ISSN:0003-2654    

DOI:10.1039/AN96287FX045

出版商:RSC    

年代:1962

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96287BX047

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

iv THE; ANALYST jDcceinber, 1962THE SOCIETY FOR ANALYTICAL CHEMISTRYFORMERLY THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTSFOUNDED 1874. INCORPORATED 1907.THE objects of the Society are to encourage, assist and extend the knowledge and study ofanalytical chemistry and of all questions relating to the analysis, nature and compositionof natural and manufactured materials by promoting lectures, demonstrations, discussionsand conferences and by publishing journals, reports and books.The Society includes members of the following classes :-(a) Ordinary Members whoare persons of not less than 21 years of age and who are or have been engaged in analytical,consulting or professional chemistry ; ( b ) Junior Members who are persons between the agesof 18 and 27 years and who are or have been engaged in analytical, consulting or professionalchemistry or bona jide full-time or part-time stuclcnts of chemistry.Each candidate forelection must be proposed by three Ordinary lllcinbers of the Society. If the Council intheir discretion think fit, such sponsorship may bc dispensed with in the case of a candidatenot residing in the United Kingdom. E\,ery application is placed before the Council andthe Council have the power in their absolute discretion to elect candidates or to suspend orreject any application. Subject to the approval of Council, any Junior Member above theage of 21 may become an Ordinary Member if he so wishes. A member ceases to be a JuniorMember on the 31st day of December in -the year in which he attains the age of 27 years.Junior Members may attend all meetings, but are not entitled to vote.The Entrance Fee for Ordinary Members is El 1s.and the Annual Subscription is E3 3s.Junior Members are not required to pay an Entrance Fee and their Annual Subscription isL1 1s. No Entrance Fee is payable by a Junior Member on transferring to Ordinary Member-ship. The Entrance Fee (where applicable) and first year’s Subscription must accompanythe completed Form of Application for Membership. Subscriptions are due on January 1stof each year.Scientific Meetings of the Society arc: usually held in October, November, December,February, April and May, in London, but from time to time meetings are arranged in otherparts of the country.All members of the Society have the privilege of using the Library of The ChemicalSociety. Full details about this facility can be obtained from the Librarian, The ChemicalSociety, Burlington House, Piccadilly, London, W.1.The Analyst, the official organ of the Society, which has a world-wide distribution, isissued monthly to all Ordinary and Junior Members, and contains original papers and notes,information about analytical methods, Government reports, reviews of books and reportsof the proceedings of the Society. In addition, all Ordinary Members receive AnalyticalAbstracts, providing a reliable index to the analytical literature of the world.Forms of application for membership of the Society may be obtained from the Secretary,The Society for Analytical Chemistry, 14 Belgrave Square, London, S.W.l.Notices of all meetings are sent to members by post.LOCAL SECTIONS AND SUBJECT GROUPSTHE North of England, Scottish, Western and Midlands Sections were formed to promote theaims and interests of the Society among the members in those areas. The Microchemistry,Physical Methods and Biological Methods Groups have been formed within the Society tofurther the study of the application of microchemical, physical and biological methods ofanalysis. All members of the Society are eligible for membership of the Groups.The Sections and Groups hold their own meetings from time to time in different places.There is no extra subscription for membership of a Section or Group. Application forregistration as a member should be made to the Secretary

ISSN:0003-2654    

DOI:10.1039/AN96287FP281

出版商:RSC    

年代:1962

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96287BP293

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

DECEMBER, 1962 Val. 87, No. 1041 THE ANALYST PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETING AN Ordinary Meeting of the Society was held at 7 p.m. on Wednesday, December 5th, 1962, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. A. J. Amos, O.B.E., BSc., F.R.I.C. The subject of the meeting was “Applications of X-ray Fluorescence,” and the following papers were presented and discussed : “The Applications of X-ray Fluorescence Spectrometry in the Steel Industry,” by D. F. Sermin, A.Met. ; “The Determination of Lead in Air Filters, Vanadium - Nickel Ratios in Oil Ashes and Strontium in Tap Water by the X-ray Fluorescence Spectrometer,” by R. G. Stone, BSc., A.R.I.C.; “X-ray Fluorescence in Archaeology at the Museum Laboratory,” by E.T. Hall, M.A., D.Phi1. NEW hlEMBERS ORDINARY MEMBERS Attilio Bosticco; Eric Richard Brown, F.P.S., F.R.I.C.; Eric Langley Bush, A.R.I.C. ; Brian Richard Chamberlain; Eric Crowther; Barrie Dawson, B.Sc.(Sheff.) ; Harvey Diehl, B.Sc., Ph.D.(Michigan) ; Archibald Cameron Docherty, B.Sc., Ph.D.(Edin.), F.R.I.C. ; Ronald Jack Hall, F.I.M.L.T. ; Fredi Jakob, B.S.(New York), Ph.D.(Rutgers) ; Ronald Mark Johnson, B.Sc.(Lond.), A.R.I.C. ; Richard Lockyer, B.Sc.(Lond.), F.R.I.C. ; George Newlands, BSc. (Glas.), A.R.I.C. ; Fosco Provvedi, L.C. (Firenze) ; Jacobus Martinus Rellage ; Andrew Robert- son, B.Sc.(Lond.), Michael Lewis Sheppard, A.R.I.C. ; John McPhail Skinner, BSc., Ph.D. (Glas.), F.R.I.C. ; Gordon Howard Smith, B.Sc.(Wales), M.I.Bio1.; Ralph Williamson; Joseph John Wilson, B.Sc.(Lond.), A.R.I.C. ; ;\fatthew Wilson. JUNIOR MEMBERS Thomas Edward Forster; George McGuire, A.H.-W.C. ; John Morton, A.M.Inst.S.P., A.R.S.H. DEA4THS Harold William Christian Mary Corner. WE record with regret the deaths of NORTH OF ENGLAND SECTION A JOINT Meeting of the North of England Section and the Tees-side Section of the Royal Institute of Chemistry was held at 8 p.m. on Monday, November 12th, 1962, at the Con- stantine Technical College, Middlesbrough. The Chair was taken by the Chairman of the Tees-side Section, Dr. D. G. Jones, BSc., A.R.C.S., D.I.C., F.R.I.C. The following paper was presented and discussed : “Solvent Extraction of Inorganic Compounds: Some Recent Developments,” by H.XI. N. H. Irving, &LA., D.Phil., D.Sc., F.R.I.C., L.R.A.M. 92 1922 PROCEEDINGS SCOTTISH SECTION [Vol. 87 A JOINT Meeting of the Scottish Section and the Glasgow Section of the Society of Chemical Industry was held at 6 p.m. on Friday, Novernber 23rd, 1962, in Room 24, Royal College of Science and Technology, George Street, Glasgow, (2.1. The Chair was taken by the Chairman of the Glasgow Section, Professor J. Monteath Robertson, C.B.E., M.A., Ph.D., DSc., F.R.S. The following papers were presented and discussed : “Applications of Cellulose Ethers,” by F. C. Hall, Ph.D., MSc., A.M.I.Chem.E., F.R.I.C. ; “Analysis of Cellulose Ethers,” by A. F. Williams, BSc., F.R.I.C. WESTERS’ SECTION A JOINT Meeting of the Western Section with the Bristol and District Section of the Royal Institute of Chemistry was held at 7 p.m. on Thursday, November 15th, 1962. The meeting took the form of a visit to the Factory of the British Nylon Spinners Ltd., Brockworth, Gloucester. MIDLANDS SECTION AN Ordinary Meeting of the Section was held at 7 p.m. on Thursday, November 8th, 1962, at the Technical College, The Butts, Coventqy. The Chair was taken by the Chairman of the Section, Dr. H. C. Smith, MSc., F.R.I.C., Dip.Ed. The following paper was presented and discussed: “Solvent Extraction,” by T. B. Pierce, RSc., M.A., D.Phi1.

ISSN:0003-2654    

DOI:10.1039/AN9628700921

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

December, 19621 HERRINGSHAW AND KASSIR: MERCURY-CATHODE ELECTROLYSIS. PART I 923 Mercury-cathode Electrolysis Part I. Effect of Current Density BY J. F. HERRINGSHAW AND 2. M. KASSIR (Chemistry Department, Imperial College of Science and Technology, London, S. W.7) The rate of deposition of copper, cadmium, zinc and iron from 0.3 N sulphuric acid on to a mercury cathode has been studied a t various current densities. Simultaneous evolution of hydrogen increases the rate of metal deposition above the value obtained by controlled-potential electrolysis. This increase is caused by an increase in the rate of transport of metal ions to the electrode. It cannot be accounted for by an increase in electrolytic transport, and must therefore be due mainly to an increase in convection and diffusion processes, i.e., stirring.Consequently, the rate of deposition per unit area is not greatly affected by applied stirring, but is dependent on the area of the electrode. IN recent years there has been a revival of interest in electrolysis at a large mercury cathode without control of the cathode potential, and the present state of development has been reviewed.1 Mercury-cathode electrolysis is a complicated system because it changes so much as deposition proceeds. Few of the many variables are controlled, notable exceptions being current density, acid concentration and, occasionally, temperature. Many workers have shown that an increase in the current density at the cathode causes an increase in the amount of metal deposited in a given time. It is not immediately obvious that this should be so.Under the conditions normally tried, at the lowest current densities, the cathode potential is extremely negative, far beyond the deposition potential of the metal, and a considerable amount of hydrogen is being evolved. For almost all metals, deposition is transport-controlled, i.e., it is limited by the rate of arrival of metal ions at the electrode surface by diffusion, convection (bulk transfer) and electrolytic transport. As the contribu- tion of the last is usually small, one might expect that the maximum rate of deposition would be substantially equal to that in a controlled-potential electrolysis ; any increase in current density should merely increase the amount of hydrogen evolved. This is clearly not so, and Coriou, HurC and Coursier2 have shown directly that when hydrogen is simultaneously evolved there is a marked exaltation in the rate of metal deposi- tion.They attributed this solely to an increase in the electrolytic transport of the metal ions, a conclusion that we regard as erroneous. Below, we present results to show the extent of this exaltation in conditions more akin to normal practice than those used by Coriou, HurC and Coursier, and we also provide evidence to show that the effect is caused mainly by an increase in convection and the consequent increase in diffusion. GENERAL PROCEDURE- Electrolytic reductions of 0.01 M copper, cadmium, zinc and ferrous sulphates in 0.3 N sulphuric acid were studied. The cell initially used was of the type shown in Fig.1 with B equal to 1.2 cm and C to 2.3 cm. Later, the area of the mercury surface was altered by changing the diameter B, and the depth of electrolyte for constant volume was altered by changing the diameter C. For larger cathode areas, simple U-tubes with B equal to C were used; for cathode areas greater than 20 sq. cm, beaker-cells analogous to those used by Chirn- side, Dauncey and Proffitt3 were used. In calculating the area of the mercury electrode, account was taken of the curvature of the surface. The smaller cells were kept in a thermo- statically controlled bath a t 25" C throughout : the larger cells were immersed in a cooling-bath, and the temperature of the electrolyte was kept a t 25" After the addition of sufficient mercury to bring the level to the top of tube B, the solution to be electrolysed (5 ml) was added, and the mercury level re-adjusted if necessary.A rubber stopper carrying the platinum anode, the tip of the saturated-calomel reference electrode and the nitrogen delivery tube was positioned as indicated. In the controlled- potential electrolyses, the anode was shielded by placing it in a tube, at the lower end of which was a sintered-glass disc, that contained 0.3 N sulphuric acid. This was to prevent the 1" C.9-24 HERRINGSHAW’ AND KSSSIR 1 MERCURY-CITHODE ELECTROLYSIS. PART I [VOl. 8; oxygen produced at the anode from reaching the cathode. In the controlled-current elec- trolyses, omission of this shield had no effect on the rate of deposition when the total current density was greater than amp per sq.cm; the reference electrode was used only occasionally. Kitrogen was passed through the soluticln for 15 minutes before electrolysis to expel most of the dissolved oxgyen, and a stream of nitrogen was normally used for stirring during the electrolysis. The electrical supply was obtained from the potentiostat described previ~usly.~ Cathode lead\ I 1- C +To reference / electrode A n o L N ltrogen lead Fig. 1. Electrolysis cell In the cell initially used the dimensions cf the cathode were chosen so that the current densities required were obtained without appreciably heating the solution. For all the cells the volume of the electrolyte was chosen so tEat the volume per unit area of mercury corre- sponded approximately to that used in normal practice.Other conditions were chosen so that complications caused by saturation of the amalgam or by build-up of solid metal, re- dissolution of the metal on washing at the end of the electrolysis, etc., were minimised. At the end of the electrolysis, the electrolyte was removed by suction without disconnect- ing the electrical supply. The cell and mercury were rinsed twice with distilled water and the rinsings removed by suction as before. The electrolyte and rinsings were then diluted accurately to a suitable volume. Solutions of copper, cadmium and zinc were analysed polarographically. Iron was determined absorptiometrically as the ferrous - o-phenanthroline complex. In experiments in which the amount of residual metal was very small, an aliquot was withdrawn before interrupting the electrolysis ; copper was determined absorptiometrically after extraction with zinc dibenzyldithiocarbamate in carbon tetrachloride and zinc by means of dithizone.PRESENTATION OF RESULTS- In the controlled-potential electrolyses, the current, corrected for the residual current, obeyed the relationship- it = i,,.lO-kt where t is the time in seconds, i, and i t are the currents passing at time 0 and t seconds, respectively, and k is a constant. The number of coulombs passed corresponded closely to the analytically determined amount of chemical change. From the above, it can be shown without further assumptions that- Ct = C‘,.lO-ktDecember, 1962; HERRINGSHAW AND KASSIR : MERCURY-CATHODE ELECTROLYSIS. PART I 925 where C, and Ct are the concentrations of the reacting species at a time 0 and t seconds, respectively.Other things being equal, the value of k is proportional to the area of the electrode,* and so we can write- where A is the area of the cathode in sq. cm. In order to compare the rates of deposition for different conditions, values of k’ were calculated. The time, in seconds, taken to reduce the initial concentration to one tenth is given by l/k‘A. It is not suggested that k’ is constant over a wide range of concentration, indeed it is almost certain that k’ becomes smaller at extremely low concentrations, but, as all the concentration changes studied were similar, comparisons based on k’ are justified. In the controlled-current electrolyses, the concentration of the metal ion was measured at the beginning and at the end of the electrolysis.This period usually amounted to 10 minutes. For some conditions, the electrolyses were carried out for various times up to 20 minutes, when it was found that the concentration decayed exponentially with the time of electrolysis. This decay was assumed to hold for all such electrolyses and, as before, values of k’ were calculated. amp per sq. cm, the rate of decrease of the concentra- tion was not truly exponential and tended towards linearity. Here, the rate of evolution of hydrogen is small. A small decrease in the rate of deposition of metal is accompanied by an equal increase of hydrogen evolution, and this increase is proportionately large. Other things being equal, an increase in the rate of hydrogen evolution is accompanied by an increase in the rate of deposition of metal; consequently, when the rate of hydrogen evolution is small, the rate of metal deposition will decrease with time more slowly than exponentially, and the rate of decrease of concentration will tend to be linear.At higher current densities, the rate of hydrogen evolution is sufficiently large for this effect to be negligible. The results at lower current densities are not vitiated because, with sufficiently small values of kt logelo, the linear and exponential decreases are indistinguishable. From equation (l), it follows that the rate of metal deposition, expressed as a current density, I, is proportional to the concentration, i.e.- where ~t is the number of electrons involved in the electrochemical change, and I‘ the volume in ml.Other things being equal, k is a function of V , and an extended form of equation (1) is- .. * * (1) ct = CJO-k’At . . . . .. At current densities less than No assumption that the deposition is transport-controlled has been made. It = log,lO.~Fk’VCt If I is in amps per sq. cm then C is in moles per ml. RESULTS- Results are shown in Table I. The values of k’ quoted are the mean of two, which were always within f 5 per cent. of the mean. The cathode potential of the mercury was measured in selected depositions of all the metals. If allowance was made for the portion of the current that effected metal deposition, the cathode potential was essentially the same as that obtained when discharging hydrogen on mercury in pure 0-3 x sulphuric acid.MECHANISM OF THE EFFECT- The results are essentially in agreement with those of Coriou, Hurk and Coursier.2 The values of k‘ are closely similar for all four metals, although their deposition characteristics are different. Copper, cadmium and zinc are electro-deposited almost reversibly, but have widely different deposition potentials ; iron is deposited irreversibly. Further, the rates of the two controlled-potential electrolyses are similar and are approximately equal to the value calculated on the Nernst model of diffusion and con~ection.~ Therefore, in all the electrolyses studied, the rate of deposition is not controlled by any irreversibility in the deposition of the metal, and the increase in the rate of deposition is not caused by the (negative) increase of the cathode potential.* It is shown on p. 927 that k’ is not always independent of A : for most of the results, the area of the electrode is 1.13 sq. cm and no error is incurred in expressing these results on the basis of 1.00 sq. cm.926 HERRINGSHAW AND KASSIR: MERCURY-CATHODE ELECTROLYSIS. PART I [VOI. 87 TABLE I EFFECT OF TOTAL CURRENT DENSITY ON THE RATE OF METAL DEPOSITION Cathode area: 1.13 sq. cm throughout. Electrolyte: 5 ml containing initially 3.235mg of copper, 5.60 mg of cadmium, 3.37 mg of zinc or 2.82 mg of ferrous iron, as sulphate, in 0.3 N sulphuric acid. Nitrogen stirring Current Time of Metal mA amp per sq. cm seconds mg k‘ Deposition of copper , . Controlled potential 4200 1.57 6.61 x Current, densit-$, electrolysis, remaining, (-0.3 volt against S.C.E.) 5.65 5 x 10-8 600 2.25 2.33 x 10-4 10 8.85 x 1.0-8 600 1.63 4.36 x 10-4 56.5 5 x 110-2 600 0.322 1.48 x 10-3 100 8.85 x 600 0.174 1.88 x 10-8 280 2.48 x 1.0-1 600 0.161 1.92 x lows Deposition of cadmium Controlled potential (-0.9 volt against S.C.E.) 10 8.85 x loe8 56.5 5 x 10-2 100 8.85 x 10 8.85 x 10-8 56.5 5 x 10-2 100 8.85 x 10-2 Deposition of iron* .. 10 8.85 x 56.5 5 x 10-2 100 8.85 x Deposition of zinc* . . 5.65 6 x 4800 600 600 600 900 600 600 600 600 600 600 2.67 2.79 0.345 0.276 1.78 1-58 0.188 0.138 1.31 0.254 0.21 5.93 x 10-5 4.46 x 10-4 1.79 x 10-3 2.74 x 10-4 4.83 x 10-4 4.90 x 10-4 1-66 x 10-3 1.93 x 1.86 x 2.04 x 1.54 X * No controlled-potential depositions of zinc or iron were made, because at the potentials necessary, hydrogen was evolved a t an appreciable rate.The rate of deposition of the metals must therefore be controlled by the rate of arrival of metal ions a t the electrode surface, and it is this rate that is increased when hydrogen is simultaneously evolved. Coriou et aL2 attributed the increase in the rate of metal deposition solely to an increase in the rate of arrival of metal ions at the electrode by electrolytic transport. An increase in the migration current does take place, but its calculated value accounts for only a few per cent. of the increase they observed in the rate of metal deposition. In our experiments, the same is true. The initial value of the transport number of the metal ions is less than 0.02 and this will decrease in proportion as the concentration of metal ion decreases.The con- tribution of electrolytic transport to k‘ is approximately 7 x 10-41’, where I’ is the total current density in amp per sq. cm.* Coriou et al. suggested that the mobilities of the relevant ions might be altered by the simultaneous rapid evolution of hydrogen, but it is difficult to imagine how this might happen. It should be noted that they regard the diffusion current as constant and disregard convection entirely. Further, they show that a ten-fold increase in the concentration of sulphuric acid decreases the value of their quantity A (equivalent to our k) only by about 10 per cent.; the expected decrease in the migration current would be ten-fold. Further evidence against the views of Coriou et al. was obtained from a study of the deposition of bismuth from chloride solutions.Bismuth chloride, 0.01 M in N hydrochloric acid, was electrolysed at controlled potential and controlled current as before ; the results are shown in Table 11. Despite the use of chloride solutions, no effect of dissolution of the platinum anode was observed. In the controlled-potential electrolyses, some of the bismuth was deposited initially as free metal, but at the end of the electrolysis only the amalgam was apparent. In such a solution, the bismuth is almost entirely in anionic forms; consequently, any increase in the electrolytic transport of the bismuth should cause a decrease in the rate of deposition. The results show that there is an increase in the rate of metal deposition similar * At sufficiently high metal-ion concentrations or a t excessively high current densities, the transport effect may become more important.December, 19621 HERRINGSHAW AKD KASSIR : MERCURY-CATHODE ELECTROLYSIS. PART I 927 TABLE I1 DEPOSITION OF BISMUTH Electrolyte: 5 ml of solution containing 9.78 mg of bismuth.Cathode area: 1.13 sq. cm. Nitrogen stirring Time of Metal Current, Current density, electrolysis, remaining, mA amp per sq. cm seconds mg k‘ Controlled potential 4800 0.775 2.03 x 10-4 (-.0.35 volt against S.C.E.) 100 8.85 X 600 0.46 1.95 x to that observed for other metals. Electrolytic transport effects are therefore not responsible for the increase in the rate of metal deposition. This increase must therefore be caused mainly by an increase in convection, i.e., bulk transfer, together with the consequent increase in rate of diffusion.I t is well known that enhanced stirring increases the rate of most transport-controlled processes and that continued increase of stirring has progressively less effect. It is difficult to imagine a method of stirring the liquid near the cathode more efficient than by electrolytic evolution of hydrogen at the cathode at a rapid rate. The greater the rate of hydrogen evolution the better the stirring, and therefore the greater the rate of metal deposition. This is borne out by the results, which also show that continued increase in the rate of hydrogen evolution has progressively less effect on the rate of metal deposition. If this view is correct, it would be expected that applied stirring* of the electrolyte would have less effect in controlled-current depositions, and that for a given current density the stirring caused by hydrogen evolution and hence the rate of metal deposition would be dependent on the geometry of the cell.EFFECT OF APPLIED STIRRING OF THE ELECTROLYTE- The deposition of copper and cadmium was studied with no applied stirring and with stirring by a stream of nitrogen or by a vibrating glass rod; the results are shown in Table 111. Other conditions were as before. TABLE I11 EFFECT OF APPLIED STIRRING Values of k’ for- Applied stirring -= copper Controlled potential . . . . . . None 4.3 x 10-5 3.7 x 10-6 Nitrogen 6.6 x 5.9 x 10-6 Vibrator 1.1 x 10-4 1.8 x 10-4 Controlled current . . .. . , None 1.52 x lo-* 1.83 X lo-* (I’ = 8.85 x amp per sq.cm) Nitrogen 1-88 x 10-3 1-93 x Vibrator 2.18 x 10-3 2.36 x 10-3 Applied stirring of the electrolyte has a considerable effect in the controlled-potential electrolyses, but its effect in the controlled-current electrolyses is much smaller. EFFECT OF CATHODE AREA- Electrolyses of unstirred copper solutions were carried out at a current density of 8.85 x amp per sq. cm in a series of cells with different cathode areas, but with the ratio of cathode area to volume of electrolyte kept constant. Under such conditions, the rate of metal deposition per unit area might be expected to be the same, because the larger cells could be regarded as the appropriate number of smaller cells in parallel; the values of k‘, which are based on concentration changes per unit area, would then be inversely pro- portional to the cathode area ( A ) .Hence, the value of k’A would be expected to be constant. However, as the results in Table IV show, this is not so; the value of k’A decreases as the area increases, although all the cells used were simple cylinders. It was observed in all * In this paper a distinction is made between stirring caused by evolution of hydrogen and stirring imposed by the operator, e.g., by means of rotating paddles; the latter is referred to as applied stirring.928 HERRINGSHAW AND KASSIR : MERCUEY-CATHODE ELECTROLYSIS. PART I [VOl. 8 i the electrolyses that the hydrogen tended to be evolved preferentially around the edges of the electrode; since the ratio of circumference to area of the electrode decreases as the area increases, this may account for the simultaneous decrease in k’A.Evidence for the im- portance of such edge effects was obtained from an electrolysis of 100 ml of copper solution in a rectangular vessel. This was 0.95 cm wide and 22.2 cm long and so had an area of 21.1 sq. cm, but its edges amounted to 46.3 cm; the beaker with the same area had a cir- cumference of 16.3 cm. Electrolysis in the rectangular cell gave k’ = 0.053 x and 10s k’A = 1.12, a value larger than that for the beaker. TABLE IV DEPOSITIOK OF COPPER Current density: 8.85 >: amp per sq. cm Diameter, Cell cm U-tube . . . . 1.2 U-tube . . . . 1.6 U-tube . . . . 2.5 U-tube . . . . 3.2 Beaker . . . . 5.2 Beaker . . . . 10.0 Cathode area ( A ) , sq. cm 1.13 2.01 4.91 8.04 21.2 78.6 Volume of electrolyte, ml 10Sk‘ 1OSk‘A 6 1-56 1.76 10 0.72 1.44 22 0.26 1.27 36 0.117 0.94 100 0.035 0.76 350 0.0062 0.48 Because of the reduced rate of deposition in larger cells, the effect of applied stirring in the larger cells and the extent of the area effect a t different current densities were investigated.A summary of the results is shown in Table IT; the conditions were as above. In general, if an increase in area causes a large decrease in k’A applied stirring serves to recover an appreciable proportion of the loss. Alternatively, this can be achieved by increasing the current density. TABLE V VALUES OF lo3 k’A FOR COPPER DEPOSITIOK Cathode area Current density, 1.13 2.01 4.91 8.04 amp per sq. cm sq. cm sq. cm sq. cm sq. cm 2.36 x lo-’ U - - 1.56 - S 2.08 8.85 x U 1.76 1.44 1.27 0.94 S 2.03 1.54 1.50 - 5 x 10-2 s 1.67 - 1.50 1.55 8.85 x U 0.42 - - - - - - - - - S 0.49 21.2 78.6 sq.cm sq. cm 1.12 - 0.75 0.48 0.91* 0*72* 0.27 - 0.31* - - - - - U = Unstirred * Stirred by an L-shaped paddle revolving at -300 r.p.m. S = Stirred by a stream of nitrogen. It should be stressed that, if the volume and current density are kept constant, an increase in the cathode area will cause an increase in k’A (3r k), i.e., in the rate of deposition of the metal, but the proportional increase in k will be less than that in A . If k is inversely proportional to the volume (vide iwfra), then, for 100ml of unstirred electrolyte and a current density of 8.85 x amp per sq. cm, we obtain- A , sq. cm . . 1.13 21.2 78.6 k, seconds-1 .. . . . . 8.8 x lo-; 7.5 x 10-4 1.68 x MISCELLANEOUS EXPERIMEKTS- amp per sq. cm to determine the effect of changing the depth and volume of the electrolyte. In the small cell ( A = 1-13 sq. cm), a change of depth of electrolyte from about 1.5 to 5 cm with the volume kept constant had virtually no effect on k’ in unstirred solutions. In the beaker-cell with a cathode area of 21.2 sq. cm, a reduction in the volume of the electrolyte (unstirred) from 100 to 50 ml resulted in k’ being almost exactly doubled, as would be expected. It seems, therefore, that at this current density the stirring caused by hydrogen evolution is sufficiently violent to accommodate these changes. A few experiments were made at a current density of 8.85 xDecember, 19621 HERRINGSHAW AND KASSIR : MERCURY-CATHODE ELECTROLYSIS.PART I 929 CONCLUSION- The increase in the rate of metal deposition obtained by an increase in the cathodic current density is caused almost entirely by an increase in the stirring of the catholyte. The contribution to the effect of electrolytic transport is small, except possibly in extreme conditions. to 10-3 M, occasionally down to -5 x M), corresponding to the middle part of the elec- trolysis. However, there is evidence that they are essentially valid down to concentrations of about 5 x 10-6 M. Values of k’A, after correction for volume, calculated from the results of Coriou et a1.,2 Bock and Hackstein,‘ Parks, Johnson, and Lykkens and Page, Simpson, and Graham,g show good agreement with our values. M, there is abundant evidence that the effective rate of deposition is much slower than expected.At the concentration levels studied, with an electrode area of 1.13 sq. cm, no useful increase in k‘ is obtained by increasing the current density beyond about 10-1 amp per sq. cm. However, at larger initial concentrations of metal ions this “optimum” current density will be higher, because the rate of metal deposition will be greater and the rate of hydrogen evolu- tion correspondingly less. If the initial concentration were ten times as large, h., 0.1 M, the optimum current density is estimated as about 5 x 10-1 amp per sq. cm. With electrodes of greater area, the optimum current density is much larger; the current required may be too large to be practicable. If the optimum current density is used, applied stirring of the electrolyte is hardly worth- while unless it is so vigorous as to be impracticable.However, if the optimum current density is not used, e.g., because the current is limited and the initial metal concentration or the area used is large, then applied stirring may be of value, provided that it is vigorous. The values of k’ for controlled-potential electrolyses can be increased by making the applied stirring more violent. Thus, at an electrode of surface area 1.13 sq. cm with 5 ml of electrolyte, extremely.vigorous stirring gave values of k’ of about 4 x but this is still below the value of k’ obtained by controlled-current electrolysis at the optimum current density. At larger electrodes, similar considerations apply, but in each type of electrolysis it is more difficult to achieve the same values of k’.Whatever the conditions, applied stirring alone cannot be as effective as the stirring caused by hydrogen evolution a t the optimum rate. The variables, current density, area and applied stirring, are interdependent and the optimum value for one depends upon the value of the other two and the concentration of metal ion. The value of the latter decreases during the electrolysis and if, for example, we electrolyse 0.1 M copper sulphate in 0.3 N sulphuric acid with a cathode area of 20 sq. cm at 6 amps with no applied stirring, conditions at the start are far from optimum; but when the concentration of copper has been reduced to 0-01 M, conditions from then on are nearly optimum. The initial stages of the deposition can be expedited by a moderate increase in temperature, either from the heating effect of the current or by using a “hot-start” method.3 Our values of k’ were obtained over a limited range of concentration of metal ion At concentrations below about REFERENCES 1. Page, J. A., Maxwell, J. A., and Graham, R. P., Analyst, 1962, 87, 245. 2. Coriou, H., Hur6, J., and Coursier, J., Anal. Chim. Acta, 1957, 16, 357. 3. Chirnside, R. C., Dauncey, L. A., and Proffitt, P. M., Analyst, 1940, 65, 446. 4. Herringshaw, J. F., and Halfhide, P. F., Ibid., 1960, 85, 69. 5. Bircumshaw, L. L., and Riddiford, A. C., Quart. Rev., 1952, 6, 157. 6. Newman, L., and Hume, D. N., J . Amer. Chem. Sac., 1957, 79, 4576. 7. Bock, R., and Hackstein, K.-G., 2. anal. Chem., 1953, 138, 339. 8. Parks, T. D., Johnson, H. O., and Lykken, L., Anal. Chem., 1948, 20, 148. 9. Page, J. A., Simpson, D. H., and Graham, R. P., Anal. Chim. Acta, 1957, 16, 194. Received July 17th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700923

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

930 MATTOCK: PROPERTIES OF TWO HIGHLY SELECTIVE p o l . 8 i The Properties of Two Highly Selective Sodium Ion-responsive Electrode Glasses BY G. MATTOCK (Electronic Imtvuments Lid., Richmond, Surrey) The properties of two sodium ion-responsive electrode glasses, designated BH68 and BH104, have been examined to assess their usefulness for laboratory and plant stream analysis. The working range of these glass electrodes has been found to be from molar (and possibly stronger) to nearly 10-5 molar, the response being logarithmic to concentration, as it is with pH electrodes. The selectivities to sodium ions in the presence of a variety of cations and anions, including hydrogen, potassium, ammonium, lithium, calcium, mag- nesium, sulphate and phosphate, have been found; only lithium has a marked influence, pH being unimportant above a value of 7 and at concentrations of sodium above Reproducibilities are similar to those for pH glass electrodes, but the speed of response (about 2 to 6 minutes) is less. Temperature effects have also been studied.It is concluded that the BHBS glass is suitable for general purpose laboratory and plant measurements, whereas the BH104 provides more acxrate results in laboratory measure- ments of strong sodium solutions. MOST work on electrode glass compositions has, until recently, been concentrated on pH properties, and investigations have mostly dealt with attempts to improve the pH range of glass electrodes. The earlier literature, well reviewed by Dole,l shows that certain ranges of glass compositions provide a measure of alltalimetal ion response, and in particular the work of Lengyel and Blum2 demonstrated the possibilities.However, the development of alkali-metal ion response with suppression of other cation reponse has only been pursued in recent years, notably by Eisenman, Rudin and Casby3 in the U.S.A., and by Nikol’skii and Shul’ts436@37 and G~remyltin*~~ and their co-workers in the U.S.S.R. Mattock has published a preliminary study on an earlier glass,10 and Leonard has also described a sodium- responsive glass.ll A part of the work described below was presented to the 1961 IMEKO Conference in Budapest. Probably the most widely used glass to date has been the NAS 11-18 composition of Eisenman, Rudin and C a ~ b y . ~ This glass has good properties, but suffers from the dis- advantage that it is highly refractory, making it difficult both to melt it to a homogeneous composition and to fabricate it into satisfactory electrodes.This paper reports on two of the most successful glasses that have been made and investigated in this laboratory in a prolonged study aimed a t improving both wclrkability and reproducibility without loss of selectivity. The work has been carried out jn full co-operation with Dr. Eisenman, and the writer is pleased to acknowledge the many valuable ideas, theories and practical sug- gestions that have been made freely available to him as a result of extensive correspondence. (Accounts of the Eisenman electrode selectivity theory are given elsewhere.12J3J4) The result has been the formulation of two glasses, coded BH68 and BH104, which are com- mercially practical* in their properties and offer some performance improvements over NAS 11-18.The BH68 has its main application in general industrial work, for which its extremely high selectivity is a major advantage: the BH104 has been found to be particularly applicable to accurate measurements on relatively strong sodium solutions. Many spheres of application are open to selective systems capable of giving a continuous non-destructive measure of sodium concentrations. Inasmuch as the glass electrode fulfils these demands, it is of obvious value in many laboratory and plant applications for which the conventional flame photometer has certain disadvantages. One particularly useful aspect of selective sodium-responsive electrodes is their sensitivity to wide ranges of concentrations, thereby permitting their use in a variety of analytical problems.Some of the many applica- tions to which BH68 and BH104 electrodes have already been put have been reviewed e1~ewhere.l~ to 1 0 - 4 ~ . * Now commercially available from Electronic Instruments Ltd., Richmond, Surrey (see E.I.L. Technical Data Sheets ELECT 9, 10 and 12).December, 19621 SODIUM 10s-RESPONSIVE ELECTRODE GLSSSES 93 1 EXPERIMENTAL ELECTRODES- Both glasses are easy to handle in a glassblower’s flame, and no difficulty was experienced in constructing electrodes of an orthodox kind that had resistances of the order of 50 megohms at room temperature. An inner reference system of a silver - silver chloride electrode dipping in either 0.1 11 hydrochloric acid or a hydrochloric acid - sodium acetate mixture was used; both were equally satisfactory, and measurements were made with reference to a conven- tional calomel electrode having a potassium chloride salt bridge with or without a secondary salt bridge of the buffer medium used in the test solution.Activation of the electrodes was effected by soaking the sensitive membranes in 8.1 M sodium chloride medium, in which they were also stored. It was found that if the BH68 electrodes (particularly) were allowed to dry out their responses to sodium suffered markedly: all electrodes were therefore kept moist at all times. SOLUTIOKS- The solutions employed were made up from analytical-reagent grade materials, when available (otherwise laboratory-reagent grade), and were stored and used in polythene con- tainers when sodium or other alkali metal that could be leached from glass vessels would have caused uncertainties. Since the electrodes respond, in principle, to ionic activities, it was considered desirable to keep the background medium constant in a given series of experi- ments so as to avoid the complications that would have arisen with variations in activity coefficient.By keeping activity coefficients constant (except with the strongest sodium solutions), the sodium response was easily studied as a function of the sodium concentrations of the solutions. The particular media employed were either ethanolamine or triethanol- amine, adjusted to the appropriate pH value for the experiment by the addition of hydro- chloric or nitric acids.MEASURING TECHNIQUE- If the electrodes respond solely to sodium ions, then a Sernst response equation should apply- 2.3826RT F where E is the potential of the electrode, aNs+ is the sodium ion activity, R is the gas constant, T is the absolute temperature, F is the Faraday and Eo is a zero term. Provided the activity coefficient of the sodium ions is held constant, this equation can be written- log aKaA E=EO+ 2.302611T E = EO’ + F 1%CN.+, where cNB+ is the concentration of the sodium ions in gram ions per litre. to introduce a term, pya, analogous to pH, and in this instance to define it as- Then- It is convenient pNa = - log cNB+ 2.3026RT F pKa E = EO’ - The form of this e.m.f. - pNa relationship is similar to that for the Nernst pH equation; it should therefore be possible to employ a pH meter for the measurements, adopting the pH calibrations as pNa ones, and applying exactly the same slope factor compensations for temperature changes.Accordingly, standardisations of the electrodes were made by arbitrary definition of a M sodium solution as having a pNa of 2.00, and adjustment of the pH meter controls to read pH = 2.00 when the electrodes were placed in the pNa 2.08 solution. Theoretical behaviour would then be manifested by indications on the instrument of pH values corre- sponding to the defined pNa of other sodium solutions. In fact, glass electrodes may respond in a mixed manner to other cations present, to an extent governed by the relative concentrations of these cations and by the inherent selectivity characteristics of the glass.The glasses described here were examined by measurement ~- (2.30:RT)932 MATTOCK: PROPERTIES OF TWO HIGHLY SELECTIVE [Vol. 87 of their response characteristics in the presence of various cations and anions to identify the degree of sodium selectivity. Measurements were made on E.I.L. pH meters, models 23A, 48A and 33B + C33B, according to the discrimination required. When close control of temperature was required a thermostat controlling to FO.05" C was used. RESULTS Most of the results obtained are shown graphically; the Figures give results derived from arbitrarily selected electrodes of those examined. The properties of individual glass electrodes, including even the best pH electrodes, vary somewhat from one to another, but it is believed that the graphs presented here give a good picture of the general properties of the glasses.Certainly a large amount of experience subsequent to this particular study has confirmed the essential validity of the results given. Concentration 'of sodium p.p.rn. -Log CNa+ Fig. 1 . Effect of pH on the sodium response of BH68 electrodes. Teniperature: 20" to 22'C. Standardisation effected in M sodium ions (-log CN&+ = 2.00), defined aspNa = 2.00. Curve A, pH 10.2 (1.25 M ethanolsmine plus hydrochloric acid medium) ; curve B, pH 9.0 (1.25 M ethanolamine plus hydrochloric acid medium); curve C, pH 8.1 (0.5 M triethanolamine plus hydrochloric acid medium): curve D, pH 6.9 (0.5 AT triethanolamine ~ Z U S hydrochloric acid medium); curve E, pH 6.0 (0.5 M triethanolamine plus hydrochloric acid medium); curve F, pH 3.13 (0.5 M triethanolamine plus hydrochloric acid medium).The theoretical (Nernst) slope is given by the broken diagonal lines. (The curves are displaced vertically to aid comparisclns)December, 19621 SODIUM ION-RESPONSIVE ELECTRODE GLASSES 933 SODIUM RESPONSE AT DIFFERENT pH VALUES- (a) BH68-Fig. 1 shows the response to change in pNa at different constant background pH values ; the diagonal broken lines indicate here, as in other similar diagrams, the theoretical response. It can be seen that, as the pNa increases, the response falls off, but that, down to pH values of about 7 to 8, it is linear in the pNa range 1 to 4 (within 0.05 pNa unit). The slight deviations at pNa = 0 may be accounted for in terms of a change in activity coefficients, since here the background medium would not be capable of keeping these constant.The response was found to be nearly theoretical (90 to 95 per cent.) with the electrodes examined; the background medium may exert an influence, and individual electrodes may vary. Be- tween pNa 4 and 5 response is still obtained, with some curvature, but measurement of response at these weaker levels is rendered difficult because of interferences caused by the presence of sodium impurities in the background media. Extensive field experience has shown that the practical lower limit of concentration is about 1 p.p.m. of sodium, corresponding to a pNa of 4.4. At lower concentrations, sodium changes are followed, but only qualitatively.Further, for the most dilute solutions it is highly desirable to maintain a constant pH of 8 to 10, and a constant background medium. Fig. 2. Sodium response range and effect of potassium on BH104 electrodes. Temperature : 20‘ C. Standardisation effected in M sodium ions (-log CN&+ = 2.00), defined as pNa = 2.00 (0.75 M triethanolamine plus nitric acid medium; pH 7.5). Curve A, no potassium; curve B, 0.1 M potassium. The theoretical (Nernst) slope is given by the broken diagonal lines. (,The curves are displaced vertically to aid comparison) A further series of experiments was carried out to test whether or not a change in the “standard potential” of the electrodes was occurring at different pH levels, such that, although linearity in sodium response was being achieved, a displacement might be resulting from a change in pH.Fig. 3 shows some of the results obtained when the pH was changed in solutions at constant pKa. Results are also included for NAS 11-18 glass, from which it can be seen that the BH68 shows some improvement in terms of insensitivity to pH at the low medium concentrations. I t is apparent that there is no marked zero shift as a result of pH change other than that shown in Fig. 1, since the two diagrams show essentially similar results. Similar experiments carried out by adding hydrochloric acid to 0.01 M sodium hydroxide (it?., pNa = 2) indicated, however, that without the buffer medium slightly greater sensitivity to pH is exhibited. The results can be summarised by stating that pH change has only a slight effect on the pNa response when the pH is greater than 7 and the ph’a is less than 4, with a constant934 MATTOCK : PROPERTIES OF TWO HIGHLY SELECTIVE [Vol.87 ionic background. 0.1 pSa unit in a constant medium. ion levels. range (up to p y a = 4). for BH68. At pNa 4, a change in pH from 10 to 7 brings a change of only about A useful guide is the ratio of hydrogen ion to sodium negligible pH effects occur in the linear response (b) BH104-The sodium ion-response range can be seen from Fig. 2 to be similar to that When this does not exceed In curved response regions, pH control is important. The effect of change in pH is shown in Fig. 3. DH Fig. 3. Effect of pHI on sodium response of BH68, EHlO4 and NAS 11-18 electrodes: 0, BH68 (1.25 M ethanolamine plus hydrochloric acid medium); 0, BH104 (0.75 M triethanolamine plus nitric acid medium) ; V , NhS 11-18 (1.25 M ethanol- amine plus hydrochloric acid medium) THE EFFECT OF CATIONS ON SODIUM RESPOXSE- 1, Potassium- (a) BH68-Fig.4 shows the effect of a background of potassium chloride on the pKa response curves; it is apparent that a 100-fold excess of potassium over sodium does not affect the response at the 0.02 to 0.05 pNa unit level of precision in the pNa region 0 to 4. Confirmation of this was obtained by adding potassium chloride to solutions of constant ionic strength a t constant pNa, when essentially the same results were obtained. However, it should be noted that transfer of an electrode from a given solution to another of similar concentration but containing potassium may cause a zero shift of up to 0.1 pNa unit, but further changes in concentration of potassium have no effect.A transient shift may occur on adding more potassium, but the original rettding is regained after 2 to 3 minutes. This phenomenon has also been observed by Friednian et al. with NAS 11-18 gPass.l6 ( b ) BHlOPFig. 2 shows that potassium has a slightly greater effect on this glass, interference becoming apparent with a 50-fold excess over sodium. 2. Ammonium- Essentially the same results were obtained as with potassium, except that the smaller sodium impurity level of the ammonium chloride employed permitted use of larger ratios of ammonium ions to sodium. From the experiments it was apparent that a 200-fold excess could be tolerated without effect on the pNa response in the region 0 to 4.December, 19621 SODIUM ION-RESPONSIVE ELECTRODE GLASSES Concentration of sodium I 0.1 0 .0 1 ~ 23 2.3 0.23 0.023 p.p.m 935 8 1 2 3 4 5 6 Fig. 4. -Log CN,+ Effect of potassium on sodium response of BH68 electrodes. Temperature: 20" to 22" C. Standardisation effected in M sodium ions (-log CN%+ = 2.00), defined as pNa = 2.00. Curve A, pH 10.2 (1.25 M ethanolamine plus hydrochloric acid medibm); curve B, pH 10.2 (1.25 M ethanol- amine plus hydrochloric acid plus 0.01 M potassium chloride medium) ; curve C, pH 6.9 (0.5 M triethanol- amine plus hydrochloric acid plus 0.01 M potassium chloride medium). The theoretical (Nernst) slope is given by the broken diagonal lines.(The curves are displaced vertically to aid compansons) 3. Lithium- Lithium was found to have a stronger influence on the pNa response for BH68 electrodes It appears that a 10-fold excess of lithium over sodium can be tolerated without (see Fig. 6 ) . significant interference. 4. Alkaline eayths- Strong solutions of calcium chloride and magnesium sulphate were added to sodium solutions of constant ionic strength at pNa levels of 2 and 4. For both BH68 and BH104 glasses no changes were observed with 100-fold excesses of both cations. Higher excesses may be tolerable, but sodium impurities in the alkaline-earth reagents limited experimental definition. THE EFFECT OF ANIONS ON SODIUM RESPONSE- Sulphate ions in 1000-fold excess were found to have no effect on either BH68 or BH104 electrodes in sodium solutions of constant ionic strength.Phosphate ions were tolerable at 50-fold excesses, in the region of pH 7 ; beyond this phosphate excess constancy of the sodium ion activity coefficient was probably not maintained, so no definite information of tolerance limits was obtainable. SPEED OF RESPONSE- The response time, i.e., the time taken t o reach an equilibrium reading within given limits is generally longer with sodium-responsive electrodes than with pH electrodes, other factors, such as the concentration change to which the electrodes are subjected, being equal.936 Fig. 5 . Effect of lithkim on sodium response of BH68 electrodes. Temperature: 22" C. Standardisa- tioneffectedin10-2~sodiumions (-logcN,+ = 2.00), defined as pNa = 2.00.Curve A, pH 6.9 (0.5 M triethanolamine plus hydrochloric acid medium- no lithium chloride); curve B, pH 6.9 (0.5 M tri- ethanolamine plus hydrochloric acid plus 0.01 M lithium chloride medium) ; curve C, pH 6.7 (0.5 M tri- ethanolamine plus hydrochloric acid plus M lithium chloride medium). The theoretical (Nernst) slope is given by the broken diagonal lines. (The curves are displaced vertically to aid comparisons) [Vol. 8' The technique of electrode usage can affect the response time significantly. Thus, washing the electrode with water during transference in sample measurements increases the response time markedly; the preferred technique is simply to wipe the membrane when the concen- tration transition is not larger than a factor of about 10, or intermediate washing with a sample of the new solution. Also, it was found that BH68 electrodes develop sluggishness if they are allowed to dry out, presumably from lattice deterioration of the glass.M sodium solutions buffered to pH 7 in 0.1 M triethanolamine - hydrochloric acid medium, stability was achieved to 0.01 pNa unit in 2 to 3 minutes with BH104 and NAS 11-18 electrodes, and in 3 to 4 minutes with BH68 electrodes. To achieve the 0.005 pNa unit level the BH104 electrodes required 4 to 5 minutes and the BH68 electrodes required 10 to 15 minutes, the NAS 11-18 times falling between the two, However, response times do depend on the concentration of sodium in the solution, equilibration in 10-1 M sodium solutions requiring about half the time needed in 10-5 or l o - 4 ~ solutions; the BH104 electrodes were generally faster than the BH68 ones.Prolonged contact of an electrode with a very dilute sodium solution will increase its response time, but re-activation can be achieved by soaking the electrode in 0.1 to 0.2 M sodium chloride, which is also a good storage medium. In continuous plant operation with concen- trations of sodium of the order of 1 p.p.m., response to step changes of sodium level is prac- tically instantaneous, even though a final value may not be achieved for several minutes. In the presence of an excess of potassium over sodium, response times are increased by a factor of 2 or 3 over those in the corresponding sodium solution. When alternating between, for example, and 1.5 xDecember, 1962: SODIUM ION-RESPONSIVE ELECTRODE GLASSES 937 SOLUBILITIES AND REPRODUCIBILITIES- As with pH electrodes, the properties of sodium-responsive electrodes vary with their history and the prevailing conditions, but some generalisations can be made based on experi- mental observations.Thus in concentrated sodium solutions, down to M, the order of drift is 0.05 pSa unit per day for both types of electrode, whereas in the lo-* M region it may be 0.1 pNa unit per day. In a given solution most of this drift occurs in the first few hours. (On re-standardising, it should be remembered that the initial standardisation after storage may be erroneous, and it is advisable to repeat the measurement.) A dependence on flow was observed, in agreement with the findings of other workers who have used NAS 11-18 g l a ~ s .~ ~ J ~ J * This effect demands that accurate measurements be made with static samples or under constant flow conditions. However, it is not significant at the 0.02 to 0.05 pSa unit level of precision with bulb electrodes in normal circum- stances, and it is not important in plant applications down to 1 p.p.m. of sodium. An idea of Ihe magnitude of the effect was gained by causing a M sodium solution to flow through a 2-mm bore capillary electrode approximately 50 mm in length; at flow velocities of 15 cm per second there was a displacement of approximately 0.1 pya unit to a lower reading. Reproducibilities depend on the sodium levels a t a given temperature-the stronger the solution the better the result.The BH104 electrodes gave better results than the BH68 and ?;AS 11-18 types, being capable of concentration reproducibilities of f l per cent. in 10-1 to 1 0 - 3 ~ sodium solutions. In general, however, + 2 per cent. would seem to be the achievable level; in the presence of potassium in similar amounts to the sodium repro- ducibility is usually f 4 per cent. The BH68 electrodes give slightly less reproducible results, a figure of f 4 per cent. being practical. Both glasses have poorer reproducibilities at higher temperatures (e.g., 40" C). Results also depend on the magnitude of the pNa transition- the larger the transition the poorer the reproducibility. In a run of 33 measurements taken during cycling of the electrodes between different sodium solutions, BH68 electrodes showed reproducibilities of k0.04 pNa unit about a mean value in pNa 1 and 3 solutions, and j0.15 pKa unit about a mean value in pNa 4.7 solutions (=04 p.p.m.of sodium). TEMPERATURE EFFECTS- In the theoretical response equation- 2.303RT E = EO' - ~ F pNa there are two temperature-dependent terms, apart from a possible true p y a - temperature coefficient,; the zero term EO' and the slope factor 2*303RT. -- Zero shift constitutes the change in electrode potential with temperature when pNa = 0; the slope shift is theoretically predictable, and should be the same as for pH glass electrodes. (a) Slope factor shift-This was studied between 20" and 40" C for both BH68 and BH104 electrodes. Although the slope factor itself was not theoretical its change with temperature was found to be in accordance with theoretical prediction, within the limits of experimental error ( & 3 mV).( b ) Zero shif&--Experimental observation between 20" and 40" C for both BH68 and BH104 electrodes, in conjunction with a calomel reference electrode at constant temperature, indicated a zero shift of +0.1 to $0.2 mV per "C increase in temperature. (The reproduci- bility of results with the BH68 electrodes when measurements were made sequentially at different temperatures was better than that for the BH104 types.) It is apparent that normal slope factor compensation, as employed on commercial pH meters, is satisfactory. Further, since saturated potassium chloride calomel reference elec- trodes also have a zero shift temperature coefficient of +0.2 mV per O C , 1 9 total zero shift will cancel out to zero in ambient temperature conditions.Thus drifts due to small ambient temperature fluctuations (up to 10" C) can be overcome by slope compensation alone. F938 MATTOCK: PROPERTIES OF TWO HIGHLY SELECTIVE [Vol. 137 DISCUSSION OF THE WORK In contrast to pH, which is normally measured as a parameter of empirical interest in its own right, pNa is probably of most value in so far as it indicates concentrations of sodium ion. The pKa is logarithmically related to the concentration of sodium ion, so there is an inherent measurement insensitivity so far as concentrations or activities are concerned. However, the importance of this insensitivity depends largely on the sodium ion level a t which measurements are to be made.In the 0 to 4 pNa region, both BH68 and BH104 electrodes have response slopes close to the theoretical, and in these circumstances an error in pNa units can be converted to a percentage error, as shown below- +O.Ol pNa unit = f 2 . 3 per cent. k0.02 pNa unit 3 k4.5 per cent. 10.05 pNa unit = +11*0 per cent. k0.10 pNa unit 3 420.5 per cent. Obviously, in the higher concentration regions it is desirable to achieve a precision of &O.Ol to &0.02 pNa unit if the measurements are to be analytically useful. This can be achieved in the 0 to 3 pNa region; the more concentrated the sodium solution and the smaller the range the better the precision. In the more dilute solutions, less precision is usually per- missible. This may be inferred from Table [, in which correction for an assumed slope equal to 50 per cent.of the theoretical has been made a t pNa 5. TABLE I LEVELS OF PRECISION Error corresponding to (= - log Sodium ions present, h0.05 pNa unit precision, p.p.in. p.p.m. 3 23 h2.1 4 2.3 & 0.2 6 0.23 0.05 The precision achievable with solutions weaker than pNa 4 depends on the response slope, which itself is somewhat dependent on pH, so it is not possible to be specific without stating the conditions. COMPARISON OF BH68 4ND BH104 ELECTRODES- The results reported in this paper, and field experience, lead to the conclusion that BH68 glass is probably the more satisfactory of the two for general purpose electrodes, because of its superior selectivity to sodium in the presence of excesses of other cations.On the other hand, BH104 appears to be a little faster in response and is capable of providing highly reproducible results, particularly when relatively concentrated sodium solutions are involved. Consequently, the BH68 electrodes have been applied in general plant and laboratory practice, whereas the BH104 electrodes have found use in more accurate laboratory work. Plant applications of BH68 electrodes have included the monitoring of tidal influences on river-water supplies, measurement of the saline content of well waters and wash waters, and investigations are proceeding into their possible use in the monitoring of boiler waters. Identification of ion-exchange break-throughs is possible provided a buffering reagent ( e g . , triethanolamine + hydrochloric acid) is added to prevent the variations in pH from signalling an apparent pNa change a t the extremely 101~ concentrations of sodium to be monitored.Laboratory applications have included measurements of sodium in beers, yeasts and natural waters. The BH104 electrodes have been used mainly so far in the biochemical sphere; sodium determinations have been performed on blood, urine and tissue fluids. They have also been successfully applied to sodium measurements in the laboratory on natural waters. COMMEXTS ON USE OF ELECTRODES- In making determinations, it is advisable to employ standardising solutions having as similar a background medium as possible to the test solutions to be examined. In laboratory practice this usually means the addition of a suitable buffer containing a large organic cation, such as triethanolamine plus hydrochloric or nitric acid to give a pH of about 7 and an ionic strength (e.g., 0.5 to 1.0 M) sufficiently large to swamp any variations in sodium activitvDecember, 19621 SODIUM ION-RESPOXSIVE ELECTRODE GLASSES 939 coefficient that may occur owing to the presence of variable amounts of other cations.During the course of a day’s run, the electrode can be held between measurements in the same medium containing 10-1 or For overnight and longer storage a 10-1 M or more concentrated solution of sodium chloride is preferable. When measurements are to be made a t different temperatures, BH68 electrodes are preferable; either BH68 or BH104 can be employed for constant or ambient temperature conditions and slope temperature compensation only need be applied when a saturated potassium chloride - calomel electrode is used as reference, at the same temperature as the glass electrode.It is also desirable for laboratory measurements that an intermediate bridge solution be used between the reference electrode and the test solution. This intermediate solution should consist of the constant ionic strength buffer or medium similar to that of the test solution. When the reference electrode is at constant temperature, a zero shift of ‘0.2 mV per O C increase (equivalent to 1 isopotential pH unitz0) must be allowed for over the temperature range 20” to 40” C. Obviously, for the most precise work, temperature control should be employed, as it is with pH glass electrodes. I thank Mr. R. Uncles, Nrs. E. Williams and Mr. A. H. Gunn for assistance with some of the experimental work, and the Directors of Electronic Instruments Ltd. for permission to publish this paper. M sodium. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. REFERENCES Dole, M., “The Glass Electrode,” John Wiley & Sons Inc., New York, 1941, chapter 7. Lengyel, B., and Blum, E., Trans. Faraday SOC., 1934, 30, 461. Eiseman. G., Rudin, D. O., and Casby, J. U., Science, 1957, 126, 831. Nikol’skii, B. P., Shul’ts, M. M., and Peshekhonova, N. V., Zhuv. Fiz. Khim., 1958, 32, 19 and 262; Shul’ts, M. M., and Aio, L. G., Vestnik-Leningrad Univ., 1955, So. 3, 153. Shul’ts, M. M., and Parfenov, A. I., Ibid., 1958, No. 16, 118. Shul’ts, M. M., Chem. Abstr., 1956, 49, 74241. Goremykin, V. E., Gidrokhim. Mat., 1957, 26, 218; 1959, 29, 205. Goremykin, V. E., and Krykov, P. A,, Izvest. Akad. iVauk SSSR, Otdel K h i m . lVauk, 1957, 1385. Mattock, G., in “Proceedings of the International Symposium on Microchemistry, 1958,” Per- gamon Press Ltd., 1960, p. 424. Leonard, J. E., Beckman Instruments Preprint R-6148, Fullerton, California. Isard, J. O., Nature, 1959, 184, 1615. Mattock, G., “pH Measurement and Titration,” Heywood & Co., London, 1961, pp. 130 to 131. Eisenman, G., paper presented a t the International Biophysics Congress, Stockholm, 1961. Mattock, G., paper presented a t the Feigl Anniversary Symposium, Birmingham, 1962. Friedman, S. M., Jamieson, J. D., Nakashima, M., and Friedman, C. L., Sczence, 1959, 130, 1252. I’ortnoy, H. D., Thomas, L. M,, and Gurdjian, E. S., Talanta, 1962, 9, 119. __ - __ , J . Appl. Physiol., 1962, 17, 175. Mattbck, G., “pH Measurement and Titration,” Heywood & Co., London, 1961, p. 154. -- , op. cit., pp. 190 to 197. 1959, 33, 1922. Received May 31st, 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700930

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

940 SULLY : PROCEDURE FOR DETE:RMINING HYDROXYL GROUPS [Vol. 87 A Procedure for Determining Hydroxyl Groups BY B. DUDLEY SULLY ( A . Boake, Roberts and Go. Ltd., London, E.15) A new method is described for detmermining hydroxyl groups in primary alcohols, secondary alcohols and phenols. The procedure is to esterify the hydroxyl group with a known weight of stearic anhydride in boiling xylene as solvent, and to decompose the excess of stearic anhydride with water by using either sodium stearate or pyridine as catalyst. The unreactecl stearic acid is then determined by titration and the reacted acid is calculated by difference. The procedure is rapid arid accurate and can be applied to a wide range of hydroxylic compounds. THERE are several possible sources of error when acetic anhydride is used in determining hydroxyl groups.The most consistent results are obtained by the procedure in which the hydroxyl group is acetylated and the resulting product is isolated; the hydroxyl group is then determined from the saponification value measured on a weighed sample of the acetate. At the moment, this appears to be the method most commonly used under routine conditions in analytical control laboratories. The second general procedure is to esterify the hydroxyl group with a known weight of acetic anhydride in pyridine and then to measure the excess of anhydride by titration with alkali. At fint sight this would appear to be an accurate and convenient method, but general experience shows that it is less reliable, a t least under routine conditions. To some extent this may be due to difficulty in measuring a pyridine solution accurately by volume, and many authorities recommend that it should be weighed ; it is thought, however, that other unknown factors operate.The reactions involved may not always proceed quantitatively, for it has been observed that dry or impure pyridine combines with acetic anhydride to form a resinous compound, and that the best results are obtained when the pyridine contains between 0.3 and 0.5 per cent. of water; larger amounts reduce the efficiency of the acety1ation.l Phthalic anhydride has been used instead of acetic anhydride, but it has no outstanding advantage, although it is claimed to be less reactive and so can be used for some compounds that otherwise decompose under the more vigorous acetylation conditions.The Zerewitinoff method with a Grignard reagent (methyl mag- nesium iodide) or the newer method with lithium aluminium hydride have usefui specific applications, but are inconvenient for general use owing to the special reagents required and the need for the removal of every trace of water from the material under investigation. A comprehensive review of the methods for determining hydroxyl groups has been published.2 The experiments described here show that stearic anhydride in boiling m-xylene forms a stearate quantitatively with a wide range of hydroxylic compounds in less than 90 minutes. Stearic anhydride is readily prepared and is surprisingly stable when exposed to air; samples stored for several months without special precautions to prevent the ingress of moisture have shown no deterioration.It also has an extremely low vapour pressure and is easy to weigh owing to its high equivalent weight. In the recommended procedure the hydroxylic com- pound to be determined is weighed into a flask together with a weighed amount of stearic anhydride, equivalent to a titration of 25 to 510 ml of normal alkali solution, and 10 ml of m-xylene is added as a solvent. An excess of the anhydride is essential, and the recommended excess of 50 per cent. above the theoretical means that the back titration should be two-thirds that of the blank titration. The stearoylation reaction is often substantially complete for primary and secondary hydroxyl groups after 15 minutes a t the boiling-point, but boiling for 30 minutes is recommended.The volume of m-xylene used is important, because an excess will hinder the decomposition of the unreacted stearic anhydride in the final stage of the determination. When the stearoylation is complete the excess of stearic anhydride is decomposed either by boiling with aqueous 3pyridine or by boiling with water containing neutral sodium stearate as an emulsifying agent. If the latter method is used then it is necessary to make the mixture homogeneous after decomposing the stearic anhydride by ad- ding neutral saponification spirit. In both procedures the titration is made a t the boiling- point of the solvent in order to keep the stearic acid in solution, otherwise a satisfactory end-point cannot be obtained.December, 19621 94 1 In an industrial laboratory the stearic anhydride method has been found rapid, accurate and convenient for determining the equivalent weights of plasticiser alcohols, and the alternative procedure that avoids the unpleasant smell of pyridine is favoured. The plasti- ciser alcohols are of the iso-octanol type and are mixtures of primary alcohols synthesised TABLE I DETERMINATION OF A STERICALLY HINDERED PHENOL SULLY : PROCEDURE FOR DETERMINING HYDROXYL GROUPS Weight of sample, s 2.2075 2.2044 2,0032 2.0086 2.2028 2.2057 2,2070 2,2073 Titre equivalent Weight of stearic Reaction to the weight of stearic anhydride, time.anhydride used, g hours ml of N NaOH 46.85 46.85 46.81 46.83 46.75 46.77 46.68 46.72 0.6 1 3 6 13.0104 13.0066 13.0026 13.0120 13.0042 13.0082 12.9877 12.9964 Substance Citronellol .. . . Geraniol . . . . Geraniol (commercial). . Cetyl alcohol . . . . Benzyl alcohol.. , . Phenylethyl alcohol (8) Menthol . . . . Glycerol monostearate TABLE I1 DETERMINATIOX OF ALCOHOLIC HYDROXYL . . . . -1 ::{ / . . . . . . Propylene glycol monostearate Acetoglyceride (sample A) . . .4cetoglyceride (sample B) . . Castor oil . . . . . . IVeight g 2.3371 2.2832 0.6863 1.0905 1.2404 1.9966 1.2257 2.0650 2.0372 2.1526 2.0488 1.2487 1.0278 2.4795 4.1265 4.0978 1.2312 2.1867 2.7103 1,4406 2.5757 2.6004 1.2380 1.2797 1.2624 1.4487 2.1980 0.9173 1.4362 2,7324 2.9078 3.2016 3,4089 2.5008 2.7974 3.0018 3.0156 Procedure x x A A A A B A A -1. A B €3 A ,2 x A A A A A x B x -4 A B A A A A A A A A A Reaction time, minutes 30 30 30 60 60 120 30 15 30 30 60 30 30 30 30 30 30 30 30 30 30 30 30 30 60 120 120 60 60 30 60 30 60 30 30 60 60 'I'itre of sample, ml 31.65 31.05 31.65 31.40 29.10 29.00 28.85 28.85 GROUPS Excess of stearic anhydride.% 57.0 47.2 80.3 72.5 54.1 69.4 39.3 62.8 62.1 55.1 69.7 61.3 71.8 54.2 32.9 34.5 105.5 12.7 1.1 11.2 38.2 36.6 38.1 92.9 97.4 73.0 68.7 91.5 20.8 90.3 93.6 105.8 92.9 69.4 40.3 37.4 48.3 Calculated purity, 84.1 87.6 92.5 93.8 97.9 98.5 98.7 99.0 % Purity by stearoylation, % 100.1 99.9 100.4 100.2 100.3 99.6 100.2 100.4 100.2 100.0 100.4 99.9 100.1 96.8* 100.5 100.0 100.3 99.8 99.6 99.8 99.7 99.9 100.5 82.3t 86.0t 87.37 88.lt 100.5: 100.5: 126.6s 128.5s 98.5s 158.65 156.8s 153.48 153.73 99.7s * The purity determined by saponification of the isolated acetate was 96.9 per cent.t The 1,2-monostearin content determined by the periodic acid method was 86.9 per cent. $ Purified by molecular distillation. f Hydroxyl value (British Standard 684: 1958).042 SULLY : PROCEDURE TOR DETERMIKIhG HYDROXYL GROUPS [Vol. 8 i with raw materials from petroleum sources. The method has also been used for determining the purity of menthol, fatty alcohols and certain perfumery alcohols, such as phenylethyl alcohol, geraniol and citronellol. Tertiary alcohols react slowly under the conditions des- cribed, and the results do not appear to be quantitative. The method cannot be used for sugars or compounds not soluble in the xylene - stearic anhydride mixture, although glycerol reacted quantitatively after boiling for 3 hours.Traces of water do not appear to react with the stearic anhydride because they are rapidly carried into the condenser by the boiling xylene. Phenol, eugenol and p-chloro-m-cresol reacted quantitatively with st earic anhydride after boiling under reflux for 30 minutes; the stearic anhydride initially contained 3 per cent. of free stearic acid. S o catalyst was added. A longer reaction time is required, however, if the phenol is sterically hindered, as, for example, with 2,5-xylenol, a jmre sample of which gave the results shown in Table I. The results obtained with mixed clommercial xylenols are comparable with those obtained by conductimetric titration or by c zlculation from gas - liquid chromatographic analysis. Various polyhydric alcohols, partly esterified with a fatty add such as stearic acid, are widely used as emulsifying agents, as, for example, in cosmetics and foodstuffs.The stearic anhydride procedure is particularly useful for compounds in this class, including those based on various glycols, as, for example, propylene glycol monostearate, for which at the moment there is no entirely satisfactory method. For determining glycerol monostearate it has an advantage over the periodic acid method in that it measures both the u and /3 mono- glycerides. Propylene glycol monostearate rexts quantitatively to give the distearate in 30 minutes, but for glycerol monostearate the reaction takes about 60 minutes. The procedure has also been used successfully for determining butanol ( N 2 per cent. w/w) in tributyl phosphate, and for fatty alcohols (less than 5 per cent.w/w) in commercial esters. Some results for determinations of alcoholic hydroxyl groups are shown in Table I1 and for phenols in Table 111. The method has been found useful for determining phenols. TABLE I11 DETERMINATION OF PHENOLS Carried out by Procedure R ; the reaction time was 30 minutes Substance g % /a Excess of stearic \Veight, anhydride, Purity', by stearoylation, 01 Phenol . . . . . . . 1.1006 484 99.4 Eugenol . . . . . . 1,4721 54.8 96.9 3.5-Xylenol . . . . . . 2.0178 41.9 97.5 3,5-Xylenol . . . . . . 2.0062 42.5 98.0 p-Chloro-nz-cresol . . . . 1.2666 3 6 4 98.3 METHOD REAGENTS- S t e a k adzydride-This is prepared by boiling under reflux for 8 hours a mixture con- taining 1000 g of stearic acid and 550 g of acetic anhydride.The excess of acetic anhydride and the acetic acid formed during the reaction are removed under vacuum, care being taken to ensure that the temperature does not rise above about 135" C. The stearic anhydride is purified by crystallisation from 1500 g of light petroleum, boiling range 60" to 80" C. A commercial grade of stearic acid is suitable, provided that it has a good heat stability, an iodine value below 4 and crystallises well. The product may contain up to 10 per cent. of stearic acid, but this is unimportant. Experience has shown however that the crystals are not uniform in composition and it is essential to grind them to a powder to ensure homo- geneity before use in analysis. The purity is determined by titration of a sample after boiling for 15 minutes with an excess of methanol, which produces one molecular equivalent of methyl stearate and one of stearic acid from each molecular equivalent of stearic anhydride.In a parallel experiment the stearic anhydride is decomposed to stearic acid by boiling with aqueous pyridine before titration with alkali. 'The free stearic acid in the stearic anhydride is calculated from the two titres. The stearic anhydride used in the experiments described in this paper contained between 2 and 4 per cent. of free stearic acid.December, 19621 SULLY: PROCEDURE: FOR DETERMISIXG EYDROXYL GROUPS 943 Xylene-Analytical-reagent grade. Pure m-xylene was used in the experiments des- cribed, but similar results are obtained with analytical-reagent grade xylene containing mixed isomers.Pyridine-Analytical-reagent grade. Sodium hydroxide solution, w-Aqueous K sodium hydroxide was used in the experiments described, but N alcoholic sodium hydroxide solution is said to give sharper end-points (personal communication from Dr. G. W. Ferguson). PKOCEDURE- Weigh the sample of hydroxylic compound into a round-bottomed 250-ml flask with 12 to 13 g of stearic anhydride, also weighed accurately, and add 10 ml of m-xylene. Stearoylate by boiling under gentle reflux for 15 to 30 minutes with the use of a water-cooled condenser with a ground joint. Two alternative procedures are available for decomposing the unreacted stearic anhydride. Procedure A involves the use of pyridine and is perhaps the better for alcohols of unknown chemical constitution or under conditions in which the alcohol is accompanied by unstable compounds.In this procedure, the xylene solution is cooled after the stearoylation reaction and to it is added 40 ml of pyridine containing 4 ml of water. The mixture is then boiled again for 15 minutes to decompose the excess of stearic anhydride. Phenolphthalein indicator is added, and the stearic acid is titrated with N sodium hydroxide solution, the solution being kept hot to avoid crystallisation. In procedure B, water and sodium stearate are used to decompose the excess of stearic anhydride. The recommended method is to cool the xylene solution at the end of the stearoylation reaction and then to add 4 mi of water and 0.6 g of sodium stearate, and to boil again under reflux for 15 to 30 minutes.The mixture is diluted with 40 ml of saponification spirit and, as before, titrated while still hot, with phenolphthalein as indicator. For both procedures the weight of the sample should be such that the titre is approxi- mately two-thirds of that equivalent to the stearic anhydride used, which corresponds t o a 50 per cent. excess over the theoretical. A blank determination is made to find the volume of standard alkali equivalent to the weight of stearic anhydride used in the deter- mination. The phenolphthalein indicator must not be added before the stearic anhydride is completely decomposed, since it is normally made up in spirit solution. I t has also been observed that the presence of unchanged stearic anhydride will cause the phenolphthalein to bleach during the titration with alkali; this is a warning that the determination must be abandoned. I thank Mrs. J. McCarthy, Miss E. Speller and Mr. R. H. Spencer for assistance with the experimental work and the Directors of A. Boake, Roberts and Co. Ltd. for permission to publish this paper. REFERENCES 1. 2. Wilson, H. N.. and Hughes, W. C., J . SOC. Chew. Ind., 1939, 58, 74. Mitchell, J., jun., Kolthoff, I. M., Proskauer, E. S., and Weissberger, A., Editors, “Organic Analysis,” Received July Sth, 1962 Interscience Publishers Inc., New York and London, 1953, Volume I, Chapter 1.

ISSN:0003-2654    

DOI:10.1039/AN9628700940

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

944 KERR, COOMBER -4ND LEWIS: USE O F BISMUTH [Vol. 87 The Use of Bismuth Radionuclides in Analysis Part II.* The Determination of Radon in Waters BY J. R. W. KERR, D. I. COOMBER BND D. T. LEWIS (Department of Scient@c and Industrial Research, Labovatory of Government Chemist, Clement's Iian Strand, London, W.C.2) Passage, A rapid radiochemical method is described for determining radon in water, based on the equilibrium of bismuth-214 with its parent radon. This bismuth isotope is precipitated from solution by n-propyl gallate, with inactive bismuth as a carrier. The precipitate of bismuth propyl gallate is collected on a filter, dried, mounted anti p-counted on a conventional instru- ment. The lower limit of sensitivity is about 20 picocuries per litre when a simple Geiger assembly is used.The storage in polythene containers of water samples taken for analysis is not recommended, as this polymeric material can absorb radon gas from aqueous solutions. RADIOCHEMICAL DETERMINATION OF RADON VIA BISMUTH-214- THE radioactive disintegration products of natural radium are produced according to the decay scheme shown in Table I, the daughter nuclei being indicated in descending order. The minor branching at radium C by a emission to thallium-210 has been omitted from the Table, since 99.96 per cent. of the disintegrations are ,& decays, the branching ratio being 2500 in favour of polonium-214. TABLE I : DECAY SCHEME FOR NATURAL RADIUM Original name Radium . . . . Radon . . . . Radium A . . . . Radium B . . . . Radium C . . . . Radium C' .. . . Radium D . . . . Radium E . . . . Radium F . . .. Atomic symbol Emission I . zzaRa U . . zaaRn U . . Z'SPO a . . 214Pb B B . . zloPb B B . . ZlOPo U . . Z14Bi . , z l r P ~ U . . 21OBi Half-life 1620 years 3.82 days 3.05 minutes 26.8 minutes 19.9 minutes 1.49 x 10-4seconds 22 years 5.0 days 138 days Energy, MeV 4-79 5.49 5.99 0.65 1.6; 3.15 7.68 0.025 1.17 5-30 The radon daughters can obviously reach equilibrium with the parent isotope in about 4 hours, and one of the standard methods for determining radon in air depends on filtering a known volume of air through a filter pervious to radon and then determining the activity of the daughter-product particulates on the filter with an ionisation chamber or similar device.1 One-hundred per cent. retention of the particulates is unfortunately difficult to achieve experimentally.Jacobi2 described an ionisation-chamber method for determining radon in natural waters, the radioactive gas being purged from solution by a stream of argon and trapped on silica gel at the temperature of liquid oxygen. For the year 1947, figures ranging from 0.26 pC per litret for the River Thames (Sutton Courtenay) to 166 pC per litre for the Pump Room water at Bath are quoted. An improved detection instrument has been described by Bryant and Mi~haelis.~ Turner, Radley and Mayneord4 have made a comprehensive survey of the natural c( activity of drinking waters drawn from various 1J.K. localities, and quote radon and daughter figures from 1 to 10,000 pC per litre at the time of sampling.The water samples were assessed by being placed in contact with a zinc - cadmium sulphide phosphor, which absorbed the daughter products of radon-222 (see Rosholt6). No rapid radiochemical method for determining radon in water has been described in the literature. A method has been developed in the laboratory based on the equilibrium * For details of Part I of this series, see reference list, p. 948. t 1 picocurie (pC) = lo-'* curies.December, 1962: R.iDIO?r’UCLIDES IN ANALYSIS. PART I1 945 of bismuth-214 with radon and its precipitation in dilute nitric acid solution in presence of 10 to 15mg of bismuth carrier. The bismuth is precipitated with 100 per cent. chemical yield as the yellow salt of propyl gallate, C,,H,,O,Bi, containing 0.4790 g of bismuth per g.A related method for determining bismuth410 in samples of lead and lead piping has already been described,, and one of the advantages of the gallate reagent is that it can separate bismuth cleanly from abundant lead. There appears to be no internationally accepted figure for the maximum permissible level of radon gas in potable waters,’ but in the United Statess a figure of 2000 pC per litre for radon $us daughter products has been suggested as a maximum figure for water supplies. The object of this investigation was to perfect a simple laboratory routine method for determining radon in a large number of samples. It can be seen from Table I that lead-214 and polonium-218 are the natural precursors of bismuth-214 and have half-lives of the same order of magnitude.When considering the removal of bismuth radionuclides by carrier precipitation with propyl gallate, it will be seen that the lead-214 will be producing bismuth-214 a t a fairly rapid rate in the supernatant solution, and it was a matter of conjecture to what extent this reaction would contaminate the first bismuth precipitate. Laboratory tests have established that the interference produced by the 3-minute polonium-218 is extremely small, i.e., experiments carried out with radium D.E.F. tracer have shown that polonium-210 is only carried bg7 the bismuth precipitate to a small extent (1 per cent.), and this factor no doubt applies to the polonium-218 isotope. In practice also we have found that the bismuth precipitate decays with a normal half-life of 19.9 minutes, and this suggests that any interference is small.It has also been experimentally confirmed that interference from the bismuth produced by the lead-214 is small; it is nevertheless desirable to maintain the conditions under which the water and the radium water standards used are compared as constant as possible. In practice, we always use 2-litre samples for the determinations and heat them to boiling under the same conditions before adding the bismuth carrier, the time of addition being accepted as zero time for the purpose of constructing the bismuth decay curve. A little lead hold-back carrier is deliberately added, but there is no evidence of the 26-minute lead-214 being carried on propyl gallate precipitates. The precipitate also appears to remain un- contaminated by natural radioactive elements, i.e., uranium, thorium, radium and polonium, which do not appear to be co-precipitated. Radon occurs naturally mainly in underground waters, in which the concentration of fission products from atomic weapons is usually negligible.Nevertheless, some experiments were carried out wherein bismuth radionuclide was precipitated in the presence of about 0.28 pC of 120-dag7 old fission products from the slow neutron fission of uranium-235. It was found that only about 0.1 per cent. of the beta - gamma of the fission products was carried on the precipitate. One experiment with a sample of natural water containing 750pC of radon, (a) with and ( b ) without fission products, gave the same bismuth-214 decay line after correction for the fission products (tail) had been applied to the decay line from (a).EXPERIMENTS WITH STANDARD RADIVM SOLUTIOXS- One-litre of standard radium solution, containing 14,200 pC of radium-226, that had been stored for sufficient time for the daughter radon to have reached equilibrium with its parent, was treated as described under “Method,” p. 947. The composite decay curve for the bismuth-210 and bismuth-214 is shown in Fig. 1. Extrapolation of the resolved bismuth-214 curve to zero time gave a concentration of this radionuclide equivalent to the radon in solution, the slope of the line corresponding to the distinctive 20-minute half-life of bismuth-214. The success of the proposed method depends essentially on the experimentally observed fact that the radon in solution is only rapidly expelled between 90” and 100” C, i.e., it remains in equilibrium with the bismuth until zero time is reached. When a solution of radon was maintained at 90” C for periods of up to 3 hours and the radon was determined as the bismuth daughter, the results were- Bismuth-214 present, % .. . . 100 89 76 68 40 14 Time, minutes . . .. .. 0 20 45 80 105 150946 KERR, COOMBER AND LEWIS: USE OF BISMUTH [Vol. 87 It must be agreed that even on the basis of these figures an error of a few minutes in the zero time is possible; nevertheless, our tests established that the method is one of the most rapid and simple techniques so far developed foI the routine determination of radon in water down to 20 pC per litre. Indeed, we found thLat if (a) a portion of a solution containing a high concentration of radium is added to 1 litrl- of water a t 90" C and the resulting solution is brought rapidly to the boil and (b) one litre of water containing an equivalent concentration of radium at 20" C is brought similarly to the boil, then within experimental error, no differences are observed in the amounts of bismuth814 obtained, as gallate, from (a) or (b).ABSOKPTION OF RADON BY POLYTHENE- In some early experiments freshly boiled aliquots of radium solution were stored in a series of polythene bottles and allowed to equilibrate; it was found that the equilibrium levels of radon determined by the bismuth method were about half the expected values, although the solution when stored in glass bottle:; gave the expected results.This suggested that polythene is a suspect container for radon solution. It is well known that radon is soluble in olive oil and is strongly absorbed by paraffin wax.9 The absorption of radon by polythene might thus have been expected. The possibility that the radon had been lost from the polythene bottles by diffusion through the plastic was eliminated by carrying out experiments in which large pellets of polythene were added to radon solutions in glass bottles. Here again, anomalous results were obtained. INVESTIGATIONS OF THE RADON GROWTH CURVE- It was considered that examination by the bismuth-214 technique of the radon growth curve in freshly boiled radium solutions would establish unequivocally the applicability of the proposed method. A solution containing 0.712 pC of radium was boiled for 4 hours, cooled, and then made up to one litre.From this litre a series of 50-ml portions were prepared in glass-stoppered 50-ml bottles having little headspace. Immediate examination of a 20-ml portion from one of the bottles diluted to 1 litre with radon-free water showed that the gallate procedure indicates only the presence of bismuth-210, ie., the solutions were initially free from radon. Further samples from the 50-ml bottles were then examined at regular intervals. The radon activity determined from the extrapolation of bismuth-214 decay curves increased with the time of siorage of the solutiop as the radon approached equilibrium with the radium, giving a growth curve for radon agreeing exactly with the well known theoretical curve.The radium standards used in this work were supplied and calibrated by U.K.A.E.A. Establishment. Amersham.December, 1M2] RADIONUCLIDES I N ANALYSIS. PART I1 947 METHOD REAGENTS- of N nitric acid, and dilute to 500 ml with water. add 50 ml of N nitric acid, and dilute to 500 nil with water. acid; warm to dissolve the solid. solution before use. Bismuth carrier solzbtion-Dissolve 2.32 g of bismuth nitrate, Bi(N0,),.5H20, in 50 ml Lead hold-back carrier solution-Dissolve 0.80 g of lead nitrate, Pb(NO,),, in water, n-Propyl gallate solwtion-Dissolve 1.0 g of n-propyl gallate in 100 ml of 0.01 N nitric If crystals form on storage, redissolve by warming the 1 ml = 2 mg of bismuth. 1 ml = 1 mg of lead. Nitric acid, 0.1 and 0.01 N. Bromocresol green indicator solwtion., Gelva solution-Dissolve 5 g of vinyl acetate polymer (Gelva 25) in 100 ml of methanol.Ethanol, absolute. PROCEDURE FOR NATURAL WATERS- Transfer 5 ml of lead hold-back carrier and 0.2 ml of bismuth carrier solutions to a 2-litre beaker. Add 1 litre of sample and then several drops of bromocresol green indicator solution, and adjust the pH of the solution to the indicator change-point by the cautious addition of 0.1 N nitric acid. Add a further 40 ml of 0.1 N nitric acid, and heat the solution rapidly to the boiling-point. At the boiling-point add immediately 5 ml of n-propyl gallate solution and then 7.3 ml of bismuth carrier solution. Continue the boiling for several minutes until the precipitate coagulates. Note the time a t which the reagent is added, i.e., zero time.Cool the solution to about 70" C in a bath of cold running water, and filter, with suction, through 21-mm Whatman No. 541 filter-paper contained in a Perspex filter stick. Wash the precipitate with 0.01 N nitric acid and then with ethanol. Dry by passing air through the precipitate, and then transfer the filter-paper and its contents to a 1-inch aluminium planchet. Mount the planchet close to the counter window in a conventional p-counting assembly, or in an anti-coincidence p-counter if the radon level is very low, the sample being covered with a 5 mg per sq. cm aluminium absorber to eliminate interference by cr-emission penetrating the mica window. Kote the counting rates at approximately 10-minute intervals over a period of 1 or 2 hours, and make one or two final observations of the counting rate after a period of 8 hours or so, to give the residual count due to the bismuth-210.If the level of the radon is low, it will be necessary for each counting period to be a full half-life of bismuth-214 for a reasonable count to be registered. The small correction, amounting to about 2 per cent., due to decay during the counting can generally be neglected.lO Plot a graph of counting rate against the mid-time of the count on semi-log paper, and extrapolate to the time of precipitation of bismuth. Subtract the activity of the residual bismuth-210 from all the points. Check the slope of the corrected curve for the bismuth-214 half-life, and convert the counts at zero time to bismuth-214 activity and hence to that of the parent radon by reference to a radium - radon standard treated under the same conditions. This standardisation procedure not only compensates for the effect of self-absorption in the precipitate, but corrects automatically for back-scatter phenomena and provides a reasonably unchanged counting geometry.Although the effect is small, it is important to note that any possible variation in the assessment of zero time is also quantitatively compensated for by comparison with a radon standard treated identically during the determination. ACCURACY OF THE METHOD- The accuracy depends on the volume of water used, the background of the fl-counter and the tail to the decay curve due to interfering radionuclides; also on the time taken for the chemical separation and mounting of the source. With a low background /3-counter having a background counting rate of 2.5 counts per minute, a sample volume of 1 litre and a counting efficiency for bismuth-214 of 25 per cent., and assuming that the sample is processed and mounted for counting in 20 minutes, the limit of detection would be 6 pC per litre. Table I1 shows the order of accuracy based on counting statistics covering four determinations at different levels of radon under the con- ditions stated above, four counts of 20 minutes' duration being made on each sample; the standard deviation of the half-life calculated from the four measurements is also shown.Add five drops of Gelva mounting solution, and dry under an infrared lamp.048 KERR, COORIEEE: AND LEWIS [Vol.87 TABLE 11: ACCURACY OF THE METHOD Standard deviation of Radon activity, Standard deviation, bismuth-214 half-life, pC per litre PC minutes 360 18 0.53 180 12 0.67 90 7.6 0.83 45 4.6 0.96 From the results it can be seen that 50 pC could be determined with 95 per cent. prob- ability to k20 per cent., and the half-life of the bismuth-214 would be determined to the same probability with a standard deviation of k2.0 minutes. The accuracy would be less when residual bismuth-210 activity was present, as this would have the effect of increasing the counter background value. If members of the thorium chain were also present, the 60-minute bismuth-212 would produce appreciable changes in the decay curve. I t may readily be calculated for a solution containing approximately 100 pC of radon per litre that, if 20 pC of bismuth-212 were present, the apparent half-life would be raised from 19.9 to 31 minutes.In our experience with most natural waters, thorium contamination has rarely been encountered. TYPICAL RESULTS- only 104 pC per litre. Table I11 shows the results obtained on natural water containing radon at a level of TABLE 111: TYPICAL RESULTS Mid-time of count from zero time, Duration of count, Counts minutes Counts minutes per minute 30 565 20 28.3 50 304 20 15.2 50 211 20 10.5 90 150 20 7.5 110 138 20 6.9 Counts per minute corrected for background and bismuth-210 24.1 11.0 6.3 3.3 2.7 The final count taken after 4 hours was 4.2 counts per minute comprising the counter background of 2.9 counts per minute and the bismuth-210 tail of 1.3 counts per minute.From these results the extrapolated zero count rate for the solution becomes 63.1 4.5 counts per minute, corresponding to the calculated initial activity of 104 k 7.4 pC of radon per litre. The experimental half-life of bismuth-214 obtained from this decay curve was found t o be 21.0 k 0.9 minutes. CONCLUSION The method described is rapid and reasonably exact for determining radon in natural waters in the range 20 to 10,000 pC per litre. It is known from research papers already publishedll that the ratio of radium to radon varies considerably for natural waters, and is generally of the order of 200 to 5000. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Courtier, G. B., Natuve, 1957, 180, 382. Jacobi, R. B., J . Chew. SOC., 1949, S.314. Bryant, J , , and Michaelis, M., U.K.A.E.A., Aniersham Research Report, RCC/R.26, 1952. Turner, R. C., Radley, J. M., and Mayneord, W. V., Nature, 1961, 189, 348. Rosholt, H. N., Anal. Chew., 1957, 29, 1398. Kerr, J. R. W., Coomber, D. I., and Lewis, D. T., Nature, 1961, 192, 547. International Commission of Radiological Protection Committee (Z), Pergamon Press, Oxford, U S . National Committee of Ra$ation Protection, N.B.S. Handbook, Washington, 1961. Hevesy, G., and Paneth, F. 4., Cook, G. B., and Duncan, J . F., “Modern Radiochemical Practice,” Oxford University Press, Grune, W’. K , Waf. G. Sewage Wks, 1962, 109, 26. Note-Reference 6 is to Part I of this series. London, New York and Paris, 1959. Radioactivity,” Oxford University Press, 1926, p. 178. 1952, p. 57. First received February 15th, 1962 Amended, July 25t12, 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700944

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

December, 19621 MONTGOMERY, DYMOCK AND THOM 949 The Rapid Colorimetric Determination of Organic Acids and their Salts in Sewage-sludge Liquor BY H. A. C. MONTGOMERY, JOAX F. DYMOCK AND N. S. THOM (Water Pollutiota Research Laboratory, Eldev Way, Stevenage, Herts.) The colorimetric composite determination of the carboxylic acids and their salts in sewage-sludge liquor is described. Since the most intense colours are given by the lower fatty acids, which also predominate in sludge liquor, the proposed method gives similar results to existing methods for determining “volatile acids.” After the removal of suspended matter, the sample (0.5 ml) is treated with ethylene glycol in the presence of sulphuric acid, and the resulting mixture of esters is determined by the ferric hydroxa- mate reaction.For acetic acid the calibration is linear to a t least 1 per cent. The method is more convenient and more precise than the distillation method of Frook. In a comparison of existing methods, the superiority of DiLallo and hlbertson’s recent titration method over the distillation procedure was con- firmed, and accurate correction factors were evaluated for different concen- trations of pure acetic acid. However, the proposed method is more con- venient experimentally, especially when a number of simultaneous deter- minations is required. The proposed method was also used successfully for determining “organic acids” in an effluent from a bacon factory. SIGNIFICANT concentrations of salts of the lower fatty acids (the “volatile acids”) are formed in the anaerobic digestion of sewage sludge and other organic wastes.During normal operation the volatile acids are converted to methane and carbon dioxide. Break-down of the digestion process is often accompanied by a sudden rise in the concentration of volatile acids, and frequent checks of this quantity are made. The individual acids are seldom determined separately in current practice, and it is usual to obtain a composite figure by titrating the distillate from an acidified sample of sludge or sludge liquor. The importance of the volatile acids in the particulate matter of whole sludge is not at present clear. H. A. Painter, in a personal communication, stated that he had found that 91.5 per cent. of the volatile ether-extractable acids in a sample of crude sludge was in solution or fine suspension; in a digested sludge the proportion was 60 per cent.The work described in this paper deals exclusively with sludge liquor, which is much more easily analysed than whole sludge. The distillation methods a t present in use are rather tedious and require large samples; moreover, the proportions of individual volatile acids recovered depend on the experimental conditions. Acetic acid, which does not form an azeotropic mixture with water, is particularly difficult to recover. The chromatographic method of Mueller, Buswell and Larsonl affords a fairly rapid separation of the principal volatile acids, but Buswell, Boring and Milam2 later reported that this method did not give reproducible results without preliminary concentration of the sample. The paper-chromatographic procedure described by the latter workers is only semi-quantitative.Gas chromatography does not seem to have been applied to sludge liquor. A method based on ether-extraction3 is time-consuming. DiLallo and Albertson4 have recently described a rapid procedure in which the acidified carbon dioxide free sample is titrated with alkali from pH 4.00 to 7.00. An empirical con- version factor is used to calculate the volatile-acid concentration. In the work described here DiLallo and Albertson’s results were confirmed, and more accurate conversion factors were evaluated. The proposed method, which is also empirical in nature, is based on the well known colorimetric ferric hydroxamate method for determining carboxylic esters.The sample is heated with ethylene glycol and sulphuric acid, and the resulting mixture of esters is allowed to react with hydroxylamine. The hydroxamic acids formed are converted to their ferric950 MONTGOMERY, DYMOCK AND THOM : RAPID COLORIMETRIC [Vol. 87 complexes and determined by optical-density measurements at 500 mp. Under the experi- mental conditions chosen, the most intense colours are given by the volatile acids. Although the ferric hydroxamate reaction has been used6 for determining anhydrous carboxylic acids, after treatment with diazomethane (or thionyl chloride and methanol), this appears to be the first time it has been used for determining the volatile acids and their salts in aqueous solution. EXPERIMENTAL ESTERIFICATION- A solution of pure sodium acetate, equivalent to 2000 p.p.m.of acetic acid, was used in studying the esterification. The simplest approach appeared to be to use a large excess of a non-volatile alcohol that was miscible with water in all proportions, with sulphuric acid as catalyst. Glycerol, ethylene glycol and 2-methoxyethanol were tested ; ethylene glycol was finally selected since it gave good yields of ester and low blank values. Purification of the glycol was usually necessary for consistent results. Good yields of 2-hydroxyethyl acetate were obtained with a ratio of glycol to water of 2.5 to 1 by volume and a sulphuric acid concentration of 0.9 M. Since large increases in the amounts of glycol and sulphuric acid gave only small increases in the yield of ester, these conditions were chosen as standard, for convenience. It is necessary to heat the mixture.When 2.2 ml of the mixture described above are heated in a boiling-water bath the yield of ester nearly reaches its maximum after 2 minutes, and after 4 minutes a gradual decay begins. The best results were obtained by heating for 3 minutes, and it was also found advantageous to cool the mixture rapidly to room temperature. Different volumes of reaction mixture have different temperature - time relationships, leading to different yields of ester, so that for consistent results the volumes used must be as specified under “Procedure” (see p. 953). The degree of esterification was determined after the conditions for colour development had been standardised. The optical density at 500 mp given by 0.01 M acetic acid was com- pared with the maximum possible optical density, obtained by adding an equivalent amount of 2-hydroxyethyl acetate dissolved in the hydroxyammonium sulphate reagent instead of in the sample solution.From the results in Table I, it can be calculated that the yield of ester is 69 per cent. of theoretical. By carrying out the whole procedure on 0.01 M 2-hydroxy- ethyl acetate solution, it was also shown that the equilibrium concentration of ester is nearly reached. TABLE I COMPARISON OF YIELD OF FERRIC HYDROXAMATE COMPLEX FROM ACETIC ACID AND FROM 2-HYDROXYETHYL ACETATE Optical density a t 500 mp Test solution (4-cm cells) Acetic acid, 0.01 M . . . . . . . . . . . . . . . . 0,440, 0.441 2-Hydroxyethyl acetate, 0.01 M . .* . . . . . . . 0.442, 0.472 } 0.433, 0.432 Alkaline hydrolysate from 2-hydroxyethyl acetate, 0.01 M (as check on . . . . . . . . . . . . . . purity) . . * . 0.636, 0.645 Same weight of 2-hydroxyethyl acetate added with hydroxyammonium 1 . . . . . . .’I sulphate reagent . . . . . . FORMATION OF SODIUM HYDROXAMATE- The hydroxamate is formed almost instantaneously without external heating in the presence of hydroxylamine (not less than 0.1 M) and excess of alkali (a slight time factor was observed when comparatively high concentrations of acetate were being determined). Glycine, which has been used as a buffer in a mixture of organic solvents,B was found to inhibit the reaction in the aqueous medium used in the work described here, and the best results were obtained with excess of alkali.December, 19621 COLOUR DEVELOPMENT- stances, including sulphate, that form ferric complexes.involved are given below- DETERMINATION OF ORGANIC ACIDS IN SEWAGE-SLUDGE LIQUOR 95 1 The optical density obtained depends on the acidity and on the concentration of sub- The results of a study of the variables 1. When the pH value of the solution (before final adjustment of the volume to 26 ml) exceeded 2.0, brown solutions and high blank values were obtained; suppression of the colour became severe below pH 1.2. Between pH 2.0 and 1.2 the stability of the colour improved as the pH was lowered. 2. For a given pH value and concentration of ferric chloride, more intense colours and slightly higher blank values were obtained with perchloric than with sulphuric acid.The maximum colour intensity could also be obtained with sulphuric acid by using a large excess of ferric chloride. Blank values were high when ferric perchlorate was used. 3. Effervescence, believed to have been due to the oxidation of hydroxylamine by the ferric salt, was least at low pH values and in the presence of sulphuric acid. Destruction of the excess of hydroxylamine by preliminary oxidation or oxime formation always gave high blank values. 4. The use of glycine as a buffera led to high blank values with perchloric acid and to suppression of the colour with sulphuric acid. The procedure finally adopted was the addition of the hydroxamate solution to a large excess of ferric chloride solution containing sufficient sulphuric acid to give a final pH value of 1.6 (before dilution).Effervescence is not serious, and a change of 0.1 pH unit alters the optical density by less than 2 per cent., an error that is insignificant in the analysis of sewage sludge. Since sulphuric acid itself has considerable buffering action at this level, a pH value of 1.6 0.1 is easily achieved, provided that reasonable care is taken in preparing and measuring reagents. The colour obtained is fairly stable, decaying at a rate of 1.3 per cent. per hour, measured over 3 hours or 0.5 per cent. per hour, measured over 19 hours. RESPONSE OF PROPIONIC AND BUTYRIC ACIDS- Several workers197 have shown that acetic, propionic and butyric acids constitute by far the greatest proportion of the volatile acids present in digesting sludge. For example, in a study7 of the effect of the rate of loading on digester performance, as the loading was increased the proportion of acetic acid (possibly containing a little pyruvic acid) fell from 68.9 to 47.3 per cent.of the volatile acids determined. The proportions of propionic and butyric acids rose from 23.0 and 4-4 per cent. t o 32.5 and 19.7 per cent., respectively. The other acids determined contributed less than 4 per cent. to the total. In the proposed method, the optical densities at 500 mp for 0.01 K solutions of calcium propionate and n-butyric acid were, respectively, 88 and 87 per cent. of that obtained from 0.01 K acetic acid. Moderate changes in the ratio of the three principal volatile acids will therefore have only a small effect on the results. RESPONSE OF OTHER POSSIBLE CONSTITUEXTS- Buckles and Thelens have described the behaviour of a large number of compounds in their qualitative test for esters by the hydroxamate reaction.Their results suggest that amides and imides would be the only substances likely to interfere under the conditions of the work described in this paper, apart from traces of dissolved esters and coloured and colloidal material. We have determined the responses of a number of representative compounds, and the results are shown in Table 11. The sample solutions were 0.01 N ; in preparing the solutions the equivalent weight of each compound was taken as the molecular weight divided by the number of carboxyl or amide groups, e g . , the citric acid was 0.0033 M and the sodium hippurate 0.005 M.Support for the view that non-volatile carboxylic acids do not occur in large quantities in sludge liquor is given by Painter’s work at the Water Pollution Research Laboratory; he found that the ether-extractable non-volatile acidity in a crude sludge liquor amounted952 MONTGOMERY, DYMOCK AND THOM : RAPID COLORIMETRIC TABLE I1 [Vol. 87 RESPONSE OF VARIOUS CARBOXYLIC COMPOUNDS IN THE PROPOSED METHOD Optical density a t 500 mp, as a percentage of that obtained for acetic acid Test substance Formic acid . . . . . . 42 Calcium lactate . . . . . . 58 Pyruvic acid . . . . . . 31 Tartaric acid . . . . . . 30 Citric acid . . . . . . . . 17 Oxalic acid . . . . . . 3s Malonic acid . . . . . . 46 Succinic acid . . . . . . 64 Sodium stearate .. . . . . Precipitation occurred Glycine . . . . . . . . 7 Glutamic acid . . . . . . 3 Acetamide . . . . . . . . 34 Sodium phthalimide . . . . 11 Sodium hippurate . . . . 28 to less than one-sixth of the volatile acidity (both expressed as acetic acid). In a digested sludge liquor the proportion was even less. Confirmation is afforded by the fact that values for volatile acids determined by the potentiometric-titration method of DiLallo and Albertson were not excessively higher than those obtained by distillation (this was not so with a bacon- factory effluent, for which the titration method gave extremely high results). An attempt was made to assess the Contribution of coloured and colloidal impurities in a filtered crude sludge liquor, containing 960 p p.m. of volatile acids, by treating the sample as described in the proposed method, but with the omission of the hydroxyammonium sulphate reagent.For a digested sludge, the close agreement, shown in Table 111, between the three methods suggests that the contribution of impurities was extremely small. By using appropriate experimental conditions, the method could probably be adapted for determining a large number of carboxylic acids in dilute aqueous solution. RANGE AKD ACCURACY OF METHOD- If purified ethylene glycol is used, 13 p.p.m. of acetic acid (corresponding to twice the standard deviation of the blank value) can be detected in pure solution. The sensitivity in sewage sludge, however, is limited by the presence of coloured and colloidal matter, as described above.The calibration is linear to a t least 1 per cent. of acetic acid (added as sodium acetate), beyond which no measurements were made. Coefficients of variation of approximately 2 per cent. were obtained from multiple determinations on pure solutions of sodium acetate at three different concentrations. The optical density obtained was equivalent to 75 p.p.m. of acetic acid. METHOD REAGENTS- Sulphcuric acid, diluted-Mix exactly equal vdumes of analytical-reagent grade sulphuric acid, spgr. 1.84, and water. Acidic ethyle?ze glycol reagent-Mix 30 ml of ethylene glycol with 4 ml of the diluted sulphuric acid. Prepare this reagent freshly each day. If the blank value exceeds 200 p.p.m. (as acetic acid), it is essential to purify the ethylene glycol by distillation from sodium hydroxide under reduced pressure.Sodium hydroxide, 4.5 x-Dissolve 180 g of sodium hydroxide, from a freshly opened bottle, in water, and dilute to 1 litre. Hydroxyammonium szdphate solution, 10 per cent. Hydroxylamiize reageizt-Mix 20 ml of 4.5 N sodium hydroxide with 5 ml of hydroxy- Acidic ferric chloride reagent-Dissolve 20 g of ferric chloride hexahydrate in 500 ml of ammonium sulphate solution just before use. water, add exactly 20 ml of concentrated sulphuric acid, and dilute to 1 litre.December, 19621 DETERMINATIOK OF ORGANIC ACIDS IN SEWAGE-SLUDGE LIQUOR 953 PROCEDURE- Clarify the sample by filtration with the use of filter-aid; in favourable cases adequate clarification can be obtained by spinning the sample in a centrifuge. Measure 0.5ml of the liquor into a dry test-tube (12.5 x 1.5 cm). Add 1.7 ml of acidic ethylene glycol reagent from a burette, and mix thoroughly (alternatively, add 1.5 ml of ethylene glycol and 0.2 ml of diluted sulphuric acid).Heat in a boiling-water bath for 3 minutes, avoiding direct contact between the test-tube and the heating element, then immediately cool the test-tube in cold water. Add 2.5 ml of hydroxylamine reagent (or 0.5 ml of hydroxyammonium sulphate solution and 2.0 ml of 4.5 N sodium hydroxide), and mix. If the concentration of volatile acids is expected to exceed 5000 p.p.m., set aside for 1 minute. Xeasure 10 ml of acidic ferric chloride reagent into a 25-ml calibrated flask, and add the solution from the test-tube; use water to rinse out the last traces.Make up to the mark with water, and shake the flask vigorously. Set aside for 5 minutes without the stopper (to facilitate the escape of dissolved gases), then measure the optical density at 500 mp, taking care to avoid the formation of gas bubbles in the optical cell. Subtract the blank value obtained with 0.5 ml of distilled water, and ascertain the organic acid concentration from a calibration graph prepared with pure solutions of acetic acid, Take the reading within 1 hour of colour development. RESULTS The proposed method was compared directly with the distillation method of Frookg and the titration method of DiLallo and Albertson,* which was modified slightly in two respects- 1. The boiling to remove carbon dioxide was carried out under reflux, to eliminate the risk of losing butyric acid.2. Since the range of the potentiometric titration, from pH 4.00 to 7.00, is arbitrary, a conversion factor must be used in calculating the results. The conversion factor varies with concentration, as the degree of ionisation of weak acids is dependent on concentration. DiLallo and Albertson, who state that their pH meter gave high readings in the region of pH 7.00, used only two conversion factors, one for concentrations above and one for concen- trations below 250 p.p.m. We have determined the conversion factor for six concentrations of pure acetic acid with an accurate pH meter; the results are shown in Fig. 1. Similar results were obtained with butyric acid. These conversion factors, which improved the accuracy of / I I I 500 I000 1500 2000 Acetic acid found by titration, p.p.m.5 8 i V 8 2 5 Fig. 1. Conversion factors for determining organic acids by potentiometric titration between pH 4.00 and 7.00954 MONTGOMERY, DYMOCK AND THOM : RAPID COLORIMETRIC [Vol. 87 the results for sewage sludge, differ from those used by DiLallo and Albertson in that they are applied to the observed concentrations instead of to the “volatile acid alkalinity.” The importance of measuring the initial pH value (4.00) as accurately as possible was confirmed. The effect of phosphate has been disc~ssed.~ Results obtained by the three methods on sludge samples, some of which were fortified with known amounts of pure sodium acetate solution, are shown in Table 111. Each figure is the mean of two or three determinations. TABLE I11 COMPARISON OF THREE METHODS FOR DETERMINING ORGANIC ACIDS I N ANAEROEXC MEDIA (Concentrations are expressed as p.p.m.of acetic acid) Results obtained by- 7 ~~ Treatment proposed method titration distillation of sample (pA-, r-A-, containing Added Added Added Acetate added acetate acetate acetate Sample added acetate* Found recovered Found recovered Found recovered f : F C 1190 1120 - 1190 960 - 1210 250 - 1430 470 - 1480 290 - F 1380 C 1580 1670 480 - C F 240 sewage Effluent 0 { 60: factory 600 F 160 C 150 F C F 600 C F C 810 F 1390 C 1460 F C - 160 330 540 770 1710 2390 - 190 425 665 - - - - 670 150 360 210 450 300 820 670 - 170 - 360 - 610 - 1360 1870 490 - 680 - * F = Filtered with use of Hyflo filter-aid; coagulants was also used in samples for distillation.C = Spun in a centrifuge at 7000 r.p.m. for 10 minutes. Since the results for samples that had been spun in a centrifuge were not significantly different from those for filtered samples, the mean values were used in calculating the recoveries of added acetate. An approximate assessment of the precision of each method was made by calculating the coefficients of variation of the total concentrations found. Each individual value was first expressed as a percentage of the mean expected value; a single coefficient of variation was then calculated for each group of three fortified samples. The results by the proposed method were 2.7 per cent. for the crude sludge and 5.5 per cent. for the digested sludge. The corresponding figures for the titration method were 1.4 and 3.8 per cent., while the distillation method gave 6.0 and 17.6 per cent. These results, though obtained on a limited number of samples, suggest that the titration procedure is the most precise.The precision of the distillation method, though poor, is probably adequate for the routine control of sludge digestion-experience has shown that it is rather difficult to stop the distillation a t exactly the right time. The bias in the colorimetric and distillation methods is less than the coefficient of variation and is probably not significant. An apparent bias in the results for crude sludge by the titration method may be due to a low initial value. The agreement between the three methods is surprisingly good, considering that, in each, different sample constituents are being included.December, 19621 DETERMINATION OF ORGANIC ACIDS IN SEWAGE-SLUDGE LIQUOR 955 A single determination by the proposed procedure takes 25 minutes, includiag time for preparation of the sample, and 16 determinations can be completed in 2 hours.A deter- mination by the titrimetric procedure takes 30 minutes unless it can be combined with a determination of total alkalinity, as described by the authors.4 The proposed procedure is much more suitable for multiple determinations than the titrimetric method ; each is more rapid than the distillation procedure. Another advantage of the proposed procedure is that the volume of sample required is smaller and therefore less malodorous. The small volume of sample is particularly useful in research work when unlimited amounts may not be available. CONCLUSIONS The proposed procedure is suitable for determining organic acids and their salts in sewage-sludge liquor, and the results are an approximate measure of the concentration of volatile acids. The procedure is quicker, more convenient and more precise than the dis- tillation procedure in which phosphoric acid is used, and it is also quicker and more suitable for routine work than a recently described titrimetric method. This paper is published by permission of the Department of Scientific and Industrial Research. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. Mueller, H. F., Buswell, A. M., and Larson, T. E., Sewuge Ind. Wastes, 1956, 28, 255, Buswell, A. M., Boring, J. R., and Milam, J. R., J . Wat. Pollut. Control Fed., 1960, 32, 721. Thomas, J . F., Wherry, C. R., and Pearson, E. 4., Proc. 10th Industr. Waste Conf., Purdue Univ., DiLallo, R., and Albertson, 0. E., J . Wat. Pollut. Control Fed., 1961, 33, 356. Hill, U. T., Ind. Eng. Chem., Anal. Ed., 1946, 18, 317. Gey, K. F., and Schon, H., Hoppe-Seyl. Z., 1956, 305, 149. Mueller, L. E., Hindin, E., Lunsford, J. V., and Dunstan, G. H., Sewage Iwd. Wastes, 1959,31, 669. Buckles, R. E., and Thelen, C. J., Anal. Chem., 1950, 22, 676. Frook, J. E., Sewage Ind. Wastes, 1957, 29, 18. 1965, 267. Received ,/ifly i l k . 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700949

出版商:RSC    

年代:1962

数据来源: RSC