ISSN:0003-2654    

DOI:10.1039/AN95479FX051

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479BX053

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479FP151

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

lil~:?llSI', D.Sr. or A.R.I.C.. wit11 mine ~ n ~ I v t i ? ~ i l ekpeii-crncr, reqi'ired ior research section of 1;ibor;ilory uf well-h r m v n I.ondon inaniif.ictiiring c o i q h ~ n y . The xork willinvolve developnient of new pro?lirts and pmceE'rs a n d theiiiiproveiiient of existing ones. lliere IS a good tenin spirit;Ind initiative and the acceptance of responaihility isonlance with qli.ililic;itions andj. 'l'irr: A X A L V ~ T , ii, (;rrsh:imStrret, I.ontlon, E.C.2.1x1 I'E Rl.11, CH El1 ICAL I X UVSTI<I 1 . 5 1.1 hllTl<l>,eiir I<edcar. ha5 v:Lc:InCIe) for hperiencednts for cheiiric.il analytic.il work on daystes ihoiild have a S.rtion:d Certificate inCheiiiistry or a Generd Certificate of l<rlocation \vith fivepasses ;it Ordinary Level, including Ilnglish, 91:ithb.arid aScirncr subject. Writc for nppliration for enipluyment lorinto the Stztli Ollirer. Iinperial Cheniiral lndiistrirs I.~mited,rh,, .\liddlesbrough. or .ipply to the ne.irestt 1:sch;inge. quotirig rtd\erli~eiiieiil referericeANALYST xi1;SEAKCH ASSISTASTS. Viscose l<.iyon factory iii R+ib .:in-Werriirgtoii .Area has vacancies fur 2 ItesearchAssistants. C)iialiiicatioiis-H.S.C. Chemistry or equivalent.Salary (wilh H.S.C. above age 3.i), 2l!l/- prr week. .4contrihtitory pension scheme 1s in operatioil. Good canteenand recreational facilities. Write Bon So. X i i , TME.ixni.vsr, . l i , Grcsh.iii! Street, I.ondon, li.C.3.ii period vi trvo ve'irs i n thc first instance. 'She personappointed will lie required to participate in a specld researchproject proreeding i n the Uepartriient. Applir.itions (threecnpxrs).shting chte of birth, qiialificationb .ind cxpcrienre,together w i t h thr rtainrs of three referees. shoiild reach theI<egiatr.ir. T h e t:iiiverstty, I.ccd~, P (from whoin furtherpartiriilxrs inay be obt.iineil). nut Liter th,rir :ioth Novernber,1!151.~ . A I < I ~ I ~ I ' I ~ I ~ I.I>LITED have a vxancy for an AnalyticalChenii\t 111 tlie Work5 ~lcclmic,tl Uepartmcnt a t theirUiriitin&?h.itii Works. Applicants shoiild h.ive :in lhonoiirsdegree and pi-eference will be given to tliose with PostGrarlu.itc traning in Analvtical Chernistrv. The peraonappointed will I,? reipon4le; rindrr the hc:irl.of t h e ChemicalInspection Section.for tlie miproveinent of test proceduresand t h r drveloprnen! o i new an,ilvtic:il techniqiies in theiield of \yrrtIictlr re4n prndiicts. Sqiia1lfic:itionb . i d <.sperience. Appqiidifications .itid experience trSupcrintcndcnt. lialirlite I-ini~ted. Redfern I<o;id \Vurks,T\-selev, Birnliilgh,uii, 1 1 .'AMBERLITE'ION EXCHANGE RESINSSTANDARDGRADESFORLABORATORYUSLot the analvticalIn addition to the speciallyprepared analytical grades,standard grades of theAmberlite resins are nowavailable for laboratory use.These are supplied in generalin the exhausted forms, butafter several cycles of re-generation, which free themfrom 'fines' and solublematerials, they assume asimilar condition to thatgrades as sumlied.The B.D.H. booklet 'ION 'EXCHANGERESINS' describes both ' Permutit' and'Anibcrlitc' resins. Copies of the booklet andprice lists may he ottainecl on request.T H E B R I T I S H D R U G H O U S E S L T D .B.D.H. LABORATORY CHEMICALS G R O U PP O O L E D O R S E TTelephone : Prole 962 ( 6 Irnes) Telegrams : Tetradome Poolefirmhll CIS4

ISSN:0003-2654    

DOI:10.1039/AN95479BP159

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

NOVEMBER, 1954 Voi. 79. No. 944 THE ANALYST PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY NEW MEMBERS ORDINARY MEMBERS John Walter Bowler, M.P.S.; Leslie Coggon; James Corrin Cain, B.Sc. (Manc.), Ph.D. (Liv.), A.R.I.C. ; -Florence Mary Lucy Evans, A.R.I.C. ; Cyril Hall; Adam Sinclair Inglis, M.Sc. (Melbourne) ; Geoffrey Stewart Ingram, A.R.I.C. ; John Waring Lucas, B.Sc. (Lond.), A.R.I.C. ; Kenneth Rogers, Ph.C. ; Richard Alan Savidge; Leslie Zaccheus Skelton, B.Sc. (Lond.), Dip.Bact. ; Kenneth Alan Wickham, BSc. (Lond.) ; George Robert Wilkinson, M.P.S., Ph.C.; Rodney James Binham Williams; Alan Ronald Witty, D.F.C., B.Sc. (Lond.), A.R.I.C. JUNIOR MEMBERS Irene Ladden, B.Pharm. (Lond.) ; Cecilia Julie Lloyd, B.Sc. (Notts.) ; Peter George Marshall, B.A. (Oxon.) ; Anthony Harold Morris, B.Sc. (Dunelm.) ; Leon E. Solomon, B.Sc. (Manc.). DEATHS WE regret to record the deaths of v John Knaggs Harold Blythen Stevens. 66 L

ISSN:0003-2654    

DOI:10.1039/AN954790661a

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

662 STOCK : THE SPECTROPHOTOMETRIC ESTIMATION OF TOTAL [Vol. 79 The Spectrophotornetric Estimation of Total Penicillins by Conversion to Penicillenic Acid and the Importance of Copper in Controlling the Reaction BY F. G. STOCK (Presented at the meeting of the Society on Wednesday, May 5th, 1954) A method for the estimation of total penicillins is proposed; it is a modification of that described by Herriott. The importance of the presence and concentration of copper is demonstrated and the effects of variations in experimental conditions are described. The application of the method to the analysis of oral tablets of penicillin is described in detail with many examples. A comparison is made of the results found by the method described with those by biological assay. A METHOD for the estimation of total penicillins was proposed by Herriott,l who made use of the fact that acid aqueous solutions of penicillin develop an absorption band at 322 mp owing to the formation of what he termed an intermediate degradation product.An empirical technique was used in which the aqueous solution of penicillin was heated with 0.4 M acetate buffer of pH 4.6 for 15 minutes in a water-bath and then rapidly cooled. The increase in absorption a t 322 mp was measured and compared with a blank, which had been kept at room temperature. Herriott outlined procedures for high and low potency material. (a) High potency material containing 35 to 500 units per mLTo 2 ml of such a solution was added 6 ml of 0.4 M acetate buffer of pH 4-6 (prepared by mixing equal quantities of 0.4 M sodium acetate and 0.4 M acetic acid) , and 4 ml of the mixture were transferred by pipette to a 30 to 50-1nl test tube, which was placed in a bath of boiling water for 15 minutes and then cooled rapidly to room temperature. The E values a t 322 mp for the heated and unheated mixtures were determined, and the increase in absorption was calculated.An accuracy of within &5 per cent. was claimed for the method.Nov., 19541 PENICILLINS BY CONVERSION TO PENICILLENIC ACID 663 (71) Low Potency material containing 8 to 80 units per mG-To 4 ml of such a solution was added 0-25ml of 5 M acetate buffer and the same procedure was adopted as for (a), except that the blank was determined on the same solution before heating. A small funnel was placed in the test tube to act as a condenser during heating, and the E value kt 322 mp of the unheated half of the sample was measured whde the other half was heating.The absorption value at 322 mp of the unheated penicillin - buffer mixture at 26” C increased each hour by 3 to 4 per cent. of that shown by the heated mixture; the increase in half an hour was therefore negligible. After heating and cooling the penicillin - buffer mixture, there was no further change in the absorption value for several hours. The maintenance of standard conditions was found to be important, unless a standard penicillin was assayed at the same time. Herriott suggested a constant volume of boiling water and a depth of immersion of the test tube sufficient to bring the level of its contents just below that of the water in the bath.He also recommended that the volume of the water in the bath should be sufficient to ensure that its temperature did not fall more than 2” to 3” C when the tube was placed in it. According to Herriott, Beer’s law is not obeyed; the value of E:Fm at 322 mp when determined at different concentrations of penicillin is not constant and this makes one of three methods of approach necessary- (i) Use of empirical curves to determine the correct value of E:Fm at 322mp for a given concentration of penicillin. (ki) An approximate method in which it is assumed that there is no alteration in the value of E:Fm at 322 mp. When an average value is used, the error is said to be about *lo per cent. (iii) Analysis of standards simultaneously.Herriott’s results showed that penicillins G, X, K and F gave nearly identical curves, and that, when mixtures of these penicillins were analysed, the results for total penicillins were within 5 per cent. of the theoretical values. I t was also stated that occasionally a value was obtained that was out of line and for which no adequate explanation could be given; this was not a frequent occurrence and was probably due to variations in copper concentration not recognised by Herriott at the time. The variables of the method and their effects upon the E value at 322 mp were described, the three most important being the time of heating in the water-bath, the pH of the buffer solution and the molar concentration of the buffer. As will be shown later, the effect of an increase in the molar concentration of the buffer was probably partly due to the accompanying increase in the copper concentration.We in this laboratory tried Herriott’s method for making an approximate check of the potencies of penicillin oral tablets, in which the presence of lactose interferes with an iodimetric assay. We were unsuccessful because great difficulty was experienced in attaining repro- ducible results. ‘There seems to be very little doubt, judging by the literature,2 that what Herriott termed “an intermediate degradation product” was a penicillenic acid. Realising the possible importance of the effect of trace metals on such a reaction, we made an investigation of the method and found that, under certain conditions, the addition of copper to the buffer solution effected an exact reproducibility of results.It seems probable that Herriott un- wittingly had sufficient copper present, derived from the sodium acetate and acetic acid used in his buffer solutions, and this resulted in some degree of reproducibility being attained. If analytical-reagent grade sodium acetate and acetic acid are used in the preparation of the buffer solution, results are so erratic that the method would appear to be of little use; this defect can be remedied by the inclusion of copper in the buffer. The method is simple, rapid and of reasonable accuracy. EXPERIMENTAL To minimise errors in the use of a pipette, the following standard procedure was adopted: A solution of penicillin was made to contain approximately 200i.u.per ml; 10ml of this solution were diluted with 30 ml of buffer and 4 ml of the mixture were transferred by pipette to each of six Pyrex-glass 6-inch x 2-inch tubes for heating in the water-bath. The con- centration of penicillin in the penicillin - buffer mixture was therefore approximately 50 i.u. per ml, which is very convenient for the determination of the increase in the E value at 322mp after heating. For the recording of standard results, it is convenient to calculate the value of E ~ o c ~ p e r m l at 322 mp for the final heated buffered solution (i.e., 200 i.u. per ml864 STOCK : THE SPECTROPHOTOMETRIC ESTIMATION OF TOTAL [Vol. 79 diluted (1 + 3) with 0.4 M acetate buffer) and also to specify the copper concentration in p.p.m. and the actual penicillin concentration at which the determination was made.EFFECT OF THE ADDITION OF COPPER- In these experiments a solution of penicillin containing 400 i.u. per ml was prepared; 25 ml of this solution were added to 160 ml of 0.4 M acetate buffer, and the volume was made up to 200 ml after the addition of the required amount of copper. The E values determined for different copper concentrations are shown in Table I ; each value is the mean of six determinations. It will be observed that the addition of a very small amount of copper produces a rise in the E value and leads to closer agreement between replicates. The fiducial limits of the mean fall rapidly when the added copper content is increased from 0.02 to 0.10 p.p.m. It is a useful routine procedure to carry out determinations in six tubes with TABLE I VARIATION OF E6,tF per ml AT 322 ~p WITH DIFFERENT CONCENTRATIONS OF COPPER Fiducial limits expressed as percentage of the Added copper, mean of six readings mean (P = 0.95) at 322 mp, Ef;;a per ml p.p.m.0.02 0.05 0.10 0.20 0.30 0.40 0.50 0.60 1.00 0.460 13-0 0.695 1.6 0.736 practically nil 0-765 79 0.769 99 0.767 97 0.754 79 0-702 99 0.582 2 9.0 no added copper, for if the fiducial limits of the mean of the six values determined are about k 2 per cent. or more and the mean E value is less than the corresponding E value determined when copper is added, then it is reasonable to assume that the original copper content of the solution was approximately O.lOp.p.m. or less. The curve relating E value to copper content rises rapidly to a maximum at approximately 0.40p.p.m., and between 0.30 and 0.50 p.p.m.there is little significant difference between the values of E; the values for the range 0.20 to 0.60 p.p.m. of copper are reasonably constant. The addition of 0-45 p.p.m. of copper to the original buffer solution prepared from analytical-reagent grade sodium acetate and analytical-reagent grade acetic acid produces the best result, the copper content of the penicillin - buffer being approximately 0.33 p.p.m. With solid penicillin the original copper content is so small that, at penicillin concentrations of approximately 50 i.u. per ml, its effect is negligible. For oral tablets containing above 10,000 i.u. the original copper content is again negligible, but an additional source of copper is the lactose.If, however, the lactose conforms to the B.P. 1953 limit of 3 p.p.m., its copper content at the dilution used will also be negligible. The final dilution of a 10,000 i:u. tablet would be 200 ml, and assuming it to contain 1 g of lactose with 3 p.p.m. of copper, the concentration of this metal derived from lactose would be merely 3/200p.p.m. 'The effect of lactose on the determination was found to be negligible; its addition to a solution of penicillin caused no alteration in the value found for E. EFFECT OF VARIATIONS IN THE EXPERIMENTAL CONDITIONS- (a) Variation of the time of immersion in the water-bath-The results in Table I1 demon- strate the fact that the maximum E value was developed after 15 minutes. A detailed examination over the range 14 to 164 minutes showed that the time of immersion was not critical to within + minute or even more.Experiments were made to ascertain the effect of various copper concentrations on the time taken for development of the maximum E value, as it was thought that the effect of a higher copper concentration might result in a more rapid development of the maximum E value. It was found, however, that whatever the concentration of copper, the maximum E value developed after about 15 minutes.Nov., 19541 PENICILLINS BY CONVERSION TO PENICILLENIC ACID TABLE I1 VARIATION OF E VALUE WITH TIME OF IMMERSION IN WATER-BATH FOR SODIUM BENZYLPENICILLIN AT A CONCENTRATION OF 44-9i.u. PER ml WITH A COPPER CONCENTRATION OF 0.33 p.p.m. Time, minutes Value of E at 322 m p 5 0.405 10 0.61 7 14 0.687 0.693 0.692 0.692 0.689 0.692 20 0.67 7 25 0.652 30 0.632 2 2 665 (b) Variation of the water-bath conditions-Table I11 shows results for solutions of penicillin heated simultaneously in different water-baths.Little difference could be observed in the E values determined, and it appeared that there was good reproducibility of results under slightly different conditions of heating in the water-bath. As a water-bath we used a 2-litre Pyrex-glass beaker filled with water to within about 2 inches from the top and heated on a gas-ring to a gentle boil. The test tubes used were Pyrex-glass 6-inch x 3-inch, six of which were used in every determination. During heating a small funnel was placed in each test tube to act as a condenser. The tubes were placed in a copper test-tube holder, which was circular in shape with six holes and fitted easily into the 2-litre beaker so that the holder with the tubes could be immersed or removed from the bath in its entirety.We found that the level of the water in the bath was of very little importance for reproducibility of results. If the penicillin - buffer mixture was heated under a reflux condenser over a burner instead of being heated in a bath, the result was much lower; in one experiment an E value of 0.214 was found for a solution that gave a value of 0.700 after immersion in a water-bath. TABLE I11 VARIATION OF E VALUE WITH VARIOUS WATER-BATH CONDITIONS FOR SODIUM BENZYLPENICILLIN AT A CONCENTRATION OF 44.9i.u. PER ml WITH A COPPER CONCENTRATION OF 0.33 p.p.m.(a) S i x tubes, small water-bath, value of E at 322 mp: 0.700, 0.694, 0+700, 0.700, 0.691, 0.688 Mean 04396 (b) Eighteen tubes, large water-bath, value of E at 322 mp: (i) inside tubes 0.688, 0.686, 0.692, 0.694, 0.692, 0.694 Mean 0.691 (ii) outside tubes 0.673, 0.696, 0.679, 0.699, 0.703, 0.703, 0.689, 0.703, 0.700, 0.689, 0.682, 0.694 Mean 0.693 (c) Variation of the volume of penicillin - bufer mixture-The specification of 4 ml as the volume of penicillin - buffer mixture to be immersed in the water-bath was far from critical, as the results in Table IV show. Any volume between 3 and 10ml resulted in approximately the same E value. Four millilitres was a convenient volume, because it is ideal for use in the cell of the spectrophotometer, but as an alternative to the use of a pipette, the tubes could be marked a t the 4-ml level or some other level if thought convenient, or a burette could be used.TABLE IV VARIATION OF THE VALUE OF E AT 322mp WITH VARIOUS VOLUMES OF PENICILLIN - BUFFER SOLUTION HEATED I N THE WATER-BATH Volume in tube, ml . . 3 4 5 6 7 8 9 10 Value of E at 322mp . . 0-690 0.691 0.694 0.685 0.679 0.682 0.680 0.680666 STOCK : THE SPECTROPHOTOMETliIC ESTIMATION OF TOTAL [Vol. 79 BEER'S LAW- Herriott showed that apparently Beer's law is not obeyed. In order to gain some idea of the errors involved by the use of different concentrations of,penicillin to determine the E::: per ml value at 322 mp, an investigation was made. I t was found that the Ei°CLu per m1 value at 322 mp as determined at a concentration of 50 i.u.per ml was increased by approxi- mately 3 per cent. if the concentration was doubled, and decreased by approximately the same amount if it was halved. It was concluded that the error introduced by disregarding this variation for concentrations between 40 and 50 i.u. per ml would not exceed 1 per cent. STANDARD RESULTS- A number of values determined for EI\&Perml at 322 mp with a copper concentration of 0.33 p.p.m. and a penicillin concentration of 40 to 50 i.u. per ml are shown in Table V. The mean value was 0.760 The benzylpenicillin used for these determinations was of recent manufacture and was found by biological assay to contain 1650 i.u. per mg for the sodium salt and 1600 i.u. per mg for the potassium salt. 0.010 (1.3 per cent.) for P = 0.95.TABLE V VALUES DETERMINED FOR E I 0 Z p o r m 1 AT 322 mp AT A COPPER CONCENTRATION OF 0.33 p.p.m. Penicillin concentration a t which determined, E f F per m1 a t 322 m p i.u. per rnl r 0.759 42 Sodium benzylpenicillh . . 0.752 0.762 0.772 0.755 0-768 40 43 45 44 46 0-762 60 "{ 0.754 50 Potassium benzylpenicillin Mean .. .. 0,760 The decay curve for a 50,000 i.u. per ml solution of penicillin kept a t 25" C was deter- There can be no doubt that only biologically mined and the results are shown in Table VI. active penicillin is measured by this method. TABLE VI DECAY CURVE RESULTS FOR POTASSIUM BENZYLPENICILLIN AT A CONCENTRATION OF 60,000 i.U. PER ml AND A TEMPERATURE OF 25" c Days 0 1 2 3 4 6 7 E60 i.n. per ml lcm a t 322mp 0.762 0.529 0.166 0.064 0.035 0.018 0.016 E/Emax 1.00 0.694 0.218 0.084 0.046 0.024 0.021 METHOD As a general routine, for samples of solid penicillin and oral tablets containing 10,000 i.u.or more, a solution containing 160 to 200 i.u. per ml was prepared. Four tablets were used for the assay of the oral preparation because of possible variations in potency, and a suitable dilution was made. Ten millilitres of the penicillin solution were added to 30 ml of 0.4 A4 acetate buffer (prepared from analytical-reagent grade material) containing 0.45 p.p.m. of added copper. Replicate determinations were made in six tubes each containing 4ml of the mixture; the mean value for the increase in absorption at 322 mp was calculated, and from the standard expression E:O,kPerml at 322 mp = 0.760, the penicillin potency was determined.The determination was repeated with acetate buffer containing no addedNov., 19541 PENICILLINS BY CONVERSION TO PENICILLENIC ACID 667 copper, and if the mean value so found was lower than that found on the addition of copper, and if also the fiducial limits of the mean of the six values determined were in the region of &2 per cent. or more, then the copper originally present in the penicillin - buffer mixture could be assumed to be approximately 0.10p.p.m. or less and was of no consequence. APPLICATION OF THE METHOD TO A NUMBER OF SAMPLES OF ORAL TABLETS OF PENICILLIN The results in Tables VII, VIII and IX indicate the type of values attainable by this method. The results are for tablets stated to contain 100,000, 200,000 and 500,000i.u.of penicillin per tablet. Four tablets were dissolved per litre of solution, and suitable aliquots were taken and diluted so that all determinations were made on penicillin - buffer solutions nominally containing 50i.u. per ml. The effect of the addition of copper in raising the E TABLE VII TO CONTAIN 100,OOOi.u. PER TABLET RESULTS OF THE ANALYSIS OF ORAL TABLETS OF PENICILLIN DECLARED Buffer with copper Buffer without copper P A 7 Sample number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Value of E at 322 mp, mean of 6 replicates 0.809 0.639 0.662 0.463 0-5 11 0.740 0.717 0.551 0.816 0.949 0-781 0.803 0.657 0.544 0.793 0.813 0.868 0.781 0.657 0.715 0.824 2 X 0-757 0.764 0.737 0.805 0-700 0.706 0.346 0.078 0.068 0.712 0.932 0.695 0.751 0.699 0.708 0.770 0,375 0.751 0.831 0.772 0.654 0.648 Fiducial limits ' Fiducial limits expressed as Value of expressed as percentage of E at 322 mp, percentage of the mean mean of the mean (P = 0.95) 6 replicates (P = 0-95) practically nil 0.400 n 0.417 99 0.412 11 0.288 31 0.329 99 0.583 99 0.630 99 0.279 21 0,630 91 0.660 19 0.333 91 0.582 11 0-249 99 0.291 99 0.536 91 0.593 91 0-552 1.2 0.475 practically nil 0.495 9? 0.434 11 0.763 91 2 x 0.365 9, 0-304 91 0.534 19 0.603 99 0.445 97 0.466 99 0.245 19 0.052 77 0.022 97 0.504 19 0.676 19 0.5 12 99 0-542 99 0.493 99 0.47 1 2 2.0 0.541 practically nil 0.213 99 0.582 1) 0.595 19 0.575 Y9 0-495 1 ) 0.443 + 7.1 5 3.2 & 2.5 - + 4.6 + 4.7 2.5 2.8 & 4.3 & 5.0 & 3.3 2 5-9 f- 3.5 & 9.4 2 5.4 f- 4.3 & 9.0 & 8.5 about & 1.0 10.3 12-6 & 8.1 f- 1-6 + 8.1 4-6 about 1.0 practically nil + 5.6 4.6 practically nil 9 1 & 4.1 f- 3.5 + 2.2 5 3-6 & 2.3 f- 1.5 & 3.0 f- 2.6 f 2.5 + 2.6 about 1.0 rfi 5.0 practically nil Penicillin content per .tablet, 106,400 84,100 87,100 60,900 67,200 97,400 94,300 72,500 107,400 124,800 102,800 105,600 73,300 71,600 104,300 107,000 114,200 102,800 86,500 94,100 108,400 199,200 100,600 97,000 105,900 92,100 92,900 45,400 10,300 8,900 96,700 122,700 91,500 98,800 92,000 93,200 101,400 49,300 98,800 109,300 101,600 86,100 85,200 l.u.Excess or deficiency, % 6.4 - 15.9 - 12.9 -39.1 - 32.8 - 2.6 - 6.7 - 27.5 7.4 24.8 2.8 5.6 - 26.7 - 28.4 4.3 7-0 14.2 2.8 - 13.6 - 5.9 8-4 100.0 0.6 - 3.0 5.9 - 7.9 - 7.1 - 54.6 - 89.7 -91.1 - 3-3 22-7 - 8.5 - 1.2 - 8.0 - 6.8 1.4 - 60.7 - 1.2 9.3 1.6 - 13.9 - 14.8668 STOCK : THE SPECTROPHOTOMETRIC ESTIMATIOK OF TOTAL [Vol.79 value and the resulting agreement between replicates is striking. As a test of the method, a number of the samples were submitted for analysis to a large laboratory specialising in the biological assay of penicillin; the agreement between the results is illustrated by the values in Table X. Even though the method gave almost perfectly reproducible results, as illustrated by the “typical analysis” in Table XI, it does not of necessity follow that the estimate of TABLE VIII RESULTS OF THE ANALYSIS OF ORAL TABLETS OF PENICILLIN DECLARED TO CONTAIN 200,000 j.u. PER TABLET Sample number 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 Sample number 77 78 79 80 81 Buffer with copper Buffer without copper T i t s 7-‘ZiZZit s Value of expressed as Value of expressed as E a t 322 mp, percentage of E a t 322 mp, percentage of Penicillin mean of the mean mean of the mean content 6 replicates (P = 0.95) 6 replicates (P = 0.95) per tablet, 0.776 0.773 0.771 0.812 0.840 0.887 0-715 0.794 0.618 0.017 0.791 0.790 0.810 0.726 0-798 0.770 0.786 0.785 0.728 0.795 0.052 0.582 0-832 0.780 0.800 0.815 0.820 0.779 0.774 0.748 0.808 0.798 0-652 practically nil 9 , 9) 7, 9, 3, 19 9 9 9 1 99 0-632 0.536 0.490 0.583 0.620 0.570 0.497 0,482 0.342 0.547 0.772 0.510 0.539 0.547 0.626 0.646 0.464 0.512 0.662 0.036 0.328 0.547 0.468 0.502 0.503 0-614 0.498 0.538 0.457 0.597 0.520 0.491 TABLE i.u.+ 1.1 204,200 4.8 203,400 -t 6.1 202,900 3.9 213,200 i 8-7 221,100 2 4.0 233,400 2 2.7 188,200 * ‘3.4 208,900 * 8.4 166,100 * 3.9 208,600 practically nil 207,900 - + 3.9 213,100 * 3.7 191,000 & 2.3 210,000 :tbout & 1.0 202,600 +. 7.5 206,800 * 10.3 206,500 * 4.7 191,500 * 2.1 209,200 about & 1.0 6800 i 8.7 153,100 -1 5.7 219,000 9.6 205,200 i 4.0 210,500 & 6.9 214,600 * 4.9 215,800 -t - 4.7 205,000 2.9 203,600 & 3.6 196,800 -t 3.0 212,600 5.6 210,000 - + 2.5 171,300 practically nil IX Excess or deficiency, % 2.1 1.7 1-5 6-6 10.6 16.7 - 5.9 4.5 - 17.0 - 100.0 4.3 4.0 6.6 - 4-5 5.0 1-3 3.4 3-3 - 4.3 4.6 - 96.6 - 23.5 9.5 2.6 5.3 7.3 7.9 2.5 1.8 - 1.6 6.3 5.0 - 14.3 RESULTS OF THE ANALYSIS OF ORAL TABLETS OF PENICILLIN DECLARED TO CONTAIN 500,000 i..u.PER TABLET Buffer with copper Value of expressed as mean of the mean 6 replicates (P = 0.95) -7 F i d u c i a l m i t s percentage of E a t 322 mp, 0.620 practically nil 0.780 19 0.749 >, 0.755 ,, 0.828 51 Buffer wirhout copper --its Value of expressed as percentage of mean of the mean 6 replicates (P = 0.95) E at 322 mp, 0.489 + 3.5 0.445 6-1 0.501 i- 4.7 0.570 + 4.1 0.526 j 3 2 . 1 Penicillin content Excess or per tablet, deficiency, i.u. Yo 513,300 2.7 408,000 - 18.4 492,800 - 1.4 496,700 - 0.7 544.800 9.0Nov., 19541 PENICILLINS BY CONVERSION TO PENICILLENIC ACID 669 the penicillin potency was a true one. From a consideration of the results in Table X and the known variance of the biological method, it could be deduced that there was no evidence of a difference in accuracy between the two methods.In arriving at this estimate the results on samples 64, 28, 29, 38 and 30 were omitted because they showed deficiencies in their declared penicillin contents exceeding 50 per cent. and, as all solutions prepared for chemical assay were made on the assumption that the potency was as declared, the resulting solutions contained less than 25 i.u. per ml and the E values were much smaller than those normally determined. For samples exhibiting such deficiencies, a fresh solution should be TABLE X COMPARISON OF THE RESULTS BY THE SPECTROPHOTOMETRIC METHOD WITH THOSE BY Bacillus subtiliis PLATE ASSAY Results by A r \ Sample Chemical Biological number method method 77 2 3 4 5 8 52 13 14 6 50 7 79 57 19 64 65 22 28 29 32 76 38 42 43 30 408,000 84,100 87,100 60,900 67,200 72,500 166,100 73,300 71,600 97,400 188,200 94,300 492,800 191,000 86,500 6800 153,100 199,200 45,400 10,300 122,700 171,300 49,300 86,100 85,200 8900 434,000 86,000 93,000 57,000 66,400 72,900 177,300 83,000 74,400 88,000 187,000 92,000 482,000 179,000 72,100 6470 155,800 208,800 30,000 3160 116,200 156,000 46,400 80,100 81,100 7360 Ratio of result by chemical method to result by biological method 0.94 0.98 0.94 1.07 1.01 1 -00 0.94 0.88 0.96 1.11 1.01 1.03 1.02 1-07 1.20 1.05 0.98 0-95 1.51 3-26 1-06 1.10 1.06 1.08 1.05 1.21 TABLE XI TYPICAL ANALYSIS OF A SAMPLE OF 100 TABLETS DECLARED TO CONTAIN 500,000 i.U.OF PENICILLIN PER TABLET Test 1 Test 2 Test 3 - - - Buffer Buffer Buffer Buffer Buffer Buffer with without with without with without copper Blank.. . . 0.026 Results . . 0.842 0-840 0.840 0.840 0.840 0.837 Mean . . . . 0.814 Variation . . practically nil copper 0.025 0.610 0.589 0.579 0.590 0.608 0.584 0.568 f 2.4% copper 0.022 0.848 0.845 0.848 0.845 0.845 0.850 0.824 practically nil copper 0.026 0-690 0.695 0.700 0.678 0.688 0.692 0.665 It 1.2% copper 0.020 0.837 0.832 0.840 0.835 0.837 0.835 0-816 practically nil copper 0.024 0.690 0.698 0.678 0.698 0.704 0.696 0.670 f 1.4% Grand mean in presence of copper buffer: 0.818, equivalent to 538,000 i.u. of penicillin per tablet. Two sets of 4 tablets wefe biologically assayed and found to contain 517,200 i.u.and 520,000 i.u.of penicillin per tablet, giving a mean of 518,600 i.u.670 STOCK THE SPECTROPHOTOMETRIC ESTIMATION OF TOTAL [Vol. 79 prepared so that the final penicillin - buffer mixture contains approximately 50 i.u. per ml; this was impossible with some of our samples because of the small amount available. For sample 29 the biological result was 3160i.u. compared with 10,300 i.u. by the chemical method; the tablets were declared to contain lOO,OOOi.u., and hence the chemical result showed a deficiency of 90 per cent. compared with 97 per cent. biologically. DISCUSSION OF RESULTS The value of 0.760 determined for Ef",per ml at 322 m p corresponds to an E::, value of 251 and 243, respectively, for sodium and potassium benzylpenicillins. This is higher than the value of 227 determined by Herriott.A value of 227 for the EYbm value corresponds to an E60 1 cm i.u per ml value of approximately 0.690, representing a sub-optimum copper con- centration of about 0.10 p.p.m., but nevertheless sufficient to give a reasonable reproducibility of results. To ascertain if the shape of a particular curve is normal, the ratio of E at 300mp to E at 322 mp and the ratio of E at 340 mp to E at 322 mp can be determined; in the samples assayed these ratios were constant at approximately 0600 and 0.600, respectively. The method is extremely simple and has the advantage of speed, a result being available within an hour. It is without doubt of exceptional value as a sorting test, and bearing in mind the limits of error of the biological assay and the fact that these limits vary with the experience of and refinements added to the method by different laboratories, we hold that the spectrophotometric assay would give more reliable results in smaller laboratories not possessing the advantages of those specialking in biological assay work.It enabled us to sort a number of samples submitted under the Food and Drugs Act in Birmingham into those that were genuine and those showing major deficiencies; the latter were then submitted to a specialist laboratory for biological assay. What was even more satisfying was that the biological assays always confirmed our findings by the chemical method. Owing to the high cost of biological assays, it may well be that further study of the penicillin - penicillenic acid reaction and its use in a spectrophotometric assay of penicillin would be well repaid.I desire to thank Imperial Chemical Industries Ltd. for carrying out the biological assays. I am indebted to Mr. H. H. Bagnall for his interest and for the encouragement and suggestions made by him during the writing of this paper; acknowledgment is also made to Mr. L. W. Hinson, who assisted in the chemical analysis. REFERENCES 1. 2. Herriott, R. M., J . Bid. Chem., 1946, 164, 725. "The Chemistry of Penicillin," 1949, Princeton University Press, New Jersey. CITY ANALYST'S LABORATORY BIRMINGHAM December Ist, 1953 DISCUSSION DR. D. C. GARRATT asked the speaker whether he had considered the use of the simple titration of oral tablets and similar preparations after addition of penicillinase.This method took only half an hour, was highly specific and did not require spectrophotometric apparatus; it was therefore useful in quite small laboratories without such facilities. DR. E. C. WOOD asked whether, in view of the effect of copper, the author had investigated the influence of other metals that might be present in traces, either in the preparations assayed or in the reagents. DR. R. E. STUCKEY said that, in view of the discrepancies in some of the results shown in Table k, i t did not seem that the method could be more than a sorting test. He asked if the author could explain why the solutions did not obey Beer's law. MR. STOCK replied that the spectrophotometric method could be used in circumstances in which the estimation mentioned by Dr. Garratt could not; there was no advantage in time with the method quoted as compared with the physical method. The fundamental point of the paper was that Herriott's spectro- photometric method was not satisfactory if analp-tical-reagent grade chemicals were used ; the inclusion of copper remedied this defect and the use of the method in the assay of oral tablets of penicillin was only one example of its application. In answer to Dr. Wood, he said that the question of other metals could be considered under two headings: first, whether there was a more suitable catalyst than copper and, secondly, whether anyNov., 19541 PENICILLINS BY CONVERSION TO PENICILLENIC ACID 67 1 other metals present in the sample under test were likely to affect the result. Copper seemed to be satis- factory as a catalyst; in oral tablets of penicillin, they had not encountered any other metals, but if the method was used in other instances when metals were likely to be present, then Dr. Wood’s point would have to be borne in mind. He did not agree with Dr. Stuckey in his assessment of the method, because a statistical examination of the results in Table X indicated the precisions and accuracies of the physical and biological methods to be about equal. As they were therefore comparing two methods each with a possible error in the region of f 6 per cent., Table X was what would be expected; there was no evidence of a bias. He had no explanation for the divergence from Beer’s law.

ISSN:0003-2654    

DOI:10.1039/AN9547900662

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Nov., 19541 PENICILLINS BY CONVERSION TO PENICILLENIC ACID 67 1 A Critical Study of the Determination of Methoxyl Groups BY A. E. HERON, R. H. REED, H. E. STAGG AND H. WATSON When the method described by Easterbrook and Hamilton for the determination of methoxyl groups in cellulose ethers was used for pure methoxy compounds, the recoveries were erratic. The cause of these erratic results was traced to the sodium thiosulphate in the wash solution used to remove hydrogen iodide from the stream of gas carrying methyl iodide into the absorber. A 25 per cent. aqueous sodium acetate solution is a satisfactory wash solution. Conditions are described in which the average recovery from vanillin and a-methyl-D-glucoside is 100 per cent., with a coefficient of variation of 0.21. THE Zeisel reaction, whereby ethers are split by the action of hydriodic acid, 3s well known as the basis of methods for the determination of methoxyl groups. Such methods usually conform to the following scheme.The sample is decomposed by boiling it with hydriodic acid, when an equivalent amount of methyl iodide is formed. The liberated methyl iodide is separated from the accompanying hydrogen iodide by passing the gases through a condenser and then a trap or scrubber designed to retain the hydrogen iodide. The methyl iodide is quantitatively absorbed and converted either to silver iodide by reaction with alcoholic silver nitrate, to an ionised iodide by reaction with an amine, or to iodate by oxidation with a solution of bromine in acetic acid. The iodide or iodate is then determined by a standard method.Many minor variations of the method have been proposed; for example, some workers make use of phenol as a solvent in the digestion flask, whilst others use the constant-boiling hydriodic acid alone; sometimes a small amount of hypophosphorous acid is added to the hydriodic acid, presumably to prevent the formation of free iodine during the digestion. Various reagents have been used to remove the hydriodic acid; these include water, red phosphorus suspended in water and a solution containing cadmium sulphate, sodium thiosulphate and sodium acetate. Several unsatisfactory features of the method have been examined systematically by Easterbrook and Hamilton,l who have designed an improved apparatus and devised a procedure especially suitable for the determination of alkoxyl groups in cellulose ethers.Extensive use of their method on pure compounds showed that, if the prescribed conditions were followed carefully, the results were usually acceptable, although occasionally they were low for no apparent reason. The information available in the literature was so confusing that it was decided to examine every stage of the method, with the object of establishing a procedure with an accuracy of at least 1 in 100 units of the theoretical methoxyl content, even on the micro- analytical scale. EXPERIMENTAL APPARATUS- described by Easterbrook and Hamilton1 was chosen. Many forms of apparatus have been used, but for this investigation the all-glass apparatus However, the calibrated jet-unit used672 HERON, REED, STAGG AND WATSON: A CRITICAL STUDY OF by them for controlling the rate of flow of the inert gas was replaced by a White - Wright flowmeter (British Standard 1428 : Part A1 : 1950).[Vol. 79 DEVELOPMENT OF A METHOD I N WHICH MONOISTHANOLAMINE IS USED AS ABSORBENT- Amines and solutions of bromine in acetic acid have both been used as absorbents for methyl iodide; for the initial investigation monoethanolamine was chosen because it is an efficient absorber for methyl iodide, with which it reacts to form an ionisable quaternary iodide. To confirm that methyl iodide can be quantitatively determined by this means, sealed capillary tubes containing weighed amounts of 25 to 30 mg of analytical-reagent grade methyl iodide were broken under the surface of 5 ml of monoethanolamine. After 30 minutes, the mixture was neutralised with 3 M sulphuric acid and a 20-ml excess of acid was added; then 1 g of ammonium nitrate was added and the solution was titrated electrometrically with 0.02 N silver nitrate solution, a valve potentiometer being used.Results were between 99.5 and 100.5 per cent. of the calculated values. Attempts were made to show that methyl iodide, introduced into the digestion flask of the apparatus in sealed capillary tubes, could be transferred to the absorber and recovered quantitatively by titration with silver nitrate solution. The results were so erratic that it was suspected that losses of methyl iodide occurred during the manipulation necessary for the breaking of the capillary tubes in the flask; no simple means of avoiding these losses could be found, and so it was decided to use a, purified methoxy compound as a source of methyl iodide.The substance used for this purpose was a-methyl-D-glucoside, recrystallised to constant specific optical rotation ; the calculated methoxyl content of this substance was 15.98 per cent. Weighed samples of a-methyl-D-glucoside were introduced into the digestion flask of the apparatus and their methoxyl content was determined, with monoethanolamine as absorbent and with water at room temperature as wash liquid in the trap. Recoveries were between 94 and 98 per cent. of the theoretical values, and it was demonstrated by qualitative tests that some of the methyl iodide had been retained by the wate-r in the trap.It seemed probable that increasing the temperature of the water in the trap would decrease the retention of methyl iodide, but the way in which this would affect the absorption of hydrogen iodide was unknown. Hence a series of determinations were made in the absence of a methoxy compound, and the amount of hydrogen iodide that passed through the trap and reached the absorber was determined by titration with silver nitrate solution. The water in the trap was maintained to within & 1" C of a given temperature during the determination and experiments were made over the range 30" to 50" C. Typical results are shown in Table I. TABLE I EFFECTIVENESS OF WATER IN THE ' r u r AT VARIOUS TEMPERATURES Temperature of water in trap, O C 30 40 42 42 42 50 Specific gravity of hydriodic acid in digestion flask .. . . . . 1.7 1.7 1.7 1.8 1-9 1.7 Volume of 0.02 N silver nitrate solution required for titration of contents of absorber, ml . . <0*05 (0.05 (0.05 0-12 0.15 0.1 to 0.9 Volume of 0.02 N silver nitrate solution required for titration of contents of trap, ml . . .. - -- <Om05 (0.05 > l o - From these results it can be seen that a trap containing water at 42" C (Le., at the boiling point of methyl iodide) is effective in retaining hydrogen iodide even when a comparatively large quantity of the gas reaches the trap, as happens when the hydriodic acid used is stronger than the constant-boiling mixture of specific gravity 1-7. Determinations were then made on the standard substance, a-methyl-D-glucoside (15.98 per cent.of methoxyl), 3 to 4 ml of water maintained at 42" & 1" C being used in the trap, and the results were as follows- Weight of sample taken, mg . . .. . . 29.0 25.8 24.5 Methoxyl found, per cent. . . .. . . 15.90 15-98 16.02 Recovery, per cent. . . .. .. . . 99.4 100.0 100.2Nov., 19541 THE DETERMINATION OF METHOXYL GROUPS 673 These results show that the first objective, ie., a method giving results with an accuracy of 1 in 100 units, was achieved. The cause of the low results occasionally found with Easterbrook and Hamilton's procedure had not been identified at this stage, although it was suspected that the sodium thiosulphate in the wash solution might be responsible. This problem is discussed below. DEVELOPMENT OF A METHOD IN WHICH A SOLUTION OF BROMINE IN ACETIC ACID IS USED AS Since its introduction by Viebock and Schwappach,2 a solution of bromine and potassium acetate in glacial acetic acid has been used extensively for absorbing methyl iodide.This reagent has two definite advantages over monoethanolamine : first, the methyl iodide is converted to iodate and six equivalents of iodine are finally liberated for each molecule of methyl iodide absorbed, so that the titre is reasonably large even when only a small amount of sample is available and, secondly, there is no need to make provision for the removal of hydrogen sulphide generated from sulphur-containing compounds. Reference has been made to the difficulty of completely oxidising the sample3s4 and so the conditions necessary for complete oxidation were investigated.The bromine solution used was that described by Easterbrook and Hamilton and was prepared by dissolving 24 g of bromine and 70 g of potassium hydroxide in 1 litre of glacial acetic acid. Sealed capillary tubes containing known weights of analytical-reagent grade methyl iodide were broken below the surface of the reagent, the proportions of bromine solution t o methyl iodide being varied from 0.5 to 2-0 ml of reagent per milligram of methyl iodide. The mixture was set aside for 5 minutes, an equal volume of 25 per cent. sodium acetate solution was added and the excess of bromine was destroyed by dropwise addition of formic acid. Potassium iodide and 3 M hydrochloric acid were added and the liberated iodine was titrated with 0-02 N sodium thiosulphate solution.When the ratio of bromine reagent to methyl iodide was 1.0 ml per mg or more, recovery of iodide was 99.6 to 100.0 per cent., but when the ratio was only 0-3 ml per mg, oxidation was incomplete; a brown colour persisted after the addition of formic acid and the recovery was only about 60 per cent. of theoretical. The theoretical amount of this bromine solution required to oxidise 1 mg of methyl iodide is only 0.15 ml, but these experiments showed that a considerable excess of reagent must be used. As the circumstances in which methyl iodide could be quantitatively oxidised had been ascertained, determinations were made on pure or-methyl-D-glucoside, with 2 to 4 ml of bromine solution per milligram of methyl iodide liberated and with water at 42" C in the trap; the results are shown in Table 11.ABSORBENT- TABLE I1 DETERMINATION OF METHOXYL CONTENT OF a-METHYL-D-GLUCOSIDE, WITH A SOLUTION OF BROMINE IN ACETIC ACID FOR ABSORPTION AND WITH WATER AT 42°C IN THE TRAP Weight of sample Methoxyl found, per Recovery, per cent.. . 99.3 99.9 99.9 98.9 100.1 99.8 99-3 99.3 99.6 Average recovery = 99-45 per cent. taken, mg . . . . 5.450 4-960 5.056 3.830 6.791 5.513 4.976 5.993 6.445 cent. . . . . 15-87 15.96 15.96 15.82 16-00 15.95 15-89 15.87 15.91 STUDY OF SCRUBBING LIQUIDS- The reason for the low results by Easterbrook and Hamilton's method was then sought. A likely source of such errors is the use of a solution containing sodium thiosulphate, which has been shown to react with methyl iodide by Slate9 and White.6 The effect of replacing water at 42" C by solutions of sodium thiosulphate, both alone and in admixture with other substances, is shown in Table 111.In all these experiments the technique and conditions used were those described in the procedure, the only deliberate variation being in the composition (not the volume) of the liquid in the trap. The conclusion from these results is that the thiosulphate absorbs methyl and ethyl iodide vapour and prevents a proportion of the alkyl halide from reaching the absorber. It has already been shown that a high accuracy can be achieved with a trap containing waterTABLE I11 EFFECT OF COMPOSITION OF SCRUBBING -LIQUID ON THE. DETERMINATION OF METHOXYL AND ETHOXYL GROUPS Na2S20,.5H,0 Na,S20,. 5H,O MgSz0,.6H,0 Mean recovery Temperature as percentage of liquid Substance of calculated Number of Composition of liquid in trap in trap, used* value tests "C (Methoxyl) .... .. .. .. .. .. 20 V 84-4 2 {:::g :: .. .. .. .. .. .. .. 20 V 74.5 2 2.5%; CHsCOONa.3H20: 12.5% . . .. .. .. 20 MG 95.4 2 2.5% . . .. .. .. .. .. .. .. 20 V 84.2 2 5.0y0 . . . . . . .. .. .. .. .. 20 V 77.8 2 10.0% . . * . . . .. .. .. .. .. 20 V 67.2 2 Na2S,0,.5H,0: 2.5% ; CdSO,: 2.5% . . . . . . .. . . . . - _ 211 fia2S20,.5H20: 5.0%; CdSO,: 2.5% . . . . .. .. .. . . 25 f .. 13 to 15 Na2S2O3.5H2O: 1.25%; CdSO,: 1.25%; CH,COONa.3H20: 6.25% 20 25 NazS,0,.5H20: 2.5%; CdSO,: 2.5% . . .. .. .. .. .. 20 Na,S20,.5Hz0: nil; CdSO,: 5.0y0 . . .. . . . . .. . . 22 Na2S,0,.5H,0: 5.0y0; CdSO,: 2.5% . . .. . . .. .. .. 25 . . .. .... .. .. .. 20 .. .. . . .. .. . . .. 20 . . .. .. .. .. .. .. 20 v V MG MG MG P P P P P P 99-75 96.4 99.3 97.3 97.1 (Ethoxyl) 98.75 96-4 99.9 96.7 98-8 98.0 A r 2 3 4 3 4 2 2 2 2 2 w Difference 2 between - highest and lowest result 8 -8 0.6 1.0 0 1.0 0 * 2 U 4.4 1.0 6.0 4 5 6-5 1-4 Z .. 1.0 * 0 1.9 2-2 $ 8 0.5 % 1.3 3 1.4 d U 4 0 2.6 r 4.1 0.9 * V = vanillin (20.38 per cent. of methoxyl). MG = a-methyl-D-glucoside (15.98 per cent. of methoxyl). P = phenacetin (25.14 per cent. of ethoxyl).Nov., 19541 THE DETERMINATION OF METHOXYL GROUPS 675 at 42" C, but the need to maintain the trap at this temperature complicates the method. The results in Table I11 suggest that losses of methyl iodide caused by thiosulphate are reduced by the addition of other electrolytes, such as cadmium sulphate and sodium acetate, to the solution, and experiments were made to see whether an aqueous solution of sodium acetate, without any other ingredient, could be used at normal laboratory temperature.The results are shown in Table IV. TABLE IV RESULTS WHEN A 25 PER CENT. SOLUTION OF SODIUM ACETATE IS USED IN THE TRAP Substance Methoxyl recovered, Temperature of percentage of trap liquid, calculated value " C Vanillin . . .. .. . . .. .. 21 21 21 22 22 25 25 21 21 a-Methyl-D-glucoside . . .. .. .. 21 Dimethoxyphenylpropionic acid . . .. 21 Average recovery = 100.0 per cent. Difference between highest and lowest figure = 0.9 per cent, 99-7 99.7 99.4 99.9 100-5 99.6 100.2 100.5 99.9 100.0 99.7 Comparison of the figures shown in Table IV with those in Table 11, when water at 42" C was used in the trap, suggests that the use of 25 per cent.sodium acetate not only allows the determination to be made with a trap at laboratory temperature, but also gives greater accuracy and precision. The effect of sodium acetate is not specific and is probably due to depression of the solubility of methyl iodide in water; it has been found by experiment that a similar con- centration of disodium hydrogen phosphate is equally effective. EFFECT OF FLOW-RATE- In all the foregoing experiments the rate at which carbon dioxide was passed through the apparatus was maintained at 6 ml per minute. It is, however, desirable to know to what extent this speed can be increased without affecting the accuracy of the results, and a number of experiments were made to test this point both on the micro and semi-micro scale.On the micro scale it was found that results were high when the flow-rate was increased substantially above 6 ml per minute, but that the incorporation of a simple bubbler, following the normal spiral bubbler and filled with 25 per cent. sodium acetate solution, allowed TABLE V EFFECT OF FLOW-RATE ON RECOVERY OF METHOXYL IN VANILLIN ON THE MICRO SCALE, 25 PER CENT. SODIUM ACETATE SOLUTION BEING USED AS SCRUBBING LIQUID ml per minute bubblers used Methoxyl found, Mean recovery, Flow-ra te , Number of % % 6 1 20.40, 20.24, 20.50, 20.31, 20.35 99.9 16 1 21.03, 21.35, 20.48, 21.34 103.3 25 1 20.56, 20-59, 20.65, 20.92 101.5 6 2 20-39 100.0 9 2 20.35, 20.41 100.0 12 2 20.30 99.6 15 2 20.47, 20.43 100.3 20 2 20-44, 20.44 100-3 Vanillin contains 20.38 per cent.of methoxyl.676 HERON, REED, STAGG AND WATSON: A CRITICAL STUDY OF [Vol. 79 flow-rates up to 20 ml per minute to be used without affecting the results significantly (see Table V). On the semi-micro scale (25 to 30 mg), however, increase of the flow-rate to 20 ml per minute, when only the normal spiral bubbler containing 25 per cent. sodium acetate solution at 20" C was used, was found to have no significant effect on the results (see Table VI). TABLE VI SEMI-MICRO DETERMINATIONS OF METHOXYL IN a-METHYL-D-GLUCOSIDE, A SPIRAL TRAP CONTAINING 25 PER CENT. SODIUM ACETATE SOLUTION AT ROOM TEMPERATURE BEING USED Weight of sample, Flow-rate, IMethoxyl found, Recovery, mg ml per minute % Yo 22.49 6 15.90 99.5 28.65 6 15.91 99.6 29.52 20 15.95 99-8 28-09 20 15.95 99.9 a-Methyl-D-glucoside contains 15.98 per cent.of methoxyl. (In these experiments the increased amount of bromine and stronger standard thiosulphate solutions specified in the semi-micro method were used.) DISCUSSION OF RESULTS- The most important conclusion to be drawn from the results shown in Table I11 is that the use of sodium thiosulphate as a component of the scrubbing liquid in methoxyl determinations is undesirable. In some circumstances quite a large proportion of the methyl iodide formed in the reaction vessel is prevented from reaching the absorber by the presence of thiosulphate in the scrubber and, although it is possible to suppress this effect by adding electrolytes and by keeping the temperature of the solution below 15" C, it is still difficult to get results of the highest accuracy when this substance is used.* Moreover, the low blanks and accurate results when water at 42" C and 25 per cent.sodium acetate solution a t 20" C were used as wash solutions demonstrate that no useful purpose is served by the inclusion of sodium thiosulphate in the solution. The experiments in which methyl iodide was oxidised with bromine emphasise the need for a large excess of bromine to achieve complete oxidation to iodate, and it is for this reason that a more concentrated solution of bromine is used in the semi-micro procedure. The bromine in acetic acid reagent is advantageous in that there is a comparatively large titre for a small amount of methyl iodide and that it can absorb a considerable amount of hydrogen sulphide without effect on the results; hence it is unnecessary to take any special precaution to remove hydrogen sulphide in the scrubber, as would be necessary when the classical method of absorption in an alcoholic solution of silver nitrate is used.It is better to use 25 to 30mg of sample rather than the 3 to 5mg commonly used in micro-analysis, and this is illustrated by the effect of the flow-rate of carbon dioxide on the results; on the larger scale the flow-rate can be increased to 20 ml per minute without affecting the results, whereas on the micro scale this can be done only if a second scrubber is incor- porated. Both the semi-micro and micro methods are described, but the former procedure is recommended for use whenever the quantity of sample is sufficient. METHODS Conditions are described for both the micro and semi-micro scales. The semi-micro procedure is similar to the micro, except that a larger volume of absorbent, containing a larger proportion of bromine, is used, and 0.1 AT instead of 0.01 N sodium thiosulphate is used for the final titration.As the results are unaffected by quite large variations in the flow-rate of the inert gas, the semi-micro method is recommended whenever a sufficient amount of sample is available. The apparatus required is the same for both methods. * The results quoted for phenacetin in Table I11 suggest that the same effect arises in the determination of ethoxyl group ; in this respect we have failed to confir:m the findings of White.6Nov., 19541 THE DETERMINATION OF METHOXYL GROUPS 677 MICRO METHOD APPARATUS- The digestion flask, A , is fitted with (i) a water condenser, B, fused to the neck and terminating in a B14 standard ground joint, N, and (iz) a side-arm reaching to within a few millimetres of the bottom of the flask with a safety bulb, C.The digestion flask should have a thin wall and base to minimise bumping during ebullition. The scrubbing trap, D, is fitted with a glass spiral and above it, a small bulb, E, to prevent the liquid in the trap from being sucked back. The side-arm of the trap is sealed at F with The apparatus is shown in Fig. 1. Fig. 1. Assembled apparatus a rubber stopper, and the trap is connected to the condenser, B, and the absorber, G, by means of appropriate ground-glass joints.The absorber is constructed with a tap at the bottom and an enclosed head, and it contains a glass spiral. The spiral has 12 to 14 turns per 10 cm of length, the capacity of the absorber being 13 to 14 ml measured from 0 to P. Carbon dioxide may be supplied from a suitable generator at a constant pressure, eg., a carbon dioxide generator unit for use with solid carbon dioxide. Alternatively, nitrogen from a cylinder can be used. A lute containing mercury to a depth of 80 to 100mm is inserted between the source of gas and the screw clip at the entry to the flowmeter to provide a constant gas pressure. A White - Wright flowmeter (British Standard 1428 : Part A1 : 1950) is recommended, but any other suitable type may be used; the flowmeter should be calibrated over the range678 1 to 12 ml per minute. calcium chloride, R.HERON, REED, STAGG AND WATSON: A CRITICAL STUDY OF [Vol. 79 The gas is passed through the flowmeter and a drying tube containing A micro-burner with a fine jet is used to heat the digestion flask. A 10-ml reservoir burette (British Standard 1428: Part D1) is required. Hydriodic acid, N.A.R., s9.g~. 1.7-This acid should be purchased in 5-ml ampoules. Hypophos$horous acid, 30 per cent.-Free from sulphate. Sodium acetate soZution , 25 $er cent.-Dissolve 25 g of analytical-reagent grade sodium acetate, CH3COONa.3H,O, in 100 ml of water. Bromine solution-Dissolve 17.6 g of potassium hydroxide in 227 ml of glacial acetic acid and add 2 ml of bromine, all of analytical-reagent grade.Starch solution, 1 per cent.-Dissolve 1 g of soluble starch in 10 ml of cold water and pour the paste into 90 ml of boiling water. PhenoG-Analytical-reagent grade. Formic acid-Analytical-reagent grade. Sodium thiosulphate solution , 0.01 N-Standardise the solution immediately before use with pure dry potassium iodate as described below. Place 3 to 4 mg of potassium iodate, accurattely weighed on a micro-balance, in a 250-ml stoppered conical flask and add about 100ml OF water. Warm the flask on a steam-bath for 10 to 15 minutes to dissolve the potassium iodate, cool the solution, wash the stopper with 1 to 2ml of water, allowing the washings to fall into the flask, and wash down the inside of the flask with a further 1 to 2 ml of water..To the solution add 5 ml of 2 N sulphuric acid and a few crystals of potassium iodide. Replace the stopper, set the flask aside, with occasional swirling, for 10 minutes and then titrate the liberated iodine with the 0.01 N sodium thiosulphate solution, 1 per cent. starch. solution being used as indicator. REAGENTS- Then- mg of potassium iodate - factor. ml of 0.01 N sodium thiosulphate solution x 0.3567 - PROCEDURE- Preparation of the ap$aratus-Clean the appropriate parts of the apparatus with a cold concentrated solution of sodium dichromate in concentrated sulphuric acid and then wash them with water. Dry the apparatus, except the absorption tube, by heating it in a current of dry air. Conditioning o f the apparatus-This procedure must be followed with each fresh ampoule of hydriodic acid. Transfer the contents of a 5-ml.ampoule of hydriodic acid, M.A.R., sp.gr.1.7, through the condenser tube into the digestion flask, and add 0.1 ml of 30 per cent. hypophosphorous acid and 2-5 g of phenol. Connect the condenser by means of the B14 standard glass joint to the scrubbing trap at N, and ensure that the joint is securely fitted. Introduce 4 ml of 25 per cent. sodium acetate solution through the side limb, F, into the scrubbing trap, D, and replace the rubber stopper. Allow cold .water to flow through the condenser. With the absorption tube, G, detached, pass a stream of carbon dioxide or nitrogen through the apparatus at 1 ml per minute, heat the digestion flask and, when the contents begin to boil, increase the flow-rate to 6ml per minute, then continue to boil the digestion mixture for 30 minutes.BZank determination-When the conditioning of the apparatus is complete, add 5 ml of bromine solution to the absorber, then connect the absorber to the apparatus by means of the glass joint and secure it in position with i x suitable spring or rubber band. Boil the digestion mixture and adjust the gas flow-rate first to 1 ml per minute and then to 6x111 per minute, as directed above. After boiling the mixture for 1 hour, remove the absorber, pour the contents into a 250-ml glass-stoppered flask containing 5ml of 25 per cent. sodium acetate solution and wash out the receiver thoroughly with water. Add formic acid dropwise to the solution in the flask until the smell of bromine can no longer be detected.Cool the flask slightly, carefully remove the stopper and add a small quantity of methyl red indicator solution from a fine capillary tube or from the end of a glass thread. If the indicator is decolourised, add a further drop of formic acid. Add 2 ml of 10 per cent. potassium iodide solution and 5 ml Assemble the apparatus as shown in Fig. 1. Insert the stopper in the flask and shake it thoroughly.Nov., 19541 THE DETERMINATION OF METHOXYL GROUPS 679 of 10 per cent. sulphuric acid, shake the flask and allow it to stand for 10 minutes. Titrate the liberated iodine with 0.01 N sodium thiosulphate, 1 per cent. starch solution being used as indicator. The volume of 0.01 N sodium thiosulphate required should not exceed 0.2 ml. The conditioned apparatus can be used for seven or eight determinations of methoxyl groups without renewal of the hydriodic acid.The 25 per cent. sodium acetate solution in the trap, D, should be changed when the hydriodic acid is renewed. Control determination-Before the apparatus is used for routine determinations, control determinations should be carried out with a standard substance. A suitable standard substance is a-methyl-D-mannoside or a-methyl-D-glucoside. When the determined per- centage agrees within 0.1 of that calculated, the apparatus may be regarded as suitable for routine work. Carry out the control determinations as described below. Determination-Take a weight of dried sample in accordance with the following values- Methoxyl content, per cent. . . .. 0 to 10 10 to 26 25 to 50 > 50 Weight of dried sample, mg . . .. 5 to 8 4 to 6 3 to 4 2 to 3 Accurately weighing spoon weigh the appropriate amount of the dried sample in the tared long-handled or weighing bottle, shown in Fig. 2. Transfer the weighing vessel with the Fig, 2. (a) Weighing spoon; (b) weighing bottle sample to the digestion flask of the conditioned apparatus. Allow the weighing vessel to remain in the reaction flask during the test, as this serves to promote even ebullition. When the sample is a liquid of low vapour pressure and is contained in a capillary tube, transfer it to the digestion flask of the conditioned apparatus, crush the capillary tube under the reagent by means of a glass rod and allow the glass rod to remain in the apparatus during the determination, If the sample has a high vapour pressure and is contained in a capillary tube, transfer it to the digestion flask of the conditioned apparatus and cool the bulb of the flask in ice or solid carbon dioxide; when the bulb has been thoroughly cooled, crush the capillary tube under the reagent by means of a glass rod and allow the glass rod to remain in the apparatus during the determination.B 10 I socket A I I! I 2-mm Fig. 3. Simple bubbler for use as sub- sidiary washer apparatus at N and ensure that the joint is securely fitted. for the blank determination. Connect the condenser immediately by means of the B14 standard glass joint to the Then proceed as described Let xml = volume of 0.01 N sodium thiosulphate solution required in test, y ml = volume of 0.01 N sodium thiosulphate solution required in blank, and W = weight of sample taken, in mg.680 HERON, REED, STAGG AND WATSON Then- [Vol.79 (’ - y ) ”” = percentage of methoxyl. W NOTE- The flow-rate can be increased to 20 ml per minute with safety, provided that a simple bubbler (Fig. 3) containing 25 per cent. sodium acetate solution is inserted between the scrubbing trap, D, and the absorber, G. SEMI-MICRO METHOD REAGENTS- As for the micro method, but with the following modified solutions. Bromine solution-Dissolve 55 g of potassium hydroxide in 100 ml of glacial acetic acid Sodium thiosuZ$hate solutiort, 0.1 N-Standardise as described for the 0.01 N , but weigh and add 1-5 ml of bromine, all of analytical-reagent grade. accurately 30 to 35 mg of potassium iodate into the conical flask instead of 3 to 4 mg.PROCEDURE- Blank determination-Proceed as described in the micro method except that 12 ml of the modified bromine solution and 0.1 N instead of 0.01 N sodium thiosulphate should be used. The volume of 0.1 N sodium thiosulphate required should not exceed 0.05 ml. The conditioned apparatus can be used for four determinations (excluding the blank), after which the contents of the digestion flask and the trap should be renewed. Determination-Grind a portion of prepared sample to pass completely a 150-mesh B.S. sieve, and weigh accurately an amount of thi:; material equivalent to not more than 10 ml of 0.1 N sodium thiosulphate. Methoxyl content, per cent. . . . . 30 20 10 Weight of sample, mg . . .. .. 15 25 50 Make the determination as described in the micro method except that 12ml of the modified bromine solution and 0.1 N instead of 0.01 N sodium thiosulphate should be used. Let xml = volume of 0.1 N sodium thiosulphate solution required in test, y ml = volume of 0.1 N sodium thjosulphate solution required in blank, and W Then- The following values indicate a suitable weight- = weight of sample taken, in mg. = percentage of methoxyl. (X - y) x 51.7 W NOTE- to 12 ml per minute without effect on the acciiracy of the results. The flow-rate of 6 ml per minute is used a:< in the micro method, but it can be increased REFERENCES 1 . 2. 3. 4. 9. 6. IMPERIAL CHEMICAL INDUSTRIES LTD. Easterbrook, W. C., and Hamilton, J., Anaiyst, 1853, 78, 551. Viebock, F., and Schwappach, A., Ber., 1930, 63, 2818. Kinsman, S., and Noller, C. R., Ind. Eng. Chem., Anal. Ed., 1938, 10, 424. Hoffman, D. O., and Wolfram, M. L., Anal. Chem., 1947, 19, 225. Slater, A., J . Chem. SOC., 1904, 85, 1291. White, E. P., Ind. Eng. Chem., Anal. Ed., 1944, 16, 207. RESEARCH DEPARTMENT BILLINGHAM DIVISION BILLINGHAM, Co. DURHAM and ANALYTICAL LABORATORIES IMPERIAL CHEMICAL INDUSTRIES LTD. HEXAGON HOUSE, BLACKLEY DYESTUFFS DIVISION MANCHESTER, 9 May 13tk, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900671

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Nov., 19541 KAWERAU 681 Electrophoresis of Serum and Urine Proteins on Filter-paper Strips and Agar Jelly with the Bridge Unit BY E. KAWERAU The bridge unit has been adapted for electrophoresis, and the present technique for the electrophoresis of serum proteins on filter-paper has been investigated with special reference to the effect of temperature, composition of the buffer solution and type of filter-paper used. The results have shown that even at high voltages the temperature rise is negligible in this apparatus and that the resolving power of the process depends amongst other factors on the cellulose to protein ratio and the density of the paper. A standard method for known amounts of serum and urine proteins is suggested, and a supplementary method in which agar jelly is used instead of filter-paper and capable of dealing with 1 millilitre of serum is described.THE results of filter-paper electrophoresis of serum and urine proteins in different laboratories should be comparable. The work described was directed to discover sources of artefact and the optimum conditions that will resolve the serum proteins into a maximum number of fractions. Since such high resolving power has not been claimed for most of the present techniques, the apparatus and method are described in detail. As many as eight distinct fractions have separated from some sera. EXPERIMENTAL APPARATUS*- One unit takes only one 2-cm filter-paper strip and, although this appears to be a dis- advantage, it is not so. Conditions for operation are more closely approaching the ideal, e g ., high voltages can be applied with only a negligible rise of temperature, the reserve of buffer solution is large and so electrophoresis can be carried out on many strips in succession, long strips can be used should a procedure demand it, media other than paper can be used and the all-glass construction prevents leaks and facilitates cleaning. The bridge unit has been describedl and the only alteration in design is to the flasks that accommodate the electrodes (see Fig. 1). Fig. 1. The bridge unit FZasks-Side-arms rise steeply from the shoulders of the flasks; they are short and have a B14 socket into which the electrode holder fits. To use the apparatus for chromatography instead of electrophoresis, this side-arm is closed with a stopper.EZectrode holder (Fig. 2)-The object of this separate compartment is to prevent changes of the buffer solution in the immediate vicinity of the electrode from affecting the main part of the buffer solution. The buffer solution enters the holder through seven holes at the bottom of the tube. An extra hole is provided at the neck of the holder to prevent blockage * Made to the author’s specification by Messrs. Quickfit & Quartz, Stone, Staffordshire, England.682 by air. at elevated temperatures, evaporation is negligible. effected by placing a short length of rubber tubing around the electrode stem. KAWERAU : ELECTROPHORESIS OF SERUM AND URINE PROTEINS [Vol. 79 There is no tight closure around the electrodes and, unless operations are conducted Complete closure, however, can be Electrodes-The platinum electrode, shown in Fig.2, consists of a short piece of stout 7 x I t mm Holes Cotton-wool Plug Carbon electrode Electrode holder Piatinurn electrode Fig. 2. Electrodes and electrode holder platinum wire wound around the lower end of a hollow glass stem. The platinum wire pierces the tube and is joined inside it to a copper wire that runs to the top of the electrode where it is fused to a brass cap. The use of these platinum electrodes is essential for work with high voltages. Resin-impregnated non-porous graphite electrodes are satisfactory for paper electrophoresis. PERFORMANCE TESTS- Efect of temperature-A rise of temperature during electrophoresis affects the process in two distinct ways : the improved conductivitly makes the electromigration faster, and the evaporation of solvent from the centre of the strip causes buffer solution to be drawn up from the reservoir and this tends to impede electromigration.2 Substantial rises of tempera- ture have been recorded by Foste? and by all workers who used wide strips or multiple strips of filter-paper in one compartment; it has caused them to adopt some form of cooling device.*4s far as the author is aware, no worker has measured the temperature of the filter- paper. It was therefore decided to make three independent temperature measurements : (i) of the paper, with the aid of a sensitive thermocouple, (ii) of the air inside the apparatus and (iii) of the air close to the apparatus; the last two measurements were made with a sensitive mercury thermometer. The thermocouple was attached to the paper in the following manner.The free end of the thermocouple was coated with a thin layer of Perspex; it was then laid against the dry filter-paper and pressed tightly against it, and then held in position with a short strip of adhesive cellulose tape; another piece of tape was placed in the same position on the back of the filter-paper and two staples were driven through from the front across the thermocouple, so fixing it without making contact. The three temperature recordings are shown graphically in Fig. 3. A potential of 600 volts for this test was supplied by coupled high-tension dry batteries. There is a steady rise of temperature during the first 3 hours, but then the temperature remains constant.Comparison of the three curves indicates that the small amount of heat generated in the filter-paper has a steadying effect on the temperature both of the filter-paper and of the air inside the apparatus, i.e., it minimises the effect of fluctuations in room temperature. In further experiments, the circulation of air in the apparatus was promoted; this did not prevent the initial rise in temperature, but the rise was reduced and stable conditions were attained sooner. When more forcible methods of cooling were used and the apparatus was in a much colder room, the stabilised temperature was lower. At lower potentials the rise of temperature was correspondingly reduced and at 200 volts the rate of loss of heat was equal to the amount of heat generated.Nov., 19541 ON FILTER-PAPER STRIPS AND AGAR JELLY WITH THE BRIDGE UNIT 683 Efect of the type of filter-Paper-As different filter-papers might have different resolving powers and an abnormal protein band might not be revealed with one paper, whereas it might show up with another and be mistaken for an artefact, the following Whatman filter- papers were tested: Nos.1,5,20, 40, 540,642, 3MM, 100, 120, E12, El3 and E17. To study the effect of the direction of the fibres, the papers were tested cut both with and against the machine direction. All papers were tested with the same pathological serum as well as with sera from different patients. With most thin and acid-washed papers, 0.01 ml of serum was found to be too much and the results A uniform volume of 0.02 ml was used on thick papers.0 0 22 21 20 9 - 19 !! f 18 t- 17 U 0. 5 16 15 FAN ON I ' FAN OfF - \' -t---f I 2 3 4 5 6 7 8 9 10 I I 12 Time, hours Fig. 3. Heat development during electrophoresis a t 600 volts, on Whatman No. 100 filter-paper, with a barbitone buffer of pH 8-6 Continuous line : paper temperature Broken line: room temperature close to the unit Dotted line : air temperature inside apparatus were streaky. But when 0.01 ml of serum was used, protein fractions of low concentration did not show on the thin papers. For this reason, only the results with the thicker papers are summarised in Table I and Fig. 4. It will be noted that the resolving power of the paper depends both on its thickness and density, increasing in direct proportion with the former and inversely with the latter.The filter-paper No. El7 is nearly as thick as cardboard, its high resolving power being due to the loose structure and great thickness. The electro-endosmotic flow is comparatively large and results in a good separation of the y-globulin fractions. Table I shows that the resolving power of the papers increases as the thickness increases. Electrophoresis carried out at right angles to the main fibre direction did not give such a good over-all separation and it caused the protein bands to be more compact in appearance, but there was a concomittant loss of resolving power of closely associated proteins. Resolving power and over-all separation are greatest with papers cut in the machine direction, but there is a tendency for the protein to go to the edge of the paper under the directional influence of the fibres.Efect ofpH-The pH of the barbitone buffer was checked after each experiment by means of the Cambridge. glass-electrode pH meter. After the first three successive experi- ments with the current flowing in the same direction, the difference in pH between the two electrodes for any subsequent experiment over a 16 to 20-hour period is as much as 0.1 to 0.2 of a unit of pH. It is therefore of the greatest importance to reverse the polarity of the684 KAWERAU: ELECTROPHORESIS OF SERUM AND URINE PROTEINS [Vol. 79 electrodes after each experiment. To study the effect of unbalanced pH, a barbitone buffer of identical molarity but with a pH of 8.4 was tested against the buffer of pH 8.6, being placed in turn in the anode and in the cathode flask.In these unbalanced systems the y-globulins behave as in the normal balanced buffer but the remainder of the pattern is altered. With the more acid buffer in the cathode flask there is less over-all separation, this being most noticeable in the lack of separation between the y- and /?-globulins. When the more Grade number E l 7 El3 E l 2 120 3MM 100 20 5 40 TABLE I PHYSICAL PROPERTIES OF WHATMAN FILTER-PAPERS USED IN THE ELECTROPHORETIC ANALYSIS OF SERUM PROTEINS Apparent Fibre Weight, density, development Handling g per sq. cm g per cu. cm 4.78 0.51 high fragile when wet; tends t o crack and split 4.35 0-51 high similar t o E17; splitting less marked 2.05 0.48 high soft surface : slight tendency to split 2-90 0.72 high smooth surface ; does not split; good wet strength 1.75 0.56 high good wet strength; slight tendency to split 1.90 0-54 low t o smooth surface; 0.94 0.63 high good wet strength 1.01 0.49 high excellent wet 0.93 0-46 medium excellent wet medium good wet strength strength strength Resolving power excellent, particu- larly in the y- globulin region very good, but not as high as E l 7 very good, but not qu&e as good as E l 3 good good good fair (too thin for fair (but too thin) fair (too thin) 0.02 ml of serum) The values for weight, apparent density and fibre development were kindly supplied by Mr.J. N. Balston of the Whatman Paper Mills and are gratefully acknowledged. acid buffer is in the anode flask, the gap between. the protein fractions widens, this being particularly noticeable between albumin and a,-globulin.In a balanced system, however, the buffer of pH 8.4 appeared to be as effective as the standard buffer of pH 8.6. Efect of the volume of bufer solution-If the buffer solution is not exactly level in both flasks, it will cross the bridge by capillarity until both levels are equal. Hence a strip of thick paper left in the unit overnight before the experiment ensures perfect levelling. However, slight differences in height were found to have little significance; for example, after starting electrophoresis, as much as 100ml of buffer solution had to be removed from one flask before any strong effect on the final pattern could be noticed. METHOD FOR SERUM ON FILTER-PAPER The method, although well known, is described to indicate modifications.APPARATUS- Three bridge units have been found sufficient, in our experience, to meet normal hospital demands. Filter-paper-For routine work use either Whatman No. 100, 120 or 3MM filter-paper cut along the machine direction into strips 2 cm wide. The E series of papers can be used if they are carefully handled and allowance is made for their higher resolving power in the interpretation of results. Only a very short tongue of filter-paper should dip into the buffer solution. A buffer volume of 350ml takes a 56-cm strip when the centre section of the bridge consists of three adaptors. Bufleer solution-The barbitone buffer is made by dissolving 10-3 g of sodium diethyl- barbiturate and 1.84g of diethylbarbituric acid in 1 litre of distilled water.This bufferNOV., 19541 ON FILTER-PAPER STRIPS AND AGAR JELLY WITH THE BRIDGE UNIT 685 solution is to be preferred to the Michaelis buffer solution used by Grassmann and Hannig,” as freedom from chloride ions is essential when agar is used. Staining reugent-Purified naphthalene black 12B 200 is made up as a saturated solution in acetone containing 10 per cent. of acetic acid. Staining with naphthalene black has been criticised as it is difficult to remove from the cellulose.6 With the present procedure, complete removal of the dye is ensured. Naphthalene black has also been criticised owing to the impurities frequently encountered in the commercial product. Red, slate black and green pigments are found and they contribute to the protein staining in various proportion.Acetone is not nearly as good a solvent for these impurities as methanol and in most instances it will be found unnecessary to purify the commercial dye. But if the dyed strip shows any tint other than a pure blue, it means that the commercial product is impure and it is not sufficient to recrystallise it from water. The dye must be purified by isolating its free acid as follows. Dissolve 10 g of the dye (sodium salt of P-nitrobenzeneazo-3 :6-disulpho-l-amino-8- naphtholazobenzene) in 900 ml of water and add 100 ml of concentrated hydrochloric acid to precipitate the free acid. Remove the precipitate by filtration with suction on a large Buchner funnel and wash the precipitate with 1 litre of 10 per cent. v/v hydrochloric acid.This removes most of the red contaminant. Continue to wash the precipitate with ice-cold distilled water until the filtrate changes from almost colourless to a deep blue. Then place the precipitate in a beaker and add water to make a thin suspension. By careful titration with sodium hydroxide, convert the acid to the sodium salt. When a pH meter indicates a stable pH of 6.5, heat the suspension to boiling, cool it, re-check the pH and add more alkali if necessary, If the solution is at a pH value lower than 6.8, there will be no free alkali in the final product. This is important, because naphthalene black decomposes in the presence of free alkali. When a pH value of between 6-5 and 6-8 has been reached, transfer the suspension to an evaporating dish and evaporate it to dryness on a boiling water-bath. The staining solution can be made from the dry powder by dissolving it in acetone without further recrystallisation.Washing soZution-A mixture of cheap denatured methanol, water and N hydrochloric acid in the proportion of 90 : 10 : 1 by volume is used. Denatured methanol alone is used for the final wash. Impregnating &z’d-Pure benzyl alcohol, n D 1-5396, is used. Latners suggested methyl salicylate, nD 1.535 to 1.538, and Grassmann, Hannig and Knede17 a mixture of a-bromo- naphthalene and liquid paraffin, nD 1-51, as impregnating fluid; the former has the disadvantage of having a clinging smell and the latter mixture can only be removed by washing the paper in ether. Benzyl alcohol can be removed by warming at 105” C or by leaving for some time at room temperature.PROCEDURE- Serum free from haemolysis should be used, preferably without delay, although little change has been observed in serum that has been stored in a refrigerator for a few days. Plasma is not suitable because the fibrinogen obscures the 7,-globulin ; serum prepared from recalcified plasma is also unsuitable because the precipitation of fibrinogen may be incomplete. Before carrying out electrophoresis, the total protein content of the serum should be deter- mined by the biuret or Kjeldahl methods. The work with different filter-papers has shown that there is an optimum cellulose to protein ratio for electrophoresis. Fig. 4 shows the variation in resolution with different filter-papers, and Fig.6 shows the similar effect produced by varying the amount of serum protein when the same type of filter-paper is used. For Whatman No. 100 filter-paper this ideal amount is about 0.7 mg of protein, or 0.01 ml of a serum containing 7 per cent. of protein. As the optimum ratio is not critically defined, sera of higher protein content need only approximate adjustment to come within this limit. For Whatman No. 100 filter- paper, 0.2 ml of serum is mixed with buffer solution according to the protein concentration as follows- Concentration of serum protein (f0-5%), per cent. 7 8 9 10 11 Then 0-01 ml of the mixed serum and buffer solution is applied to the filter-paper. When the concentration of protein in the serum is less than 6-5 per cent., larger volumes of undiluted serum should be used, e.g., 0.014 ml of a serum containing 5 per cent.of protein. Volume of buffer solution, ml . . .. . . 0.0 0.025 0.050 0.075 0.100686 KAWERAU ELECTROPHORESIS OF SERUM AND URINE PROTEINS [Vol. 79 If the results are to be uniform it is important that the serum is placed on the exact centre of the filter-paper. The calculated volume of serum is applied to the centre of the filter-paper strip with a micro pipette held vertically and drawn evenly across the filter-paper to within 2 mm of its edge. The strip is now drawn through some of the buffer solution in a watch-glass, up to the serum area, excess of buffer being removed immediately by blotting with dry filter-paper. Fill the flasks with 350ml of buffer solution and introduce the strip through the slot in the T-piece.Thread on the adaptors and introduce the other end of the strip through the slot of the T-piece into the second flask to complete the bridge. The double cone adaptor in the centre of the bridge assists in placing the strip in the centre of the whole apparatus (see Fig. 1). Equilibrium is reached within 10 minutes; then connect the source of current-a potential of 200 volts ensures a smooth direct current of approximately 0.2 mA per cm of Whatman No. 100 filter-paper. Electromigratioii for 16 to 18 hours gives an albumin - globulin separation of 15 cm. Reverse the polarity of the electrodes after each experiment ; after ten experiments, recombine and redistribute the buffer. After removal from the apparatus, dry the strip in a horizontal position at 105" C for 20 minutes.The stain is a t a maximum within 20 minutes, and the strip is then washed. Three to four changes of wash fluid produce a completely white background. If photo-electric scanning is used for the quantitative estimation of the protein fractions, the strip must be rendered translucent by impregnation with benzyl alcohol in a high vacuum provided by an oil-pump. The density values are plotted on graph paper, as shown in Fig. 9. For details concerning the quantitative' evaluation of these curves, the reader is referred to work of Grassmann, Hannig and Knedel' and Crook, Harris, Hassan and Warren.8 Keeping the extreme ends of the strip dry facilitates handling. Transfer the hot strip immediately to the staining bath. METHOD FOR URINE ON FILTER-PAPER Urinary and cerebro-spinal fluid proteins cannot be submitted to electrophoresis without Dialysis is time-consuming and involves the risk of bacterial Precipitation of protein is the quickest procedure and is precise. The least preliminary concentration.decomposition. harmful precipitant was found to be the Folin tungstate - sulphuric acid reagent. PROCEDURE- Estimate the protein in a diluted sample of the urine by the turbidimetric salicylsulphonic acid procedure used for cerebro-spinal fluid protein^.^ Take a volume of urine containing 50 mg of protein and add an equal volume of distilled water. If this diluted urine is alkaline, then make it just acid, with litmus paper as indicator, by the addition of a few drops of dilute acetic acid.To every 10 ml of this mixture add 1.0 ml of 10 per cent. sodium tungstate and, after stirring, an equal volume of 0.66 N sulphuric acid. Allow the mixture to stand for 10 minutes then centrifuge it in a 250-ml centrifuge tube at 3000r.p.m. and 20cm radius for 10 minutes. Decant the supernatant liquid and invert the tube to drain. Re-dissolve the bulky protein precipitate in 0.1 ml of N sodium hydroxide. Another 0.1 ml of N sodium hydroxide may be added if there is still much undissolved material after thorough stirring. Any undissolved material remaining after the second addition of alkali is mucoprotein, and it should be removed by centrifuging. When large volumes of urine have been used, a large amount of mucoprotein may be present; some of it dissolves and appears on the diagram in the position of the y,-globulin (Fig.6, l b ) . The dissolved protein is made up to a volume of 2.5 ml with the barbitone buffer. Apply 0.02 ml of this solution to the filter-paper in the manner described and, in addition, apply 0.01 ml to the reverse side of the paper in exactly the same place. The amount of protein examined is equivalent to a 6 per cent. concentration of serum protein. DISCUSSION- Absence of trailing of the protein and the fact that the mobilities of the individual fractions appear to be identical with those found :€or the corresponding serum proteins (see Fig. 6) shows that the precipitant does not materially alter the protein. The electrophoresis diagrams of urine proteins determined by Slater and Kunkel,lo who used dialysis, showed Filter the urine through tightly packed moist glass-wool.1 No.El7 No. E l 3 No. El2 No. 120 No. 3MM No. 100 I Fig. 4. Resolving power of different Whatman filter-papers. All the strips received the same quantity, 0.02 ml, of a pathological serum Fig. 5. Effect of various amounts of protein on Whatman KO. 100 filter-paper. Strip (a) has twice the amount of protein as ( b ) , which has the correct amount. Note the resolution of the 8-globulin fraction and the pencil marks at the edge of the paper, where the protein was applied. The albumin band is on t!ie extreme left and the other bands are in the same order as in Fig. 4I l a I b 2 2a 3 3a 4 Fig. 6. Electrophoresis of serum and urine proteins: 1, the patient’s serum protein pattern; la, a small amount of serum from this patient was added to a normal urine and allowed to stand for 30 minutes and the protein was then recovered by the method described: insufficient protein was talcen and the C I ~ - and 8-globulin bands were just perceptible to the naked eye but cannot be seen in the figure.The albumin and y-globulins occupy the same position as in 1 ; l b , 300 ml of the patient’s urine were used and the band is due to y-globulin (augmented by mucoprotein); 2, pattern of the serum proteins from a case of acute nephritis; 2an, the same patient’s urine proteins; 3, pattern of the serum proteins from a case of chronic nephritis: note the poor albumin band and deficient /%globulin band; 3a, the same patient’s urine proteins: note the large albumin fraction and significant amount of P-globulin; 4, a case of “plain” albuminuriat Fig.8. The bridge unit filled with agar jelly. The agar is hot and translucent, and so the position of the glass rod can be seenStart 2 hours 4 hours 6 hours 8 hours 10 hours Fig. 9. Analysis by the paper-strip method of serum protein fractions obtained by continuous electroplioresis in agar jellyNOV., 19541 ON FILTER-PAPER STRIPS AND AGAR JELLY WITH THE BRIDGE UNIT 687 considerable trailing. By dialysis higher concentration of protein can be attained than by the method described in this paper; but for clinical use the information found by this rapid quantitative method outweighs the minor disadvantage. METHOD FOR SERUM ON AGAR JELLY A glass rod can be introduced through the ventilator hole in the ground joint of one of the adaptors to form a compartment in the agar while it is setting (see Fig.8); this compartment holds about 1 ml of fluid (serum). During electrophoresis samples can be withdrawn without interrupting the process. The whole centre section of the bridge can be filled with agar jelly. MATERIALS AND APPARATUS- pack capable of supplying 400 volts d.c. at a current of 50 to 100 mA is required. flasks to the bridge. shown in Fig. 7. No additional apparatus is required, but platinum electrodes are essential. A power Agar-Finest quality New Zealand agar was used as a 2 per cent. solution in water. FiZter-paper-This is required to act as a lead for the current from the buffer in the It is cut from Whatman No. 120 paper in self-retaining strips, as A bundle (9, 10 or 11) of these strips is pushed through the slot of the T-piece to make a tight fit. .I I I E -; " I ?t_, - f I I I I I I I I I I I I 20 cm I I I I I I I I I I I I I I 8 I I 1 I I Fig. 7. Self-retaining filter-paper strip The bufer solution-The same barbitone buffer solution as before. Staining and washing reagents-The same reagents that have been described can be used, but each should be diluted with an equal volume of water. PROCEDURE- Inorganic ions must be reduced to a minimum in the agar before it can be used for this work. Weigh 4 g of powdered agar and put it into a 250-ml beaker that has been marked at the 200-ml level, fill the beaker with distilled water and allow the agar to soak overnight. The next morning wash the agar on a large sintered-glass filter with distilled water without applying suction.Continue this operation until the filtrate does not give a reaction for sulphate or chloride ions; then return the agar to the beaker and add 2.06 g of sodium diethylbarbiturate, 0.368 g of diethylbarbituric acid and distilled water to the 200-ml mark. Immerse the beaker in a bath of boiling water to dissolve the agar and buffer salts. Fill the flasks with buffer as in the filter-paper method, set the T-piece on one flask and push the bundle of self- retaining filter-papers through the slot to make a tight fit, and then moisten their protruding Purify the agar as follows. Assemble the bridge unit to receive the hot agar.688 KAWE RAU [Vol. 79 tongues with buffer.Next set the three assembled adaptors on to the end of this T-piece and fix the apparatus in a tilted position with the aid of a circular piece of wood and a clamp, as shown in Fig. 8. A glass rod is inserted three-quarters of the way through the hole in the ground joint between the second and third adaptor and fixed in position by a short sleeve of rubber tubing. When the agar has dissolved, cool it to approximately 60" C, then pour the agar into the bridge and fill it completely. While the agar is still hot and fluid, push the second 1-piece, which is blocked with filter-paper, into position. The displaced agar overflows, as shown in Fig. 8. Allow the agar to cool and set with the bridge unit in the tilted position. A small amount of air is drawn in as the agar contracts and these bubbles tend to collect above the lip of the second T-piece where they do not interfere.When the agar is cold, lower the apparatus to the horizontal and set it on the second flask. The glass rod is now removed from the agar by careful manipulation to avoid injury to the agar. The opening in the glass is closed with a strip of adhesive cellulose tape during operation. Make the electrode farthest away from the compartment the positive pole. The result of electrophoresis of 1.0 ml of serum in agar for 10 hours a t 400 volts is shown in Fig. 9. A small amount of serum (0-02 ml) was removed every 2 hours and examined by the paper method. The results show that the fastest fractions, the albumin and yz- globulin, disappear first, and no albumin remains after electrophoresis for 8 hours.The slow fractions or those whose mobility is equal to the endosmotic flow remain longest in the compartment. As there is a steady fall in volume in the compartment, there may be an apparent increase in the concentration of these latter fractions. This procedure has been used to confirm the presence of abnormal protein fractions in pathological sera; for example, it could be shown that the 7,-fraction shown in Fig. 4 was not an artefact but a separate entity, as it possessed a mobility in agar that differed from that of either its neighbouring p- or 7,-fractions. Recovery of material from the agar-The whole agar cast can be extruded complete by the following method. Detach the T-pieces with their filter-paper tongues from the agar by gentle rotation at the joint, some to-and-fro movement and a sudden pull. This leaves the three agar-filled adaptors. With sufficient cooling the agar shrinks so that it can be pushed out in one piece. Cut the agar into pieces according to the position of the material and freeze it. When the agar thaws, it loses its structure and becomes like a sponge; the fluid that (drips out can be collected. The material can then be further purified by dialysis and finally stored after freeze-drying. A hole is so formed in the agar and it should have a capacity of 1 ml. Place this tubular section in a refrigerator. The author wishes to record his thanks to the planning department of Messrs. Quickfit & Quartz for assistance with the design of the electrodes and to Messrs. W. & R. Balston Ltd., for a generous supply of specially cut filter-paper strips of various grades and specifications. The work was greatly helped by advice and encouragement from Dr. G. Roche Lynch, O.B.E., whose kind assistance with the manuscript is also gratefully acknowledged. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES ' Kawerau, E., Biochem. J., 1951, 48, 281. Durrum, E. L., Science, 1951, 113, 66. Foster, A. B., Chem. G. Ind., 1952, 1050. Grassmann, W., and Hannig, K., Hoppe-Seyl. Z., 1952, 290, 1. Loeffler, W., and Wunderly, C., J . Clin. Path., 1953, 6, 282. Latner, A. L., Biochem. J . , 1952, 52, xxxii. Grassmann, W., Hannig, I<., and Knedel, M., Dtsch. med. Wschr., 1951, 76, 333. Crook, E. M., Harris, H., Hassan, F., and Warren, F. L., Biochern. J . , 1954, 56, 434. King, E. J., and Haslewood, G. A. D., Lancet, 1936, ii, 1153. Slater, R. J., and Kunkel, H. G., J . Lab. Clin. &red., 1953, 41, 619. DEPARTMENT OF CHEMICAL PATHOLOGY LONDON, W.2 ST. MARY'S HOSPITAL April 21st, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900681

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Nov., 19541 HASLAM AND SQUIRRELL 689 Observations on an Automatic Titrimeter BY J. HASLAM AND D. C. M. SQUIRRELL An account is given of experimental work on the performance of a modern prototype automatic titrimeter for general analytical work. Titra- tions that have been made with this titrimeter include those of sulphuric, phosphoric and boric acids with standard alkali, of chlorides, bromides and iodides with standard silver nitrate solution, and of iodine, liberated from potassium iodide by potassium permanganate, with standard sodium thio- sulphate solution. Observations are made on the use of the titrimeter in the maintenance of a definite pH during the titration of fluorides with standard thorium nitrate solution. WE have recently had an opportunity of examining in the laboratory an automatic titrimeter.Our main consideration has not been that of the electronic circuits of the particular apparatus, but rather its use and adaptation to the analytical problems arising in the laboratory. The prototype instrument, shown in Fig. 1, was manufactured by Messrs. Electronic Instruments Limited, of Richmond, and consists essentially of- (2) A burette system, which is connected to a solenoid-operated valve that has, in effect, three operational positions in which (a) the burette delivers liquid at the maximum rate, (b) the burette delivers liquid a t a slow drop-rate and (c) the flow of liquid from the burette is stopped altogether. (zi) The instrument, which is connected to the solenoid-valve burette tap. It is so designed that by use of the appropriate controls, titration of, for example, a known amount of acid can be carried out to a definite pH value with standard alkali delivered from the burette.The controls are such that most of the alkali required in the titration is added from the burette at the maximum rate to a pre-arranged point that is approximately the end-point. Finally, the rest of the alkali required in the titration is added at a slow drop-rate, the instrument being so designed that at the appropriate pH value chosen for the end-point, the flow of liquid from the burette automatically ceases. For example, when acid is titrated with alkali to pH 7, it is possible to use the controls in such a way that the slow drop-rate com- mences a t about pH 4 or 5. As titration to a definite pH value is, in effect, titration to a definite potential difference between sources of variable and constant potential, the instrument is capable of dealing with titrations involving other e.m.f.differences, for example, the titration of reducing agents with oxidising agents and the titration of halides with standard silver solution. We have made preliminary experiments on the use of this apparatus for the titration of standard sulphuric, phosphoric and boric acids with standard sodium hydroxide solution, of chloride, bromide and iodide with standard silver solution, and of iodine, liberated by interaction of potassium pennanganate and potassium iodide, with standard sodium thio- sulphate solution. Further, we have used the titrimeter for the automatic maintenance of pH at a definite value during the titration of fluoride with thorium nitrate solution. The particular burette system that we used was such that at the fast flow-rate it delivered 25 ml of liquid in 68-5 seconds and 50 ml in 165 seconds; this compared very well with the time of flow of a 50-ml burette to British Standard 846:1952.The slow drop-rate was such that when filled with liquid to the 25-ml mark, the burette delivered 200 drops, equivalent to 6.6 ml, in 105 seconds. The burette and titrimeter were first assembled as shown schematically in the diagram, and the instrument was used in the following manner for the acid - base titrations. After the instrument had warmed up, the “function” switch was moved to SET ZERO and the indicator needle was adjusted to zero by the “set zero” control.The “function” switch was now moved to SET BUFFER and, with the electrodes immersed in the selected buffer and the instrument dial set to the particular buffer value, i.e., 4-0 or 7.0, the indicator needle was again adjusted to zero with the “set buffer’’ control. This adjustment was repeated to check the settings before the solution to be titrated was placed in position and690 HASLAM AND SQUIRRELL: OBSERVATIONS ON [Vol. 79 the stirrer started. The “function” switch was now moved to the MV FALLING or MV RISING position according to the particular titration and the “fast - slow change-over” control was moved to the desired number of mV before the end-point that the slow drop-rate was to start. The automatic switch was then moved to the FAST position and the titration was completed automatically.- PROTI I MAINS FUSES VOLTAGE SELECTOR % ENGRAVING DETAILS OF TOP PANEL CONNECTION TO BURETTE t TED ELECTRODES Fig. 1. The automatic titrimeter TITRATION OF STANDARD SULPHURIC ACID WITH STANDARD SODIUM HYDROXIDE SOLUTION Standard acid and alkali solutions were prepared so as to be 0.1 N by the ordinary methods of standardisation to indicators given in the I.C.I. scheme of standardisation.1 Titrations of 25-ml aliquots of 0.1 N sulphuric acid solution with 0.1 N sodium hydroxide solution were made in triplicate, the automatic titrimeter being used at different settings of the “fast -slow change-over” control. All titrations were made to a pH setting of 7-0, with an initial volume of 100 ml in the titration vessel.The results are shown in Table I. Titrations of 40-ml aliquots of 0.1 N sulphuric: acid were similarly carried out to pH settings of 7-0 and 4-0, and the results, shown in Table 11, were determined at change-over control settings of 200 and 75mV, respectively.Nov., 19541 AN AUTOMATIC TITRIMETER TABLE I 691 TITRATION Change-over setting, mV 300 200 100 pH setting 7.0 4.0 OF 25-ml ALIQUOTS OF STANDARD SULPHURIC ACID WITH STANDARD SODIUM HYDROXIDE Number of drops Volume of 0.1 N added slowly to pH immediately sodium hydroxide end-point at end-point used, 261 8-35 25.07 269 8-92 25.1 1 265 8.61 25-1 1 16 8-7 1 25-1 1 15 8-62 25.10 16 8.47 25.12 1 9.10 25-2 1 2 9.15 25-20 1 9.20 25.22 ml TABLE I1 VARIATION OF TITRE WITH pH SETTING Volume of 0-1 N sodium hydroxide ml pH at end-point used, 1 { 4qoo:;: Immediately after end-point, pH 8.5, falling in about 1 minute to pH 7-2 pH at end-point remained reasonabJy constant (a) 4.3 (a) 39-95 (b) 4.2 (b) 39.99 EFFECT OF CONCENTRATION OF THE SOLUTION IN THE BURETTE- The effect of the dilution of the alkali on the titration of the same weight of 0.1 N sulphuric acid is shown in Table 111.The titrations were made at a pH setting of 7.0 with the “fast - slow change-over” control set at 200 mV and an initial volume of 100 ml in the titration vessel. The titrations were performed on 20.00 g of the 0.1 N sulphuric acid solution. The pH at the end-point was read after the solution had equilibrated for 1 minute. TABLE 111 TITRATION OF 20.00g OF 0-1 N SULPHURIC ACID WITH SODIUM HYDROXIDE SOLUTIONS OF VARIOUS CONCENTRATIONS Normality of pH value of sodium hydroxide end-point 0.1 0.05 0.025 7.2 7.3 7.35 7.12 7-20 7.12 7-10 7.12 7.20 Volume of alkali used, ml 20.1 1 20.10 20.10 40.2 1 40-20 40.2 1 80-44 80-46 80-40 We have the following observations to make on the results in Tables I, I1 and 111.The instrument gives excellent precision in duplicate titrations under fixed conditions. The use of more dilute solutions in the burette does not appear t o improve the extent of the overshoot in terms of pH or equivalent amount of sodium hydroxide, when titrating to a pH setting of 7.0. The cut-out at the pH setting of 7.0 corresponds to the colour-change end-point of phenol red (pH 6.8 to 8.4). The cut-out at the pH setting of 4-0 corresponds to the colour-change end-point of methyl orange (pH 3.1 to 4.4), which was used as indicator when the standard acid and alkali solutions were standardised.In the above titrations, owing to the size and rate of flow of the drops, the automatic titrimeter overshoots the indicator end-point by about 2 drops.692 HASLAM AND SQUIRRELL : OBSERVATIONS ON [Vol. 79 TITRATION OF STANDARD PHOSPHORIC ACID WITH STANDARD SODIUM HYDROXIDE A solution of phosphoric acid was prepared slo as to contain 9 4 0 g of phosphoric acid per litre as determined by standardisation by titration to sodium dihydrogen phosphate with 0.1 N sodium hydroxide solution, methyl orange (pH 3.1 to 4.4) being used as indicator. Titrations of 20-ml aliquots of the phosphoric acid with 0.1 N sodium hydroxide solution were made in triplicate, the automatic titrimeter being used at different settings of the ‘‘ fast - slow change-over ” control.The titrations were made to a pH setting of 4.5, with an initial volume of 100ml in the titration vessel. The results are shown in Table IV. TITRATION OF 20-ml ALIQUOTS OF STANDARD PHOSPHORIC ACID WITH STANDARD SODIUM HYDROXIDE Number of drops Change-over added slowly to setting, end-point mV 100 50 25 pH at end-point 4-72 4.82 4-77 4.88 4-89 4-85 5.00 4-89 4.75 Volume of 0.1 N sodium hydroxide added, ml 20.30 20.32 20.30 20.30 20.30 20.25 20.32 20-30 20-34 The agreement between titrations is again good and the pH of the end-point is quite stable. TITRATION OF BORIC ACID WITH STANDARD SODIUM HYDROXIDE Then 10 and 20-ml aliquots of this solution were treated with 20ml of 0.1 N sulphuric acid and this acid solution was titrated to a pH setting of 7.1, the automatic instrument being used.Next, 5.0 g of mannitol were added and the titration of the mannitoboric acid was made to a pH setting of 8.3. The difference in the two titres corresponds to the boric acid in solution. The results are shown in Table V. A solution of borax, Na,B,O,.lOH,O, containing 3.8143 g per litre, was prepared. TABLE V TITRATION OF BORIC ACID WITH STANDARD SODIUM HYDROXIDE Volume of pH a t Volume of 0.1 N Theoretical borax solution, pH setting end-point sodium hydroxide, Difference difference ml ml 10 10 20 7-1 7.8 8.3 9.5 7.1 8.1 8.3 9.6 7-1 7.5 8.3 9.3 4.0 4-0 8.0 Although the difference between the two titrations to pH 7.1 and 8.3 is near the theoretical value, this is probably due to the fact that both titrations are overshot by the same volume.The apparatus could no doubt be used for such applications by suitable adjustments of the pH setting and by use of an improved tap, so that there is less chance of overshooting the end-point at each set pH value. TITRATION OF FLUORIDES WITH STANDARD THORIUM NITRATE SOLUTION As we have indicated previously,2 we are interested in the determination of fluorine in organic compounds. Our method involves fusion of the compound with sodium peroxide followed by distillation of the fusion product with perchloric acid and subsequent titration of the fluoride in the distillate with standard tho:rium solution, after suitable adjustment of the pH of the test solution.Nov., 19541 AN AUTOMATIC TITRIMETER 693 During the thorium titration, which is carried out at a pH value of 3, the solution tends to become acid as the thorium solution is added and it is necessary to add sodium hydroxide to maintain the pH at 3 whilst the thorium solution is still being added.In the past we have had to keep the pH constant manually, but we find that with the automatic titrimeter the operation can be performed automatically. We have tried the test under ordinary experi- mental conditions and, with the instrument set at a pH of 3, we found that when the titration solution attained a pH of 2.9, automatic slow addition of sodium hydroxide solution occurred until the pH was re-adjusted to 3.0.Hence, by means of the automatic instrument it was possible to maintain the pH of the solution being titrated automatically so that it never dropped below 2-9. This seemed to us to be extremely useful as, with this automatic arrange- ment, we were able to concentrate on the real purpose of the titration, i.e., the determination of the colorimetric end-point when all the fluoride had reacted with the thorium nitrate solution. The details of the titration and the instrument adjustments are as follows. The instrument was first adjusted to zero and standardised to a known pH of 4.0, as described above. The solution to be titrated, i.e., the distillate containing the fluoride, was placed in position in the apparatus and the burette was filled with 0.1 N sodium hydroxide solution.The stirrer was started and, with the instrument used simply as a pH meter, the pH of the solution was adjusted to 6.0 by the addition of N sodium hydroxide from a second burette. The pH was now adjusted to 3.3 by means of 0.1 N hydrochloric acid added from a third burette, and then 1 ml of alizarin red S indicator was added. The dial of the instrument was set at pH 3.0 and the “function” switch was turned to pH RISING. With the “fast - slow change-over” control set at zero, the automatic switch was moved to the SLOW position. The titration of the fluoride with 0.05 N thorium nitrate solution from a fourth burette was now started and continued until the pH of the solution fell below 3, which automatically started the dropwise flow of the 0.1 N sodium hydroxide from the automatic burette.The addition of thorium solution was now stopped and not recommenced until the automatic instrument had re-adjusted the pH to 3.0, i.e., after the addition of 8 to 10 drops of the 0.1 N sodium hydroxide solution. This procedure was continued until the end-point was shown by the colour change of the indicator. This mode of operation will result in a considerable saving of time. TITRATION OF CHLORIDES WITH STANDARD SILVER NITRATE SOLUTION For several years the rapid determination of chlorine in organic compounds has been extremely important in this laboratory. It is of particular value in the examination of polyvinyl chloride and related copolymers. The method that has been used successfully over the past eight years is that of Haslam and S ~ p p e t .~ By this method the sample is fused with sodium peroxide and a catalyst in a stainless-steel bomb and the fusion product is dissolved in water. Excess of peroxide is removed and the solution is finally adjusted to a definite acidity with nitric acid before titration. In this titration a silver wire, in contact with a definite concentration of silver ions, is used as a source of constant potential and another silver wire, in contact with the solution being titrated, as the source of variable potential. With a known amount of water containing a definite amount of nitric acid, the potential difference between the two electrodes is balanced to give a zero reading on the galvanometer. After this adjustment, the dilute nitric acid is replaced by the solution of the fusion products. This chloride solution is now titrated with standard silver nitrate solution to the same zero reading on the galvanometer.Precautions are taken t o wash the electrodes with water just before the final end-point is reached. Although this method has always been satisfactory, particular care has to be exercised so that the end-point is not overshot when the amount of chlorine in a sample is unknown. This is liable to occur when dealing with all kinds of polymers and compounds containing chlorine. It is true that such difficulties may be avoided after overshooting the end-point by the addition of a known amount of standard sodium chloride solution before completing the titration. Nevertheless, it has always been considered desirable to keep the method as simple as possible and to reduce the number of standard solutions to the minimum.It seemed to us that it would be possible to use the automatic titrimeter in such a way as to avoid all difficulties of this kind even on absolutely unknown samples. Accordingly, preliminary experiments were undertaken, as a result of which the following procedure was devised for use with the same electrode system.3694 HASLAM AND SQUIRRELL : OBSERVATIONS ON [Vol. 79 A weighed sample of AnalaR sodium chloride was fused with sodium peroxide and a catalyst in a stainless-steel bomb, and the fusion product was prepared in the usual manner for analysis with the addition of 1.5 ml of 2 N ni.tric acid in excess. The instrument was then adjusted to zero in the following manner.A solution consisting of 150ml of distilled water and 1.5 ml of 2 N nitric acid was placed in position. The stirrer was switched off, the function switch turned to SET ZERO and the “set zero” control adjusted until the needle was at zero. The dilute nitric acid solution was now withdrawn and the sample solution, of total volume of about 150 ml, was put in position. With the dial set to -50 mV and the “fast - slow change-over” control set at 30 mV, the automatic titration was started by moving the automatic switch to FAST. When this selected potential difference, -50 mV, was reached, the flow of silver nitrate solution stopped automatically and the control switch was moved to OFF, the stirrer was stopped and the electrodes were withdrawn and washed.After they had been put into position again, the stirrer was :re-started, the dial was set to zero and the automatic switch wasput on SLOW. The titration was now completed automatically to the zero potential and the burette reading was taken. The final potential difference was measured to find the amount by which the end-point had been overshot in terms of mV. The procedure was tested on amounts of AnalaR sodium chloride unknown to the operator at the time of test. All these samples were fused with sodium peroxide. The silver nitrate solution was standardised against known weights of the AnalaR sodium chloride, which had been submitted to the complete fusion process. Appropriate corrections were made for blanks on the reagents. Further, the chlorine was determined by means of the above procedure on a standard sample of polyvinyl chloride, the chlorine content of which had already been determined by the conventional potentiometric pr~cedure.~ The results were as follows- STANDARDISATION- Determination of the blank for the catalyst, sodium peroxide and the bomb (automatic titrimeter on SLOW)- Volume of 0.1 A- silver nitrate Overshooting of Set end-point, used, end-point, mV ml mV 0 0.2 1 0 0 0-18 +9 0 0.1 1 0 The average blank was 0.16 ml of 0.1 N silver nitrate.Titration of known weights of AnalaR sodium chloride- Volume of 0.1 N silver nitrate Amount of Overshooting of Set end-point , used, sodium chloride, end-point, mV ml g mV 0 40.44 0-2362 +I1 0 35.04 0.2049 + 8 Then 1 ml of silver nitrate solution SE 0.005877‘2 g of sodium chloride or 1 ml of silver nitrate solution _= 0.003545 g of chloride.UNKNOWN SAMPLES- Setting to stop for washing electrodes: “fast - slow,’’ 30 mV; wash-point, -50 mV Setting for final end-point: 0 mV Volume of Sodium Sodium Drops to Drops to Potential silver nitrate chloride chloride wash-point end-point at end-point, - blank, found, taken, 32 12 - 41.96 0.2451 0.2443 0.0652 11.23 0.0656 30 13 + 22 35-23 0.2059 0-2057 20.54 0.1200 0.1192 29 12 + 10 23 11 + 16 mV ml g g Standard sample of polyvinyl chloride- Instrument settings as above. Chlorine content Volume of determined by silver nitrate Potential Chlorine Average potentiometric Sample, - blank, at end-point, content, value method, 0,2890 45.92 + 2 56-32 56.35 56.5 g ml mV % % 56-39} 0.301 1 47.90 + 11Nov., 19541 AN AUTOMATIC TITRIMETER 695 In our view this modification represents a considerable improvement and it is now possible to determine chlorine in a large number of polymeric materials and organic compounds. with considerable accuracy in about 1 hour per sample.In the titration with silver nitrate there is no danger of the end-point being overshot. TITRATION OF BROMIDES AND IODIDES WITH STANDARD SILVER NITRATE SOLUTION The performance of the automatic titrimeter was then examined for the titration of aqueous solutions of AnalaR potassium iodide and bromide with silver nitrate solution, the same electrode system and zero adjustment procedure being used as for the titration of chloride described above. From an examination of the potential curves of the titrations, however, the following equivalence points were noted and used for the end-point setting on the dial for automatic titrations.Equivalence point for bromide, -50 mV. Equivalence point -for iodide, -200 mV. BROMIDE TITRATION- Standaydisation- The results were as follows- Setting for wash stop: “fast - slow,” 20mV wash-point, - 150 mV Setting for end-point: -50 mV Drops to Drops to Potential Volume of Potassium wash-point end-point a t end-point, silver nitrate, bromide taken, 1 10 + 80 39.25 0.4058 mV ml g Then 1 ml of silver nitrate solution = 0.01185 g of potassium bromide. Setting for wash stop: “fast - slow,” 30 mV wash-point, - 150 mV Setting for end-point : - 50 mV Drops to Drops to Potential Volume of bromide bromide Unknown samfdes- Potassium Potassium wash-point end-point a t end-point, silver nitrate, found, taken, 0.3598 0-3596 22 19 + 40 43.93 0.5207 0-5216 mV ml g g 7 9 + 70 30.36 IODIDE TITRATION- Standardisation- Setting for wash stop: “fast - slow,” 30 mV Setting for end-point: -200 mV wash-point, -420 mV Potassium Drops to Drops to Potential Volume of iodide wash-point end-point a t end-point, silver nittate, taken, 50 36 - 30 15.44 0.2560 mV ml g Then 1 ml of silver nitrate solution = 0.01658 g of potassium iodide.Settings as above. Drops to Drops to Potential Volume of iodide iodide wash-point end-point a t end-point, silver nitrate, found, taken, 74 41 + 75 21.45 0.3556 0.3525 104 55 - 50 27-60 0.4576 0.4574 Unknown samples- Potassium Potassium mV ml g g OXIDATION - REDUCTION TITRATIONS Finally a method was developed for the use of the titrimeter in the titration of iodine, liberated by the reaction of a known amount of standard potassium permanganate solution696 HASLAM AND SQUIRRELL [Vol.79 with acidified potassium iodide, with sodium t hiosulphate solution. The procedure and results are described below. The electrode system, namely a calomel electrode as reference electrode and platinum foil as indicator electrode, were connected to the terminals in the automatic titrimeter marked REFERENCE and GLASS ELECTRODE, respectively. With the “function” switch on SET ZERO the indicator needle was adjusted to zero with the “set zero” control. A known amount of 0.1 N potassium permanganate solution was placed in a 250-ml beaker and 100ml of 0.3 N sulphuric acid and then 0.3 g of potassium iodide were added.The solution was diluted to about 150 ml with water and stirred in the apparatus for 1 to 2 minutes to allow the potassium iodide to react completely with the potassium permanganate. The “function” switch was now moved to MV FALLING and the following dial and “fast - slow” change-over settings were made. These settings were chosen after a study of the full potential curve of the titration- “Fast - slow change-over,” 50 mV. End-point, +250 mV. With the burette filled with 0.1 N sodium thiosulphate solution, the “automatic” switch was moved to FAST and the titration was completed automatically. The sodium thiosulphate solution was first standardised by titration in the instrument of 25.0 ml of the 0-1 N potassium permanganate solution and then 2 volumes of the solution, unknown to the operator at the time of test, were similarly titrated.The results were as follows- STANDARDISATION- Setting for “fast - slow change-over” : 50 mV Setting for end-point: +250 mV Volume of Volume of 0.1 N potassium Drops to Potential 0.1 N sodium permanganate end-point a t end-point, thiosulphate, taken, mV ml ml 11 + 209 25-12 25.0 10 ‘ + 209 25-23 25.0 ’ Then 1 ml of sodium thiosulphate solution = 0.9931 ml of potassium permanganate solution. UNKNOWN SAMPLES- Settings as above. Volume of Volume of Volume of 0-1 N potassium 0.1 N potassium Drops to Potential 0.1 N sodium permanganate permanganate end-point a t end-point, thiosulp hate, found, taken, 20 + 203 25-7 1 25-54 25.60 23 + 198 35-66 35-41 35-50 mV ml ml ml The above results are quite satisfactory, and it should be realised that it is possible to use the automatic titrimeter for other oxidation - reduction titrations if the instrument is initially set to a zero reading on the dial to correspond with a previously titrated solution. Our purpose in preparing this paper is to d.raw the attention of other analysts to the great possibilities of this type of automatic instrument in general industrial work. Many modifications may be made; for example, valves of different slow drop-rates can be used for different purposes. Our preliminary experiments have convinced us that instruments of this type, perhaps modified, can be used in many ways for more accurate and speedier analytical titrations. We are indebted to Mr. 2. J. Bujwid for hk assistance in these trials. REFERENCES 1. 2. 3. Analytical Chemists’ Committee of Imperial Chemical Industries Limited, Analyst, 1950, 75, 577. Haslam, J., and Soppet, W. W., J . SOC. Chem Ind., 1948, 67, 33. Haslam, J., and Whettem, S. M. A., J . A$@. iShem., 1952, 2, 339. IMPERIAL CHEMICAL INDUSTRIES LIMITED WELWVN GARDEN CITY, HERTS. PLASTICS DIVISION Aflril 12th, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900689

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

Nov., 1954.1 BUSH 697 A Rapid Method for the Determination of Lead in Steel BY G. H. BUSH A rapid method is described for the determination of lead in steel, alloy steel and cast iron. By this method the absorption due to the presence of colloidal lead sulphide in alkaline solution is measured with the Spekker absorptiometer, the interference of iron being prevented by complexing it as the ferrocyanide. The effect of individual constituents in alloy steel is considered briefly. The lead content of a steel can be determined in about 30 minutes on a 0-2-g sample, compared with the 1 to 2 days and 6 to 10-g sample required for the normal methods. IN an investigation into the segregation of lead in leaded steel billets, it became necessary to make large numbers of lead determinations, for which an unpublished method due to A.T. Etheridge was used. Essentially this method consists in dissolving the steel in hydrochloric acid, removing the greater part of the iron as ferric chloride by extraction with ether, con- verting the lead to sulphate and finally determining it as lead molybdate. Although this method is accurate, it is tedious and time-consuming. Other methods for the determination of lead in steel are equally lengthy, generally involving initial precipitation of the lead as sulphide in slightly acid or neutral solution and its conversion to lead sulphate with or without final determination of the lead as lead molybdate. An unpublished method due to B. S. Evans is somewhat shorter and involves precipitation of the lead with ammonium sulphide in alkaline solution after the iron has been complexed with potassium cyanide, the lead being estimated as lead sulphate.In all these methods the amount of sample used is 5 g or more and the time required for a determination is between 1 and 2 days. It was evident that for work in which a large number of determinations are required, such as an exploration of the variation of lead content over the surface of a thin section of a steel billet, a more rapid method suitable for use with a. much smaller sample was required. As lead in low concentrations can be estimated absorptiometrically as the sulphide in alkaline solution, attention was directed to obviating the effects of such elements as iron and manganese in an alkaline solution of a steel when treated with sodium sulphide.I t was found that by complexing the iron as ferrocyanide under controlled conditions and removing any manganese ferrocyanide by filtration, the lead could be determined by comparing the absorption of one half of the resultant solution after the addition of sodium sulphide with that of the remaining half of the solution in which the lead sulphide colour had not been developed. The results by this method have been satisfactory, and a determination of lead in steel, alloy steel and cast iron can be made in about 30 minutes on a 0-2-g sample. By slight modifications the method can be adapted to the determination of lead in steel on the semi- micro scale. The lead content of leaded steels normally lies within the range 0 to 0.5 per cent.and, whilst the method to be described covers this range, the upper limit can be extended con- siderably by a suitable choice of sample weight. EXPERIMENTAL Preliminary experiments were made with a stock solution of ferrous chloride prepared by dissolving 2.5 g of steel in diluted hydrochloric acid (1 + l), removing any carbon by filtration and diluting to a volume of 500 ml with distilled water. A 20-ml aliquot of this solution was placed in a small conical flask, 10 ml of 50 per cent. w/v citric acid were added to prevent subsequent precipitation of iron and then 20 per cent. w/v sodium carbonate was added until the liquid was distinctly alkaline. Next, 2 g of potassium cyanide were added and the solution was boiled gently for 5 minutes. With some samples it was found that the solution became almost colourless, whilst with others it became coloured in various shades of pale yellow, depending on the time of698 BUSH: A RAPID METHOD FOR THE [Vol.79 boiling and, more particularly, on the length of time that the ferrous chloride solution had been prepared before use. The solution was cooled and filtered into a 100-ml calibrated flask, 3 ml of standard lead chloride solution (1 ml = 0-0001 g of lead) were added, and then 5 ml of 5 per cent. w/v sodium sulphide, $after which the solution was diluted to the mark with distilled water. Although the resultant solutions contained the same amount of lead, they varied in colour from dark green to brown, which indicated that the conversion of the iron to ferro- cyanide was incomplete owing to the presence of ferric salts, and the results were disappointing.The effect was more noticeable when the stock solution of ferrous chloride had been left for some time before use, and it was clear that complete reduction of the iron was essential before addition of the cyanide. Whereas B. S. Evans used sulphur dioxide to reduce iron before its complete conversion to ferrocyanide, Scott and Furmanl recommend the use of sodium metabisulphite. In subsequent experiments when lead was added at an early stage and sulphur dioxide used to ensure reduction of the iron, conversion of the iron to ferrocyanide was complete and there was no green colour formed on the addition of sodium sulphide solution ; but results for the determination of lead were erratic.Further investigation indicated that the presence of sulphuric acid forined by oxidation of some of the sulphurous acid had resulted in a reaction with the lead to form the sulphate, which, owing to its insolubility, did not give a colour with sodium sulphide. A similar effect, but even more pronounced, was observed in experiments in which the lead was added at an early stage and was partly removed as lead sulphate during filtration. Therefore it was necessary to reduce the iron with a reagent other than sulphur dioxide, and stannous chloride, titanous chloride, sodium hypophosphite, hydrazine and hydroxylamine hydrochloride were tried, but with little success. Small amounts of iron were not reduced and consequently the final solution was coloured green. It was thought that the use of any reducing agent could be avoided, provided that the iron did not become oxidised at any stage before the addition of potassium cyanide, and this was considered to be possible if air could be excluded from the flask during the preliininary st ages.Accordingly, a 0-2-g sample of steel was placed in a 100-ml conical flask, 8 ml of diluted hydrochloric acid (1 + 1) were added and the flask was fitted with a small Contat - Gockel valve, as shown in Fig. 1. The valve consists of a pear-shaped glass vessel fitted with a short stem connecting with an internal syphon. The stem of the valve is fitted with a rubber stopper of a size suitable for the 100-ml flask. The flask was placed on a hot-plate and a mixture of 15 ml of 50 per cent. w/v sodium citrate and 15ml of 25 per cent.w/v sodium carbonate was added to the valve. After gentle boiling for about 5 minutes, the sample dissolved, and the apparatils was removed from the hot-plate and allowed to cool. As the steam in the flask condensed, the sodium citrate - sodium carbonate solution in the valve was drawn into the flask, a little at a time, and the carbon dioxide generated by the subsequent reaction maintained a non-oxidising atmosphere. As soon as the violent evolution of carbor,: dioxide had subsided, the rubber stopper holding the valve was eased from the flask, the contents of the valve were allowed to drain into the flask and 10 ml of 20 per cent. w/v potassium cyanide were added immediately. After standing for about 5 minutes, the flask and contents were heated to incipient boiling on a hot-plate and then cooled in a bath of cold water.With this procedure, the conversion of the iron to ferrocyanide was complete and the solution after filtering was alinost colourless. In later experiments in which standard lead solution was added to the steel before solution, it was found that after the iron had been complexed, the solution was slightly brown and occasionally slightly turbid, depending on the amount of lead present. This turbidity was due to lead sulphide formed from a small amount of sodium sulphide in the solution; the sodium sulphide was formed by the sodium citrate - sodium carbonate solution in the valve absorbing the hydrogen sulphide evolved during solution of the sample. Hence, to avoid loss of lead, it wcis necessary to decompose the lead sulphide before filtration, and 1 drop of 20-volume hydrogen peroxide was added to the cold solution after the iron had been complexed.A slight yellow colour that developed at this stage could be discharged by the addition of sodium sulphide or sodium dithionite (Na,S,O,) and, in the absence of lead, the addition of either reagent resulted in an almost colourless solution.Nov., 19541 DETERMINATION OF LEAD IN STEEL 699 I n the determination of lead, the solution, after the addition of hydrogen peroxide, was divided into two equal parts, solution A and solution B. Sodium sulphide was added to solution B to produce the lead sulphide colour and a solution of sodium dithionite was added to solution A to discharge any yellow colour without developing any colour due to the lead.The difference in the optical density of a series of solutions A and B containing various amounts of lead was found to be directly proportional to the amount of lead present, provided that the same solutions were used and the same conditions of development observed. With colloidal solutions, however, the optical density depends on the particle size, which in turn depends on the conditions of development, so that a calibration curve prepared with one series of solutions under one set of conditions and resulting in a straight-line relationship may differ from a second curve prepared on a subsequent occasion with different solutions. Fig. 1. Flask fitted with Contat - Gockel valve It was decided that a standard steel of known lead content should be treated as described in the method with every batch of samples to be analysed, the lead content of the samples being determined from the difference in the optical density of solutions A and B by simple proport ion.STABILITY OF THE LEAD SULPHIDE COLOUR- The brown colour of colloidal lead sulphide was found to be stable for about 15 to 20 minutes without the addition of a stabiliser, such as gum acacia or gelatin; under normal conditions, this time is adequate for measurements to be made. To increase the stability, 5 ml of a 1 per cent. w/v solution of gum acacia or gelatin can be added to both the blank and the sample after the addition of sodium dithionite and sodium sulphide, respectively, and before the dilution to the 100-ml mark.The method described below does not include the use of a stabiliser, and its use is left to the discretion of the operator. EFFECT OF ALLOY ELEMENTS IN STEEL- Alloying elements normally found in alloy steels do not develop a colour with sodium sulphide under the conditions of the method, and the base colour imparted to the solution by certain elements affects equally the absorption of the blank and that of the sample and presents no difficulty in the determination of lead, which depends on the difference in the optical density of the two solutions. When a steel containing appreciable amounts of tungsten was treated with the acid used for solution of the sample, the tungsten remained undissolved as a blackish precipitate and was subsequently removed by filtration. It was found that when this precipitate is washed, a diluted solution (5 per cent.v/v) of the sodium citrate - sodium carbonate reagent used i n the valve should replace the distilled water normally used, so as to avoid the tendency of this type of precipitate to pass through the pulp filter.700 BUSH: A RAPID METHOD FOR THE [Vol. 79 To assess the effect of alloy elements on the determination of lead, weighed quantities of the elements were added to 0-2-g samples of a plain carbon steel and the experiments were performed as described below. The equivalent of 0.3 per cent. of lead was added to each solution after the iron had been converted to ferrocyanide. In addition, the determina- tion of lead was performed on 0.2-g samples of certain B.C.S.alloy steels to which were added the equivalent of 0.3 per cent. of lead. The results of these experiments are shown :in Table I. TABLE I DETERMINATION OF LEAD I N THE PRESENCE OF THE ALLOYING ELEMENTS OF STEEL WITH 0.30 PER CENT. OF LEAD ADDED Drum reading 7-- Element added Nickel . . .. . . Chromium . . .. Titanium . . .. Vanadium . . .. Molybdenum . . .. Aluminium . . . . Tungsten . . .. Copper. .. . . Manganese . . . . Tin . . .. .. Cobalt . . .. .. B.C.S. steel No. “W” . . .. .. Cr - V steel “V” No. 165 H.S. steel No. 220 . . Stainless steel No. 209 .. . . .. . . . . . . . . .. .. .. .. Equivalent addition, 3.0 3.5 1.0 1.0 1.0 4.0 4-0 1.0 5-0 0.5 1.0 % Blank. (solution A) 0.890 0.847 0.880 0.866 0.897 0.922 0.913 0-924 0.922 0-921 0.888 Ni 0.44 { $:iiii ] 0.768 CO, 4.76 Cr, 0.86 W, 6.74 Cr* 0’41 } 0.897 “K” { Mo, 0.36 Ni - Cr - Mo cast iron Sample (solution B) 0.510 0.476 0-520 0.602 0.639 0.559 0.555 0.547 0.534 0.567 0-500 Difference (A-B) 0.380 0.37 1 0.360 0.364 0.358 0.363 0.358 0.377 0.388 0.354 0.388 0.394 0.374 0.522 0.462 0.186 0.535 0.388 0.363 0.380 0.362 Lead found, 0.308 0.300 0.291 0-295 0.290 0-294 0-290 0.305 0.312 0.287 0.312 % 0.306 0.318 0-298 0.312 0.297 It will be noticed that the presence of 0.5 per cent.of tin, 5.0 per cent. of manganese, 1.0 per cent. of copper, 4.0 per cent. of tungsten and 4.0 per cent. of aluminium has little effect on the determination, as the blank (solution A) gives approximately the same drum reading as that of a carbon steel (blank (solution A) in Table 11).With elements that produce coloured solutions, the difference between the drum reading for the blank and that for the sample corresponds to the amount of lead present, within the accuracy of the method. By adding known amounts of lead to 0-2-g samples of lead-free steel and then determining lead by the method described below, the amount of lead that could be determined con- veniently was found to be within the range 0 to 005 per cent. The results were as follows- RANGE AND REPRODUCIBILITY OF THE METHOD-- 0.500 Lead added, per cent. . . 0.050 0.150 0.250 Lead found, per cent. . . 0.054 0.149 0.249 0.408 0.508 0.400 Normal requirements do not entail the determination of lead in steel outside this range, but by reducing the weight of sample used, the upper limit of the range can be extended considerably.Nov., 19541 DETERMINATION OF LEAD I N STEEL 701 The reproducibility af the method is indicated by the results shown in Table 11, which were determined for B.C.S.steel No. 212; a lead-free steel to which a known amount of lead had been added was used as standard. TABLE I1 REPRODUCIBILITY OF RESULTS FOR A LEAD-BEARING STEEL (B.c.s. NO. 212) CONTAINING 0.28 Drum reading A r \ Blank Sample (solution A) (solution B) 0.920 0.924 0.919 0-922 0.922 0.924 0.920 0-921 0.919 0,923 0.559 0.566 0.57 1 0-576 0.580 0.571 0.554 0.562 0.561 0.568 PER CENT. OF LEAD Difference (A--B) 0.361 0.368 0-348 0-346 0-342 0.353 0.366 0.359 0.358 0.356 Lead found, 0.286 0.284 0-276 0.274 0.272 0-280 0-290 0.286 0.284 0-282 % Mean lead content = 0.281 per cent.Standard deviation = 0.006 per cent. METHOD REAGENTS- All reagents should be of recognised analytical grade. Hydrochloric acid, diluted (1 + 1). Sodium citrate - sodium carbonate solution for the Contat - Gockel valve-Dissolve 250 g of sodium citrate and 125 g of anhydrous sodium carbonate in distilled water, pass the solution through a pulp filter and dilute it to 1 litre with distilled water. Potassium cyanide solution-A freshly prepared 20 per cent. w/v solution. Hydrogen peroxide, 20-volume. Sodium sulfihide solution-Prepare a 10 per cent. w/v solution and filter it through pulp before use. Sodium dithionite solution-Prepare a fresh 10 per cent. w/v solution as required. Dis- solve 10 g of sodium dithionite (Na,S,O,) in 100 ml of distilled water, boil the solution for about 2 minutes to remove the slight brown colour, then cool and filter it.PROCEDURE- Weigh accurately 0.2 g of sample into a dry 100-ml conical flask and add S ml of diluted hydrochloric acid (1 + 1). Fit the flask with a Contat - Gockel valve, as shown in Fig. 1, place the flask on the cooler part of a hot-plate and add 30 ml of the sodium citrate - sodium carbonate mixture to the valve. Heat the flask gently until the sample dissolves, taking care not to touch it with cold fingers or unduly disturb it so as to avoid the contents of the valve sucking back into the flask. When the sample has dissolved, boil the solution briskly for about 5 seconds and then remove the flask from the hot-plate. After the violent evolution of gas in the flask has subsided, loosen the valve and allow the contents to run into the flask.Add 15ml of 20 per cent. w/v potassium cyanide immediately, with gentle shaking, and allow the solution to stand for 5 minutes. Warm the flask and contents to incipient boiling on a hot-plate, i.e., to about 85" to 90" C, when the liquid will have become pale straw coloured with a slight brown turbidity due to the presence of a small amount of lead sulphide. Cool the solution thoroughly, add 1 drop of 20-volume hydrogen peroxide, allow it to stand for 5 minutes and filter it through a pulp filter into a 100-ml calibrated flask, washing the filter and flask four times with about 50ml of cold water. Dilute to the mark with water and mix the contents thoroughly. Transfer 50 ml of the solution to a second 100-ml calibrated flask and mark the two solutions A and B.To solution B, add 10 ml of 10 per cent. w/v sodium sulphide and to solution A, 5ml of 10 per cent. w/v sodium dithionite (Na,S,O,) and dilute each solution to the mark with water.702 HALES AND KYNASTON: THE PREPARATION OF PRESSED [Vol. 79 Measure the absorption of the two solutions with the Spekker absorptiometer, a 4-cm cell, Ilford No. 601 spectrum violet filter and a water setting of 1.0. Determine the lead content from the difference between the absorption of the blank (solution A) and the sample (solution B) by proportionality from similar figures determined on a standard steel of known lead content. If a spectrophotometer is used, the optical density of solution B is measured relative to solution A in a 4-cm cell at a wavelength of 430 mp. This value for the optical density is proportional to the amount of lead present and is evaluated by reference to a similar figure for a standard steel of known lead content. The reagents should be measured with either a pipette or burette, except the valve solution, for which a measuring cylinder is satisfactory. A standard sample should be examined with every batch of samples. Acknowledgment is made to the Chief Scientist, Ministry of Supply, for permission to publish this paper. REFERENCE 1. 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, Volume I, p. 621. MINISTRY OF SUPPLY ARMAMENT RESEARCH ESTABLISHMENT FORT HALSTEAD, SEVENOAKS, KENT Jusze lst, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900697

出版商:RSC    

年代:1954

数据来源: RSC