, Elvander and Chipman% have recently shown that hydrogen sulphide in the gases evolved from the dissolution of iron in acids can be accurately determined by first washing the gases to remove acid vapours, drying with “Anhydrone” and then absorbing and weighing the hydrogen sulphide directly in a bulb of “Ascarite.” This method was used to determine the sulphur content of a sample of manganese sulphide prepared according to the method described by Biltz and Vl’iechrr~ann.*~* The samples used for the direct weighing method weighed 0-25 g, and were dissolved in the evolution flask in the usual way.Samples of the same specimen of manganese sulphide weighing 0.05 g were similarly dissolved and the hydrogen sulphide determined by the volumetric method as described in this paper. The results, expressed as percentage of sulphur- in the specimen of manganese sulphide, were as follows- Weight of sample Sulphur, Mean % By direct weighing of hydrogen sulphide . . 0.26 g 34.4, 34.4 34.4 By hypochlorite titration . . .. . . 0.05 g 34.1, 34-7, 34.2 34.3 From these figures it is concluded that the hypochlorite titration method is an accurate It gives results that are probably correct to within means of determining hydrogen sulphide.0.3 per cent. of the true value. COMPARISONS WITH THE STANDARD GRAVIMETRIC (BARIUM SULPHATE) METHOD- Borings from a special iron - sulphur ingot, prepared in the Chemistry Laboratory of the British Iron and Steel Research Association, were sampled by mixing and quartering. Twelve determinations by the new volumetric method gave a mean of 0-384(5) per cent. of sulphur. Independent umpire analyses on 5-g samples carried out by Miss M. Dalziel in * It is noteworthy that Biltz and Wiechmann claimed that manganese sulphide so prepared is stoicheio- metric. The evidence, however, was only that the manganese content was correct for manganese sulphide; the sulphide content of their sample may well have been low, as was that of the manganese sulphide used for the present test.The above test was not, of course, dependent upon having a stoicheiometric sulphur content. Possibly some S (= 32) is replaced by (OH), (= 34).Sept., 19511 DETERMINATION OF HYDROGEN SULPHIDE AND SOLUBLE SULPHIDES 515 the Advanced Analytical Laboratories of this Department, using the standard A.S.T.M. gravimetric method, gave 0.387 per cent. Similar comparisons have also been made with mixtures of zinc sulphide and zinc (volumetric 1-14 per cent.; gravimetric 1.15 per cent.) and of ferrous sulphide and iron (volumetric 1.01 per cent. ; gravimetric 1.01(5) per cent.). These tests confirm that the evolution - volumetric method described here gives results for sulphur in iron and other metals that are in good agreement with the gravimetric method. There can be no doubt that it is reliable within the required limits of + 1 per cent.; it is probably better than this.The data appear to suggest that the volumetric method may be giving results about 0.6 per cent. lower than the gravimetric method, but even the standard barium sulphate method is probably not of the highest absolute accuracy. The present authors do not know of any unambiguous tests of the ultimate reliability of the barium sulphate determination, whereas its liability to errors is well known.2925 TIME REQUIRED FOR A DETERMINATION AND LIMIT OF SENSITIVITY A single determination of sulphur in iron can be completed in about 1i hours, allowing 1 hour for the dissolution of the sample.For the analysis of mixtures of hydrogen sulphide and hydrogen the time required is only about half an hour. As two determinations can easily be carried out simultaneously, an experienced worker with several sets of apparatus could perform 8 to 12 analyses in a working day. Since the hypochlorite method is more than four times as sensitive as the ordinary method by virtue of its stoicheiometry and its better end-point, it is worth considering how small a quantity of sulphur as hydrogen sulphide or soluble sulphide can be detected. Some tests made with a small absorption flask and an ordinary 5-ml micro-burette showed that the end-point is sensitive to +0.005 ml of 0.062 N sodium thiosulphate. Hence, allowing this uncertainty on both the unknown and the blank, the uncertainty of the difference would be +O-Ol ml.This is a limiting sensitivity of 0-0024 mg of sulphur, which would be 0.00024 per cent. by weight in a l-g sample of iron, or 0.00017 per cent. by volume as hydrogen sulphide in 1 litre of a gas mixture at N.T.P. The above estimate gives the sensitivity; the smallest amount of sulphur that could be determined with a given accuracy is, of course, correspondingly greater. For example, 0.01 per cent. of sulphur in a l-g sample of iron could be readily determined correct to the nearest 0.001 per cent. With larger amounts of sample or higher sulphur contents the full accuracy of the method (i.e., within +O-3 to 0.6 per cent. of the true sulphur content) can be achieved.Two of us (A. L. and D. A. S.) thank the British Iron and Steel Research Association for B.I.S.R.A. Bursaries, during the tenure of which the above work was done. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Kitchener, J. A., Bockris, J. O’M., and Liberman, A., Disc. Farad. SOC., 1948, No. 4, 49. Lundell, G. E. F., Hoffman‘ J. I., and Bright, H. A., “Chemical Analysis of Iron and Steel,” John Wiley and Sons Inc., New York, 1931, chap. XII. Scott, W. W., and Furman, N. H., “Standard Methods of Chemical Analysis,” Fifth Edition, D. Van Nostrand Co. Inc., New York, and the Technical Press Ltd., London, 1939, pp. 1442-1444. -- , Ibid., p. 914. Luniell, G. E. F., Hoffman, J. I., and Bright, H. A., op. cit., p. 242. Treadwell, F. P., and Hall, W. T., “Analytical Chemistry,” John Wiley and Sons Inc., New York, Hoar, T.P., and Eyles, G. E. S., Analyst, 1939, 64, 666. Brunck, O., 2. anal. Chem., 1906, 45, 541. Etheridge, A. T., in Mitchell, C. A. (Editor), “Recent Advances in Analytical Chemistry,” Shaw, J . A., I n d . Eng. Chem., Anal. Ed., 1940, 12, 668. Treadwell, F. P., and Hall, W. T., 09. cit., p. 557. Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Macmillan Willard, H. H., and Cake, W. E., J . Amer. Chem. SOC., 1921, 43, 1610. Pepkowitz, L. P., Anal. Chem., 1948, 20, 968. Mellor, J. W., “Comprehensive Treatise of Inorganic and Theoretical Chemistry,” Longmans, 1930, p. 583. J. & A. Churchill Ltd., London, 1931, Volume 11, p. 174. and Co., Ltd., London, 1943, p.588. Green and Co., Ltd., London, 1922, Vol. 11, p. 252.516 KITCHENER, LIBERMAN AND SPRATT [Vol. 76 16. 17. Foerster, F., 2. Elektrochem., 1917, 23, 138. 18. 19. SO. 21. 22. Cullen, G. E., and Hubbard, R. S., J . BioZ. C h e w , 1919, 37, 511. Gmelin’s Handbuch dev Anorg. Chem., 1927 (8), 6, 265. Fuchs, P., Bodenkunde u. P’anzenerniihr., 1942, 28, 385. Vogel, A. I., “Textbook of Quantitative Inorganic Analysis,” Longmans, Green and Co., Ltd., London, 1943, p. 459. Bendall, J. R., Mann, F. G., and Purdie, D., J . Chern. Soc., 1942, 157. See, for example, Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” John Wiley and Sons Inc., New York, 1929, p. 207; :Scott, W. W., and Furman, N. H., op. cit., p. 201, footnote. Sherman, C. W., Elvander, H.I., and Chipman, J., J . Metals, 1950, 188, 334. Biltz, W., and Wiechmann, F., 2. anorg. Chem. 1936, 228, 268. Kolthoff, I. M., and Sandell, E. B., op. cit., chajp. XX and pp. 718-719. 23. 24. 25. DEPARTMENT OF INORGANIC AND PHYSICAL CHEMISTRY IMPERIAL COLLEGE LONDON, S.W.7 DISCUSSION MR. F. L. OKELL said that the authors were to be congratulated upon having removed one of the three long-standing difficulties of the volumetric evolution methods for determining hydrogen sulphide, namely, the uncertainty that attended the iodimetric titration of metallic sulphides. The authors’ opinion that the reaction was not stoicheiometric was confirmed by his own experience of the method, By applying the alkaline hypochlorite oxidation to this determination they had improved its scientific aspect, as distinct from manipulative technique, and so increased our knowledge of this attractive but hitherto disappointing method.With this improvement the method should become of service for many purposes other than those of the foundry chemist. He asked if the all-glass apparatus described by the author was indispensable to the method and mentioned that S. G. Clarke (Analyst, 1931,56, 436) had used corks and black rubber tubing in an evolution method for hydrogen sulphide on samples weighing but 0.1 g and got good results. DR. KITCHENER, in reply, thanked Mr. Okell for his remarks and said that, as an all-glass apparatus eliminated one possible source of trouble, they had not investigated the suitability of cork and rubber tubing. However, it was worth mentioning that in other work they had found polythene tubing satis- factory and had proved by direct test that, unlike rubber, it did not absorb hydrogen sulphide.Polyvinyl chloride tubing was also available and was superior to rubber for many purposes in the laboratory, e.g., in the polarograph. MR. J. HASLAM, although not necessarily implying that the test results would be affected, rather doubted the stoicheiometry of the reaction between hypochlorite, iodide and acid. That seemed to him to be taken account of in the best methods of standardisation of hypochlorite. In these methods what ultimately happened was that a slight deficiency of arsenite solution was added to a known amount of hypochlorite, and only a t this stage, when a very small amount of hypochlorite was present in excess, was iodide added, the iodine then liberated being titrated by the addition of a further small volume of standard arseni te . MR. R. F. MILTON said that in his experience titration of hypochlorites with potassium iodide and thiosulphate was unsatisfactory owing to the end-point being indefinite and variable. He had found that the arsenite and iodine titration was to be preferred. DR. KITCHENER replied that although the arsenite method was no doubt preferable in some instances for determining hypochlorite, especially when chlorates were present (e.g., in bleaching powder), they considered the direct potassium iodide reaction entirely :satisfactory for pure solutions of hypochlorites. This opinion was supported by the literature (see reference 12, above) and by the reproducibility of their results (Table I). DR. J. H. HAMENCE asked if the method could be used to estimate hydrogen sulphide in water directly, without distillation. MR. T. MCLACHLAN said that he did not think it would be possible to apply this method to the deter- mination of traces of hydrogen sulphide in waters on account of the large amount of dissolved air. More- over, when they were present in traces, sulphur dioxide was frequently there in addition to hydrogen sulphide. With regard to the reliability of the methods of Dr. Kitchener and his co-workers. it should be remembered that in the determination of available chlorine in sodium hypochlorite solution, the analyst was concerned only with the amount of hypochlorite, whereas Dr. Kitchener was indifferent to the forms in which the chlorine was present, as long as he could eventually determine the total amount after liberation with acid and iodide. The end-point in their titrations was very sharp. DR. KITCHENER concurred.

ISSN:0003-2654    

DOI:10.1039/AN9517600509

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

Sept., 19511 CLULEY 51 7 The Determination of Germanium The following three papers were presented at the Meeting of the Society on Wednesday, May 2nd, 1951. Part I. Titration of Mannito-Germanic Acid* BY H. J. CLULEY In aqueous solution germanium dioxide reacts with mannitol to form a strong, complex acid ; germanium can be determined volumetrically by titration with sodium hydroxide solution of the mannito-germanic acid so formed. A method has been evolved in which the volumetric procedure is applied after a preliminary separation of the germanium by precipitation. as sulphide. By a simple modification the interference of arsenic is readily obviated. GERMANIUM, long a neglected element, has in recent years achieved prominence owing to its extensive application as a semi-conductor in crystal rectifiers and similar devices.To meet the consequent demand germanium is now being produced in this country by extraction from flue dusts occurring in the producer systems of gasworks using certain coals in which germanium is a trace c0nstituent.l The analytical difficulties encountered during work on the extraction of germanium from flue dusts encouraged a search for new methods of deter- mination of the element and the results of this analytical investigation are reported in this and the following two papers (pp. 523 and 530). The majority of published procedures for the determination of germanium suffer from lack of selectivity and, in particular, from interference from arsenic, with which germanium is commonly associated. This is a serious objection, as a complete separation of germanium from arsenic is not readily achieved. The procedure most widely used for the determination of germanium is the tannin method, with which a separation from all elements other than tantalum, niobium and tungsten is stated to be theoretically possible.2 Initially, the tannin method was investigated, the conditions of precipitation recom- mended by Davies and Morgan2 being used.I t was found that at low concentrations the germanium - tannin precipitate tended to become colloidal, with consequent low results, and it was concluded from this investigation that the method was unlikely to be wholly satisfactory for the small amount of germanium it was required to determine. The method of precipitation employed by Davies and Morgan has subsequently been criticised by HolnessJ3 who advocated precipitation from an oxalic acid solution in preference to sulphuric acid solution.A further disadvantage of the tannin method is the tedious preliminary ignition at 600" C, with repeated nitric acid oxidation, that is necessary to ensure the absence of the volatile germanous oxide before the final ignition at 900" C. This paper describes a volumetric method applied to the determination of germanium ; the development of an absorptiometric method and its subsequent application to the determination of germanium in flue dust, coal and coke are described in the following two papers. Attention was then directed to volumetric and absorptiometric methods. EXISTING VOLUMETRIC METHODS There appear to be only two published procedures for the volumetric determination of germanium.Willard and Zeuhkle4 proposed a method based on the formation of a thio- germanate in an acetate-buffered solution. The thiogermanate was oxidised with standard iodine solution the excess of which was determined by titration with thiosulphate. The authors state that the method is of limited application. The other method is due to TchakirianJ5 who found that germanium dioxide in aqueous solution reacted with mannitol (and with other polyhydric alcohols) to form a strong complex acid. The mannito-germanic acid so formed could be titrated with sodium hydroxide solution * Taken from a Thesis submitted t o the University of London for the degree of M.Sc.518 CLULEY: THE DETERMINATION OF GERMANIUM [Vol.76 to the phenolphthalein end-point, one molecule of sodium hydroxide being equivalent to one atom of germanium. The mannito-germanic acid also reacted with an iodate - iodide mixture and the liberated iodine could be titrated with thiosulph’ate. Tchakirian stated that the alkalimetric procedure could not be applied to germanium solutions containing strong acids, and there is no evidence from his oraginal5 or subsequent6y7 papers that either of his methods had been used except for pure germanium solutions. The alkalimetric titration of boric acid in the presence of mannitol is a procedure universally used, and the corresponding method for the determination of germanium appeared worthy of investigation. In particular, the method for boron is applicable to solutions containing strong acids and this cast doubt on Tchakirian’s statement that the method for germanium was not applicable under these conditions.An investigation was therefore made in an attempt to verify Tchakirian’s observations and to extend the usefulness of his met hod. THE TITRATION OF MANNITO-GERMANIC ACID Preliminary experiments were performed on the formation of mannito-germanic acid, and pH curves were drawn for the neutralisation of the acid with sodium hydroxide. These curves gave inflection points at about pH 7.8 and confirmed that 1 molecule of sodium hydroxide was equivalent to 1 atom of germanium. Three different titration procedures were then examined, germanium solutions containing free sulphuric acid being used. (a) Calcium carbonate neutralisation method-The solution was neutralised with calcium carbonate and, after filtration, mannitol was added and the solution was titrated with sodium hydroxide solution to the phenolphthalein end-point.This procedure was based on the Wherry method for determination of boron.* (b) Double indicator method-The solution was adjusted with the sodium hydroxide solution to the P-nitrophenol end-point ; after addition of mannitol the solution was titrated to the phenolphthalein end-point. This type of procedure is commonly used for titration of boric acid.g (c) Fixed pH method-This was a modification of the foregoing method (6) in which the pH value of the solution before addition of the mannitol and the pH value at the end of the titration are the same.This procedure has also been used for the titration of boric acid. lo y1l It was found that all three methods were suitable for the determination of 1 to 10-mg quantities of germanium, approximately 0.02 N sodium hydroxide being used. Hence it was established that an alkalimetric procedure could be applied to the determination of germanium in solutions containing a strong acid. Methods (b) and (c) are more rapid as they avoid the necessity for filtration. The following theoretical comparison of methods (b} and (c) shows that the latter is likely to be less influenced by other substances present in solution. In the double indicator method (b) the pH value at the P-nitrophenol end-point is about 6.0, falling to about 4.0 on addition of mannitol and increasing to about 8.4 at the phenol- phthalein end-point.In this type of titration, therefore, the formation and subsequent neutralisation of the mannito-germanic acid are accompanied by a net change of pH value from 6.0 to 8.4. If substances exerting any buffering action are present the amount of sodium hydroxide required to effect this net pH change will be increased, with consequent error in the determination. In method (c) the neutralisation of the mannito-germanic acid is carried only to the stage where the pH value is the same as that before addition of mannitol, the net pH change is zero and the error due to any bufEer effect vanishes. In practice, buffering agents can in fact be tolerated provided that they ,are not present in sufficient concentration to render the titration end-point indefinite.In this type of titration the neutralisation of the mannito-germanic acid is obviously incomplete and an empirical relationship exists between the germanium and the sodium hydroxide solution. This relationship must be established by titration of known amounts of germanium under closely standardised conditions. The fixed pH method, (c), was therefore preferred as it appeared less susceptible to errors caused by the ionic environment of the germanium. A further advantage is the reduction in the size of the blank, a matter of some importance for the small titrations involved. The fixed pH value selected was 6.2, and the standardisation titrations were carried out in the following manner.Sept., 19511 PART I. TITRATION OF MANNITO-GERMANIC ACID 519 To the weakly acid germanium solutions, of volume about 80 ml, 7 drops of bromocresol purple indicator were added.Carbonate-free sodium hydroxide solution, 0.01 85 N , was then added until a pH of 6.2 was reached, as indicated by colour comparison with pH 6.2 buffer solution containing indicator. Then l o g of mannitol were added and the solutions were titrated with the sodium hydroxide solution until a pH of 6.2 was again attained. The usual correction for the blank was applied. TABLE I FIXED pH TITRATION METHOD : STANDARDISATION TITRATIONS Corrected titre of Germanium taken, sodium hydroxide, mg ml 0.0185 N 1 0.67, 0.70 3 2-19, 2-20 6 4-37, 4-42 10 7-35, 7.37 20 14.57, 14-59 . Weight of germanium equivalent to 1 ml of sodium hydroxide, mg 1.461 1.367 1.366 1-359 1-'371 NOTE-For complete neutralisation of mannito-germanic acid the theoretical relationship is 1 ml of 0.0185 N sodium hydroxide = 1.343 mg of germanium.From these titrations the germanium equivalence of the sodium hydroxide solution was calculated for each weight of germanium titrated, as shown in Table I. Theoretically the germanium equivalence of the titrant will vary slightly with the weight of germanium titrated, but it is clear that for the small amounts of germanium under consideration the use of a mean equivalence factor will occasion little error. It will be seen that the equivalence factor thus established empirically for this method of titration differs only slightly from the theoretical relationship obtaining for complete neutralisation of the mannito-germanic acid, so that the loss of sensitivity is very small.To illustrate this type of titration a pH curve was prepared of the titration of 10mg of germanium under the above conditions, and this is shown in Fig. 1. The dotted portion Fig. 1. Graph of a fixed pH titration of the first curve shows how the titration continues if mannitol is not added; the dotted portion of the second curve shows the completion of the neutralisation of the mannito- germanic acid after pH 6.2.520 CLULEY: THE DETERMINATION OF GERMANIUM [Vol. 76 TITRATION OF GERMANIUM AFTER: SEPARATION AS SULPHIDE Substances that would be expected to interfere with the titration of germanium fall (a) Buffering agents, if they are present in a sufficiently high concentration to render the titration end-points indefinite.(b) Bases precipitated by sodium hydroxide within the pH range of the titration. (This is a special case of buffering.) (c) Compounds which react similarly with mannitol, e.g., boric acid. For the general application of the volumetric method a preliminary separation of the germanium will normally be necessary. Precipitation as sulphide serves to separate germanium from boron and from the majority of bases of type (b). Experiments were therefore carried out on the application of the volumetric method after separation of germanium in this manner. Known amounts of germanium were precipitated as sulphide and the filtered precipitates were dissolved in ammonium hydroxide solution and oxidised with hydrogen peroxide.After boiling with sodium hydroxide solution to eliminate ammonia and hydrogen peroxide the solutions were just acidified with sulphuric acid and the fixed pH titration method applied. At first difficulties were encountered owing to the effect on the indicator of oxidising substances still present in solution. This trouble was overcome by boiling with an excess of sulphuric acid after the elimination of ammonia and hydrogen peroxide, then making the solution just acid and applying the volumetric method. The results obtained in this manner are shown in Table 11. into three classes- TABLE ICI TITRATION OF GERMANIUM AFTER PRECIPITATION AS SULPHIDE Germanium taken, Germanium found, mg mg 1 1.00, 0.98 3 2-98, 3-04 6 6.03, 6.01 10 9.96, 9.90, 9.96, 9.85 These results show that the volumetric procedure can successfully be applied after separation of germanium as sulphide from pure solutions.However, in practice, other elements of the analytical sub-group IIB will accompany germanium in this separation. Of these elements arsenic is commonly associated. with germanium and interferes with most methods for its determination. For these reasons the effect of arsenic on the volumetric method was examined in some detail. TITRATION OF GERMANIUM IN THE PRESENCE OF ARSENIC Titration by the fixed pH method of known amounts of germanium in the presence of arsenic established that as much as 100 mg of tervalent arsenic could be tolerated without detriment to the determination of 10 mg or less of germanium. Similar amounts of quin- quevalent arsenic were found to interfere owing to pronounced buffering of the solution.If the sulphide separation procedure previously examined were used, the co-precipitated arsenic would be oxidised by the hydrogen peroxide to the quinquevalent state and would therefore interfere with the final titration. The nm-interference of tervalent arsenic suggested that this difficulty could be resolved by reduction of the arsenic to the tervalent state before titration. For this purpose reduction with sulphur dioxide was used because the excess of reducing agent could readily be removed by boiling. This reduction was applied to arsenate solutions containing germanium and in the subsequent titrations no interference from buffering occurred. Finally, this modification was applied to the volumetric determination of germanium after precipitation as sulphide from solutions containing arsenic. The results of these experiments are shown in Table 111.Sept., 19511 PART I.TITRATION OF MANNITO-GERMANIC ACID 521 TABLE I11 DETERMINATION OF GERMANIUM IN THE PRESENCE OF 100mg OF ARSENIC Quinquevalent arsenic reduced to the tervalent state before titration Germanium added to 100 mg of arsenic, mg . . .. 1 3 5 5 6 Germanium found, mg . . .. .. .. . . 0.98 2.93 4.98 4.94 5.92 The results in Table I11 show that small amounts of germanium can be determined in the presence of substantial amounts of arsenic. The effect of other elements has not been extensively investigated. Traces of antimony and tin can be tolerated, but large amounts interfere owing to the formation of insoluble hydroxides or basic salts.RECOMMENDED METHOD FOR THE VOLUMETRIC DETERMINATION OF GERMANIUM APPARATUS- It is necessary to use beakers or flasks made from boron-free glass or platinum apparatus where strongly alkaline solutions are employed, as any boron dissolved from borosilicate glassware will be titrated with the germanium. SPECIAL REAGENTS- Sodium germanate solution-Transfer to a platinum crucible 1.4408 g of pure ignited germanium dioxide, fuse with 5 g of sodium carbonate and dissolve the cold melt in hot water; just acidify the solution with dilute sulphuric acid, boil to eliminate carbon dioxide, cool and dilute to 1000 ml. 1 ml of solution G 1 mg of germanium. Bufler solution, p H 6.2-Mix 33.9 ml of 0.1 M citric acid with 66.1 ml of 0.2 M di-sodium hydrogen phosphate. STANDARDISATION OF THE SODIUM HYDROXIDE SOLUTION- Transfer volumes of standard sodium germanate solution, covering the range of 1 to 20 mg of germanium, to 250-ml flasks.Dilute each volume to about 50 ml, boil for 5 minutes to eliminate any carbon dioxide and then cool. Add 7 drops of bromocresol purple indicator and add from a burette carbonate-free sodium hydroxide solution, approximately 0.02 N , until a pH value of 6-2 is reached as indicated by colour comparison with an equal volume of pH 6.2 buffer solution containing 7 drops of indicator. Add 10 g of mannitol and titrate with the sodium hydroxide solution until a pH value of 6.2 is again reached. Carry out a blank determination and correct the titration figures accordingly.From the corrected titration figures calculate the equivalence of the sodium hydroxide solution, in terms of mg of germanium per ml, for each weight of germanium titrated. It will be observed that the germanium equivalence of the sodium hydroxide solution differs slightly for each weight of germanium titrated, but for the small weights of germanium used it may be permissible to use a mean equivalence factor. PROCEDURE- The sample solution, of volume about 100 ml and containing 1 to 20 mg of germanium, should not contain large amounts of antimony or tin, and sulphuric acid should be the only acid present. Transfer the sample solution to a conical flask and add sufficient diluted sulphuric acid (1 + 1) to give a sulphuric acid concentration of about 5-5 N .Pass a rapid stream of hydrogen sulphide through the solution for 30 minutes, cork the flask and allow it to stand overnight. Filter the solution through a close filter-paper and wash the precipitate well with 5.5 N sulphuric acid saturated with hydrogen sulphide. Dissolve the precipitate through the paper into a boron-free flask (or platinum dish) with three successive 8-ml portions of diluted ammonium hydroxide solution (2 + 1) and wash the paper first with dilute ammonium522 CLULEY: THE DETERMINATION OF GERMANIUM [Vol. 76 hydroxide solution (1 + 9) and finally with hot water. Add 20ml of 6 per cent. w/v hydrogen peroxide and allow to stand for 10 minutes in the cold to ensure complete oxidation of the sulphides. Add 5 ml of freshly prepared 5 N sodium hydroxide and boil the solution vigorously until all ammonia and hydrogen peroxide has been evolved, adding hot water from time to time to maintain the volume of the solution.Add 8ml of dilute sulphuric acid (1 + 6 ) and boil for 10 minutes to remove any oxides of nitrogen formed from oxidation of the ammonia. If desired, the solution may be transferred to ordinary glassware at this stage. Pass a rapid stream of sulphur dioxide into the hot solution until it is cold; this takes about 30 minutes. Boil off the excess of sulphur dioxide, ensuring complete removal by continuing the boiling for 5 minutes after the smell of sulphur dioxide can no longer be detected. The treatment with sulphur dioxide may be omitted if arsenic is absent. Add 2 drops of bromocresol purple indicator and then add freshly prepared 5 N sodium hydroxide until the solution is just alkaline.Add N sulphuric acid until the indicator just turns yellow, dilute to about 80m1, boil for 5 minutes to eliminate any carbon dioxide and cool. Add 5 more drops of bromocresol purple indicator, adjust to pH 6.2, add 10 g of mannitol and titrate to pH 6-2 as in the standardisation titrations. Perform a blank determination, correct the sample titration figure and deduce the germanium content of the sample solution. NOTES- The presence of germanium in reagents is unlikely, but it is necessary to carry out blank determinations because of the possibility of deriving traces of boron from reagents or glassware. The titrations may be carried out with the aid of a pH meter if required, but it is advantageous to retain the use of the indicator so that readings need only be taken when the indicator colour shows proximity to pH 6-2.The method is equally satisfactory for large amounts of germanium, for which a stronger solution of sodium hydroxide should be used. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Chirnside, R. C., Times Review of Industry, July, 1950. Davies, G. R., and Morgan, G. T., Analyst, 1938, 63, 388. Holness, H., Anal. Chim. Acta, 1948, 2, 254. Willard, H. H., and Zeuhlke, C. W., Ind. Eng. Chem., Anal. Ed., 1944, 16, 322. Tchakirian, A., Compt. Rend., 1928, 187, 229. -, Bull. SOC. Chim. France, 1932, [4] 51, 846. Wherry, E. T., J . Amer. Chem. SOC., 1908, 30, 1687. Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” John Wiley and Sons Foote, F.J., Ind. Eng. Chem., Anal. Ed., 1932, 4, 39. Hollander, M., and Rieman, W., Ibid., 1946, 18, 788. -, Ibid., 1943, [5] 10, 98. Inc., New York, p. 611. RESEARCH LABORATORIES THE GENERAL ELECTRIC COMPANY LIMITED WEMBLEY, MIDDLESEX and SIR JOHN CASS COLLEGE JEWRY STREET LONDON, E.C.3Sept., 19511 PART 11. ABSORPTIOMETRIC DETERMINATIOX WITH PHENYLFLUORONE 523 Part 11. Absorptiometric Determination with Phenylfluorone* BY H. J. CLULEY An absorptiometric method is described for the determination of germanium by means of 2 : 3 : 7-tnhydroxy-9-pheny1-6-fluorone (phenyl- fluorone). The method is about four times as sensitive as the molybdenum blue method. By a simple distillation from hydrochloric acid solution the germanium is readily separated from the few elements that have been found to interfere with the phenylfluorone method.ABSORPTIOMETRIC methods for the determination of germanium have been investigated to only a slight extent. The only methods at present available are those based on the formation of the yellow germano-molybdic acid1 or on its subsequent reduction to molybdenum blue.2t3 These procedures have the disadvantage that similar reactions are given by silicon, phosphorus and arsenic, and a complete separation of germanium from arsenic in particular is not readily achieved. A search of the literature was carried out in an endeavour to find a more selective colour reaction for germanium that might form the basis of an absorptiometric method. Gillis, Hoste and Claeys? who have examined a number of derivatives of fluorone for their potential value as analytical reagents, have stated that 2 :3 :7-trihydroxy-9-phenyl- 6-fluorone (phenylfluorone, Fig.1) is a specific reagent for the detection of germanium. C H I d Fig. 1. Phenylfluorone Their spot test is carried out on a test paper pre-treated with an alcoholic solution of the reagent acidified with hydrochloric acid; a drop of a test solution containing quadrivalent germanium gives a pink colouration that does not disappear on treatment with 6 N nitric acid. The reaction of germanium with phenylfluorone has now been studied as the basis of an absorptiometric method. PRELIMINARY EXPERIMENTS Phenylfluorone was prepared from tri-acetylhydroxy-hydroquinone and benzaldehyde (see Procedure) by the method employed by Gillis, Hoste and Claey~.~ The reagent is only slightly soluble in alcohol and similar solvents, but its solubility in alcohol is increased by adding a small volume of dilute hydrochloric or sulphuric acid, which appears to convert phenylfluorone to the corresponding salt.For preliminary experiments a solution of 0-05 g of phenylfluorone in a mixture of 95 ml of alcohol and 5 ml of dilute sulphuric acid (1 + 6) was used, and this is subsequently referred to as a 0.05 per cent. solution. This reagent solution, yellow in colour, gave a colour reaction with quadrivalent ger- manium in dilute hydrochloric or sulphuric acid solution. The colour so formed was orange, presumably due to a combination of the yellow colour of the reagent and the pink colour of the germanium complex.On standing, the germanium compound tended to precipitate, but the use of gum arabic as a protective colloid was effective in stabilising the colour, and this was used throughout the subsequent experiments. Owing to the strong colour of phenylfluorone itself in acid solution, it was found necessary for absorptiometric measurements to use as a reference solution a blank containing the reagent but no germanium. Under these conditions of measurement the maximum absorption of * Taken in part from a Thesis submitted to the University of London for the degree of MSc.524 CLULEY: THE DETERMINATLOK OF GERMAKIUM [Vol. 76 the germanium colour was in the blue - green regAon. The initial absorptiometric measure- ments showed that the intensity of the germanium colour tended to increase with time.In their spot test, Gillis, Hoste and Claeys4 had found it necessary to treat the colour spot with 6 N nitric acid to destroy the colours produced by certain interfering elements. Similar treatment of germanium colours developed in dilute sulphuric acid solution was found to result in destruction of the reagent. 'This was possibly due to the presence of nitrous acid, but the use of nitric acid was not pursued. COLOUR DEVELOPMENT IN SULPHURIC ACID SOLUTION Systematic experiments were carried out to (establish conditions suitable for the rapid development of the germanium - phenylfluorone colour. For this work a sulphuric acid medium was used, as it was considered that in the preparation of a sample solution the use of hydrochloric acid should normally be avoided owing to the volatility of germanium tetrachloride. The concentration of sulphuric acid, reagenl and germanium were varied, but in all experiments 5 ml of 0.5 per cent.w/v gum arabic solution were added and the final volume was 50 ml. The absorptiometric measurements were carried out on a Spekker absorptiometer with 1-cm cells, the tungsten lamp and Ilford No. 603 blue - green filters, and as a reference solution a blank was used, prepared in the same manner as the test solution except that no germanium was added. The rates of colour development were found to increase with increasing reagent con- centrations and to decrease with increasing sulphuric acid concentrations. At low concen- trations of acid precipitation of phenylfluorone occurred.The use of 15 ml of a 0.02 per cent. reagent solution, and of 10 ml of dilute sulphuric acid (1 + 6), giving an acid concentration of about 1.05N in the final volume of 50m1, proved satisfactory and ensured complete colour development within 30 minutes. The relationship between absorption and germanium concentration with these conditions of colour development is shown in Table I. There is a slight deviation from Beer's law that becomes more marked with increasing germanium concentration. The sensitivity of the reaction is ,apparent. TABLE 1 EFFECT OF GERMANIUM CONCENTRATION ON COLOUR DEVELOPMENT IN Sulphuric acid concentration 1.05 N; 15 ml of 0.02 per cent. reagent solution used Germanium, pg .. . . * . . . 10 20 30 40 50 Drum reading, measured after 30 minutes 0.200 0.405 0.608 0.800 0.945 With the conditions outlined above, absorptiometric determinations of 25-pg quantities of germanium were carried out in the presence of a large number of individual elements to assess their effect on the method. Initially the weight of each element added was 25 mg (1000 times the weight of germanium) except for calcium and boron, where low solubility necessitated the use of smaller weights, and for sodium, potassium and ammonium, where 1-g quantities were used. This work showed that the following ions did not interfere: NH,', Na', K', Li', Cu", Ag', Be", Mg", Ca", Zn", Cd", Hg", Al"', Cr"', Mn", Fe", Co", Ni", B03"', PO4"' and Cl'. The following ions were found to cause interference, the numbers in brackets giving the maximum permissible concentration of the element expressed as a multiple of the germanium concentration: Ga"' (< l), Ti"" (2), Sn" (< l ) , Sn"" (< l), As"' (50), AsO,"' (loo), Sb"' (< I), Bi"' (lo), MOO," (< 1) and Fe"' (10).It will be observed that the interference of arsenic was very slight, but that that of gallium, tin, antimony and molybdenum was most marked. Strong oxidising agents such as dichromate and permanganate were also found to interfere by destroying the reagent. Under the conditions used the absorptiometric procedure was, therefore, selective but not specific for germanium. Modification of the conditions of colour development showed little promise of eradicating the interferences observed and it was clear that a preliminary separation of the germanium would be necessary for the general application of the method.The interference of arsenic, antimony, bismuth, molybdenum and tin precluded thc SULPHURIC ACID SOLUTIONSept., 19511 PART 11. ABSORPTIOMETRIC DETERMINATIOK WITH PHENYLFLUORONE 525 use of the sulphide method for separation of the germanium, but the alternative procedure of distillation of the tetrachloride appeared to be promising. Arsenic is the element that normally causes most difficulty in this separation, and the use of fractional distillation in a current of chlorine has been proposed as a method of avoiding the partial co-distillation of a r ~ e n i c . ~ However, the relatively low sensitivity of the reagent to arsenic suggested that a simple distillation without any such precautions might effect an adequate separation.The distillation is normally carried out from solutions containing at least half their volume of concentrated hydrochloric acid and similar concentrations of acid are present in the distillate. Before examining this method of separation it was therefore necessary to establish conditions suitable for the application of the absorptiometric procedure to hydro- chloric acid solutions. COLOUR DEVELOPMENT IX HYDROCHLORIC ACID SOLUTION The work on colour development in hydrochloric acid solution was carried out by the same general procedure as was used for the study of the reaction in sulphuric acid medium. The results led directly to the conditions finally adopted for the absorptiometric determination of germanium, and it is considered useful to record these results in some detail.EFFECT OF PHENYLFLUORONE CONCEKTRATION- With 25-pg quantities of germanium and a final hydrochloric acid concentration of 1.0 K , the effect of reagent concentration was studied for the range 2 to 10 ml of the 0.05 per cent. solution. In each experiment sufficient alcohol was added to bring the total volume of alcohol present to 15 ml. The results recorded in Table I1 again show the marked dependence of rate of colour development on reagent concentration and it will be seen that, for the acid concentration used, only the 8 and 10-ml quantities of the 0.05 per cent. phenylfluorone solution could effect complete development of the colour within 30 minutes.For subsequent work the intermediate quantity, 9 ml, was used, but to avoid separate addition of alcohol this was added as 15ml of a 0.03 per cent. solution. Owing to the significant absorption of light by the reagent solution alone, it is a disadvantage to use quantities of reagent greatly in excess of the amount required to give an adequate rate of colour development. TABLE I1 EFFECT OF PHENYLFLUORONE CONCENTRATION ON COLOUR DEVELOPMENT IN 25 pg of germanium used; final acid concentration 1.0 N HYDROCHLORIC ACID SOLUTION Drum readings, measured after 0.05% reagent, Alcohol, I -l ml ml 15 min. 30 min. 60 min. 90 min. 2 13 0.039 0.060 0.070 0.106 4 11 0.348 0.410 0,445 0.453 6 9 0.478 0.480 0.488 0.498 7 8 0.462 0.469 0.479 0.483 8 7 0.472 0.478 0.476 0,482 10 6 0.484 0.483 0.484 0.486 A EFFECT OF HYDROCHLORIC ACID COXCENTRATION- With 25-pg quantities of germanium and 15 ml of 0.03 per cent.phenylfluorone solution the effect of hydrochloric acid concentration was studied for the range 0.25 to 5.0 N. The results, recorded in Table 111, again show that the rate of colour development decreases with increasing acid concentration. For the quantity of reagent used, the acid range of 0.25 to 1.5 N is effective in producing rapid colour development and in giving almost identical drum readings for the fully developed colours. However, this does not mean that, within this range, the acid concentration need not be closely controlled. It was found that the absorption of the reagent solution itself vaned somewhat with the acid concentration and in the acid range 0.25 to 1.5 N consistent drum readings are obtained only if the acid con- centration in the blank or reference solution is essentially the same as that in the germanium solution,526 CLULEY: THE DETERMINATION OF GERMANIUM [Vol.76 For application of the absorptiometric method to highly acid distillates it was desirable to use the highest acid concentration consistent with the requirements of rapid colour Germanium, yg Fig. 2. Calibration graph for Determination of Gmnanium in Hydrochloric Acid Solution development. between 1.0 and 1.5 N . For this purpose it was decided to use an acid concentration intermediate TABLE I11 EFFECT OF HYDROCHLORIC ACID CONCENTRATION ON COLOUR DEVELOPMEST I N HYDROCHLORIC ACID SOLUTION 25 pg of germanium and 15 ml of 0.03 per cent. reagent solution used Drum xeadings measured after Hydrochloric acid I A \ concentration 15 min.30 min. 60 min. 90 min. 0.25 N 0.470 0.470 0.475 0.473 1.0 N 0.472 0.475 0.474 0.475 1.5 N 0.468 0.474 0.477 0.473 2.0 N 0,442 0.463 0.472 0.477 3.0 N 0.177 0.237 0.333 0,377 5.0 N 0.014 0.022 0.030 0.03 1 EFFECT OF GERMANIUM CONCENTRATION- With the quantity of phenylfluorone previously decided upon, viz., 15ml of 0.03 per cent. solution, colours were developed and measured for amounts of 10 to 50 pg of germanium. TABLE I V EFFECT OF GERMANIUM CONCENTRATION ON COLOUR DEVELOPMENT IN HYDROCHLORIC ACID SOLUTION Hydrochloric acid concentration 1.15 N ; 15 ml of 0.03 per cent. reagent used Drum reading Germanium, r A > rg Series 1 Series 2 Series 3 Series 4 Mean 10 0.208 0.204 0.209 0.208 0.207 20 0.402 0.403 0.387 0.397 0,397 30 0.577 0.582 0.566 0.575 0.575 40 0,739 0.741 0.725 0.742 0.737 50 0.866 0.871 0.872 0,880 0.872Sept., 19511 PART 11.ABSORPTIOMETRIC DETERMINATION WITH PHENYLFLUORONE 527 For convenience the volume of hydrochloric acid, sp.gr. 1.18, used was 5.0 ml, giving a con- centration of about 1.15 N in the final volume of 50 ml. The results of four such series of measurements, made at intervals over a period of about four months, are shown in Table IV. A calibration graph, prepared from the mean values, is shown in Fig. 2. The deviation from Beer’s law is more marked than was observed for the determination in sulphuric acid solution. STABILITY OF THE DEVELOPED COLOURS- The absorptiometric measurements recorded in Table IV were made after the solutions had stood in the cold for 30 minutes.Subsequent measurements confirmed that the colours were completely developed within this time and showed that their absorption remained constant for at least 14 hours. I t was usually observed that after 2 or 3 days precipitation of the germanium - phenylfluorone compound commenced. EFFECT OF OTHER ELEMENTS- A hydrochloric acid distillation would clearly separate germanium from most elements and it was therefore considered unnecessary to examine the effect of a large number of elements on the absorptiometric determination in hydrochloric acid solution. The effect of arsenic, which would be expected partly to co-distil with the germanium, and of some of the other elements previously found to interfere, was investigated.Under the conditions selected for the absorptiometric determination in hydrochloric acid solution it was found that arsenic did not interfere, that the interference of tin and titanium was diminished but that molybdenum and antimony still interfered strongly. DETERMINATION OF GERMANIUM AFTER SEPARATION BY HYDROCHLORIC ACID DISTILLATION Initially the simple distillation of germanium from pure solutions containing 50 per cent. by volume of hydrochloric acid was examined. This acid concentration, being very nearly the constant boiling composition, was convenient in that the acid concentration of the distillate would approximate to the same known strength throughout the distillation. With solutions of this acid concentration and of initial volume of 50 ml, the absorptio- metric procedure was applied to successive fractions of the distillate.It was established that not less than 95 per cent. of the germanium distilled over in the first 10m1, and that for complete recovery of the germanium it was necessary to distil only 20 ml. In this manner a number of determinations was made on pure germanium solutions, using an initial volume of 50 ml and applying the absorptiometric procedure to an appropriate aliquot of the 20 ml of distillate collected, with the results shown in Table V. A rate of distillation of about 2 ml per minute was used, so that a single distillation was complete in about 15 minutes. TABLE V ABSORPTIOMETRIC DETERMINATION OF GERMANIUM AFTER DISTILLATION FROM HYDROCHLORIC ACID SOLUTION Aliquot of distillate Germanium taken, used Germanium found, Pg PLg 10 50 _ _ 150 500 1000 25/60 25/50 10/50 20/250 20/500 i n - _ 50 148 510 1006 The same procedure was then applied to binary mixtures of germanium with other elements.These included arsenic, titanium and stannic tin, all of which form anhydrous chlorides with relatively low boiling-points, although it was considered unlikely that the last two elements would co-distil under the conditions used. The separations from molyb- denum and antimony were also examined individually as these two elements were known to exert the greatest interference with the absorptiometric determination. The results in528 CLULEY: THE DETERMINATIOii OF GERMANIUM [Vol. 76 Table VI show that none of these five elements interferes with the procedure even when initially present to the extent of 200 times the weight of germanium. TABLE VI ABSORPTIOMETRIC DETERMINATION OF GERMANIUN AFTER DISTILLATION FROM SOLUTIONS CONTAINING OTHER ELEMENTS Weight of Germanium mg /% Arsenic, as As,O, .. .. . . . . .. 10 50 Tin, as SnC14.6H,0 . . . . . . . . . . 10 60 Titanium, as Ti(SO,), . . .. .. . . . . 10 50 Antimony, as SbCl, .. . . 10 50 Molybdenum, as (NH4j~Mo,O,;.4H,d ' . . . . 10 60 Element added element, taken, Germanium found, Pg 60 50 51 49 49 The separation of germanium from other individual elements by this method of distillation was not examined extensively as, apart from the elements discussed above, none of the common elements forms chlorides likely to distil under the conditions employed.However, as a final test, the procedure was applied to known amounts of germanium in the presence of a complex mixture prepared to simulate a flue dust and containing the following ions: Na', K', Cu", Ag', Mg", Zn", BO,"', Al"', Ga"', SO,", Sn"", Pb", Ti"", PO4"', As ", Sb"', VO,', Cr"', Moo4'', Mn", Fe"' and Ni". Each ion was present in amount equivalent to 10 mg of the corresponding oxide except sodium, for which the equivalent of 1 g of the carbonate was added. The results in Table VII show good recovery of the germanium and no evidence of interference. These experiments were designed to assess the possibilities of the method for the determination of germanium in flue dusts, but they serve to show that the method should be applicable to a wide range of materials.TABLE YII DETERMINATION OF GERMANIUM IN SYNTHETIC FLUE DUST MIXTURES Elements present in the mixture Germanium added, Germanium found, Pg Na, K, Cu, Ag, Mg, Zn, B, 20 Al, Ga, Si, Sn, Pb, Ti, P, As, Sb, V, Cr, Mo, Mn, Fe, Ni 20 79 293 RECOMMENDED METHOD FOR THE ABSORPTIOMETRIC DETERMINATION OF GERMANIUM WITH PHENYLFLUORONE PREPARATION OF PHENYLFLUORONE- Dissolve 25 g of tri-acetylhydroxyhydroquinone by warming with a mixture of 150 ml of alcohol, 130 ml of water and 40 ml of diluted sulphuric acid (1 + 1). Add 25 g (about 24 ml) of benzaldehyde and allow the mixture to stand for 8 days, with occasional stirring. Filter by suction the yellow precipitate of phenylfluorone sulphate so obtained and wash the precipitate with the solution mixture.Suspend the precipitate in about 300ml of water, add sufficient sodium hydroxide solution to give a pH value of about 4 to ensure complete hydrolysis to phenylfluorone, stir and allow the mixture to stand overnight. Filter the precipitate of phenylfluorone by suction, and then wash first with water and finally three times with alcohol to remove the last traces of benzaldehyde. Dry the precipitate in a vacuum desiccator. SPECIAL REAGENTS- Phenyl$uorone solution-Dissolve 0.030 g of phenylfluorone by warming with a mixture of 85 ml of alcohol and 5 ml of dilute sulphuric acid (1 + 6), cool and dilute to 100 ml with alcohol. Standard sodium germanate solutiort (stock solution)-Transfer 14408 g of pure ignited germanium dioxide to a platinum crucible, fuse with 5 g of sodium carbonate and dissolveSept., 19511 PART 11.ABSORPTIOMETRIC DETERMINATION WITH PHENYLFLUORONE 529 the cold melt in hot water; just acidify the solution with dilute sulphuric acid, boil to eliminate carbon dioxide, cool and dilute to 1000 ml. 1 ml of stock solution = 1 mg of germanium. Standard sodium germanate solution (working solution)-Prepare freshly when required by 100-fold dilution of the stock solution. 1 ml of working solution = 10 pg of germanium. Gum arabic solution-Dissolve 1.0 g of gum arabic in 200 ml of hot water and cool. APPARATUS- The distillation apparatus consists essentially of a 100-ml distilling flask, a water-cooled condenser and a 25-ml measuring cylinder as receiver. The use of components with ground- glass joints facilitates the assembly and avoids trouble due to attack by hydrochloric acid on rubber bungs.PREPARATION OF CALIBRATION GRAPH- To 50-ml graduated flasks add amounts of 0 to 5 ml of the dilute (working) germanate solution, to cover the range 0 to 50 pg of germanium. Add sufficient water to give a volume of 20m1, then add 5ml of the gum arabic solution and 5.0ml of hydrochloric acid, sp.gr. 1.18, and mix. Cool to about 20" C and add from a pipette 15 ml of the phenylfluorone solution, dilute to 50 ml and mix well. Allow the solutions to stand at about 20" C for 30 minutes. Measure the germanium solutions on the Spekker absorptiometer or similar instrument with 1-cm cells, the tungsten lamp and Ilford No. 603 blue - green filters, using as a reference solution the solution containing no added germanium.Prepare a calibration graph from the results. PROCEDURE- The sample solution should contain 10 to 1000 pg of germanium and should be free from nitrates and other oxidants likely to liberate chlorine from strong hydrochloric acid solutions. Transfer the neutral or alkaline sample solution, or a suitable aliquot, to the 100-ml distillation flask, neutralise if necessary with hydrochloric acid and dilute to 25 ml as indicated by an appropriate line previously marked on the flask. Add 25 ml of hydrochloric acid, sp.gr. 1.18, swirl once and immediately connect the flask to the distillation apparatus. Heat the solution to boiling over a period of about 5 minutes and then distil at a rate of about 2 ml per minute.The distillate should be quite cold. Collect exactly 20 ml of distillate, remove the receiver and stop the distillation. Dilute the distillate to a known volume and transfer to a 50-ml graduated flask an aliquot that should not exceed half the distillate and is expected to contain 5 to 50 pg of germanium. With the same quantities of reagents carry out a blank distillation in the same manner and to a second 50-ml graduated flask transfer the corresponding aliquot of the blank distillate. Treat both the test and blank aliquots as follows, mixing after each addition. Add 5ml of the gum arabic solution followed by sufficient hydrochloric acid, sp.gr. 1.18, to make the total acid in the solution equivalent to 5.0 ml of hydrochloric acid. For this purpose it may be assumed that the 20 ml of distillate contained the equivalent of 10 ml of the concentrated acid.Add sufficient water to give a total volume of about 30 ml and adjust the temperature to about 20" C. Add with a pipette 15 ml of the phenylfluorone solution, dilute to 50 ml and allow to stand at about 20" C for 30 minutes. With the blank as a reference solution, measure the absorption of the test solution under the conditions used in the preparation of the calibration graph. From the graph deduce the germanium content of the measured solution and calculate the germanium content of the sample. DISCUSSION OF METHOD The experimental work had established that the distillation provided a satisfactory separation of germanium from any common element likely to interfere by colour reaction530 CLULEY : THE DETERMINATION OF GERMANIUM [Vol.76 with phenylfluorone. The reagent is, however, destroyed by strong oxidising agents, and for this reason the sample solution should not contain any substance that might liberate chlorine from the hydrochloric acid during the distillation. The practice of carrying out a blank distillation was adopted mainly to ensure that the acid concentrations in the blank and test aliquots were virtually identical. Any interfering elements present in reagents used prior to the distillation are also removed in this manner. Hitherto the absorptiometric determination of germanium has been based on the forma- tion of the yellow germano-molybdic acid or on its subsequent reduction to molybdenum blue, the latter method having the greater sensitivity and selectivity. Comparison with the published data on the molybdenum blue method2y3 shows that the phenylfluorone procedure, as applied in hydrochloric acid solution, is of similar selectivity and is about four times as sensitive.The main advantages of the phenylfluorone procedure are that it can be applied directly to a hydrochloric acid distillate with the minimum of manipulation, and that the distillation can be simply performed without the precautions necessary to attain complete separation from arsenic. After preparation of the sample solutions, the distillation and absorptiometric deter- mination with phenylfluorone can be completed in 2 to 24 hours for duplicate determinations. Hence this procedure should effect a considerable saving of time over such gravimetric methods as determination with tannin.This YS illustrated in the following paper dealing with the application of the phenylfluorone procedure to the determination of germanium in flue dusts. REFERENCES 1. 2. 3. Kitson, R. E., and Mellon, M. G., Ind. Eng. Chem., Anal. Ed., 1944, 16, 128. Hybbinette, A. G., and Sandell, E. B., Ibid., 1942, 14, 715. Boltz, D. F., and Mellon, hl. G., Anal. Chem., 1987, 19, 873. 4. 5. THE GENERAL ELECTRIC COMPANY LIMITED Gillis, J., Hoste, J., and Claeys, A,, Anal. Chim. Acta, 1947, 1 , 302. Dennis, L. M., and Johnson, E. B., J . Amer. Chem. Soc., 1923, 45, 1380. RESEARCH LABORATORIES WEMBLEY, MIDDLESEX and SIR JOHN CASS COLLEGE JEWRY STREET LONDON, E.C.3 Part 111.Determination in Flue Dust, Coal and Coke BY H. J. CLULEY The author's absorptiometric method (p. 523) for the determination of germanium with phenylfluorone has been applied to the analysis of flue dusts. The high sensitivity of the absorptiometric procedure permits the use of samples weighing 0.1 g or less and this greatly facilitates the decom- position of the sample by sodium carbonate fusion. Duplicate determinations can be completed in 3 to 3+ hours. The absorptiometric procedure has also been applied to the determination of germanium in coal and coke. In one method the sample is decomposed by a procedure similar to the Eschka method for the determination of sulphur in coal. I n a second method the sample is decomposed by combustion in a bomb calorimeter.Samples of coal and coke examined by these methods were found to contain 7 to 12 parts of germanium per million. THE available information on the determination of germanium in flue dusts and similar materials is somewhat meagre. One method for flue dusts that has been used at the Chemical Research Laboratory1 consists in fusion of the sample with sodium hydroxide, separation of the germanium from the majority of constituents, by precipitation as sulphide and finally gravimetric determination with tannin. Apart from being time-consuming, one apparentSept., 19511 PART III. DETERMISATION IN FLUE DUST, COAL AKD COKE 531 disadvantage of this procedure appears to be that for low germanium contents it may be desirable to carry out the fusion on several grams of sample.Alimarin and his co-workers have published a number of procedures for the similar problem of determining germanium in coal ash. These methods involve decomposition of the sample by treatment with hydrofluoric and sulphuric acid^,^$^ by fusion with sodium peroxide2p4 or by fusion with sodium carbonate and sulphur5; the germanium is then either separated by distillation of the tetrachloride and determined by precipitation as sulphide,2J or separated as sulphide and determined with tannin.415 Of these methods those involving a hydrofluoric acid attack appear open to criticism owing to the risk of loss of volatile germanium tetrafluoride. The methods involving fusion with an alkaline flux and final determination with tannin are similar to the method outlined above.The fact that Alimarin and his colleagues have within a short time proposed four successive methods might suggest that the earlier procedures were not wholly satisfactory. The previous work of the author on the absorptiometric determination of germanium with phenylfluorone (p. 523) indicated that this procedure might be applied with advantage to flue dusts. It had already been established that the procedure could readily be applied to germanium in the presence of substantial proportions of the large number of elements normally encountered in flue dusts. The high sensitivity of the method showed that a maximum sample weight of 0.1 g would be required and the use of such small weights should greatly facilitate the initial decomposition by fusion. The speed of the absorptiometric procedure also gave promise of a rapid method.This paper describes the application of the absorptiometric determination of germanium with phenylfluorone to flue dusts and subsequently to coal and coke. THE DETERMINATION OF GERMANIUM IN FLUE DUST Two flue dusts, samples 90 and 44, of different origins and of contrasting types, were selected for the initial experiments. In sample 44 most of the germanium was in a readily accessible form, as shown by the relatively high yield of germanium resulting from direct treatment with hydrochloric acid, whereas sample 90 had yielded only a small proportion of its germanium with this treatment. The germanium contents of these two samples were determined in the following manner. A 0.1-g portion of the ground sample was fused with 1 g of sodium carbonate in a platinum crucible, the melt being finally heated at 1000" C for 15 minutes. The melt was disintegrated with water and the mixture was transferred to a 100-ml distillation flask, neutralised with hydrochloric acid and diluted to 25 ml.A volume of 25 ml of hydrochloric acid was then added, and the distillation and the subsequent absorptiometric determination on an aliquot of the distillate were carried out as previously described.6 To investigate the possibility of loss of germanium by volatilisation during the fusion, further determinations were carried out in a similar manner but fusing at 1200" C, a treatment calculated to aggravate any such loss. TABLE I DETERMIXATIONS OF GERMANIUM IN FLUE DUSTS : COMPARISON OF PHENYLFLUORONE METHOD WITH OTHER METHODS Spectrographic Germanium estimate of Sample Method* found, Mean, germanium, Y O % Y O 0.69, 0.72 0.70, 0.70 B 0.65, 0.67 -4, 0.92, 0.94 44 A2 0.93, 0.94 B 0.85, 0.87 4 90 A2 0.71 0.70 0.8 0.66 0.93 0.94 1.0 0.86 * The methods of determination mere as follows- Method A,-Sodium carbonate fusion a t 1000" C, separation of germanium by hydrochloric acid Method A,-As A,, but fusion a t 1200" C.Method B-Sodium hydroxide fusion, separation of germanium as sulphide, final gravimetric deter- distillation, final absorptiometric determination with phenylfluorone. mination with tannin.532 CLULEY: THE DETERMINATION OF GERITASIUM [voi. 76 Complete solution was always obtained prior to the distillation; this showed that the methods of carbonate fusion as used resulted in complete decomposition of the samples, The results, shown in Table I, methods A, and ,Az, indicated good agreement between the individual determinations on each sample, and there was no evidence of loss of germanium during the fusion.As an independent chemical check on the results obtained absorptiometrically, deter- minations were also made by a procedure similar to the Chemical Research Laboratory method previously outlined. The results so obtained are also shown in Table I, method B. In addition, spectrographic estimations, based 011 comparison with synthetically prepared standards, were made on the two samples. Taking into consideration the difficulties of determining a minor constituent in such a complex material as flue dust, the agreement between the two chemical methods is reasonable. The results of both chemical methods are in approximate agreement with the spectrographic data, for which the expected accuracy would be 20.1 per cent.of germanium. The over-all mean of the chemical results for the two samples, 90 and 44, are 0.68 and 0.91 per cent. respectively, and when the completely independent chemical nature of the two chemical methods is borne in mind, it must be considered Ihat these values are very near to the true germanium contents of the two samples. Further trials of the method were made with samples of flue dust chosen to cover a range of germanium contents and to represent a number of different sources of the material. For these determinations fusion at the full heat of a Meker burner was used and the complete solution attained in all determinations proved thzt this method of decomposition was satis- factory. The results of the further trials are recorded in Table 11, which, for completeness, also contains the results obtained earlier for samples 90 and 44. It will be seen that the precision of the method is of a high order and that for each sample the value for the germanium content is consistent with the spectrographic result based on comparison with synthetic standards.TABLE I1 ABSORPTIOMETRIC DETERMINATIONS OF GERMANIUM IN FLUE DUSTS Sample 61 38 60 65 90 55 44 43 Germanium found, % 0.021, 0.021 0.14, 0.145 0.235, 0.23 0.30, 0.305 0.69, 0.72, 0.70, 0.70 0.825, 0.81 0.93, 0.94, 0.92, 0.94 1.15, 1-15 Mean, % 0,021 0.14 0.23 0.30 0.70 0.82 0.93 1.15 Spectrographic estimate of germanium, negligible % 0.1 0.2 0.3 0.8 0.7 1.0 > 1.0 Comparison with the tannin method shows the absorptiometric procedure to have a number of advantages.These include the ability to use small weights of sample, which greatly facilitates the decomposition by fusion an'd permits a more elegant procedure. The high sensitivity of the phenylfluorone reaction renders the absorptiometric method particularly suitable for samples containing small amounts of germanium; for example, 0.01 per cent. of germanium can readily be determined on 0.1 g of sample. The time required for duplicate determinations is 3 to 34 hours, compared with 13 to 2 days for the tannin method. The length of this latter method is largely due to the necessity for allowing the sulphide precipitate to stand overnight and to the lengthy low temperature ignition and nitric acid treatment of the tannin precipitate.RECOMMENDED METHOD FOR THE DETERMINATION OF GERMANIUM I N FLUE DUST APPARATUS AND REAGENTS- The distillation apparatus and the special reagents required for the final absorptiometric determination are described on p. 528 of the preceding paper on the absorptiometric deter- mination of germanium with phenylfluorone.Sept., 19511 PART 111. DETERhIINATIOh' 1 s FLUE DUST, COAL AKD COKE PROCEDURE- 533 Weigh 0.01 to 0.1 g of the ground sample, depending on expected germanium content, into a platinum crucible. Mix the sample intimately in the crucible with 0.5g of sodium carbonate and cover the charge with a further 0 6 g of sodium carbonate.Heat the crucible gently over a Meker burner, increasing the heat gradually over a period of about 5 minutes until a temperature of about 1000" C is reached. Continue heating in this manner, with occasional swirling of the melt, for a further 15 minutes. To the cool melt add about 10 ml of hot water and allow the crucible to stand on the edge of a hot-plate until the melt is completely disintegrated. Cool the crucible in running water and then transfer the contents to the 100-ml distilling flask. Add to the crucible 4 ml of diluted hydrochloric acid (1 $- l), swirl to dissolve any solid particles still adhering to the crucible and then transfer the acid solution to the main solution contained in the flask.Swirl the flask to promote the liberation of carbon dioxide and dilute the solution to 25 ml as indicated by an appropriate line previously marked on the flask. Add 25 ml of hydrochloric acid, sp.gr. 1.18, and carry out the distillation and the subsequent absorptio- metric determination on a suitable aliquot of distillate as described on p. 529 of the preceding paper,6 using a blank taken through the method as the reference solution. XOTE- that the prepared sample is representative and homogeneous. owing to the small weight of the sample. Owing to the heterogeneous nature of flue dusts, precautions must be taken to ensure This is particularly important THE DETERMINATIOK OF GERMANIUM IN C09L AND COKE The facility with which germanium in flue dust can be determined by the above method encouraged an attempt to apply the phenylfluorone procedure to the determination of ger- manium in coal and coke.The germanium content of so-called "germaniferous" coal appears to be of the order of 10 parts per million, and other workers7~* have ashed the coal to effect a concentration of the germanium before determination. I t has been stated, however, that unless considerable care is taken to maintain strongly oxidising conditions during the ashing, substantial loss of germanium by volatilisation may It was considered that the sensitivity of the phenylfluorone reaction might permit a direct determination on the coal or coke without recourse to preliminary ashing. Initial experiments were carried out on a sample of Boldon coal and on a sample of coke from this coal.The procedure used for the decomposition of the sample was similar to that used in the Eschka method for determination of sulphur. Sodium carbonate alone was used in place of the usual admixture with magnesium oxide, as it was required to fuse the residue remaining after destruction of the carbonaceous matter. This destruction was carried out at 600" C instead of the customary 800" C in order to prevent loss by volatilisation of germanous oxide, which is stated to be volatile at 700" C. The determinations were completed by distillation and absorptiometric measurement as used for the determination of germanium in flue dusts. As a check on the results of these determinations a different method of decomposition was used. Samples were decomposed in a bomb calorimeter by precisely the same method a s used for sulphur determinations, and the contents of the bomb were washed into a platinum dish.Sodium carbonate was added to neutralise the acids formed during the combustion and the solution was evaporated to dryness. The residue was then fused with a further quantity of sodium carbonate and the determination completed by distillation and absorptio- metric measurement. The results of determinations by these two procedures are shown in Table 111, which also includes results by the sodium carbonate - ignition method on further samples of coal and coke. Although the combustion in the bomb should reduce the risk of loss of germanium by volatilisation to a minimum, it will be observed that the results so obtained are slightly lower than those of the sodium carbonate - ignition method.This difference is attributed to some loss of ash incurred by fusion into the silica crucible used to contain the sample for combustion. I t is likely that this loss could be avoided by using a platinum container534 CLULEY: THE DETERMINATION O F GERMAXIUM [l'ol 76 if work of higher accuracy were desired. It is interesting to note that spectrographic examina- tion of the four samples, made before the chemical determinations, showed that the most sensitive lines of germanium could only just be tdetected, and from this it was concluded that the germanium content of the samples was approximately of the order of 10 parts per million. TABLE I11 DETERMINlTIONS OF GERMANIlJM I N COAL AND COKE Sodium carbonate - ignition method Bomb method r-y---------7 7 Germanium found, Mean, Germanium found, Mean, p.p.m.P P.m. p.p.m. p.p.m. Boldon coal . . .. . . 7.0, 8.0 7.5 6.6, 6.6 6.5 Coke from Boldon coal. . . . 7.0, 8.0 1.5 6.1, 6.4 6 Harton coal .. . . . . 7.2, 9.2 8 Coke from Harton coal . . 11.0, 12.5 12 ,4 comparison of the two procedures shows that the ignition method requires considerably less manipulation and has the advantage that a number of samples can be decomposed simultaneously. However, the bomb method has a greater sensitivity as it uses a 1-g sample in contrast to a maximum of 0.5 g for the ignition method. The limitation of the size of the sample in the latter method is associated with the quantity of sodium carbonate required, as large concentrations of sodium salts result in precipitation of sodium chloride from the strong hydrochloric acid solution used for the distillation. During these experiments it was found that the effective sensitivity of both methods could be enhanced by collecting a smaller volume of distillate and applying the absorptio- metric procedure directly to this without subdivision. The loss of germanium resulting from this modification was found to be negligible for the very small amounts of germanium involved.In this manner it should be possible to determine germanium down to 4 p.p.m. with the ignition method and 2 p.p.m. with the bomb method. Although complete data are not available for an accurate assessment of the efficiency with which germanium is deposited in flue dusts, it may be of interest to record the information resulting from the above determinations.This shows that on conversion of the Boldon coal to coke, some 60 per cent. of the germanium was retained by the coke; on use of this coke in producers, a much smaller proportion of germanium was deposited in the flue dusts in the waste heat system, although these dusts contained sufficient to justify extraction of the element. - - - - THE TWO METHODS FOR THE DETERMINATION OF GERMANIUM I N COAL AKD COKE APPARATUS AND SPECIAL REAGENTS- The distillation apparatus and the special reagents required for the final absorptiometric determination are described in the preceding paper on the absorptiometric determination of germanium with phenylfluorone.6 The bomb calorimeter required for the bomb method is of the conventional type used for the determinations of calorific value and sulphur in coal and coke.SODIUM CARBONATE - IGNITION METHOD- Weigh into a platinum crucible 0.5 g of the sample previously ground to pass a 120 B.S. screen. Mix the sample intimately in the crucible with 1.5 g of sodium carbonate (do not tap down) and cover the charge with a further 0.5 g of sodium carbonate. Transfer the uncovered crucible to a muffle furnace and, allowing access of air, raise the temperature to 600" C over a period of about an hour. Maintain this temperature for 14 hours, or longer if carbonaceous matter is still visible. Remove the crucible from the furnace, stir the contents and return the crucible to the furnace for a further hour to destroy any last traces of carbonaceous matter.Fuse the contents of the crucible, disintegrate the melt and prepare the solution for distillation in the same manner as used in the method for flue dusts, except that the volume of diluted hydrochloric acid (1 + 1) used should be 8 ml to neutralise the larger weight of sodium carbonate used.Sept., 19511 PART 111. DETERMINATIOK I S FLUE DUST, COAL ASD COKE 535 Distil at the rate of about 2 ml per minute, collecting only the first 10 ml of distillate. Apply the absorptiometric procedure directly to this distillate without subdivision and to a blank distillate similarly prepared as described in the preceding paper.6 The 10ml of distillate contains the appropriate amount of hydrochloric acid for the absorptiometric determination and no further addition of acid is required.BOMB METHOD- Place 10 ml of water in the bomb and carry out the combustion of the pellet in the normal manner, with an oxygen pressure of 25 atmospheres. After the combustion allow the bomb to stand for 30 minutes to permit the acid mist to settle and then slowly release the pressure. Wash the solution and ash into a platinum dish, add 0.2g of sodium carbonate to make the solution alkaline and then evaporate to dryness. Add 1 g of sodium carbonate, fuse, disintegrate the melt and prepare the solution for distillation exactly as in the method for flue dusts. Perform the distillation and the final absorptiometric determination exactly as in the ignition method. REFERENCES Make about 1 g of the ground sample into a pellet and weigh it.1. 2. 3. 4. Alimarin, I. P., and Aleksewa, 0. A., Ibid., 1940, 13, 1393. 5. 6. 7. 8. THE GENERAL ELECTRIC COMPANY LIMITED “Chemical Research, 1938-46,” H.M. Stationery Office, London, p. 30. Alimarin, I. P., Ivanov-Emin, B. N., Medvedeva, 0. A., and Yanovskaya, C. Y., Zauod. Lab., Alimarin, I. P., and Ivanov-Emin, B. N., J . A@$. Chem., U.S.S.R., 1940, 13, 951. Alimarin, I. P., Trudy Vsesoyuz Konferentsii Anal. Khim., 1943, 2, 371. Cluley, H. J,, Analyst, 1951, 76, 523. Goldschmidt, V. M., J . Chem. SOC., 1937, 655. Morgan, G. T., and Davies, G. R., Chem. and Ind., 1937, 16, 717. 1940, 9, 271. RESEARCH LABORATORIES WEMBLEY, MIDDLESEX and SIR JOHN CASS COLLEGE JEWRY STREET LONDON, E.C.3 DISCUSSIOK ON THE ABOVE THREE PAPERS MR. W.H. BENNETT asked whether fluorine interfered with the phenylfluorone colour reaction, and if it did, had the author considered the possibility of the interference of fluorine in the determination of germanium in coals,. some of which contained small amounts of fluorine. MR. CLULEY replied that he had no specific information on the effect of fluorine on the colour reaction, (Later work has shown that as much as 2.5 mg of fluorine may be present as fluoride in the final solution without effect on the colour reaction, although substantially greater amounts of fluorine cause interference by reducing the intensity of the germanium colour, presumably owing to partial complexing of the germanium. When the maximum sample weight of 1 g of coal was used, the amount of fluorine in the coal would not exceed 0.2 mg, so that no interference from this source would be expected.) MR.R. F. MILTON drew attention to the author’s statement that reduction of germanomolybdate to molybdenum blue was not satisfactory in the presence of arsenic, phosphorus or silicon. In point of fact, each one of this group of elements could be estimated in the presence qf the others, if cognisance were taken of the conditions under which the complex molybdate formation occurred. Thus, silicomolybdate formed in 0.1 N sulphuric acid and the reduction occurred in 2 N sulphuric acid with stannous chloride. The molybdenum blue could be extracted with butyl solvent and estimated. If the residual solution were neutralised to 1.2 N sulphuric acid, phosphomolybdate would reduce and the colour could then be extracted and estimated. The reduction of acidity to 0.9 N would allow germanomolybdate to reduce and to be estimated. If arsenic, which had been maintained in the tervalent state, were then oxidised, it would form the arsenomolybdate, which could then be estimated after reduction with stannous chloride. He had two questions to put to the author. First, he asked if it was necessary in the determination of germanium in flue dust to make a fusion of the material before the hydrochloric acid distillation. In his experience, when arsenic and germanium were present together, the germanium usually distilled very readily with the arsenic when heated with hydrochloric acid. Was it possible to distil the flue dust directly with hydrochloric acid? Secondly, he asked whether the author could give any idea of the sensitivity of the colour reaction DR. J. H. HAMENCE said he had listened t o the papers with great interest.DETERMIXATION OF THE GRADE STRENGTH OF PECTINS [Vol. 76 536 between phenylfluorone and molybdenum. Detection of traces of molybdenum had now assumed con- siderable importance in agricultural work, and new sensitive tests for this element were always welcome. Finally, Dr. Hamence thanked the author for having shown in a most conclusive manner that the distillation process for the separation of arsenic from antimony and tin was in fact a most efficient method, and that neither of these elements distilled over with the arsenic. In the past this had always been assumed to be a fact, but to the best of his knowledge it had not previously been confirmed experimentally. MR. CLULEY replied that although some flue dusts gave an almost quantitative yield of the germanium on direct distillation with hydrochloric acid, other dusts afforded only a small proportion of their germanium with this treatment. In the latter type of dust the germanium was presumably mainly present in a chemical form in which it was not readily soluble in hydrochloric acid. Decomposition of the sample by fusion was therefore necessary to ensure complete recovery of the germanium in the determination. The sensitivity of the colour reaction of phenylfluorone with molybdenum was of the same high order as the germanium reaction, and phenylfluorone could probably be used for the detection of molybdenum if required. Gillis, Claeys and Hostel have stated that the associated reagent o-hydroxyphenylfluorone is, under appropriate conditions, a specific reagent for the detection of molybdenum, REFERENCE TO DISCUSSION 1. Gillis, J , , Claeys, A,, and Hoste, J., Anal. Chim. Acta, 1947, 1, 421.

ISSN:0003-2654    

DOI:10.1039/AN9517600517

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

536 DETERMINATION OF THE GRADE STRENGTH OF PECTINS [Vol. 76 Determination of the Grade Strength of Pectins REPORT OF THE PECTIN SUB-COMMITTEE OF THE JAM PANEL, BRITISH FOOD MANUFACTURING INDUSTRIES RESEARCH ASSOCIATION THE Jam Panel of the British Food Manufacturing Industries Research Association in May, 1948, appointed a Sub-committee with the terms of reference- “To devise standard methods for the determination of the grade strength of pectins as commercially supplied for jam manufacture.” The members of the Sub-committee were Miss M. Olliver (Chivers and Sons Ltd.), Messrs. L. M. Adams (Scott, Preserve Makers Ltd.), A. C. Francis (Rowntree and Co. Ltd.), H. Threadgold (Wm. P. Hartley Ltd.) and C. L. Hinton (B.F.M.I.R.A.), with the late Mr. T. Rendle (Chivers and Sons Ltd.) as Chairman.A considerable amount of experimental work on the subject was carried out by members of the Sub-committee, which also had the benefit of hearing detailed results of work since published by Miss M. Olliver.192 The following is the Report of the Sub-committee that was formally adopted by the Jam Panel of the Research Association in May, 1950. The system of expressing the jellying value of commercial pectins by means of a grade number was introduced in the U.S.A. SO far as can be ascertained, no precise definition of “grade” has been laid down by any authoritative body, though a Committee of the American Institute of Food Technologists is at present studying the question. From statements that have appeared in the literature, however, it appears that the grade of a pectin is generally understood to be the weight of sugar with which one part by weight of the pectin will, under suitable conditions, form a satisfactory jelly.The conditions implied require a soluble solids content of 65 per cent. in the jelly, but apart from that, the conditions for the com- position and preparation of the jelly are very imprecise. There has been a further lack of precision on what is to be regarded as a jelly of satisfactory firmness. The Sub-committee holds very strongly that the conditions under which the grading of pectins in the United Kingdom is determined should be related as closely as possible to the conditions under which the pectin is to be commercially used in jam manufacture in this country. In the proposed method of this Report an attempt has been made to implement this view as far as possible, though it is appreciated that the composition of different jams in so far as this affects the jellying behaviour of the pectin, is somewhat variable.The conditions of composition of the standard jelly have been chosen, after a survey of typical jams on the market, as conforming most closely with the composition of the majority of commercial jams in this respect. At the same time, it is the opinion of the Sub-committee that the recommended method should as far as possible, for commercial reasons, reproduce US. grade values, provided thatSept., 19511 DETERMINATION OF THE GRADE STRENGTH OF PECTINS 537 they can be brought into conformity with the conditions of use of pectins in British manu- facture. It appears that if a rather weak jelly is accepted as the standard of jelly strength to satisfy the U.S.grade definition, then the results of grading by the method proposed by this Sub-committee fall into line on the whole with U.S. grade figures (see Appendix, p. 539). For the purpose of the practical determination of jelly strength, however, a rather stronger jelly is preferred, prepared from a slightly larger proportion of pectin. The grade value found, however, is fundamentally referred to the definition of grade that was adopted by the Sub-committee, as follows- “The grade of a pectin is defined as the ratio of total soluble solids to pectin in a jelly of standard strength prepared in a standard manner with total soluble solids between 70 and 71 per cent.” The Sub-committee fully appreciates that there is much in the behaviour of pectins of various kinds that is not yet known or understood and that may affect their value in practical use in a sense not indicated by the proposed standard test.The Sub-committee does not put forward the test as finally satisfactory in all its details, and in fact anticipates that it may be capable of improvement as further knowledge is gained. For the measurement of jelly strength, the Sub-committee recommends for the present two instruments as standard, viz., the “B.A.R.” Jelly Tester3 and the California Exchange Ridgeli~neter.~ Measurements with other jelly-testing instruments could be accepted when data are available to relate their readings of the strength of the standard jelly to those obtained with the two instruments mentioned, I t should be pointed out, however, that these two instruments essentially measure the elastic property of the jelly.Some other instruments, such as those that measure the breaking strain of the jelly, while perhaps giving valuable information on other properties of the jelly, may not be so suitable for use in measuring the jelly strength for the purpose of the proposed test. This, and other aspects of the standardisation of pectin quality, require further and continued investigation, and it is the view of the Sub-committee that it, or some similar body, should be kept in being by the Panel for the purpose of making any recommendations that may be desirable from time to time. April 27th, 1950.Provisional Standard Method for Determination of the Grade Strength of Pectins 1. PREPARATION OF A SOLUTION OF THE PECTIN- (a) The pectin, if powdered, is made up with sugar syrup, on the basis of its alleged grade strength, to give a solution corresponding to a 5-grade liquid pectin, as follows- (b) Into 75 g of cold 66 to 67 per cent. sugar syrup stir 15 g of 100-grade pectin, or an equivalent amount of any other grade, and distribute evenly. Then stir in 200ml of hot distilled water, cool, and adjust the weight to 300 g, mixing well, and ensuring that no lumps remain undispersed. Allow to stand for at least 1 hour, and before using. stir again to break up any lumps that are still visible. Cover the vessel during the standing period, otherwise a skin may form that is not easy to dissolve.(c) Dispersion of the pectin must be complete and it is recommended that if possible examination under the microscope should be made to confirm that no particles remain. Should the measures recommended in 1 (b) be inadequate, more drastic methods may be adopted, such as the use of hot syrup or mechanical stirring or both for the initial dis- persion. (d) A larger or smaller quantity of pectin solution, with ingredients in the same proportions, may be prepared if desired, e.g., to enable duplicate or repeat tests to be made. (e) Liquid pectin usually has a grade strength of about 5 and should be used without dilution. At no stage should the pectin solution be boiled. 2. PREPARATION OF THE BUFFER SOLUTION- 220 ml of N potassium carbonate.This solution, undiluted, should have a pH of 2.82. (a) Dissolve 100 g of pure monohydrate citric acid in 600 ml of distilled water and add Roil to remove carbon dioxide, cool, and dilute to 1 litre.538 DETERMINATION OF THE GRADE STRENGTH OF PECTINS [Vol. 76 (b) I n testing rapid-set pectins with a potassium buffer it is sometimes found that air entangled in the boiling jelly does not escape before setting occurs. In such circum- stances it is permissible to use a buffer prepared as in 2 (a) but with N sodium carbonate in place of the potassium carbonate provided that the grade is ultimately determined from a calibration graph (see 6 ( b ) ) obtained with jellies prepared with sodium buffers in the same way. (c) If a stock of buffer solution is kept for any length of time, thymol may be added as a preservative unless the jellies are to be subsequently used for flavour tests.3. SUGAR- A high grade refined sugar (sucrose) should be used for the test. 4. PREPARATION OF THE JELLY- (a) Weigh 926 g of the pectin solution or of the liquid pectin sample into a beaker and add 200 to 300ml of distilled water, mixing well. Determine the pH by means of a glass electrode immersed in the liquid. If this is not in the range of 3.05 to 3.1 (or lower if sulphur dioxide is present), add with stirring 0.1 N or N hydrochloric acid or 0.1 N sodium hydroxide as required, so that the necessary pH range is reached. Remove the glass electrode and stirrer, washing them thoroughly with a jet of distilled water. Transfer the contents of the beaker to a weighed copper, aluminium or stainless steel boiling pan of about 8 litres capacity, washing out the beaker well with distilled water.Make up to such a weight with distilled water that the time taken to boil down to the required 600g of jelly (see below) will be approximately 10 minutes. The weight after addition of this distilled water should not be less than 500 g and may be higher according to size and shape of pan and type of gas-ring used. To the mixture then add 25 ml of the buffer solution, and 417 g of sugar (minus the solids in 92-5 g of pectin solution). Heat the mixture over an efficient gas-ring, with constant stirring by means of a wooden spoon or paddle or thick glass rod. Boil until the weight of the mixture is reduced to 600 g in approximately 10 minutes' actual boiling time.Check the finish of the boiling as this is approached by lifting the pan on to a scale and weighing to within 1 g. Pour the boiled mixture into a square box (3 x 3 x 3& inches deep) if a B.A.R. Jelly Tester is to be used, or into two standard glasses fitted with paper collars for the Exchange Ridgelimeter test. Cover the surface of the jellies at once with waxed paper or, for jellies for the B.A.R. test, with a thin layei- of high-grade liquid paraffin. (b) A larger quantity of jelly may be prepared, e g . , sufficient to fill three boxes for the B.A.R. Jelly Tester or the corresponding number of glasses for the Ridgelimeter, with amounts of pectin solution, buffer solution, and sugar in the same proportions as in 4 (a) with respect to the weight of finished jelly.The amount of water added should be adjusted so that the boiling time is approximately 10 minutes. (c) The amount of sugar used is calculated to allow for the slight increase in weight from inversion. Should the percentage of soluble solids in the jelly when finally tested not fall within the range 70 to 71 per cent., the amount of sugar used in subsequent tests should be suitably adjusted. 5. COOLING, STORAGE ,4ND TESTING OF THE JELLY- (a) Allow the vessel containing the jelly to remain on the bench in air at room tempera- ture (but as near to 20" C as possible) for 2 hours. Then place it in an incubator at 30" C for 22 hours. After removing the jelly from the incubator take off the wax paper, or remove the paraffin, and test the jelly strength at once without first cooling.After testing, determine the soluble solids content of the jelly by refractometer (this should be 70.5 per cent. & 0.5 per cent.) and the pH of a 50 per cent. w/v solution of the jelly (this should be 3.10 4 0.05). (b) If the soluble solids content is not within the specified range, repeat the preparation of the jelly after suitably adjusting the amount of sugar as mentioned under 4 (c). (c) If the pH is not within the specified limits, repeat the preparation of the jelly after suitably adjusting the pH of the buffer solution to the extent by which the pH of the jelly is either low or high.Sept., 19511 DETERMINATION OF THE GRADE STRENGTH OF PECTINS 539 (d) The B.A.R. Jelly Tester used in testing the jelly should have a pulley with a diameter (inside the grooves) of 4-32 cm 0-06 cm, and a square vane with a side of 2 cm & 0.01 cm, and diagonals of 2.83 cm rt_ 0.02 cm.(e) The Ridgelimeter glasses. used in testing the jelly should conform with the specification given by Cox and Higby4, vix., Hazel Atlas No. 85 glasses with a depth of 3.125 inches a t the centre. 6. CALCULATION OF THE GRADE STRENGTH- (a) If the jelly strength obtained is 16 ml with the B.A.R. Jelly Tester (after deducting the blank), or 24.4 per cent. sag with the Exchange Ridgelimeter, calculate the grade strength of the pectin by the formula- Grade strength = 77 Percentage of pectin sample in jelly’ (b) If the jelly strength obtained is not as specified in 6 (a), but lies between 10 and 22 ml with the B.A.R.Jelly Tester or 28.6 per cent. and 21.5 per cent. sag with the Exchange Ridgelimeter, calculate the approximate grade strength from a calibration graph obtained as follows from jellies prepared from varying percentages of a standard pectin whose grade strength satisfies the conditions in 6 (a)- From this solution, using amounts from 80 to, say, 115 g, prepare a number of jellies as in 4 (a), and test as described in 5 (a). Plot the jelly strengths obtained with the B.A.R. Jelly Tester or the percentages of sag obtained with the Exchange Ridgelimeter against the percentages of 100-grade (or equivalent 100-grade) pectin in the jellies. To calculate the grade strength of the sample under test, take the reading obtained with it as under 5 (a) and from the graph find the percentages of 100-grade pectin corresponding to the reading. Suppose it is x per cent.and that the jelly prepared as in 4 contained p per cent. of the sample. Then the approximate grade strength of the latter is given by- Prepare a 5-grade solution of the standard pectin as directed in 1 (b). 100 x x Approx. grade strength = - P ‘ (c) To obtain a more accurate value for the grade strength, prepare a second jelly with the weight of pectin solution adjusted to correspond with the approximate grade strength found from the graph. If this second jelly has a strength equal to the required standard as in 6 (a), calculate the grade strength from the amount of pectin used, by the formula in 6 (a). If the jelly has a strength only slightly different from the required standard, calculate the grade strength from the calibration graph and the amount of pectin used, as in 6 (b).APPENDIX RELATION TO THE U.S. SYSTEM OF DEFINING GRADE STRENGTH The grade strength of a pectin is defined as the ratio of total soluble solids to pectin in a jelly of standard composition and strength prepared by a standard procedure. The standard procedure and composition for the British standardisation have been defined in the foregoing description. For the purpose of the definition it is also necessary to define a standard strength for the jell!.. I t has been found that when commercial pectins of specific U.S. grade strength, as tested under the rather different conditions of accepted U.S. practice, are tested under the conditions laid down above, with the percentage of pectin in the jelly equal to- 70-5 grade strength of pectin’ (k., the jelly corresponds strictly to the grade strength definition given above), the strength of the jelly when tested with the B.A.R.Tester averages approximately 12 ml (blank deducted) or approximately 27 per cent. sag with the Ridgelimeter. Such a jelly, in the opinion of the Sub-committee, is rather too weak to be taken as a standard jelly. I t is, however, desirable to assume that it satisfies the “standard strength” of the definition in order not to conflict with accepted U.S. grading practice. For the proposed standard test, a rather firmer jelly is prescribed, having a strength of 16 ml with the B.A.R. Jelly Tester or 24.4 per cent. sag540 CUTHBERTSON, PEGLER, QUADLING AND HERBERT [Vol. 76 with the Exchange Ridgelimeter. It has been found that to produce a jelly of this strength, the pectin must be used in approximately 12/11 times the concentration required for the “standard strength.” Hence the grade strength of a pectin is given by the proposed standard method as follows- 70.5 Grade strength = -~ 11 (percentage of pectin in jelly) 12 77 (percentage of pectin in jelly)’ - - as in 6 (a). This Report is published by permission of the Department of Scientific and Industrial Research and of the Council of the British Food Manufacturing Industries Research Association. REFERENCES 1. Olliver, M., Food Technol., 1950, 4, 370. 2. 3. 4. - , J . Sci. Food and Agric., 1950, 1, 329. Campbell, L. E., J. SOC. Chem. Ind., 1938, 57, 413. Cox, R. E., and Higby, R. H., Food Ind., 1944, 16, 441, 505. LABORATORIES OF THE BRITISH FOOD MANUFACTURING INDUSTRIES RESEARCH ASSOCIATION RANDALLS ROAD, LEATHERHEAD, SURREY Director of Research: L. E. CAMPBELL February, 1951

ISSN:0003-2654    

DOI:10.1039/AN9517600536

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

540 CUTHBERTSON, PEGLER, QUADLING AND HERBERT [Vol. 76 The Assay of Vitamin Part v Some Substances that Interfere with the B12 Response of an Escherichia coli Mutant to Vitamin B12 BY W. F. J. CUTHBERTSON, H. F. PEGLER, C. QUADLING AND VALERIE HERBERT (Presented at the meeting of the Biological Methods Group on Tuesday, December 19th, 1950) The effects of various substances on the Escherichia coli plate assay of vitamin B,, have been investigated and permissible limits have been defined. For satisfactory assay C must be less than KN, where C is the maximum permissible concentration of the interfering substance expressed as a per- centage, K is a constant depending on the nature of the interfering substance and N is the concentration of vitamin B,, in pg per ml. Some reagents increase while others decrease the apparent vitamin-B,, response.IN a preliminary communication1 from these laboratories, it was shown that a mutant of Escherichia coli could be used as an assay organism for the measurement of vitamin B12. A more detailed description of the precise procedure recommended is in course of preparation ; meanwhile it may be of assistance to others working in this field if we give an account of some essential preliminary work, for without its completion a routine method could not have been established. This work was concerned with the effect of various reagents on the response of the mutant (which we call variant M 200) when used in the cup-plate assay for vitamin BI2. The reagents tested were chosen for a number of reasons; ( a ) they mightSept., 19511 THE ASSAY OF VITAMIS Bit.PART V 541 TABLE I EFFECT OF INTERFERING AGENTS ON RESPONSE TO VITAMIN B,, Solutions containing 0.2 pg of vitamin B,, per ml Response decreased Added Apparent ethylene glycol, vitamin B,, content, A I \ % v/v Pg Per 0.0 0.3 25 0.168 50 0.145 75 0.133 Response increased CuS0,.5H20, vitamin B,, content, I h 7 Added Apparent PQ Per ml 0-2 %&+ 0.01 0-226 0.05 0-236 1.0 0.258 be used as preservatives for vitamin B,, concentrates, (b) they might be used in the preparation of extracts, and (c) they might be expected to replace vitamin B,, in the nutrition of the test organism. Different concentrations of the reagents studied were placed on the assay plates in absence or presence of the vitamin. When the reagents were tested in its presence the concentration of the vitamin used was always kept at 0.2 pg per ml, Le., the same as that TABLE I1 Substance VALUES OF CONSTANT, Qualitative action A r In absence of In presence of vitamin B,, vitamin BIZ Response Alcohol slight reduces response reduced Acetone nil 9 9 Ethylene glycol nil 9 9 Propylene glycol nil 3, Ascorbic acid inhibits at 5% 9 , Formalin, 400,6 inhibits at 0.2y0.91 v/\- Values of K for \. 2 per cent. error error 5 per cent. 5 12.5 0.2 0-5 1 2.5 0.00," 0.0053 0.5 1.0 10 30 Response Butanol unaltered Toluene at concentra- Sodium formate tions tested MnS0,.4H20 FeS0,.7Hz0 Choline Betaine nil nil nil below 0.204 stimulates at nil up to O.Olyo stimulates at stimulates at 0.5 9.6 ; inhibib above 1% nil up to 20% 0.003 to o*lyo 0.05 yo nil - - nil - - nil below 0.2% > 1 > 2 nil up to 1.0% > 5 > 10 nil up to 0.01% > 0.05 > 0.1 nil up to 0.5% > 2-5 > 5 nil up to 0.5% > 2.5 > 5 nil up to 20% > 100 > 100 Response Methionine stimulates causes faint very very increased outer growth variable, variable,.zones - 0.007 - 0.015 Thioglycollic acid inhibits above increases response 0.5 1.7 0.5% Phenol inhibits above 5% 9 9 CuS04.5H,0 stimulates at Sodium chloride no effect below 3, o*oo5~0 10 to 20% 0.25 0-5 0.005 0.015 10 20 t o 40 Effect variable Homocystine stimulates at variable 0.2 0-5 0.2:; ; inhibits a t 1 t o 2% 0.2 to 0.5 0.2 to 0.5 Potassium inhibition above inhibition of cyanide 0.1% growth Cyanide in concentrations above 0.1 per cent. causes inhibition and prevents growth over the whole plate.542 NOAKES: THE DETERMINATION OF SELENIUM [Vol.76 used in the provisional assay technique. In all experiments the growth zones were compared qualitatively with those caused by the vitamin at concentrations of 0.02 and 0.2 pg per ml on the same plate. If the test mixtures produced growth qualitatively similar to that obtained with the vitamin alone then the diameters of the growth zones were measured and the apparent concentrations of vitamin B,, in the test mixtures were calculated from them. In this way the effect of interfering agents on the apparent activity could be determined. In some instances the interfering agent caused an apparent decrease in response, in others an increase. Typical examples of each kind of effect are shown in Table I.From these and similar results it is possible by graphical interpolation to determine the permissible concentration of the different substances allowable for a specified error. In the assay procedure used all solutions are diluted to approximately 0.2 pg per ml before assay ; clearly, the concentration of interfering agent producing a given error increases with the concentration of the original vitamin B,, solutions. Thus the allowable concentration of interfering substances may be expressed in the form C = KN per cent. where C is the permitted concentration in the test solution expressed as a percentage, N is the vitamin B,, concentration (in pg per ml) and K is a constant depending on the interfering substance. In Table I1 are given the values of K at levels leading to a 2 per cent. or a 5 per cent. error. This table shows whether the interfering agents cause an increase or a decrease in the apparent vitamin-B,, activity of the solution; the effects of the agents in the absence of the vitamin are also shown. A number of amino-acids have been tested by Bessell and Lees, to determine whether they interfere in this assay. Histidine, arginine and isoleucine tested at a level of 5 mg per ml were found to promote growth corresponding to 0.002, 0.001 and less than 0.005 pg of vitamin B,, per ml, respectively; none of these substances at this concentration interfered with the assay of solutions containing more than 0-2 pg of vitamin B,, per ml. REFERENCES 1. 2. GREENFORD, MIDDLESEX Bessel, C. J., Harrison, E., and Lees, K. A., Chem. and Ind., 1950, 561. Bessel, C. J., and Lees, K. A., private communication. GLAXO LABORATORIES LTD.

ISSN:0003-2654    

DOI:10.1039/AN9517600540

出版商:RSC    

年代:1951

数据来源: RSC