ISSN:0003-2654    

DOI:10.1039/AN95681FX043

出版商:RSC    

年代:1956

数据来源: RSC

摘要:

No. 40 September, 1956 THE SOCIETY FOR ANALYTICAL CHEMISTRY BULLETIN FORTHCOMING MEETINGS Ordinary Meeting of the Society, October 3rd, 1956 AK Ordinary Meeting of the Society will be held at 7 p.m. on Wednesday, October 3rd, 1956, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The subject of the Meeting will be “Chromatography” and the following papers will be presented and discussed :-- “The Determination of Vitamin D and Related Compounds. Part I : Introduction and Preparation of Compounds in the Irradiation Series. Part I1 : Analysis of Irradiation Products,” by W. H. C. Shaw, F.P.S., F.R.I.C., J. P. Jefferies, BSc., A.R.I.C., and T. E. Holt, BSc., A.R.I.C. “Some Examples of the Use of Paper Chromatography in Toxicological Analysis,” by A.S. Curry, MA., Ph.D. Joint Meeting of the North of England Section and the Physical Methods Group, October 19th, 1956 A JOINT Meeting of the North of England Section and the Physical Methods Group will be held at (5.30 p.m. on Friday, October 19th, 1956, at the Robinson Laboratory, The University, Manchester. The subject of the meeting will be “Ion Exchange.” Joint Meeting of the Scottish Section with the Methods of Analysis Panel (Glasgow), September 28th, 1956 A JOINT Meeting of the Scottish Section and the Methods of Analysis Panel (Glasgow) will be held at 7.15 p.m. on Friday, September ZSth, 1956, at the Central Station Hotel, Glasgow. The following papers will be presented and discussed:- “The Analysis of Titanium and its Alloys,” by W. T. Elwell, F.R.I.C.“The Estimation of Molybdenum,” by R. Kerr, BSc., A.R.I.C. Joint Meeting of the Scottish Section with the Stirlingshire and District Sections of the Royal Institute of Chemistry and the Society of Chemical Industry, October 23rd, 1956 A JOINT Meeting of the Scottish Section with the Stirlingshire and District Sections of the Royal Institute of Chemistry and the Society of Chemical Industry will be held at 7.30 p.m. onTuesday, October 23rd, 1956, at the Lea Park Restaurant, Falkirk. The Meeting will take the form of an Exhibition of Special Analytical Apparatus. Joint Meeting of the Western Section with the S.W. Counties Section of the Royal Institute of Chemistry, October 6th, 1956 THE South-Western Counties Section of the Royal Institute of Chemistry will hold a Sympos- ium on Saturday October, 6th, 1956, at the Technical College, Plymouth, in which the Western Section of the Society has been invited to participate.The subject of the Symposium will be “Chemistry as a Career.”Joint Meeting of the Western Section with the Society of Chemical Industry, October loth, 1956 JOIXT Neeting of the Western Section with the Society of Chemical Industry will be held at 7 p.m. on Wednesday, October loth, 1956, at the Technical College, Newport. The following lecture will be given :- “Sequestering Agents in Analytical Chemistry,” by Dr. I ~ . L. Smith. Ordinary Meeting of the Midlands Section, September 27th, 1956 a?j Ordinary Meeting of the Section will be held at 7.15 p.m. on Thursday, September Z t h , 1956, at the Lecture Theatre, Derby and District College of Art, Green Lane, Derby.The following lecture will be given :- “High-precision Absorptiometry,” by \V. T. L. Seal, B.A., A.R.I.C. Joint Meeting of the Midlands Section with the Microchemistry Group, October 5th, 1956 A JOIST Meeting of the Midlands Section with the Microchemistry Group will be held at 6.30 p.m. on Friday, October 5th, 1956, at the University Chemical Laboratory, Pembroke Street, Cambridge. The subject of the Meeting will be “Sub-micro Methods in Inorganic and Organic Analysis” and the following papers will be presented and discussed :- Introduction by R. Belcher, Ph.D., D.Sc., F.R.I.C. “General Review of Sub-micro Methods,” by T. S. West, BSc., Ph.D., A.R.I.C. “The Determination of Alkoxyl,” by M. K. Bhatty, MSc., A.R.I.C.“The Determination of Xitrogen,” by M. Williams, B.Sc., A.R.I.C. “The Determination of Iodine,” by A. R. Shah, MSc., A.R.I.C. The Meeting will be preceded at 2 p.m. by a visit by kind permission of Messrs. Fisons Pest Control Ltd. to their Research Station at Chesterford Park. Ordinary Meeting of the Midlands Section, October 25th, 1956 AN Ordinary Meeting of the Section will be held a t 7 p.m. on Thursday, October 25th, 1956, at the Gas Showrooms, Nottingham. The following lecture will be given:- “Recent Advances in Ion-exchange Resins,” by D. K. Hale. BRITISH STANDARDS INSTITUTION DRAFT SPECIFICATIONS A FEW copies of the following draft specifications, issued for comment only, are available to members of the Society, and can be obtained from the Secretary, The Society for -4nalytical Chemistry, 7-8 Idol Lane, London, E.C.3.Draft Specifications prepared by Technical Committee FHC/4-Solvents and Allied Products. CW(FHC)7028-Draft B.S. for Di-n-butyl Phthalate (Revision of B.S. 573). CW(FHC)7029-Draft B.S. for Diethyl Phthalate (Revision of B.S. 574). CW(FHC)7030-Draft B.S. for Acetone (Revision of B.S. 509). CW(FHC)7031-Draft B.S. for Diacetone Alcohol (Revision of B.S. 549). CU’(FHC)7032-Draft B.S. for o-Dichlorobenzene (Grades A and B). CW(FHC)7033-Draft B.S. for Hexachloroethane (Revision of B.S. 577). Draft Specification prepared by Technical Committee LBCjll-Microchemical Apparatus. CW(LBC)6476-Draft B.S. for Vaporimetric Molecular Weight Determination Apparatus (Part K1 of R.S. 1428, Microchemical Apparatus).Draft Specification prepared by Technical Committee LBC/l l/l-Apparatus for Micro- Cl%’(LBC)6812-Draft B.S.for Rapid Method Combustion Tubes (Belcher and Draft Specification prepared by Technical Committee OSC/4-Essential Oils. Draft Specification prepared by Technical Committee DIC/l-Chemicals for Chemical C\V(DIC)7081-Draft B.S. for Chemical Preparations for Chemical Closets (Re- chemical Combustion Methods. Ingram Type) (Part -45 of B.S. 1428). CW(OSC,)’il72-Draft B.S. for Essential Oils. Closets. circulation and Portable Types) for Aircraft. Draft Specification prepared bq- the Conference on Artists’ Materials M/36. CW’(M)7003-Draft B.S. Definitions of Powder Pigments for Artists’ Paints. Draft Specification prepared by Technical Committee ISE/18-Sampling and Analysis CIV(ISE)6505--Draft US.methods for the Analysis of Iron and Steel (Determination of Iron and Steel. of Sickel and Gravimetric Determination of Sulphur). Draft Specification prepared by Technical Committee PCC/I. CLV(PHC)7616-Draft B.S. for Recommended Common Names for Pest Control Products (Addition to Revision of B.S. 1831). COMMUNICATIONS ACCEPTED FOR PUBLICATION IN THE ANALYST THE following communications have been accepted for publication in The AnaZyst, and are expected to appear in the near future. “The Determination of Mercury in the ;\tmosphere: a Modified Method,” by R. G. Drew and E. King. “Turbidimetric Determination of Chlorine in Titanium,” by H. J. G. Challis and J. T. Jones. “X Continuous Ether Extractor,” by K. J. Jensen and R.W. Bane. “The Determination of Vitamin D and Related Compounds. (Apparatus.) Part I : Introduction, and Preparation of Compounds in the Irradiation Series,” by W. H. C. Shaw, J. P. Jefferies and T. E. Holt. “The Determination of Vitamin I> and Related Compounds. Part 11: Analysis of Irradiation Products,” by LV. H. C. Shaw and J. P. Jefferies. “The Determination of Moisture in Dieldrin and Aldrin by the Karl Fischer Titration Method,” by K. 1;. Sporek. “The Determination of Iron in Iron Ores, Slags and Refractories by Thioacetamide Reduction,” by P. H. Scholes. “The Determination of Mercury in Air,” by G. A. Sergeant, B. E. Dixon and R. G. Lidzey. “Determination of Uranium with -4inmonium Thiosulphate and Sodium Hyposulphite,” by H. X. Ray and S. P. Bhattacharayya. “*in Improved End-point in Iodimetric Titrations,” by G. J. Kakabadse, B. Manohin and RII. M. Crowder. (Note.) “Developments in the Micro Vacuum Fusion Method with Particular Reference to the Determination of Oxygen, Nitrogen and Hydrogen in Beryllium, Titanium, Zircon- ium, Thorium and Uranium,” by E. Booth, F. J. Bryant and A. Parker. (Kote.) “Decomposition of Oxide Minerals by Fusion with Borax,” by P. G. Jeffery. (Note.) PRINTED BY W. HEFFER & SONS LTD.. CAMBRIDGE. ENGLAND

ISSN:0003-2654    

DOI:10.1039/AN956810X045

出版商:RSC    

年代:1956

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95681BX048

出版商:RSC    

年代:1956

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95681FP105

出版商:RSC    

年代:1956

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95681BP113

出版商:RSC    

年代:1956

数据来源: RSC

摘要:

SEPTEMBER, 1956 Vol. 81, No. 966 THE ANALYST PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY DEATH William Wilson. WE record with regret the death of NORTH OF ENGLAND SECTION THE Nineteenth Summer Meeting of the Section was held at the Imperial Hotel, Llandudno, from Friday, June 15th, to Monday, June 18th, 1956. The Chairman of the Section, Mr. J. R. Walmesley, A.M.C.T., F.P.S., F.R.I.C., presided. On the morning of Saturday, June 16th, Mr. F. L. Okell, F.R.I.C., Advisory Editor of The Agzalyst, gave a talk entitled “Some Memories of the Last Fifty Years,” in which he outlined the part played by the Society and its members in the advancement of analytical chemistry during the last half-century, WESTERN SECTION THE Summer Meeting of the Section was held at Bath from Friday, June lst, to Monday, June 4th, 1956.The scientific sessioii was opened on Saturday, June Znd, in the Pump Room by the Mayor of Bath, Councillor S. A. Smith. Mr. H. N. Wilson, F.R.I.C., introduced the film “The Technique of Sampling” (Imperial Chemical Industries Ltd.) , which shows the general principles of sampling and the application of those principles to manual and automatic sampling. When the film had been shown, the following papers were presented and discussed: “Sampling with Regard to Foods and Drugs,’’ by A. Tyler, M.B.E., F.R.S.H., F.S.I.A. (Chief Sanitary Inspector of the City of Bath); “A Few Comments on the Administration of the Fertilisers and Feeding Stuffs Act,” by C. J. Sears, M.I.W.M.A. (Chief Inspector of Weights and Measures, Wiltshire County Council). After lunch in the Pump Room, a visit was paid to the Roman Baths and Guildhall, and in the evening a dinner was held at the Vineyard, Colerne, near Bath. On the Sunday afternoon, members and their guests were taken on a coach tour of Limpley Stoke, Bradford- on-Avon, Lacock and Lacock Abbey, returning via Castle Coombe to Bath. 505

ISSN:0003-2654    

DOI:10.1039/AN9568100505

出版商:RSC    

年代:1956

数据来源: RSC

摘要:

506 FERKETT, MILNER, SHALGOSKY ANL, SLEE: A COMPAKATIVE [Vol. 81 A Comparative Study of Three Recently Developed Plolarographs BY D. J. FERRETT, G. W. C. MILNER., H. I. SHALGOSKY AND L. J. SLEE (Presented at the meeting of the Physical Methods Group on Tuesday, February 14th, 1956) The recently developed instruments include the cathode-ray polarograph, the square-wave polarograph and the Cambridge Univector polarograph unit. These instruments have been tested to obtain information on their relative merits and details of the results obtained are described. The tests included the following aspects : (i) sensitivity for reversible and irreversible reductions a t the dropping-mercury electrode, (ii) resolution for elements with half-wave potential values very close together, (iii) effects of the reduction of a major constituent a t a more positive potenti.al on the determination of a minor constituent, and (iv) speed of applicatjon, reproducibility and usefulness in analytical chemistry.DESPITE statements about the relative merits of high-sensitivity derivative polarographs,l y 2 no comparative study of their behaviour has been made. L4 series of solutions has, therefore, been examined by using three polarographs that are more sensitive than those hitherto available commercially. These instruments are- (a) the Univector unit, manufactured by Cambridge Instrument Co. Ltd., (6) the Mervyn - Harwell square-wave pclarograph, manufactured by Mervyn Instru- ments, and (c) the single-sweep cathode-ray polarograph, manufactured by Southern Instruments Ltd.These will be referred to as CU, SWP and CKP, respectively, in this paper. The CU unit in conjunction with a coiiventional polarograph produces derivative polarograms, and the Cambridge Instrument Co. Ltd. claim that it increases the sensitivity of such an instrument by a factor of up to t ~ e n t y f o l d . ~ For these tests, the CU was coupled to a Cambridge pen-recording polarograph. The SWP produces only derivative polarograms. The CRP used was the original instrument constructed by Davis and S e a b ~ r n , ~ but an almost identical instrument is now manufactured by Southern Instruments Ltd. This polarograph will give both its own typical polarogram (which is in fact semi-derivative) and the derivative of this. Whenever possible, both types of waves were obtained and recordings were made by using a 35-mm oscilloscope camera.EXPERIMENTAL For derivative polarograms it is not necessary to remove oxygen from the solutions except for the high-sensitivity settings of all three instruments. For normal polarograms on the CRP and Cambridge polarographs, however, the solutions must be free from dissolved oxygen. This was ensured by passing hydrogen through the solution for at least 10 minutes before polarograms were recorded. For examination with the SWP and CU, the base electrolyte must not be less than M , so that the electrical resistance of the cell shall be small. The CU and CRP use any ordinary polarographic cells, but, for the best results at low concentrations, the SWP requires a specially constructed cell in which the drops of mercury do not fall into the anode.The CRP and SWP require the leads to the electrodes to be .screened. SENSITIVITY- hydrochloric acid was studied, the results being shown in Fig. 1. cadmium waves measured were- RESULTS (a) For reversible reductions-A solution containing 0.18 pg of cadmium per ml in M The peak heights of the507 Sept., 19561 STUDY OF THREE RECENTLY DEVELOPED POLAROGRAPHS SWP . I . . . . 240 mm at maximum sensitivity derivative CRP . . . . . . 5 mm at maximum sensitivity cu . . , . . . 20 mm at maximum sensitivity normal CRP . . . . . . 40 mm at 1/5 maximum sensitivity normal Cambridge polarograph . . The base-line with the normal CRP presentation at 1/5 maximum sensitivity sloped so steeply that it was not possible to measure the peak height at any sensitivity greater just visible, say 1 mm, a t maximum sensitivity. i ” ” - 0.4 - 0.6 Potential against mercury-pool anode, volts 1 h --I_ .- 5 0 > ._ U YL .- 5 24 v U c aJ t -0’5 -0‘6 -0‘7 -0.8 Potential against mercury-pool anode, volts Fig.1 ( b ) , part 1. Cathode-ray polarogram for a 0.1 8 pg per ml solution of cadmium in M hydrochloric acid. Normal polarogram lh Cd 2 -0‘3 -0’4 -0’5 -0‘6 -0’7 Potential against mercury-pool anode, volts 3 u Fig. 1 ( a ) . Square-wave Fig. 1 ( b ) , part 2. Cathode-ray polaro- polarogram for a 0.18 p g per gram for a 0.18 pg per ml solution of ml solution of cadmium in M cadmium in M hydrochloric acid. Derivative polarogram hydrochloric acid Damping = 6’0 Derivative u Normal J I I - 0‘4 -- 0’8 Potential against mercury-pool anode, volts -0’4 - C‘8 Potential against mercury-pool anode, volts Fig.1 (c). Univector polarograms for a 0.18 pg per ml solution of cadmium in M hydrochloric acid508 FERRETT, MILNER, SHALGOSKY AND SLEE A COMPARATIVE [Vol. 81 than this. (One cause of the sloping base line is the presence of impurities in the base electrolyte and since for comparative tests all solutions were M such interference was un- avoidable.) With more dilute electrolytes it is possible to increase the usable peak height. The ratios of peak heights measured were: normal polarograph to derivative CRP to CU to normal CRP to SWP = 1 to 5 to 20 to 40 to 240. (b) For irreversible reductions- (i) A solution containing 5 pg of nickel per ml in M potassium chloride solution was studied.The peak heights measured were- SWP . . . . . . 90 min at maximum sensitivity CU . . . . . . - 2 rnm at maximum sensitivity derivative CRP . . . . . . - 2 mm at maximum sensitivity. On the CU, the derivative CRP and the SWP the wave had a very broad base. On the normal circuit of the CRP the wave did not have the usual peaked shape but was more normal CRP . . . . . . 50 min at 1/5 maximum sensitivity like the stepped shape of the conventional polarogram. 7- , 1 - 0'9 -1-0 -1.1 Potential against mercury-pool anode, volts Fig. 2 (a). Square-wave polarogram for a 5 pg per ml solution of nickel in M potassium chloride I .- -1. X I Potential against mercury-pool anode, volts Fig. 2 ( b ) , part 1. Cathode-ray polarogram for a 5 pg per ml solution of nickel in M potassium chloride. Normal polarogram L C -0'9 -1.0 - 1 .1 -1.2 5 2 W Potential against mercury-pool anode, volts Fig. 2 ( b ) , part 2. Cathode-ray polaro- gram for a 5 pg per ml solution of nickel in M potassium chloride. Derivative polarogram With the CRP there is a considerable base-line slope even at one-fifth maximum sensi- tivity and it is apparent that the CRP and S'AJP have about similar sensitivities with this reduction (see Fig. 2 ) . (ii) A solution of Pontachrome violet SW (2-hydroxy-5-sulpho-a-benzeneazo-2-naphthol) containing 2 pg of aluminium per ml was used.Sept., 19561 STUDY OF THREE RECENTLY DEVELOPED POLAROGRAPHS 509 The peak heights obtained were- CU .. . . . . 29 mm at 1/10 maximum sensitivity SWP . .. . . . 180 mm at 1/20 maximum sensitivity normal CRP . . . . . , 27 mm at 1/1000 maximum sensitivity. This is obviously a highly irreversible reduction, as the CRP is more sensitive than the SWP, despite the fact that four electrons are involved in the reduction of the aluminium dye. This should favour the SWP, for the peak height with this instrument varies with n2, whereas with the CRP it varies with n3/2. RESOLUTION- (a) A solution containing 20 pg of cadmium per ml and 20 pg of indium per ml in M potassium chloride-For a 1 to 1 mixture the degree of separation of peaks with the SWP and the derivative CRP appeared to be the same and both were measurable. The double peak was not resolved by the CU. For a cadmium to indium ratio of 10 to 1 the indium wave could be detected as a swelling at the base of the cadmium wave with both the SWP and the CU.With the derivative CRP the peaks were clearly separated. It would be possible in both cases to make an approximate estimate of the indium concentration. (b) A solution containing 10 pg of thallium per ml and 10 pg of lead per ml in M hydrochloric acid-For a 1 to 1 mixture the peaks were resolved by both the CRP and SWP. At 10 to 1 and 1 to 10 the peaks could be distinguished but not measured. We estimate that 7 to 1 would be about the maximum ratio that could be measured. Once again the peaks were not resolved by the CU, although the presence of two reductions was evident from the shape of the peak. EFFECT OF PREVIOUS REDUCTIONS- (a) Fe"' to Cu ratio of 20 to 1 (1 mg of iron per ml in M hydrochloric acid)-On all three instruments the copper wave was readily measured.The normal circuit of the CRP, like most conventional polarographs, can compensate for a previous reduction 50 times larger than that being measured. Such polarograms from conventional instruments are, however, difficult to measure, because of the large current variations due to drop growth and fall. As the CRP trace is carried out in the lifetime of a single drop, no such interference occurs. (b) Fe'" to Ni ratio of 100 to 1 (1 mg of iron per ml in M hydrochloric acid)-Here, the nickel peak was observed with all three instruments. The derivative circuit of the CRP was of course used for this and subsequent experiments in this series. (c) Fe"' to Cu ratio of 333 to 1 (1 mg of iron per ml in M hydrocMoric acid)-The copper wave was visible on both the CU and the SWP, neither of which was working at its maximum sensitivity. In fact, the copper wave has been detected with the SWP when the iron to copper ratio was 5000 to 1.The CRP could not be used above one-fifth of its maximum sensitivity because the cell current due to the iron reduction was so large that the voltage drop in the higher values of cell-load resistors was greater than could be compensated by the voltage-sweep circuit. ( d ) Fe"' to Bi ratio of 1000 to 1 (1 mg of iron per ml in M hydrochloric acid)-With the CRP the bismuth peak was visible but not measurable, since, owing to its very early reduction potential, the yave occurs on the part of the trace that is distorted when the potential passes through zero applied volts.The peak was well defined and measurable with the SWP but not with the CU. TYPICAL APPLICATIONS (a) A solution of a bronze containing copper, zinc and nickel in the ratios 86 to 1-86 to 0.04 was prepared in a M ammonium hydroxide - 0.2 M ammonium chloride solution. An initial solution in which the copper level was 5 mg per ml did not give good results with any of the instruments. A tenfold dilution produced peaks for nickel and zinc on both the SWP and CRP, although, since the derivative circuit was used, the SWP was more sensitive than the CRP by a factor of 20. The nickel wave was not visible on the CU, as might be expected, since the concentration of nickel was only 0.5 pg per ml. With all three instruments the Cu" and Pb" peaks were easily measured.With the CRP, the (b) A Mazak alloy containing copper, lead, cadmium and indium was examined.510 FERRETT, MILNER, SHALGOSKY AND SLEE: A COMPARATIVE [Vol. 81 indium peak was sufficiently resolved to permit a direct measurement of the indium concen- tration to be made with an accuracy of about 10 per cent. The resolution with the SWP and the CU was not sufficient to permit a direct measurement (see Fig. 3). - h > .- .- u .- ?4 E z - U C u' 70 - 60 - 50 - 40- 30 - 20 - i i 1 I I I - 101 ' I 1 - 0'54 -0'64 -0'74 Potential against mercury-pool anode, volts Fig. 3 (a). Square-wave polarogram for Mazak alloy in chloride solution U c 0, L 3 U Cd a Potential against mercury-pool anode, volts Fig. 3 (c). Univector polarogram for Mazak alloy in chloride solution DISCUSSION In the analysis of solutions containing a single cation at dilutions up to ten times those normally employed in polarography, there is little to choose between the CRP and the SWP.At greater dilutions, however, the slope of the base line of the CRP trace increases and finally sets a limit to the amplifier gain that can usefully be employed. This slope may be largely eliminated by means of the derivative circuit, but the signal then obtained is about 25 times less than before. At very low concentrations there is insufficient reserve of gain available in the present model to enable the signal loss to be fully overcome. Considerations of ease and accuracy of peak measurement may well, however, dictate the use of the derivative technique. The CU suffers from base-line irregularities at high sensitivities, but it must be pointed out that the manufacturers recommend that sensitivities greater than one-fifth maximum should not be employed.The derivative trace was, however, much easier to measure than that produced by the normal polarograph. The SWP is virtually free from base-lime difficulties, even a t maximum sensitivity. With the SWP and CU the base electrolyte imust be extremely pure as a concentration of 1 molar has to be used. With the CRP the base electrolyte can be very dilute, with the consequent reduction in the difficulties caused by impurities.Sept., 19561 STUDY OF THREE RECENTLY DEVELOPED POLAROGRAPHS 51 1 SPEED- The CRP has an advantage in that an immediate scan over its potential range of 0-5 volt is made every 7 seconds.The greatest care over deoxygenation is necessary only at high sensitivities of the CU and CRP. Bubbling pure hydrogen through for 20 minutes is generally essential to ensure freedom from oxygen interference. With the SWP it is rarely necessary to deoxygenate for more than 3 minutes. Thus the total time for polarography of a single solution is about the same for the CRP and SWP. If polarograms have to be recorded for many solutions, however, the solutions may be deoxygenated simultaneously and ten solutions may readily be dealt with in 30 minutes on the CRP. The CU and SWP have a slow recorder type presentation. ACCURACY OF PEAK-HEIGHT MEASUREMENT- The peak height can therefore be measured over a maximum height of 100 mm with an accuracy of & 0.5 per cent.The SWP, with a thin ink line over 280mm, is obviously more accurate than this. The CU has a maximum chart height of 75 mm, and for the best resolution damping should not be used. However, when care was taken, the results appeared to be as accurate as those produced by the other two presentations. The CRP trace can have a width as small as approximately 0.2 mm. REPRODUCIBILITY- With all three instruments, the limitation in reproducibility appears to be due to variations of the capillary characteristics. Thus the SWP and the CU (on sensitivities less than one-fifth) may give slightly differing results from successive polarograms of the same solution, The CRP presents a trace that appears to be identical for successive drops.If the drop-time should change, then the drop fall is observed to occur at a different place on the screen. I t is, therefore, easier to observe a change in the drop-time with the CRP than with the SWP or CU. At sensitivities greater than one-fifth the CU results may vary much more. MAINTENANCE- down. unlikely to be more troublesome than normal polarographs. but experience has shown that 6 months is a reasonable trouble-free period. The CRP has been in everyday use at Woolwich for 5 years with only one major break- The CU and pen-recording polarograph both have electronic amplifiers, but are The SWP is a more recent development and long term testing has not been possible, CONCLUSIONS The sensitivity and resolution of three recently designed polarographs have been compared.Typically reversible and irreversible reductions have been studied. While it is hoped that the results are representative, it must be borne in mind that the relative behaviour of these instruments may vary considerably from one electrode reduction to another. SENSITIVITY- The SWP is the most sensitive of the polarographs for the study of a reversible reduction, e g . , cadmium. It is 6 times as sensitive as the normal CRP trace, 12 times as sensitive as the CU, 40 to 50 times as sensitive as the derivative CRP and more than 200 times as sensitive as a conventional polarograph. The CRP loses less in sensitivity when studying irreversible reductions than does the SWP. RESOLUTION- The CRP on its derivative circuit has a rather better resolving power than the SWP. EFFECT OF PREVIOUS REDUCTIONS- The SWP is less affected by earlier reductions than the CRP and CU and can deal5 with The maximum ratio that can be examined with the CRP is 400 to 1. ratios of 20,000 to 1. With the CU the maximum ratio is about 800 to 1.512 SHALGOSKY : THE POLAROGRAPHIC [Vol. 81 REFERENCES 1. 2. Lamb, B., Evershed News, 1956, 4, 16. 3. 4. 5. Reynolds, G. F., J. Polarogruphic Soc., 1955, 1, 17. “The Cambridge Univector Polarograph Unit,” Cambridge Instrument Co. Ltd., London, 1953. Davis, H. M., and Seaborn, J. F., Electronic Eng., 1953, 25, 314. Ferrett, D. J., and Milner, G. W. C., Analyst, 1956, 81, 193. ANALYTICAL CHEMISTRY GROUP ATOMIC ENERGY RESEARCH ESTABLISHMENT HARWELL, NR. DIDCOT, BERKS. UNITED KINGDOM ATOMIC ENERGY AUTHORITY and RESEARCH GROUP WOOLWICH OUTSTATION, WOOLWICH, S.E.18 Mavch 22nd, 1956

ISSN:0003-2654    

DOI:10.1039/AN9568100506

出版商:RSC    

年代:1956

数据来源: RSC

摘要:

512 SHALGOSKY : THE POLAROGRAPHIC [Vol. 81 The Polarographic Determination of Uranium BY H. I. SHALGOSKY (Presented at the meeting of the Physical Methods Group on Tuesday, February 14th, 1.956) A published method for the polarographic determination of uranium in an acid tartrate medium has been examined. Under the conditions that were adopted to achieve maximum sensitivity, the method has been shown to give low results. This is due to the effecl: of heating uranium in concentrated sulphuric acid solution. Satisfactory results are obtained when perchloric acid is used for the chemical treatment and the dilute sulphuric acid is added to the cold solution immediately before the polarographic determination. The sensitivity of the final method is shown to be similar to those of other polarographic methods for the d.etermination of uranium, and the factors that affect sensitivity are discussed.A METHOD for the polarographic determination of uranium in ores was recently published by Legge,l who claims that his method is more sensitive than others. He rejects the acid tartrate medium recommended by Lewis and Overton2 on the grounds that it lacks sensitivity, even though fewer elements interfere with the determination in this medium. However, the data he presents to justify his choice of base el.ectrolyte appear to be theoretically unsound. Since a sensitive method for the determination of uranium was required, it was decided to investigate this matter. Some of the experirnental work will be given in detail, because of the unusual nature of the difficulties encountered.EXPERIME~NTAL The Cambridge pen-recording polarograph used had a maximum sensitivity of 0.0026 pA per mm. The dropping-mercury electrode had a capillary characteristic of 2-68 rng2j3 sec.-1/2 in 2 N potassium chloride with no applied potential. All solutions were deoxygenated with electrolytically generated hydrogen and polarograms were recorded at 25" 0.1" C, a mercury pool, being used as anode. All reagients were of analytical-reagent quality and were polarographically checked for purity. Doubly distilled water was used throughout. A standard solution containing 100mg of uranium per ml was available. This had been prepared by dissolving pure uranium metal in nitric acid. Uranyl sulphate solutions were prepared by heating aliquots of the nitrate solution to fuming with sulphuric acid and diluting to volume with water. A 0.1 per cent.w/v sollition of proteose peptone containing 0.2 per cent. w/v of redistilled phenol as preservative was used as a maximum suppressor, being diluted 100-fold in use. A tartrate stock solutiori was prepared by dissolving 15 g of Rochelle salt in water. adding 10ml of saturated Dotassium chloride solution and 10 ml of proteose ACID BASE ELECTROLYTES- the equation- peptone solution, agd making up to look1 with water. RELATIVE SENSITIVITIES OF THE TARTRATE AND OXALIC The diffusion current constant, I , which is given bySept., 19561 DETERMINATION OF URANIUM 513 where id = diffusion current in pA, c = concentration of uranium in millimoles per litre, and m2I3 t1I6 = capillary characteristic as mg 2/3 sec.-lI2, was measured, with use of the two base electrolytes that were to be compared.A 1.0-ml aliquot of a standard uranyl nitrate solution containing 100 pg of uranium per ml was placed in a 5-ml calibrated flask, and 1.0 ml of 5 M sulphuric acid and 0.5 ml of tartrate stock solution were added. The solution was made up to the mark and a polarogram was recorded. The diffusion constant was 2.66. The second solution contained 1 pg of uranium per ml, and was 0.5 M in oxalic acid and 0.9 M in sulphuric acid. The diffusion-current constant at this concentration of uranium was 2.31. Hence, the tartrate base electrolyte is of comparable sensitivity with the oxalic acid base recommended by Legge, and further work was therefore confined to the tartrate medium, as fewer elements interfere.EFFECT OF BASE-ELECTROLYTE COMPOSITION- A calibration graph was prepared with use of the tartrate base electrolyte and the procedure outlined above, and samples containing known amounts of uranium were then examined. In order to eliminate interference from the gross amounts of nitrate ion present, the samples were evaporated to fuming with 1.0 ml of 5 M sulphuric acid and washed into a 5-ml calibrated flask, 0.5 ml of tartrate stock solution was added, the solution was made up to the mark and polarograms were recorded. As the results were not consistent, the effect of variations of the concentration of sulphuric acid in the final solution on the diffusion current due to 20 pg of uranium per ml was studied, with the results shown in Table I.TABLE I EFFECT OF SULPHURIC ACID CONCENTRATION ON DIFFUSION CURRENT Concentration of sulphuric acid, M . .On2 0.4 0.6 0.8 1.0 2.0 3.0 Diffusion current, pA . . .. . .0*494 0.540 0.572 0.591 0.598 0.572 0.520 The optimum concentration of sulphuric acid for maximum sensitivity was therefore 1.0M. It was obvious, however, that care must be taken to ensure that as little acid as possible was lost by fuming in order to maintain a uniform acid concentration. A 10 per cent. variation in the concentration of the tartrate solution was found to have no measurable effect on the uranium diffusion current. Determinations carried out on the sample of known content were now about 10 per cent. low. Since the only difference between the sample and standard solutions was that the former had been heated with sulphuric acid, the experiments described below were carried out.EFFECT OF HEATING- Aliquots of uranyl nitrate solution were heated with 1 ml of 5 M sulphuric acid at 400" C in an air-bath for about 20 minutes. The solutions were then washed into 5-ml calibrated flasks, 0-5 ml of tartrate stock solution was added to each, and the mixtures were made up to the mark and polarograms were recorded. Duplicate results are shown in Table 11, together with those for a series of solutions that had not been heated. TABLE I1 EFFECT OF HEATING URANIUM WITH SULPHURIC ACID Uranium recovered, pg Uranium r A \ added, Unheated Heated Heated Pg solution solution solution 60 62 58 59 100 102 92 70 180 176 142 157 Standard additions of uranium to solutions giving low recoveries, and to unheated solutions, gave almost identical increases in height, equal to those expected from unheated solutions.No difference was detected between heated and unheated solutions containing5 14 SHALGOSKY: THE POLAROGRAPHIC [Vol. 81 no uranium. Addition of sodium nitrate to solutions that had been heated did not affect the uranium diffusion current ; nor did substitution of uranyl sulphate for uranyl nitrate. It was noted during these experiments, however, that the diffusion current from heated solutions slowly increased with time, and the results in Table I11 were therefore obtained. TABLE ]:I1 VARIATION WITH TIME OF STEP HEIGHT FOR HEATED SOLUTIONS Time after dilution, minutes .. 0 20 120 140 Unheated Diffusion current, pA . . . . 0.462 0-514 0.553 0.565 0.565 solution In order to test whether this phenomenon occurred when the base electrolyte recom- mended by Leggel was used, a heated uranium in sulphuric acid solution was made up with oxalic acid such that the final solution was 0.5 M in oxalic acid and 0-9 M in sulphuric acid. The diffusion current obtained was only two-thirds of that found from a similar solution prepared without heating, but increased to the same value after the solution had been set aside overnight. As it was undesirable to wait for nearly 3 hours to obtain reproducible results, the use of sulphuric acid for the removal of nitrate ion was abandoned in favour of 0.5 ml of 60 per cent. perchloric acid, which is known to be a non-complexing medium for uranium.The sulphuric acid, which must still be present, however, was now added to the cold sample solution in the 5-ml calibrated flask after the tartrate and uranium solutions had been mixed. The presence of perchloric acid precluded the use of potassium salts in the base electrolyte, and a new stock solution was prepared for use in all subsequent experiments involving the tartrate medium. This stock solution contained 12 g of disodium tartrate, Na,C4H,0,.2H,O, 10 ml of proteose peptone solution and 0.293 g of sodium chloride per 100 ml. The effect of variations in the composition of the base electrolyte on the diffusion current due to 2Opg of uranium per ml is shown in Table IV. TABLE IV EFFECT OF BASE-ELECTROLYTE COMPOSITION Amount of 60 per cent.Amount of Amount of perchloric acid, 5 M sulphuric acid, tartrate solution, ml ml ml 0.25 1.0 0.5 0.5 1.0 0.5 1.0 1.0 0.5 0.5 0.9 0.5 0.5 1.1 0.5 0.5 1.0 0.45 0.5 1-0 0-55 Diffusion current, PA 0.546 0-579 0.59 1 0.546 0.572 0.565 0.559 Since the concentration of perchloric acid in the final solution is important, the pro- cedure adopted to remove nitrate ion was to evaporate the uranium in perchloric acid solution down to 0.1 ml and add 0.4 ml of 60 per cent. perchloric acid. After this treatment the diffusion current fell to 0.494 PA, but in contrast to the effect of sulphuric acid, it was quite reproducible. Further, no change was observed with time, even when the solution was set aside overnight. A calibration graph was prepared by using solutions of uranium in perchloric acid that had been heated.The graph was slightly curved at its lower end, but was almost linear over the range 20 to 2OOpg of uranium per ml. Determinations on samples of known uranium content were satisfactory and agreed with results obtained absorptiometrically . INTERFERENCES- It was known that bismuth, copper, lead and molybdenum are reduced at potentials close to that of uranium in the tartrate medium and the half-wave potentials of these elements are shown in Table V, together with the values for uranium in the oxalic acid base electrolyte and in N sulphurk acid.Sept., 19561 DETERMINATION OF URANIUM 515 Only the molybdenum wave was sufficiently close to that of uranium t o interfere, and therefore it must not be present in sample solutions.the concentration of nitrate ion in the final solution must be less than 0-1 M. Lewis and Overton2 reported TABLE V HALF-WAVE POTENTIALS Eg, volts against the saturated- Element Medium calomel electrode Uranium tartrate Bismuth tartrate Copper tartrate Molybdenum tartrate Lead tartrate Uranium oxalic acid Uranium AT sulphuric acid - 0.44 - 0.28 - 0.29 - 0.49 - 0.66 - 0.33 - 0.42 METHOD Sulphuric acid, 5 M-Add 100 ml of 18 M sulphuric acid to 200 ml of cold water, Proteose peptone solution, 0.1 per cent. w/v-Dissolve 0.1 g of proteose peptone REAGENTS- and dilute to 360ml with water. 0 . 2 ~ of phenol in 100ml of water. that cool and -Disidium tartrate stock solution-Dissolve 12.0 g of disodium tartrate, Na,C4H406.2H,0, and 0.293 g of sodium chloride in water, add 10 ml of 0.1 per cent.proteose peptone solution and dilute to 100ml with water. PROCEDURE- The uranium should be in perchloric acid solution and free from molybdenum. Evaporate the perchloric acid down to 0.1 ml and add 0.4 ml of 60 per cent. perchloric acid. Transfer the solution to a 5-ml calibrated flask and add washings of the sample container, using a total of about 2 ml of water (a dropper pipette is most convenient for this). Add 0-5 ml of disodium tartrate stock solution to the calibrated flask, set aside for about 2 minutes and then add 1.0 ml of 5 M sulphuric acid. Make up to the mark with further washings of the sample container, place the stopper in the flask and shake. Transfer the solution to a polarographic cell, deoxygenate it, and record a polarogram over the range 0 to -0-6 volt, using the mercury pool as anode. Measure the height of the uranium step, which occurs at a half-wave potential of -0.35 volt, using the intersection method of step-height measure- ment. Read off the amount of uranium in the sample from a calibration graph prepared by using known amounts of uranium in perchloric acid solution and proceeding exactly as described above.RESULTS With the Cambridge pen-recording polarograph, a step-height of at least 40 mm (0.104 PA) can be obtained except with less than about 20 pg of uranium. Steps can be measured to within kO.5 mm (0.0013 PA), which limits the precision of the method to f 1-5 per cent. This was the maximum variation observed for six determinations of 1OOpg of uranium.The precision falls to t-3 per cent. with 10 pg of uranium, giving a step-height of 20 0-5 mm (0-052 0.0013 PA), and this is considered to be the limit of determination of the method. The limit of detection is about 2 pg of uranium. Determinations have been carried out on solutions of uranium separated from ores of known composition, and the results will be given elsewhere,3 with full details of the method, including the separation. SENSITIVITY- of electrons, n, involved in a reduction at the dropping-mercury electrode. IlkoviE equation- DISCUSSION OF RESULTS * From the value of the diffusion current constant, I , it is possible to calculate the number From the I 605 D%' - - - id =605 D* m213 t l k516 SHALGOSKY : THE POLAROGRAPHIC [Vol.81 where D = diffusion coefficient. If the value of the diffusion coefficient of uranium is assumed to be 0.60 x cma sec.-l, then from the value I = 1.02 in acid tartrate given by Legge, n = 0.69. We find n = 1.8 from the value I = 2.66 in tartrate medium, which compares well with Legge's value I = 2-73 in oxalic acid medium. The discrepancy in Legge's data may be partly due to the effect of heating uranium with sulphuric acid, although it is unlikely that such treatment would reduce the diffusion current to a point where the number of electrons taking part in the reaction fell below 1.0. This is absurdly small. BASE-ELECTROLYTE COMPOSITION- Sheel and Watters4 have shown that in su:lphuric acid alone the diffusion current of uranium is dependent on the concentration of the sulphate ion, and that the optimum concentration of the acid for maximum diffusion current is 1.5 N .The variations of diffusion current for uranium shown in Table I were not, therefore, entirely unexpected, although Lewis and Overton reported that sulphate ion had no effect on the tartrate method. It is possible that the variations observed were due solely to differences in hydrogen-ion concentration, but the results in Table IV indicate that both the sulphate-ion concentration and the hydrogen-ion concentration affect the diffusion current. It is obvious that for a routine method, the sulphuric acid and perchloric acid concen- trations should be carefully controlled. The diffusion-current constants for the three media, sulphuric acid alone, and with oxalic acid and disodium tartrate present are remarkably similar, being 2-75, 2-73 and 2.66, respectively.can be followed by the disproportionation- This disproportionation occurs more readily if the products are stabilised by complex formation, but this will also cause a shift in the half-wave potential of reduction (1) if the Uv species resulting from this reduction remain complexed, then the prime factor in driving reaction (2) to the right becomes the complexing of the UIV species. Hence, although the Uvl is complexed to different extents in the three media considered (as shown by the half-wave potential in Table V), the similarity of the diffusion-current constants indicates the the Urn sulphate complex is stronger than those with oxalic acid and disodium tartrate .In support of this, Legge has shown that the diffusion current of uranium in the presence of sulphuric acid is independent of the oxalic acid concentration over the range 0.1 to 0.5 M . Also, the diffusion-current constant of uranium in approximately 0.01 M hydrofluoric acid solution4, which is known5 to form a strong complex with UIV (K = 3.7 x 10-14), is 2.48. It is concluded that the addition of further complexing agents, such as oxalic acid, tartrate and citrate, to uranium in sulphate solutions assists mainly in the separation of interfering waves and hardly affects the diffusion current or the mechanism of the reactions at the dropping-mercury electrode. Kern and Orlemanne have shown that in a non-complexing medium the rate of dis- proportionation is proportional to the square of the UO,' concentration. The non-linearity of our calibration graph indicates that this relationship also holds in complexing media.As long as the n value of the reduction is less than 2, this non-linearity should be observed, but as the concentration of uranium is increased, so n will approach 2, and the graph will become linear. Our graph was curved over the range lop6 to 10-4M uranium and linear over the range The other constituents of the base electrolyte, vix., proteose peptone and sodium chloride, take no part in the reactions. The chloride ion permits a mercury pool to be used as anode, which is convenient in a routine method. The proteose peptone stock solution is quite stable and the tartrate stock solution has given reproducible results for 3 months, although it became cloudy towards the end of this period.EFFECT OF HEATING- Further work is required before the cause of the apparent loss of uranium when it is heated with concentrated sulphuric acid can be known with any certainty. A possible explanation is that a very stable uranyl su1phat:e complex is formed or, what amounts to the same thing, the normal uranyl species is dehydrated and the water replaced with sulphate Now the reduction of the simple uranyl ion- . . - - (1) .. * * (2) UO," + e - UO,' .. .. . . 2UO,'+ H'- UO," + UO OH' .. to 10-sM uranium.Sept., 19561 DETERMINATION OF URANIUM 517 ions. Such a complex, after reduction, might not disproportionate as rapidly as the normal species, but some difference in the half-wave potentials of the normal U” and the complexed Uvl species might be expected.No such shift has, however, been observed in these solutions. It is not known whether the smaller and reproducible decrease of the diffusion current that occurs when uranium is heated with perchloric acid is due to a similar reason to that mentioned above. Whatever the cause, however, the effect is very important in the detenni- nation of uranium. It must be stressed that, when sulphuric acid was used, it was not possible to produce the lowering of the diffusion current every time, and there may be some critical temperature, concentration, time or other factor of which we are unaware. CONCLUSIONS As a result of this investigation, which has been directed towards the production of a reliable and sensitive method for the determination of uranium, the following recommenda- tions are made- (a) Uranium salts in concentrated sulphuric acid solution should not be strongly heated immediately before the polarographic determination.(b) The preparation of calibration and sample solutions should be as similar as possible, and for extreme accuracy as many points as possible at the lower end of the calibration graph should be obtained. The tartrate base electrolyte should be used to avoid interfering waves. These recommendations have been incorporated in a method3 for the determination of uranium in ores, which is rapid, accurate and covers a wide range of uranium contents. (c) I am indebted to Messrs. E. C. Hunt, E. A. Terry and F. E. Wild for carrying out separa- tions of uranium from ores and to Mr.G. J. Hunter for helpful discussion. REFERENCES 1. Legge, D. I., =1nal. Chem., 1954, 26, 1617. 2. 3. 4. 5. 6. Lewis, J . -4., and Overton, K. C., Department of Scientific and Industrial Research, Chemical Wild, F. E., Atomic Energy Research Establishment Report C/R 1868. Sheel, S. W., and Watters, J. I., cited in Rodden, C. J., “Analytical Chemistry of the Manhatten Latimer, W. M., “The Oxidation States of the Elements and their Potentials in Aqueous Solutions,” Kern, D. M. H., and Orlemann, E. F., J . Amer. Chern. SOC., 1949, 71, 2102. Research Laboratory, Scientific Report CRL/AE/41, 1949. Project,” McGraw-Hill Book Co. Inc., New York, 1950, Volume I, pp. 596 to 610. Second Edition, Prentice-Hall Inc., New York, 1952, p. 303.UNITED KINGDOM ATOMIC ENERGY AUTHORITY RESEARCH GROUP WOOLWICH OUTSTATION WOOLWICH, S.E.18 May 4th, 1956 DISCUSSION ON THE FOREGOING TWO PAPERS MR. A. F. WILLIAMS asked Mr. Milner for his opinion on the superiority or otherwise of the cathode-ray polarograph over the square-wave polarograph for organic determination. He would also like to know why a change in peak height occurred as the screen was traversed and whether this change was a drawback in the use of the instrument. MR. MILNER answered that for most organic reductions the normal circuit of the cathode-ray polaro- graph was possibly more sensitive than the square-wave polarograph. In reply to the second point, he said that the peak current measured with a cathode-ray polarograph was proportional to tP2l3, where t , was the time at which the peak occurred, measured from the start of drop growth.If a peak was measured at different places on the screen, then t , would vary, and the peak current also. For accurate quantitative work it was necessary, therefore, to use the same start potential whenever measuring ,a particular wave. This was not considered to be a drawback in the use of the instrument and, for routine determinations by unskilled operators, it was probably an advantage. DR. FURNESS remarked that, when the cathode-ray polarograph waves due to the reduction of substances at the dropping-mercury electrode were displayed, the fly-back of polarising potential from negative to more positive values was triggered by the cessation of diffusion current a t the instant of detachment of a drop from the capillary.For a wave due to oxidation of a substance at the dropping-mercury electrode, the change in diffusion current would be zero when the drop became detached, if the sweep were from positive to more negative potentials. He would like to know what provision was then made for triggering the fly-back, and to what extent would the shape of the wave be influenced by the diffusion current reaching518 SHERRATT : DETERMINATION OF [Vol. 81 a minimum value at the instant of maximum drop size. If there were difficulties, could they be overcome by applying a voltage sweep in the direction of negative to more positive potentials, or were such difficulties avoidable simply by reversing the leads to the polarogra,phic cell.MR. SHALGOSKY stated that the simplest method of achieving synchronisation when the wave shape did not allow it to occur automatically was to arrange for the drop-fall to occur in the fly-back period of the instrument. It was hoped to publish a paper describing the wave shapes obtained with reversed potential sweeps. MR. T. R. DAVIES, on behalf of the manufacturers, stated that it should be possible to modify the cathode-ray polarograph for work on anodic oxidations. DR. J. A. HUNTER said that, if adequate synchronisation (with cathode-ray polarography) could not otherwise be obtained in the recording of anodic waves or others in which the final current was small, it was then possible to achieve the synchronisation by manual detachment of the drop at the end of the trace.In order to obtain reproducible peak heights, in view of the dependence of peak heights on the instant during the sweep at which the peak occurred, it was essential that the trace should be accurately framed within the same voltage lines in all experiments, as otherwise the instant of occurrence of the peak would not always correspond to the same stage in the life of the drop. DR. J. E. PAGE wished to know how sensitive the cathode-ray polarograph was in the measurement of catalytic waves, e.g., cysteine and cystine in an ammoniacal cobalt buffer solution. MR. T. R. DAVIES replied that he had done some qualitative work with cystine and cysteine, and catalytic waves due to the evolution of hydrogen had been produced in the form of peaks.The medium had been an ammoniacal solution containing a bivalent cobalt salt. MR. D. R. CURRY asked Mr. Milner whether the square-wave polarograph was reversible in respect of voltage scanning and what the advatages of reverse scanning were. He enquired about the advantages of increased chart size (Honeywell - Brown) as oscillations enlarged pro vatu. MR. G. W. C. MILNER, in reply, said that with the square-wave polarograph it was possible to use reversible voltage scanning, which was a convenience when it was necessary to record the same peak several times. Several scanning rates were available with this instrument, the maximum rate corresponding to 0.575 volt per minute and the minimum rate to 0.038 volt per minute. The Honeywell - Brown recorder in the Univector polarograph being demonstrated was not chosen purely for the convenience of its great chart size, but mainly because it contained its own amplification unit. MR. W. J. PARKER, referring to Mr. Shalgosky’s suggested method for the polarographic determination of uranium, which would require a supporting electrolyte containing hydrofluoric acid, stated that although such a method was impracticable when a glass dropping-mercury electrode was used as cathode, it might well be that some of the new polarographic electrodes might permit the successful application of this promising method. It might be helpful to Mr. Shalgosky if he mentioned that a considerable volume of work had been carried out both in America and in this country on the development of polarographic elec- trodes suitable for use in hydrofluoric acid media. Although some of these electrodes were based on principles other than that of the dropping-mercury electrode, nevertheless these electrodes might well have a special application to such methods as that suggested by Mr. Shalgosky. References to some of these electrodes were to be found in papers by V. S. G-riffiths and W. J. Parker (Research, 1954, 7, s46, and in Anal. Chem. Acta, 1956, 14, 194). He had found this to be very satisfactory.

ISSN:0003-2654    

DOI:10.1039/AN9568100512

出版商:RSC    

年代:1956

数据来源: RSC

摘要:

518 SHERRATT : DETERMINATION OF Determination of Volatile Oil in [Vol. 81 Effluents BY J. G. SHERRATT An account is given of a method for determining traces of neutral volatile oils, such as fuel oils, in waters and effluents. The solubility of certain types of fuel oil in water is considered, and some of the effects of pollution of natural waters by oil are briefly described. THE determination of the glycerides of fatty acids and of mixed high-boiling hydrocarbon oils in effluents and river waters within limits of accuracy that are acceptable for most purposes does not present intrinsic difficulty, since these classes of compounds can usually be extracted by appropriate solvents, from which they can ’be recovered and ultimately weighed. For obvious reasons mixtures of volatile oils, such as occur in fuel oils and similar commercial products, cannot be determined in this way, nor can methods that depend upon the principle of distillation and subsequent volumetric measurement of the separated “oil” be made sufficiently sensitive to determine a few parts per million in effluents or river waters.ManySept., 19561 VOLATILE OIL I N EFFLUENTS 519 excellent met hods based upon absorption spectrophotometry have been proposed for the determination of individual hydrocarbons, but these methods are invariably too specific to be of value to the water analyst for routine examinations for oil. Despite the fact that the ever-increasing use of fuel oil has already created serious problems by polluting rivers, estuaries and, indeed, the oceans, very few methods have been published for determining traces of volatile oils. A method based upon gasometric measure- ment has been proposed by Webber and Burkesl for the determination of light hydrocarbons (including butadiene) in rubber-works waste waters, but its application is restricted, since many effluents and most river waters contain dissolved inert gases that would interfere. The danger of discharging relatively large quantities of volatile hydrocarbons into sewers is obvious, but it is not always appreciated that a few parts per million of commercial fuel oils may cause serious damage in rivers, not only by collecting behind weirs and on the surface of slow-flowing water and thereby interfering with atmospheric oxygen-exchange,2,3 but also by their specific toxicity to fish3 and their relatively high biochemical oxygen demand (B.O.D.). For example, during some recent tests in this laboratory, it was found that the addition of two drops of either petrol, white spirit or kerosene (paraffin burning oil) to 1 litre of a fairly clean river water raised the B.O.D.from 3.5 to over 10 p.p.m. The effect was not dependent upon “seeding” the water with extraneous bacteria. Clearly there is a need for a widely applicable method of analysis that will detect and determine neutral volatile oils in the range of 5 to 500 p.p.m. in any kind of water (sewage, effluents and river waters). For practical purposes extreme accuracy is rarely essential, but the method should be capable of demonstrating increments of about 2 to 3 mg per litre in the lower range and 20 to 30 mg per litre in the higher.EXPERIMENTAL The heterogeneous character of commercial fuel oils and the extreme stability of aliphatic hydrocarbons militate against methods of analysis based upon chemical reaction. Of the physical properties, specific-gravity and refractive-index measurements are impracticable in the lower ranges of concentration, and spectrophotometric absorption is too specialised to form the basis of a general method. Most commercial oils fluoresce under ultra-violet excitation, but an attempt to utilise this effect soon indicated that it was too variable and too subject to extraneous influence to be of value for general quantitative use. Among the properties characteristic of volatile oils is their miscibility with acetone and their relative insolubility in water.If a dilute solution of oil in acetone is mixed with an excess of water, the oil is thrown out of the solution in the form of a white turbidity; under suitable conditions the turbidity may be made proportional to the quantity of oil and the effect used to determine commercial oil dissolved in acetone. If the dilution in acetone is made with water containing a trace of synthetic detergent, the emulsion is stable for many hours and if, in addition, the water is acid, basic substances will not usually interfere. Trials of the sensitivity of the effect showed that as little as 0.001 ml of kerosene or white spirit, dispersed in 50 ml of liquid, give a faint but perceptible turbidity and additional increments of the same order could easily be distinguished from one another.With com- mercial petrol and the lower-boiling hydrocarbons the limit for detection was about 0-003 to 0-004 ml at this dilution, but if the final volume was reduced to 10 ml instead of 50, 0.001 ml of petrol was just perceptible by careful comparison with a blank and 0.002 ml was unmistakable. Before this method of determination of volatile oils could be applied to effluents or natural waters, a method of separating and concentrating the oils had to be devised. Since extraction with organic solvents cannot be used, some form of selective adsorption appeared to offer the most readily available means. After a number of trials activated carbon was selected as being a suitable absorbent, and a brand kindly supplied by Messrs.Sutcliffe, Speakman & Co. Ltd., of Leigh, Lancs., was used for the purposes described in this paper. One gram of the carbon was supported on a plug of glass-wool in a small cylindrical funnel and 0.04 ml of kerosene was measured from a micro-burette on to it. Then 50 ml of water were passed through the carbon and allowed to drain away, excess of water finally being removed by gentle blowing. Five successive 1-ml volumes of acetone were passed through the carbon and collected in a graduated measuring cylinder. The last few drops of acetone were displaced from the carbon by blowing through the funnel. The acetone was diluted to exactly 50 ml with a solution containing 1 ml of sulphuric acid and 1 ml of Teepol per litre.520 SHERRATT : DETERMINATION OF [Vol.81 The resultant turbidity was compared with a standard containing 0.04 ml of kerosene dissolved in 5 ml of acetone and diluted to 50 ml with the Teepol solution. There was no visible differ- ence between the turbidity in the two tubes and the amount of light absorbed or dispersed by both liquids, measured in a 1-cm cell in a photometer at 5500 A, was practically identical. This experiment indicated that traces of oil adsorbed on activated carbon could be recovered quantitatively by extracting it with acetone, but when 0-04ml of oil was shaken with 1 litre of water and the whole of the liquid filtered through the carbon, only 30 to 40 per cent. of the quantity added could be recovered. The loss was not due to oil being left in the original vessel, which was washed out with acetone after all the water had been filtered.Several attempts were made to improve the percentage recovery by steam-distilling the oil in an apparatus arranged so that the condensed liquid passed through the carbon before reaching the free atmosphere. By replacing the carbon filters successively after increments of 50 ml of water had been distilled and finally extracting each filter with acetone and testing the extracts in the manner already described, it was found that no more oil could be recovered on carbon after 250 ml of water had passed over from an original volume of 1 litre of water and 0.05ml of kerosene. It was also proved, by stopping the inflow of steam, continuing the distillation almost to dryness, filtering the remaining aqueous liquid through carbon and subsequently extracting the flask and the carbon with acetone, that no oil remained unvolatilised.Nevertheless, in a series of experiments, the average quantity of kerosene recovered varied between 40 and 50 per cent. of the quantity originally added to the distilla- tion flask. Rather better percentage recoveries were obtained if the experimental oil was petrol, all of which was found to pass from the distillation flask in the first 100 ml. During the course of these trials it was noticed that the percentage of oil recovered varied inversely with the quantity of water distilled, which suggested that some of the oil was re-extracted from the carbon by water. This theory was confirmed by adsorbing measured volumes of volatile oil on carbon and washing it with various quantities of water before extracting the carbon with acetone and measuring the turbid.ity of the oil - water dispersion.In every instance, loss of oil increased proportionately with the volume of water that was filtered through the carbon. The loss was not appreciably diminished by first cooling the washing water to approximately 3" C. In view of this hitherto unsuspected solubility factor, it became apparent that quantita- tive recovery of small amounts of volatile oil ,would not be obtained by filtering aqueous liquids through carbon, with or without intermediate distillation. Attempts were made, therefore, to volatilise the oil in a current of air. This procedure was much more successful in recovering the oil and so the method that follows was finally adopted.METHOD A4PPARATUS- A small cylindrical or thistle funnel with a stem about 18 to 20 cm long and 0.5 crn in diameter. The stem is fitted with a rubber stopper that will fit tightly into the upper end of a reflux condenser. A 2-necked $ask (the "sulphur dioxide" flask of Monier-Williams) of about 1.5 litres capacity, fitted with a reflux condenser in one neck and with an air-delivery tube, reaching to the bottom of the flask, in the other. A small aquarium aerator, or other means of directing air through the apparatus. Test tubes (6 inch x Q inch), graduated at 2 ml and 10 ml. Acid Teepol-A distilled water solution containing 1 ml of concentrated sulphuric acid Activated carbon-Obtainable from Messrs. Sutcliff, Speakman & Co.Ltd., Leigh, Lancs., Sodium hydroxide solution-A 10 per cent. w/v solution of the analytical-reagent grade A cetone-Analytical-reagent grade. PROCEDURE FOR DETERMINING VOLATILE OIL IN THE RANGE 3 TO 15mg PER LITRE- Weigh 0.2 g of activated carbon and pack it lightly between tight plugs of glass-wool in the narrow stem of the small funnel (absorption tube). Arrange the flask and condenser REAGENTS- and 1 ml of Teepol per litre. quality No. 207, type B, mesh 20 to 40 B.S.S. material in distilled water. Cool to between 5" and 10" C before use.Sept., 19561 VOLATILE OIL IN EFFLUENTS 521 for reflux distillation in a current of air and fit the absorption tube into the outlet end of the apparatus (Fig. 1). Samples of effluent or river water in which oils are to be determined should be collected with precautions to ensure that they are representative and should be contained in wide- necked bottles of clear glass, holding approximately 1 litre.* Mark the level of the liquid on the outside of the bottle, for subsequent determination of the volume, and pour all the sample into the distillation flask.Rinse out the bottle with two portions of 2.5 ml of acetone, Fig. 1. Adsorption apparatus taking care that the acetone dissolves any traces of oil left adhering to the inside of the sample bottle; pour the acetone into the distillation flask. Add sodium hydroxide solution to raise the pH of the sample to approximately 10. Connect the gas-delivery tube to the flask and pass a current of air through the sample, at such a rate that the bubbles can just be counted; gently boil the liquid.About 10 minutes after drops of acetone can no longer be observed forming in the condenser, discontinue heating the flask and remove the small funnel containing the carbon. Wash the carbon with acetone, added drop by drop, and collect the filtrate in a graduated test tube. When exactly 2 ml of acetone have been collected, dilute the filtrate to 10 ml with the acid Teepol solution. PROCEDURE FOR PREPARING STANDARDS- Prepare a solution containing 1.00ml of petrol, sp.gr. 0-73, or other volatile oil (see note below), dissolved in and made up with acetone to 100ml. From a micro-burette measure into a series of matched test tubes 0.4, 0.8, 1-2, 1.6 and 2.0 ml of this solution and * Sampling of liquids that contain visible floating oil requires specialised equipment and technique.A method is recommended by the A.B.C.M. - S.A.C. Joint Committee on Methods for the Analysis of Trade Effluents (Analyst, 1956, 81, 492).522 SHERRATT: DETERMINATION OF [Vol. 81 dilute the first four of these to exactly 2 ml with acetone. Finally make each of the standards up to 10 ml with acid Teepol solution and match the sample visually against these standards. NOTE- I t is shown subsequently (Table 11) that equal weights (or volumes) of different classes of commercial fuel oils do not give the same degrees of turbidity when subjected to the above procedure, but that the variations between different oils of roughly the same category, e.g., kerosene and white spirit, or two different brands of petrol, are insignificant.For the most accurate work, therefore, it is necessary to prepare com- parison standards with oil of the type that is present in the sample. Fortunately the classes of oil likely to be present in effluents can usually be closely identified by their odour. An alternative procedure, if identification is impossible, is to refer the sample to some arbitrary standard such as petrol or white spirit, on the same principle that “phenols” in water are usually referred to hydroxybenzene. This may lead to some sacrifice of accuracy, which, however, is not likely to be of practical importance. MODIFIED PROCEDURE FOR DETERMINING OILS IN THE RANGE 12 TO 40mg PER LITRE- The apparatus and reagents are as described above, but 50-ml graduated cylinders should be used instead of test tubes.Pack 1 g of carbon into the stem of the small funnel and transfer the oil to the carbon by volatilising it in the manner already prescribed. At the end of this operation extract the oil from the carbon with successive 1-ml quantities of acetone, collecting the filtrate in a 50-ml graduated cylinder. When exactly 5 ml of filtrate have passed through the carbon, dilute the acetone to 50ml with acid Teepol solution. Prepare standards as before containing 1, 2, 3 4 and 5 ml of a 1 per cent. v/v solution of the oil in acetone. Add pure acetone to the lirst four standards to bring the volumes to 5ml and then dilute each to 50ml with acid Teepol solution. Measure the light absorbed or dispersed by the standards and the sample, or match the sample against the standards visually if a suitable photometer is not available.PROCEDURE FOR QUANTITIES OF OILS IN EXCESS OF 40mg PER LITRE- If the quantities of volatile oil are in excess of 40 mg per litre, the final turbidity will be too deep to measure. As it is impossible to distribute floating oil uniformly, the difficulty cannot be overcome by making the determination on only a portion of the sample. A procedure depending upon extracting the whole of the sample with chloroform and subse- quently hydrolysing an aliquot of the chloroform *with an excess of sodium hydroxide was first developed, but a much better method, suggested to me by Dr. J. Haslam, is to adsorb all the volatile oil on the carbon, by the method already described, and then to extract the carbon with repeated sinall volumes of acetone, finally making the acetone solution up to a measured volume and determining the turbidity developed from an aliquot of this solution.The above methods may be combined with an extraction procedure to determine the total oil in effluents and similar samples that may contain lubricating oil or diesel fuel of relatively high boiling point. Under the conditions prescribed only about 20 to 30 per cent. of diesel fuel, with an initial boiling point of about 180” C, is recovered on the carbon, the balance remaining in the flask. After all the volatile oil has been removed the liquid remaining in the flask is cooled, acidified with acetic acid or hydrochloric acid and extracted twice with light petroleum.The light petroleum extracts are combined, washed with a little water and shaken with 2 g of anhydrous sodium sulphate. The light petroleum is then filtered and transferred to a weighed flask. The solvent is evaporated and the residue dried and weighed. Subsequently it is dissolved in neutral alcohol and any acid is determined by titration with 0.02 N sodium hydroxide, phenolphthalein being used as indicator. In this way neutral volatile oil, non-volatile oil and fatty acid, free or combined with inorganic ions, e.g., as soap, may be determined separately in one sample. RECOVERY OF KNOWN QUANTITIES OF DIFFERENT FUEL OILS- In Table I are given some of the results obtained by the above procedure when known quantities of different commercial volatile oils were added to 1 litre of water.In each case the final turbidity was matched against standards prepared from the oil under examination.Sept., 19561 VOLATILE OIL IN EFFLUENTS 523 As it is impossible to measure very small quantities of volatile oil accurately and is incon- venient to weigh them, a 1 per cent. solution of oil in propylene glycol” was used to measure the added oil in the range 0 to 30 mg. Quantities in excess of 30 mg were measured directly from a micro-burette. TABLE I RECOVERY OF FUEL OILS Description of oil Quantity of oil added, Quantity of oil recovered, mg mg Cigarette lighter fuel, sp.gr. 0.70 { 2?5 between 3 and 4 (visual) 19 (instrument) Pirst-grade petrol brand “A,” J 4.4 4 (visual) 3.5 (visual) Commercial petrol brand “B,” sp.gr.0-73 1 36 30 (instrument) 21 (instrument) sp.gr. 0-73 White spirit, sp.gr. 0-785 17.3 (instrument) 34 (instrument) 3.1 2.3 (visual) 6.3 5-5 (visual) 15.5 12 (instrument) 31 29 (instrument) 160 130 (visual, after chloroform extraction) volatile non-volatile by extrac- (” tion and weighing 10 i Kerosene brand “A,” sp.gr. 0.785 Diesel fuel, initial boiling point 180°C . . .. .. .. VARIATION OF TURBIDITY WITH EQUAL WEIGHTS OF DIFFERENT COMMERCIAL OILS- different commercial oils at a wavelength of 5500 A with a light path of 17 mm. Table I1 sets out the light dispersed or absorbed in a photo-electric photometer by TABLE I1 LIGHT DISPERSED OK ABSORBED BY DIFFERENT OILS Volume of Photometer readings, logarithmic scale, for oil dispersed - 1 in 50 ml of acid light lighter petrol petrol white kerosene kerosene Teepol solution, petroleum, fuel, brand “A,” brand “B,” spirit, brand “A,” brand “,B” ml sp.gr.0.646 sp.gr. 0.705 sp.gr. 0.730 sp.gr. 0.730 sp.gr. 0.785 sp.gr. 0.785 sp.gr. 0.785 0.005 0.005 - - - 0.068 0.095 0.100 0.010 0.015 0-032 0.021 0.026 0.168 0.166 0.165 0.020 0.050 0.095 0-069 0.060 0.31 8 0.336 0-338 0.660 0.030 0.066 0.176 0.185 0.180 0-545 0.565 0.040 0.09*5 0.290 0.280 0.286 0.662 0.660 0.050 0.050 - 0.390 0.450 0.430 - - - It may be noted that the turbidity, as measured by light interference, usually increases with the specific gravity of the oil and, hence, with‘the proportion of hydrocarbons of higher molecular weight. This may be due to the dual effect of differences in refractivity and solubility.Very volatile hydrocarbons, such as light petroleum, are not likely to be met in practice and, although measurement made against arbitrary standards prepared, for example, from petrol or kerosense will impair the accuracy of the method if the oil in the sample is of a markedly different grade, it will probably be convenient to use them for routine examinations. INTERFERING SUBSTANCES The term “oil” lacks precision and any method of general application for the determina- tion of oil cannot be specific. The qualifications “volatile” and “neutral” are useful in imply- ing some restriction on the classes of compounds that are to be determined, and the method is designed not to include as “oil” certain volatile substances, e.g., phenols, that are not usually so described.Clearly this method of determination will include substances that, under the conditions of the test, are volatile, neutral, soluble in acetone and insoluble in water if, in addition, they are also adsorbed on activated carbon. As there may be differences * -Acetone was used as a solvent for diesel fuel instead of propylene glycol.524 SHERRATT: DETERMINATION OF [Vol. 81 of opinion as to whether, e.g., the inclusion of nitrobenzene or certain other synthetic organic compounds as “oil” is legitimate, it is difficult to decide what would constitute interference in the direction of over-estimation. This difficulty is not, of course, peculiar t o the proposed method ; it applies equally to the determination of non-volatile oils by extraction procedures and to distillation procedures involving measurement of separated oil.In theory, at least, the number of organic compounds that may be present in an industrial effluent is almost infinite : within practical limits, therefore, investigation of possible interference in the direction of over-estimation must be confined to those classes of compound that may be present in effluents such as sewage, leaving the ra.re and exceptional problem to be considered on its merits. Among non-oily substances that are frequently encountered in effluents are hydrogen sulphide and alkaline sulphides and polysulphides : sulphur may be produced from these compounds under the conditions of the test. Sulphur is volatile in steam, soluble in acetone and insoluble in water. Prima facie, therefore, sulphur and sulphide compounds might interfere with the method of analysis, but tests with sulphur and polysulphides have indicated that in concentrations up to 1OOOmg per litre these substances do not interfere.On the negative side no substance has been encountered that leads directly to low results if the unavoidable limitations of a method that depends upon very low solubilities are taken into account. Clearly, compounds that are appreciably soluble in water, e g . , benzene, would tend to be under-estimated; and, although it would be easy to modify or adapt the principles involved to meet many specific problems, a detailed consideration of individual cases would be inappropriate in describing a method primarily designed to estimate volatile oils. In the sphere of river pollution, volatile oils are usually aliphatic hydrocarbons of low solubility.In the presence of volatile-oil solvents, such as ketones, the procedure of transferring the oil from the sample to the carbon was found to take longer than usual, presumably because the reflux effect permits traces of the oil to be dissolved in the condensing liquid and returned to the flask. By continuing the passage of air until drops of condensing organic solvents were no longer visible in the condenser, satisfactory recovery of traces of oil in the presence of up to 2000 mg of acetone per litre was obtained. SOLUBILITY IN WATER OF SOME COMMERCIAL OILS- fuel oils. 30 minutes and then allowed to settle in a separating funnel overnight. was withdrawn and filtered through a wet No.31 Whatman filter-paper. examined for volatile oil by the method described; the following results were obtained- The method has been used to determine the solubility in water of some commercial An excess of the test oil was shaken with tap water in a shaking machine for The aqueous layer The filtrate was Sample number 1 2 3 4 5 6 7 8 9 10 Solubility, nig per litre First-grade petrol brand “A” .. . . .. 23 White spirit . . .. .. . . . . . . 17 Kerosene (repeat) . . . . .. . . . . 3 TABLE I11 VOLATILE OIL FOUND IN WATER SAMPLES Commercial petrol brand “B” . . . . .. 17 Kerosene , . .. .. .. . . .. 3 Volatile oil, mg per litre (expressed as Description of sample kerosene) Remarks River water from highly industrialised district . . 5.0 Known to contain surface drainage As No. 1, after coagulation, settlement and filtration 0.0 - Crude sewage from large military aerodrome , . 8-0 - As No. 3, after treatment . . .. . . .. 0.0 - Clean water from upland river . . .. .. 0.0 B.O.D. 1 p.p.m. As No. 5, after receiving heavy pollution from organic chemical works . . .. .. .. .. 0.0 B.O.D. 10 p.p.m. 0.0 Drv weather As No. 7, after treatment . . .. .. . . 0-0 Dry weather Drainage from open-cast coal mining, sample A . . 0.0 Contained 10 p.p.m. of Drainage from open-cast coal mining, sample B . . 6.0 Visible traces of oil and Crude sewage from rural area .. .. .. non-volatile oil smell of diesel fuelSept., 19561 VOLATILE OIL I N EFFLUENTS 525 Table 111 shows some of the results that have been obtained on certain samples sent in for routine analysis, including samples of crude and treated sewage and of river waters that might be expected to contain traces of volatile oils derived from gasworks effluents and the surface drainage of industrial districts. REFERENCES 1. 2. 3. Webber, L. A., and Burkes, C. E., Anal. Chem., 1952, 24, 1086. Anon., J . Boston SOC. Civ. Eng., 1924, 11, 237. Department of the Interior, Washington, D.C., U.S.X., Special Report, Fisheries No. 1, August, 1949. PUBLIC L % ~ ~ ~ ~ ~ ~ ’ ~ LABORATORY \V A R R I N GT o N , LA ?c’ c s . FLAG LANE A p d 16th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9568100518

出版商:RSC    

年代:1956

数据来源: RSC

摘要:

Sept., 19561 VOLATILE OIL I N EFFLUENTS 525 The Determination of Selenium in Effluents BY D. N. FOGG AND N. T. WLKINSON The American Public Health Association method for the determination of selenium has been investigated and found unsatisfactory, because the colour of the turbidity produced by the precipitated selenium differs in the sample solution and in the standards. In the proposed method the selenium is first separated by distillation with hydrobromic acid and bromine in the presence of sulphuric acid. The selenium is then precipitated with ascorbic acid as a turbidity under standard conditions of acidity in both the sample and standard distillates. Under the conditions of the test reproducible colours and accurate results are obtained. In the presence of organic matter full recovery of selenium is not attained when a wet destruction of the organic matter is carried out with nitric and sulphuric acids.Recovery is complete, however, if in addition to the nitric and sulphuric acids a small volume sf perchloric acid is added. THE method for the determination of selenium described by the American Public Health Association1 is stated to be tentative, since it is conceded that the method gives low results. We have examined the method and consider it to be unsatisfactory, because the same conditions are not maintained for both the preparation of the standard colours and the sample under test. These different conditions cause the colour turbidities to vary in tint from yellow to pink. From the results of our experimental work we have developed a method requiring standard conditions and we have replaced the two reducing agents used in the A.P.H.A.met hod-sulphur dioxide and h ydroxylamine-by one reducing agent, ascorbic acid.2 This reagent reduces both the excess of bromine and precipitates the selenium consistently in a pink colloidal form, which permits easy matching of the sample solution with standards. I n the destruction of any organic matter by wet digestion with nitric and sulphuric acids, loss of selenium occurs. If, however, the digestion is done in the presence of a small amount of perchloric acid, complete recovery of selenium is obtained. EXPERIMENTAL Our first experiments were directed to ascertaining the accuracy of the A.P.H.A.l met hod. A 5-ml portion of a standard selenium solution (1 ml = 0.00010 g of selenium) was added to 100 ml of distilled water and the selenium was determined by the procedure described by the A.P.H.A.Comparison of the test solution was made against a series of standards prepared from a standard distillate as described by the A.P.H.A. The colour of the turbidity produced in the test solution was yellowish pink, but in each of the standard solutions it was reddish pink, which made accurate matching impossible. It should be noted that the standards are prepared by first precipitating the whole of the selenium in the standard distillate and then diluting aliquots of the latter. The experiment described above was repeated, but this time the standard distillate was diluted to 100 ml in a calibrated flask. Aliquots of the solution, e g ., 1.0, 2.5, 5 0, 7.5526 FOGG AND WILKINSON : ’THE DETERMINATIOS [Vol. 81 and 10-0m1, were placed into a series of 100-ml Nessler cylinders and each solution was diluted to 75 ml with distilled water. Each solution was then treated with sulphur dioxide and hydroxylamine hydrochloride as described by the A.P.H.A. No precipitate of selenium was obtained in any of the standards during the l-hour standing time. The result of this experiment appeared to indicate that the acidity of the solution is of importance when the reduction is carried out with sulphur dioxide and hydroxylamine. EFFECT OF ACIDITY ON THE PRECIPITATION OF SELENIUM WITH SULPHUR DIOXIDE AND HYDROXYLAMINE- A 2.0-ml portion of the standard selenium solution was added to each of five Nessler cylinders (100 ml capacity).Volumes of hydrcibromic acid, vix., 10, 20, 30, 40 and 50 ml, were added to each of the cylinders and the solutions were diluted to 75 ml with distilled water. Sulphur dioxide gas was passed into each solution for 5 seconds, and 2 ml of 10 per cent. hydroxylamine hydrochloride and then 1.0 ml of 5 per cent. gum arabic solution were added. Each solution was mixed and set aside for 1 hour. The results are shown in Table I. TABLE: I EFFECT OF ACIDITY ON SELLENIUM PRECIPITATION Volume of hydrobromic acid in test solution, Result ml 10 No turbidity 20 Trace of turbidity 30 Partial development of turbidity 40 Full development of turbidity* 50 Full development of turbidity* * The turbidities produced in solutions t o which 40 ml and 50 ml of hydrobromic acid had been added were alike in colour and intensity, and were therefore regarded as having “full development of turbidity.” The result of this experiment shows that it is necessary t o have present a minimum of 40ml of 48 per cent. hydrobromic acid in a volume of 78ml in order to obtain complete precipitation of the selenium.Many experiments were carried out in an endeavour to standardise the conditions of the test with sulphur dioxide and hydroxylamine hydrochloride as reducing agents so that the same shade of turbidity was reproducible. PRECIPITATION O F SELENIUM BY SULPHUR DIOXIDE AND HYDROXYLAMINE I N SIX IDENTICAL SOLUTIONS- Six identical solutions were prepared in 50-ml Nessler cylinders, each containing 20 ml of hydrobromic acid, 5 ml of hydrobromic acid - bromine reagent, 2.5 ml of the standard selenium solution and 10 ml of distilled water.The reagents added to each cylinder were of the same quality and taken from the same bottles. Sulphur dioxide gas was passed into each solution until the bromine was reduced and for 5 seconds longer, and then 2 ml of 10 per cent. w/v hydroxylamine hydrochloride solution and 1.0 ml of 5 per cent. w/v gum arabic solution were added. The solutions were mixed and set aside for 1 hour. As far as could be judged, the reduction with sulphur dioxide was carried out in the same manner for each solution. A reddish-pink turbidity was produced in each of three solutions, a yellowish-pink turbidity was produced in each of two solutions and an intermediate tint of pink and yellow turbidity was produced in the remaining solution.In view of the difficulty of standardising the conditions for the precipitation of selenium with sulphur dioxide, experiments were next carried out with ascorbic acid2 as reducing agent. PRECIPITATION OF SELENIUM WITH ASCORBIC ACID I N SIX IDENTICAL SOLUTIONS- Six identical solutions were prepared in 50-ml Nessler cylinders, each containing 20 ml of hydrobromic acid, 5 ml of hydrobromic acid - bromine reagent, 2.5 ml of the standard selenium solution and 10 ml of distilled water. The reduction was carried out by the addition of 0.4 g of ascorbic acid to each solution, the solution being mixed until the ascorbic acid had dissolved. On standing for 20 minutes and for at least 1 hour thereafter each solution produced coloured turbidities of the same intensity and tint.Sept., 19561 OF SELENIUM IN EFFLUENTS 527 EFFECT OF ACIDITY ON THE PRECIPITATION OF SELENIUM WITH ASCORBIC ACID- A 1.5-ml portion of the standard selenium solution was placed in each of six 50-ml Nessler cylinders.Different volumes of 48 per cent. redistilled hydrobromic acid were added (5 ml, 10 ml, 15 ml, 20 ml, 25 ml and 30 ml) and the solutions were diluted to 50 ml. Then 0.4g of ascorbic acid was added and the solution was stirred until the ascorbic acid had completely dissolved. There was a full development of turbidity in each solution, as all the solutions were alike in the amounts of selenium precipitated. They were different however in that with increasing acidity the colour of the turbidity changed gradually from yellowish pink to almost pure pink.The experiments were repeated with 0.5 ml of the standard selenium solution for each test, and similar results were obtained. It was decided to adopt ascorbic acid as the reagent for precipitating selenium. SEPARATION OF SELENIUM BY DISTILLATION- Distillations of solutions containing known quantities of selenium were carried out, including the distillation of a “bulk” volume of standard selenium solution. For the artificial sample solutions, 0.5, 2.0 and 2.5 ml of the standard selenium solution were each diluted to 25 ml with distilled water, and a distillation was carried out on each solution as follows. To the 25 ml of solution contained in a 300-ml round-bottomed flask having a ground- glass socket 50 ml of hydrobromic acid and 6 ml of hydrobromic acid - bromine reagent were added.Then 25 ml of sulphuric acid, spgr. 1.84, were added carefully, the flask being cooled under running water during the addition. A small boiling rod was placed in the flask, and an adaptor and water-cooled condenser were fitted; the condenser had a long stem with narrow jet at the outlet. A 100-ml conical flask with a graduation mark etched at 80 ml was placed at the condenser outlet. Four millilitres of hydrobromic acid - bromine reagent were added to the conical flask, which was tilted so that the tip of the condenser outlet was below the level of the solution in the receiver. The contents of the distillation flask were heated, gently at first until air was displaced and afterwards more strongly, and the distillation was allowed to proceed until the volume of solution in the receiver was SO ml.The conical flask was removed and the distillation was stopped. The distillate was transferred to a 100-ml calibrated flask, diluted to the mark with distilled water and the contents of the flask were mixed. Distillation of 25 ml of the standard selenium solution was then carried out exactly as described above, and the distillate was also diluted to 100ml with mixing. A determination of the acidity of each sample distillate and the standard distillate was carried out as follows. Five millilitres of each distillate were diluted with 100ml of water and boiled until free from bromine. Each solution was cooled and titrated with N sodium hydroxide, with methyl orange as indicator.The acidity of each distillate was similar, 5 ml requiring 26-0 ml of N sodium hydroxide. The concentration of the hydrobromic acid (48 per cent.) was also determined by titration with N sodium hydroxide and found to be 9 N. Fifty millilitres of each artificial sample distillate were transferred to 50-ml Nessler cylinders. The acidity of the 50 ml of distillate in each of the Nessler cylinders was equivalent to 260/9 = 29 ml of hydrobromic acid. Suit able aliquots of the standard distillate were measured into 50-ml Nessler cylinders and sufficient hydrobromic acid was added so that the acidity of each standard solution was the same as that of 50 ml of the artificial sample distillate, ie., = 260 ml of N acid.Each standard was diluted to 50 ml with distilled water. Then 0.4 g of ascorbic acid was added to each standard and artificial sample solutions, each solution being mixed by stirring until the ascorbic acid had dissolved. The turbidities produced in the sample solutions compared well in intensity and tint with the corresponding standard solutions containing the same amount of selenium, i.e., 0*000025, 0.00010 and 0-000125 g of selenium. A series of standards was also prepared from the standard selenium solution without distillation ; 0.25, 1.0 and 1.25-ml portions of the standard selenium solution were measured into 50-ml Nessler cylinders, 29 ml of 48 per cent. hydrobromic acid were added to each and the solutions were diluted to 50 ml with distilled water. Reduction with ascorbic acid Each solution was then set aside for 30 minutes.528 FOGG AND WILKINSON THE DETERMINATION [Vol.81 was then carried out as described above. The turbidities produced were exactly comparable with the corresponding sample and standard distillate solutions described above. Recovery of selenium was therefore quantitative by the distillation procedure and the various bromine contents of the distillates did not affect the subsequent reduction with ascorbic acid, provided that the sample and standard solutions both contained hydrobromic acid equivalent to 260 ml of N acid. EFFECT O F ORGANIC MATTER- Destruction of organic matter with a mixturc: of nitric and sulphuuric acids-*4 1-0-ml and a 2.5-ml portion of the standard selenium solution were placed in 250-ml beakers; each solution was diluted to 50 ml with distilled water and 0.2 g of Quebracho tannin was added.Then 5 ml of nitric acid, sp.gr. 1.42, were added and the solutions were evaporated to 10 ml, cooled and 5 ml of sulphuric acid were added to each. The solutions were evaporated on a sand-bath until white fumes of sulphur trioxide appeared. This treatment destroyed the organic matter. The solutions were cooled and 10 ml of distilled water were added to each, and the solutions were again evaporated until white fumes of sulphur trioxide appeared. The treatment with distilled water and evaporation was repeated once more. Each solution was transferred to a distillation flask, 25 ml of distilled water being used for the transfer, and 50 ml of hydrobromic acid and 6 ml of hydrobromic acid - bromine reagent were added, and then 25ml of sulphiiric acid, sp.gr.1.84, were added carefully, the flask being cooled under running water during the addition. The distillation of this solution was carried out as described previously and the distillate was diluted to 100 ml in a calibrated flask. A standard distillate was prepared by distillation of 25 ml of standard selenium solution after addition of 50ml of hydrobromic acid and so on, and the distillate was diluted to 100ml in a calibrated flask. Fifty millilitres or a smaller volume of the artificial sample distillates were transferred to 50-ml Nessler cylinders. Aliquots of the standard distillate were measured into similar Nessler cylinders. The acidity of the solutions was, when necessary, made equal to that of 50ml of the sample distilJate by adding hydrobromic acid and then diluting to 50ml.The selenium was then precipitated by addition of 0.4 g of ascorbic acid to each. The results obtained by visual comparison are given in Table 11. TABLE; I1 EFFECT OF OXIDATION O F ORGANIC MATTER WITH NITRIC - SULPHURIC ACID ON SELENIUM RECOVERY Volume of aliquot of sample distillate taken, ml 50 50 50 50 10 50 Selenium added in aliquot, g 0.00005 0*000125 0.00005 0.0001 25 0~00020 0~00020 Selenium found in aliquot, g 0.000038 0~00010 0.000038 0.000075 0.00016 0.00006 We thought that the low results might be due to the precipitation of selenium by nitrous acid formed during the wet-oxidation procedure. We therefore decided to carry out a treatment of selenium solution in the absence of organic matter with nitric and sulphuric acids and then carry out the test for selenium by the method described above.The results are given in Table 111. TABLE I11 EFFECT OF TREATMENT WITH NITRIC - SULPHURIC ACID ON SELENIUM RECOVERY Volume of aliquot of Selenium added Selenium found sample distillate taken, in aliquot, in aliquot, ml g g 50 0.000025 0.000025 50 0~00010 0+00010 50 0.000125 0.0001 25Sept., 1956; O F SELENIUM I N EFFLUENTS 529 These results supported our opinion that nitric acid was not a strong enough oxidising agent to prevent loss of selenium during a wet oxidation of organic matter with nitric and sulphuric acids. Further confirmation of this was obtained in another experiment.After destruction of organic matter by wet digestion with nitric and sulphuric acids the nitric - sulphuric acid complex that forms in such a digestion was decomposed by addition of 10 ml of a saturated solution of ammonium oxalate and the solution was evaporated until white fumes of sulphur trioxide appeared. Elemental selenium was observed in the solution. Destruction of organic matter with a mixture of nitric, perchloric and sulphuric acids- Experiments were carried out to ascertain what effect a small addition of perchloric acid would have in helping to prevent loss of selenium during destruction of organic matter by wet oxidation. Different volumes of the standard selenium solution were placed in 250-ml beakers and each solution was diluted to 50 ml with distilled water and 0.2 g of Quebracho tannin was added to each. Then 5 ml of nitric acid, sp.gr.1-42, and 1 ml of 60 per cent. perchloric acid were added to each solution, which was then evaporated to 10 ml. Each was cooled and 5 ml of sulphuric acid were added; the solutions were then evaporated on a sand-bath until white fumes of sulphur trioxide appeared. The solutions were cooled, 10 ml of distilled water were added to each, and they were evaporated until white fumes again appeared. The treatment with distilled water and evaporation was repeated once more. Each solution was transferred to a distillation flask, 25 ml of distilled water being used for the transfer. From this point the procedure was exactly as described for the previous experiment. The results are given in Table IV.TABLE IV EFFECT OF OXIDATION OF ORGANIC MATTER WITH NITRIC - PERCHLORIC - SULPHURIC ACID ON SELENIUM RECOVERY Volume of aliquot of sample distillate taken, ml 50 50 50 10 50 Selenium added in aliquot, g 0*000025 0*00010 0.000125 0~00020 0*00002 Selenium found in aliquot, g 0.000025 0*00010 0.000125 0*00020 0~00002 The method was then tried out on solutions each containing 0.003 g of calcium, 0.0008 g of magnesium, 0-005 g of sulphate, 0-002 g of chloride, 0.0002 g of tin, 0.00015 g of arsenic and 0.5 g of tannin in a volume of 50 ml. Different volumes of the standard selenium solution containing 0.00025 g, 0.0025 gJ 0.0020 g and 0.00004 g of selenium were added to four solutions of the above composition. A blank test was also performed. The selenium in the solutions was then determined as described under “Method.” The results are given in Table V.TABLE V EFFECT OF ADDED SUBSTANCES ON SELENIUM RECOVERY Volume of aliquot of sample distillate taken, ml 50 50 10 10 50 Selenium added in aliquot, g nil 0.000125 0.00025 0~00020 0~00002 Selenium found in aliquot, g nil 0.000125 0*00025 0~00020 0*00002 The experiments were repeated in the presence of the following amounts of impurities in each test: 0.03 g of calcium, 0.01 g of magnesium, 0.05 g of sulphate, 0.02 g of chloride, 0.002 g of tin, 0.002 g of arsenic and 0.5 g of tannin in a volume of 50 ml. Different volumes of the standard selenium solution containing 0.00002 g, 0.00010 g, 0.00060 g and 0.0025 g of selenium were added to four solutions of the above composition.530 FOGG AND WILKINSON THE DETERMINATION [Vol.81 A blank test was also performed. described under “Method.” The selenium in the solutions was then determined as The results are given in Table VI. TABLE VI EFFECT OF ADDED SUBSTANCES ON SELENIUM RECOVERY Volume of aliquot of sample distillate taken, ml 50 50 50 25 10 Selenium added in aliquot, g nil 0~00001 0.00005 0*00015 0*00025 Selenium found in aliquot, .!3 nil 0~00001 0-00005 0.000 15 0*00025 METHOD REAGENTS- Nitric acid, sp.gr. 1.42. Sulphuric acid, sp.gr. 1-84, Perchloric acid, 60 eer cent. Methyl orange indicator solution. Sodium hydroxide, 1-0 N. Hydrobromic acid, redistilled-Purify the acid by distillation. Collect the col ourless Determine the hydrogen bromide content of the collected Hydrobromic acid - bromine reagent-Mix 3 ml of bromine with 197 ml of redistilled Standard selenium solution-(a) Stock solutzon: Weigh 1.405 g of selenium dioxide and Transfer the solution to a 1-litre calibrated flask, add 80ml (b) Dilute solution: Add 10ml middle fraction of the distillate.fraction by titrating 5 ml with N sodium hydroxide, using methyl orange as indicator. hydrobromic acid. dissolve it in distilled water. of redistilled hydrobromic acid and dilute the solution to 1 litre. Measure 100 ml of the selenium stock solution into a 1-litre calibrated flask. of redistilled hydrobromic acid and dilute the solution to 1 litre. 1 ml = 0.00010 g of selenium. Ascorbic acid. Measure a suitable volume of the sample into a beaker. PROCEDURE FOR TREATING THE SAMPLE- Acidify the solution by the addition of nitric acid and add 5 ml in excess; then add 1.0 ml of 60 per cent.perchloric acid. Evaporate the sample to about 10ml and then cool it. Add 5ml of sulphuric acid and heat the solution on a sand-bath until white fumes appear, If the organic matter is not completely destroyed at this stage, add a further 1 ml of nitric acid and heat the solution again until white fumes appear. Allow the solution to cool, then add 10 ml of distilled water and heat the solution until white fumes appear. Repeat the treatment with distilled water and evaporate a second time. PROCEDURE FOR DISTILLING THE SELENIUM- Cool the solution and transfer it to the distillation flask, using 25 ml of distilled water for the transfer. Add 50 ml of hydrobromic acid and 6 ml of the hydrobromic acid - bromine reagent.Then slowly and carefully add 25 ml of sulphuric acid to the contents of the flask, cooling the flask during the addition. Fit an adaptor and water-cooled condenser to the flask and place a 100-ml conical flask, graduated at 80 ml and containing 4 ml of hydrobromic acid - bromine reagent, at the condenser outlet; tilt the conical flask so that the tip of the Condenser lies below the level of the reagent. Heat the contents of the distillation flask, gently at first until the solution is boiling and afterwards more strongly, and allow the distillation to proceed until the volume of solution in the receiver is 80ml. Transfer the distillate to a 100-ml calibrated flask, dilute the solution to the mark with distilled water and mix.PROCEDURE FOR PREPARING A STANDARD SELENIUM DISTILLATE- Measure into a distillation flask similar to that used for the sample solution, 25 ml of the dilute selenium solution (1 ml = 0.00010 g of selenium) and follow the same procedure Place a boiling rod inside the distillation flask. Remove the receiver and discontinue the distillation.Sept., 19561 O F SELENIUM I N EFFLUENTS 531 for the distillation of selenium as described above, commencing with the addition of 50ml of hydrobromic acid. Allow the distillation to proceed until the volume of solution in the receiver is 80 ml. Transfer the distillate to a 100-ml calibrated flask and dilute the solution to the mark with distilled water and mix. PROCEDURE FOR COLORIMETRIC DETERMINATION- Determine the acidity of the sample and standard distillate solutions on an aliquot of each solution as follows. Measure 5 ml of the solution, add 100 ml of distilled water and boil the solution free from bromine, but not below 50ml. Cool the solution, add 2 drops of methyl orange indicator and titrate with N sodium hydroxide. Measure aliquots of the standard distillate solution covering the range 0 to 10 ml (Le., 0 to 0*00025 g of selenium) in steps of 1 ml into 50-ml Nessler cylinders. From the results of the above titrations add a calculated volume of hydrobromic acid to each cylinder so that the acid content of each standard is the same as that of 50ml of sample distillate solution. Dilute each standard to 50ml with distilled water. Precipitate the selenium in the sample and standard solutions by adding 0.4 g of ascorbic acid to each solution and stir the solutions until the ascorbic acid has dissolved. Set the solutions aside for 30 minutes, and then visually match the coloured turbidity in the sample solution against the coloured turbidities in the standard solutions. Measure 50 ml of the sample distillate solution into a 50-ml Nessler cylinder. REFERENCES 1. 2. American Public Health Association, “Standard Methods for the Examination of Water, Sewage Rudra, M. N., and Rudra, Siuli, Curr. Sci., 1952, 21, 229. and Industrial Wastes,” Tenth Edition, New York, 1955, p. 179. IMPERIAL CHEMICAL INDUSTRIES LIMITED RESEARCH DEPARTMENT ALKALI DIVISION WINNINGTON, VORTHWICH, CHESHIRE April 26th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9568100525

出版商:RSC    

年代:1956

数据来源: RSC